diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,53934 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 14.25925925925926,
+  "eval_steps": 500,
+  "global_step": 7700,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.001851851851851852,
+      "grad_norm": 16.01453742088244,
+      "learning_rate": 3.0864197530864196e-10,
+      "loss": 1.8896,
+      "step": 1
+    },
+    {
+      "epoch": 0.003703703703703704,
+      "grad_norm": 15.375592963986762,
+      "learning_rate": 6.172839506172839e-10,
+      "loss": 1.9114,
+      "step": 2
+    },
+    {
+      "epoch": 0.005555555555555556,
+      "grad_norm": 15.405659955056148,
+      "learning_rate": 9.259259259259258e-10,
+      "loss": 1.8899,
+      "step": 3
+    },
+    {
+      "epoch": 0.007407407407407408,
+      "grad_norm": 15.613867338083537,
+      "learning_rate": 1.2345679012345679e-09,
+      "loss": 1.8778,
+      "step": 4
+    },
+    {
+      "epoch": 0.009259259259259259,
+      "grad_norm": 16.24258898972724,
+      "learning_rate": 1.5432098765432097e-09,
+      "loss": 1.8796,
+      "step": 5
+    },
+    {
+      "epoch": 0.011111111111111112,
+      "grad_norm": 15.967516821922272,
+      "learning_rate": 1.8518518518518516e-09,
+      "loss": 1.8563,
+      "step": 6
+    },
+    {
+      "epoch": 0.012962962962962963,
+      "grad_norm": 16.330852281159242,
+      "learning_rate": 2.1604938271604934e-09,
+      "loss": 1.862,
+      "step": 7
+    },
+    {
+      "epoch": 0.014814814814814815,
+      "grad_norm": 15.376267659430235,
+      "learning_rate": 2.4691358024691357e-09,
+      "loss": 1.8683,
+      "step": 8
+    },
+    {
+      "epoch": 0.016666666666666666,
+      "grad_norm": 15.328960258881947,
+      "learning_rate": 2.7777777777777776e-09,
+      "loss": 1.8845,
+      "step": 9
+    },
+    {
+      "epoch": 0.018518518518518517,
+      "grad_norm": 15.39795148685497,
+      "learning_rate": 3.0864197530864194e-09,
+      "loss": 1.8821,
+      "step": 10
+    },
+    {
+      "epoch": 0.020370370370370372,
+      "grad_norm": 15.89711552333428,
+      "learning_rate": 3.3950617283950613e-09,
+      "loss": 1.8833,
+      "step": 11
+    },
+    {
+      "epoch": 0.022222222222222223,
+      "grad_norm": 15.436268793133458,
+      "learning_rate": 3.703703703703703e-09,
+      "loss": 1.886,
+      "step": 12
+    },
+    {
+      "epoch": 0.024074074074074074,
+      "grad_norm": 15.58921411331668,
+      "learning_rate": 4.012345679012345e-09,
+      "loss": 1.8843,
+      "step": 13
+    },
+    {
+      "epoch": 0.025925925925925925,
+      "grad_norm": 16.350936505218403,
+      "learning_rate": 4.320987654320987e-09,
+      "loss": 1.8674,
+      "step": 14
+    },
+    {
+      "epoch": 0.027777777777777776,
+      "grad_norm": 15.59724033876366,
+      "learning_rate": 4.6296296296296295e-09,
+      "loss": 1.8857,
+      "step": 15
+    },
+    {
+      "epoch": 0.02962962962962963,
+      "grad_norm": 15.806538261793046,
+      "learning_rate": 4.938271604938271e-09,
+      "loss": 1.8755,
+      "step": 16
+    },
+    {
+      "epoch": 0.03148148148148148,
+      "grad_norm": 15.902812899049493,
+      "learning_rate": 5.246913580246913e-09,
+      "loss": 1.8819,
+      "step": 17
+    },
+    {
+      "epoch": 0.03333333333333333,
+      "grad_norm": 16.448802476502646,
+      "learning_rate": 5.555555555555555e-09,
+      "loss": 1.8983,
+      "step": 18
+    },
+    {
+      "epoch": 0.03518518518518519,
+      "grad_norm": 15.990295108515035,
+      "learning_rate": 5.864197530864197e-09,
+      "loss": 1.8844,
+      "step": 19
+    },
+    {
+      "epoch": 0.037037037037037035,
+      "grad_norm": 15.174397301064433,
+      "learning_rate": 6.172839506172839e-09,
+      "loss": 1.916,
+      "step": 20
+    },
+    {
+      "epoch": 0.03888888888888889,
+      "grad_norm": 15.769790859011602,
+      "learning_rate": 6.481481481481481e-09,
+      "loss": 1.8974,
+      "step": 21
+    },
+    {
+      "epoch": 0.040740740740740744,
+      "grad_norm": 15.612263076501023,
+      "learning_rate": 6.7901234567901226e-09,
+      "loss": 1.8726,
+      "step": 22
+    },
+    {
+      "epoch": 0.04259259259259259,
+      "grad_norm": 15.051733205144405,
+      "learning_rate": 7.0987654320987644e-09,
+      "loss": 1.8778,
+      "step": 23
+    },
+    {
+      "epoch": 0.044444444444444446,
+      "grad_norm": 15.635262076445171,
+      "learning_rate": 7.407407407407406e-09,
+      "loss": 1.8593,
+      "step": 24
+    },
+    {
+      "epoch": 0.046296296296296294,
+      "grad_norm": 15.497782004107918,
+      "learning_rate": 7.716049382716048e-09,
+      "loss": 1.8676,
+      "step": 25
+    },
+    {
+      "epoch": 0.04814814814814815,
+      "grad_norm": 16.193103961933446,
+      "learning_rate": 8.02469135802469e-09,
+      "loss": 1.8722,
+      "step": 26
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.309127038724816,
+      "learning_rate": 8.333333333333332e-09,
+      "loss": 1.9183,
+      "step": 27
+    },
+    {
+      "epoch": 0.05185185185185185,
+      "grad_norm": 15.65547088924166,
+      "learning_rate": 8.641975308641974e-09,
+      "loss": 1.8873,
+      "step": 28
+    },
+    {
+      "epoch": 0.053703703703703705,
+      "grad_norm": 16.00856397449651,
+      "learning_rate": 8.950617283950616e-09,
+      "loss": 1.8726,
+      "step": 29
+    },
+    {
+      "epoch": 0.05555555555555555,
+      "grad_norm": 15.60440723586894,
+      "learning_rate": 9.259259259259259e-09,
+      "loss": 1.8856,
+      "step": 30
+    },
+    {
+      "epoch": 0.05740740740740741,
+      "grad_norm": 15.371658124387729,
+      "learning_rate": 9.567901234567901e-09,
+      "loss": 1.8822,
+      "step": 31
+    },
+    {
+      "epoch": 0.05925925925925926,
+      "grad_norm": 16.084314624459886,
+      "learning_rate": 9.876543209876543e-09,
+      "loss": 1.8913,
+      "step": 32
+    },
+    {
+      "epoch": 0.06111111111111111,
+      "grad_norm": 15.580766097420982,
+      "learning_rate": 1.0185185185185185e-08,
+      "loss": 1.856,
+      "step": 33
+    },
+    {
+      "epoch": 0.06296296296296296,
+      "grad_norm": 16.17880487954311,
+      "learning_rate": 1.0493827160493827e-08,
+      "loss": 1.888,
+      "step": 34
+    },
+    {
+      "epoch": 0.06481481481481481,
+      "grad_norm": 15.936455325209984,
+      "learning_rate": 1.0802469135802468e-08,
+      "loss": 1.8952,
+      "step": 35
+    },
+    {
+      "epoch": 0.06666666666666667,
+      "grad_norm": 15.298388504567042,
+      "learning_rate": 1.111111111111111e-08,
+      "loss": 1.8959,
+      "step": 36
+    },
+    {
+      "epoch": 0.06851851851851852,
+      "grad_norm": 15.144670267344019,
+      "learning_rate": 1.1419753086419752e-08,
+      "loss": 1.8926,
+      "step": 37
+    },
+    {
+      "epoch": 0.07037037037037037,
+      "grad_norm": 15.960938559417837,
+      "learning_rate": 1.1728395061728394e-08,
+      "loss": 1.8889,
+      "step": 38
+    },
+    {
+      "epoch": 0.07222222222222222,
+      "grad_norm": 15.777294203296226,
+      "learning_rate": 1.2037037037037036e-08,
+      "loss": 1.8995,
+      "step": 39
+    },
+    {
+      "epoch": 0.07407407407407407,
+      "grad_norm": 15.497973406924203,
+      "learning_rate": 1.2345679012345678e-08,
+      "loss": 1.8804,
+      "step": 40
+    },
+    {
+      "epoch": 0.07592592592592592,
+      "grad_norm": 16.141837983841427,
+      "learning_rate": 1.2654320987654321e-08,
+      "loss": 1.8894,
+      "step": 41
+    },
+    {
+      "epoch": 0.07777777777777778,
+      "grad_norm": 14.684290707896258,
+      "learning_rate": 1.2962962962962961e-08,
+      "loss": 1.8917,
+      "step": 42
+    },
+    {
+      "epoch": 0.07962962962962963,
+      "grad_norm": 15.634733612368853,
+      "learning_rate": 1.3271604938271605e-08,
+      "loss": 1.9091,
+      "step": 43
+    },
+    {
+      "epoch": 0.08148148148148149,
+      "grad_norm": 15.961623048553635,
+      "learning_rate": 1.3580246913580245e-08,
+      "loss": 1.8841,
+      "step": 44
+    },
+    {
+      "epoch": 0.08333333333333333,
+      "grad_norm": 15.7446567956114,
+      "learning_rate": 1.3888888888888889e-08,
+      "loss": 1.8975,
+      "step": 45
+    },
+    {
+      "epoch": 0.08518518518518518,
+      "grad_norm": 15.949740281110246,
+      "learning_rate": 1.4197530864197529e-08,
+      "loss": 1.892,
+      "step": 46
+    },
+    {
+      "epoch": 0.08703703703703704,
+      "grad_norm": 15.43925634774645,
+      "learning_rate": 1.4506172839506172e-08,
+      "loss": 1.8864,
+      "step": 47
+    },
+    {
+      "epoch": 0.08888888888888889,
+      "grad_norm": 15.396877731257465,
+      "learning_rate": 1.4814814814814813e-08,
+      "loss": 1.8773,
+      "step": 48
+    },
+    {
+      "epoch": 0.09074074074074075,
+      "grad_norm": 15.230593923513473,
+      "learning_rate": 1.5123456790123456e-08,
+      "loss": 1.9074,
+      "step": 49
+    },
+    {
+      "epoch": 0.09259259259259259,
+      "grad_norm": 15.758712834103486,
+      "learning_rate": 1.5432098765432096e-08,
+      "loss": 1.8853,
+      "step": 50
+    },
+    {
+      "epoch": 0.09444444444444444,
+      "grad_norm": 15.207502212719666,
+      "learning_rate": 1.574074074074074e-08,
+      "loss": 1.8784,
+      "step": 51
+    },
+    {
+      "epoch": 0.0962962962962963,
+      "grad_norm": 15.123948116692935,
+      "learning_rate": 1.604938271604938e-08,
+      "loss": 1.8772,
+      "step": 52
+    },
+    {
+      "epoch": 0.09814814814814815,
+      "grad_norm": 15.278280361005264,
+      "learning_rate": 1.6358024691358024e-08,
+      "loss": 1.8965,
+      "step": 53
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 15.839011867588935,
+      "learning_rate": 1.6666666666666664e-08,
+      "loss": 1.8967,
+      "step": 54
+    },
+    {
+      "epoch": 0.10185185185185185,
+      "grad_norm": 16.65603378069207,
+      "learning_rate": 1.6975308641975307e-08,
+      "loss": 1.8939,
+      "step": 55
+    },
+    {
+      "epoch": 0.1037037037037037,
+      "grad_norm": 15.816691700766857,
+      "learning_rate": 1.7283950617283947e-08,
+      "loss": 1.8847,
+      "step": 56
+    },
+    {
+      "epoch": 0.10555555555555556,
+      "grad_norm": 15.798316407389207,
+      "learning_rate": 1.759259259259259e-08,
+      "loss": 1.887,
+      "step": 57
+    },
+    {
+      "epoch": 0.10740740740740741,
+      "grad_norm": 15.47639797900752,
+      "learning_rate": 1.790123456790123e-08,
+      "loss": 1.8748,
+      "step": 58
+    },
+    {
+      "epoch": 0.10925925925925926,
+      "grad_norm": 15.27598829120491,
+      "learning_rate": 1.8209876543209875e-08,
+      "loss": 1.8878,
+      "step": 59
+    },
+    {
+      "epoch": 0.1111111111111111,
+      "grad_norm": 15.563651931865994,
+      "learning_rate": 1.8518518518518518e-08,
+      "loss": 1.8734,
+      "step": 60
+    },
+    {
+      "epoch": 0.11296296296296296,
+      "grad_norm": 15.480857981942341,
+      "learning_rate": 1.882716049382716e-08,
+      "loss": 1.8833,
+      "step": 61
+    },
+    {
+      "epoch": 0.11481481481481481,
+      "grad_norm": 15.464224881293541,
+      "learning_rate": 1.9135802469135802e-08,
+      "loss": 1.8972,
+      "step": 62
+    },
+    {
+      "epoch": 0.11666666666666667,
+      "grad_norm": 15.90152313570589,
+      "learning_rate": 1.9444444444444445e-08,
+      "loss": 1.87,
+      "step": 63
+    },
+    {
+      "epoch": 0.11851851851851852,
+      "grad_norm": 14.983814185373543,
+      "learning_rate": 1.9753086419753086e-08,
+      "loss": 1.8687,
+      "step": 64
+    },
+    {
+      "epoch": 0.12037037037037036,
+      "grad_norm": 15.356093556702819,
+      "learning_rate": 2.006172839506173e-08,
+      "loss": 1.8886,
+      "step": 65
+    },
+    {
+      "epoch": 0.12222222222222222,
+      "grad_norm": 15.982885648456714,
+      "learning_rate": 2.037037037037037e-08,
+      "loss": 1.8817,
+      "step": 66
+    },
+    {
+      "epoch": 0.12407407407407407,
+      "grad_norm": 15.130325195138646,
+      "learning_rate": 2.0679012345679013e-08,
+      "loss": 1.8789,
+      "step": 67
+    },
+    {
+      "epoch": 0.1259259259259259,
+      "grad_norm": 15.614819205673005,
+      "learning_rate": 2.0987654320987653e-08,
+      "loss": 1.8949,
+      "step": 68
+    },
+    {
+      "epoch": 0.12777777777777777,
+      "grad_norm": 15.860795548772547,
+      "learning_rate": 2.1296296296296297e-08,
+      "loss": 1.9031,
+      "step": 69
+    },
+    {
+      "epoch": 0.12962962962962962,
+      "grad_norm": 15.463036496331627,
+      "learning_rate": 2.1604938271604937e-08,
+      "loss": 1.9112,
+      "step": 70
+    },
+    {
+      "epoch": 0.13148148148148148,
+      "grad_norm": 16.012861099884283,
+      "learning_rate": 2.191358024691358e-08,
+      "loss": 1.8721,
+      "step": 71
+    },
+    {
+      "epoch": 0.13333333333333333,
+      "grad_norm": 14.842117485417482,
+      "learning_rate": 2.222222222222222e-08,
+      "loss": 1.8864,
+      "step": 72
+    },
+    {
+      "epoch": 0.13518518518518519,
+      "grad_norm": 15.066224179737386,
+      "learning_rate": 2.2530864197530864e-08,
+      "loss": 1.8667,
+      "step": 73
+    },
+    {
+      "epoch": 0.13703703703703704,
+      "grad_norm": 15.373961458497948,
+      "learning_rate": 2.2839506172839504e-08,
+      "loss": 1.8816,
+      "step": 74
+    },
+    {
+      "epoch": 0.1388888888888889,
+      "grad_norm": 15.784324151509434,
+      "learning_rate": 2.3148148148148148e-08,
+      "loss": 1.8444,
+      "step": 75
+    },
+    {
+      "epoch": 0.14074074074074075,
+      "grad_norm": 15.196481277715948,
+      "learning_rate": 2.3456790123456788e-08,
+      "loss": 1.8624,
+      "step": 76
+    },
+    {
+      "epoch": 0.1425925925925926,
+      "grad_norm": 15.761145080537604,
+      "learning_rate": 2.376543209876543e-08,
+      "loss": 1.8882,
+      "step": 77
+    },
+    {
+      "epoch": 0.14444444444444443,
+      "grad_norm": 15.184276001289748,
+      "learning_rate": 2.407407407407407e-08,
+      "loss": 1.8815,
+      "step": 78
+    },
+    {
+      "epoch": 0.14629629629629629,
+      "grad_norm": 15.867557039261923,
+      "learning_rate": 2.4382716049382715e-08,
+      "loss": 1.8617,
+      "step": 79
+    },
+    {
+      "epoch": 0.14814814814814814,
+      "grad_norm": 15.528652016760974,
+      "learning_rate": 2.4691358024691355e-08,
+      "loss": 1.8606,
+      "step": 80
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 16.54859029739373,
+      "learning_rate": 2.5e-08,
+      "loss": 1.8925,
+      "step": 81
+    },
+    {
+      "epoch": 0.15185185185185185,
+      "grad_norm": 14.920939542797587,
+      "learning_rate": 2.5308641975308642e-08,
+      "loss": 1.8635,
+      "step": 82
+    },
+    {
+      "epoch": 0.1537037037037037,
+      "grad_norm": 15.386169707489259,
+      "learning_rate": 2.5617283950617286e-08,
+      "loss": 1.8643,
+      "step": 83
+    },
+    {
+      "epoch": 0.15555555555555556,
+      "grad_norm": 14.923801795083323,
+      "learning_rate": 2.5925925925925923e-08,
+      "loss": 1.8584,
+      "step": 84
+    },
+    {
+      "epoch": 0.1574074074074074,
+      "grad_norm": 16.573869792629477,
+      "learning_rate": 2.6234567901234566e-08,
+      "loss": 1.8974,
+      "step": 85
+    },
+    {
+      "epoch": 0.15925925925925927,
+      "grad_norm": 15.467925489897883,
+      "learning_rate": 2.654320987654321e-08,
+      "loss": 1.8619,
+      "step": 86
+    },
+    {
+      "epoch": 0.16111111111111112,
+      "grad_norm": 15.300386666285116,
+      "learning_rate": 2.6851851851851853e-08,
+      "loss": 1.8603,
+      "step": 87
+    },
+    {
+      "epoch": 0.16296296296296298,
+      "grad_norm": 15.605978677185178,
+      "learning_rate": 2.716049382716049e-08,
+      "loss": 1.8666,
+      "step": 88
+    },
+    {
+      "epoch": 0.1648148148148148,
+      "grad_norm": 15.459323029806637,
+      "learning_rate": 2.7469135802469134e-08,
+      "loss": 1.877,
+      "step": 89
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 15.37123605504257,
+      "learning_rate": 2.7777777777777777e-08,
+      "loss": 1.8859,
+      "step": 90
+    },
+    {
+      "epoch": 0.1685185185185185,
+      "grad_norm": 15.062010266255461,
+      "learning_rate": 2.808641975308642e-08,
+      "loss": 1.8555,
+      "step": 91
+    },
+    {
+      "epoch": 0.17037037037037037,
+      "grad_norm": 15.243190194498046,
+      "learning_rate": 2.8395061728395058e-08,
+      "loss": 1.8615,
+      "step": 92
+    },
+    {
+      "epoch": 0.17222222222222222,
+      "grad_norm": 15.299465623497353,
+      "learning_rate": 2.87037037037037e-08,
+      "loss": 1.8746,
+      "step": 93
+    },
+    {
+      "epoch": 0.17407407407407408,
+      "grad_norm": 16.017789825984863,
+      "learning_rate": 2.9012345679012345e-08,
+      "loss": 1.8422,
+      "step": 94
+    },
+    {
+      "epoch": 0.17592592592592593,
+      "grad_norm": 15.316176955655463,
+      "learning_rate": 2.9320987654320988e-08,
+      "loss": 1.8897,
+      "step": 95
+    },
+    {
+      "epoch": 0.17777777777777778,
+      "grad_norm": 15.120821230346957,
+      "learning_rate": 2.9629629629629625e-08,
+      "loss": 1.8606,
+      "step": 96
+    },
+    {
+      "epoch": 0.17962962962962964,
+      "grad_norm": 15.038634971379201,
+      "learning_rate": 2.993827160493827e-08,
+      "loss": 1.8772,
+      "step": 97
+    },
+    {
+      "epoch": 0.1814814814814815,
+      "grad_norm": 15.6967389857466,
+      "learning_rate": 3.024691358024691e-08,
+      "loss": 1.8485,
+      "step": 98
+    },
+    {
+      "epoch": 0.18333333333333332,
+      "grad_norm": 15.451950300728415,
+      "learning_rate": 3.0555555555555556e-08,
+      "loss": 1.8665,
+      "step": 99
+    },
+    {
+      "epoch": 0.18518518518518517,
+      "grad_norm": 15.212053142047973,
+      "learning_rate": 3.086419753086419e-08,
+      "loss": 1.8867,
+      "step": 100
+    },
+    {
+      "epoch": 0.18703703703703703,
+      "grad_norm": 15.702086924496298,
+      "learning_rate": 3.1172839506172836e-08,
+      "loss": 1.8727,
+      "step": 101
+    },
+    {
+      "epoch": 0.18888888888888888,
+      "grad_norm": 15.96363973798426,
+      "learning_rate": 3.148148148148148e-08,
+      "loss": 1.8869,
+      "step": 102
+    },
+    {
+      "epoch": 0.19074074074074074,
+      "grad_norm": 15.144738909008389,
+      "learning_rate": 3.179012345679012e-08,
+      "loss": 1.8564,
+      "step": 103
+    },
+    {
+      "epoch": 0.1925925925925926,
+      "grad_norm": 15.327783569924641,
+      "learning_rate": 3.209876543209876e-08,
+      "loss": 1.8837,
+      "step": 104
+    },
+    {
+      "epoch": 0.19444444444444445,
+      "grad_norm": 15.015824132699844,
+      "learning_rate": 3.2407407407407403e-08,
+      "loss": 1.8836,
+      "step": 105
+    },
+    {
+      "epoch": 0.1962962962962963,
+      "grad_norm": 16.079740454552617,
+      "learning_rate": 3.271604938271605e-08,
+      "loss": 1.8739,
+      "step": 106
+    },
+    {
+      "epoch": 0.19814814814814816,
+      "grad_norm": 15.469412631203738,
+      "learning_rate": 3.302469135802469e-08,
+      "loss": 1.8541,
+      "step": 107
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 15.148696319429204,
+      "learning_rate": 3.333333333333333e-08,
+      "loss": 1.8431,
+      "step": 108
+    },
+    {
+      "epoch": 0.20185185185185187,
+      "grad_norm": 15.167549682839613,
+      "learning_rate": 3.364197530864197e-08,
+      "loss": 1.8696,
+      "step": 109
+    },
+    {
+      "epoch": 0.2037037037037037,
+      "grad_norm": 15.555184045371803,
+      "learning_rate": 3.3950617283950614e-08,
+      "loss": 1.861,
+      "step": 110
+    },
+    {
+      "epoch": 0.20555555555555555,
+      "grad_norm": 15.649091725683697,
+      "learning_rate": 3.425925925925926e-08,
+      "loss": 1.8514,
+      "step": 111
+    },
+    {
+      "epoch": 0.2074074074074074,
+      "grad_norm": 15.162869233515563,
+      "learning_rate": 3.4567901234567895e-08,
+      "loss": 1.8784,
+      "step": 112
+    },
+    {
+      "epoch": 0.20925925925925926,
+      "grad_norm": 15.73018581787736,
+      "learning_rate": 3.487654320987654e-08,
+      "loss": 1.8587,
+      "step": 113
+    },
+    {
+      "epoch": 0.2111111111111111,
+      "grad_norm": 14.870954626807174,
+      "learning_rate": 3.518518518518518e-08,
+      "loss": 1.8494,
+      "step": 114
+    },
+    {
+      "epoch": 0.21296296296296297,
+      "grad_norm": 15.143299331711805,
+      "learning_rate": 3.5493827160493825e-08,
+      "loss": 1.8211,
+      "step": 115
+    },
+    {
+      "epoch": 0.21481481481481482,
+      "grad_norm": 14.94353813265894,
+      "learning_rate": 3.580246913580246e-08,
+      "loss": 1.8483,
+      "step": 116
+    },
+    {
+      "epoch": 0.21666666666666667,
+      "grad_norm": 15.097702504420209,
+      "learning_rate": 3.6111111111111106e-08,
+      "loss": 1.836,
+      "step": 117
+    },
+    {
+      "epoch": 0.21851851851851853,
+      "grad_norm": 14.714407171183284,
+      "learning_rate": 3.641975308641975e-08,
+      "loss": 1.8255,
+      "step": 118
+    },
+    {
+      "epoch": 0.22037037037037038,
+      "grad_norm": 14.923597793071034,
+      "learning_rate": 3.672839506172839e-08,
+      "loss": 1.84,
+      "step": 119
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 15.348980889915634,
+      "learning_rate": 3.7037037037037036e-08,
+      "loss": 1.8453,
+      "step": 120
+    },
+    {
+      "epoch": 0.22407407407407406,
+      "grad_norm": 15.227747932391384,
+      "learning_rate": 3.734567901234568e-08,
+      "loss": 1.8052,
+      "step": 121
+    },
+    {
+      "epoch": 0.22592592592592592,
+      "grad_norm": 14.746296661183148,
+      "learning_rate": 3.765432098765432e-08,
+      "loss": 1.8264,
+      "step": 122
+    },
+    {
+      "epoch": 0.22777777777777777,
+      "grad_norm": 16.131628947607084,
+      "learning_rate": 3.796296296296296e-08,
+      "loss": 1.8502,
+      "step": 123
+    },
+    {
+      "epoch": 0.22962962962962963,
+      "grad_norm": 15.319469480247056,
+      "learning_rate": 3.8271604938271604e-08,
+      "loss": 1.8288,
+      "step": 124
+    },
+    {
+      "epoch": 0.23148148148148148,
+      "grad_norm": 15.177552741329016,
+      "learning_rate": 3.858024691358025e-08,
+      "loss": 1.83,
+      "step": 125
+    },
+    {
+      "epoch": 0.23333333333333334,
+      "grad_norm": 14.83299348919801,
+      "learning_rate": 3.888888888888889e-08,
+      "loss": 1.8164,
+      "step": 126
+    },
+    {
+      "epoch": 0.2351851851851852,
+      "grad_norm": 14.75159373435331,
+      "learning_rate": 3.9197530864197534e-08,
+      "loss": 1.8466,
+      "step": 127
+    },
+    {
+      "epoch": 0.23703703703703705,
+      "grad_norm": 14.804098646037755,
+      "learning_rate": 3.950617283950617e-08,
+      "loss": 1.8236,
+      "step": 128
+    },
+    {
+      "epoch": 0.2388888888888889,
+      "grad_norm": 14.709991449008557,
+      "learning_rate": 3.9814814814814815e-08,
+      "loss": 1.8148,
+      "step": 129
+    },
+    {
+      "epoch": 0.24074074074074073,
+      "grad_norm": 14.853414636656769,
+      "learning_rate": 4.012345679012346e-08,
+      "loss": 1.831,
+      "step": 130
+    },
+    {
+      "epoch": 0.24259259259259258,
+      "grad_norm": 15.651472651182399,
+      "learning_rate": 4.04320987654321e-08,
+      "loss": 1.8288,
+      "step": 131
+    },
+    {
+      "epoch": 0.24444444444444444,
+      "grad_norm": 15.965201353715011,
+      "learning_rate": 4.074074074074074e-08,
+      "loss": 1.8194,
+      "step": 132
+    },
+    {
+      "epoch": 0.2462962962962963,
+      "grad_norm": 15.18102590434218,
+      "learning_rate": 4.104938271604938e-08,
+      "loss": 1.8254,
+      "step": 133
+    },
+    {
+      "epoch": 0.24814814814814815,
+      "grad_norm": 14.932468681124538,
+      "learning_rate": 4.1358024691358026e-08,
+      "loss": 1.8061,
+      "step": 134
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 15.397229031460373,
+      "learning_rate": 4.166666666666667e-08,
+      "loss": 1.8223,
+      "step": 135
+    },
+    {
+      "epoch": 0.2518518518518518,
+      "grad_norm": 14.98960680651719,
+      "learning_rate": 4.1975308641975306e-08,
+      "loss": 1.8094,
+      "step": 136
+    },
+    {
+      "epoch": 0.2537037037037037,
+      "grad_norm": 14.814808212375743,
+      "learning_rate": 4.228395061728395e-08,
+      "loss": 1.8249,
+      "step": 137
+    },
+    {
+      "epoch": 0.25555555555555554,
+      "grad_norm": 15.486470872876474,
+      "learning_rate": 4.259259259259259e-08,
+      "loss": 1.8442,
+      "step": 138
+    },
+    {
+      "epoch": 0.2574074074074074,
+      "grad_norm": 15.689354082865808,
+      "learning_rate": 4.2901234567901237e-08,
+      "loss": 1.8146,
+      "step": 139
+    },
+    {
+      "epoch": 0.25925925925925924,
+      "grad_norm": 14.83414161988492,
+      "learning_rate": 4.3209876543209874e-08,
+      "loss": 1.8466,
+      "step": 140
+    },
+    {
+      "epoch": 0.2611111111111111,
+      "grad_norm": 17.77942177387046,
+      "learning_rate": 4.351851851851852e-08,
+      "loss": 1.8304,
+      "step": 141
+    },
+    {
+      "epoch": 0.26296296296296295,
+      "grad_norm": 14.836118142473255,
+      "learning_rate": 4.382716049382716e-08,
+      "loss": 1.8189,
+      "step": 142
+    },
+    {
+      "epoch": 0.26481481481481484,
+      "grad_norm": 15.30036179717283,
+      "learning_rate": 4.4135802469135804e-08,
+      "loss": 1.8343,
+      "step": 143
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 14.738970792481036,
+      "learning_rate": 4.444444444444444e-08,
+      "loss": 1.8143,
+      "step": 144
+    },
+    {
+      "epoch": 0.26851851851851855,
+      "grad_norm": 14.542012053556816,
+      "learning_rate": 4.4753086419753084e-08,
+      "loss": 1.8194,
+      "step": 145
+    },
+    {
+      "epoch": 0.27037037037037037,
+      "grad_norm": 14.484645134463825,
+      "learning_rate": 4.506172839506173e-08,
+      "loss": 1.8177,
+      "step": 146
+    },
+    {
+      "epoch": 0.2722222222222222,
+      "grad_norm": 15.685240605752165,
+      "learning_rate": 4.537037037037037e-08,
+      "loss": 1.8179,
+      "step": 147
+    },
+    {
+      "epoch": 0.2740740740740741,
+      "grad_norm": 14.990947152614467,
+      "learning_rate": 4.567901234567901e-08,
+      "loss": 1.8319,
+      "step": 148
+    },
+    {
+      "epoch": 0.2759259259259259,
+      "grad_norm": 15.451965202760704,
+      "learning_rate": 4.598765432098765e-08,
+      "loss": 1.8357,
+      "step": 149
+    },
+    {
+      "epoch": 0.2777777777777778,
+      "grad_norm": 15.204525042476222,
+      "learning_rate": 4.6296296296296295e-08,
+      "loss": 1.8303,
+      "step": 150
+    },
+    {
+      "epoch": 0.2796296296296296,
+      "grad_norm": 15.104514218766667,
+      "learning_rate": 4.660493827160494e-08,
+      "loss": 1.7974,
+      "step": 151
+    },
+    {
+      "epoch": 0.2814814814814815,
+      "grad_norm": 15.105548365392572,
+      "learning_rate": 4.6913580246913576e-08,
+      "loss": 1.8249,
+      "step": 152
+    },
+    {
+      "epoch": 0.2833333333333333,
+      "grad_norm": 15.747532055963093,
+      "learning_rate": 4.722222222222222e-08,
+      "loss": 1.8008,
+      "step": 153
+    },
+    {
+      "epoch": 0.2851851851851852,
+      "grad_norm": 15.353991895527601,
+      "learning_rate": 4.753086419753086e-08,
+      "loss": 1.8251,
+      "step": 154
+    },
+    {
+      "epoch": 0.28703703703703703,
+      "grad_norm": 15.209788532634054,
+      "learning_rate": 4.7839506172839506e-08,
+      "loss": 1.809,
+      "step": 155
+    },
+    {
+      "epoch": 0.28888888888888886,
+      "grad_norm": 15.80173994023578,
+      "learning_rate": 4.814814814814814e-08,
+      "loss": 1.8013,
+      "step": 156
+    },
+    {
+      "epoch": 0.29074074074074074,
+      "grad_norm": 16.048047139021342,
+      "learning_rate": 4.845679012345679e-08,
+      "loss": 1.7916,
+      "step": 157
+    },
+    {
+      "epoch": 0.29259259259259257,
+      "grad_norm": 15.962081332069088,
+      "learning_rate": 4.876543209876543e-08,
+      "loss": 1.7817,
+      "step": 158
+    },
+    {
+      "epoch": 0.29444444444444445,
+      "grad_norm": 16.466556952911027,
+      "learning_rate": 4.9074074074074074e-08,
+      "loss": 1.7185,
+      "step": 159
+    },
+    {
+      "epoch": 0.2962962962962963,
+      "grad_norm": 16.223198762951395,
+      "learning_rate": 4.938271604938271e-08,
+      "loss": 1.7481,
+      "step": 160
+    },
+    {
+      "epoch": 0.29814814814814816,
+      "grad_norm": 15.887450875336,
+      "learning_rate": 4.9691358024691354e-08,
+      "loss": 1.7135,
+      "step": 161
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 15.365124448323291,
+      "learning_rate": 5e-08,
+      "loss": 1.7141,
+      "step": 162
+    },
+    {
+      "epoch": 0.30185185185185187,
+      "grad_norm": 16.476719186450662,
+      "learning_rate": 5.030864197530864e-08,
+      "loss": 1.712,
+      "step": 163
+    },
+    {
+      "epoch": 0.3037037037037037,
+      "grad_norm": 15.080000212695525,
+      "learning_rate": 5.0617283950617285e-08,
+      "loss": 1.7059,
+      "step": 164
+    },
+    {
+      "epoch": 0.3055555555555556,
+      "grad_norm": 16.143726335461462,
+      "learning_rate": 5.092592592592593e-08,
+      "loss": 1.6906,
+      "step": 165
+    },
+    {
+      "epoch": 0.3074074074074074,
+      "grad_norm": 14.6141471939126,
+      "learning_rate": 5.123456790123457e-08,
+      "loss": 1.7233,
+      "step": 166
+    },
+    {
+      "epoch": 0.30925925925925923,
+      "grad_norm": 15.057721002857242,
+      "learning_rate": 5.15432098765432e-08,
+      "loss": 1.6839,
+      "step": 167
+    },
+    {
+      "epoch": 0.3111111111111111,
+      "grad_norm": 14.889380937788225,
+      "learning_rate": 5.1851851851851846e-08,
+      "loss": 1.6885,
+      "step": 168
+    },
+    {
+      "epoch": 0.31296296296296294,
+      "grad_norm": 14.851576464414487,
+      "learning_rate": 5.216049382716049e-08,
+      "loss": 1.6857,
+      "step": 169
+    },
+    {
+      "epoch": 0.3148148148148148,
+      "grad_norm": 15.295223573712184,
+      "learning_rate": 5.246913580246913e-08,
+      "loss": 1.7006,
+      "step": 170
+    },
+    {
+      "epoch": 0.31666666666666665,
+      "grad_norm": 14.803728888120778,
+      "learning_rate": 5.2777777777777776e-08,
+      "loss": 1.6766,
+      "step": 171
+    },
+    {
+      "epoch": 0.31851851851851853,
+      "grad_norm": 15.086524537977429,
+      "learning_rate": 5.308641975308642e-08,
+      "loss": 1.6815,
+      "step": 172
+    },
+    {
+      "epoch": 0.32037037037037036,
+      "grad_norm": 15.802758931254427,
+      "learning_rate": 5.339506172839506e-08,
+      "loss": 1.6949,
+      "step": 173
+    },
+    {
+      "epoch": 0.32222222222222224,
+      "grad_norm": 15.247930974320342,
+      "learning_rate": 5.3703703703703707e-08,
+      "loss": 1.6797,
+      "step": 174
+    },
+    {
+      "epoch": 0.32407407407407407,
+      "grad_norm": 15.27233097586875,
+      "learning_rate": 5.401234567901234e-08,
+      "loss": 1.6789,
+      "step": 175
+    },
+    {
+      "epoch": 0.32592592592592595,
+      "grad_norm": 15.380931053883348,
+      "learning_rate": 5.432098765432098e-08,
+      "loss": 1.6681,
+      "step": 176
+    },
+    {
+      "epoch": 0.3277777777777778,
+      "grad_norm": 14.755118321502803,
+      "learning_rate": 5.4629629629629624e-08,
+      "loss": 1.6586,
+      "step": 177
+    },
+    {
+      "epoch": 0.3296296296296296,
+      "grad_norm": 15.535643296681924,
+      "learning_rate": 5.493827160493827e-08,
+      "loss": 1.6645,
+      "step": 178
+    },
+    {
+      "epoch": 0.3314814814814815,
+      "grad_norm": 15.16851840528597,
+      "learning_rate": 5.524691358024691e-08,
+      "loss": 1.6719,
+      "step": 179
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 15.071491861044388,
+      "learning_rate": 5.5555555555555555e-08,
+      "loss": 1.6874,
+      "step": 180
+    },
+    {
+      "epoch": 0.3351851851851852,
+      "grad_norm": 15.0838502603752,
+      "learning_rate": 5.58641975308642e-08,
+      "loss": 1.6684,
+      "step": 181
+    },
+    {
+      "epoch": 0.337037037037037,
+      "grad_norm": 15.035761965661147,
+      "learning_rate": 5.617283950617284e-08,
+      "loss": 1.6646,
+      "step": 182
+    },
+    {
+      "epoch": 0.3388888888888889,
+      "grad_norm": 15.3909017297271,
+      "learning_rate": 5.648148148148147e-08,
+      "loss": 1.6647,
+      "step": 183
+    },
+    {
+      "epoch": 0.34074074074074073,
+      "grad_norm": 15.389123026027889,
+      "learning_rate": 5.6790123456790115e-08,
+      "loss": 1.6439,
+      "step": 184
+    },
+    {
+      "epoch": 0.3425925925925926,
+      "grad_norm": 15.538843832757056,
+      "learning_rate": 5.709876543209876e-08,
+      "loss": 1.6501,
+      "step": 185
+    },
+    {
+      "epoch": 0.34444444444444444,
+      "grad_norm": 16.121666139011364,
+      "learning_rate": 5.74074074074074e-08,
+      "loss": 1.6504,
+      "step": 186
+    },
+    {
+      "epoch": 0.34629629629629627,
+      "grad_norm": 15.597442964448913,
+      "learning_rate": 5.7716049382716046e-08,
+      "loss": 1.6622,
+      "step": 187
+    },
+    {
+      "epoch": 0.34814814814814815,
+      "grad_norm": 16.13349804618981,
+      "learning_rate": 5.802469135802469e-08,
+      "loss": 1.6236,
+      "step": 188
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 15.863696027424748,
+      "learning_rate": 5.833333333333333e-08,
+      "loss": 1.642,
+      "step": 189
+    },
+    {
+      "epoch": 0.35185185185185186,
+      "grad_norm": 16.4252638158174,
+      "learning_rate": 5.8641975308641976e-08,
+      "loss": 1.6307,
+      "step": 190
+    },
+    {
+      "epoch": 0.3537037037037037,
+      "grad_norm": 15.841001838915014,
+      "learning_rate": 5.895061728395061e-08,
+      "loss": 1.6103,
+      "step": 191
+    },
+    {
+      "epoch": 0.35555555555555557,
+      "grad_norm": 15.673573923427517,
+      "learning_rate": 5.925925925925925e-08,
+      "loss": 1.602,
+      "step": 192
+    },
+    {
+      "epoch": 0.3574074074074074,
+      "grad_norm": 16.441184669156055,
+      "learning_rate": 5.9567901234567894e-08,
+      "loss": 1.602,
+      "step": 193
+    },
+    {
+      "epoch": 0.3592592592592593,
+      "grad_norm": 16.36704864970962,
+      "learning_rate": 5.987654320987654e-08,
+      "loss": 1.6094,
+      "step": 194
+    },
+    {
+      "epoch": 0.3611111111111111,
+      "grad_norm": 16.143726546373465,
+      "learning_rate": 6.018518518518519e-08,
+      "loss": 1.6203,
+      "step": 195
+    },
+    {
+      "epoch": 0.362962962962963,
+      "grad_norm": 15.767462442066334,
+      "learning_rate": 6.049382716049382e-08,
+      "loss": 1.5872,
+      "step": 196
+    },
+    {
+      "epoch": 0.3648148148148148,
+      "grad_norm": 16.183360356848596,
+      "learning_rate": 6.080246913580247e-08,
+      "loss": 1.6067,
+      "step": 197
+    },
+    {
+      "epoch": 0.36666666666666664,
+      "grad_norm": 15.6568073944227,
+      "learning_rate": 6.111111111111111e-08,
+      "loss": 1.5832,
+      "step": 198
+    },
+    {
+      "epoch": 0.3685185185185185,
+      "grad_norm": 15.031914020275904,
+      "learning_rate": 6.141975308641975e-08,
+      "loss": 1.5804,
+      "step": 199
+    },
+    {
+      "epoch": 0.37037037037037035,
+      "grad_norm": 13.901553800473613,
+      "learning_rate": 6.172839506172839e-08,
+      "loss": 1.5736,
+      "step": 200
+    },
+    {
+      "epoch": 0.37222222222222223,
+      "grad_norm": 13.529862363036859,
+      "learning_rate": 6.203703703703704e-08,
+      "loss": 1.5735,
+      "step": 201
+    },
+    {
+      "epoch": 0.37407407407407406,
+      "grad_norm": 13.769584697382156,
+      "learning_rate": 6.234567901234567e-08,
+      "loss": 1.5502,
+      "step": 202
+    },
+    {
+      "epoch": 0.37592592592592594,
+      "grad_norm": 13.283164609686715,
+      "learning_rate": 6.265432098765432e-08,
+      "loss": 1.5369,
+      "step": 203
+    },
+    {
+      "epoch": 0.37777777777777777,
+      "grad_norm": 13.219835975939468,
+      "learning_rate": 6.296296296296296e-08,
+      "loss": 1.5405,
+      "step": 204
+    },
+    {
+      "epoch": 0.37962962962962965,
+      "grad_norm": 12.823211358025947,
+      "learning_rate": 6.327160493827161e-08,
+      "loss": 1.5308,
+      "step": 205
+    },
+    {
+      "epoch": 0.3814814814814815,
+      "grad_norm": 12.77571724349778,
+      "learning_rate": 6.358024691358025e-08,
+      "loss": 1.51,
+      "step": 206
+    },
+    {
+      "epoch": 0.38333333333333336,
+      "grad_norm": 12.684182910096494,
+      "learning_rate": 6.388888888888888e-08,
+      "loss": 1.5063,
+      "step": 207
+    },
+    {
+      "epoch": 0.3851851851851852,
+      "grad_norm": 11.306595481582432,
+      "learning_rate": 6.419753086419752e-08,
+      "loss": 1.5226,
+      "step": 208
+    },
+    {
+      "epoch": 0.387037037037037,
+      "grad_norm": 11.70198984938417,
+      "learning_rate": 6.450617283950617e-08,
+      "loss": 1.5003,
+      "step": 209
+    },
+    {
+      "epoch": 0.3888888888888889,
+      "grad_norm": 12.39623859805833,
+      "learning_rate": 6.481481481481481e-08,
+      "loss": 1.506,
+      "step": 210
+    },
+    {
+      "epoch": 0.3907407407407407,
+      "grad_norm": 11.180344111163487,
+      "learning_rate": 6.512345679012346e-08,
+      "loss": 1.4859,
+      "step": 211
+    },
+    {
+      "epoch": 0.3925925925925926,
+      "grad_norm": 10.71706262975008,
+      "learning_rate": 6.54320987654321e-08,
+      "loss": 1.4878,
+      "step": 212
+    },
+    {
+      "epoch": 0.39444444444444443,
+      "grad_norm": 10.475139159343728,
+      "learning_rate": 6.574074074074074e-08,
+      "loss": 1.477,
+      "step": 213
+    },
+    {
+      "epoch": 0.3962962962962963,
+      "grad_norm": 10.935818149952677,
+      "learning_rate": 6.604938271604938e-08,
+      "loss": 1.4568,
+      "step": 214
+    },
+    {
+      "epoch": 0.39814814814814814,
+      "grad_norm": 10.847247788731964,
+      "learning_rate": 6.635802469135802e-08,
+      "loss": 1.4365,
+      "step": 215
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.499892118117987,
+      "learning_rate": 6.666666666666665e-08,
+      "loss": 1.4678,
+      "step": 216
+    },
+    {
+      "epoch": 0.40185185185185185,
+      "grad_norm": 9.807924411811511,
+      "learning_rate": 6.69753086419753e-08,
+      "loss": 1.4558,
+      "step": 217
+    },
+    {
+      "epoch": 0.40370370370370373,
+      "grad_norm": 10.627392105783526,
+      "learning_rate": 6.728395061728394e-08,
+      "loss": 1.4156,
+      "step": 218
+    },
+    {
+      "epoch": 0.40555555555555556,
+      "grad_norm": 10.453639635708425,
+      "learning_rate": 6.759259259259259e-08,
+      "loss": 1.4222,
+      "step": 219
+    },
+    {
+      "epoch": 0.4074074074074074,
+      "grad_norm": 10.436967114672623,
+      "learning_rate": 6.790123456790123e-08,
+      "loss": 1.4234,
+      "step": 220
+    },
+    {
+      "epoch": 0.40925925925925927,
+      "grad_norm": 11.502093278036753,
+      "learning_rate": 6.820987654320988e-08,
+      "loss": 1.4168,
+      "step": 221
+    },
+    {
+      "epoch": 0.4111111111111111,
+      "grad_norm": 12.109663297773258,
+      "learning_rate": 6.851851851851852e-08,
+      "loss": 1.364,
+      "step": 222
+    },
+    {
+      "epoch": 0.412962962962963,
+      "grad_norm": 10.288026845785605,
+      "learning_rate": 6.882716049382715e-08,
+      "loss": 1.4177,
+      "step": 223
+    },
+    {
+      "epoch": 0.4148148148148148,
+      "grad_norm": 10.420630646475221,
+      "learning_rate": 6.913580246913579e-08,
+      "loss": 1.4085,
+      "step": 224
+    },
+    {
+      "epoch": 0.4166666666666667,
+      "grad_norm": 9.784391847322885,
+      "learning_rate": 6.944444444444444e-08,
+      "loss": 1.405,
+      "step": 225
+    },
+    {
+      "epoch": 0.4185185185185185,
+      "grad_norm": 9.671341865243887,
+      "learning_rate": 6.975308641975308e-08,
+      "loss": 1.3689,
+      "step": 226
+    },
+    {
+      "epoch": 0.4203703703703704,
+      "grad_norm": 10.416339402031978,
+      "learning_rate": 7.006172839506173e-08,
+      "loss": 1.3843,
+      "step": 227
+    },
+    {
+      "epoch": 0.4222222222222222,
+      "grad_norm": 9.67646153902481,
+      "learning_rate": 7.037037037037036e-08,
+      "loss": 1.3711,
+      "step": 228
+    },
+    {
+      "epoch": 0.42407407407407405,
+      "grad_norm": 9.224987510519476,
+      "learning_rate": 7.067901234567901e-08,
+      "loss": 1.3723,
+      "step": 229
+    },
+    {
+      "epoch": 0.42592592592592593,
+      "grad_norm": 9.468803750953535,
+      "learning_rate": 7.098765432098765e-08,
+      "loss": 1.3412,
+      "step": 230
+    },
+    {
+      "epoch": 0.42777777777777776,
+      "grad_norm": 9.347082531477147,
+      "learning_rate": 7.129629629629629e-08,
+      "loss": 1.3381,
+      "step": 231
+    },
+    {
+      "epoch": 0.42962962962962964,
+      "grad_norm": 8.898812306680487,
+      "learning_rate": 7.160493827160492e-08,
+      "loss": 1.3566,
+      "step": 232
+    },
+    {
+      "epoch": 0.43148148148148147,
+      "grad_norm": 9.416233114849978,
+      "learning_rate": 7.191358024691357e-08,
+      "loss": 1.3309,
+      "step": 233
+    },
+    {
+      "epoch": 0.43333333333333335,
+      "grad_norm": 9.378481244087665,
+      "learning_rate": 7.222222222222221e-08,
+      "loss": 1.2955,
+      "step": 234
+    },
+    {
+      "epoch": 0.4351851851851852,
+      "grad_norm": 9.590341847604815,
+      "learning_rate": 7.253086419753086e-08,
+      "loss": 1.2931,
+      "step": 235
+    },
+    {
+      "epoch": 0.43703703703703706,
+      "grad_norm": 9.204696036060932,
+      "learning_rate": 7.28395061728395e-08,
+      "loss": 1.3017,
+      "step": 236
+    },
+    {
+      "epoch": 0.4388888888888889,
+      "grad_norm": 9.107752164095162,
+      "learning_rate": 7.314814814814815e-08,
+      "loss": 1.2897,
+      "step": 237
+    },
+    {
+      "epoch": 0.44074074074074077,
+      "grad_norm": 9.003981478035788,
+      "learning_rate": 7.345679012345679e-08,
+      "loss": 1.3058,
+      "step": 238
+    },
+    {
+      "epoch": 0.4425925925925926,
+      "grad_norm": 9.058466378046083,
+      "learning_rate": 7.376543209876542e-08,
+      "loss": 1.2834,
+      "step": 239
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 8.705122307100883,
+      "learning_rate": 7.407407407407407e-08,
+      "loss": 1.2818,
+      "step": 240
+    },
+    {
+      "epoch": 0.4462962962962963,
+      "grad_norm": 9.205049637998936,
+      "learning_rate": 7.438271604938271e-08,
+      "loss": 1.2662,
+      "step": 241
+    },
+    {
+      "epoch": 0.44814814814814813,
+      "grad_norm": 8.785421682988352,
+      "learning_rate": 7.469135802469136e-08,
+      "loss": 1.2836,
+      "step": 242
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.191977042098788,
+      "learning_rate": 7.5e-08,
+      "loss": 1.2499,
+      "step": 243
+    },
+    {
+      "epoch": 0.45185185185185184,
+      "grad_norm": 8.598255115042189,
+      "learning_rate": 7.530864197530863e-08,
+      "loss": 1.2633,
+      "step": 244
+    },
+    {
+      "epoch": 0.4537037037037037,
+      "grad_norm": 9.512153148550793,
+      "learning_rate": 7.561728395061728e-08,
+      "loss": 1.2379,
+      "step": 245
+    },
+    {
+      "epoch": 0.45555555555555555,
+      "grad_norm": 8.724604329013456,
+      "learning_rate": 7.592592592592592e-08,
+      "loss": 1.2412,
+      "step": 246
+    },
+    {
+      "epoch": 0.45740740740740743,
+      "grad_norm": 7.934232264621604,
+      "learning_rate": 7.623456790123457e-08,
+      "loss": 1.2831,
+      "step": 247
+    },
+    {
+      "epoch": 0.45925925925925926,
+      "grad_norm": 7.935473144855036,
+      "learning_rate": 7.654320987654321e-08,
+      "loss": 1.2201,
+      "step": 248
+    },
+    {
+      "epoch": 0.46111111111111114,
+      "grad_norm": 8.099377891087682,
+      "learning_rate": 7.685185185185184e-08,
+      "loss": 1.2542,
+      "step": 249
+    },
+    {
+      "epoch": 0.46296296296296297,
+      "grad_norm": 8.406755975217708,
+      "learning_rate": 7.71604938271605e-08,
+      "loss": 1.1994,
+      "step": 250
+    },
+    {
+      "epoch": 0.4648148148148148,
+      "grad_norm": 8.201107335726807,
+      "learning_rate": 7.746913580246913e-08,
+      "loss": 1.2031,
+      "step": 251
+    },
+    {
+      "epoch": 0.4666666666666667,
+      "grad_norm": 8.196934615673412,
+      "learning_rate": 7.777777777777778e-08,
+      "loss": 1.1864,
+      "step": 252
+    },
+    {
+      "epoch": 0.4685185185185185,
+      "grad_norm": 7.230676191162654,
+      "learning_rate": 7.808641975308642e-08,
+      "loss": 1.2167,
+      "step": 253
+    },
+    {
+      "epoch": 0.4703703703703704,
+      "grad_norm": 7.426414216347639,
+      "learning_rate": 7.839506172839507e-08,
+      "loss": 1.1935,
+      "step": 254
+    },
+    {
+      "epoch": 0.4722222222222222,
+      "grad_norm": 7.243389599418745,
+      "learning_rate": 7.87037037037037e-08,
+      "loss": 1.235,
+      "step": 255
+    },
+    {
+      "epoch": 0.4740740740740741,
+      "grad_norm": 7.634374622663417,
+      "learning_rate": 7.901234567901234e-08,
+      "loss": 1.1815,
+      "step": 256
+    },
+    {
+      "epoch": 0.4759259259259259,
+      "grad_norm": 6.941994252703506,
+      "learning_rate": 7.932098765432098e-08,
+      "loss": 1.2203,
+      "step": 257
+    },
+    {
+      "epoch": 0.4777777777777778,
+      "grad_norm": 6.666200938968774,
+      "learning_rate": 7.962962962962963e-08,
+      "loss": 1.2104,
+      "step": 258
+    },
+    {
+      "epoch": 0.47962962962962963,
+      "grad_norm": 7.164672262499252,
+      "learning_rate": 7.993827160493827e-08,
+      "loss": 1.1653,
+      "step": 259
+    },
+    {
+      "epoch": 0.48148148148148145,
+      "grad_norm": 7.227692093630713,
+      "learning_rate": 8.024691358024692e-08,
+      "loss": 1.1453,
+      "step": 260
+    },
+    {
+      "epoch": 0.48333333333333334,
+      "grad_norm": 6.7590989140523465,
+      "learning_rate": 8.055555555555555e-08,
+      "loss": 1.1861,
+      "step": 261
+    },
+    {
+      "epoch": 0.48518518518518516,
+      "grad_norm": 6.595710379020176,
+      "learning_rate": 8.08641975308642e-08,
+      "loss": 1.137,
+      "step": 262
+    },
+    {
+      "epoch": 0.48703703703703705,
+      "grad_norm": 6.367587250699385,
+      "learning_rate": 8.117283950617284e-08,
+      "loss": 1.1767,
+      "step": 263
+    },
+    {
+      "epoch": 0.4888888888888889,
+      "grad_norm": 6.2427486232307094,
+      "learning_rate": 8.148148148148148e-08,
+      "loss": 1.1324,
+      "step": 264
+    },
+    {
+      "epoch": 0.49074074074074076,
+      "grad_norm": 6.079506123745785,
+      "learning_rate": 8.179012345679011e-08,
+      "loss": 1.1635,
+      "step": 265
+    },
+    {
+      "epoch": 0.4925925925925926,
+      "grad_norm": 6.5345425826060195,
+      "learning_rate": 8.209876543209876e-08,
+      "loss": 1.1588,
+      "step": 266
+    },
+    {
+      "epoch": 0.49444444444444446,
+      "grad_norm": 5.767447481234529,
+      "learning_rate": 8.24074074074074e-08,
+      "loss": 1.1006,
+      "step": 267
+    },
+    {
+      "epoch": 0.4962962962962963,
+      "grad_norm": 5.503571960558135,
+      "learning_rate": 8.271604938271605e-08,
+      "loss": 1.1806,
+      "step": 268
+    },
+    {
+      "epoch": 0.4981481481481482,
+      "grad_norm": 5.413380853058772,
+      "learning_rate": 8.302469135802469e-08,
+      "loss": 1.1344,
+      "step": 269
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.1633535564884285,
+      "learning_rate": 8.333333333333334e-08,
+      "loss": 1.1065,
+      "step": 270
+    },
+    {
+      "epoch": 0.5018518518518519,
+      "grad_norm": 5.096040714664006,
+      "learning_rate": 8.364197530864198e-08,
+      "loss": 1.1796,
+      "step": 271
+    },
+    {
+      "epoch": 0.5037037037037037,
+      "grad_norm": 5.148802562497922,
+      "learning_rate": 8.395061728395061e-08,
+      "loss": 1.0927,
+      "step": 272
+    },
+    {
+      "epoch": 0.5055555555555555,
+      "grad_norm": 4.819744826698795,
+      "learning_rate": 8.425925925925925e-08,
+      "loss": 1.1802,
+      "step": 273
+    },
+    {
+      "epoch": 0.5074074074074074,
+      "grad_norm": 6.940152223659339,
+      "learning_rate": 8.45679012345679e-08,
+      "loss": 1.1431,
+      "step": 274
+    },
+    {
+      "epoch": 0.5092592592592593,
+      "grad_norm": 4.8899229426671935,
+      "learning_rate": 8.487654320987654e-08,
+      "loss": 1.1424,
+      "step": 275
+    },
+    {
+      "epoch": 0.5111111111111111,
+      "grad_norm": 4.633461899062297,
+      "learning_rate": 8.518518518518519e-08,
+      "loss": 1.1732,
+      "step": 276
+    },
+    {
+      "epoch": 0.512962962962963,
+      "grad_norm": 5.366341922696669,
+      "learning_rate": 8.549382716049382e-08,
+      "loss": 1.1262,
+      "step": 277
+    },
+    {
+      "epoch": 0.5148148148148148,
+      "grad_norm": 5.407296170169389,
+      "learning_rate": 8.580246913580247e-08,
+      "loss": 1.1421,
+      "step": 278
+    },
+    {
+      "epoch": 0.5166666666666667,
+      "grad_norm": 4.64055400633339,
+      "learning_rate": 8.611111111111111e-08,
+      "loss": 1.1244,
+      "step": 279
+    },
+    {
+      "epoch": 0.5185185185185185,
+      "grad_norm": 4.790759190356316,
+      "learning_rate": 8.641975308641975e-08,
+      "loss": 1.0795,
+      "step": 280
+    },
+    {
+      "epoch": 0.5203703703703704,
+      "grad_norm": 4.508466333612143,
+      "learning_rate": 8.672839506172838e-08,
+      "loss": 1.1168,
+      "step": 281
+    },
+    {
+      "epoch": 0.5222222222222223,
+      "grad_norm": 4.601535680356002,
+      "learning_rate": 8.703703703703703e-08,
+      "loss": 1.0751,
+      "step": 282
+    },
+    {
+      "epoch": 0.524074074074074,
+      "grad_norm": 4.356061978717215,
+      "learning_rate": 8.734567901234567e-08,
+      "loss": 1.1383,
+      "step": 283
+    },
+    {
+      "epoch": 0.5259259259259259,
+      "grad_norm": 4.739645434419619,
+      "learning_rate": 8.765432098765432e-08,
+      "loss": 1.0895,
+      "step": 284
+    },
+    {
+      "epoch": 0.5277777777777778,
+      "grad_norm": 4.557295878083224,
+      "learning_rate": 8.796296296296296e-08,
+      "loss": 1.0586,
+      "step": 285
+    },
+    {
+      "epoch": 0.5296296296296297,
+      "grad_norm": 4.983263941075996,
+      "learning_rate": 8.827160493827161e-08,
+      "loss": 1.1093,
+      "step": 286
+    },
+    {
+      "epoch": 0.5314814814814814,
+      "grad_norm": 4.322416983003166,
+      "learning_rate": 8.858024691358025e-08,
+      "loss": 1.0785,
+      "step": 287
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 4.421637263273455,
+      "learning_rate": 8.888888888888888e-08,
+      "loss": 1.1444,
+      "step": 288
+    },
+    {
+      "epoch": 0.5351851851851852,
+      "grad_norm": 4.222520171046782,
+      "learning_rate": 8.919753086419752e-08,
+      "loss": 1.1139,
+      "step": 289
+    },
+    {
+      "epoch": 0.5370370370370371,
+      "grad_norm": 4.283460517774991,
+      "learning_rate": 8.950617283950617e-08,
+      "loss": 1.1247,
+      "step": 290
+    },
+    {
+      "epoch": 0.5388888888888889,
+      "grad_norm": 4.182758065374512,
+      "learning_rate": 8.98148148148148e-08,
+      "loss": 1.0799,
+      "step": 291
+    },
+    {
+      "epoch": 0.5407407407407407,
+      "grad_norm": 4.475415379775168,
+      "learning_rate": 9.012345679012346e-08,
+      "loss": 1.0955,
+      "step": 292
+    },
+    {
+      "epoch": 0.5425925925925926,
+      "grad_norm": 4.568332544127409,
+      "learning_rate": 9.043209876543209e-08,
+      "loss": 1.0592,
+      "step": 293
+    },
+    {
+      "epoch": 0.5444444444444444,
+      "grad_norm": 4.117536951160185,
+      "learning_rate": 9.074074074074074e-08,
+      "loss": 1.1324,
+      "step": 294
+    },
+    {
+      "epoch": 0.5462962962962963,
+      "grad_norm": 4.316537569625379,
+      "learning_rate": 9.104938271604938e-08,
+      "loss": 1.107,
+      "step": 295
+    },
+    {
+      "epoch": 0.5481481481481482,
+      "grad_norm": 4.1795506768589155,
+      "learning_rate": 9.135802469135802e-08,
+      "loss": 1.0782,
+      "step": 296
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.359117337896192,
+      "learning_rate": 9.166666666666665e-08,
+      "loss": 1.0395,
+      "step": 297
+    },
+    {
+      "epoch": 0.5518518518518518,
+      "grad_norm": 4.11711181502351,
+      "learning_rate": 9.19753086419753e-08,
+      "loss": 1.0698,
+      "step": 298
+    },
+    {
+      "epoch": 0.5537037037037037,
+      "grad_norm": 6.256660607973502,
+      "learning_rate": 9.228395061728394e-08,
+      "loss": 1.0683,
+      "step": 299
+    },
+    {
+      "epoch": 0.5555555555555556,
+      "grad_norm": 4.040085947010745,
+      "learning_rate": 9.259259259259259e-08,
+      "loss": 1.055,
+      "step": 300
+    },
+    {
+      "epoch": 0.5574074074074075,
+      "grad_norm": 4.0753999882391385,
+      "learning_rate": 9.290123456790123e-08,
+      "loss": 1.0595,
+      "step": 301
+    },
+    {
+      "epoch": 0.5592592592592592,
+      "grad_norm": 3.8715528015682525,
+      "learning_rate": 9.320987654320988e-08,
+      "loss": 1.0637,
+      "step": 302
+    },
+    {
+      "epoch": 0.5611111111111111,
+      "grad_norm": 4.147891280688802,
+      "learning_rate": 9.351851851851851e-08,
+      "loss": 1.0428,
+      "step": 303
+    },
+    {
+      "epoch": 0.562962962962963,
+      "grad_norm": 4.475004281259303,
+      "learning_rate": 9.382716049382715e-08,
+      "loss": 1.0994,
+      "step": 304
+    },
+    {
+      "epoch": 0.5648148148148148,
+      "grad_norm": 3.804844933042943,
+      "learning_rate": 9.413580246913579e-08,
+      "loss": 1.0936,
+      "step": 305
+    },
+    {
+      "epoch": 0.5666666666666667,
+      "grad_norm": 3.93283281444395,
+      "learning_rate": 9.444444444444444e-08,
+      "loss": 1.0874,
+      "step": 306
+    },
+    {
+      "epoch": 0.5685185185185185,
+      "grad_norm": 3.8545783040385437,
+      "learning_rate": 9.475308641975308e-08,
+      "loss": 1.0828,
+      "step": 307
+    },
+    {
+      "epoch": 0.5703703703703704,
+      "grad_norm": 3.879617339937958,
+      "learning_rate": 9.506172839506173e-08,
+      "loss": 1.0993,
+      "step": 308
+    },
+    {
+      "epoch": 0.5722222222222222,
+      "grad_norm": 3.853722588826157,
+      "learning_rate": 9.537037037037036e-08,
+      "loss": 1.0596,
+      "step": 309
+    },
+    {
+      "epoch": 0.5740740740740741,
+      "grad_norm": 4.4180795472700485,
+      "learning_rate": 9.567901234567901e-08,
+      "loss": 1.056,
+      "step": 310
+    },
+    {
+      "epoch": 0.575925925925926,
+      "grad_norm": 4.13332656152739,
+      "learning_rate": 9.598765432098765e-08,
+      "loss": 1.0385,
+      "step": 311
+    },
+    {
+      "epoch": 0.5777777777777777,
+      "grad_norm": 3.8698195381240224,
+      "learning_rate": 9.629629629629629e-08,
+      "loss": 1.0905,
+      "step": 312
+    },
+    {
+      "epoch": 0.5796296296296296,
+      "grad_norm": 3.7649247371434913,
+      "learning_rate": 9.660493827160492e-08,
+      "loss": 1.0244,
+      "step": 313
+    },
+    {
+      "epoch": 0.5814814814814815,
+      "grad_norm": 3.886218344399538,
+      "learning_rate": 9.691358024691357e-08,
+      "loss": 1.0756,
+      "step": 314
+    },
+    {
+      "epoch": 0.5833333333333334,
+      "grad_norm": 3.832552287760325,
+      "learning_rate": 9.722222222222221e-08,
+      "loss": 1.0114,
+      "step": 315
+    },
+    {
+      "epoch": 0.5851851851851851,
+      "grad_norm": 3.6146573722704267,
+      "learning_rate": 9.753086419753086e-08,
+      "loss": 1.0214,
+      "step": 316
+    },
+    {
+      "epoch": 0.587037037037037,
+      "grad_norm": 3.7899711682732704,
+      "learning_rate": 9.78395061728395e-08,
+      "loss": 1.0346,
+      "step": 317
+    },
+    {
+      "epoch": 0.5888888888888889,
+      "grad_norm": 3.8241333661862056,
+      "learning_rate": 9.814814814814815e-08,
+      "loss": 1.0269,
+      "step": 318
+    },
+    {
+      "epoch": 0.5907407407407408,
+      "grad_norm": 3.9222450614025743,
+      "learning_rate": 9.845679012345678e-08,
+      "loss": 1.036,
+      "step": 319
+    },
+    {
+      "epoch": 0.5925925925925926,
+      "grad_norm": 3.6406348375553756,
+      "learning_rate": 9.876543209876542e-08,
+      "loss": 1.0167,
+      "step": 320
+    },
+    {
+      "epoch": 0.5944444444444444,
+      "grad_norm": 3.6373113117333564,
+      "learning_rate": 9.907407407407407e-08,
+      "loss": 1.0365,
+      "step": 321
+    },
+    {
+      "epoch": 0.5962962962962963,
+      "grad_norm": 3.841999292920067,
+      "learning_rate": 9.938271604938271e-08,
+      "loss": 0.9952,
+      "step": 322
+    },
+    {
+      "epoch": 0.5981481481481481,
+      "grad_norm": 3.5190549270282756,
+      "learning_rate": 9.969135802469136e-08,
+      "loss": 1.0176,
+      "step": 323
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.9277792826794435,
+      "learning_rate": 1e-07,
+      "loss": 1.0132,
+      "step": 324
+    },
+    {
+      "epoch": 0.6018518518518519,
+      "grad_norm": 3.4991407608885914,
+      "learning_rate": 1.0030864197530863e-07,
+      "loss": 1.068,
+      "step": 325
+    },
+    {
+      "epoch": 0.6037037037037037,
+      "grad_norm": 3.500378688791865,
+      "learning_rate": 1.0061728395061728e-07,
+      "loss": 1.0703,
+      "step": 326
+    },
+    {
+      "epoch": 0.6055555555555555,
+      "grad_norm": 4.193052022493057,
+      "learning_rate": 1.0092592592592592e-07,
+      "loss": 0.9992,
+      "step": 327
+    },
+    {
+      "epoch": 0.6074074074074074,
+      "grad_norm": 4.091013817038656,
+      "learning_rate": 1.0123456790123457e-07,
+      "loss": 1.0896,
+      "step": 328
+    },
+    {
+      "epoch": 0.6092592592592593,
+      "grad_norm": 3.7112977635689233,
+      "learning_rate": 1.015432098765432e-07,
+      "loss": 0.984,
+      "step": 329
+    },
+    {
+      "epoch": 0.6111111111111112,
+      "grad_norm": 3.360409883844207,
+      "learning_rate": 1.0185185185185186e-07,
+      "loss": 1.035,
+      "step": 330
+    },
+    {
+      "epoch": 0.6129629629629629,
+      "grad_norm": 3.433642670718037,
+      "learning_rate": 1.021604938271605e-07,
+      "loss": 1.035,
+      "step": 331
+    },
+    {
+      "epoch": 0.6148148148148148,
+      "grad_norm": 3.212656245186043,
+      "learning_rate": 1.0246913580246914e-07,
+      "loss": 1.0085,
+      "step": 332
+    },
+    {
+      "epoch": 0.6166666666666667,
+      "grad_norm": 3.254282617277017,
+      "learning_rate": 1.0277777777777777e-07,
+      "loss": 1.0448,
+      "step": 333
+    },
+    {
+      "epoch": 0.6185185185185185,
+      "grad_norm": 3.518638459412198,
+      "learning_rate": 1.030864197530864e-07,
+      "loss": 0.9822,
+      "step": 334
+    },
+    {
+      "epoch": 0.6203703703703703,
+      "grad_norm": 3.203534842703008,
+      "learning_rate": 1.0339506172839505e-07,
+      "loss": 1.0235,
+      "step": 335
+    },
+    {
+      "epoch": 0.6222222222222222,
+      "grad_norm": 3.2226059542304553,
+      "learning_rate": 1.0370370370370369e-07,
+      "loss": 1.0092,
+      "step": 336
+    },
+    {
+      "epoch": 0.6240740740740741,
+      "grad_norm": 3.1951991177522103,
+      "learning_rate": 1.0401234567901234e-07,
+      "loss": 1.0156,
+      "step": 337
+    },
+    {
+      "epoch": 0.6259259259259259,
+      "grad_norm": 3.1456877871563607,
+      "learning_rate": 1.0432098765432098e-07,
+      "loss": 0.9946,
+      "step": 338
+    },
+    {
+      "epoch": 0.6277777777777778,
+      "grad_norm": 3.754516469146756,
+      "learning_rate": 1.0462962962962963e-07,
+      "loss": 0.9868,
+      "step": 339
+    },
+    {
+      "epoch": 0.6296296296296297,
+      "grad_norm": 3.7534332564859816,
+      "learning_rate": 1.0493827160493827e-07,
+      "loss": 1.0737,
+      "step": 340
+    },
+    {
+      "epoch": 0.6314814814814815,
+      "grad_norm": 3.3612600856994677,
+      "learning_rate": 1.0524691358024692e-07,
+      "loss": 0.9916,
+      "step": 341
+    },
+    {
+      "epoch": 0.6333333333333333,
+      "grad_norm": 3.0886146650751676,
+      "learning_rate": 1.0555555555555555e-07,
+      "loss": 1.0262,
+      "step": 342
+    },
+    {
+      "epoch": 0.6351851851851852,
+      "grad_norm": 10.782510450941533,
+      "learning_rate": 1.058641975308642e-07,
+      "loss": 0.9866,
+      "step": 343
+    },
+    {
+      "epoch": 0.6370370370370371,
+      "grad_norm": 3.142360207609036,
+      "learning_rate": 1.0617283950617284e-07,
+      "loss": 0.9486,
+      "step": 344
+    },
+    {
+      "epoch": 0.6388888888888888,
+      "grad_norm": 3.0910722749301893,
+      "learning_rate": 1.0648148148148149e-07,
+      "loss": 0.9855,
+      "step": 345
+    },
+    {
+      "epoch": 0.6407407407407407,
+      "grad_norm": 3.0858788934690446,
+      "learning_rate": 1.0679012345679013e-07,
+      "loss": 1.0133,
+      "step": 346
+    },
+    {
+      "epoch": 0.6425925925925926,
+      "grad_norm": 3.0911926600551696,
+      "learning_rate": 1.0709876543209878e-07,
+      "loss": 0.9406,
+      "step": 347
+    },
+    {
+      "epoch": 0.6444444444444445,
+      "grad_norm": 3.0597800798170414,
+      "learning_rate": 1.0740740740740741e-07,
+      "loss": 0.9874,
+      "step": 348
+    },
+    {
+      "epoch": 0.6462962962962963,
+      "grad_norm": 2.981092097330975,
+      "learning_rate": 1.0771604938271604e-07,
+      "loss": 0.9965,
+      "step": 349
+    },
+    {
+      "epoch": 0.6481481481481481,
+      "grad_norm": 2.9987104420125354,
+      "learning_rate": 1.0802469135802467e-07,
+      "loss": 0.9332,
+      "step": 350
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.04519271249131,
+      "learning_rate": 1.0833333333333332e-07,
+      "loss": 1.0534,
+      "step": 351
+    },
+    {
+      "epoch": 0.6518518518518519,
+      "grad_norm": 2.849398460345178,
+      "learning_rate": 1.0864197530864196e-07,
+      "loss": 0.9925,
+      "step": 352
+    },
+    {
+      "epoch": 0.6537037037037037,
+      "grad_norm": 2.8476217984678374,
+      "learning_rate": 1.0895061728395061e-07,
+      "loss": 1.0026,
+      "step": 353
+    },
+    {
+      "epoch": 0.6555555555555556,
+      "grad_norm": 3.2860186349924936,
+      "learning_rate": 1.0925925925925925e-07,
+      "loss": 0.9693,
+      "step": 354
+    },
+    {
+      "epoch": 0.6574074074074074,
+      "grad_norm": 2.9055661435857685,
+      "learning_rate": 1.095679012345679e-07,
+      "loss": 0.9959,
+      "step": 355
+    },
+    {
+      "epoch": 0.6592592592592592,
+      "grad_norm": 2.8698105783205876,
+      "learning_rate": 1.0987654320987653e-07,
+      "loss": 1.0042,
+      "step": 356
+    },
+    {
+      "epoch": 0.6611111111111111,
+      "grad_norm": 2.8994083090789986,
+      "learning_rate": 1.1018518518518519e-07,
+      "loss": 0.9425,
+      "step": 357
+    },
+    {
+      "epoch": 0.662962962962963,
+      "grad_norm": 3.0711753606889958,
+      "learning_rate": 1.1049382716049382e-07,
+      "loss": 0.9382,
+      "step": 358
+    },
+    {
+      "epoch": 0.6648148148148149,
+      "grad_norm": 3.055805679959846,
+      "learning_rate": 1.1080246913580247e-07,
+      "loss": 0.9751,
+      "step": 359
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 2.7437744731379006,
+      "learning_rate": 1.1111111111111111e-07,
+      "loss": 0.9771,
+      "step": 360
+    },
+    {
+      "epoch": 0.6685185185185185,
+      "grad_norm": 3.2083309317148307,
+      "learning_rate": 1.1141975308641976e-07,
+      "loss": 0.9216,
+      "step": 361
+    },
+    {
+      "epoch": 0.6703703703703704,
+      "grad_norm": 2.9211725325299085,
+      "learning_rate": 1.117283950617284e-07,
+      "loss": 0.992,
+      "step": 362
+    },
+    {
+      "epoch": 0.6722222222222223,
+      "grad_norm": 3.3679385542846547,
+      "learning_rate": 1.1203703703703705e-07,
+      "loss": 0.9802,
+      "step": 363
+    },
+    {
+      "epoch": 0.674074074074074,
+      "grad_norm": 2.6680671454091813,
+      "learning_rate": 1.1234567901234568e-07,
+      "loss": 0.9565,
+      "step": 364
+    },
+    {
+      "epoch": 0.6759259259259259,
+      "grad_norm": 2.9327345491165775,
+      "learning_rate": 1.1265432098765431e-07,
+      "loss": 0.9714,
+      "step": 365
+    },
+    {
+      "epoch": 0.6777777777777778,
+      "grad_norm": 2.914073529883788,
+      "learning_rate": 1.1296296296296294e-07,
+      "loss": 0.9827,
+      "step": 366
+    },
+    {
+      "epoch": 0.6796296296296296,
+      "grad_norm": 2.886811903293114,
+      "learning_rate": 1.132716049382716e-07,
+      "loss": 0.911,
+      "step": 367
+    },
+    {
+      "epoch": 0.6814814814814815,
+      "grad_norm": 3.0308042131753994,
+      "learning_rate": 1.1358024691358023e-07,
+      "loss": 1.0101,
+      "step": 368
+    },
+    {
+      "epoch": 0.6833333333333333,
+      "grad_norm": 2.7958369287491625,
+      "learning_rate": 1.1388888888888888e-07,
+      "loss": 0.8902,
+      "step": 369
+    },
+    {
+      "epoch": 0.6851851851851852,
+      "grad_norm": 2.6122766390774563,
+      "learning_rate": 1.1419753086419752e-07,
+      "loss": 0.9456,
+      "step": 370
+    },
+    {
+      "epoch": 0.687037037037037,
+      "grad_norm": 2.662744553113895,
+      "learning_rate": 1.1450617283950617e-07,
+      "loss": 0.9444,
+      "step": 371
+    },
+    {
+      "epoch": 0.6888888888888889,
+      "grad_norm": 2.819124116450683,
+      "learning_rate": 1.148148148148148e-07,
+      "loss": 0.9955,
+      "step": 372
+    },
+    {
+      "epoch": 0.6907407407407408,
+      "grad_norm": 2.900655599234632,
+      "learning_rate": 1.1512345679012345e-07,
+      "loss": 1.0019,
+      "step": 373
+    },
+    {
+      "epoch": 0.6925925925925925,
+      "grad_norm": 2.706102039336673,
+      "learning_rate": 1.1543209876543209e-07,
+      "loss": 1.0713,
+      "step": 374
+    },
+    {
+      "epoch": 0.6944444444444444,
+      "grad_norm": 2.7726424374425425,
+      "learning_rate": 1.1574074074074074e-07,
+      "loss": 1.0292,
+      "step": 375
+    },
+    {
+      "epoch": 0.6962962962962963,
+      "grad_norm": 2.7566159571889886,
+      "learning_rate": 1.1604938271604938e-07,
+      "loss": 0.9611,
+      "step": 376
+    },
+    {
+      "epoch": 0.6981481481481482,
+      "grad_norm": 3.1510749199541443,
+      "learning_rate": 1.1635802469135803e-07,
+      "loss": 0.9945,
+      "step": 377
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7425882172844007,
+      "learning_rate": 1.1666666666666667e-07,
+      "loss": 0.9356,
+      "step": 378
+    },
+    {
+      "epoch": 0.7018518518518518,
+      "grad_norm": 2.4352316404435252,
+      "learning_rate": 1.1697530864197532e-07,
+      "loss": 0.9164,
+      "step": 379
+    },
+    {
+      "epoch": 0.7037037037037037,
+      "grad_norm": 2.3459367599905896,
+      "learning_rate": 1.1728395061728395e-07,
+      "loss": 0.9506,
+      "step": 380
+    },
+    {
+      "epoch": 0.7055555555555556,
+      "grad_norm": 3.0041608313424173,
+      "learning_rate": 1.1759259259259258e-07,
+      "loss": 0.9725,
+      "step": 381
+    },
+    {
+      "epoch": 0.7074074074074074,
+      "grad_norm": 2.742575853590362,
+      "learning_rate": 1.1790123456790121e-07,
+      "loss": 0.9795,
+      "step": 382
+    },
+    {
+      "epoch": 0.7092592592592593,
+      "grad_norm": 4.840488192646905,
+      "learning_rate": 1.1820987654320986e-07,
+      "loss": 0.9137,
+      "step": 383
+    },
+    {
+      "epoch": 0.7111111111111111,
+      "grad_norm": 2.479164235734708,
+      "learning_rate": 1.185185185185185e-07,
+      "loss": 0.9409,
+      "step": 384
+    },
+    {
+      "epoch": 0.7129629629629629,
+      "grad_norm": 2.4522212362192892,
+      "learning_rate": 1.1882716049382715e-07,
+      "loss": 0.9407,
+      "step": 385
+    },
+    {
+      "epoch": 0.7148148148148148,
+      "grad_norm": 2.3959924395157777,
+      "learning_rate": 1.1913580246913579e-07,
+      "loss": 0.9627,
+      "step": 386
+    },
+    {
+      "epoch": 0.7166666666666667,
+      "grad_norm": 2.403811670042839,
+      "learning_rate": 1.1944444444444442e-07,
+      "loss": 0.9516,
+      "step": 387
+    },
+    {
+      "epoch": 0.7185185185185186,
+      "grad_norm": 2.3419572615787816,
+      "learning_rate": 1.1975308641975307e-07,
+      "loss": 0.9268,
+      "step": 388
+    },
+    {
+      "epoch": 0.7203703703703703,
+      "grad_norm": 2.497470338646794,
+      "learning_rate": 1.2006172839506172e-07,
+      "loss": 0.9109,
+      "step": 389
+    },
+    {
+      "epoch": 0.7222222222222222,
+      "grad_norm": 2.324528256770774,
+      "learning_rate": 1.2037037037037037e-07,
+      "loss": 0.9467,
+      "step": 390
+    },
+    {
+      "epoch": 0.7240740740740741,
+      "grad_norm": 2.3451110299913736,
+      "learning_rate": 1.20679012345679e-07,
+      "loss": 0.9169,
+      "step": 391
+    },
+    {
+      "epoch": 0.725925925925926,
+      "grad_norm": 3.316057861621452,
+      "learning_rate": 1.2098765432098765e-07,
+      "loss": 0.9395,
+      "step": 392
+    },
+    {
+      "epoch": 0.7277777777777777,
+      "grad_norm": 2.9385499228836838,
+      "learning_rate": 1.212962962962963e-07,
+      "loss": 0.9744,
+      "step": 393
+    },
+    {
+      "epoch": 0.7296296296296296,
+      "grad_norm": 2.2885592797725165,
+      "learning_rate": 1.2160493827160495e-07,
+      "loss": 0.9238,
+      "step": 394
+    },
+    {
+      "epoch": 0.7314814814814815,
+      "grad_norm": 2.2054451654866707,
+      "learning_rate": 1.2191358024691357e-07,
+      "loss": 0.9436,
+      "step": 395
+    },
+    {
+      "epoch": 0.7333333333333333,
+      "grad_norm": 2.5514270934723067,
+      "learning_rate": 1.2222222222222222e-07,
+      "loss": 0.9315,
+      "step": 396
+    },
+    {
+      "epoch": 0.7351851851851852,
+      "grad_norm": 2.2368117556543456,
+      "learning_rate": 1.2253086419753085e-07,
+      "loss": 0.9638,
+      "step": 397
+    },
+    {
+      "epoch": 0.737037037037037,
+      "grad_norm": 2.266035335465804,
+      "learning_rate": 1.228395061728395e-07,
+      "loss": 0.8805,
+      "step": 398
+    },
+    {
+      "epoch": 0.7388888888888889,
+      "grad_norm": 2.2358935123241,
+      "learning_rate": 1.2314814814814815e-07,
+      "loss": 0.9705,
+      "step": 399
+    },
+    {
+      "epoch": 0.7407407407407407,
+      "grad_norm": 2.31380824764338,
+      "learning_rate": 1.2345679012345677e-07,
+      "loss": 0.9071,
+      "step": 400
+    },
+    {
+      "epoch": 0.7425925925925926,
+      "grad_norm": 2.5111441323147985,
+      "learning_rate": 1.2376543209876542e-07,
+      "loss": 0.9094,
+      "step": 401
+    },
+    {
+      "epoch": 0.7444444444444445,
+      "grad_norm": 2.3304081052081282,
+      "learning_rate": 1.2407407407407407e-07,
+      "loss": 0.93,
+      "step": 402
+    },
+    {
+      "epoch": 0.7462962962962963,
+      "grad_norm": 2.37840030150777,
+      "learning_rate": 1.2438271604938272e-07,
+      "loss": 0.9203,
+      "step": 403
+    },
+    {
+      "epoch": 0.7481481481481481,
+      "grad_norm": 2.432117775628416,
+      "learning_rate": 1.2469135802469134e-07,
+      "loss": 0.904,
+      "step": 404
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0092139480408537,
+      "learning_rate": 1.25e-07,
+      "loss": 0.9309,
+      "step": 405
+    },
+    {
+      "epoch": 0.7518518518518519,
+      "grad_norm": 3.672761262710248,
+      "learning_rate": 1.2530864197530864e-07,
+      "loss": 0.9027,
+      "step": 406
+    },
+    {
+      "epoch": 0.7537037037037037,
+      "grad_norm": 2.1249398432733253,
+      "learning_rate": 1.2561728395061727e-07,
+      "loss": 0.9258,
+      "step": 407
+    },
+    {
+      "epoch": 0.7555555555555555,
+      "grad_norm": 2.361832315773115,
+      "learning_rate": 1.2592592592592592e-07,
+      "loss": 0.8982,
+      "step": 408
+    },
+    {
+      "epoch": 0.7574074074074074,
+      "grad_norm": 2.2076222382736828,
+      "learning_rate": 1.2623456790123457e-07,
+      "loss": 0.9559,
+      "step": 409
+    },
+    {
+      "epoch": 0.7592592592592593,
+      "grad_norm": 2.319232858121409,
+      "learning_rate": 1.2654320987654322e-07,
+      "loss": 0.9057,
+      "step": 410
+    },
+    {
+      "epoch": 0.7611111111111111,
+      "grad_norm": 2.227871917953055,
+      "learning_rate": 1.2685185185185184e-07,
+      "loss": 0.9421,
+      "step": 411
+    },
+    {
+      "epoch": 0.762962962962963,
+      "grad_norm": 2.223934637384949,
+      "learning_rate": 1.271604938271605e-07,
+      "loss": 0.9403,
+      "step": 412
+    },
+    {
+      "epoch": 0.7648148148148148,
+      "grad_norm": 2.142718818730283,
+      "learning_rate": 1.2746913580246914e-07,
+      "loss": 0.91,
+      "step": 413
+    },
+    {
+      "epoch": 0.7666666666666667,
+      "grad_norm": 2.0779692106131264,
+      "learning_rate": 1.2777777777777777e-07,
+      "loss": 0.9485,
+      "step": 414
+    },
+    {
+      "epoch": 0.7685185185185185,
+      "grad_norm": 2.342474251086406,
+      "learning_rate": 1.2808641975308642e-07,
+      "loss": 0.9255,
+      "step": 415
+    },
+    {
+      "epoch": 0.7703703703703704,
+      "grad_norm": 2.329358870521356,
+      "learning_rate": 1.2839506172839504e-07,
+      "loss": 0.9434,
+      "step": 416
+    },
+    {
+      "epoch": 0.7722222222222223,
+      "grad_norm": 2.21863092723048,
+      "learning_rate": 1.287037037037037e-07,
+      "loss": 0.9363,
+      "step": 417
+    },
+    {
+      "epoch": 0.774074074074074,
+      "grad_norm": 2.033704886221468,
+      "learning_rate": 1.2901234567901234e-07,
+      "loss": 0.8838,
+      "step": 418
+    },
+    {
+      "epoch": 0.7759259259259259,
+      "grad_norm": 1.9977248258436775,
+      "learning_rate": 1.29320987654321e-07,
+      "loss": 0.8925,
+      "step": 419
+    },
+    {
+      "epoch": 0.7777777777777778,
+      "grad_norm": 2.04238549943764,
+      "learning_rate": 1.2962962962962961e-07,
+      "loss": 0.9343,
+      "step": 420
+    },
+    {
+      "epoch": 0.7796296296296297,
+      "grad_norm": 2.2302422155062276,
+      "learning_rate": 1.2993827160493826e-07,
+      "loss": 0.9192,
+      "step": 421
+    },
+    {
+      "epoch": 0.7814814814814814,
+      "grad_norm": 1.9950899526602213,
+      "learning_rate": 1.3024691358024691e-07,
+      "loss": 0.8803,
+      "step": 422
+    },
+    {
+      "epoch": 0.7833333333333333,
+      "grad_norm": 2.2831484969299587,
+      "learning_rate": 1.3055555555555556e-07,
+      "loss": 0.8651,
+      "step": 423
+    },
+    {
+      "epoch": 0.7851851851851852,
+      "grad_norm": 2.344415429236907,
+      "learning_rate": 1.308641975308642e-07,
+      "loss": 0.8864,
+      "step": 424
+    },
+    {
+      "epoch": 0.7870370370370371,
+      "grad_norm": 2.222725365830128,
+      "learning_rate": 1.3117283950617284e-07,
+      "loss": 0.9135,
+      "step": 425
+    },
+    {
+      "epoch": 0.7888888888888889,
+      "grad_norm": 2.2566021695397906,
+      "learning_rate": 1.314814814814815e-07,
+      "loss": 0.9597,
+      "step": 426
+    },
+    {
+      "epoch": 0.7907407407407407,
+      "grad_norm": 2.237807988628356,
+      "learning_rate": 1.3179012345679014e-07,
+      "loss": 0.8831,
+      "step": 427
+    },
+    {
+      "epoch": 0.7925925925925926,
+      "grad_norm": 2.297155195766305,
+      "learning_rate": 1.3209876543209876e-07,
+      "loss": 0.9147,
+      "step": 428
+    },
+    {
+      "epoch": 0.7944444444444444,
+      "grad_norm": 2.809965128994331,
+      "learning_rate": 1.324074074074074e-07,
+      "loss": 0.9325,
+      "step": 429
+    },
+    {
+      "epoch": 0.7962962962962963,
+      "grad_norm": 2.025045575533884,
+      "learning_rate": 1.3271604938271604e-07,
+      "loss": 0.9587,
+      "step": 430
+    },
+    {
+      "epoch": 0.7981481481481482,
+      "grad_norm": 2.0159452459757143,
+      "learning_rate": 1.3302469135802469e-07,
+      "loss": 0.9072,
+      "step": 431
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.029223716341751,
+      "learning_rate": 1.333333333333333e-07,
+      "loss": 0.8628,
+      "step": 432
+    },
+    {
+      "epoch": 0.8018518518518518,
+      "grad_norm": 1.921147791939492,
+      "learning_rate": 1.3364197530864196e-07,
+      "loss": 0.8962,
+      "step": 433
+    },
+    {
+      "epoch": 0.8037037037037037,
+      "grad_norm": 1.998351573317739,
+      "learning_rate": 1.339506172839506e-07,
+      "loss": 0.9269,
+      "step": 434
+    },
+    {
+      "epoch": 0.8055555555555556,
+      "grad_norm": 2.202907873704923,
+      "learning_rate": 1.3425925925925926e-07,
+      "loss": 0.9064,
+      "step": 435
+    },
+    {
+      "epoch": 0.8074074074074075,
+      "grad_norm": 2.070570348046798,
+      "learning_rate": 1.3456790123456788e-07,
+      "loss": 0.9168,
+      "step": 436
+    },
+    {
+      "epoch": 0.8092592592592592,
+      "grad_norm": 2.0174231832876517,
+      "learning_rate": 1.3487654320987653e-07,
+      "loss": 0.8804,
+      "step": 437
+    },
+    {
+      "epoch": 0.8111111111111111,
+      "grad_norm": 2.246933453381705,
+      "learning_rate": 1.3518518518518518e-07,
+      "loss": 0.9697,
+      "step": 438
+    },
+    {
+      "epoch": 0.812962962962963,
+      "grad_norm": 2.199080672800522,
+      "learning_rate": 1.3549382716049383e-07,
+      "loss": 0.9057,
+      "step": 439
+    },
+    {
+      "epoch": 0.8148148148148148,
+      "grad_norm": 2.2364283143669286,
+      "learning_rate": 1.3580246913580246e-07,
+      "loss": 0.8524,
+      "step": 440
+    },
+    {
+      "epoch": 0.8166666666666667,
+      "grad_norm": 4.024158243823565,
+      "learning_rate": 1.361111111111111e-07,
+      "loss": 0.9327,
+      "step": 441
+    },
+    {
+      "epoch": 0.8185185185185185,
+      "grad_norm": 2.044294128111021,
+      "learning_rate": 1.3641975308641976e-07,
+      "loss": 0.8726,
+      "step": 442
+    },
+    {
+      "epoch": 0.8203703703703704,
+      "grad_norm": 1.8158630448309119,
+      "learning_rate": 1.367283950617284e-07,
+      "loss": 0.8908,
+      "step": 443
+    },
+    {
+      "epoch": 0.8222222222222222,
+      "grad_norm": 2.4331116688216783,
+      "learning_rate": 1.3703703703703703e-07,
+      "loss": 0.8874,
+      "step": 444
+    },
+    {
+      "epoch": 0.8240740740740741,
+      "grad_norm": 1.9981430466466217,
+      "learning_rate": 1.3734567901234568e-07,
+      "loss": 0.869,
+      "step": 445
+    },
+    {
+      "epoch": 0.825925925925926,
+      "grad_norm": 1.9795801077027648,
+      "learning_rate": 1.376543209876543e-07,
+      "loss": 0.8912,
+      "step": 446
+    },
+    {
+      "epoch": 0.8277777777777777,
+      "grad_norm": 1.9397925528326283,
+      "learning_rate": 1.3796296296296296e-07,
+      "loss": 0.9536,
+      "step": 447
+    },
+    {
+      "epoch": 0.8296296296296296,
+      "grad_norm": 2.2321691004910305,
+      "learning_rate": 1.3827160493827158e-07,
+      "loss": 0.9704,
+      "step": 448
+    },
+    {
+      "epoch": 0.8314814814814815,
+      "grad_norm": 1.876055348972833,
+      "learning_rate": 1.3858024691358023e-07,
+      "loss": 0.8835,
+      "step": 449
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 1.9096879514198053,
+      "learning_rate": 1.3888888888888888e-07,
+      "loss": 0.8971,
+      "step": 450
+    },
+    {
+      "epoch": 0.8351851851851851,
+      "grad_norm": 1.8105515638937926,
+      "learning_rate": 1.3919753086419753e-07,
+      "loss": 0.9556,
+      "step": 451
+    },
+    {
+      "epoch": 0.837037037037037,
+      "grad_norm": 2.0514395378516252,
+      "learning_rate": 1.3950617283950615e-07,
+      "loss": 0.8561,
+      "step": 452
+    },
+    {
+      "epoch": 0.8388888888888889,
+      "grad_norm": 1.8642844364032094,
+      "learning_rate": 1.398148148148148e-07,
+      "loss": 0.8433,
+      "step": 453
+    },
+    {
+      "epoch": 0.8407407407407408,
+      "grad_norm": 1.804313908556319,
+      "learning_rate": 1.4012345679012345e-07,
+      "loss": 0.8654,
+      "step": 454
+    },
+    {
+      "epoch": 0.8425925925925926,
+      "grad_norm": 1.8953109256785983,
+      "learning_rate": 1.404320987654321e-07,
+      "loss": 0.9814,
+      "step": 455
+    },
+    {
+      "epoch": 0.8444444444444444,
+      "grad_norm": 2.1225876682364273,
+      "learning_rate": 1.4074074074074073e-07,
+      "loss": 0.9057,
+      "step": 456
+    },
+    {
+      "epoch": 0.8462962962962963,
+      "grad_norm": 2.2016444974916323,
+      "learning_rate": 1.4104938271604938e-07,
+      "loss": 0.8201,
+      "step": 457
+    },
+    {
+      "epoch": 0.8481481481481481,
+      "grad_norm": 2.0432606819951524,
+      "learning_rate": 1.4135802469135803e-07,
+      "loss": 0.8842,
+      "step": 458
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1280328243286752,
+      "learning_rate": 1.4166666666666668e-07,
+      "loss": 0.8402,
+      "step": 459
+    },
+    {
+      "epoch": 0.8518518518518519,
+      "grad_norm": 2.000472025214132,
+      "learning_rate": 1.419753086419753e-07,
+      "loss": 0.8819,
+      "step": 460
+    },
+    {
+      "epoch": 0.8537037037037037,
+      "grad_norm": 1.8512466548066744,
+      "learning_rate": 1.4228395061728395e-07,
+      "loss": 0.9432,
+      "step": 461
+    },
+    {
+      "epoch": 0.8555555555555555,
+      "grad_norm": 2.0106659004010137,
+      "learning_rate": 1.4259259259259258e-07,
+      "loss": 0.8719,
+      "step": 462
+    },
+    {
+      "epoch": 0.8574074074074074,
+      "grad_norm": 1.8848461393858795,
+      "learning_rate": 1.4290123456790123e-07,
+      "loss": 0.9284,
+      "step": 463
+    },
+    {
+      "epoch": 0.8592592592592593,
+      "grad_norm": 2.530928394391292,
+      "learning_rate": 1.4320987654320985e-07,
+      "loss": 0.9414,
+      "step": 464
+    },
+    {
+      "epoch": 0.8611111111111112,
+      "grad_norm": 1.7563339526281114,
+      "learning_rate": 1.435185185185185e-07,
+      "loss": 0.839,
+      "step": 465
+    },
+    {
+      "epoch": 0.8629629629629629,
+      "grad_norm": 1.7548405700159353,
+      "learning_rate": 1.4382716049382715e-07,
+      "loss": 0.8824,
+      "step": 466
+    },
+    {
+      "epoch": 0.8648148148148148,
+      "grad_norm": 5.659133892657317,
+      "learning_rate": 1.441358024691358e-07,
+      "loss": 0.8713,
+      "step": 467
+    },
+    {
+      "epoch": 0.8666666666666667,
+      "grad_norm": 3.3047969228291567,
+      "learning_rate": 1.4444444444444442e-07,
+      "loss": 0.9101,
+      "step": 468
+    },
+    {
+      "epoch": 0.8685185185185185,
+      "grad_norm": 1.8286881417606347,
+      "learning_rate": 1.4475308641975307e-07,
+      "loss": 0.8855,
+      "step": 469
+    },
+    {
+      "epoch": 0.8703703703703703,
+      "grad_norm": 1.713564429231544,
+      "learning_rate": 1.4506172839506172e-07,
+      "loss": 0.8755,
+      "step": 470
+    },
+    {
+      "epoch": 0.8722222222222222,
+      "grad_norm": 1.7522479788067298,
+      "learning_rate": 1.4537037037037037e-07,
+      "loss": 0.8801,
+      "step": 471
+    },
+    {
+      "epoch": 0.8740740740740741,
+      "grad_norm": 1.7913855849127627,
+      "learning_rate": 1.45679012345679e-07,
+      "loss": 0.8893,
+      "step": 472
+    },
+    {
+      "epoch": 0.8759259259259259,
+      "grad_norm": 1.8590698519635387,
+      "learning_rate": 1.4598765432098765e-07,
+      "loss": 0.8477,
+      "step": 473
+    },
+    {
+      "epoch": 0.8777777777777778,
+      "grad_norm": 1.9168882323220686,
+      "learning_rate": 1.462962962962963e-07,
+      "loss": 0.8839,
+      "step": 474
+    },
+    {
+      "epoch": 0.8796296296296297,
+      "grad_norm": 1.8287509487852251,
+      "learning_rate": 1.4660493827160495e-07,
+      "loss": 0.934,
+      "step": 475
+    },
+    {
+      "epoch": 0.8814814814814815,
+      "grad_norm": 1.7318355238854954,
+      "learning_rate": 1.4691358024691357e-07,
+      "loss": 0.8592,
+      "step": 476
+    },
+    {
+      "epoch": 0.8833333333333333,
+      "grad_norm": 1.8155966159699268,
+      "learning_rate": 1.4722222222222222e-07,
+      "loss": 0.9373,
+      "step": 477
+    },
+    {
+      "epoch": 0.8851851851851852,
+      "grad_norm": 1.8778776325197515,
+      "learning_rate": 1.4753086419753085e-07,
+      "loss": 0.8686,
+      "step": 478
+    },
+    {
+      "epoch": 0.8870370370370371,
+      "grad_norm": 2.1537192152388185,
+      "learning_rate": 1.478395061728395e-07,
+      "loss": 0.8472,
+      "step": 479
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 1.8763174725335232,
+      "learning_rate": 1.4814814814814815e-07,
+      "loss": 0.809,
+      "step": 480
+    },
+    {
+      "epoch": 0.8907407407407407,
+      "grad_norm": 1.8861836499017712,
+      "learning_rate": 1.4845679012345677e-07,
+      "loss": 0.8237,
+      "step": 481
+    },
+    {
+      "epoch": 0.8925925925925926,
+      "grad_norm": 1.8513811134477731,
+      "learning_rate": 1.4876543209876542e-07,
+      "loss": 0.9212,
+      "step": 482
+    },
+    {
+      "epoch": 0.8944444444444445,
+      "grad_norm": 1.7768570889746993,
+      "learning_rate": 1.4907407407407407e-07,
+      "loss": 0.8951,
+      "step": 483
+    },
+    {
+      "epoch": 0.8962962962962963,
+      "grad_norm": 1.7852128710682684,
+      "learning_rate": 1.4938271604938272e-07,
+      "loss": 0.9029,
+      "step": 484
+    },
+    {
+      "epoch": 0.8981481481481481,
+      "grad_norm": 3.4913964921006713,
+      "learning_rate": 1.4969135802469134e-07,
+      "loss": 0.9091,
+      "step": 485
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3549257572774924,
+      "learning_rate": 1.5e-07,
+      "loss": 0.8579,
+      "step": 486
+    },
+    {
+      "epoch": 0.9018518518518519,
+      "grad_norm": 1.7330307330370882,
+      "learning_rate": 1.5030864197530864e-07,
+      "loss": 0.9276,
+      "step": 487
+    },
+    {
+      "epoch": 0.9037037037037037,
+      "grad_norm": 1.8259607839651515,
+      "learning_rate": 1.5061728395061727e-07,
+      "loss": 0.891,
+      "step": 488
+    },
+    {
+      "epoch": 0.9055555555555556,
+      "grad_norm": 1.7047001942211497,
+      "learning_rate": 1.5092592592592592e-07,
+      "loss": 0.8804,
+      "step": 489
+    },
+    {
+      "epoch": 0.9074074074074074,
+      "grad_norm": 1.9490305282427514,
+      "learning_rate": 1.5123456790123457e-07,
+      "loss": 0.8768,
+      "step": 490
+    },
+    {
+      "epoch": 0.9092592592592592,
+      "grad_norm": 1.658719915297525,
+      "learning_rate": 1.5154320987654322e-07,
+      "loss": 0.858,
+      "step": 491
+    },
+    {
+      "epoch": 0.9111111111111111,
+      "grad_norm": 1.9402965103571035,
+      "learning_rate": 1.5185185185185184e-07,
+      "loss": 0.8649,
+      "step": 492
+    },
+    {
+      "epoch": 0.912962962962963,
+      "grad_norm": 1.718571810022627,
+      "learning_rate": 1.521604938271605e-07,
+      "loss": 0.8937,
+      "step": 493
+    },
+    {
+      "epoch": 0.9148148148148149,
+      "grad_norm": 1.6297690955124837,
+      "learning_rate": 1.5246913580246914e-07,
+      "loss": 0.828,
+      "step": 494
+    },
+    {
+      "epoch": 0.9166666666666666,
+      "grad_norm": 1.8414582414718768,
+      "learning_rate": 1.5277777777777776e-07,
+      "loss": 0.9126,
+      "step": 495
+    },
+    {
+      "epoch": 0.9185185185185185,
+      "grad_norm": 1.7802538734709807,
+      "learning_rate": 1.5308641975308642e-07,
+      "loss": 0.9056,
+      "step": 496
+    },
+    {
+      "epoch": 0.9203703703703704,
+      "grad_norm": 5.228694141231115,
+      "learning_rate": 1.5339506172839504e-07,
+      "loss": 0.8687,
+      "step": 497
+    },
+    {
+      "epoch": 0.9222222222222223,
+      "grad_norm": 1.8464622991838917,
+      "learning_rate": 1.537037037037037e-07,
+      "loss": 0.9016,
+      "step": 498
+    },
+    {
+      "epoch": 0.924074074074074,
+      "grad_norm": 2.4297371592733694,
+      "learning_rate": 1.5401234567901234e-07,
+      "loss": 0.8035,
+      "step": 499
+    },
+    {
+      "epoch": 0.9259259259259259,
+      "grad_norm": 2.0737176515541664,
+      "learning_rate": 1.54320987654321e-07,
+      "loss": 0.898,
+      "step": 500
+    },
+    {
+      "epoch": 0.9277777777777778,
+      "grad_norm": 1.8183579167972654,
+      "learning_rate": 1.546296296296296e-07,
+      "loss": 0.8512,
+      "step": 501
+    },
+    {
+      "epoch": 0.9296296296296296,
+      "grad_norm": 1.816892174063631,
+      "learning_rate": 1.5493827160493826e-07,
+      "loss": 0.8708,
+      "step": 502
+    },
+    {
+      "epoch": 0.9314814814814815,
+      "grad_norm": 1.6465826736383782,
+      "learning_rate": 1.552469135802469e-07,
+      "loss": 0.8607,
+      "step": 503
+    },
+    {
+      "epoch": 0.9333333333333333,
+      "grad_norm": 2.2704556478446714,
+      "learning_rate": 1.5555555555555556e-07,
+      "loss": 0.8925,
+      "step": 504
+    },
+    {
+      "epoch": 0.9351851851851852,
+      "grad_norm": 1.7467114879223669,
+      "learning_rate": 1.558641975308642e-07,
+      "loss": 0.8675,
+      "step": 505
+    },
+    {
+      "epoch": 0.937037037037037,
+      "grad_norm": 1.6062240697663543,
+      "learning_rate": 1.5617283950617284e-07,
+      "loss": 0.7859,
+      "step": 506
+    },
+    {
+      "epoch": 0.9388888888888889,
+      "grad_norm": 1.640348857593773,
+      "learning_rate": 1.564814814814815e-07,
+      "loss": 0.8543,
+      "step": 507
+    },
+    {
+      "epoch": 0.9407407407407408,
+      "grad_norm": 1.550816377843016,
+      "learning_rate": 1.5679012345679014e-07,
+      "loss": 0.8109,
+      "step": 508
+    },
+    {
+      "epoch": 0.9425925925925925,
+      "grad_norm": 2.0305085131556058,
+      "learning_rate": 1.5709876543209876e-07,
+      "loss": 0.8309,
+      "step": 509
+    },
+    {
+      "epoch": 0.9444444444444444,
+      "grad_norm": 1.568766105297353,
+      "learning_rate": 1.574074074074074e-07,
+      "loss": 0.8161,
+      "step": 510
+    },
+    {
+      "epoch": 0.9462962962962963,
+      "grad_norm": 1.7340153277737653,
+      "learning_rate": 1.5771604938271603e-07,
+      "loss": 0.9177,
+      "step": 511
+    },
+    {
+      "epoch": 0.9481481481481482,
+      "grad_norm": 1.885988879145682,
+      "learning_rate": 1.5802469135802468e-07,
+      "loss": 0.8394,
+      "step": 512
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.29569659414859,
+      "learning_rate": 1.583333333333333e-07,
+      "loss": 0.8561,
+      "step": 513
+    },
+    {
+      "epoch": 0.9518518518518518,
+      "grad_norm": 1.7335115853805168,
+      "learning_rate": 1.5864197530864196e-07,
+      "loss": 0.9011,
+      "step": 514
+    },
+    {
+      "epoch": 0.9537037037037037,
+      "grad_norm": 1.675001315928241,
+      "learning_rate": 1.589506172839506e-07,
+      "loss": 0.858,
+      "step": 515
+    },
+    {
+      "epoch": 0.9555555555555556,
+      "grad_norm": 1.9824008234504722,
+      "learning_rate": 1.5925925925925926e-07,
+      "loss": 0.8752,
+      "step": 516
+    },
+    {
+      "epoch": 0.9574074074074074,
+      "grad_norm": 1.5903658251420252,
+      "learning_rate": 1.5956790123456788e-07,
+      "loss": 0.8516,
+      "step": 517
+    },
+    {
+      "epoch": 0.9592592592592593,
+      "grad_norm": 3.004633637313183,
+      "learning_rate": 1.5987654320987653e-07,
+      "loss": 0.8521,
+      "step": 518
+    },
+    {
+      "epoch": 0.9611111111111111,
+      "grad_norm": 2.5953683287541898,
+      "learning_rate": 1.6018518518518518e-07,
+      "loss": 0.8415,
+      "step": 519
+    },
+    {
+      "epoch": 0.9629629629629629,
+      "grad_norm": 1.6725515794786023,
+      "learning_rate": 1.6049382716049383e-07,
+      "loss": 0.8566,
+      "step": 520
+    },
+    {
+      "epoch": 0.9648148148148148,
+      "grad_norm": 3.2869740744171803,
+      "learning_rate": 1.6080246913580246e-07,
+      "loss": 0.8579,
+      "step": 521
+    },
+    {
+      "epoch": 0.9666666666666667,
+      "grad_norm": 1.6233912867628748,
+      "learning_rate": 1.611111111111111e-07,
+      "loss": 0.8437,
+      "step": 522
+    },
+    {
+      "epoch": 0.9685185185185186,
+      "grad_norm": 1.6642946681133515,
+      "learning_rate": 1.6141975308641976e-07,
+      "loss": 0.8328,
+      "step": 523
+    },
+    {
+      "epoch": 0.9703703703703703,
+      "grad_norm": 1.6643806957389826,
+      "learning_rate": 1.617283950617284e-07,
+      "loss": 0.8804,
+      "step": 524
+    },
+    {
+      "epoch": 0.9722222222222222,
+      "grad_norm": 1.6615537227064379,
+      "learning_rate": 1.6203703703703703e-07,
+      "loss": 0.8534,
+      "step": 525
+    },
+    {
+      "epoch": 0.9740740740740741,
+      "grad_norm": 1.574154420468767,
+      "learning_rate": 1.6234567901234568e-07,
+      "loss": 0.8414,
+      "step": 526
+    },
+    {
+      "epoch": 0.975925925925926,
+      "grad_norm": 2.2283760433888857,
+      "learning_rate": 1.626543209876543e-07,
+      "loss": 0.8231,
+      "step": 527
+    },
+    {
+      "epoch": 0.9777777777777777,
+      "grad_norm": 1.7037218290467242,
+      "learning_rate": 1.6296296296296295e-07,
+      "loss": 0.8769,
+      "step": 528
+    },
+    {
+      "epoch": 0.9796296296296296,
+      "grad_norm": 2.754981651435544,
+      "learning_rate": 1.6327160493827158e-07,
+      "loss": 0.8094,
+      "step": 529
+    },
+    {
+      "epoch": 0.9814814814814815,
+      "grad_norm": 3.3500242013818915,
+      "learning_rate": 1.6358024691358023e-07,
+      "loss": 0.8842,
+      "step": 530
+    },
+    {
+      "epoch": 0.9833333333333333,
+      "grad_norm": 1.6897837295849563,
+      "learning_rate": 1.6388888888888888e-07,
+      "loss": 0.8687,
+      "step": 531
+    },
+    {
+      "epoch": 0.9851851851851852,
+      "grad_norm": 1.541496447171207,
+      "learning_rate": 1.6419753086419753e-07,
+      "loss": 0.8503,
+      "step": 532
+    },
+    {
+      "epoch": 0.987037037037037,
+      "grad_norm": 1.7026082175637889,
+      "learning_rate": 1.6450617283950615e-07,
+      "loss": 0.8877,
+      "step": 533
+    },
+    {
+      "epoch": 0.9888888888888889,
+      "grad_norm": 1.6431224642792341,
+      "learning_rate": 1.648148148148148e-07,
+      "loss": 0.8548,
+      "step": 534
+    },
+    {
+      "epoch": 0.9907407407407407,
+      "grad_norm": 2.1178964677350605,
+      "learning_rate": 1.6512345679012345e-07,
+      "loss": 0.8484,
+      "step": 535
+    },
+    {
+      "epoch": 0.9925925925925926,
+      "grad_norm": 3.250841692000507,
+      "learning_rate": 1.654320987654321e-07,
+      "loss": 0.8962,
+      "step": 536
+    },
+    {
+      "epoch": 0.9944444444444445,
+      "grad_norm": 1.66266098194129,
+      "learning_rate": 1.6574074074074073e-07,
+      "loss": 0.8529,
+      "step": 537
+    },
+    {
+      "epoch": 0.9962962962962963,
+      "grad_norm": 1.5252470823405428,
+      "learning_rate": 1.6604938271604938e-07,
+      "loss": 0.8466,
+      "step": 538
+    },
+    {
+      "epoch": 0.9981481481481481,
+      "grad_norm": 2.9922623261242984,
+      "learning_rate": 1.6635802469135803e-07,
+      "loss": 0.758,
+      "step": 539
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.934430205759249,
+      "learning_rate": 1.6666666666666668e-07,
+      "loss": 0.8866,
+      "step": 540
+    },
+    {
+      "epoch": 1.0018518518518518,
+      "grad_norm": 1.8883347084868174,
+      "learning_rate": 1.669753086419753e-07,
+      "loss": 0.8547,
+      "step": 541
+    },
+    {
+      "epoch": 1.0037037037037038,
+      "grad_norm": 1.5748500720605405,
+      "learning_rate": 1.6728395061728395e-07,
+      "loss": 0.87,
+      "step": 542
+    },
+    {
+      "epoch": 1.0055555555555555,
+      "grad_norm": 1.560708327560299,
+      "learning_rate": 1.6759259259259257e-07,
+      "loss": 0.8182,
+      "step": 543
+    },
+    {
+      "epoch": 1.0074074074074073,
+      "grad_norm": 2.0051977214364407,
+      "learning_rate": 1.6790123456790122e-07,
+      "loss": 0.8718,
+      "step": 544
+    },
+    {
+      "epoch": 1.0092592592592593,
+      "grad_norm": 1.6191423023001053,
+      "learning_rate": 1.6820987654320985e-07,
+      "loss": 0.8248,
+      "step": 545
+    },
+    {
+      "epoch": 1.011111111111111,
+      "grad_norm": 1.5051175720206331,
+      "learning_rate": 1.685185185185185e-07,
+      "loss": 0.8916,
+      "step": 546
+    },
+    {
+      "epoch": 1.012962962962963,
+      "grad_norm": 2.9158430296966045,
+      "learning_rate": 1.6882716049382715e-07,
+      "loss": 0.8395,
+      "step": 547
+    },
+    {
+      "epoch": 1.0148148148148148,
+      "grad_norm": 1.4991342659178246,
+      "learning_rate": 1.691358024691358e-07,
+      "loss": 0.8426,
+      "step": 548
+    },
+    {
+      "epoch": 1.0166666666666666,
+      "grad_norm": 1.6995730111540799,
+      "learning_rate": 1.6944444444444442e-07,
+      "loss": 0.8478,
+      "step": 549
+    },
+    {
+      "epoch": 1.0185185185185186,
+      "grad_norm": 1.4665838725695703,
+      "learning_rate": 1.6975308641975307e-07,
+      "loss": 0.8478,
+      "step": 550
+    },
+    {
+      "epoch": 1.0203703703703704,
+      "grad_norm": 1.5934504149694464,
+      "learning_rate": 1.7006172839506172e-07,
+      "loss": 0.8472,
+      "step": 551
+    },
+    {
+      "epoch": 1.0222222222222221,
+      "grad_norm": 1.5776774920082102,
+      "learning_rate": 1.7037037037037037e-07,
+      "loss": 0.8501,
+      "step": 552
+    },
+    {
+      "epoch": 1.0240740740740741,
+      "grad_norm": 1.8108062171862294,
+      "learning_rate": 1.70679012345679e-07,
+      "loss": 0.7989,
+      "step": 553
+    },
+    {
+      "epoch": 1.025925925925926,
+      "grad_norm": 1.5609766897230202,
+      "learning_rate": 1.7098765432098765e-07,
+      "loss": 0.8796,
+      "step": 554
+    },
+    {
+      "epoch": 1.0277777777777777,
+      "grad_norm": 1.5421333063522007,
+      "learning_rate": 1.712962962962963e-07,
+      "loss": 0.8144,
+      "step": 555
+    },
+    {
+      "epoch": 1.0296296296296297,
+      "grad_norm": 1.5906872588547103,
+      "learning_rate": 1.7160493827160495e-07,
+      "loss": 0.8231,
+      "step": 556
+    },
+    {
+      "epoch": 1.0314814814814814,
+      "grad_norm": 1.55563369873077,
+      "learning_rate": 1.7191358024691357e-07,
+      "loss": 0.8441,
+      "step": 557
+    },
+    {
+      "epoch": 1.0333333333333334,
+      "grad_norm": 2.034442281969645,
+      "learning_rate": 1.7222222222222222e-07,
+      "loss": 0.8555,
+      "step": 558
+    },
+    {
+      "epoch": 1.0351851851851852,
+      "grad_norm": 1.6802229628138714,
+      "learning_rate": 1.7253086419753084e-07,
+      "loss": 0.8632,
+      "step": 559
+    },
+    {
+      "epoch": 1.037037037037037,
+      "grad_norm": 1.6434079268881443,
+      "learning_rate": 1.728395061728395e-07,
+      "loss": 0.9062,
+      "step": 560
+    },
+    {
+      "epoch": 1.038888888888889,
+      "grad_norm": 10.170009592221364,
+      "learning_rate": 1.7314814814814814e-07,
+      "loss": 0.8269,
+      "step": 561
+    },
+    {
+      "epoch": 1.0407407407407407,
+      "grad_norm": 1.5129567429071522,
+      "learning_rate": 1.7345679012345677e-07,
+      "loss": 0.7992,
+      "step": 562
+    },
+    {
+      "epoch": 1.0425925925925925,
+      "grad_norm": 1.4693189752768574,
+      "learning_rate": 1.7376543209876542e-07,
+      "loss": 0.8625,
+      "step": 563
+    },
+    {
+      "epoch": 1.0444444444444445,
+      "grad_norm": 1.4837637561015875,
+      "learning_rate": 1.7407407407407407e-07,
+      "loss": 0.8337,
+      "step": 564
+    },
+    {
+      "epoch": 1.0462962962962963,
+      "grad_norm": 1.4800586250040941,
+      "learning_rate": 1.7438271604938272e-07,
+      "loss": 0.8073,
+      "step": 565
+    },
+    {
+      "epoch": 1.048148148148148,
+      "grad_norm": 1.5010334413639592,
+      "learning_rate": 1.7469135802469134e-07,
+      "loss": 0.8723,
+      "step": 566
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.5246163461303288,
+      "learning_rate": 1.75e-07,
+      "loss": 0.9139,
+      "step": 567
+    },
+    {
+      "epoch": 1.0518518518518518,
+      "grad_norm": 1.5096409785175586,
+      "learning_rate": 1.7530864197530864e-07,
+      "loss": 0.7871,
+      "step": 568
+    },
+    {
+      "epoch": 1.0537037037037038,
+      "grad_norm": 1.7033015989867157,
+      "learning_rate": 1.7561728395061727e-07,
+      "loss": 0.8503,
+      "step": 569
+    },
+    {
+      "epoch": 1.0555555555555556,
+      "grad_norm": 1.8331304797783068,
+      "learning_rate": 1.7592592592592592e-07,
+      "loss": 0.8303,
+      "step": 570
+    },
+    {
+      "epoch": 1.0574074074074074,
+      "grad_norm": 2.657686596984486,
+      "learning_rate": 1.7623456790123457e-07,
+      "loss": 0.8945,
+      "step": 571
+    },
+    {
+      "epoch": 1.0592592592592593,
+      "grad_norm": 1.4015798649304902,
+      "learning_rate": 1.7654320987654322e-07,
+      "loss": 0.7581,
+      "step": 572
+    },
+    {
+      "epoch": 1.0611111111111111,
+      "grad_norm": 2.6653805849934242,
+      "learning_rate": 1.7685185185185184e-07,
+      "loss": 0.8767,
+      "step": 573
+    },
+    {
+      "epoch": 1.0629629629629629,
+      "grad_norm": 1.5964624363478677,
+      "learning_rate": 1.771604938271605e-07,
+      "loss": 0.8748,
+      "step": 574
+    },
+    {
+      "epoch": 1.0648148148148149,
+      "grad_norm": 1.656722307672164,
+      "learning_rate": 1.7746913580246914e-07,
+      "loss": 0.8251,
+      "step": 575
+    },
+    {
+      "epoch": 1.0666666666666667,
+      "grad_norm": 3.2261514557816664,
+      "learning_rate": 1.7777777777777776e-07,
+      "loss": 0.7857,
+      "step": 576
+    },
+    {
+      "epoch": 1.0685185185185184,
+      "grad_norm": 1.4007949119888583,
+      "learning_rate": 1.7808641975308641e-07,
+      "loss": 0.7776,
+      "step": 577
+    },
+    {
+      "epoch": 1.0703703703703704,
+      "grad_norm": 1.6083368995055551,
+      "learning_rate": 1.7839506172839504e-07,
+      "loss": 0.915,
+      "step": 578
+    },
+    {
+      "epoch": 1.0722222222222222,
+      "grad_norm": 1.6301023545495026,
+      "learning_rate": 1.787037037037037e-07,
+      "loss": 0.8663,
+      "step": 579
+    },
+    {
+      "epoch": 1.074074074074074,
+      "grad_norm": 1.5182789492763844,
+      "learning_rate": 1.7901234567901234e-07,
+      "loss": 0.832,
+      "step": 580
+    },
+    {
+      "epoch": 1.075925925925926,
+      "grad_norm": 1.7588871797491485,
+      "learning_rate": 1.79320987654321e-07,
+      "loss": 0.8315,
+      "step": 581
+    },
+    {
+      "epoch": 1.0777777777777777,
+      "grad_norm": 1.529148150776432,
+      "learning_rate": 1.796296296296296e-07,
+      "loss": 0.8357,
+      "step": 582
+    },
+    {
+      "epoch": 1.0796296296296297,
+      "grad_norm": 1.4254554207617063,
+      "learning_rate": 1.7993827160493826e-07,
+      "loss": 0.8107,
+      "step": 583
+    },
+    {
+      "epoch": 1.0814814814814815,
+      "grad_norm": 1.5027242983018325,
+      "learning_rate": 1.802469135802469e-07,
+      "loss": 0.8475,
+      "step": 584
+    },
+    {
+      "epoch": 1.0833333333333333,
+      "grad_norm": 1.4657323461202916,
+      "learning_rate": 1.8055555555555556e-07,
+      "loss": 0.8119,
+      "step": 585
+    },
+    {
+      "epoch": 1.0851851851851853,
+      "grad_norm": 1.4079806350705697,
+      "learning_rate": 1.8086419753086419e-07,
+      "loss": 0.8334,
+      "step": 586
+    },
+    {
+      "epoch": 1.087037037037037,
+      "grad_norm": 1.6242605991898924,
+      "learning_rate": 1.8117283950617284e-07,
+      "loss": 0.7631,
+      "step": 587
+    },
+    {
+      "epoch": 1.0888888888888888,
+      "grad_norm": 1.8769687310635081,
+      "learning_rate": 1.8148148148148149e-07,
+      "loss": 0.7859,
+      "step": 588
+    },
+    {
+      "epoch": 1.0907407407407408,
+      "grad_norm": 1.934757438979972,
+      "learning_rate": 1.8179012345679014e-07,
+      "loss": 0.8405,
+      "step": 589
+    },
+    {
+      "epoch": 1.0925925925925926,
+      "grad_norm": 1.5275657148498663,
+      "learning_rate": 1.8209876543209876e-07,
+      "loss": 0.8969,
+      "step": 590
+    },
+    {
+      "epoch": 1.0944444444444446,
+      "grad_norm": 1.535143711770453,
+      "learning_rate": 1.824074074074074e-07,
+      "loss": 0.7807,
+      "step": 591
+    },
+    {
+      "epoch": 1.0962962962962963,
+      "grad_norm": 1.7247098676492747,
+      "learning_rate": 1.8271604938271603e-07,
+      "loss": 0.8063,
+      "step": 592
+    },
+    {
+      "epoch": 1.098148148148148,
+      "grad_norm": 1.8582360291923699,
+      "learning_rate": 1.8302469135802468e-07,
+      "loss": 0.8432,
+      "step": 593
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.4459341873170923,
+      "learning_rate": 1.833333333333333e-07,
+      "loss": 0.8294,
+      "step": 594
+    },
+    {
+      "epoch": 1.1018518518518519,
+      "grad_norm": 1.6402085048812947,
+      "learning_rate": 1.8364197530864196e-07,
+      "loss": 0.8303,
+      "step": 595
+    },
+    {
+      "epoch": 1.1037037037037036,
+      "grad_norm": 1.5383980751937993,
+      "learning_rate": 1.839506172839506e-07,
+      "loss": 0.843,
+      "step": 596
+    },
+    {
+      "epoch": 1.1055555555555556,
+      "grad_norm": 1.816409667660388,
+      "learning_rate": 1.8425925925925926e-07,
+      "loss": 0.8793,
+      "step": 597
+    },
+    {
+      "epoch": 1.1074074074074074,
+      "grad_norm": 1.6956364381623301,
+      "learning_rate": 1.8456790123456788e-07,
+      "loss": 0.9443,
+      "step": 598
+    },
+    {
+      "epoch": 1.1092592592592592,
+      "grad_norm": 2.328299425083654,
+      "learning_rate": 1.8487654320987653e-07,
+      "loss": 0.8211,
+      "step": 599
+    },
+    {
+      "epoch": 1.1111111111111112,
+      "grad_norm": 1.6009889909653843,
+      "learning_rate": 1.8518518518518518e-07,
+      "loss": 0.8219,
+      "step": 600
+    },
+    {
+      "epoch": 1.112962962962963,
+      "grad_norm": 1.7262702252944861,
+      "learning_rate": 1.8549382716049383e-07,
+      "loss": 0.8141,
+      "step": 601
+    },
+    {
+      "epoch": 1.1148148148148147,
+      "grad_norm": 2.175237774059329,
+      "learning_rate": 1.8580246913580246e-07,
+      "loss": 0.8183,
+      "step": 602
+    },
+    {
+      "epoch": 1.1166666666666667,
+      "grad_norm": 1.6705397958700674,
+      "learning_rate": 1.861111111111111e-07,
+      "loss": 0.8491,
+      "step": 603
+    },
+    {
+      "epoch": 1.1185185185185185,
+      "grad_norm": 1.4166684807479268,
+      "learning_rate": 1.8641975308641976e-07,
+      "loss": 0.7582,
+      "step": 604
+    },
+    {
+      "epoch": 1.1203703703703705,
+      "grad_norm": 4.789621688237432,
+      "learning_rate": 1.867283950617284e-07,
+      "loss": 0.8323,
+      "step": 605
+    },
+    {
+      "epoch": 1.1222222222222222,
+      "grad_norm": 1.542870095454017,
+      "learning_rate": 1.8703703703703703e-07,
+      "loss": 0.82,
+      "step": 606
+    },
+    {
+      "epoch": 1.124074074074074,
+      "grad_norm": 1.4251956480393917,
+      "learning_rate": 1.8734567901234568e-07,
+      "loss": 0.8415,
+      "step": 607
+    },
+    {
+      "epoch": 1.125925925925926,
+      "grad_norm": 1.6725642490505486,
+      "learning_rate": 1.876543209876543e-07,
+      "loss": 0.8563,
+      "step": 608
+    },
+    {
+      "epoch": 1.1277777777777778,
+      "grad_norm": 1.8001724023147403,
+      "learning_rate": 1.8796296296296295e-07,
+      "loss": 0.7856,
+      "step": 609
+    },
+    {
+      "epoch": 1.1296296296296295,
+      "grad_norm": 1.5777165127139179,
+      "learning_rate": 1.8827160493827158e-07,
+      "loss": 0.8428,
+      "step": 610
+    },
+    {
+      "epoch": 1.1314814814814815,
+      "grad_norm": 1.5761579111337034,
+      "learning_rate": 1.8858024691358023e-07,
+      "loss": 0.8513,
+      "step": 611
+    },
+    {
+      "epoch": 1.1333333333333333,
+      "grad_norm": 1.8242570335542432,
+      "learning_rate": 1.8888888888888888e-07,
+      "loss": 0.8156,
+      "step": 612
+    },
+    {
+      "epoch": 1.1351851851851853,
+      "grad_norm": 1.530691819395211,
+      "learning_rate": 1.8919753086419753e-07,
+      "loss": 0.7881,
+      "step": 613
+    },
+    {
+      "epoch": 1.137037037037037,
+      "grad_norm": 1.9164515154073438,
+      "learning_rate": 1.8950617283950615e-07,
+      "loss": 0.8257,
+      "step": 614
+    },
+    {
+      "epoch": 1.1388888888888888,
+      "grad_norm": 1.4929066752543092,
+      "learning_rate": 1.898148148148148e-07,
+      "loss": 0.8359,
+      "step": 615
+    },
+    {
+      "epoch": 1.1407407407407408,
+      "grad_norm": 1.4081027554450505,
+      "learning_rate": 1.9012345679012345e-07,
+      "loss": 0.8767,
+      "step": 616
+    },
+    {
+      "epoch": 1.1425925925925926,
+      "grad_norm": 1.5770738488789227,
+      "learning_rate": 1.904320987654321e-07,
+      "loss": 0.8134,
+      "step": 617
+    },
+    {
+      "epoch": 1.1444444444444444,
+      "grad_norm": 1.7184457454114574,
+      "learning_rate": 1.9074074074074073e-07,
+      "loss": 0.8596,
+      "step": 618
+    },
+    {
+      "epoch": 1.1462962962962964,
+      "grad_norm": 1.4055190812277196,
+      "learning_rate": 1.9104938271604938e-07,
+      "loss": 0.823,
+      "step": 619
+    },
+    {
+      "epoch": 1.1481481481481481,
+      "grad_norm": 1.348126776083916,
+      "learning_rate": 1.9135802469135803e-07,
+      "loss": 0.8183,
+      "step": 620
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.5088465269448812,
+      "learning_rate": 1.9166666666666668e-07,
+      "loss": 0.7821,
+      "step": 621
+    },
+    {
+      "epoch": 1.151851851851852,
+      "grad_norm": 1.7632370443667849,
+      "learning_rate": 1.919753086419753e-07,
+      "loss": 0.84,
+      "step": 622
+    },
+    {
+      "epoch": 1.1537037037037037,
+      "grad_norm": 1.9255514334517234,
+      "learning_rate": 1.9228395061728395e-07,
+      "loss": 0.7746,
+      "step": 623
+    },
+    {
+      "epoch": 1.1555555555555554,
+      "grad_norm": 1.6286763962754056,
+      "learning_rate": 1.9259259259259257e-07,
+      "loss": 0.8576,
+      "step": 624
+    },
+    {
+      "epoch": 1.1574074074074074,
+      "grad_norm": 1.6633722663137955,
+      "learning_rate": 1.9290123456790122e-07,
+      "loss": 0.8155,
+      "step": 625
+    },
+    {
+      "epoch": 1.1592592592592592,
+      "grad_norm": 1.4192426771058053,
+      "learning_rate": 1.9320987654320985e-07,
+      "loss": 0.8083,
+      "step": 626
+    },
+    {
+      "epoch": 1.1611111111111112,
+      "grad_norm": 1.4440661199091922,
+      "learning_rate": 1.935185185185185e-07,
+      "loss": 0.8555,
+      "step": 627
+    },
+    {
+      "epoch": 1.162962962962963,
+      "grad_norm": 1.6233104448677302,
+      "learning_rate": 1.9382716049382715e-07,
+      "loss": 0.8894,
+      "step": 628
+    },
+    {
+      "epoch": 1.1648148148148147,
+      "grad_norm": 1.4165073974240723,
+      "learning_rate": 1.941358024691358e-07,
+      "loss": 0.8253,
+      "step": 629
+    },
+    {
+      "epoch": 1.1666666666666667,
+      "grad_norm": 1.628051655179612,
+      "learning_rate": 1.9444444444444442e-07,
+      "loss": 0.8778,
+      "step": 630
+    },
+    {
+      "epoch": 1.1685185185185185,
+      "grad_norm": 1.5499826920909507,
+      "learning_rate": 1.9475308641975307e-07,
+      "loss": 0.8489,
+      "step": 631
+    },
+    {
+      "epoch": 1.1703703703703703,
+      "grad_norm": 1.962979237347802,
+      "learning_rate": 1.9506172839506172e-07,
+      "loss": 0.7925,
+      "step": 632
+    },
+    {
+      "epoch": 1.1722222222222223,
+      "grad_norm": 1.7591379566308454,
+      "learning_rate": 1.9537037037037037e-07,
+      "loss": 0.8667,
+      "step": 633
+    },
+    {
+      "epoch": 1.174074074074074,
+      "grad_norm": 2.460868892243051,
+      "learning_rate": 1.95679012345679e-07,
+      "loss": 0.8494,
+      "step": 634
+    },
+    {
+      "epoch": 1.175925925925926,
+      "grad_norm": 1.8877059014847026,
+      "learning_rate": 1.9598765432098765e-07,
+      "loss": 0.8017,
+      "step": 635
+    },
+    {
+      "epoch": 1.1777777777777778,
+      "grad_norm": 1.7045657591829686,
+      "learning_rate": 1.962962962962963e-07,
+      "loss": 0.8192,
+      "step": 636
+    },
+    {
+      "epoch": 1.1796296296296296,
+      "grad_norm": 1.5452640381715912,
+      "learning_rate": 1.9660493827160495e-07,
+      "loss": 0.7966,
+      "step": 637
+    },
+    {
+      "epoch": 1.1814814814814816,
+      "grad_norm": 1.732657264409647,
+      "learning_rate": 1.9691358024691357e-07,
+      "loss": 0.8075,
+      "step": 638
+    },
+    {
+      "epoch": 1.1833333333333333,
+      "grad_norm": 1.4526087769238891,
+      "learning_rate": 1.9722222222222222e-07,
+      "loss": 0.8596,
+      "step": 639
+    },
+    {
+      "epoch": 1.1851851851851851,
+      "grad_norm": 1.601317179763928,
+      "learning_rate": 1.9753086419753084e-07,
+      "loss": 0.7948,
+      "step": 640
+    },
+    {
+      "epoch": 1.1870370370370371,
+      "grad_norm": 1.5937199109006792,
+      "learning_rate": 1.978395061728395e-07,
+      "loss": 0.8487,
+      "step": 641
+    },
+    {
+      "epoch": 1.1888888888888889,
+      "grad_norm": 1.418344467800995,
+      "learning_rate": 1.9814814814814814e-07,
+      "loss": 0.8215,
+      "step": 642
+    },
+    {
+      "epoch": 1.1907407407407407,
+      "grad_norm": 1.3326736296022426,
+      "learning_rate": 1.9845679012345677e-07,
+      "loss": 0.7929,
+      "step": 643
+    },
+    {
+      "epoch": 1.1925925925925926,
+      "grad_norm": 1.8151537792443233,
+      "learning_rate": 1.9876543209876542e-07,
+      "loss": 0.8282,
+      "step": 644
+    },
+    {
+      "epoch": 1.1944444444444444,
+      "grad_norm": 1.6758174444592453,
+      "learning_rate": 1.9907407407407407e-07,
+      "loss": 0.769,
+      "step": 645
+    },
+    {
+      "epoch": 1.1962962962962962,
+      "grad_norm": 1.4288624957309835,
+      "learning_rate": 1.9938271604938272e-07,
+      "loss": 0.812,
+      "step": 646
+    },
+    {
+      "epoch": 1.1981481481481482,
+      "grad_norm": 2.2318233017324633,
+      "learning_rate": 1.9969135802469134e-07,
+      "loss": 0.7757,
+      "step": 647
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.673189714917265,
+      "learning_rate": 2e-07,
+      "loss": 0.8142,
+      "step": 648
+    },
+    {
+      "epoch": 1.201851851851852,
+      "grad_norm": 1.81829191197008,
+      "learning_rate": 1.9999999887586419e-07,
+      "loss": 0.7731,
+      "step": 649
+    },
+    {
+      "epoch": 1.2037037037037037,
+      "grad_norm": 1.676121940820495,
+      "learning_rate": 1.999999955034568e-07,
+      "loss": 0.827,
+      "step": 650
+    },
+    {
+      "epoch": 1.2055555555555555,
+      "grad_norm": 1.3725717260554386,
+      "learning_rate": 1.9999998988277795e-07,
+      "loss": 0.7714,
+      "step": 651
+    },
+    {
+      "epoch": 1.2074074074074075,
+      "grad_norm": 1.5384999249290476,
+      "learning_rate": 1.9999998201382774e-07,
+      "loss": 0.8222,
+      "step": 652
+    },
+    {
+      "epoch": 1.2092592592592593,
+      "grad_norm": 1.3696578068453862,
+      "learning_rate": 1.9999997189660628e-07,
+      "loss": 0.8165,
+      "step": 653
+    },
+    {
+      "epoch": 1.211111111111111,
+      "grad_norm": 1.855368937170554,
+      "learning_rate": 1.9999995953111393e-07,
+      "loss": 0.8613,
+      "step": 654
+    },
+    {
+      "epoch": 1.212962962962963,
+      "grad_norm": 1.2370747321595208,
+      "learning_rate": 1.9999994491735084e-07,
+      "loss": 0.768,
+      "step": 655
+    },
+    {
+      "epoch": 1.2148148148148148,
+      "grad_norm": 1.5050274168084548,
+      "learning_rate": 1.9999992805531741e-07,
+      "loss": 0.796,
+      "step": 656
+    },
+    {
+      "epoch": 1.2166666666666668,
+      "grad_norm": 1.3984355995590978,
+      "learning_rate": 1.99999908945014e-07,
+      "loss": 0.8355,
+      "step": 657
+    },
+    {
+      "epoch": 1.2185185185185186,
+      "grad_norm": 2.0841481858806574,
+      "learning_rate": 1.9999988758644104e-07,
+      "loss": 0.8257,
+      "step": 658
+    },
+    {
+      "epoch": 1.2203703703703703,
+      "grad_norm": 1.280201023265319,
+      "learning_rate": 1.9999986397959903e-07,
+      "loss": 0.8096,
+      "step": 659
+    },
+    {
+      "epoch": 1.2222222222222223,
+      "grad_norm": 1.4407652952391021,
+      "learning_rate": 1.9999983812448847e-07,
+      "loss": 0.8393,
+      "step": 660
+    },
+    {
+      "epoch": 1.224074074074074,
+      "grad_norm": 4.297211053403042,
+      "learning_rate": 1.9999981002110996e-07,
+      "loss": 0.8565,
+      "step": 661
+    },
+    {
+      "epoch": 1.2259259259259259,
+      "grad_norm": 1.3281484149102734,
+      "learning_rate": 1.999997796694641e-07,
+      "loss": 0.8095,
+      "step": 662
+    },
+    {
+      "epoch": 1.2277777777777779,
+      "grad_norm": 1.4225231105051765,
+      "learning_rate": 1.999997470695516e-07,
+      "loss": 0.8483,
+      "step": 663
+    },
+    {
+      "epoch": 1.2296296296296296,
+      "grad_norm": 2.041529020910693,
+      "learning_rate": 1.999997122213732e-07,
+      "loss": 0.8252,
+      "step": 664
+    },
+    {
+      "epoch": 1.2314814814814814,
+      "grad_norm": 1.3320722685618107,
+      "learning_rate": 1.9999967512492969e-07,
+      "loss": 0.8237,
+      "step": 665
+    },
+    {
+      "epoch": 1.2333333333333334,
+      "grad_norm": 1.6871437488594683,
+      "learning_rate": 1.9999963578022186e-07,
+      "loss": 0.851,
+      "step": 666
+    },
+    {
+      "epoch": 1.2351851851851852,
+      "grad_norm": 1.5737732007069218,
+      "learning_rate": 1.9999959418725066e-07,
+      "loss": 0.8535,
+      "step": 667
+    },
+    {
+      "epoch": 1.237037037037037,
+      "grad_norm": 2.0286472424418567,
+      "learning_rate": 1.9999955034601696e-07,
+      "loss": 0.8725,
+      "step": 668
+    },
+    {
+      "epoch": 1.238888888888889,
+      "grad_norm": 3.61800278877449,
+      "learning_rate": 1.9999950425652178e-07,
+      "loss": 0.7998,
+      "step": 669
+    },
+    {
+      "epoch": 1.2407407407407407,
+      "grad_norm": 2.8285737528671437,
+      "learning_rate": 1.9999945591876616e-07,
+      "loss": 0.834,
+      "step": 670
+    },
+    {
+      "epoch": 1.2425925925925925,
+      "grad_norm": 1.6329172129507732,
+      "learning_rate": 1.9999940533275115e-07,
+      "loss": 0.7917,
+      "step": 671
+    },
+    {
+      "epoch": 1.2444444444444445,
+      "grad_norm": 4.083465952931434,
+      "learning_rate": 1.9999935249847796e-07,
+      "loss": 0.8115,
+      "step": 672
+    },
+    {
+      "epoch": 1.2462962962962962,
+      "grad_norm": 1.5603826600388242,
+      "learning_rate": 1.999992974159477e-07,
+      "loss": 0.8261,
+      "step": 673
+    },
+    {
+      "epoch": 1.2481481481481482,
+      "grad_norm": 2.1610248517678112,
+      "learning_rate": 1.999992400851616e-07,
+      "loss": 0.8175,
+      "step": 674
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.301137053762421,
+      "learning_rate": 1.9999918050612107e-07,
+      "loss": 0.7972,
+      "step": 675
+    },
+    {
+      "epoch": 1.2518518518518518,
+      "grad_norm": 1.4350718372391504,
+      "learning_rate": 1.999991186788273e-07,
+      "loss": 0.8086,
+      "step": 676
+    },
+    {
+      "epoch": 1.2537037037037038,
+      "grad_norm": 1.3996355983978452,
+      "learning_rate": 1.999990546032818e-07,
+      "loss": 0.8132,
+      "step": 677
+    },
+    {
+      "epoch": 1.2555555555555555,
+      "grad_norm": 2.046076241863195,
+      "learning_rate": 1.999989882794859e-07,
+      "loss": 0.825,
+      "step": 678
+    },
+    {
+      "epoch": 1.2574074074074075,
+      "grad_norm": 1.9174275837783201,
+      "learning_rate": 1.9999891970744122e-07,
+      "loss": 0.8017,
+      "step": 679
+    },
+    {
+      "epoch": 1.2592592592592593,
+      "grad_norm": 1.5740862949497096,
+      "learning_rate": 1.9999884888714916e-07,
+      "loss": 0.8454,
+      "step": 680
+    },
+    {
+      "epoch": 1.261111111111111,
+      "grad_norm": 1.573943685824295,
+      "learning_rate": 1.9999877581861147e-07,
+      "loss": 0.8092,
+      "step": 681
+    },
+    {
+      "epoch": 1.262962962962963,
+      "grad_norm": 1.858044964623011,
+      "learning_rate": 1.9999870050182963e-07,
+      "loss": 0.8688,
+      "step": 682
+    },
+    {
+      "epoch": 1.2648148148148148,
+      "grad_norm": 1.608219461922296,
+      "learning_rate": 1.9999862293680544e-07,
+      "loss": 0.8168,
+      "step": 683
+    },
+    {
+      "epoch": 1.2666666666666666,
+      "grad_norm": 1.3532272377834922,
+      "learning_rate": 1.9999854312354063e-07,
+      "loss": 0.825,
+      "step": 684
+    },
+    {
+      "epoch": 1.2685185185185186,
+      "grad_norm": 1.5895271404188862,
+      "learning_rate": 1.9999846106203698e-07,
+      "loss": 0.8429,
+      "step": 685
+    },
+    {
+      "epoch": 1.2703703703703704,
+      "grad_norm": 17.48759201834969,
+      "learning_rate": 1.999983767522963e-07,
+      "loss": 0.7783,
+      "step": 686
+    },
+    {
+      "epoch": 1.2722222222222221,
+      "grad_norm": 1.5828554984901237,
+      "learning_rate": 1.9999829019432056e-07,
+      "loss": 0.7948,
+      "step": 687
+    },
+    {
+      "epoch": 1.2740740740740741,
+      "grad_norm": 2.701786831731794,
+      "learning_rate": 1.9999820138811164e-07,
+      "loss": 0.767,
+      "step": 688
+    },
+    {
+      "epoch": 1.275925925925926,
+      "grad_norm": 1.5710332993008467,
+      "learning_rate": 1.9999811033367155e-07,
+      "loss": 0.8305,
+      "step": 689
+    },
+    {
+      "epoch": 1.2777777777777777,
+      "grad_norm": 1.6722669093862321,
+      "learning_rate": 1.9999801703100236e-07,
+      "loss": 0.8075,
+      "step": 690
+    },
+    {
+      "epoch": 1.2796296296296297,
+      "grad_norm": 1.3571600013004477,
+      "learning_rate": 1.9999792148010616e-07,
+      "loss": 0.785,
+      "step": 691
+    },
+    {
+      "epoch": 1.2814814814814814,
+      "grad_norm": 1.7584995967761679,
+      "learning_rate": 1.999978236809851e-07,
+      "loss": 0.8275,
+      "step": 692
+    },
+    {
+      "epoch": 1.2833333333333332,
+      "grad_norm": 1.6908576105463775,
+      "learning_rate": 1.9999772363364138e-07,
+      "loss": 0.7764,
+      "step": 693
+    },
+    {
+      "epoch": 1.2851851851851852,
+      "grad_norm": 1.8336220147272548,
+      "learning_rate": 1.9999762133807723e-07,
+      "loss": 0.7871,
+      "step": 694
+    },
+    {
+      "epoch": 1.287037037037037,
+      "grad_norm": 1.5197665661563555,
+      "learning_rate": 1.9999751679429496e-07,
+      "loss": 0.8317,
+      "step": 695
+    },
+    {
+      "epoch": 1.2888888888888888,
+      "grad_norm": 1.392921843686158,
+      "learning_rate": 1.9999741000229693e-07,
+      "loss": 0.8176,
+      "step": 696
+    },
+    {
+      "epoch": 1.2907407407407407,
+      "grad_norm": 1.7142334810776394,
+      "learning_rate": 1.9999730096208553e-07,
+      "loss": 0.7566,
+      "step": 697
+    },
+    {
+      "epoch": 1.2925925925925925,
+      "grad_norm": 1.33257801620074,
+      "learning_rate": 1.999971896736632e-07,
+      "loss": 0.8202,
+      "step": 698
+    },
+    {
+      "epoch": 1.2944444444444445,
+      "grad_norm": 1.357846537448651,
+      "learning_rate": 1.9999707613703246e-07,
+      "loss": 0.8475,
+      "step": 699
+    },
+    {
+      "epoch": 1.2962962962962963,
+      "grad_norm": 2.933054090341049,
+      "learning_rate": 1.999969603521959e-07,
+      "loss": 0.8341,
+      "step": 700
+    },
+    {
+      "epoch": 1.2981481481481483,
+      "grad_norm": 1.346188543492007,
+      "learning_rate": 1.9999684231915606e-07,
+      "loss": 0.763,
+      "step": 701
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.3967016006284687,
+      "learning_rate": 1.9999672203791561e-07,
+      "loss": 0.8189,
+      "step": 702
+    },
+    {
+      "epoch": 1.3018518518518518,
+      "grad_norm": 1.5955327686122693,
+      "learning_rate": 1.9999659950847728e-07,
+      "loss": 0.8343,
+      "step": 703
+    },
+    {
+      "epoch": 1.3037037037037038,
+      "grad_norm": 1.5976141423974135,
+      "learning_rate": 1.999964747308438e-07,
+      "loss": 0.8329,
+      "step": 704
+    },
+    {
+      "epoch": 1.3055555555555556,
+      "grad_norm": 1.4392796070897462,
+      "learning_rate": 1.9999634770501798e-07,
+      "loss": 0.7379,
+      "step": 705
+    },
+    {
+      "epoch": 1.3074074074074074,
+      "grad_norm": 1.584308070732416,
+      "learning_rate": 1.9999621843100267e-07,
+      "loss": 0.8115,
+      "step": 706
+    },
+    {
+      "epoch": 1.3092592592592593,
+      "grad_norm": 2.702347813899695,
+      "learning_rate": 1.999960869088008e-07,
+      "loss": 0.7978,
+      "step": 707
+    },
+    {
+      "epoch": 1.3111111111111111,
+      "grad_norm": 1.4954508281545915,
+      "learning_rate": 1.999959531384153e-07,
+      "loss": 0.8064,
+      "step": 708
+    },
+    {
+      "epoch": 1.3129629629629629,
+      "grad_norm": 1.7314897367970046,
+      "learning_rate": 1.9999581711984916e-07,
+      "loss": 0.8648,
+      "step": 709
+    },
+    {
+      "epoch": 1.3148148148148149,
+      "grad_norm": 1.8234726133694985,
+      "learning_rate": 1.9999567885310552e-07,
+      "loss": 0.7563,
+      "step": 710
+    },
+    {
+      "epoch": 1.3166666666666667,
+      "grad_norm": 2.0815322421777607,
+      "learning_rate": 1.999955383381874e-07,
+      "loss": 0.8034,
+      "step": 711
+    },
+    {
+      "epoch": 1.3185185185185184,
+      "grad_norm": 1.452416477738848,
+      "learning_rate": 1.99995395575098e-07,
+      "loss": 0.8275,
+      "step": 712
+    },
+    {
+      "epoch": 1.3203703703703704,
+      "grad_norm": 1.385547178875855,
+      "learning_rate": 1.9999525056384052e-07,
+      "loss": 0.7986,
+      "step": 713
+    },
+    {
+      "epoch": 1.3222222222222222,
+      "grad_norm": 1.6018152115518913,
+      "learning_rate": 1.9999510330441826e-07,
+      "loss": 0.7761,
+      "step": 714
+    },
+    {
+      "epoch": 1.324074074074074,
+      "grad_norm": 2.613760755025956,
+      "learning_rate": 1.9999495379683444e-07,
+      "loss": 0.7934,
+      "step": 715
+    },
+    {
+      "epoch": 1.325925925925926,
+      "grad_norm": 1.4548932715290424,
+      "learning_rate": 1.9999480204109251e-07,
+      "loss": 0.7822,
+      "step": 716
+    },
+    {
+      "epoch": 1.3277777777777777,
+      "grad_norm": 1.3911296980557761,
+      "learning_rate": 1.9999464803719587e-07,
+      "loss": 0.789,
+      "step": 717
+    },
+    {
+      "epoch": 1.3296296296296295,
+      "grad_norm": 1.383933108423946,
+      "learning_rate": 1.9999449178514792e-07,
+      "loss": 0.7718,
+      "step": 718
+    },
+    {
+      "epoch": 1.3314814814814815,
+      "grad_norm": 3.1847143238699864,
+      "learning_rate": 1.999943332849522e-07,
+      "loss": 0.814,
+      "step": 719
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 1.594065175286285,
+      "learning_rate": 1.9999417253661234e-07,
+      "loss": 0.825,
+      "step": 720
+    },
+    {
+      "epoch": 1.3351851851851853,
+      "grad_norm": 1.5185822785528786,
+      "learning_rate": 1.9999400954013188e-07,
+      "loss": 0.8408,
+      "step": 721
+    },
+    {
+      "epoch": 1.337037037037037,
+      "grad_norm": 1.6331028949400026,
+      "learning_rate": 1.999938442955145e-07,
+      "loss": 0.7784,
+      "step": 722
+    },
+    {
+      "epoch": 1.338888888888889,
+      "grad_norm": 1.4885723201648686,
+      "learning_rate": 1.999936768027639e-07,
+      "loss": 0.8108,
+      "step": 723
+    },
+    {
+      "epoch": 1.3407407407407408,
+      "grad_norm": 1.6253958747394246,
+      "learning_rate": 1.999935070618839e-07,
+      "loss": 0.7606,
+      "step": 724
+    },
+    {
+      "epoch": 1.3425925925925926,
+      "grad_norm": 2.100753153108363,
+      "learning_rate": 1.9999333507287827e-07,
+      "loss": 0.7585,
+      "step": 725
+    },
+    {
+      "epoch": 1.3444444444444446,
+      "grad_norm": 1.711567579429311,
+      "learning_rate": 1.999931608357509e-07,
+      "loss": 0.8569,
+      "step": 726
+    },
+    {
+      "epoch": 1.3462962962962963,
+      "grad_norm": 2.731792389961523,
+      "learning_rate": 1.9999298435050568e-07,
+      "loss": 0.8126,
+      "step": 727
+    },
+    {
+      "epoch": 1.348148148148148,
+      "grad_norm": 1.6302884387160579,
+      "learning_rate": 1.999928056171466e-07,
+      "loss": 0.7948,
+      "step": 728
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.765748181345404,
+      "learning_rate": 1.999926246356777e-07,
+      "loss": 0.8708,
+      "step": 729
+    },
+    {
+      "epoch": 1.3518518518518519,
+      "grad_norm": 1.9756705535899186,
+      "learning_rate": 1.9999244140610298e-07,
+      "loss": 0.8435,
+      "step": 730
+    },
+    {
+      "epoch": 1.3537037037037036,
+      "grad_norm": 1.4228246423515123,
+      "learning_rate": 1.9999225592842665e-07,
+      "loss": 0.7985,
+      "step": 731
+    },
+    {
+      "epoch": 1.3555555555555556,
+      "grad_norm": 1.744488402830874,
+      "learning_rate": 1.999920682026528e-07,
+      "loss": 0.8642,
+      "step": 732
+    },
+    {
+      "epoch": 1.3574074074074074,
+      "grad_norm": 1.4106173918746343,
+      "learning_rate": 1.9999187822878568e-07,
+      "loss": 0.7934,
+      "step": 733
+    },
+    {
+      "epoch": 1.3592592592592592,
+      "grad_norm": 1.5603012222840908,
+      "learning_rate": 1.9999168600682958e-07,
+      "loss": 0.8443,
+      "step": 734
+    },
+    {
+      "epoch": 1.3611111111111112,
+      "grad_norm": 1.9684145208270543,
+      "learning_rate": 1.999914915367888e-07,
+      "loss": 0.6867,
+      "step": 735
+    },
+    {
+      "epoch": 1.362962962962963,
+      "grad_norm": 1.9078147431897725,
+      "learning_rate": 1.999912948186677e-07,
+      "loss": 0.835,
+      "step": 736
+    },
+    {
+      "epoch": 1.3648148148148147,
+      "grad_norm": 1.3041471586294773,
+      "learning_rate": 1.9999109585247075e-07,
+      "loss": 0.7823,
+      "step": 737
+    },
+    {
+      "epoch": 1.3666666666666667,
+      "grad_norm": 1.4188184952644083,
+      "learning_rate": 1.9999089463820237e-07,
+      "loss": 0.7751,
+      "step": 738
+    },
+    {
+      "epoch": 1.3685185185185185,
+      "grad_norm": 2.444346601690097,
+      "learning_rate": 1.9999069117586712e-07,
+      "loss": 0.8326,
+      "step": 739
+    },
+    {
+      "epoch": 1.3703703703703702,
+      "grad_norm": 1.3501257826270001,
+      "learning_rate": 1.9999048546546953e-07,
+      "loss": 0.8196,
+      "step": 740
+    },
+    {
+      "epoch": 1.3722222222222222,
+      "grad_norm": 2.3372204727965236,
+      "learning_rate": 1.9999027750701429e-07,
+      "loss": 0.8256,
+      "step": 741
+    },
+    {
+      "epoch": 1.374074074074074,
+      "grad_norm": 2.025133239373389,
+      "learning_rate": 1.9999006730050602e-07,
+      "loss": 0.7684,
+      "step": 742
+    },
+    {
+      "epoch": 1.375925925925926,
+      "grad_norm": 1.4560990570388062,
+      "learning_rate": 1.999898548459495e-07,
+      "loss": 0.8069,
+      "step": 743
+    },
+    {
+      "epoch": 1.3777777777777778,
+      "grad_norm": 1.587066567252244,
+      "learning_rate": 1.9998964014334944e-07,
+      "loss": 0.7644,
+      "step": 744
+    },
+    {
+      "epoch": 1.3796296296296298,
+      "grad_norm": 1.5048099316236232,
+      "learning_rate": 1.9998942319271075e-07,
+      "loss": 0.7822,
+      "step": 745
+    },
+    {
+      "epoch": 1.3814814814814815,
+      "grad_norm": 1.5456315308946729,
+      "learning_rate": 1.9998920399403822e-07,
+      "loss": 0.8821,
+      "step": 746
+    },
+    {
+      "epoch": 1.3833333333333333,
+      "grad_norm": 2.0098488030872033,
+      "learning_rate": 1.9998898254733683e-07,
+      "loss": 0.8151,
+      "step": 747
+    },
+    {
+      "epoch": 1.3851851851851853,
+      "grad_norm": 1.5106633218657948,
+      "learning_rate": 1.9998875885261156e-07,
+      "loss": 0.8001,
+      "step": 748
+    },
+    {
+      "epoch": 1.387037037037037,
+      "grad_norm": 1.4314783784907348,
+      "learning_rate": 1.999885329098674e-07,
+      "loss": 0.826,
+      "step": 749
+    },
+    {
+      "epoch": 1.3888888888888888,
+      "grad_norm": 1.7604892244147972,
+      "learning_rate": 1.9998830471910945e-07,
+      "loss": 0.8427,
+      "step": 750
+    },
+    {
+      "epoch": 1.3907407407407408,
+      "grad_norm": 1.4427224781548,
+      "learning_rate": 1.9998807428034285e-07,
+      "loss": 0.812,
+      "step": 751
+    },
+    {
+      "epoch": 1.3925925925925926,
+      "grad_norm": 1.3919650093002605,
+      "learning_rate": 1.999878415935728e-07,
+      "loss": 0.8232,
+      "step": 752
+    },
+    {
+      "epoch": 1.3944444444444444,
+      "grad_norm": 1.5123998153492724,
+      "learning_rate": 1.999876066588045e-07,
+      "loss": 0.7973,
+      "step": 753
+    },
+    {
+      "epoch": 1.3962962962962964,
+      "grad_norm": 2.023370613360696,
+      "learning_rate": 1.9998736947604323e-07,
+      "loss": 0.8064,
+      "step": 754
+    },
+    {
+      "epoch": 1.3981481481481481,
+      "grad_norm": 2.571057131225494,
+      "learning_rate": 1.9998713004529434e-07,
+      "loss": 0.7671,
+      "step": 755
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.2689081255007357,
+      "learning_rate": 1.9998688836656322e-07,
+      "loss": 0.761,
+      "step": 756
+    },
+    {
+      "epoch": 1.401851851851852,
+      "grad_norm": 1.4377393351999066,
+      "learning_rate": 1.9998664443985524e-07,
+      "loss": 0.745,
+      "step": 757
+    },
+    {
+      "epoch": 1.4037037037037037,
+      "grad_norm": 2.1939721261948253,
+      "learning_rate": 1.99986398265176e-07,
+      "loss": 0.7851,
+      "step": 758
+    },
+    {
+      "epoch": 1.4055555555555554,
+      "grad_norm": 1.4503461755590286,
+      "learning_rate": 1.999861498425309e-07,
+      "loss": 0.7936,
+      "step": 759
+    },
+    {
+      "epoch": 1.4074074074074074,
+      "grad_norm": 1.8394565547533492,
+      "learning_rate": 1.9998589917192564e-07,
+      "loss": 0.7988,
+      "step": 760
+    },
+    {
+      "epoch": 1.4092592592592592,
+      "grad_norm": 1.468303491569509,
+      "learning_rate": 1.999856462533658e-07,
+      "loss": 0.7708,
+      "step": 761
+    },
+    {
+      "epoch": 1.411111111111111,
+      "grad_norm": 2.0549287530631526,
+      "learning_rate": 1.9998539108685708e-07,
+      "loss": 0.7606,
+      "step": 762
+    },
+    {
+      "epoch": 1.412962962962963,
+      "grad_norm": 1.4655881735425278,
+      "learning_rate": 1.9998513367240523e-07,
+      "loss": 0.8103,
+      "step": 763
+    },
+    {
+      "epoch": 1.4148148148148147,
+      "grad_norm": 1.3635001556298394,
+      "learning_rate": 1.9998487401001597e-07,
+      "loss": 0.761,
+      "step": 764
+    },
+    {
+      "epoch": 1.4166666666666667,
+      "grad_norm": 1.3484868745303848,
+      "learning_rate": 1.9998461209969522e-07,
+      "loss": 0.7891,
+      "step": 765
+    },
+    {
+      "epoch": 1.4185185185185185,
+      "grad_norm": 2.381878464718892,
+      "learning_rate": 1.9998434794144884e-07,
+      "loss": 0.7819,
+      "step": 766
+    },
+    {
+      "epoch": 1.4203703703703705,
+      "grad_norm": 1.7386202947061842,
+      "learning_rate": 1.999840815352827e-07,
+      "loss": 0.8041,
+      "step": 767
+    },
+    {
+      "epoch": 1.4222222222222223,
+      "grad_norm": 1.3368181347470822,
+      "learning_rate": 1.9998381288120295e-07,
+      "loss": 0.8021,
+      "step": 768
+    },
+    {
+      "epoch": 1.424074074074074,
+      "grad_norm": 3.1464029983683677,
+      "learning_rate": 1.9998354197921548e-07,
+      "loss": 0.7841,
+      "step": 769
+    },
+    {
+      "epoch": 1.425925925925926,
+      "grad_norm": 1.8305186362597654,
+      "learning_rate": 1.9998326882932643e-07,
+      "loss": 0.7639,
+      "step": 770
+    },
+    {
+      "epoch": 1.4277777777777778,
+      "grad_norm": 1.2428686113792757,
+      "learning_rate": 1.9998299343154198e-07,
+      "loss": 0.7986,
+      "step": 771
+    },
+    {
+      "epoch": 1.4296296296296296,
+      "grad_norm": 1.392108608033091,
+      "learning_rate": 1.9998271578586827e-07,
+      "loss": 0.8157,
+      "step": 772
+    },
+    {
+      "epoch": 1.4314814814814816,
+      "grad_norm": 1.5540628226099635,
+      "learning_rate": 1.9998243589231153e-07,
+      "loss": 0.7705,
+      "step": 773
+    },
+    {
+      "epoch": 1.4333333333333333,
+      "grad_norm": 1.543105889651524,
+      "learning_rate": 1.9998215375087813e-07,
+      "loss": 0.7952,
+      "step": 774
+    },
+    {
+      "epoch": 1.4351851851851851,
+      "grad_norm": 1.3917823867151182,
+      "learning_rate": 1.9998186936157436e-07,
+      "loss": 0.8254,
+      "step": 775
+    },
+    {
+      "epoch": 1.4370370370370371,
+      "grad_norm": 1.2977219318503417,
+      "learning_rate": 1.999815827244066e-07,
+      "loss": 0.811,
+      "step": 776
+    },
+    {
+      "epoch": 1.4388888888888889,
+      "grad_norm": 1.4195135393158569,
+      "learning_rate": 1.9998129383938133e-07,
+      "loss": 0.7993,
+      "step": 777
+    },
+    {
+      "epoch": 1.4407407407407407,
+      "grad_norm": 1.332976070867669,
+      "learning_rate": 1.9998100270650502e-07,
+      "loss": 0.7471,
+      "step": 778
+    },
+    {
+      "epoch": 1.4425925925925926,
+      "grad_norm": 1.5295871560467964,
+      "learning_rate": 1.9998070932578422e-07,
+      "loss": 0.7723,
+      "step": 779
+    },
+    {
+      "epoch": 1.4444444444444444,
+      "grad_norm": 1.3320993191659307,
+      "learning_rate": 1.9998041369722555e-07,
+      "loss": 0.7905,
+      "step": 780
+    },
+    {
+      "epoch": 1.4462962962962962,
+      "grad_norm": 2.5443449059167342,
+      "learning_rate": 1.9998011582083564e-07,
+      "loss": 0.8171,
+      "step": 781
+    },
+    {
+      "epoch": 1.4481481481481482,
+      "grad_norm": 1.4307395946540447,
+      "learning_rate": 1.9997981569662116e-07,
+      "loss": 0.7868,
+      "step": 782
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.458921852302464,
+      "learning_rate": 1.9997951332458888e-07,
+      "loss": 0.8283,
+      "step": 783
+    },
+    {
+      "epoch": 1.4518518518518517,
+      "grad_norm": 1.3580288423255023,
+      "learning_rate": 1.9997920870474563e-07,
+      "loss": 0.862,
+      "step": 784
+    },
+    {
+      "epoch": 1.4537037037037037,
+      "grad_norm": 1.5810332384189063,
+      "learning_rate": 1.9997890183709824e-07,
+      "loss": 0.8137,
+      "step": 785
+    },
+    {
+      "epoch": 1.4555555555555555,
+      "grad_norm": 1.3458762592152296,
+      "learning_rate": 1.9997859272165354e-07,
+      "loss": 0.7453,
+      "step": 786
+    },
+    {
+      "epoch": 1.4574074074074075,
+      "grad_norm": 4.059567578444875,
+      "learning_rate": 1.9997828135841856e-07,
+      "loss": 0.7874,
+      "step": 787
+    },
+    {
+      "epoch": 1.4592592592592593,
+      "grad_norm": 1.9165797879044952,
+      "learning_rate": 1.999779677474003e-07,
+      "loss": 0.7749,
+      "step": 788
+    },
+    {
+      "epoch": 1.4611111111111112,
+      "grad_norm": 1.3348032667290637,
+      "learning_rate": 1.9997765188860577e-07,
+      "loss": 0.7775,
+      "step": 789
+    },
+    {
+      "epoch": 1.462962962962963,
+      "grad_norm": 1.2852606649162384,
+      "learning_rate": 1.999773337820421e-07,
+      "loss": 0.7656,
+      "step": 790
+    },
+    {
+      "epoch": 1.4648148148148148,
+      "grad_norm": 1.3187710472389504,
+      "learning_rate": 1.9997701342771642e-07,
+      "loss": 0.8221,
+      "step": 791
+    },
+    {
+      "epoch": 1.4666666666666668,
+      "grad_norm": 1.3164764119936931,
+      "learning_rate": 1.9997669082563595e-07,
+      "loss": 0.8143,
+      "step": 792
+    },
+    {
+      "epoch": 1.4685185185185186,
+      "grad_norm": 1.5709538616240728,
+      "learning_rate": 1.9997636597580794e-07,
+      "loss": 0.7649,
+      "step": 793
+    },
+    {
+      "epoch": 1.4703703703703703,
+      "grad_norm": 1.5347942294432204,
+      "learning_rate": 1.999760388782397e-07,
+      "loss": 0.7905,
+      "step": 794
+    },
+    {
+      "epoch": 1.4722222222222223,
+      "grad_norm": 6.655542854853261,
+      "learning_rate": 1.999757095329386e-07,
+      "loss": 0.7781,
+      "step": 795
+    },
+    {
+      "epoch": 1.474074074074074,
+      "grad_norm": 1.5316136147155046,
+      "learning_rate": 1.9997537793991194e-07,
+      "loss": 0.8179,
+      "step": 796
+    },
+    {
+      "epoch": 1.4759259259259259,
+      "grad_norm": 1.7833710510859966,
+      "learning_rate": 1.9997504409916733e-07,
+      "loss": 0.7802,
+      "step": 797
+    },
+    {
+      "epoch": 1.4777777777777779,
+      "grad_norm": 5.27321204738106,
+      "learning_rate": 1.9997470801071218e-07,
+      "loss": 0.7445,
+      "step": 798
+    },
+    {
+      "epoch": 1.4796296296296296,
+      "grad_norm": 1.6615755275028636,
+      "learning_rate": 1.9997436967455404e-07,
+      "loss": 0.7525,
+      "step": 799
+    },
+    {
+      "epoch": 1.4814814814814814,
+      "grad_norm": 1.3827060496974195,
+      "learning_rate": 1.9997402909070057e-07,
+      "loss": 0.8076,
+      "step": 800
+    },
+    {
+      "epoch": 1.4833333333333334,
+      "grad_norm": 1.515486984875506,
+      "learning_rate": 1.999736862591594e-07,
+      "loss": 0.7582,
+      "step": 801
+    },
+    {
+      "epoch": 1.4851851851851852,
+      "grad_norm": 2.1911955532337553,
+      "learning_rate": 1.9997334117993823e-07,
+      "loss": 0.7468,
+      "step": 802
+    },
+    {
+      "epoch": 1.487037037037037,
+      "grad_norm": 1.3137153001241144,
+      "learning_rate": 1.9997299385304483e-07,
+      "loss": 0.849,
+      "step": 803
+    },
+    {
+      "epoch": 1.488888888888889,
+      "grad_norm": 1.4033594381191727,
+      "learning_rate": 1.9997264427848698e-07,
+      "loss": 0.7699,
+      "step": 804
+    },
+    {
+      "epoch": 1.4907407407407407,
+      "grad_norm": 1.616486331551188,
+      "learning_rate": 1.999722924562726e-07,
+      "loss": 0.8119,
+      "step": 805
+    },
+    {
+      "epoch": 1.4925925925925925,
+      "grad_norm": 1.611378589097031,
+      "learning_rate": 1.9997193838640955e-07,
+      "loss": 0.7491,
+      "step": 806
+    },
+    {
+      "epoch": 1.4944444444444445,
+      "grad_norm": 1.8885748958730557,
+      "learning_rate": 1.9997158206890582e-07,
+      "loss": 0.7771,
+      "step": 807
+    },
+    {
+      "epoch": 1.4962962962962962,
+      "grad_norm": 1.2975278134562398,
+      "learning_rate": 1.9997122350376938e-07,
+      "loss": 0.8015,
+      "step": 808
+    },
+    {
+      "epoch": 1.4981481481481482,
+      "grad_norm": 1.5747918569426438,
+      "learning_rate": 1.9997086269100833e-07,
+      "loss": 0.8423,
+      "step": 809
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.8186690943705288,
+      "learning_rate": 1.999704996306308e-07,
+      "loss": 0.8081,
+      "step": 810
+    },
+    {
+      "epoch": 1.501851851851852,
+      "grad_norm": 1.3900499893368639,
+      "learning_rate": 1.999701343226449e-07,
+      "loss": 0.7831,
+      "step": 811
+    },
+    {
+      "epoch": 1.5037037037037035,
+      "grad_norm": 1.414262466865652,
+      "learning_rate": 1.9996976676705886e-07,
+      "loss": 0.7735,
+      "step": 812
+    },
+    {
+      "epoch": 1.5055555555555555,
+      "grad_norm": 1.3854954652081493,
+      "learning_rate": 1.9996939696388092e-07,
+      "loss": 0.8117,
+      "step": 813
+    },
+    {
+      "epoch": 1.5074074074074075,
+      "grad_norm": 1.3352991112748513,
+      "learning_rate": 1.9996902491311948e-07,
+      "loss": 0.759,
+      "step": 814
+    },
+    {
+      "epoch": 1.5092592592592593,
+      "grad_norm": 1.4606364588246132,
+      "learning_rate": 1.999686506147828e-07,
+      "loss": 0.7859,
+      "step": 815
+    },
+    {
+      "epoch": 1.511111111111111,
+      "grad_norm": 1.3072406803222494,
+      "learning_rate": 1.999682740688794e-07,
+      "loss": 0.7727,
+      "step": 816
+    },
+    {
+      "epoch": 1.512962962962963,
+      "grad_norm": 1.577789969006411,
+      "learning_rate": 1.9996789527541766e-07,
+      "loss": 0.7997,
+      "step": 817
+    },
+    {
+      "epoch": 1.5148148148148148,
+      "grad_norm": 1.3342314218130178,
+      "learning_rate": 1.9996751423440612e-07,
+      "loss": 0.7569,
+      "step": 818
+    },
+    {
+      "epoch": 1.5166666666666666,
+      "grad_norm": 1.4337330612761081,
+      "learning_rate": 1.9996713094585334e-07,
+      "loss": 0.8181,
+      "step": 819
+    },
+    {
+      "epoch": 1.5185185185185186,
+      "grad_norm": 1.4425600988022205,
+      "learning_rate": 1.9996674540976797e-07,
+      "loss": 0.8009,
+      "step": 820
+    },
+    {
+      "epoch": 1.5203703703703704,
+      "grad_norm": 1.3742195597196458,
+      "learning_rate": 1.9996635762615863e-07,
+      "loss": 0.8167,
+      "step": 821
+    },
+    {
+      "epoch": 1.5222222222222221,
+      "grad_norm": 1.5195254091486206,
+      "learning_rate": 1.999659675950341e-07,
+      "loss": 0.7807,
+      "step": 822
+    },
+    {
+      "epoch": 1.5240740740740741,
+      "grad_norm": 1.3234522103246977,
+      "learning_rate": 1.9996557531640312e-07,
+      "loss": 0.8112,
+      "step": 823
+    },
+    {
+      "epoch": 1.525925925925926,
+      "grad_norm": 1.4747702928793858,
+      "learning_rate": 1.9996518079027446e-07,
+      "loss": 0.7531,
+      "step": 824
+    },
+    {
+      "epoch": 1.5277777777777777,
+      "grad_norm": 1.4713620896887833,
+      "learning_rate": 1.9996478401665708e-07,
+      "loss": 0.7493,
+      "step": 825
+    },
+    {
+      "epoch": 1.5296296296296297,
+      "grad_norm": 1.332434865024635,
+      "learning_rate": 1.9996438499555982e-07,
+      "loss": 0.786,
+      "step": 826
+    },
+    {
+      "epoch": 1.5314814814814814,
+      "grad_norm": 1.3230372557402514,
+      "learning_rate": 1.999639837269917e-07,
+      "loss": 0.7977,
+      "step": 827
+    },
+    {
+      "epoch": 1.5333333333333332,
+      "grad_norm": 1.667462527734816,
+      "learning_rate": 1.9996358021096172e-07,
+      "loss": 0.7757,
+      "step": 828
+    },
+    {
+      "epoch": 1.5351851851851852,
+      "grad_norm": 1.4809443789626107,
+      "learning_rate": 1.99963174447479e-07,
+      "loss": 0.8128,
+      "step": 829
+    },
+    {
+      "epoch": 1.5370370370370372,
+      "grad_norm": 1.4015098922549134,
+      "learning_rate": 1.9996276643655258e-07,
+      "loss": 0.7927,
+      "step": 830
+    },
+    {
+      "epoch": 1.5388888888888888,
+      "grad_norm": 6.474043305591081,
+      "learning_rate": 1.999623561781917e-07,
+      "loss": 0.7773,
+      "step": 831
+    },
+    {
+      "epoch": 1.5407407407407407,
+      "grad_norm": 1.3965801879089659,
+      "learning_rate": 1.9996194367240554e-07,
+      "loss": 0.7864,
+      "step": 832
+    },
+    {
+      "epoch": 1.5425925925925927,
+      "grad_norm": 1.6855553465095316,
+      "learning_rate": 1.9996152891920342e-07,
+      "loss": 0.7943,
+      "step": 833
+    },
+    {
+      "epoch": 1.5444444444444443,
+      "grad_norm": 1.3397043658891168,
+      "learning_rate": 1.999611119185946e-07,
+      "loss": 0.7993,
+      "step": 834
+    },
+    {
+      "epoch": 1.5462962962962963,
+      "grad_norm": 1.6467530903268044,
+      "learning_rate": 1.9996069267058851e-07,
+      "loss": 0.8165,
+      "step": 835
+    },
+    {
+      "epoch": 1.5481481481481483,
+      "grad_norm": 1.379256390059123,
+      "learning_rate": 1.9996027117519458e-07,
+      "loss": 0.7634,
+      "step": 836
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.4035934399022019,
+      "learning_rate": 1.9995984743242226e-07,
+      "loss": 0.8012,
+      "step": 837
+    },
+    {
+      "epoch": 1.5518518518518518,
+      "grad_norm": 1.3337453757666426,
+      "learning_rate": 1.9995942144228108e-07,
+      "loss": 0.8022,
+      "step": 838
+    },
+    {
+      "epoch": 1.5537037037037038,
+      "grad_norm": 1.7528515777046711,
+      "learning_rate": 1.999589932047806e-07,
+      "loss": 0.76,
+      "step": 839
+    },
+    {
+      "epoch": 1.5555555555555556,
+      "grad_norm": 1.9684089615083693,
+      "learning_rate": 1.999585627199305e-07,
+      "loss": 0.8091,
+      "step": 840
+    },
+    {
+      "epoch": 1.5574074074074074,
+      "grad_norm": 1.427520378401312,
+      "learning_rate": 1.9995812998774038e-07,
+      "loss": 0.833,
+      "step": 841
+    },
+    {
+      "epoch": 1.5592592592592593,
+      "grad_norm": 1.3881160996675828,
+      "learning_rate": 1.9995769500822006e-07,
+      "loss": 0.7653,
+      "step": 842
+    },
+    {
+      "epoch": 1.5611111111111111,
+      "grad_norm": 2.2195578266559184,
+      "learning_rate": 1.9995725778137924e-07,
+      "loss": 0.8062,
+      "step": 843
+    },
+    {
+      "epoch": 1.5629629629629629,
+      "grad_norm": 1.2614323844099824,
+      "learning_rate": 1.9995681830722782e-07,
+      "loss": 0.786,
+      "step": 844
+    },
+    {
+      "epoch": 1.5648148148148149,
+      "grad_norm": 1.5655955039081937,
+      "learning_rate": 1.9995637658577562e-07,
+      "loss": 0.7901,
+      "step": 845
+    },
+    {
+      "epoch": 1.5666666666666667,
+      "grad_norm": 1.3021620993914482,
+      "learning_rate": 1.999559326170326e-07,
+      "loss": 0.8106,
+      "step": 846
+    },
+    {
+      "epoch": 1.5685185185185184,
+      "grad_norm": 1.442966557857159,
+      "learning_rate": 1.9995548640100878e-07,
+      "loss": 0.7262,
+      "step": 847
+    },
+    {
+      "epoch": 1.5703703703703704,
+      "grad_norm": 1.8039328844248892,
+      "learning_rate": 1.999550379377141e-07,
+      "loss": 0.7614,
+      "step": 848
+    },
+    {
+      "epoch": 1.5722222222222222,
+      "grad_norm": 1.5717220513889545,
+      "learning_rate": 1.999545872271587e-07,
+      "loss": 0.7749,
+      "step": 849
+    },
+    {
+      "epoch": 1.574074074074074,
+      "grad_norm": 1.4449487934438003,
+      "learning_rate": 1.999541342693527e-07,
+      "loss": 0.7997,
+      "step": 850
+    },
+    {
+      "epoch": 1.575925925925926,
+      "grad_norm": 2.3902711381662884,
+      "learning_rate": 1.9995367906430634e-07,
+      "loss": 0.7744,
+      "step": 851
+    },
+    {
+      "epoch": 1.5777777777777777,
+      "grad_norm": 1.5521337852002064,
+      "learning_rate": 1.9995322161202974e-07,
+      "loss": 0.7766,
+      "step": 852
+    },
+    {
+      "epoch": 1.5796296296296295,
+      "grad_norm": 1.4810952028164313,
+      "learning_rate": 1.999527619125333e-07,
+      "loss": 0.792,
+      "step": 853
+    },
+    {
+      "epoch": 1.5814814814814815,
+      "grad_norm": 1.873965974787256,
+      "learning_rate": 1.9995229996582727e-07,
+      "loss": 0.7844,
+      "step": 854
+    },
+    {
+      "epoch": 1.5833333333333335,
+      "grad_norm": 1.5975769776683897,
+      "learning_rate": 1.999518357719221e-07,
+      "loss": 0.7781,
+      "step": 855
+    },
+    {
+      "epoch": 1.585185185185185,
+      "grad_norm": 1.3433087398314394,
+      "learning_rate": 1.9995136933082818e-07,
+      "loss": 0.7814,
+      "step": 856
+    },
+    {
+      "epoch": 1.587037037037037,
+      "grad_norm": 1.4733235690859368,
+      "learning_rate": 1.9995090064255601e-07,
+      "loss": 0.8272,
+      "step": 857
+    },
+    {
+      "epoch": 1.588888888888889,
+      "grad_norm": 1.538597131002794,
+      "learning_rate": 1.9995042970711614e-07,
+      "loss": 0.7559,
+      "step": 858
+    },
+    {
+      "epoch": 1.5907407407407408,
+      "grad_norm": 1.8744026521188375,
+      "learning_rate": 1.9994995652451916e-07,
+      "loss": 0.735,
+      "step": 859
+    },
+    {
+      "epoch": 1.5925925925925926,
+      "grad_norm": 1.3854505744333534,
+      "learning_rate": 1.9994948109477567e-07,
+      "loss": 0.8418,
+      "step": 860
+    },
+    {
+      "epoch": 1.5944444444444446,
+      "grad_norm": 1.4661803675537535,
+      "learning_rate": 1.9994900341789643e-07,
+      "loss": 0.7467,
+      "step": 861
+    },
+    {
+      "epoch": 1.5962962962962963,
+      "grad_norm": 1.3205550748055843,
+      "learning_rate": 1.9994852349389208e-07,
+      "loss": 0.812,
+      "step": 862
+    },
+    {
+      "epoch": 1.598148148148148,
+      "grad_norm": 1.453642466089374,
+      "learning_rate": 1.999480413227735e-07,
+      "loss": 0.762,
+      "step": 863
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.5143196961255514,
+      "learning_rate": 1.999475569045515e-07,
+      "loss": 0.7903,
+      "step": 864
+    },
+    {
+      "epoch": 1.6018518518518519,
+      "grad_norm": 1.3319493284980677,
+      "learning_rate": 1.9994707023923694e-07,
+      "loss": 0.797,
+      "step": 865
+    },
+    {
+      "epoch": 1.6037037037037036,
+      "grad_norm": 1.8022799954583217,
+      "learning_rate": 1.999465813268408e-07,
+      "loss": 0.8108,
+      "step": 866
+    },
+    {
+      "epoch": 1.6055555555555556,
+      "grad_norm": 1.3501851646630176,
+      "learning_rate": 1.9994609016737407e-07,
+      "loss": 0.7945,
+      "step": 867
+    },
+    {
+      "epoch": 1.6074074074074074,
+      "grad_norm": 1.524640301680115,
+      "learning_rate": 1.999455967608478e-07,
+      "loss": 0.7771,
+      "step": 868
+    },
+    {
+      "epoch": 1.6092592592592592,
+      "grad_norm": 2.3873074893284825,
+      "learning_rate": 1.9994510110727302e-07,
+      "loss": 0.7603,
+      "step": 869
+    },
+    {
+      "epoch": 1.6111111111111112,
+      "grad_norm": 3.7068813222005312,
+      "learning_rate": 1.9994460320666095e-07,
+      "loss": 0.7635,
+      "step": 870
+    },
+    {
+      "epoch": 1.612962962962963,
+      "grad_norm": 1.4231771848430872,
+      "learning_rate": 1.9994410305902277e-07,
+      "loss": 0.7609,
+      "step": 871
+    },
+    {
+      "epoch": 1.6148148148148147,
+      "grad_norm": 1.531615697390392,
+      "learning_rate": 1.9994360066436967e-07,
+      "loss": 0.819,
+      "step": 872
+    },
+    {
+      "epoch": 1.6166666666666667,
+      "grad_norm": 1.317558498348599,
+      "learning_rate": 1.9994309602271299e-07,
+      "loss": 0.7676,
+      "step": 873
+    },
+    {
+      "epoch": 1.6185185185185185,
+      "grad_norm": 2.2429093846450203,
+      "learning_rate": 1.999425891340641e-07,
+      "loss": 0.78,
+      "step": 874
+    },
+    {
+      "epoch": 1.6203703703703702,
+      "grad_norm": 1.3729813366205148,
+      "learning_rate": 1.9994207999843434e-07,
+      "loss": 0.82,
+      "step": 875
+    },
+    {
+      "epoch": 1.6222222222222222,
+      "grad_norm": 1.340530718512813,
+      "learning_rate": 1.999415686158352e-07,
+      "loss": 0.7205,
+      "step": 876
+    },
+    {
+      "epoch": 1.6240740740740742,
+      "grad_norm": 1.6237245193881744,
+      "learning_rate": 1.9994105498627816e-07,
+      "loss": 0.8091,
+      "step": 877
+    },
+    {
+      "epoch": 1.6259259259259258,
+      "grad_norm": 1.4879353413850342,
+      "learning_rate": 1.9994053910977475e-07,
+      "loss": 0.8508,
+      "step": 878
+    },
+    {
+      "epoch": 1.6277777777777778,
+      "grad_norm": 1.8625800229898672,
+      "learning_rate": 1.999400209863366e-07,
+      "loss": 0.7712,
+      "step": 879
+    },
+    {
+      "epoch": 1.6296296296296298,
+      "grad_norm": 1.3024264983148957,
+      "learning_rate": 1.9993950061597534e-07,
+      "loss": 0.8029,
+      "step": 880
+    },
+    {
+      "epoch": 1.6314814814814815,
+      "grad_norm": 1.3147706316257266,
+      "learning_rate": 1.9993897799870266e-07,
+      "loss": 0.7875,
+      "step": 881
+    },
+    {
+      "epoch": 1.6333333333333333,
+      "grad_norm": 1.3194809596557366,
+      "learning_rate": 1.9993845313453038e-07,
+      "loss": 0.7883,
+      "step": 882
+    },
+    {
+      "epoch": 1.6351851851851853,
+      "grad_norm": 1.4183417869141677,
+      "learning_rate": 1.999379260234702e-07,
+      "loss": 0.7552,
+      "step": 883
+    },
+    {
+      "epoch": 1.637037037037037,
+      "grad_norm": 1.3947153543970918,
+      "learning_rate": 1.9993739666553402e-07,
+      "loss": 0.7563,
+      "step": 884
+    },
+    {
+      "epoch": 1.6388888888888888,
+      "grad_norm": 1.3268726506713047,
+      "learning_rate": 1.9993686506073373e-07,
+      "loss": 0.7818,
+      "step": 885
+    },
+    {
+      "epoch": 1.6407407407407408,
+      "grad_norm": 1.4612349845390744,
+      "learning_rate": 1.9993633120908128e-07,
+      "loss": 0.7743,
+      "step": 886
+    },
+    {
+      "epoch": 1.6425925925925926,
+      "grad_norm": 1.4598335751397802,
+      "learning_rate": 1.9993579511058872e-07,
+      "loss": 0.7799,
+      "step": 887
+    },
+    {
+      "epoch": 1.6444444444444444,
+      "grad_norm": 1.5354530703186837,
+      "learning_rate": 1.9993525676526804e-07,
+      "loss": 0.8388,
+      "step": 888
+    },
+    {
+      "epoch": 1.6462962962962964,
+      "grad_norm": 1.4793914337304956,
+      "learning_rate": 1.999347161731314e-07,
+      "loss": 0.776,
+      "step": 889
+    },
+    {
+      "epoch": 1.6481481481481481,
+      "grad_norm": 1.6079250075291351,
+      "learning_rate": 1.9993417333419088e-07,
+      "loss": 0.7595,
+      "step": 890
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.3685160632787292,
+      "learning_rate": 1.9993362824845874e-07,
+      "loss": 0.8028,
+      "step": 891
+    },
+    {
+      "epoch": 1.651851851851852,
+      "grad_norm": 1.4776133538145957,
+      "learning_rate": 1.9993308091594722e-07,
+      "loss": 0.7373,
+      "step": 892
+    },
+    {
+      "epoch": 1.6537037037037037,
+      "grad_norm": 1.3876764163078121,
+      "learning_rate": 1.9993253133666864e-07,
+      "loss": 0.7876,
+      "step": 893
+    },
+    {
+      "epoch": 1.6555555555555554,
+      "grad_norm": 1.5186057063342262,
+      "learning_rate": 1.9993197951063534e-07,
+      "loss": 0.7707,
+      "step": 894
+    },
+    {
+      "epoch": 1.6574074074074074,
+      "grad_norm": 1.6561840526852425,
+      "learning_rate": 1.999314254378597e-07,
+      "loss": 0.7678,
+      "step": 895
+    },
+    {
+      "epoch": 1.6592592592592592,
+      "grad_norm": 3.340669311151622,
+      "learning_rate": 1.9993086911835424e-07,
+      "loss": 0.7829,
+      "step": 896
+    },
+    {
+      "epoch": 1.661111111111111,
+      "grad_norm": 1.5364577004626374,
+      "learning_rate": 1.9993031055213142e-07,
+      "loss": 0.767,
+      "step": 897
+    },
+    {
+      "epoch": 1.662962962962963,
+      "grad_norm": 1.494286308386131,
+      "learning_rate": 1.999297497392038e-07,
+      "loss": 0.7709,
+      "step": 898
+    },
+    {
+      "epoch": 1.664814814814815,
+      "grad_norm": 1.7165139560096818,
+      "learning_rate": 1.9992918667958404e-07,
+      "loss": 0.75,
+      "step": 899
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 1.5151312622380881,
+      "learning_rate": 1.999286213732847e-07,
+      "loss": 0.7655,
+      "step": 900
+    },
+    {
+      "epoch": 1.6685185185185185,
+      "grad_norm": 1.4393055887815804,
+      "learning_rate": 1.999280538203186e-07,
+      "loss": 0.7877,
+      "step": 901
+    },
+    {
+      "epoch": 1.6703703703703705,
+      "grad_norm": 1.3909681715905493,
+      "learning_rate": 1.9992748402069846e-07,
+      "loss": 0.7931,
+      "step": 902
+    },
+    {
+      "epoch": 1.6722222222222223,
+      "grad_norm": 1.3367514800859155,
+      "learning_rate": 1.9992691197443703e-07,
+      "loss": 0.7675,
+      "step": 903
+    },
+    {
+      "epoch": 1.674074074074074,
+      "grad_norm": 1.5660279955796936,
+      "learning_rate": 1.9992633768154726e-07,
+      "loss": 0.7281,
+      "step": 904
+    },
+    {
+      "epoch": 1.675925925925926,
+      "grad_norm": 1.6593436189179602,
+      "learning_rate": 1.99925761142042e-07,
+      "loss": 0.803,
+      "step": 905
+    },
+    {
+      "epoch": 1.6777777777777778,
+      "grad_norm": 1.4859403459535097,
+      "learning_rate": 1.9992518235593424e-07,
+      "loss": 0.7751,
+      "step": 906
+    },
+    {
+      "epoch": 1.6796296296296296,
+      "grad_norm": 1.4123037224546764,
+      "learning_rate": 1.99924601323237e-07,
+      "loss": 0.7416,
+      "step": 907
+    },
+    {
+      "epoch": 1.6814814814814816,
+      "grad_norm": 1.2990577938387011,
+      "learning_rate": 1.9992401804396334e-07,
+      "loss": 0.7485,
+      "step": 908
+    },
+    {
+      "epoch": 1.6833333333333333,
+      "grad_norm": 1.3759901083208825,
+      "learning_rate": 1.9992343251812634e-07,
+      "loss": 0.7779,
+      "step": 909
+    },
+    {
+      "epoch": 1.6851851851851851,
+      "grad_norm": 1.6708555846734454,
+      "learning_rate": 1.999228447457392e-07,
+      "loss": 0.6831,
+      "step": 910
+    },
+    {
+      "epoch": 1.6870370370370371,
+      "grad_norm": 1.3920984218182832,
+      "learning_rate": 1.9992225472681513e-07,
+      "loss": 0.788,
+      "step": 911
+    },
+    {
+      "epoch": 1.6888888888888889,
+      "grad_norm": 1.3539725437171173,
+      "learning_rate": 1.9992166246136736e-07,
+      "loss": 0.745,
+      "step": 912
+    },
+    {
+      "epoch": 1.6907407407407407,
+      "grad_norm": 1.3179970990736476,
+      "learning_rate": 1.9992106794940925e-07,
+      "loss": 0.7424,
+      "step": 913
+    },
+    {
+      "epoch": 1.6925925925925926,
+      "grad_norm": 1.6647135105039546,
+      "learning_rate": 1.9992047119095418e-07,
+      "loss": 0.7891,
+      "step": 914
+    },
+    {
+      "epoch": 1.6944444444444444,
+      "grad_norm": 1.390235185114431,
+      "learning_rate": 1.999198721860155e-07,
+      "loss": 0.7372,
+      "step": 915
+    },
+    {
+      "epoch": 1.6962962962962962,
+      "grad_norm": 1.250125589913379,
+      "learning_rate": 1.9991927093460674e-07,
+      "loss": 0.7532,
+      "step": 916
+    },
+    {
+      "epoch": 1.6981481481481482,
+      "grad_norm": 1.432752558537211,
+      "learning_rate": 1.9991866743674137e-07,
+      "loss": 0.7855,
+      "step": 917
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.414140842812916,
+      "learning_rate": 1.9991806169243298e-07,
+      "loss": 0.7831,
+      "step": 918
+    },
+    {
+      "epoch": 1.7018518518518517,
+      "grad_norm": 2.3584733494694605,
+      "learning_rate": 1.9991745370169523e-07,
+      "loss": 0.7428,
+      "step": 919
+    },
+    {
+      "epoch": 1.7037037037037037,
+      "grad_norm": 1.5400767488738312,
+      "learning_rate": 1.9991684346454168e-07,
+      "loss": 0.8048,
+      "step": 920
+    },
+    {
+      "epoch": 1.7055555555555557,
+      "grad_norm": 1.5216429489119565,
+      "learning_rate": 1.9991623098098617e-07,
+      "loss": 0.7573,
+      "step": 921
+    },
+    {
+      "epoch": 1.7074074074074073,
+      "grad_norm": 1.3402999091218104,
+      "learning_rate": 1.9991561625104242e-07,
+      "loss": 0.8006,
+      "step": 922
+    },
+    {
+      "epoch": 1.7092592592592593,
+      "grad_norm": 1.6367634390472985,
+      "learning_rate": 1.9991499927472425e-07,
+      "loss": 0.7638,
+      "step": 923
+    },
+    {
+      "epoch": 1.7111111111111112,
+      "grad_norm": 1.3246648877724783,
+      "learning_rate": 1.9991438005204552e-07,
+      "loss": 0.749,
+      "step": 924
+    },
+    {
+      "epoch": 1.7129629629629628,
+      "grad_norm": 1.637132614117255,
+      "learning_rate": 1.9991375858302018e-07,
+      "loss": 0.7832,
+      "step": 925
+    },
+    {
+      "epoch": 1.7148148148148148,
+      "grad_norm": 1.329745756943769,
+      "learning_rate": 1.9991313486766215e-07,
+      "loss": 0.7881,
+      "step": 926
+    },
+    {
+      "epoch": 1.7166666666666668,
+      "grad_norm": 1.39331932578327,
+      "learning_rate": 1.9991250890598553e-07,
+      "loss": 0.7573,
+      "step": 927
+    },
+    {
+      "epoch": 1.7185185185185186,
+      "grad_norm": 1.3887427303224111,
+      "learning_rate": 1.9991188069800431e-07,
+      "loss": 0.8576,
+      "step": 928
+    },
+    {
+      "epoch": 1.7203703703703703,
+      "grad_norm": 1.30847382261147,
+      "learning_rate": 1.9991125024373267e-07,
+      "loss": 0.7765,
+      "step": 929
+    },
+    {
+      "epoch": 1.7222222222222223,
+      "grad_norm": 1.459885257803674,
+      "learning_rate": 1.999106175431848e-07,
+      "loss": 0.7201,
+      "step": 930
+    },
+    {
+      "epoch": 1.724074074074074,
+      "grad_norm": 1.2439278216824947,
+      "learning_rate": 1.9990998259637486e-07,
+      "loss": 0.7867,
+      "step": 931
+    },
+    {
+      "epoch": 1.7259259259259259,
+      "grad_norm": 1.3240204787554097,
+      "learning_rate": 1.9990934540331717e-07,
+      "loss": 0.7553,
+      "step": 932
+    },
+    {
+      "epoch": 1.7277777777777779,
+      "grad_norm": 1.6480773658958212,
+      "learning_rate": 1.9990870596402602e-07,
+      "loss": 0.7713,
+      "step": 933
+    },
+    {
+      "epoch": 1.7296296296296296,
+      "grad_norm": 1.4536599548496305,
+      "learning_rate": 1.9990806427851585e-07,
+      "loss": 0.7525,
+      "step": 934
+    },
+    {
+      "epoch": 1.7314814814814814,
+      "grad_norm": 1.2636993216040457,
+      "learning_rate": 1.9990742034680105e-07,
+      "loss": 0.7284,
+      "step": 935
+    },
+    {
+      "epoch": 1.7333333333333334,
+      "grad_norm": 2.5749205613917225,
+      "learning_rate": 1.9990677416889606e-07,
+      "loss": 0.8042,
+      "step": 936
+    },
+    {
+      "epoch": 1.7351851851851852,
+      "grad_norm": 1.49376272212408,
+      "learning_rate": 1.9990612574481547e-07,
+      "loss": 0.7817,
+      "step": 937
+    },
+    {
+      "epoch": 1.737037037037037,
+      "grad_norm": 1.7261974542179666,
+      "learning_rate": 1.9990547507457383e-07,
+      "loss": 0.7779,
+      "step": 938
+    },
+    {
+      "epoch": 1.738888888888889,
+      "grad_norm": 1.4497238464789381,
+      "learning_rate": 1.999048221581858e-07,
+      "loss": 0.7577,
+      "step": 939
+    },
+    {
+      "epoch": 1.7407407407407407,
+      "grad_norm": 1.509865668190673,
+      "learning_rate": 1.9990416699566596e-07,
+      "loss": 0.7903,
+      "step": 940
+    },
+    {
+      "epoch": 1.7425925925925925,
+      "grad_norm": 1.3596773090201275,
+      "learning_rate": 1.9990350958702918e-07,
+      "loss": 0.7765,
+      "step": 941
+    },
+    {
+      "epoch": 1.7444444444444445,
+      "grad_norm": 1.423302817350013,
+      "learning_rate": 1.9990284993229012e-07,
+      "loss": 0.8126,
+      "step": 942
+    },
+    {
+      "epoch": 1.7462962962962965,
+      "grad_norm": 1.239754724121764,
+      "learning_rate": 1.9990218803146368e-07,
+      "loss": 0.7453,
+      "step": 943
+    },
+    {
+      "epoch": 1.748148148148148,
+      "grad_norm": 1.3817845223584746,
+      "learning_rate": 1.9990152388456471e-07,
+      "loss": 0.7379,
+      "step": 944
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.2875603320167348,
+      "learning_rate": 1.9990085749160818e-07,
+      "loss": 0.8182,
+      "step": 945
+    },
+    {
+      "epoch": 1.751851851851852,
+      "grad_norm": 2.0181658079683737,
+      "learning_rate": 1.9990018885260905e-07,
+      "loss": 0.8163,
+      "step": 946
+    },
+    {
+      "epoch": 1.7537037037037035,
+      "grad_norm": 1.4952035292359285,
+      "learning_rate": 1.9989951796758234e-07,
+      "loss": 0.7634,
+      "step": 947
+    },
+    {
+      "epoch": 1.7555555555555555,
+      "grad_norm": 1.340738450776537,
+      "learning_rate": 1.998988448365431e-07,
+      "loss": 0.7872,
+      "step": 948
+    },
+    {
+      "epoch": 1.7574074074074075,
+      "grad_norm": 1.399537935293725,
+      "learning_rate": 1.9989816945950657e-07,
+      "loss": 0.7716,
+      "step": 949
+    },
+    {
+      "epoch": 1.7592592592592593,
+      "grad_norm": 1.535868346196315,
+      "learning_rate": 1.9989749183648785e-07,
+      "loss": 0.8115,
+      "step": 950
+    },
+    {
+      "epoch": 1.761111111111111,
+      "grad_norm": 1.433963445421848,
+      "learning_rate": 1.998968119675022e-07,
+      "loss": 0.8099,
+      "step": 951
+    },
+    {
+      "epoch": 1.762962962962963,
+      "grad_norm": 1.3873999924942957,
+      "learning_rate": 1.998961298525649e-07,
+      "loss": 0.7789,
+      "step": 952
+    },
+    {
+      "epoch": 1.7648148148148148,
+      "grad_norm": 1.3212520883435643,
+      "learning_rate": 1.9989544549169127e-07,
+      "loss": 0.7405,
+      "step": 953
+    },
+    {
+      "epoch": 1.7666666666666666,
+      "grad_norm": 1.3038838284160543,
+      "learning_rate": 1.9989475888489672e-07,
+      "loss": 0.7388,
+      "step": 954
+    },
+    {
+      "epoch": 1.7685185185185186,
+      "grad_norm": 1.4028278928487332,
+      "learning_rate": 1.998940700321967e-07,
+      "loss": 0.7484,
+      "step": 955
+    },
+    {
+      "epoch": 1.7703703703703704,
+      "grad_norm": 1.437825392929185,
+      "learning_rate": 1.9989337893360667e-07,
+      "loss": 0.7538,
+      "step": 956
+    },
+    {
+      "epoch": 1.7722222222222221,
+      "grad_norm": 1.3295417487043837,
+      "learning_rate": 1.9989268558914217e-07,
+      "loss": 0.7452,
+      "step": 957
+    },
+    {
+      "epoch": 1.7740740740740741,
+      "grad_norm": 1.6626271303687088,
+      "learning_rate": 1.9989198999881878e-07,
+      "loss": 0.8013,
+      "step": 958
+    },
+    {
+      "epoch": 1.775925925925926,
+      "grad_norm": 3.698332225453685,
+      "learning_rate": 1.9989129216265214e-07,
+      "loss": 0.7616,
+      "step": 959
+    },
+    {
+      "epoch": 1.7777777777777777,
+      "grad_norm": 1.3816989748673836,
+      "learning_rate": 1.99890592080658e-07,
+      "loss": 0.722,
+      "step": 960
+    },
+    {
+      "epoch": 1.7796296296296297,
+      "grad_norm": 1.3046144809430562,
+      "learning_rate": 1.9988988975285204e-07,
+      "loss": 0.7361,
+      "step": 961
+    },
+    {
+      "epoch": 1.7814814814814814,
+      "grad_norm": 1.4518523974344757,
+      "learning_rate": 1.9988918517925005e-07,
+      "loss": 0.8193,
+      "step": 962
+    },
+    {
+      "epoch": 1.7833333333333332,
+      "grad_norm": 2.3750512495840965,
+      "learning_rate": 1.9988847835986786e-07,
+      "loss": 0.7587,
+      "step": 963
+    },
+    {
+      "epoch": 1.7851851851851852,
+      "grad_norm": 2.192672507249222,
+      "learning_rate": 1.998877692947214e-07,
+      "loss": 0.7587,
+      "step": 964
+    },
+    {
+      "epoch": 1.7870370370370372,
+      "grad_norm": 1.5302685333436545,
+      "learning_rate": 1.9988705798382661e-07,
+      "loss": 0.7876,
+      "step": 965
+    },
+    {
+      "epoch": 1.7888888888888888,
+      "grad_norm": 1.418036599899706,
+      "learning_rate": 1.9988634442719948e-07,
+      "loss": 0.7684,
+      "step": 966
+    },
+    {
+      "epoch": 1.7907407407407407,
+      "grad_norm": 2.419996407593016,
+      "learning_rate": 1.99885628624856e-07,
+      "loss": 0.6958,
+      "step": 967
+    },
+    {
+      "epoch": 1.7925925925925927,
+      "grad_norm": 1.4944578188848778,
+      "learning_rate": 1.9988491057681232e-07,
+      "loss": 0.7705,
+      "step": 968
+    },
+    {
+      "epoch": 1.7944444444444443,
+      "grad_norm": 3.0693779890514,
+      "learning_rate": 1.9988419028308457e-07,
+      "loss": 0.739,
+      "step": 969
+    },
+    {
+      "epoch": 1.7962962962962963,
+      "grad_norm": 1.4341933458154759,
+      "learning_rate": 1.9988346774368894e-07,
+      "loss": 0.7418,
+      "step": 970
+    },
+    {
+      "epoch": 1.7981481481481483,
+      "grad_norm": 1.5736459192659946,
+      "learning_rate": 1.9988274295864167e-07,
+      "loss": 0.7601,
+      "step": 971
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.4929965237519107,
+      "learning_rate": 1.9988201592795905e-07,
+      "loss": 0.803,
+      "step": 972
+    },
+    {
+      "epoch": 1.8018518518518518,
+      "grad_norm": 2.133063903152309,
+      "learning_rate": 1.9988128665165746e-07,
+      "loss": 0.7051,
+      "step": 973
+    },
+    {
+      "epoch": 1.8037037037037038,
+      "grad_norm": 1.4119194601919294,
+      "learning_rate": 1.9988055512975329e-07,
+      "loss": 0.7737,
+      "step": 974
+    },
+    {
+      "epoch": 1.8055555555555556,
+      "grad_norm": 1.3911171054460025,
+      "learning_rate": 1.9987982136226292e-07,
+      "loss": 0.8068,
+      "step": 975
+    },
+    {
+      "epoch": 1.8074074074074074,
+      "grad_norm": 1.9381069385259089,
+      "learning_rate": 1.9987908534920291e-07,
+      "loss": 0.7539,
+      "step": 976
+    },
+    {
+      "epoch": 1.8092592592592593,
+      "grad_norm": 1.6411214166656005,
+      "learning_rate": 1.9987834709058983e-07,
+      "loss": 0.7811,
+      "step": 977
+    },
+    {
+      "epoch": 1.8111111111111111,
+      "grad_norm": 1.507315883841706,
+      "learning_rate": 1.9987760658644023e-07,
+      "loss": 0.7836,
+      "step": 978
+    },
+    {
+      "epoch": 1.8129629629629629,
+      "grad_norm": 1.869801800461327,
+      "learning_rate": 1.998768638367707e-07,
+      "loss": 0.7749,
+      "step": 979
+    },
+    {
+      "epoch": 1.8148148148148149,
+      "grad_norm": 1.7565505719473904,
+      "learning_rate": 1.998761188415981e-07,
+      "loss": 0.7967,
+      "step": 980
+    },
+    {
+      "epoch": 1.8166666666666667,
+      "grad_norm": 1.6345579550760483,
+      "learning_rate": 1.9987537160093907e-07,
+      "loss": 0.792,
+      "step": 981
+    },
+    {
+      "epoch": 1.8185185185185184,
+      "grad_norm": 1.9607225127855206,
+      "learning_rate": 1.9987462211481038e-07,
+      "loss": 0.7928,
+      "step": 982
+    },
+    {
+      "epoch": 1.8203703703703704,
+      "grad_norm": 1.5188361447171843,
+      "learning_rate": 1.9987387038322898e-07,
+      "loss": 0.8108,
+      "step": 983
+    },
+    {
+      "epoch": 1.8222222222222222,
+      "grad_norm": 1.5413453361100633,
+      "learning_rate": 1.998731164062117e-07,
+      "loss": 0.7977,
+      "step": 984
+    },
+    {
+      "epoch": 1.824074074074074,
+      "grad_norm": 1.3757713481914136,
+      "learning_rate": 1.998723601837755e-07,
+      "loss": 0.749,
+      "step": 985
+    },
+    {
+      "epoch": 1.825925925925926,
+      "grad_norm": 1.4354818948523813,
+      "learning_rate": 1.9987160171593743e-07,
+      "loss": 0.8036,
+      "step": 986
+    },
+    {
+      "epoch": 1.8277777777777777,
+      "grad_norm": 1.597649565887714,
+      "learning_rate": 1.9987084100271448e-07,
+      "loss": 0.7323,
+      "step": 987
+    },
+    {
+      "epoch": 1.8296296296296295,
+      "grad_norm": 1.1936510229152042,
+      "learning_rate": 1.998700780441238e-07,
+      "loss": 0.8061,
+      "step": 988
+    },
+    {
+      "epoch": 1.8314814814814815,
+      "grad_norm": 1.7299553561365208,
+      "learning_rate": 1.9986931284018248e-07,
+      "loss": 0.7887,
+      "step": 989
+    },
+    {
+      "epoch": 1.8333333333333335,
+      "grad_norm": 1.64664118210252,
+      "learning_rate": 1.9986854539090781e-07,
+      "loss": 0.7784,
+      "step": 990
+    },
+    {
+      "epoch": 1.835185185185185,
+      "grad_norm": 1.9177963561964584,
+      "learning_rate": 1.99867775696317e-07,
+      "loss": 0.774,
+      "step": 991
+    },
+    {
+      "epoch": 1.837037037037037,
+      "grad_norm": 1.3093476778746875,
+      "learning_rate": 1.998670037564273e-07,
+      "loss": 0.7363,
+      "step": 992
+    },
+    {
+      "epoch": 1.838888888888889,
+      "grad_norm": 1.319367963893485,
+      "learning_rate": 1.9986622957125618e-07,
+      "loss": 0.7447,
+      "step": 993
+    },
+    {
+      "epoch": 1.8407407407407408,
+      "grad_norm": 2.676685381876739,
+      "learning_rate": 1.9986545314082097e-07,
+      "loss": 0.7519,
+      "step": 994
+    },
+    {
+      "epoch": 1.8425925925925926,
+      "grad_norm": 1.5326917801294218,
+      "learning_rate": 1.9986467446513915e-07,
+      "loss": 0.7681,
+      "step": 995
+    },
+    {
+      "epoch": 1.8444444444444446,
+      "grad_norm": 1.8419702936033213,
+      "learning_rate": 1.998638935442282e-07,
+      "loss": 0.7142,
+      "step": 996
+    },
+    {
+      "epoch": 1.8462962962962963,
+      "grad_norm": 1.3323039771053324,
+      "learning_rate": 1.998631103781057e-07,
+      "loss": 0.7286,
+      "step": 997
+    },
+    {
+      "epoch": 1.848148148148148,
+      "grad_norm": 1.3394903259952737,
+      "learning_rate": 1.9986232496678928e-07,
+      "loss": 0.753,
+      "step": 998
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.7055420272114223,
+      "learning_rate": 1.9986153731029653e-07,
+      "loss": 0.7844,
+      "step": 999
+    },
+    {
+      "epoch": 1.8518518518518519,
+      "grad_norm": 2.34274316574407,
+      "learning_rate": 1.9986074740864525e-07,
+      "loss": 0.7619,
+      "step": 1000
+    },
+    {
+      "epoch": 1.8537037037037036,
+      "grad_norm": 1.2647293607810843,
+      "learning_rate": 1.9985995526185313e-07,
+      "loss": 0.7645,
+      "step": 1001
+    },
+    {
+      "epoch": 1.8555555555555556,
+      "grad_norm": 1.325895710474424,
+      "learning_rate": 1.99859160869938e-07,
+      "loss": 0.728,
+      "step": 1002
+    },
+    {
+      "epoch": 1.8574074074074074,
+      "grad_norm": 1.6743053959927081,
+      "learning_rate": 1.9985836423291771e-07,
+      "loss": 0.8377,
+      "step": 1003
+    },
+    {
+      "epoch": 1.8592592592592592,
+      "grad_norm": 1.914293654241324,
+      "learning_rate": 1.998575653508102e-07,
+      "loss": 0.7333,
+      "step": 1004
+    },
+    {
+      "epoch": 1.8611111111111112,
+      "grad_norm": 4.547147360648242,
+      "learning_rate": 1.9985676422363338e-07,
+      "loss": 0.7765,
+      "step": 1005
+    },
+    {
+      "epoch": 1.862962962962963,
+      "grad_norm": 1.3432555614109056,
+      "learning_rate": 1.9985596085140533e-07,
+      "loss": 0.7946,
+      "step": 1006
+    },
+    {
+      "epoch": 1.8648148148148147,
+      "grad_norm": 3.539767209438379,
+      "learning_rate": 1.9985515523414404e-07,
+      "loss": 0.7472,
+      "step": 1007
+    },
+    {
+      "epoch": 1.8666666666666667,
+      "grad_norm": 1.4946296123352325,
+      "learning_rate": 1.9985434737186769e-07,
+      "loss": 0.7934,
+      "step": 1008
+    },
+    {
+      "epoch": 1.8685185185185185,
+      "grad_norm": 1.4855968713815713,
+      "learning_rate": 1.998535372645944e-07,
+      "loss": 0.7589,
+      "step": 1009
+    },
+    {
+      "epoch": 1.8703703703703702,
+      "grad_norm": 1.2654104905041015,
+      "learning_rate": 1.9985272491234238e-07,
+      "loss": 0.7541,
+      "step": 1010
+    },
+    {
+      "epoch": 1.8722222222222222,
+      "grad_norm": 1.3082185718578754,
+      "learning_rate": 1.9985191031512993e-07,
+      "loss": 0.8079,
+      "step": 1011
+    },
+    {
+      "epoch": 1.8740740740740742,
+      "grad_norm": 1.6232591014864821,
+      "learning_rate": 1.9985109347297534e-07,
+      "loss": 0.8203,
+      "step": 1012
+    },
+    {
+      "epoch": 1.8759259259259258,
+      "grad_norm": 1.5545942041103336,
+      "learning_rate": 1.9985027438589696e-07,
+      "loss": 0.768,
+      "step": 1013
+    },
+    {
+      "epoch": 1.8777777777777778,
+      "grad_norm": 1.6783166712367008,
+      "learning_rate": 1.9984945305391323e-07,
+      "loss": 0.7649,
+      "step": 1014
+    },
+    {
+      "epoch": 1.8796296296296298,
+      "grad_norm": 1.4934396970379067,
+      "learning_rate": 1.998486294770426e-07,
+      "loss": 0.7254,
+      "step": 1015
+    },
+    {
+      "epoch": 1.8814814814814815,
+      "grad_norm": 1.8125299677321407,
+      "learning_rate": 1.998478036553036e-07,
+      "loss": 0.8056,
+      "step": 1016
+    },
+    {
+      "epoch": 1.8833333333333333,
+      "grad_norm": 1.785947678012694,
+      "learning_rate": 1.998469755887148e-07,
+      "loss": 0.756,
+      "step": 1017
+    },
+    {
+      "epoch": 1.8851851851851853,
+      "grad_norm": 1.5501905077907572,
+      "learning_rate": 1.9984614527729482e-07,
+      "loss": 0.7913,
+      "step": 1018
+    },
+    {
+      "epoch": 1.887037037037037,
+      "grad_norm": 1.4673739574958995,
+      "learning_rate": 1.9984531272106228e-07,
+      "loss": 0.786,
+      "step": 1019
+    },
+    {
+      "epoch": 1.8888888888888888,
+      "grad_norm": 1.5149206975977105,
+      "learning_rate": 1.9984447792003597e-07,
+      "loss": 0.7519,
+      "step": 1020
+    },
+    {
+      "epoch": 1.8907407407407408,
+      "grad_norm": 1.5332146548019172,
+      "learning_rate": 1.9984364087423463e-07,
+      "loss": 0.8084,
+      "step": 1021
+    },
+    {
+      "epoch": 1.8925925925925926,
+      "grad_norm": 1.6798081001422076,
+      "learning_rate": 1.9984280158367703e-07,
+      "loss": 0.7245,
+      "step": 1022
+    },
+    {
+      "epoch": 1.8944444444444444,
+      "grad_norm": 1.3687354583030944,
+      "learning_rate": 1.998419600483821e-07,
+      "loss": 0.8057,
+      "step": 1023
+    },
+    {
+      "epoch": 1.8962962962962964,
+      "grad_norm": 1.4728179311724792,
+      "learning_rate": 1.9984111626836875e-07,
+      "loss": 0.7358,
+      "step": 1024
+    },
+    {
+      "epoch": 1.8981481481481481,
+      "grad_norm": 1.4682529755028493,
+      "learning_rate": 1.9984027024365596e-07,
+      "loss": 0.7193,
+      "step": 1025
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.353931026811784,
+      "learning_rate": 1.9983942197426268e-07,
+      "loss": 0.782,
+      "step": 1026
+    },
+    {
+      "epoch": 1.901851851851852,
+      "grad_norm": 1.323413112380447,
+      "learning_rate": 1.998385714602081e-07,
+      "loss": 0.7608,
+      "step": 1027
+    },
+    {
+      "epoch": 1.9037037037037037,
+      "grad_norm": 1.349649237885736,
+      "learning_rate": 1.998377187015112e-07,
+      "loss": 0.7525,
+      "step": 1028
+    },
+    {
+      "epoch": 1.9055555555555554,
+      "grad_norm": 1.5133026470740185,
+      "learning_rate": 1.9983686369819126e-07,
+      "loss": 0.7252,
+      "step": 1029
+    },
+    {
+      "epoch": 1.9074074074074074,
+      "grad_norm": 1.529895027913531,
+      "learning_rate": 1.998360064502675e-07,
+      "loss": 0.7322,
+      "step": 1030
+    },
+    {
+      "epoch": 1.9092592592592592,
+      "grad_norm": 1.3683455309371708,
+      "learning_rate": 1.9983514695775915e-07,
+      "loss": 0.7654,
+      "step": 1031
+    },
+    {
+      "epoch": 1.911111111111111,
+      "grad_norm": 1.386781752640885,
+      "learning_rate": 1.9983428522068553e-07,
+      "loss": 0.7802,
+      "step": 1032
+    },
+    {
+      "epoch": 1.912962962962963,
+      "grad_norm": 7.267501283575071,
+      "learning_rate": 1.9983342123906603e-07,
+      "loss": 0.7591,
+      "step": 1033
+    },
+    {
+      "epoch": 1.914814814814815,
+      "grad_norm": 1.5374512127495408,
+      "learning_rate": 1.9983255501292012e-07,
+      "loss": 0.7546,
+      "step": 1034
+    },
+    {
+      "epoch": 1.9166666666666665,
+      "grad_norm": 1.3894365469249104,
+      "learning_rate": 1.998316865422672e-07,
+      "loss": 0.7425,
+      "step": 1035
+    },
+    {
+      "epoch": 1.9185185185185185,
+      "grad_norm": 1.3874291485993693,
+      "learning_rate": 1.9983081582712683e-07,
+      "loss": 0.7832,
+      "step": 1036
+    },
+    {
+      "epoch": 1.9203703703703705,
+      "grad_norm": 1.3064464904357003,
+      "learning_rate": 1.9982994286751858e-07,
+      "loss": 0.7978,
+      "step": 1037
+    },
+    {
+      "epoch": 1.9222222222222223,
+      "grad_norm": 1.582241919580081,
+      "learning_rate": 1.9982906766346205e-07,
+      "loss": 0.7479,
+      "step": 1038
+    },
+    {
+      "epoch": 1.924074074074074,
+      "grad_norm": 1.3279186913188272,
+      "learning_rate": 1.99828190214977e-07,
+      "loss": 0.7035,
+      "step": 1039
+    },
+    {
+      "epoch": 1.925925925925926,
+      "grad_norm": 1.3699480195385578,
+      "learning_rate": 1.9982731052208307e-07,
+      "loss": 0.7945,
+      "step": 1040
+    },
+    {
+      "epoch": 1.9277777777777778,
+      "grad_norm": 1.7259479474815946,
+      "learning_rate": 1.998264285848001e-07,
+      "loss": 0.7482,
+      "step": 1041
+    },
+    {
+      "epoch": 1.9296296296296296,
+      "grad_norm": 1.4844346643748427,
+      "learning_rate": 1.998255444031479e-07,
+      "loss": 0.7675,
+      "step": 1042
+    },
+    {
+      "epoch": 1.9314814814814816,
+      "grad_norm": 1.3975117459361475,
+      "learning_rate": 1.9982465797714632e-07,
+      "loss": 0.7678,
+      "step": 1043
+    },
+    {
+      "epoch": 1.9333333333333333,
+      "grad_norm": 1.3202089517147364,
+      "learning_rate": 1.998237693068153e-07,
+      "loss": 0.819,
+      "step": 1044
+    },
+    {
+      "epoch": 1.9351851851851851,
+      "grad_norm": 1.299947303921116,
+      "learning_rate": 1.9982287839217483e-07,
+      "loss": 0.7774,
+      "step": 1045
+    },
+    {
+      "epoch": 1.9370370370370371,
+      "grad_norm": 1.4987729976260935,
+      "learning_rate": 1.9982198523324497e-07,
+      "loss": 0.7579,
+      "step": 1046
+    },
+    {
+      "epoch": 1.9388888888888889,
+      "grad_norm": 1.5560900777954654,
+      "learning_rate": 1.9982108983004573e-07,
+      "loss": 0.7562,
+      "step": 1047
+    },
+    {
+      "epoch": 1.9407407407407407,
+      "grad_norm": 1.3422556176886826,
+      "learning_rate": 1.9982019218259728e-07,
+      "loss": 0.7272,
+      "step": 1048
+    },
+    {
+      "epoch": 1.9425925925925926,
+      "grad_norm": 1.478719884379062,
+      "learning_rate": 1.9981929229091982e-07,
+      "loss": 0.7812,
+      "step": 1049
+    },
+    {
+      "epoch": 1.9444444444444444,
+      "grad_norm": 1.2455833949829016,
+      "learning_rate": 1.9981839015503356e-07,
+      "loss": 0.73,
+      "step": 1050
+    },
+    {
+      "epoch": 1.9462962962962962,
+      "grad_norm": 1.6427296811374335,
+      "learning_rate": 1.9981748577495882e-07,
+      "loss": 0.7531,
+      "step": 1051
+    },
+    {
+      "epoch": 1.9481481481481482,
+      "grad_norm": 1.3744901950496324,
+      "learning_rate": 1.9981657915071586e-07,
+      "loss": 0.753,
+      "step": 1052
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.5270359146528198,
+      "learning_rate": 1.9981567028232512e-07,
+      "loss": 0.7574,
+      "step": 1053
+    },
+    {
+      "epoch": 1.9518518518518517,
+      "grad_norm": 1.4389730046470506,
+      "learning_rate": 1.99814759169807e-07,
+      "loss": 0.7533,
+      "step": 1054
+    },
+    {
+      "epoch": 1.9537037037037037,
+      "grad_norm": 1.3096995681968853,
+      "learning_rate": 1.99813845813182e-07,
+      "loss": 0.7648,
+      "step": 1055
+    },
+    {
+      "epoch": 1.9555555555555557,
+      "grad_norm": 1.4176783980098466,
+      "learning_rate": 1.9981293021247067e-07,
+      "loss": 0.7046,
+      "step": 1056
+    },
+    {
+      "epoch": 1.9574074074074073,
+      "grad_norm": 1.2281733478512156,
+      "learning_rate": 1.998120123676936e-07,
+      "loss": 0.7684,
+      "step": 1057
+    },
+    {
+      "epoch": 1.9592592592592593,
+      "grad_norm": 1.2189469708483014,
+      "learning_rate": 1.998110922788714e-07,
+      "loss": 0.7526,
+      "step": 1058
+    },
+    {
+      "epoch": 1.9611111111111112,
+      "grad_norm": 1.4521186933405126,
+      "learning_rate": 1.998101699460247e-07,
+      "loss": 0.7919,
+      "step": 1059
+    },
+    {
+      "epoch": 1.9629629629629628,
+      "grad_norm": 1.4628029193341177,
+      "learning_rate": 1.9980924536917436e-07,
+      "loss": 0.7534,
+      "step": 1060
+    },
+    {
+      "epoch": 1.9648148148148148,
+      "grad_norm": 1.2616788752733763,
+      "learning_rate": 1.998083185483411e-07,
+      "loss": 0.7254,
+      "step": 1061
+    },
+    {
+      "epoch": 1.9666666666666668,
+      "grad_norm": 1.3730086237523185,
+      "learning_rate": 1.9980738948354574e-07,
+      "loss": 0.7416,
+      "step": 1062
+    },
+    {
+      "epoch": 1.9685185185185186,
+      "grad_norm": 1.310619553346444,
+      "learning_rate": 1.998064581748092e-07,
+      "loss": 0.7976,
+      "step": 1063
+    },
+    {
+      "epoch": 1.9703703703703703,
+      "grad_norm": 1.4891024950328993,
+      "learning_rate": 1.9980552462215243e-07,
+      "loss": 0.8111,
+      "step": 1064
+    },
+    {
+      "epoch": 1.9722222222222223,
+      "grad_norm": 2.7226386028805214,
+      "learning_rate": 1.9980458882559642e-07,
+      "loss": 0.7614,
+      "step": 1065
+    },
+    {
+      "epoch": 1.974074074074074,
+      "grad_norm": 1.791562733109985,
+      "learning_rate": 1.9980365078516212e-07,
+      "loss": 0.7618,
+      "step": 1066
+    },
+    {
+      "epoch": 1.9759259259259259,
+      "grad_norm": 1.4722941810114145,
+      "learning_rate": 1.9980271050087074e-07,
+      "loss": 0.7998,
+      "step": 1067
+    },
+    {
+      "epoch": 1.9777777777777779,
+      "grad_norm": 1.421312923217099,
+      "learning_rate": 1.9980176797274334e-07,
+      "loss": 0.7356,
+      "step": 1068
+    },
+    {
+      "epoch": 1.9796296296296296,
+      "grad_norm": 4.404404168548993,
+      "learning_rate": 1.9980082320080115e-07,
+      "loss": 0.7729,
+      "step": 1069
+    },
+    {
+      "epoch": 1.9814814814814814,
+      "grad_norm": 2.807841897993627,
+      "learning_rate": 1.997998761850654e-07,
+      "loss": 0.7031,
+      "step": 1070
+    },
+    {
+      "epoch": 1.9833333333333334,
+      "grad_norm": 1.2146651225346319,
+      "learning_rate": 1.9979892692555737e-07,
+      "loss": 0.7458,
+      "step": 1071
+    },
+    {
+      "epoch": 1.9851851851851852,
+      "grad_norm": 1.6151498113086764,
+      "learning_rate": 1.9979797542229843e-07,
+      "loss": 0.7849,
+      "step": 1072
+    },
+    {
+      "epoch": 1.987037037037037,
+      "grad_norm": 1.9101424059321606,
+      "learning_rate": 1.9979702167530993e-07,
+      "loss": 0.7276,
+      "step": 1073
+    },
+    {
+      "epoch": 1.988888888888889,
+      "grad_norm": 1.8527770325311863,
+      "learning_rate": 1.9979606568461337e-07,
+      "loss": 0.7872,
+      "step": 1074
+    },
+    {
+      "epoch": 1.9907407407407407,
+      "grad_norm": 1.6534336218889853,
+      "learning_rate": 1.997951074502302e-07,
+      "loss": 0.715,
+      "step": 1075
+    },
+    {
+      "epoch": 1.9925925925925925,
+      "grad_norm": 1.5402678844152873,
+      "learning_rate": 1.9979414697218198e-07,
+      "loss": 0.7416,
+      "step": 1076
+    },
+    {
+      "epoch": 1.9944444444444445,
+      "grad_norm": 1.180553748820107,
+      "learning_rate": 1.997931842504903e-07,
+      "loss": 0.7176,
+      "step": 1077
+    },
+    {
+      "epoch": 1.9962962962962965,
+      "grad_norm": 1.308310925597982,
+      "learning_rate": 1.9979221928517677e-07,
+      "loss": 0.8047,
+      "step": 1078
+    },
+    {
+      "epoch": 1.998148148148148,
+      "grad_norm": 1.4086706085230458,
+      "learning_rate": 1.9979125207626315e-07,
+      "loss": 0.7973,
+      "step": 1079
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.439532492057802,
+      "learning_rate": 1.9979028262377116e-07,
+      "loss": 0.7057,
+      "step": 1080
+    },
+    {
+      "epoch": 2.001851851851852,
+      "grad_norm": 3.742312300553071,
+      "learning_rate": 1.9978931092772258e-07,
+      "loss": 0.769,
+      "step": 1081
+    },
+    {
+      "epoch": 2.0037037037037035,
+      "grad_norm": 1.8585457575425635,
+      "learning_rate": 1.9978833698813926e-07,
+      "loss": 0.7508,
+      "step": 1082
+    },
+    {
+      "epoch": 2.0055555555555555,
+      "grad_norm": 1.5030387887218244,
+      "learning_rate": 1.997873608050431e-07,
+      "loss": 0.7611,
+      "step": 1083
+    },
+    {
+      "epoch": 2.0074074074074075,
+      "grad_norm": 1.300856313413514,
+      "learning_rate": 1.9978638237845606e-07,
+      "loss": 0.7669,
+      "step": 1084
+    },
+    {
+      "epoch": 2.009259259259259,
+      "grad_norm": 1.8991658561495355,
+      "learning_rate": 1.9978540170840016e-07,
+      "loss": 0.7982,
+      "step": 1085
+    },
+    {
+      "epoch": 2.011111111111111,
+      "grad_norm": 1.3891993955211341,
+      "learning_rate": 1.9978441879489737e-07,
+      "loss": 0.789,
+      "step": 1086
+    },
+    {
+      "epoch": 2.012962962962963,
+      "grad_norm": 1.2713680021678209,
+      "learning_rate": 1.9978343363796983e-07,
+      "loss": 0.7645,
+      "step": 1087
+    },
+    {
+      "epoch": 2.0148148148148146,
+      "grad_norm": 1.6233373509493494,
+      "learning_rate": 1.9978244623763974e-07,
+      "loss": 0.7466,
+      "step": 1088
+    },
+    {
+      "epoch": 2.0166666666666666,
+      "grad_norm": 1.2854174566083847,
+      "learning_rate": 1.9978145659392923e-07,
+      "loss": 0.7605,
+      "step": 1089
+    },
+    {
+      "epoch": 2.0185185185185186,
+      "grad_norm": 1.5034516341365665,
+      "learning_rate": 1.9978046470686056e-07,
+      "loss": 0.7603,
+      "step": 1090
+    },
+    {
+      "epoch": 2.0203703703703706,
+      "grad_norm": 1.3607129822819393,
+      "learning_rate": 1.9977947057645604e-07,
+      "loss": 0.7587,
+      "step": 1091
+    },
+    {
+      "epoch": 2.022222222222222,
+      "grad_norm": 1.4252312779020693,
+      "learning_rate": 1.9977847420273805e-07,
+      "loss": 0.7339,
+      "step": 1092
+    },
+    {
+      "epoch": 2.024074074074074,
+      "grad_norm": 2.2850255266168475,
+      "learning_rate": 1.9977747558572898e-07,
+      "loss": 0.7495,
+      "step": 1093
+    },
+    {
+      "epoch": 2.025925925925926,
+      "grad_norm": 1.280053501160805,
+      "learning_rate": 1.9977647472545121e-07,
+      "loss": 0.7192,
+      "step": 1094
+    },
+    {
+      "epoch": 2.0277777777777777,
+      "grad_norm": 1.357419608222813,
+      "learning_rate": 1.9977547162192737e-07,
+      "loss": 0.7597,
+      "step": 1095
+    },
+    {
+      "epoch": 2.0296296296296297,
+      "grad_norm": 1.34589559526732,
+      "learning_rate": 1.9977446627517988e-07,
+      "loss": 0.729,
+      "step": 1096
+    },
+    {
+      "epoch": 2.0314814814814817,
+      "grad_norm": 4.632316368137749,
+      "learning_rate": 1.9977345868523144e-07,
+      "loss": 0.7543,
+      "step": 1097
+    },
+    {
+      "epoch": 2.033333333333333,
+      "grad_norm": 1.3253998226400878,
+      "learning_rate": 1.9977244885210464e-07,
+      "loss": 0.7171,
+      "step": 1098
+    },
+    {
+      "epoch": 2.035185185185185,
+      "grad_norm": 1.3671055208615106,
+      "learning_rate": 1.9977143677582222e-07,
+      "loss": 0.7301,
+      "step": 1099
+    },
+    {
+      "epoch": 2.037037037037037,
+      "grad_norm": 1.7310213780935957,
+      "learning_rate": 1.9977042245640697e-07,
+      "loss": 0.7895,
+      "step": 1100
+    },
+    {
+      "epoch": 2.0388888888888888,
+      "grad_norm": 2.7648618399315814,
+      "learning_rate": 1.997694058938816e-07,
+      "loss": 0.7887,
+      "step": 1101
+    },
+    {
+      "epoch": 2.0407407407407407,
+      "grad_norm": 1.4513820653490532,
+      "learning_rate": 1.9976838708826903e-07,
+      "loss": 0.7843,
+      "step": 1102
+    },
+    {
+      "epoch": 2.0425925925925927,
+      "grad_norm": 1.462192271315431,
+      "learning_rate": 1.9976736603959217e-07,
+      "loss": 0.6909,
+      "step": 1103
+    },
+    {
+      "epoch": 2.0444444444444443,
+      "grad_norm": 1.5338282090107127,
+      "learning_rate": 1.997663427478739e-07,
+      "loss": 0.6944,
+      "step": 1104
+    },
+    {
+      "epoch": 2.0462962962962963,
+      "grad_norm": 1.4166834039190126,
+      "learning_rate": 1.9976531721313737e-07,
+      "loss": 0.7619,
+      "step": 1105
+    },
+    {
+      "epoch": 2.0481481481481483,
+      "grad_norm": 1.3709847244996538,
+      "learning_rate": 1.9976428943540549e-07,
+      "loss": 0.7621,
+      "step": 1106
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 3.404650291829635,
+      "learning_rate": 1.9976325941470145e-07,
+      "loss": 0.7368,
+      "step": 1107
+    },
+    {
+      "epoch": 2.051851851851852,
+      "grad_norm": 2.3702115184846546,
+      "learning_rate": 1.9976222715104836e-07,
+      "loss": 0.7805,
+      "step": 1108
+    },
+    {
+      "epoch": 2.053703703703704,
+      "grad_norm": 2.6013403883405126,
+      "learning_rate": 1.9976119264446946e-07,
+      "loss": 0.7705,
+      "step": 1109
+    },
+    {
+      "epoch": 2.0555555555555554,
+      "grad_norm": 1.4160560317452116,
+      "learning_rate": 1.9976015589498805e-07,
+      "loss": 0.7694,
+      "step": 1110
+    },
+    {
+      "epoch": 2.0574074074074074,
+      "grad_norm": 1.5769278435884768,
+      "learning_rate": 1.9975911690262733e-07,
+      "loss": 0.778,
+      "step": 1111
+    },
+    {
+      "epoch": 2.0592592592592593,
+      "grad_norm": 1.500735852563375,
+      "learning_rate": 1.9975807566741073e-07,
+      "loss": 0.7356,
+      "step": 1112
+    },
+    {
+      "epoch": 2.061111111111111,
+      "grad_norm": 1.5896967539997306,
+      "learning_rate": 1.997570321893617e-07,
+      "loss": 0.765,
+      "step": 1113
+    },
+    {
+      "epoch": 2.062962962962963,
+      "grad_norm": 1.3020079809979694,
+      "learning_rate": 1.997559864685036e-07,
+      "loss": 0.7594,
+      "step": 1114
+    },
+    {
+      "epoch": 2.064814814814815,
+      "grad_norm": 1.4196410980954213,
+      "learning_rate": 1.9975493850486003e-07,
+      "loss": 0.7762,
+      "step": 1115
+    },
+    {
+      "epoch": 2.066666666666667,
+      "grad_norm": 1.5604632360783146,
+      "learning_rate": 1.9975388829845446e-07,
+      "loss": 0.7745,
+      "step": 1116
+    },
+    {
+      "epoch": 2.0685185185185184,
+      "grad_norm": 1.4043201527727218,
+      "learning_rate": 1.997528358493106e-07,
+      "loss": 0.742,
+      "step": 1117
+    },
+    {
+      "epoch": 2.0703703703703704,
+      "grad_norm": 1.3973675755732033,
+      "learning_rate": 1.9975178115745204e-07,
+      "loss": 0.714,
+      "step": 1118
+    },
+    {
+      "epoch": 2.0722222222222224,
+      "grad_norm": 1.2952777302609368,
+      "learning_rate": 1.9975072422290252e-07,
+      "loss": 0.7583,
+      "step": 1119
+    },
+    {
+      "epoch": 2.074074074074074,
+      "grad_norm": 1.556133467251769,
+      "learning_rate": 1.997496650456858e-07,
+      "loss": 0.7359,
+      "step": 1120
+    },
+    {
+      "epoch": 2.075925925925926,
+      "grad_norm": 3.4386429960327054,
+      "learning_rate": 1.997486036258257e-07,
+      "loss": 0.7409,
+      "step": 1121
+    },
+    {
+      "epoch": 2.077777777777778,
+      "grad_norm": 1.4861788414353911,
+      "learning_rate": 1.997475399633461e-07,
+      "loss": 0.7835,
+      "step": 1122
+    },
+    {
+      "epoch": 2.0796296296296295,
+      "grad_norm": 1.388262414056101,
+      "learning_rate": 1.9974647405827088e-07,
+      "loss": 0.7216,
+      "step": 1123
+    },
+    {
+      "epoch": 2.0814814814814815,
+      "grad_norm": 1.2440384428919482,
+      "learning_rate": 1.9974540591062398e-07,
+      "loss": 0.6982,
+      "step": 1124
+    },
+    {
+      "epoch": 2.0833333333333335,
+      "grad_norm": 1.631293916240114,
+      "learning_rate": 1.9974433552042947e-07,
+      "loss": 0.7499,
+      "step": 1125
+    },
+    {
+      "epoch": 2.085185185185185,
+      "grad_norm": 1.298597813965211,
+      "learning_rate": 1.9974326288771143e-07,
+      "loss": 0.7482,
+      "step": 1126
+    },
+    {
+      "epoch": 2.087037037037037,
+      "grad_norm": 2.0400880225703513,
+      "learning_rate": 1.997421880124939e-07,
+      "loss": 0.7982,
+      "step": 1127
+    },
+    {
+      "epoch": 2.088888888888889,
+      "grad_norm": 1.7138634485578874,
+      "learning_rate": 1.997411108948011e-07,
+      "loss": 0.7892,
+      "step": 1128
+    },
+    {
+      "epoch": 2.0907407407407406,
+      "grad_norm": 1.435262908773315,
+      "learning_rate": 1.9974003153465725e-07,
+      "loss": 0.7813,
+      "step": 1129
+    },
+    {
+      "epoch": 2.0925925925925926,
+      "grad_norm": 1.3592285851029247,
+      "learning_rate": 1.9973894993208658e-07,
+      "loss": 0.7445,
+      "step": 1130
+    },
+    {
+      "epoch": 2.0944444444444446,
+      "grad_norm": 2.179842417002354,
+      "learning_rate": 1.9973786608711342e-07,
+      "loss": 0.7464,
+      "step": 1131
+    },
+    {
+      "epoch": 2.096296296296296,
+      "grad_norm": 1.2926719020528166,
+      "learning_rate": 1.9973677999976218e-07,
+      "loss": 0.8023,
+      "step": 1132
+    },
+    {
+      "epoch": 2.098148148148148,
+      "grad_norm": 1.6511375747136297,
+      "learning_rate": 1.9973569167005722e-07,
+      "loss": 0.7284,
+      "step": 1133
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.6715280736077083,
+      "learning_rate": 1.9973460109802304e-07,
+      "loss": 0.6956,
+      "step": 1134
+    },
+    {
+      "epoch": 2.1018518518518516,
+      "grad_norm": 1.6447842784651854,
+      "learning_rate": 1.9973350828368414e-07,
+      "loss": 0.754,
+      "step": 1135
+    },
+    {
+      "epoch": 2.1037037037037036,
+      "grad_norm": 1.3897705242109888,
+      "learning_rate": 1.9973241322706512e-07,
+      "loss": 0.7037,
+      "step": 1136
+    },
+    {
+      "epoch": 2.1055555555555556,
+      "grad_norm": 2.161984534013614,
+      "learning_rate": 1.997313159281906e-07,
+      "loss": 0.7844,
+      "step": 1137
+    },
+    {
+      "epoch": 2.1074074074074076,
+      "grad_norm": 1.4229790022471303,
+      "learning_rate": 1.9973021638708518e-07,
+      "loss": 0.7621,
+      "step": 1138
+    },
+    {
+      "epoch": 2.109259259259259,
+      "grad_norm": 1.3590593506100692,
+      "learning_rate": 1.9972911460377365e-07,
+      "loss": 0.7357,
+      "step": 1139
+    },
+    {
+      "epoch": 2.111111111111111,
+      "grad_norm": 1.408787627489393,
+      "learning_rate": 1.9972801057828079e-07,
+      "loss": 0.7683,
+      "step": 1140
+    },
+    {
+      "epoch": 2.112962962962963,
+      "grad_norm": 1.2304459761898376,
+      "learning_rate": 1.9972690431063138e-07,
+      "loss": 0.738,
+      "step": 1141
+    },
+    {
+      "epoch": 2.1148148148148147,
+      "grad_norm": 3.5184333962917296,
+      "learning_rate": 1.997257958008503e-07,
+      "loss": 0.7544,
+      "step": 1142
+    },
+    {
+      "epoch": 2.1166666666666667,
+      "grad_norm": 1.2813847531083853,
+      "learning_rate": 1.997246850489625e-07,
+      "loss": 0.734,
+      "step": 1143
+    },
+    {
+      "epoch": 2.1185185185185187,
+      "grad_norm": 1.763551807969625,
+      "learning_rate": 1.997235720549929e-07,
+      "loss": 0.7159,
+      "step": 1144
+    },
+    {
+      "epoch": 2.1203703703703702,
+      "grad_norm": 1.330678602342504,
+      "learning_rate": 1.9972245681896657e-07,
+      "loss": 0.7259,
+      "step": 1145
+    },
+    {
+      "epoch": 2.1222222222222222,
+      "grad_norm": 1.506622127641233,
+      "learning_rate": 1.9972133934090858e-07,
+      "loss": 0.7881,
+      "step": 1146
+    },
+    {
+      "epoch": 2.1240740740740742,
+      "grad_norm": 2.4467463972573182,
+      "learning_rate": 1.9972021962084401e-07,
+      "loss": 0.7291,
+      "step": 1147
+    },
+    {
+      "epoch": 2.1259259259259258,
+      "grad_norm": 3.60433915632854,
+      "learning_rate": 1.997190976587981e-07,
+      "loss": 0.7796,
+      "step": 1148
+    },
+    {
+      "epoch": 2.1277777777777778,
+      "grad_norm": 1.4115383709763014,
+      "learning_rate": 1.9971797345479605e-07,
+      "loss": 0.7447,
+      "step": 1149
+    },
+    {
+      "epoch": 2.1296296296296298,
+      "grad_norm": 1.681619890488606,
+      "learning_rate": 1.9971684700886308e-07,
+      "loss": 0.7545,
+      "step": 1150
+    },
+    {
+      "epoch": 2.1314814814814813,
+      "grad_norm": 2.0417354930949325,
+      "learning_rate": 1.997157183210246e-07,
+      "loss": 0.6984,
+      "step": 1151
+    },
+    {
+      "epoch": 2.1333333333333333,
+      "grad_norm": 1.9867810299370199,
+      "learning_rate": 1.9971458739130595e-07,
+      "loss": 0.7509,
+      "step": 1152
+    },
+    {
+      "epoch": 2.1351851851851853,
+      "grad_norm": 1.295106427854352,
+      "learning_rate": 1.9971345421973256e-07,
+      "loss": 0.7768,
+      "step": 1153
+    },
+    {
+      "epoch": 2.137037037037037,
+      "grad_norm": 1.5979162158023175,
+      "learning_rate": 1.997123188063299e-07,
+      "loss": 0.7344,
+      "step": 1154
+    },
+    {
+      "epoch": 2.138888888888889,
+      "grad_norm": 2.377197014235787,
+      "learning_rate": 1.9971118115112344e-07,
+      "loss": 0.7425,
+      "step": 1155
+    },
+    {
+      "epoch": 2.140740740740741,
+      "grad_norm": 1.2497616842818906,
+      "learning_rate": 1.9971004125413886e-07,
+      "loss": 0.7177,
+      "step": 1156
+    },
+    {
+      "epoch": 2.1425925925925924,
+      "grad_norm": 1.4003956100181234,
+      "learning_rate": 1.9970889911540176e-07,
+      "loss": 0.7516,
+      "step": 1157
+    },
+    {
+      "epoch": 2.1444444444444444,
+      "grad_norm": 1.6348103457723056,
+      "learning_rate": 1.9970775473493778e-07,
+      "loss": 0.7801,
+      "step": 1158
+    },
+    {
+      "epoch": 2.1462962962962964,
+      "grad_norm": 1.4136450512382126,
+      "learning_rate": 1.9970660811277268e-07,
+      "loss": 0.6966,
+      "step": 1159
+    },
+    {
+      "epoch": 2.148148148148148,
+      "grad_norm": 1.6142455764087282,
+      "learning_rate": 1.9970545924893223e-07,
+      "loss": 0.7495,
+      "step": 1160
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.6178997612498114,
+      "learning_rate": 1.9970430814344228e-07,
+      "loss": 0.7966,
+      "step": 1161
+    },
+    {
+      "epoch": 2.151851851851852,
+      "grad_norm": 1.4232253204221887,
+      "learning_rate": 1.9970315479632866e-07,
+      "loss": 0.7422,
+      "step": 1162
+    },
+    {
+      "epoch": 2.153703703703704,
+      "grad_norm": 1.5044370898039416,
+      "learning_rate": 1.9970199920761735e-07,
+      "loss": 0.7694,
+      "step": 1163
+    },
+    {
+      "epoch": 2.1555555555555554,
+      "grad_norm": 1.3437644544948428,
+      "learning_rate": 1.9970084137733428e-07,
+      "loss": 0.7736,
+      "step": 1164
+    },
+    {
+      "epoch": 2.1574074074074074,
+      "grad_norm": 1.6890754307428033,
+      "learning_rate": 1.9969968130550554e-07,
+      "loss": 0.7808,
+      "step": 1165
+    },
+    {
+      "epoch": 2.1592592592592594,
+      "grad_norm": 1.361561129812041,
+      "learning_rate": 1.9969851899215716e-07,
+      "loss": 0.8053,
+      "step": 1166
+    },
+    {
+      "epoch": 2.161111111111111,
+      "grad_norm": 4.838929318922867,
+      "learning_rate": 1.9969735443731533e-07,
+      "loss": 0.7754,
+      "step": 1167
+    },
+    {
+      "epoch": 2.162962962962963,
+      "grad_norm": 1.5476326656490054,
+      "learning_rate": 1.996961876410062e-07,
+      "loss": 0.7765,
+      "step": 1168
+    },
+    {
+      "epoch": 2.164814814814815,
+      "grad_norm": 1.4084510628858158,
+      "learning_rate": 1.9969501860325597e-07,
+      "loss": 0.7676,
+      "step": 1169
+    },
+    {
+      "epoch": 2.1666666666666665,
+      "grad_norm": 1.4163701593789408,
+      "learning_rate": 1.9969384732409097e-07,
+      "loss": 0.7505,
+      "step": 1170
+    },
+    {
+      "epoch": 2.1685185185185185,
+      "grad_norm": 1.771755355140723,
+      "learning_rate": 1.996926738035375e-07,
+      "loss": 0.695,
+      "step": 1171
+    },
+    {
+      "epoch": 2.1703703703703705,
+      "grad_norm": 1.4669304605030435,
+      "learning_rate": 1.99691498041622e-07,
+      "loss": 0.7229,
+      "step": 1172
+    },
+    {
+      "epoch": 2.172222222222222,
+      "grad_norm": 1.3267890524408694,
+      "learning_rate": 1.9969032003837084e-07,
+      "loss": 0.7234,
+      "step": 1173
+    },
+    {
+      "epoch": 2.174074074074074,
+      "grad_norm": 1.4194547818969117,
+      "learning_rate": 1.9968913979381052e-07,
+      "loss": 0.7217,
+      "step": 1174
+    },
+    {
+      "epoch": 2.175925925925926,
+      "grad_norm": 1.5153391098990912,
+      "learning_rate": 1.9968795730796758e-07,
+      "loss": 0.7904,
+      "step": 1175
+    },
+    {
+      "epoch": 2.1777777777777776,
+      "grad_norm": 1.2940548838760708,
+      "learning_rate": 1.9968677258086865e-07,
+      "loss": 0.7094,
+      "step": 1176
+    },
+    {
+      "epoch": 2.1796296296296296,
+      "grad_norm": 1.490808723435155,
+      "learning_rate": 1.996855856125403e-07,
+      "loss": 0.7833,
+      "step": 1177
+    },
+    {
+      "epoch": 2.1814814814814816,
+      "grad_norm": 2.396912831421858,
+      "learning_rate": 1.9968439640300926e-07,
+      "loss": 0.7285,
+      "step": 1178
+    },
+    {
+      "epoch": 2.183333333333333,
+      "grad_norm": 2.6549758199448843,
+      "learning_rate": 1.9968320495230223e-07,
+      "loss": 0.6962,
+      "step": 1179
+    },
+    {
+      "epoch": 2.185185185185185,
+      "grad_norm": 1.5147539697111214,
+      "learning_rate": 1.9968201126044605e-07,
+      "loss": 0.7599,
+      "step": 1180
+    },
+    {
+      "epoch": 2.187037037037037,
+      "grad_norm": 1.4378491090761631,
+      "learning_rate": 1.996808153274675e-07,
+      "loss": 0.7822,
+      "step": 1181
+    },
+    {
+      "epoch": 2.188888888888889,
+      "grad_norm": 1.4882275792897923,
+      "learning_rate": 1.996796171533935e-07,
+      "loss": 0.789,
+      "step": 1182
+    },
+    {
+      "epoch": 2.1907407407407407,
+      "grad_norm": 1.4527471550256748,
+      "learning_rate": 1.99678416738251e-07,
+      "loss": 0.7387,
+      "step": 1183
+    },
+    {
+      "epoch": 2.1925925925925926,
+      "grad_norm": 1.4190799171358848,
+      "learning_rate": 1.9967721408206694e-07,
+      "loss": 0.6648,
+      "step": 1184
+    },
+    {
+      "epoch": 2.1944444444444446,
+      "grad_norm": 1.583024267678944,
+      "learning_rate": 1.9967600918486838e-07,
+      "loss": 0.7173,
+      "step": 1185
+    },
+    {
+      "epoch": 2.196296296296296,
+      "grad_norm": 1.512672592951793,
+      "learning_rate": 1.9967480204668244e-07,
+      "loss": 0.7384,
+      "step": 1186
+    },
+    {
+      "epoch": 2.198148148148148,
+      "grad_norm": 1.3230071571320559,
+      "learning_rate": 1.9967359266753623e-07,
+      "loss": 0.7779,
+      "step": 1187
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.527666430187374,
+      "learning_rate": 1.9967238104745695e-07,
+      "loss": 0.7709,
+      "step": 1188
+    },
+    {
+      "epoch": 2.2018518518518517,
+      "grad_norm": 1.5321498753215999,
+      "learning_rate": 1.996711671864718e-07,
+      "loss": 0.7489,
+      "step": 1189
+    },
+    {
+      "epoch": 2.2037037037037037,
+      "grad_norm": 1.62265937753039,
+      "learning_rate": 1.9966995108460816e-07,
+      "loss": 0.689,
+      "step": 1190
+    },
+    {
+      "epoch": 2.2055555555555557,
+      "grad_norm": 1.4880134468881105,
+      "learning_rate": 1.996687327418933e-07,
+      "loss": 0.7238,
+      "step": 1191
+    },
+    {
+      "epoch": 2.2074074074074073,
+      "grad_norm": 1.6630532477636921,
+      "learning_rate": 1.9966751215835465e-07,
+      "loss": 0.75,
+      "step": 1192
+    },
+    {
+      "epoch": 2.2092592592592593,
+      "grad_norm": 1.9684648966364346,
+      "learning_rate": 1.996662893340196e-07,
+      "loss": 0.7486,
+      "step": 1193
+    },
+    {
+      "epoch": 2.2111111111111112,
+      "grad_norm": 1.4030132391334154,
+      "learning_rate": 1.996650642689157e-07,
+      "loss": 0.7315,
+      "step": 1194
+    },
+    {
+      "epoch": 2.212962962962963,
+      "grad_norm": 1.911094077117738,
+      "learning_rate": 1.9966383696307045e-07,
+      "loss": 0.7229,
+      "step": 1195
+    },
+    {
+      "epoch": 2.214814814814815,
+      "grad_norm": 1.5840556545521352,
+      "learning_rate": 1.996626074165115e-07,
+      "loss": 0.7349,
+      "step": 1196
+    },
+    {
+      "epoch": 2.216666666666667,
+      "grad_norm": 1.756288854094917,
+      "learning_rate": 1.9966137562926643e-07,
+      "loss": 0.7007,
+      "step": 1197
+    },
+    {
+      "epoch": 2.2185185185185183,
+      "grad_norm": 1.3073936842333558,
+      "learning_rate": 1.9966014160136296e-07,
+      "loss": 0.7532,
+      "step": 1198
+    },
+    {
+      "epoch": 2.2203703703703703,
+      "grad_norm": 4.035589368652393,
+      "learning_rate": 1.9965890533282887e-07,
+      "loss": 0.7112,
+      "step": 1199
+    },
+    {
+      "epoch": 2.2222222222222223,
+      "grad_norm": 1.4732753028362748,
+      "learning_rate": 1.9965766682369185e-07,
+      "loss": 0.7302,
+      "step": 1200
+    },
+    {
+      "epoch": 2.224074074074074,
+      "grad_norm": 2.763755114368663,
+      "learning_rate": 1.9965642607397987e-07,
+      "loss": 0.7505,
+      "step": 1201
+    },
+    {
+      "epoch": 2.225925925925926,
+      "grad_norm": 1.339919554346128,
+      "learning_rate": 1.996551830837207e-07,
+      "loss": 0.6981,
+      "step": 1202
+    },
+    {
+      "epoch": 2.227777777777778,
+      "grad_norm": 1.4443921967016284,
+      "learning_rate": 1.9965393785294244e-07,
+      "loss": 0.7439,
+      "step": 1203
+    },
+    {
+      "epoch": 2.2296296296296294,
+      "grad_norm": 1.4073117309434156,
+      "learning_rate": 1.9965269038167295e-07,
+      "loss": 0.7162,
+      "step": 1204
+    },
+    {
+      "epoch": 2.2314814814814814,
+      "grad_norm": 1.5692579691249378,
+      "learning_rate": 1.9965144066994033e-07,
+      "loss": 0.7359,
+      "step": 1205
+    },
+    {
+      "epoch": 2.2333333333333334,
+      "grad_norm": 1.5463640968234122,
+      "learning_rate": 1.9965018871777271e-07,
+      "loss": 0.6792,
+      "step": 1206
+    },
+    {
+      "epoch": 2.2351851851851854,
+      "grad_norm": 1.4976390579987924,
+      "learning_rate": 1.9964893452519816e-07,
+      "loss": 0.7765,
+      "step": 1207
+    },
+    {
+      "epoch": 2.237037037037037,
+      "grad_norm": 1.43084691544816,
+      "learning_rate": 1.9964767809224494e-07,
+      "loss": 0.7612,
+      "step": 1208
+    },
+    {
+      "epoch": 2.238888888888889,
+      "grad_norm": 1.6311366341923859,
+      "learning_rate": 1.996464194189413e-07,
+      "loss": 0.7094,
+      "step": 1209
+    },
+    {
+      "epoch": 2.240740740740741,
+      "grad_norm": 2.576922359812976,
+      "learning_rate": 1.9964515850531548e-07,
+      "loss": 0.7372,
+      "step": 1210
+    },
+    {
+      "epoch": 2.2425925925925925,
+      "grad_norm": 2.2331007759152834,
+      "learning_rate": 1.9964389535139588e-07,
+      "loss": 0.768,
+      "step": 1211
+    },
+    {
+      "epoch": 2.2444444444444445,
+      "grad_norm": 1.2417765903652613,
+      "learning_rate": 1.9964262995721085e-07,
+      "loss": 0.7452,
+      "step": 1212
+    },
+    {
+      "epoch": 2.2462962962962965,
+      "grad_norm": 1.3297827731211054,
+      "learning_rate": 1.9964136232278893e-07,
+      "loss": 0.7143,
+      "step": 1213
+    },
+    {
+      "epoch": 2.248148148148148,
+      "grad_norm": 1.4082222752683888,
+      "learning_rate": 1.9964009244815854e-07,
+      "loss": 0.734,
+      "step": 1214
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.3311351221397543,
+      "learning_rate": 1.9963882033334824e-07,
+      "loss": 0.7364,
+      "step": 1215
+    },
+    {
+      "epoch": 2.251851851851852,
+      "grad_norm": 1.5331442442302905,
+      "learning_rate": 1.9963754597838666e-07,
+      "loss": 0.7347,
+      "step": 1216
+    },
+    {
+      "epoch": 2.2537037037037035,
+      "grad_norm": 1.5942800975078237,
+      "learning_rate": 1.9963626938330242e-07,
+      "loss": 0.7264,
+      "step": 1217
+    },
+    {
+      "epoch": 2.2555555555555555,
+      "grad_norm": 1.4532867896154087,
+      "learning_rate": 1.9963499054812422e-07,
+      "loss": 0.8134,
+      "step": 1218
+    },
+    {
+      "epoch": 2.2574074074074075,
+      "grad_norm": 1.4106077017962686,
+      "learning_rate": 1.9963370947288087e-07,
+      "loss": 0.7602,
+      "step": 1219
+    },
+    {
+      "epoch": 2.259259259259259,
+      "grad_norm": 1.4617468746958917,
+      "learning_rate": 1.996324261576011e-07,
+      "loss": 0.7383,
+      "step": 1220
+    },
+    {
+      "epoch": 2.261111111111111,
+      "grad_norm": 1.3703956515571203,
+      "learning_rate": 1.996311406023138e-07,
+      "loss": 0.7622,
+      "step": 1221
+    },
+    {
+      "epoch": 2.262962962962963,
+      "grad_norm": 1.4280626801500316,
+      "learning_rate": 1.9962985280704787e-07,
+      "loss": 0.7503,
+      "step": 1222
+    },
+    {
+      "epoch": 2.264814814814815,
+      "grad_norm": 1.4779680443888599,
+      "learning_rate": 1.996285627718323e-07,
+      "loss": 0.7718,
+      "step": 1223
+    },
+    {
+      "epoch": 2.2666666666666666,
+      "grad_norm": 1.4319549088731647,
+      "learning_rate": 1.9962727049669598e-07,
+      "loss": 0.7636,
+      "step": 1224
+    },
+    {
+      "epoch": 2.2685185185185186,
+      "grad_norm": 1.5659256944422904,
+      "learning_rate": 1.9962597598166807e-07,
+      "loss": 0.7127,
+      "step": 1225
+    },
+    {
+      "epoch": 2.2703703703703706,
+      "grad_norm": 1.382465822642072,
+      "learning_rate": 1.9962467922677763e-07,
+      "loss": 0.6671,
+      "step": 1226
+    },
+    {
+      "epoch": 2.272222222222222,
+      "grad_norm": 1.9137542357073167,
+      "learning_rate": 1.9962338023205385e-07,
+      "loss": 0.7209,
+      "step": 1227
+    },
+    {
+      "epoch": 2.274074074074074,
+      "grad_norm": 1.3210331280527465,
+      "learning_rate": 1.9962207899752588e-07,
+      "loss": 0.741,
+      "step": 1228
+    },
+    {
+      "epoch": 2.275925925925926,
+      "grad_norm": 1.2074600758268053,
+      "learning_rate": 1.9962077552322301e-07,
+      "loss": 0.7324,
+      "step": 1229
+    },
+    {
+      "epoch": 2.2777777777777777,
+      "grad_norm": 1.400795484828022,
+      "learning_rate": 1.9961946980917453e-07,
+      "loss": 0.7263,
+      "step": 1230
+    },
+    {
+      "epoch": 2.2796296296296297,
+      "grad_norm": 2.166761449238475,
+      "learning_rate": 1.9961816185540985e-07,
+      "loss": 0.7601,
+      "step": 1231
+    },
+    {
+      "epoch": 2.2814814814814817,
+      "grad_norm": 1.3370869307741051,
+      "learning_rate": 1.996168516619583e-07,
+      "loss": 0.7751,
+      "step": 1232
+    },
+    {
+      "epoch": 2.283333333333333,
+      "grad_norm": 1.4225726500809228,
+      "learning_rate": 1.9961553922884935e-07,
+      "loss": 0.7314,
+      "step": 1233
+    },
+    {
+      "epoch": 2.285185185185185,
+      "grad_norm": 1.5971809673079405,
+      "learning_rate": 1.9961422455611257e-07,
+      "loss": 0.691,
+      "step": 1234
+    },
+    {
+      "epoch": 2.287037037037037,
+      "grad_norm": 1.8156131681572563,
+      "learning_rate": 1.9961290764377745e-07,
+      "loss": 0.7631,
+      "step": 1235
+    },
+    {
+      "epoch": 2.2888888888888888,
+      "grad_norm": 1.8245386969298583,
+      "learning_rate": 1.996115884918736e-07,
+      "loss": 0.7428,
+      "step": 1236
+    },
+    {
+      "epoch": 2.2907407407407407,
+      "grad_norm": 1.6666003394322357,
+      "learning_rate": 1.9961026710043075e-07,
+      "loss": 0.7396,
+      "step": 1237
+    },
+    {
+      "epoch": 2.2925925925925927,
+      "grad_norm": 3.2775651045587724,
+      "learning_rate": 1.9960894346947855e-07,
+      "loss": 0.7456,
+      "step": 1238
+    },
+    {
+      "epoch": 2.2944444444444443,
+      "grad_norm": 1.313254787643138,
+      "learning_rate": 1.9960761759904675e-07,
+      "loss": 0.7371,
+      "step": 1239
+    },
+    {
+      "epoch": 2.2962962962962963,
+      "grad_norm": 1.4514041104248467,
+      "learning_rate": 1.9960628948916516e-07,
+      "loss": 0.7223,
+      "step": 1240
+    },
+    {
+      "epoch": 2.2981481481481483,
+      "grad_norm": 1.6307566339317496,
+      "learning_rate": 1.9960495913986367e-07,
+      "loss": 0.7649,
+      "step": 1241
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.3757045434901785,
+      "learning_rate": 1.9960362655117216e-07,
+      "loss": 0.7395,
+      "step": 1242
+    },
+    {
+      "epoch": 2.301851851851852,
+      "grad_norm": 1.3181530481192756,
+      "learning_rate": 1.9960229172312064e-07,
+      "loss": 0.7341,
+      "step": 1243
+    },
+    {
+      "epoch": 2.303703703703704,
+      "grad_norm": 1.3159019446998637,
+      "learning_rate": 1.9960095465573909e-07,
+      "loss": 0.7108,
+      "step": 1244
+    },
+    {
+      "epoch": 2.3055555555555554,
+      "grad_norm": 1.439739292413786,
+      "learning_rate": 1.995996153490575e-07,
+      "loss": 0.7208,
+      "step": 1245
+    },
+    {
+      "epoch": 2.3074074074074074,
+      "grad_norm": 1.5639247460941688,
+      "learning_rate": 1.9959827380310613e-07,
+      "loss": 0.7648,
+      "step": 1246
+    },
+    {
+      "epoch": 2.3092592592592593,
+      "grad_norm": 1.3664602578227352,
+      "learning_rate": 1.99596930017915e-07,
+      "loss": 0.7544,
+      "step": 1247
+    },
+    {
+      "epoch": 2.311111111111111,
+      "grad_norm": 2.074099644035599,
+      "learning_rate": 1.995955839935144e-07,
+      "loss": 0.7118,
+      "step": 1248
+    },
+    {
+      "epoch": 2.312962962962963,
+      "grad_norm": 1.465785641487809,
+      "learning_rate": 1.995942357299346e-07,
+      "loss": 0.7399,
+      "step": 1249
+    },
+    {
+      "epoch": 2.314814814814815,
+      "grad_norm": 2.4772583402368995,
+      "learning_rate": 1.9959288522720588e-07,
+      "loss": 0.7644,
+      "step": 1250
+    },
+    {
+      "epoch": 2.3166666666666664,
+      "grad_norm": 1.4168197817198627,
+      "learning_rate": 1.995915324853586e-07,
+      "loss": 0.7667,
+      "step": 1251
+    },
+    {
+      "epoch": 2.3185185185185184,
+      "grad_norm": 1.315761213911486,
+      "learning_rate": 1.9959017750442316e-07,
+      "loss": 0.7354,
+      "step": 1252
+    },
+    {
+      "epoch": 2.3203703703703704,
+      "grad_norm": 1.6012055837160273,
+      "learning_rate": 1.9958882028443007e-07,
+      "loss": 0.7702,
+      "step": 1253
+    },
+    {
+      "epoch": 2.3222222222222224,
+      "grad_norm": 3.869447565213466,
+      "learning_rate": 1.9958746082540984e-07,
+      "loss": 0.7378,
+      "step": 1254
+    },
+    {
+      "epoch": 2.324074074074074,
+      "grad_norm": 1.3896457987577644,
+      "learning_rate": 1.99586099127393e-07,
+      "loss": 0.7938,
+      "step": 1255
+    },
+    {
+      "epoch": 2.325925925925926,
+      "grad_norm": 1.3285227542943618,
+      "learning_rate": 1.9958473519041014e-07,
+      "loss": 0.7564,
+      "step": 1256
+    },
+    {
+      "epoch": 2.327777777777778,
+      "grad_norm": 1.5622805090178404,
+      "learning_rate": 1.99583369014492e-07,
+      "loss": 0.7465,
+      "step": 1257
+    },
+    {
+      "epoch": 2.3296296296296295,
+      "grad_norm": 1.7133870464560872,
+      "learning_rate": 1.9958200059966928e-07,
+      "loss": 0.7404,
+      "step": 1258
+    },
+    {
+      "epoch": 2.3314814814814815,
+      "grad_norm": 1.226662033422179,
+      "learning_rate": 1.9958062994597272e-07,
+      "loss": 0.7507,
+      "step": 1259
+    },
+    {
+      "epoch": 2.3333333333333335,
+      "grad_norm": 1.3357050930059793,
+      "learning_rate": 1.995792570534331e-07,
+      "loss": 0.7348,
+      "step": 1260
+    },
+    {
+      "epoch": 2.335185185185185,
+      "grad_norm": 1.4116054751822662,
+      "learning_rate": 1.9957788192208137e-07,
+      "loss": 0.7208,
+      "step": 1261
+    },
+    {
+      "epoch": 2.337037037037037,
+      "grad_norm": 1.4944051441425161,
+      "learning_rate": 1.9957650455194836e-07,
+      "loss": 0.7459,
+      "step": 1262
+    },
+    {
+      "epoch": 2.338888888888889,
+      "grad_norm": 1.3274322361818032,
+      "learning_rate": 1.9957512494306513e-07,
+      "loss": 0.7394,
+      "step": 1263
+    },
+    {
+      "epoch": 2.3407407407407406,
+      "grad_norm": 1.2784429905748773,
+      "learning_rate": 1.9957374309546263e-07,
+      "loss": 0.7193,
+      "step": 1264
+    },
+    {
+      "epoch": 2.3425925925925926,
+      "grad_norm": 1.347064083367779,
+      "learning_rate": 1.9957235900917195e-07,
+      "loss": 0.7014,
+      "step": 1265
+    },
+    {
+      "epoch": 2.3444444444444446,
+      "grad_norm": 1.3100819833547577,
+      "learning_rate": 1.995709726842242e-07,
+      "loss": 0.7503,
+      "step": 1266
+    },
+    {
+      "epoch": 2.346296296296296,
+      "grad_norm": 2.040831965584761,
+      "learning_rate": 1.9956958412065056e-07,
+      "loss": 0.7427,
+      "step": 1267
+    },
+    {
+      "epoch": 2.348148148148148,
+      "grad_norm": 5.09687394405009,
+      "learning_rate": 1.9956819331848226e-07,
+      "loss": 0.7324,
+      "step": 1268
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.436584781908511,
+      "learning_rate": 1.995668002777505e-07,
+      "loss": 0.7291,
+      "step": 1269
+    },
+    {
+      "epoch": 2.351851851851852,
+      "grad_norm": 1.390596417737228,
+      "learning_rate": 1.9956540499848668e-07,
+      "loss": 0.7284,
+      "step": 1270
+    },
+    {
+      "epoch": 2.3537037037037036,
+      "grad_norm": 2.217339272217483,
+      "learning_rate": 1.9956400748072214e-07,
+      "loss": 0.726,
+      "step": 1271
+    },
+    {
+      "epoch": 2.3555555555555556,
+      "grad_norm": 1.3019079412972838,
+      "learning_rate": 1.9956260772448833e-07,
+      "loss": 0.771,
+      "step": 1272
+    },
+    {
+      "epoch": 2.3574074074074076,
+      "grad_norm": 1.4091549947469788,
+      "learning_rate": 1.9956120572981666e-07,
+      "loss": 0.6765,
+      "step": 1273
+    },
+    {
+      "epoch": 2.359259259259259,
+      "grad_norm": 1.7272101322961766,
+      "learning_rate": 1.9955980149673867e-07,
+      "loss": 0.7287,
+      "step": 1274
+    },
+    {
+      "epoch": 2.361111111111111,
+      "grad_norm": 1.7592990089891896,
+      "learning_rate": 1.9955839502528597e-07,
+      "loss": 0.7055,
+      "step": 1275
+    },
+    {
+      "epoch": 2.362962962962963,
+      "grad_norm": 1.2522616599956227,
+      "learning_rate": 1.9955698631549015e-07,
+      "loss": 0.71,
+      "step": 1276
+    },
+    {
+      "epoch": 2.3648148148148147,
+      "grad_norm": 1.4979857231808944,
+      "learning_rate": 1.9955557536738284e-07,
+      "loss": 0.775,
+      "step": 1277
+    },
+    {
+      "epoch": 2.3666666666666667,
+      "grad_norm": 1.9229564453154884,
+      "learning_rate": 1.9955416218099585e-07,
+      "loss": 0.7172,
+      "step": 1278
+    },
+    {
+      "epoch": 2.3685185185185187,
+      "grad_norm": 1.7020988661355843,
+      "learning_rate": 1.9955274675636092e-07,
+      "loss": 0.7076,
+      "step": 1279
+    },
+    {
+      "epoch": 2.3703703703703702,
+      "grad_norm": 1.3211692785207134,
+      "learning_rate": 1.9955132909350983e-07,
+      "loss": 0.7127,
+      "step": 1280
+    },
+    {
+      "epoch": 2.3722222222222222,
+      "grad_norm": 1.3728623978053636,
+      "learning_rate": 1.9954990919247447e-07,
+      "loss": 0.7479,
+      "step": 1281
+    },
+    {
+      "epoch": 2.3740740740740742,
+      "grad_norm": 1.4129051162117727,
+      "learning_rate": 1.9954848705328682e-07,
+      "loss": 0.706,
+      "step": 1282
+    },
+    {
+      "epoch": 2.3759259259259258,
+      "grad_norm": 1.4864202567030413,
+      "learning_rate": 1.995470626759788e-07,
+      "loss": 0.7829,
+      "step": 1283
+    },
+    {
+      "epoch": 2.3777777777777778,
+      "grad_norm": 1.4322383635756901,
+      "learning_rate": 1.9954563606058243e-07,
+      "loss": 0.7504,
+      "step": 1284
+    },
+    {
+      "epoch": 2.3796296296296298,
+      "grad_norm": 1.2738809704083554,
+      "learning_rate": 1.9954420720712982e-07,
+      "loss": 0.7072,
+      "step": 1285
+    },
+    {
+      "epoch": 2.3814814814814813,
+      "grad_norm": 1.504796334831779,
+      "learning_rate": 1.9954277611565307e-07,
+      "loss": 0.7309,
+      "step": 1286
+    },
+    {
+      "epoch": 2.3833333333333333,
+      "grad_norm": 1.3528167079894706,
+      "learning_rate": 1.9954134278618432e-07,
+      "loss": 0.7315,
+      "step": 1287
+    },
+    {
+      "epoch": 2.3851851851851853,
+      "grad_norm": 1.544388573843145,
+      "learning_rate": 1.995399072187559e-07,
+      "loss": 0.7539,
+      "step": 1288
+    },
+    {
+      "epoch": 2.387037037037037,
+      "grad_norm": 1.3101024883482584,
+      "learning_rate": 1.995384694134e-07,
+      "loss": 0.7848,
+      "step": 1289
+    },
+    {
+      "epoch": 2.388888888888889,
+      "grad_norm": 2.739151201465498,
+      "learning_rate": 1.9953702937014892e-07,
+      "loss": 0.7352,
+      "step": 1290
+    },
+    {
+      "epoch": 2.390740740740741,
+      "grad_norm": 1.594452795133058,
+      "learning_rate": 1.9953558708903513e-07,
+      "loss": 0.768,
+      "step": 1291
+    },
+    {
+      "epoch": 2.3925925925925924,
+      "grad_norm": 1.4577210153692408,
+      "learning_rate": 1.9953414257009098e-07,
+      "loss": 0.7099,
+      "step": 1292
+    },
+    {
+      "epoch": 2.3944444444444444,
+      "grad_norm": 1.4523890645160162,
+      "learning_rate": 1.9953269581334898e-07,
+      "loss": 0.7724,
+      "step": 1293
+    },
+    {
+      "epoch": 2.3962962962962964,
+      "grad_norm": 2.38572112543369,
+      "learning_rate": 1.9953124681884166e-07,
+      "loss": 0.8075,
+      "step": 1294
+    },
+    {
+      "epoch": 2.398148148148148,
+      "grad_norm": 1.5545941224078266,
+      "learning_rate": 1.995297955866016e-07,
+      "loss": 0.7447,
+      "step": 1295
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.3918212878154332,
+      "learning_rate": 1.9952834211666138e-07,
+      "loss": 0.7315,
+      "step": 1296
+    },
+    {
+      "epoch": 2.401851851851852,
+      "grad_norm": 1.5217648568059665,
+      "learning_rate": 1.9952688640905375e-07,
+      "loss": 0.718,
+      "step": 1297
+    },
+    {
+      "epoch": 2.403703703703704,
+      "grad_norm": 1.3679870358225215,
+      "learning_rate": 1.995254284638114e-07,
+      "loss": 0.6739,
+      "step": 1298
+    },
+    {
+      "epoch": 2.4055555555555554,
+      "grad_norm": 1.4514544719088338,
+      "learning_rate": 1.9952396828096707e-07,
+      "loss": 0.7539,
+      "step": 1299
+    },
+    {
+      "epoch": 2.4074074074074074,
+      "grad_norm": 1.5798593765026705,
+      "learning_rate": 1.9952250586055367e-07,
+      "loss": 0.6912,
+      "step": 1300
+    },
+    {
+      "epoch": 2.4092592592592594,
+      "grad_norm": 1.323997059452625,
+      "learning_rate": 1.9952104120260405e-07,
+      "loss": 0.7257,
+      "step": 1301
+    },
+    {
+      "epoch": 2.411111111111111,
+      "grad_norm": 2.0748887970357517,
+      "learning_rate": 1.995195743071511e-07,
+      "loss": 0.7233,
+      "step": 1302
+    },
+    {
+      "epoch": 2.412962962962963,
+      "grad_norm": 1.2538093316094987,
+      "learning_rate": 1.9951810517422783e-07,
+      "loss": 0.7338,
+      "step": 1303
+    },
+    {
+      "epoch": 2.414814814814815,
+      "grad_norm": 1.3034081544174396,
+      "learning_rate": 1.9951663380386727e-07,
+      "loss": 0.7689,
+      "step": 1304
+    },
+    {
+      "epoch": 2.4166666666666665,
+      "grad_norm": 1.4305553163289428,
+      "learning_rate": 1.995151601961025e-07,
+      "loss": 0.7043,
+      "step": 1305
+    },
+    {
+      "epoch": 2.4185185185185185,
+      "grad_norm": 1.3918206507554012,
+      "learning_rate": 1.9951368435096666e-07,
+      "loss": 0.6981,
+      "step": 1306
+    },
+    {
+      "epoch": 2.4203703703703705,
+      "grad_norm": 2.115482621301663,
+      "learning_rate": 1.995122062684929e-07,
+      "loss": 0.7181,
+      "step": 1307
+    },
+    {
+      "epoch": 2.422222222222222,
+      "grad_norm": 1.5353132536292329,
+      "learning_rate": 1.995107259487145e-07,
+      "loss": 0.7267,
+      "step": 1308
+    },
+    {
+      "epoch": 2.424074074074074,
+      "grad_norm": 1.3220417503313013,
+      "learning_rate": 1.995092433916647e-07,
+      "loss": 0.742,
+      "step": 1309
+    },
+    {
+      "epoch": 2.425925925925926,
+      "grad_norm": 1.735163383524067,
+      "learning_rate": 1.9950775859737683e-07,
+      "loss": 0.7582,
+      "step": 1310
+    },
+    {
+      "epoch": 2.4277777777777776,
+      "grad_norm": 1.4669934710497194,
+      "learning_rate": 1.995062715658843e-07,
+      "loss": 0.7234,
+      "step": 1311
+    },
+    {
+      "epoch": 2.4296296296296296,
+      "grad_norm": 1.3589612687031636,
+      "learning_rate": 1.9950478229722054e-07,
+      "loss": 0.7331,
+      "step": 1312
+    },
+    {
+      "epoch": 2.4314814814814816,
+      "grad_norm": 1.4510512231537043,
+      "learning_rate": 1.9950329079141904e-07,
+      "loss": 0.7162,
+      "step": 1313
+    },
+    {
+      "epoch": 2.4333333333333336,
+      "grad_norm": 6.43609263841738,
+      "learning_rate": 1.995017970485133e-07,
+      "loss": 0.7439,
+      "step": 1314
+    },
+    {
+      "epoch": 2.435185185185185,
+      "grad_norm": 1.4550032112648212,
+      "learning_rate": 1.995003010685369e-07,
+      "loss": 0.7496,
+      "step": 1315
+    },
+    {
+      "epoch": 2.437037037037037,
+      "grad_norm": 1.28730307738335,
+      "learning_rate": 1.994988028515235e-07,
+      "loss": 0.7367,
+      "step": 1316
+    },
+    {
+      "epoch": 2.438888888888889,
+      "grad_norm": 1.493038642709901,
+      "learning_rate": 1.994973023975068e-07,
+      "loss": 0.7887,
+      "step": 1317
+    },
+    {
+      "epoch": 2.4407407407407407,
+      "grad_norm": 1.6992223520238683,
+      "learning_rate": 1.994957997065205e-07,
+      "loss": 0.6892,
+      "step": 1318
+    },
+    {
+      "epoch": 2.4425925925925926,
+      "grad_norm": 1.3734996204251164,
+      "learning_rate": 1.9949429477859838e-07,
+      "loss": 0.771,
+      "step": 1319
+    },
+    {
+      "epoch": 2.4444444444444446,
+      "grad_norm": 1.3051818716362529,
+      "learning_rate": 1.9949278761377428e-07,
+      "loss": 0.689,
+      "step": 1320
+    },
+    {
+      "epoch": 2.446296296296296,
+      "grad_norm": 1.5177936272081083,
+      "learning_rate": 1.9949127821208212e-07,
+      "loss": 0.7089,
+      "step": 1321
+    },
+    {
+      "epoch": 2.448148148148148,
+      "grad_norm": 1.403616389139682,
+      "learning_rate": 1.994897665735558e-07,
+      "loss": 0.7246,
+      "step": 1322
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.7081335790237482,
+      "learning_rate": 1.9948825269822933e-07,
+      "loss": 0.7574,
+      "step": 1323
+    },
+    {
+      "epoch": 2.4518518518518517,
+      "grad_norm": 2.498650256362956,
+      "learning_rate": 1.9948673658613675e-07,
+      "loss": 0.7201,
+      "step": 1324
+    },
+    {
+      "epoch": 2.4537037037037037,
+      "grad_norm": 1.3487085473802438,
+      "learning_rate": 1.994852182373121e-07,
+      "loss": 0.736,
+      "step": 1325
+    },
+    {
+      "epoch": 2.4555555555555557,
+      "grad_norm": 1.3782873827988964,
+      "learning_rate": 1.9948369765178955e-07,
+      "loss": 0.735,
+      "step": 1326
+    },
+    {
+      "epoch": 2.4574074074074073,
+      "grad_norm": 1.621600873009482,
+      "learning_rate": 1.994821748296033e-07,
+      "loss": 0.7499,
+      "step": 1327
+    },
+    {
+      "epoch": 2.4592592592592593,
+      "grad_norm": 1.3240194327308723,
+      "learning_rate": 1.9948064977078755e-07,
+      "loss": 0.7066,
+      "step": 1328
+    },
+    {
+      "epoch": 2.4611111111111112,
+      "grad_norm": 1.4321939897417713,
+      "learning_rate": 1.9947912247537665e-07,
+      "loss": 0.7502,
+      "step": 1329
+    },
+    {
+      "epoch": 2.462962962962963,
+      "grad_norm": 1.476595812814863,
+      "learning_rate": 1.9947759294340485e-07,
+      "loss": 0.6964,
+      "step": 1330
+    },
+    {
+      "epoch": 2.464814814814815,
+      "grad_norm": 1.6128769573405382,
+      "learning_rate": 1.9947606117490657e-07,
+      "loss": 0.7027,
+      "step": 1331
+    },
+    {
+      "epoch": 2.466666666666667,
+      "grad_norm": 2.436776070200861,
+      "learning_rate": 1.994745271699163e-07,
+      "loss": 0.7566,
+      "step": 1332
+    },
+    {
+      "epoch": 2.4685185185185183,
+      "grad_norm": 2.201402659636539,
+      "learning_rate": 1.994729909284685e-07,
+      "loss": 0.7294,
+      "step": 1333
+    },
+    {
+      "epoch": 2.4703703703703703,
+      "grad_norm": 1.4525911019143531,
+      "learning_rate": 1.9947145245059765e-07,
+      "loss": 0.7577,
+      "step": 1334
+    },
+    {
+      "epoch": 2.4722222222222223,
+      "grad_norm": 1.47649127251014,
+      "learning_rate": 1.9946991173633841e-07,
+      "loss": 0.7068,
+      "step": 1335
+    },
+    {
+      "epoch": 2.474074074074074,
+      "grad_norm": 1.3796124237134613,
+      "learning_rate": 1.9946836878572543e-07,
+      "loss": 0.7753,
+      "step": 1336
+    },
+    {
+      "epoch": 2.475925925925926,
+      "grad_norm": 1.3455624677273677,
+      "learning_rate": 1.9946682359879333e-07,
+      "loss": 0.7383,
+      "step": 1337
+    },
+    {
+      "epoch": 2.477777777777778,
+      "grad_norm": 1.9675019541056298,
+      "learning_rate": 1.994652761755769e-07,
+      "loss": 0.732,
+      "step": 1338
+    },
+    {
+      "epoch": 2.4796296296296294,
+      "grad_norm": 2.2608283646835274,
+      "learning_rate": 1.9946372651611092e-07,
+      "loss": 0.7165,
+      "step": 1339
+    },
+    {
+      "epoch": 2.4814814814814814,
+      "grad_norm": 1.4798478832314137,
+      "learning_rate": 1.9946217462043023e-07,
+      "loss": 0.7298,
+      "step": 1340
+    },
+    {
+      "epoch": 2.4833333333333334,
+      "grad_norm": 1.5757807776503756,
+      "learning_rate": 1.994606204885697e-07,
+      "loss": 0.7077,
+      "step": 1341
+    },
+    {
+      "epoch": 2.485185185185185,
+      "grad_norm": 1.7699428796798822,
+      "learning_rate": 1.9945906412056432e-07,
+      "loss": 0.7411,
+      "step": 1342
+    },
+    {
+      "epoch": 2.487037037037037,
+      "grad_norm": 1.5205513973035123,
+      "learning_rate": 1.9945750551644905e-07,
+      "loss": 0.7566,
+      "step": 1343
+    },
+    {
+      "epoch": 2.488888888888889,
+      "grad_norm": 1.4683735155241477,
+      "learning_rate": 1.9945594467625892e-07,
+      "loss": 0.7031,
+      "step": 1344
+    },
+    {
+      "epoch": 2.490740740740741,
+      "grad_norm": 1.665967937097463,
+      "learning_rate": 1.9945438160002902e-07,
+      "loss": 0.7377,
+      "step": 1345
+    },
+    {
+      "epoch": 2.4925925925925925,
+      "grad_norm": 1.4088919029718936,
+      "learning_rate": 1.9945281628779453e-07,
+      "loss": 0.782,
+      "step": 1346
+    },
+    {
+      "epoch": 2.4944444444444445,
+      "grad_norm": 2.5187312198219125,
+      "learning_rate": 1.994512487395906e-07,
+      "loss": 0.7331,
+      "step": 1347
+    },
+    {
+      "epoch": 2.4962962962962965,
+      "grad_norm": 1.9306301542382656,
+      "learning_rate": 1.9944967895545253e-07,
+      "loss": 0.7386,
+      "step": 1348
+    },
+    {
+      "epoch": 2.498148148148148,
+      "grad_norm": 2.3790078314720136,
+      "learning_rate": 1.9944810693541556e-07,
+      "loss": 0.7261,
+      "step": 1349
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.487959192167854,
+      "learning_rate": 1.9944653267951503e-07,
+      "loss": 0.722,
+      "step": 1350
+    },
+    {
+      "epoch": 2.501851851851852,
+      "grad_norm": 1.3848128439458334,
+      "learning_rate": 1.9944495618778635e-07,
+      "loss": 0.7249,
+      "step": 1351
+    },
+    {
+      "epoch": 2.5037037037037035,
+      "grad_norm": 1.3473791688552903,
+      "learning_rate": 1.99443377460265e-07,
+      "loss": 0.737,
+      "step": 1352
+    },
+    {
+      "epoch": 2.5055555555555555,
+      "grad_norm": 1.7181077632032955,
+      "learning_rate": 1.994417964969864e-07,
+      "loss": 0.6886,
+      "step": 1353
+    },
+    {
+      "epoch": 2.5074074074074075,
+      "grad_norm": 2.25765993896238,
+      "learning_rate": 1.9944021329798616e-07,
+      "loss": 0.7461,
+      "step": 1354
+    },
+    {
+      "epoch": 2.5092592592592595,
+      "grad_norm": 3.2947546451288794,
+      "learning_rate": 1.9943862786329988e-07,
+      "loss": 0.7655,
+      "step": 1355
+    },
+    {
+      "epoch": 2.511111111111111,
+      "grad_norm": 1.66835857253872,
+      "learning_rate": 1.9943704019296313e-07,
+      "loss": 0.7143,
+      "step": 1356
+    },
+    {
+      "epoch": 2.512962962962963,
+      "grad_norm": 1.3785416513400173,
+      "learning_rate": 1.9943545028701166e-07,
+      "loss": 0.7233,
+      "step": 1357
+    },
+    {
+      "epoch": 2.514814814814815,
+      "grad_norm": 1.2805502946902554,
+      "learning_rate": 1.9943385814548122e-07,
+      "loss": 0.755,
+      "step": 1358
+    },
+    {
+      "epoch": 2.5166666666666666,
+      "grad_norm": 1.419040120188716,
+      "learning_rate": 1.9943226376840755e-07,
+      "loss": 0.7479,
+      "step": 1359
+    },
+    {
+      "epoch": 2.5185185185185186,
+      "grad_norm": 1.4365460621979755,
+      "learning_rate": 1.9943066715582658e-07,
+      "loss": 0.6918,
+      "step": 1360
+    },
+    {
+      "epoch": 2.5203703703703706,
+      "grad_norm": 1.6644583860050746,
+      "learning_rate": 1.9942906830777412e-07,
+      "loss": 0.733,
+      "step": 1361
+    },
+    {
+      "epoch": 2.522222222222222,
+      "grad_norm": 2.8935566897237788,
+      "learning_rate": 1.994274672242862e-07,
+      "loss": 0.7312,
+      "step": 1362
+    },
+    {
+      "epoch": 2.524074074074074,
+      "grad_norm": 1.4212290242919747,
+      "learning_rate": 1.9942586390539877e-07,
+      "loss": 0.7392,
+      "step": 1363
+    },
+    {
+      "epoch": 2.525925925925926,
+      "grad_norm": 1.377153985466174,
+      "learning_rate": 1.994242583511479e-07,
+      "loss": 0.7452,
+      "step": 1364
+    },
+    {
+      "epoch": 2.5277777777777777,
+      "grad_norm": 2.3282075988119817,
+      "learning_rate": 1.994226505615696e-07,
+      "loss": 0.7243,
+      "step": 1365
+    },
+    {
+      "epoch": 2.5296296296296297,
+      "grad_norm": 1.6438608755254531,
+      "learning_rate": 1.9942104053670015e-07,
+      "loss": 0.7521,
+      "step": 1366
+    },
+    {
+      "epoch": 2.5314814814814817,
+      "grad_norm": 1.5157406787614836,
+      "learning_rate": 1.9941942827657567e-07,
+      "loss": 0.6806,
+      "step": 1367
+    },
+    {
+      "epoch": 2.533333333333333,
+      "grad_norm": 1.836797133163215,
+      "learning_rate": 1.994178137812324e-07,
+      "loss": 0.757,
+      "step": 1368
+    },
+    {
+      "epoch": 2.535185185185185,
+      "grad_norm": 1.4177530089585573,
+      "learning_rate": 1.9941619705070672e-07,
+      "loss": 0.728,
+      "step": 1369
+    },
+    {
+      "epoch": 2.537037037037037,
+      "grad_norm": 1.4183179301649709,
+      "learning_rate": 1.9941457808503485e-07,
+      "loss": 0.7261,
+      "step": 1370
+    },
+    {
+      "epoch": 2.5388888888888888,
+      "grad_norm": 1.2933395766770701,
+      "learning_rate": 1.994129568842533e-07,
+      "loss": 0.7443,
+      "step": 1371
+    },
+    {
+      "epoch": 2.5407407407407407,
+      "grad_norm": 1.377139936520184,
+      "learning_rate": 1.9941133344839847e-07,
+      "loss": 0.7305,
+      "step": 1372
+    },
+    {
+      "epoch": 2.5425925925925927,
+      "grad_norm": 1.447004496252634,
+      "learning_rate": 1.9940970777750685e-07,
+      "loss": 0.7111,
+      "step": 1373
+    },
+    {
+      "epoch": 2.5444444444444443,
+      "grad_norm": 1.4073204308846854,
+      "learning_rate": 1.9940807987161504e-07,
+      "loss": 0.7492,
+      "step": 1374
+    },
+    {
+      "epoch": 2.5462962962962963,
+      "grad_norm": 1.3087732721771723,
+      "learning_rate": 1.9940644973075956e-07,
+      "loss": 0.711,
+      "step": 1375
+    },
+    {
+      "epoch": 2.5481481481481483,
+      "grad_norm": 1.489240328186331,
+      "learning_rate": 1.9940481735497716e-07,
+      "loss": 0.7501,
+      "step": 1376
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.7552615420474227,
+      "learning_rate": 1.9940318274430447e-07,
+      "loss": 0.7166,
+      "step": 1377
+    },
+    {
+      "epoch": 2.551851851851852,
+      "grad_norm": 1.4688327907631578,
+      "learning_rate": 1.9940154589877823e-07,
+      "loss": 0.6983,
+      "step": 1378
+    },
+    {
+      "epoch": 2.553703703703704,
+      "grad_norm": 1.4345992939228318,
+      "learning_rate": 1.993999068184353e-07,
+      "loss": 0.7107,
+      "step": 1379
+    },
+    {
+      "epoch": 2.5555555555555554,
+      "grad_norm": 1.962583253413522,
+      "learning_rate": 1.993982655033125e-07,
+      "loss": 0.6895,
+      "step": 1380
+    },
+    {
+      "epoch": 2.5574074074074074,
+      "grad_norm": 1.4584206018477688,
+      "learning_rate": 1.993966219534467e-07,
+      "loss": 0.7353,
+      "step": 1381
+    },
+    {
+      "epoch": 2.5592592592592593,
+      "grad_norm": 7.344289544183646,
+      "learning_rate": 1.9939497616887488e-07,
+      "loss": 0.7287,
+      "step": 1382
+    },
+    {
+      "epoch": 2.561111111111111,
+      "grad_norm": 1.606620569367819,
+      "learning_rate": 1.9939332814963405e-07,
+      "loss": 0.7304,
+      "step": 1383
+    },
+    {
+      "epoch": 2.562962962962963,
+      "grad_norm": 1.601226709074676,
+      "learning_rate": 1.9939167789576128e-07,
+      "loss": 0.748,
+      "step": 1384
+    },
+    {
+      "epoch": 2.564814814814815,
+      "grad_norm": 1.2828542326409174,
+      "learning_rate": 1.9939002540729365e-07,
+      "loss": 0.7691,
+      "step": 1385
+    },
+    {
+      "epoch": 2.5666666666666664,
+      "grad_norm": 2.049625059439289,
+      "learning_rate": 1.993883706842683e-07,
+      "loss": 0.7137,
+      "step": 1386
+    },
+    {
+      "epoch": 2.5685185185185184,
+      "grad_norm": 1.8496576138698904,
+      "learning_rate": 1.993867137267224e-07,
+      "loss": 0.6919,
+      "step": 1387
+    },
+    {
+      "epoch": 2.5703703703703704,
+      "grad_norm": 1.311962953399448,
+      "learning_rate": 1.993850545346933e-07,
+      "loss": 0.757,
+      "step": 1388
+    },
+    {
+      "epoch": 2.572222222222222,
+      "grad_norm": 1.2462191087023502,
+      "learning_rate": 1.993833931082182e-07,
+      "loss": 0.7222,
+      "step": 1389
+    },
+    {
+      "epoch": 2.574074074074074,
+      "grad_norm": 2.1964495667189095,
+      "learning_rate": 1.9938172944733454e-07,
+      "loss": 0.7353,
+      "step": 1390
+    },
+    {
+      "epoch": 2.575925925925926,
+      "grad_norm": 1.4278099404405515,
+      "learning_rate": 1.9938006355207966e-07,
+      "loss": 0.7526,
+      "step": 1391
+    },
+    {
+      "epoch": 2.5777777777777775,
+      "grad_norm": 1.4407481870984324,
+      "learning_rate": 1.9937839542249106e-07,
+      "loss": 0.7132,
+      "step": 1392
+    },
+    {
+      "epoch": 2.5796296296296295,
+      "grad_norm": 1.6406838100182555,
+      "learning_rate": 1.9937672505860622e-07,
+      "loss": 0.7211,
+      "step": 1393
+    },
+    {
+      "epoch": 2.5814814814814815,
+      "grad_norm": 1.5264800298000034,
+      "learning_rate": 1.993750524604627e-07,
+      "loss": 0.7311,
+      "step": 1394
+    },
+    {
+      "epoch": 2.5833333333333335,
+      "grad_norm": 2.2747301835962737,
+      "learning_rate": 1.993733776280981e-07,
+      "loss": 0.7087,
+      "step": 1395
+    },
+    {
+      "epoch": 2.585185185185185,
+      "grad_norm": 1.237741273066771,
+      "learning_rate": 1.9937170056155006e-07,
+      "loss": 0.7222,
+      "step": 1396
+    },
+    {
+      "epoch": 2.587037037037037,
+      "grad_norm": 1.3213607402135585,
+      "learning_rate": 1.993700212608563e-07,
+      "loss": 0.755,
+      "step": 1397
+    },
+    {
+      "epoch": 2.588888888888889,
+      "grad_norm": 1.4684444180742577,
+      "learning_rate": 1.993683397260546e-07,
+      "loss": 0.704,
+      "step": 1398
+    },
+    {
+      "epoch": 2.590740740740741,
+      "grad_norm": 1.2577211850671282,
+      "learning_rate": 1.9936665595718273e-07,
+      "loss": 0.7304,
+      "step": 1399
+    },
+    {
+      "epoch": 2.5925925925925926,
+      "grad_norm": 1.4732745439814985,
+      "learning_rate": 1.9936496995427857e-07,
+      "loss": 0.7101,
+      "step": 1400
+    },
+    {
+      "epoch": 2.5944444444444446,
+      "grad_norm": 1.4041608303881208,
+      "learning_rate": 1.9936328171738e-07,
+      "loss": 0.7039,
+      "step": 1401
+    },
+    {
+      "epoch": 2.5962962962962965,
+      "grad_norm": 1.4545056768760085,
+      "learning_rate": 1.99361591246525e-07,
+      "loss": 0.7094,
+      "step": 1402
+    },
+    {
+      "epoch": 2.598148148148148,
+      "grad_norm": 1.4346120185530165,
+      "learning_rate": 1.9935989854175156e-07,
+      "loss": 0.7105,
+      "step": 1403
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.5121752691030814,
+      "learning_rate": 1.9935820360309775e-07,
+      "loss": 0.7099,
+      "step": 1404
+    },
+    {
+      "epoch": 2.601851851851852,
+      "grad_norm": 2.8549940448132385,
+      "learning_rate": 1.9935650643060168e-07,
+      "loss": 0.7345,
+      "step": 1405
+    },
+    {
+      "epoch": 2.6037037037037036,
+      "grad_norm": 1.3666763697107338,
+      "learning_rate": 1.993548070243015e-07,
+      "loss": 0.7059,
+      "step": 1406
+    },
+    {
+      "epoch": 2.6055555555555556,
+      "grad_norm": 1.4739599529807506,
+      "learning_rate": 1.9935310538423538e-07,
+      "loss": 0.739,
+      "step": 1407
+    },
+    {
+      "epoch": 2.6074074074074076,
+      "grad_norm": 1.4019653185119212,
+      "learning_rate": 1.9935140151044164e-07,
+      "loss": 0.76,
+      "step": 1408
+    },
+    {
+      "epoch": 2.609259259259259,
+      "grad_norm": 1.9540045908761232,
+      "learning_rate": 1.9934969540295853e-07,
+      "loss": 0.7597,
+      "step": 1409
+    },
+    {
+      "epoch": 2.611111111111111,
+      "grad_norm": 1.3182232284969628,
+      "learning_rate": 1.9934798706182448e-07,
+      "loss": 0.6808,
+      "step": 1410
+    },
+    {
+      "epoch": 2.612962962962963,
+      "grad_norm": 1.6506572134468076,
+      "learning_rate": 1.993462764870778e-07,
+      "loss": 0.7151,
+      "step": 1411
+    },
+    {
+      "epoch": 2.6148148148148147,
+      "grad_norm": 1.4244205286532778,
+      "learning_rate": 1.9934456367875706e-07,
+      "loss": 0.7234,
+      "step": 1412
+    },
+    {
+      "epoch": 2.6166666666666667,
+      "grad_norm": 1.5765445142615166,
+      "learning_rate": 1.993428486369007e-07,
+      "loss": 0.72,
+      "step": 1413
+    },
+    {
+      "epoch": 2.6185185185185187,
+      "grad_norm": 1.3517511430814697,
+      "learning_rate": 1.9934113136154726e-07,
+      "loss": 0.7293,
+      "step": 1414
+    },
+    {
+      "epoch": 2.6203703703703702,
+      "grad_norm": 1.4218381796670414,
+      "learning_rate": 1.9933941185273543e-07,
+      "loss": 0.714,
+      "step": 1415
+    },
+    {
+      "epoch": 2.6222222222222222,
+      "grad_norm": 1.6556829414793377,
+      "learning_rate": 1.9933769011050378e-07,
+      "loss": 0.73,
+      "step": 1416
+    },
+    {
+      "epoch": 2.6240740740740742,
+      "grad_norm": 1.6957755935795544,
+      "learning_rate": 1.993359661348911e-07,
+      "loss": 0.7166,
+      "step": 1417
+    },
+    {
+      "epoch": 2.6259259259259258,
+      "grad_norm": 1.441535191694854,
+      "learning_rate": 1.9933423992593607e-07,
+      "loss": 0.7141,
+      "step": 1418
+    },
+    {
+      "epoch": 2.6277777777777778,
+      "grad_norm": 1.647606304559682,
+      "learning_rate": 1.9933251148367755e-07,
+      "loss": 0.7395,
+      "step": 1419
+    },
+    {
+      "epoch": 2.6296296296296298,
+      "grad_norm": 1.7169659387407783,
+      "learning_rate": 1.993307808081544e-07,
+      "loss": 0.7043,
+      "step": 1420
+    },
+    {
+      "epoch": 2.6314814814814813,
+      "grad_norm": 1.648043118822608,
+      "learning_rate": 1.993290478994055e-07,
+      "loss": 0.6766,
+      "step": 1421
+    },
+    {
+      "epoch": 2.6333333333333333,
+      "grad_norm": 1.8411238108635852,
+      "learning_rate": 1.9932731275746984e-07,
+      "loss": 0.7535,
+      "step": 1422
+    },
+    {
+      "epoch": 2.6351851851851853,
+      "grad_norm": 1.3398947894551478,
+      "learning_rate": 1.9932557538238642e-07,
+      "loss": 0.7046,
+      "step": 1423
+    },
+    {
+      "epoch": 2.637037037037037,
+      "grad_norm": 1.5219901478819284,
+      "learning_rate": 1.9932383577419428e-07,
+      "loss": 0.7018,
+      "step": 1424
+    },
+    {
+      "epoch": 2.638888888888889,
+      "grad_norm": 1.8535331870555205,
+      "learning_rate": 1.9932209393293258e-07,
+      "loss": 0.6998,
+      "step": 1425
+    },
+    {
+      "epoch": 2.640740740740741,
+      "grad_norm": 1.3948215769488879,
+      "learning_rate": 1.9932034985864043e-07,
+      "loss": 0.7281,
+      "step": 1426
+    },
+    {
+      "epoch": 2.6425925925925924,
+      "grad_norm": 1.4687380050627368,
+      "learning_rate": 1.993186035513571e-07,
+      "loss": 0.6579,
+      "step": 1427
+    },
+    {
+      "epoch": 2.6444444444444444,
+      "grad_norm": 1.5444439131304921,
+      "learning_rate": 1.9931685501112182e-07,
+      "loss": 0.7495,
+      "step": 1428
+    },
+    {
+      "epoch": 2.6462962962962964,
+      "grad_norm": 1.4573829658949196,
+      "learning_rate": 1.9931510423797384e-07,
+      "loss": 0.7507,
+      "step": 1429
+    },
+    {
+      "epoch": 2.648148148148148,
+      "grad_norm": 1.5164027382102139,
+      "learning_rate": 1.9931335123195262e-07,
+      "loss": 0.729,
+      "step": 1430
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.7807591373454794,
+      "learning_rate": 1.9931159599309755e-07,
+      "loss": 0.6606,
+      "step": 1431
+    },
+    {
+      "epoch": 2.651851851851852,
+      "grad_norm": 1.9537125393507286,
+      "learning_rate": 1.9930983852144804e-07,
+      "loss": 0.7524,
+      "step": 1432
+    },
+    {
+      "epoch": 2.6537037037037035,
+      "grad_norm": 1.5076329470963121,
+      "learning_rate": 1.9930807881704368e-07,
+      "loss": 0.7292,
+      "step": 1433
+    },
+    {
+      "epoch": 2.6555555555555554,
+      "grad_norm": 1.7149101274722025,
+      "learning_rate": 1.9930631687992396e-07,
+      "loss": 0.7527,
+      "step": 1434
+    },
+    {
+      "epoch": 2.6574074074074074,
+      "grad_norm": 1.9147017997919344,
+      "learning_rate": 1.993045527101285e-07,
+      "loss": 0.7356,
+      "step": 1435
+    },
+    {
+      "epoch": 2.659259259259259,
+      "grad_norm": 1.4184128796472453,
+      "learning_rate": 1.9930278630769704e-07,
+      "loss": 0.7204,
+      "step": 1436
+    },
+    {
+      "epoch": 2.661111111111111,
+      "grad_norm": 1.3583863839665171,
+      "learning_rate": 1.9930101767266923e-07,
+      "loss": 0.7351,
+      "step": 1437
+    },
+    {
+      "epoch": 2.662962962962963,
+      "grad_norm": 1.604272902725679,
+      "learning_rate": 1.9929924680508484e-07,
+      "loss": 0.7012,
+      "step": 1438
+    },
+    {
+      "epoch": 2.664814814814815,
+      "grad_norm": 1.6204049473591053,
+      "learning_rate": 1.9929747370498368e-07,
+      "loss": 0.7633,
+      "step": 1439
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 1.4753625913045245,
+      "learning_rate": 1.9929569837240564e-07,
+      "loss": 0.7193,
+      "step": 1440
+    },
+    {
+      "epoch": 2.6685185185185185,
+      "grad_norm": 1.3481003940160432,
+      "learning_rate": 1.9929392080739057e-07,
+      "loss": 0.7271,
+      "step": 1441
+    },
+    {
+      "epoch": 2.6703703703703705,
+      "grad_norm": 1.5428727208360309,
+      "learning_rate": 1.9929214100997854e-07,
+      "loss": 0.721,
+      "step": 1442
+    },
+    {
+      "epoch": 2.6722222222222225,
+      "grad_norm": 1.3178025410945733,
+      "learning_rate": 1.9929035898020948e-07,
+      "loss": 0.7191,
+      "step": 1443
+    },
+    {
+      "epoch": 2.674074074074074,
+      "grad_norm": 1.376741882777622,
+      "learning_rate": 1.9928857471812347e-07,
+      "loss": 0.7082,
+      "step": 1444
+    },
+    {
+      "epoch": 2.675925925925926,
+      "grad_norm": 1.3999587100507738,
+      "learning_rate": 1.992867882237607e-07,
+      "loss": 0.736,
+      "step": 1445
+    },
+    {
+      "epoch": 2.677777777777778,
+      "grad_norm": 1.4187994381577271,
+      "learning_rate": 1.9928499949716122e-07,
+      "loss": 0.688,
+      "step": 1446
+    },
+    {
+      "epoch": 2.6796296296296296,
+      "grad_norm": 1.3688063742951624,
+      "learning_rate": 1.9928320853836529e-07,
+      "loss": 0.7288,
+      "step": 1447
+    },
+    {
+      "epoch": 2.6814814814814816,
+      "grad_norm": 1.5174559167512862,
+      "learning_rate": 1.992814153474132e-07,
+      "loss": 0.7429,
+      "step": 1448
+    },
+    {
+      "epoch": 2.6833333333333336,
+      "grad_norm": 1.345651670511597,
+      "learning_rate": 1.9927961992434526e-07,
+      "loss": 0.6998,
+      "step": 1449
+    },
+    {
+      "epoch": 2.685185185185185,
+      "grad_norm": 1.2843568364875257,
+      "learning_rate": 1.992778222692018e-07,
+      "loss": 0.7112,
+      "step": 1450
+    },
+    {
+      "epoch": 2.687037037037037,
+      "grad_norm": 1.3332390141321087,
+      "learning_rate": 1.9927602238202326e-07,
+      "loss": 0.7223,
+      "step": 1451
+    },
+    {
+      "epoch": 2.688888888888889,
+      "grad_norm": 1.22297008094667,
+      "learning_rate": 1.9927422026285012e-07,
+      "loss": 0.7217,
+      "step": 1452
+    },
+    {
+      "epoch": 2.6907407407407407,
+      "grad_norm": 2.744603741519903,
+      "learning_rate": 1.9927241591172287e-07,
+      "loss": 0.6965,
+      "step": 1453
+    },
+    {
+      "epoch": 2.6925925925925926,
+      "grad_norm": 1.5452297436865128,
+      "learning_rate": 1.992706093286821e-07,
+      "loss": 0.7409,
+      "step": 1454
+    },
+    {
+      "epoch": 2.6944444444444446,
+      "grad_norm": 1.3773033529222274,
+      "learning_rate": 1.9926880051376841e-07,
+      "loss": 0.7949,
+      "step": 1455
+    },
+    {
+      "epoch": 2.696296296296296,
+      "grad_norm": 1.4404363061410415,
+      "learning_rate": 1.9926698946702251e-07,
+      "loss": 0.7191,
+      "step": 1456
+    },
+    {
+      "epoch": 2.698148148148148,
+      "grad_norm": 1.5785980737770442,
+      "learning_rate": 1.9926517618848506e-07,
+      "loss": 0.6962,
+      "step": 1457
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.9138041993716404,
+      "learning_rate": 1.9926336067819682e-07,
+      "loss": 0.7411,
+      "step": 1458
+    },
+    {
+      "epoch": 2.7018518518518517,
+      "grad_norm": 1.2699927395404789,
+      "learning_rate": 1.9926154293619868e-07,
+      "loss": 0.6818,
+      "step": 1459
+    },
+    {
+      "epoch": 2.7037037037037037,
+      "grad_norm": 2.652407051472237,
+      "learning_rate": 1.9925972296253144e-07,
+      "loss": 0.7473,
+      "step": 1460
+    },
+    {
+      "epoch": 2.7055555555555557,
+      "grad_norm": 1.753958906352326,
+      "learning_rate": 1.9925790075723605e-07,
+      "loss": 0.7087,
+      "step": 1461
+    },
+    {
+      "epoch": 2.7074074074074073,
+      "grad_norm": 1.3529865864936963,
+      "learning_rate": 1.9925607632035345e-07,
+      "loss": 0.6758,
+      "step": 1462
+    },
+    {
+      "epoch": 2.7092592592592593,
+      "grad_norm": 1.37313979976493,
+      "learning_rate": 1.9925424965192473e-07,
+      "loss": 0.7214,
+      "step": 1463
+    },
+    {
+      "epoch": 2.7111111111111112,
+      "grad_norm": 1.407332563273367,
+      "learning_rate": 1.9925242075199088e-07,
+      "loss": 0.7252,
+      "step": 1464
+    },
+    {
+      "epoch": 2.712962962962963,
+      "grad_norm": 1.4053114667167617,
+      "learning_rate": 1.9925058962059302e-07,
+      "loss": 0.6868,
+      "step": 1465
+    },
+    {
+      "epoch": 2.714814814814815,
+      "grad_norm": 1.3720977567276125,
+      "learning_rate": 1.9924875625777238e-07,
+      "loss": 0.7308,
+      "step": 1466
+    },
+    {
+      "epoch": 2.716666666666667,
+      "grad_norm": 2.031860874764236,
+      "learning_rate": 1.9924692066357012e-07,
+      "loss": 0.724,
+      "step": 1467
+    },
+    {
+      "epoch": 2.7185185185185183,
+      "grad_norm": 1.6766940600295293,
+      "learning_rate": 1.9924508283802755e-07,
+      "loss": 0.7078,
+      "step": 1468
+    },
+    {
+      "epoch": 2.7203703703703703,
+      "grad_norm": 1.702889744510947,
+      "learning_rate": 1.9924324278118597e-07,
+      "loss": 0.7516,
+      "step": 1469
+    },
+    {
+      "epoch": 2.7222222222222223,
+      "grad_norm": 1.4106313576934042,
+      "learning_rate": 1.9924140049308673e-07,
+      "loss": 0.7098,
+      "step": 1470
+    },
+    {
+      "epoch": 2.724074074074074,
+      "grad_norm": 1.5248315352087267,
+      "learning_rate": 1.9923955597377131e-07,
+      "loss": 0.7323,
+      "step": 1471
+    },
+    {
+      "epoch": 2.725925925925926,
+      "grad_norm": 2.0992228365384964,
+      "learning_rate": 1.9923770922328112e-07,
+      "loss": 0.7515,
+      "step": 1472
+    },
+    {
+      "epoch": 2.727777777777778,
+      "grad_norm": 1.8377903683184047,
+      "learning_rate": 1.9923586024165767e-07,
+      "loss": 0.7468,
+      "step": 1473
+    },
+    {
+      "epoch": 2.7296296296296294,
+      "grad_norm": 2.1622808185079907,
+      "learning_rate": 1.9923400902894261e-07,
+      "loss": 0.7584,
+      "step": 1474
+    },
+    {
+      "epoch": 2.7314814814814814,
+      "grad_norm": 1.380384955132983,
+      "learning_rate": 1.992321555851775e-07,
+      "loss": 0.702,
+      "step": 1475
+    },
+    {
+      "epoch": 2.7333333333333334,
+      "grad_norm": 1.3547755070212955,
+      "learning_rate": 1.99230299910404e-07,
+      "loss": 0.6587,
+      "step": 1476
+    },
+    {
+      "epoch": 2.735185185185185,
+      "grad_norm": 1.6097112785705026,
+      "learning_rate": 1.992284420046639e-07,
+      "loss": 0.6953,
+      "step": 1477
+    },
+    {
+      "epoch": 2.737037037037037,
+      "grad_norm": 1.2901271040411313,
+      "learning_rate": 1.9922658186799887e-07,
+      "loss": 0.7578,
+      "step": 1478
+    },
+    {
+      "epoch": 2.738888888888889,
+      "grad_norm": 1.4345757814223434,
+      "learning_rate": 1.992247195004508e-07,
+      "loss": 0.6525,
+      "step": 1479
+    },
+    {
+      "epoch": 2.7407407407407405,
+      "grad_norm": 1.3629627367709394,
+      "learning_rate": 1.9922285490206154e-07,
+      "loss": 0.7624,
+      "step": 1480
+    },
+    {
+      "epoch": 2.7425925925925925,
+      "grad_norm": 1.3013276237592246,
+      "learning_rate": 1.9922098807287304e-07,
+      "loss": 0.6946,
+      "step": 1481
+    },
+    {
+      "epoch": 2.7444444444444445,
+      "grad_norm": 1.5417261700302196,
+      "learning_rate": 1.9921911901292722e-07,
+      "loss": 0.7015,
+      "step": 1482
+    },
+    {
+      "epoch": 2.7462962962962965,
+      "grad_norm": 1.429741192286043,
+      "learning_rate": 1.9921724772226618e-07,
+      "loss": 0.7385,
+      "step": 1483
+    },
+    {
+      "epoch": 2.748148148148148,
+      "grad_norm": 1.3542948761957248,
+      "learning_rate": 1.992153742009319e-07,
+      "loss": 0.6911,
+      "step": 1484
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.367272080925192,
+      "learning_rate": 1.9921349844896654e-07,
+      "loss": 0.7095,
+      "step": 1485
+    },
+    {
+      "epoch": 2.751851851851852,
+      "grad_norm": 1.4296083497101486,
+      "learning_rate": 1.9921162046641228e-07,
+      "loss": 0.738,
+      "step": 1486
+    },
+    {
+      "epoch": 2.7537037037037035,
+      "grad_norm": 1.3315970825011056,
+      "learning_rate": 1.9920974025331138e-07,
+      "loss": 0.7091,
+      "step": 1487
+    },
+    {
+      "epoch": 2.7555555555555555,
+      "grad_norm": 1.4888919175325739,
+      "learning_rate": 1.99207857809706e-07,
+      "loss": 0.6967,
+      "step": 1488
+    },
+    {
+      "epoch": 2.7574074074074075,
+      "grad_norm": 1.611702710474475,
+      "learning_rate": 1.992059731356386e-07,
+      "loss": 0.6973,
+      "step": 1489
+    },
+    {
+      "epoch": 2.7592592592592595,
+      "grad_norm": 1.5556533935399386,
+      "learning_rate": 1.992040862311514e-07,
+      "loss": 0.7161,
+      "step": 1490
+    },
+    {
+      "epoch": 2.761111111111111,
+      "grad_norm": 1.888419657741896,
+      "learning_rate": 1.9920219709628699e-07,
+      "loss": 0.7144,
+      "step": 1491
+    },
+    {
+      "epoch": 2.762962962962963,
+      "grad_norm": 2.711646101270849,
+      "learning_rate": 1.9920030573108772e-07,
+      "loss": 0.7775,
+      "step": 1492
+    },
+    {
+      "epoch": 2.764814814814815,
+      "grad_norm": 1.7408632639415766,
+      "learning_rate": 1.9919841213559615e-07,
+      "loss": 0.6922,
+      "step": 1493
+    },
+    {
+      "epoch": 2.7666666666666666,
+      "grad_norm": 1.6941950561119365,
+      "learning_rate": 1.9919651630985487e-07,
+      "loss": 0.6921,
+      "step": 1494
+    },
+    {
+      "epoch": 2.7685185185185186,
+      "grad_norm": 1.3737057114402744,
+      "learning_rate": 1.9919461825390646e-07,
+      "loss": 0.6744,
+      "step": 1495
+    },
+    {
+      "epoch": 2.7703703703703706,
+      "grad_norm": 1.9139374928838895,
+      "learning_rate": 1.9919271796779367e-07,
+      "loss": 0.7123,
+      "step": 1496
+    },
+    {
+      "epoch": 2.772222222222222,
+      "grad_norm": 1.6811225389365376,
+      "learning_rate": 1.9919081545155916e-07,
+      "loss": 0.7564,
+      "step": 1497
+    },
+    {
+      "epoch": 2.774074074074074,
+      "grad_norm": 1.4332878176164237,
+      "learning_rate": 1.991889107052457e-07,
+      "loss": 0.7419,
+      "step": 1498
+    },
+    {
+      "epoch": 2.775925925925926,
+      "grad_norm": 1.4935347248276818,
+      "learning_rate": 1.9918700372889616e-07,
+      "loss": 0.7517,
+      "step": 1499
+    },
+    {
+      "epoch": 2.7777777777777777,
+      "grad_norm": 1.384083210231503,
+      "learning_rate": 1.9918509452255336e-07,
+      "loss": 0.7019,
+      "step": 1500
+    },
+    {
+      "epoch": 2.7796296296296297,
+      "grad_norm": 1.413106769413716,
+      "learning_rate": 1.991831830862603e-07,
+      "loss": 0.7631,
+      "step": 1501
+    },
+    {
+      "epoch": 2.7814814814814817,
+      "grad_norm": 1.4970556202287384,
+      "learning_rate": 1.9918126942005985e-07,
+      "loss": 0.7165,
+      "step": 1502
+    },
+    {
+      "epoch": 2.783333333333333,
+      "grad_norm": 1.7728976124533433,
+      "learning_rate": 1.9917935352399512e-07,
+      "loss": 0.7331,
+      "step": 1503
+    },
+    {
+      "epoch": 2.785185185185185,
+      "grad_norm": 1.4171460514978556,
+      "learning_rate": 1.991774353981092e-07,
+      "loss": 0.7258,
+      "step": 1504
+    },
+    {
+      "epoch": 2.787037037037037,
+      "grad_norm": 1.8826535957849944,
+      "learning_rate": 1.991755150424451e-07,
+      "loss": 0.7141,
+      "step": 1505
+    },
+    {
+      "epoch": 2.7888888888888888,
+      "grad_norm": 1.6579873958720395,
+      "learning_rate": 1.9917359245704607e-07,
+      "loss": 0.7554,
+      "step": 1506
+    },
+    {
+      "epoch": 2.7907407407407407,
+      "grad_norm": 1.489400972138541,
+      "learning_rate": 1.9917166764195535e-07,
+      "loss": 0.7218,
+      "step": 1507
+    },
+    {
+      "epoch": 2.7925925925925927,
+      "grad_norm": 2.313111244149141,
+      "learning_rate": 1.991697405972162e-07,
+      "loss": 0.7267,
+      "step": 1508
+    },
+    {
+      "epoch": 2.7944444444444443,
+      "grad_norm": 1.3952291567025221,
+      "learning_rate": 1.9916781132287194e-07,
+      "loss": 0.7303,
+      "step": 1509
+    },
+    {
+      "epoch": 2.7962962962962963,
+      "grad_norm": 1.5890406608455308,
+      "learning_rate": 1.9916587981896596e-07,
+      "loss": 0.7241,
+      "step": 1510
+    },
+    {
+      "epoch": 2.7981481481481483,
+      "grad_norm": 1.3718581660900853,
+      "learning_rate": 1.9916394608554166e-07,
+      "loss": 0.7481,
+      "step": 1511
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.5160270562013836,
+      "learning_rate": 1.991620101226425e-07,
+      "loss": 0.6682,
+      "step": 1512
+    },
+    {
+      "epoch": 2.801851851851852,
+      "grad_norm": 1.3118551173038357,
+      "learning_rate": 1.9916007193031205e-07,
+      "loss": 0.7204,
+      "step": 1513
+    },
+    {
+      "epoch": 2.803703703703704,
+      "grad_norm": 1.2949838309786599,
+      "learning_rate": 1.9915813150859385e-07,
+      "loss": 0.7178,
+      "step": 1514
+    },
+    {
+      "epoch": 2.8055555555555554,
+      "grad_norm": 1.7004135791952835,
+      "learning_rate": 1.9915618885753158e-07,
+      "loss": 0.7734,
+      "step": 1515
+    },
+    {
+      "epoch": 2.8074074074074074,
+      "grad_norm": 1.352637727783801,
+      "learning_rate": 1.9915424397716886e-07,
+      "loss": 0.6849,
+      "step": 1516
+    },
+    {
+      "epoch": 2.8092592592592593,
+      "grad_norm": 1.5473451725920722,
+      "learning_rate": 1.9915229686754946e-07,
+      "loss": 0.6832,
+      "step": 1517
+    },
+    {
+      "epoch": 2.811111111111111,
+      "grad_norm": 1.3287342631789687,
+      "learning_rate": 1.9915034752871711e-07,
+      "loss": 0.69,
+      "step": 1518
+    },
+    {
+      "epoch": 2.812962962962963,
+      "grad_norm": 1.3720750580165226,
+      "learning_rate": 1.9914839596071566e-07,
+      "loss": 0.7501,
+      "step": 1519
+    },
+    {
+      "epoch": 2.814814814814815,
+      "grad_norm": 1.4232971879299827,
+      "learning_rate": 1.99146442163589e-07,
+      "loss": 0.7237,
+      "step": 1520
+    },
+    {
+      "epoch": 2.8166666666666664,
+      "grad_norm": 1.5821146734751461,
+      "learning_rate": 1.9914448613738103e-07,
+      "loss": 0.7185,
+      "step": 1521
+    },
+    {
+      "epoch": 2.8185185185185184,
+      "grad_norm": 1.4414412852138414,
+      "learning_rate": 1.9914252788213574e-07,
+      "loss": 0.6824,
+      "step": 1522
+    },
+    {
+      "epoch": 2.8203703703703704,
+      "grad_norm": 1.7144281826093233,
+      "learning_rate": 1.9914056739789716e-07,
+      "loss": 0.7105,
+      "step": 1523
+    },
+    {
+      "epoch": 2.822222222222222,
+      "grad_norm": 1.69747855090801,
+      "learning_rate": 1.9913860468470937e-07,
+      "loss": 0.7183,
+      "step": 1524
+    },
+    {
+      "epoch": 2.824074074074074,
+      "grad_norm": 1.9398439185782745,
+      "learning_rate": 1.9913663974261648e-07,
+      "loss": 0.6894,
+      "step": 1525
+    },
+    {
+      "epoch": 2.825925925925926,
+      "grad_norm": 3.6921179461241813,
+      "learning_rate": 1.9913467257166269e-07,
+      "loss": 0.6879,
+      "step": 1526
+    },
+    {
+      "epoch": 2.8277777777777775,
+      "grad_norm": 2.158776601881717,
+      "learning_rate": 1.991327031718922e-07,
+      "loss": 0.7145,
+      "step": 1527
+    },
+    {
+      "epoch": 2.8296296296296295,
+      "grad_norm": 1.3493932056090379,
+      "learning_rate": 1.9913073154334928e-07,
+      "loss": 0.7385,
+      "step": 1528
+    },
+    {
+      "epoch": 2.8314814814814815,
+      "grad_norm": 1.2785099004157228,
+      "learning_rate": 1.9912875768607832e-07,
+      "loss": 0.7401,
+      "step": 1529
+    },
+    {
+      "epoch": 2.8333333333333335,
+      "grad_norm": 1.8081616343782652,
+      "learning_rate": 1.9912678160012364e-07,
+      "loss": 0.7125,
+      "step": 1530
+    },
+    {
+      "epoch": 2.835185185185185,
+      "grad_norm": 1.5246545046562165,
+      "learning_rate": 1.991248032855297e-07,
+      "loss": 0.6957,
+      "step": 1531
+    },
+    {
+      "epoch": 2.837037037037037,
+      "grad_norm": 1.4880472415222572,
+      "learning_rate": 1.9912282274234094e-07,
+      "loss": 0.707,
+      "step": 1532
+    },
+    {
+      "epoch": 2.838888888888889,
+      "grad_norm": 1.383381944786622,
+      "learning_rate": 1.9912083997060192e-07,
+      "loss": 0.7503,
+      "step": 1533
+    },
+    {
+      "epoch": 2.840740740740741,
+      "grad_norm": 1.3895322485746548,
+      "learning_rate": 1.9911885497035724e-07,
+      "loss": 0.6786,
+      "step": 1534
+    },
+    {
+      "epoch": 2.8425925925925926,
+      "grad_norm": 1.393414344160297,
+      "learning_rate": 1.9911686774165145e-07,
+      "loss": 0.7138,
+      "step": 1535
+    },
+    {
+      "epoch": 2.8444444444444446,
+      "grad_norm": 1.4157120923346107,
+      "learning_rate": 1.991148782845293e-07,
+      "loss": 0.7342,
+      "step": 1536
+    },
+    {
+      "epoch": 2.8462962962962965,
+      "grad_norm": 1.4113867767478832,
+      "learning_rate": 1.991128865990355e-07,
+      "loss": 0.7532,
+      "step": 1537
+    },
+    {
+      "epoch": 2.848148148148148,
+      "grad_norm": 1.2641943736247099,
+      "learning_rate": 1.991108926852148e-07,
+      "loss": 0.7414,
+      "step": 1538
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.3477940433471907,
+      "learning_rate": 1.9910889654311208e-07,
+      "loss": 0.7254,
+      "step": 1539
+    },
+    {
+      "epoch": 2.851851851851852,
+      "grad_norm": 1.4723486302190898,
+      "learning_rate": 1.9910689817277215e-07,
+      "loss": 0.7422,
+      "step": 1540
+    },
+    {
+      "epoch": 2.8537037037037036,
+      "grad_norm": 1.4767158522435488,
+      "learning_rate": 1.9910489757424004e-07,
+      "loss": 0.744,
+      "step": 1541
+    },
+    {
+      "epoch": 2.8555555555555556,
+      "grad_norm": 1.4095226595398418,
+      "learning_rate": 1.9910289474756062e-07,
+      "loss": 0.7678,
+      "step": 1542
+    },
+    {
+      "epoch": 2.8574074074074076,
+      "grad_norm": 1.5217756251669254,
+      "learning_rate": 1.9910088969277895e-07,
+      "loss": 0.7445,
+      "step": 1543
+    },
+    {
+      "epoch": 2.859259259259259,
+      "grad_norm": 1.3469934114055042,
+      "learning_rate": 1.9909888240994017e-07,
+      "loss": 0.7406,
+      "step": 1544
+    },
+    {
+      "epoch": 2.861111111111111,
+      "grad_norm": 1.631959340029328,
+      "learning_rate": 1.9909687289908935e-07,
+      "loss": 0.7059,
+      "step": 1545
+    },
+    {
+      "epoch": 2.862962962962963,
+      "grad_norm": 1.4958800264921208,
+      "learning_rate": 1.9909486116027168e-07,
+      "loss": 0.7058,
+      "step": 1546
+    },
+    {
+      "epoch": 2.8648148148148147,
+      "grad_norm": 1.4236333734618511,
+      "learning_rate": 1.990928471935324e-07,
+      "loss": 0.7094,
+      "step": 1547
+    },
+    {
+      "epoch": 2.8666666666666667,
+      "grad_norm": 1.5178503788104258,
+      "learning_rate": 1.990908309989168e-07,
+      "loss": 0.7443,
+      "step": 1548
+    },
+    {
+      "epoch": 2.8685185185185187,
+      "grad_norm": 1.5105425013016076,
+      "learning_rate": 1.9908881257647016e-07,
+      "loss": 0.7396,
+      "step": 1549
+    },
+    {
+      "epoch": 2.8703703703703702,
+      "grad_norm": 2.355577851446489,
+      "learning_rate": 1.990867919262379e-07,
+      "loss": 0.7447,
+      "step": 1550
+    },
+    {
+      "epoch": 2.8722222222222222,
+      "grad_norm": 1.4677595499689686,
+      "learning_rate": 1.9908476904826547e-07,
+      "loss": 0.759,
+      "step": 1551
+    },
+    {
+      "epoch": 2.8740740740740742,
+      "grad_norm": 1.552627302641364,
+      "learning_rate": 1.9908274394259828e-07,
+      "loss": 0.751,
+      "step": 1552
+    },
+    {
+      "epoch": 2.8759259259259258,
+      "grad_norm": 1.6555852580999029,
+      "learning_rate": 1.9908071660928193e-07,
+      "loss": 0.7355,
+      "step": 1553
+    },
+    {
+      "epoch": 2.8777777777777778,
+      "grad_norm": 1.3723374885598025,
+      "learning_rate": 1.9907868704836194e-07,
+      "loss": 0.7222,
+      "step": 1554
+    },
+    {
+      "epoch": 2.8796296296296298,
+      "grad_norm": 1.5900472687022837,
+      "learning_rate": 1.9907665525988402e-07,
+      "loss": 0.7413,
+      "step": 1555
+    },
+    {
+      "epoch": 2.8814814814814813,
+      "grad_norm": 3.5880722849062674,
+      "learning_rate": 1.9907462124389377e-07,
+      "loss": 0.734,
+      "step": 1556
+    },
+    {
+      "epoch": 2.8833333333333333,
+      "grad_norm": 3.1509341016854746,
+      "learning_rate": 1.9907258500043694e-07,
+      "loss": 0.7355,
+      "step": 1557
+    },
+    {
+      "epoch": 2.8851851851851853,
+      "grad_norm": 1.5360821041028956,
+      "learning_rate": 1.9907054652955932e-07,
+      "loss": 0.6681,
+      "step": 1558
+    },
+    {
+      "epoch": 2.887037037037037,
+      "grad_norm": 1.4547394731485015,
+      "learning_rate": 1.9906850583130677e-07,
+      "loss": 0.7625,
+      "step": 1559
+    },
+    {
+      "epoch": 2.888888888888889,
+      "grad_norm": 1.709178223262143,
+      "learning_rate": 1.9906646290572513e-07,
+      "loss": 0.7002,
+      "step": 1560
+    },
+    {
+      "epoch": 2.890740740740741,
+      "grad_norm": 1.2054478082821016,
+      "learning_rate": 1.9906441775286035e-07,
+      "loss": 0.6804,
+      "step": 1561
+    },
+    {
+      "epoch": 2.8925925925925924,
+      "grad_norm": 1.666213109998135,
+      "learning_rate": 1.9906237037275843e-07,
+      "loss": 0.7851,
+      "step": 1562
+    },
+    {
+      "epoch": 2.8944444444444444,
+      "grad_norm": 1.6798417204777845,
+      "learning_rate": 1.9906032076546532e-07,
+      "loss": 0.6579,
+      "step": 1563
+    },
+    {
+      "epoch": 2.8962962962962964,
+      "grad_norm": 1.3826982573769029,
+      "learning_rate": 1.990582689310272e-07,
+      "loss": 0.7176,
+      "step": 1564
+    },
+    {
+      "epoch": 2.898148148148148,
+      "grad_norm": 1.4657547018707673,
+      "learning_rate": 1.9905621486949013e-07,
+      "loss": 0.7429,
+      "step": 1565
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 2.126039634312253,
+      "learning_rate": 1.9905415858090032e-07,
+      "loss": 0.7256,
+      "step": 1566
+    },
+    {
+      "epoch": 2.901851851851852,
+      "grad_norm": 1.3237019339778255,
+      "learning_rate": 1.9905210006530398e-07,
+      "loss": 0.7053,
+      "step": 1567
+    },
+    {
+      "epoch": 2.9037037037037035,
+      "grad_norm": 1.3647018670925397,
+      "learning_rate": 1.9905003932274746e-07,
+      "loss": 0.7146,
+      "step": 1568
+    },
+    {
+      "epoch": 2.9055555555555554,
+      "grad_norm": 1.458602935413366,
+      "learning_rate": 1.9904797635327702e-07,
+      "loss": 0.7084,
+      "step": 1569
+    },
+    {
+      "epoch": 2.9074074074074074,
+      "grad_norm": 1.3781877310261341,
+      "learning_rate": 1.9904591115693905e-07,
+      "loss": 0.7504,
+      "step": 1570
+    },
+    {
+      "epoch": 2.909259259259259,
+      "grad_norm": 2.221669754642774,
+      "learning_rate": 1.9904384373378e-07,
+      "loss": 0.7876,
+      "step": 1571
+    },
+    {
+      "epoch": 2.911111111111111,
+      "grad_norm": 1.3812586932654014,
+      "learning_rate": 1.9904177408384633e-07,
+      "loss": 0.7444,
+      "step": 1572
+    },
+    {
+      "epoch": 2.912962962962963,
+      "grad_norm": 1.6747140864461514,
+      "learning_rate": 1.990397022071846e-07,
+      "loss": 0.6335,
+      "step": 1573
+    },
+    {
+      "epoch": 2.914814814814815,
+      "grad_norm": 1.4317123943799357,
+      "learning_rate": 1.9903762810384138e-07,
+      "loss": 0.7324,
+      "step": 1574
+    },
+    {
+      "epoch": 2.9166666666666665,
+      "grad_norm": 2.0302283230660456,
+      "learning_rate": 1.990355517738633e-07,
+      "loss": 0.6766,
+      "step": 1575
+    },
+    {
+      "epoch": 2.9185185185185185,
+      "grad_norm": 1.5482722587521691,
+      "learning_rate": 1.9903347321729699e-07,
+      "loss": 0.7131,
+      "step": 1576
+    },
+    {
+      "epoch": 2.9203703703703705,
+      "grad_norm": 1.6357874902641074,
+      "learning_rate": 1.990313924341893e-07,
+      "loss": 0.6799,
+      "step": 1577
+    },
+    {
+      "epoch": 2.9222222222222225,
+      "grad_norm": 1.4018801800672824,
+      "learning_rate": 1.9902930942458692e-07,
+      "loss": 0.7061,
+      "step": 1578
+    },
+    {
+      "epoch": 2.924074074074074,
+      "grad_norm": 1.565864855508223,
+      "learning_rate": 1.990272241885367e-07,
+      "loss": 0.7363,
+      "step": 1579
+    },
+    {
+      "epoch": 2.925925925925926,
+      "grad_norm": 1.4209708460174861,
+      "learning_rate": 1.9902513672608552e-07,
+      "loss": 0.7049,
+      "step": 1580
+    },
+    {
+      "epoch": 2.927777777777778,
+      "grad_norm": 1.601211762325991,
+      "learning_rate": 1.9902304703728034e-07,
+      "loss": 0.6928,
+      "step": 1581
+    },
+    {
+      "epoch": 2.9296296296296296,
+      "grad_norm": 1.941811892938044,
+      "learning_rate": 1.9902095512216808e-07,
+      "loss": 0.7122,
+      "step": 1582
+    },
+    {
+      "epoch": 2.9314814814814816,
+      "grad_norm": 1.563309407291449,
+      "learning_rate": 1.9901886098079582e-07,
+      "loss": 0.7014,
+      "step": 1583
+    },
+    {
+      "epoch": 2.9333333333333336,
+      "grad_norm": 2.25580973584275,
+      "learning_rate": 1.9901676461321067e-07,
+      "loss": 0.7151,
+      "step": 1584
+    },
+    {
+      "epoch": 2.935185185185185,
+      "grad_norm": 1.3662637398853443,
+      "learning_rate": 1.990146660194597e-07,
+      "loss": 0.7043,
+      "step": 1585
+    },
+    {
+      "epoch": 2.937037037037037,
+      "grad_norm": 2.8500493066284487,
+      "learning_rate": 1.990125651995901e-07,
+      "loss": 0.7456,
+      "step": 1586
+    },
+    {
+      "epoch": 2.938888888888889,
+      "grad_norm": 1.4101988677031434,
+      "learning_rate": 1.9901046215364917e-07,
+      "loss": 0.74,
+      "step": 1587
+    },
+    {
+      "epoch": 2.9407407407407407,
+      "grad_norm": 1.544313036377568,
+      "learning_rate": 1.9900835688168408e-07,
+      "loss": 0.6936,
+      "step": 1588
+    },
+    {
+      "epoch": 2.9425925925925926,
+      "grad_norm": 1.4074380588460953,
+      "learning_rate": 1.9900624938374227e-07,
+      "loss": 0.658,
+      "step": 1589
+    },
+    {
+      "epoch": 2.9444444444444446,
+      "grad_norm": 1.3162581344440665,
+      "learning_rate": 1.990041396598711e-07,
+      "loss": 0.6656,
+      "step": 1590
+    },
+    {
+      "epoch": 2.946296296296296,
+      "grad_norm": 1.520048816301999,
+      "learning_rate": 1.9900202771011795e-07,
+      "loss": 0.7155,
+      "step": 1591
+    },
+    {
+      "epoch": 2.948148148148148,
+      "grad_norm": 1.4340900834721495,
+      "learning_rate": 1.989999135345303e-07,
+      "loss": 0.7333,
+      "step": 1592
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 11.286555927192733,
+      "learning_rate": 1.9899779713315575e-07,
+      "loss": 0.7405,
+      "step": 1593
+    },
+    {
+      "epoch": 2.9518518518518517,
+      "grad_norm": 1.5293603892505057,
+      "learning_rate": 1.9899567850604185e-07,
+      "loss": 0.7062,
+      "step": 1594
+    },
+    {
+      "epoch": 2.9537037037037037,
+      "grad_norm": 1.3140294351980084,
+      "learning_rate": 1.9899355765323623e-07,
+      "loss": 0.7083,
+      "step": 1595
+    },
+    {
+      "epoch": 2.9555555555555557,
+      "grad_norm": 1.3405231484968647,
+      "learning_rate": 1.9899143457478655e-07,
+      "loss": 0.6909,
+      "step": 1596
+    },
+    {
+      "epoch": 2.9574074074074073,
+      "grad_norm": 1.3672900881444934,
+      "learning_rate": 1.9898930927074055e-07,
+      "loss": 0.712,
+      "step": 1597
+    },
+    {
+      "epoch": 2.9592592592592593,
+      "grad_norm": 1.703842596725543,
+      "learning_rate": 1.9898718174114605e-07,
+      "loss": 0.7021,
+      "step": 1598
+    },
+    {
+      "epoch": 2.9611111111111112,
+      "grad_norm": 1.7375989794196072,
+      "learning_rate": 1.9898505198605088e-07,
+      "loss": 0.758,
+      "step": 1599
+    },
+    {
+      "epoch": 2.962962962962963,
+      "grad_norm": 1.6224928533143612,
+      "learning_rate": 1.9898292000550289e-07,
+      "loss": 0.7396,
+      "step": 1600
+    },
+    {
+      "epoch": 2.964814814814815,
+      "grad_norm": 1.6303459865409737,
+      "learning_rate": 1.9898078579955002e-07,
+      "loss": 0.6431,
+      "step": 1601
+    },
+    {
+      "epoch": 2.966666666666667,
+      "grad_norm": 1.654190860525658,
+      "learning_rate": 1.9897864936824024e-07,
+      "loss": 0.6561,
+      "step": 1602
+    },
+    {
+      "epoch": 2.9685185185185183,
+      "grad_norm": 1.788151948461757,
+      "learning_rate": 1.989765107116216e-07,
+      "loss": 0.7187,
+      "step": 1603
+    },
+    {
+      "epoch": 2.9703703703703703,
+      "grad_norm": 1.4738733364351084,
+      "learning_rate": 1.9897436982974223e-07,
+      "loss": 0.7513,
+      "step": 1604
+    },
+    {
+      "epoch": 2.9722222222222223,
+      "grad_norm": 1.4176979977823627,
+      "learning_rate": 1.9897222672265017e-07,
+      "loss": 0.7311,
+      "step": 1605
+    },
+    {
+      "epoch": 2.974074074074074,
+      "grad_norm": 1.2131457107880457,
+      "learning_rate": 1.9897008139039366e-07,
+      "loss": 0.6931,
+      "step": 1606
+    },
+    {
+      "epoch": 2.975925925925926,
+      "grad_norm": 1.4790144518991266,
+      "learning_rate": 1.9896793383302094e-07,
+      "loss": 0.6802,
+      "step": 1607
+    },
+    {
+      "epoch": 2.977777777777778,
+      "grad_norm": 1.7385149317862878,
+      "learning_rate": 1.9896578405058025e-07,
+      "loss": 0.7609,
+      "step": 1608
+    },
+    {
+      "epoch": 2.9796296296296294,
+      "grad_norm": 1.3892458405453747,
+      "learning_rate": 1.9896363204311996e-07,
+      "loss": 0.7373,
+      "step": 1609
+    },
+    {
+      "epoch": 2.9814814814814814,
+      "grad_norm": 1.3627744163643796,
+      "learning_rate": 1.9896147781068842e-07,
+      "loss": 0.7432,
+      "step": 1610
+    },
+    {
+      "epoch": 2.9833333333333334,
+      "grad_norm": 1.3074709448165318,
+      "learning_rate": 1.989593213533341e-07,
+      "loss": 0.6811,
+      "step": 1611
+    },
+    {
+      "epoch": 2.985185185185185,
+      "grad_norm": 1.448861395054024,
+      "learning_rate": 1.9895716267110548e-07,
+      "loss": 0.652,
+      "step": 1612
+    },
+    {
+      "epoch": 2.987037037037037,
+      "grad_norm": 1.2823434727948166,
+      "learning_rate": 1.9895500176405106e-07,
+      "loss": 0.6876,
+      "step": 1613
+    },
+    {
+      "epoch": 2.988888888888889,
+      "grad_norm": 1.2754660886216431,
+      "learning_rate": 1.9895283863221945e-07,
+      "loss": 0.697,
+      "step": 1614
+    },
+    {
+      "epoch": 2.9907407407407405,
+      "grad_norm": 1.9914142825761545,
+      "learning_rate": 1.9895067327565927e-07,
+      "loss": 0.74,
+      "step": 1615
+    },
+    {
+      "epoch": 2.9925925925925925,
+      "grad_norm": 1.505506782590237,
+      "learning_rate": 1.989485056944192e-07,
+      "loss": 0.6696,
+      "step": 1616
+    },
+    {
+      "epoch": 2.9944444444444445,
+      "grad_norm": 3.1005063199012457,
+      "learning_rate": 1.98946335888548e-07,
+      "loss": 0.7325,
+      "step": 1617
+    },
+    {
+      "epoch": 2.9962962962962965,
+      "grad_norm": 1.348441663216238,
+      "learning_rate": 1.9894416385809442e-07,
+      "loss": 0.7448,
+      "step": 1618
+    },
+    {
+      "epoch": 2.998148148148148,
+      "grad_norm": 2.1157934108696135,
+      "learning_rate": 1.9894198960310736e-07,
+      "loss": 0.7123,
+      "step": 1619
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.5400644539695503,
+      "learning_rate": 1.989398131236356e-07,
+      "loss": 0.7479,
+      "step": 1620
+    },
+    {
+      "epoch": 3.001851851851852,
+      "grad_norm": 1.3465290403646841,
+      "learning_rate": 1.9893763441972812e-07,
+      "loss": 0.7176,
+      "step": 1621
+    },
+    {
+      "epoch": 3.0037037037037035,
+      "grad_norm": 2.5116095724844363,
+      "learning_rate": 1.9893545349143393e-07,
+      "loss": 0.7012,
+      "step": 1622
+    },
+    {
+      "epoch": 3.0055555555555555,
+      "grad_norm": 1.3581760933892026,
+      "learning_rate": 1.9893327033880203e-07,
+      "loss": 0.6863,
+      "step": 1623
+    },
+    {
+      "epoch": 3.0074074074074075,
+      "grad_norm": 1.2994520920497343,
+      "learning_rate": 1.9893108496188155e-07,
+      "loss": 0.716,
+      "step": 1624
+    },
+    {
+      "epoch": 3.009259259259259,
+      "grad_norm": 1.3055433011569353,
+      "learning_rate": 1.9892889736072154e-07,
+      "loss": 0.6888,
+      "step": 1625
+    },
+    {
+      "epoch": 3.011111111111111,
+      "grad_norm": 1.2499442093256048,
+      "learning_rate": 1.9892670753537124e-07,
+      "loss": 0.6973,
+      "step": 1626
+    },
+    {
+      "epoch": 3.012962962962963,
+      "grad_norm": 1.3692448570675324,
+      "learning_rate": 1.9892451548587986e-07,
+      "loss": 0.68,
+      "step": 1627
+    },
+    {
+      "epoch": 3.0148148148148146,
+      "grad_norm": 5.254663109492828,
+      "learning_rate": 1.9892232121229672e-07,
+      "loss": 0.7396,
+      "step": 1628
+    },
+    {
+      "epoch": 3.0166666666666666,
+      "grad_norm": 1.4458957757007047,
+      "learning_rate": 1.9892012471467114e-07,
+      "loss": 0.7068,
+      "step": 1629
+    },
+    {
+      "epoch": 3.0185185185185186,
+      "grad_norm": 1.4568042426347534,
+      "learning_rate": 1.9891792599305245e-07,
+      "loss": 0.708,
+      "step": 1630
+    },
+    {
+      "epoch": 3.0203703703703706,
+      "grad_norm": 1.5898842089100773,
+      "learning_rate": 1.9891572504749016e-07,
+      "loss": 0.6825,
+      "step": 1631
+    },
+    {
+      "epoch": 3.022222222222222,
+      "grad_norm": 1.4093596408780869,
+      "learning_rate": 1.9891352187803373e-07,
+      "loss": 0.741,
+      "step": 1632
+    },
+    {
+      "epoch": 3.024074074074074,
+      "grad_norm": 1.4536944611496536,
+      "learning_rate": 1.9891131648473266e-07,
+      "loss": 0.752,
+      "step": 1633
+    },
+    {
+      "epoch": 3.025925925925926,
+      "grad_norm": 1.4395753321416864,
+      "learning_rate": 1.989091088676366e-07,
+      "loss": 0.6715,
+      "step": 1634
+    },
+    {
+      "epoch": 3.0277777777777777,
+      "grad_norm": 1.4961611160424322,
+      "learning_rate": 1.9890689902679507e-07,
+      "loss": 0.7228,
+      "step": 1635
+    },
+    {
+      "epoch": 3.0296296296296297,
+      "grad_norm": 1.4922576209506542,
+      "learning_rate": 1.9890468696225788e-07,
+      "loss": 0.7549,
+      "step": 1636
+    },
+    {
+      "epoch": 3.0314814814814817,
+      "grad_norm": 1.6955670865230463,
+      "learning_rate": 1.9890247267407468e-07,
+      "loss": 0.7276,
+      "step": 1637
+    },
+    {
+      "epoch": 3.033333333333333,
+      "grad_norm": 1.4550193857556193,
+      "learning_rate": 1.9890025616229526e-07,
+      "loss": 0.7127,
+      "step": 1638
+    },
+    {
+      "epoch": 3.035185185185185,
+      "grad_norm": 1.249933019312746,
+      "learning_rate": 1.9889803742696952e-07,
+      "loss": 0.6699,
+      "step": 1639
+    },
+    {
+      "epoch": 3.037037037037037,
+      "grad_norm": 1.5506750348497658,
+      "learning_rate": 1.9889581646814725e-07,
+      "loss": 0.7337,
+      "step": 1640
+    },
+    {
+      "epoch": 3.0388888888888888,
+      "grad_norm": 2.755592884413332,
+      "learning_rate": 1.9889359328587848e-07,
+      "loss": 0.7405,
+      "step": 1641
+    },
+    {
+      "epoch": 3.0407407407407407,
+      "grad_norm": 1.6330365000153635,
+      "learning_rate": 1.988913678802131e-07,
+      "loss": 0.73,
+      "step": 1642
+    },
+    {
+      "epoch": 3.0425925925925927,
+      "grad_norm": 1.5619368172557622,
+      "learning_rate": 1.988891402512012e-07,
+      "loss": 0.7563,
+      "step": 1643
+    },
+    {
+      "epoch": 3.0444444444444443,
+      "grad_norm": 1.4887327335824572,
+      "learning_rate": 1.9888691039889285e-07,
+      "loss": 0.7684,
+      "step": 1644
+    },
+    {
+      "epoch": 3.0462962962962963,
+      "grad_norm": 1.7746831756420167,
+      "learning_rate": 1.9888467832333817e-07,
+      "loss": 0.7693,
+      "step": 1645
+    },
+    {
+      "epoch": 3.0481481481481483,
+      "grad_norm": 1.4558601823973059,
+      "learning_rate": 1.9888244402458737e-07,
+      "loss": 0.6964,
+      "step": 1646
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 2.730373109051131,
+      "learning_rate": 1.9888020750269066e-07,
+      "loss": 0.6868,
+      "step": 1647
+    },
+    {
+      "epoch": 3.051851851851852,
+      "grad_norm": 1.3171058015474684,
+      "learning_rate": 1.9887796875769835e-07,
+      "loss": 0.6741,
+      "step": 1648
+    },
+    {
+      "epoch": 3.053703703703704,
+      "grad_norm": 1.6828570889766399,
+      "learning_rate": 1.988757277896607e-07,
+      "loss": 0.6984,
+      "step": 1649
+    },
+    {
+      "epoch": 3.0555555555555554,
+      "grad_norm": 1.7102743906255844,
+      "learning_rate": 1.988734845986282e-07,
+      "loss": 0.7288,
+      "step": 1650
+    },
+    {
+      "epoch": 3.0574074074074074,
+      "grad_norm": 1.4648711735671105,
+      "learning_rate": 1.988712391846512e-07,
+      "loss": 0.7536,
+      "step": 1651
+    },
+    {
+      "epoch": 3.0592592592592593,
+      "grad_norm": 2.7389505677907153,
+      "learning_rate": 1.9886899154778023e-07,
+      "loss": 0.7341,
+      "step": 1652
+    },
+    {
+      "epoch": 3.061111111111111,
+      "grad_norm": 1.3866068178133961,
+      "learning_rate": 1.988667416880658e-07,
+      "loss": 0.6619,
+      "step": 1653
+    },
+    {
+      "epoch": 3.062962962962963,
+      "grad_norm": 1.542935055861403,
+      "learning_rate": 1.9886448960555852e-07,
+      "loss": 0.7048,
+      "step": 1654
+    },
+    {
+      "epoch": 3.064814814814815,
+      "grad_norm": 1.5361354972717536,
+      "learning_rate": 1.9886223530030897e-07,
+      "loss": 0.7004,
+      "step": 1655
+    },
+    {
+      "epoch": 3.066666666666667,
+      "grad_norm": 1.6880469698301295,
+      "learning_rate": 1.9885997877236786e-07,
+      "loss": 0.7209,
+      "step": 1656
+    },
+    {
+      "epoch": 3.0685185185185184,
+      "grad_norm": 1.7682470421796426,
+      "learning_rate": 1.9885772002178594e-07,
+      "loss": 0.7509,
+      "step": 1657
+    },
+    {
+      "epoch": 3.0703703703703704,
+      "grad_norm": 2.0197081624474857,
+      "learning_rate": 1.9885545904861401e-07,
+      "loss": 0.7501,
+      "step": 1658
+    },
+    {
+      "epoch": 3.0722222222222224,
+      "grad_norm": 1.3349564461482262,
+      "learning_rate": 1.988531958529028e-07,
+      "loss": 0.7179,
+      "step": 1659
+    },
+    {
+      "epoch": 3.074074074074074,
+      "grad_norm": 1.6653903828205585,
+      "learning_rate": 1.9885093043470334e-07,
+      "loss": 0.7032,
+      "step": 1660
+    },
+    {
+      "epoch": 3.075925925925926,
+      "grad_norm": 1.7863874143951282,
+      "learning_rate": 1.9884866279406647e-07,
+      "loss": 0.7324,
+      "step": 1661
+    },
+    {
+      "epoch": 3.077777777777778,
+      "grad_norm": 1.5414246900775297,
+      "learning_rate": 1.9884639293104318e-07,
+      "loss": 0.7215,
+      "step": 1662
+    },
+    {
+      "epoch": 3.0796296296296295,
+      "grad_norm": 1.3500687796087663,
+      "learning_rate": 1.988441208456845e-07,
+      "loss": 0.7436,
+      "step": 1663
+    },
+    {
+      "epoch": 3.0814814814814815,
+      "grad_norm": 1.5677478826583047,
+      "learning_rate": 1.9884184653804157e-07,
+      "loss": 0.7004,
+      "step": 1664
+    },
+    {
+      "epoch": 3.0833333333333335,
+      "grad_norm": 1.4095622928721916,
+      "learning_rate": 1.9883957000816545e-07,
+      "loss": 0.6872,
+      "step": 1665
+    },
+    {
+      "epoch": 3.085185185185185,
+      "grad_norm": 1.5182424388915532,
+      "learning_rate": 1.9883729125610734e-07,
+      "loss": 0.7276,
+      "step": 1666
+    },
+    {
+      "epoch": 3.087037037037037,
+      "grad_norm": 1.4318940099260271,
+      "learning_rate": 1.988350102819185e-07,
+      "loss": 0.6919,
+      "step": 1667
+    },
+    {
+      "epoch": 3.088888888888889,
+      "grad_norm": 1.8798896662429445,
+      "learning_rate": 1.988327270856502e-07,
+      "loss": 0.713,
+      "step": 1668
+    },
+    {
+      "epoch": 3.0907407407407406,
+      "grad_norm": 1.7396830969265804,
+      "learning_rate": 1.9883044166735375e-07,
+      "loss": 0.7096,
+      "step": 1669
+    },
+    {
+      "epoch": 3.0925925925925926,
+      "grad_norm": 1.4675500179598062,
+      "learning_rate": 1.9882815402708058e-07,
+      "loss": 0.6673,
+      "step": 1670
+    },
+    {
+      "epoch": 3.0944444444444446,
+      "grad_norm": 3.908630997338015,
+      "learning_rate": 1.9882586416488208e-07,
+      "loss": 0.6757,
+      "step": 1671
+    },
+    {
+      "epoch": 3.096296296296296,
+      "grad_norm": 1.5773921026996343,
+      "learning_rate": 1.9882357208080977e-07,
+      "loss": 0.7608,
+      "step": 1672
+    },
+    {
+      "epoch": 3.098148148148148,
+      "grad_norm": 1.5807244741734738,
+      "learning_rate": 1.9882127777491514e-07,
+      "loss": 0.7144,
+      "step": 1673
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 1.8922916047260412,
+      "learning_rate": 1.9881898124724978e-07,
+      "loss": 0.6948,
+      "step": 1674
+    },
+    {
+      "epoch": 3.1018518518518516,
+      "grad_norm": 1.3615325912481913,
+      "learning_rate": 1.9881668249786533e-07,
+      "loss": 0.751,
+      "step": 1675
+    },
+    {
+      "epoch": 3.1037037037037036,
+      "grad_norm": 1.4705548481990303,
+      "learning_rate": 1.988143815268135e-07,
+      "loss": 0.6496,
+      "step": 1676
+    },
+    {
+      "epoch": 3.1055555555555556,
+      "grad_norm": 2.33415194489758,
+      "learning_rate": 1.9881207833414599e-07,
+      "loss": 0.7061,
+      "step": 1677
+    },
+    {
+      "epoch": 3.1074074074074076,
+      "grad_norm": 2.5969211621662427,
+      "learning_rate": 1.9880977291991456e-07,
+      "loss": 0.688,
+      "step": 1678
+    },
+    {
+      "epoch": 3.109259259259259,
+      "grad_norm": 1.5417996015192457,
+      "learning_rate": 1.9880746528417112e-07,
+      "loss": 0.6758,
+      "step": 1679
+    },
+    {
+      "epoch": 3.111111111111111,
+      "grad_norm": 3.269981042236833,
+      "learning_rate": 1.9880515542696746e-07,
+      "loss": 0.7392,
+      "step": 1680
+    },
+    {
+      "epoch": 3.112962962962963,
+      "grad_norm": 1.4054886995259201,
+      "learning_rate": 1.988028433483556e-07,
+      "loss": 0.6951,
+      "step": 1681
+    },
+    {
+      "epoch": 3.1148148148148147,
+      "grad_norm": 2.15587450450883,
+      "learning_rate": 1.9880052904838744e-07,
+      "loss": 0.672,
+      "step": 1682
+    },
+    {
+      "epoch": 3.1166666666666667,
+      "grad_norm": 1.5399876800245342,
+      "learning_rate": 1.9879821252711507e-07,
+      "loss": 0.6865,
+      "step": 1683
+    },
+    {
+      "epoch": 3.1185185185185187,
+      "grad_norm": 1.4940979279637496,
+      "learning_rate": 1.987958937845905e-07,
+      "loss": 0.7134,
+      "step": 1684
+    },
+    {
+      "epoch": 3.1203703703703702,
+      "grad_norm": 1.6600346744212056,
+      "learning_rate": 1.9879357282086596e-07,
+      "loss": 0.7183,
+      "step": 1685
+    },
+    {
+      "epoch": 3.1222222222222222,
+      "grad_norm": 1.4661779934176216,
+      "learning_rate": 1.987912496359936e-07,
+      "loss": 0.6695,
+      "step": 1686
+    },
+    {
+      "epoch": 3.1240740740740742,
+      "grad_norm": 1.4292786918613043,
+      "learning_rate": 1.9878892423002562e-07,
+      "loss": 0.7261,
+      "step": 1687
+    },
+    {
+      "epoch": 3.1259259259259258,
+      "grad_norm": 1.5135291921133704,
+      "learning_rate": 1.987865966030143e-07,
+      "loss": 0.7349,
+      "step": 1688
+    },
+    {
+      "epoch": 3.1277777777777778,
+      "grad_norm": 1.465639785609288,
+      "learning_rate": 1.98784266755012e-07,
+      "loss": 0.6916,
+      "step": 1689
+    },
+    {
+      "epoch": 3.1296296296296298,
+      "grad_norm": 1.292903294793804,
+      "learning_rate": 1.987819346860711e-07,
+      "loss": 0.6983,
+      "step": 1690
+    },
+    {
+      "epoch": 3.1314814814814813,
+      "grad_norm": 1.5769108742797056,
+      "learning_rate": 1.98779600396244e-07,
+      "loss": 0.7236,
+      "step": 1691
+    },
+    {
+      "epoch": 3.1333333333333333,
+      "grad_norm": 1.439158811004055,
+      "learning_rate": 1.9877726388558322e-07,
+      "loss": 0.708,
+      "step": 1692
+    },
+    {
+      "epoch": 3.1351851851851853,
+      "grad_norm": 2.7838377971471076,
+      "learning_rate": 1.9877492515414125e-07,
+      "loss": 0.6605,
+      "step": 1693
+    },
+    {
+      "epoch": 3.137037037037037,
+      "grad_norm": 1.770285195097917,
+      "learning_rate": 1.987725842019707e-07,
+      "loss": 0.7156,
+      "step": 1694
+    },
+    {
+      "epoch": 3.138888888888889,
+      "grad_norm": 1.7884590387188937,
+      "learning_rate": 1.987702410291242e-07,
+      "loss": 0.7023,
+      "step": 1695
+    },
+    {
+      "epoch": 3.140740740740741,
+      "grad_norm": 1.3955629912551473,
+      "learning_rate": 1.9876789563565443e-07,
+      "loss": 0.7182,
+      "step": 1696
+    },
+    {
+      "epoch": 3.1425925925925924,
+      "grad_norm": 1.4431664841162344,
+      "learning_rate": 1.987655480216141e-07,
+      "loss": 0.6731,
+      "step": 1697
+    },
+    {
+      "epoch": 3.1444444444444444,
+      "grad_norm": 1.3261493602966676,
+      "learning_rate": 1.9876319818705602e-07,
+      "loss": 0.646,
+      "step": 1698
+    },
+    {
+      "epoch": 3.1462962962962964,
+      "grad_norm": 1.3959952532044104,
+      "learning_rate": 1.9876084613203302e-07,
+      "loss": 0.6796,
+      "step": 1699
+    },
+    {
+      "epoch": 3.148148148148148,
+      "grad_norm": 2.031273973137189,
+      "learning_rate": 1.9875849185659795e-07,
+      "loss": 0.6788,
+      "step": 1700
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 1.5551745296788908,
+      "learning_rate": 1.9875613536080377e-07,
+      "loss": 0.7309,
+      "step": 1701
+    },
+    {
+      "epoch": 3.151851851851852,
+      "grad_norm": 1.6289415891619674,
+      "learning_rate": 1.9875377664470345e-07,
+      "loss": 0.7677,
+      "step": 1702
+    },
+    {
+      "epoch": 3.153703703703704,
+      "grad_norm": 1.4390675140572768,
+      "learning_rate": 1.9875141570835004e-07,
+      "loss": 0.7382,
+      "step": 1703
+    },
+    {
+      "epoch": 3.1555555555555554,
+      "grad_norm": 1.372965455625381,
+      "learning_rate": 1.9874905255179657e-07,
+      "loss": 0.7105,
+      "step": 1704
+    },
+    {
+      "epoch": 3.1574074074074074,
+      "grad_norm": 1.6015794601427793,
+      "learning_rate": 1.987466871750962e-07,
+      "loss": 0.6916,
+      "step": 1705
+    },
+    {
+      "epoch": 3.1592592592592594,
+      "grad_norm": 1.5890351606935418,
+      "learning_rate": 1.987443195783021e-07,
+      "loss": 0.7614,
+      "step": 1706
+    },
+    {
+      "epoch": 3.161111111111111,
+      "grad_norm": 1.5303485579252196,
+      "learning_rate": 1.9874194976146755e-07,
+      "loss": 0.6869,
+      "step": 1707
+    },
+    {
+      "epoch": 3.162962962962963,
+      "grad_norm": 1.5551718907204275,
+      "learning_rate": 1.9873957772464576e-07,
+      "loss": 0.6605,
+      "step": 1708
+    },
+    {
+      "epoch": 3.164814814814815,
+      "grad_norm": 1.4360847819466684,
+      "learning_rate": 1.9873720346789008e-07,
+      "loss": 0.738,
+      "step": 1709
+    },
+    {
+      "epoch": 3.1666666666666665,
+      "grad_norm": 1.4184993620680044,
+      "learning_rate": 1.9873482699125393e-07,
+      "loss": 0.6959,
+      "step": 1710
+    },
+    {
+      "epoch": 3.1685185185185185,
+      "grad_norm": 1.9224815358750618,
+      "learning_rate": 1.987324482947907e-07,
+      "loss": 0.74,
+      "step": 1711
+    },
+    {
+      "epoch": 3.1703703703703705,
+      "grad_norm": 1.5359644124916254,
+      "learning_rate": 1.9873006737855386e-07,
+      "loss": 0.6937,
+      "step": 1712
+    },
+    {
+      "epoch": 3.172222222222222,
+      "grad_norm": 1.6687035934098722,
+      "learning_rate": 1.9872768424259698e-07,
+      "loss": 0.7322,
+      "step": 1713
+    },
+    {
+      "epoch": 3.174074074074074,
+      "grad_norm": 1.7354524302271312,
+      "learning_rate": 1.987252988869736e-07,
+      "loss": 0.6876,
+      "step": 1714
+    },
+    {
+      "epoch": 3.175925925925926,
+      "grad_norm": 1.8078074991161097,
+      "learning_rate": 1.987229113117374e-07,
+      "loss": 0.7763,
+      "step": 1715
+    },
+    {
+      "epoch": 3.1777777777777776,
+      "grad_norm": 1.5431668606476954,
+      "learning_rate": 1.9872052151694202e-07,
+      "loss": 0.7032,
+      "step": 1716
+    },
+    {
+      "epoch": 3.1796296296296296,
+      "grad_norm": 2.0257639254774085,
+      "learning_rate": 1.9871812950264117e-07,
+      "loss": 0.69,
+      "step": 1717
+    },
+    {
+      "epoch": 3.1814814814814816,
+      "grad_norm": 1.7816118494622915,
+      "learning_rate": 1.9871573526888869e-07,
+      "loss": 0.6912,
+      "step": 1718
+    },
+    {
+      "epoch": 3.183333333333333,
+      "grad_norm": 1.367941302477703,
+      "learning_rate": 1.987133388157383e-07,
+      "loss": 0.6725,
+      "step": 1719
+    },
+    {
+      "epoch": 3.185185185185185,
+      "grad_norm": 1.9654016251306183,
+      "learning_rate": 1.9871094014324402e-07,
+      "loss": 0.6604,
+      "step": 1720
+    },
+    {
+      "epoch": 3.187037037037037,
+      "grad_norm": 1.548416789303848,
+      "learning_rate": 1.9870853925145968e-07,
+      "loss": 0.7079,
+      "step": 1721
+    },
+    {
+      "epoch": 3.188888888888889,
+      "grad_norm": 1.3323893288004114,
+      "learning_rate": 1.9870613614043933e-07,
+      "loss": 0.6793,
+      "step": 1722
+    },
+    {
+      "epoch": 3.1907407407407407,
+      "grad_norm": 1.3496211227864021,
+      "learning_rate": 1.9870373081023691e-07,
+      "loss": 0.7095,
+      "step": 1723
+    },
+    {
+      "epoch": 3.1925925925925926,
+      "grad_norm": 1.5871603174102669,
+      "learning_rate": 1.9870132326090656e-07,
+      "loss": 0.7386,
+      "step": 1724
+    },
+    {
+      "epoch": 3.1944444444444446,
+      "grad_norm": 1.4320257269335934,
+      "learning_rate": 1.986989134925024e-07,
+      "loss": 0.7216,
+      "step": 1725
+    },
+    {
+      "epoch": 3.196296296296296,
+      "grad_norm": 1.5076149315040464,
+      "learning_rate": 1.9869650150507859e-07,
+      "loss": 0.7277,
+      "step": 1726
+    },
+    {
+      "epoch": 3.198148148148148,
+      "grad_norm": 1.4735141461354706,
+      "learning_rate": 1.986940872986894e-07,
+      "loss": 0.7014,
+      "step": 1727
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 1.4614415608173328,
+      "learning_rate": 1.9869167087338905e-07,
+      "loss": 0.6927,
+      "step": 1728
+    },
+    {
+      "epoch": 3.2018518518518517,
+      "grad_norm": 1.3521199024968569,
+      "learning_rate": 1.9868925222923192e-07,
+      "loss": 0.6945,
+      "step": 1729
+    },
+    {
+      "epoch": 3.2037037037037037,
+      "grad_norm": 1.3345681083578147,
+      "learning_rate": 1.9868683136627235e-07,
+      "loss": 0.7484,
+      "step": 1730
+    },
+    {
+      "epoch": 3.2055555555555557,
+      "grad_norm": 1.4216257424639744,
+      "learning_rate": 1.9868440828456477e-07,
+      "loss": 0.718,
+      "step": 1731
+    },
+    {
+      "epoch": 3.2074074074074073,
+      "grad_norm": 1.5742081718620373,
+      "learning_rate": 1.9868198298416368e-07,
+      "loss": 0.7128,
+      "step": 1732
+    },
+    {
+      "epoch": 3.2092592592592593,
+      "grad_norm": 1.660929669372897,
+      "learning_rate": 1.986795554651236e-07,
+      "loss": 0.7197,
+      "step": 1733
+    },
+    {
+      "epoch": 3.2111111111111112,
+      "grad_norm": 1.418589702033337,
+      "learning_rate": 1.986771257274991e-07,
+      "loss": 0.6862,
+      "step": 1734
+    },
+    {
+      "epoch": 3.212962962962963,
+      "grad_norm": 1.4879585482571616,
+      "learning_rate": 1.9867469377134483e-07,
+      "loss": 0.7015,
+      "step": 1735
+    },
+    {
+      "epoch": 3.214814814814815,
+      "grad_norm": 1.4155096043374973,
+      "learning_rate": 1.986722595967154e-07,
+      "loss": 0.7119,
+      "step": 1736
+    },
+    {
+      "epoch": 3.216666666666667,
+      "grad_norm": 1.6719920845608907,
+      "learning_rate": 1.9866982320366564e-07,
+      "loss": 0.6801,
+      "step": 1737
+    },
+    {
+      "epoch": 3.2185185185185183,
+      "grad_norm": 1.917493852131688,
+      "learning_rate": 1.9866738459225023e-07,
+      "loss": 0.693,
+      "step": 1738
+    },
+    {
+      "epoch": 3.2203703703703703,
+      "grad_norm": 1.4048596837328573,
+      "learning_rate": 1.9866494376252406e-07,
+      "loss": 0.6552,
+      "step": 1739
+    },
+    {
+      "epoch": 3.2222222222222223,
+      "grad_norm": 1.3107955464321916,
+      "learning_rate": 1.9866250071454198e-07,
+      "loss": 0.7368,
+      "step": 1740
+    },
+    {
+      "epoch": 3.224074074074074,
+      "grad_norm": 1.6415296097464296,
+      "learning_rate": 1.9866005544835894e-07,
+      "loss": 0.6746,
+      "step": 1741
+    },
+    {
+      "epoch": 3.225925925925926,
+      "grad_norm": 1.3668511561598151,
+      "learning_rate": 1.9865760796402986e-07,
+      "loss": 0.7314,
+      "step": 1742
+    },
+    {
+      "epoch": 3.227777777777778,
+      "grad_norm": 1.278809798682975,
+      "learning_rate": 1.9865515826160982e-07,
+      "loss": 0.6867,
+      "step": 1743
+    },
+    {
+      "epoch": 3.2296296296296294,
+      "grad_norm": 1.3961839263779474,
+      "learning_rate": 1.9865270634115387e-07,
+      "loss": 0.7131,
+      "step": 1744
+    },
+    {
+      "epoch": 3.2314814814814814,
+      "grad_norm": 1.4091498954729598,
+      "learning_rate": 1.9865025220271715e-07,
+      "loss": 0.6982,
+      "step": 1745
+    },
+    {
+      "epoch": 3.2333333333333334,
+      "grad_norm": 1.5630621979796469,
+      "learning_rate": 1.9864779584635484e-07,
+      "loss": 0.7226,
+      "step": 1746
+    },
+    {
+      "epoch": 3.2351851851851854,
+      "grad_norm": 1.7604298806648988,
+      "learning_rate": 1.9864533727212214e-07,
+      "loss": 0.7291,
+      "step": 1747
+    },
+    {
+      "epoch": 3.237037037037037,
+      "grad_norm": 1.3320347511429715,
+      "learning_rate": 1.9864287648007436e-07,
+      "loss": 0.714,
+      "step": 1748
+    },
+    {
+      "epoch": 3.238888888888889,
+      "grad_norm": 1.6794335700204426,
+      "learning_rate": 1.986404134702668e-07,
+      "loss": 0.6914,
+      "step": 1749
+    },
+    {
+      "epoch": 3.240740740740741,
+      "grad_norm": 3.0510458140298247,
+      "learning_rate": 1.9863794824275483e-07,
+      "loss": 0.7281,
+      "step": 1750
+    },
+    {
+      "epoch": 3.2425925925925925,
+      "grad_norm": 1.3230940795455612,
+      "learning_rate": 1.9863548079759388e-07,
+      "loss": 0.7322,
+      "step": 1751
+    },
+    {
+      "epoch": 3.2444444444444445,
+      "grad_norm": 1.4313794759766358,
+      "learning_rate": 1.9863301113483946e-07,
+      "loss": 0.7064,
+      "step": 1752
+    },
+    {
+      "epoch": 3.2462962962962965,
+      "grad_norm": 1.4628379933589932,
+      "learning_rate": 1.9863053925454705e-07,
+      "loss": 0.6756,
+      "step": 1753
+    },
+    {
+      "epoch": 3.248148148148148,
+      "grad_norm": 1.406276310367317,
+      "learning_rate": 1.9862806515677227e-07,
+      "loss": 0.7186,
+      "step": 1754
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 1.4343961238620486,
+      "learning_rate": 1.9862558884157065e-07,
+      "loss": 0.6865,
+      "step": 1755
+    },
+    {
+      "epoch": 3.251851851851852,
+      "grad_norm": 2.366793024973306,
+      "learning_rate": 1.9862311030899798e-07,
+      "loss": 0.6603,
+      "step": 1756
+    },
+    {
+      "epoch": 3.2537037037037035,
+      "grad_norm": 1.6388609763325115,
+      "learning_rate": 1.9862062955910992e-07,
+      "loss": 0.6923,
+      "step": 1757
+    },
+    {
+      "epoch": 3.2555555555555555,
+      "grad_norm": 2.0986713998445925,
+      "learning_rate": 1.9861814659196228e-07,
+      "loss": 0.7106,
+      "step": 1758
+    },
+    {
+      "epoch": 3.2574074074074075,
+      "grad_norm": 1.2530713132208617,
+      "learning_rate": 1.9861566140761083e-07,
+      "loss": 0.666,
+      "step": 1759
+    },
+    {
+      "epoch": 3.259259259259259,
+      "grad_norm": 1.7827017227512207,
+      "learning_rate": 1.9861317400611148e-07,
+      "loss": 0.7117,
+      "step": 1760
+    },
+    {
+      "epoch": 3.261111111111111,
+      "grad_norm": 1.6668804613145347,
+      "learning_rate": 1.986106843875202e-07,
+      "loss": 0.6844,
+      "step": 1761
+    },
+    {
+      "epoch": 3.262962962962963,
+      "grad_norm": 2.433122523000563,
+      "learning_rate": 1.9860819255189285e-07,
+      "loss": 0.6974,
+      "step": 1762
+    },
+    {
+      "epoch": 3.264814814814815,
+      "grad_norm": 1.5368541544360192,
+      "learning_rate": 1.9860569849928552e-07,
+      "loss": 0.7131,
+      "step": 1763
+    },
+    {
+      "epoch": 3.2666666666666666,
+      "grad_norm": 1.5907604703887104,
+      "learning_rate": 1.986032022297543e-07,
+      "loss": 0.6813,
+      "step": 1764
+    },
+    {
+      "epoch": 3.2685185185185186,
+      "grad_norm": 2.054360580524189,
+      "learning_rate": 1.986007037433553e-07,
+      "loss": 0.718,
+      "step": 1765
+    },
+    {
+      "epoch": 3.2703703703703706,
+      "grad_norm": 1.9050273075340438,
+      "learning_rate": 1.9859820304014466e-07,
+      "loss": 0.7575,
+      "step": 1766
+    },
+    {
+      "epoch": 3.272222222222222,
+      "grad_norm": 1.6060824521122723,
+      "learning_rate": 1.9859570012017866e-07,
+      "loss": 0.6988,
+      "step": 1767
+    },
+    {
+      "epoch": 3.274074074074074,
+      "grad_norm": 1.4382538556506632,
+      "learning_rate": 1.985931949835135e-07,
+      "loss": 0.7233,
+      "step": 1768
+    },
+    {
+      "epoch": 3.275925925925926,
+      "grad_norm": 1.7742874379466718,
+      "learning_rate": 1.9859068763020557e-07,
+      "loss": 0.6661,
+      "step": 1769
+    },
+    {
+      "epoch": 3.2777777777777777,
+      "grad_norm": 1.6377630137693746,
+      "learning_rate": 1.985881780603112e-07,
+      "loss": 0.7553,
+      "step": 1770
+    },
+    {
+      "epoch": 3.2796296296296297,
+      "grad_norm": 1.5376859283539732,
+      "learning_rate": 1.9858566627388681e-07,
+      "loss": 0.6796,
+      "step": 1771
+    },
+    {
+      "epoch": 3.2814814814814817,
+      "grad_norm": 1.5021536429969475,
+      "learning_rate": 1.9858315227098892e-07,
+      "loss": 0.7195,
+      "step": 1772
+    },
+    {
+      "epoch": 3.283333333333333,
+      "grad_norm": 1.4369446812207705,
+      "learning_rate": 1.98580636051674e-07,
+      "loss": 0.7122,
+      "step": 1773
+    },
+    {
+      "epoch": 3.285185185185185,
+      "grad_norm": 1.6431245894661268,
+      "learning_rate": 1.9857811761599864e-07,
+      "loss": 0.7535,
+      "step": 1774
+    },
+    {
+      "epoch": 3.287037037037037,
+      "grad_norm": 1.4250330847004862,
+      "learning_rate": 1.9857559696401948e-07,
+      "loss": 0.6714,
+      "step": 1775
+    },
+    {
+      "epoch": 3.2888888888888888,
+      "grad_norm": 1.722137182459568,
+      "learning_rate": 1.9857307409579317e-07,
+      "loss": 0.7529,
+      "step": 1776
+    },
+    {
+      "epoch": 3.2907407407407407,
+      "grad_norm": 1.47834174654245,
+      "learning_rate": 1.985705490113764e-07,
+      "loss": 0.6833,
+      "step": 1777
+    },
+    {
+      "epoch": 3.2925925925925927,
+      "grad_norm": 1.6717925145228725,
+      "learning_rate": 1.98568021710826e-07,
+      "loss": 0.6778,
+      "step": 1778
+    },
+    {
+      "epoch": 3.2944444444444443,
+      "grad_norm": 1.3389523840169226,
+      "learning_rate": 1.9856549219419878e-07,
+      "loss": 0.7261,
+      "step": 1779
+    },
+    {
+      "epoch": 3.2962962962962963,
+      "grad_norm": 1.508202190548861,
+      "learning_rate": 1.9856296046155158e-07,
+      "loss": 0.6902,
+      "step": 1780
+    },
+    {
+      "epoch": 3.2981481481481483,
+      "grad_norm": 1.3055339718961279,
+      "learning_rate": 1.9856042651294132e-07,
+      "loss": 0.7292,
+      "step": 1781
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 1.3784260823512247,
+      "learning_rate": 1.9855789034842504e-07,
+      "loss": 0.69,
+      "step": 1782
+    },
+    {
+      "epoch": 3.301851851851852,
+      "grad_norm": 1.4540692938786228,
+      "learning_rate": 1.9855535196805966e-07,
+      "loss": 0.662,
+      "step": 1783
+    },
+    {
+      "epoch": 3.303703703703704,
+      "grad_norm": 1.4179602530351605,
+      "learning_rate": 1.985528113719023e-07,
+      "loss": 0.7272,
+      "step": 1784
+    },
+    {
+      "epoch": 3.3055555555555554,
+      "grad_norm": 1.431388432484835,
+      "learning_rate": 1.9855026856001007e-07,
+      "loss": 0.6861,
+      "step": 1785
+    },
+    {
+      "epoch": 3.3074074074074074,
+      "grad_norm": 1.474845184737349,
+      "learning_rate": 1.9854772353244016e-07,
+      "loss": 0.7601,
+      "step": 1786
+    },
+    {
+      "epoch": 3.3092592592592593,
+      "grad_norm": 1.4115749619499567,
+      "learning_rate": 1.9854517628924977e-07,
+      "loss": 0.6965,
+      "step": 1787
+    },
+    {
+      "epoch": 3.311111111111111,
+      "grad_norm": 2.031703428532122,
+      "learning_rate": 1.9854262683049616e-07,
+      "loss": 0.7051,
+      "step": 1788
+    },
+    {
+      "epoch": 3.312962962962963,
+      "grad_norm": 1.4210085566074337,
+      "learning_rate": 1.9854007515623667e-07,
+      "loss": 0.6889,
+      "step": 1789
+    },
+    {
+      "epoch": 3.314814814814815,
+      "grad_norm": 1.4557990808669314,
+      "learning_rate": 1.9853752126652864e-07,
+      "loss": 0.7347,
+      "step": 1790
+    },
+    {
+      "epoch": 3.3166666666666664,
+      "grad_norm": 1.367150528394596,
+      "learning_rate": 1.9853496516142955e-07,
+      "loss": 0.7006,
+      "step": 1791
+    },
+    {
+      "epoch": 3.3185185185185184,
+      "grad_norm": 2.470382011775795,
+      "learning_rate": 1.9853240684099682e-07,
+      "loss": 0.7246,
+      "step": 1792
+    },
+    {
+      "epoch": 3.3203703703703704,
+      "grad_norm": 1.4075732692730825,
+      "learning_rate": 1.9852984630528792e-07,
+      "loss": 0.6332,
+      "step": 1793
+    },
+    {
+      "epoch": 3.3222222222222224,
+      "grad_norm": 1.377107831679937,
+      "learning_rate": 1.9852728355436052e-07,
+      "loss": 0.7314,
+      "step": 1794
+    },
+    {
+      "epoch": 3.324074074074074,
+      "grad_norm": 1.5179482731981948,
+      "learning_rate": 1.9852471858827219e-07,
+      "loss": 0.6982,
+      "step": 1795
+    },
+    {
+      "epoch": 3.325925925925926,
+      "grad_norm": 2.01220396343066,
+      "learning_rate": 1.985221514070806e-07,
+      "loss": 0.7429,
+      "step": 1796
+    },
+    {
+      "epoch": 3.327777777777778,
+      "grad_norm": 1.3123577076162922,
+      "learning_rate": 1.9851958201084345e-07,
+      "loss": 0.6921,
+      "step": 1797
+    },
+    {
+      "epoch": 3.3296296296296295,
+      "grad_norm": 1.4965491778271642,
+      "learning_rate": 1.985170103996185e-07,
+      "loss": 0.7382,
+      "step": 1798
+    },
+    {
+      "epoch": 3.3314814814814815,
+      "grad_norm": 1.4850693258187,
+      "learning_rate": 1.9851443657346364e-07,
+      "loss": 0.7577,
+      "step": 1799
+    },
+    {
+      "epoch": 3.3333333333333335,
+      "grad_norm": 1.7904392962818572,
+      "learning_rate": 1.9851186053243664e-07,
+      "loss": 0.6895,
+      "step": 1800
+    },
+    {
+      "epoch": 3.335185185185185,
+      "grad_norm": 2.8057665234037055,
+      "learning_rate": 1.9850928227659547e-07,
+      "loss": 0.6865,
+      "step": 1801
+    },
+    {
+      "epoch": 3.337037037037037,
+      "grad_norm": 1.371401778740363,
+      "learning_rate": 1.9850670180599808e-07,
+      "loss": 0.7548,
+      "step": 1802
+    },
+    {
+      "epoch": 3.338888888888889,
+      "grad_norm": 2.215297675271137,
+      "learning_rate": 1.985041191207025e-07,
+      "loss": 0.6867,
+      "step": 1803
+    },
+    {
+      "epoch": 3.3407407407407406,
+      "grad_norm": 1.6420758470306913,
+      "learning_rate": 1.9850153422076682e-07,
+      "loss": 0.6951,
+      "step": 1804
+    },
+    {
+      "epoch": 3.3425925925925926,
+      "grad_norm": 1.4527864639750439,
+      "learning_rate": 1.9849894710624912e-07,
+      "loss": 0.7336,
+      "step": 1805
+    },
+    {
+      "epoch": 3.3444444444444446,
+      "grad_norm": 1.3217069089705615,
+      "learning_rate": 1.9849635777720755e-07,
+      "loss": 0.6662,
+      "step": 1806
+    },
+    {
+      "epoch": 3.346296296296296,
+      "grad_norm": 2.067910615618579,
+      "learning_rate": 1.9849376623370034e-07,
+      "loss": 0.6142,
+      "step": 1807
+    },
+    {
+      "epoch": 3.348148148148148,
+      "grad_norm": 1.6148500108855797,
+      "learning_rate": 1.9849117247578575e-07,
+      "loss": 0.7426,
+      "step": 1808
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 1.5223554238171622,
+      "learning_rate": 1.9848857650352212e-07,
+      "loss": 0.7149,
+      "step": 1809
+    },
+    {
+      "epoch": 3.351851851851852,
+      "grad_norm": 1.5893635024278807,
+      "learning_rate": 1.984859783169678e-07,
+      "loss": 0.6712,
+      "step": 1810
+    },
+    {
+      "epoch": 3.3537037037037036,
+      "grad_norm": 2.0638628133874355,
+      "learning_rate": 1.984833779161812e-07,
+      "loss": 0.713,
+      "step": 1811
+    },
+    {
+      "epoch": 3.3555555555555556,
+      "grad_norm": 1.3625841468778455,
+      "learning_rate": 1.984807753012208e-07,
+      "loss": 0.7031,
+      "step": 1812
+    },
+    {
+      "epoch": 3.3574074074074076,
+      "grad_norm": 1.4926336279072059,
+      "learning_rate": 1.9847817047214508e-07,
+      "loss": 0.7021,
+      "step": 1813
+    },
+    {
+      "epoch": 3.359259259259259,
+      "grad_norm": 1.4858279133201107,
+      "learning_rate": 1.9847556342901265e-07,
+      "loss": 0.6981,
+      "step": 1814
+    },
+    {
+      "epoch": 3.361111111111111,
+      "grad_norm": 1.3795031843352072,
+      "learning_rate": 1.9847295417188207e-07,
+      "loss": 0.741,
+      "step": 1815
+    },
+    {
+      "epoch": 3.362962962962963,
+      "grad_norm": 2.2109787933302876,
+      "learning_rate": 1.9847034270081206e-07,
+      "loss": 0.6689,
+      "step": 1816
+    },
+    {
+      "epoch": 3.3648148148148147,
+      "grad_norm": 1.8045167707521255,
+      "learning_rate": 1.984677290158613e-07,
+      "loss": 0.7523,
+      "step": 1817
+    },
+    {
+      "epoch": 3.3666666666666667,
+      "grad_norm": 1.4517747645748975,
+      "learning_rate": 1.9846511311708855e-07,
+      "loss": 0.7491,
+      "step": 1818
+    },
+    {
+      "epoch": 3.3685185185185187,
+      "grad_norm": 1.2931889574317914,
+      "learning_rate": 1.9846249500455263e-07,
+      "loss": 0.6776,
+      "step": 1819
+    },
+    {
+      "epoch": 3.3703703703703702,
+      "grad_norm": 1.4044752792037491,
+      "learning_rate": 1.984598746783124e-07,
+      "loss": 0.6783,
+      "step": 1820
+    },
+    {
+      "epoch": 3.3722222222222222,
+      "grad_norm": 1.6478546800324836,
+      "learning_rate": 1.9845725213842676e-07,
+      "loss": 0.7151,
+      "step": 1821
+    },
+    {
+      "epoch": 3.3740740740740742,
+      "grad_norm": 1.8568294350567034,
+      "learning_rate": 1.9845462738495473e-07,
+      "loss": 0.7111,
+      "step": 1822
+    },
+    {
+      "epoch": 3.3759259259259258,
+      "grad_norm": 1.4979035309359159,
+      "learning_rate": 1.9845200041795523e-07,
+      "loss": 0.6891,
+      "step": 1823
+    },
+    {
+      "epoch": 3.3777777777777778,
+      "grad_norm": 1.4950173832590299,
+      "learning_rate": 1.984493712374874e-07,
+      "loss": 0.6792,
+      "step": 1824
+    },
+    {
+      "epoch": 3.3796296296296298,
+      "grad_norm": 1.4108846458026787,
+      "learning_rate": 1.984467398436103e-07,
+      "loss": 0.778,
+      "step": 1825
+    },
+    {
+      "epoch": 3.3814814814814813,
+      "grad_norm": 1.5839148313364935,
+      "learning_rate": 1.984441062363831e-07,
+      "loss": 0.703,
+      "step": 1826
+    },
+    {
+      "epoch": 3.3833333333333333,
+      "grad_norm": 1.4385794392575448,
+      "learning_rate": 1.9844147041586507e-07,
+      "loss": 0.7153,
+      "step": 1827
+    },
+    {
+      "epoch": 3.3851851851851853,
+      "grad_norm": 1.5272095334171423,
+      "learning_rate": 1.9843883238211538e-07,
+      "loss": 0.7068,
+      "step": 1828
+    },
+    {
+      "epoch": 3.387037037037037,
+      "grad_norm": 2.6314228093941163,
+      "learning_rate": 1.984361921351934e-07,
+      "loss": 0.7182,
+      "step": 1829
+    },
+    {
+      "epoch": 3.388888888888889,
+      "grad_norm": 1.4089460445025883,
+      "learning_rate": 1.984335496751585e-07,
+      "loss": 0.6674,
+      "step": 1830
+    },
+    {
+      "epoch": 3.390740740740741,
+      "grad_norm": 3.0085161164177805,
+      "learning_rate": 1.9843090500207e-07,
+      "loss": 0.7527,
+      "step": 1831
+    },
+    {
+      "epoch": 3.3925925925925924,
+      "grad_norm": 1.4130940415035835,
+      "learning_rate": 1.9842825811598747e-07,
+      "loss": 0.7204,
+      "step": 1832
+    },
+    {
+      "epoch": 3.3944444444444444,
+      "grad_norm": 1.5105516145504305,
+      "learning_rate": 1.9842560901697035e-07,
+      "loss": 0.6699,
+      "step": 1833
+    },
+    {
+      "epoch": 3.3962962962962964,
+      "grad_norm": 1.491999933569889,
+      "learning_rate": 1.9842295770507822e-07,
+      "loss": 0.6861,
+      "step": 1834
+    },
+    {
+      "epoch": 3.398148148148148,
+      "grad_norm": 1.253748570846862,
+      "learning_rate": 1.984203041803707e-07,
+      "loss": 0.6806,
+      "step": 1835
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 1.5545753659869026,
+      "learning_rate": 1.9841764844290743e-07,
+      "loss": 0.6927,
+      "step": 1836
+    },
+    {
+      "epoch": 3.401851851851852,
+      "grad_norm": 1.501761073419926,
+      "learning_rate": 1.984149904927481e-07,
+      "loss": 0.7158,
+      "step": 1837
+    },
+    {
+      "epoch": 3.403703703703704,
+      "grad_norm": 1.7265965770255634,
+      "learning_rate": 1.9841233032995252e-07,
+      "loss": 0.6741,
+      "step": 1838
+    },
+    {
+      "epoch": 3.4055555555555554,
+      "grad_norm": 1.6305566817488533,
+      "learning_rate": 1.9840966795458045e-07,
+      "loss": 0.6849,
+      "step": 1839
+    },
+    {
+      "epoch": 3.4074074074074074,
+      "grad_norm": 1.4081321041940684,
+      "learning_rate": 1.984070033666918e-07,
+      "loss": 0.7227,
+      "step": 1840
+    },
+    {
+      "epoch": 3.4092592592592594,
+      "grad_norm": 1.5194649070203468,
+      "learning_rate": 1.984043365663464e-07,
+      "loss": 0.6884,
+      "step": 1841
+    },
+    {
+      "epoch": 3.411111111111111,
+      "grad_norm": 1.7708797081830638,
+      "learning_rate": 1.984016675536043e-07,
+      "loss": 0.677,
+      "step": 1842
+    },
+    {
+      "epoch": 3.412962962962963,
+      "grad_norm": 1.387747704612861,
+      "learning_rate": 1.9839899632852545e-07,
+      "loss": 0.7512,
+      "step": 1843
+    },
+    {
+      "epoch": 3.414814814814815,
+      "grad_norm": 2.0993706208511282,
+      "learning_rate": 1.983963228911699e-07,
+      "loss": 0.6885,
+      "step": 1844
+    },
+    {
+      "epoch": 3.4166666666666665,
+      "grad_norm": 1.3781961699995935,
+      "learning_rate": 1.9839364724159776e-07,
+      "loss": 0.7207,
+      "step": 1845
+    },
+    {
+      "epoch": 3.4185185185185185,
+      "grad_norm": 1.5712083425661263,
+      "learning_rate": 1.9839096937986922e-07,
+      "loss": 0.6917,
+      "step": 1846
+    },
+    {
+      "epoch": 3.4203703703703705,
+      "grad_norm": 1.7443941577490623,
+      "learning_rate": 1.9838828930604444e-07,
+      "loss": 0.7186,
+      "step": 1847
+    },
+    {
+      "epoch": 3.422222222222222,
+      "grad_norm": 1.4627056735055046,
+      "learning_rate": 1.9838560702018373e-07,
+      "loss": 0.7947,
+      "step": 1848
+    },
+    {
+      "epoch": 3.424074074074074,
+      "grad_norm": 1.4323519976183854,
+      "learning_rate": 1.9838292252234734e-07,
+      "loss": 0.7513,
+      "step": 1849
+    },
+    {
+      "epoch": 3.425925925925926,
+      "grad_norm": 1.5282606769898337,
+      "learning_rate": 1.9838023581259564e-07,
+      "loss": 0.6961,
+      "step": 1850
+    },
+    {
+      "epoch": 3.4277777777777776,
+      "grad_norm": 1.5200764897120511,
+      "learning_rate": 1.9837754689098903e-07,
+      "loss": 0.6774,
+      "step": 1851
+    },
+    {
+      "epoch": 3.4296296296296296,
+      "grad_norm": 1.5836343448700176,
+      "learning_rate": 1.9837485575758805e-07,
+      "loss": 0.7075,
+      "step": 1852
+    },
+    {
+      "epoch": 3.4314814814814816,
+      "grad_norm": 1.533404345674884,
+      "learning_rate": 1.9837216241245306e-07,
+      "loss": 0.6937,
+      "step": 1853
+    },
+    {
+      "epoch": 3.4333333333333336,
+      "grad_norm": 1.7884097474333618,
+      "learning_rate": 1.983694668556447e-07,
+      "loss": 0.7169,
+      "step": 1854
+    },
+    {
+      "epoch": 3.435185185185185,
+      "grad_norm": 1.689564754648096,
+      "learning_rate": 1.983667690872236e-07,
+      "loss": 0.7194,
+      "step": 1855
+    },
+    {
+      "epoch": 3.437037037037037,
+      "grad_norm": 1.584534178578136,
+      "learning_rate": 1.983640691072503e-07,
+      "loss": 0.7312,
+      "step": 1856
+    },
+    {
+      "epoch": 3.438888888888889,
+      "grad_norm": 2.6423491864346618,
+      "learning_rate": 1.9836136691578562e-07,
+      "loss": 0.6967,
+      "step": 1857
+    },
+    {
+      "epoch": 3.4407407407407407,
+      "grad_norm": 1.5964527218129614,
+      "learning_rate": 1.9835866251289023e-07,
+      "loss": 0.6772,
+      "step": 1858
+    },
+    {
+      "epoch": 3.4425925925925926,
+      "grad_norm": 1.4982229630711585,
+      "learning_rate": 1.98355955898625e-07,
+      "loss": 0.6997,
+      "step": 1859
+    },
+    {
+      "epoch": 3.4444444444444446,
+      "grad_norm": 1.4673660259822097,
+      "learning_rate": 1.9835324707305074e-07,
+      "loss": 0.6836,
+      "step": 1860
+    },
+    {
+      "epoch": 3.446296296296296,
+      "grad_norm": 1.842385103664792,
+      "learning_rate": 1.9835053603622838e-07,
+      "loss": 0.7167,
+      "step": 1861
+    },
+    {
+      "epoch": 3.448148148148148,
+      "grad_norm": 1.5177275684482423,
+      "learning_rate": 1.9834782278821883e-07,
+      "loss": 0.6853,
+      "step": 1862
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 2.167833033772285,
+      "learning_rate": 1.983451073290831e-07,
+      "loss": 0.7273,
+      "step": 1863
+    },
+    {
+      "epoch": 3.4518518518518517,
+      "grad_norm": 1.560681261677744,
+      "learning_rate": 1.983423896588823e-07,
+      "loss": 0.6453,
+      "step": 1864
+    },
+    {
+      "epoch": 3.4537037037037037,
+      "grad_norm": 1.5380173541655122,
+      "learning_rate": 1.983396697776775e-07,
+      "loss": 0.7512,
+      "step": 1865
+    },
+    {
+      "epoch": 3.4555555555555557,
+      "grad_norm": 1.4693529725852381,
+      "learning_rate": 1.983369476855298e-07,
+      "loss": 0.7107,
+      "step": 1866
+    },
+    {
+      "epoch": 3.4574074074074073,
+      "grad_norm": 1.4125957262066153,
+      "learning_rate": 1.983342233825004e-07,
+      "loss": 0.7222,
+      "step": 1867
+    },
+    {
+      "epoch": 3.4592592592592593,
+      "grad_norm": 1.5180714260911108,
+      "learning_rate": 1.9833149686865066e-07,
+      "loss": 0.7133,
+      "step": 1868
+    },
+    {
+      "epoch": 3.4611111111111112,
+      "grad_norm": 1.3970107683126725,
+      "learning_rate": 1.9832876814404175e-07,
+      "loss": 0.6759,
+      "step": 1869
+    },
+    {
+      "epoch": 3.462962962962963,
+      "grad_norm": 1.3742674833403816,
+      "learning_rate": 1.983260372087351e-07,
+      "loss": 0.689,
+      "step": 1870
+    },
+    {
+      "epoch": 3.464814814814815,
+      "grad_norm": 1.533264797025038,
+      "learning_rate": 1.983233040627921e-07,
+      "loss": 0.6807,
+      "step": 1871
+    },
+    {
+      "epoch": 3.466666666666667,
+      "grad_norm": 1.5775174685083997,
+      "learning_rate": 1.9832056870627416e-07,
+      "loss": 0.6532,
+      "step": 1872
+    },
+    {
+      "epoch": 3.4685185185185183,
+      "grad_norm": 1.57956971869636,
+      "learning_rate": 1.983178311392428e-07,
+      "loss": 0.7,
+      "step": 1873
+    },
+    {
+      "epoch": 3.4703703703703703,
+      "grad_norm": 1.60845502213651,
+      "learning_rate": 1.983150913617596e-07,
+      "loss": 0.711,
+      "step": 1874
+    },
+    {
+      "epoch": 3.4722222222222223,
+      "grad_norm": 1.5212273014649118,
+      "learning_rate": 1.983123493738861e-07,
+      "loss": 0.7242,
+      "step": 1875
+    },
+    {
+      "epoch": 3.474074074074074,
+      "grad_norm": 1.8248193052095798,
+      "learning_rate": 1.9830960517568395e-07,
+      "loss": 0.7044,
+      "step": 1876
+    },
+    {
+      "epoch": 3.475925925925926,
+      "grad_norm": 1.4933086652691223,
+      "learning_rate": 1.9830685876721488e-07,
+      "loss": 0.668,
+      "step": 1877
+    },
+    {
+      "epoch": 3.477777777777778,
+      "grad_norm": 1.409484200081249,
+      "learning_rate": 1.9830411014854068e-07,
+      "loss": 0.7498,
+      "step": 1878
+    },
+    {
+      "epoch": 3.4796296296296294,
+      "grad_norm": 1.450000499176415,
+      "learning_rate": 1.9830135931972304e-07,
+      "loss": 0.7118,
+      "step": 1879
+    },
+    {
+      "epoch": 3.4814814814814814,
+      "grad_norm": 1.3893754012226047,
+      "learning_rate": 1.982986062808239e-07,
+      "loss": 0.7373,
+      "step": 1880
+    },
+    {
+      "epoch": 3.4833333333333334,
+      "grad_norm": 1.3333801218890933,
+      "learning_rate": 1.982958510319051e-07,
+      "loss": 0.6864,
+      "step": 1881
+    },
+    {
+      "epoch": 3.485185185185185,
+      "grad_norm": 1.6738239699454398,
+      "learning_rate": 1.9829309357302857e-07,
+      "loss": 0.681,
+      "step": 1882
+    },
+    {
+      "epoch": 3.487037037037037,
+      "grad_norm": 1.8996939479255008,
+      "learning_rate": 1.9829033390425635e-07,
+      "loss": 0.6832,
+      "step": 1883
+    },
+    {
+      "epoch": 3.488888888888889,
+      "grad_norm": 1.4297045913556865,
+      "learning_rate": 1.982875720256505e-07,
+      "loss": 0.6724,
+      "step": 1884
+    },
+    {
+      "epoch": 3.490740740740741,
+      "grad_norm": 1.3175346513259658,
+      "learning_rate": 1.9828480793727308e-07,
+      "loss": 0.7191,
+      "step": 1885
+    },
+    {
+      "epoch": 3.4925925925925925,
+      "grad_norm": 1.2809625303647758,
+      "learning_rate": 1.982820416391862e-07,
+      "loss": 0.6955,
+      "step": 1886
+    },
+    {
+      "epoch": 3.4944444444444445,
+      "grad_norm": 1.7030736431840674,
+      "learning_rate": 1.9827927313145214e-07,
+      "loss": 0.6785,
+      "step": 1887
+    },
+    {
+      "epoch": 3.4962962962962965,
+      "grad_norm": 1.4782137077450757,
+      "learning_rate": 1.9827650241413308e-07,
+      "loss": 0.6843,
+      "step": 1888
+    },
+    {
+      "epoch": 3.498148148148148,
+      "grad_norm": 11.343940043858082,
+      "learning_rate": 1.9827372948729132e-07,
+      "loss": 0.67,
+      "step": 1889
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 2.4214805025973365,
+      "learning_rate": 1.9827095435098923e-07,
+      "loss": 0.6929,
+      "step": 1890
+    },
+    {
+      "epoch": 3.501851851851852,
+      "grad_norm": 1.446837042767488,
+      "learning_rate": 1.9826817700528918e-07,
+      "loss": 0.7051,
+      "step": 1891
+    },
+    {
+      "epoch": 3.5037037037037035,
+      "grad_norm": 1.525076726251059,
+      "learning_rate": 1.9826539745025364e-07,
+      "loss": 0.7174,
+      "step": 1892
+    },
+    {
+      "epoch": 3.5055555555555555,
+      "grad_norm": 1.8971822298167618,
+      "learning_rate": 1.9826261568594505e-07,
+      "loss": 0.7369,
+      "step": 1893
+    },
+    {
+      "epoch": 3.5074074074074075,
+      "grad_norm": 1.402419883821451,
+      "learning_rate": 1.98259831712426e-07,
+      "loss": 0.7071,
+      "step": 1894
+    },
+    {
+      "epoch": 3.5092592592592595,
+      "grad_norm": 1.4348258278852095,
+      "learning_rate": 1.9825704552975905e-07,
+      "loss": 0.7685,
+      "step": 1895
+    },
+    {
+      "epoch": 3.511111111111111,
+      "grad_norm": 1.3758144198481757,
+      "learning_rate": 1.9825425713800684e-07,
+      "loss": 0.7234,
+      "step": 1896
+    },
+    {
+      "epoch": 3.512962962962963,
+      "grad_norm": 1.432153376179934,
+      "learning_rate": 1.9825146653723212e-07,
+      "loss": 0.683,
+      "step": 1897
+    },
+    {
+      "epoch": 3.514814814814815,
+      "grad_norm": 1.592847660149842,
+      "learning_rate": 1.9824867372749754e-07,
+      "loss": 0.7368,
+      "step": 1898
+    },
+    {
+      "epoch": 3.5166666666666666,
+      "grad_norm": 1.4260888262639537,
+      "learning_rate": 1.9824587870886599e-07,
+      "loss": 0.7514,
+      "step": 1899
+    },
+    {
+      "epoch": 3.5185185185185186,
+      "grad_norm": 1.425142455196794,
+      "learning_rate": 1.9824308148140018e-07,
+      "loss": 0.6946,
+      "step": 1900
+    },
+    {
+      "epoch": 3.5203703703703706,
+      "grad_norm": 1.486037666687816,
+      "learning_rate": 1.9824028204516313e-07,
+      "loss": 0.687,
+      "step": 1901
+    },
+    {
+      "epoch": 3.522222222222222,
+      "grad_norm": 1.5990733904107073,
+      "learning_rate": 1.9823748040021772e-07,
+      "loss": 0.7399,
+      "step": 1902
+    },
+    {
+      "epoch": 3.524074074074074,
+      "grad_norm": 1.9209588481351703,
+      "learning_rate": 1.982346765466269e-07,
+      "loss": 0.6501,
+      "step": 1903
+    },
+    {
+      "epoch": 3.525925925925926,
+      "grad_norm": 1.4512384059639374,
+      "learning_rate": 1.982318704844538e-07,
+      "loss": 0.7175,
+      "step": 1904
+    },
+    {
+      "epoch": 3.5277777777777777,
+      "grad_norm": 1.6544460474443168,
+      "learning_rate": 1.9822906221376142e-07,
+      "loss": 0.6819,
+      "step": 1905
+    },
+    {
+      "epoch": 3.5296296296296297,
+      "grad_norm": 2.256026112534218,
+      "learning_rate": 1.98226251734613e-07,
+      "loss": 0.7003,
+      "step": 1906
+    },
+    {
+      "epoch": 3.5314814814814817,
+      "grad_norm": 1.595422499695949,
+      "learning_rate": 1.9822343904707157e-07,
+      "loss": 0.705,
+      "step": 1907
+    },
+    {
+      "epoch": 3.533333333333333,
+      "grad_norm": 1.7061819333454802,
+      "learning_rate": 1.9822062415120051e-07,
+      "loss": 0.7048,
+      "step": 1908
+    },
+    {
+      "epoch": 3.535185185185185,
+      "grad_norm": 1.4195221640502955,
+      "learning_rate": 1.9821780704706305e-07,
+      "loss": 0.7488,
+      "step": 1909
+    },
+    {
+      "epoch": 3.537037037037037,
+      "grad_norm": 1.9156954767277399,
+      "learning_rate": 1.9821498773472256e-07,
+      "loss": 0.7194,
+      "step": 1910
+    },
+    {
+      "epoch": 3.5388888888888888,
+      "grad_norm": 1.8031470161670127,
+      "learning_rate": 1.9821216621424237e-07,
+      "loss": 0.7232,
+      "step": 1911
+    },
+    {
+      "epoch": 3.5407407407407407,
+      "grad_norm": 1.5315746107696702,
+      "learning_rate": 1.9820934248568593e-07,
+      "loss": 0.6812,
+      "step": 1912
+    },
+    {
+      "epoch": 3.5425925925925927,
+      "grad_norm": 1.3025150919659598,
+      "learning_rate": 1.9820651654911676e-07,
+      "loss": 0.7151,
+      "step": 1913
+    },
+    {
+      "epoch": 3.5444444444444443,
+      "grad_norm": 1.6177111951062892,
+      "learning_rate": 1.9820368840459836e-07,
+      "loss": 0.7453,
+      "step": 1914
+    },
+    {
+      "epoch": 3.5462962962962963,
+      "grad_norm": 1.7753496097386425,
+      "learning_rate": 1.9820085805219433e-07,
+      "loss": 0.6847,
+      "step": 1915
+    },
+    {
+      "epoch": 3.5481481481481483,
+      "grad_norm": 1.4141943625649132,
+      "learning_rate": 1.981980254919683e-07,
+      "loss": 0.6573,
+      "step": 1916
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 1.3753426273588385,
+      "learning_rate": 1.9819519072398395e-07,
+      "loss": 0.7115,
+      "step": 1917
+    },
+    {
+      "epoch": 3.551851851851852,
+      "grad_norm": 1.6651880352440698,
+      "learning_rate": 1.9819235374830505e-07,
+      "loss": 0.7062,
+      "step": 1918
+    },
+    {
+      "epoch": 3.553703703703704,
+      "grad_norm": 5.42353531434544,
+      "learning_rate": 1.9818951456499532e-07,
+      "loss": 0.7279,
+      "step": 1919
+    },
+    {
+      "epoch": 3.5555555555555554,
+      "grad_norm": 1.442599430680827,
+      "learning_rate": 1.9818667317411864e-07,
+      "loss": 0.6872,
+      "step": 1920
+    },
+    {
+      "epoch": 3.5574074074074074,
+      "grad_norm": 1.596581035572579,
+      "learning_rate": 1.9818382957573887e-07,
+      "loss": 0.6706,
+      "step": 1921
+    },
+    {
+      "epoch": 3.5592592592592593,
+      "grad_norm": 1.5801005818380645,
+      "learning_rate": 1.9818098376991995e-07,
+      "loss": 0.7154,
+      "step": 1922
+    },
+    {
+      "epoch": 3.561111111111111,
+      "grad_norm": 1.348122080031413,
+      "learning_rate": 1.9817813575672587e-07,
+      "loss": 0.6542,
+      "step": 1923
+    },
+    {
+      "epoch": 3.562962962962963,
+      "grad_norm": 1.5730064508346542,
+      "learning_rate": 1.981752855362206e-07,
+      "loss": 0.7175,
+      "step": 1924
+    },
+    {
+      "epoch": 3.564814814814815,
+      "grad_norm": 1.2527586950293443,
+      "learning_rate": 1.9817243310846833e-07,
+      "loss": 0.6598,
+      "step": 1925
+    },
+    {
+      "epoch": 3.5666666666666664,
+      "grad_norm": 1.8362728130364492,
+      "learning_rate": 1.981695784735331e-07,
+      "loss": 0.7132,
+      "step": 1926
+    },
+    {
+      "epoch": 3.5685185185185184,
+      "grad_norm": 1.436897812936575,
+      "learning_rate": 1.9816672163147913e-07,
+      "loss": 0.6979,
+      "step": 1927
+    },
+    {
+      "epoch": 3.5703703703703704,
+      "grad_norm": 1.3818273524498794,
+      "learning_rate": 1.9816386258237065e-07,
+      "loss": 0.6888,
+      "step": 1928
+    },
+    {
+      "epoch": 3.572222222222222,
+      "grad_norm": 1.8411447032643546,
+      "learning_rate": 1.981610013262719e-07,
+      "loss": 0.6785,
+      "step": 1929
+    },
+    {
+      "epoch": 3.574074074074074,
+      "grad_norm": 1.7121233462109091,
+      "learning_rate": 1.981581378632473e-07,
+      "loss": 0.7334,
+      "step": 1930
+    },
+    {
+      "epoch": 3.575925925925926,
+      "grad_norm": 1.561830720118593,
+      "learning_rate": 1.9815527219336113e-07,
+      "loss": 0.7199,
+      "step": 1931
+    },
+    {
+      "epoch": 3.5777777777777775,
+      "grad_norm": 3.4500124348032752,
+      "learning_rate": 1.9815240431667782e-07,
+      "loss": 0.7113,
+      "step": 1932
+    },
+    {
+      "epoch": 3.5796296296296295,
+      "grad_norm": 1.5484682542410215,
+      "learning_rate": 1.9814953423326193e-07,
+      "loss": 0.7341,
+      "step": 1933
+    },
+    {
+      "epoch": 3.5814814814814815,
+      "grad_norm": 1.4238796151975872,
+      "learning_rate": 1.9814666194317793e-07,
+      "loss": 0.6976,
+      "step": 1934
+    },
+    {
+      "epoch": 3.5833333333333335,
+      "grad_norm": 2.283827498050039,
+      "learning_rate": 1.9814378744649041e-07,
+      "loss": 0.6982,
+      "step": 1935
+    },
+    {
+      "epoch": 3.585185185185185,
+      "grad_norm": 1.4087371610510286,
+      "learning_rate": 1.9814091074326396e-07,
+      "loss": 0.7009,
+      "step": 1936
+    },
+    {
+      "epoch": 3.587037037037037,
+      "grad_norm": 1.3477722141793254,
+      "learning_rate": 1.9813803183356335e-07,
+      "loss": 0.6668,
+      "step": 1937
+    },
+    {
+      "epoch": 3.588888888888889,
+      "grad_norm": 1.8745696148294126,
+      "learning_rate": 1.9813515071745322e-07,
+      "loss": 0.6847,
+      "step": 1938
+    },
+    {
+      "epoch": 3.590740740740741,
+      "grad_norm": 1.8021310741794647,
+      "learning_rate": 1.9813226739499836e-07,
+      "loss": 0.7186,
+      "step": 1939
+    },
+    {
+      "epoch": 3.5925925925925926,
+      "grad_norm": 1.509633072480398,
+      "learning_rate": 1.981293818662636e-07,
+      "loss": 0.7159,
+      "step": 1940
+    },
+    {
+      "epoch": 3.5944444444444446,
+      "grad_norm": 1.4675348428612094,
+      "learning_rate": 1.9812649413131383e-07,
+      "loss": 0.692,
+      "step": 1941
+    },
+    {
+      "epoch": 3.5962962962962965,
+      "grad_norm": 1.4097829290374755,
+      "learning_rate": 1.9812360419021395e-07,
+      "loss": 0.6862,
+      "step": 1942
+    },
+    {
+      "epoch": 3.598148148148148,
+      "grad_norm": 1.5781638537942753,
+      "learning_rate": 1.9812071204302897e-07,
+      "loss": 0.7129,
+      "step": 1943
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 1.5196045953144095,
+      "learning_rate": 1.981178176898239e-07,
+      "loss": 0.7054,
+      "step": 1944
+    },
+    {
+      "epoch": 3.601851851851852,
+      "grad_norm": 1.8758151840317552,
+      "learning_rate": 1.9811492113066378e-07,
+      "loss": 0.7297,
+      "step": 1945
+    },
+    {
+      "epoch": 3.6037037037037036,
+      "grad_norm": 1.5253808813624143,
+      "learning_rate": 1.9811202236561377e-07,
+      "loss": 0.6904,
+      "step": 1946
+    },
+    {
+      "epoch": 3.6055555555555556,
+      "grad_norm": 1.698065489781154,
+      "learning_rate": 1.9810912139473904e-07,
+      "loss": 0.6856,
+      "step": 1947
+    },
+    {
+      "epoch": 3.6074074074074076,
+      "grad_norm": 3.5386847166983766,
+      "learning_rate": 1.9810621821810475e-07,
+      "loss": 0.6707,
+      "step": 1948
+    },
+    {
+      "epoch": 3.609259259259259,
+      "grad_norm": 1.7630963583130548,
+      "learning_rate": 1.9810331283577628e-07,
+      "loss": 0.6897,
+      "step": 1949
+    },
+    {
+      "epoch": 3.611111111111111,
+      "grad_norm": 1.3857742277969796,
+      "learning_rate": 1.9810040524781888e-07,
+      "loss": 0.7382,
+      "step": 1950
+    },
+    {
+      "epoch": 3.612962962962963,
+      "grad_norm": 1.3734946341593384,
+      "learning_rate": 1.9809749545429791e-07,
+      "loss": 0.6744,
+      "step": 1951
+    },
+    {
+      "epoch": 3.6148148148148147,
+      "grad_norm": 2.303107556487947,
+      "learning_rate": 1.9809458345527883e-07,
+      "loss": 0.6553,
+      "step": 1952
+    },
+    {
+      "epoch": 3.6166666666666667,
+      "grad_norm": 1.4950940892456774,
+      "learning_rate": 1.980916692508271e-07,
+      "loss": 0.6463,
+      "step": 1953
+    },
+    {
+      "epoch": 3.6185185185185187,
+      "grad_norm": 1.7925375390372997,
+      "learning_rate": 1.9808875284100824e-07,
+      "loss": 0.6826,
+      "step": 1954
+    },
+    {
+      "epoch": 3.6203703703703702,
+      "grad_norm": 1.480093862909119,
+      "learning_rate": 1.9808583422588775e-07,
+      "loss": 0.692,
+      "step": 1955
+    },
+    {
+      "epoch": 3.6222222222222222,
+      "grad_norm": 1.6878061639490665,
+      "learning_rate": 1.9808291340553139e-07,
+      "loss": 0.7004,
+      "step": 1956
+    },
+    {
+      "epoch": 3.6240740740740742,
+      "grad_norm": 1.6328820596569467,
+      "learning_rate": 1.9807999038000468e-07,
+      "loss": 0.7131,
+      "step": 1957
+    },
+    {
+      "epoch": 3.6259259259259258,
+      "grad_norm": 1.4413178657128076,
+      "learning_rate": 1.9807706514937344e-07,
+      "loss": 0.6968,
+      "step": 1958
+    },
+    {
+      "epoch": 3.6277777777777778,
+      "grad_norm": 1.4861856575483776,
+      "learning_rate": 1.980741377137034e-07,
+      "loss": 0.7087,
+      "step": 1959
+    },
+    {
+      "epoch": 3.6296296296296298,
+      "grad_norm": 1.450398711422783,
+      "learning_rate": 1.980712080730604e-07,
+      "loss": 0.7298,
+      "step": 1960
+    },
+    {
+      "epoch": 3.6314814814814813,
+      "grad_norm": 1.7686534728701246,
+      "learning_rate": 1.9806827622751026e-07,
+      "loss": 0.6793,
+      "step": 1961
+    },
+    {
+      "epoch": 3.6333333333333333,
+      "grad_norm": 2.0503800147413265,
+      "learning_rate": 1.9806534217711892e-07,
+      "loss": 0.704,
+      "step": 1962
+    },
+    {
+      "epoch": 3.6351851851851853,
+      "grad_norm": 1.4919802579266688,
+      "learning_rate": 1.9806240592195235e-07,
+      "loss": 0.7447,
+      "step": 1963
+    },
+    {
+      "epoch": 3.637037037037037,
+      "grad_norm": 1.3158391307142547,
+      "learning_rate": 1.9805946746207655e-07,
+      "loss": 0.6951,
+      "step": 1964
+    },
+    {
+      "epoch": 3.638888888888889,
+      "grad_norm": 1.387093836355547,
+      "learning_rate": 1.9805652679755763e-07,
+      "loss": 0.6782,
+      "step": 1965
+    },
+    {
+      "epoch": 3.640740740740741,
+      "grad_norm": 1.58426465857652,
+      "learning_rate": 1.9805358392846164e-07,
+      "loss": 0.7507,
+      "step": 1966
+    },
+    {
+      "epoch": 3.6425925925925924,
+      "grad_norm": 1.4062073987628834,
+      "learning_rate": 1.9805063885485476e-07,
+      "loss": 0.6919,
+      "step": 1967
+    },
+    {
+      "epoch": 3.6444444444444444,
+      "grad_norm": 1.4713974122016362,
+      "learning_rate": 1.9804769157680325e-07,
+      "loss": 0.7027,
+      "step": 1968
+    },
+    {
+      "epoch": 3.6462962962962964,
+      "grad_norm": 1.4662203585245022,
+      "learning_rate": 1.9804474209437333e-07,
+      "loss": 0.7188,
+      "step": 1969
+    },
+    {
+      "epoch": 3.648148148148148,
+      "grad_norm": 1.3029363415877,
+      "learning_rate": 1.9804179040763132e-07,
+      "loss": 0.7262,
+      "step": 1970
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 1.405182602453171,
+      "learning_rate": 1.9803883651664357e-07,
+      "loss": 0.7192,
+      "step": 1971
+    },
+    {
+      "epoch": 3.651851851851852,
+      "grad_norm": 1.5528130029290417,
+      "learning_rate": 1.9803588042147653e-07,
+      "loss": 0.7816,
+      "step": 1972
+    },
+    {
+      "epoch": 3.6537037037037035,
+      "grad_norm": 1.5805543499120351,
+      "learning_rate": 1.9803292212219662e-07,
+      "loss": 0.6829,
+      "step": 1973
+    },
+    {
+      "epoch": 3.6555555555555554,
+      "grad_norm": 1.5424612361575634,
+      "learning_rate": 1.980299616188704e-07,
+      "loss": 0.7299,
+      "step": 1974
+    },
+    {
+      "epoch": 3.6574074074074074,
+      "grad_norm": 1.4060507652410217,
+      "learning_rate": 1.9802699891156436e-07,
+      "loss": 0.6992,
+      "step": 1975
+    },
+    {
+      "epoch": 3.659259259259259,
+      "grad_norm": 1.7471079807195133,
+      "learning_rate": 1.9802403400034516e-07,
+      "loss": 0.6756,
+      "step": 1976
+    },
+    {
+      "epoch": 3.661111111111111,
+      "grad_norm": 3.0085525503831794,
+      "learning_rate": 1.9802106688527945e-07,
+      "loss": 0.6703,
+      "step": 1977
+    },
+    {
+      "epoch": 3.662962962962963,
+      "grad_norm": 2.1842143747557836,
+      "learning_rate": 1.9801809756643394e-07,
+      "loss": 0.6757,
+      "step": 1978
+    },
+    {
+      "epoch": 3.664814814814815,
+      "grad_norm": 1.634080689590214,
+      "learning_rate": 1.9801512604387535e-07,
+      "loss": 0.7206,
+      "step": 1979
+    },
+    {
+      "epoch": 3.6666666666666665,
+      "grad_norm": 1.56745649562272,
+      "learning_rate": 1.9801215231767056e-07,
+      "loss": 0.7439,
+      "step": 1980
+    },
+    {
+      "epoch": 3.6685185185185185,
+      "grad_norm": 1.3999167261626693,
+      "learning_rate": 1.9800917638788636e-07,
+      "loss": 0.7085,
+      "step": 1981
+    },
+    {
+      "epoch": 3.6703703703703705,
+      "grad_norm": 1.464422738842721,
+      "learning_rate": 1.980061982545897e-07,
+      "loss": 0.7237,
+      "step": 1982
+    },
+    {
+      "epoch": 3.6722222222222225,
+      "grad_norm": 1.520337377827549,
+      "learning_rate": 1.9800321791784754e-07,
+      "loss": 0.6891,
+      "step": 1983
+    },
+    {
+      "epoch": 3.674074074074074,
+      "grad_norm": 1.4504553429395952,
+      "learning_rate": 1.9800023537772686e-07,
+      "loss": 0.7088,
+      "step": 1984
+    },
+    {
+      "epoch": 3.675925925925926,
+      "grad_norm": 1.31576516299285,
+      "learning_rate": 1.9799725063429472e-07,
+      "loss": 0.7121,
+      "step": 1985
+    },
+    {
+      "epoch": 3.677777777777778,
+      "grad_norm": 1.647316184829491,
+      "learning_rate": 1.979942636876182e-07,
+      "loss": 0.6851,
+      "step": 1986
+    },
+    {
+      "epoch": 3.6796296296296296,
+      "grad_norm": 1.5264392525036654,
+      "learning_rate": 1.979912745377645e-07,
+      "loss": 0.7135,
+      "step": 1987
+    },
+    {
+      "epoch": 3.6814814814814816,
+      "grad_norm": 1.4004447779462403,
+      "learning_rate": 1.9798828318480084e-07,
+      "loss": 0.7534,
+      "step": 1988
+    },
+    {
+      "epoch": 3.6833333333333336,
+      "grad_norm": 1.421737883389671,
+      "learning_rate": 1.979852896287944e-07,
+      "loss": 0.6945,
+      "step": 1989
+    },
+    {
+      "epoch": 3.685185185185185,
+      "grad_norm": 1.3769733359853031,
+      "learning_rate": 1.9798229386981256e-07,
+      "loss": 0.6964,
+      "step": 1990
+    },
+    {
+      "epoch": 3.687037037037037,
+      "grad_norm": 1.380796306197519,
+      "learning_rate": 1.979792959079226e-07,
+      "loss": 0.6818,
+      "step": 1991
+    },
+    {
+      "epoch": 3.688888888888889,
+      "grad_norm": 1.507949694068586,
+      "learning_rate": 1.9797629574319199e-07,
+      "loss": 0.6516,
+      "step": 1992
+    },
+    {
+      "epoch": 3.6907407407407407,
+      "grad_norm": 1.3504008732820652,
+      "learning_rate": 1.9797329337568815e-07,
+      "loss": 0.6784,
+      "step": 1993
+    },
+    {
+      "epoch": 3.6925925925925926,
+      "grad_norm": 1.8646438247757418,
+      "learning_rate": 1.9797028880547857e-07,
+      "loss": 0.6611,
+      "step": 1994
+    },
+    {
+      "epoch": 3.6944444444444446,
+      "grad_norm": 1.5331222234717918,
+      "learning_rate": 1.9796728203263084e-07,
+      "loss": 0.7425,
+      "step": 1995
+    },
+    {
+      "epoch": 3.696296296296296,
+      "grad_norm": 1.4674305006650243,
+      "learning_rate": 1.979642730572125e-07,
+      "loss": 0.7261,
+      "step": 1996
+    },
+    {
+      "epoch": 3.698148148148148,
+      "grad_norm": 1.433639727653537,
+      "learning_rate": 1.9796126187929127e-07,
+      "loss": 0.7131,
+      "step": 1997
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 1.4766776329393654,
+      "learning_rate": 1.9795824849893478e-07,
+      "loss": 0.7274,
+      "step": 1998
+    },
+    {
+      "epoch": 3.7018518518518517,
+      "grad_norm": 1.5498696619591807,
+      "learning_rate": 1.9795523291621084e-07,
+      "loss": 0.6378,
+      "step": 1999
+    },
+    {
+      "epoch": 3.7037037037037037,
+      "grad_norm": 2.056798683658681,
+      "learning_rate": 1.979522151311872e-07,
+      "loss": 0.7054,
+      "step": 2000
+    },
+    {
+      "epoch": 3.7055555555555557,
+      "grad_norm": 1.9168652757739209,
+      "learning_rate": 1.9794919514393174e-07,
+      "loss": 0.6955,
+      "step": 2001
+    },
+    {
+      "epoch": 3.7074074074074073,
+      "grad_norm": 1.3427405856092864,
+      "learning_rate": 1.9794617295451236e-07,
+      "loss": 0.6803,
+      "step": 2002
+    },
+    {
+      "epoch": 3.7092592592592593,
+      "grad_norm": 1.4538624910105042,
+      "learning_rate": 1.97943148562997e-07,
+      "loss": 0.7037,
+      "step": 2003
+    },
+    {
+      "epoch": 3.7111111111111112,
+      "grad_norm": 1.609870531030301,
+      "learning_rate": 1.9794012196945359e-07,
+      "loss": 0.6755,
+      "step": 2004
+    },
+    {
+      "epoch": 3.712962962962963,
+      "grad_norm": 1.3067156590234197,
+      "learning_rate": 1.979370931739503e-07,
+      "loss": 0.6655,
+      "step": 2005
+    },
+    {
+      "epoch": 3.714814814814815,
+      "grad_norm": 1.4877041837675509,
+      "learning_rate": 1.9793406217655514e-07,
+      "loss": 0.7153,
+      "step": 2006
+    },
+    {
+      "epoch": 3.716666666666667,
+      "grad_norm": 1.4331058594994008,
+      "learning_rate": 1.9793102897733628e-07,
+      "loss": 0.6593,
+      "step": 2007
+    },
+    {
+      "epoch": 3.7185185185185183,
+      "grad_norm": 1.4166386806150968,
+      "learning_rate": 1.979279935763619e-07,
+      "loss": 0.6745,
+      "step": 2008
+    },
+    {
+      "epoch": 3.7203703703703703,
+      "grad_norm": 1.5513582114294344,
+      "learning_rate": 1.9792495597370028e-07,
+      "loss": 0.6914,
+      "step": 2009
+    },
+    {
+      "epoch": 3.7222222222222223,
+      "grad_norm": 1.6670143178468493,
+      "learning_rate": 1.9792191616941967e-07,
+      "loss": 0.7328,
+      "step": 2010
+    },
+    {
+      "epoch": 3.724074074074074,
+      "grad_norm": 1.57401772482662,
+      "learning_rate": 1.9791887416358843e-07,
+      "loss": 0.7114,
+      "step": 2011
+    },
+    {
+      "epoch": 3.725925925925926,
+      "grad_norm": 1.521960752005683,
+      "learning_rate": 1.9791582995627497e-07,
+      "loss": 0.6835,
+      "step": 2012
+    },
+    {
+      "epoch": 3.727777777777778,
+      "grad_norm": 1.5065661637057683,
+      "learning_rate": 1.9791278354754773e-07,
+      "loss": 0.688,
+      "step": 2013
+    },
+    {
+      "epoch": 3.7296296296296294,
+      "grad_norm": 1.943834273685275,
+      "learning_rate": 1.9790973493747515e-07,
+      "loss": 0.7396,
+      "step": 2014
+    },
+    {
+      "epoch": 3.7314814814814814,
+      "grad_norm": 1.8792028376394398,
+      "learning_rate": 1.9790668412612584e-07,
+      "loss": 0.7555,
+      "step": 2015
+    },
+    {
+      "epoch": 3.7333333333333334,
+      "grad_norm": 1.4512496006177689,
+      "learning_rate": 1.9790363111356836e-07,
+      "loss": 0.7064,
+      "step": 2016
+    },
+    {
+      "epoch": 3.735185185185185,
+      "grad_norm": 1.521998551107538,
+      "learning_rate": 1.9790057589987135e-07,
+      "loss": 0.707,
+      "step": 2017
+    },
+    {
+      "epoch": 3.737037037037037,
+      "grad_norm": 2.4848097369119615,
+      "learning_rate": 1.9789751848510349e-07,
+      "loss": 0.6751,
+      "step": 2018
+    },
+    {
+      "epoch": 3.738888888888889,
+      "grad_norm": 1.5320525142519865,
+      "learning_rate": 1.978944588693335e-07,
+      "loss": 0.6554,
+      "step": 2019
+    },
+    {
+      "epoch": 3.7407407407407405,
+      "grad_norm": 1.5317323963504297,
+      "learning_rate": 1.9789139705263025e-07,
+      "loss": 0.7245,
+      "step": 2020
+    },
+    {
+      "epoch": 3.7425925925925925,
+      "grad_norm": 1.4848651542471398,
+      "learning_rate": 1.978883330350625e-07,
+      "loss": 0.6909,
+      "step": 2021
+    },
+    {
+      "epoch": 3.7444444444444445,
+      "grad_norm": 1.3011226506723446,
+      "learning_rate": 1.9788526681669918e-07,
+      "loss": 0.7162,
+      "step": 2022
+    },
+    {
+      "epoch": 3.7462962962962965,
+      "grad_norm": 2.0603608025329914,
+      "learning_rate": 1.9788219839760918e-07,
+      "loss": 0.6976,
+      "step": 2023
+    },
+    {
+      "epoch": 3.748148148148148,
+      "grad_norm": 1.5721552562851024,
+      "learning_rate": 1.9787912777786154e-07,
+      "loss": 0.693,
+      "step": 2024
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 1.776457099525263,
+      "learning_rate": 1.9787605495752527e-07,
+      "loss": 0.6921,
+      "step": 2025
+    },
+    {
+      "epoch": 3.751851851851852,
+      "grad_norm": 1.4995937563146968,
+      "learning_rate": 1.9787297993666943e-07,
+      "loss": 0.68,
+      "step": 2026
+    },
+    {
+      "epoch": 3.7537037037037035,
+      "grad_norm": 1.5205568125459694,
+      "learning_rate": 1.9786990271536318e-07,
+      "loss": 0.7253,
+      "step": 2027
+    },
+    {
+      "epoch": 3.7555555555555555,
+      "grad_norm": 1.5091086882895302,
+      "learning_rate": 1.9786682329367575e-07,
+      "loss": 0.712,
+      "step": 2028
+    },
+    {
+      "epoch": 3.7574074074074075,
+      "grad_norm": 1.7870643678695348,
+      "learning_rate": 1.978637416716763e-07,
+      "loss": 0.6854,
+      "step": 2029
+    },
+    {
+      "epoch": 3.7592592592592595,
+      "grad_norm": 1.911121254209067,
+      "learning_rate": 1.9786065784943415e-07,
+      "loss": 0.696,
+      "step": 2030
+    },
+    {
+      "epoch": 3.761111111111111,
+      "grad_norm": 3.5116070925072425,
+      "learning_rate": 1.9785757182701862e-07,
+      "loss": 0.6973,
+      "step": 2031
+    },
+    {
+      "epoch": 3.762962962962963,
+      "grad_norm": 1.288370880172143,
+      "learning_rate": 1.9785448360449912e-07,
+      "loss": 0.717,
+      "step": 2032
+    },
+    {
+      "epoch": 3.764814814814815,
+      "grad_norm": 1.4654645664311308,
+      "learning_rate": 1.9785139318194504e-07,
+      "loss": 0.6982,
+      "step": 2033
+    },
+    {
+      "epoch": 3.7666666666666666,
+      "grad_norm": 1.4875349563135496,
+      "learning_rate": 1.978483005594259e-07,
+      "loss": 0.6417,
+      "step": 2034
+    },
+    {
+      "epoch": 3.7685185185185186,
+      "grad_norm": 1.4903407458896933,
+      "learning_rate": 1.978452057370112e-07,
+      "loss": 0.706,
+      "step": 2035
+    },
+    {
+      "epoch": 3.7703703703703706,
+      "grad_norm": 1.3498656727464402,
+      "learning_rate": 1.9784210871477053e-07,
+      "loss": 0.7015,
+      "step": 2036
+    },
+    {
+      "epoch": 3.772222222222222,
+      "grad_norm": 1.4561177158041902,
+      "learning_rate": 1.978390094927735e-07,
+      "loss": 0.7258,
+      "step": 2037
+    },
+    {
+      "epoch": 3.774074074074074,
+      "grad_norm": 1.5056345130887145,
+      "learning_rate": 1.9783590807108984e-07,
+      "loss": 0.6856,
+      "step": 2038
+    },
+    {
+      "epoch": 3.775925925925926,
+      "grad_norm": 1.3636144780062167,
+      "learning_rate": 1.9783280444978924e-07,
+      "loss": 0.6788,
+      "step": 2039
+    },
+    {
+      "epoch": 3.7777777777777777,
+      "grad_norm": 1.5763413064558116,
+      "learning_rate": 1.9782969862894148e-07,
+      "loss": 0.7254,
+      "step": 2040
+    },
+    {
+      "epoch": 3.7796296296296297,
+      "grad_norm": 1.5831828661719225,
+      "learning_rate": 1.9782659060861642e-07,
+      "loss": 0.7085,
+      "step": 2041
+    },
+    {
+      "epoch": 3.7814814814814817,
+      "grad_norm": 1.559301490312729,
+      "learning_rate": 1.9782348038888387e-07,
+      "loss": 0.7009,
+      "step": 2042
+    },
+    {
+      "epoch": 3.783333333333333,
+      "grad_norm": 1.9807510684118732,
+      "learning_rate": 1.9782036796981383e-07,
+      "loss": 0.7166,
+      "step": 2043
+    },
+    {
+      "epoch": 3.785185185185185,
+      "grad_norm": 1.9191446790891853,
+      "learning_rate": 1.9781725335147624e-07,
+      "loss": 0.6643,
+      "step": 2044
+    },
+    {
+      "epoch": 3.787037037037037,
+      "grad_norm": 3.1126646602583503,
+      "learning_rate": 1.978141365339411e-07,
+      "loss": 0.6856,
+      "step": 2045
+    },
+    {
+      "epoch": 3.7888888888888888,
+      "grad_norm": 2.337472437667415,
+      "learning_rate": 1.978110175172785e-07,
+      "loss": 0.7104,
+      "step": 2046
+    },
+    {
+      "epoch": 3.7907407407407407,
+      "grad_norm": 1.7197943963713567,
+      "learning_rate": 1.9780789630155862e-07,
+      "loss": 0.6728,
+      "step": 2047
+    },
+    {
+      "epoch": 3.7925925925925927,
+      "grad_norm": 1.6202242251905143,
+      "learning_rate": 1.9780477288685158e-07,
+      "loss": 0.6419,
+      "step": 2048
+    },
+    {
+      "epoch": 3.7944444444444443,
+      "grad_norm": 1.4939723533886449,
+      "learning_rate": 1.9780164727322758e-07,
+      "loss": 0.6997,
+      "step": 2049
+    },
+    {
+      "epoch": 3.7962962962962963,
+      "grad_norm": 1.600820817558738,
+      "learning_rate": 1.9779851946075695e-07,
+      "loss": 0.7045,
+      "step": 2050
+    },
+    {
+      "epoch": 3.7981481481481483,
+      "grad_norm": 1.4179458361924653,
+      "learning_rate": 1.9779538944951e-07,
+      "loss": 0.6433,
+      "step": 2051
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 1.408718324432222,
+      "learning_rate": 1.9779225723955707e-07,
+      "loss": 0.7133,
+      "step": 2052
+    },
+    {
+      "epoch": 3.801851851851852,
+      "grad_norm": 1.705288026674248,
+      "learning_rate": 1.977891228309686e-07,
+      "loss": 0.7153,
+      "step": 2053
+    },
+    {
+      "epoch": 3.803703703703704,
+      "grad_norm": 1.5061348743499523,
+      "learning_rate": 1.9778598622381503e-07,
+      "loss": 0.7049,
+      "step": 2054
+    },
+    {
+      "epoch": 3.8055555555555554,
+      "grad_norm": 1.4265299231778683,
+      "learning_rate": 1.9778284741816692e-07,
+      "loss": 0.6546,
+      "step": 2055
+    },
+    {
+      "epoch": 3.8074074074074074,
+      "grad_norm": 1.3183671470338603,
+      "learning_rate": 1.9777970641409485e-07,
+      "loss": 0.7073,
+      "step": 2056
+    },
+    {
+      "epoch": 3.8092592592592593,
+      "grad_norm": 1.5032706601594712,
+      "learning_rate": 1.977765632116694e-07,
+      "loss": 0.7246,
+      "step": 2057
+    },
+    {
+      "epoch": 3.811111111111111,
+      "grad_norm": 1.3855457654491812,
+      "learning_rate": 1.9777341781096127e-07,
+      "loss": 0.734,
+      "step": 2058
+    },
+    {
+      "epoch": 3.812962962962963,
+      "grad_norm": 1.6538440986071663,
+      "learning_rate": 1.9777027021204112e-07,
+      "loss": 0.6826,
+      "step": 2059
+    },
+    {
+      "epoch": 3.814814814814815,
+      "grad_norm": 1.4233937776207124,
+      "learning_rate": 1.9776712041497977e-07,
+      "loss": 0.7046,
+      "step": 2060
+    },
+    {
+      "epoch": 3.8166666666666664,
+      "grad_norm": 2.088124984233603,
+      "learning_rate": 1.9776396841984804e-07,
+      "loss": 0.6707,
+      "step": 2061
+    },
+    {
+      "epoch": 3.8185185185185184,
+      "grad_norm": 1.5372764521498996,
+      "learning_rate": 1.9776081422671677e-07,
+      "loss": 0.7223,
+      "step": 2062
+    },
+    {
+      "epoch": 3.8203703703703704,
+      "grad_norm": 1.6259154880893407,
+      "learning_rate": 1.977576578356569e-07,
+      "loss": 0.731,
+      "step": 2063
+    },
+    {
+      "epoch": 3.822222222222222,
+      "grad_norm": 1.35635294757172,
+      "learning_rate": 1.9775449924673932e-07,
+      "loss": 0.6952,
+      "step": 2064
+    },
+    {
+      "epoch": 3.824074074074074,
+      "grad_norm": 1.412001533714791,
+      "learning_rate": 1.9775133846003515e-07,
+      "loss": 0.7235,
+      "step": 2065
+    },
+    {
+      "epoch": 3.825925925925926,
+      "grad_norm": 1.4062812757225207,
+      "learning_rate": 1.9774817547561538e-07,
+      "loss": 0.723,
+      "step": 2066
+    },
+    {
+      "epoch": 3.8277777777777775,
+      "grad_norm": 1.6023306834626792,
+      "learning_rate": 1.9774501029355116e-07,
+      "loss": 0.7002,
+      "step": 2067
+    },
+    {
+      "epoch": 3.8296296296296295,
+      "grad_norm": 1.4951150910164601,
+      "learning_rate": 1.9774184291391362e-07,
+      "loss": 0.704,
+      "step": 2068
+    },
+    {
+      "epoch": 3.8314814814814815,
+      "grad_norm": 1.5231815140564111,
+      "learning_rate": 1.97738673336774e-07,
+      "loss": 0.7131,
+      "step": 2069
+    },
+    {
+      "epoch": 3.8333333333333335,
+      "grad_norm": 1.5957682017314692,
+      "learning_rate": 1.9773550156220354e-07,
+      "loss": 0.6992,
+      "step": 2070
+    },
+    {
+      "epoch": 3.835185185185185,
+      "grad_norm": 2.0166025999711525,
+      "learning_rate": 1.9773232759027355e-07,
+      "loss": 0.6928,
+      "step": 2071
+    },
+    {
+      "epoch": 3.837037037037037,
+      "grad_norm": 1.571719310867269,
+      "learning_rate": 1.9772915142105542e-07,
+      "loss": 0.6729,
+      "step": 2072
+    },
+    {
+      "epoch": 3.838888888888889,
+      "grad_norm": 1.3956578120181538,
+      "learning_rate": 1.9772597305462054e-07,
+      "loss": 0.7294,
+      "step": 2073
+    },
+    {
+      "epoch": 3.840740740740741,
+      "grad_norm": 1.4021165919239085,
+      "learning_rate": 1.9772279249104037e-07,
+      "loss": 0.727,
+      "step": 2074
+    },
+    {
+      "epoch": 3.8425925925925926,
+      "grad_norm": 1.4928251728343354,
+      "learning_rate": 1.9771960973038638e-07,
+      "loss": 0.698,
+      "step": 2075
+    },
+    {
+      "epoch": 3.8444444444444446,
+      "grad_norm": 1.3185730537474556,
+      "learning_rate": 1.9771642477273018e-07,
+      "loss": 0.7006,
+      "step": 2076
+    },
+    {
+      "epoch": 3.8462962962962965,
+      "grad_norm": 1.2587927776882453,
+      "learning_rate": 1.9771323761814337e-07,
+      "loss": 0.7173,
+      "step": 2077
+    },
+    {
+      "epoch": 3.848148148148148,
+      "grad_norm": 1.4150570154501556,
+      "learning_rate": 1.9771004826669757e-07,
+      "loss": 0.6956,
+      "step": 2078
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 1.3471200353892092,
+      "learning_rate": 1.9770685671846453e-07,
+      "loss": 0.7201,
+      "step": 2079
+    },
+    {
+      "epoch": 3.851851851851852,
+      "grad_norm": 1.6364455422132873,
+      "learning_rate": 1.9770366297351598e-07,
+      "loss": 0.7436,
+      "step": 2080
+    },
+    {
+      "epoch": 3.8537037037037036,
+      "grad_norm": 1.5439290186514478,
+      "learning_rate": 1.9770046703192372e-07,
+      "loss": 0.7223,
+      "step": 2081
+    },
+    {
+      "epoch": 3.8555555555555556,
+      "grad_norm": 1.3750221231896445,
+      "learning_rate": 1.9769726889375962e-07,
+      "loss": 0.6436,
+      "step": 2082
+    },
+    {
+      "epoch": 3.8574074074074076,
+      "grad_norm": 1.4648862148395114,
+      "learning_rate": 1.9769406855909556e-07,
+      "loss": 0.6973,
+      "step": 2083
+    },
+    {
+      "epoch": 3.859259259259259,
+      "grad_norm": 1.5109337836145946,
+      "learning_rate": 1.9769086602800353e-07,
+      "loss": 0.713,
+      "step": 2084
+    },
+    {
+      "epoch": 3.861111111111111,
+      "grad_norm": 1.369789548190783,
+      "learning_rate": 1.9768766130055548e-07,
+      "loss": 0.7304,
+      "step": 2085
+    },
+    {
+      "epoch": 3.862962962962963,
+      "grad_norm": 1.4954309097823089,
+      "learning_rate": 1.976844543768235e-07,
+      "loss": 0.6516,
+      "step": 2086
+    },
+    {
+      "epoch": 3.8648148148148147,
+      "grad_norm": 1.3887369520444244,
+      "learning_rate": 1.9768124525687968e-07,
+      "loss": 0.7136,
+      "step": 2087
+    },
+    {
+      "epoch": 3.8666666666666667,
+      "grad_norm": 1.4408895085668612,
+      "learning_rate": 1.9767803394079615e-07,
+      "loss": 0.7069,
+      "step": 2088
+    },
+    {
+      "epoch": 3.8685185185185187,
+      "grad_norm": 1.9529059632425914,
+      "learning_rate": 1.9767482042864514e-07,
+      "loss": 0.6825,
+      "step": 2089
+    },
+    {
+      "epoch": 3.8703703703703702,
+      "grad_norm": 1.9033095152552635,
+      "learning_rate": 1.976716047204989e-07,
+      "loss": 0.6625,
+      "step": 2090
+    },
+    {
+      "epoch": 3.8722222222222222,
+      "grad_norm": 1.4464385013475276,
+      "learning_rate": 1.976683868164297e-07,
+      "loss": 0.6889,
+      "step": 2091
+    },
+    {
+      "epoch": 3.8740740740740742,
+      "grad_norm": 1.7461217796006685,
+      "learning_rate": 1.976651667165099e-07,
+      "loss": 0.687,
+      "step": 2092
+    },
+    {
+      "epoch": 3.8759259259259258,
+      "grad_norm": 1.5172342151832177,
+      "learning_rate": 1.976619444208119e-07,
+      "loss": 0.7089,
+      "step": 2093
+    },
+    {
+      "epoch": 3.8777777777777778,
+      "grad_norm": 1.9882867582913393,
+      "learning_rate": 1.9765871992940814e-07,
+      "loss": 0.7023,
+      "step": 2094
+    },
+    {
+      "epoch": 3.8796296296296298,
+      "grad_norm": 1.3665506673574688,
+      "learning_rate": 1.976554932423711e-07,
+      "loss": 0.687,
+      "step": 2095
+    },
+    {
+      "epoch": 3.8814814814814813,
+      "grad_norm": 1.39428078518463,
+      "learning_rate": 1.976522643597734e-07,
+      "loss": 0.7245,
+      "step": 2096
+    },
+    {
+      "epoch": 3.8833333333333333,
+      "grad_norm": 1.6418300696502135,
+      "learning_rate": 1.9764903328168754e-07,
+      "loss": 0.6876,
+      "step": 2097
+    },
+    {
+      "epoch": 3.8851851851851853,
+      "grad_norm": 1.4648251915385586,
+      "learning_rate": 1.9764580000818621e-07,
+      "loss": 0.7176,
+      "step": 2098
+    },
+    {
+      "epoch": 3.887037037037037,
+      "grad_norm": 1.4366369843777507,
+      "learning_rate": 1.9764256453934206e-07,
+      "loss": 0.6957,
+      "step": 2099
+    },
+    {
+      "epoch": 3.888888888888889,
+      "grad_norm": 1.532200763932793,
+      "learning_rate": 1.9763932687522793e-07,
+      "loss": 0.707,
+      "step": 2100
+    },
+    {
+      "epoch": 3.890740740740741,
+      "grad_norm": 1.754310847101926,
+      "learning_rate": 1.9763608701591652e-07,
+      "loss": 0.6561,
+      "step": 2101
+    },
+    {
+      "epoch": 3.8925925925925924,
+      "grad_norm": 1.7106571761424132,
+      "learning_rate": 1.9763284496148067e-07,
+      "loss": 0.7514,
+      "step": 2102
+    },
+    {
+      "epoch": 3.8944444444444444,
+      "grad_norm": 1.33344946077013,
+      "learning_rate": 1.9762960071199333e-07,
+      "loss": 0.6714,
+      "step": 2103
+    },
+    {
+      "epoch": 3.8962962962962964,
+      "grad_norm": 1.3480036027435867,
+      "learning_rate": 1.9762635426752738e-07,
+      "loss": 0.6785,
+      "step": 2104
+    },
+    {
+      "epoch": 3.898148148148148,
+      "grad_norm": 1.535732902513717,
+      "learning_rate": 1.9762310562815588e-07,
+      "loss": 0.7003,
+      "step": 2105
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 2.3342955607009253,
+      "learning_rate": 1.9761985479395177e-07,
+      "loss": 0.6553,
+      "step": 2106
+    },
+    {
+      "epoch": 3.901851851851852,
+      "grad_norm": 1.4494541848641362,
+      "learning_rate": 1.9761660176498825e-07,
+      "loss": 0.7365,
+      "step": 2107
+    },
+    {
+      "epoch": 3.9037037037037035,
+      "grad_norm": 1.5628178216199449,
+      "learning_rate": 1.9761334654133835e-07,
+      "loss": 0.7124,
+      "step": 2108
+    },
+    {
+      "epoch": 3.9055555555555554,
+      "grad_norm": 2.2697898769289284,
+      "learning_rate": 1.9761008912307533e-07,
+      "loss": 0.6823,
+      "step": 2109
+    },
+    {
+      "epoch": 3.9074074074074074,
+      "grad_norm": 1.781464621555349,
+      "learning_rate": 1.976068295102724e-07,
+      "loss": 0.665,
+      "step": 2110
+    },
+    {
+      "epoch": 3.909259259259259,
+      "grad_norm": 1.3942453117716922,
+      "learning_rate": 1.9760356770300285e-07,
+      "loss": 0.7116,
+      "step": 2111
+    },
+    {
+      "epoch": 3.911111111111111,
+      "grad_norm": 1.4152386658428937,
+      "learning_rate": 1.9760030370134e-07,
+      "loss": 0.6577,
+      "step": 2112
+    },
+    {
+      "epoch": 3.912962962962963,
+      "grad_norm": 1.4363132682118671,
+      "learning_rate": 1.9759703750535727e-07,
+      "loss": 0.7228,
+      "step": 2113
+    },
+    {
+      "epoch": 3.914814814814815,
+      "grad_norm": 1.856928142706703,
+      "learning_rate": 1.9759376911512805e-07,
+      "loss": 0.7051,
+      "step": 2114
+    },
+    {
+      "epoch": 3.9166666666666665,
+      "grad_norm": 1.5644776469398753,
+      "learning_rate": 1.9759049853072582e-07,
+      "loss": 0.6926,
+      "step": 2115
+    },
+    {
+      "epoch": 3.9185185185185185,
+      "grad_norm": 1.8180639546252566,
+      "learning_rate": 1.9758722575222416e-07,
+      "loss": 0.7115,
+      "step": 2116
+    },
+    {
+      "epoch": 3.9203703703703705,
+      "grad_norm": 1.5593164389046483,
+      "learning_rate": 1.9758395077969664e-07,
+      "loss": 0.7199,
+      "step": 2117
+    },
+    {
+      "epoch": 3.9222222222222225,
+      "grad_norm": 1.4475018495526304,
+      "learning_rate": 1.9758067361321681e-07,
+      "loss": 0.6925,
+      "step": 2118
+    },
+    {
+      "epoch": 3.924074074074074,
+      "grad_norm": 1.456726814484005,
+      "learning_rate": 1.9757739425285847e-07,
+      "loss": 0.6887,
+      "step": 2119
+    },
+    {
+      "epoch": 3.925925925925926,
+      "grad_norm": 1.45533907653281,
+      "learning_rate": 1.9757411269869527e-07,
+      "loss": 0.7021,
+      "step": 2120
+    },
+    {
+      "epoch": 3.927777777777778,
+      "grad_norm": 1.32284728498338,
+      "learning_rate": 1.97570828950801e-07,
+      "loss": 0.7028,
+      "step": 2121
+    },
+    {
+      "epoch": 3.9296296296296296,
+      "grad_norm": 1.7725585450558663,
+      "learning_rate": 1.975675430092495e-07,
+      "loss": 0.6658,
+      "step": 2122
+    },
+    {
+      "epoch": 3.9314814814814816,
+      "grad_norm": 1.4115818603619108,
+      "learning_rate": 1.9756425487411463e-07,
+      "loss": 0.6653,
+      "step": 2123
+    },
+    {
+      "epoch": 3.9333333333333336,
+      "grad_norm": 1.5111266002908923,
+      "learning_rate": 1.9756096454547038e-07,
+      "loss": 0.6809,
+      "step": 2124
+    },
+    {
+      "epoch": 3.935185185185185,
+      "grad_norm": 2.1484609200607245,
+      "learning_rate": 1.9755767202339063e-07,
+      "loss": 0.6965,
+      "step": 2125
+    },
+    {
+      "epoch": 3.937037037037037,
+      "grad_norm": 1.9629451950219259,
+      "learning_rate": 1.9755437730794946e-07,
+      "loss": 0.6781,
+      "step": 2126
+    },
+    {
+      "epoch": 3.938888888888889,
+      "grad_norm": 1.4073283972031392,
+      "learning_rate": 1.9755108039922094e-07,
+      "loss": 0.679,
+      "step": 2127
+    },
+    {
+      "epoch": 3.9407407407407407,
+      "grad_norm": 1.5044559613728987,
+      "learning_rate": 1.9754778129727917e-07,
+      "loss": 0.6975,
+      "step": 2128
+    },
+    {
+      "epoch": 3.9425925925925926,
+      "grad_norm": 1.4781379310574028,
+      "learning_rate": 1.9754448000219837e-07,
+      "loss": 0.6524,
+      "step": 2129
+    },
+    {
+      "epoch": 3.9444444444444446,
+      "grad_norm": 1.7175765794120874,
+      "learning_rate": 1.9754117651405271e-07,
+      "loss": 0.7085,
+      "step": 2130
+    },
+    {
+      "epoch": 3.946296296296296,
+      "grad_norm": 1.3991352643221722,
+      "learning_rate": 1.9753787083291652e-07,
+      "loss": 0.7439,
+      "step": 2131
+    },
+    {
+      "epoch": 3.948148148148148,
+      "grad_norm": 1.3898289272509385,
+      "learning_rate": 1.9753456295886403e-07,
+      "loss": 0.7004,
+      "step": 2132
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 1.549627942835458,
+      "learning_rate": 1.975312528919697e-07,
+      "loss": 0.6927,
+      "step": 2133
+    },
+    {
+      "epoch": 3.9518518518518517,
+      "grad_norm": 1.4378032393098668,
+      "learning_rate": 1.975279406323079e-07,
+      "loss": 0.6969,
+      "step": 2134
+    },
+    {
+      "epoch": 3.9537037037037037,
+      "grad_norm": 1.3791988285978989,
+      "learning_rate": 1.975246261799531e-07,
+      "loss": 0.6965,
+      "step": 2135
+    },
+    {
+      "epoch": 3.9555555555555557,
+      "grad_norm": 1.481814109095302,
+      "learning_rate": 1.9752130953497986e-07,
+      "loss": 0.6934,
+      "step": 2136
+    },
+    {
+      "epoch": 3.9574074074074073,
+      "grad_norm": 1.5090948159860287,
+      "learning_rate": 1.975179906974627e-07,
+      "loss": 0.7053,
+      "step": 2137
+    },
+    {
+      "epoch": 3.9592592592592593,
+      "grad_norm": 1.4225543850617801,
+      "learning_rate": 1.9751466966747628e-07,
+      "loss": 0.7022,
+      "step": 2138
+    },
+    {
+      "epoch": 3.9611111111111112,
+      "grad_norm": 1.6541347671954156,
+      "learning_rate": 1.975113464450952e-07,
+      "loss": 0.6823,
+      "step": 2139
+    },
+    {
+      "epoch": 3.962962962962963,
+      "grad_norm": 1.6594650372192261,
+      "learning_rate": 1.975080210303943e-07,
+      "loss": 0.719,
+      "step": 2140
+    },
+    {
+      "epoch": 3.964814814814815,
+      "grad_norm": 1.9309095868089667,
+      "learning_rate": 1.9750469342344815e-07,
+      "loss": 0.7146,
+      "step": 2141
+    },
+    {
+      "epoch": 3.966666666666667,
+      "grad_norm": 1.6333450369742117,
+      "learning_rate": 1.9750136362433174e-07,
+      "loss": 0.6571,
+      "step": 2142
+    },
+    {
+      "epoch": 3.9685185185185183,
+      "grad_norm": 1.7329951546550835,
+      "learning_rate": 1.9749803163311985e-07,
+      "loss": 0.6623,
+      "step": 2143
+    },
+    {
+      "epoch": 3.9703703703703703,
+      "grad_norm": 1.5532786104958647,
+      "learning_rate": 1.974946974498874e-07,
+      "loss": 0.7035,
+      "step": 2144
+    },
+    {
+      "epoch": 3.9722222222222223,
+      "grad_norm": 1.6323745480348946,
+      "learning_rate": 1.9749136107470936e-07,
+      "loss": 0.6795,
+      "step": 2145
+    },
+    {
+      "epoch": 3.974074074074074,
+      "grad_norm": 1.3989398802253061,
+      "learning_rate": 1.9748802250766075e-07,
+      "loss": 0.6797,
+      "step": 2146
+    },
+    {
+      "epoch": 3.975925925925926,
+      "grad_norm": 1.7248875519213498,
+      "learning_rate": 1.974846817488166e-07,
+      "loss": 0.7201,
+      "step": 2147
+    },
+    {
+      "epoch": 3.977777777777778,
+      "grad_norm": 1.3932903335531583,
+      "learning_rate": 1.9748133879825206e-07,
+      "loss": 0.6729,
+      "step": 2148
+    },
+    {
+      "epoch": 3.9796296296296294,
+      "grad_norm": 1.5604056151476517,
+      "learning_rate": 1.9747799365604228e-07,
+      "loss": 0.6769,
+      "step": 2149
+    },
+    {
+      "epoch": 3.9814814814814814,
+      "grad_norm": 1.548387860802549,
+      "learning_rate": 1.9747464632226242e-07,
+      "loss": 0.7067,
+      "step": 2150
+    },
+    {
+      "epoch": 3.9833333333333334,
+      "grad_norm": 1.5999215115914402,
+      "learning_rate": 1.9747129679698783e-07,
+      "loss": 0.7391,
+      "step": 2151
+    },
+    {
+      "epoch": 3.985185185185185,
+      "grad_norm": 1.2913727032002287,
+      "learning_rate": 1.9746794508029368e-07,
+      "loss": 0.6915,
+      "step": 2152
+    },
+    {
+      "epoch": 3.987037037037037,
+      "grad_norm": 1.7733173594819152,
+      "learning_rate": 1.9746459117225548e-07,
+      "loss": 0.6658,
+      "step": 2153
+    },
+    {
+      "epoch": 3.988888888888889,
+      "grad_norm": 1.7798418460077214,
+      "learning_rate": 1.9746123507294852e-07,
+      "loss": 0.6677,
+      "step": 2154
+    },
+    {
+      "epoch": 3.9907407407407405,
+      "grad_norm": 1.5272738685494236,
+      "learning_rate": 1.974578767824483e-07,
+      "loss": 0.684,
+      "step": 2155
+    },
+    {
+      "epoch": 3.9925925925925925,
+      "grad_norm": 1.6031112127581029,
+      "learning_rate": 1.9745451630083032e-07,
+      "loss": 0.6936,
+      "step": 2156
+    },
+    {
+      "epoch": 3.9944444444444445,
+      "grad_norm": 1.5080577083063602,
+      "learning_rate": 1.9745115362817013e-07,
+      "loss": 0.6984,
+      "step": 2157
+    },
+    {
+      "epoch": 3.9962962962962965,
+      "grad_norm": 1.5029217673104314,
+      "learning_rate": 1.9744778876454333e-07,
+      "loss": 0.7014,
+      "step": 2158
+    },
+    {
+      "epoch": 3.998148148148148,
+      "grad_norm": 1.5298356743188082,
+      "learning_rate": 1.974444217100256e-07,
+      "loss": 0.7117,
+      "step": 2159
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.6547944841668067,
+      "learning_rate": 1.974410524646926e-07,
+      "loss": 0.6649,
+      "step": 2160
+    },
+    {
+      "epoch": 4.0018518518518515,
+      "grad_norm": 1.4615931784732643,
+      "learning_rate": 1.9743768102862008e-07,
+      "loss": 0.7127,
+      "step": 2161
+    },
+    {
+      "epoch": 4.003703703703704,
+      "grad_norm": 1.6981235427227295,
+      "learning_rate": 1.9743430740188388e-07,
+      "loss": 0.7041,
+      "step": 2162
+    },
+    {
+      "epoch": 4.0055555555555555,
+      "grad_norm": 1.579588220028123,
+      "learning_rate": 1.9743093158455978e-07,
+      "loss": 0.681,
+      "step": 2163
+    },
+    {
+      "epoch": 4.007407407407407,
+      "grad_norm": 1.6718687893799213,
+      "learning_rate": 1.9742755357672375e-07,
+      "loss": 0.7043,
+      "step": 2164
+    },
+    {
+      "epoch": 4.0092592592592595,
+      "grad_norm": 1.5367334488998505,
+      "learning_rate": 1.9742417337845173e-07,
+      "loss": 0.7381,
+      "step": 2165
+    },
+    {
+      "epoch": 4.011111111111111,
+      "grad_norm": 1.5151069959415868,
+      "learning_rate": 1.9742079098981966e-07,
+      "loss": 0.6807,
+      "step": 2166
+    },
+    {
+      "epoch": 4.012962962962963,
+      "grad_norm": 4.656958498627862,
+      "learning_rate": 1.9741740641090364e-07,
+      "loss": 0.6555,
+      "step": 2167
+    },
+    {
+      "epoch": 4.014814814814815,
+      "grad_norm": 1.3947326597756573,
+      "learning_rate": 1.974140196417797e-07,
+      "loss": 0.7133,
+      "step": 2168
+    },
+    {
+      "epoch": 4.016666666666667,
+      "grad_norm": 1.4680308663458679,
+      "learning_rate": 1.9741063068252406e-07,
+      "loss": 0.6907,
+      "step": 2169
+    },
+    {
+      "epoch": 4.018518518518518,
+      "grad_norm": 1.5330549305361343,
+      "learning_rate": 1.974072395332129e-07,
+      "loss": 0.6882,
+      "step": 2170
+    },
+    {
+      "epoch": 4.020370370370371,
+      "grad_norm": 1.4826289867980418,
+      "learning_rate": 1.9740384619392238e-07,
+      "loss": 0.6796,
+      "step": 2171
+    },
+    {
+      "epoch": 4.022222222222222,
+      "grad_norm": 2.895970942377341,
+      "learning_rate": 1.9740045066472888e-07,
+      "loss": 0.6617,
+      "step": 2172
+    },
+    {
+      "epoch": 4.024074074074074,
+      "grad_norm": 1.715138901443854,
+      "learning_rate": 1.9739705294570872e-07,
+      "loss": 0.6921,
+      "step": 2173
+    },
+    {
+      "epoch": 4.025925925925926,
+      "grad_norm": 1.661284148376113,
+      "learning_rate": 1.9739365303693827e-07,
+      "loss": 0.677,
+      "step": 2174
+    },
+    {
+      "epoch": 4.027777777777778,
+      "grad_norm": 1.401716886576476,
+      "learning_rate": 1.97390250938494e-07,
+      "loss": 0.6531,
+      "step": 2175
+    },
+    {
+      "epoch": 4.029629629629629,
+      "grad_norm": 1.468490670459241,
+      "learning_rate": 1.9738684665045237e-07,
+      "loss": 0.7035,
+      "step": 2176
+    },
+    {
+      "epoch": 4.031481481481482,
+      "grad_norm": 1.5817868846639522,
+      "learning_rate": 1.9738344017288996e-07,
+      "loss": 0.6696,
+      "step": 2177
+    },
+    {
+      "epoch": 4.033333333333333,
+      "grad_norm": 1.671296048466896,
+      "learning_rate": 1.973800315058833e-07,
+      "loss": 0.6557,
+      "step": 2178
+    },
+    {
+      "epoch": 4.035185185185185,
+      "grad_norm": 1.452127251710725,
+      "learning_rate": 1.9737662064950905e-07,
+      "loss": 0.7101,
+      "step": 2179
+    },
+    {
+      "epoch": 4.037037037037037,
+      "grad_norm": 1.4707461133550865,
+      "learning_rate": 1.9737320760384387e-07,
+      "loss": 0.7459,
+      "step": 2180
+    },
+    {
+      "epoch": 4.038888888888889,
+      "grad_norm": 1.4109725872193846,
+      "learning_rate": 1.9736979236896457e-07,
+      "loss": 0.6851,
+      "step": 2181
+    },
+    {
+      "epoch": 4.040740740740741,
+      "grad_norm": 1.2909741955074516,
+      "learning_rate": 1.9736637494494787e-07,
+      "loss": 0.6944,
+      "step": 2182
+    },
+    {
+      "epoch": 4.042592592592593,
+      "grad_norm": 1.4539069373199944,
+      "learning_rate": 1.9736295533187062e-07,
+      "loss": 0.6853,
+      "step": 2183
+    },
+    {
+      "epoch": 4.044444444444444,
+      "grad_norm": 1.4381460541649138,
+      "learning_rate": 1.9735953352980967e-07,
+      "loss": 0.6961,
+      "step": 2184
+    },
+    {
+      "epoch": 4.046296296296297,
+      "grad_norm": 1.5695544992835817,
+      "learning_rate": 1.97356109538842e-07,
+      "loss": 0.7177,
+      "step": 2185
+    },
+    {
+      "epoch": 4.048148148148148,
+      "grad_norm": 1.6211469964245664,
+      "learning_rate": 1.973526833590446e-07,
+      "loss": 0.6877,
+      "step": 2186
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 1.2867753143903538,
+      "learning_rate": 1.9734925499049444e-07,
+      "loss": 0.6489,
+      "step": 2187
+    },
+    {
+      "epoch": 4.051851851851852,
+      "grad_norm": 1.909640708661203,
+      "learning_rate": 1.9734582443326862e-07,
+      "loss": 0.6674,
+      "step": 2188
+    },
+    {
+      "epoch": 4.053703703703704,
+      "grad_norm": 2.1438291234846467,
+      "learning_rate": 1.9734239168744433e-07,
+      "loss": 0.7147,
+      "step": 2189
+    },
+    {
+      "epoch": 4.055555555555555,
+      "grad_norm": 2.0191059311475974,
+      "learning_rate": 1.9733895675309865e-07,
+      "loss": 0.6962,
+      "step": 2190
+    },
+    {
+      "epoch": 4.057407407407408,
+      "grad_norm": 1.3273764517973292,
+      "learning_rate": 1.973355196303089e-07,
+      "loss": 0.7137,
+      "step": 2191
+    },
+    {
+      "epoch": 4.059259259259259,
+      "grad_norm": 1.403757029300068,
+      "learning_rate": 1.9733208031915226e-07,
+      "loss": 0.6906,
+      "step": 2192
+    },
+    {
+      "epoch": 4.061111111111111,
+      "grad_norm": 1.3260719766271234,
+      "learning_rate": 1.9732863881970612e-07,
+      "loss": 0.679,
+      "step": 2193
+    },
+    {
+      "epoch": 4.062962962962963,
+      "grad_norm": 1.7334559671238368,
+      "learning_rate": 1.9732519513204785e-07,
+      "loss": 0.6547,
+      "step": 2194
+    },
+    {
+      "epoch": 4.064814814814815,
+      "grad_norm": 1.4306997267146155,
+      "learning_rate": 1.9732174925625485e-07,
+      "loss": 0.6502,
+      "step": 2195
+    },
+    {
+      "epoch": 4.066666666666666,
+      "grad_norm": 1.7511762120130396,
+      "learning_rate": 1.9731830119240462e-07,
+      "loss": 0.6647,
+      "step": 2196
+    },
+    {
+      "epoch": 4.068518518518519,
+      "grad_norm": 1.6338369370694825,
+      "learning_rate": 1.9731485094057466e-07,
+      "loss": 0.7323,
+      "step": 2197
+    },
+    {
+      "epoch": 4.07037037037037,
+      "grad_norm": 2.1227467076699456,
+      "learning_rate": 1.9731139850084254e-07,
+      "loss": 0.7332,
+      "step": 2198
+    },
+    {
+      "epoch": 4.072222222222222,
+      "grad_norm": 1.5372721559203275,
+      "learning_rate": 1.9730794387328587e-07,
+      "loss": 0.6605,
+      "step": 2199
+    },
+    {
+      "epoch": 4.074074074074074,
+      "grad_norm": 1.5656828327346766,
+      "learning_rate": 1.9730448705798237e-07,
+      "loss": 0.6717,
+      "step": 2200
+    },
+    {
+      "epoch": 4.075925925925926,
+      "grad_norm": 2.1763961935398024,
+      "learning_rate": 1.9730102805500973e-07,
+      "loss": 0.6928,
+      "step": 2201
+    },
+    {
+      "epoch": 4.0777777777777775,
+      "grad_norm": 1.5406319524000287,
+      "learning_rate": 1.972975668644457e-07,
+      "loss": 0.7133,
+      "step": 2202
+    },
+    {
+      "epoch": 4.07962962962963,
+      "grad_norm": 1.5334006906455309,
+      "learning_rate": 1.9729410348636811e-07,
+      "loss": 0.7402,
+      "step": 2203
+    },
+    {
+      "epoch": 4.0814814814814815,
+      "grad_norm": 1.3507887777965895,
+      "learning_rate": 1.972906379208548e-07,
+      "loss": 0.6864,
+      "step": 2204
+    },
+    {
+      "epoch": 4.083333333333333,
+      "grad_norm": 1.4709371853705544,
+      "learning_rate": 1.9728717016798378e-07,
+      "loss": 0.728,
+      "step": 2205
+    },
+    {
+      "epoch": 4.0851851851851855,
+      "grad_norm": 1.6176567857381396,
+      "learning_rate": 1.9728370022783286e-07,
+      "loss": 0.6588,
+      "step": 2206
+    },
+    {
+      "epoch": 4.087037037037037,
+      "grad_norm": 1.4503358353816915,
+      "learning_rate": 1.9728022810048022e-07,
+      "loss": 0.6808,
+      "step": 2207
+    },
+    {
+      "epoch": 4.088888888888889,
+      "grad_norm": 1.3604464627227404,
+      "learning_rate": 1.972767537860038e-07,
+      "loss": 0.7081,
+      "step": 2208
+    },
+    {
+      "epoch": 4.090740740740741,
+      "grad_norm": 1.4538034743653498,
+      "learning_rate": 1.9727327728448175e-07,
+      "loss": 0.6807,
+      "step": 2209
+    },
+    {
+      "epoch": 4.092592592592593,
+      "grad_norm": 1.4474918488198876,
+      "learning_rate": 1.9726979859599226e-07,
+      "loss": 0.696,
+      "step": 2210
+    },
+    {
+      "epoch": 4.094444444444444,
+      "grad_norm": 1.6549450936720331,
+      "learning_rate": 1.9726631772061352e-07,
+      "loss": 0.6814,
+      "step": 2211
+    },
+    {
+      "epoch": 4.0962962962962965,
+      "grad_norm": 1.3512251985750467,
+      "learning_rate": 1.9726283465842377e-07,
+      "loss": 0.7127,
+      "step": 2212
+    },
+    {
+      "epoch": 4.098148148148148,
+      "grad_norm": 1.327772691807005,
+      "learning_rate": 1.9725934940950136e-07,
+      "loss": 0.691,
+      "step": 2213
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 1.7885523371542662,
+      "learning_rate": 1.972558619739246e-07,
+      "loss": 0.6471,
+      "step": 2214
+    },
+    {
+      "epoch": 4.101851851851852,
+      "grad_norm": 1.5540435558714403,
+      "learning_rate": 1.9725237235177193e-07,
+      "loss": 0.7084,
+      "step": 2215
+    },
+    {
+      "epoch": 4.103703703703704,
+      "grad_norm": 1.6317095980122032,
+      "learning_rate": 1.9724888054312177e-07,
+      "loss": 0.692,
+      "step": 2216
+    },
+    {
+      "epoch": 4.105555555555555,
+      "grad_norm": 1.5196913536493002,
+      "learning_rate": 1.9724538654805267e-07,
+      "loss": 0.6981,
+      "step": 2217
+    },
+    {
+      "epoch": 4.107407407407408,
+      "grad_norm": 1.390276381124955,
+      "learning_rate": 1.9724189036664317e-07,
+      "loss": 0.6412,
+      "step": 2218
+    },
+    {
+      "epoch": 4.109259259259259,
+      "grad_norm": 1.619629918950123,
+      "learning_rate": 1.972383919989719e-07,
+      "loss": 0.6927,
+      "step": 2219
+    },
+    {
+      "epoch": 4.111111111111111,
+      "grad_norm": 1.3732543235182741,
+      "learning_rate": 1.972348914451174e-07,
+      "loss": 0.6737,
+      "step": 2220
+    },
+    {
+      "epoch": 4.112962962962963,
+      "grad_norm": 1.380189554354143,
+      "learning_rate": 1.972313887051585e-07,
+      "loss": 0.6825,
+      "step": 2221
+    },
+    {
+      "epoch": 4.114814814814815,
+      "grad_norm": 1.490360712284871,
+      "learning_rate": 1.9722788377917393e-07,
+      "loss": 0.7127,
+      "step": 2222
+    },
+    {
+      "epoch": 4.116666666666666,
+      "grad_norm": 1.5537586492854871,
+      "learning_rate": 1.9722437666724243e-07,
+      "loss": 0.7089,
+      "step": 2223
+    },
+    {
+      "epoch": 4.118518518518519,
+      "grad_norm": 1.4089764960492903,
+      "learning_rate": 1.972208673694429e-07,
+      "loss": 0.7038,
+      "step": 2224
+    },
+    {
+      "epoch": 4.12037037037037,
+      "grad_norm": 1.6411428446735306,
+      "learning_rate": 1.9721735588585422e-07,
+      "loss": 0.6955,
+      "step": 2225
+    },
+    {
+      "epoch": 4.122222222222222,
+      "grad_norm": 1.4940716163272492,
+      "learning_rate": 1.9721384221655535e-07,
+      "loss": 0.6954,
+      "step": 2226
+    },
+    {
+      "epoch": 4.124074074074074,
+      "grad_norm": 2.131987017236616,
+      "learning_rate": 1.9721032636162528e-07,
+      "loss": 0.6622,
+      "step": 2227
+    },
+    {
+      "epoch": 4.125925925925926,
+      "grad_norm": 1.4513611427607525,
+      "learning_rate": 1.9720680832114303e-07,
+      "loss": 0.7035,
+      "step": 2228
+    },
+    {
+      "epoch": 4.127777777777778,
+      "grad_norm": 1.4567607267379143,
+      "learning_rate": 1.9720328809518777e-07,
+      "loss": 0.6919,
+      "step": 2229
+    },
+    {
+      "epoch": 4.12962962962963,
+      "grad_norm": 1.6848523295755233,
+      "learning_rate": 1.9719976568383854e-07,
+      "loss": 0.7328,
+      "step": 2230
+    },
+    {
+      "epoch": 4.131481481481481,
+      "grad_norm": 1.4987194252598426,
+      "learning_rate": 1.9719624108717458e-07,
+      "loss": 0.6528,
+      "step": 2231
+    },
+    {
+      "epoch": 4.133333333333334,
+      "grad_norm": 1.4437971921970107,
+      "learning_rate": 1.9719271430527518e-07,
+      "loss": 0.6745,
+      "step": 2232
+    },
+    {
+      "epoch": 4.135185185185185,
+      "grad_norm": 1.5568687632915494,
+      "learning_rate": 1.9718918533821958e-07,
+      "loss": 0.6705,
+      "step": 2233
+    },
+    {
+      "epoch": 4.137037037037037,
+      "grad_norm": 1.4253078383756228,
+      "learning_rate": 1.9718565418608712e-07,
+      "loss": 0.651,
+      "step": 2234
+    },
+    {
+      "epoch": 4.138888888888889,
+      "grad_norm": 1.3987986571385564,
+      "learning_rate": 1.9718212084895723e-07,
+      "loss": 0.616,
+      "step": 2235
+    },
+    {
+      "epoch": 4.140740740740741,
+      "grad_norm": 1.5825004287965694,
+      "learning_rate": 1.9717858532690927e-07,
+      "loss": 0.702,
+      "step": 2236
+    },
+    {
+      "epoch": 4.142592592592592,
+      "grad_norm": 2.9471120870232492,
+      "learning_rate": 1.9717504762002283e-07,
+      "loss": 0.7082,
+      "step": 2237
+    },
+    {
+      "epoch": 4.144444444444445,
+      "grad_norm": 1.4925028987116182,
+      "learning_rate": 1.9717150772837737e-07,
+      "loss": 0.656,
+      "step": 2238
+    },
+    {
+      "epoch": 4.146296296296296,
+      "grad_norm": 1.2629375992874476,
+      "learning_rate": 1.9716796565205252e-07,
+      "loss": 0.6757,
+      "step": 2239
+    },
+    {
+      "epoch": 4.148148148148148,
+      "grad_norm": 1.456162169204436,
+      "learning_rate": 1.971644213911279e-07,
+      "loss": 0.6712,
+      "step": 2240
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 2.0523464747360287,
+      "learning_rate": 1.9716087494568314e-07,
+      "loss": 0.7397,
+      "step": 2241
+    },
+    {
+      "epoch": 4.151851851851852,
+      "grad_norm": 1.6850372818349062,
+      "learning_rate": 1.971573263157981e-07,
+      "loss": 0.7001,
+      "step": 2242
+    },
+    {
+      "epoch": 4.1537037037037035,
+      "grad_norm": 1.3856317262839872,
+      "learning_rate": 1.9715377550155243e-07,
+      "loss": 0.6192,
+      "step": 2243
+    },
+    {
+      "epoch": 4.155555555555556,
+      "grad_norm": 2.027055331601776,
+      "learning_rate": 1.9715022250302605e-07,
+      "loss": 0.6701,
+      "step": 2244
+    },
+    {
+      "epoch": 4.157407407407407,
+      "grad_norm": 1.4407227930368782,
+      "learning_rate": 1.971466673202988e-07,
+      "loss": 0.6737,
+      "step": 2245
+    },
+    {
+      "epoch": 4.159259259259259,
+      "grad_norm": 1.369143827499616,
+      "learning_rate": 1.9714310995345063e-07,
+      "loss": 0.7217,
+      "step": 2246
+    },
+    {
+      "epoch": 4.161111111111111,
+      "grad_norm": 1.5866031852302096,
+      "learning_rate": 1.9713955040256153e-07,
+      "loss": 0.6896,
+      "step": 2247
+    },
+    {
+      "epoch": 4.162962962962963,
+      "grad_norm": 2.3982661927193547,
+      "learning_rate": 1.9713598866771148e-07,
+      "loss": 0.6616,
+      "step": 2248
+    },
+    {
+      "epoch": 4.1648148148148145,
+      "grad_norm": 1.5310348420037856,
+      "learning_rate": 1.971324247489806e-07,
+      "loss": 0.6302,
+      "step": 2249
+    },
+    {
+      "epoch": 4.166666666666667,
+      "grad_norm": 1.3852294639169036,
+      "learning_rate": 1.9712885864644899e-07,
+      "loss": 0.7009,
+      "step": 2250
+    },
+    {
+      "epoch": 4.1685185185185185,
+      "grad_norm": 1.949168338344836,
+      "learning_rate": 1.9712529036019685e-07,
+      "loss": 0.6668,
+      "step": 2251
+    },
+    {
+      "epoch": 4.17037037037037,
+      "grad_norm": 1.403538975222234,
+      "learning_rate": 1.971217198903044e-07,
+      "loss": 0.6814,
+      "step": 2252
+    },
+    {
+      "epoch": 4.1722222222222225,
+      "grad_norm": 1.4063020007965452,
+      "learning_rate": 1.971181472368519e-07,
+      "loss": 0.6849,
+      "step": 2253
+    },
+    {
+      "epoch": 4.174074074074074,
+      "grad_norm": 1.502607814919187,
+      "learning_rate": 1.9711457239991968e-07,
+      "loss": 0.6965,
+      "step": 2254
+    },
+    {
+      "epoch": 4.175925925925926,
+      "grad_norm": 1.4549386071020158,
+      "learning_rate": 1.9711099537958816e-07,
+      "loss": 0.6769,
+      "step": 2255
+    },
+    {
+      "epoch": 4.177777777777778,
+      "grad_norm": 1.6408425489752825,
+      "learning_rate": 1.9710741617593766e-07,
+      "loss": 0.6872,
+      "step": 2256
+    },
+    {
+      "epoch": 4.17962962962963,
+      "grad_norm": 1.3878875237278199,
+      "learning_rate": 1.971038347890487e-07,
+      "loss": 0.7287,
+      "step": 2257
+    },
+    {
+      "epoch": 4.181481481481481,
+      "grad_norm": 1.4814270605062365,
+      "learning_rate": 1.9710025121900182e-07,
+      "loss": 0.6812,
+      "step": 2258
+    },
+    {
+      "epoch": 4.183333333333334,
+      "grad_norm": 1.6718393882272573,
+      "learning_rate": 1.9709666546587758e-07,
+      "loss": 0.703,
+      "step": 2259
+    },
+    {
+      "epoch": 4.185185185185185,
+      "grad_norm": 1.5794060442929199,
+      "learning_rate": 1.9709307752975658e-07,
+      "loss": 0.6779,
+      "step": 2260
+    },
+    {
+      "epoch": 4.187037037037037,
+      "grad_norm": 1.4340291266180283,
+      "learning_rate": 1.9708948741071952e-07,
+      "loss": 0.7084,
+      "step": 2261
+    },
+    {
+      "epoch": 4.188888888888889,
+      "grad_norm": 1.4022833261570278,
+      "learning_rate": 1.9708589510884708e-07,
+      "loss": 0.6374,
+      "step": 2262
+    },
+    {
+      "epoch": 4.190740740740741,
+      "grad_norm": 1.6988093889320932,
+      "learning_rate": 1.9708230062422002e-07,
+      "loss": 0.6267,
+      "step": 2263
+    },
+    {
+      "epoch": 4.192592592592592,
+      "grad_norm": 2.7673642160659644,
+      "learning_rate": 1.9707870395691916e-07,
+      "loss": 0.7112,
+      "step": 2264
+    },
+    {
+      "epoch": 4.194444444444445,
+      "grad_norm": 1.4308799160643817,
+      "learning_rate": 1.970751051070254e-07,
+      "loss": 0.7114,
+      "step": 2265
+    },
+    {
+      "epoch": 4.196296296296296,
+      "grad_norm": 1.479304820147132,
+      "learning_rate": 1.9707150407461964e-07,
+      "loss": 0.6767,
+      "step": 2266
+    },
+    {
+      "epoch": 4.198148148148148,
+      "grad_norm": 1.581126447377077,
+      "learning_rate": 1.970679008597828e-07,
+      "loss": 0.6829,
+      "step": 2267
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 1.6758079381412818,
+      "learning_rate": 1.970642954625959e-07,
+      "loss": 0.7111,
+      "step": 2268
+    },
+    {
+      "epoch": 4.201851851851852,
+      "grad_norm": 1.7290307164664305,
+      "learning_rate": 1.9706068788314005e-07,
+      "loss": 0.6948,
+      "step": 2269
+    },
+    {
+      "epoch": 4.203703703703703,
+      "grad_norm": 2.5065337782890453,
+      "learning_rate": 1.970570781214963e-07,
+      "loss": 0.7169,
+      "step": 2270
+    },
+    {
+      "epoch": 4.205555555555556,
+      "grad_norm": 1.4876006766849879,
+      "learning_rate": 1.9705346617774582e-07,
+      "loss": 0.7442,
+      "step": 2271
+    },
+    {
+      "epoch": 4.207407407407407,
+      "grad_norm": 1.6868008505828358,
+      "learning_rate": 1.9704985205196987e-07,
+      "loss": 0.7037,
+      "step": 2272
+    },
+    {
+      "epoch": 4.209259259259259,
+      "grad_norm": 1.7897595947542868,
+      "learning_rate": 1.9704623574424964e-07,
+      "loss": 0.7647,
+      "step": 2273
+    },
+    {
+      "epoch": 4.211111111111111,
+      "grad_norm": 1.4062866460885806,
+      "learning_rate": 1.9704261725466648e-07,
+      "loss": 0.7046,
+      "step": 2274
+    },
+    {
+      "epoch": 4.212962962962963,
+      "grad_norm": 2.4129085814698557,
+      "learning_rate": 1.9703899658330168e-07,
+      "loss": 0.7162,
+      "step": 2275
+    },
+    {
+      "epoch": 4.214814814814815,
+      "grad_norm": 1.9701160169061898,
+      "learning_rate": 1.9703537373023668e-07,
+      "loss": 0.6995,
+      "step": 2276
+    },
+    {
+      "epoch": 4.216666666666667,
+      "grad_norm": 1.3513334182313512,
+      "learning_rate": 1.9703174869555296e-07,
+      "loss": 0.7123,
+      "step": 2277
+    },
+    {
+      "epoch": 4.218518518518518,
+      "grad_norm": 1.7334622199352077,
+      "learning_rate": 1.97028121479332e-07,
+      "loss": 0.6776,
+      "step": 2278
+    },
+    {
+      "epoch": 4.220370370370371,
+      "grad_norm": 2.191188979558819,
+      "learning_rate": 1.9702449208165532e-07,
+      "loss": 0.7168,
+      "step": 2279
+    },
+    {
+      "epoch": 4.222222222222222,
+      "grad_norm": 1.50460988072586,
+      "learning_rate": 1.9702086050260454e-07,
+      "loss": 0.7014,
+      "step": 2280
+    },
+    {
+      "epoch": 4.224074074074074,
+      "grad_norm": 1.6722983569725687,
+      "learning_rate": 1.9701722674226135e-07,
+      "loss": 0.6265,
+      "step": 2281
+    },
+    {
+      "epoch": 4.225925925925926,
+      "grad_norm": 1.5421617775623717,
+      "learning_rate": 1.9701359080070738e-07,
+      "loss": 0.6733,
+      "step": 2282
+    },
+    {
+      "epoch": 4.227777777777778,
+      "grad_norm": 1.3824095268494505,
+      "learning_rate": 1.9700995267802444e-07,
+      "loss": 0.68,
+      "step": 2283
+    },
+    {
+      "epoch": 4.229629629629629,
+      "grad_norm": 1.415004352752854,
+      "learning_rate": 1.9700631237429426e-07,
+      "loss": 0.7003,
+      "step": 2284
+    },
+    {
+      "epoch": 4.231481481481482,
+      "grad_norm": 1.56431904802866,
+      "learning_rate": 1.970026698895987e-07,
+      "loss": 0.6722,
+      "step": 2285
+    },
+    {
+      "epoch": 4.233333333333333,
+      "grad_norm": 1.5869417264614845,
+      "learning_rate": 1.969990252240197e-07,
+      "loss": 0.6537,
+      "step": 2286
+    },
+    {
+      "epoch": 4.235185185185185,
+      "grad_norm": 1.4883632494014178,
+      "learning_rate": 1.9699537837763912e-07,
+      "loss": 0.6926,
+      "step": 2287
+    },
+    {
+      "epoch": 4.237037037037037,
+      "grad_norm": 1.4891195908400743,
+      "learning_rate": 1.9699172935053902e-07,
+      "loss": 0.6835,
+      "step": 2288
+    },
+    {
+      "epoch": 4.238888888888889,
+      "grad_norm": 1.3910281952779788,
+      "learning_rate": 1.9698807814280145e-07,
+      "loss": 0.6574,
+      "step": 2289
+    },
+    {
+      "epoch": 4.2407407407407405,
+      "grad_norm": 1.6819731681986096,
+      "learning_rate": 1.9698442475450843e-07,
+      "loss": 0.6688,
+      "step": 2290
+    },
+    {
+      "epoch": 4.242592592592593,
+      "grad_norm": 1.5872465106139797,
+      "learning_rate": 1.9698076918574214e-07,
+      "loss": 0.6432,
+      "step": 2291
+    },
+    {
+      "epoch": 4.2444444444444445,
+      "grad_norm": 1.4686013989363231,
+      "learning_rate": 1.9697711143658477e-07,
+      "loss": 0.7397,
+      "step": 2292
+    },
+    {
+      "epoch": 4.246296296296296,
+      "grad_norm": 2.4159677052324504,
+      "learning_rate": 1.9697345150711854e-07,
+      "loss": 0.6663,
+      "step": 2293
+    },
+    {
+      "epoch": 4.2481481481481485,
+      "grad_norm": 4.980204864651794,
+      "learning_rate": 1.9696978939742577e-07,
+      "loss": 0.675,
+      "step": 2294
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 1.4060698365820736,
+      "learning_rate": 1.9696612510758874e-07,
+      "loss": 0.7257,
+      "step": 2295
+    },
+    {
+      "epoch": 4.2518518518518515,
+      "grad_norm": 1.4089277172815022,
+      "learning_rate": 1.9696245863768987e-07,
+      "loss": 0.6574,
+      "step": 2296
+    },
+    {
+      "epoch": 4.253703703703704,
+      "grad_norm": 1.6803499585189763,
+      "learning_rate": 1.9695878998781158e-07,
+      "loss": 0.6952,
+      "step": 2297
+    },
+    {
+      "epoch": 4.2555555555555555,
+      "grad_norm": 1.5430778345737401,
+      "learning_rate": 1.9695511915803638e-07,
+      "loss": 0.6854,
+      "step": 2298
+    },
+    {
+      "epoch": 4.257407407407407,
+      "grad_norm": 1.5882620877329388,
+      "learning_rate": 1.9695144614844676e-07,
+      "loss": 0.6717,
+      "step": 2299
+    },
+    {
+      "epoch": 4.2592592592592595,
+      "grad_norm": 1.4311850445392638,
+      "learning_rate": 1.9694777095912533e-07,
+      "loss": 0.6974,
+      "step": 2300
+    },
+    {
+      "epoch": 4.261111111111111,
+      "grad_norm": 1.3753003535916093,
+      "learning_rate": 1.9694409359015465e-07,
+      "loss": 0.6978,
+      "step": 2301
+    },
+    {
+      "epoch": 4.262962962962963,
+      "grad_norm": 1.6863316782220699,
+      "learning_rate": 1.969404140416175e-07,
+      "loss": 0.727,
+      "step": 2302
+    },
+    {
+      "epoch": 4.264814814814815,
+      "grad_norm": 1.456873810713547,
+      "learning_rate": 1.9693673231359652e-07,
+      "loss": 0.6628,
+      "step": 2303
+    },
+    {
+      "epoch": 4.266666666666667,
+      "grad_norm": 1.6201139473330586,
+      "learning_rate": 1.9693304840617456e-07,
+      "loss": 0.6928,
+      "step": 2304
+    },
+    {
+      "epoch": 4.268518518518518,
+      "grad_norm": 1.4839274081238842,
+      "learning_rate": 1.9692936231943438e-07,
+      "loss": 0.721,
+      "step": 2305
+    },
+    {
+      "epoch": 4.270370370370371,
+      "grad_norm": 1.3895999291296606,
+      "learning_rate": 1.9692567405345888e-07,
+      "loss": 0.7082,
+      "step": 2306
+    },
+    {
+      "epoch": 4.272222222222222,
+      "grad_norm": 1.3761618374147762,
+      "learning_rate": 1.9692198360833098e-07,
+      "loss": 0.6738,
+      "step": 2307
+    },
+    {
+      "epoch": 4.274074074074074,
+      "grad_norm": 1.692909362625955,
+      "learning_rate": 1.9691829098413368e-07,
+      "loss": 0.7318,
+      "step": 2308
+    },
+    {
+      "epoch": 4.275925925925926,
+      "grad_norm": 2.2654075167340655,
+      "learning_rate": 1.9691459618094996e-07,
+      "loss": 0.7193,
+      "step": 2309
+    },
+    {
+      "epoch": 4.277777777777778,
+      "grad_norm": 1.4717886113152736,
+      "learning_rate": 1.969108991988629e-07,
+      "loss": 0.7267,
+      "step": 2310
+    },
+    {
+      "epoch": 4.279629629629629,
+      "grad_norm": 1.8551498631186467,
+      "learning_rate": 1.969072000379556e-07,
+      "loss": 0.6772,
+      "step": 2311
+    },
+    {
+      "epoch": 4.281481481481482,
+      "grad_norm": 1.31833327253509,
+      "learning_rate": 1.9690349869831125e-07,
+      "loss": 0.7001,
+      "step": 2312
+    },
+    {
+      "epoch": 4.283333333333333,
+      "grad_norm": 1.4921738146631278,
+      "learning_rate": 1.968997951800131e-07,
+      "loss": 0.6843,
+      "step": 2313
+    },
+    {
+      "epoch": 4.285185185185185,
+      "grad_norm": 1.4869023237836838,
+      "learning_rate": 1.9689608948314433e-07,
+      "loss": 0.7066,
+      "step": 2314
+    },
+    {
+      "epoch": 4.287037037037037,
+      "grad_norm": 1.7499097730414759,
+      "learning_rate": 1.9689238160778833e-07,
+      "loss": 0.6995,
+      "step": 2315
+    },
+    {
+      "epoch": 4.288888888888889,
+      "grad_norm": 1.5476176483612694,
+      "learning_rate": 1.9688867155402843e-07,
+      "loss": 0.7139,
+      "step": 2316
+    },
+    {
+      "epoch": 4.290740740740741,
+      "grad_norm": 1.6543335984688297,
+      "learning_rate": 1.9688495932194805e-07,
+      "loss": 0.6553,
+      "step": 2317
+    },
+    {
+      "epoch": 4.292592592592593,
+      "grad_norm": 3.4176797230006306,
+      "learning_rate": 1.9688124491163066e-07,
+      "loss": 0.6984,
+      "step": 2318
+    },
+    {
+      "epoch": 4.294444444444444,
+      "grad_norm": 1.4890040045378254,
+      "learning_rate": 1.9687752832315973e-07,
+      "loss": 0.7351,
+      "step": 2319
+    },
+    {
+      "epoch": 4.296296296296296,
+      "grad_norm": 1.4150723142801551,
+      "learning_rate": 1.9687380955661888e-07,
+      "loss": 0.6748,
+      "step": 2320
+    },
+    {
+      "epoch": 4.298148148148148,
+      "grad_norm": 1.3442858635744965,
+      "learning_rate": 1.9687008861209167e-07,
+      "loss": 0.7559,
+      "step": 2321
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 1.5278787688267936,
+      "learning_rate": 1.9686636548966176e-07,
+      "loss": 0.7076,
+      "step": 2322
+    },
+    {
+      "epoch": 4.301851851851852,
+      "grad_norm": 1.3617132550771704,
+      "learning_rate": 1.968626401894129e-07,
+      "loss": 0.626,
+      "step": 2323
+    },
+    {
+      "epoch": 4.303703703703704,
+      "grad_norm": 3.2242961736469318,
+      "learning_rate": 1.968589127114288e-07,
+      "loss": 0.6766,
+      "step": 2324
+    },
+    {
+      "epoch": 4.305555555555555,
+      "grad_norm": 1.7128904080471745,
+      "learning_rate": 1.9685518305579327e-07,
+      "loss": 0.668,
+      "step": 2325
+    },
+    {
+      "epoch": 4.307407407407408,
+      "grad_norm": 1.6107532461968255,
+      "learning_rate": 1.9685145122259016e-07,
+      "loss": 0.6854,
+      "step": 2326
+    },
+    {
+      "epoch": 4.309259259259259,
+      "grad_norm": 1.5890173961707887,
+      "learning_rate": 1.9684771721190342e-07,
+      "loss": 0.6377,
+      "step": 2327
+    },
+    {
+      "epoch": 4.311111111111111,
+      "grad_norm": 1.323674044479915,
+      "learning_rate": 1.9684398102381692e-07,
+      "loss": 0.7093,
+      "step": 2328
+    },
+    {
+      "epoch": 4.312962962962963,
+      "grad_norm": 3.675952489110242,
+      "learning_rate": 1.9684024265841474e-07,
+      "loss": 0.6596,
+      "step": 2329
+    },
+    {
+      "epoch": 4.314814814814815,
+      "grad_norm": 2.0034725577657424,
+      "learning_rate": 1.9683650211578088e-07,
+      "loss": 0.6319,
+      "step": 2330
+    },
+    {
+      "epoch": 4.316666666666666,
+      "grad_norm": 1.5845396526248812,
+      "learning_rate": 1.9683275939599942e-07,
+      "loss": 0.7091,
+      "step": 2331
+    },
+    {
+      "epoch": 4.318518518518519,
+      "grad_norm": 1.4877926277331357,
+      "learning_rate": 1.9682901449915455e-07,
+      "loss": 0.6735,
+      "step": 2332
+    },
+    {
+      "epoch": 4.32037037037037,
+      "grad_norm": 1.4784240311293557,
+      "learning_rate": 1.9682526742533045e-07,
+      "loss": 0.6897,
+      "step": 2333
+    },
+    {
+      "epoch": 4.322222222222222,
+      "grad_norm": 1.3816827709237887,
+      "learning_rate": 1.9682151817461138e-07,
+      "loss": 0.6699,
+      "step": 2334
+    },
+    {
+      "epoch": 4.324074074074074,
+      "grad_norm": 1.5334864373471044,
+      "learning_rate": 1.968177667470816e-07,
+      "loss": 0.6998,
+      "step": 2335
+    },
+    {
+      "epoch": 4.325925925925926,
+      "grad_norm": 1.5550643485124138,
+      "learning_rate": 1.9681401314282548e-07,
+      "loss": 0.7168,
+      "step": 2336
+    },
+    {
+      "epoch": 4.3277777777777775,
+      "grad_norm": 1.706395438568234,
+      "learning_rate": 1.9681025736192742e-07,
+      "loss": 0.6919,
+      "step": 2337
+    },
+    {
+      "epoch": 4.32962962962963,
+      "grad_norm": 1.5992700670608937,
+      "learning_rate": 1.9680649940447182e-07,
+      "loss": 0.6784,
+      "step": 2338
+    },
+    {
+      "epoch": 4.3314814814814815,
+      "grad_norm": 1.6981882215391304,
+      "learning_rate": 1.968027392705432e-07,
+      "loss": 0.6916,
+      "step": 2339
+    },
+    {
+      "epoch": 4.333333333333333,
+      "grad_norm": 1.4934205886203187,
+      "learning_rate": 1.9679897696022607e-07,
+      "loss": 0.7427,
+      "step": 2340
+    },
+    {
+      "epoch": 4.3351851851851855,
+      "grad_norm": 1.4002997648779245,
+      "learning_rate": 1.9679521247360504e-07,
+      "loss": 0.6595,
+      "step": 2341
+    },
+    {
+      "epoch": 4.337037037037037,
+      "grad_norm": 2.047164484611014,
+      "learning_rate": 1.9679144581076477e-07,
+      "loss": 0.6527,
+      "step": 2342
+    },
+    {
+      "epoch": 4.338888888888889,
+      "grad_norm": 1.4305475391234135,
+      "learning_rate": 1.9678767697178988e-07,
+      "loss": 0.6839,
+      "step": 2343
+    },
+    {
+      "epoch": 4.340740740740741,
+      "grad_norm": 1.5038768583705346,
+      "learning_rate": 1.9678390595676516e-07,
+      "loss": 0.6979,
+      "step": 2344
+    },
+    {
+      "epoch": 4.342592592592593,
+      "grad_norm": 1.8953918709655337,
+      "learning_rate": 1.9678013276577538e-07,
+      "loss": 0.6268,
+      "step": 2345
+    },
+    {
+      "epoch": 4.344444444444444,
+      "grad_norm": 1.890318725555,
+      "learning_rate": 1.9677635739890534e-07,
+      "loss": 0.6675,
+      "step": 2346
+    },
+    {
+      "epoch": 4.3462962962962965,
+      "grad_norm": 2.031671680129892,
+      "learning_rate": 1.9677257985623995e-07,
+      "loss": 0.7113,
+      "step": 2347
+    },
+    {
+      "epoch": 4.348148148148148,
+      "grad_norm": 1.599502066015545,
+      "learning_rate": 1.9676880013786416e-07,
+      "loss": 0.6859,
+      "step": 2348
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 1.5070218514048113,
+      "learning_rate": 1.9676501824386292e-07,
+      "loss": 0.6969,
+      "step": 2349
+    },
+    {
+      "epoch": 4.351851851851852,
+      "grad_norm": 1.4977923178284103,
+      "learning_rate": 1.9676123417432123e-07,
+      "loss": 0.6923,
+      "step": 2350
+    },
+    {
+      "epoch": 4.353703703703704,
+      "grad_norm": 1.3998476686446415,
+      "learning_rate": 1.9675744792932424e-07,
+      "loss": 0.7088,
+      "step": 2351
+    },
+    {
+      "epoch": 4.355555555555555,
+      "grad_norm": 1.6322499275784297,
+      "learning_rate": 1.96753659508957e-07,
+      "loss": 0.6873,
+      "step": 2352
+    },
+    {
+      "epoch": 4.357407407407408,
+      "grad_norm": 1.5679297419678981,
+      "learning_rate": 1.967498689133047e-07,
+      "loss": 0.7184,
+      "step": 2353
+    },
+    {
+      "epoch": 4.359259259259259,
+      "grad_norm": 1.686332798631044,
+      "learning_rate": 1.967460761424526e-07,
+      "loss": 0.6904,
+      "step": 2354
+    },
+    {
+      "epoch": 4.361111111111111,
+      "grad_norm": 1.490096183947135,
+      "learning_rate": 1.9674228119648596e-07,
+      "loss": 0.6771,
+      "step": 2355
+    },
+    {
+      "epoch": 4.362962962962963,
+      "grad_norm": 1.558047028061556,
+      "learning_rate": 1.9673848407549007e-07,
+      "loss": 0.6961,
+      "step": 2356
+    },
+    {
+      "epoch": 4.364814814814815,
+      "grad_norm": 1.7441474155082488,
+      "learning_rate": 1.9673468477955032e-07,
+      "loss": 0.6923,
+      "step": 2357
+    },
+    {
+      "epoch": 4.366666666666666,
+      "grad_norm": 1.7117247371163942,
+      "learning_rate": 1.9673088330875215e-07,
+      "loss": 0.6831,
+      "step": 2358
+    },
+    {
+      "epoch": 4.368518518518519,
+      "grad_norm": 1.5542845877712854,
+      "learning_rate": 1.9672707966318099e-07,
+      "loss": 0.6569,
+      "step": 2359
+    },
+    {
+      "epoch": 4.37037037037037,
+      "grad_norm": 1.5844531527933063,
+      "learning_rate": 1.9672327384292237e-07,
+      "loss": 0.6756,
+      "step": 2360
+    },
+    {
+      "epoch": 4.372222222222222,
+      "grad_norm": 2.1076030936767762,
+      "learning_rate": 1.9671946584806188e-07,
+      "loss": 0.6679,
+      "step": 2361
+    },
+    {
+      "epoch": 4.374074074074074,
+      "grad_norm": 1.3780602984216406,
+      "learning_rate": 1.967156556786851e-07,
+      "loss": 0.6928,
+      "step": 2362
+    },
+    {
+      "epoch": 4.375925925925926,
+      "grad_norm": 1.7865773329558863,
+      "learning_rate": 1.9671184333487769e-07,
+      "loss": 0.7067,
+      "step": 2363
+    },
+    {
+      "epoch": 4.377777777777778,
+      "grad_norm": 1.6991029406573863,
+      "learning_rate": 1.9670802881672538e-07,
+      "loss": 0.6782,
+      "step": 2364
+    },
+    {
+      "epoch": 4.37962962962963,
+      "grad_norm": 1.7640894888355967,
+      "learning_rate": 1.9670421212431396e-07,
+      "loss": 0.6872,
+      "step": 2365
+    },
+    {
+      "epoch": 4.381481481481481,
+      "grad_norm": 1.6820068347164712,
+      "learning_rate": 1.9670039325772918e-07,
+      "loss": 0.6815,
+      "step": 2366
+    },
+    {
+      "epoch": 4.383333333333334,
+      "grad_norm": 1.50621941463131,
+      "learning_rate": 1.9669657221705692e-07,
+      "loss": 0.6503,
+      "step": 2367
+    },
+    {
+      "epoch": 4.385185185185185,
+      "grad_norm": 1.5467105191025625,
+      "learning_rate": 1.966927490023831e-07,
+      "loss": 0.6681,
+      "step": 2368
+    },
+    {
+      "epoch": 4.387037037037037,
+      "grad_norm": 2.07453642348771,
+      "learning_rate": 1.9668892361379364e-07,
+      "loss": 0.696,
+      "step": 2369
+    },
+    {
+      "epoch": 4.388888888888889,
+      "grad_norm": 1.4732809889821752,
+      "learning_rate": 1.9668509605137463e-07,
+      "loss": 0.6973,
+      "step": 2370
+    },
+    {
+      "epoch": 4.390740740740741,
+      "grad_norm": 1.3780632992932822,
+      "learning_rate": 1.9668126631521203e-07,
+      "loss": 0.7168,
+      "step": 2371
+    },
+    {
+      "epoch": 4.392592592592592,
+      "grad_norm": 3.092274387794323,
+      "learning_rate": 1.96677434405392e-07,
+      "loss": 0.749,
+      "step": 2372
+    },
+    {
+      "epoch": 4.394444444444445,
+      "grad_norm": 1.5498985181523233,
+      "learning_rate": 1.9667360032200064e-07,
+      "loss": 0.6788,
+      "step": 2373
+    },
+    {
+      "epoch": 4.396296296296296,
+      "grad_norm": 1.5431896323403844,
+      "learning_rate": 1.966697640651242e-07,
+      "loss": 0.6823,
+      "step": 2374
+    },
+    {
+      "epoch": 4.398148148148148,
+      "grad_norm": 1.3448848916045388,
+      "learning_rate": 1.9666592563484893e-07,
+      "loss": 0.6743,
+      "step": 2375
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 1.6670249929742256,
+      "learning_rate": 1.966620850312611e-07,
+      "loss": 0.695,
+      "step": 2376
+    },
+    {
+      "epoch": 4.401851851851852,
+      "grad_norm": 1.4748960946861083,
+      "learning_rate": 1.9665824225444707e-07,
+      "loss": 0.7143,
+      "step": 2377
+    },
+    {
+      "epoch": 4.4037037037037035,
+      "grad_norm": 1.4431618700190834,
+      "learning_rate": 1.9665439730449325e-07,
+      "loss": 0.7015,
+      "step": 2378
+    },
+    {
+      "epoch": 4.405555555555556,
+      "grad_norm": 1.5125610840472394,
+      "learning_rate": 1.9665055018148603e-07,
+      "loss": 0.6946,
+      "step": 2379
+    },
+    {
+      "epoch": 4.407407407407407,
+      "grad_norm": 1.4962978890567262,
+      "learning_rate": 1.9664670088551198e-07,
+      "loss": 0.6597,
+      "step": 2380
+    },
+    {
+      "epoch": 4.409259259259259,
+      "grad_norm": 1.66247046530707,
+      "learning_rate": 1.966428494166576e-07,
+      "loss": 0.7001,
+      "step": 2381
+    },
+    {
+      "epoch": 4.411111111111111,
+      "grad_norm": 1.6837769840799373,
+      "learning_rate": 1.9663899577500947e-07,
+      "loss": 0.635,
+      "step": 2382
+    },
+    {
+      "epoch": 4.412962962962963,
+      "grad_norm": 1.51486886127756,
+      "learning_rate": 1.9663513996065425e-07,
+      "loss": 0.7089,
+      "step": 2383
+    },
+    {
+      "epoch": 4.4148148148148145,
+      "grad_norm": 1.3122943212886047,
+      "learning_rate": 1.9663128197367864e-07,
+      "loss": 0.6537,
+      "step": 2384
+    },
+    {
+      "epoch": 4.416666666666667,
+      "grad_norm": 2.6176071359062556,
+      "learning_rate": 1.9662742181416936e-07,
+      "loss": 0.7263,
+      "step": 2385
+    },
+    {
+      "epoch": 4.4185185185185185,
+      "grad_norm": 1.9221524390865365,
+      "learning_rate": 1.966235594822132e-07,
+      "loss": 0.6741,
+      "step": 2386
+    },
+    {
+      "epoch": 4.42037037037037,
+      "grad_norm": 1.554543842041304,
+      "learning_rate": 1.96619694977897e-07,
+      "loss": 0.7177,
+      "step": 2387
+    },
+    {
+      "epoch": 4.4222222222222225,
+      "grad_norm": 1.520867093901909,
+      "learning_rate": 1.9661582830130762e-07,
+      "loss": 0.6715,
+      "step": 2388
+    },
+    {
+      "epoch": 4.424074074074074,
+      "grad_norm": 1.3633850987431106,
+      "learning_rate": 1.9661195945253203e-07,
+      "loss": 0.6953,
+      "step": 2389
+    },
+    {
+      "epoch": 4.425925925925926,
+      "grad_norm": 1.621114779937641,
+      "learning_rate": 1.966080884316572e-07,
+      "loss": 0.668,
+      "step": 2390
+    },
+    {
+      "epoch": 4.427777777777778,
+      "grad_norm": 1.6283779267691725,
+      "learning_rate": 1.966042152387702e-07,
+      "loss": 0.664,
+      "step": 2391
+    },
+    {
+      "epoch": 4.42962962962963,
+      "grad_norm": 1.7732568278325749,
+      "learning_rate": 1.9660033987395802e-07,
+      "loss": 0.7128,
+      "step": 2392
+    },
+    {
+      "epoch": 4.431481481481481,
+      "grad_norm": 2.005239505823147,
+      "learning_rate": 1.9659646233730782e-07,
+      "loss": 0.6819,
+      "step": 2393
+    },
+    {
+      "epoch": 4.433333333333334,
+      "grad_norm": 1.4806031086417617,
+      "learning_rate": 1.965925826289068e-07,
+      "loss": 0.7168,
+      "step": 2394
+    },
+    {
+      "epoch": 4.435185185185185,
+      "grad_norm": 2.7011701540418165,
+      "learning_rate": 1.9658870074884223e-07,
+      "loss": 0.7213,
+      "step": 2395
+    },
+    {
+      "epoch": 4.437037037037037,
+      "grad_norm": 1.6615366133170555,
+      "learning_rate": 1.965848166972013e-07,
+      "loss": 0.6156,
+      "step": 2396
+    },
+    {
+      "epoch": 4.438888888888889,
+      "grad_norm": 1.3741393099098673,
+      "learning_rate": 1.9658093047407136e-07,
+      "loss": 0.7057,
+      "step": 2397
+    },
+    {
+      "epoch": 4.440740740740741,
+      "grad_norm": 1.5457289951528703,
+      "learning_rate": 1.9657704207953983e-07,
+      "loss": 0.6635,
+      "step": 2398
+    },
+    {
+      "epoch": 4.442592592592592,
+      "grad_norm": 1.781929065650584,
+      "learning_rate": 1.9657315151369408e-07,
+      "loss": 0.6263,
+      "step": 2399
+    },
+    {
+      "epoch": 4.444444444444445,
+      "grad_norm": 1.6173288177732108,
+      "learning_rate": 1.965692587766216e-07,
+      "loss": 0.6296,
+      "step": 2400
+    },
+    {
+      "epoch": 4.446296296296296,
+      "grad_norm": 1.6761171449830787,
+      "learning_rate": 1.9656536386840987e-07,
+      "loss": 0.7233,
+      "step": 2401
+    },
+    {
+      "epoch": 4.448148148148148,
+      "grad_norm": 1.3702406511156924,
+      "learning_rate": 1.965614667891465e-07,
+      "loss": 0.6626,
+      "step": 2402
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 1.6119917256295062,
+      "learning_rate": 1.9655756753891915e-07,
+      "loss": 0.703,
+      "step": 2403
+    },
+    {
+      "epoch": 4.451851851851852,
+      "grad_norm": 1.5402894568752385,
+      "learning_rate": 1.9655366611781538e-07,
+      "loss": 0.691,
+      "step": 2404
+    },
+    {
+      "epoch": 4.453703703703704,
+      "grad_norm": 1.54251161451274,
+      "learning_rate": 1.96549762525923e-07,
+      "loss": 0.7081,
+      "step": 2405
+    },
+    {
+      "epoch": 4.455555555555556,
+      "grad_norm": 1.5856002467827401,
+      "learning_rate": 1.965458567633297e-07,
+      "loss": 0.7057,
+      "step": 2406
+    },
+    {
+      "epoch": 4.457407407407407,
+      "grad_norm": 1.5166706394239964,
+      "learning_rate": 1.9654194883012335e-07,
+      "loss": 0.639,
+      "step": 2407
+    },
+    {
+      "epoch": 4.459259259259259,
+      "grad_norm": 1.4020674626160166,
+      "learning_rate": 1.9653803872639175e-07,
+      "loss": 0.6922,
+      "step": 2408
+    },
+    {
+      "epoch": 4.461111111111111,
+      "grad_norm": 1.6831364244895552,
+      "learning_rate": 1.9653412645222288e-07,
+      "loss": 0.707,
+      "step": 2409
+    },
+    {
+      "epoch": 4.462962962962963,
+      "grad_norm": 1.4110691733242287,
+      "learning_rate": 1.9653021200770467e-07,
+      "loss": 0.6761,
+      "step": 2410
+    },
+    {
+      "epoch": 4.464814814814815,
+      "grad_norm": 1.7237659498575442,
+      "learning_rate": 1.9652629539292512e-07,
+      "loss": 0.6972,
+      "step": 2411
+    },
+    {
+      "epoch": 4.466666666666667,
+      "grad_norm": 1.4932202921523328,
+      "learning_rate": 1.9652237660797225e-07,
+      "loss": 0.6749,
+      "step": 2412
+    },
+    {
+      "epoch": 4.468518518518518,
+      "grad_norm": 1.6116619910054157,
+      "learning_rate": 1.9651845565293423e-07,
+      "loss": 0.6347,
+      "step": 2413
+    },
+    {
+      "epoch": 4.470370370370371,
+      "grad_norm": 1.6772409168750315,
+      "learning_rate": 1.965145325278992e-07,
+      "loss": 0.7371,
+      "step": 2414
+    },
+    {
+      "epoch": 4.472222222222222,
+      "grad_norm": 1.5878064543705943,
+      "learning_rate": 1.9651060723295535e-07,
+      "loss": 0.6431,
+      "step": 2415
+    },
+    {
+      "epoch": 4.474074074074074,
+      "grad_norm": 1.3830616204809014,
+      "learning_rate": 1.965066797681909e-07,
+      "loss": 0.6777,
+      "step": 2416
+    },
+    {
+      "epoch": 4.475925925925926,
+      "grad_norm": 1.7002639460923434,
+      "learning_rate": 1.9650275013369421e-07,
+      "loss": 0.7034,
+      "step": 2417
+    },
+    {
+      "epoch": 4.477777777777778,
+      "grad_norm": 2.038646674254075,
+      "learning_rate": 1.964988183295536e-07,
+      "loss": 0.6578,
+      "step": 2418
+    },
+    {
+      "epoch": 4.479629629629629,
+      "grad_norm": 1.7307788126039625,
+      "learning_rate": 1.9649488435585743e-07,
+      "loss": 0.6944,
+      "step": 2419
+    },
+    {
+      "epoch": 4.481481481481482,
+      "grad_norm": 1.5922951431882066,
+      "learning_rate": 1.9649094821269424e-07,
+      "loss": 0.6489,
+      "step": 2420
+    },
+    {
+      "epoch": 4.483333333333333,
+      "grad_norm": 1.3216260980940011,
+      "learning_rate": 1.964870099001524e-07,
+      "loss": 0.6504,
+      "step": 2421
+    },
+    {
+      "epoch": 4.485185185185185,
+      "grad_norm": 1.4984329468265338,
+      "learning_rate": 1.9648306941832055e-07,
+      "loss": 0.6495,
+      "step": 2422
+    },
+    {
+      "epoch": 4.487037037037037,
+      "grad_norm": 1.5446642960928283,
+      "learning_rate": 1.9647912676728729e-07,
+      "loss": 0.7122,
+      "step": 2423
+    },
+    {
+      "epoch": 4.488888888888889,
+      "grad_norm": 2.543488707428639,
+      "learning_rate": 1.9647518194714122e-07,
+      "loss": 0.719,
+      "step": 2424
+    },
+    {
+      "epoch": 4.4907407407407405,
+      "grad_norm": 2.060011821320633,
+      "learning_rate": 1.96471234957971e-07,
+      "loss": 0.6642,
+      "step": 2425
+    },
+    {
+      "epoch": 4.492592592592593,
+      "grad_norm": 1.587620858457138,
+      "learning_rate": 1.9646728579986542e-07,
+      "loss": 0.6551,
+      "step": 2426
+    },
+    {
+      "epoch": 4.4944444444444445,
+      "grad_norm": 1.6116222103706,
+      "learning_rate": 1.964633344729133e-07,
+      "loss": 0.6835,
+      "step": 2427
+    },
+    {
+      "epoch": 4.496296296296296,
+      "grad_norm": 1.4478623102876287,
+      "learning_rate": 1.9645938097720336e-07,
+      "loss": 0.6453,
+      "step": 2428
+    },
+    {
+      "epoch": 4.4981481481481485,
+      "grad_norm": 1.6068415350856775,
+      "learning_rate": 1.964554253128246e-07,
+      "loss": 0.7145,
+      "step": 2429
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 1.4256365709305527,
+      "learning_rate": 1.9645146747986589e-07,
+      "loss": 0.6522,
+      "step": 2430
+    },
+    {
+      "epoch": 4.5018518518518515,
+      "grad_norm": 1.5353351608677912,
+      "learning_rate": 1.9644750747841622e-07,
+      "loss": 0.7083,
+      "step": 2431
+    },
+    {
+      "epoch": 4.503703703703704,
+      "grad_norm": 1.7544991379226202,
+      "learning_rate": 1.9644354530856465e-07,
+      "loss": 0.7151,
+      "step": 2432
+    },
+    {
+      "epoch": 4.5055555555555555,
+      "grad_norm": 1.5087673143829372,
+      "learning_rate": 1.9643958097040024e-07,
+      "loss": 0.6671,
+      "step": 2433
+    },
+    {
+      "epoch": 4.507407407407407,
+      "grad_norm": 1.356781894671453,
+      "learning_rate": 1.9643561446401217e-07,
+      "loss": 0.684,
+      "step": 2434
+    },
+    {
+      "epoch": 4.5092592592592595,
+      "grad_norm": 1.5841396202681788,
+      "learning_rate": 1.964316457894895e-07,
+      "loss": 0.6815,
+      "step": 2435
+    },
+    {
+      "epoch": 4.511111111111111,
+      "grad_norm": 2.159185886689002,
+      "learning_rate": 1.9642767494692158e-07,
+      "loss": 0.7013,
+      "step": 2436
+    },
+    {
+      "epoch": 4.512962962962963,
+      "grad_norm": 1.6160052270632668,
+      "learning_rate": 1.964237019363976e-07,
+      "loss": 0.6552,
+      "step": 2437
+    },
+    {
+      "epoch": 4.514814814814815,
+      "grad_norm": 1.4558896411201991,
+      "learning_rate": 1.9641972675800695e-07,
+      "loss": 0.6823,
+      "step": 2438
+    },
+    {
+      "epoch": 4.516666666666667,
+      "grad_norm": 1.695200750044827,
+      "learning_rate": 1.9641574941183893e-07,
+      "loss": 0.7192,
+      "step": 2439
+    },
+    {
+      "epoch": 4.518518518518518,
+      "grad_norm": 1.7358590588009553,
+      "learning_rate": 1.9641176989798302e-07,
+      "loss": 0.7085,
+      "step": 2440
+    },
+    {
+      "epoch": 4.520370370370371,
+      "grad_norm": 1.5119955027098257,
+      "learning_rate": 1.964077882165287e-07,
+      "loss": 0.6718,
+      "step": 2441
+    },
+    {
+      "epoch": 4.522222222222222,
+      "grad_norm": 1.6872179603981665,
+      "learning_rate": 1.9640380436756544e-07,
+      "loss": 0.6804,
+      "step": 2442
+    },
+    {
+      "epoch": 4.524074074074074,
+      "grad_norm": 1.7023356501940974,
+      "learning_rate": 1.963998183511828e-07,
+      "loss": 0.7048,
+      "step": 2443
+    },
+    {
+      "epoch": 4.525925925925926,
+      "grad_norm": 1.864878210703174,
+      "learning_rate": 1.9639583016747048e-07,
+      "loss": 0.7453,
+      "step": 2444
+    },
+    {
+      "epoch": 4.527777777777778,
+      "grad_norm": 1.7407177542937717,
+      "learning_rate": 1.9639183981651808e-07,
+      "loss": 0.6842,
+      "step": 2445
+    },
+    {
+      "epoch": 4.52962962962963,
+      "grad_norm": 1.7883034287823847,
+      "learning_rate": 1.963878472984153e-07,
+      "loss": 0.6785,
+      "step": 2446
+    },
+    {
+      "epoch": 4.531481481481482,
+      "grad_norm": 1.5591214579620831,
+      "learning_rate": 1.9638385261325192e-07,
+      "loss": 0.6729,
+      "step": 2447
+    },
+    {
+      "epoch": 4.533333333333333,
+      "grad_norm": 1.6080240317581804,
+      "learning_rate": 1.9637985576111778e-07,
+      "loss": 0.685,
+      "step": 2448
+    },
+    {
+      "epoch": 4.535185185185185,
+      "grad_norm": 1.9957364959103105,
+      "learning_rate": 1.9637585674210274e-07,
+      "loss": 0.661,
+      "step": 2449
+    },
+    {
+      "epoch": 4.537037037037037,
+      "grad_norm": 1.5215072428715826,
+      "learning_rate": 1.9637185555629663e-07,
+      "loss": 0.6713,
+      "step": 2450
+    },
+    {
+      "epoch": 4.538888888888889,
+      "grad_norm": 1.643521573232913,
+      "learning_rate": 1.963678522037895e-07,
+      "loss": 0.6935,
+      "step": 2451
+    },
+    {
+      "epoch": 4.540740740740741,
+      "grad_norm": 1.4458233435771315,
+      "learning_rate": 1.963638466846713e-07,
+      "loss": 0.7033,
+      "step": 2452
+    },
+    {
+      "epoch": 4.542592592592593,
+      "grad_norm": 1.4529875669676566,
+      "learning_rate": 1.9635983899903215e-07,
+      "loss": 0.7,
+      "step": 2453
+    },
+    {
+      "epoch": 4.544444444444444,
+      "grad_norm": 1.535621846755058,
+      "learning_rate": 1.963558291469621e-07,
+      "loss": 0.7377,
+      "step": 2454
+    },
+    {
+      "epoch": 4.546296296296296,
+      "grad_norm": 1.3917112409585297,
+      "learning_rate": 1.963518171285513e-07,
+      "loss": 0.6688,
+      "step": 2455
+    },
+    {
+      "epoch": 4.548148148148148,
+      "grad_norm": 2.145977150141421,
+      "learning_rate": 1.9634780294388995e-07,
+      "loss": 0.6811,
+      "step": 2456
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 1.634622934844824,
+      "learning_rate": 1.9634378659306832e-07,
+      "loss": 0.6966,
+      "step": 2457
+    },
+    {
+      "epoch": 4.551851851851852,
+      "grad_norm": 1.4789292012956052,
+      "learning_rate": 1.963397680761767e-07,
+      "loss": 0.6946,
+      "step": 2458
+    },
+    {
+      "epoch": 4.553703703703704,
+      "grad_norm": 1.4459577812881277,
+      "learning_rate": 1.9633574739330546e-07,
+      "loss": 0.671,
+      "step": 2459
+    },
+    {
+      "epoch": 4.555555555555555,
+      "grad_norm": 1.3421470003354312,
+      "learning_rate": 1.9633172454454495e-07,
+      "loss": 0.7119,
+      "step": 2460
+    },
+    {
+      "epoch": 4.557407407407408,
+      "grad_norm": 1.4777860469181126,
+      "learning_rate": 1.9632769952998566e-07,
+      "loss": 0.7256,
+      "step": 2461
+    },
+    {
+      "epoch": 4.559259259259259,
+      "grad_norm": 1.3293392775929562,
+      "learning_rate": 1.9632367234971802e-07,
+      "loss": 0.7326,
+      "step": 2462
+    },
+    {
+      "epoch": 4.561111111111111,
+      "grad_norm": 1.7657205146975035,
+      "learning_rate": 1.9631964300383267e-07,
+      "loss": 0.7225,
+      "step": 2463
+    },
+    {
+      "epoch": 4.562962962962963,
+      "grad_norm": 1.6070353156946582,
+      "learning_rate": 1.963156114924201e-07,
+      "loss": 0.6556,
+      "step": 2464
+    },
+    {
+      "epoch": 4.564814814814815,
+      "grad_norm": 1.8722185653369976,
+      "learning_rate": 1.96311577815571e-07,
+      "loss": 0.6969,
+      "step": 2465
+    },
+    {
+      "epoch": 4.566666666666666,
+      "grad_norm": 1.5852800887906722,
+      "learning_rate": 1.9630754197337608e-07,
+      "loss": 0.7098,
+      "step": 2466
+    },
+    {
+      "epoch": 4.568518518518519,
+      "grad_norm": 2.5997157514137337,
+      "learning_rate": 1.9630350396592604e-07,
+      "loss": 0.6959,
+      "step": 2467
+    },
+    {
+      "epoch": 4.57037037037037,
+      "grad_norm": 1.4523933697811655,
+      "learning_rate": 1.9629946379331167e-07,
+      "loss": 0.6616,
+      "step": 2468
+    },
+    {
+      "epoch": 4.572222222222222,
+      "grad_norm": 1.5856203419210637,
+      "learning_rate": 1.962954214556238e-07,
+      "loss": 0.6888,
+      "step": 2469
+    },
+    {
+      "epoch": 4.574074074074074,
+      "grad_norm": 1.610935755469287,
+      "learning_rate": 1.9629137695295334e-07,
+      "loss": 0.7206,
+      "step": 2470
+    },
+    {
+      "epoch": 4.575925925925926,
+      "grad_norm": 1.6159209436172026,
+      "learning_rate": 1.962873302853912e-07,
+      "loss": 0.6918,
+      "step": 2471
+    },
+    {
+      "epoch": 4.5777777777777775,
+      "grad_norm": 1.4612138613264,
+      "learning_rate": 1.9628328145302834e-07,
+      "loss": 0.6529,
+      "step": 2472
+    },
+    {
+      "epoch": 4.57962962962963,
+      "grad_norm": 1.6277979455846525,
+      "learning_rate": 1.9627923045595582e-07,
+      "loss": 0.654,
+      "step": 2473
+    },
+    {
+      "epoch": 4.5814814814814815,
+      "grad_norm": 1.4594466889751476,
+      "learning_rate": 1.9627517729426476e-07,
+      "loss": 0.6605,
+      "step": 2474
+    },
+    {
+      "epoch": 4.583333333333333,
+      "grad_norm": 1.5986337471335546,
+      "learning_rate": 1.9627112196804618e-07,
+      "loss": 0.6503,
+      "step": 2475
+    },
+    {
+      "epoch": 4.5851851851851855,
+      "grad_norm": 1.7207611157671368,
+      "learning_rate": 1.962670644773913e-07,
+      "loss": 0.6656,
+      "step": 2476
+    },
+    {
+      "epoch": 4.587037037037037,
+      "grad_norm": 1.7472845925595935,
+      "learning_rate": 1.9626300482239139e-07,
+      "loss": 0.615,
+      "step": 2477
+    },
+    {
+      "epoch": 4.588888888888889,
+      "grad_norm": 1.5655867417505582,
+      "learning_rate": 1.9625894300313766e-07,
+      "loss": 0.683,
+      "step": 2478
+    },
+    {
+      "epoch": 4.590740740740741,
+      "grad_norm": 1.4673176152861458,
+      "learning_rate": 1.9625487901972146e-07,
+      "loss": 0.6685,
+      "step": 2479
+    },
+    {
+      "epoch": 4.592592592592593,
+      "grad_norm": 1.3277572844281333,
+      "learning_rate": 1.9625081287223417e-07,
+      "loss": 0.6612,
+      "step": 2480
+    },
+    {
+      "epoch": 4.594444444444444,
+      "grad_norm": 1.5876659714430752,
+      "learning_rate": 1.9624674456076717e-07,
+      "loss": 0.7101,
+      "step": 2481
+    },
+    {
+      "epoch": 4.5962962962962965,
+      "grad_norm": 1.6105766675469035,
+      "learning_rate": 1.9624267408541197e-07,
+      "loss": 0.6791,
+      "step": 2482
+    },
+    {
+      "epoch": 4.598148148148148,
+      "grad_norm": 1.6155759236205214,
+      "learning_rate": 1.9623860144626007e-07,
+      "loss": 0.716,
+      "step": 2483
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 1.3439939259020257,
+      "learning_rate": 1.9623452664340303e-07,
+      "loss": 0.6676,
+      "step": 2484
+    },
+    {
+      "epoch": 4.601851851851852,
+      "grad_norm": 1.4679818582936264,
+      "learning_rate": 1.9623044967693242e-07,
+      "loss": 0.6811,
+      "step": 2485
+    },
+    {
+      "epoch": 4.603703703703704,
+      "grad_norm": 1.559172559240203,
+      "learning_rate": 1.9622637054694e-07,
+      "loss": 0.6472,
+      "step": 2486
+    },
+    {
+      "epoch": 4.605555555555555,
+      "grad_norm": 1.6036210917023734,
+      "learning_rate": 1.9622228925351737e-07,
+      "loss": 0.6962,
+      "step": 2487
+    },
+    {
+      "epoch": 4.607407407407408,
+      "grad_norm": 1.3561314436215313,
+      "learning_rate": 1.9621820579675637e-07,
+      "loss": 0.637,
+      "step": 2488
+    },
+    {
+      "epoch": 4.609259259259259,
+      "grad_norm": 1.3972570028452092,
+      "learning_rate": 1.9621412017674877e-07,
+      "loss": 0.7311,
+      "step": 2489
+    },
+    {
+      "epoch": 4.611111111111111,
+      "grad_norm": 1.3740122124621938,
+      "learning_rate": 1.9621003239358648e-07,
+      "loss": 0.7166,
+      "step": 2490
+    },
+    {
+      "epoch": 4.612962962962963,
+      "grad_norm": 1.689510697769527,
+      "learning_rate": 1.962059424473613e-07,
+      "loss": 0.7339,
+      "step": 2491
+    },
+    {
+      "epoch": 4.614814814814815,
+      "grad_norm": 1.4247349983545494,
+      "learning_rate": 1.962018503381653e-07,
+      "loss": 0.6459,
+      "step": 2492
+    },
+    {
+      "epoch": 4.616666666666667,
+      "grad_norm": 1.6077183637468233,
+      "learning_rate": 1.9619775606609036e-07,
+      "loss": 0.689,
+      "step": 2493
+    },
+    {
+      "epoch": 4.618518518518519,
+      "grad_norm": 1.621950037448575,
+      "learning_rate": 1.9619365963122862e-07,
+      "loss": 0.6728,
+      "step": 2494
+    },
+    {
+      "epoch": 4.62037037037037,
+      "grad_norm": 1.588792512223338,
+      "learning_rate": 1.9618956103367216e-07,
+      "loss": 0.6673,
+      "step": 2495
+    },
+    {
+      "epoch": 4.622222222222222,
+      "grad_norm": 1.3558640415594747,
+      "learning_rate": 1.9618546027351311e-07,
+      "loss": 0.6748,
+      "step": 2496
+    },
+    {
+      "epoch": 4.624074074074074,
+      "grad_norm": 1.4596811440019657,
+      "learning_rate": 1.961813573508437e-07,
+      "loss": 0.6683,
+      "step": 2497
+    },
+    {
+      "epoch": 4.625925925925926,
+      "grad_norm": 1.5338412161859356,
+      "learning_rate": 1.9617725226575615e-07,
+      "loss": 0.6908,
+      "step": 2498
+    },
+    {
+      "epoch": 4.627777777777778,
+      "grad_norm": 1.610252367001902,
+      "learning_rate": 1.9617314501834273e-07,
+      "loss": 0.6915,
+      "step": 2499
+    },
+    {
+      "epoch": 4.62962962962963,
+      "grad_norm": 1.6332266311468415,
+      "learning_rate": 1.9616903560869583e-07,
+      "loss": 0.7113,
+      "step": 2500
+    },
+    {
+      "epoch": 4.631481481481481,
+      "grad_norm": 1.6114485279437283,
+      "learning_rate": 1.961649240369078e-07,
+      "loss": 0.6711,
+      "step": 2501
+    },
+    {
+      "epoch": 4.633333333333333,
+      "grad_norm": 1.406623520026227,
+      "learning_rate": 1.961608103030711e-07,
+      "loss": 0.6527,
+      "step": 2502
+    },
+    {
+      "epoch": 4.635185185185185,
+      "grad_norm": 1.9476433823561836,
+      "learning_rate": 1.9615669440727825e-07,
+      "loss": 0.677,
+      "step": 2503
+    },
+    {
+      "epoch": 4.637037037037037,
+      "grad_norm": 1.451848427975165,
+      "learning_rate": 1.961525763496217e-07,
+      "loss": 0.6561,
+      "step": 2504
+    },
+    {
+      "epoch": 4.638888888888889,
+      "grad_norm": 1.4690683263411852,
+      "learning_rate": 1.9614845613019413e-07,
+      "loss": 0.6891,
+      "step": 2505
+    },
+    {
+      "epoch": 4.640740740740741,
+      "grad_norm": 1.473266870790206,
+      "learning_rate": 1.961443337490881e-07,
+      "loss": 0.6572,
+      "step": 2506
+    },
+    {
+      "epoch": 4.642592592592592,
+      "grad_norm": 1.3814093429351664,
+      "learning_rate": 1.9614020920639633e-07,
+      "loss": 0.6416,
+      "step": 2507
+    },
+    {
+      "epoch": 4.644444444444445,
+      "grad_norm": 1.4335432358850984,
+      "learning_rate": 1.9613608250221156e-07,
+      "loss": 0.6912,
+      "step": 2508
+    },
+    {
+      "epoch": 4.646296296296296,
+      "grad_norm": 1.54130799663505,
+      "learning_rate": 1.9613195363662656e-07,
+      "loss": 0.7293,
+      "step": 2509
+    },
+    {
+      "epoch": 4.648148148148148,
+      "grad_norm": 1.4140725259659122,
+      "learning_rate": 1.9612782260973414e-07,
+      "loss": 0.6871,
+      "step": 2510
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 1.3933266209765902,
+      "learning_rate": 1.9612368942162717e-07,
+      "loss": 0.6884,
+      "step": 2511
+    },
+    {
+      "epoch": 4.651851851851852,
+      "grad_norm": 1.6801208131532885,
+      "learning_rate": 1.961195540723986e-07,
+      "loss": 0.6966,
+      "step": 2512
+    },
+    {
+      "epoch": 4.6537037037037035,
+      "grad_norm": 1.5616389800348724,
+      "learning_rate": 1.9611541656214142e-07,
+      "loss": 0.7646,
+      "step": 2513
+    },
+    {
+      "epoch": 4.655555555555556,
+      "grad_norm": 1.6018904201339035,
+      "learning_rate": 1.961112768909486e-07,
+      "loss": 0.7127,
+      "step": 2514
+    },
+    {
+      "epoch": 4.657407407407407,
+      "grad_norm": 1.536136422995422,
+      "learning_rate": 1.9610713505891327e-07,
+      "loss": 0.661,
+      "step": 2515
+    },
+    {
+      "epoch": 4.659259259259259,
+      "grad_norm": 1.2916127592865736,
+      "learning_rate": 1.9610299106612848e-07,
+      "loss": 0.6235,
+      "step": 2516
+    },
+    {
+      "epoch": 4.661111111111111,
+      "grad_norm": 1.4296912572085654,
+      "learning_rate": 1.9609884491268748e-07,
+      "loss": 0.716,
+      "step": 2517
+    },
+    {
+      "epoch": 4.662962962962963,
+      "grad_norm": 1.3975080778540827,
+      "learning_rate": 1.9609469659868344e-07,
+      "loss": 0.6644,
+      "step": 2518
+    },
+    {
+      "epoch": 4.6648148148148145,
+      "grad_norm": 1.5320929309325129,
+      "learning_rate": 1.9609054612420963e-07,
+      "loss": 0.6219,
+      "step": 2519
+    },
+    {
+      "epoch": 4.666666666666667,
+      "grad_norm": 1.5315430129556165,
+      "learning_rate": 1.9608639348935936e-07,
+      "loss": 0.7452,
+      "step": 2520
+    },
+    {
+      "epoch": 4.6685185185185185,
+      "grad_norm": 1.7254608249682457,
+      "learning_rate": 1.9608223869422597e-07,
+      "loss": 0.7178,
+      "step": 2521
+    },
+    {
+      "epoch": 4.67037037037037,
+      "grad_norm": 1.504428069918678,
+      "learning_rate": 1.9607808173890294e-07,
+      "loss": 0.696,
+      "step": 2522
+    },
+    {
+      "epoch": 4.6722222222222225,
+      "grad_norm": 1.5322461300382144,
+      "learning_rate": 1.960739226234837e-07,
+      "loss": 0.7157,
+      "step": 2523
+    },
+    {
+      "epoch": 4.674074074074074,
+      "grad_norm": 1.5900607180789752,
+      "learning_rate": 1.960697613480617e-07,
+      "loss": 0.6546,
+      "step": 2524
+    },
+    {
+      "epoch": 4.675925925925926,
+      "grad_norm": 1.5930014641807357,
+      "learning_rate": 1.960655979127306e-07,
+      "loss": 0.7054,
+      "step": 2525
+    },
+    {
+      "epoch": 4.677777777777778,
+      "grad_norm": 1.5777043461019717,
+      "learning_rate": 1.960614323175839e-07,
+      "loss": 0.745,
+      "step": 2526
+    },
+    {
+      "epoch": 4.67962962962963,
+      "grad_norm": 1.672902950794046,
+      "learning_rate": 1.960572645627153e-07,
+      "loss": 0.6302,
+      "step": 2527
+    },
+    {
+      "epoch": 4.681481481481481,
+      "grad_norm": 1.7492836158072649,
+      "learning_rate": 1.9605309464821852e-07,
+      "loss": 0.6689,
+      "step": 2528
+    },
+    {
+      "epoch": 4.683333333333334,
+      "grad_norm": 1.4531047783738247,
+      "learning_rate": 1.960489225741873e-07,
+      "loss": 0.6562,
+      "step": 2529
+    },
+    {
+      "epoch": 4.685185185185185,
+      "grad_norm": 1.7522197282750123,
+      "learning_rate": 1.9604474834071542e-07,
+      "loss": 0.642,
+      "step": 2530
+    },
+    {
+      "epoch": 4.687037037037037,
+      "grad_norm": 1.9035753566690667,
+      "learning_rate": 1.9604057194789676e-07,
+      "loss": 0.7133,
+      "step": 2531
+    },
+    {
+      "epoch": 4.688888888888889,
+      "grad_norm": 2.249287265532964,
+      "learning_rate": 1.960363933958252e-07,
+      "loss": 0.7063,
+      "step": 2532
+    },
+    {
+      "epoch": 4.690740740740741,
+      "grad_norm": 2.000160447310834,
+      "learning_rate": 1.9603221268459468e-07,
+      "loss": 0.7022,
+      "step": 2533
+    },
+    {
+      "epoch": 4.692592592592592,
+      "grad_norm": 1.5689898123125854,
+      "learning_rate": 1.960280298142992e-07,
+      "loss": 0.6588,
+      "step": 2534
+    },
+    {
+      "epoch": 4.694444444444445,
+      "grad_norm": 1.4839313249873542,
+      "learning_rate": 1.9602384478503283e-07,
+      "loss": 0.6588,
+      "step": 2535
+    },
+    {
+      "epoch": 4.696296296296296,
+      "grad_norm": 1.404681618289382,
+      "learning_rate": 1.960196575968896e-07,
+      "loss": 0.6782,
+      "step": 2536
+    },
+    {
+      "epoch": 4.698148148148148,
+      "grad_norm": 1.5440557451402022,
+      "learning_rate": 1.9601546824996369e-07,
+      "loss": 0.7025,
+      "step": 2537
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 1.5209307620580186,
+      "learning_rate": 1.9601127674434928e-07,
+      "loss": 0.7049,
+      "step": 2538
+    },
+    {
+      "epoch": 4.701851851851852,
+      "grad_norm": 1.4882912604297536,
+      "learning_rate": 1.9600708308014057e-07,
+      "loss": 0.7011,
+      "step": 2539
+    },
+    {
+      "epoch": 4.703703703703704,
+      "grad_norm": 1.376471744192453,
+      "learning_rate": 1.9600288725743192e-07,
+      "loss": 0.7226,
+      "step": 2540
+    },
+    {
+      "epoch": 4.705555555555556,
+      "grad_norm": 1.395341320189781,
+      "learning_rate": 1.9599868927631765e-07,
+      "loss": 0.6449,
+      "step": 2541
+    },
+    {
+      "epoch": 4.707407407407407,
+      "grad_norm": 1.523134352032048,
+      "learning_rate": 1.9599448913689204e-07,
+      "loss": 0.7144,
+      "step": 2542
+    },
+    {
+      "epoch": 4.709259259259259,
+      "grad_norm": 2.0961339458867516,
+      "learning_rate": 1.9599028683924964e-07,
+      "loss": 0.703,
+      "step": 2543
+    },
+    {
+      "epoch": 4.711111111111111,
+      "grad_norm": 1.3300624922406747,
+      "learning_rate": 1.959860823834849e-07,
+      "loss": 0.6339,
+      "step": 2544
+    },
+    {
+      "epoch": 4.712962962962963,
+      "grad_norm": 1.302116166969961,
+      "learning_rate": 1.9598187576969234e-07,
+      "loss": 0.7167,
+      "step": 2545
+    },
+    {
+      "epoch": 4.714814814814815,
+      "grad_norm": 1.4664818998034495,
+      "learning_rate": 1.959776669979665e-07,
+      "loss": 0.7121,
+      "step": 2546
+    },
+    {
+      "epoch": 4.716666666666667,
+      "grad_norm": 1.4684768506851666,
+      "learning_rate": 1.9597345606840206e-07,
+      "loss": 0.6771,
+      "step": 2547
+    },
+    {
+      "epoch": 4.718518518518518,
+      "grad_norm": 1.3619799791310647,
+      "learning_rate": 1.9596924298109365e-07,
+      "loss": 0.6889,
+      "step": 2548
+    },
+    {
+      "epoch": 4.72037037037037,
+      "grad_norm": 1.9874509884656524,
+      "learning_rate": 1.9596502773613602e-07,
+      "loss": 0.6676,
+      "step": 2549
+    },
+    {
+      "epoch": 4.722222222222222,
+      "grad_norm": 1.4666539531597036,
+      "learning_rate": 1.9596081033362393e-07,
+      "loss": 0.7081,
+      "step": 2550
+    },
+    {
+      "epoch": 4.724074074074074,
+      "grad_norm": 1.4591212775113622,
+      "learning_rate": 1.959565907736522e-07,
+      "loss": 0.6821,
+      "step": 2551
+    },
+    {
+      "epoch": 4.725925925925926,
+      "grad_norm": 1.6105275042909308,
+      "learning_rate": 1.9595236905631567e-07,
+      "loss": 0.6451,
+      "step": 2552
+    },
+    {
+      "epoch": 4.727777777777778,
+      "grad_norm": 1.5352243609040608,
+      "learning_rate": 1.959481451817093e-07,
+      "loss": 0.6831,
+      "step": 2553
+    },
+    {
+      "epoch": 4.729629629629629,
+      "grad_norm": 6.049561837551794,
+      "learning_rate": 1.9594391914992803e-07,
+      "loss": 0.6992,
+      "step": 2554
+    },
+    {
+      "epoch": 4.731481481481482,
+      "grad_norm": 1.6117055707567023,
+      "learning_rate": 1.9593969096106688e-07,
+      "loss": 0.6822,
+      "step": 2555
+    },
+    {
+      "epoch": 4.733333333333333,
+      "grad_norm": 1.6100100616769308,
+      "learning_rate": 1.959354606152209e-07,
+      "loss": 0.7079,
+      "step": 2556
+    },
+    {
+      "epoch": 4.735185185185185,
+      "grad_norm": 1.3680011324512609,
+      "learning_rate": 1.9593122811248522e-07,
+      "loss": 0.633,
+      "step": 2557
+    },
+    {
+      "epoch": 4.737037037037037,
+      "grad_norm": 1.533417437300952,
+      "learning_rate": 1.95926993452955e-07,
+      "loss": 0.6672,
+      "step": 2558
+    },
+    {
+      "epoch": 4.738888888888889,
+      "grad_norm": 4.351594857068372,
+      "learning_rate": 1.9592275663672543e-07,
+      "loss": 0.6345,
+      "step": 2559
+    },
+    {
+      "epoch": 4.7407407407407405,
+      "grad_norm": 1.5497296381413772,
+      "learning_rate": 1.9591851766389174e-07,
+      "loss": 0.7001,
+      "step": 2560
+    },
+    {
+      "epoch": 4.742592592592593,
+      "grad_norm": 1.3588472471020898,
+      "learning_rate": 1.959142765345493e-07,
+      "loss": 0.6969,
+      "step": 2561
+    },
+    {
+      "epoch": 4.7444444444444445,
+      "grad_norm": 2.144818587608991,
+      "learning_rate": 1.9591003324879338e-07,
+      "loss": 0.6832,
+      "step": 2562
+    },
+    {
+      "epoch": 4.746296296296296,
+      "grad_norm": 1.3072838602291186,
+      "learning_rate": 1.9590578780671945e-07,
+      "loss": 0.6666,
+      "step": 2563
+    },
+    {
+      "epoch": 4.7481481481481485,
+      "grad_norm": 1.3331818712867731,
+      "learning_rate": 1.9590154020842294e-07,
+      "loss": 0.6936,
+      "step": 2564
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 1.699866272058813,
+      "learning_rate": 1.9589729045399933e-07,
+      "loss": 0.7237,
+      "step": 2565
+    },
+    {
+      "epoch": 4.7518518518518515,
+      "grad_norm": 1.4264088677582054,
+      "learning_rate": 1.9589303854354416e-07,
+      "loss": 0.7232,
+      "step": 2566
+    },
+    {
+      "epoch": 4.753703703703704,
+      "grad_norm": 1.5549432734879838,
+      "learning_rate": 1.9588878447715306e-07,
+      "loss": 0.7117,
+      "step": 2567
+    },
+    {
+      "epoch": 4.7555555555555555,
+      "grad_norm": 1.4472904339605825,
+      "learning_rate": 1.9588452825492168e-07,
+      "loss": 0.7106,
+      "step": 2568
+    },
+    {
+      "epoch": 4.757407407407407,
+      "grad_norm": 1.5394126519144269,
+      "learning_rate": 1.9588026987694565e-07,
+      "loss": 0.6925,
+      "step": 2569
+    },
+    {
+      "epoch": 4.7592592592592595,
+      "grad_norm": 2.971944118039773,
+      "learning_rate": 1.9587600934332075e-07,
+      "loss": 0.6522,
+      "step": 2570
+    },
+    {
+      "epoch": 4.761111111111111,
+      "grad_norm": 1.3407780155578763,
+      "learning_rate": 1.9587174665414282e-07,
+      "loss": 0.6632,
+      "step": 2571
+    },
+    {
+      "epoch": 4.762962962962963,
+      "grad_norm": 1.6055322955615947,
+      "learning_rate": 1.9586748180950758e-07,
+      "loss": 0.6814,
+      "step": 2572
+    },
+    {
+      "epoch": 4.764814814814815,
+      "grad_norm": 1.4152320630850797,
+      "learning_rate": 1.95863214809511e-07,
+      "loss": 0.6733,
+      "step": 2573
+    },
+    {
+      "epoch": 4.766666666666667,
+      "grad_norm": 1.4929362577876002,
+      "learning_rate": 1.9585894565424898e-07,
+      "loss": 0.6939,
+      "step": 2574
+    },
+    {
+      "epoch": 4.768518518518518,
+      "grad_norm": 1.460521891075413,
+      "learning_rate": 1.9585467434381751e-07,
+      "loss": 0.6585,
+      "step": 2575
+    },
+    {
+      "epoch": 4.770370370370371,
+      "grad_norm": 1.461293730109857,
+      "learning_rate": 1.9585040087831265e-07,
+      "loss": 0.6135,
+      "step": 2576
+    },
+    {
+      "epoch": 4.772222222222222,
+      "grad_norm": 2.3813280478966883,
+      "learning_rate": 1.9584612525783042e-07,
+      "loss": 0.6652,
+      "step": 2577
+    },
+    {
+      "epoch": 4.774074074074074,
+      "grad_norm": 1.8412587455644527,
+      "learning_rate": 1.95841847482467e-07,
+      "loss": 0.6484,
+      "step": 2578
+    },
+    {
+      "epoch": 4.775925925925926,
+      "grad_norm": 1.4172860772977958,
+      "learning_rate": 1.9583756755231853e-07,
+      "loss": 0.6529,
+      "step": 2579
+    },
+    {
+      "epoch": 4.777777777777778,
+      "grad_norm": 1.5530412211210471,
+      "learning_rate": 1.9583328546748125e-07,
+      "loss": 0.6268,
+      "step": 2580
+    },
+    {
+      "epoch": 4.77962962962963,
+      "grad_norm": 1.535534316611214,
+      "learning_rate": 1.9582900122805145e-07,
+      "loss": 0.6641,
+      "step": 2581
+    },
+    {
+      "epoch": 4.781481481481482,
+      "grad_norm": 2.1748549856854567,
+      "learning_rate": 1.9582471483412543e-07,
+      "loss": 0.655,
+      "step": 2582
+    },
+    {
+      "epoch": 4.783333333333333,
+      "grad_norm": 1.6841573333989606,
+      "learning_rate": 1.9582042628579956e-07,
+      "loss": 0.6864,
+      "step": 2583
+    },
+    {
+      "epoch": 4.785185185185185,
+      "grad_norm": 1.5311357546678117,
+      "learning_rate": 1.9581613558317027e-07,
+      "loss": 0.6467,
+      "step": 2584
+    },
+    {
+      "epoch": 4.787037037037037,
+      "grad_norm": 2.8057207953903953,
+      "learning_rate": 1.95811842726334e-07,
+      "loss": 0.6969,
+      "step": 2585
+    },
+    {
+      "epoch": 4.788888888888889,
+      "grad_norm": 1.718009726212732,
+      "learning_rate": 1.9580754771538732e-07,
+      "loss": 0.6807,
+      "step": 2586
+    },
+    {
+      "epoch": 4.790740740740741,
+      "grad_norm": 1.4090495628613642,
+      "learning_rate": 1.9580325055042674e-07,
+      "loss": 0.7133,
+      "step": 2587
+    },
+    {
+      "epoch": 4.792592592592593,
+      "grad_norm": 2.0765659934498637,
+      "learning_rate": 1.9579895123154887e-07,
+      "loss": 0.7212,
+      "step": 2588
+    },
+    {
+      "epoch": 4.794444444444444,
+      "grad_norm": 1.5506387328654636,
+      "learning_rate": 1.9579464975885042e-07,
+      "loss": 0.6659,
+      "step": 2589
+    },
+    {
+      "epoch": 4.796296296296296,
+      "grad_norm": 1.3343560966769126,
+      "learning_rate": 1.9579034613242804e-07,
+      "loss": 0.6534,
+      "step": 2590
+    },
+    {
+      "epoch": 4.798148148148148,
+      "grad_norm": 1.5716720780290347,
+      "learning_rate": 1.9578604035237856e-07,
+      "loss": 0.691,
+      "step": 2591
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 1.3554224563661759,
+      "learning_rate": 1.957817324187987e-07,
+      "loss": 0.6639,
+      "step": 2592
+    },
+    {
+      "epoch": 4.801851851851852,
+      "grad_norm": 1.28732952185562,
+      "learning_rate": 1.9577742233178536e-07,
+      "loss": 0.6832,
+      "step": 2593
+    },
+    {
+      "epoch": 4.803703703703704,
+      "grad_norm": 1.5309043659866546,
+      "learning_rate": 1.9577311009143548e-07,
+      "loss": 0.6947,
+      "step": 2594
+    },
+    {
+      "epoch": 4.805555555555555,
+      "grad_norm": 1.453191398942573,
+      "learning_rate": 1.9576879569784592e-07,
+      "loss": 0.6971,
+      "step": 2595
+    },
+    {
+      "epoch": 4.807407407407408,
+      "grad_norm": 1.4143822486270237,
+      "learning_rate": 1.9576447915111377e-07,
+      "loss": 0.6854,
+      "step": 2596
+    },
+    {
+      "epoch": 4.809259259259259,
+      "grad_norm": 1.5755946097336484,
+      "learning_rate": 1.95760160451336e-07,
+      "loss": 0.6974,
+      "step": 2597
+    },
+    {
+      "epoch": 4.811111111111111,
+      "grad_norm": 1.4725985076973205,
+      "learning_rate": 1.9575583959860975e-07,
+      "loss": 0.6929,
+      "step": 2598
+    },
+    {
+      "epoch": 4.812962962962963,
+      "grad_norm": 2.002995867864467,
+      "learning_rate": 1.9575151659303218e-07,
+      "loss": 0.7016,
+      "step": 2599
+    },
+    {
+      "epoch": 4.814814814814815,
+      "grad_norm": 1.8102165871063058,
+      "learning_rate": 1.9574719143470042e-07,
+      "loss": 0.6662,
+      "step": 2600
+    },
+    {
+      "epoch": 4.816666666666666,
+      "grad_norm": 1.621177989958412,
+      "learning_rate": 1.9574286412371177e-07,
+      "loss": 0.7199,
+      "step": 2601
+    },
+    {
+      "epoch": 4.818518518518519,
+      "grad_norm": 1.5148548287176322,
+      "learning_rate": 1.9573853466016354e-07,
+      "loss": 0.6733,
+      "step": 2602
+    },
+    {
+      "epoch": 4.82037037037037,
+      "grad_norm": 1.8660690254127292,
+      "learning_rate": 1.95734203044153e-07,
+      "loss": 0.6594,
+      "step": 2603
+    },
+    {
+      "epoch": 4.822222222222222,
+      "grad_norm": 1.9403288070156381,
+      "learning_rate": 1.9572986927577754e-07,
+      "loss": 0.6994,
+      "step": 2604
+    },
+    {
+      "epoch": 4.824074074074074,
+      "grad_norm": 1.5574392648529414,
+      "learning_rate": 1.9572553335513465e-07,
+      "loss": 0.6442,
+      "step": 2605
+    },
+    {
+      "epoch": 4.825925925925926,
+      "grad_norm": 1.9651121053242995,
+      "learning_rate": 1.9572119528232175e-07,
+      "loss": 0.6684,
+      "step": 2606
+    },
+    {
+      "epoch": 4.8277777777777775,
+      "grad_norm": 1.4836938331881797,
+      "learning_rate": 1.957168550574364e-07,
+      "loss": 0.6564,
+      "step": 2607
+    },
+    {
+      "epoch": 4.82962962962963,
+      "grad_norm": 2.074361031300887,
+      "learning_rate": 1.9571251268057626e-07,
+      "loss": 0.7036,
+      "step": 2608
+    },
+    {
+      "epoch": 4.8314814814814815,
+      "grad_norm": 1.3913488668062026,
+      "learning_rate": 1.957081681518388e-07,
+      "loss": 0.6491,
+      "step": 2609
+    },
+    {
+      "epoch": 4.833333333333333,
+      "grad_norm": 1.6698805886455452,
+      "learning_rate": 1.9570382147132185e-07,
+      "loss": 0.6576,
+      "step": 2610
+    },
+    {
+      "epoch": 4.8351851851851855,
+      "grad_norm": 1.6692744102885133,
+      "learning_rate": 1.9569947263912302e-07,
+      "loss": 0.6918,
+      "step": 2611
+    },
+    {
+      "epoch": 4.837037037037037,
+      "grad_norm": 2.250237785541846,
+      "learning_rate": 1.9569512165534015e-07,
+      "loss": 0.6858,
+      "step": 2612
+    },
+    {
+      "epoch": 4.838888888888889,
+      "grad_norm": 2.0258299823379997,
+      "learning_rate": 1.9569076852007104e-07,
+      "loss": 0.6792,
+      "step": 2613
+    },
+    {
+      "epoch": 4.840740740740741,
+      "grad_norm": 1.3166064995721136,
+      "learning_rate": 1.9568641323341355e-07,
+      "loss": 0.6862,
+      "step": 2614
+    },
+    {
+      "epoch": 4.842592592592593,
+      "grad_norm": 1.6378534252343364,
+      "learning_rate": 1.9568205579546561e-07,
+      "loss": 0.6476,
+      "step": 2615
+    },
+    {
+      "epoch": 4.844444444444444,
+      "grad_norm": 1.7411619440573904,
+      "learning_rate": 1.956776962063252e-07,
+      "loss": 0.6766,
+      "step": 2616
+    },
+    {
+      "epoch": 4.8462962962962965,
+      "grad_norm": 1.4902231341231,
+      "learning_rate": 1.956733344660903e-07,
+      "loss": 0.698,
+      "step": 2617
+    },
+    {
+      "epoch": 4.848148148148148,
+      "grad_norm": 1.6826732544785905,
+      "learning_rate": 1.95668970574859e-07,
+      "loss": 0.6925,
+      "step": 2618
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 1.6132862339391547,
+      "learning_rate": 1.9566460453272943e-07,
+      "loss": 0.6426,
+      "step": 2619
+    },
+    {
+      "epoch": 4.851851851851852,
+      "grad_norm": 1.509499724322264,
+      "learning_rate": 1.9566023633979974e-07,
+      "loss": 0.6677,
+      "step": 2620
+    },
+    {
+      "epoch": 4.853703703703704,
+      "grad_norm": 1.4286323318046328,
+      "learning_rate": 1.956558659961681e-07,
+      "loss": 0.7127,
+      "step": 2621
+    },
+    {
+      "epoch": 4.855555555555555,
+      "grad_norm": 1.4604074391120105,
+      "learning_rate": 1.956514935019328e-07,
+      "loss": 0.7076,
+      "step": 2622
+    },
+    {
+      "epoch": 4.857407407407408,
+      "grad_norm": 1.3116592330715366,
+      "learning_rate": 1.9564711885719214e-07,
+      "loss": 0.6899,
+      "step": 2623
+    },
+    {
+      "epoch": 4.859259259259259,
+      "grad_norm": 1.4730802042284514,
+      "learning_rate": 1.9564274206204447e-07,
+      "loss": 0.6729,
+      "step": 2624
+    },
+    {
+      "epoch": 4.861111111111111,
+      "grad_norm": 1.4718604988630335,
+      "learning_rate": 1.9563836311658822e-07,
+      "loss": 0.7056,
+      "step": 2625
+    },
+    {
+      "epoch": 4.862962962962963,
+      "grad_norm": 1.4359698303780324,
+      "learning_rate": 1.956339820209218e-07,
+      "loss": 0.6728,
+      "step": 2626
+    },
+    {
+      "epoch": 4.864814814814815,
+      "grad_norm": 1.6351348722401102,
+      "learning_rate": 1.9562959877514374e-07,
+      "loss": 0.7104,
+      "step": 2627
+    },
+    {
+      "epoch": 4.866666666666667,
+      "grad_norm": 1.5378185101318458,
+      "learning_rate": 1.9562521337935253e-07,
+      "loss": 0.6885,
+      "step": 2628
+    },
+    {
+      "epoch": 4.868518518518519,
+      "grad_norm": 4.065624679685904,
+      "learning_rate": 1.9562082583364686e-07,
+      "loss": 0.6934,
+      "step": 2629
+    },
+    {
+      "epoch": 4.87037037037037,
+      "grad_norm": 1.5330305515150862,
+      "learning_rate": 1.956164361381253e-07,
+      "loss": 0.6545,
+      "step": 2630
+    },
+    {
+      "epoch": 4.872222222222222,
+      "grad_norm": 1.5792460026013166,
+      "learning_rate": 1.9561204429288657e-07,
+      "loss": 0.6585,
+      "step": 2631
+    },
+    {
+      "epoch": 4.874074074074074,
+      "grad_norm": 1.6852842892947653,
+      "learning_rate": 1.956076502980294e-07,
+      "loss": 0.6856,
+      "step": 2632
+    },
+    {
+      "epoch": 4.875925925925926,
+      "grad_norm": 1.446362265985485,
+      "learning_rate": 1.9560325415365257e-07,
+      "loss": 0.6691,
+      "step": 2633
+    },
+    {
+      "epoch": 4.877777777777778,
+      "grad_norm": 1.478944105102043,
+      "learning_rate": 1.9559885585985497e-07,
+      "loss": 0.6462,
+      "step": 2634
+    },
+    {
+      "epoch": 4.87962962962963,
+      "grad_norm": 1.3951419744936036,
+      "learning_rate": 1.9559445541673543e-07,
+      "loss": 0.6808,
+      "step": 2635
+    },
+    {
+      "epoch": 4.881481481481481,
+      "grad_norm": 1.754978855300922,
+      "learning_rate": 1.9559005282439292e-07,
+      "loss": 0.6429,
+      "step": 2636
+    },
+    {
+      "epoch": 4.883333333333333,
+      "grad_norm": 1.449261360830208,
+      "learning_rate": 1.9558564808292636e-07,
+      "loss": 0.6344,
+      "step": 2637
+    },
+    {
+      "epoch": 4.885185185185185,
+      "grad_norm": 1.5792897252697125,
+      "learning_rate": 1.9558124119243486e-07,
+      "loss": 0.6766,
+      "step": 2638
+    },
+    {
+      "epoch": 4.887037037037037,
+      "grad_norm": 1.7367672539313557,
+      "learning_rate": 1.9557683215301743e-07,
+      "loss": 0.66,
+      "step": 2639
+    },
+    {
+      "epoch": 4.888888888888889,
+      "grad_norm": 1.5862960703696574,
+      "learning_rate": 1.9557242096477327e-07,
+      "loss": 0.6994,
+      "step": 2640
+    },
+    {
+      "epoch": 4.890740740740741,
+      "grad_norm": 1.6922835335513207,
+      "learning_rate": 1.955680076278015e-07,
+      "loss": 0.694,
+      "step": 2641
+    },
+    {
+      "epoch": 4.892592592592592,
+      "grad_norm": 1.82944800147624,
+      "learning_rate": 1.9556359214220137e-07,
+      "loss": 0.705,
+      "step": 2642
+    },
+    {
+      "epoch": 4.894444444444445,
+      "grad_norm": 1.621980112509395,
+      "learning_rate": 1.955591745080721e-07,
+      "loss": 0.6569,
+      "step": 2643
+    },
+    {
+      "epoch": 4.896296296296296,
+      "grad_norm": 3.320650322721205,
+      "learning_rate": 1.9555475472551311e-07,
+      "loss": 0.6734,
+      "step": 2644
+    },
+    {
+      "epoch": 4.898148148148148,
+      "grad_norm": 1.5335604450726346,
+      "learning_rate": 1.9555033279462367e-07,
+      "loss": 0.6422,
+      "step": 2645
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 1.3712288995221626,
+      "learning_rate": 1.9554590871550327e-07,
+      "loss": 0.6917,
+      "step": 2646
+    },
+    {
+      "epoch": 4.901851851851852,
+      "grad_norm": 1.349011385329623,
+      "learning_rate": 1.955414824882513e-07,
+      "loss": 0.6918,
+      "step": 2647
+    },
+    {
+      "epoch": 4.9037037037037035,
+      "grad_norm": 1.2876496306287544,
+      "learning_rate": 1.9553705411296735e-07,
+      "loss": 0.7059,
+      "step": 2648
+    },
+    {
+      "epoch": 4.905555555555556,
+      "grad_norm": 1.480916643859869,
+      "learning_rate": 1.9553262358975094e-07,
+      "loss": 0.643,
+      "step": 2649
+    },
+    {
+      "epoch": 4.907407407407407,
+      "grad_norm": 1.565604936070093,
+      "learning_rate": 1.955281909187017e-07,
+      "loss": 0.7096,
+      "step": 2650
+    },
+    {
+      "epoch": 4.909259259259259,
+      "grad_norm": 1.4676536126054007,
+      "learning_rate": 1.9552375609991927e-07,
+      "loss": 0.6893,
+      "step": 2651
+    },
+    {
+      "epoch": 4.911111111111111,
+      "grad_norm": 1.3380304793967999,
+      "learning_rate": 1.9551931913350334e-07,
+      "loss": 0.6505,
+      "step": 2652
+    },
+    {
+      "epoch": 4.912962962962963,
+      "grad_norm": 1.5615450651318896,
+      "learning_rate": 1.9551488001955373e-07,
+      "loss": 0.6665,
+      "step": 2653
+    },
+    {
+      "epoch": 4.9148148148148145,
+      "grad_norm": 1.7988459133966521,
+      "learning_rate": 1.9551043875817017e-07,
+      "loss": 0.709,
+      "step": 2654
+    },
+    {
+      "epoch": 4.916666666666667,
+      "grad_norm": 1.3876556704870973,
+      "learning_rate": 1.9550599534945259e-07,
+      "loss": 0.6868,
+      "step": 2655
+    },
+    {
+      "epoch": 4.9185185185185185,
+      "grad_norm": 1.5617115311718706,
+      "learning_rate": 1.955015497935008e-07,
+      "loss": 0.6962,
+      "step": 2656
+    },
+    {
+      "epoch": 4.92037037037037,
+      "grad_norm": 1.8783542449190045,
+      "learning_rate": 1.9549710209041478e-07,
+      "loss": 0.6585,
+      "step": 2657
+    },
+    {
+      "epoch": 4.9222222222222225,
+      "grad_norm": 1.434083253111606,
+      "learning_rate": 1.9549265224029456e-07,
+      "loss": 0.663,
+      "step": 2658
+    },
+    {
+      "epoch": 4.924074074074074,
+      "grad_norm": 1.4424141133200097,
+      "learning_rate": 1.9548820024324017e-07,
+      "loss": 0.6924,
+      "step": 2659
+    },
+    {
+      "epoch": 4.925925925925926,
+      "grad_norm": 1.5150646349713737,
+      "learning_rate": 1.954837460993517e-07,
+      "loss": 0.685,
+      "step": 2660
+    },
+    {
+      "epoch": 4.927777777777778,
+      "grad_norm": 1.927555098185982,
+      "learning_rate": 1.9547928980872929e-07,
+      "loss": 0.7084,
+      "step": 2661
+    },
+    {
+      "epoch": 4.92962962962963,
+      "grad_norm": 1.6846270214958476,
+      "learning_rate": 1.9547483137147312e-07,
+      "loss": 0.6817,
+      "step": 2662
+    },
+    {
+      "epoch": 4.931481481481481,
+      "grad_norm": 1.5251097022343931,
+      "learning_rate": 1.9547037078768342e-07,
+      "loss": 0.6996,
+      "step": 2663
+    },
+    {
+      "epoch": 4.933333333333334,
+      "grad_norm": 2.3336481867084107,
+      "learning_rate": 1.9546590805746051e-07,
+      "loss": 0.6445,
+      "step": 2664
+    },
+    {
+      "epoch": 4.935185185185185,
+      "grad_norm": 1.4766742961198567,
+      "learning_rate": 1.954614431809047e-07,
+      "loss": 0.6804,
+      "step": 2665
+    },
+    {
+      "epoch": 4.937037037037037,
+      "grad_norm": 1.7973668305011095,
+      "learning_rate": 1.9545697615811637e-07,
+      "loss": 0.6712,
+      "step": 2666
+    },
+    {
+      "epoch": 4.938888888888889,
+      "grad_norm": 1.4628137784980484,
+      "learning_rate": 1.9545250698919597e-07,
+      "loss": 0.7261,
+      "step": 2667
+    },
+    {
+      "epoch": 4.940740740740741,
+      "grad_norm": 1.4439564852227373,
+      "learning_rate": 1.9544803567424398e-07,
+      "loss": 0.6986,
+      "step": 2668
+    },
+    {
+      "epoch": 4.942592592592592,
+      "grad_norm": 1.6770071853074,
+      "learning_rate": 1.954435622133609e-07,
+      "loss": 0.7004,
+      "step": 2669
+    },
+    {
+      "epoch": 4.944444444444445,
+      "grad_norm": 1.4701736895259265,
+      "learning_rate": 1.9543908660664733e-07,
+      "loss": 0.6944,
+      "step": 2670
+    },
+    {
+      "epoch": 4.946296296296296,
+      "grad_norm": 1.6089138863677834,
+      "learning_rate": 1.9543460885420386e-07,
+      "loss": 0.6573,
+      "step": 2671
+    },
+    {
+      "epoch": 4.948148148148148,
+      "grad_norm": 1.4658356954218936,
+      "learning_rate": 1.954301289561312e-07,
+      "loss": 0.6548,
+      "step": 2672
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 2.055443747313343,
+      "learning_rate": 1.9542564691253007e-07,
+      "loss": 0.7419,
+      "step": 2673
+    },
+    {
+      "epoch": 4.951851851851852,
+      "grad_norm": 1.407779024281975,
+      "learning_rate": 1.954211627235012e-07,
+      "loss": 0.6733,
+      "step": 2674
+    },
+    {
+      "epoch": 4.953703703703704,
+      "grad_norm": 1.737115123653284,
+      "learning_rate": 1.9541667638914546e-07,
+      "loss": 0.6543,
+      "step": 2675
+    },
+    {
+      "epoch": 4.955555555555556,
+      "grad_norm": 1.4803068485974478,
+      "learning_rate": 1.9541218790956367e-07,
+      "loss": 0.7033,
+      "step": 2676
+    },
+    {
+      "epoch": 4.957407407407407,
+      "grad_norm": 1.5013824918581729,
+      "learning_rate": 1.9540769728485674e-07,
+      "loss": 0.7117,
+      "step": 2677
+    },
+    {
+      "epoch": 4.959259259259259,
+      "grad_norm": 1.3156214150479348,
+      "learning_rate": 1.9540320451512567e-07,
+      "loss": 0.6867,
+      "step": 2678
+    },
+    {
+      "epoch": 4.961111111111111,
+      "grad_norm": 1.4407186739701892,
+      "learning_rate": 1.953987096004715e-07,
+      "loss": 0.6437,
+      "step": 2679
+    },
+    {
+      "epoch": 4.962962962962963,
+      "grad_norm": 1.5999497591386125,
+      "learning_rate": 1.9539421254099516e-07,
+      "loss": 0.7113,
+      "step": 2680
+    },
+    {
+      "epoch": 4.964814814814815,
+      "grad_norm": 1.4142931520112962,
+      "learning_rate": 1.9538971333679787e-07,
+      "loss": 0.7224,
+      "step": 2681
+    },
+    {
+      "epoch": 4.966666666666667,
+      "grad_norm": 1.3331528856433439,
+      "learning_rate": 1.9538521198798076e-07,
+      "loss": 0.701,
+      "step": 2682
+    },
+    {
+      "epoch": 4.968518518518518,
+      "grad_norm": 1.6423688070666986,
+      "learning_rate": 1.95380708494645e-07,
+      "loss": 0.6539,
+      "step": 2683
+    },
+    {
+      "epoch": 4.97037037037037,
+      "grad_norm": 1.3822785224406957,
+      "learning_rate": 1.953762028568919e-07,
+      "loss": 0.6442,
+      "step": 2684
+    },
+    {
+      "epoch": 4.972222222222222,
+      "grad_norm": 1.4700579032771612,
+      "learning_rate": 1.9537169507482267e-07,
+      "loss": 0.7062,
+      "step": 2685
+    },
+    {
+      "epoch": 4.974074074074074,
+      "grad_norm": 1.7922159796105215,
+      "learning_rate": 1.9536718514853873e-07,
+      "loss": 0.6698,
+      "step": 2686
+    },
+    {
+      "epoch": 4.975925925925926,
+      "grad_norm": 1.4652734939706755,
+      "learning_rate": 1.9536267307814147e-07,
+      "loss": 0.6257,
+      "step": 2687
+    },
+    {
+      "epoch": 4.977777777777778,
+      "grad_norm": 1.8360566393794944,
+      "learning_rate": 1.953581588637323e-07,
+      "loss": 0.709,
+      "step": 2688
+    },
+    {
+      "epoch": 4.979629629629629,
+      "grad_norm": 1.7996465469168061,
+      "learning_rate": 1.9535364250541274e-07,
+      "loss": 0.6548,
+      "step": 2689
+    },
+    {
+      "epoch": 4.981481481481482,
+      "grad_norm": 1.5514553935132824,
+      "learning_rate": 1.9534912400328433e-07,
+      "loss": 0.6584,
+      "step": 2690
+    },
+    {
+      "epoch": 4.983333333333333,
+      "grad_norm": 1.5941852534699035,
+      "learning_rate": 1.9534460335744864e-07,
+      "loss": 0.697,
+      "step": 2691
+    },
+    {
+      "epoch": 4.985185185185185,
+      "grad_norm": 1.5088265875512994,
+      "learning_rate": 1.953400805680073e-07,
+      "loss": 0.6538,
+      "step": 2692
+    },
+    {
+      "epoch": 4.987037037037037,
+      "grad_norm": 1.7594699460268615,
+      "learning_rate": 1.9533555563506202e-07,
+      "loss": 0.7083,
+      "step": 2693
+    },
+    {
+      "epoch": 4.988888888888889,
+      "grad_norm": 2.3528463802717994,
+      "learning_rate": 1.9533102855871454e-07,
+      "loss": 0.6566,
+      "step": 2694
+    },
+    {
+      "epoch": 4.9907407407407405,
+      "grad_norm": 1.6394322103731769,
+      "learning_rate": 1.953264993390666e-07,
+      "loss": 0.6996,
+      "step": 2695
+    },
+    {
+      "epoch": 4.992592592592593,
+      "grad_norm": 1.5742827726219282,
+      "learning_rate": 1.9532196797622005e-07,
+      "loss": 0.6638,
+      "step": 2696
+    },
+    {
+      "epoch": 4.9944444444444445,
+      "grad_norm": 1.4852115191028312,
+      "learning_rate": 1.9531743447027678e-07,
+      "loss": 0.6553,
+      "step": 2697
+    },
+    {
+      "epoch": 4.996296296296296,
+      "grad_norm": 1.7456692094689956,
+      "learning_rate": 1.953128988213387e-07,
+      "loss": 0.6787,
+      "step": 2698
+    },
+    {
+      "epoch": 4.9981481481481485,
+      "grad_norm": 1.4942473152661144,
+      "learning_rate": 1.9530836102950777e-07,
+      "loss": 0.6967,
+      "step": 2699
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 1.5898572034515874,
+      "learning_rate": 1.9530382109488609e-07,
+      "loss": 0.6387,
+      "step": 2700
+    },
+    {
+      "epoch": 5.0018518518518515,
+      "grad_norm": 1.8622424932831836,
+      "learning_rate": 1.9529927901757563e-07,
+      "loss": 0.6643,
+      "step": 2701
+    },
+    {
+      "epoch": 5.003703703703704,
+      "grad_norm": 1.7352849687357825,
+      "learning_rate": 1.9529473479767855e-07,
+      "loss": 0.6157,
+      "step": 2702
+    },
+    {
+      "epoch": 5.0055555555555555,
+      "grad_norm": 1.538745971169652,
+      "learning_rate": 1.95290188435297e-07,
+      "loss": 0.6936,
+      "step": 2703
+    },
+    {
+      "epoch": 5.007407407407407,
+      "grad_norm": 1.5123732026287093,
+      "learning_rate": 1.9528563993053325e-07,
+      "loss": 0.6456,
+      "step": 2704
+    },
+    {
+      "epoch": 5.0092592592592595,
+      "grad_norm": 1.7103623939784156,
+      "learning_rate": 1.9528108928348948e-07,
+      "loss": 0.671,
+      "step": 2705
+    },
+    {
+      "epoch": 5.011111111111111,
+      "grad_norm": 2.2441659113562724,
+      "learning_rate": 1.9527653649426807e-07,
+      "loss": 0.6841,
+      "step": 2706
+    },
+    {
+      "epoch": 5.012962962962963,
+      "grad_norm": 2.013139402638732,
+      "learning_rate": 1.9527198156297133e-07,
+      "loss": 0.6311,
+      "step": 2707
+    },
+    {
+      "epoch": 5.014814814814815,
+      "grad_norm": 1.4998899794777623,
+      "learning_rate": 1.952674244897017e-07,
+      "loss": 0.6905,
+      "step": 2708
+    },
+    {
+      "epoch": 5.016666666666667,
+      "grad_norm": 1.58898736856241,
+      "learning_rate": 1.9526286527456163e-07,
+      "loss": 0.6762,
+      "step": 2709
+    },
+    {
+      "epoch": 5.018518518518518,
+      "grad_norm": 1.4504511132858373,
+      "learning_rate": 1.9525830391765363e-07,
+      "loss": 0.6571,
+      "step": 2710
+    },
+    {
+      "epoch": 5.020370370370371,
+      "grad_norm": 1.941280591241212,
+      "learning_rate": 1.952537404190802e-07,
+      "loss": 0.6777,
+      "step": 2711
+    },
+    {
+      "epoch": 5.022222222222222,
+      "grad_norm": 1.5970643986501012,
+      "learning_rate": 1.9524917477894402e-07,
+      "loss": 0.6477,
+      "step": 2712
+    },
+    {
+      "epoch": 5.024074074074074,
+      "grad_norm": 3.47689999741679,
+      "learning_rate": 1.9524460699734768e-07,
+      "loss": 0.6798,
+      "step": 2713
+    },
+    {
+      "epoch": 5.025925925925926,
+      "grad_norm": 1.4961824294695965,
+      "learning_rate": 1.952400370743939e-07,
+      "loss": 0.6993,
+      "step": 2714
+    },
+    {
+      "epoch": 5.027777777777778,
+      "grad_norm": 1.4592406550752999,
+      "learning_rate": 1.9523546501018542e-07,
+      "loss": 0.6578,
+      "step": 2715
+    },
+    {
+      "epoch": 5.029629629629629,
+      "grad_norm": 1.5687043212999647,
+      "learning_rate": 1.95230890804825e-07,
+      "loss": 0.7169,
+      "step": 2716
+    },
+    {
+      "epoch": 5.031481481481482,
+      "grad_norm": 1.4784765478469364,
+      "learning_rate": 1.9522631445841559e-07,
+      "loss": 0.711,
+      "step": 2717
+    },
+    {
+      "epoch": 5.033333333333333,
+      "grad_norm": 1.4771210302151068,
+      "learning_rate": 1.9522173597105995e-07,
+      "loss": 0.7465,
+      "step": 2718
+    },
+    {
+      "epoch": 5.035185185185185,
+      "grad_norm": 2.294481126889156,
+      "learning_rate": 1.9521715534286104e-07,
+      "loss": 0.6801,
+      "step": 2719
+    },
+    {
+      "epoch": 5.037037037037037,
+      "grad_norm": 1.4992248393087069,
+      "learning_rate": 1.9521257257392192e-07,
+      "loss": 0.6614,
+      "step": 2720
+    },
+    {
+      "epoch": 5.038888888888889,
+      "grad_norm": 1.652181246645023,
+      "learning_rate": 1.9520798766434553e-07,
+      "loss": 0.6695,
+      "step": 2721
+    },
+    {
+      "epoch": 5.040740740740741,
+      "grad_norm": 1.5520728647514515,
+      "learning_rate": 1.9520340061423505e-07,
+      "loss": 0.6692,
+      "step": 2722
+    },
+    {
+      "epoch": 5.042592592592593,
+      "grad_norm": 1.642697153147436,
+      "learning_rate": 1.9519881142369353e-07,
+      "loss": 0.7678,
+      "step": 2723
+    },
+    {
+      "epoch": 5.044444444444444,
+      "grad_norm": 1.8413925747381803,
+      "learning_rate": 1.9519422009282417e-07,
+      "loss": 0.6902,
+      "step": 2724
+    },
+    {
+      "epoch": 5.046296296296297,
+      "grad_norm": 1.558458270045277,
+      "learning_rate": 1.9518962662173019e-07,
+      "loss": 0.644,
+      "step": 2725
+    },
+    {
+      "epoch": 5.048148148148148,
+      "grad_norm": 1.506289546349195,
+      "learning_rate": 1.951850310105149e-07,
+      "loss": 0.692,
+      "step": 2726
+    },
+    {
+      "epoch": 5.05,
+      "grad_norm": 1.742068937604263,
+      "learning_rate": 1.9518043325928156e-07,
+      "loss": 0.6747,
+      "step": 2727
+    },
+    {
+      "epoch": 5.051851851851852,
+      "grad_norm": 1.6323158604015076,
+      "learning_rate": 1.951758333681336e-07,
+      "loss": 0.718,
+      "step": 2728
+    },
+    {
+      "epoch": 5.053703703703704,
+      "grad_norm": 1.792627807649043,
+      "learning_rate": 1.951712313371744e-07,
+      "loss": 0.6745,
+      "step": 2729
+    },
+    {
+      "epoch": 5.055555555555555,
+      "grad_norm": 1.8264906424021514,
+      "learning_rate": 1.9516662716650742e-07,
+      "loss": 0.7001,
+      "step": 2730
+    },
+    {
+      "epoch": 5.057407407407408,
+      "grad_norm": 1.6061210899553822,
+      "learning_rate": 1.951620208562362e-07,
+      "loss": 0.6726,
+      "step": 2731
+    },
+    {
+      "epoch": 5.059259259259259,
+      "grad_norm": 1.4979117063166532,
+      "learning_rate": 1.9515741240646431e-07,
+      "loss": 0.6287,
+      "step": 2732
+    },
+    {
+      "epoch": 5.061111111111111,
+      "grad_norm": 2.266713071108368,
+      "learning_rate": 1.9515280181729536e-07,
+      "loss": 0.6875,
+      "step": 2733
+    },
+    {
+      "epoch": 5.062962962962963,
+      "grad_norm": 1.517888437424714,
+      "learning_rate": 1.95148189088833e-07,
+      "loss": 0.6352,
+      "step": 2734
+    },
+    {
+      "epoch": 5.064814814814815,
+      "grad_norm": 1.5294411305699602,
+      "learning_rate": 1.9514357422118088e-07,
+      "loss": 0.6371,
+      "step": 2735
+    },
+    {
+      "epoch": 5.066666666666666,
+      "grad_norm": 1.4703201295182762,
+      "learning_rate": 1.9513895721444284e-07,
+      "loss": 0.6878,
+      "step": 2736
+    },
+    {
+      "epoch": 5.068518518518519,
+      "grad_norm": 1.9091504529435155,
+      "learning_rate": 1.9513433806872266e-07,
+      "loss": 0.6675,
+      "step": 2737
+    },
+    {
+      "epoch": 5.07037037037037,
+      "grad_norm": 1.535683250640905,
+      "learning_rate": 1.9512971678412416e-07,
+      "loss": 0.6354,
+      "step": 2738
+    },
+    {
+      "epoch": 5.072222222222222,
+      "grad_norm": 1.3864964391489514,
+      "learning_rate": 1.9512509336075124e-07,
+      "loss": 0.6704,
+      "step": 2739
+    },
+    {
+      "epoch": 5.074074074074074,
+      "grad_norm": 1.7633170844137405,
+      "learning_rate": 1.951204677987079e-07,
+      "loss": 0.6622,
+      "step": 2740
+    },
+    {
+      "epoch": 5.075925925925926,
+      "grad_norm": 1.5562407546239936,
+      "learning_rate": 1.9511584009809808e-07,
+      "loss": 0.635,
+      "step": 2741
+    },
+    {
+      "epoch": 5.0777777777777775,
+      "grad_norm": 2.5331213736466687,
+      "learning_rate": 1.9511121025902585e-07,
+      "loss": 0.6841,
+      "step": 2742
+    },
+    {
+      "epoch": 5.07962962962963,
+      "grad_norm": 1.8154606730327534,
+      "learning_rate": 1.951065782815953e-07,
+      "loss": 0.6569,
+      "step": 2743
+    },
+    {
+      "epoch": 5.0814814814814815,
+      "grad_norm": 1.5684281231840984,
+      "learning_rate": 1.9510194416591057e-07,
+      "loss": 0.6702,
+      "step": 2744
+    },
+    {
+      "epoch": 5.083333333333333,
+      "grad_norm": 1.440241814918552,
+      "learning_rate": 1.950973079120758e-07,
+      "loss": 0.6678,
+      "step": 2745
+    },
+    {
+      "epoch": 5.0851851851851855,
+      "grad_norm": 1.5155773370058676,
+      "learning_rate": 1.950926695201953e-07,
+      "loss": 0.6527,
+      "step": 2746
+    },
+    {
+      "epoch": 5.087037037037037,
+      "grad_norm": 2.3661705015095427,
+      "learning_rate": 1.9508802899037334e-07,
+      "loss": 0.6836,
+      "step": 2747
+    },
+    {
+      "epoch": 5.088888888888889,
+      "grad_norm": 1.5130410631065652,
+      "learning_rate": 1.950833863227142e-07,
+      "loss": 0.6919,
+      "step": 2748
+    },
+    {
+      "epoch": 5.090740740740741,
+      "grad_norm": 1.6021419486188553,
+      "learning_rate": 1.9507874151732234e-07,
+      "loss": 0.7322,
+      "step": 2749
+    },
+    {
+      "epoch": 5.092592592592593,
+      "grad_norm": 2.0235869546472274,
+      "learning_rate": 1.950740945743021e-07,
+      "loss": 0.7066,
+      "step": 2750
+    },
+    {
+      "epoch": 5.094444444444444,
+      "grad_norm": 1.8970633116807134,
+      "learning_rate": 1.95069445493758e-07,
+      "loss": 0.667,
+      "step": 2751
+    },
+    {
+      "epoch": 5.0962962962962965,
+      "grad_norm": 1.6674380557253778,
+      "learning_rate": 1.9506479427579454e-07,
+      "loss": 0.7001,
+      "step": 2752
+    },
+    {
+      "epoch": 5.098148148148148,
+      "grad_norm": 1.430605359305374,
+      "learning_rate": 1.9506014092051637e-07,
+      "loss": 0.682,
+      "step": 2753
+    },
+    {
+      "epoch": 5.1,
+      "grad_norm": 1.4845448059912651,
+      "learning_rate": 1.9505548542802802e-07,
+      "loss": 0.6929,
+      "step": 2754
+    },
+    {
+      "epoch": 5.101851851851852,
+      "grad_norm": 1.4359422915676057,
+      "learning_rate": 1.950508277984342e-07,
+      "loss": 0.7202,
+      "step": 2755
+    },
+    {
+      "epoch": 5.103703703703704,
+      "grad_norm": 1.5038846098969982,
+      "learning_rate": 1.950461680318396e-07,
+      "loss": 0.6521,
+      "step": 2756
+    },
+    {
+      "epoch": 5.105555555555555,
+      "grad_norm": 1.5236839483000373,
+      "learning_rate": 1.9504150612834899e-07,
+      "loss": 0.6454,
+      "step": 2757
+    },
+    {
+      "epoch": 5.107407407407408,
+      "grad_norm": 1.46801847253954,
+      "learning_rate": 1.9503684208806722e-07,
+      "loss": 0.7031,
+      "step": 2758
+    },
+    {
+      "epoch": 5.109259259259259,
+      "grad_norm": 1.3871955504440538,
+      "learning_rate": 1.950321759110991e-07,
+      "loss": 0.7016,
+      "step": 2759
+    },
+    {
+      "epoch": 5.111111111111111,
+      "grad_norm": 1.4539471871797784,
+      "learning_rate": 1.950275075975496e-07,
+      "loss": 0.6648,
+      "step": 2760
+    },
+    {
+      "epoch": 5.112962962962963,
+      "grad_norm": 1.615297523484419,
+      "learning_rate": 1.950228371475236e-07,
+      "loss": 0.6567,
+      "step": 2761
+    },
+    {
+      "epoch": 5.114814814814815,
+      "grad_norm": 1.474463806413928,
+      "learning_rate": 1.9501816456112618e-07,
+      "loss": 0.6704,
+      "step": 2762
+    },
+    {
+      "epoch": 5.116666666666666,
+      "grad_norm": 2.012067588383928,
+      "learning_rate": 1.9501348983846232e-07,
+      "loss": 0.6597,
+      "step": 2763
+    },
+    {
+      "epoch": 5.118518518518519,
+      "grad_norm": 1.5565270378519152,
+      "learning_rate": 1.9500881297963718e-07,
+      "loss": 0.708,
+      "step": 2764
+    },
+    {
+      "epoch": 5.12037037037037,
+      "grad_norm": 1.7055643886980758,
+      "learning_rate": 1.9500413398475587e-07,
+      "loss": 0.6618,
+      "step": 2765
+    },
+    {
+      "epoch": 5.122222222222222,
+      "grad_norm": 1.5020234475682968,
+      "learning_rate": 1.949994528539236e-07,
+      "loss": 0.6551,
+      "step": 2766
+    },
+    {
+      "epoch": 5.124074074074074,
+      "grad_norm": 1.5178357896780215,
+      "learning_rate": 1.9499476958724563e-07,
+      "loss": 0.6813,
+      "step": 2767
+    },
+    {
+      "epoch": 5.125925925925926,
+      "grad_norm": 1.4922108144510435,
+      "learning_rate": 1.9499008418482724e-07,
+      "loss": 0.6298,
+      "step": 2768
+    },
+    {
+      "epoch": 5.127777777777778,
+      "grad_norm": 1.6501580057463796,
+      "learning_rate": 1.9498539664677374e-07,
+      "loss": 0.7192,
+      "step": 2769
+    },
+    {
+      "epoch": 5.12962962962963,
+      "grad_norm": 1.823136672728746,
+      "learning_rate": 1.9498070697319057e-07,
+      "loss": 0.669,
+      "step": 2770
+    },
+    {
+      "epoch": 5.131481481481481,
+      "grad_norm": 1.6124739557201277,
+      "learning_rate": 1.9497601516418314e-07,
+      "loss": 0.6677,
+      "step": 2771
+    },
+    {
+      "epoch": 5.133333333333334,
+      "grad_norm": 1.6127603035957159,
+      "learning_rate": 1.9497132121985694e-07,
+      "loss": 0.6689,
+      "step": 2772
+    },
+    {
+      "epoch": 5.135185185185185,
+      "grad_norm": 1.459456301463151,
+      "learning_rate": 1.949666251403175e-07,
+      "loss": 0.6022,
+      "step": 2773
+    },
+    {
+      "epoch": 5.137037037037037,
+      "grad_norm": 1.7149872704751348,
+      "learning_rate": 1.949619269256704e-07,
+      "loss": 0.6667,
+      "step": 2774
+    },
+    {
+      "epoch": 5.138888888888889,
+      "grad_norm": 1.4993614782027365,
+      "learning_rate": 1.9495722657602128e-07,
+      "loss": 0.6647,
+      "step": 2775
+    },
+    {
+      "epoch": 5.140740740740741,
+      "grad_norm": 1.460664552811264,
+      "learning_rate": 1.9495252409147578e-07,
+      "loss": 0.6561,
+      "step": 2776
+    },
+    {
+      "epoch": 5.142592592592592,
+      "grad_norm": 1.4331919020484565,
+      "learning_rate": 1.9494781947213967e-07,
+      "loss": 0.6601,
+      "step": 2777
+    },
+    {
+      "epoch": 5.144444444444445,
+      "grad_norm": 1.3402165234324948,
+      "learning_rate": 1.949431127181187e-07,
+      "loss": 0.6605,
+      "step": 2778
+    },
+    {
+      "epoch": 5.146296296296296,
+      "grad_norm": 1.4228384370319394,
+      "learning_rate": 1.9493840382951866e-07,
+      "loss": 0.674,
+      "step": 2779
+    },
+    {
+      "epoch": 5.148148148148148,
+      "grad_norm": 1.4340539119694766,
+      "learning_rate": 1.9493369280644552e-07,
+      "loss": 0.7383,
+      "step": 2780
+    },
+    {
+      "epoch": 5.15,
+      "grad_norm": 1.479137910591616,
+      "learning_rate": 1.949289796490051e-07,
+      "loss": 0.7143,
+      "step": 2781
+    },
+    {
+      "epoch": 5.151851851851852,
+      "grad_norm": 1.376411319707628,
+      "learning_rate": 1.949242643573034e-07,
+      "loss": 0.6821,
+      "step": 2782
+    },
+    {
+      "epoch": 5.1537037037037035,
+      "grad_norm": 1.578345934052031,
+      "learning_rate": 1.9491954693144642e-07,
+      "loss": 0.6961,
+      "step": 2783
+    },
+    {
+      "epoch": 5.155555555555556,
+      "grad_norm": 1.2859267621520243,
+      "learning_rate": 1.9491482737154018e-07,
+      "loss": 0.6728,
+      "step": 2784
+    },
+    {
+      "epoch": 5.157407407407407,
+      "grad_norm": 1.4219097751728964,
+      "learning_rate": 1.9491010567769092e-07,
+      "loss": 0.6543,
+      "step": 2785
+    },
+    {
+      "epoch": 5.159259259259259,
+      "grad_norm": 1.9218915234661444,
+      "learning_rate": 1.9490538185000465e-07,
+      "loss": 0.6895,
+      "step": 2786
+    },
+    {
+      "epoch": 5.161111111111111,
+      "grad_norm": 2.6321554666272706,
+      "learning_rate": 1.949006558885877e-07,
+      "loss": 0.6907,
+      "step": 2787
+    },
+    {
+      "epoch": 5.162962962962963,
+      "grad_norm": 1.6683796734423295,
+      "learning_rate": 1.948959277935462e-07,
+      "loss": 0.6757,
+      "step": 2788
+    },
+    {
+      "epoch": 5.1648148148148145,
+      "grad_norm": 3.164475198559798,
+      "learning_rate": 1.9489119756498653e-07,
+      "loss": 0.7005,
+      "step": 2789
+    },
+    {
+      "epoch": 5.166666666666667,
+      "grad_norm": 1.3947709393826655,
+      "learning_rate": 1.94886465203015e-07,
+      "loss": 0.6643,
+      "step": 2790
+    },
+    {
+      "epoch": 5.1685185185185185,
+      "grad_norm": 1.407047543237564,
+      "learning_rate": 1.9488173070773805e-07,
+      "loss": 0.7031,
+      "step": 2791
+    },
+    {
+      "epoch": 5.17037037037037,
+      "grad_norm": 1.724143373380897,
+      "learning_rate": 1.948769940792621e-07,
+      "loss": 0.7153,
+      "step": 2792
+    },
+    {
+      "epoch": 5.1722222222222225,
+      "grad_norm": 1.6040764749658936,
+      "learning_rate": 1.9487225531769363e-07,
+      "loss": 0.6709,
+      "step": 2793
+    },
+    {
+      "epoch": 5.174074074074074,
+      "grad_norm": 1.4574123916461896,
+      "learning_rate": 1.948675144231392e-07,
+      "loss": 0.6957,
+      "step": 2794
+    },
+    {
+      "epoch": 5.175925925925926,
+      "grad_norm": 2.11707214898699,
+      "learning_rate": 1.948627713957054e-07,
+      "loss": 0.6914,
+      "step": 2795
+    },
+    {
+      "epoch": 5.177777777777778,
+      "grad_norm": 1.293401916482147,
+      "learning_rate": 1.948580262354988e-07,
+      "loss": 0.6873,
+      "step": 2796
+    },
+    {
+      "epoch": 5.17962962962963,
+      "grad_norm": 1.216865363540106,
+      "learning_rate": 1.948532789426262e-07,
+      "loss": 0.6683,
+      "step": 2797
+    },
+    {
+      "epoch": 5.181481481481481,
+      "grad_norm": 1.5937996438551358,
+      "learning_rate": 1.9484852951719425e-07,
+      "loss": 0.611,
+      "step": 2798
+    },
+    {
+      "epoch": 5.183333333333334,
+      "grad_norm": 1.7804555710582373,
+      "learning_rate": 1.9484377795930976e-07,
+      "loss": 0.7335,
+      "step": 2799
+    },
+    {
+      "epoch": 5.185185185185185,
+      "grad_norm": 1.6367260097845768,
+      "learning_rate": 1.9483902426907952e-07,
+      "loss": 0.7165,
+      "step": 2800
+    },
+    {
+      "epoch": 5.187037037037037,
+      "grad_norm": 1.9573853835410369,
+      "learning_rate": 1.9483426844661049e-07,
+      "loss": 0.642,
+      "step": 2801
+    },
+    {
+      "epoch": 5.188888888888889,
+      "grad_norm": 1.2969647223491638,
+      "learning_rate": 1.9482951049200952e-07,
+      "loss": 0.7149,
+      "step": 2802
+    },
+    {
+      "epoch": 5.190740740740741,
+      "grad_norm": 1.4353197914430602,
+      "learning_rate": 1.9482475040538358e-07,
+      "loss": 0.6546,
+      "step": 2803
+    },
+    {
+      "epoch": 5.192592592592592,
+      "grad_norm": 1.3875427752324039,
+      "learning_rate": 1.9481998818683972e-07,
+      "loss": 0.6609,
+      "step": 2804
+    },
+    {
+      "epoch": 5.194444444444445,
+      "grad_norm": 1.766264695603854,
+      "learning_rate": 1.9481522383648498e-07,
+      "loss": 0.6881,
+      "step": 2805
+    },
+    {
+      "epoch": 5.196296296296296,
+      "grad_norm": 1.6005034787564454,
+      "learning_rate": 1.948104573544265e-07,
+      "loss": 0.6429,
+      "step": 2806
+    },
+    {
+      "epoch": 5.198148148148148,
+      "grad_norm": 2.1307872920913824,
+      "learning_rate": 1.9480568874077146e-07,
+      "loss": 0.7005,
+      "step": 2807
+    },
+    {
+      "epoch": 5.2,
+      "grad_norm": 1.5677577449065927,
+      "learning_rate": 1.9480091799562703e-07,
+      "loss": 0.6611,
+      "step": 2808
+    },
+    {
+      "epoch": 5.201851851851852,
+      "grad_norm": 3.864663361425221,
+      "learning_rate": 1.9479614511910052e-07,
+      "loss": 0.6512,
+      "step": 2809
+    },
+    {
+      "epoch": 5.203703703703703,
+      "grad_norm": 1.7471123822403607,
+      "learning_rate": 1.9479137011129918e-07,
+      "loss": 0.7004,
+      "step": 2810
+    },
+    {
+      "epoch": 5.205555555555556,
+      "grad_norm": 1.5362430901262045,
+      "learning_rate": 1.9478659297233039e-07,
+      "loss": 0.6252,
+      "step": 2811
+    },
+    {
+      "epoch": 5.207407407407407,
+      "grad_norm": 1.477713954296486,
+      "learning_rate": 1.9478181370230153e-07,
+      "loss": 0.7105,
+      "step": 2812
+    },
+    {
+      "epoch": 5.209259259259259,
+      "grad_norm": 1.7323259525957793,
+      "learning_rate": 1.9477703230132012e-07,
+      "loss": 0.6576,
+      "step": 2813
+    },
+    {
+      "epoch": 5.211111111111111,
+      "grad_norm": 1.768586985527259,
+      "learning_rate": 1.9477224876949358e-07,
+      "loss": 0.6646,
+      "step": 2814
+    },
+    {
+      "epoch": 5.212962962962963,
+      "grad_norm": 1.5020262897631673,
+      "learning_rate": 1.9476746310692952e-07,
+      "loss": 0.7046,
+      "step": 2815
+    },
+    {
+      "epoch": 5.214814814814815,
+      "grad_norm": 2.3730736753357586,
+      "learning_rate": 1.9476267531373547e-07,
+      "loss": 0.6391,
+      "step": 2816
+    },
+    {
+      "epoch": 5.216666666666667,
+      "grad_norm": 1.7314570247601258,
+      "learning_rate": 1.9475788539001914e-07,
+      "loss": 0.6438,
+      "step": 2817
+    },
+    {
+      "epoch": 5.218518518518518,
+      "grad_norm": 1.484058269346073,
+      "learning_rate": 1.9475309333588815e-07,
+      "loss": 0.6714,
+      "step": 2818
+    },
+    {
+      "epoch": 5.220370370370371,
+      "grad_norm": 1.6954949408772848,
+      "learning_rate": 1.9474829915145033e-07,
+      "loss": 0.6579,
+      "step": 2819
+    },
+    {
+      "epoch": 5.222222222222222,
+      "grad_norm": 1.8071272843783646,
+      "learning_rate": 1.9474350283681337e-07,
+      "loss": 0.6646,
+      "step": 2820
+    },
+    {
+      "epoch": 5.224074074074074,
+      "grad_norm": 1.5677656925350874,
+      "learning_rate": 1.9473870439208514e-07,
+      "loss": 0.6783,
+      "step": 2821
+    },
+    {
+      "epoch": 5.225925925925926,
+      "grad_norm": 1.5609470747266903,
+      "learning_rate": 1.9473390381737355e-07,
+      "loss": 0.6711,
+      "step": 2822
+    },
+    {
+      "epoch": 5.227777777777778,
+      "grad_norm": 1.9426868122481387,
+      "learning_rate": 1.947291011127865e-07,
+      "loss": 0.6641,
+      "step": 2823
+    },
+    {
+      "epoch": 5.229629629629629,
+      "grad_norm": 1.4698191919185866,
+      "learning_rate": 1.9472429627843202e-07,
+      "loss": 0.7239,
+      "step": 2824
+    },
+    {
+      "epoch": 5.231481481481482,
+      "grad_norm": 1.4336401393811464,
+      "learning_rate": 1.9471948931441803e-07,
+      "loss": 0.6024,
+      "step": 2825
+    },
+    {
+      "epoch": 5.233333333333333,
+      "grad_norm": 1.7134130567823629,
+      "learning_rate": 1.947146802208527e-07,
+      "loss": 0.6364,
+      "step": 2826
+    },
+    {
+      "epoch": 5.235185185185185,
+      "grad_norm": 1.8008859986136447,
+      "learning_rate": 1.947098689978441e-07,
+      "loss": 0.631,
+      "step": 2827
+    },
+    {
+      "epoch": 5.237037037037037,
+      "grad_norm": 1.3728328385954804,
+      "learning_rate": 1.9470505564550045e-07,
+      "loss": 0.6792,
+      "step": 2828
+    },
+    {
+      "epoch": 5.238888888888889,
+      "grad_norm": 2.083104488316351,
+      "learning_rate": 1.947002401639299e-07,
+      "loss": 0.6897,
+      "step": 2829
+    },
+    {
+      "epoch": 5.2407407407407405,
+      "grad_norm": 1.7336016394842522,
+      "learning_rate": 1.9469542255324077e-07,
+      "loss": 0.6883,
+      "step": 2830
+    },
+    {
+      "epoch": 5.242592592592593,
+      "grad_norm": 1.4560158522070599,
+      "learning_rate": 1.9469060281354134e-07,
+      "loss": 0.6773,
+      "step": 2831
+    },
+    {
+      "epoch": 5.2444444444444445,
+      "grad_norm": 1.739413226781611,
+      "learning_rate": 1.9468578094493997e-07,
+      "loss": 0.6408,
+      "step": 2832
+    },
+    {
+      "epoch": 5.246296296296296,
+      "grad_norm": 1.5705140806790685,
+      "learning_rate": 1.946809569475451e-07,
+      "loss": 0.6874,
+      "step": 2833
+    },
+    {
+      "epoch": 5.2481481481481485,
+      "grad_norm": 1.7579573129028792,
+      "learning_rate": 1.9467613082146516e-07,
+      "loss": 0.6819,
+      "step": 2834
+    },
+    {
+      "epoch": 5.25,
+      "grad_norm": 1.5817169405323315,
+      "learning_rate": 1.9467130256680866e-07,
+      "loss": 0.6829,
+      "step": 2835
+    },
+    {
+      "epoch": 5.2518518518518515,
+      "grad_norm": 2.244142383353803,
+      "learning_rate": 1.9466647218368417e-07,
+      "loss": 0.6884,
+      "step": 2836
+    },
+    {
+      "epoch": 5.253703703703704,
+      "grad_norm": 1.6363122054091424,
+      "learning_rate": 1.9466163967220028e-07,
+      "loss": 0.7068,
+      "step": 2837
+    },
+    {
+      "epoch": 5.2555555555555555,
+      "grad_norm": 1.6144176275240842,
+      "learning_rate": 1.946568050324656e-07,
+      "loss": 0.716,
+      "step": 2838
+    },
+    {
+      "epoch": 5.257407407407407,
+      "grad_norm": 1.3674793886344967,
+      "learning_rate": 1.946519682645889e-07,
+      "loss": 0.6745,
+      "step": 2839
+    },
+    {
+      "epoch": 5.2592592592592595,
+      "grad_norm": 1.3938275450288142,
+      "learning_rate": 1.9464712936867885e-07,
+      "loss": 0.6904,
+      "step": 2840
+    },
+    {
+      "epoch": 5.261111111111111,
+      "grad_norm": 1.4049026712184929,
+      "learning_rate": 1.9464228834484428e-07,
+      "loss": 0.7159,
+      "step": 2841
+    },
+    {
+      "epoch": 5.262962962962963,
+      "grad_norm": 1.5771338785707232,
+      "learning_rate": 1.9463744519319402e-07,
+      "loss": 0.6726,
+      "step": 2842
+    },
+    {
+      "epoch": 5.264814814814815,
+      "grad_norm": 2.9542976555884404,
+      "learning_rate": 1.94632599913837e-07,
+      "loss": 0.6779,
+      "step": 2843
+    },
+    {
+      "epoch": 5.266666666666667,
+      "grad_norm": 1.3404559966669218,
+      "learning_rate": 1.9462775250688205e-07,
+      "loss": 0.6808,
+      "step": 2844
+    },
+    {
+      "epoch": 5.268518518518518,
+      "grad_norm": 1.3822389079831359,
+      "learning_rate": 1.9462290297243826e-07,
+      "loss": 0.715,
+      "step": 2845
+    },
+    {
+      "epoch": 5.270370370370371,
+      "grad_norm": 1.3824869495723504,
+      "learning_rate": 1.9461805131061463e-07,
+      "loss": 0.6578,
+      "step": 2846
+    },
+    {
+      "epoch": 5.272222222222222,
+      "grad_norm": 1.4505436950571033,
+      "learning_rate": 1.9461319752152022e-07,
+      "loss": 0.6757,
+      "step": 2847
+    },
+    {
+      "epoch": 5.274074074074074,
+      "grad_norm": 1.3475952155856883,
+      "learning_rate": 1.9460834160526413e-07,
+      "loss": 0.6874,
+      "step": 2848
+    },
+    {
+      "epoch": 5.275925925925926,
+      "grad_norm": 1.4689952822308374,
+      "learning_rate": 1.946034835619556e-07,
+      "loss": 0.6566,
+      "step": 2849
+    },
+    {
+      "epoch": 5.277777777777778,
+      "grad_norm": 1.7297581775706106,
+      "learning_rate": 1.945986233917038e-07,
+      "loss": 0.6957,
+      "step": 2850
+    },
+    {
+      "epoch": 5.279629629629629,
+      "grad_norm": 1.659480326424263,
+      "learning_rate": 1.9459376109461804e-07,
+      "loss": 0.7062,
+      "step": 2851
+    },
+    {
+      "epoch": 5.281481481481482,
+      "grad_norm": 1.4682629536535758,
+      "learning_rate": 1.9458889667080763e-07,
+      "loss": 0.6386,
+      "step": 2852
+    },
+    {
+      "epoch": 5.283333333333333,
+      "grad_norm": 1.5439377986448244,
+      "learning_rate": 1.945840301203819e-07,
+      "loss": 0.6881,
+      "step": 2853
+    },
+    {
+      "epoch": 5.285185185185185,
+      "grad_norm": 1.3735687875444373,
+      "learning_rate": 1.945791614434503e-07,
+      "loss": 0.6853,
+      "step": 2854
+    },
+    {
+      "epoch": 5.287037037037037,
+      "grad_norm": 1.5970107063251036,
+      "learning_rate": 1.9457429064012225e-07,
+      "loss": 0.6729,
+      "step": 2855
+    },
+    {
+      "epoch": 5.288888888888889,
+      "grad_norm": 1.5541939396683784,
+      "learning_rate": 1.9456941771050733e-07,
+      "loss": 0.7082,
+      "step": 2856
+    },
+    {
+      "epoch": 5.290740740740741,
+      "grad_norm": 1.4020851123494547,
+      "learning_rate": 1.9456454265471502e-07,
+      "loss": 0.6781,
+      "step": 2857
+    },
+    {
+      "epoch": 5.292592592592593,
+      "grad_norm": 1.6694653739531418,
+      "learning_rate": 1.9455966547285497e-07,
+      "loss": 0.6626,
+      "step": 2858
+    },
+    {
+      "epoch": 5.294444444444444,
+      "grad_norm": 1.577576229774534,
+      "learning_rate": 1.9455478616503683e-07,
+      "loss": 0.6745,
+      "step": 2859
+    },
+    {
+      "epoch": 5.296296296296296,
+      "grad_norm": 1.8847629395762089,
+      "learning_rate": 1.9454990473137026e-07,
+      "loss": 0.6914,
+      "step": 2860
+    },
+    {
+      "epoch": 5.298148148148148,
+      "grad_norm": 1.503833375461945,
+      "learning_rate": 1.9454502117196508e-07,
+      "loss": 0.6536,
+      "step": 2861
+    },
+    {
+      "epoch": 5.3,
+      "grad_norm": 1.358700152789321,
+      "learning_rate": 1.94540135486931e-07,
+      "loss": 0.6216,
+      "step": 2862
+    },
+    {
+      "epoch": 5.301851851851852,
+      "grad_norm": 1.4050809930502495,
+      "learning_rate": 1.9453524767637792e-07,
+      "loss": 0.6893,
+      "step": 2863
+    },
+    {
+      "epoch": 5.303703703703704,
+      "grad_norm": 1.8220424979567904,
+      "learning_rate": 1.9453035774041573e-07,
+      "loss": 0.6506,
+      "step": 2864
+    },
+    {
+      "epoch": 5.305555555555555,
+      "grad_norm": 1.401604670019065,
+      "learning_rate": 1.9452546567915435e-07,
+      "loss": 0.6716,
+      "step": 2865
+    },
+    {
+      "epoch": 5.307407407407408,
+      "grad_norm": 1.6118042619586435,
+      "learning_rate": 1.9452057149270377e-07,
+      "loss": 0.6495,
+      "step": 2866
+    },
+    {
+      "epoch": 5.309259259259259,
+      "grad_norm": 1.5542989499933388,
+      "learning_rate": 1.9451567518117405e-07,
+      "loss": 0.6912,
+      "step": 2867
+    },
+    {
+      "epoch": 5.311111111111111,
+      "grad_norm": 1.5381582135417735,
+      "learning_rate": 1.9451077674467523e-07,
+      "loss": 0.6593,
+      "step": 2868
+    },
+    {
+      "epoch": 5.312962962962963,
+      "grad_norm": 1.5042316006003649,
+      "learning_rate": 1.9450587618331747e-07,
+      "loss": 0.6998,
+      "step": 2869
+    },
+    {
+      "epoch": 5.314814814814815,
+      "grad_norm": 1.5120526154181062,
+      "learning_rate": 1.9450097349721094e-07,
+      "loss": 0.6759,
+      "step": 2870
+    },
+    {
+      "epoch": 5.316666666666666,
+      "grad_norm": 1.51614906203917,
+      "learning_rate": 1.9449606868646587e-07,
+      "loss": 0.7016,
+      "step": 2871
+    },
+    {
+      "epoch": 5.318518518518519,
+      "grad_norm": 1.428776845741125,
+      "learning_rate": 1.944911617511925e-07,
+      "loss": 0.6558,
+      "step": 2872
+    },
+    {
+      "epoch": 5.32037037037037,
+      "grad_norm": 1.6854191845790605,
+      "learning_rate": 1.944862526915012e-07,
+      "loss": 0.6837,
+      "step": 2873
+    },
+    {
+      "epoch": 5.322222222222222,
+      "grad_norm": 1.7341208906467187,
+      "learning_rate": 1.9448134150750233e-07,
+      "loss": 0.6367,
+      "step": 2874
+    },
+    {
+      "epoch": 5.324074074074074,
+      "grad_norm": 1.5528532968680224,
+      "learning_rate": 1.944764281993063e-07,
+      "loss": 0.6569,
+      "step": 2875
+    },
+    {
+      "epoch": 5.325925925925926,
+      "grad_norm": 1.3061477004565407,
+      "learning_rate": 1.9447151276702354e-07,
+      "loss": 0.6999,
+      "step": 2876
+    },
+    {
+      "epoch": 5.3277777777777775,
+      "grad_norm": 1.4335630851649692,
+      "learning_rate": 1.9446659521076462e-07,
+      "loss": 0.7029,
+      "step": 2877
+    },
+    {
+      "epoch": 5.32962962962963,
+      "grad_norm": 2.954479068314854,
+      "learning_rate": 1.9446167553064004e-07,
+      "loss": 0.6866,
+      "step": 2878
+    },
+    {
+      "epoch": 5.3314814814814815,
+      "grad_norm": 1.5355574747257594,
+      "learning_rate": 1.9445675372676047e-07,
+      "loss": 0.6518,
+      "step": 2879
+    },
+    {
+      "epoch": 5.333333333333333,
+      "grad_norm": 1.7808925484710638,
+      "learning_rate": 1.9445182979923653e-07,
+      "loss": 0.667,
+      "step": 2880
+    },
+    {
+      "epoch": 5.3351851851851855,
+      "grad_norm": 1.5361518503757934,
+      "learning_rate": 1.9444690374817892e-07,
+      "loss": 0.6729,
+      "step": 2881
+    },
+    {
+      "epoch": 5.337037037037037,
+      "grad_norm": 1.5419605479996106,
+      "learning_rate": 1.9444197557369838e-07,
+      "loss": 0.6257,
+      "step": 2882
+    },
+    {
+      "epoch": 5.338888888888889,
+      "grad_norm": 1.9984880547190327,
+      "learning_rate": 1.9443704527590578e-07,
+      "loss": 0.682,
+      "step": 2883
+    },
+    {
+      "epoch": 5.340740740740741,
+      "grad_norm": 1.4725431370509865,
+      "learning_rate": 1.9443211285491188e-07,
+      "loss": 0.6995,
+      "step": 2884
+    },
+    {
+      "epoch": 5.342592592592593,
+      "grad_norm": 1.8200180770753691,
+      "learning_rate": 1.944271783108276e-07,
+      "loss": 0.7545,
+      "step": 2885
+    },
+    {
+      "epoch": 5.344444444444444,
+      "grad_norm": 1.8547336472576708,
+      "learning_rate": 1.9442224164376393e-07,
+      "loss": 0.6793,
+      "step": 2886
+    },
+    {
+      "epoch": 5.3462962962962965,
+      "grad_norm": 1.6449937348992367,
+      "learning_rate": 1.9441730285383179e-07,
+      "loss": 0.6754,
+      "step": 2887
+    },
+    {
+      "epoch": 5.348148148148148,
+      "grad_norm": 1.476239413484283,
+      "learning_rate": 1.9441236194114225e-07,
+      "loss": 0.686,
+      "step": 2888
+    },
+    {
+      "epoch": 5.35,
+      "grad_norm": 1.564057599915203,
+      "learning_rate": 1.944074189058064e-07,
+      "loss": 0.7177,
+      "step": 2889
+    },
+    {
+      "epoch": 5.351851851851852,
+      "grad_norm": 1.6232443065500868,
+      "learning_rate": 1.9440247374793538e-07,
+      "loss": 0.6862,
+      "step": 2890
+    },
+    {
+      "epoch": 5.353703703703704,
+      "grad_norm": 1.6742158305027384,
+      "learning_rate": 1.9439752646764034e-07,
+      "loss": 0.6966,
+      "step": 2891
+    },
+    {
+      "epoch": 5.355555555555555,
+      "grad_norm": 1.3510746479521587,
+      "learning_rate": 1.9439257706503253e-07,
+      "loss": 0.6462,
+      "step": 2892
+    },
+    {
+      "epoch": 5.357407407407408,
+      "grad_norm": 1.483323814899246,
+      "learning_rate": 1.943876255402232e-07,
+      "loss": 0.7131,
+      "step": 2893
+    },
+    {
+      "epoch": 5.359259259259259,
+      "grad_norm": 1.9853380402379108,
+      "learning_rate": 1.9438267189332375e-07,
+      "loss": 0.6562,
+      "step": 2894
+    },
+    {
+      "epoch": 5.361111111111111,
+      "grad_norm": 1.5641062492327031,
+      "learning_rate": 1.9437771612444544e-07,
+      "loss": 0.6914,
+      "step": 2895
+    },
+    {
+      "epoch": 5.362962962962963,
+      "grad_norm": 1.5793841290274553,
+      "learning_rate": 1.9437275823369976e-07,
+      "loss": 0.7162,
+      "step": 2896
+    },
+    {
+      "epoch": 5.364814814814815,
+      "grad_norm": 5.693226334836632,
+      "learning_rate": 1.943677982211982e-07,
+      "loss": 0.6706,
+      "step": 2897
+    },
+    {
+      "epoch": 5.366666666666666,
+      "grad_norm": 1.7172205682507975,
+      "learning_rate": 1.943628360870522e-07,
+      "loss": 0.6726,
+      "step": 2898
+    },
+    {
+      "epoch": 5.368518518518519,
+      "grad_norm": 1.7366661289176768,
+      "learning_rate": 1.9435787183137335e-07,
+      "loss": 0.6228,
+      "step": 2899
+    },
+    {
+      "epoch": 5.37037037037037,
+      "grad_norm": 1.4233524879901576,
+      "learning_rate": 1.9435290545427327e-07,
+      "loss": 0.6165,
+      "step": 2900
+    },
+    {
+      "epoch": 5.372222222222222,
+      "grad_norm": 1.586204624208911,
+      "learning_rate": 1.9434793695586365e-07,
+      "loss": 0.6723,
+      "step": 2901
+    },
+    {
+      "epoch": 5.374074074074074,
+      "grad_norm": 2.100788991398898,
+      "learning_rate": 1.9434296633625614e-07,
+      "loss": 0.6123,
+      "step": 2902
+    },
+    {
+      "epoch": 5.375925925925926,
+      "grad_norm": 1.707196280231772,
+      "learning_rate": 1.943379935955625e-07,
+      "loss": 0.7042,
+      "step": 2903
+    },
+    {
+      "epoch": 5.377777777777778,
+      "grad_norm": 1.6607282413862121,
+      "learning_rate": 1.943330187338946e-07,
+      "loss": 0.6789,
+      "step": 2904
+    },
+    {
+      "epoch": 5.37962962962963,
+      "grad_norm": 1.7423264171455726,
+      "learning_rate": 1.9432804175136418e-07,
+      "loss": 0.6171,
+      "step": 2905
+    },
+    {
+      "epoch": 5.381481481481481,
+      "grad_norm": 1.460691236447556,
+      "learning_rate": 1.9432306264808322e-07,
+      "loss": 0.6685,
+      "step": 2906
+    },
+    {
+      "epoch": 5.383333333333334,
+      "grad_norm": 1.6251796806998862,
+      "learning_rate": 1.9431808142416362e-07,
+      "loss": 0.6527,
+      "step": 2907
+    },
+    {
+      "epoch": 5.385185185185185,
+      "grad_norm": 1.5083887460518863,
+      "learning_rate": 1.9431309807971738e-07,
+      "loss": 0.6948,
+      "step": 2908
+    },
+    {
+      "epoch": 5.387037037037037,
+      "grad_norm": 1.4483643763724703,
+      "learning_rate": 1.9430811261485656e-07,
+      "loss": 0.6867,
+      "step": 2909
+    },
+    {
+      "epoch": 5.388888888888889,
+      "grad_norm": 1.825427469411476,
+      "learning_rate": 1.9430312502969324e-07,
+      "loss": 0.6893,
+      "step": 2910
+    },
+    {
+      "epoch": 5.390740740740741,
+      "grad_norm": 1.4174216598914398,
+      "learning_rate": 1.9429813532433953e-07,
+      "loss": 0.6768,
+      "step": 2911
+    },
+    {
+      "epoch": 5.392592592592592,
+      "grad_norm": 1.6698229745333968,
+      "learning_rate": 1.9429314349890761e-07,
+      "loss": 0.67,
+      "step": 2912
+    },
+    {
+      "epoch": 5.394444444444445,
+      "grad_norm": 1.4357508555177596,
+      "learning_rate": 1.9428814955350976e-07,
+      "loss": 0.6722,
+      "step": 2913
+    },
+    {
+      "epoch": 5.396296296296296,
+      "grad_norm": 1.5923082468288956,
+      "learning_rate": 1.942831534882582e-07,
+      "loss": 0.6738,
+      "step": 2914
+    },
+    {
+      "epoch": 5.398148148148148,
+      "grad_norm": 1.6860674908288755,
+      "learning_rate": 1.942781553032653e-07,
+      "loss": 0.6478,
+      "step": 2915
+    },
+    {
+      "epoch": 5.4,
+      "grad_norm": 1.4974631615691536,
+      "learning_rate": 1.942731549986434e-07,
+      "loss": 0.7059,
+      "step": 2916
+    },
+    {
+      "epoch": 5.401851851851852,
+      "grad_norm": 2.2473767368622055,
+      "learning_rate": 1.9426815257450494e-07,
+      "loss": 0.6999,
+      "step": 2917
+    },
+    {
+      "epoch": 5.4037037037037035,
+      "grad_norm": 2.5227875253793886,
+      "learning_rate": 1.942631480309624e-07,
+      "loss": 0.6515,
+      "step": 2918
+    },
+    {
+      "epoch": 5.405555555555556,
+      "grad_norm": 1.6563997976461724,
+      "learning_rate": 1.9425814136812825e-07,
+      "loss": 0.6917,
+      "step": 2919
+    },
+    {
+      "epoch": 5.407407407407407,
+      "grad_norm": 1.597575290589142,
+      "learning_rate": 1.9425313258611508e-07,
+      "loss": 0.6379,
+      "step": 2920
+    },
+    {
+      "epoch": 5.409259259259259,
+      "grad_norm": 1.3576372341251943,
+      "learning_rate": 1.9424812168503554e-07,
+      "loss": 0.6315,
+      "step": 2921
+    },
+    {
+      "epoch": 5.411111111111111,
+      "grad_norm": 1.5506275926302855,
+      "learning_rate": 1.9424310866500223e-07,
+      "loss": 0.6518,
+      "step": 2922
+    },
+    {
+      "epoch": 5.412962962962963,
+      "grad_norm": 1.4749643270065134,
+      "learning_rate": 1.9423809352612786e-07,
+      "loss": 0.6838,
+      "step": 2923
+    },
+    {
+      "epoch": 5.4148148148148145,
+      "grad_norm": 2.6489837459389065,
+      "learning_rate": 1.942330762685252e-07,
+      "loss": 0.6824,
+      "step": 2924
+    },
+    {
+      "epoch": 5.416666666666667,
+      "grad_norm": 1.315890563185257,
+      "learning_rate": 1.9422805689230705e-07,
+      "loss": 0.7033,
+      "step": 2925
+    },
+    {
+      "epoch": 5.4185185185185185,
+      "grad_norm": 1.395458411256235,
+      "learning_rate": 1.9422303539758625e-07,
+      "loss": 0.6984,
+      "step": 2926
+    },
+    {
+      "epoch": 5.42037037037037,
+      "grad_norm": 1.5319872281239346,
+      "learning_rate": 1.9421801178447574e-07,
+      "loss": 0.6623,
+      "step": 2927
+    },
+    {
+      "epoch": 5.4222222222222225,
+      "grad_norm": 1.3846258145233896,
+      "learning_rate": 1.9421298605308845e-07,
+      "loss": 0.6561,
+      "step": 2928
+    },
+    {
+      "epoch": 5.424074074074074,
+      "grad_norm": 1.4989289725082253,
+      "learning_rate": 1.9420795820353733e-07,
+      "loss": 0.6482,
+      "step": 2929
+    },
+    {
+      "epoch": 5.425925925925926,
+      "grad_norm": 1.393535443972757,
+      "learning_rate": 1.9420292823593543e-07,
+      "loss": 0.718,
+      "step": 2930
+    },
+    {
+      "epoch": 5.427777777777778,
+      "grad_norm": 1.3144700050476068,
+      "learning_rate": 1.9419789615039588e-07,
+      "loss": 0.6396,
+      "step": 2931
+    },
+    {
+      "epoch": 5.42962962962963,
+      "grad_norm": 1.3998071811253625,
+      "learning_rate": 1.9419286194703177e-07,
+      "loss": 0.6484,
+      "step": 2932
+    },
+    {
+      "epoch": 5.431481481481481,
+      "grad_norm": 1.5983662613344372,
+      "learning_rate": 1.941878256259563e-07,
+      "loss": 0.7362,
+      "step": 2933
+    },
+    {
+      "epoch": 5.433333333333334,
+      "grad_norm": 1.4044020828608224,
+      "learning_rate": 1.9418278718728272e-07,
+      "loss": 0.671,
+      "step": 2934
+    },
+    {
+      "epoch": 5.435185185185185,
+      "grad_norm": 1.4399794269790391,
+      "learning_rate": 1.9417774663112426e-07,
+      "loss": 0.5834,
+      "step": 2935
+    },
+    {
+      "epoch": 5.437037037037037,
+      "grad_norm": 1.7459385014787752,
+      "learning_rate": 1.9417270395759426e-07,
+      "loss": 0.6903,
+      "step": 2936
+    },
+    {
+      "epoch": 5.438888888888889,
+      "grad_norm": 1.5243699475353711,
+      "learning_rate": 1.9416765916680615e-07,
+      "loss": 0.6505,
+      "step": 2937
+    },
+    {
+      "epoch": 5.440740740740741,
+      "grad_norm": 1.3793364691779872,
+      "learning_rate": 1.9416261225887326e-07,
+      "loss": 0.6662,
+      "step": 2938
+    },
+    {
+      "epoch": 5.442592592592592,
+      "grad_norm": 2.0953323288135444,
+      "learning_rate": 1.9415756323390914e-07,
+      "loss": 0.6106,
+      "step": 2939
+    },
+    {
+      "epoch": 5.444444444444445,
+      "grad_norm": 1.4290060655734755,
+      "learning_rate": 1.9415251209202728e-07,
+      "loss": 0.6317,
+      "step": 2940
+    },
+    {
+      "epoch": 5.446296296296296,
+      "grad_norm": 1.4455754787910795,
+      "learning_rate": 1.941474588333412e-07,
+      "loss": 0.6952,
+      "step": 2941
+    },
+    {
+      "epoch": 5.448148148148148,
+      "grad_norm": 1.4285820996856942,
+      "learning_rate": 1.941424034579646e-07,
+      "loss": 0.6396,
+      "step": 2942
+    },
+    {
+      "epoch": 5.45,
+      "grad_norm": 1.5889285596447713,
+      "learning_rate": 1.94137345966011e-07,
+      "loss": 0.6912,
+      "step": 2943
+    },
+    {
+      "epoch": 5.451851851851852,
+      "grad_norm": 1.4167829590991743,
+      "learning_rate": 1.9413228635759424e-07,
+      "loss": 0.6863,
+      "step": 2944
+    },
+    {
+      "epoch": 5.453703703703704,
+      "grad_norm": 1.3428993035324839,
+      "learning_rate": 1.94127224632828e-07,
+      "loss": 0.6498,
+      "step": 2945
+    },
+    {
+      "epoch": 5.455555555555556,
+      "grad_norm": 1.3497550044754816,
+      "learning_rate": 1.9412216079182614e-07,
+      "loss": 0.6878,
+      "step": 2946
+    },
+    {
+      "epoch": 5.457407407407407,
+      "grad_norm": 1.3612805883911834,
+      "learning_rate": 1.9411709483470244e-07,
+      "loss": 0.6614,
+      "step": 2947
+    },
+    {
+      "epoch": 5.459259259259259,
+      "grad_norm": 1.7452345996108638,
+      "learning_rate": 1.9411202676157087e-07,
+      "loss": 0.7028,
+      "step": 2948
+    },
+    {
+      "epoch": 5.461111111111111,
+      "grad_norm": 1.3947965069952588,
+      "learning_rate": 1.9410695657254528e-07,
+      "loss": 0.6493,
+      "step": 2949
+    },
+    {
+      "epoch": 5.462962962962963,
+      "grad_norm": 1.3409229335695583,
+      "learning_rate": 1.9410188426773976e-07,
+      "loss": 0.627,
+      "step": 2950
+    },
+    {
+      "epoch": 5.464814814814815,
+      "grad_norm": 1.3318809602071386,
+      "learning_rate": 1.9409680984726828e-07,
+      "loss": 0.6494,
+      "step": 2951
+    },
+    {
+      "epoch": 5.466666666666667,
+      "grad_norm": 1.8300137133764702,
+      "learning_rate": 1.9409173331124498e-07,
+      "loss": 0.6994,
+      "step": 2952
+    },
+    {
+      "epoch": 5.468518518518518,
+      "grad_norm": 1.500389161582396,
+      "learning_rate": 1.940866546597839e-07,
+      "loss": 0.7004,
+      "step": 2953
+    },
+    {
+      "epoch": 5.470370370370371,
+      "grad_norm": 3.6801809309329516,
+      "learning_rate": 1.9408157389299938e-07,
+      "loss": 0.6904,
+      "step": 2954
+    },
+    {
+      "epoch": 5.472222222222222,
+      "grad_norm": 1.5258044809238842,
+      "learning_rate": 1.940764910110055e-07,
+      "loss": 0.6866,
+      "step": 2955
+    },
+    {
+      "epoch": 5.474074074074074,
+      "grad_norm": 1.902625106246917,
+      "learning_rate": 1.9407140601391657e-07,
+      "loss": 0.7115,
+      "step": 2956
+    },
+    {
+      "epoch": 5.475925925925926,
+      "grad_norm": 1.5513732363924757,
+      "learning_rate": 1.94066318901847e-07,
+      "loss": 0.684,
+      "step": 2957
+    },
+    {
+      "epoch": 5.477777777777778,
+      "grad_norm": 2.095748644244585,
+      "learning_rate": 1.9406122967491108e-07,
+      "loss": 0.6706,
+      "step": 2958
+    },
+    {
+      "epoch": 5.479629629629629,
+      "grad_norm": 1.9262204957794142,
+      "learning_rate": 1.9405613833322324e-07,
+      "loss": 0.6594,
+      "step": 2959
+    },
+    {
+      "epoch": 5.481481481481482,
+      "grad_norm": 1.444565092012416,
+      "learning_rate": 1.9405104487689797e-07,
+      "loss": 0.6749,
+      "step": 2960
+    },
+    {
+      "epoch": 5.483333333333333,
+      "grad_norm": 1.5424753216598721,
+      "learning_rate": 1.9404594930604975e-07,
+      "loss": 0.6772,
+      "step": 2961
+    },
+    {
+      "epoch": 5.485185185185185,
+      "grad_norm": 1.5047592108047783,
+      "learning_rate": 1.9404085162079317e-07,
+      "loss": 0.6924,
+      "step": 2962
+    },
+    {
+      "epoch": 5.487037037037037,
+      "grad_norm": 1.4706940258969183,
+      "learning_rate": 1.9403575182124286e-07,
+      "loss": 0.68,
+      "step": 2963
+    },
+    {
+      "epoch": 5.488888888888889,
+      "grad_norm": 1.6688407574218445,
+      "learning_rate": 1.9403064990751343e-07,
+      "loss": 0.668,
+      "step": 2964
+    },
+    {
+      "epoch": 5.4907407407407405,
+      "grad_norm": 1.449326482274683,
+      "learning_rate": 1.9402554587971961e-07,
+      "loss": 0.6346,
+      "step": 2965
+    },
+    {
+      "epoch": 5.492592592592593,
+      "grad_norm": 1.4919651190182803,
+      "learning_rate": 1.9402043973797615e-07,
+      "loss": 0.64,
+      "step": 2966
+    },
+    {
+      "epoch": 5.4944444444444445,
+      "grad_norm": 1.5165074081158856,
+      "learning_rate": 1.9401533148239785e-07,
+      "loss": 0.6545,
+      "step": 2967
+    },
+    {
+      "epoch": 5.496296296296296,
+      "grad_norm": 1.7157365599221481,
+      "learning_rate": 1.9401022111309958e-07,
+      "loss": 0.6837,
+      "step": 2968
+    },
+    {
+      "epoch": 5.4981481481481485,
+      "grad_norm": 1.8374082642507168,
+      "learning_rate": 1.9400510863019616e-07,
+      "loss": 0.6841,
+      "step": 2969
+    },
+    {
+      "epoch": 5.5,
+      "grad_norm": 1.7782057848141277,
+      "learning_rate": 1.9399999403380262e-07,
+      "loss": 0.6537,
+      "step": 2970
+    },
+    {
+      "epoch": 5.5018518518518515,
+      "grad_norm": 1.6193677417046048,
+      "learning_rate": 1.939948773240339e-07,
+      "loss": 0.6722,
+      "step": 2971
+    },
+    {
+      "epoch": 5.503703703703704,
+      "grad_norm": 1.839276451566974,
+      "learning_rate": 1.9398975850100509e-07,
+      "loss": 0.6238,
+      "step": 2972
+    },
+    {
+      "epoch": 5.5055555555555555,
+      "grad_norm": 1.8725145797813199,
+      "learning_rate": 1.939846375648312e-07,
+      "loss": 0.7245,
+      "step": 2973
+    },
+    {
+      "epoch": 5.507407407407407,
+      "grad_norm": 1.3441020382176563,
+      "learning_rate": 1.9397951451562737e-07,
+      "loss": 0.6669,
+      "step": 2974
+    },
+    {
+      "epoch": 5.5092592592592595,
+      "grad_norm": 1.8091126481893145,
+      "learning_rate": 1.9397438935350885e-07,
+      "loss": 0.667,
+      "step": 2975
+    },
+    {
+      "epoch": 5.511111111111111,
+      "grad_norm": 1.910275571247403,
+      "learning_rate": 1.9396926207859085e-07,
+      "loss": 0.6747,
+      "step": 2976
+    },
+    {
+      "epoch": 5.512962962962963,
+      "grad_norm": 1.9830925759769786,
+      "learning_rate": 1.9396413269098857e-07,
+      "loss": 0.6754,
+      "step": 2977
+    },
+    {
+      "epoch": 5.514814814814815,
+      "grad_norm": 1.606592644759462,
+      "learning_rate": 1.939590011908174e-07,
+      "loss": 0.6865,
+      "step": 2978
+    },
+    {
+      "epoch": 5.516666666666667,
+      "grad_norm": 1.3742644681080447,
+      "learning_rate": 1.939538675781927e-07,
+      "loss": 0.6367,
+      "step": 2979
+    },
+    {
+      "epoch": 5.518518518518518,
+      "grad_norm": 1.8783109195767198,
+      "learning_rate": 1.9394873185322988e-07,
+      "loss": 0.6959,
+      "step": 2980
+    },
+    {
+      "epoch": 5.520370370370371,
+      "grad_norm": 1.5921200834309328,
+      "learning_rate": 1.939435940160444e-07,
+      "loss": 0.6891,
+      "step": 2981
+    },
+    {
+      "epoch": 5.522222222222222,
+      "grad_norm": 1.612228922987339,
+      "learning_rate": 1.939384540667518e-07,
+      "loss": 0.6371,
+      "step": 2982
+    },
+    {
+      "epoch": 5.524074074074074,
+      "grad_norm": 2.7696971499542316,
+      "learning_rate": 1.9393331200546764e-07,
+      "loss": 0.7107,
+      "step": 2983
+    },
+    {
+      "epoch": 5.525925925925926,
+      "grad_norm": 1.5503760663612347,
+      "learning_rate": 1.9392816783230745e-07,
+      "loss": 0.6602,
+      "step": 2984
+    },
+    {
+      "epoch": 5.527777777777778,
+      "grad_norm": 1.808780330076644,
+      "learning_rate": 1.9392302154738695e-07,
+      "loss": 0.7219,
+      "step": 2985
+    },
+    {
+      "epoch": 5.52962962962963,
+      "grad_norm": 2.316889557786525,
+      "learning_rate": 1.9391787315082188e-07,
+      "loss": 0.6877,
+      "step": 2986
+    },
+    {
+      "epoch": 5.531481481481482,
+      "grad_norm": 1.6347250733106344,
+      "learning_rate": 1.939127226427279e-07,
+      "loss": 0.6808,
+      "step": 2987
+    },
+    {
+      "epoch": 5.533333333333333,
+      "grad_norm": 1.4205303371170537,
+      "learning_rate": 1.939075700232209e-07,
+      "loss": 0.6629,
+      "step": 2988
+    },
+    {
+      "epoch": 5.535185185185185,
+      "grad_norm": 1.5062129418245878,
+      "learning_rate": 1.9390241529241664e-07,
+      "loss": 0.6631,
+      "step": 2989
+    },
+    {
+      "epoch": 5.537037037037037,
+      "grad_norm": 1.31335146159003,
+      "learning_rate": 1.9389725845043104e-07,
+      "loss": 0.6749,
+      "step": 2990
+    },
+    {
+      "epoch": 5.538888888888889,
+      "grad_norm": 1.8628955335828814,
+      "learning_rate": 1.9389209949738006e-07,
+      "loss": 0.6761,
+      "step": 2991
+    },
+    {
+      "epoch": 5.540740740740741,
+      "grad_norm": 1.669216218729797,
+      "learning_rate": 1.9388693843337968e-07,
+      "loss": 0.6742,
+      "step": 2992
+    },
+    {
+      "epoch": 5.542592592592593,
+      "grad_norm": 1.3340010398856033,
+      "learning_rate": 1.9388177525854595e-07,
+      "loss": 0.6662,
+      "step": 2993
+    },
+    {
+      "epoch": 5.544444444444444,
+      "grad_norm": 1.4617330410696125,
+      "learning_rate": 1.938766099729949e-07,
+      "loss": 0.6174,
+      "step": 2994
+    },
+    {
+      "epoch": 5.546296296296296,
+      "grad_norm": 1.313019414908476,
+      "learning_rate": 1.9387144257684272e-07,
+      "loss": 0.6588,
+      "step": 2995
+    },
+    {
+      "epoch": 5.548148148148148,
+      "grad_norm": 1.613916167300068,
+      "learning_rate": 1.9386627307020554e-07,
+      "loss": 0.645,
+      "step": 2996
+    },
+    {
+      "epoch": 5.55,
+      "grad_norm": 2.122689499817843,
+      "learning_rate": 1.938611014531996e-07,
+      "loss": 0.6673,
+      "step": 2997
+    },
+    {
+      "epoch": 5.551851851851852,
+      "grad_norm": 1.641597250070579,
+      "learning_rate": 1.9385592772594118e-07,
+      "loss": 0.6734,
+      "step": 2998
+    },
+    {
+      "epoch": 5.553703703703704,
+      "grad_norm": 1.465225389793985,
+      "learning_rate": 1.938507518885466e-07,
+      "loss": 0.7056,
+      "step": 2999
+    },
+    {
+      "epoch": 5.555555555555555,
+      "grad_norm": 1.3560676765894122,
+      "learning_rate": 1.9384557394113226e-07,
+      "loss": 0.6356,
+      "step": 3000
+    },
+    {
+      "epoch": 5.557407407407408,
+      "grad_norm": 1.5017648465486404,
+      "learning_rate": 1.938403938838145e-07,
+      "loss": 0.6855,
+      "step": 3001
+    },
+    {
+      "epoch": 5.559259259259259,
+      "grad_norm": 1.5082982066661013,
+      "learning_rate": 1.9383521171670982e-07,
+      "loss": 0.6598,
+      "step": 3002
+    },
+    {
+      "epoch": 5.561111111111111,
+      "grad_norm": 1.590000135989519,
+      "learning_rate": 1.9383002743993473e-07,
+      "loss": 0.6623,
+      "step": 3003
+    },
+    {
+      "epoch": 5.562962962962963,
+      "grad_norm": 1.477823356411927,
+      "learning_rate": 1.9382484105360576e-07,
+      "loss": 0.6555,
+      "step": 3004
+    },
+    {
+      "epoch": 5.564814814814815,
+      "grad_norm": 1.5761015836830574,
+      "learning_rate": 1.9381965255783957e-07,
+      "loss": 0.6403,
+      "step": 3005
+    },
+    {
+      "epoch": 5.566666666666666,
+      "grad_norm": 1.4465205071979121,
+      "learning_rate": 1.9381446195275277e-07,
+      "loss": 0.6442,
+      "step": 3006
+    },
+    {
+      "epoch": 5.568518518518519,
+      "grad_norm": 1.5656541107552768,
+      "learning_rate": 1.9380926923846207e-07,
+      "loss": 0.6626,
+      "step": 3007
+    },
+    {
+      "epoch": 5.57037037037037,
+      "grad_norm": 1.8849170633009384,
+      "learning_rate": 1.9380407441508424e-07,
+      "loss": 0.6529,
+      "step": 3008
+    },
+    {
+      "epoch": 5.572222222222222,
+      "grad_norm": 1.566007676480427,
+      "learning_rate": 1.9379887748273602e-07,
+      "loss": 0.6427,
+      "step": 3009
+    },
+    {
+      "epoch": 5.574074074074074,
+      "grad_norm": 1.5031049804944971,
+      "learning_rate": 1.9379367844153429e-07,
+      "loss": 0.744,
+      "step": 3010
+    },
+    {
+      "epoch": 5.575925925925926,
+      "grad_norm": 1.607387330475157,
+      "learning_rate": 1.9378847729159593e-07,
+      "loss": 0.6593,
+      "step": 3011
+    },
+    {
+      "epoch": 5.5777777777777775,
+      "grad_norm": 2.6693408806417747,
+      "learning_rate": 1.9378327403303788e-07,
+      "loss": 0.6268,
+      "step": 3012
+    },
+    {
+      "epoch": 5.57962962962963,
+      "grad_norm": 1.6469348887414648,
+      "learning_rate": 1.937780686659771e-07,
+      "loss": 0.669,
+      "step": 3013
+    },
+    {
+      "epoch": 5.5814814814814815,
+      "grad_norm": 1.8577305173064904,
+      "learning_rate": 1.9377286119053066e-07,
+      "loss": 0.6673,
+      "step": 3014
+    },
+    {
+      "epoch": 5.583333333333333,
+      "grad_norm": 2.3770772668479903,
+      "learning_rate": 1.937676516068156e-07,
+      "loss": 0.6682,
+      "step": 3015
+    },
+    {
+      "epoch": 5.5851851851851855,
+      "grad_norm": 1.386629585628963,
+      "learning_rate": 1.9376243991494907e-07,
+      "loss": 0.6466,
+      "step": 3016
+    },
+    {
+      "epoch": 5.587037037037037,
+      "grad_norm": 1.5880745414668402,
+      "learning_rate": 1.9375722611504823e-07,
+      "loss": 0.6361,
+      "step": 3017
+    },
+    {
+      "epoch": 5.588888888888889,
+      "grad_norm": 1.75952976542882,
+      "learning_rate": 1.937520102072303e-07,
+      "loss": 0.6849,
+      "step": 3018
+    },
+    {
+      "epoch": 5.590740740740741,
+      "grad_norm": 1.5641833439271742,
+      "learning_rate": 1.9374679219161257e-07,
+      "loss": 0.687,
+      "step": 3019
+    },
+    {
+      "epoch": 5.592592592592593,
+      "grad_norm": 2.111528469463874,
+      "learning_rate": 1.9374157206831233e-07,
+      "loss": 0.7033,
+      "step": 3020
+    },
+    {
+      "epoch": 5.594444444444444,
+      "grad_norm": 1.7181832468719178,
+      "learning_rate": 1.9373634983744695e-07,
+      "loss": 0.6586,
+      "step": 3021
+    },
+    {
+      "epoch": 5.5962962962962965,
+      "grad_norm": 2.360895119629398,
+      "learning_rate": 1.937311254991339e-07,
+      "loss": 0.6948,
+      "step": 3022
+    },
+    {
+      "epoch": 5.598148148148148,
+      "grad_norm": 1.5390042183476398,
+      "learning_rate": 1.937258990534905e-07,
+      "loss": 0.6548,
+      "step": 3023
+    },
+    {
+      "epoch": 5.6,
+      "grad_norm": 1.514027297406205,
+      "learning_rate": 1.9372067050063436e-07,
+      "loss": 0.6698,
+      "step": 3024
+    },
+    {
+      "epoch": 5.601851851851852,
+      "grad_norm": 1.402929899519556,
+      "learning_rate": 1.9371543984068302e-07,
+      "loss": 0.6732,
+      "step": 3025
+    },
+    {
+      "epoch": 5.603703703703704,
+      "grad_norm": 1.7746518325099097,
+      "learning_rate": 1.9371020707375403e-07,
+      "loss": 0.6702,
+      "step": 3026
+    },
+    {
+      "epoch": 5.605555555555555,
+      "grad_norm": 1.5342597824255881,
+      "learning_rate": 1.9370497219996508e-07,
+      "loss": 0.6569,
+      "step": 3027
+    },
+    {
+      "epoch": 5.607407407407408,
+      "grad_norm": 1.465997532523462,
+      "learning_rate": 1.9369973521943388e-07,
+      "loss": 0.6735,
+      "step": 3028
+    },
+    {
+      "epoch": 5.609259259259259,
+      "grad_norm": 2.206975163870859,
+      "learning_rate": 1.9369449613227813e-07,
+      "loss": 0.6644,
+      "step": 3029
+    },
+    {
+      "epoch": 5.611111111111111,
+      "grad_norm": 1.5014885917314045,
+      "learning_rate": 1.9368925493861566e-07,
+      "loss": 0.6706,
+      "step": 3030
+    },
+    {
+      "epoch": 5.612962962962963,
+      "grad_norm": 1.8209765441989565,
+      "learning_rate": 1.9368401163856426e-07,
+      "loss": 0.6654,
+      "step": 3031
+    },
+    {
+      "epoch": 5.614814814814815,
+      "grad_norm": 1.491323596112947,
+      "learning_rate": 1.9367876623224184e-07,
+      "loss": 0.6973,
+      "step": 3032
+    },
+    {
+      "epoch": 5.616666666666667,
+      "grad_norm": 1.9744024188488531,
+      "learning_rate": 1.9367351871976632e-07,
+      "loss": 0.7124,
+      "step": 3033
+    },
+    {
+      "epoch": 5.618518518518519,
+      "grad_norm": 1.6289982305355026,
+      "learning_rate": 1.936682691012557e-07,
+      "loss": 0.6658,
+      "step": 3034
+    },
+    {
+      "epoch": 5.62037037037037,
+      "grad_norm": 1.6225262331838626,
+      "learning_rate": 1.93663017376828e-07,
+      "loss": 0.6636,
+      "step": 3035
+    },
+    {
+      "epoch": 5.622222222222222,
+      "grad_norm": 1.8898997313628014,
+      "learning_rate": 1.9365776354660125e-07,
+      "loss": 0.6458,
+      "step": 3036
+    },
+    {
+      "epoch": 5.624074074074074,
+      "grad_norm": 1.9593879943134564,
+      "learning_rate": 1.9365250761069363e-07,
+      "loss": 0.6788,
+      "step": 3037
+    },
+    {
+      "epoch": 5.625925925925926,
+      "grad_norm": 5.840143770684436,
+      "learning_rate": 1.9364724956922328e-07,
+      "loss": 0.6522,
+      "step": 3038
+    },
+    {
+      "epoch": 5.627777777777778,
+      "grad_norm": 2.0273225284926957,
+      "learning_rate": 1.9364198942230844e-07,
+      "loss": 0.683,
+      "step": 3039
+    },
+    {
+      "epoch": 5.62962962962963,
+      "grad_norm": 2.083897234532407,
+      "learning_rate": 1.9363672717006732e-07,
+      "loss": 0.6732,
+      "step": 3040
+    },
+    {
+      "epoch": 5.631481481481481,
+      "grad_norm": 1.8030036642060996,
+      "learning_rate": 1.9363146281261827e-07,
+      "loss": 0.6749,
+      "step": 3041
+    },
+    {
+      "epoch": 5.633333333333333,
+      "grad_norm": 1.5549134621890215,
+      "learning_rate": 1.9362619635007963e-07,
+      "loss": 0.7169,
+      "step": 3042
+    },
+    {
+      "epoch": 5.635185185185185,
+      "grad_norm": 1.6499408971479643,
+      "learning_rate": 1.9362092778256983e-07,
+      "loss": 0.7181,
+      "step": 3043
+    },
+    {
+      "epoch": 5.637037037037037,
+      "grad_norm": 1.6541788016346053,
+      "learning_rate": 1.9361565711020728e-07,
+      "loss": 0.6681,
+      "step": 3044
+    },
+    {
+      "epoch": 5.638888888888889,
+      "grad_norm": 2.5772332485467793,
+      "learning_rate": 1.936103843331105e-07,
+      "loss": 0.6896,
+      "step": 3045
+    },
+    {
+      "epoch": 5.640740740740741,
+      "grad_norm": 1.862076521665792,
+      "learning_rate": 1.9360510945139806e-07,
+      "loss": 0.6752,
+      "step": 3046
+    },
+    {
+      "epoch": 5.642592592592592,
+      "grad_norm": 1.7795509222981707,
+      "learning_rate": 1.935998324651885e-07,
+      "loss": 0.7239,
+      "step": 3047
+    },
+    {
+      "epoch": 5.644444444444445,
+      "grad_norm": 1.6288779946017031,
+      "learning_rate": 1.935945533746005e-07,
+      "loss": 0.6827,
+      "step": 3048
+    },
+    {
+      "epoch": 5.646296296296296,
+      "grad_norm": 1.7353777879347545,
+      "learning_rate": 1.9358927217975276e-07,
+      "loss": 0.6292,
+      "step": 3049
+    },
+    {
+      "epoch": 5.648148148148148,
+      "grad_norm": 1.8564949579085719,
+      "learning_rate": 1.93583988880764e-07,
+      "loss": 0.6373,
+      "step": 3050
+    },
+    {
+      "epoch": 5.65,
+      "grad_norm": 1.939407507598791,
+      "learning_rate": 1.9357870347775297e-07,
+      "loss": 0.6486,
+      "step": 3051
+    },
+    {
+      "epoch": 5.651851851851852,
+      "grad_norm": 1.9194230505513739,
+      "learning_rate": 1.9357341597083857e-07,
+      "loss": 0.6855,
+      "step": 3052
+    },
+    {
+      "epoch": 5.6537037037037035,
+      "grad_norm": 1.94071649557587,
+      "learning_rate": 1.9356812636013963e-07,
+      "loss": 0.692,
+      "step": 3053
+    },
+    {
+      "epoch": 5.655555555555556,
+      "grad_norm": 1.6718659366554096,
+      "learning_rate": 1.9356283464577506e-07,
+      "loss": 0.6662,
+      "step": 3054
+    },
+    {
+      "epoch": 5.657407407407407,
+      "grad_norm": 2.0050321572741545,
+      "learning_rate": 1.9355754082786384e-07,
+      "loss": 0.6783,
+      "step": 3055
+    },
+    {
+      "epoch": 5.659259259259259,
+      "grad_norm": 1.5337280438773258,
+      "learning_rate": 1.9355224490652504e-07,
+      "loss": 0.6768,
+      "step": 3056
+    },
+    {
+      "epoch": 5.661111111111111,
+      "grad_norm": 1.5448094146964897,
+      "learning_rate": 1.9354694688187768e-07,
+      "loss": 0.6696,
+      "step": 3057
+    },
+    {
+      "epoch": 5.662962962962963,
+      "grad_norm": 1.860632010249602,
+      "learning_rate": 1.935416467540409e-07,
+      "loss": 0.6922,
+      "step": 3058
+    },
+    {
+      "epoch": 5.6648148148148145,
+      "grad_norm": 1.6359189036459534,
+      "learning_rate": 1.9353634452313382e-07,
+      "loss": 0.6665,
+      "step": 3059
+    },
+    {
+      "epoch": 5.666666666666667,
+      "grad_norm": 1.434756173356067,
+      "learning_rate": 1.9353104018927565e-07,
+      "loss": 0.6484,
+      "step": 3060
+    },
+    {
+      "epoch": 5.6685185185185185,
+      "grad_norm": 1.948186153162462,
+      "learning_rate": 1.9352573375258575e-07,
+      "loss": 0.67,
+      "step": 3061
+    },
+    {
+      "epoch": 5.67037037037037,
+      "grad_norm": 1.5982433520538097,
+      "learning_rate": 1.9352042521318327e-07,
+      "loss": 0.6913,
+      "step": 3062
+    },
+    {
+      "epoch": 5.6722222222222225,
+      "grad_norm": 1.380339450597154,
+      "learning_rate": 1.9351511457118765e-07,
+      "loss": 0.6742,
+      "step": 3063
+    },
+    {
+      "epoch": 5.674074074074074,
+      "grad_norm": 1.4197555959778585,
+      "learning_rate": 1.9350980182671827e-07,
+      "loss": 0.6668,
+      "step": 3064
+    },
+    {
+      "epoch": 5.675925925925926,
+      "grad_norm": 1.5029882107411736,
+      "learning_rate": 1.9350448697989462e-07,
+      "loss": 0.672,
+      "step": 3065
+    },
+    {
+      "epoch": 5.677777777777778,
+      "grad_norm": 1.6434691743536687,
+      "learning_rate": 1.9349917003083611e-07,
+      "loss": 0.6807,
+      "step": 3066
+    },
+    {
+      "epoch": 5.67962962962963,
+      "grad_norm": 1.5870592522694542,
+      "learning_rate": 1.9349385097966233e-07,
+      "loss": 0.6708,
+      "step": 3067
+    },
+    {
+      "epoch": 5.681481481481481,
+      "grad_norm": 1.6101027911119095,
+      "learning_rate": 1.9348852982649283e-07,
+      "loss": 0.67,
+      "step": 3068
+    },
+    {
+      "epoch": 5.683333333333334,
+      "grad_norm": 1.3936503593260219,
+      "learning_rate": 1.934832065714473e-07,
+      "loss": 0.6972,
+      "step": 3069
+    },
+    {
+      "epoch": 5.685185185185185,
+      "grad_norm": 1.7572130240760366,
+      "learning_rate": 1.9347788121464537e-07,
+      "loss": 0.6467,
+      "step": 3070
+    },
+    {
+      "epoch": 5.687037037037037,
+      "grad_norm": 1.547500074074858,
+      "learning_rate": 1.934725537562068e-07,
+      "loss": 0.7068,
+      "step": 3071
+    },
+    {
+      "epoch": 5.688888888888889,
+      "grad_norm": 1.4973576040083854,
+      "learning_rate": 1.9346722419625135e-07,
+      "loss": 0.6773,
+      "step": 3072
+    },
+    {
+      "epoch": 5.690740740740741,
+      "grad_norm": 1.3716341793583535,
+      "learning_rate": 1.9346189253489884e-07,
+      "loss": 0.682,
+      "step": 3073
+    },
+    {
+      "epoch": 5.692592592592592,
+      "grad_norm": 1.7507547486927795,
+      "learning_rate": 1.9345655877226913e-07,
+      "loss": 0.6852,
+      "step": 3074
+    },
+    {
+      "epoch": 5.694444444444445,
+      "grad_norm": 1.3766742209370904,
+      "learning_rate": 1.9345122290848218e-07,
+      "loss": 0.6749,
+      "step": 3075
+    },
+    {
+      "epoch": 5.696296296296296,
+      "grad_norm": 2.4314264555509375,
+      "learning_rate": 1.9344588494365794e-07,
+      "loss": 0.6577,
+      "step": 3076
+    },
+    {
+      "epoch": 5.698148148148148,
+      "grad_norm": 1.3732823957425164,
+      "learning_rate": 1.9344054487791638e-07,
+      "loss": 0.6797,
+      "step": 3077
+    },
+    {
+      "epoch": 5.7,
+      "grad_norm": 2.441507971640275,
+      "learning_rate": 1.934352027113776e-07,
+      "loss": 0.669,
+      "step": 3078
+    },
+    {
+      "epoch": 5.701851851851852,
+      "grad_norm": 1.7895974928068723,
+      "learning_rate": 1.9342985844416173e-07,
+      "loss": 0.6864,
+      "step": 3079
+    },
+    {
+      "epoch": 5.703703703703704,
+      "grad_norm": 1.515956115089075,
+      "learning_rate": 1.9342451207638888e-07,
+      "loss": 0.6937,
+      "step": 3080
+    },
+    {
+      "epoch": 5.705555555555556,
+      "grad_norm": 1.547323622781393,
+      "learning_rate": 1.9341916360817921e-07,
+      "loss": 0.7071,
+      "step": 3081
+    },
+    {
+      "epoch": 5.707407407407407,
+      "grad_norm": 1.655857180601228,
+      "learning_rate": 1.9341381303965308e-07,
+      "loss": 0.6191,
+      "step": 3082
+    },
+    {
+      "epoch": 5.709259259259259,
+      "grad_norm": 1.484552359747053,
+      "learning_rate": 1.934084603709307e-07,
+      "loss": 0.6599,
+      "step": 3083
+    },
+    {
+      "epoch": 5.711111111111111,
+      "grad_norm": 1.5035470451854136,
+      "learning_rate": 1.9340310560213244e-07,
+      "loss": 0.6687,
+      "step": 3084
+    },
+    {
+      "epoch": 5.712962962962963,
+      "grad_norm": 1.3915232327243305,
+      "learning_rate": 1.9339774873337868e-07,
+      "loss": 0.6757,
+      "step": 3085
+    },
+    {
+      "epoch": 5.714814814814815,
+      "grad_norm": 1.7470536229939362,
+      "learning_rate": 1.9339238976478986e-07,
+      "loss": 0.6644,
+      "step": 3086
+    },
+    {
+      "epoch": 5.716666666666667,
+      "grad_norm": 1.7876411304128315,
+      "learning_rate": 1.9338702869648648e-07,
+      "loss": 0.6545,
+      "step": 3087
+    },
+    {
+      "epoch": 5.718518518518518,
+      "grad_norm": 1.5541025370563417,
+      "learning_rate": 1.9338166552858906e-07,
+      "loss": 0.6564,
+      "step": 3088
+    },
+    {
+      "epoch": 5.72037037037037,
+      "grad_norm": 1.6010514811084968,
+      "learning_rate": 1.9337630026121818e-07,
+      "loss": 0.6891,
+      "step": 3089
+    },
+    {
+      "epoch": 5.722222222222222,
+      "grad_norm": 1.5908301722684488,
+      "learning_rate": 1.9337093289449446e-07,
+      "loss": 0.6584,
+      "step": 3090
+    },
+    {
+      "epoch": 5.724074074074074,
+      "grad_norm": 1.6367252369865792,
+      "learning_rate": 1.9336556342853857e-07,
+      "loss": 0.6718,
+      "step": 3091
+    },
+    {
+      "epoch": 5.725925925925926,
+      "grad_norm": 1.7485577159550505,
+      "learning_rate": 1.9336019186347125e-07,
+      "loss": 0.7073,
+      "step": 3092
+    },
+    {
+      "epoch": 5.727777777777778,
+      "grad_norm": 1.4950591970547817,
+      "learning_rate": 1.9335481819941326e-07,
+      "loss": 0.6573,
+      "step": 3093
+    },
+    {
+      "epoch": 5.729629629629629,
+      "grad_norm": 1.377075708431481,
+      "learning_rate": 1.9334944243648538e-07,
+      "loss": 0.6629,
+      "step": 3094
+    },
+    {
+      "epoch": 5.731481481481482,
+      "grad_norm": 1.307829341741579,
+      "learning_rate": 1.9334406457480853e-07,
+      "loss": 0.6168,
+      "step": 3095
+    },
+    {
+      "epoch": 5.733333333333333,
+      "grad_norm": 1.4463013563843563,
+      "learning_rate": 1.9333868461450358e-07,
+      "loss": 0.6969,
+      "step": 3096
+    },
+    {
+      "epoch": 5.735185185185185,
+      "grad_norm": 1.7536236222455248,
+      "learning_rate": 1.933333025556915e-07,
+      "loss": 0.6263,
+      "step": 3097
+    },
+    {
+      "epoch": 5.737037037037037,
+      "grad_norm": 1.4390663806014874,
+      "learning_rate": 1.9332791839849328e-07,
+      "loss": 0.6141,
+      "step": 3098
+    },
+    {
+      "epoch": 5.738888888888889,
+      "grad_norm": 1.3569344420250968,
+      "learning_rate": 1.9332253214302997e-07,
+      "loss": 0.6828,
+      "step": 3099
+    },
+    {
+      "epoch": 5.7407407407407405,
+      "grad_norm": 3.348758968827594,
+      "learning_rate": 1.933171437894227e-07,
+      "loss": 0.6576,
+      "step": 3100
+    },
+    {
+      "epoch": 5.742592592592593,
+      "grad_norm": 1.805984983846325,
+      "learning_rate": 1.9331175333779257e-07,
+      "loss": 0.7046,
+      "step": 3101
+    },
+    {
+      "epoch": 5.7444444444444445,
+      "grad_norm": 1.3755396359386274,
+      "learning_rate": 1.933063607882608e-07,
+      "loss": 0.6598,
+      "step": 3102
+    },
+    {
+      "epoch": 5.746296296296296,
+      "grad_norm": 1.5206046602809848,
+      "learning_rate": 1.9330096614094864e-07,
+      "loss": 0.6279,
+      "step": 3103
+    },
+    {
+      "epoch": 5.7481481481481485,
+      "grad_norm": 1.9727222083918303,
+      "learning_rate": 1.9329556939597733e-07,
+      "loss": 0.6818,
+      "step": 3104
+    },
+    {
+      "epoch": 5.75,
+      "grad_norm": 1.4488535226291053,
+      "learning_rate": 1.9329017055346828e-07,
+      "loss": 0.6259,
+      "step": 3105
+    },
+    {
+      "epoch": 5.7518518518518515,
+      "grad_norm": 1.8723026657408528,
+      "learning_rate": 1.9328476961354277e-07,
+      "loss": 0.6493,
+      "step": 3106
+    },
+    {
+      "epoch": 5.753703703703704,
+      "grad_norm": 1.6088414825401331,
+      "learning_rate": 1.932793665763223e-07,
+      "loss": 0.6763,
+      "step": 3107
+    },
+    {
+      "epoch": 5.7555555555555555,
+      "grad_norm": 1.7677963540157626,
+      "learning_rate": 1.9327396144192835e-07,
+      "loss": 0.6611,
+      "step": 3108
+    },
+    {
+      "epoch": 5.757407407407407,
+      "grad_norm": 1.844647273626407,
+      "learning_rate": 1.932685542104824e-07,
+      "loss": 0.7264,
+      "step": 3109
+    },
+    {
+      "epoch": 5.7592592592592595,
+      "grad_norm": 1.6325255691915002,
+      "learning_rate": 1.9326314488210603e-07,
+      "loss": 0.6503,
+      "step": 3110
+    },
+    {
+      "epoch": 5.761111111111111,
+      "grad_norm": 1.3434655203475394,
+      "learning_rate": 1.932577334569209e-07,
+      "loss": 0.6529,
+      "step": 3111
+    },
+    {
+      "epoch": 5.762962962962963,
+      "grad_norm": 1.6255110816283564,
+      "learning_rate": 1.932523199350486e-07,
+      "loss": 0.6508,
+      "step": 3112
+    },
+    {
+      "epoch": 5.764814814814815,
+      "grad_norm": 1.6089545124745757,
+      "learning_rate": 1.932469043166109e-07,
+      "loss": 0.6728,
+      "step": 3113
+    },
+    {
+      "epoch": 5.766666666666667,
+      "grad_norm": 1.7591660200264743,
+      "learning_rate": 1.9324148660172952e-07,
+      "loss": 0.6494,
+      "step": 3114
+    },
+    {
+      "epoch": 5.768518518518518,
+      "grad_norm": 1.479934145887186,
+      "learning_rate": 1.9323606679052627e-07,
+      "loss": 0.6833,
+      "step": 3115
+    },
+    {
+      "epoch": 5.770370370370371,
+      "grad_norm": 2.0597281586506795,
+      "learning_rate": 1.9323064488312303e-07,
+      "loss": 0.6371,
+      "step": 3116
+    },
+    {
+      "epoch": 5.772222222222222,
+      "grad_norm": 1.4814336960370498,
+      "learning_rate": 1.932252208796417e-07,
+      "loss": 0.7106,
+      "step": 3117
+    },
+    {
+      "epoch": 5.774074074074074,
+      "grad_norm": 1.5476183995742245,
+      "learning_rate": 1.932197947802042e-07,
+      "loss": 0.7032,
+      "step": 3118
+    },
+    {
+      "epoch": 5.775925925925926,
+      "grad_norm": 1.5465393185472458,
+      "learning_rate": 1.932143665849325e-07,
+      "loss": 0.6848,
+      "step": 3119
+    },
+    {
+      "epoch": 5.777777777777778,
+      "grad_norm": 1.5982362981341727,
+      "learning_rate": 1.9320893629394873e-07,
+      "loss": 0.6983,
+      "step": 3120
+    },
+    {
+      "epoch": 5.77962962962963,
+      "grad_norm": 1.5609548423717912,
+      "learning_rate": 1.9320350390737484e-07,
+      "loss": 0.6436,
+      "step": 3121
+    },
+    {
+      "epoch": 5.781481481481482,
+      "grad_norm": 1.482513581745419,
+      "learning_rate": 1.931980694253331e-07,
+      "loss": 0.6482,
+      "step": 3122
+    },
+    {
+      "epoch": 5.783333333333333,
+      "grad_norm": 1.5968450403624015,
+      "learning_rate": 1.9319263284794564e-07,
+      "loss": 0.6284,
+      "step": 3123
+    },
+    {
+      "epoch": 5.785185185185185,
+      "grad_norm": 1.4210885808309044,
+      "learning_rate": 1.931871941753347e-07,
+      "loss": 0.6655,
+      "step": 3124
+    },
+    {
+      "epoch": 5.787037037037037,
+      "grad_norm": 1.5514072248636095,
+      "learning_rate": 1.931817534076225e-07,
+      "loss": 0.7029,
+      "step": 3125
+    },
+    {
+      "epoch": 5.788888888888889,
+      "grad_norm": 1.6207300826295732,
+      "learning_rate": 1.9317631054493142e-07,
+      "loss": 0.6508,
+      "step": 3126
+    },
+    {
+      "epoch": 5.790740740740741,
+      "grad_norm": 1.61700088372504,
+      "learning_rate": 1.9317086558738383e-07,
+      "loss": 0.6496,
+      "step": 3127
+    },
+    {
+      "epoch": 5.792592592592593,
+      "grad_norm": 1.6775660047990084,
+      "learning_rate": 1.931654185351021e-07,
+      "loss": 0.676,
+      "step": 3128
+    },
+    {
+      "epoch": 5.794444444444444,
+      "grad_norm": 1.4122099768719387,
+      "learning_rate": 1.9315996938820876e-07,
+      "loss": 0.7009,
+      "step": 3129
+    },
+    {
+      "epoch": 5.796296296296296,
+      "grad_norm": 1.593078724967063,
+      "learning_rate": 1.931545181468263e-07,
+      "loss": 0.6658,
+      "step": 3130
+    },
+    {
+      "epoch": 5.798148148148148,
+      "grad_norm": 1.989623385071106,
+      "learning_rate": 1.9314906481107722e-07,
+      "loss": 0.6589,
+      "step": 3131
+    },
+    {
+      "epoch": 5.8,
+      "grad_norm": 1.8452054582470114,
+      "learning_rate": 1.9314360938108424e-07,
+      "loss": 0.6918,
+      "step": 3132
+    },
+    {
+      "epoch": 5.801851851851852,
+      "grad_norm": 1.6139587687033337,
+      "learning_rate": 1.931381518569699e-07,
+      "loss": 0.6327,
+      "step": 3133
+    },
+    {
+      "epoch": 5.803703703703704,
+      "grad_norm": 1.8197311604550148,
+      "learning_rate": 1.9313269223885692e-07,
+      "loss": 0.6453,
+      "step": 3134
+    },
+    {
+      "epoch": 5.805555555555555,
+      "grad_norm": 2.2888817665587418,
+      "learning_rate": 1.9312723052686812e-07,
+      "loss": 0.7086,
+      "step": 3135
+    },
+    {
+      "epoch": 5.807407407407408,
+      "grad_norm": 1.4006108452242423,
+      "learning_rate": 1.9312176672112626e-07,
+      "loss": 0.6337,
+      "step": 3136
+    },
+    {
+      "epoch": 5.809259259259259,
+      "grad_norm": 1.4705205932652263,
+      "learning_rate": 1.9311630082175415e-07,
+      "loss": 0.6773,
+      "step": 3137
+    },
+    {
+      "epoch": 5.811111111111111,
+      "grad_norm": 1.663415331183204,
+      "learning_rate": 1.931108328288747e-07,
+      "loss": 0.697,
+      "step": 3138
+    },
+    {
+      "epoch": 5.812962962962963,
+      "grad_norm": 1.9988591218228136,
+      "learning_rate": 1.9310536274261083e-07,
+      "loss": 0.6556,
+      "step": 3139
+    },
+    {
+      "epoch": 5.814814814814815,
+      "grad_norm": 1.6198445757075786,
+      "learning_rate": 1.9309989056308553e-07,
+      "loss": 0.6886,
+      "step": 3140
+    },
+    {
+      "epoch": 5.816666666666666,
+      "grad_norm": 1.7800971799868306,
+      "learning_rate": 1.9309441629042186e-07,
+      "loss": 0.7088,
+      "step": 3141
+    },
+    {
+      "epoch": 5.818518518518519,
+      "grad_norm": 1.5991795396771342,
+      "learning_rate": 1.9308893992474285e-07,
+      "loss": 0.6969,
+      "step": 3142
+    },
+    {
+      "epoch": 5.82037037037037,
+      "grad_norm": 1.530754210435707,
+      "learning_rate": 1.9308346146617164e-07,
+      "loss": 0.6736,
+      "step": 3143
+    },
+    {
+      "epoch": 5.822222222222222,
+      "grad_norm": 1.7214837627237491,
+      "learning_rate": 1.9307798091483143e-07,
+      "loss": 0.651,
+      "step": 3144
+    },
+    {
+      "epoch": 5.824074074074074,
+      "grad_norm": 1.4735048569022127,
+      "learning_rate": 1.930724982708454e-07,
+      "loss": 0.6376,
+      "step": 3145
+    },
+    {
+      "epoch": 5.825925925925926,
+      "grad_norm": 1.606932832107485,
+      "learning_rate": 1.9306701353433685e-07,
+      "loss": 0.6213,
+      "step": 3146
+    },
+    {
+      "epoch": 5.8277777777777775,
+      "grad_norm": 1.472813998472956,
+      "learning_rate": 1.9306152670542906e-07,
+      "loss": 0.7253,
+      "step": 3147
+    },
+    {
+      "epoch": 5.82962962962963,
+      "grad_norm": 1.8671947681157124,
+      "learning_rate": 1.9305603778424537e-07,
+      "loss": 0.6667,
+      "step": 3148
+    },
+    {
+      "epoch": 5.8314814814814815,
+      "grad_norm": 1.5151768405053028,
+      "learning_rate": 1.9305054677090924e-07,
+      "loss": 0.6783,
+      "step": 3149
+    },
+    {
+      "epoch": 5.833333333333333,
+      "grad_norm": 1.6116926068269934,
+      "learning_rate": 1.930450536655441e-07,
+      "loss": 0.6297,
+      "step": 3150
+    },
+    {
+      "epoch": 5.8351851851851855,
+      "grad_norm": 1.6594230998501391,
+      "learning_rate": 1.9303955846827343e-07,
+      "loss": 0.741,
+      "step": 3151
+    },
+    {
+      "epoch": 5.837037037037037,
+      "grad_norm": 1.6451829488166798,
+      "learning_rate": 1.930340611792208e-07,
+      "loss": 0.6389,
+      "step": 3152
+    },
+    {
+      "epoch": 5.838888888888889,
+      "grad_norm": 1.5208046282215946,
+      "learning_rate": 1.930285617985098e-07,
+      "loss": 0.6726,
+      "step": 3153
+    },
+    {
+      "epoch": 5.840740740740741,
+      "grad_norm": 1.399174385757347,
+      "learning_rate": 1.9302306032626404e-07,
+      "loss": 0.6238,
+      "step": 3154
+    },
+    {
+      "epoch": 5.842592592592593,
+      "grad_norm": 1.6934333327737037,
+      "learning_rate": 1.9301755676260729e-07,
+      "loss": 0.6697,
+      "step": 3155
+    },
+    {
+      "epoch": 5.844444444444444,
+      "grad_norm": 1.6870100432688353,
+      "learning_rate": 1.9301205110766322e-07,
+      "loss": 0.688,
+      "step": 3156
+    },
+    {
+      "epoch": 5.8462962962962965,
+      "grad_norm": 1.3879402122138877,
+      "learning_rate": 1.930065433615556e-07,
+      "loss": 0.6294,
+      "step": 3157
+    },
+    {
+      "epoch": 5.848148148148148,
+      "grad_norm": 1.8457281896677984,
+      "learning_rate": 1.930010335244083e-07,
+      "loss": 0.6578,
+      "step": 3158
+    },
+    {
+      "epoch": 5.85,
+      "grad_norm": 1.735811771558393,
+      "learning_rate": 1.9299552159634515e-07,
+      "loss": 0.6395,
+      "step": 3159
+    },
+    {
+      "epoch": 5.851851851851852,
+      "grad_norm": 1.4172907329530469,
+      "learning_rate": 1.9299000757749013e-07,
+      "loss": 0.6818,
+      "step": 3160
+    },
+    {
+      "epoch": 5.853703703703704,
+      "grad_norm": 1.3985172088259052,
+      "learning_rate": 1.929844914679672e-07,
+      "loss": 0.6377,
+      "step": 3161
+    },
+    {
+      "epoch": 5.855555555555555,
+      "grad_norm": 1.3511012971509335,
+      "learning_rate": 1.929789732679003e-07,
+      "loss": 0.6998,
+      "step": 3162
+    },
+    {
+      "epoch": 5.857407407407408,
+      "grad_norm": 1.4673256331251812,
+      "learning_rate": 1.9297345297741362e-07,
+      "loss": 0.6731,
+      "step": 3163
+    },
+    {
+      "epoch": 5.859259259259259,
+      "grad_norm": 1.5154356547355405,
+      "learning_rate": 1.9296793059663118e-07,
+      "loss": 0.6685,
+      "step": 3164
+    },
+    {
+      "epoch": 5.861111111111111,
+      "grad_norm": 1.471094740838117,
+      "learning_rate": 1.9296240612567716e-07,
+      "loss": 0.6708,
+      "step": 3165
+    },
+    {
+      "epoch": 5.862962962962963,
+      "grad_norm": 1.6823911525751123,
+      "learning_rate": 1.9295687956467577e-07,
+      "loss": 0.7071,
+      "step": 3166
+    },
+    {
+      "epoch": 5.864814814814815,
+      "grad_norm": 1.549637030960512,
+      "learning_rate": 1.9295135091375126e-07,
+      "loss": 0.7085,
+      "step": 3167
+    },
+    {
+      "epoch": 5.866666666666667,
+      "grad_norm": 1.4823879490741958,
+      "learning_rate": 1.9294582017302794e-07,
+      "loss": 0.7546,
+      "step": 3168
+    },
+    {
+      "epoch": 5.868518518518519,
+      "grad_norm": 1.6035790223586877,
+      "learning_rate": 1.9294028734263014e-07,
+      "loss": 0.6597,
+      "step": 3169
+    },
+    {
+      "epoch": 5.87037037037037,
+      "grad_norm": 1.3380787951200714,
+      "learning_rate": 1.9293475242268223e-07,
+      "loss": 0.655,
+      "step": 3170
+    },
+    {
+      "epoch": 5.872222222222222,
+      "grad_norm": 1.6254994848876927,
+      "learning_rate": 1.929292154133087e-07,
+      "loss": 0.6769,
+      "step": 3171
+    },
+    {
+      "epoch": 5.874074074074074,
+      "grad_norm": 1.6421715807954698,
+      "learning_rate": 1.9292367631463402e-07,
+      "loss": 0.6455,
+      "step": 3172
+    },
+    {
+      "epoch": 5.875925925925926,
+      "grad_norm": 1.6381023825796484,
+      "learning_rate": 1.9291813512678272e-07,
+      "loss": 0.6654,
+      "step": 3173
+    },
+    {
+      "epoch": 5.877777777777778,
+      "grad_norm": 1.8532628640486033,
+      "learning_rate": 1.9291259184987938e-07,
+      "loss": 0.6371,
+      "step": 3174
+    },
+    {
+      "epoch": 5.87962962962963,
+      "grad_norm": 1.6330188547559685,
+      "learning_rate": 1.9290704648404863e-07,
+      "loss": 0.6715,
+      "step": 3175
+    },
+    {
+      "epoch": 5.881481481481481,
+      "grad_norm": 1.504033728320947,
+      "learning_rate": 1.9290149902941515e-07,
+      "loss": 0.6551,
+      "step": 3176
+    },
+    {
+      "epoch": 5.883333333333333,
+      "grad_norm": 2.0499396534022436,
+      "learning_rate": 1.9289594948610364e-07,
+      "loss": 0.6942,
+      "step": 3177
+    },
+    {
+      "epoch": 5.885185185185185,
+      "grad_norm": 1.5782185557272341,
+      "learning_rate": 1.9289039785423888e-07,
+      "loss": 0.6325,
+      "step": 3178
+    },
+    {
+      "epoch": 5.887037037037037,
+      "grad_norm": 1.448502024093899,
+      "learning_rate": 1.9288484413394567e-07,
+      "loss": 0.6532,
+      "step": 3179
+    },
+    {
+      "epoch": 5.888888888888889,
+      "grad_norm": 1.3941981189264658,
+      "learning_rate": 1.9287928832534893e-07,
+      "loss": 0.6546,
+      "step": 3180
+    },
+    {
+      "epoch": 5.890740740740741,
+      "grad_norm": 2.0057391865140755,
+      "learning_rate": 1.9287373042857356e-07,
+      "loss": 0.6574,
+      "step": 3181
+    },
+    {
+      "epoch": 5.892592592592592,
+      "grad_norm": 2.0301992019694723,
+      "learning_rate": 1.9286817044374443e-07,
+      "loss": 0.6781,
+      "step": 3182
+    },
+    {
+      "epoch": 5.894444444444445,
+      "grad_norm": 1.6096883955639443,
+      "learning_rate": 1.9286260837098662e-07,
+      "loss": 0.7255,
+      "step": 3183
+    },
+    {
+      "epoch": 5.896296296296296,
+      "grad_norm": 1.4246677287421317,
+      "learning_rate": 1.9285704421042516e-07,
+      "loss": 0.6526,
+      "step": 3184
+    },
+    {
+      "epoch": 5.898148148148148,
+      "grad_norm": 1.7999254664651674,
+      "learning_rate": 1.9285147796218516e-07,
+      "loss": 0.6591,
+      "step": 3185
+    },
+    {
+      "epoch": 5.9,
+      "grad_norm": 1.7472846916172196,
+      "learning_rate": 1.9284590962639176e-07,
+      "loss": 0.6475,
+      "step": 3186
+    },
+    {
+      "epoch": 5.901851851851852,
+      "grad_norm": 1.5182371952362772,
+      "learning_rate": 1.9284033920317013e-07,
+      "loss": 0.6531,
+      "step": 3187
+    },
+    {
+      "epoch": 5.9037037037037035,
+      "grad_norm": 1.399734562440533,
+      "learning_rate": 1.9283476669264554e-07,
+      "loss": 0.6824,
+      "step": 3188
+    },
+    {
+      "epoch": 5.905555555555556,
+      "grad_norm": 1.4505831566307255,
+      "learning_rate": 1.9282919209494323e-07,
+      "loss": 0.6564,
+      "step": 3189
+    },
+    {
+      "epoch": 5.907407407407407,
+      "grad_norm": 1.6983977132524728,
+      "learning_rate": 1.9282361541018858e-07,
+      "loss": 0.6363,
+      "step": 3190
+    },
+    {
+      "epoch": 5.909259259259259,
+      "grad_norm": 1.462861434244709,
+      "learning_rate": 1.9281803663850699e-07,
+      "loss": 0.6915,
+      "step": 3191
+    },
+    {
+      "epoch": 5.911111111111111,
+      "grad_norm": 1.6590387438118874,
+      "learning_rate": 1.9281245578002378e-07,
+      "loss": 0.6384,
+      "step": 3192
+    },
+    {
+      "epoch": 5.912962962962963,
+      "grad_norm": 1.5221404311931686,
+      "learning_rate": 1.9280687283486455e-07,
+      "loss": 0.7088,
+      "step": 3193
+    },
+    {
+      "epoch": 5.9148148148148145,
+      "grad_norm": 1.6158752464994992,
+      "learning_rate": 1.9280128780315472e-07,
+      "loss": 0.6702,
+      "step": 3194
+    },
+    {
+      "epoch": 5.916666666666667,
+      "grad_norm": 1.4166402620577003,
+      "learning_rate": 1.927957006850199e-07,
+      "loss": 0.6814,
+      "step": 3195
+    },
+    {
+      "epoch": 5.9185185185185185,
+      "grad_norm": 1.4815897112790115,
+      "learning_rate": 1.9279011148058572e-07,
+      "loss": 0.6381,
+      "step": 3196
+    },
+    {
+      "epoch": 5.92037037037037,
+      "grad_norm": 1.4596040691377503,
+      "learning_rate": 1.9278452018997779e-07,
+      "loss": 0.68,
+      "step": 3197
+    },
+    {
+      "epoch": 5.9222222222222225,
+      "grad_norm": 1.3901922117362056,
+      "learning_rate": 1.9277892681332183e-07,
+      "loss": 0.6459,
+      "step": 3198
+    },
+    {
+      "epoch": 5.924074074074074,
+      "grad_norm": 1.9296858392826894,
+      "learning_rate": 1.9277333135074368e-07,
+      "loss": 0.6415,
+      "step": 3199
+    },
+    {
+      "epoch": 5.925925925925926,
+      "grad_norm": 1.5138859044023336,
+      "learning_rate": 1.92767733802369e-07,
+      "loss": 0.6552,
+      "step": 3200
+    },
+    {
+      "epoch": 5.927777777777778,
+      "grad_norm": 1.7278625766545153,
+      "learning_rate": 1.9276213416832374e-07,
+      "loss": 0.6726,
+      "step": 3201
+    },
+    {
+      "epoch": 5.92962962962963,
+      "grad_norm": 1.4943273887276074,
+      "learning_rate": 1.927565324487338e-07,
+      "loss": 0.6791,
+      "step": 3202
+    },
+    {
+      "epoch": 5.931481481481481,
+      "grad_norm": 1.6859281712062053,
+      "learning_rate": 1.9275092864372504e-07,
+      "loss": 0.6731,
+      "step": 3203
+    },
+    {
+      "epoch": 5.933333333333334,
+      "grad_norm": 1.5249263227319036,
+      "learning_rate": 1.9274532275342352e-07,
+      "loss": 0.6411,
+      "step": 3204
+    },
+    {
+      "epoch": 5.935185185185185,
+      "grad_norm": 1.3001713793707412,
+      "learning_rate": 1.9273971477795525e-07,
+      "loss": 0.6836,
+      "step": 3205
+    },
+    {
+      "epoch": 5.937037037037037,
+      "grad_norm": 1.8766222587097663,
+      "learning_rate": 1.9273410471744632e-07,
+      "loss": 0.6625,
+      "step": 3206
+    },
+    {
+      "epoch": 5.938888888888889,
+      "grad_norm": 1.567684611784952,
+      "learning_rate": 1.9272849257202283e-07,
+      "loss": 0.673,
+      "step": 3207
+    },
+    {
+      "epoch": 5.940740740740741,
+      "grad_norm": 2.072647953912945,
+      "learning_rate": 1.92722878341811e-07,
+      "loss": 0.6637,
+      "step": 3208
+    },
+    {
+      "epoch": 5.942592592592592,
+      "grad_norm": 1.3276483101642031,
+      "learning_rate": 1.92717262026937e-07,
+      "loss": 0.67,
+      "step": 3209
+    },
+    {
+      "epoch": 5.944444444444445,
+      "grad_norm": 1.6291842182860714,
+      "learning_rate": 1.9271164362752718e-07,
+      "loss": 0.6295,
+      "step": 3210
+    },
+    {
+      "epoch": 5.946296296296296,
+      "grad_norm": 1.6531296674625306,
+      "learning_rate": 1.9270602314370778e-07,
+      "loss": 0.7137,
+      "step": 3211
+    },
+    {
+      "epoch": 5.948148148148148,
+      "grad_norm": 1.5011625741864751,
+      "learning_rate": 1.927004005756052e-07,
+      "loss": 0.6711,
+      "step": 3212
+    },
+    {
+      "epoch": 5.95,
+      "grad_norm": 2.2772641566950367,
+      "learning_rate": 1.9269477592334587e-07,
+      "loss": 0.6786,
+      "step": 3213
+    },
+    {
+      "epoch": 5.951851851851852,
+      "grad_norm": 2.3301676799243016,
+      "learning_rate": 1.926891491870562e-07,
+      "loss": 0.6043,
+      "step": 3214
+    },
+    {
+      "epoch": 5.953703703703704,
+      "grad_norm": 2.1565166126831206,
+      "learning_rate": 1.9268352036686274e-07,
+      "loss": 0.6381,
+      "step": 3215
+    },
+    {
+      "epoch": 5.955555555555556,
+      "grad_norm": 1.3965667144729528,
+      "learning_rate": 1.9267788946289199e-07,
+      "loss": 0.6936,
+      "step": 3216
+    },
+    {
+      "epoch": 5.957407407407407,
+      "grad_norm": 1.668933927186015,
+      "learning_rate": 1.9267225647527058e-07,
+      "loss": 0.6989,
+      "step": 3217
+    },
+    {
+      "epoch": 5.959259259259259,
+      "grad_norm": 1.6814189012613956,
+      "learning_rate": 1.9266662140412515e-07,
+      "loss": 0.6764,
+      "step": 3218
+    },
+    {
+      "epoch": 5.961111111111111,
+      "grad_norm": 1.6573762483619858,
+      "learning_rate": 1.926609842495824e-07,
+      "loss": 0.6587,
+      "step": 3219
+    },
+    {
+      "epoch": 5.962962962962963,
+      "grad_norm": 1.5925934986670407,
+      "learning_rate": 1.9265534501176905e-07,
+      "loss": 0.6646,
+      "step": 3220
+    },
+    {
+      "epoch": 5.964814814814815,
+      "grad_norm": 1.7386825765795433,
+      "learning_rate": 1.926497036908119e-07,
+      "loss": 0.7053,
+      "step": 3221
+    },
+    {
+      "epoch": 5.966666666666667,
+      "grad_norm": 1.5513124747883649,
+      "learning_rate": 1.9264406028683778e-07,
+      "loss": 0.6114,
+      "step": 3222
+    },
+    {
+      "epoch": 5.968518518518518,
+      "grad_norm": 1.741564320479593,
+      "learning_rate": 1.9263841479997357e-07,
+      "loss": 0.6316,
+      "step": 3223
+    },
+    {
+      "epoch": 5.97037037037037,
+      "grad_norm": 1.5200633163767672,
+      "learning_rate": 1.926327672303462e-07,
+      "loss": 0.6797,
+      "step": 3224
+    },
+    {
+      "epoch": 5.972222222222222,
+      "grad_norm": 1.577544443816808,
+      "learning_rate": 1.9262711757808262e-07,
+      "loss": 0.6952,
+      "step": 3225
+    },
+    {
+      "epoch": 5.974074074074074,
+      "grad_norm": 1.5480104319689156,
+      "learning_rate": 1.9262146584330987e-07,
+      "loss": 0.6544,
+      "step": 3226
+    },
+    {
+      "epoch": 5.975925925925926,
+      "grad_norm": 1.5372275894781748,
+      "learning_rate": 1.92615812026155e-07,
+      "loss": 0.681,
+      "step": 3227
+    },
+    {
+      "epoch": 5.977777777777778,
+      "grad_norm": 1.700402767559671,
+      "learning_rate": 1.9261015612674515e-07,
+      "loss": 0.687,
+      "step": 3228
+    },
+    {
+      "epoch": 5.979629629629629,
+      "grad_norm": 1.5279800214160841,
+      "learning_rate": 1.9260449814520747e-07,
+      "loss": 0.6141,
+      "step": 3229
+    },
+    {
+      "epoch": 5.981481481481482,
+      "grad_norm": 1.6164714346226579,
+      "learning_rate": 1.9259883808166918e-07,
+      "loss": 0.6804,
+      "step": 3230
+    },
+    {
+      "epoch": 5.983333333333333,
+      "grad_norm": 1.5886694542372342,
+      "learning_rate": 1.9259317593625747e-07,
+      "loss": 0.6719,
+      "step": 3231
+    },
+    {
+      "epoch": 5.985185185185185,
+      "grad_norm": 1.49270136076904,
+      "learning_rate": 1.925875117090997e-07,
+      "loss": 0.6736,
+      "step": 3232
+    },
+    {
+      "epoch": 5.987037037037037,
+      "grad_norm": 1.5241185598741427,
+      "learning_rate": 1.9258184540032322e-07,
+      "loss": 0.6624,
+      "step": 3233
+    },
+    {
+      "epoch": 5.988888888888889,
+      "grad_norm": 1.4837138337792564,
+      "learning_rate": 1.925761770100554e-07,
+      "loss": 0.6876,
+      "step": 3234
+    },
+    {
+      "epoch": 5.9907407407407405,
+      "grad_norm": 1.6546796169015763,
+      "learning_rate": 1.9257050653842369e-07,
+      "loss": 0.6527,
+      "step": 3235
+    },
+    {
+      "epoch": 5.992592592592593,
+      "grad_norm": 1.5188511838123737,
+      "learning_rate": 1.9256483398555553e-07,
+      "loss": 0.6526,
+      "step": 3236
+    },
+    {
+      "epoch": 5.9944444444444445,
+      "grad_norm": 1.5601513374579616,
+      "learning_rate": 1.9255915935157857e-07,
+      "loss": 0.6818,
+      "step": 3237
+    },
+    {
+      "epoch": 5.996296296296296,
+      "grad_norm": 1.5844502052633924,
+      "learning_rate": 1.9255348263662027e-07,
+      "loss": 0.6483,
+      "step": 3238
+    },
+    {
+      "epoch": 5.9981481481481485,
+      "grad_norm": 1.695389321959413,
+      "learning_rate": 1.9254780384080833e-07,
+      "loss": 0.6888,
+      "step": 3239
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 1.506239263979736,
+      "learning_rate": 1.925421229642704e-07,
+      "loss": 0.6766,
+      "step": 3240
+    },
+    {
+      "epoch": 6.0018518518518515,
+      "grad_norm": 1.4448384705576789,
+      "learning_rate": 1.9253644000713422e-07,
+      "loss": 0.658,
+      "step": 3241
+    },
+    {
+      "epoch": 6.003703703703704,
+      "grad_norm": 1.5068671226051193,
+      "learning_rate": 1.9253075496952754e-07,
+      "loss": 0.663,
+      "step": 3242
+    },
+    {
+      "epoch": 6.0055555555555555,
+      "grad_norm": 1.9336947608505894,
+      "learning_rate": 1.9252506785157817e-07,
+      "loss": 0.667,
+      "step": 3243
+    },
+    {
+      "epoch": 6.007407407407407,
+      "grad_norm": 1.4939981006308398,
+      "learning_rate": 1.9251937865341397e-07,
+      "loss": 0.6319,
+      "step": 3244
+    },
+    {
+      "epoch": 6.0092592592592595,
+      "grad_norm": 1.5241668308773684,
+      "learning_rate": 1.9251368737516286e-07,
+      "loss": 0.6379,
+      "step": 3245
+    },
+    {
+      "epoch": 6.011111111111111,
+      "grad_norm": 1.9521554429013426,
+      "learning_rate": 1.925079940169528e-07,
+      "loss": 0.6845,
+      "step": 3246
+    },
+    {
+      "epoch": 6.012962962962963,
+      "grad_norm": 1.5794745158179309,
+      "learning_rate": 1.925022985789118e-07,
+      "loss": 0.6585,
+      "step": 3247
+    },
+    {
+      "epoch": 6.014814814814815,
+      "grad_norm": 1.9881908666505577,
+      "learning_rate": 1.924966010611679e-07,
+      "loss": 0.661,
+      "step": 3248
+    },
+    {
+      "epoch": 6.016666666666667,
+      "grad_norm": 1.402943567940391,
+      "learning_rate": 1.9249090146384918e-07,
+      "loss": 0.6384,
+      "step": 3249
+    },
+    {
+      "epoch": 6.018518518518518,
+      "grad_norm": 1.4462048670835508,
+      "learning_rate": 1.9248519978708377e-07,
+      "loss": 0.6747,
+      "step": 3250
+    },
+    {
+      "epoch": 6.020370370370371,
+      "grad_norm": 1.6993090392276184,
+      "learning_rate": 1.9247949603099988e-07,
+      "loss": 0.6143,
+      "step": 3251
+    },
+    {
+      "epoch": 6.022222222222222,
+      "grad_norm": 1.5586666410459007,
+      "learning_rate": 1.924737901957258e-07,
+      "loss": 0.6645,
+      "step": 3252
+    },
+    {
+      "epoch": 6.024074074074074,
+      "grad_norm": 1.5249833849458387,
+      "learning_rate": 1.9246808228138972e-07,
+      "loss": 0.6642,
+      "step": 3253
+    },
+    {
+      "epoch": 6.025925925925926,
+      "grad_norm": 1.6049132156048016,
+      "learning_rate": 1.9246237228812002e-07,
+      "loss": 0.6373,
+      "step": 3254
+    },
+    {
+      "epoch": 6.027777777777778,
+      "grad_norm": 1.665420901538134,
+      "learning_rate": 1.9245666021604507e-07,
+      "loss": 0.6542,
+      "step": 3255
+    },
+    {
+      "epoch": 6.029629629629629,
+      "grad_norm": 1.5923442851880225,
+      "learning_rate": 1.9245094606529332e-07,
+      "loss": 0.6684,
+      "step": 3256
+    },
+    {
+      "epoch": 6.031481481481482,
+      "grad_norm": 1.547366313597984,
+      "learning_rate": 1.9244522983599315e-07,
+      "loss": 0.6337,
+      "step": 3257
+    },
+    {
+      "epoch": 6.033333333333333,
+      "grad_norm": 2.153715134123784,
+      "learning_rate": 1.924395115282732e-07,
+      "loss": 0.6686,
+      "step": 3258
+    },
+    {
+      "epoch": 6.035185185185185,
+      "grad_norm": 1.4563291619516572,
+      "learning_rate": 1.924337911422619e-07,
+      "loss": 0.6763,
+      "step": 3259
+    },
+    {
+      "epoch": 6.037037037037037,
+      "grad_norm": 1.6643606281204921,
+      "learning_rate": 1.9242806867808795e-07,
+      "loss": 0.6638,
+      "step": 3260
+    },
+    {
+      "epoch": 6.038888888888889,
+      "grad_norm": 1.548870741599961,
+      "learning_rate": 1.9242234413588002e-07,
+      "loss": 0.6497,
+      "step": 3261
+    },
+    {
+      "epoch": 6.040740740740741,
+      "grad_norm": 2.9327633456583677,
+      "learning_rate": 1.9241661751576676e-07,
+      "loss": 0.6674,
+      "step": 3262
+    },
+    {
+      "epoch": 6.042592592592593,
+      "grad_norm": 1.4901888064230668,
+      "learning_rate": 1.9241088881787694e-07,
+      "loss": 0.6462,
+      "step": 3263
+    },
+    {
+      "epoch": 6.044444444444444,
+      "grad_norm": 1.4065076667917518,
+      "learning_rate": 1.9240515804233936e-07,
+      "loss": 0.6399,
+      "step": 3264
+    },
+    {
+      "epoch": 6.046296296296297,
+      "grad_norm": 1.350434313806594,
+      "learning_rate": 1.9239942518928286e-07,
+      "loss": 0.6731,
+      "step": 3265
+    },
+    {
+      "epoch": 6.048148148148148,
+      "grad_norm": 1.4158879874539907,
+      "learning_rate": 1.9239369025883628e-07,
+      "loss": 0.6648,
+      "step": 3266
+    },
+    {
+      "epoch": 6.05,
+      "grad_norm": 1.9114391748446615,
+      "learning_rate": 1.9238795325112868e-07,
+      "loss": 0.67,
+      "step": 3267
+    },
+    {
+      "epoch": 6.051851851851852,
+      "grad_norm": 1.6138123657063617,
+      "learning_rate": 1.9238221416628892e-07,
+      "loss": 0.6532,
+      "step": 3268
+    },
+    {
+      "epoch": 6.053703703703704,
+      "grad_norm": 1.4063975574167729,
+      "learning_rate": 1.923764730044461e-07,
+      "loss": 0.6801,
+      "step": 3269
+    },
+    {
+      "epoch": 6.055555555555555,
+      "grad_norm": 1.466317451244446,
+      "learning_rate": 1.923707297657293e-07,
+      "loss": 0.681,
+      "step": 3270
+    },
+    {
+      "epoch": 6.057407407407408,
+      "grad_norm": 2.0148717854543308,
+      "learning_rate": 1.9236498445026753e-07,
+      "loss": 0.7087,
+      "step": 3271
+    },
+    {
+      "epoch": 6.059259259259259,
+      "grad_norm": 1.6243257950685415,
+      "learning_rate": 1.9235923705819013e-07,
+      "loss": 0.6456,
+      "step": 3272
+    },
+    {
+      "epoch": 6.061111111111111,
+      "grad_norm": 1.6272678880330924,
+      "learning_rate": 1.9235348758962622e-07,
+      "loss": 0.6736,
+      "step": 3273
+    },
+    {
+      "epoch": 6.062962962962963,
+      "grad_norm": 2.1243603306086842,
+      "learning_rate": 1.9234773604470507e-07,
+      "loss": 0.6277,
+      "step": 3274
+    },
+    {
+      "epoch": 6.064814814814815,
+      "grad_norm": 1.4682004796994141,
+      "learning_rate": 1.9234198242355602e-07,
+      "loss": 0.7142,
+      "step": 3275
+    },
+    {
+      "epoch": 6.066666666666666,
+      "grad_norm": 1.4514682179896041,
+      "learning_rate": 1.923362267263084e-07,
+      "loss": 0.6753,
+      "step": 3276
+    },
+    {
+      "epoch": 6.068518518518519,
+      "grad_norm": 1.4264938055029956,
+      "learning_rate": 1.923304689530916e-07,
+      "loss": 0.6361,
+      "step": 3277
+    },
+    {
+      "epoch": 6.07037037037037,
+      "grad_norm": 1.3774428955600742,
+      "learning_rate": 1.923247091040351e-07,
+      "loss": 0.6927,
+      "step": 3278
+    },
+    {
+      "epoch": 6.072222222222222,
+      "grad_norm": 1.8117583015479766,
+      "learning_rate": 1.9231894717926842e-07,
+      "loss": 0.671,
+      "step": 3279
+    },
+    {
+      "epoch": 6.074074074074074,
+      "grad_norm": 1.3815959151207655,
+      "learning_rate": 1.9231318317892106e-07,
+      "loss": 0.6762,
+      "step": 3280
+    },
+    {
+      "epoch": 6.075925925925926,
+      "grad_norm": 1.519901756847425,
+      "learning_rate": 1.923074171031226e-07,
+      "loss": 0.7097,
+      "step": 3281
+    },
+    {
+      "epoch": 6.0777777777777775,
+      "grad_norm": 1.703388260101047,
+      "learning_rate": 1.9230164895200271e-07,
+      "loss": 0.6981,
+      "step": 3282
+    },
+    {
+      "epoch": 6.07962962962963,
+      "grad_norm": 1.3739243694967267,
+      "learning_rate": 1.9229587872569105e-07,
+      "loss": 0.6897,
+      "step": 3283
+    },
+    {
+      "epoch": 6.0814814814814815,
+      "grad_norm": 2.1783638495690867,
+      "learning_rate": 1.9229010642431742e-07,
+      "loss": 0.6794,
+      "step": 3284
+    },
+    {
+      "epoch": 6.083333333333333,
+      "grad_norm": 1.6497742497369867,
+      "learning_rate": 1.9228433204801148e-07,
+      "loss": 0.6836,
+      "step": 3285
+    },
+    {
+      "epoch": 6.0851851851851855,
+      "grad_norm": 2.0938996057325054,
+      "learning_rate": 1.9227855559690314e-07,
+      "loss": 0.6669,
+      "step": 3286
+    },
+    {
+      "epoch": 6.087037037037037,
+      "grad_norm": 1.4443119554837531,
+      "learning_rate": 1.9227277707112226e-07,
+      "loss": 0.6875,
+      "step": 3287
+    },
+    {
+      "epoch": 6.088888888888889,
+      "grad_norm": 3.909007861073493,
+      "learning_rate": 1.9226699647079872e-07,
+      "loss": 0.6662,
+      "step": 3288
+    },
+    {
+      "epoch": 6.090740740740741,
+      "grad_norm": 1.5774446082880966,
+      "learning_rate": 1.9226121379606254e-07,
+      "loss": 0.6785,
+      "step": 3289
+    },
+    {
+      "epoch": 6.092592592592593,
+      "grad_norm": 1.5570381030724414,
+      "learning_rate": 1.9225542904704365e-07,
+      "loss": 0.6661,
+      "step": 3290
+    },
+    {
+      "epoch": 6.094444444444444,
+      "grad_norm": 1.7722882775673867,
+      "learning_rate": 1.922496422238722e-07,
+      "loss": 0.6532,
+      "step": 3291
+    },
+    {
+      "epoch": 6.0962962962962965,
+      "grad_norm": 1.4153832125684849,
+      "learning_rate": 1.922438533266782e-07,
+      "loss": 0.7035,
+      "step": 3292
+    },
+    {
+      "epoch": 6.098148148148148,
+      "grad_norm": 2.0858320184451853,
+      "learning_rate": 1.9223806235559189e-07,
+      "loss": 0.647,
+      "step": 3293
+    },
+    {
+      "epoch": 6.1,
+      "grad_norm": 1.6764944847519188,
+      "learning_rate": 1.922322693107434e-07,
+      "loss": 0.6528,
+      "step": 3294
+    },
+    {
+      "epoch": 6.101851851851852,
+      "grad_norm": 1.7162558996793766,
+      "learning_rate": 1.9222647419226297e-07,
+      "loss": 0.6384,
+      "step": 3295
+    },
+    {
+      "epoch": 6.103703703703704,
+      "grad_norm": 1.6493952329606891,
+      "learning_rate": 1.9222067700028095e-07,
+      "loss": 0.6685,
+      "step": 3296
+    },
+    {
+      "epoch": 6.105555555555555,
+      "grad_norm": 1.481788648070295,
+      "learning_rate": 1.9221487773492764e-07,
+      "loss": 0.6973,
+      "step": 3297
+    },
+    {
+      "epoch": 6.107407407407408,
+      "grad_norm": 1.4495162455776365,
+      "learning_rate": 1.9220907639633342e-07,
+      "loss": 0.6299,
+      "step": 3298
+    },
+    {
+      "epoch": 6.109259259259259,
+      "grad_norm": 1.555265097779018,
+      "learning_rate": 1.9220327298462874e-07,
+      "loss": 0.7063,
+      "step": 3299
+    },
+    {
+      "epoch": 6.111111111111111,
+      "grad_norm": 1.6035271386899408,
+      "learning_rate": 1.9219746749994403e-07,
+      "loss": 0.6269,
+      "step": 3300
+    },
+    {
+      "epoch": 6.112962962962963,
+      "grad_norm": 1.3245827542119324,
+      "learning_rate": 1.9219165994240987e-07,
+      "loss": 0.695,
+      "step": 3301
+    },
+    {
+      "epoch": 6.114814814814815,
+      "grad_norm": 1.5082492443585795,
+      "learning_rate": 1.9218585031215682e-07,
+      "loss": 0.6433,
+      "step": 3302
+    },
+    {
+      "epoch": 6.116666666666666,
+      "grad_norm": 2.3193651821465053,
+      "learning_rate": 1.9218003860931548e-07,
+      "loss": 0.6656,
+      "step": 3303
+    },
+    {
+      "epoch": 6.118518518518519,
+      "grad_norm": 1.517198966669395,
+      "learning_rate": 1.921742248340165e-07,
+      "loss": 0.6111,
+      "step": 3304
+    },
+    {
+      "epoch": 6.12037037037037,
+      "grad_norm": 1.638136947163177,
+      "learning_rate": 1.921684089863906e-07,
+      "loss": 0.6501,
+      "step": 3305
+    },
+    {
+      "epoch": 6.122222222222222,
+      "grad_norm": 1.565969798221956,
+      "learning_rate": 1.9216259106656855e-07,
+      "loss": 0.6786,
+      "step": 3306
+    },
+    {
+      "epoch": 6.124074074074074,
+      "grad_norm": 1.7694339421775493,
+      "learning_rate": 1.9215677107468113e-07,
+      "loss": 0.6658,
+      "step": 3307
+    },
+    {
+      "epoch": 6.125925925925926,
+      "grad_norm": 1.6803687136826442,
+      "learning_rate": 1.921509490108592e-07,
+      "loss": 0.6685,
+      "step": 3308
+    },
+    {
+      "epoch": 6.127777777777778,
+      "grad_norm": 1.4359361649591924,
+      "learning_rate": 1.921451248752337e-07,
+      "loss": 0.6363,
+      "step": 3309
+    },
+    {
+      "epoch": 6.12962962962963,
+      "grad_norm": 1.4338033274768724,
+      "learning_rate": 1.921392986679355e-07,
+      "loss": 0.7184,
+      "step": 3310
+    },
+    {
+      "epoch": 6.131481481481481,
+      "grad_norm": 1.5341599626288873,
+      "learning_rate": 1.921334703890956e-07,
+      "loss": 0.6685,
+      "step": 3311
+    },
+    {
+      "epoch": 6.133333333333334,
+      "grad_norm": 1.7760408779864978,
+      "learning_rate": 1.921276400388451e-07,
+      "loss": 0.6626,
+      "step": 3312
+    },
+    {
+      "epoch": 6.135185185185185,
+      "grad_norm": 1.701511965565045,
+      "learning_rate": 1.92121807617315e-07,
+      "loss": 0.7086,
+      "step": 3313
+    },
+    {
+      "epoch": 6.137037037037037,
+      "grad_norm": 1.5488584755189525,
+      "learning_rate": 1.9211597312463648e-07,
+      "loss": 0.6388,
+      "step": 3314
+    },
+    {
+      "epoch": 6.138888888888889,
+      "grad_norm": 1.6314173807773145,
+      "learning_rate": 1.921101365609407e-07,
+      "loss": 0.6746,
+      "step": 3315
+    },
+    {
+      "epoch": 6.140740740740741,
+      "grad_norm": 1.8598083034126227,
+      "learning_rate": 1.9210429792635888e-07,
+      "loss": 0.684,
+      "step": 3316
+    },
+    {
+      "epoch": 6.142592592592592,
+      "grad_norm": 1.5424155430017175,
+      "learning_rate": 1.9209845722102228e-07,
+      "loss": 0.6425,
+      "step": 3317
+    },
+    {
+      "epoch": 6.144444444444445,
+      "grad_norm": 1.6110149182668472,
+      "learning_rate": 1.9209261444506227e-07,
+      "loss": 0.6346,
+      "step": 3318
+    },
+    {
+      "epoch": 6.146296296296296,
+      "grad_norm": 1.444019835423443,
+      "learning_rate": 1.9208676959861012e-07,
+      "loss": 0.6801,
+      "step": 3319
+    },
+    {
+      "epoch": 6.148148148148148,
+      "grad_norm": 1.6515702559174343,
+      "learning_rate": 1.920809226817973e-07,
+      "loss": 0.672,
+      "step": 3320
+    },
+    {
+      "epoch": 6.15,
+      "grad_norm": 1.7143028084078356,
+      "learning_rate": 1.9207507369475527e-07,
+      "loss": 0.6689,
+      "step": 3321
+    },
+    {
+      "epoch": 6.151851851851852,
+      "grad_norm": 1.51699561976835,
+      "learning_rate": 1.920692226376155e-07,
+      "loss": 0.6571,
+      "step": 3322
+    },
+    {
+      "epoch": 6.1537037037037035,
+      "grad_norm": 1.3188794834709554,
+      "learning_rate": 1.9206336951050952e-07,
+      "loss": 0.6946,
+      "step": 3323
+    },
+    {
+      "epoch": 6.155555555555556,
+      "grad_norm": 1.7836721402771154,
+      "learning_rate": 1.9205751431356897e-07,
+      "loss": 0.6682,
+      "step": 3324
+    },
+    {
+      "epoch": 6.157407407407407,
+      "grad_norm": 1.695294532941119,
+      "learning_rate": 1.920516570469255e-07,
+      "loss": 0.6956,
+      "step": 3325
+    },
+    {
+      "epoch": 6.159259259259259,
+      "grad_norm": 1.7260113051081005,
+      "learning_rate": 1.9204579771071075e-07,
+      "loss": 0.6461,
+      "step": 3326
+    },
+    {
+      "epoch": 6.161111111111111,
+      "grad_norm": 1.4013776990874338,
+      "learning_rate": 1.9203993630505648e-07,
+      "loss": 0.6735,
+      "step": 3327
+    },
+    {
+      "epoch": 6.162962962962963,
+      "grad_norm": 1.606015622454816,
+      "learning_rate": 1.9203407283009444e-07,
+      "loss": 0.67,
+      "step": 3328
+    },
+    {
+      "epoch": 6.1648148148148145,
+      "grad_norm": 3.0184006279633704,
+      "learning_rate": 1.920282072859565e-07,
+      "loss": 0.6759,
+      "step": 3329
+    },
+    {
+      "epoch": 6.166666666666667,
+      "grad_norm": 1.6341492547872276,
+      "learning_rate": 1.9202233967277454e-07,
+      "loss": 0.6684,
+      "step": 3330
+    },
+    {
+      "epoch": 6.1685185185185185,
+      "grad_norm": 1.5650332345731641,
+      "learning_rate": 1.9201646999068042e-07,
+      "loss": 0.6342,
+      "step": 3331
+    },
+    {
+      "epoch": 6.17037037037037,
+      "grad_norm": 1.5776098419792035,
+      "learning_rate": 1.9201059823980615e-07,
+      "loss": 0.6507,
+      "step": 3332
+    },
+    {
+      "epoch": 6.1722222222222225,
+      "grad_norm": 1.3912193707984237,
+      "learning_rate": 1.9200472442028375e-07,
+      "loss": 0.6561,
+      "step": 3333
+    },
+    {
+      "epoch": 6.174074074074074,
+      "grad_norm": 1.4712625538489306,
+      "learning_rate": 1.9199884853224525e-07,
+      "loss": 0.6567,
+      "step": 3334
+    },
+    {
+      "epoch": 6.175925925925926,
+      "grad_norm": 1.7647268734713886,
+      "learning_rate": 1.9199297057582278e-07,
+      "loss": 0.6435,
+      "step": 3335
+    },
+    {
+      "epoch": 6.177777777777778,
+      "grad_norm": 2.1072799249397223,
+      "learning_rate": 1.9198709055114848e-07,
+      "loss": 0.6861,
+      "step": 3336
+    },
+    {
+      "epoch": 6.17962962962963,
+      "grad_norm": 1.4598892021068544,
+      "learning_rate": 1.9198120845835453e-07,
+      "loss": 0.6537,
+      "step": 3337
+    },
+    {
+      "epoch": 6.181481481481481,
+      "grad_norm": 1.9817834624125323,
+      "learning_rate": 1.9197532429757324e-07,
+      "loss": 0.6801,
+      "step": 3338
+    },
+    {
+      "epoch": 6.183333333333334,
+      "grad_norm": 1.5218399752139922,
+      "learning_rate": 1.9196943806893683e-07,
+      "loss": 0.6722,
+      "step": 3339
+    },
+    {
+      "epoch": 6.185185185185185,
+      "grad_norm": 1.5618825449371072,
+      "learning_rate": 1.9196354977257766e-07,
+      "loss": 0.6953,
+      "step": 3340
+    },
+    {
+      "epoch": 6.187037037037037,
+      "grad_norm": 1.4268630152187416,
+      "learning_rate": 1.919576594086281e-07,
+      "loss": 0.6756,
+      "step": 3341
+    },
+    {
+      "epoch": 6.188888888888889,
+      "grad_norm": 1.5081163502280814,
+      "learning_rate": 1.9195176697722066e-07,
+      "loss": 0.6499,
+      "step": 3342
+    },
+    {
+      "epoch": 6.190740740740741,
+      "grad_norm": 1.4403414514101327,
+      "learning_rate": 1.9194587247848775e-07,
+      "loss": 0.6882,
+      "step": 3343
+    },
+    {
+      "epoch": 6.192592592592592,
+      "grad_norm": 1.4113805168701803,
+      "learning_rate": 1.9193997591256188e-07,
+      "loss": 0.702,
+      "step": 3344
+    },
+    {
+      "epoch": 6.194444444444445,
+      "grad_norm": 1.537333510024604,
+      "learning_rate": 1.9193407727957566e-07,
+      "loss": 0.6628,
+      "step": 3345
+    },
+    {
+      "epoch": 6.196296296296296,
+      "grad_norm": 1.8040425820769224,
+      "learning_rate": 1.919281765796617e-07,
+      "loss": 0.632,
+      "step": 3346
+    },
+    {
+      "epoch": 6.198148148148148,
+      "grad_norm": 1.442920474064865,
+      "learning_rate": 1.9192227381295263e-07,
+      "loss": 0.6866,
+      "step": 3347
+    },
+    {
+      "epoch": 6.2,
+      "grad_norm": 1.633535299852807,
+      "learning_rate": 1.919163689795812e-07,
+      "loss": 0.6948,
+      "step": 3348
+    },
+    {
+      "epoch": 6.201851851851852,
+      "grad_norm": 1.5353752964391654,
+      "learning_rate": 1.919104620796802e-07,
+      "loss": 0.6664,
+      "step": 3349
+    },
+    {
+      "epoch": 6.203703703703703,
+      "grad_norm": 1.306342887143042,
+      "learning_rate": 1.919045531133823e-07,
+      "loss": 0.6205,
+      "step": 3350
+    },
+    {
+      "epoch": 6.205555555555556,
+      "grad_norm": 1.642092118730288,
+      "learning_rate": 1.9189864208082048e-07,
+      "loss": 0.6901,
+      "step": 3351
+    },
+    {
+      "epoch": 6.207407407407407,
+      "grad_norm": 1.5140846820955263,
+      "learning_rate": 1.9189272898212764e-07,
+      "loss": 0.6551,
+      "step": 3352
+    },
+    {
+      "epoch": 6.209259259259259,
+      "grad_norm": 1.8087419917613925,
+      "learning_rate": 1.9188681381743662e-07,
+      "loss": 0.671,
+      "step": 3353
+    },
+    {
+      "epoch": 6.211111111111111,
+      "grad_norm": 1.5323281068740797,
+      "learning_rate": 1.9188089658688048e-07,
+      "loss": 0.6753,
+      "step": 3354
+    },
+    {
+      "epoch": 6.212962962962963,
+      "grad_norm": 1.4914873329476706,
+      "learning_rate": 1.9187497729059224e-07,
+      "loss": 0.6522,
+      "step": 3355
+    },
+    {
+      "epoch": 6.214814814814815,
+      "grad_norm": 1.5152457738013503,
+      "learning_rate": 1.9186905592870497e-07,
+      "loss": 0.6804,
+      "step": 3356
+    },
+    {
+      "epoch": 6.216666666666667,
+      "grad_norm": 1.6805050649653726,
+      "learning_rate": 1.9186313250135182e-07,
+      "loss": 0.7242,
+      "step": 3357
+    },
+    {
+      "epoch": 6.218518518518518,
+      "grad_norm": 2.016589350831726,
+      "learning_rate": 1.9185720700866597e-07,
+      "loss": 0.6773,
+      "step": 3358
+    },
+    {
+      "epoch": 6.220370370370371,
+      "grad_norm": 1.5867126132779024,
+      "learning_rate": 1.9185127945078058e-07,
+      "loss": 0.629,
+      "step": 3359
+    },
+    {
+      "epoch": 6.222222222222222,
+      "grad_norm": 2.555724963147106,
+      "learning_rate": 1.91845349827829e-07,
+      "loss": 0.6583,
+      "step": 3360
+    },
+    {
+      "epoch": 6.224074074074074,
+      "grad_norm": 1.4505794963276681,
+      "learning_rate": 1.9183941813994454e-07,
+      "loss": 0.6393,
+      "step": 3361
+    },
+    {
+      "epoch": 6.225925925925926,
+      "grad_norm": 5.767031031874181,
+      "learning_rate": 1.9183348438726048e-07,
+      "loss": 0.6557,
+      "step": 3362
+    },
+    {
+      "epoch": 6.227777777777778,
+      "grad_norm": 1.61949811131035,
+      "learning_rate": 1.9182754856991028e-07,
+      "loss": 0.6476,
+      "step": 3363
+    },
+    {
+      "epoch": 6.229629629629629,
+      "grad_norm": 1.410102440854385,
+      "learning_rate": 1.918216106880274e-07,
+      "loss": 0.6442,
+      "step": 3364
+    },
+    {
+      "epoch": 6.231481481481482,
+      "grad_norm": 1.5828107819617578,
+      "learning_rate": 1.9181567074174532e-07,
+      "loss": 0.6822,
+      "step": 3365
+    },
+    {
+      "epoch": 6.233333333333333,
+      "grad_norm": 1.4482426048154045,
+      "learning_rate": 1.9180972873119758e-07,
+      "loss": 0.6839,
+      "step": 3366
+    },
+    {
+      "epoch": 6.235185185185185,
+      "grad_norm": 1.5254398064451258,
+      "learning_rate": 1.9180378465651782e-07,
+      "loss": 0.6432,
+      "step": 3367
+    },
+    {
+      "epoch": 6.237037037037037,
+      "grad_norm": 1.5712406047304206,
+      "learning_rate": 1.9179783851783963e-07,
+      "loss": 0.6447,
+      "step": 3368
+    },
+    {
+      "epoch": 6.238888888888889,
+      "grad_norm": 1.6101021332698338,
+      "learning_rate": 1.917918903152967e-07,
+      "loss": 0.6204,
+      "step": 3369
+    },
+    {
+      "epoch": 6.2407407407407405,
+      "grad_norm": 2.074215443583755,
+      "learning_rate": 1.917859400490228e-07,
+      "loss": 0.6996,
+      "step": 3370
+    },
+    {
+      "epoch": 6.242592592592593,
+      "grad_norm": 1.5712661098611045,
+      "learning_rate": 1.9177998771915164e-07,
+      "loss": 0.6581,
+      "step": 3371
+    },
+    {
+      "epoch": 6.2444444444444445,
+      "grad_norm": 1.6320109272099848,
+      "learning_rate": 1.9177403332581712e-07,
+      "loss": 0.64,
+      "step": 3372
+    },
+    {
+      "epoch": 6.246296296296296,
+      "grad_norm": 1.4119372007155595,
+      "learning_rate": 1.9176807686915307e-07,
+      "loss": 0.7032,
+      "step": 3373
+    },
+    {
+      "epoch": 6.2481481481481485,
+      "grad_norm": 1.5996084972469322,
+      "learning_rate": 1.917621183492934e-07,
+      "loss": 0.6708,
+      "step": 3374
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 1.5482380105102012,
+      "learning_rate": 1.917561577663721e-07,
+      "loss": 0.6754,
+      "step": 3375
+    },
+    {
+      "epoch": 6.2518518518518515,
+      "grad_norm": 1.6278065338317644,
+      "learning_rate": 1.9175019512052315e-07,
+      "loss": 0.6495,
+      "step": 3376
+    },
+    {
+      "epoch": 6.253703703703704,
+      "grad_norm": 1.427443309089482,
+      "learning_rate": 1.9174423041188064e-07,
+      "loss": 0.7212,
+      "step": 3377
+    },
+    {
+      "epoch": 6.2555555555555555,
+      "grad_norm": 1.5541245674329187,
+      "learning_rate": 1.9173826364057867e-07,
+      "loss": 0.6534,
+      "step": 3378
+    },
+    {
+      "epoch": 6.257407407407407,
+      "grad_norm": 1.4798706511108735,
+      "learning_rate": 1.9173229480675136e-07,
+      "loss": 0.6872,
+      "step": 3379
+    },
+    {
+      "epoch": 6.2592592592592595,
+      "grad_norm": 1.5741355778396606,
+      "learning_rate": 1.9172632391053293e-07,
+      "loss": 0.6242,
+      "step": 3380
+    },
+    {
+      "epoch": 6.261111111111111,
+      "grad_norm": 2.1736452190263127,
+      "learning_rate": 1.9172035095205762e-07,
+      "loss": 0.6251,
+      "step": 3381
+    },
+    {
+      "epoch": 6.262962962962963,
+      "grad_norm": 1.5680883386519138,
+      "learning_rate": 1.9171437593145966e-07,
+      "loss": 0.6979,
+      "step": 3382
+    },
+    {
+      "epoch": 6.264814814814815,
+      "grad_norm": 1.7182523419892173,
+      "learning_rate": 1.917083988488735e-07,
+      "loss": 0.672,
+      "step": 3383
+    },
+    {
+      "epoch": 6.266666666666667,
+      "grad_norm": 2.085816581645039,
+      "learning_rate": 1.917024197044334e-07,
+      "loss": 0.6975,
+      "step": 3384
+    },
+    {
+      "epoch": 6.268518518518518,
+      "grad_norm": 1.3934471906208208,
+      "learning_rate": 1.916964384982739e-07,
+      "loss": 0.7055,
+      "step": 3385
+    },
+    {
+      "epoch": 6.270370370370371,
+      "grad_norm": 1.4409719661635803,
+      "learning_rate": 1.9169045523052941e-07,
+      "loss": 0.6744,
+      "step": 3386
+    },
+    {
+      "epoch": 6.272222222222222,
+      "grad_norm": 1.9071123625518116,
+      "learning_rate": 1.9168446990133442e-07,
+      "loss": 0.6614,
+      "step": 3387
+    },
+    {
+      "epoch": 6.274074074074074,
+      "grad_norm": 2.5389267641304727,
+      "learning_rate": 1.9167848251082358e-07,
+      "loss": 0.6764,
+      "step": 3388
+    },
+    {
+      "epoch": 6.275925925925926,
+      "grad_norm": 1.3045398433591127,
+      "learning_rate": 1.9167249305913144e-07,
+      "loss": 0.6419,
+      "step": 3389
+    },
+    {
+      "epoch": 6.277777777777778,
+      "grad_norm": 1.4265062862635696,
+      "learning_rate": 1.9166650154639267e-07,
+      "loss": 0.655,
+      "step": 3390
+    },
+    {
+      "epoch": 6.279629629629629,
+      "grad_norm": 1.4740999658806446,
+      "learning_rate": 1.9166050797274198e-07,
+      "loss": 0.7062,
+      "step": 3391
+    },
+    {
+      "epoch": 6.281481481481482,
+      "grad_norm": 2.99327903327846,
+      "learning_rate": 1.9165451233831413e-07,
+      "loss": 0.6196,
+      "step": 3392
+    },
+    {
+      "epoch": 6.283333333333333,
+      "grad_norm": 1.489313161249133,
+      "learning_rate": 1.9164851464324392e-07,
+      "loss": 0.6732,
+      "step": 3393
+    },
+    {
+      "epoch": 6.285185185185185,
+      "grad_norm": 1.8268014928067233,
+      "learning_rate": 1.9164251488766618e-07,
+      "loss": 0.6937,
+      "step": 3394
+    },
+    {
+      "epoch": 6.287037037037037,
+      "grad_norm": 1.4010652760287288,
+      "learning_rate": 1.916365130717158e-07,
+      "loss": 0.6728,
+      "step": 3395
+    },
+    {
+      "epoch": 6.288888888888889,
+      "grad_norm": 1.4955383789510575,
+      "learning_rate": 1.9163050919552776e-07,
+      "loss": 0.7062,
+      "step": 3396
+    },
+    {
+      "epoch": 6.290740740740741,
+      "grad_norm": 1.8298734868077546,
+      "learning_rate": 1.9162450325923697e-07,
+      "loss": 0.6779,
+      "step": 3397
+    },
+    {
+      "epoch": 6.292592592592593,
+      "grad_norm": 1.3846873769370864,
+      "learning_rate": 1.9161849526297852e-07,
+      "loss": 0.709,
+      "step": 3398
+    },
+    {
+      "epoch": 6.294444444444444,
+      "grad_norm": 1.659133186209603,
+      "learning_rate": 1.9161248520688745e-07,
+      "loss": 0.6388,
+      "step": 3399
+    },
+    {
+      "epoch": 6.296296296296296,
+      "grad_norm": 1.5794134707910168,
+      "learning_rate": 1.916064730910989e-07,
+      "loss": 0.6755,
+      "step": 3400
+    },
+    {
+      "epoch": 6.298148148148148,
+      "grad_norm": 1.417430610217626,
+      "learning_rate": 1.9160045891574806e-07,
+      "loss": 0.655,
+      "step": 3401
+    },
+    {
+      "epoch": 6.3,
+      "grad_norm": 1.669536168966088,
+      "learning_rate": 1.915944426809701e-07,
+      "loss": 0.6259,
+      "step": 3402
+    },
+    {
+      "epoch": 6.301851851851852,
+      "grad_norm": 1.6540145764154959,
+      "learning_rate": 1.9158842438690028e-07,
+      "loss": 0.7047,
+      "step": 3403
+    },
+    {
+      "epoch": 6.303703703703704,
+      "grad_norm": 1.581507051187245,
+      "learning_rate": 1.9158240403367396e-07,
+      "loss": 0.6847,
+      "step": 3404
+    },
+    {
+      "epoch": 6.305555555555555,
+      "grad_norm": 1.8494292893526751,
+      "learning_rate": 1.9157638162142646e-07,
+      "loss": 0.6586,
+      "step": 3405
+    },
+    {
+      "epoch": 6.307407407407408,
+      "grad_norm": 1.6001401114536582,
+      "learning_rate": 1.9157035715029317e-07,
+      "loss": 0.6206,
+      "step": 3406
+    },
+    {
+      "epoch": 6.309259259259259,
+      "grad_norm": 1.9812505126575766,
+      "learning_rate": 1.9156433062040956e-07,
+      "loss": 0.6366,
+      "step": 3407
+    },
+    {
+      "epoch": 6.311111111111111,
+      "grad_norm": 2.277901288785482,
+      "learning_rate": 1.915583020319111e-07,
+      "loss": 0.6157,
+      "step": 3408
+    },
+    {
+      "epoch": 6.312962962962963,
+      "grad_norm": 1.707848761298398,
+      "learning_rate": 1.9155227138493335e-07,
+      "loss": 0.6422,
+      "step": 3409
+    },
+    {
+      "epoch": 6.314814814814815,
+      "grad_norm": 1.6746797176861448,
+      "learning_rate": 1.915462386796119e-07,
+      "loss": 0.6668,
+      "step": 3410
+    },
+    {
+      "epoch": 6.316666666666666,
+      "grad_norm": 1.6696580726154555,
+      "learning_rate": 1.9154020391608236e-07,
+      "loss": 0.672,
+      "step": 3411
+    },
+    {
+      "epoch": 6.318518518518519,
+      "grad_norm": 1.8529384731080532,
+      "learning_rate": 1.9153416709448043e-07,
+      "loss": 0.6763,
+      "step": 3412
+    },
+    {
+      "epoch": 6.32037037037037,
+      "grad_norm": 1.6689760975037833,
+      "learning_rate": 1.915281282149418e-07,
+      "loss": 0.6466,
+      "step": 3413
+    },
+    {
+      "epoch": 6.322222222222222,
+      "grad_norm": 1.426932617951893,
+      "learning_rate": 1.9152208727760227e-07,
+      "loss": 0.6564,
+      "step": 3414
+    },
+    {
+      "epoch": 6.324074074074074,
+      "grad_norm": 2.129197528045345,
+      "learning_rate": 1.9151604428259765e-07,
+      "loss": 0.6852,
+      "step": 3415
+    },
+    {
+      "epoch": 6.325925925925926,
+      "grad_norm": 1.785447640339465,
+      "learning_rate": 1.9150999923006378e-07,
+      "loss": 0.6749,
+      "step": 3416
+    },
+    {
+      "epoch": 6.3277777777777775,
+      "grad_norm": 1.4903525606466723,
+      "learning_rate": 1.9150395212013662e-07,
+      "loss": 0.6073,
+      "step": 3417
+    },
+    {
+      "epoch": 6.32962962962963,
+      "grad_norm": 1.4565758031313258,
+      "learning_rate": 1.9149790295295206e-07,
+      "loss": 0.6496,
+      "step": 3418
+    },
+    {
+      "epoch": 6.3314814814814815,
+      "grad_norm": 1.6642606339460744,
+      "learning_rate": 1.9149185172864616e-07,
+      "loss": 0.6908,
+      "step": 3419
+    },
+    {
+      "epoch": 6.333333333333333,
+      "grad_norm": 1.5488521842397405,
+      "learning_rate": 1.9148579844735495e-07,
+      "loss": 0.6404,
+      "step": 3420
+    },
+    {
+      "epoch": 6.3351851851851855,
+      "grad_norm": 1.497859446088985,
+      "learning_rate": 1.914797431092145e-07,
+      "loss": 0.6602,
+      "step": 3421
+    },
+    {
+      "epoch": 6.337037037037037,
+      "grad_norm": 1.6853275632443632,
+      "learning_rate": 1.9147368571436097e-07,
+      "loss": 0.6597,
+      "step": 3422
+    },
+    {
+      "epoch": 6.338888888888889,
+      "grad_norm": 1.698456252837249,
+      "learning_rate": 1.914676262629306e-07,
+      "loss": 0.5796,
+      "step": 3423
+    },
+    {
+      "epoch": 6.340740740740741,
+      "grad_norm": 1.56287755830112,
+      "learning_rate": 1.9146156475505952e-07,
+      "loss": 0.684,
+      "step": 3424
+    },
+    {
+      "epoch": 6.342592592592593,
+      "grad_norm": 1.5510770876196125,
+      "learning_rate": 1.9145550119088404e-07,
+      "loss": 0.6434,
+      "step": 3425
+    },
+    {
+      "epoch": 6.344444444444444,
+      "grad_norm": 1.3610545902526667,
+      "learning_rate": 1.9144943557054053e-07,
+      "loss": 0.5937,
+      "step": 3426
+    },
+    {
+      "epoch": 6.3462962962962965,
+      "grad_norm": 1.4976729337028059,
+      "learning_rate": 1.9144336789416534e-07,
+      "loss": 0.6803,
+      "step": 3427
+    },
+    {
+      "epoch": 6.348148148148148,
+      "grad_norm": 1.5556266803670558,
+      "learning_rate": 1.9143729816189488e-07,
+      "loss": 0.6959,
+      "step": 3428
+    },
+    {
+      "epoch": 6.35,
+      "grad_norm": 1.833288957501599,
+      "learning_rate": 1.9143122637386564e-07,
+      "loss": 0.6675,
+      "step": 3429
+    },
+    {
+      "epoch": 6.351851851851852,
+      "grad_norm": 1.7097929956779876,
+      "learning_rate": 1.9142515253021404e-07,
+      "loss": 0.6574,
+      "step": 3430
+    },
+    {
+      "epoch": 6.353703703703704,
+      "grad_norm": 1.4826663772887276,
+      "learning_rate": 1.9141907663107677e-07,
+      "loss": 0.6822,
+      "step": 3431
+    },
+    {
+      "epoch": 6.355555555555555,
+      "grad_norm": 1.4284192455158202,
+      "learning_rate": 1.9141299867659034e-07,
+      "loss": 0.6872,
+      "step": 3432
+    },
+    {
+      "epoch": 6.357407407407408,
+      "grad_norm": 1.6063410989501834,
+      "learning_rate": 1.9140691866689144e-07,
+      "loss": 0.6284,
+      "step": 3433
+    },
+    {
+      "epoch": 6.359259259259259,
+      "grad_norm": 1.3710009448688145,
+      "learning_rate": 1.9140083660211675e-07,
+      "loss": 0.6845,
+      "step": 3434
+    },
+    {
+      "epoch": 6.361111111111111,
+      "grad_norm": 1.37125158829942,
+      "learning_rate": 1.91394752482403e-07,
+      "loss": 0.6778,
+      "step": 3435
+    },
+    {
+      "epoch": 6.362962962962963,
+      "grad_norm": 1.7920836689663495,
+      "learning_rate": 1.91388666307887e-07,
+      "loss": 0.6614,
+      "step": 3436
+    },
+    {
+      "epoch": 6.364814814814815,
+      "grad_norm": 1.6169473332441728,
+      "learning_rate": 1.9138257807870555e-07,
+      "loss": 0.6898,
+      "step": 3437
+    },
+    {
+      "epoch": 6.366666666666666,
+      "grad_norm": 1.4511184099836392,
+      "learning_rate": 1.9137648779499558e-07,
+      "loss": 0.6488,
+      "step": 3438
+    },
+    {
+      "epoch": 6.368518518518519,
+      "grad_norm": 3.2756727132891124,
+      "learning_rate": 1.9137039545689397e-07,
+      "loss": 0.6264,
+      "step": 3439
+    },
+    {
+      "epoch": 6.37037037037037,
+      "grad_norm": 1.6320513703683337,
+      "learning_rate": 1.9136430106453774e-07,
+      "loss": 0.6493,
+      "step": 3440
+    },
+    {
+      "epoch": 6.372222222222222,
+      "grad_norm": 1.372712142012325,
+      "learning_rate": 1.9135820461806386e-07,
+      "loss": 0.6551,
+      "step": 3441
+    },
+    {
+      "epoch": 6.374074074074074,
+      "grad_norm": 1.4233008427160678,
+      "learning_rate": 1.913521061176094e-07,
+      "loss": 0.6752,
+      "step": 3442
+    },
+    {
+      "epoch": 6.375925925925926,
+      "grad_norm": 1.60618672177728,
+      "learning_rate": 1.913460055633115e-07,
+      "loss": 0.6168,
+      "step": 3443
+    },
+    {
+      "epoch": 6.377777777777778,
+      "grad_norm": 1.763724534295357,
+      "learning_rate": 1.9133990295530732e-07,
+      "loss": 0.6238,
+      "step": 3444
+    },
+    {
+      "epoch": 6.37962962962963,
+      "grad_norm": 1.4017229732786478,
+      "learning_rate": 1.9133379829373405e-07,
+      "loss": 0.6375,
+      "step": 3445
+    },
+    {
+      "epoch": 6.381481481481481,
+      "grad_norm": 2.3358142122918957,
+      "learning_rate": 1.913276915787289e-07,
+      "loss": 0.6765,
+      "step": 3446
+    },
+    {
+      "epoch": 6.383333333333334,
+      "grad_norm": 1.5248172607926531,
+      "learning_rate": 1.9132158281042923e-07,
+      "loss": 0.6469,
+      "step": 3447
+    },
+    {
+      "epoch": 6.385185185185185,
+      "grad_norm": 1.5012034867503168,
+      "learning_rate": 1.9131547198897233e-07,
+      "loss": 0.6709,
+      "step": 3448
+    },
+    {
+      "epoch": 6.387037037037037,
+      "grad_norm": 2.0285030171956313,
+      "learning_rate": 1.9130935911449563e-07,
+      "loss": 0.6705,
+      "step": 3449
+    },
+    {
+      "epoch": 6.388888888888889,
+      "grad_norm": 1.4929696543172324,
+      "learning_rate": 1.9130324418713654e-07,
+      "loss": 0.5815,
+      "step": 3450
+    },
+    {
+      "epoch": 6.390740740740741,
+      "grad_norm": 1.748207755802049,
+      "learning_rate": 1.9129712720703254e-07,
+      "loss": 0.6531,
+      "step": 3451
+    },
+    {
+      "epoch": 6.392592592592592,
+      "grad_norm": 1.6102653597159946,
+      "learning_rate": 1.9129100817432115e-07,
+      "loss": 0.6631,
+      "step": 3452
+    },
+    {
+      "epoch": 6.394444444444445,
+      "grad_norm": 1.4354879816857333,
+      "learning_rate": 1.9128488708913997e-07,
+      "loss": 0.6617,
+      "step": 3453
+    },
+    {
+      "epoch": 6.396296296296296,
+      "grad_norm": 1.5323256088910597,
+      "learning_rate": 1.912787639516266e-07,
+      "loss": 0.6822,
+      "step": 3454
+    },
+    {
+      "epoch": 6.398148148148148,
+      "grad_norm": 1.975147709669045,
+      "learning_rate": 1.9127263876191871e-07,
+      "loss": 0.6353,
+      "step": 3455
+    },
+    {
+      "epoch": 6.4,
+      "grad_norm": 1.7335395939795983,
+      "learning_rate": 1.91266511520154e-07,
+      "loss": 0.6685,
+      "step": 3456
+    },
+    {
+      "epoch": 6.401851851851852,
+      "grad_norm": 1.470428367473061,
+      "learning_rate": 1.9126038222647024e-07,
+      "loss": 0.7029,
+      "step": 3457
+    },
+    {
+      "epoch": 6.4037037037037035,
+      "grad_norm": 1.6491674131842289,
+      "learning_rate": 1.9125425088100525e-07,
+      "loss": 0.6468,
+      "step": 3458
+    },
+    {
+      "epoch": 6.405555555555556,
+      "grad_norm": 1.7837052191056182,
+      "learning_rate": 1.9124811748389681e-07,
+      "loss": 0.6621,
+      "step": 3459
+    },
+    {
+      "epoch": 6.407407407407407,
+      "grad_norm": 1.4773470602095442,
+      "learning_rate": 1.912419820352829e-07,
+      "loss": 0.6694,
+      "step": 3460
+    },
+    {
+      "epoch": 6.409259259259259,
+      "grad_norm": 2.41615950812344,
+      "learning_rate": 1.9123584453530142e-07,
+      "loss": 0.6541,
+      "step": 3461
+    },
+    {
+      "epoch": 6.411111111111111,
+      "grad_norm": 1.6551929261096554,
+      "learning_rate": 1.9122970498409035e-07,
+      "loss": 0.7091,
+      "step": 3462
+    },
+    {
+      "epoch": 6.412962962962963,
+      "grad_norm": 1.6496518259984632,
+      "learning_rate": 1.9122356338178773e-07,
+      "loss": 0.6238,
+      "step": 3463
+    },
+    {
+      "epoch": 6.4148148148148145,
+      "grad_norm": 2.4245246179191136,
+      "learning_rate": 1.9121741972853168e-07,
+      "loss": 0.64,
+      "step": 3464
+    },
+    {
+      "epoch": 6.416666666666667,
+      "grad_norm": 1.4971063033915384,
+      "learning_rate": 1.9121127402446024e-07,
+      "loss": 0.6356,
+      "step": 3465
+    },
+    {
+      "epoch": 6.4185185185185185,
+      "grad_norm": 1.6169379581882568,
+      "learning_rate": 1.9120512626971168e-07,
+      "loss": 0.6235,
+      "step": 3466
+    },
+    {
+      "epoch": 6.42037037037037,
+      "grad_norm": 1.4174126049171334,
+      "learning_rate": 1.9119897646442412e-07,
+      "loss": 0.6566,
+      "step": 3467
+    },
+    {
+      "epoch": 6.4222222222222225,
+      "grad_norm": 1.6776269916317015,
+      "learning_rate": 1.9119282460873593e-07,
+      "loss": 0.6309,
+      "step": 3468
+    },
+    {
+      "epoch": 6.424074074074074,
+      "grad_norm": 1.3895209628854723,
+      "learning_rate": 1.9118667070278532e-07,
+      "loss": 0.6894,
+      "step": 3469
+    },
+    {
+      "epoch": 6.425925925925926,
+      "grad_norm": 1.69209431763122,
+      "learning_rate": 1.9118051474671074e-07,
+      "loss": 0.6196,
+      "step": 3470
+    },
+    {
+      "epoch": 6.427777777777778,
+      "grad_norm": 1.8103124746368358,
+      "learning_rate": 1.9117435674065052e-07,
+      "loss": 0.6652,
+      "step": 3471
+    },
+    {
+      "epoch": 6.42962962962963,
+      "grad_norm": 1.7201860399694917,
+      "learning_rate": 1.9116819668474316e-07,
+      "loss": 0.6523,
+      "step": 3472
+    },
+    {
+      "epoch": 6.431481481481481,
+      "grad_norm": 1.5894130993481135,
+      "learning_rate": 1.9116203457912709e-07,
+      "loss": 0.6768,
+      "step": 3473
+    },
+    {
+      "epoch": 6.433333333333334,
+      "grad_norm": 1.43929195989301,
+      "learning_rate": 1.9115587042394092e-07,
+      "loss": 0.6382,
+      "step": 3474
+    },
+    {
+      "epoch": 6.435185185185185,
+      "grad_norm": 1.8346544278283612,
+      "learning_rate": 1.9114970421932322e-07,
+      "loss": 0.6961,
+      "step": 3475
+    },
+    {
+      "epoch": 6.437037037037037,
+      "grad_norm": 1.7034550503509098,
+      "learning_rate": 1.9114353596541254e-07,
+      "loss": 0.6615,
+      "step": 3476
+    },
+    {
+      "epoch": 6.438888888888889,
+      "grad_norm": 1.3651589031920115,
+      "learning_rate": 1.911373656623477e-07,
+      "loss": 0.6859,
+      "step": 3477
+    },
+    {
+      "epoch": 6.440740740740741,
+      "grad_norm": 1.6360328518097047,
+      "learning_rate": 1.9113119331026733e-07,
+      "loss": 0.6176,
+      "step": 3478
+    },
+    {
+      "epoch": 6.442592592592592,
+      "grad_norm": 1.5022692946560237,
+      "learning_rate": 1.9112501890931022e-07,
+      "loss": 0.6492,
+      "step": 3479
+    },
+    {
+      "epoch": 6.444444444444445,
+      "grad_norm": 1.6584397649180205,
+      "learning_rate": 1.911188424596152e-07,
+      "loss": 0.7087,
+      "step": 3480
+    },
+    {
+      "epoch": 6.446296296296296,
+      "grad_norm": 3.9558014621885182,
+      "learning_rate": 1.9111266396132116e-07,
+      "loss": 0.6575,
+      "step": 3481
+    },
+    {
+      "epoch": 6.448148148148148,
+      "grad_norm": 1.491705065355962,
+      "learning_rate": 1.9110648341456693e-07,
+      "loss": 0.6141,
+      "step": 3482
+    },
+    {
+      "epoch": 6.45,
+      "grad_norm": 1.8074815983797556,
+      "learning_rate": 1.9110030081949153e-07,
+      "loss": 0.6846,
+      "step": 3483
+    },
+    {
+      "epoch": 6.451851851851852,
+      "grad_norm": 1.568925684188308,
+      "learning_rate": 1.9109411617623391e-07,
+      "loss": 0.6168,
+      "step": 3484
+    },
+    {
+      "epoch": 6.453703703703704,
+      "grad_norm": 1.3328503187688294,
+      "learning_rate": 1.910879294849332e-07,
+      "loss": 0.6754,
+      "step": 3485
+    },
+    {
+      "epoch": 6.455555555555556,
+      "grad_norm": 1.339476004343173,
+      "learning_rate": 1.9108174074572842e-07,
+      "loss": 0.6873,
+      "step": 3486
+    },
+    {
+      "epoch": 6.457407407407407,
+      "grad_norm": 1.6843423899888483,
+      "learning_rate": 1.9107554995875875e-07,
+      "loss": 0.6704,
+      "step": 3487
+    },
+    {
+      "epoch": 6.459259259259259,
+      "grad_norm": 1.4159829128260302,
+      "learning_rate": 1.9106935712416333e-07,
+      "loss": 0.6825,
+      "step": 3488
+    },
+    {
+      "epoch": 6.461111111111111,
+      "grad_norm": 1.4616694142510485,
+      "learning_rate": 1.9106316224208143e-07,
+      "loss": 0.5907,
+      "step": 3489
+    },
+    {
+      "epoch": 6.462962962962963,
+      "grad_norm": 1.4261532937922716,
+      "learning_rate": 1.9105696531265235e-07,
+      "loss": 0.647,
+      "step": 3490
+    },
+    {
+      "epoch": 6.464814814814815,
+      "grad_norm": 1.4342219018111817,
+      "learning_rate": 1.9105076633601533e-07,
+      "loss": 0.6576,
+      "step": 3491
+    },
+    {
+      "epoch": 6.466666666666667,
+      "grad_norm": 1.3907042454793173,
+      "learning_rate": 1.9104456531230985e-07,
+      "loss": 0.6505,
+      "step": 3492
+    },
+    {
+      "epoch": 6.468518518518518,
+      "grad_norm": 1.4112376101651003,
+      "learning_rate": 1.910383622416752e-07,
+      "loss": 0.6942,
+      "step": 3493
+    },
+    {
+      "epoch": 6.470370370370371,
+      "grad_norm": 1.5109287401691613,
+      "learning_rate": 1.9103215712425097e-07,
+      "loss": 0.6151,
+      "step": 3494
+    },
+    {
+      "epoch": 6.472222222222222,
+      "grad_norm": 1.552618521467297,
+      "learning_rate": 1.910259499601766e-07,
+      "loss": 0.617,
+      "step": 3495
+    },
+    {
+      "epoch": 6.474074074074074,
+      "grad_norm": 1.8608266173216272,
+      "learning_rate": 1.9101974074959164e-07,
+      "loss": 0.6969,
+      "step": 3496
+    },
+    {
+      "epoch": 6.475925925925926,
+      "grad_norm": 1.37219659410563,
+      "learning_rate": 1.910135294926357e-07,
+      "loss": 0.6549,
+      "step": 3497
+    },
+    {
+      "epoch": 6.477777777777778,
+      "grad_norm": 1.4335034329481964,
+      "learning_rate": 1.9100731618944844e-07,
+      "loss": 0.6518,
+      "step": 3498
+    },
+    {
+      "epoch": 6.479629629629629,
+      "grad_norm": 1.4625744892098476,
+      "learning_rate": 1.9100110084016952e-07,
+      "loss": 0.6817,
+      "step": 3499
+    },
+    {
+      "epoch": 6.481481481481482,
+      "grad_norm": 1.4956408438940878,
+      "learning_rate": 1.909948834449387e-07,
+      "loss": 0.6449,
+      "step": 3500
+    },
+    {
+      "epoch": 6.483333333333333,
+      "grad_norm": 1.496066186014099,
+      "learning_rate": 1.9098866400389578e-07,
+      "loss": 0.7182,
+      "step": 3501
+    },
+    {
+      "epoch": 6.485185185185185,
+      "grad_norm": 1.9098552202998413,
+      "learning_rate": 1.9098244251718055e-07,
+      "loss": 0.6873,
+      "step": 3502
+    },
+    {
+      "epoch": 6.487037037037037,
+      "grad_norm": 1.6937796441403588,
+      "learning_rate": 1.9097621898493293e-07,
+      "loss": 0.6474,
+      "step": 3503
+    },
+    {
+      "epoch": 6.488888888888889,
+      "grad_norm": 1.5185315334522649,
+      "learning_rate": 1.909699934072928e-07,
+      "loss": 0.6832,
+      "step": 3504
+    },
+    {
+      "epoch": 6.4907407407407405,
+      "grad_norm": 1.5695052033692563,
+      "learning_rate": 1.9096376578440016e-07,
+      "loss": 0.6771,
+      "step": 3505
+    },
+    {
+      "epoch": 6.492592592592593,
+      "grad_norm": 1.3800690951375854,
+      "learning_rate": 1.90957536116395e-07,
+      "loss": 0.6784,
+      "step": 3506
+    },
+    {
+      "epoch": 6.4944444444444445,
+      "grad_norm": 1.544958976020382,
+      "learning_rate": 1.909513044034174e-07,
+      "loss": 0.6474,
+      "step": 3507
+    },
+    {
+      "epoch": 6.496296296296296,
+      "grad_norm": 1.5256162306385321,
+      "learning_rate": 1.9094507064560746e-07,
+      "loss": 0.645,
+      "step": 3508
+    },
+    {
+      "epoch": 6.4981481481481485,
+      "grad_norm": 1.3915880481965817,
+      "learning_rate": 1.9093883484310533e-07,
+      "loss": 0.6614,
+      "step": 3509
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 1.378096839744085,
+      "learning_rate": 1.909325969960512e-07,
+      "loss": 0.6216,
+      "step": 3510
+    },
+    {
+      "epoch": 6.5018518518518515,
+      "grad_norm": 1.6025282410958077,
+      "learning_rate": 1.9092635710458533e-07,
+      "loss": 0.6386,
+      "step": 3511
+    },
+    {
+      "epoch": 6.503703703703704,
+      "grad_norm": 1.5917477300156002,
+      "learning_rate": 1.9092011516884798e-07,
+      "loss": 0.6416,
+      "step": 3512
+    },
+    {
+      "epoch": 6.5055555555555555,
+      "grad_norm": 1.6282878708885695,
+      "learning_rate": 1.9091387118897953e-07,
+      "loss": 0.6758,
+      "step": 3513
+    },
+    {
+      "epoch": 6.507407407407407,
+      "grad_norm": 1.311585953143118,
+      "learning_rate": 1.909076251651203e-07,
+      "loss": 0.6643,
+      "step": 3514
+    },
+    {
+      "epoch": 6.5092592592592595,
+      "grad_norm": 1.5683270155089606,
+      "learning_rate": 1.9090137709741081e-07,
+      "loss": 0.634,
+      "step": 3515
+    },
+    {
+      "epoch": 6.511111111111111,
+      "grad_norm": 1.7821499159645384,
+      "learning_rate": 1.9089512698599145e-07,
+      "loss": 0.6604,
+      "step": 3516
+    },
+    {
+      "epoch": 6.512962962962963,
+      "grad_norm": 1.3659166617388496,
+      "learning_rate": 1.9088887483100277e-07,
+      "loss": 0.646,
+      "step": 3517
+    },
+    {
+      "epoch": 6.514814814814815,
+      "grad_norm": 1.52707141954496,
+      "learning_rate": 1.908826206325853e-07,
+      "loss": 0.621,
+      "step": 3518
+    },
+    {
+      "epoch": 6.516666666666667,
+      "grad_norm": 1.5315816528514754,
+      "learning_rate": 1.9087636439087973e-07,
+      "loss": 0.6574,
+      "step": 3519
+    },
+    {
+      "epoch": 6.518518518518518,
+      "grad_norm": 1.552779362275153,
+      "learning_rate": 1.9087010610602666e-07,
+      "loss": 0.654,
+      "step": 3520
+    },
+    {
+      "epoch": 6.520370370370371,
+      "grad_norm": 1.537784393880499,
+      "learning_rate": 1.908638457781668e-07,
+      "loss": 0.6445,
+      "step": 3521
+    },
+    {
+      "epoch": 6.522222222222222,
+      "grad_norm": 1.5310748957344984,
+      "learning_rate": 1.9085758340744092e-07,
+      "loss": 0.6498,
+      "step": 3522
+    },
+    {
+      "epoch": 6.524074074074074,
+      "grad_norm": 1.4601432378948946,
+      "learning_rate": 1.9085131899398976e-07,
+      "loss": 0.6578,
+      "step": 3523
+    },
+    {
+      "epoch": 6.525925925925926,
+      "grad_norm": 1.4413110373869804,
+      "learning_rate": 1.9084505253795424e-07,
+      "loss": 0.6807,
+      "step": 3524
+    },
+    {
+      "epoch": 6.527777777777778,
+      "grad_norm": 1.6464147492343302,
+      "learning_rate": 1.908387840394752e-07,
+      "loss": 0.6839,
+      "step": 3525
+    },
+    {
+      "epoch": 6.52962962962963,
+      "grad_norm": 1.6304665083380474,
+      "learning_rate": 1.908325134986936e-07,
+      "loss": 0.6613,
+      "step": 3526
+    },
+    {
+      "epoch": 6.531481481481482,
+      "grad_norm": 1.3851430322764795,
+      "learning_rate": 1.9082624091575034e-07,
+      "loss": 0.6326,
+      "step": 3527
+    },
+    {
+      "epoch": 6.533333333333333,
+      "grad_norm": 1.404006762516901,
+      "learning_rate": 1.9081996629078654e-07,
+      "loss": 0.6728,
+      "step": 3528
+    },
+    {
+      "epoch": 6.535185185185185,
+      "grad_norm": 1.4498564116386556,
+      "learning_rate": 1.9081368962394322e-07,
+      "loss": 0.6824,
+      "step": 3529
+    },
+    {
+      "epoch": 6.537037037037037,
+      "grad_norm": 1.3628743587318175,
+      "learning_rate": 1.9080741091536153e-07,
+      "loss": 0.6304,
+      "step": 3530
+    },
+    {
+      "epoch": 6.538888888888889,
+      "grad_norm": 1.3221319601036166,
+      "learning_rate": 1.9080113016518263e-07,
+      "loss": 0.6433,
+      "step": 3531
+    },
+    {
+      "epoch": 6.540740740740741,
+      "grad_norm": 1.5930103250396523,
+      "learning_rate": 1.9079484737354768e-07,
+      "loss": 0.6647,
+      "step": 3532
+    },
+    {
+      "epoch": 6.542592592592593,
+      "grad_norm": 1.4803340619749408,
+      "learning_rate": 1.9078856254059797e-07,
+      "loss": 0.7207,
+      "step": 3533
+    },
+    {
+      "epoch": 6.544444444444444,
+      "grad_norm": 1.583338190977288,
+      "learning_rate": 1.907822756664748e-07,
+      "loss": 0.6541,
+      "step": 3534
+    },
+    {
+      "epoch": 6.546296296296296,
+      "grad_norm": 1.3597349589840217,
+      "learning_rate": 1.9077598675131948e-07,
+      "loss": 0.6915,
+      "step": 3535
+    },
+    {
+      "epoch": 6.548148148148148,
+      "grad_norm": 1.659837047491889,
+      "learning_rate": 1.9076969579527348e-07,
+      "loss": 0.6493,
+      "step": 3536
+    },
+    {
+      "epoch": 6.55,
+      "grad_norm": 2.021272565884714,
+      "learning_rate": 1.907634027984782e-07,
+      "loss": 0.6426,
+      "step": 3537
+    },
+    {
+      "epoch": 6.551851851851852,
+      "grad_norm": 1.7785422519535794,
+      "learning_rate": 1.907571077610751e-07,
+      "loss": 0.6715,
+      "step": 3538
+    },
+    {
+      "epoch": 6.553703703703704,
+      "grad_norm": 1.342462330809328,
+      "learning_rate": 1.9075081068320572e-07,
+      "loss": 0.6402,
+      "step": 3539
+    },
+    {
+      "epoch": 6.555555555555555,
+      "grad_norm": 1.4327846649043527,
+      "learning_rate": 1.907445115650116e-07,
+      "loss": 0.6529,
+      "step": 3540
+    },
+    {
+      "epoch": 6.557407407407408,
+      "grad_norm": 1.4304568368638708,
+      "learning_rate": 1.9073821040663446e-07,
+      "loss": 0.6411,
+      "step": 3541
+    },
+    {
+      "epoch": 6.559259259259259,
+      "grad_norm": 1.603628663431344,
+      "learning_rate": 1.9073190720821592e-07,
+      "loss": 0.6416,
+      "step": 3542
+    },
+    {
+      "epoch": 6.561111111111111,
+      "grad_norm": 1.7769364542069177,
+      "learning_rate": 1.9072560196989764e-07,
+      "loss": 0.6461,
+      "step": 3543
+    },
+    {
+      "epoch": 6.562962962962963,
+      "grad_norm": 1.9872393557421308,
+      "learning_rate": 1.9071929469182142e-07,
+      "loss": 0.7213,
+      "step": 3544
+    },
+    {
+      "epoch": 6.564814814814815,
+      "grad_norm": 1.640454704186781,
+      "learning_rate": 1.9071298537412907e-07,
+      "loss": 0.639,
+      "step": 3545
+    },
+    {
+      "epoch": 6.566666666666666,
+      "grad_norm": 1.4313170831766509,
+      "learning_rate": 1.9070667401696247e-07,
+      "loss": 0.6702,
+      "step": 3546
+    },
+    {
+      "epoch": 6.568518518518519,
+      "grad_norm": 1.4968096501907018,
+      "learning_rate": 1.9070036062046343e-07,
+      "loss": 0.6095,
+      "step": 3547
+    },
+    {
+      "epoch": 6.57037037037037,
+      "grad_norm": 1.5418600560473843,
+      "learning_rate": 1.9069404518477397e-07,
+      "loss": 0.6482,
+      "step": 3548
+    },
+    {
+      "epoch": 6.572222222222222,
+      "grad_norm": 1.557955934344611,
+      "learning_rate": 1.9068772771003604e-07,
+      "loss": 0.6443,
+      "step": 3549
+    },
+    {
+      "epoch": 6.574074074074074,
+      "grad_norm": 2.5980352575194696,
+      "learning_rate": 1.9068140819639172e-07,
+      "loss": 0.6083,
+      "step": 3550
+    },
+    {
+      "epoch": 6.575925925925926,
+      "grad_norm": 1.6872405238305705,
+      "learning_rate": 1.9067508664398303e-07,
+      "loss": 0.6354,
+      "step": 3551
+    },
+    {
+      "epoch": 6.5777777777777775,
+      "grad_norm": 1.5996374547292114,
+      "learning_rate": 1.9066876305295214e-07,
+      "loss": 0.6386,
+      "step": 3552
+    },
+    {
+      "epoch": 6.57962962962963,
+      "grad_norm": 1.9227447752325662,
+      "learning_rate": 1.906624374234412e-07,
+      "loss": 0.6105,
+      "step": 3553
+    },
+    {
+      "epoch": 6.5814814814814815,
+      "grad_norm": 1.381125193874707,
+      "learning_rate": 1.9065610975559241e-07,
+      "loss": 0.6662,
+      "step": 3554
+    },
+    {
+      "epoch": 6.583333333333333,
+      "grad_norm": 1.6763112740820068,
+      "learning_rate": 1.906497800495481e-07,
+      "loss": 0.6725,
+      "step": 3555
+    },
+    {
+      "epoch": 6.5851851851851855,
+      "grad_norm": 1.5304111243140848,
+      "learning_rate": 1.9064344830545046e-07,
+      "loss": 0.6392,
+      "step": 3556
+    },
+    {
+      "epoch": 6.587037037037037,
+      "grad_norm": 1.9180125056119757,
+      "learning_rate": 1.90637114523442e-07,
+      "loss": 0.6653,
+      "step": 3557
+    },
+    {
+      "epoch": 6.588888888888889,
+      "grad_norm": 3.1404883221884057,
+      "learning_rate": 1.90630778703665e-07,
+      "loss": 0.6594,
+      "step": 3558
+    },
+    {
+      "epoch": 6.590740740740741,
+      "grad_norm": 1.5320784304267354,
+      "learning_rate": 1.9062444084626192e-07,
+      "loss": 0.6627,
+      "step": 3559
+    },
+    {
+      "epoch": 6.592592592592593,
+      "grad_norm": 1.698347706238461,
+      "learning_rate": 1.9061810095137532e-07,
+      "loss": 0.6518,
+      "step": 3560
+    },
+    {
+      "epoch": 6.594444444444444,
+      "grad_norm": 1.8503443559882269,
+      "learning_rate": 1.9061175901914767e-07,
+      "loss": 0.6625,
+      "step": 3561
+    },
+    {
+      "epoch": 6.5962962962962965,
+      "grad_norm": 1.502712207905687,
+      "learning_rate": 1.906054150497216e-07,
+      "loss": 0.679,
+      "step": 3562
+    },
+    {
+      "epoch": 6.598148148148148,
+      "grad_norm": 2.2420178811252027,
+      "learning_rate": 1.905990690432397e-07,
+      "loss": 0.6714,
+      "step": 3563
+    },
+    {
+      "epoch": 6.6,
+      "grad_norm": 1.426132603147637,
+      "learning_rate": 1.9059272099984466e-07,
+      "loss": 0.6355,
+      "step": 3564
+    },
+    {
+      "epoch": 6.601851851851852,
+      "grad_norm": 3.2311387759751464,
+      "learning_rate": 1.9058637091967923e-07,
+      "loss": 0.6385,
+      "step": 3565
+    },
+    {
+      "epoch": 6.603703703703704,
+      "grad_norm": 1.464351256965262,
+      "learning_rate": 1.9058001880288614e-07,
+      "loss": 0.6815,
+      "step": 3566
+    },
+    {
+      "epoch": 6.605555555555555,
+      "grad_norm": 1.59727281260324,
+      "learning_rate": 1.9057366464960822e-07,
+      "loss": 0.6669,
+      "step": 3567
+    },
+    {
+      "epoch": 6.607407407407408,
+      "grad_norm": 1.555545966934019,
+      "learning_rate": 1.9056730845998833e-07,
+      "loss": 0.6608,
+      "step": 3568
+    },
+    {
+      "epoch": 6.609259259259259,
+      "grad_norm": 1.3354981240292743,
+      "learning_rate": 1.9056095023416936e-07,
+      "loss": 0.5987,
+      "step": 3569
+    },
+    {
+      "epoch": 6.611111111111111,
+      "grad_norm": 1.3640399136781258,
+      "learning_rate": 1.9055458997229426e-07,
+      "loss": 0.666,
+      "step": 3570
+    },
+    {
+      "epoch": 6.612962962962963,
+      "grad_norm": 1.304527838523876,
+      "learning_rate": 1.9054822767450602e-07,
+      "loss": 0.6297,
+      "step": 3571
+    },
+    {
+      "epoch": 6.614814814814815,
+      "grad_norm": 2.028089013989317,
+      "learning_rate": 1.9054186334094772e-07,
+      "loss": 0.6668,
+      "step": 3572
+    },
+    {
+      "epoch": 6.616666666666667,
+      "grad_norm": 1.412820961063143,
+      "learning_rate": 1.9053549697176244e-07,
+      "loss": 0.6453,
+      "step": 3573
+    },
+    {
+      "epoch": 6.618518518518519,
+      "grad_norm": 1.519384259862545,
+      "learning_rate": 1.9052912856709328e-07,
+      "loss": 0.6152,
+      "step": 3574
+    },
+    {
+      "epoch": 6.62037037037037,
+      "grad_norm": 1.5701999244434224,
+      "learning_rate": 1.9052275812708344e-07,
+      "loss": 0.6794,
+      "step": 3575
+    },
+    {
+      "epoch": 6.622222222222222,
+      "grad_norm": 1.5601762010325246,
+      "learning_rate": 1.9051638565187616e-07,
+      "loss": 0.6816,
+      "step": 3576
+    },
+    {
+      "epoch": 6.624074074074074,
+      "grad_norm": 1.549440481439686,
+      "learning_rate": 1.9051001114161463e-07,
+      "loss": 0.6436,
+      "step": 3577
+    },
+    {
+      "epoch": 6.625925925925926,
+      "grad_norm": 1.434176175710681,
+      "learning_rate": 1.905036345964423e-07,
+      "loss": 0.6433,
+      "step": 3578
+    },
+    {
+      "epoch": 6.627777777777778,
+      "grad_norm": 1.6055836311587846,
+      "learning_rate": 1.9049725601650244e-07,
+      "loss": 0.645,
+      "step": 3579
+    },
+    {
+      "epoch": 6.62962962962963,
+      "grad_norm": 1.5525979025070047,
+      "learning_rate": 1.9049087540193845e-07,
+      "loss": 0.6428,
+      "step": 3580
+    },
+    {
+      "epoch": 6.631481481481481,
+      "grad_norm": 3.4383964362447075,
+      "learning_rate": 1.9048449275289382e-07,
+      "loss": 0.6159,
+      "step": 3581
+    },
+    {
+      "epoch": 6.633333333333333,
+      "grad_norm": 1.593866565104126,
+      "learning_rate": 1.9047810806951206e-07,
+      "loss": 0.687,
+      "step": 3582
+    },
+    {
+      "epoch": 6.635185185185185,
+      "grad_norm": 1.490724425194913,
+      "learning_rate": 1.9047172135193666e-07,
+      "loss": 0.6403,
+      "step": 3583
+    },
+    {
+      "epoch": 6.637037037037037,
+      "grad_norm": 1.5031928217828185,
+      "learning_rate": 1.904653326003113e-07,
+      "loss": 0.6476,
+      "step": 3584
+    },
+    {
+      "epoch": 6.638888888888889,
+      "grad_norm": 1.5238868620861905,
+      "learning_rate": 1.9045894181477952e-07,
+      "loss": 0.6851,
+      "step": 3585
+    },
+    {
+      "epoch": 6.640740740740741,
+      "grad_norm": 1.4788607025563554,
+      "learning_rate": 1.9045254899548507e-07,
+      "loss": 0.6539,
+      "step": 3586
+    },
+    {
+      "epoch": 6.642592592592592,
+      "grad_norm": 1.3643331152600338,
+      "learning_rate": 1.9044615414257165e-07,
+      "loss": 0.6786,
+      "step": 3587
+    },
+    {
+      "epoch": 6.644444444444445,
+      "grad_norm": 1.5213643119843239,
+      "learning_rate": 1.9043975725618302e-07,
+      "loss": 0.6691,
+      "step": 3588
+    },
+    {
+      "epoch": 6.646296296296296,
+      "grad_norm": 1.8462854866239424,
+      "learning_rate": 1.90433358336463e-07,
+      "loss": 0.6895,
+      "step": 3589
+    },
+    {
+      "epoch": 6.648148148148148,
+      "grad_norm": 1.9198249255725623,
+      "learning_rate": 1.9042695738355552e-07,
+      "loss": 0.7524,
+      "step": 3590
+    },
+    {
+      "epoch": 6.65,
+      "grad_norm": 1.6441331792564244,
+      "learning_rate": 1.9042055439760442e-07,
+      "loss": 0.687,
+      "step": 3591
+    },
+    {
+      "epoch": 6.651851851851852,
+      "grad_norm": 2.226794634160272,
+      "learning_rate": 1.9041414937875368e-07,
+      "loss": 0.6534,
+      "step": 3592
+    },
+    {
+      "epoch": 6.6537037037037035,
+      "grad_norm": 1.5938911055605058,
+      "learning_rate": 1.904077423271473e-07,
+      "loss": 0.6493,
+      "step": 3593
+    },
+    {
+      "epoch": 6.655555555555556,
+      "grad_norm": 1.5257997175416245,
+      "learning_rate": 1.9040133324292932e-07,
+      "loss": 0.6844,
+      "step": 3594
+    },
+    {
+      "epoch": 6.657407407407407,
+      "grad_norm": 1.4899952780270258,
+      "learning_rate": 1.9039492212624387e-07,
+      "loss": 0.6552,
+      "step": 3595
+    },
+    {
+      "epoch": 6.659259259259259,
+      "grad_norm": 1.65106697519256,
+      "learning_rate": 1.9038850897723503e-07,
+      "loss": 0.6443,
+      "step": 3596
+    },
+    {
+      "epoch": 6.661111111111111,
+      "grad_norm": 1.460551136135121,
+      "learning_rate": 1.9038209379604703e-07,
+      "loss": 0.6533,
+      "step": 3597
+    },
+    {
+      "epoch": 6.662962962962963,
+      "grad_norm": 1.404160694039314,
+      "learning_rate": 1.9037567658282407e-07,
+      "loss": 0.682,
+      "step": 3598
+    },
+    {
+      "epoch": 6.6648148148148145,
+      "grad_norm": 1.5239984959272146,
+      "learning_rate": 1.9036925733771048e-07,
+      "loss": 0.6504,
+      "step": 3599
+    },
+    {
+      "epoch": 6.666666666666667,
+      "grad_norm": 1.5171195025470456,
+      "learning_rate": 1.9036283606085054e-07,
+      "loss": 0.6393,
+      "step": 3600
+    },
+    {
+      "epoch": 6.6685185185185185,
+      "grad_norm": 1.590037615031641,
+      "learning_rate": 1.9035641275238858e-07,
+      "loss": 0.6761,
+      "step": 3601
+    },
+    {
+      "epoch": 6.67037037037037,
+      "grad_norm": 1.4477573761909848,
+      "learning_rate": 1.903499874124691e-07,
+      "loss": 0.6504,
+      "step": 3602
+    },
+    {
+      "epoch": 6.6722222222222225,
+      "grad_norm": 1.3808780730625259,
+      "learning_rate": 1.903435600412365e-07,
+      "loss": 0.6613,
+      "step": 3603
+    },
+    {
+      "epoch": 6.674074074074074,
+      "grad_norm": 1.7024238755083634,
+      "learning_rate": 1.903371306388353e-07,
+      "loss": 0.6665,
+      "step": 3604
+    },
+    {
+      "epoch": 6.675925925925926,
+      "grad_norm": 1.6534460938459639,
+      "learning_rate": 1.9033069920541007e-07,
+      "loss": 0.6325,
+      "step": 3605
+    },
+    {
+      "epoch": 6.677777777777778,
+      "grad_norm": 1.7680368366360395,
+      "learning_rate": 1.9032426574110535e-07,
+      "loss": 0.6625,
+      "step": 3606
+    },
+    {
+      "epoch": 6.67962962962963,
+      "grad_norm": 1.6637677084100126,
+      "learning_rate": 1.9031783024606587e-07,
+      "loss": 0.6501,
+      "step": 3607
+    },
+    {
+      "epoch": 6.681481481481481,
+      "grad_norm": 1.6660782812856678,
+      "learning_rate": 1.9031139272043623e-07,
+      "loss": 0.68,
+      "step": 3608
+    },
+    {
+      "epoch": 6.683333333333334,
+      "grad_norm": 1.67232389224004,
+      "learning_rate": 1.9030495316436118e-07,
+      "loss": 0.6674,
+      "step": 3609
+    },
+    {
+      "epoch": 6.685185185185185,
+      "grad_norm": 1.708669316071796,
+      "learning_rate": 1.9029851157798556e-07,
+      "loss": 0.64,
+      "step": 3610
+    },
+    {
+      "epoch": 6.687037037037037,
+      "grad_norm": 1.5381193120897307,
+      "learning_rate": 1.9029206796145413e-07,
+      "loss": 0.6742,
+      "step": 3611
+    },
+    {
+      "epoch": 6.688888888888889,
+      "grad_norm": 1.557536196178727,
+      "learning_rate": 1.902856223149118e-07,
+      "loss": 0.6127,
+      "step": 3612
+    },
+    {
+      "epoch": 6.690740740740741,
+      "grad_norm": 1.5339706282702676,
+      "learning_rate": 1.9027917463850347e-07,
+      "loss": 0.6371,
+      "step": 3613
+    },
+    {
+      "epoch": 6.692592592592592,
+      "grad_norm": 1.5591672007068569,
+      "learning_rate": 1.902727249323741e-07,
+      "loss": 0.63,
+      "step": 3614
+    },
+    {
+      "epoch": 6.694444444444445,
+      "grad_norm": 1.451981717697895,
+      "learning_rate": 1.902662731966687e-07,
+      "loss": 0.6193,
+      "step": 3615
+    },
+    {
+      "epoch": 6.696296296296296,
+      "grad_norm": 1.5182456827043982,
+      "learning_rate": 1.9025981943153228e-07,
+      "loss": 0.6551,
+      "step": 3616
+    },
+    {
+      "epoch": 6.698148148148148,
+      "grad_norm": 1.8914097147039861,
+      "learning_rate": 1.9025336363711002e-07,
+      "loss": 0.6821,
+      "step": 3617
+    },
+    {
+      "epoch": 6.7,
+      "grad_norm": 1.8119385081592834,
+      "learning_rate": 1.9024690581354698e-07,
+      "loss": 0.642,
+      "step": 3618
+    },
+    {
+      "epoch": 6.701851851851852,
+      "grad_norm": 1.564284701728419,
+      "learning_rate": 1.902404459609884e-07,
+      "loss": 0.6594,
+      "step": 3619
+    },
+    {
+      "epoch": 6.703703703703704,
+      "grad_norm": 1.6367265966760483,
+      "learning_rate": 1.9023398407957957e-07,
+      "loss": 0.6896,
+      "step": 3620
+    },
+    {
+      "epoch": 6.705555555555556,
+      "grad_norm": 1.9223921672916777,
+      "learning_rate": 1.9022752016946565e-07,
+      "loss": 0.6774,
+      "step": 3621
+    },
+    {
+      "epoch": 6.707407407407407,
+      "grad_norm": 1.6486392685417746,
+      "learning_rate": 1.90221054230792e-07,
+      "loss": 0.6648,
+      "step": 3622
+    },
+    {
+      "epoch": 6.709259259259259,
+      "grad_norm": 1.659731780565725,
+      "learning_rate": 1.9021458626370403e-07,
+      "loss": 0.6911,
+      "step": 3623
+    },
+    {
+      "epoch": 6.711111111111111,
+      "grad_norm": 1.4729867084364658,
+      "learning_rate": 1.9020811626834719e-07,
+      "loss": 0.7183,
+      "step": 3624
+    },
+    {
+      "epoch": 6.712962962962963,
+      "grad_norm": 1.6601496240955846,
+      "learning_rate": 1.9020164424486685e-07,
+      "loss": 0.6424,
+      "step": 3625
+    },
+    {
+      "epoch": 6.714814814814815,
+      "grad_norm": 1.565639553086471,
+      "learning_rate": 1.9019517019340856e-07,
+      "loss": 0.6663,
+      "step": 3626
+    },
+    {
+      "epoch": 6.716666666666667,
+      "grad_norm": 1.620155777521443,
+      "learning_rate": 1.9018869411411788e-07,
+      "loss": 0.6577,
+      "step": 3627
+    },
+    {
+      "epoch": 6.718518518518518,
+      "grad_norm": 1.5013395508326475,
+      "learning_rate": 1.9018221600714043e-07,
+      "loss": 0.6677,
+      "step": 3628
+    },
+    {
+      "epoch": 6.72037037037037,
+      "grad_norm": 1.4475065971540078,
+      "learning_rate": 1.901757358726218e-07,
+      "loss": 0.6266,
+      "step": 3629
+    },
+    {
+      "epoch": 6.722222222222222,
+      "grad_norm": 1.5601766751913895,
+      "learning_rate": 1.9016925371070776e-07,
+      "loss": 0.6363,
+      "step": 3630
+    },
+    {
+      "epoch": 6.724074074074074,
+      "grad_norm": 1.5927068616657944,
+      "learning_rate": 1.9016276952154398e-07,
+      "loss": 0.6368,
+      "step": 3631
+    },
+    {
+      "epoch": 6.725925925925926,
+      "grad_norm": 1.5776497160193574,
+      "learning_rate": 1.9015628330527625e-07,
+      "loss": 0.6527,
+      "step": 3632
+    },
+    {
+      "epoch": 6.727777777777778,
+      "grad_norm": 1.6473345836128068,
+      "learning_rate": 1.901497950620504e-07,
+      "loss": 0.6347,
+      "step": 3633
+    },
+    {
+      "epoch": 6.729629629629629,
+      "grad_norm": 1.570245606739739,
+      "learning_rate": 1.9014330479201234e-07,
+      "loss": 0.6591,
+      "step": 3634
+    },
+    {
+      "epoch": 6.731481481481482,
+      "grad_norm": 1.45709132223943,
+      "learning_rate": 1.9013681249530795e-07,
+      "loss": 0.6541,
+      "step": 3635
+    },
+    {
+      "epoch": 6.733333333333333,
+      "grad_norm": 1.4581327408399674,
+      "learning_rate": 1.9013031817208322e-07,
+      "loss": 0.6507,
+      "step": 3636
+    },
+    {
+      "epoch": 6.735185185185185,
+      "grad_norm": 1.8865447386482699,
+      "learning_rate": 1.9012382182248413e-07,
+      "loss": 0.61,
+      "step": 3637
+    },
+    {
+      "epoch": 6.737037037037037,
+      "grad_norm": 1.5314080752434895,
+      "learning_rate": 1.901173234466568e-07,
+      "loss": 0.6661,
+      "step": 3638
+    },
+    {
+      "epoch": 6.738888888888889,
+      "grad_norm": 1.696917313911211,
+      "learning_rate": 1.901108230447472e-07,
+      "loss": 0.6447,
+      "step": 3639
+    },
+    {
+      "epoch": 6.7407407407407405,
+      "grad_norm": 1.559217656282439,
+      "learning_rate": 1.9010432061690163e-07,
+      "loss": 0.6322,
+      "step": 3640
+    },
+    {
+      "epoch": 6.742592592592593,
+      "grad_norm": 1.631279597257817,
+      "learning_rate": 1.9009781616326616e-07,
+      "loss": 0.6121,
+      "step": 3641
+    },
+    {
+      "epoch": 6.7444444444444445,
+      "grad_norm": 1.6531603742333165,
+      "learning_rate": 1.900913096839871e-07,
+      "loss": 0.6272,
+      "step": 3642
+    },
+    {
+      "epoch": 6.746296296296296,
+      "grad_norm": 1.5893929247462644,
+      "learning_rate": 1.9008480117921075e-07,
+      "loss": 0.6904,
+      "step": 3643
+    },
+    {
+      "epoch": 6.7481481481481485,
+      "grad_norm": 1.6028986652736323,
+      "learning_rate": 1.900782906490834e-07,
+      "loss": 0.6486,
+      "step": 3644
+    },
+    {
+      "epoch": 6.75,
+      "grad_norm": 1.5784819772103595,
+      "learning_rate": 1.9007177809375138e-07,
+      "loss": 0.6534,
+      "step": 3645
+    },
+    {
+      "epoch": 6.7518518518518515,
+      "grad_norm": 1.4843783997718978,
+      "learning_rate": 1.9006526351336118e-07,
+      "loss": 0.6357,
+      "step": 3646
+    },
+    {
+      "epoch": 6.753703703703704,
+      "grad_norm": 1.4347707211685412,
+      "learning_rate": 1.9005874690805926e-07,
+      "loss": 0.6379,
+      "step": 3647
+    },
+    {
+      "epoch": 6.7555555555555555,
+      "grad_norm": 1.8082011737370267,
+      "learning_rate": 1.900522282779921e-07,
+      "loss": 0.6775,
+      "step": 3648
+    },
+    {
+      "epoch": 6.757407407407407,
+      "grad_norm": 1.4323001375174702,
+      "learning_rate": 1.9004570762330627e-07,
+      "loss": 0.6421,
+      "step": 3649
+    },
+    {
+      "epoch": 6.7592592592592595,
+      "grad_norm": 1.4598483569102236,
+      "learning_rate": 1.9003918494414838e-07,
+      "loss": 0.6975,
+      "step": 3650
+    },
+    {
+      "epoch": 6.761111111111111,
+      "grad_norm": 1.5324125968973363,
+      "learning_rate": 1.9003266024066504e-07,
+      "loss": 0.7096,
+      "step": 3651
+    },
+    {
+      "epoch": 6.762962962962963,
+      "grad_norm": 1.4122644275958263,
+      "learning_rate": 1.9002613351300298e-07,
+      "loss": 0.7087,
+      "step": 3652
+    },
+    {
+      "epoch": 6.764814814814815,
+      "grad_norm": 1.6553660384444624,
+      "learning_rate": 1.9001960476130895e-07,
+      "loss": 0.6454,
+      "step": 3653
+    },
+    {
+      "epoch": 6.766666666666667,
+      "grad_norm": 1.6761040730676293,
+      "learning_rate": 1.9001307398572972e-07,
+      "loss": 0.6815,
+      "step": 3654
+    },
+    {
+      "epoch": 6.768518518518518,
+      "grad_norm": 1.5579402446549757,
+      "learning_rate": 1.9000654118641212e-07,
+      "loss": 0.6891,
+      "step": 3655
+    },
+    {
+      "epoch": 6.770370370370371,
+      "grad_norm": 1.5285465413775523,
+      "learning_rate": 1.90000006363503e-07,
+      "loss": 0.6922,
+      "step": 3656
+    },
+    {
+      "epoch": 6.772222222222222,
+      "grad_norm": 1.558157121276329,
+      "learning_rate": 1.899934695171493e-07,
+      "loss": 0.6595,
+      "step": 3657
+    },
+    {
+      "epoch": 6.774074074074074,
+      "grad_norm": 1.468871471995222,
+      "learning_rate": 1.8998693064749798e-07,
+      "loss": 0.6886,
+      "step": 3658
+    },
+    {
+      "epoch": 6.775925925925926,
+      "grad_norm": 1.5910103800078768,
+      "learning_rate": 1.8998038975469608e-07,
+      "loss": 0.7032,
+      "step": 3659
+    },
+    {
+      "epoch": 6.777777777777778,
+      "grad_norm": 1.7030079963166107,
+      "learning_rate": 1.8997384683889063e-07,
+      "loss": 0.6726,
+      "step": 3660
+    },
+    {
+      "epoch": 6.77962962962963,
+      "grad_norm": 1.7253881426431534,
+      "learning_rate": 1.899673019002287e-07,
+      "loss": 0.6056,
+      "step": 3661
+    },
+    {
+      "epoch": 6.781481481481482,
+      "grad_norm": 1.5759875449971485,
+      "learning_rate": 1.8996075493885754e-07,
+      "loss": 0.663,
+      "step": 3662
+    },
+    {
+      "epoch": 6.783333333333333,
+      "grad_norm": 1.4675711051542797,
+      "learning_rate": 1.8995420595492426e-07,
+      "loss": 0.6151,
+      "step": 3663
+    },
+    {
+      "epoch": 6.785185185185185,
+      "grad_norm": 1.7946600064410578,
+      "learning_rate": 1.899476549485761e-07,
+      "loss": 0.6387,
+      "step": 3664
+    },
+    {
+      "epoch": 6.787037037037037,
+      "grad_norm": 1.5098823524452794,
+      "learning_rate": 1.8994110191996038e-07,
+      "loss": 0.6712,
+      "step": 3665
+    },
+    {
+      "epoch": 6.788888888888889,
+      "grad_norm": 1.7301205099205819,
+      "learning_rate": 1.899345468692244e-07,
+      "loss": 0.6464,
+      "step": 3666
+    },
+    {
+      "epoch": 6.790740740740741,
+      "grad_norm": 1.8317356681476744,
+      "learning_rate": 1.8992798979651555e-07,
+      "loss": 0.6296,
+      "step": 3667
+    },
+    {
+      "epoch": 6.792592592592593,
+      "grad_norm": 1.6819194696226583,
+      "learning_rate": 1.8992143070198122e-07,
+      "loss": 0.6691,
+      "step": 3668
+    },
+    {
+      "epoch": 6.794444444444444,
+      "grad_norm": 2.4692521962781586,
+      "learning_rate": 1.8991486958576893e-07,
+      "loss": 0.6427,
+      "step": 3669
+    },
+    {
+      "epoch": 6.796296296296296,
+      "grad_norm": 1.576866072527621,
+      "learning_rate": 1.899083064480262e-07,
+      "loss": 0.6642,
+      "step": 3670
+    },
+    {
+      "epoch": 6.798148148148148,
+      "grad_norm": 1.3835151645033728,
+      "learning_rate": 1.8990174128890053e-07,
+      "loss": 0.6744,
+      "step": 3671
+    },
+    {
+      "epoch": 6.8,
+      "grad_norm": 1.7439390373733747,
+      "learning_rate": 1.8989517410853953e-07,
+      "loss": 0.6354,
+      "step": 3672
+    },
+    {
+      "epoch": 6.801851851851852,
+      "grad_norm": 1.6657889949287623,
+      "learning_rate": 1.8988860490709087e-07,
+      "loss": 0.678,
+      "step": 3673
+    },
+    {
+      "epoch": 6.803703703703704,
+      "grad_norm": 1.689354563387892,
+      "learning_rate": 1.8988203368470223e-07,
+      "loss": 0.6813,
+      "step": 3674
+    },
+    {
+      "epoch": 6.805555555555555,
+      "grad_norm": 1.4532169248299824,
+      "learning_rate": 1.8987546044152138e-07,
+      "loss": 0.6091,
+      "step": 3675
+    },
+    {
+      "epoch": 6.807407407407408,
+      "grad_norm": 1.4390991828113617,
+      "learning_rate": 1.8986888517769607e-07,
+      "loss": 0.6739,
+      "step": 3676
+    },
+    {
+      "epoch": 6.809259259259259,
+      "grad_norm": 3.563430021277078,
+      "learning_rate": 1.8986230789337413e-07,
+      "loss": 0.6605,
+      "step": 3677
+    },
+    {
+      "epoch": 6.811111111111111,
+      "grad_norm": 1.877626494712757,
+      "learning_rate": 1.8985572858870346e-07,
+      "loss": 0.6101,
+      "step": 3678
+    },
+    {
+      "epoch": 6.812962962962963,
+      "grad_norm": 2.4495109425371644,
+      "learning_rate": 1.89849147263832e-07,
+      "loss": 0.6777,
+      "step": 3679
+    },
+    {
+      "epoch": 6.814814814814815,
+      "grad_norm": 1.6373458280822453,
+      "learning_rate": 1.8984256391890763e-07,
+      "loss": 0.6637,
+      "step": 3680
+    },
+    {
+      "epoch": 6.816666666666666,
+      "grad_norm": 2.0339961095190215,
+      "learning_rate": 1.8983597855407843e-07,
+      "loss": 0.693,
+      "step": 3681
+    },
+    {
+      "epoch": 6.818518518518519,
+      "grad_norm": 1.3967124979792362,
+      "learning_rate": 1.8982939116949247e-07,
+      "loss": 0.6882,
+      "step": 3682
+    },
+    {
+      "epoch": 6.82037037037037,
+      "grad_norm": 1.4535407344970497,
+      "learning_rate": 1.8982280176529777e-07,
+      "loss": 0.7006,
+      "step": 3683
+    },
+    {
+      "epoch": 6.822222222222222,
+      "grad_norm": 1.7322387700173034,
+      "learning_rate": 1.8981621034164256e-07,
+      "loss": 0.6533,
+      "step": 3684
+    },
+    {
+      "epoch": 6.824074074074074,
+      "grad_norm": 1.8912976923510763,
+      "learning_rate": 1.89809616898675e-07,
+      "loss": 0.6459,
+      "step": 3685
+    },
+    {
+      "epoch": 6.825925925925926,
+      "grad_norm": 1.5542267896306647,
+      "learning_rate": 1.8980302143654333e-07,
+      "loss": 0.6561,
+      "step": 3686
+    },
+    {
+      "epoch": 6.8277777777777775,
+      "grad_norm": 1.6631186467969499,
+      "learning_rate": 1.8979642395539584e-07,
+      "loss": 0.6461,
+      "step": 3687
+    },
+    {
+      "epoch": 6.82962962962963,
+      "grad_norm": 1.4249597361328639,
+      "learning_rate": 1.8978982445538088e-07,
+      "loss": 0.671,
+      "step": 3688
+    },
+    {
+      "epoch": 6.8314814814814815,
+      "grad_norm": 1.4921100682037167,
+      "learning_rate": 1.8978322293664677e-07,
+      "loss": 0.674,
+      "step": 3689
+    },
+    {
+      "epoch": 6.833333333333333,
+      "grad_norm": 1.5023788231474966,
+      "learning_rate": 1.8977661939934198e-07,
+      "loss": 0.6674,
+      "step": 3690
+    },
+    {
+      "epoch": 6.8351851851851855,
+      "grad_norm": 1.9942026792163534,
+      "learning_rate": 1.8977001384361496e-07,
+      "loss": 0.6784,
+      "step": 3691
+    },
+    {
+      "epoch": 6.837037037037037,
+      "grad_norm": 1.497253795214737,
+      "learning_rate": 1.8976340626961418e-07,
+      "loss": 0.6552,
+      "step": 3692
+    },
+    {
+      "epoch": 6.838888888888889,
+      "grad_norm": 1.6422850609691322,
+      "learning_rate": 1.8975679667748826e-07,
+      "loss": 0.6449,
+      "step": 3693
+    },
+    {
+      "epoch": 6.840740740740741,
+      "grad_norm": 1.3947561721057105,
+      "learning_rate": 1.8975018506738576e-07,
+      "loss": 0.6595,
+      "step": 3694
+    },
+    {
+      "epoch": 6.842592592592593,
+      "grad_norm": 2.75657147715647,
+      "learning_rate": 1.8974357143945537e-07,
+      "loss": 0.696,
+      "step": 3695
+    },
+    {
+      "epoch": 6.844444444444444,
+      "grad_norm": 1.857394533609168,
+      "learning_rate": 1.8973695579384573e-07,
+      "loss": 0.6461,
+      "step": 3696
+    },
+    {
+      "epoch": 6.8462962962962965,
+      "grad_norm": 1.997427286217202,
+      "learning_rate": 1.897303381307056e-07,
+      "loss": 0.6719,
+      "step": 3697
+    },
+    {
+      "epoch": 6.848148148148148,
+      "grad_norm": 1.416733749823453,
+      "learning_rate": 1.8972371845018377e-07,
+      "loss": 0.6154,
+      "step": 3698
+    },
+    {
+      "epoch": 6.85,
+      "grad_norm": 1.5585819483769643,
+      "learning_rate": 1.897170967524291e-07,
+      "loss": 0.679,
+      "step": 3699
+    },
+    {
+      "epoch": 6.851851851851852,
+      "grad_norm": 1.5393001317437378,
+      "learning_rate": 1.897104730375904e-07,
+      "loss": 0.6614,
+      "step": 3700
+    },
+    {
+      "epoch": 6.853703703703704,
+      "grad_norm": 1.661770609922812,
+      "learning_rate": 1.897038473058166e-07,
+      "loss": 0.6709,
+      "step": 3701
+    },
+    {
+      "epoch": 6.855555555555555,
+      "grad_norm": 1.5975764750358483,
+      "learning_rate": 1.8969721955725668e-07,
+      "loss": 0.6373,
+      "step": 3702
+    },
+    {
+      "epoch": 6.857407407407408,
+      "grad_norm": 1.3259836817862736,
+      "learning_rate": 1.8969058979205968e-07,
+      "loss": 0.6434,
+      "step": 3703
+    },
+    {
+      "epoch": 6.859259259259259,
+      "grad_norm": 1.5721941674254143,
+      "learning_rate": 1.896839580103746e-07,
+      "loss": 0.7058,
+      "step": 3704
+    },
+    {
+      "epoch": 6.861111111111111,
+      "grad_norm": 1.6728291920538325,
+      "learning_rate": 1.8967732421235057e-07,
+      "loss": 0.6341,
+      "step": 3705
+    },
+    {
+      "epoch": 6.862962962962963,
+      "grad_norm": 1.5774156773087555,
+      "learning_rate": 1.8967068839813673e-07,
+      "loss": 0.6804,
+      "step": 3706
+    },
+    {
+      "epoch": 6.864814814814815,
+      "grad_norm": 1.965298544747606,
+      "learning_rate": 1.8966405056788227e-07,
+      "loss": 0.6441,
+      "step": 3707
+    },
+    {
+      "epoch": 6.866666666666667,
+      "grad_norm": 1.7513797049797886,
+      "learning_rate": 1.8965741072173644e-07,
+      "loss": 0.6133,
+      "step": 3708
+    },
+    {
+      "epoch": 6.868518518518519,
+      "grad_norm": 1.5579099413716568,
+      "learning_rate": 1.8965076885984854e-07,
+      "loss": 0.6609,
+      "step": 3709
+    },
+    {
+      "epoch": 6.87037037037037,
+      "grad_norm": 1.6450827434769868,
+      "learning_rate": 1.8964412498236785e-07,
+      "loss": 0.6378,
+      "step": 3710
+    },
+    {
+      "epoch": 6.872222222222222,
+      "grad_norm": 1.3831758194677863,
+      "learning_rate": 1.8963747908944372e-07,
+      "loss": 0.6889,
+      "step": 3711
+    },
+    {
+      "epoch": 6.874074074074074,
+      "grad_norm": 1.5677892936721385,
+      "learning_rate": 1.8963083118122562e-07,
+      "loss": 0.6804,
+      "step": 3712
+    },
+    {
+      "epoch": 6.875925925925926,
+      "grad_norm": 1.6451382518344124,
+      "learning_rate": 1.8962418125786303e-07,
+      "loss": 0.6897,
+      "step": 3713
+    },
+    {
+      "epoch": 6.877777777777778,
+      "grad_norm": 1.465058575361201,
+      "learning_rate": 1.8961752931950544e-07,
+      "loss": 0.6288,
+      "step": 3714
+    },
+    {
+      "epoch": 6.87962962962963,
+      "grad_norm": 1.8808036509802426,
+      "learning_rate": 1.8961087536630235e-07,
+      "loss": 0.6385,
+      "step": 3715
+    },
+    {
+      "epoch": 6.881481481481481,
+      "grad_norm": 1.6016158267785385,
+      "learning_rate": 1.8960421939840344e-07,
+      "loss": 0.6166,
+      "step": 3716
+    },
+    {
+      "epoch": 6.883333333333333,
+      "grad_norm": 1.6260360623218741,
+      "learning_rate": 1.895975614159583e-07,
+      "loss": 0.6185,
+      "step": 3717
+    },
+    {
+      "epoch": 6.885185185185185,
+      "grad_norm": 1.5297775695269318,
+      "learning_rate": 1.8959090141911665e-07,
+      "loss": 0.6401,
+      "step": 3718
+    },
+    {
+      "epoch": 6.887037037037037,
+      "grad_norm": 1.7343884149624085,
+      "learning_rate": 1.8958423940802816e-07,
+      "loss": 0.6745,
+      "step": 3719
+    },
+    {
+      "epoch": 6.888888888888889,
+      "grad_norm": 1.3219886578391538,
+      "learning_rate": 1.8957757538284272e-07,
+      "loss": 0.6355,
+      "step": 3720
+    },
+    {
+      "epoch": 6.890740740740741,
+      "grad_norm": 1.4618830623508936,
+      "learning_rate": 1.8957090934371007e-07,
+      "loss": 0.6661,
+      "step": 3721
+    },
+    {
+      "epoch": 6.892592592592592,
+      "grad_norm": 1.8065799642056688,
+      "learning_rate": 1.8956424129078013e-07,
+      "loss": 0.6567,
+      "step": 3722
+    },
+    {
+      "epoch": 6.894444444444445,
+      "grad_norm": 1.4928704055693034,
+      "learning_rate": 1.8955757122420275e-07,
+      "loss": 0.6622,
+      "step": 3723
+    },
+    {
+      "epoch": 6.896296296296296,
+      "grad_norm": 1.8252254396519585,
+      "learning_rate": 1.8955089914412796e-07,
+      "loss": 0.6348,
+      "step": 3724
+    },
+    {
+      "epoch": 6.898148148148148,
+      "grad_norm": 1.7925279281791,
+      "learning_rate": 1.8954422505070574e-07,
+      "loss": 0.6472,
+      "step": 3725
+    },
+    {
+      "epoch": 6.9,
+      "grad_norm": 1.6770141151811553,
+      "learning_rate": 1.8953754894408616e-07,
+      "loss": 0.6658,
+      "step": 3726
+    },
+    {
+      "epoch": 6.901851851851852,
+      "grad_norm": 1.5584695713974657,
+      "learning_rate": 1.8953087082441924e-07,
+      "loss": 0.673,
+      "step": 3727
+    },
+    {
+      "epoch": 6.9037037037037035,
+      "grad_norm": 1.3328363643814065,
+      "learning_rate": 1.8952419069185527e-07,
+      "loss": 0.6337,
+      "step": 3728
+    },
+    {
+      "epoch": 6.905555555555556,
+      "grad_norm": 1.7023790785545256,
+      "learning_rate": 1.8951750854654428e-07,
+      "loss": 0.657,
+      "step": 3729
+    },
+    {
+      "epoch": 6.907407407407407,
+      "grad_norm": 1.57593606065029,
+      "learning_rate": 1.895108243886366e-07,
+      "loss": 0.6454,
+      "step": 3730
+    },
+    {
+      "epoch": 6.909259259259259,
+      "grad_norm": 1.5176957610124138,
+      "learning_rate": 1.8950413821828248e-07,
+      "loss": 0.661,
+      "step": 3731
+    },
+    {
+      "epoch": 6.911111111111111,
+      "grad_norm": 1.6553322127203016,
+      "learning_rate": 1.8949745003563222e-07,
+      "loss": 0.623,
+      "step": 3732
+    },
+    {
+      "epoch": 6.912962962962963,
+      "grad_norm": 1.448231726102951,
+      "learning_rate": 1.8949075984083622e-07,
+      "loss": 0.6115,
+      "step": 3733
+    },
+    {
+      "epoch": 6.9148148148148145,
+      "grad_norm": 1.649715706166609,
+      "learning_rate": 1.894840676340449e-07,
+      "loss": 0.6406,
+      "step": 3734
+    },
+    {
+      "epoch": 6.916666666666667,
+      "grad_norm": 1.3708481330181648,
+      "learning_rate": 1.894773734154087e-07,
+      "loss": 0.6363,
+      "step": 3735
+    },
+    {
+      "epoch": 6.9185185185185185,
+      "grad_norm": 1.8490920386925511,
+      "learning_rate": 1.8947067718507814e-07,
+      "loss": 0.6662,
+      "step": 3736
+    },
+    {
+      "epoch": 6.92037037037037,
+      "grad_norm": 1.4857734145358497,
+      "learning_rate": 1.8946397894320375e-07,
+      "loss": 0.7231,
+      "step": 3737
+    },
+    {
+      "epoch": 6.9222222222222225,
+      "grad_norm": 1.7840043358150104,
+      "learning_rate": 1.8945727868993613e-07,
+      "loss": 0.6666,
+      "step": 3738
+    },
+    {
+      "epoch": 6.924074074074074,
+      "grad_norm": 1.867143935152978,
+      "learning_rate": 1.8945057642542591e-07,
+      "loss": 0.6939,
+      "step": 3739
+    },
+    {
+      "epoch": 6.925925925925926,
+      "grad_norm": 2.375579610071121,
+      "learning_rate": 1.894438721498238e-07,
+      "loss": 0.6594,
+      "step": 3740
+    },
+    {
+      "epoch": 6.927777777777778,
+      "grad_norm": 1.6727899561474089,
+      "learning_rate": 1.8943716586328052e-07,
+      "loss": 0.6364,
+      "step": 3741
+    },
+    {
+      "epoch": 6.92962962962963,
+      "grad_norm": 2.1328839934262414,
+      "learning_rate": 1.8943045756594685e-07,
+      "loss": 0.6822,
+      "step": 3742
+    },
+    {
+      "epoch": 6.931481481481481,
+      "grad_norm": 1.9963577592510184,
+      "learning_rate": 1.894237472579736e-07,
+      "loss": 0.6465,
+      "step": 3743
+    },
+    {
+      "epoch": 6.933333333333334,
+      "grad_norm": 1.7912782459950478,
+      "learning_rate": 1.8941703493951163e-07,
+      "loss": 0.6431,
+      "step": 3744
+    },
+    {
+      "epoch": 6.935185185185185,
+      "grad_norm": 1.633204654362644,
+      "learning_rate": 1.8941032061071186e-07,
+      "loss": 0.6788,
+      "step": 3745
+    },
+    {
+      "epoch": 6.937037037037037,
+      "grad_norm": 1.5130681240939485,
+      "learning_rate": 1.8940360427172527e-07,
+      "loss": 0.67,
+      "step": 3746
+    },
+    {
+      "epoch": 6.938888888888889,
+      "grad_norm": 1.4701924131258375,
+      "learning_rate": 1.893968859227028e-07,
+      "loss": 0.6994,
+      "step": 3747
+    },
+    {
+      "epoch": 6.940740740740741,
+      "grad_norm": 1.5780033257574644,
+      "learning_rate": 1.8939016556379558e-07,
+      "loss": 0.633,
+      "step": 3748
+    },
+    {
+      "epoch": 6.942592592592592,
+      "grad_norm": 1.513845315573104,
+      "learning_rate": 1.8938344319515466e-07,
+      "loss": 0.6485,
+      "step": 3749
+    },
+    {
+      "epoch": 6.944444444444445,
+      "grad_norm": 1.5162303552476346,
+      "learning_rate": 1.8937671881693117e-07,
+      "loss": 0.663,
+      "step": 3750
+    },
+    {
+      "epoch": 6.946296296296296,
+      "grad_norm": 1.4872176066181257,
+      "learning_rate": 1.8936999242927625e-07,
+      "loss": 0.6435,
+      "step": 3751
+    },
+    {
+      "epoch": 6.948148148148148,
+      "grad_norm": 1.4513972699548876,
+      "learning_rate": 1.8936326403234123e-07,
+      "loss": 0.6382,
+      "step": 3752
+    },
+    {
+      "epoch": 6.95,
+      "grad_norm": 1.371152779136157,
+      "learning_rate": 1.8935653362627728e-07,
+      "loss": 0.6425,
+      "step": 3753
+    },
+    {
+      "epoch": 6.951851851851852,
+      "grad_norm": 1.5893033714815785,
+      "learning_rate": 1.893498012112358e-07,
+      "loss": 0.6855,
+      "step": 3754
+    },
+    {
+      "epoch": 6.953703703703704,
+      "grad_norm": 1.6018091762083748,
+      "learning_rate": 1.893430667873681e-07,
+      "loss": 0.6902,
+      "step": 3755
+    },
+    {
+      "epoch": 6.955555555555556,
+      "grad_norm": 1.5234220101690332,
+      "learning_rate": 1.8933633035482562e-07,
+      "loss": 0.6753,
+      "step": 3756
+    },
+    {
+      "epoch": 6.957407407407407,
+      "grad_norm": 1.5662858067571013,
+      "learning_rate": 1.893295919137598e-07,
+      "loss": 0.6766,
+      "step": 3757
+    },
+    {
+      "epoch": 6.959259259259259,
+      "grad_norm": 1.635046606984755,
+      "learning_rate": 1.8932285146432215e-07,
+      "loss": 0.6125,
+      "step": 3758
+    },
+    {
+      "epoch": 6.961111111111111,
+      "grad_norm": 1.526099270404238,
+      "learning_rate": 1.8931610900666416e-07,
+      "loss": 0.6756,
+      "step": 3759
+    },
+    {
+      "epoch": 6.962962962962963,
+      "grad_norm": 1.7483574053613806,
+      "learning_rate": 1.8930936454093752e-07,
+      "loss": 0.6631,
+      "step": 3760
+    },
+    {
+      "epoch": 6.964814814814815,
+      "grad_norm": 1.821187802000183,
+      "learning_rate": 1.8930261806729375e-07,
+      "loss": 0.6678,
+      "step": 3761
+    },
+    {
+      "epoch": 6.966666666666667,
+      "grad_norm": 1.8067898156884483,
+      "learning_rate": 1.8929586958588462e-07,
+      "loss": 0.6698,
+      "step": 3762
+    },
+    {
+      "epoch": 6.968518518518518,
+      "grad_norm": 1.4730568325665743,
+      "learning_rate": 1.892891190968618e-07,
+      "loss": 0.6941,
+      "step": 3763
+    },
+    {
+      "epoch": 6.97037037037037,
+      "grad_norm": 1.4322413224394548,
+      "learning_rate": 1.8928236660037708e-07,
+      "loss": 0.6619,
+      "step": 3764
+    },
+    {
+      "epoch": 6.972222222222222,
+      "grad_norm": 1.4437285302509855,
+      "learning_rate": 1.8927561209658227e-07,
+      "loss": 0.6906,
+      "step": 3765
+    },
+    {
+      "epoch": 6.974074074074074,
+      "grad_norm": 2.0368963930578716,
+      "learning_rate": 1.892688555856292e-07,
+      "loss": 0.6649,
+      "step": 3766
+    },
+    {
+      "epoch": 6.975925925925926,
+      "grad_norm": 1.913007026516596,
+      "learning_rate": 1.8926209706766986e-07,
+      "loss": 0.6778,
+      "step": 3767
+    },
+    {
+      "epoch": 6.977777777777778,
+      "grad_norm": 2.000599159827014,
+      "learning_rate": 1.8925533654285615e-07,
+      "loss": 0.6854,
+      "step": 3768
+    },
+    {
+      "epoch": 6.979629629629629,
+      "grad_norm": 1.5554951368903644,
+      "learning_rate": 1.8924857401134e-07,
+      "loss": 0.6495,
+      "step": 3769
+    },
+    {
+      "epoch": 6.981481481481482,
+      "grad_norm": 1.5708077309923056,
+      "learning_rate": 1.8924180947327356e-07,
+      "loss": 0.6618,
+      "step": 3770
+    },
+    {
+      "epoch": 6.983333333333333,
+      "grad_norm": 1.6402954939170293,
+      "learning_rate": 1.8923504292880885e-07,
+      "loss": 0.6049,
+      "step": 3771
+    },
+    {
+      "epoch": 6.985185185185185,
+      "grad_norm": 1.4131337947493847,
+      "learning_rate": 1.89228274378098e-07,
+      "loss": 0.6744,
+      "step": 3772
+    },
+    {
+      "epoch": 6.987037037037037,
+      "grad_norm": 1.5548263555384274,
+      "learning_rate": 1.8922150382129325e-07,
+      "loss": 0.6536,
+      "step": 3773
+    },
+    {
+      "epoch": 6.988888888888889,
+      "grad_norm": 1.4625756497529754,
+      "learning_rate": 1.8921473125854672e-07,
+      "loss": 0.6455,
+      "step": 3774
+    },
+    {
+      "epoch": 6.9907407407407405,
+      "grad_norm": 1.4586315382373547,
+      "learning_rate": 1.8920795669001076e-07,
+      "loss": 0.6196,
+      "step": 3775
+    },
+    {
+      "epoch": 6.992592592592593,
+      "grad_norm": 1.4401381843446186,
+      "learning_rate": 1.8920118011583765e-07,
+      "loss": 0.6488,
+      "step": 3776
+    },
+    {
+      "epoch": 6.9944444444444445,
+      "grad_norm": 1.976729657997227,
+      "learning_rate": 1.891944015361797e-07,
+      "loss": 0.6743,
+      "step": 3777
+    },
+    {
+      "epoch": 6.996296296296296,
+      "grad_norm": 1.9011699693512465,
+      "learning_rate": 1.891876209511894e-07,
+      "loss": 0.6468,
+      "step": 3778
+    },
+    {
+      "epoch": 6.9981481481481485,
+      "grad_norm": 1.5686409897996878,
+      "learning_rate": 1.8918083836101918e-07,
+      "loss": 0.6572,
+      "step": 3779
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 1.584022508893578,
+      "learning_rate": 1.8917405376582142e-07,
+      "loss": 0.6883,
+      "step": 3780
+    },
+    {
+      "epoch": 7.0018518518518515,
+      "grad_norm": 1.5148635158449415,
+      "learning_rate": 1.891672671657488e-07,
+      "loss": 0.6761,
+      "step": 3781
+    },
+    {
+      "epoch": 7.003703703703704,
+      "grad_norm": 1.6201043954948564,
+      "learning_rate": 1.8916047856095384e-07,
+      "loss": 0.6561,
+      "step": 3782
+    },
+    {
+      "epoch": 7.0055555555555555,
+      "grad_norm": 1.319132924367095,
+      "learning_rate": 1.891536879515891e-07,
+      "loss": 0.6457,
+      "step": 3783
+    },
+    {
+      "epoch": 7.007407407407407,
+      "grad_norm": 1.9014528315996195,
+      "learning_rate": 1.891468953378074e-07,
+      "loss": 0.6261,
+      "step": 3784
+    },
+    {
+      "epoch": 7.0092592592592595,
+      "grad_norm": 1.904225659480633,
+      "learning_rate": 1.8914010071976133e-07,
+      "loss": 0.641,
+      "step": 3785
+    },
+    {
+      "epoch": 7.011111111111111,
+      "grad_norm": 1.4401106113315392,
+      "learning_rate": 1.8913330409760368e-07,
+      "loss": 0.6585,
+      "step": 3786
+    },
+    {
+      "epoch": 7.012962962962963,
+      "grad_norm": 1.4646495180750831,
+      "learning_rate": 1.8912650547148729e-07,
+      "loss": 0.6707,
+      "step": 3787
+    },
+    {
+      "epoch": 7.014814814814815,
+      "grad_norm": 2.010739386577859,
+      "learning_rate": 1.89119704841565e-07,
+      "loss": 0.6587,
+      "step": 3788
+    },
+    {
+      "epoch": 7.016666666666667,
+      "grad_norm": 1.6411988059033766,
+      "learning_rate": 1.891129022079897e-07,
+      "loss": 0.6316,
+      "step": 3789
+    },
+    {
+      "epoch": 7.018518518518518,
+      "grad_norm": 1.6861889459666546,
+      "learning_rate": 1.8910609757091432e-07,
+      "loss": 0.6351,
+      "step": 3790
+    },
+    {
+      "epoch": 7.020370370370371,
+      "grad_norm": 1.5455541701430804,
+      "learning_rate": 1.8909929093049186e-07,
+      "loss": 0.6631,
+      "step": 3791
+    },
+    {
+      "epoch": 7.022222222222222,
+      "grad_norm": 2.4503574803448123,
+      "learning_rate": 1.8909248228687535e-07,
+      "loss": 0.6903,
+      "step": 3792
+    },
+    {
+      "epoch": 7.024074074074074,
+      "grad_norm": 1.7845738642128253,
+      "learning_rate": 1.8908567164021786e-07,
+      "loss": 0.6531,
+      "step": 3793
+    },
+    {
+      "epoch": 7.025925925925926,
+      "grad_norm": 1.6139964760581305,
+      "learning_rate": 1.8907885899067254e-07,
+      "loss": 0.6728,
+      "step": 3794
+    },
+    {
+      "epoch": 7.027777777777778,
+      "grad_norm": 2.177508652230196,
+      "learning_rate": 1.8907204433839254e-07,
+      "loss": 0.6686,
+      "step": 3795
+    },
+    {
+      "epoch": 7.029629629629629,
+      "grad_norm": 1.6185494915398189,
+      "learning_rate": 1.8906522768353103e-07,
+      "loss": 0.6415,
+      "step": 3796
+    },
+    {
+      "epoch": 7.031481481481482,
+      "grad_norm": 1.2830642330663806,
+      "learning_rate": 1.8905840902624134e-07,
+      "loss": 0.6777,
+      "step": 3797
+    },
+    {
+      "epoch": 7.033333333333333,
+      "grad_norm": 1.5491670956500094,
+      "learning_rate": 1.8905158836667673e-07,
+      "loss": 0.6218,
+      "step": 3798
+    },
+    {
+      "epoch": 7.035185185185185,
+      "grad_norm": 1.5181712339371882,
+      "learning_rate": 1.8904476570499056e-07,
+      "loss": 0.6292,
+      "step": 3799
+    },
+    {
+      "epoch": 7.037037037037037,
+      "grad_norm": 1.7479168469024826,
+      "learning_rate": 1.8903794104133616e-07,
+      "loss": 0.6496,
+      "step": 3800
+    },
+    {
+      "epoch": 7.038888888888889,
+      "grad_norm": 2.322662597891969,
+      "learning_rate": 1.8903111437586707e-07,
+      "loss": 0.6502,
+      "step": 3801
+    },
+    {
+      "epoch": 7.040740740740741,
+      "grad_norm": 1.4617329741188945,
+      "learning_rate": 1.8902428570873675e-07,
+      "loss": 0.6729,
+      "step": 3802
+    },
+    {
+      "epoch": 7.042592592592593,
+      "grad_norm": 1.5448420207988391,
+      "learning_rate": 1.8901745504009866e-07,
+      "loss": 0.6436,
+      "step": 3803
+    },
+    {
+      "epoch": 7.044444444444444,
+      "grad_norm": 1.424090275074607,
+      "learning_rate": 1.8901062237010642e-07,
+      "loss": 0.6442,
+      "step": 3804
+    },
+    {
+      "epoch": 7.046296296296297,
+      "grad_norm": 1.4587107661069902,
+      "learning_rate": 1.8900378769891363e-07,
+      "loss": 0.6727,
+      "step": 3805
+    },
+    {
+      "epoch": 7.048148148148148,
+      "grad_norm": 1.4006843536790925,
+      "learning_rate": 1.8899695102667396e-07,
+      "loss": 0.6243,
+      "step": 3806
+    },
+    {
+      "epoch": 7.05,
+      "grad_norm": 1.6143833895325665,
+      "learning_rate": 1.8899011235354113e-07,
+      "loss": 0.6358,
+      "step": 3807
+    },
+    {
+      "epoch": 7.051851851851852,
+      "grad_norm": 1.3807486570238263,
+      "learning_rate": 1.8898327167966887e-07,
+      "loss": 0.6564,
+      "step": 3808
+    },
+    {
+      "epoch": 7.053703703703704,
+      "grad_norm": 1.3984914690669097,
+      "learning_rate": 1.8897642900521103e-07,
+      "loss": 0.6382,
+      "step": 3809
+    },
+    {
+      "epoch": 7.055555555555555,
+      "grad_norm": 1.4233921684234079,
+      "learning_rate": 1.8896958433032139e-07,
+      "loss": 0.6165,
+      "step": 3810
+    },
+    {
+      "epoch": 7.057407407407408,
+      "grad_norm": 1.5496563254416587,
+      "learning_rate": 1.8896273765515383e-07,
+      "loss": 0.6366,
+      "step": 3811
+    },
+    {
+      "epoch": 7.059259259259259,
+      "grad_norm": 2.7940927374900686,
+      "learning_rate": 1.8895588897986233e-07,
+      "loss": 0.693,
+      "step": 3812
+    },
+    {
+      "epoch": 7.061111111111111,
+      "grad_norm": 1.5990704089133272,
+      "learning_rate": 1.8894903830460082e-07,
+      "loss": 0.7033,
+      "step": 3813
+    },
+    {
+      "epoch": 7.062962962962963,
+      "grad_norm": 1.4485419736912386,
+      "learning_rate": 1.8894218562952336e-07,
+      "loss": 0.6842,
+      "step": 3814
+    },
+    {
+      "epoch": 7.064814814814815,
+      "grad_norm": 1.3575841490905451,
+      "learning_rate": 1.8893533095478403e-07,
+      "loss": 0.682,
+      "step": 3815
+    },
+    {
+      "epoch": 7.066666666666666,
+      "grad_norm": 1.4247753726405616,
+      "learning_rate": 1.889284742805369e-07,
+      "loss": 0.6481,
+      "step": 3816
+    },
+    {
+      "epoch": 7.068518518518519,
+      "grad_norm": 1.5431771573911641,
+      "learning_rate": 1.8892161560693612e-07,
+      "loss": 0.6602,
+      "step": 3817
+    },
+    {
+      "epoch": 7.07037037037037,
+      "grad_norm": 1.4913059094015684,
+      "learning_rate": 1.8891475493413594e-07,
+      "loss": 0.6524,
+      "step": 3818
+    },
+    {
+      "epoch": 7.072222222222222,
+      "grad_norm": 1.4411822425317622,
+      "learning_rate": 1.8890789226229055e-07,
+      "loss": 0.6269,
+      "step": 3819
+    },
+    {
+      "epoch": 7.074074074074074,
+      "grad_norm": 1.6310596446757368,
+      "learning_rate": 1.889010275915543e-07,
+      "loss": 0.6624,
+      "step": 3820
+    },
+    {
+      "epoch": 7.075925925925926,
+      "grad_norm": 1.4582472932053814,
+      "learning_rate": 1.888941609220815e-07,
+      "loss": 0.6399,
+      "step": 3821
+    },
+    {
+      "epoch": 7.0777777777777775,
+      "grad_norm": 1.6701647403780597,
+      "learning_rate": 1.8888729225402652e-07,
+      "loss": 0.6757,
+      "step": 3822
+    },
+    {
+      "epoch": 7.07962962962963,
+      "grad_norm": 1.4741038380070752,
+      "learning_rate": 1.888804215875438e-07,
+      "loss": 0.6725,
+      "step": 3823
+    },
+    {
+      "epoch": 7.0814814814814815,
+      "grad_norm": 1.354451323753022,
+      "learning_rate": 1.8887354892278782e-07,
+      "loss": 0.6461,
+      "step": 3824
+    },
+    {
+      "epoch": 7.083333333333333,
+      "grad_norm": 1.6106308005765915,
+      "learning_rate": 1.8886667425991307e-07,
+      "loss": 0.6099,
+      "step": 3825
+    },
+    {
+      "epoch": 7.0851851851851855,
+      "grad_norm": 1.404362210362807,
+      "learning_rate": 1.888597975990741e-07,
+      "loss": 0.6516,
+      "step": 3826
+    },
+    {
+      "epoch": 7.087037037037037,
+      "grad_norm": 1.9192295923238643,
+      "learning_rate": 1.8885291894042558e-07,
+      "loss": 0.653,
+      "step": 3827
+    },
+    {
+      "epoch": 7.088888888888889,
+      "grad_norm": 2.357469502616754,
+      "learning_rate": 1.8884603828412212e-07,
+      "loss": 0.6408,
+      "step": 3828
+    },
+    {
+      "epoch": 7.090740740740741,
+      "grad_norm": 1.6600772019635135,
+      "learning_rate": 1.8883915563031837e-07,
+      "loss": 0.6639,
+      "step": 3829
+    },
+    {
+      "epoch": 7.092592592592593,
+      "grad_norm": 1.65294751230379,
+      "learning_rate": 1.8883227097916917e-07,
+      "loss": 0.6489,
+      "step": 3830
+    },
+    {
+      "epoch": 7.094444444444444,
+      "grad_norm": 1.4917645208882038,
+      "learning_rate": 1.8882538433082922e-07,
+      "loss": 0.7064,
+      "step": 3831
+    },
+    {
+      "epoch": 7.0962962962962965,
+      "grad_norm": 1.483161226487801,
+      "learning_rate": 1.888184956854534e-07,
+      "loss": 0.6752,
+      "step": 3832
+    },
+    {
+      "epoch": 7.098148148148148,
+      "grad_norm": 1.6625645182672377,
+      "learning_rate": 1.8881160504319657e-07,
+      "loss": 0.6225,
+      "step": 3833
+    },
+    {
+      "epoch": 7.1,
+      "grad_norm": 1.6628820010557193,
+      "learning_rate": 1.8880471240421364e-07,
+      "loss": 0.6582,
+      "step": 3834
+    },
+    {
+      "epoch": 7.101851851851852,
+      "grad_norm": 1.3936728465431745,
+      "learning_rate": 1.887978177686596e-07,
+      "loss": 0.6603,
+      "step": 3835
+    },
+    {
+      "epoch": 7.103703703703704,
+      "grad_norm": 3.1120663049006723,
+      "learning_rate": 1.8879092113668943e-07,
+      "loss": 0.6477,
+      "step": 3836
+    },
+    {
+      "epoch": 7.105555555555555,
+      "grad_norm": 1.3528474015908034,
+      "learning_rate": 1.8878402250845821e-07,
+      "loss": 0.6973,
+      "step": 3837
+    },
+    {
+      "epoch": 7.107407407407408,
+      "grad_norm": 1.6064333266934185,
+      "learning_rate": 1.8877712188412104e-07,
+      "loss": 0.6668,
+      "step": 3838
+    },
+    {
+      "epoch": 7.109259259259259,
+      "grad_norm": 1.4184827569603229,
+      "learning_rate": 1.8877021926383302e-07,
+      "loss": 0.6415,
+      "step": 3839
+    },
+    {
+      "epoch": 7.111111111111111,
+      "grad_norm": 2.2987675229885323,
+      "learning_rate": 1.8876331464774942e-07,
+      "loss": 0.6212,
+      "step": 3840
+    },
+    {
+      "epoch": 7.112962962962963,
+      "grad_norm": 1.6548802174221124,
+      "learning_rate": 1.887564080360254e-07,
+      "loss": 0.6581,
+      "step": 3841
+    },
+    {
+      "epoch": 7.114814814814815,
+      "grad_norm": 1.979919832157401,
+      "learning_rate": 1.8874949942881625e-07,
+      "loss": 0.6536,
+      "step": 3842
+    },
+    {
+      "epoch": 7.116666666666666,
+      "grad_norm": 2.0173517985854117,
+      "learning_rate": 1.8874258882627733e-07,
+      "loss": 0.6358,
+      "step": 3843
+    },
+    {
+      "epoch": 7.118518518518519,
+      "grad_norm": 1.5985849065778481,
+      "learning_rate": 1.8873567622856402e-07,
+      "loss": 0.684,
+      "step": 3844
+    },
+    {
+      "epoch": 7.12037037037037,
+      "grad_norm": 1.6315226586683655,
+      "learning_rate": 1.887287616358317e-07,
+      "loss": 0.6514,
+      "step": 3845
+    },
+    {
+      "epoch": 7.122222222222222,
+      "grad_norm": 1.6036906764410606,
+      "learning_rate": 1.887218450482358e-07,
+      "loss": 0.6366,
+      "step": 3846
+    },
+    {
+      "epoch": 7.124074074074074,
+      "grad_norm": 1.8699340968748386,
+      "learning_rate": 1.887149264659319e-07,
+      "loss": 0.6668,
+      "step": 3847
+    },
+    {
+      "epoch": 7.125925925925926,
+      "grad_norm": 1.4819708051686205,
+      "learning_rate": 1.8870800588907547e-07,
+      "loss": 0.6476,
+      "step": 3848
+    },
+    {
+      "epoch": 7.127777777777778,
+      "grad_norm": 1.855927403652474,
+      "learning_rate": 1.8870108331782214e-07,
+      "loss": 0.6722,
+      "step": 3849
+    },
+    {
+      "epoch": 7.12962962962963,
+      "grad_norm": 1.6502540815895659,
+      "learning_rate": 1.886941587523276e-07,
+      "loss": 0.6174,
+      "step": 3850
+    },
+    {
+      "epoch": 7.131481481481481,
+      "grad_norm": 6.27316503381482,
+      "learning_rate": 1.8868723219274742e-07,
+      "loss": 0.6554,
+      "step": 3851
+    },
+    {
+      "epoch": 7.133333333333334,
+      "grad_norm": 1.7578849986145926,
+      "learning_rate": 1.8868030363923743e-07,
+      "loss": 0.6433,
+      "step": 3852
+    },
+    {
+      "epoch": 7.135185185185185,
+      "grad_norm": 1.6964535594880703,
+      "learning_rate": 1.8867337309195336e-07,
+      "loss": 0.6472,
+      "step": 3853
+    },
+    {
+      "epoch": 7.137037037037037,
+      "grad_norm": 1.5635892556909845,
+      "learning_rate": 1.88666440551051e-07,
+      "loss": 0.6939,
+      "step": 3854
+    },
+    {
+      "epoch": 7.138888888888889,
+      "grad_norm": 1.6794114761553867,
+      "learning_rate": 1.886595060166863e-07,
+      "loss": 0.6199,
+      "step": 3855
+    },
+    {
+      "epoch": 7.140740740740741,
+      "grad_norm": 1.4944608806373647,
+      "learning_rate": 1.8865256948901506e-07,
+      "loss": 0.6907,
+      "step": 3856
+    },
+    {
+      "epoch": 7.142592592592592,
+      "grad_norm": 1.5058651462008914,
+      "learning_rate": 1.8864563096819327e-07,
+      "loss": 0.676,
+      "step": 3857
+    },
+    {
+      "epoch": 7.144444444444445,
+      "grad_norm": 1.7225456820470928,
+      "learning_rate": 1.8863869045437694e-07,
+      "loss": 0.6401,
+      "step": 3858
+    },
+    {
+      "epoch": 7.146296296296296,
+      "grad_norm": 1.8313228140413405,
+      "learning_rate": 1.8863174794772214e-07,
+      "loss": 0.6427,
+      "step": 3859
+    },
+    {
+      "epoch": 7.148148148148148,
+      "grad_norm": 1.4074709176373037,
+      "learning_rate": 1.8862480344838492e-07,
+      "loss": 0.6611,
+      "step": 3860
+    },
+    {
+      "epoch": 7.15,
+      "grad_norm": 1.4773359487564475,
+      "learning_rate": 1.8861785695652141e-07,
+      "loss": 0.6293,
+      "step": 3861
+    },
+    {
+      "epoch": 7.151851851851852,
+      "grad_norm": 1.6334113374766588,
+      "learning_rate": 1.886109084722878e-07,
+      "loss": 0.6487,
+      "step": 3862
+    },
+    {
+      "epoch": 7.1537037037037035,
+      "grad_norm": 1.5871776141561962,
+      "learning_rate": 1.8860395799584027e-07,
+      "loss": 0.6684,
+      "step": 3863
+    },
+    {
+      "epoch": 7.155555555555556,
+      "grad_norm": 1.7675143086737106,
+      "learning_rate": 1.8859700552733515e-07,
+      "loss": 0.6612,
+      "step": 3864
+    },
+    {
+      "epoch": 7.157407407407407,
+      "grad_norm": 1.5625901916181437,
+      "learning_rate": 1.8859005106692873e-07,
+      "loss": 0.6959,
+      "step": 3865
+    },
+    {
+      "epoch": 7.159259259259259,
+      "grad_norm": 1.5362406437161675,
+      "learning_rate": 1.8858309461477735e-07,
+      "loss": 0.6408,
+      "step": 3866
+    },
+    {
+      "epoch": 7.161111111111111,
+      "grad_norm": 1.4598016385756507,
+      "learning_rate": 1.885761361710374e-07,
+      "loss": 0.6798,
+      "step": 3867
+    },
+    {
+      "epoch": 7.162962962962963,
+      "grad_norm": 1.5998399467113293,
+      "learning_rate": 1.8856917573586535e-07,
+      "loss": 0.6379,
+      "step": 3868
+    },
+    {
+      "epoch": 7.1648148148148145,
+      "grad_norm": 1.4686353698298469,
+      "learning_rate": 1.885622133094177e-07,
+      "loss": 0.631,
+      "step": 3869
+    },
+    {
+      "epoch": 7.166666666666667,
+      "grad_norm": 1.536428113332608,
+      "learning_rate": 1.8855524889185093e-07,
+      "loss": 0.6599,
+      "step": 3870
+    },
+    {
+      "epoch": 7.1685185185185185,
+      "grad_norm": 1.5964128859774331,
+      "learning_rate": 1.8854828248332165e-07,
+      "loss": 0.6348,
+      "step": 3871
+    },
+    {
+      "epoch": 7.17037037037037,
+      "grad_norm": 1.5687555028743951,
+      "learning_rate": 1.8854131408398651e-07,
+      "loss": 0.6582,
+      "step": 3872
+    },
+    {
+      "epoch": 7.1722222222222225,
+      "grad_norm": 1.5339340899242777,
+      "learning_rate": 1.885343436940021e-07,
+      "loss": 0.6743,
+      "step": 3873
+    },
+    {
+      "epoch": 7.174074074074074,
+      "grad_norm": 1.413969126390241,
+      "learning_rate": 1.8852737131352528e-07,
+      "loss": 0.6057,
+      "step": 3874
+    },
+    {
+      "epoch": 7.175925925925926,
+      "grad_norm": 1.389578942011871,
+      "learning_rate": 1.8852039694271265e-07,
+      "loss": 0.6284,
+      "step": 3875
+    },
+    {
+      "epoch": 7.177777777777778,
+      "grad_norm": 1.7142042908403565,
+      "learning_rate": 1.885134205817211e-07,
+      "loss": 0.6913,
+      "step": 3876
+    },
+    {
+      "epoch": 7.17962962962963,
+      "grad_norm": 2.6299219853837736,
+      "learning_rate": 1.8850644223070745e-07,
+      "loss": 0.6222,
+      "step": 3877
+    },
+    {
+      "epoch": 7.181481481481481,
+      "grad_norm": 2.3473674730398035,
+      "learning_rate": 1.884994618898286e-07,
+      "loss": 0.6575,
+      "step": 3878
+    },
+    {
+      "epoch": 7.183333333333334,
+      "grad_norm": 1.630867989586954,
+      "learning_rate": 1.8849247955924147e-07,
+      "loss": 0.6414,
+      "step": 3879
+    },
+    {
+      "epoch": 7.185185185185185,
+      "grad_norm": 1.8145038461911651,
+      "learning_rate": 1.884854952391031e-07,
+      "loss": 0.6507,
+      "step": 3880
+    },
+    {
+      "epoch": 7.187037037037037,
+      "grad_norm": 1.5987641775808559,
+      "learning_rate": 1.8847850892957045e-07,
+      "loss": 0.6711,
+      "step": 3881
+    },
+    {
+      "epoch": 7.188888888888889,
+      "grad_norm": 2.696866273756317,
+      "learning_rate": 1.884715206308006e-07,
+      "loss": 0.6517,
+      "step": 3882
+    },
+    {
+      "epoch": 7.190740740740741,
+      "grad_norm": 8.973496167083082,
+      "learning_rate": 1.884645303429507e-07,
+      "loss": 0.6508,
+      "step": 3883
+    },
+    {
+      "epoch": 7.192592592592592,
+      "grad_norm": 1.6659351260990651,
+      "learning_rate": 1.8845753806617791e-07,
+      "loss": 0.6288,
+      "step": 3884
+    },
+    {
+      "epoch": 7.194444444444445,
+      "grad_norm": 1.4719304183891508,
+      "learning_rate": 1.8845054380063937e-07,
+      "loss": 0.6818,
+      "step": 3885
+    },
+    {
+      "epoch": 7.196296296296296,
+      "grad_norm": 1.4892986386424063,
+      "learning_rate": 1.884435475464924e-07,
+      "loss": 0.643,
+      "step": 3886
+    },
+    {
+      "epoch": 7.198148148148148,
+      "grad_norm": 1.5833156254768301,
+      "learning_rate": 1.8843654930389429e-07,
+      "loss": 0.5923,
+      "step": 3887
+    },
+    {
+      "epoch": 7.2,
+      "grad_norm": 1.5900125590073069,
+      "learning_rate": 1.8842954907300233e-07,
+      "loss": 0.6475,
+      "step": 3888
+    },
+    {
+      "epoch": 7.201851851851852,
+      "grad_norm": 1.9049173073075365,
+      "learning_rate": 1.8842254685397397e-07,
+      "loss": 0.6222,
+      "step": 3889
+    },
+    {
+      "epoch": 7.203703703703703,
+      "grad_norm": 1.2899176724558696,
+      "learning_rate": 1.884155426469666e-07,
+      "loss": 0.6403,
+      "step": 3890
+    },
+    {
+      "epoch": 7.205555555555556,
+      "grad_norm": 1.6366894540785846,
+      "learning_rate": 1.8840853645213768e-07,
+      "loss": 0.6708,
+      "step": 3891
+    },
+    {
+      "epoch": 7.207407407407407,
+      "grad_norm": 1.277917052255994,
+      "learning_rate": 1.8840152826964474e-07,
+      "loss": 0.6065,
+      "step": 3892
+    },
+    {
+      "epoch": 7.209259259259259,
+      "grad_norm": 1.9130971269001955,
+      "learning_rate": 1.8839451809964537e-07,
+      "loss": 0.6875,
+      "step": 3893
+    },
+    {
+      "epoch": 7.211111111111111,
+      "grad_norm": 1.522689540944064,
+      "learning_rate": 1.8838750594229716e-07,
+      "loss": 0.6781,
+      "step": 3894
+    },
+    {
+      "epoch": 7.212962962962963,
+      "grad_norm": 1.499668882598213,
+      "learning_rate": 1.8838049179775774e-07,
+      "loss": 0.6156,
+      "step": 3895
+    },
+    {
+      "epoch": 7.214814814814815,
+      "grad_norm": 1.5689321670663572,
+      "learning_rate": 1.8837347566618481e-07,
+      "loss": 0.643,
+      "step": 3896
+    },
+    {
+      "epoch": 7.216666666666667,
+      "grad_norm": 1.4867489809770138,
+      "learning_rate": 1.8836645754773616e-07,
+      "loss": 0.6527,
+      "step": 3897
+    },
+    {
+      "epoch": 7.218518518518518,
+      "grad_norm": 1.5481088621741936,
+      "learning_rate": 1.8835943744256958e-07,
+      "loss": 0.6546,
+      "step": 3898
+    },
+    {
+      "epoch": 7.220370370370371,
+      "grad_norm": 1.6518008488001963,
+      "learning_rate": 1.883524153508428e-07,
+      "loss": 0.6025,
+      "step": 3899
+    },
+    {
+      "epoch": 7.222222222222222,
+      "grad_norm": 1.4935660192485618,
+      "learning_rate": 1.8834539127271377e-07,
+      "loss": 0.613,
+      "step": 3900
+    },
+    {
+      "epoch": 7.224074074074074,
+      "grad_norm": 1.6446163565824925,
+      "learning_rate": 1.8833836520834044e-07,
+      "loss": 0.6513,
+      "step": 3901
+    },
+    {
+      "epoch": 7.225925925925926,
+      "grad_norm": 1.6544292092216015,
+      "learning_rate": 1.883313371578807e-07,
+      "loss": 0.6419,
+      "step": 3902
+    },
+    {
+      "epoch": 7.227777777777778,
+      "grad_norm": 1.611398208604589,
+      "learning_rate": 1.883243071214926e-07,
+      "loss": 0.6145,
+      "step": 3903
+    },
+    {
+      "epoch": 7.229629629629629,
+      "grad_norm": 1.570677382116332,
+      "learning_rate": 1.8831727509933422e-07,
+      "loss": 0.6554,
+      "step": 3904
+    },
+    {
+      "epoch": 7.231481481481482,
+      "grad_norm": 1.698686244968201,
+      "learning_rate": 1.8831024109156362e-07,
+      "loss": 0.6535,
+      "step": 3905
+    },
+    {
+      "epoch": 7.233333333333333,
+      "grad_norm": 1.3183208014599064,
+      "learning_rate": 1.8830320509833895e-07,
+      "loss": 0.6301,
+      "step": 3906
+    },
+    {
+      "epoch": 7.235185185185185,
+      "grad_norm": 1.3986975525552585,
+      "learning_rate": 1.8829616711981837e-07,
+      "loss": 0.6218,
+      "step": 3907
+    },
+    {
+      "epoch": 7.237037037037037,
+      "grad_norm": 1.366057077596558,
+      "learning_rate": 1.882891271561602e-07,
+      "loss": 0.6761,
+      "step": 3908
+    },
+    {
+      "epoch": 7.238888888888889,
+      "grad_norm": 1.436074307512177,
+      "learning_rate": 1.882820852075226e-07,
+      "loss": 0.699,
+      "step": 3909
+    },
+    {
+      "epoch": 7.2407407407407405,
+      "grad_norm": 1.7941992120343393,
+      "learning_rate": 1.88275041274064e-07,
+      "loss": 0.6536,
+      "step": 3910
+    },
+    {
+      "epoch": 7.242592592592593,
+      "grad_norm": 1.4959755979091727,
+      "learning_rate": 1.8826799535594268e-07,
+      "loss": 0.6936,
+      "step": 3911
+    },
+    {
+      "epoch": 7.2444444444444445,
+      "grad_norm": 1.4669996445640352,
+      "learning_rate": 1.8826094745331714e-07,
+      "loss": 0.6804,
+      "step": 3912
+    },
+    {
+      "epoch": 7.246296296296296,
+      "grad_norm": 1.4510489047234636,
+      "learning_rate": 1.8825389756634577e-07,
+      "loss": 0.634,
+      "step": 3913
+    },
+    {
+      "epoch": 7.2481481481481485,
+      "grad_norm": 1.507667624718648,
+      "learning_rate": 1.8824684569518707e-07,
+      "loss": 0.6412,
+      "step": 3914
+    },
+    {
+      "epoch": 7.25,
+      "grad_norm": 2.26204919181382,
+      "learning_rate": 1.8823979183999964e-07,
+      "loss": 0.6346,
+      "step": 3915
+    },
+    {
+      "epoch": 7.2518518518518515,
+      "grad_norm": 2.2553876605011234,
+      "learning_rate": 1.8823273600094202e-07,
+      "loss": 0.6458,
+      "step": 3916
+    },
+    {
+      "epoch": 7.253703703703704,
+      "grad_norm": 1.4143685791332317,
+      "learning_rate": 1.8822567817817286e-07,
+      "loss": 0.6379,
+      "step": 3917
+    },
+    {
+      "epoch": 7.2555555555555555,
+      "grad_norm": 1.3817341785013824,
+      "learning_rate": 1.8821861837185082e-07,
+      "loss": 0.6528,
+      "step": 3918
+    },
+    {
+      "epoch": 7.257407407407407,
+      "grad_norm": 2.261792719811734,
+      "learning_rate": 1.8821155658213464e-07,
+      "loss": 0.6799,
+      "step": 3919
+    },
+    {
+      "epoch": 7.2592592592592595,
+      "grad_norm": 1.6073797206949885,
+      "learning_rate": 1.882044928091831e-07,
+      "loss": 0.6158,
+      "step": 3920
+    },
+    {
+      "epoch": 7.261111111111111,
+      "grad_norm": 1.4764509201538014,
+      "learning_rate": 1.88197427053155e-07,
+      "loss": 0.6289,
+      "step": 3921
+    },
+    {
+      "epoch": 7.262962962962963,
+      "grad_norm": 1.5525193173261245,
+      "learning_rate": 1.8819035931420923e-07,
+      "loss": 0.6345,
+      "step": 3922
+    },
+    {
+      "epoch": 7.264814814814815,
+      "grad_norm": 1.3764809218195233,
+      "learning_rate": 1.881832895925046e-07,
+      "loss": 0.655,
+      "step": 3923
+    },
+    {
+      "epoch": 7.266666666666667,
+      "grad_norm": 1.7909947471470726,
+      "learning_rate": 1.8817621788820015e-07,
+      "loss": 0.6961,
+      "step": 3924
+    },
+    {
+      "epoch": 7.268518518518518,
+      "grad_norm": 1.4945584351625694,
+      "learning_rate": 1.8816914420145484e-07,
+      "loss": 0.6441,
+      "step": 3925
+    },
+    {
+      "epoch": 7.270370370370371,
+      "grad_norm": 1.4756961611301094,
+      "learning_rate": 1.881620685324277e-07,
+      "loss": 0.604,
+      "step": 3926
+    },
+    {
+      "epoch": 7.272222222222222,
+      "grad_norm": 1.6034171479543946,
+      "learning_rate": 1.8815499088127782e-07,
+      "loss": 0.6652,
+      "step": 3927
+    },
+    {
+      "epoch": 7.274074074074074,
+      "grad_norm": 1.6979320194310967,
+      "learning_rate": 1.8814791124816431e-07,
+      "loss": 0.6324,
+      "step": 3928
+    },
+    {
+      "epoch": 7.275925925925926,
+      "grad_norm": 1.488771109320376,
+      "learning_rate": 1.881408296332464e-07,
+      "loss": 0.6637,
+      "step": 3929
+    },
+    {
+      "epoch": 7.277777777777778,
+      "grad_norm": 1.7298965202641352,
+      "learning_rate": 1.8813374603668318e-07,
+      "loss": 0.6605,
+      "step": 3930
+    },
+    {
+      "epoch": 7.279629629629629,
+      "grad_norm": 1.7955800628411287,
+      "learning_rate": 1.8812666045863402e-07,
+      "loss": 0.6553,
+      "step": 3931
+    },
+    {
+      "epoch": 7.281481481481482,
+      "grad_norm": 1.536856073088939,
+      "learning_rate": 1.8811957289925817e-07,
+      "loss": 0.7155,
+      "step": 3932
+    },
+    {
+      "epoch": 7.283333333333333,
+      "grad_norm": 4.672029829624969,
+      "learning_rate": 1.88112483358715e-07,
+      "loss": 0.6273,
+      "step": 3933
+    },
+    {
+      "epoch": 7.285185185185185,
+      "grad_norm": 2.058608757288798,
+      "learning_rate": 1.8810539183716387e-07,
+      "loss": 0.6399,
+      "step": 3934
+    },
+    {
+      "epoch": 7.287037037037037,
+      "grad_norm": 1.7895664318485744,
+      "learning_rate": 1.8809829833476427e-07,
+      "loss": 0.6193,
+      "step": 3935
+    },
+    {
+      "epoch": 7.288888888888889,
+      "grad_norm": 1.6735787370584698,
+      "learning_rate": 1.8809120285167565e-07,
+      "loss": 0.6822,
+      "step": 3936
+    },
+    {
+      "epoch": 7.290740740740741,
+      "grad_norm": 1.6569132458058724,
+      "learning_rate": 1.8808410538805752e-07,
+      "loss": 0.6397,
+      "step": 3937
+    },
+    {
+      "epoch": 7.292592592592593,
+      "grad_norm": 1.6423860174800282,
+      "learning_rate": 1.880770059440695e-07,
+      "loss": 0.683,
+      "step": 3938
+    },
+    {
+      "epoch": 7.294444444444444,
+      "grad_norm": 1.928869152293549,
+      "learning_rate": 1.8806990451987112e-07,
+      "loss": 0.6339,
+      "step": 3939
+    },
+    {
+      "epoch": 7.296296296296296,
+      "grad_norm": 1.9896247663473383,
+      "learning_rate": 1.8806280111562214e-07,
+      "loss": 0.6946,
+      "step": 3940
+    },
+    {
+      "epoch": 7.298148148148148,
+      "grad_norm": 1.7409612936664953,
+      "learning_rate": 1.880556957314822e-07,
+      "loss": 0.6492,
+      "step": 3941
+    },
+    {
+      "epoch": 7.3,
+      "grad_norm": 1.7510285567339543,
+      "learning_rate": 1.8804858836761106e-07,
+      "loss": 0.6583,
+      "step": 3942
+    },
+    {
+      "epoch": 7.301851851851852,
+      "grad_norm": 1.5451859824210152,
+      "learning_rate": 1.8804147902416855e-07,
+      "loss": 0.6356,
+      "step": 3943
+    },
+    {
+      "epoch": 7.303703703703704,
+      "grad_norm": 1.362482648982307,
+      "learning_rate": 1.8803436770131444e-07,
+      "loss": 0.6415,
+      "step": 3944
+    },
+    {
+      "epoch": 7.305555555555555,
+      "grad_norm": 1.4676534218636943,
+      "learning_rate": 1.8802725439920865e-07,
+      "loss": 0.6331,
+      "step": 3945
+    },
+    {
+      "epoch": 7.307407407407408,
+      "grad_norm": 1.6048907731684041,
+      "learning_rate": 1.8802013911801108e-07,
+      "loss": 0.684,
+      "step": 3946
+    },
+    {
+      "epoch": 7.309259259259259,
+      "grad_norm": 1.6965847442519473,
+      "learning_rate": 1.880130218578818e-07,
+      "loss": 0.6736,
+      "step": 3947
+    },
+    {
+      "epoch": 7.311111111111111,
+      "grad_norm": 1.6076566054676888,
+      "learning_rate": 1.880059026189807e-07,
+      "loss": 0.6086,
+      "step": 3948
+    },
+    {
+      "epoch": 7.312962962962963,
+      "grad_norm": 1.5311607599184833,
+      "learning_rate": 1.8799878140146785e-07,
+      "loss": 0.59,
+      "step": 3949
+    },
+    {
+      "epoch": 7.314814814814815,
+      "grad_norm": 1.3828818058044876,
+      "learning_rate": 1.8799165820550344e-07,
+      "loss": 0.6213,
+      "step": 3950
+    },
+    {
+      "epoch": 7.316666666666666,
+      "grad_norm": 1.5412514403839601,
+      "learning_rate": 1.8798453303124755e-07,
+      "loss": 0.6459,
+      "step": 3951
+    },
+    {
+      "epoch": 7.318518518518519,
+      "grad_norm": 1.6610813797053923,
+      "learning_rate": 1.8797740587886043e-07,
+      "loss": 0.6128,
+      "step": 3952
+    },
+    {
+      "epoch": 7.32037037037037,
+      "grad_norm": 1.9076274754200053,
+      "learning_rate": 1.8797027674850227e-07,
+      "loss": 0.6092,
+      "step": 3953
+    },
+    {
+      "epoch": 7.322222222222222,
+      "grad_norm": 1.5040723671179161,
+      "learning_rate": 1.8796314564033335e-07,
+      "loss": 0.6535,
+      "step": 3954
+    },
+    {
+      "epoch": 7.324074074074074,
+      "grad_norm": 1.8666699580720543,
+      "learning_rate": 1.8795601255451402e-07,
+      "loss": 0.6415,
+      "step": 3955
+    },
+    {
+      "epoch": 7.325925925925926,
+      "grad_norm": 1.7952492974322916,
+      "learning_rate": 1.8794887749120464e-07,
+      "loss": 0.6422,
+      "step": 3956
+    },
+    {
+      "epoch": 7.3277777777777775,
+      "grad_norm": 1.6526668294092484,
+      "learning_rate": 1.879417404505656e-07,
+      "loss": 0.6249,
+      "step": 3957
+    },
+    {
+      "epoch": 7.32962962962963,
+      "grad_norm": 1.5771963090925472,
+      "learning_rate": 1.8793460143275742e-07,
+      "loss": 0.6532,
+      "step": 3958
+    },
+    {
+      "epoch": 7.3314814814814815,
+      "grad_norm": 1.8678657727234584,
+      "learning_rate": 1.8792746043794058e-07,
+      "loss": 0.6232,
+      "step": 3959
+    },
+    {
+      "epoch": 7.333333333333333,
+      "grad_norm": 1.7589011450543741,
+      "learning_rate": 1.879203174662756e-07,
+      "loss": 0.6621,
+      "step": 3960
+    },
+    {
+      "epoch": 7.3351851851851855,
+      "grad_norm": 1.6241955707340285,
+      "learning_rate": 1.8791317251792312e-07,
+      "loss": 0.6623,
+      "step": 3961
+    },
+    {
+      "epoch": 7.337037037037037,
+      "grad_norm": 1.9852262336269557,
+      "learning_rate": 1.8790602559304377e-07,
+      "loss": 0.6543,
+      "step": 3962
+    },
+    {
+      "epoch": 7.338888888888889,
+      "grad_norm": 1.7617850997132742,
+      "learning_rate": 1.8789887669179815e-07,
+      "loss": 0.6607,
+      "step": 3963
+    },
+    {
+      "epoch": 7.340740740740741,
+      "grad_norm": 1.6416548301433183,
+      "learning_rate": 1.878917258143471e-07,
+      "loss": 0.6322,
+      "step": 3964
+    },
+    {
+      "epoch": 7.342592592592593,
+      "grad_norm": 1.4702170252549254,
+      "learning_rate": 1.878845729608513e-07,
+      "loss": 0.6448,
+      "step": 3965
+    },
+    {
+      "epoch": 7.344444444444444,
+      "grad_norm": 2.0293644914716795,
+      "learning_rate": 1.8787741813147168e-07,
+      "loss": 0.6644,
+      "step": 3966
+    },
+    {
+      "epoch": 7.3462962962962965,
+      "grad_norm": 1.5609067834681374,
+      "learning_rate": 1.87870261326369e-07,
+      "loss": 0.6344,
+      "step": 3967
+    },
+    {
+      "epoch": 7.348148148148148,
+      "grad_norm": 1.6320911450561009,
+      "learning_rate": 1.8786310254570417e-07,
+      "loss": 0.6216,
+      "step": 3968
+    },
+    {
+      "epoch": 7.35,
+      "grad_norm": 1.530596755852607,
+      "learning_rate": 1.8785594178963816e-07,
+      "loss": 0.6444,
+      "step": 3969
+    },
+    {
+      "epoch": 7.351851851851852,
+      "grad_norm": 1.4486525552744287,
+      "learning_rate": 1.8784877905833198e-07,
+      "loss": 0.6599,
+      "step": 3970
+    },
+    {
+      "epoch": 7.353703703703704,
+      "grad_norm": 1.6468175125216373,
+      "learning_rate": 1.8784161435194666e-07,
+      "loss": 0.6535,
+      "step": 3971
+    },
+    {
+      "epoch": 7.355555555555555,
+      "grad_norm": 2.515406861929086,
+      "learning_rate": 1.8783444767064328e-07,
+      "loss": 0.6162,
+      "step": 3972
+    },
+    {
+      "epoch": 7.357407407407408,
+      "grad_norm": 2.1408575222742967,
+      "learning_rate": 1.8782727901458295e-07,
+      "loss": 0.6082,
+      "step": 3973
+    },
+    {
+      "epoch": 7.359259259259259,
+      "grad_norm": 1.5767584206411043,
+      "learning_rate": 1.8782010838392686e-07,
+      "loss": 0.6621,
+      "step": 3974
+    },
+    {
+      "epoch": 7.361111111111111,
+      "grad_norm": 1.5336110860037588,
+      "learning_rate": 1.878129357788362e-07,
+      "loss": 0.6119,
+      "step": 3975
+    },
+    {
+      "epoch": 7.362962962962963,
+      "grad_norm": 1.6740254455036625,
+      "learning_rate": 1.8780576119947228e-07,
+      "loss": 0.7035,
+      "step": 3976
+    },
+    {
+      "epoch": 7.364814814814815,
+      "grad_norm": 1.5989550949179896,
+      "learning_rate": 1.8779858464599635e-07,
+      "loss": 0.6406,
+      "step": 3977
+    },
+    {
+      "epoch": 7.366666666666666,
+      "grad_norm": 1.7439363606875136,
+      "learning_rate": 1.8779140611856977e-07,
+      "loss": 0.6565,
+      "step": 3978
+    },
+    {
+      "epoch": 7.368518518518519,
+      "grad_norm": 1.8878705061997887,
+      "learning_rate": 1.8778422561735393e-07,
+      "loss": 0.6085,
+      "step": 3979
+    },
+    {
+      "epoch": 7.37037037037037,
+      "grad_norm": 1.3416696649699746,
+      "learning_rate": 1.877770431425103e-07,
+      "loss": 0.6435,
+      "step": 3980
+    },
+    {
+      "epoch": 7.372222222222222,
+      "grad_norm": 1.6151147095024616,
+      "learning_rate": 1.8776985869420036e-07,
+      "loss": 0.6295,
+      "step": 3981
+    },
+    {
+      "epoch": 7.374074074074074,
+      "grad_norm": 1.4710999831483937,
+      "learning_rate": 1.877626722725856e-07,
+      "loss": 0.6368,
+      "step": 3982
+    },
+    {
+      "epoch": 7.375925925925926,
+      "grad_norm": 1.6531219984563443,
+      "learning_rate": 1.877554838778276e-07,
+      "loss": 0.6592,
+      "step": 3983
+    },
+    {
+      "epoch": 7.377777777777778,
+      "grad_norm": 1.530917705147961,
+      "learning_rate": 1.87748293510088e-07,
+      "loss": 0.6665,
+      "step": 3984
+    },
+    {
+      "epoch": 7.37962962962963,
+      "grad_norm": 1.7138727852005868,
+      "learning_rate": 1.8774110116952844e-07,
+      "loss": 0.6704,
+      "step": 3985
+    },
+    {
+      "epoch": 7.381481481481481,
+      "grad_norm": 1.560608128273509,
+      "learning_rate": 1.8773390685631061e-07,
+      "loss": 0.6577,
+      "step": 3986
+    },
+    {
+      "epoch": 7.383333333333334,
+      "grad_norm": 1.6224083625363526,
+      "learning_rate": 1.8772671057059628e-07,
+      "loss": 0.661,
+      "step": 3987
+    },
+    {
+      "epoch": 7.385185185185185,
+      "grad_norm": 1.5726732441112203,
+      "learning_rate": 1.8771951231254724e-07,
+      "loss": 0.6482,
+      "step": 3988
+    },
+    {
+      "epoch": 7.387037037037037,
+      "grad_norm": 1.9291963480221006,
+      "learning_rate": 1.8771231208232533e-07,
+      "loss": 0.6534,
+      "step": 3989
+    },
+    {
+      "epoch": 7.388888888888889,
+      "grad_norm": 1.5343596598027045,
+      "learning_rate": 1.8770510988009238e-07,
+      "loss": 0.6709,
+      "step": 3990
+    },
+    {
+      "epoch": 7.390740740740741,
+      "grad_norm": 1.2932356018626276,
+      "learning_rate": 1.876979057060104e-07,
+      "loss": 0.6167,
+      "step": 3991
+    },
+    {
+      "epoch": 7.392592592592592,
+      "grad_norm": 1.4489345231061077,
+      "learning_rate": 1.8769069956024128e-07,
+      "loss": 0.6377,
+      "step": 3992
+    },
+    {
+      "epoch": 7.394444444444445,
+      "grad_norm": 1.7032318991020898,
+      "learning_rate": 1.8768349144294707e-07,
+      "loss": 0.6524,
+      "step": 3993
+    },
+    {
+      "epoch": 7.396296296296296,
+      "grad_norm": 1.5067205342109846,
+      "learning_rate": 1.8767628135428985e-07,
+      "loss": 0.6809,
+      "step": 3994
+    },
+    {
+      "epoch": 7.398148148148148,
+      "grad_norm": 1.5282536219355214,
+      "learning_rate": 1.876690692944317e-07,
+      "loss": 0.6629,
+      "step": 3995
+    },
+    {
+      "epoch": 7.4,
+      "grad_norm": 1.5922591850406895,
+      "learning_rate": 1.8766185526353477e-07,
+      "loss": 0.6646,
+      "step": 3996
+    },
+    {
+      "epoch": 7.401851851851852,
+      "grad_norm": 1.5105664851385068,
+      "learning_rate": 1.8765463926176125e-07,
+      "loss": 0.6506,
+      "step": 3997
+    },
+    {
+      "epoch": 7.4037037037037035,
+      "grad_norm": 1.5136841683198523,
+      "learning_rate": 1.8764742128927335e-07,
+      "loss": 0.5898,
+      "step": 3998
+    },
+    {
+      "epoch": 7.405555555555556,
+      "grad_norm": 1.550094609955071,
+      "learning_rate": 1.8764020134623337e-07,
+      "loss": 0.6511,
+      "step": 3999
+    },
+    {
+      "epoch": 7.407407407407407,
+      "grad_norm": 1.3737429684767637,
+      "learning_rate": 1.8763297943280366e-07,
+      "loss": 0.6475,
+      "step": 4000
+    },
+    {
+      "epoch": 7.409259259259259,
+      "grad_norm": 1.7435083066201096,
+      "learning_rate": 1.8762575554914655e-07,
+      "loss": 0.6383,
+      "step": 4001
+    },
+    {
+      "epoch": 7.411111111111111,
+      "grad_norm": 1.4886156360198073,
+      "learning_rate": 1.8761852969542447e-07,
+      "loss": 0.6527,
+      "step": 4002
+    },
+    {
+      "epoch": 7.412962962962963,
+      "grad_norm": 1.5659926446768353,
+      "learning_rate": 1.8761130187179988e-07,
+      "loss": 0.6434,
+      "step": 4003
+    },
+    {
+      "epoch": 7.4148148148148145,
+      "grad_norm": 1.7098065567296454,
+      "learning_rate": 1.8760407207843526e-07,
+      "loss": 0.6387,
+      "step": 4004
+    },
+    {
+      "epoch": 7.416666666666667,
+      "grad_norm": 1.6000855220329502,
+      "learning_rate": 1.8759684031549318e-07,
+      "loss": 0.6396,
+      "step": 4005
+    },
+    {
+      "epoch": 7.4185185185185185,
+      "grad_norm": 1.465118207342098,
+      "learning_rate": 1.8758960658313626e-07,
+      "loss": 0.638,
+      "step": 4006
+    },
+    {
+      "epoch": 7.42037037037037,
+      "grad_norm": 1.552743290645851,
+      "learning_rate": 1.8758237088152705e-07,
+      "loss": 0.6299,
+      "step": 4007
+    },
+    {
+      "epoch": 7.4222222222222225,
+      "grad_norm": 1.6209530980176254,
+      "learning_rate": 1.8757513321082826e-07,
+      "loss": 0.686,
+      "step": 4008
+    },
+    {
+      "epoch": 7.424074074074074,
+      "grad_norm": 1.6088353083788376,
+      "learning_rate": 1.8756789357120266e-07,
+      "loss": 0.6762,
+      "step": 4009
+    },
+    {
+      "epoch": 7.425925925925926,
+      "grad_norm": 2.0756921880937975,
+      "learning_rate": 1.8756065196281295e-07,
+      "loss": 0.6266,
+      "step": 4010
+    },
+    {
+      "epoch": 7.427777777777778,
+      "grad_norm": 1.5774188227604389,
+      "learning_rate": 1.8755340838582198e-07,
+      "loss": 0.6501,
+      "step": 4011
+    },
+    {
+      "epoch": 7.42962962962963,
+      "grad_norm": 1.9601902196906416,
+      "learning_rate": 1.8754616284039262e-07,
+      "loss": 0.6418,
+      "step": 4012
+    },
+    {
+      "epoch": 7.431481481481481,
+      "grad_norm": 1.9357120800036363,
+      "learning_rate": 1.8753891532668775e-07,
+      "loss": 0.679,
+      "step": 4013
+    },
+    {
+      "epoch": 7.433333333333334,
+      "grad_norm": 2.043317406892933,
+      "learning_rate": 1.8753166584487028e-07,
+      "loss": 0.6242,
+      "step": 4014
+    },
+    {
+      "epoch": 7.435185185185185,
+      "grad_norm": 1.5645792562007772,
+      "learning_rate": 1.8752441439510328e-07,
+      "loss": 0.6295,
+      "step": 4015
+    },
+    {
+      "epoch": 7.437037037037037,
+      "grad_norm": 1.8893870468079825,
+      "learning_rate": 1.8751716097754966e-07,
+      "loss": 0.6574,
+      "step": 4016
+    },
+    {
+      "epoch": 7.438888888888889,
+      "grad_norm": 1.4123013040658945,
+      "learning_rate": 1.875099055923726e-07,
+      "loss": 0.6534,
+      "step": 4017
+    },
+    {
+      "epoch": 7.440740740740741,
+      "grad_norm": 1.4955145167966055,
+      "learning_rate": 1.8750264823973523e-07,
+      "loss": 0.6719,
+      "step": 4018
+    },
+    {
+      "epoch": 7.442592592592592,
+      "grad_norm": 1.370447461015703,
+      "learning_rate": 1.8749538891980062e-07,
+      "loss": 0.6786,
+      "step": 4019
+    },
+    {
+      "epoch": 7.444444444444445,
+      "grad_norm": 1.8380337806617724,
+      "learning_rate": 1.8748812763273205e-07,
+      "loss": 0.6852,
+      "step": 4020
+    },
+    {
+      "epoch": 7.446296296296296,
+      "grad_norm": 1.798180985712554,
+      "learning_rate": 1.8748086437869275e-07,
+      "loss": 0.6118,
+      "step": 4021
+    },
+    {
+      "epoch": 7.448148148148148,
+      "grad_norm": 1.5041488822934745,
+      "learning_rate": 1.8747359915784605e-07,
+      "loss": 0.6671,
+      "step": 4022
+    },
+    {
+      "epoch": 7.45,
+      "grad_norm": 1.4826549827362148,
+      "learning_rate": 1.8746633197035523e-07,
+      "loss": 0.6602,
+      "step": 4023
+    },
+    {
+      "epoch": 7.451851851851852,
+      "grad_norm": 1.534158352676765,
+      "learning_rate": 1.8745906281638375e-07,
+      "loss": 0.6479,
+      "step": 4024
+    },
+    {
+      "epoch": 7.453703703703704,
+      "grad_norm": 1.5152215479266407,
+      "learning_rate": 1.8745179169609497e-07,
+      "loss": 0.6468,
+      "step": 4025
+    },
+    {
+      "epoch": 7.455555555555556,
+      "grad_norm": 1.6936004984417925,
+      "learning_rate": 1.8744451860965243e-07,
+      "loss": 0.7017,
+      "step": 4026
+    },
+    {
+      "epoch": 7.457407407407407,
+      "grad_norm": 1.5532458965125615,
+      "learning_rate": 1.8743724355721959e-07,
+      "loss": 0.6633,
+      "step": 4027
+    },
+    {
+      "epoch": 7.459259259259259,
+      "grad_norm": 2.830682580661416,
+      "learning_rate": 1.8742996653896004e-07,
+      "loss": 0.6325,
+      "step": 4028
+    },
+    {
+      "epoch": 7.461111111111111,
+      "grad_norm": 1.5554722988885161,
+      "learning_rate": 1.874226875550374e-07,
+      "loss": 0.6591,
+      "step": 4029
+    },
+    {
+      "epoch": 7.462962962962963,
+      "grad_norm": 3.4003184089125456,
+      "learning_rate": 1.874154066056153e-07,
+      "loss": 0.6578,
+      "step": 4030
+    },
+    {
+      "epoch": 7.464814814814815,
+      "grad_norm": 1.4879026736168877,
+      "learning_rate": 1.8740812369085744e-07,
+      "loss": 0.6776,
+      "step": 4031
+    },
+    {
+      "epoch": 7.466666666666667,
+      "grad_norm": 1.8577202723662751,
+      "learning_rate": 1.8740083881092755e-07,
+      "loss": 0.6808,
+      "step": 4032
+    },
+    {
+      "epoch": 7.468518518518518,
+      "grad_norm": 1.6480056299091879,
+      "learning_rate": 1.8739355196598944e-07,
+      "loss": 0.6731,
+      "step": 4033
+    },
+    {
+      "epoch": 7.470370370370371,
+      "grad_norm": 1.4266394018792556,
+      "learning_rate": 1.8738626315620697e-07,
+      "loss": 0.6829,
+      "step": 4034
+    },
+    {
+      "epoch": 7.472222222222222,
+      "grad_norm": 1.7993016395786594,
+      "learning_rate": 1.8737897238174388e-07,
+      "loss": 0.6762,
+      "step": 4035
+    },
+    {
+      "epoch": 7.474074074074074,
+      "grad_norm": 1.7744971752791796,
+      "learning_rate": 1.8737167964276425e-07,
+      "loss": 0.6435,
+      "step": 4036
+    },
+    {
+      "epoch": 7.475925925925926,
+      "grad_norm": 1.5467923767374805,
+      "learning_rate": 1.873643849394319e-07,
+      "loss": 0.634,
+      "step": 4037
+    },
+    {
+      "epoch": 7.477777777777778,
+      "grad_norm": 1.416268413536496,
+      "learning_rate": 1.8735708827191095e-07,
+      "loss": 0.6659,
+      "step": 4038
+    },
+    {
+      "epoch": 7.479629629629629,
+      "grad_norm": 1.4636895234396003,
+      "learning_rate": 1.8734978964036537e-07,
+      "loss": 0.5964,
+      "step": 4039
+    },
+    {
+      "epoch": 7.481481481481482,
+      "grad_norm": 1.5593564915151255,
+      "learning_rate": 1.8734248904495928e-07,
+      "loss": 0.644,
+      "step": 4040
+    },
+    {
+      "epoch": 7.483333333333333,
+      "grad_norm": 2.0615276535442333,
+      "learning_rate": 1.8733518648585682e-07,
+      "loss": 0.6199,
+      "step": 4041
+    },
+    {
+      "epoch": 7.485185185185185,
+      "grad_norm": 1.5671530258906192,
+      "learning_rate": 1.873278819632222e-07,
+      "loss": 0.6332,
+      "step": 4042
+    },
+    {
+      "epoch": 7.487037037037037,
+      "grad_norm": 1.523895005421646,
+      "learning_rate": 1.873205754772196e-07,
+      "loss": 0.6585,
+      "step": 4043
+    },
+    {
+      "epoch": 7.488888888888889,
+      "grad_norm": 1.538715776387948,
+      "learning_rate": 1.8731326702801328e-07,
+      "loss": 0.655,
+      "step": 4044
+    },
+    {
+      "epoch": 7.4907407407407405,
+      "grad_norm": 1.5859609127860925,
+      "learning_rate": 1.873059566157676e-07,
+      "loss": 0.6518,
+      "step": 4045
+    },
+    {
+      "epoch": 7.492592592592593,
+      "grad_norm": 1.3829374655509887,
+      "learning_rate": 1.8729864424064685e-07,
+      "loss": 0.622,
+      "step": 4046
+    },
+    {
+      "epoch": 7.4944444444444445,
+      "grad_norm": 1.7939613736001132,
+      "learning_rate": 1.872913299028155e-07,
+      "loss": 0.6557,
+      "step": 4047
+    },
+    {
+      "epoch": 7.496296296296296,
+      "grad_norm": 1.7906214805864318,
+      "learning_rate": 1.8728401360243801e-07,
+      "loss": 0.6697,
+      "step": 4048
+    },
+    {
+      "epoch": 7.4981481481481485,
+      "grad_norm": 1.6453260599093877,
+      "learning_rate": 1.8727669533967882e-07,
+      "loss": 0.6184,
+      "step": 4049
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 1.418240400125389,
+      "learning_rate": 1.8726937511470244e-07,
+      "loss": 0.633,
+      "step": 4050
+    },
+    {
+      "epoch": 7.5018518518518515,
+      "grad_norm": 1.4344293873011618,
+      "learning_rate": 1.8726205292767353e-07,
+      "loss": 0.623,
+      "step": 4051
+    },
+    {
+      "epoch": 7.503703703703704,
+      "grad_norm": 1.5055414276790458,
+      "learning_rate": 1.8725472877875665e-07,
+      "loss": 0.7047,
+      "step": 4052
+    },
+    {
+      "epoch": 7.5055555555555555,
+      "grad_norm": 1.6205361185627871,
+      "learning_rate": 1.8724740266811652e-07,
+      "loss": 0.6511,
+      "step": 4053
+    },
+    {
+      "epoch": 7.507407407407407,
+      "grad_norm": 1.8670585084832534,
+      "learning_rate": 1.8724007459591778e-07,
+      "loss": 0.6621,
+      "step": 4054
+    },
+    {
+      "epoch": 7.5092592592592595,
+      "grad_norm": 1.3419629611741195,
+      "learning_rate": 1.8723274456232523e-07,
+      "loss": 0.6109,
+      "step": 4055
+    },
+    {
+      "epoch": 7.511111111111111,
+      "grad_norm": 2.782860920470479,
+      "learning_rate": 1.8722541256750368e-07,
+      "loss": 0.6479,
+      "step": 4056
+    },
+    {
+      "epoch": 7.512962962962963,
+      "grad_norm": 1.6492478442743528,
+      "learning_rate": 1.8721807861161795e-07,
+      "loss": 0.639,
+      "step": 4057
+    },
+    {
+      "epoch": 7.514814814814815,
+      "grad_norm": 1.6453918198706934,
+      "learning_rate": 1.8721074269483292e-07,
+      "loss": 0.6113,
+      "step": 4058
+    },
+    {
+      "epoch": 7.516666666666667,
+      "grad_norm": 2.0578754491473985,
+      "learning_rate": 1.8720340481731355e-07,
+      "loss": 0.6413,
+      "step": 4059
+    },
+    {
+      "epoch": 7.518518518518518,
+      "grad_norm": 1.596467408196277,
+      "learning_rate": 1.8719606497922475e-07,
+      "loss": 0.6481,
+      "step": 4060
+    },
+    {
+      "epoch": 7.520370370370371,
+      "grad_norm": 1.6790788253844622,
+      "learning_rate": 1.8718872318073166e-07,
+      "loss": 0.6417,
+      "step": 4061
+    },
+    {
+      "epoch": 7.522222222222222,
+      "grad_norm": 1.6472597441947012,
+      "learning_rate": 1.8718137942199924e-07,
+      "loss": 0.6134,
+      "step": 4062
+    },
+    {
+      "epoch": 7.524074074074074,
+      "grad_norm": 1.399303210350014,
+      "learning_rate": 1.871740337031926e-07,
+      "loss": 0.6688,
+      "step": 4063
+    },
+    {
+      "epoch": 7.525925925925926,
+      "grad_norm": 1.4801833157831892,
+      "learning_rate": 1.8716668602447697e-07,
+      "loss": 0.615,
+      "step": 4064
+    },
+    {
+      "epoch": 7.527777777777778,
+      "grad_norm": 1.5060934448923868,
+      "learning_rate": 1.8715933638601747e-07,
+      "loss": 0.6678,
+      "step": 4065
+    },
+    {
+      "epoch": 7.52962962962963,
+      "grad_norm": 1.4460234910383185,
+      "learning_rate": 1.8715198478797935e-07,
+      "loss": 0.6113,
+      "step": 4066
+    },
+    {
+      "epoch": 7.531481481481482,
+      "grad_norm": 1.956525870718335,
+      "learning_rate": 1.8714463123052794e-07,
+      "loss": 0.6404,
+      "step": 4067
+    },
+    {
+      "epoch": 7.533333333333333,
+      "grad_norm": 1.4309588994630034,
+      "learning_rate": 1.8713727571382853e-07,
+      "loss": 0.657,
+      "step": 4068
+    },
+    {
+      "epoch": 7.535185185185185,
+      "grad_norm": 2.568462404788745,
+      "learning_rate": 1.871299182380465e-07,
+      "loss": 0.6673,
+      "step": 4069
+    },
+    {
+      "epoch": 7.537037037037037,
+      "grad_norm": 1.4946588970444665,
+      "learning_rate": 1.871225588033473e-07,
+      "loss": 0.6267,
+      "step": 4070
+    },
+    {
+      "epoch": 7.538888888888889,
+      "grad_norm": 1.5236186631144908,
+      "learning_rate": 1.871151974098963e-07,
+      "loss": 0.6642,
+      "step": 4071
+    },
+    {
+      "epoch": 7.540740740740741,
+      "grad_norm": 1.6973899684221827,
+      "learning_rate": 1.871078340578591e-07,
+      "loss": 0.6438,
+      "step": 4072
+    },
+    {
+      "epoch": 7.542592592592593,
+      "grad_norm": 1.5599095910521932,
+      "learning_rate": 1.871004687474012e-07,
+      "loss": 0.6267,
+      "step": 4073
+    },
+    {
+      "epoch": 7.544444444444444,
+      "grad_norm": 1.5949436518509634,
+      "learning_rate": 1.8709310147868817e-07,
+      "loss": 0.6805,
+      "step": 4074
+    },
+    {
+      "epoch": 7.546296296296296,
+      "grad_norm": 1.5539873584251627,
+      "learning_rate": 1.870857322518857e-07,
+      "loss": 0.663,
+      "step": 4075
+    },
+    {
+      "epoch": 7.548148148148148,
+      "grad_norm": 1.4658836808621116,
+      "learning_rate": 1.8707836106715946e-07,
+      "loss": 0.6932,
+      "step": 4076
+    },
+    {
+      "epoch": 7.55,
+      "grad_norm": 1.7332260236788322,
+      "learning_rate": 1.8707098792467514e-07,
+      "loss": 0.6748,
+      "step": 4077
+    },
+    {
+      "epoch": 7.551851851851852,
+      "grad_norm": 1.9394567152486057,
+      "learning_rate": 1.8706361282459855e-07,
+      "loss": 0.7063,
+      "step": 4078
+    },
+    {
+      "epoch": 7.553703703703704,
+      "grad_norm": 1.574868185309456,
+      "learning_rate": 1.870562357670955e-07,
+      "loss": 0.6002,
+      "step": 4079
+    },
+    {
+      "epoch": 7.555555555555555,
+      "grad_norm": 1.5996707840386655,
+      "learning_rate": 1.870488567523318e-07,
+      "loss": 0.6647,
+      "step": 4080
+    },
+    {
+      "epoch": 7.557407407407408,
+      "grad_norm": 1.8674986538734952,
+      "learning_rate": 1.8704147578047338e-07,
+      "loss": 0.6809,
+      "step": 4081
+    },
+    {
+      "epoch": 7.559259259259259,
+      "grad_norm": 1.7562215366182878,
+      "learning_rate": 1.870340928516862e-07,
+      "loss": 0.664,
+      "step": 4082
+    },
+    {
+      "epoch": 7.561111111111111,
+      "grad_norm": 1.3540635157724146,
+      "learning_rate": 1.870267079661362e-07,
+      "loss": 0.6129,
+      "step": 4083
+    },
+    {
+      "epoch": 7.562962962962963,
+      "grad_norm": 1.3417011511580168,
+      "learning_rate": 1.870193211239895e-07,
+      "loss": 0.6198,
+      "step": 4084
+    },
+    {
+      "epoch": 7.564814814814815,
+      "grad_norm": 1.4992339490772237,
+      "learning_rate": 1.8701193232541206e-07,
+      "loss": 0.6421,
+      "step": 4085
+    },
+    {
+      "epoch": 7.566666666666666,
+      "grad_norm": 1.6067545683737503,
+      "learning_rate": 1.8700454157057008e-07,
+      "loss": 0.6315,
+      "step": 4086
+    },
+    {
+      "epoch": 7.568518518518519,
+      "grad_norm": 1.8488157713980993,
+      "learning_rate": 1.8699714885962973e-07,
+      "loss": 0.6782,
+      "step": 4087
+    },
+    {
+      "epoch": 7.57037037037037,
+      "grad_norm": 1.6190132458664848,
+      "learning_rate": 1.8698975419275716e-07,
+      "loss": 0.6534,
+      "step": 4088
+    },
+    {
+      "epoch": 7.572222222222222,
+      "grad_norm": 1.6983870556164997,
+      "learning_rate": 1.8698235757011863e-07,
+      "loss": 0.6422,
+      "step": 4089
+    },
+    {
+      "epoch": 7.574074074074074,
+      "grad_norm": 1.5614689595140214,
+      "learning_rate": 1.8697495899188053e-07,
+      "loss": 0.6724,
+      "step": 4090
+    },
+    {
+      "epoch": 7.575925925925926,
+      "grad_norm": 1.5441570201477295,
+      "learning_rate": 1.869675584582091e-07,
+      "loss": 0.6038,
+      "step": 4091
+    },
+    {
+      "epoch": 7.5777777777777775,
+      "grad_norm": 1.446960781686149,
+      "learning_rate": 1.8696015596927074e-07,
+      "loss": 0.6784,
+      "step": 4092
+    },
+    {
+      "epoch": 7.57962962962963,
+      "grad_norm": 1.7529008641910226,
+      "learning_rate": 1.869527515252319e-07,
+      "loss": 0.6176,
+      "step": 4093
+    },
+    {
+      "epoch": 7.5814814814814815,
+      "grad_norm": 1.6847089951326684,
+      "learning_rate": 1.8694534512625907e-07,
+      "loss": 0.6265,
+      "step": 4094
+    },
+    {
+      "epoch": 7.583333333333333,
+      "grad_norm": 1.4162552686373346,
+      "learning_rate": 1.8693793677251875e-07,
+      "loss": 0.6602,
+      "step": 4095
+    },
+    {
+      "epoch": 7.5851851851851855,
+      "grad_norm": 1.423651370019151,
+      "learning_rate": 1.8693052646417744e-07,
+      "loss": 0.6397,
+      "step": 4096
+    },
+    {
+      "epoch": 7.587037037037037,
+      "grad_norm": 1.5178443648516051,
+      "learning_rate": 1.8692311420140183e-07,
+      "loss": 0.654,
+      "step": 4097
+    },
+    {
+      "epoch": 7.588888888888889,
+      "grad_norm": 1.4734287879802168,
+      "learning_rate": 1.8691569998435853e-07,
+      "loss": 0.6505,
+      "step": 4098
+    },
+    {
+      "epoch": 7.590740740740741,
+      "grad_norm": 1.4322999673146037,
+      "learning_rate": 1.8690828381321425e-07,
+      "loss": 0.6345,
+      "step": 4099
+    },
+    {
+      "epoch": 7.592592592592593,
+      "grad_norm": 1.399537941239336,
+      "learning_rate": 1.8690086568813568e-07,
+      "loss": 0.6682,
+      "step": 4100
+    },
+    {
+      "epoch": 7.594444444444444,
+      "grad_norm": 1.4578955970828915,
+      "learning_rate": 1.8689344560928962e-07,
+      "loss": 0.634,
+      "step": 4101
+    },
+    {
+      "epoch": 7.5962962962962965,
+      "grad_norm": 1.3507504118187434,
+      "learning_rate": 1.8688602357684294e-07,
+      "loss": 0.6212,
+      "step": 4102
+    },
+    {
+      "epoch": 7.598148148148148,
+      "grad_norm": 1.514425950643582,
+      "learning_rate": 1.8687859959096246e-07,
+      "loss": 0.6117,
+      "step": 4103
+    },
+    {
+      "epoch": 7.6,
+      "grad_norm": 1.7487680201284967,
+      "learning_rate": 1.8687117365181512e-07,
+      "loss": 0.6439,
+      "step": 4104
+    },
+    {
+      "epoch": 7.601851851851852,
+      "grad_norm": 1.5926878763333923,
+      "learning_rate": 1.8686374575956782e-07,
+      "loss": 0.6159,
+      "step": 4105
+    },
+    {
+      "epoch": 7.603703703703704,
+      "grad_norm": 1.4476350343337396,
+      "learning_rate": 1.8685631591438762e-07,
+      "loss": 0.624,
+      "step": 4106
+    },
+    {
+      "epoch": 7.605555555555555,
+      "grad_norm": 1.4128404717725838,
+      "learning_rate": 1.868488841164415e-07,
+      "loss": 0.6421,
+      "step": 4107
+    },
+    {
+      "epoch": 7.607407407407408,
+      "grad_norm": 1.6589220959329893,
+      "learning_rate": 1.8684145036589665e-07,
+      "loss": 0.5936,
+      "step": 4108
+    },
+    {
+      "epoch": 7.609259259259259,
+      "grad_norm": 1.6706293647527808,
+      "learning_rate": 1.868340146629201e-07,
+      "loss": 0.6477,
+      "step": 4109
+    },
+    {
+      "epoch": 7.611111111111111,
+      "grad_norm": 1.3567694028042443,
+      "learning_rate": 1.8682657700767908e-07,
+      "loss": 0.7028,
+      "step": 4110
+    },
+    {
+      "epoch": 7.612962962962963,
+      "grad_norm": 1.575009583686362,
+      "learning_rate": 1.8681913740034077e-07,
+      "loss": 0.681,
+      "step": 4111
+    },
+    {
+      "epoch": 7.614814814814815,
+      "grad_norm": 1.6028449486300527,
+      "learning_rate": 1.8681169584107248e-07,
+      "loss": 0.6813,
+      "step": 4112
+    },
+    {
+      "epoch": 7.616666666666667,
+      "grad_norm": 1.6138012920798457,
+      "learning_rate": 1.868042523300415e-07,
+      "loss": 0.6642,
+      "step": 4113
+    },
+    {
+      "epoch": 7.618518518518519,
+      "grad_norm": 1.6417737668598553,
+      "learning_rate": 1.8679680686741511e-07,
+      "loss": 0.6624,
+      "step": 4114
+    },
+    {
+      "epoch": 7.62037037037037,
+      "grad_norm": 1.944018814569686,
+      "learning_rate": 1.867893594533608e-07,
+      "loss": 0.6947,
+      "step": 4115
+    },
+    {
+      "epoch": 7.622222222222222,
+      "grad_norm": 1.5583566702500589,
+      "learning_rate": 1.86781910088046e-07,
+      "loss": 0.6575,
+      "step": 4116
+    },
+    {
+      "epoch": 7.624074074074074,
+      "grad_norm": 1.6248047442394085,
+      "learning_rate": 1.8677445877163814e-07,
+      "loss": 0.6404,
+      "step": 4117
+    },
+    {
+      "epoch": 7.625925925925926,
+      "grad_norm": 1.6359576339180701,
+      "learning_rate": 1.8676700550430477e-07,
+      "loss": 0.6745,
+      "step": 4118
+    },
+    {
+      "epoch": 7.627777777777778,
+      "grad_norm": 1.3942692090006625,
+      "learning_rate": 1.8675955028621346e-07,
+      "loss": 0.6559,
+      "step": 4119
+    },
+    {
+      "epoch": 7.62962962962963,
+      "grad_norm": 1.8661747658574874,
+      "learning_rate": 1.8675209311753183e-07,
+      "loss": 0.6365,
+      "step": 4120
+    },
+    {
+      "epoch": 7.631481481481481,
+      "grad_norm": 1.7458349511732807,
+      "learning_rate": 1.8674463399842755e-07,
+      "loss": 0.63,
+      "step": 4121
+    },
+    {
+      "epoch": 7.633333333333333,
+      "grad_norm": 1.4325536750833956,
+      "learning_rate": 1.8673717292906828e-07,
+      "loss": 0.6521,
+      "step": 4122
+    },
+    {
+      "epoch": 7.635185185185185,
+      "grad_norm": 1.5807239475664143,
+      "learning_rate": 1.8672970990962178e-07,
+      "loss": 0.6401,
+      "step": 4123
+    },
+    {
+      "epoch": 7.637037037037037,
+      "grad_norm": 1.8146623204222527,
+      "learning_rate": 1.8672224494025587e-07,
+      "loss": 0.6214,
+      "step": 4124
+    },
+    {
+      "epoch": 7.638888888888889,
+      "grad_norm": 1.5746341321799902,
+      "learning_rate": 1.8671477802113832e-07,
+      "loss": 0.6383,
+      "step": 4125
+    },
+    {
+      "epoch": 7.640740740740741,
+      "grad_norm": 1.4294142776569376,
+      "learning_rate": 1.867073091524371e-07,
+      "loss": 0.6819,
+      "step": 4126
+    },
+    {
+      "epoch": 7.642592592592592,
+      "grad_norm": 1.5730825867120848,
+      "learning_rate": 1.8669983833432003e-07,
+      "loss": 0.6776,
+      "step": 4127
+    },
+    {
+      "epoch": 7.644444444444445,
+      "grad_norm": 1.6503488283002208,
+      "learning_rate": 1.8669236556695514e-07,
+      "loss": 0.6662,
+      "step": 4128
+    },
+    {
+      "epoch": 7.646296296296296,
+      "grad_norm": 1.6033761589685698,
+      "learning_rate": 1.8668489085051042e-07,
+      "loss": 0.6274,
+      "step": 4129
+    },
+    {
+      "epoch": 7.648148148148148,
+      "grad_norm": 1.5472723799331083,
+      "learning_rate": 1.8667741418515393e-07,
+      "loss": 0.6397,
+      "step": 4130
+    },
+    {
+      "epoch": 7.65,
+      "grad_norm": 1.5269383901455245,
+      "learning_rate": 1.8666993557105375e-07,
+      "loss": 0.634,
+      "step": 4131
+    },
+    {
+      "epoch": 7.651851851851852,
+      "grad_norm": 1.7745620664794373,
+      "learning_rate": 1.8666245500837802e-07,
+      "loss": 0.704,
+      "step": 4132
+    },
+    {
+      "epoch": 7.6537037037037035,
+      "grad_norm": 1.463490393733774,
+      "learning_rate": 1.8665497249729493e-07,
+      "loss": 0.6135,
+      "step": 4133
+    },
+    {
+      "epoch": 7.655555555555556,
+      "grad_norm": 1.4476115112626073,
+      "learning_rate": 1.866474880379727e-07,
+      "loss": 0.6444,
+      "step": 4134
+    },
+    {
+      "epoch": 7.657407407407407,
+      "grad_norm": 1.5550557659600037,
+      "learning_rate": 1.8664000163057965e-07,
+      "loss": 0.6334,
+      "step": 4135
+    },
+    {
+      "epoch": 7.659259259259259,
+      "grad_norm": 1.4533100743292084,
+      "learning_rate": 1.86632513275284e-07,
+      "loss": 0.6727,
+      "step": 4136
+    },
+    {
+      "epoch": 7.661111111111111,
+      "grad_norm": 1.336483695629435,
+      "learning_rate": 1.866250229722542e-07,
+      "loss": 0.6222,
+      "step": 4137
+    },
+    {
+      "epoch": 7.662962962962963,
+      "grad_norm": 1.9822843581928953,
+      "learning_rate": 1.866175307216586e-07,
+      "loss": 0.6008,
+      "step": 4138
+    },
+    {
+      "epoch": 7.6648148148148145,
+      "grad_norm": 1.5860642113070924,
+      "learning_rate": 1.8661003652366565e-07,
+      "loss": 0.6395,
+      "step": 4139
+    },
+    {
+      "epoch": 7.666666666666667,
+      "grad_norm": 1.5766500516971789,
+      "learning_rate": 1.8660254037844388e-07,
+      "loss": 0.6296,
+      "step": 4140
+    },
+    {
+      "epoch": 7.6685185185185185,
+      "grad_norm": 1.3911359736764293,
+      "learning_rate": 1.8659504228616175e-07,
+      "loss": 0.5978,
+      "step": 4141
+    },
+    {
+      "epoch": 7.67037037037037,
+      "grad_norm": 1.6035764175993417,
+      "learning_rate": 1.8658754224698788e-07,
+      "loss": 0.6507,
+      "step": 4142
+    },
+    {
+      "epoch": 7.6722222222222225,
+      "grad_norm": 1.4717955274007033,
+      "learning_rate": 1.8658004026109092e-07,
+      "loss": 0.6542,
+      "step": 4143
+    },
+    {
+      "epoch": 7.674074074074074,
+      "grad_norm": 1.469327684153371,
+      "learning_rate": 1.865725363286395e-07,
+      "loss": 0.644,
+      "step": 4144
+    },
+    {
+      "epoch": 7.675925925925926,
+      "grad_norm": 1.5363057272052651,
+      "learning_rate": 1.8656503044980235e-07,
+      "loss": 0.6804,
+      "step": 4145
+    },
+    {
+      "epoch": 7.677777777777778,
+      "grad_norm": 1.4531300272551757,
+      "learning_rate": 1.865575226247482e-07,
+      "loss": 0.6775,
+      "step": 4146
+    },
+    {
+      "epoch": 7.67962962962963,
+      "grad_norm": 1.5383319069582633,
+      "learning_rate": 1.8655001285364584e-07,
+      "loss": 0.6039,
+      "step": 4147
+    },
+    {
+      "epoch": 7.681481481481481,
+      "grad_norm": 1.7909824422740164,
+      "learning_rate": 1.8654250113666415e-07,
+      "loss": 0.6171,
+      "step": 4148
+    },
+    {
+      "epoch": 7.683333333333334,
+      "grad_norm": 1.3492724670530756,
+      "learning_rate": 1.8653498747397195e-07,
+      "loss": 0.6335,
+      "step": 4149
+    },
+    {
+      "epoch": 7.685185185185185,
+      "grad_norm": 1.553135203701596,
+      "learning_rate": 1.8652747186573822e-07,
+      "loss": 0.6299,
+      "step": 4150
+    },
+    {
+      "epoch": 7.687037037037037,
+      "grad_norm": 1.4691030321047163,
+      "learning_rate": 1.8651995431213196e-07,
+      "loss": 0.6378,
+      "step": 4151
+    },
+    {
+      "epoch": 7.688888888888889,
+      "grad_norm": 1.8708028071404206,
+      "learning_rate": 1.865124348133221e-07,
+      "loss": 0.6234,
+      "step": 4152
+    },
+    {
+      "epoch": 7.690740740740741,
+      "grad_norm": 1.3929958372307631,
+      "learning_rate": 1.8650491336947775e-07,
+      "loss": 0.6259,
+      "step": 4153
+    },
+    {
+      "epoch": 7.692592592592592,
+      "grad_norm": 1.420316610570055,
+      "learning_rate": 1.8649738998076798e-07,
+      "loss": 0.6443,
+      "step": 4154
+    },
+    {
+      "epoch": 7.694444444444445,
+      "grad_norm": 1.749485863276849,
+      "learning_rate": 1.86489864647362e-07,
+      "loss": 0.6598,
+      "step": 4155
+    },
+    {
+      "epoch": 7.696296296296296,
+      "grad_norm": 1.347132867413065,
+      "learning_rate": 1.8648233736942894e-07,
+      "loss": 0.6407,
+      "step": 4156
+    },
+    {
+      "epoch": 7.698148148148148,
+      "grad_norm": 1.9319174794968108,
+      "learning_rate": 1.8647480814713804e-07,
+      "loss": 0.6305,
+      "step": 4157
+    },
+    {
+      "epoch": 7.7,
+      "grad_norm": 1.588122532870714,
+      "learning_rate": 1.8646727698065863e-07,
+      "loss": 0.6831,
+      "step": 4158
+    },
+    {
+      "epoch": 7.701851851851852,
+      "grad_norm": 1.5731503694414677,
+      "learning_rate": 1.8645974387015998e-07,
+      "loss": 0.595,
+      "step": 4159
+    },
+    {
+      "epoch": 7.703703703703704,
+      "grad_norm": 1.703346002954575,
+      "learning_rate": 1.8645220881581144e-07,
+      "loss": 0.6737,
+      "step": 4160
+    },
+    {
+      "epoch": 7.705555555555556,
+      "grad_norm": 1.549629897948105,
+      "learning_rate": 1.8644467181778246e-07,
+      "loss": 0.6653,
+      "step": 4161
+    },
+    {
+      "epoch": 7.707407407407407,
+      "grad_norm": 1.5190009048795332,
+      "learning_rate": 1.8643713287624247e-07,
+      "loss": 0.6204,
+      "step": 4162
+    },
+    {
+      "epoch": 7.709259259259259,
+      "grad_norm": 2.102605853425928,
+      "learning_rate": 1.86429591991361e-07,
+      "loss": 0.6703,
+      "step": 4163
+    },
+    {
+      "epoch": 7.711111111111111,
+      "grad_norm": 2.224338451369923,
+      "learning_rate": 1.8642204916330755e-07,
+      "loss": 0.6727,
+      "step": 4164
+    },
+    {
+      "epoch": 7.712962962962963,
+      "grad_norm": 1.446373386823586,
+      "learning_rate": 1.8641450439225169e-07,
+      "loss": 0.6802,
+      "step": 4165
+    },
+    {
+      "epoch": 7.714814814814815,
+      "grad_norm": 1.437228662435052,
+      "learning_rate": 1.864069576783631e-07,
+      "loss": 0.6529,
+      "step": 4166
+    },
+    {
+      "epoch": 7.716666666666667,
+      "grad_norm": 1.6400586302159657,
+      "learning_rate": 1.8639940902181142e-07,
+      "loss": 0.6227,
+      "step": 4167
+    },
+    {
+      "epoch": 7.718518518518518,
+      "grad_norm": 1.4350685004590364,
+      "learning_rate": 1.8639185842276635e-07,
+      "loss": 0.66,
+      "step": 4168
+    },
+    {
+      "epoch": 7.72037037037037,
+      "grad_norm": 1.4091877959516466,
+      "learning_rate": 1.863843058813977e-07,
+      "loss": 0.6844,
+      "step": 4169
+    },
+    {
+      "epoch": 7.722222222222222,
+      "grad_norm": 1.362957417411261,
+      "learning_rate": 1.863767513978752e-07,
+      "loss": 0.6774,
+      "step": 4170
+    },
+    {
+      "epoch": 7.724074074074074,
+      "grad_norm": 1.9436180705373733,
+      "learning_rate": 1.8636919497236876e-07,
+      "loss": 0.6374,
+      "step": 4171
+    },
+    {
+      "epoch": 7.725925925925926,
+      "grad_norm": 1.530064974317294,
+      "learning_rate": 1.8636163660504822e-07,
+      "loss": 0.6205,
+      "step": 4172
+    },
+    {
+      "epoch": 7.727777777777778,
+      "grad_norm": 3.3478188353432223,
+      "learning_rate": 1.8635407629608358e-07,
+      "loss": 0.6163,
+      "step": 4173
+    },
+    {
+      "epoch": 7.729629629629629,
+      "grad_norm": 1.537655692465855,
+      "learning_rate": 1.8634651404564473e-07,
+      "loss": 0.6487,
+      "step": 4174
+    },
+    {
+      "epoch": 7.731481481481482,
+      "grad_norm": 1.4478956761419441,
+      "learning_rate": 1.8633894985390173e-07,
+      "loss": 0.6518,
+      "step": 4175
+    },
+    {
+      "epoch": 7.733333333333333,
+      "grad_norm": 1.5665318667251331,
+      "learning_rate": 1.8633138372102466e-07,
+      "loss": 0.6714,
+      "step": 4176
+    },
+    {
+      "epoch": 7.735185185185185,
+      "grad_norm": 1.561653964888794,
+      "learning_rate": 1.863238156471836e-07,
+      "loss": 0.648,
+      "step": 4177
+    },
+    {
+      "epoch": 7.737037037037037,
+      "grad_norm": 1.6655212599700395,
+      "learning_rate": 1.863162456325487e-07,
+      "loss": 0.6267,
+      "step": 4178
+    },
+    {
+      "epoch": 7.738888888888889,
+      "grad_norm": 1.8084827258756873,
+      "learning_rate": 1.863086736772902e-07,
+      "loss": 0.6411,
+      "step": 4179
+    },
+    {
+      "epoch": 7.7407407407407405,
+      "grad_norm": 1.59021261787782,
+      "learning_rate": 1.8630109978157827e-07,
+      "loss": 0.6472,
+      "step": 4180
+    },
+    {
+      "epoch": 7.742592592592593,
+      "grad_norm": 1.7005023157939847,
+      "learning_rate": 1.8629352394558328e-07,
+      "loss": 0.6206,
+      "step": 4181
+    },
+    {
+      "epoch": 7.7444444444444445,
+      "grad_norm": 1.5556903789202785,
+      "learning_rate": 1.8628594616947546e-07,
+      "loss": 0.6476,
+      "step": 4182
+    },
+    {
+      "epoch": 7.746296296296296,
+      "grad_norm": 1.6136594489433103,
+      "learning_rate": 1.8627836645342523e-07,
+      "loss": 0.6869,
+      "step": 4183
+    },
+    {
+      "epoch": 7.7481481481481485,
+      "grad_norm": 1.8900510714546466,
+      "learning_rate": 1.86270784797603e-07,
+      "loss": 0.6542,
+      "step": 4184
+    },
+    {
+      "epoch": 7.75,
+      "grad_norm": 1.593568532823711,
+      "learning_rate": 1.862632012021792e-07,
+      "loss": 0.6513,
+      "step": 4185
+    },
+    {
+      "epoch": 7.7518518518518515,
+      "grad_norm": 1.4748305020417718,
+      "learning_rate": 1.8625561566732438e-07,
+      "loss": 0.6759,
+      "step": 4186
+    },
+    {
+      "epoch": 7.753703703703704,
+      "grad_norm": 1.6213285537756228,
+      "learning_rate": 1.8624802819320902e-07,
+      "loss": 0.672,
+      "step": 4187
+    },
+    {
+      "epoch": 7.7555555555555555,
+      "grad_norm": 1.403214412600631,
+      "learning_rate": 1.8624043878000374e-07,
+      "loss": 0.6347,
+      "step": 4188
+    },
+    {
+      "epoch": 7.757407407407407,
+      "grad_norm": 1.7317484034809942,
+      "learning_rate": 1.862328474278792e-07,
+      "loss": 0.6617,
+      "step": 4189
+    },
+    {
+      "epoch": 7.7592592592592595,
+      "grad_norm": 1.4312230393163905,
+      "learning_rate": 1.8622525413700603e-07,
+      "loss": 0.639,
+      "step": 4190
+    },
+    {
+      "epoch": 7.761111111111111,
+      "grad_norm": 1.5467294525945965,
+      "learning_rate": 1.8621765890755497e-07,
+      "loss": 0.6641,
+      "step": 4191
+    },
+    {
+      "epoch": 7.762962962962963,
+      "grad_norm": 1.7597578980312507,
+      "learning_rate": 1.8621006173969675e-07,
+      "loss": 0.6333,
+      "step": 4192
+    },
+    {
+      "epoch": 7.764814814814815,
+      "grad_norm": 1.5385857974725132,
+      "learning_rate": 1.8620246263360224e-07,
+      "loss": 0.6303,
+      "step": 4193
+    },
+    {
+      "epoch": 7.766666666666667,
+      "grad_norm": 1.5603199575634885,
+      "learning_rate": 1.861948615894422e-07,
+      "loss": 0.6104,
+      "step": 4194
+    },
+    {
+      "epoch": 7.768518518518518,
+      "grad_norm": 1.7077788971930847,
+      "learning_rate": 1.8618725860738761e-07,
+      "loss": 0.616,
+      "step": 4195
+    },
+    {
+      "epoch": 7.770370370370371,
+      "grad_norm": 1.4075230011896642,
+      "learning_rate": 1.8617965368760936e-07,
+      "loss": 0.6915,
+      "step": 4196
+    },
+    {
+      "epoch": 7.772222222222222,
+      "grad_norm": 1.429929144872628,
+      "learning_rate": 1.861720468302784e-07,
+      "loss": 0.633,
+      "step": 4197
+    },
+    {
+      "epoch": 7.774074074074074,
+      "grad_norm": 1.6918948829595555,
+      "learning_rate": 1.8616443803556583e-07,
+      "loss": 0.6243,
+      "step": 4198
+    },
+    {
+      "epoch": 7.775925925925926,
+      "grad_norm": 1.4922432742782052,
+      "learning_rate": 1.8615682730364265e-07,
+      "loss": 0.6492,
+      "step": 4199
+    },
+    {
+      "epoch": 7.777777777777778,
+      "grad_norm": 1.4295920482476536,
+      "learning_rate": 1.8614921463468e-07,
+      "loss": 0.6366,
+      "step": 4200
+    },
+    {
+      "epoch": 7.77962962962963,
+      "grad_norm": 1.5561707759060341,
+      "learning_rate": 1.8614160002884902e-07,
+      "loss": 0.6569,
+      "step": 4201
+    },
+    {
+      "epoch": 7.781481481481482,
+      "grad_norm": 1.5595983391384,
+      "learning_rate": 1.8613398348632093e-07,
+      "loss": 0.6683,
+      "step": 4202
+    },
+    {
+      "epoch": 7.783333333333333,
+      "grad_norm": 1.453895198698467,
+      "learning_rate": 1.8612636500726694e-07,
+      "loss": 0.6776,
+      "step": 4203
+    },
+    {
+      "epoch": 7.785185185185185,
+      "grad_norm": 1.4813169379502018,
+      "learning_rate": 1.8611874459185837e-07,
+      "loss": 0.6284,
+      "step": 4204
+    },
+    {
+      "epoch": 7.787037037037037,
+      "grad_norm": 2.7805780326579446,
+      "learning_rate": 1.8611112224026652e-07,
+      "loss": 0.6448,
+      "step": 4205
+    },
+    {
+      "epoch": 7.788888888888889,
+      "grad_norm": 1.4077861802135274,
+      "learning_rate": 1.8610349795266273e-07,
+      "loss": 0.6505,
+      "step": 4206
+    },
+    {
+      "epoch": 7.790740740740741,
+      "grad_norm": 1.5090737509015753,
+      "learning_rate": 1.8609587172921854e-07,
+      "loss": 0.6621,
+      "step": 4207
+    },
+    {
+      "epoch": 7.792592592592593,
+      "grad_norm": 1.5741265115904144,
+      "learning_rate": 1.860882435701052e-07,
+      "loss": 0.634,
+      "step": 4208
+    },
+    {
+      "epoch": 7.794444444444444,
+      "grad_norm": 1.5084332309374784,
+      "learning_rate": 1.8608061347549442e-07,
+      "loss": 0.6274,
+      "step": 4209
+    },
+    {
+      "epoch": 7.796296296296296,
+      "grad_norm": 1.622029858928403,
+      "learning_rate": 1.8607298144555763e-07,
+      "loss": 0.5935,
+      "step": 4210
+    },
+    {
+      "epoch": 7.798148148148148,
+      "grad_norm": 1.5772689872271521,
+      "learning_rate": 1.8606534748046645e-07,
+      "loss": 0.6524,
+      "step": 4211
+    },
+    {
+      "epoch": 7.8,
+      "grad_norm": 3.2961944151240847,
+      "learning_rate": 1.8605771158039253e-07,
+      "loss": 0.6624,
+      "step": 4212
+    },
+    {
+      "epoch": 7.801851851851852,
+      "grad_norm": 1.9150620798645273,
+      "learning_rate": 1.8605007374550747e-07,
+      "loss": 0.6264,
+      "step": 4213
+    },
+    {
+      "epoch": 7.803703703703704,
+      "grad_norm": 1.8630508993229984,
+      "learning_rate": 1.860424339759831e-07,
+      "loss": 0.6749,
+      "step": 4214
+    },
+    {
+      "epoch": 7.805555555555555,
+      "grad_norm": 1.5282530718257328,
+      "learning_rate": 1.860347922719911e-07,
+      "loss": 0.6011,
+      "step": 4215
+    },
+    {
+      "epoch": 7.807407407407408,
+      "grad_norm": 1.712442421556992,
+      "learning_rate": 1.860271486337033e-07,
+      "loss": 0.6377,
+      "step": 4216
+    },
+    {
+      "epoch": 7.809259259259259,
+      "grad_norm": 2.4504626785639467,
+      "learning_rate": 1.8601950306129156e-07,
+      "loss": 0.6016,
+      "step": 4217
+    },
+    {
+      "epoch": 7.811111111111111,
+      "grad_norm": 1.7471682745128487,
+      "learning_rate": 1.8601185555492775e-07,
+      "loss": 0.6012,
+      "step": 4218
+    },
+    {
+      "epoch": 7.812962962962963,
+      "grad_norm": 1.5842894188431227,
+      "learning_rate": 1.8600420611478388e-07,
+      "loss": 0.666,
+      "step": 4219
+    },
+    {
+      "epoch": 7.814814814814815,
+      "grad_norm": 1.6603301316866073,
+      "learning_rate": 1.8599655474103182e-07,
+      "loss": 0.6541,
+      "step": 4220
+    },
+    {
+      "epoch": 7.816666666666666,
+      "grad_norm": 1.8708674570862722,
+      "learning_rate": 1.8598890143384366e-07,
+      "loss": 0.6478,
+      "step": 4221
+    },
+    {
+      "epoch": 7.818518518518519,
+      "grad_norm": 1.368562142897557,
+      "learning_rate": 1.8598124619339143e-07,
+      "loss": 0.6239,
+      "step": 4222
+    },
+    {
+      "epoch": 7.82037037037037,
+      "grad_norm": 1.4363596719162366,
+      "learning_rate": 1.8597358901984727e-07,
+      "loss": 0.6646,
+      "step": 4223
+    },
+    {
+      "epoch": 7.822222222222222,
+      "grad_norm": 1.775958451072795,
+      "learning_rate": 1.8596592991338333e-07,
+      "loss": 0.6412,
+      "step": 4224
+    },
+    {
+      "epoch": 7.824074074074074,
+      "grad_norm": 1.7252118115311468,
+      "learning_rate": 1.8595826887417183e-07,
+      "loss": 0.6202,
+      "step": 4225
+    },
+    {
+      "epoch": 7.825925925925926,
+      "grad_norm": 1.6836867453631728,
+      "learning_rate": 1.8595060590238498e-07,
+      "loss": 0.6523,
+      "step": 4226
+    },
+    {
+      "epoch": 7.8277777777777775,
+      "grad_norm": 1.4806842244711003,
+      "learning_rate": 1.8594294099819506e-07,
+      "loss": 0.6228,
+      "step": 4227
+    },
+    {
+      "epoch": 7.82962962962963,
+      "grad_norm": 3.662140315928878,
+      "learning_rate": 1.8593527416177442e-07,
+      "loss": 0.6542,
+      "step": 4228
+    },
+    {
+      "epoch": 7.8314814814814815,
+      "grad_norm": 1.7823423846598296,
+      "learning_rate": 1.8592760539329542e-07,
+      "loss": 0.6491,
+      "step": 4229
+    },
+    {
+      "epoch": 7.833333333333333,
+      "grad_norm": 2.4833834280045037,
+      "learning_rate": 1.8591993469293046e-07,
+      "loss": 0.6531,
+      "step": 4230
+    },
+    {
+      "epoch": 7.8351851851851855,
+      "grad_norm": 1.5381319220752652,
+      "learning_rate": 1.85912262060852e-07,
+      "loss": 0.6186,
+      "step": 4231
+    },
+    {
+      "epoch": 7.837037037037037,
+      "grad_norm": 1.5617546543737226,
+      "learning_rate": 1.8590458749723262e-07,
+      "loss": 0.5944,
+      "step": 4232
+    },
+    {
+      "epoch": 7.838888888888889,
+      "grad_norm": 1.7299643375012799,
+      "learning_rate": 1.8589691100224475e-07,
+      "loss": 0.6389,
+      "step": 4233
+    },
+    {
+      "epoch": 7.840740740740741,
+      "grad_norm": 1.5588826698334022,
+      "learning_rate": 1.8588923257606104e-07,
+      "loss": 0.6073,
+      "step": 4234
+    },
+    {
+      "epoch": 7.842592592592593,
+      "grad_norm": 1.707479621324081,
+      "learning_rate": 1.8588155221885408e-07,
+      "loss": 0.6309,
+      "step": 4235
+    },
+    {
+      "epoch": 7.844444444444444,
+      "grad_norm": 1.4316085111407069,
+      "learning_rate": 1.858738699307966e-07,
+      "loss": 0.6845,
+      "step": 4236
+    },
+    {
+      "epoch": 7.8462962962962965,
+      "grad_norm": 1.458372035120529,
+      "learning_rate": 1.8586618571206133e-07,
+      "loss": 0.6422,
+      "step": 4237
+    },
+    {
+      "epoch": 7.848148148148148,
+      "grad_norm": 2.0739420093556915,
+      "learning_rate": 1.8585849956282095e-07,
+      "loss": 0.6129,
+      "step": 4238
+    },
+    {
+      "epoch": 7.85,
+      "grad_norm": 1.7042284260106428,
+      "learning_rate": 1.858508114832483e-07,
+      "loss": 0.6199,
+      "step": 4239
+    },
+    {
+      "epoch": 7.851851851851852,
+      "grad_norm": 1.7217152096944712,
+      "learning_rate": 1.858431214735163e-07,
+      "loss": 0.6698,
+      "step": 4240
+    },
+    {
+      "epoch": 7.853703703703704,
+      "grad_norm": 1.5269613222959182,
+      "learning_rate": 1.8583542953379772e-07,
+      "loss": 0.6076,
+      "step": 4241
+    },
+    {
+      "epoch": 7.855555555555555,
+      "grad_norm": 1.6024419616940813,
+      "learning_rate": 1.8582773566426558e-07,
+      "loss": 0.6239,
+      "step": 4242
+    },
+    {
+      "epoch": 7.857407407407408,
+      "grad_norm": 1.792100159409612,
+      "learning_rate": 1.8582003986509282e-07,
+      "loss": 0.6887,
+      "step": 4243
+    },
+    {
+      "epoch": 7.859259259259259,
+      "grad_norm": 1.5524411399251183,
+      "learning_rate": 1.8581234213645253e-07,
+      "loss": 0.6009,
+      "step": 4244
+    },
+    {
+      "epoch": 7.861111111111111,
+      "grad_norm": 1.7138923463648534,
+      "learning_rate": 1.858046424785177e-07,
+      "loss": 0.6131,
+      "step": 4245
+    },
+    {
+      "epoch": 7.862962962962963,
+      "grad_norm": 1.6579334318768093,
+      "learning_rate": 1.8579694089146144e-07,
+      "loss": 0.6829,
+      "step": 4246
+    },
+    {
+      "epoch": 7.864814814814815,
+      "grad_norm": 1.7584239444790144,
+      "learning_rate": 1.8578923737545695e-07,
+      "loss": 0.6217,
+      "step": 4247
+    },
+    {
+      "epoch": 7.866666666666667,
+      "grad_norm": 1.658315787124268,
+      "learning_rate": 1.8578153193067743e-07,
+      "loss": 0.6696,
+      "step": 4248
+    },
+    {
+      "epoch": 7.868518518518519,
+      "grad_norm": 1.7134472877783422,
+      "learning_rate": 1.8577382455729604e-07,
+      "loss": 0.6494,
+      "step": 4249
+    },
+    {
+      "epoch": 7.87037037037037,
+      "grad_norm": 1.624983905525824,
+      "learning_rate": 1.8576611525548614e-07,
+      "loss": 0.6346,
+      "step": 4250
+    },
+    {
+      "epoch": 7.872222222222222,
+      "grad_norm": 8.312117548713308,
+      "learning_rate": 1.8575840402542107e-07,
+      "loss": 0.6388,
+      "step": 4251
+    },
+    {
+      "epoch": 7.874074074074074,
+      "grad_norm": 1.5945072361428139,
+      "learning_rate": 1.8575069086727412e-07,
+      "loss": 0.613,
+      "step": 4252
+    },
+    {
+      "epoch": 7.875925925925926,
+      "grad_norm": 1.5040493394371637,
+      "learning_rate": 1.8574297578121874e-07,
+      "loss": 0.6301,
+      "step": 4253
+    },
+    {
+      "epoch": 7.877777777777778,
+      "grad_norm": 1.3848543515523097,
+      "learning_rate": 1.857352587674284e-07,
+      "loss": 0.6116,
+      "step": 4254
+    },
+    {
+      "epoch": 7.87962962962963,
+      "grad_norm": 3.4059485554226328,
+      "learning_rate": 1.8572753982607661e-07,
+      "loss": 0.621,
+      "step": 4255
+    },
+    {
+      "epoch": 7.881481481481481,
+      "grad_norm": 1.380778060377783,
+      "learning_rate": 1.8571981895733686e-07,
+      "loss": 0.6302,
+      "step": 4256
+    },
+    {
+      "epoch": 7.883333333333333,
+      "grad_norm": 2.5197301367682257,
+      "learning_rate": 1.8571209616138277e-07,
+      "loss": 0.6529,
+      "step": 4257
+    },
+    {
+      "epoch": 7.885185185185185,
+      "grad_norm": 1.394644740304966,
+      "learning_rate": 1.8570437143838797e-07,
+      "loss": 0.6744,
+      "step": 4258
+    },
+    {
+      "epoch": 7.887037037037037,
+      "grad_norm": 1.6460818344196817,
+      "learning_rate": 1.8569664478852616e-07,
+      "loss": 0.6105,
+      "step": 4259
+    },
+    {
+      "epoch": 7.888888888888889,
+      "grad_norm": 1.5927364343792982,
+      "learning_rate": 1.85688916211971e-07,
+      "loss": 0.6865,
+      "step": 4260
+    },
+    {
+      "epoch": 7.890740740740741,
+      "grad_norm": 2.0963265659207875,
+      "learning_rate": 1.856811857088963e-07,
+      "loss": 0.6509,
+      "step": 4261
+    },
+    {
+      "epoch": 7.892592592592592,
+      "grad_norm": 1.9794374700612483,
+      "learning_rate": 1.856734532794758e-07,
+      "loss": 0.6298,
+      "step": 4262
+    },
+    {
+      "epoch": 7.894444444444445,
+      "grad_norm": 1.5525811133556973,
+      "learning_rate": 1.856657189238834e-07,
+      "loss": 0.5925,
+      "step": 4263
+    },
+    {
+      "epoch": 7.896296296296296,
+      "grad_norm": 2.0709331003014464,
+      "learning_rate": 1.85657982642293e-07,
+      "loss": 0.6437,
+      "step": 4264
+    },
+    {
+      "epoch": 7.898148148148148,
+      "grad_norm": 1.6976029855130077,
+      "learning_rate": 1.8565024443487848e-07,
+      "loss": 0.6645,
+      "step": 4265
+    },
+    {
+      "epoch": 7.9,
+      "grad_norm": 1.5529385899318902,
+      "learning_rate": 1.8564250430181384e-07,
+      "loss": 0.6513,
+      "step": 4266
+    },
+    {
+      "epoch": 7.901851851851852,
+      "grad_norm": 1.4414187730073065,
+      "learning_rate": 1.856347622432731e-07,
+      "loss": 0.6116,
+      "step": 4267
+    },
+    {
+      "epoch": 7.9037037037037035,
+      "grad_norm": 1.7384404136145046,
+      "learning_rate": 1.8562701825943035e-07,
+      "loss": 0.6572,
+      "step": 4268
+    },
+    {
+      "epoch": 7.905555555555556,
+      "grad_norm": 1.5401282393535152,
+      "learning_rate": 1.8561927235045964e-07,
+      "loss": 0.6364,
+      "step": 4269
+    },
+    {
+      "epoch": 7.907407407407407,
+      "grad_norm": 1.8818522707456258,
+      "learning_rate": 1.8561152451653517e-07,
+      "loss": 0.6609,
+      "step": 4270
+    },
+    {
+      "epoch": 7.909259259259259,
+      "grad_norm": 2.0577859345367076,
+      "learning_rate": 1.856037747578311e-07,
+      "loss": 0.6427,
+      "step": 4271
+    },
+    {
+      "epoch": 7.911111111111111,
+      "grad_norm": 1.6308335595786876,
+      "learning_rate": 1.855960230745217e-07,
+      "loss": 0.6258,
+      "step": 4272
+    },
+    {
+      "epoch": 7.912962962962963,
+      "grad_norm": 1.5851849895190326,
+      "learning_rate": 1.855882694667812e-07,
+      "loss": 0.654,
+      "step": 4273
+    },
+    {
+      "epoch": 7.9148148148148145,
+      "grad_norm": 1.4676589101905246,
+      "learning_rate": 1.8558051393478395e-07,
+      "loss": 0.632,
+      "step": 4274
+    },
+    {
+      "epoch": 7.916666666666667,
+      "grad_norm": 1.4516158575923812,
+      "learning_rate": 1.8557275647870433e-07,
+      "loss": 0.6338,
+      "step": 4275
+    },
+    {
+      "epoch": 7.9185185185185185,
+      "grad_norm": 1.6025343622657053,
+      "learning_rate": 1.855649970987167e-07,
+      "loss": 0.6355,
+      "step": 4276
+    },
+    {
+      "epoch": 7.92037037037037,
+      "grad_norm": 1.7740244786284893,
+      "learning_rate": 1.855572357949956e-07,
+      "loss": 0.6474,
+      "step": 4277
+    },
+    {
+      "epoch": 7.9222222222222225,
+      "grad_norm": 1.6227287514844153,
+      "learning_rate": 1.855494725677154e-07,
+      "loss": 0.6409,
+      "step": 4278
+    },
+    {
+      "epoch": 7.924074074074074,
+      "grad_norm": 1.5908595654365214,
+      "learning_rate": 1.8554170741705076e-07,
+      "loss": 0.6613,
+      "step": 4279
+    },
+    {
+      "epoch": 7.925925925925926,
+      "grad_norm": 1.567252608698113,
+      "learning_rate": 1.8553394034317619e-07,
+      "loss": 0.6723,
+      "step": 4280
+    },
+    {
+      "epoch": 7.927777777777778,
+      "grad_norm": 1.453275447335963,
+      "learning_rate": 1.8552617134626633e-07,
+      "loss": 0.6268,
+      "step": 4281
+    },
+    {
+      "epoch": 7.92962962962963,
+      "grad_norm": 1.4753947705197545,
+      "learning_rate": 1.8551840042649586e-07,
+      "loss": 0.6487,
+      "step": 4282
+    },
+    {
+      "epoch": 7.931481481481481,
+      "grad_norm": 1.5814961888000902,
+      "learning_rate": 1.855106275840395e-07,
+      "loss": 0.648,
+      "step": 4283
+    },
+    {
+      "epoch": 7.933333333333334,
+      "grad_norm": 1.45632632111028,
+      "learning_rate": 1.8550285281907198e-07,
+      "loss": 0.6605,
+      "step": 4284
+    },
+    {
+      "epoch": 7.935185185185185,
+      "grad_norm": 1.6794029166643385,
+      "learning_rate": 1.854950761317681e-07,
+      "loss": 0.6601,
+      "step": 4285
+    },
+    {
+      "epoch": 7.937037037037037,
+      "grad_norm": 1.395438030656467,
+      "learning_rate": 1.854872975223027e-07,
+      "loss": 0.598,
+      "step": 4286
+    },
+    {
+      "epoch": 7.938888888888889,
+      "grad_norm": 1.238066686594357,
+      "learning_rate": 1.8547951699085068e-07,
+      "loss": 0.608,
+      "step": 4287
+    },
+    {
+      "epoch": 7.940740740740741,
+      "grad_norm": 1.6229941168801458,
+      "learning_rate": 1.8547173453758695e-07,
+      "loss": 0.6941,
+      "step": 4288
+    },
+    {
+      "epoch": 7.942592592592592,
+      "grad_norm": 1.5292300190384138,
+      "learning_rate": 1.854639501626865e-07,
+      "loss": 0.631,
+      "step": 4289
+    },
+    {
+      "epoch": 7.944444444444445,
+      "grad_norm": 1.4496658060766403,
+      "learning_rate": 1.8545616386632433e-07,
+      "loss": 0.6459,
+      "step": 4290
+    },
+    {
+      "epoch": 7.946296296296296,
+      "grad_norm": 1.6171495851246431,
+      "learning_rate": 1.854483756486755e-07,
+      "loss": 0.6509,
+      "step": 4291
+    },
+    {
+      "epoch": 7.948148148148148,
+      "grad_norm": 1.6253871821133663,
+      "learning_rate": 1.8544058550991513e-07,
+      "loss": 0.6456,
+      "step": 4292
+    },
+    {
+      "epoch": 7.95,
+      "grad_norm": 1.3464307100900252,
+      "learning_rate": 1.8543279345021832e-07,
+      "loss": 0.5869,
+      "step": 4293
+    },
+    {
+      "epoch": 7.951851851851852,
+      "grad_norm": 1.7335300007136392,
+      "learning_rate": 1.8542499946976027e-07,
+      "loss": 0.6586,
+      "step": 4294
+    },
+    {
+      "epoch": 7.953703703703704,
+      "grad_norm": 2.6870364919059506,
+      "learning_rate": 1.8541720356871626e-07,
+      "loss": 0.6322,
+      "step": 4295
+    },
+    {
+      "epoch": 7.955555555555556,
+      "grad_norm": 1.5670737690153551,
+      "learning_rate": 1.854094057472615e-07,
+      "loss": 0.6173,
+      "step": 4296
+    },
+    {
+      "epoch": 7.957407407407407,
+      "grad_norm": 1.3526343505502112,
+      "learning_rate": 1.8540160600557134e-07,
+      "loss": 0.6194,
+      "step": 4297
+    },
+    {
+      "epoch": 7.959259259259259,
+      "grad_norm": 1.448531816474373,
+      "learning_rate": 1.853938043438211e-07,
+      "loss": 0.664,
+      "step": 4298
+    },
+    {
+      "epoch": 7.961111111111111,
+      "grad_norm": 2.4246803155012606,
+      "learning_rate": 1.8538600076218626e-07,
+      "loss": 0.6319,
+      "step": 4299
+    },
+    {
+      "epoch": 7.962962962962963,
+      "grad_norm": 1.7017338196614473,
+      "learning_rate": 1.853781952608422e-07,
+      "loss": 0.6544,
+      "step": 4300
+    },
+    {
+      "epoch": 7.964814814814815,
+      "grad_norm": 1.4700864334407744,
+      "learning_rate": 1.853703878399644e-07,
+      "loss": 0.6532,
+      "step": 4301
+    },
+    {
+      "epoch": 7.966666666666667,
+      "grad_norm": 1.6545321919887692,
+      "learning_rate": 1.8536257849972844e-07,
+      "loss": 0.6462,
+      "step": 4302
+    },
+    {
+      "epoch": 7.968518518518518,
+      "grad_norm": 1.565082343945817,
+      "learning_rate": 1.8535476724030987e-07,
+      "loss": 0.6448,
+      "step": 4303
+    },
+    {
+      "epoch": 7.97037037037037,
+      "grad_norm": 1.6067404281562736,
+      "learning_rate": 1.8534695406188435e-07,
+      "loss": 0.6263,
+      "step": 4304
+    },
+    {
+      "epoch": 7.972222222222222,
+      "grad_norm": 1.4730438201168865,
+      "learning_rate": 1.8533913896462745e-07,
+      "loss": 0.6265,
+      "step": 4305
+    },
+    {
+      "epoch": 7.974074074074074,
+      "grad_norm": 1.5164473424735692,
+      "learning_rate": 1.8533132194871497e-07,
+      "loss": 0.6417,
+      "step": 4306
+    },
+    {
+      "epoch": 7.975925925925926,
+      "grad_norm": 1.7426257340424622,
+      "learning_rate": 1.8532350301432258e-07,
+      "loss": 0.6304,
+      "step": 4307
+    },
+    {
+      "epoch": 7.977777777777778,
+      "grad_norm": 1.9292112457906772,
+      "learning_rate": 1.8531568216162614e-07,
+      "loss": 0.6251,
+      "step": 4308
+    },
+    {
+      "epoch": 7.979629629629629,
+      "grad_norm": 1.8134625828924371,
+      "learning_rate": 1.853078593908014e-07,
+      "loss": 0.6264,
+      "step": 4309
+    },
+    {
+      "epoch": 7.981481481481482,
+      "grad_norm": 1.6079926089118437,
+      "learning_rate": 1.8530003470202434e-07,
+      "loss": 0.6583,
+      "step": 4310
+    },
+    {
+      "epoch": 7.983333333333333,
+      "grad_norm": 1.7299718911925903,
+      "learning_rate": 1.852922080954708e-07,
+      "loss": 0.6604,
+      "step": 4311
+    },
+    {
+      "epoch": 7.985185185185185,
+      "grad_norm": 1.7540557514425157,
+      "learning_rate": 1.852843795713168e-07,
+      "loss": 0.7104,
+      "step": 4312
+    },
+    {
+      "epoch": 7.987037037037037,
+      "grad_norm": 1.5018343152050084,
+      "learning_rate": 1.852765491297383e-07,
+      "loss": 0.6364,
+      "step": 4313
+    },
+    {
+      "epoch": 7.988888888888889,
+      "grad_norm": 1.5380650953221953,
+      "learning_rate": 1.8526871677091136e-07,
+      "loss": 0.6474,
+      "step": 4314
+    },
+    {
+      "epoch": 7.9907407407407405,
+      "grad_norm": 1.879917967560461,
+      "learning_rate": 1.8526088249501212e-07,
+      "loss": 0.6191,
+      "step": 4315
+    },
+    {
+      "epoch": 7.992592592592593,
+      "grad_norm": 1.6924644344447572,
+      "learning_rate": 1.852530463022166e-07,
+      "loss": 0.6649,
+      "step": 4316
+    },
+    {
+      "epoch": 7.9944444444444445,
+      "grad_norm": 1.5033886068092717,
+      "learning_rate": 1.8524520819270114e-07,
+      "loss": 0.6652,
+      "step": 4317
+    },
+    {
+      "epoch": 7.996296296296296,
+      "grad_norm": 1.3640533603143965,
+      "learning_rate": 1.8523736816664184e-07,
+      "loss": 0.6334,
+      "step": 4318
+    },
+    {
+      "epoch": 7.9981481481481485,
+      "grad_norm": 1.4770085813986176,
+      "learning_rate": 1.85229526224215e-07,
+      "loss": 0.6362,
+      "step": 4319
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 1.6997209249669618,
+      "learning_rate": 1.8522168236559692e-07,
+      "loss": 0.6934,
+      "step": 4320
+    },
+    {
+      "epoch": 8.001851851851852,
+      "grad_norm": 1.6463305880589907,
+      "learning_rate": 1.85213836590964e-07,
+      "loss": 0.6503,
+      "step": 4321
+    },
+    {
+      "epoch": 8.003703703703703,
+      "grad_norm": 1.8935318491018889,
+      "learning_rate": 1.8520598890049254e-07,
+      "loss": 0.6494,
+      "step": 4322
+    },
+    {
+      "epoch": 8.005555555555556,
+      "grad_norm": 1.66853566427282,
+      "learning_rate": 1.8519813929435908e-07,
+      "loss": 0.6382,
+      "step": 4323
+    },
+    {
+      "epoch": 8.007407407407408,
+      "grad_norm": 1.7978058810590847,
+      "learning_rate": 1.8519028777274005e-07,
+      "loss": 0.6659,
+      "step": 4324
+    },
+    {
+      "epoch": 8.00925925925926,
+      "grad_norm": 1.320906984219989,
+      "learning_rate": 1.8518243433581194e-07,
+      "loss": 0.6796,
+      "step": 4325
+    },
+    {
+      "epoch": 8.011111111111111,
+      "grad_norm": 1.5965518375485208,
+      "learning_rate": 1.8517457898375136e-07,
+      "loss": 0.6353,
+      "step": 4326
+    },
+    {
+      "epoch": 8.012962962962963,
+      "grad_norm": 1.5946415135397438,
+      "learning_rate": 1.8516672171673495e-07,
+      "loss": 0.6317,
+      "step": 4327
+    },
+    {
+      "epoch": 8.014814814814814,
+      "grad_norm": 1.6117001839716567,
+      "learning_rate": 1.8515886253493932e-07,
+      "loss": 0.6004,
+      "step": 4328
+    },
+    {
+      "epoch": 8.016666666666667,
+      "grad_norm": 1.5769063563174675,
+      "learning_rate": 1.8515100143854116e-07,
+      "loss": 0.7085,
+      "step": 4329
+    },
+    {
+      "epoch": 8.018518518518519,
+      "grad_norm": 1.5131430070108751,
+      "learning_rate": 1.851431384277172e-07,
+      "loss": 0.6534,
+      "step": 4330
+    },
+    {
+      "epoch": 8.02037037037037,
+      "grad_norm": 1.3425115448660485,
+      "learning_rate": 1.851352735026443e-07,
+      "loss": 0.6744,
+      "step": 4331
+    },
+    {
+      "epoch": 8.022222222222222,
+      "grad_norm": 1.649345172673442,
+      "learning_rate": 1.8512740666349918e-07,
+      "loss": 0.6335,
+      "step": 4332
+    },
+    {
+      "epoch": 8.024074074074074,
+      "grad_norm": 1.768822983071256,
+      "learning_rate": 1.8511953791045877e-07,
+      "loss": 0.6469,
+      "step": 4333
+    },
+    {
+      "epoch": 8.025925925925925,
+      "grad_norm": 1.7247692819257618,
+      "learning_rate": 1.8511166724369996e-07,
+      "loss": 0.6824,
+      "step": 4334
+    },
+    {
+      "epoch": 8.027777777777779,
+      "grad_norm": 1.5123442165031116,
+      "learning_rate": 1.851037946633997e-07,
+      "loss": 0.6182,
+      "step": 4335
+    },
+    {
+      "epoch": 8.02962962962963,
+      "grad_norm": 1.4697323300137064,
+      "learning_rate": 1.8509592016973503e-07,
+      "loss": 0.7144,
+      "step": 4336
+    },
+    {
+      "epoch": 8.031481481481482,
+      "grad_norm": 1.8044024070931437,
+      "learning_rate": 1.8508804376288297e-07,
+      "loss": 0.6558,
+      "step": 4337
+    },
+    {
+      "epoch": 8.033333333333333,
+      "grad_norm": 1.665857768710196,
+      "learning_rate": 1.8508016544302055e-07,
+      "loss": 0.6221,
+      "step": 4338
+    },
+    {
+      "epoch": 8.035185185185185,
+      "grad_norm": 1.5675589029891257,
+      "learning_rate": 1.8507228521032493e-07,
+      "loss": 0.6831,
+      "step": 4339
+    },
+    {
+      "epoch": 8.037037037037036,
+      "grad_norm": 1.8239402150067279,
+      "learning_rate": 1.8506440306497333e-07,
+      "loss": 0.646,
+      "step": 4340
+    },
+    {
+      "epoch": 8.03888888888889,
+      "grad_norm": 3.501799224046458,
+      "learning_rate": 1.8505651900714289e-07,
+      "loss": 0.7225,
+      "step": 4341
+    },
+    {
+      "epoch": 8.040740740740741,
+      "grad_norm": 1.5963086775963384,
+      "learning_rate": 1.850486330370109e-07,
+      "loss": 0.6809,
+      "step": 4342
+    },
+    {
+      "epoch": 8.042592592592593,
+      "grad_norm": 1.9526566130968028,
+      "learning_rate": 1.8504074515475464e-07,
+      "loss": 0.6606,
+      "step": 4343
+    },
+    {
+      "epoch": 8.044444444444444,
+      "grad_norm": 1.6290485135049857,
+      "learning_rate": 1.8503285536055145e-07,
+      "loss": 0.6437,
+      "step": 4344
+    },
+    {
+      "epoch": 8.046296296296296,
+      "grad_norm": 1.657413998377113,
+      "learning_rate": 1.8502496365457873e-07,
+      "loss": 0.6435,
+      "step": 4345
+    },
+    {
+      "epoch": 8.048148148148147,
+      "grad_norm": 1.465077365534314,
+      "learning_rate": 1.8501707003701395e-07,
+      "loss": 0.6362,
+      "step": 4346
+    },
+    {
+      "epoch": 8.05,
+      "grad_norm": 2.0106985575683387,
+      "learning_rate": 1.850091745080345e-07,
+      "loss": 0.6334,
+      "step": 4347
+    },
+    {
+      "epoch": 8.051851851851852,
+      "grad_norm": 1.5505533501509736,
+      "learning_rate": 1.850012770678179e-07,
+      "loss": 0.6722,
+      "step": 4348
+    },
+    {
+      "epoch": 8.053703703703704,
+      "grad_norm": 1.6036963048630817,
+      "learning_rate": 1.8499337771654176e-07,
+      "loss": 0.6396,
+      "step": 4349
+    },
+    {
+      "epoch": 8.055555555555555,
+      "grad_norm": 1.6011496495248434,
+      "learning_rate": 1.8498547645438364e-07,
+      "loss": 0.6683,
+      "step": 4350
+    },
+    {
+      "epoch": 8.057407407407407,
+      "grad_norm": 1.3797810921372278,
+      "learning_rate": 1.849775732815212e-07,
+      "loss": 0.6333,
+      "step": 4351
+    },
+    {
+      "epoch": 8.059259259259258,
+      "grad_norm": 1.5533664317223292,
+      "learning_rate": 1.8496966819813213e-07,
+      "loss": 0.6564,
+      "step": 4352
+    },
+    {
+      "epoch": 8.061111111111112,
+      "grad_norm": 1.9275002885753574,
+      "learning_rate": 1.8496176120439412e-07,
+      "loss": 0.6313,
+      "step": 4353
+    },
+    {
+      "epoch": 8.062962962962963,
+      "grad_norm": 1.6328122279662558,
+      "learning_rate": 1.8495385230048499e-07,
+      "loss": 0.6389,
+      "step": 4354
+    },
+    {
+      "epoch": 8.064814814814815,
+      "grad_norm": 1.4991029471086874,
+      "learning_rate": 1.8494594148658252e-07,
+      "loss": 0.6653,
+      "step": 4355
+    },
+    {
+      "epoch": 8.066666666666666,
+      "grad_norm": 1.534128662119345,
+      "learning_rate": 1.8493802876286456e-07,
+      "loss": 0.6774,
+      "step": 4356
+    },
+    {
+      "epoch": 8.068518518518518,
+      "grad_norm": 1.9999009197071358,
+      "learning_rate": 1.8493011412950903e-07,
+      "loss": 0.6065,
+      "step": 4357
+    },
+    {
+      "epoch": 8.07037037037037,
+      "grad_norm": 1.5220124854008472,
+      "learning_rate": 1.8492219758669387e-07,
+      "loss": 0.6619,
+      "step": 4358
+    },
+    {
+      "epoch": 8.072222222222223,
+      "grad_norm": 1.9095483502041495,
+      "learning_rate": 1.8491427913459707e-07,
+      "loss": 0.6492,
+      "step": 4359
+    },
+    {
+      "epoch": 8.074074074074074,
+      "grad_norm": 1.3001501103484336,
+      "learning_rate": 1.8490635877339664e-07,
+      "loss": 0.6673,
+      "step": 4360
+    },
+    {
+      "epoch": 8.075925925925926,
+      "grad_norm": 1.4017387089671602,
+      "learning_rate": 1.8489843650327067e-07,
+      "loss": 0.6746,
+      "step": 4361
+    },
+    {
+      "epoch": 8.077777777777778,
+      "grad_norm": 1.6264662484582797,
+      "learning_rate": 1.848905123243973e-07,
+      "loss": 0.6783,
+      "step": 4362
+    },
+    {
+      "epoch": 8.079629629629629,
+      "grad_norm": 1.4885869832802863,
+      "learning_rate": 1.8488258623695462e-07,
+      "loss": 0.6114,
+      "step": 4363
+    },
+    {
+      "epoch": 8.081481481481482,
+      "grad_norm": 1.6734729503633872,
+      "learning_rate": 1.8487465824112084e-07,
+      "loss": 0.6397,
+      "step": 4364
+    },
+    {
+      "epoch": 8.083333333333334,
+      "grad_norm": 1.3722899036518412,
+      "learning_rate": 1.8486672833707425e-07,
+      "loss": 0.6543,
+      "step": 4365
+    },
+    {
+      "epoch": 8.085185185185185,
+      "grad_norm": 1.4160090447514315,
+      "learning_rate": 1.8485879652499311e-07,
+      "loss": 0.6575,
+      "step": 4366
+    },
+    {
+      "epoch": 8.087037037037037,
+      "grad_norm": 1.638746924366951,
+      "learning_rate": 1.8485086280505576e-07,
+      "loss": 0.6325,
+      "step": 4367
+    },
+    {
+      "epoch": 8.088888888888889,
+      "grad_norm": 2.07357267133277,
+      "learning_rate": 1.848429271774406e-07,
+      "loss": 0.6514,
+      "step": 4368
+    },
+    {
+      "epoch": 8.09074074074074,
+      "grad_norm": 1.388452779040256,
+      "learning_rate": 1.8483498964232593e-07,
+      "loss": 0.6956,
+      "step": 4369
+    },
+    {
+      "epoch": 8.092592592592593,
+      "grad_norm": 2.2146826901188694,
+      "learning_rate": 1.8482705019989034e-07,
+      "loss": 0.6657,
+      "step": 4370
+    },
+    {
+      "epoch": 8.094444444444445,
+      "grad_norm": 1.3897073864275475,
+      "learning_rate": 1.8481910885031224e-07,
+      "loss": 0.6304,
+      "step": 4371
+    },
+    {
+      "epoch": 8.096296296296297,
+      "grad_norm": 1.8588323917875276,
+      "learning_rate": 1.8481116559377024e-07,
+      "loss": 0.6121,
+      "step": 4372
+    },
+    {
+      "epoch": 8.098148148148148,
+      "grad_norm": 1.5909449451188151,
+      "learning_rate": 1.8480322043044287e-07,
+      "loss": 0.6391,
+      "step": 4373
+    },
+    {
+      "epoch": 8.1,
+      "grad_norm": 1.6987182691635367,
+      "learning_rate": 1.8479527336050878e-07,
+      "loss": 0.6344,
+      "step": 4374
+    },
+    {
+      "epoch": 8.101851851851851,
+      "grad_norm": 1.5486197353695765,
+      "learning_rate": 1.8478732438414666e-07,
+      "loss": 0.67,
+      "step": 4375
+    },
+    {
+      "epoch": 8.103703703703705,
+      "grad_norm": 1.3226929125277187,
+      "learning_rate": 1.847793735015352e-07,
+      "loss": 0.6706,
+      "step": 4376
+    },
+    {
+      "epoch": 8.105555555555556,
+      "grad_norm": 1.3837574224367704,
+      "learning_rate": 1.8477142071285316e-07,
+      "loss": 0.6257,
+      "step": 4377
+    },
+    {
+      "epoch": 8.107407407407408,
+      "grad_norm": 1.5315962693672416,
+      "learning_rate": 1.8476346601827936e-07,
+      "loss": 0.6336,
+      "step": 4378
+    },
+    {
+      "epoch": 8.10925925925926,
+      "grad_norm": 1.4767475905610052,
+      "learning_rate": 1.8475550941799264e-07,
+      "loss": 0.652,
+      "step": 4379
+    },
+    {
+      "epoch": 8.11111111111111,
+      "grad_norm": 1.5308153588879692,
+      "learning_rate": 1.8474755091217184e-07,
+      "loss": 0.6577,
+      "step": 4380
+    },
+    {
+      "epoch": 8.112962962962962,
+      "grad_norm": 1.2693184323832511,
+      "learning_rate": 1.8473959050099595e-07,
+      "loss": 0.6165,
+      "step": 4381
+    },
+    {
+      "epoch": 8.114814814814816,
+      "grad_norm": 1.3508101952097766,
+      "learning_rate": 1.847316281846439e-07,
+      "loss": 0.6329,
+      "step": 4382
+    },
+    {
+      "epoch": 8.116666666666667,
+      "grad_norm": 1.5248127101440063,
+      "learning_rate": 1.8472366396329476e-07,
+      "loss": 0.6677,
+      "step": 4383
+    },
+    {
+      "epoch": 8.118518518518519,
+      "grad_norm": 1.4675346734606156,
+      "learning_rate": 1.8471569783712752e-07,
+      "loss": 0.6897,
+      "step": 4384
+    },
+    {
+      "epoch": 8.12037037037037,
+      "grad_norm": 1.452606606369142,
+      "learning_rate": 1.847077298063213e-07,
+      "loss": 0.6795,
+      "step": 4385
+    },
+    {
+      "epoch": 8.122222222222222,
+      "grad_norm": 1.3186620871342463,
+      "learning_rate": 1.8469975987105526e-07,
+      "loss": 0.6273,
+      "step": 4386
+    },
+    {
+      "epoch": 8.124074074074073,
+      "grad_norm": 1.4605072670996249,
+      "learning_rate": 1.846917880315086e-07,
+      "loss": 0.6737,
+      "step": 4387
+    },
+    {
+      "epoch": 8.125925925925927,
+      "grad_norm": 1.60525991633902,
+      "learning_rate": 1.846838142878605e-07,
+      "loss": 0.5899,
+      "step": 4388
+    },
+    {
+      "epoch": 8.127777777777778,
+      "grad_norm": 1.5295405089725784,
+      "learning_rate": 1.8467583864029023e-07,
+      "loss": 0.6407,
+      "step": 4389
+    },
+    {
+      "epoch": 8.12962962962963,
+      "grad_norm": 1.8837356310494495,
+      "learning_rate": 1.8466786108897718e-07,
+      "loss": 0.6047,
+      "step": 4390
+    },
+    {
+      "epoch": 8.131481481481481,
+      "grad_norm": 1.7326375202580129,
+      "learning_rate": 1.8465988163410065e-07,
+      "loss": 0.6584,
+      "step": 4391
+    },
+    {
+      "epoch": 8.133333333333333,
+      "grad_norm": 1.9541442422159079,
+      "learning_rate": 1.8465190027584002e-07,
+      "loss": 0.6528,
+      "step": 4392
+    },
+    {
+      "epoch": 8.135185185185184,
+      "grad_norm": 1.634809818560585,
+      "learning_rate": 1.846439170143748e-07,
+      "loss": 0.6796,
+      "step": 4393
+    },
+    {
+      "epoch": 8.137037037037038,
+      "grad_norm": 1.4625602427882987,
+      "learning_rate": 1.846359318498844e-07,
+      "loss": 0.658,
+      "step": 4394
+    },
+    {
+      "epoch": 8.13888888888889,
+      "grad_norm": 1.293210269158073,
+      "learning_rate": 1.846279447825484e-07,
+      "loss": 0.6044,
+      "step": 4395
+    },
+    {
+      "epoch": 8.14074074074074,
+      "grad_norm": 1.4174705384243207,
+      "learning_rate": 1.8461995581254634e-07,
+      "loss": 0.6474,
+      "step": 4396
+    },
+    {
+      "epoch": 8.142592592592592,
+      "grad_norm": 1.565633661072717,
+      "learning_rate": 1.8461196494005788e-07,
+      "loss": 0.6497,
+      "step": 4397
+    },
+    {
+      "epoch": 8.144444444444444,
+      "grad_norm": 1.4862483121816472,
+      "learning_rate": 1.8460397216526262e-07,
+      "loss": 0.6921,
+      "step": 4398
+    },
+    {
+      "epoch": 8.146296296296295,
+      "grad_norm": 1.4742831036198445,
+      "learning_rate": 1.845959774883403e-07,
+      "loss": 0.6603,
+      "step": 4399
+    },
+    {
+      "epoch": 8.148148148148149,
+      "grad_norm": 1.5245190999193128,
+      "learning_rate": 1.8458798090947062e-07,
+      "loss": 0.6346,
+      "step": 4400
+    },
+    {
+      "epoch": 8.15,
+      "grad_norm": 2.6289947895979795,
+      "learning_rate": 1.8457998242883345e-07,
+      "loss": 0.6121,
+      "step": 4401
+    },
+    {
+      "epoch": 8.151851851851852,
+      "grad_norm": 1.5753656994842067,
+      "learning_rate": 1.845719820466085e-07,
+      "loss": 0.657,
+      "step": 4402
+    },
+    {
+      "epoch": 8.153703703703703,
+      "grad_norm": 1.6369488831914727,
+      "learning_rate": 1.8456397976297574e-07,
+      "loss": 0.6802,
+      "step": 4403
+    },
+    {
+      "epoch": 8.155555555555555,
+      "grad_norm": 1.5602197748074216,
+      "learning_rate": 1.84555975578115e-07,
+      "loss": 0.7119,
+      "step": 4404
+    },
+    {
+      "epoch": 8.157407407407407,
+      "grad_norm": 1.419753530834734,
+      "learning_rate": 1.845479694922063e-07,
+      "loss": 0.6218,
+      "step": 4405
+    },
+    {
+      "epoch": 8.15925925925926,
+      "grad_norm": 1.5965928258490798,
+      "learning_rate": 1.8453996150542964e-07,
+      "loss": 0.6473,
+      "step": 4406
+    },
+    {
+      "epoch": 8.161111111111111,
+      "grad_norm": 1.4252535019427055,
+      "learning_rate": 1.84531951617965e-07,
+      "loss": 0.6442,
+      "step": 4407
+    },
+    {
+      "epoch": 8.162962962962963,
+      "grad_norm": 1.7328641861437968,
+      "learning_rate": 1.8452393982999253e-07,
+      "loss": 0.6223,
+      "step": 4408
+    },
+    {
+      "epoch": 8.164814814814815,
+      "grad_norm": 1.4578094338045657,
+      "learning_rate": 1.845159261416923e-07,
+      "loss": 0.6682,
+      "step": 4409
+    },
+    {
+      "epoch": 8.166666666666666,
+      "grad_norm": 1.566527952702924,
+      "learning_rate": 1.8450791055324454e-07,
+      "loss": 0.6322,
+      "step": 4410
+    },
+    {
+      "epoch": 8.168518518518518,
+      "grad_norm": 1.4457926849917626,
+      "learning_rate": 1.8449989306482941e-07,
+      "loss": 0.6522,
+      "step": 4411
+    },
+    {
+      "epoch": 8.170370370370371,
+      "grad_norm": 1.5107172795680537,
+      "learning_rate": 1.8449187367662723e-07,
+      "loss": 0.6493,
+      "step": 4412
+    },
+    {
+      "epoch": 8.172222222222222,
+      "grad_norm": 1.5901504504824933,
+      "learning_rate": 1.8448385238881818e-07,
+      "loss": 0.6378,
+      "step": 4413
+    },
+    {
+      "epoch": 8.174074074074074,
+      "grad_norm": 1.6261408486277367,
+      "learning_rate": 1.8447582920158273e-07,
+      "loss": 0.6503,
+      "step": 4414
+    },
+    {
+      "epoch": 8.175925925925926,
+      "grad_norm": 1.4731146018208923,
+      "learning_rate": 1.844678041151012e-07,
+      "loss": 0.6382,
+      "step": 4415
+    },
+    {
+      "epoch": 8.177777777777777,
+      "grad_norm": 1.650804041661046,
+      "learning_rate": 1.8445977712955402e-07,
+      "loss": 0.632,
+      "step": 4416
+    },
+    {
+      "epoch": 8.17962962962963,
+      "grad_norm": 1.8744121055056089,
+      "learning_rate": 1.8445174824512165e-07,
+      "loss": 0.6274,
+      "step": 4417
+    },
+    {
+      "epoch": 8.181481481481482,
+      "grad_norm": 1.520125284346202,
+      "learning_rate": 1.844437174619846e-07,
+      "loss": 0.6348,
+      "step": 4418
+    },
+    {
+      "epoch": 8.183333333333334,
+      "grad_norm": 2.0812536694644,
+      "learning_rate": 1.8443568478032345e-07,
+      "loss": 0.6571,
+      "step": 4419
+    },
+    {
+      "epoch": 8.185185185185185,
+      "grad_norm": 1.5159981249232075,
+      "learning_rate": 1.8442765020031876e-07,
+      "loss": 0.6687,
+      "step": 4420
+    },
+    {
+      "epoch": 8.187037037037037,
+      "grad_norm": 1.5200485704478703,
+      "learning_rate": 1.8441961372215123e-07,
+      "loss": 0.6174,
+      "step": 4421
+    },
+    {
+      "epoch": 8.188888888888888,
+      "grad_norm": 1.6702124172076425,
+      "learning_rate": 1.8441157534600146e-07,
+      "loss": 0.6214,
+      "step": 4422
+    },
+    {
+      "epoch": 8.190740740740742,
+      "grad_norm": 1.558946158227551,
+      "learning_rate": 1.8440353507205026e-07,
+      "loss": 0.6494,
+      "step": 4423
+    },
+    {
+      "epoch": 8.192592592592593,
+      "grad_norm": 1.3178292131506788,
+      "learning_rate": 1.8439549290047833e-07,
+      "loss": 0.6148,
+      "step": 4424
+    },
+    {
+      "epoch": 8.194444444444445,
+      "grad_norm": 1.6258181046546634,
+      "learning_rate": 1.8438744883146653e-07,
+      "loss": 0.6137,
+      "step": 4425
+    },
+    {
+      "epoch": 8.196296296296296,
+      "grad_norm": 1.933349907683464,
+      "learning_rate": 1.8437940286519568e-07,
+      "loss": 0.6616,
+      "step": 4426
+    },
+    {
+      "epoch": 8.198148148148148,
+      "grad_norm": 1.550809557818013,
+      "learning_rate": 1.8437135500184666e-07,
+      "loss": 0.6444,
+      "step": 4427
+    },
+    {
+      "epoch": 8.2,
+      "grad_norm": 1.7348700374583959,
+      "learning_rate": 1.8436330524160044e-07,
+      "loss": 0.6476,
+      "step": 4428
+    },
+    {
+      "epoch": 8.201851851851853,
+      "grad_norm": 1.5675733841164587,
+      "learning_rate": 1.84355253584638e-07,
+      "loss": 0.6734,
+      "step": 4429
+    },
+    {
+      "epoch": 8.203703703703704,
+      "grad_norm": 1.4643498703928413,
+      "learning_rate": 1.8434720003114036e-07,
+      "loss": 0.6048,
+      "step": 4430
+    },
+    {
+      "epoch": 8.205555555555556,
+      "grad_norm": 1.7356581217539526,
+      "learning_rate": 1.8433914458128858e-07,
+      "loss": 0.6501,
+      "step": 4431
+    },
+    {
+      "epoch": 8.207407407407407,
+      "grad_norm": 1.749789735311238,
+      "learning_rate": 1.8433108723526373e-07,
+      "loss": 0.6533,
+      "step": 4432
+    },
+    {
+      "epoch": 8.209259259259259,
+      "grad_norm": 1.4760663658099382,
+      "learning_rate": 1.8432302799324703e-07,
+      "loss": 0.6839,
+      "step": 4433
+    },
+    {
+      "epoch": 8.21111111111111,
+      "grad_norm": 1.3915985742078312,
+      "learning_rate": 1.8431496685541966e-07,
+      "loss": 0.6382,
+      "step": 4434
+    },
+    {
+      "epoch": 8.212962962962964,
+      "grad_norm": 1.5337320469283482,
+      "learning_rate": 1.843069038219628e-07,
+      "loss": 0.6657,
+      "step": 4435
+    },
+    {
+      "epoch": 8.214814814814815,
+      "grad_norm": 1.9392122524922977,
+      "learning_rate": 1.842988388930578e-07,
+      "loss": 0.648,
+      "step": 4436
+    },
+    {
+      "epoch": 8.216666666666667,
+      "grad_norm": 2.6080664065494013,
+      "learning_rate": 1.8429077206888592e-07,
+      "loss": 0.598,
+      "step": 4437
+    },
+    {
+      "epoch": 8.218518518518518,
+      "grad_norm": 1.4176907829478624,
+      "learning_rate": 1.842827033496286e-07,
+      "loss": 0.604,
+      "step": 4438
+    },
+    {
+      "epoch": 8.22037037037037,
+      "grad_norm": 1.4349506000776107,
+      "learning_rate": 1.8427463273546716e-07,
+      "loss": 0.6145,
+      "step": 4439
+    },
+    {
+      "epoch": 8.222222222222221,
+      "grad_norm": 1.4518955441655228,
+      "learning_rate": 1.842665602265831e-07,
+      "loss": 0.6179,
+      "step": 4440
+    },
+    {
+      "epoch": 8.224074074074075,
+      "grad_norm": 1.4432296280533177,
+      "learning_rate": 1.842584858231579e-07,
+      "loss": 0.6196,
+      "step": 4441
+    },
+    {
+      "epoch": 8.225925925925926,
+      "grad_norm": 1.5930920736653553,
+      "learning_rate": 1.8425040952537309e-07,
+      "loss": 0.6222,
+      "step": 4442
+    },
+    {
+      "epoch": 8.227777777777778,
+      "grad_norm": 1.5437189024522038,
+      "learning_rate": 1.8424233133341025e-07,
+      "loss": 0.6446,
+      "step": 4443
+    },
+    {
+      "epoch": 8.22962962962963,
+      "grad_norm": 1.584257659466699,
+      "learning_rate": 1.8423425124745103e-07,
+      "loss": 0.6779,
+      "step": 4444
+    },
+    {
+      "epoch": 8.231481481481481,
+      "grad_norm": 2.6045755456545057,
+      "learning_rate": 1.8422616926767702e-07,
+      "loss": 0.656,
+      "step": 4445
+    },
+    {
+      "epoch": 8.233333333333333,
+      "grad_norm": 1.5696173877055712,
+      "learning_rate": 1.8421808539427002e-07,
+      "loss": 0.6494,
+      "step": 4446
+    },
+    {
+      "epoch": 8.235185185185186,
+      "grad_norm": 1.7944084779019387,
+      "learning_rate": 1.842099996274117e-07,
+      "loss": 0.6719,
+      "step": 4447
+    },
+    {
+      "epoch": 8.237037037037037,
+      "grad_norm": 1.4192688076247182,
+      "learning_rate": 1.8420191196728388e-07,
+      "loss": 0.6431,
+      "step": 4448
+    },
+    {
+      "epoch": 8.238888888888889,
+      "grad_norm": 1.615805838148867,
+      "learning_rate": 1.841938224140684e-07,
+      "loss": 0.6343,
+      "step": 4449
+    },
+    {
+      "epoch": 8.24074074074074,
+      "grad_norm": 1.4622559081780984,
+      "learning_rate": 1.8418573096794712e-07,
+      "loss": 0.6213,
+      "step": 4450
+    },
+    {
+      "epoch": 8.242592592592592,
+      "grad_norm": 1.4235125550651881,
+      "learning_rate": 1.8417763762910196e-07,
+      "loss": 0.6888,
+      "step": 4451
+    },
+    {
+      "epoch": 8.244444444444444,
+      "grad_norm": 1.5886904843159426,
+      "learning_rate": 1.841695423977149e-07,
+      "loss": 0.6277,
+      "step": 4452
+    },
+    {
+      "epoch": 8.246296296296297,
+      "grad_norm": 1.5425377616543834,
+      "learning_rate": 1.8416144527396789e-07,
+      "loss": 0.6399,
+      "step": 4453
+    },
+    {
+      "epoch": 8.248148148148148,
+      "grad_norm": 1.5471698447424052,
+      "learning_rate": 1.8415334625804303e-07,
+      "loss": 0.671,
+      "step": 4454
+    },
+    {
+      "epoch": 8.25,
+      "grad_norm": 1.4217648391956874,
+      "learning_rate": 1.8414524535012242e-07,
+      "loss": 0.6142,
+      "step": 4455
+    },
+    {
+      "epoch": 8.251851851851852,
+      "grad_norm": 1.5084265828577654,
+      "learning_rate": 1.841371425503881e-07,
+      "loss": 0.6833,
+      "step": 4456
+    },
+    {
+      "epoch": 8.253703703703703,
+      "grad_norm": 1.399182612384619,
+      "learning_rate": 1.8412903785902235e-07,
+      "loss": 0.635,
+      "step": 4457
+    },
+    {
+      "epoch": 8.255555555555556,
+      "grad_norm": 1.9357521076138207,
+      "learning_rate": 1.841209312762073e-07,
+      "loss": 0.6235,
+      "step": 4458
+    },
+    {
+      "epoch": 8.257407407407408,
+      "grad_norm": 1.5735292173773687,
+      "learning_rate": 1.841128228021253e-07,
+      "loss": 0.6613,
+      "step": 4459
+    },
+    {
+      "epoch": 8.25925925925926,
+      "grad_norm": 1.8104192216203563,
+      "learning_rate": 1.8410471243695855e-07,
+      "loss": 0.616,
+      "step": 4460
+    },
+    {
+      "epoch": 8.261111111111111,
+      "grad_norm": 1.5368353609788117,
+      "learning_rate": 1.8409660018088946e-07,
+      "loss": 0.6311,
+      "step": 4461
+    },
+    {
+      "epoch": 8.262962962962963,
+      "grad_norm": 1.7106857496061911,
+      "learning_rate": 1.840884860341004e-07,
+      "loss": 0.6851,
+      "step": 4462
+    },
+    {
+      "epoch": 8.264814814814814,
+      "grad_norm": 1.5679887869269284,
+      "learning_rate": 1.8408036999677377e-07,
+      "loss": 0.6266,
+      "step": 4463
+    },
+    {
+      "epoch": 8.266666666666667,
+      "grad_norm": 7.226378031845,
+      "learning_rate": 1.8407225206909208e-07,
+      "loss": 0.6169,
+      "step": 4464
+    },
+    {
+      "epoch": 8.268518518518519,
+      "grad_norm": 1.4476282976606532,
+      "learning_rate": 1.8406413225123784e-07,
+      "loss": 0.6375,
+      "step": 4465
+    },
+    {
+      "epoch": 8.27037037037037,
+      "grad_norm": 1.5106928677491207,
+      "learning_rate": 1.8405601054339355e-07,
+      "loss": 0.6639,
+      "step": 4466
+    },
+    {
+      "epoch": 8.272222222222222,
+      "grad_norm": 2.573263763803964,
+      "learning_rate": 1.8404788694574187e-07,
+      "loss": 0.632,
+      "step": 4467
+    },
+    {
+      "epoch": 8.274074074074074,
+      "grad_norm": 1.3784290685138725,
+      "learning_rate": 1.8403976145846543e-07,
+      "loss": 0.5972,
+      "step": 4468
+    },
+    {
+      "epoch": 8.275925925925925,
+      "grad_norm": 1.4775133270272727,
+      "learning_rate": 1.840316340817469e-07,
+      "loss": 0.6518,
+      "step": 4469
+    },
+    {
+      "epoch": 8.277777777777779,
+      "grad_norm": 1.4429689298292858,
+      "learning_rate": 1.84023504815769e-07,
+      "loss": 0.6619,
+      "step": 4470
+    },
+    {
+      "epoch": 8.27962962962963,
+      "grad_norm": 1.5634488793950807,
+      "learning_rate": 1.8401537366071452e-07,
+      "loss": 0.6402,
+      "step": 4471
+    },
+    {
+      "epoch": 8.281481481481482,
+      "grad_norm": 1.3815683476118397,
+      "learning_rate": 1.8400724061676626e-07,
+      "loss": 0.6083,
+      "step": 4472
+    },
+    {
+      "epoch": 8.283333333333333,
+      "grad_norm": 1.4814211980123844,
+      "learning_rate": 1.8399910568410705e-07,
+      "loss": 0.6206,
+      "step": 4473
+    },
+    {
+      "epoch": 8.285185185185185,
+      "grad_norm": 1.5126234557080462,
+      "learning_rate": 1.8399096886291983e-07,
+      "loss": 0.6569,
+      "step": 4474
+    },
+    {
+      "epoch": 8.287037037037036,
+      "grad_norm": 1.4905478081848376,
+      "learning_rate": 1.8398283015338748e-07,
+      "loss": 0.6752,
+      "step": 4475
+    },
+    {
+      "epoch": 8.28888888888889,
+      "grad_norm": 1.6238189415991262,
+      "learning_rate": 1.8397468955569307e-07,
+      "loss": 0.6879,
+      "step": 4476
+    },
+    {
+      "epoch": 8.290740740740741,
+      "grad_norm": 1.7677577704848157,
+      "learning_rate": 1.8396654707001952e-07,
+      "loss": 0.6402,
+      "step": 4477
+    },
+    {
+      "epoch": 8.292592592592593,
+      "grad_norm": 1.4088138604229454,
+      "learning_rate": 1.8395840269654996e-07,
+      "loss": 0.6259,
+      "step": 4478
+    },
+    {
+      "epoch": 8.294444444444444,
+      "grad_norm": 1.5362515791971183,
+      "learning_rate": 1.8395025643546748e-07,
+      "loss": 0.6472,
+      "step": 4479
+    },
+    {
+      "epoch": 8.296296296296296,
+      "grad_norm": 1.546211815592026,
+      "learning_rate": 1.8394210828695522e-07,
+      "loss": 0.6076,
+      "step": 4480
+    },
+    {
+      "epoch": 8.298148148148147,
+      "grad_norm": 1.5885515198620532,
+      "learning_rate": 1.839339582511964e-07,
+      "loss": 0.6363,
+      "step": 4481
+    },
+    {
+      "epoch": 8.3,
+      "grad_norm": 1.4788680858177672,
+      "learning_rate": 1.8392580632837423e-07,
+      "loss": 0.6375,
+      "step": 4482
+    },
+    {
+      "epoch": 8.301851851851852,
+      "grad_norm": 1.5580058305481075,
+      "learning_rate": 1.83917652518672e-07,
+      "loss": 0.6377,
+      "step": 4483
+    },
+    {
+      "epoch": 8.303703703703704,
+      "grad_norm": 1.691032618517879,
+      "learning_rate": 1.8390949682227302e-07,
+      "loss": 0.6499,
+      "step": 4484
+    },
+    {
+      "epoch": 8.305555555555555,
+      "grad_norm": 1.3978341668575072,
+      "learning_rate": 1.8390133923936064e-07,
+      "loss": 0.6636,
+      "step": 4485
+    },
+    {
+      "epoch": 8.307407407407407,
+      "grad_norm": 1.6007950023712536,
+      "learning_rate": 1.8389317977011828e-07,
+      "loss": 0.6402,
+      "step": 4486
+    },
+    {
+      "epoch": 8.309259259259258,
+      "grad_norm": 1.8494485784292016,
+      "learning_rate": 1.838850184147294e-07,
+      "loss": 0.6699,
+      "step": 4487
+    },
+    {
+      "epoch": 8.311111111111112,
+      "grad_norm": 1.8585086606793517,
+      "learning_rate": 1.838768551733775e-07,
+      "loss": 0.6507,
+      "step": 4488
+    },
+    {
+      "epoch": 8.312962962962963,
+      "grad_norm": 1.5945178824609725,
+      "learning_rate": 1.8386869004624604e-07,
+      "loss": 0.6489,
+      "step": 4489
+    },
+    {
+      "epoch": 8.314814814814815,
+      "grad_norm": 1.4984619488016442,
+      "learning_rate": 1.838605230335187e-07,
+      "loss": 0.6517,
+      "step": 4490
+    },
+    {
+      "epoch": 8.316666666666666,
+      "grad_norm": 1.9315032752479253,
+      "learning_rate": 1.83852354135379e-07,
+      "loss": 0.6131,
+      "step": 4491
+    },
+    {
+      "epoch": 8.318518518518518,
+      "grad_norm": 1.3195694125447324,
+      "learning_rate": 1.8384418335201066e-07,
+      "loss": 0.6392,
+      "step": 4492
+    },
+    {
+      "epoch": 8.32037037037037,
+      "grad_norm": 1.6595583653607129,
+      "learning_rate": 1.8383601068359735e-07,
+      "loss": 0.641,
+      "step": 4493
+    },
+    {
+      "epoch": 8.322222222222223,
+      "grad_norm": 1.3498339808571895,
+      "learning_rate": 1.8382783613032283e-07,
+      "loss": 0.6274,
+      "step": 4494
+    },
+    {
+      "epoch": 8.324074074074074,
+      "grad_norm": 1.5288211928657378,
+      "learning_rate": 1.838196596923709e-07,
+      "loss": 0.6629,
+      "step": 4495
+    },
+    {
+      "epoch": 8.325925925925926,
+      "grad_norm": 1.521044331787283,
+      "learning_rate": 1.8381148136992534e-07,
+      "loss": 0.5925,
+      "step": 4496
+    },
+    {
+      "epoch": 8.327777777777778,
+      "grad_norm": 2.1594079249317666,
+      "learning_rate": 1.8380330116317007e-07,
+      "loss": 0.6169,
+      "step": 4497
+    },
+    {
+      "epoch": 8.329629629629629,
+      "grad_norm": 1.9142829518684596,
+      "learning_rate": 1.8379511907228896e-07,
+      "loss": 0.6249,
+      "step": 4498
+    },
+    {
+      "epoch": 8.331481481481482,
+      "grad_norm": 2.190524264767541,
+      "learning_rate": 1.8378693509746602e-07,
+      "loss": 0.6115,
+      "step": 4499
+    },
+    {
+      "epoch": 8.333333333333334,
+      "grad_norm": 1.5447064729566953,
+      "learning_rate": 1.837787492388852e-07,
+      "loss": 0.6323,
+      "step": 4500
+    },
+    {
+      "epoch": 8.335185185185185,
+      "grad_norm": 1.6328414471817385,
+      "learning_rate": 1.8377056149673057e-07,
+      "loss": 0.6525,
+      "step": 4501
+    },
+    {
+      "epoch": 8.337037037037037,
+      "grad_norm": 1.3366391145866388,
+      "learning_rate": 1.8376237187118617e-07,
+      "loss": 0.6197,
+      "step": 4502
+    },
+    {
+      "epoch": 8.338888888888889,
+      "grad_norm": 1.370917703692105,
+      "learning_rate": 1.8375418036243617e-07,
+      "loss": 0.6291,
+      "step": 4503
+    },
+    {
+      "epoch": 8.34074074074074,
+      "grad_norm": 1.8541410270394698,
+      "learning_rate": 1.8374598697066475e-07,
+      "loss": 0.6369,
+      "step": 4504
+    },
+    {
+      "epoch": 8.342592592592593,
+      "grad_norm": 1.4542215035162216,
+      "learning_rate": 1.8373779169605603e-07,
+      "loss": 0.6619,
+      "step": 4505
+    },
+    {
+      "epoch": 8.344444444444445,
+      "grad_norm": 1.6191045063866245,
+      "learning_rate": 1.8372959453879438e-07,
+      "loss": 0.6086,
+      "step": 4506
+    },
+    {
+      "epoch": 8.346296296296297,
+      "grad_norm": 1.824311218026256,
+      "learning_rate": 1.8372139549906402e-07,
+      "loss": 0.6554,
+      "step": 4507
+    },
+    {
+      "epoch": 8.348148148148148,
+      "grad_norm": 1.530900501725902,
+      "learning_rate": 1.8371319457704934e-07,
+      "loss": 0.6776,
+      "step": 4508
+    },
+    {
+      "epoch": 8.35,
+      "grad_norm": 1.6540338128844,
+      "learning_rate": 1.8370499177293462e-07,
+      "loss": 0.664,
+      "step": 4509
+    },
+    {
+      "epoch": 8.351851851851851,
+      "grad_norm": 1.6417826859270062,
+      "learning_rate": 1.8369678708690436e-07,
+      "loss": 0.6686,
+      "step": 4510
+    },
+    {
+      "epoch": 8.353703703703705,
+      "grad_norm": 1.9617097612692143,
+      "learning_rate": 1.8368858051914304e-07,
+      "loss": 0.6486,
+      "step": 4511
+    },
+    {
+      "epoch": 8.355555555555556,
+      "grad_norm": 1.702950181609952,
+      "learning_rate": 1.8368037206983513e-07,
+      "loss": 0.645,
+      "step": 4512
+    },
+    {
+      "epoch": 8.357407407407408,
+      "grad_norm": 1.4557490552542212,
+      "learning_rate": 1.836721617391652e-07,
+      "loss": 0.6388,
+      "step": 4513
+    },
+    {
+      "epoch": 8.35925925925926,
+      "grad_norm": 1.5668711873375605,
+      "learning_rate": 1.8366394952731776e-07,
+      "loss": 0.6276,
+      "step": 4514
+    },
+    {
+      "epoch": 8.36111111111111,
+      "grad_norm": 1.4112367621654067,
+      "learning_rate": 1.8365573543447754e-07,
+      "loss": 0.6653,
+      "step": 4515
+    },
+    {
+      "epoch": 8.362962962962962,
+      "grad_norm": 1.604709948789727,
+      "learning_rate": 1.8364751946082918e-07,
+      "loss": 0.6596,
+      "step": 4516
+    },
+    {
+      "epoch": 8.364814814814816,
+      "grad_norm": 1.4546142157849635,
+      "learning_rate": 1.8363930160655744e-07,
+      "loss": 0.6644,
+      "step": 4517
+    },
+    {
+      "epoch": 8.366666666666667,
+      "grad_norm": 1.6605625604766467,
+      "learning_rate": 1.83631081871847e-07,
+      "loss": 0.6201,
+      "step": 4518
+    },
+    {
+      "epoch": 8.368518518518519,
+      "grad_norm": 1.4296544642826179,
+      "learning_rate": 1.836228602568827e-07,
+      "loss": 0.6216,
+      "step": 4519
+    },
+    {
+      "epoch": 8.37037037037037,
+      "grad_norm": 1.5080770207724064,
+      "learning_rate": 1.8361463676184938e-07,
+      "loss": 0.5978,
+      "step": 4520
+    },
+    {
+      "epoch": 8.372222222222222,
+      "grad_norm": 1.4510967381213307,
+      "learning_rate": 1.8360641138693195e-07,
+      "loss": 0.6804,
+      "step": 4521
+    },
+    {
+      "epoch": 8.374074074074073,
+      "grad_norm": 1.6864678245257798,
+      "learning_rate": 1.8359818413231531e-07,
+      "loss": 0.6138,
+      "step": 4522
+    },
+    {
+      "epoch": 8.375925925925927,
+      "grad_norm": 1.8309270085825362,
+      "learning_rate": 1.8358995499818442e-07,
+      "loss": 0.6183,
+      "step": 4523
+    },
+    {
+      "epoch": 8.377777777777778,
+      "grad_norm": 1.581548293298123,
+      "learning_rate": 1.8358172398472436e-07,
+      "loss": 0.6269,
+      "step": 4524
+    },
+    {
+      "epoch": 8.37962962962963,
+      "grad_norm": 1.5912797255674556,
+      "learning_rate": 1.8357349109212013e-07,
+      "loss": 0.6608,
+      "step": 4525
+    },
+    {
+      "epoch": 8.381481481481481,
+      "grad_norm": 1.4686004244078337,
+      "learning_rate": 1.8356525632055683e-07,
+      "loss": 0.6687,
+      "step": 4526
+    },
+    {
+      "epoch": 8.383333333333333,
+      "grad_norm": 1.476903957886199,
+      "learning_rate": 1.8355701967021958e-07,
+      "loss": 0.6605,
+      "step": 4527
+    },
+    {
+      "epoch": 8.385185185185184,
+      "grad_norm": 1.4229082800769943,
+      "learning_rate": 1.8354878114129364e-07,
+      "loss": 0.6205,
+      "step": 4528
+    },
+    {
+      "epoch": 8.387037037037038,
+      "grad_norm": 1.6011482064276765,
+      "learning_rate": 1.8354054073396417e-07,
+      "loss": 0.6494,
+      "step": 4529
+    },
+    {
+      "epoch": 8.38888888888889,
+      "grad_norm": 1.4849614245967997,
+      "learning_rate": 1.8353229844841644e-07,
+      "loss": 0.6178,
+      "step": 4530
+    },
+    {
+      "epoch": 8.39074074074074,
+      "grad_norm": 1.4415165403682586,
+      "learning_rate": 1.8352405428483575e-07,
+      "loss": 0.6778,
+      "step": 4531
+    },
+    {
+      "epoch": 8.392592592592592,
+      "grad_norm": 1.4330827623318043,
+      "learning_rate": 1.8351580824340752e-07,
+      "loss": 0.6649,
+      "step": 4532
+    },
+    {
+      "epoch": 8.394444444444444,
+      "grad_norm": 1.562137986180405,
+      "learning_rate": 1.8350756032431708e-07,
+      "loss": 0.7017,
+      "step": 4533
+    },
+    {
+      "epoch": 8.396296296296295,
+      "grad_norm": 1.491515292892923,
+      "learning_rate": 1.8349931052774984e-07,
+      "loss": 0.6398,
+      "step": 4534
+    },
+    {
+      "epoch": 8.398148148148149,
+      "grad_norm": 1.4577729697546422,
+      "learning_rate": 1.8349105885389138e-07,
+      "loss": 0.6289,
+      "step": 4535
+    },
+    {
+      "epoch": 8.4,
+      "grad_norm": 1.7012731865284316,
+      "learning_rate": 1.834828053029271e-07,
+      "loss": 0.6305,
+      "step": 4536
+    },
+    {
+      "epoch": 8.401851851851852,
+      "grad_norm": 1.5534133877230325,
+      "learning_rate": 1.8347454987504266e-07,
+      "loss": 0.6549,
+      "step": 4537
+    },
+    {
+      "epoch": 8.403703703703703,
+      "grad_norm": 2.06982233609931,
+      "learning_rate": 1.834662925704236e-07,
+      "loss": 0.5926,
+      "step": 4538
+    },
+    {
+      "epoch": 8.405555555555555,
+      "grad_norm": 1.3814469697110232,
+      "learning_rate": 1.834580333892556e-07,
+      "loss": 0.6392,
+      "step": 4539
+    },
+    {
+      "epoch": 8.407407407407407,
+      "grad_norm": 1.382295643754722,
+      "learning_rate": 1.8344977233172437e-07,
+      "loss": 0.665,
+      "step": 4540
+    },
+    {
+      "epoch": 8.40925925925926,
+      "grad_norm": 1.3990804594408262,
+      "learning_rate": 1.834415093980156e-07,
+      "loss": 0.6354,
+      "step": 4541
+    },
+    {
+      "epoch": 8.411111111111111,
+      "grad_norm": 1.363437766916677,
+      "learning_rate": 1.8343324458831506e-07,
+      "loss": 0.6374,
+      "step": 4542
+    },
+    {
+      "epoch": 8.412962962962963,
+      "grad_norm": 1.6314452720963881,
+      "learning_rate": 1.8342497790280857e-07,
+      "loss": 0.6608,
+      "step": 4543
+    },
+    {
+      "epoch": 8.414814814814815,
+      "grad_norm": 1.3630394580862388,
+      "learning_rate": 1.8341670934168204e-07,
+      "loss": 0.6233,
+      "step": 4544
+    },
+    {
+      "epoch": 8.416666666666666,
+      "grad_norm": 1.9049982870144648,
+      "learning_rate": 1.8340843890512127e-07,
+      "loss": 0.5932,
+      "step": 4545
+    },
+    {
+      "epoch": 8.418518518518518,
+      "grad_norm": 1.874352471556863,
+      "learning_rate": 1.834001665933123e-07,
+      "loss": 0.6537,
+      "step": 4546
+    },
+    {
+      "epoch": 8.420370370370371,
+      "grad_norm": 1.6993176651426807,
+      "learning_rate": 1.833918924064411e-07,
+      "loss": 0.6541,
+      "step": 4547
+    },
+    {
+      "epoch": 8.422222222222222,
+      "grad_norm": 1.564222617365321,
+      "learning_rate": 1.833836163446936e-07,
+      "loss": 0.631,
+      "step": 4548
+    },
+    {
+      "epoch": 8.424074074074074,
+      "grad_norm": 1.8963502602083337,
+      "learning_rate": 1.8337533840825595e-07,
+      "loss": 0.6853,
+      "step": 4549
+    },
+    {
+      "epoch": 8.425925925925926,
+      "grad_norm": 1.4688978684591096,
+      "learning_rate": 1.833670585973143e-07,
+      "loss": 0.6844,
+      "step": 4550
+    },
+    {
+      "epoch": 8.427777777777777,
+      "grad_norm": 1.595942662172928,
+      "learning_rate": 1.8335877691205472e-07,
+      "loss": 0.6047,
+      "step": 4551
+    },
+    {
+      "epoch": 8.42962962962963,
+      "grad_norm": 1.4500602902578772,
+      "learning_rate": 1.8335049335266344e-07,
+      "loss": 0.6364,
+      "step": 4552
+    },
+    {
+      "epoch": 8.431481481481482,
+      "grad_norm": 1.3893302144834807,
+      "learning_rate": 1.833422079193267e-07,
+      "loss": 0.6477,
+      "step": 4553
+    },
+    {
+      "epoch": 8.433333333333334,
+      "grad_norm": 1.4838805938301063,
+      "learning_rate": 1.8333392061223078e-07,
+      "loss": 0.6639,
+      "step": 4554
+    },
+    {
+      "epoch": 8.435185185185185,
+      "grad_norm": 1.4699031024385159,
+      "learning_rate": 1.8332563143156193e-07,
+      "loss": 0.635,
+      "step": 4555
+    },
+    {
+      "epoch": 8.437037037037037,
+      "grad_norm": 1.4374806778754223,
+      "learning_rate": 1.8331734037750664e-07,
+      "loss": 0.5841,
+      "step": 4556
+    },
+    {
+      "epoch": 8.438888888888888,
+      "grad_norm": 1.5296147639887085,
+      "learning_rate": 1.8330904745025122e-07,
+      "loss": 0.6831,
+      "step": 4557
+    },
+    {
+      "epoch": 8.440740740740742,
+      "grad_norm": 1.5381406069873516,
+      "learning_rate": 1.8330075264998218e-07,
+      "loss": 0.583,
+      "step": 4558
+    },
+    {
+      "epoch": 8.442592592592593,
+      "grad_norm": 1.7699048857280157,
+      "learning_rate": 1.8329245597688594e-07,
+      "loss": 0.6491,
+      "step": 4559
+    },
+    {
+      "epoch": 8.444444444444445,
+      "grad_norm": 1.4577225336014208,
+      "learning_rate": 1.8328415743114912e-07,
+      "loss": 0.6554,
+      "step": 4560
+    },
+    {
+      "epoch": 8.446296296296296,
+      "grad_norm": 1.5996410280927722,
+      "learning_rate": 1.832758570129582e-07,
+      "loss": 0.6051,
+      "step": 4561
+    },
+    {
+      "epoch": 8.448148148148148,
+      "grad_norm": 1.434765292793164,
+      "learning_rate": 1.8326755472249985e-07,
+      "loss": 0.7148,
+      "step": 4562
+    },
+    {
+      "epoch": 8.45,
+      "grad_norm": 1.4282866469760362,
+      "learning_rate": 1.8325925055996074e-07,
+      "loss": 0.6291,
+      "step": 4563
+    },
+    {
+      "epoch": 8.451851851851853,
+      "grad_norm": 1.6708832412222447,
+      "learning_rate": 1.8325094452552755e-07,
+      "loss": 0.6581,
+      "step": 4564
+    },
+    {
+      "epoch": 8.453703703703704,
+      "grad_norm": 1.7123286697631361,
+      "learning_rate": 1.8324263661938702e-07,
+      "loss": 0.6125,
+      "step": 4565
+    },
+    {
+      "epoch": 8.455555555555556,
+      "grad_norm": 1.4812133951355995,
+      "learning_rate": 1.832343268417259e-07,
+      "loss": 0.6699,
+      "step": 4566
+    },
+    {
+      "epoch": 8.457407407407407,
+      "grad_norm": 1.5158201209553643,
+      "learning_rate": 1.8322601519273107e-07,
+      "loss": 0.6646,
+      "step": 4567
+    },
+    {
+      "epoch": 8.459259259259259,
+      "grad_norm": 1.7434690183775805,
+      "learning_rate": 1.832177016725894e-07,
+      "loss": 0.6511,
+      "step": 4568
+    },
+    {
+      "epoch": 8.46111111111111,
+      "grad_norm": 1.8785515829188772,
+      "learning_rate": 1.8320938628148777e-07,
+      "loss": 0.653,
+      "step": 4569
+    },
+    {
+      "epoch": 8.462962962962964,
+      "grad_norm": 1.4281991813971242,
+      "learning_rate": 1.8320106901961311e-07,
+      "loss": 0.6911,
+      "step": 4570
+    },
+    {
+      "epoch": 8.464814814814815,
+      "grad_norm": 1.3679988205675195,
+      "learning_rate": 1.8319274988715248e-07,
+      "loss": 0.6357,
+      "step": 4571
+    },
+    {
+      "epoch": 8.466666666666667,
+      "grad_norm": 1.7775941656739307,
+      "learning_rate": 1.831844288842929e-07,
+      "loss": 0.6462,
+      "step": 4572
+    },
+    {
+      "epoch": 8.468518518518518,
+      "grad_norm": 1.5231018287887081,
+      "learning_rate": 1.831761060112214e-07,
+      "loss": 0.6409,
+      "step": 4573
+    },
+    {
+      "epoch": 8.47037037037037,
+      "grad_norm": 1.4144579142939049,
+      "learning_rate": 1.8316778126812515e-07,
+      "loss": 0.5929,
+      "step": 4574
+    },
+    {
+      "epoch": 8.472222222222221,
+      "grad_norm": 1.3834680242035378,
+      "learning_rate": 1.831594546551913e-07,
+      "loss": 0.6291,
+      "step": 4575
+    },
+    {
+      "epoch": 8.474074074074075,
+      "grad_norm": 1.57796722508518,
+      "learning_rate": 1.8315112617260703e-07,
+      "loss": 0.617,
+      "step": 4576
+    },
+    {
+      "epoch": 8.475925925925926,
+      "grad_norm": 1.6181525456398926,
+      "learning_rate": 1.8314279582055963e-07,
+      "loss": 0.6773,
+      "step": 4577
+    },
+    {
+      "epoch": 8.477777777777778,
+      "grad_norm": 1.5919046856276704,
+      "learning_rate": 1.8313446359923636e-07,
+      "loss": 0.6523,
+      "step": 4578
+    },
+    {
+      "epoch": 8.47962962962963,
+      "grad_norm": 1.5173886834138257,
+      "learning_rate": 1.8312612950882456e-07,
+      "loss": 0.6245,
+      "step": 4579
+    },
+    {
+      "epoch": 8.481481481481481,
+      "grad_norm": 1.371609178734613,
+      "learning_rate": 1.831177935495116e-07,
+      "loss": 0.6377,
+      "step": 4580
+    },
+    {
+      "epoch": 8.483333333333333,
+      "grad_norm": 1.7343617199354378,
+      "learning_rate": 1.8310945572148491e-07,
+      "loss": 0.6296,
+      "step": 4581
+    },
+    {
+      "epoch": 8.485185185185186,
+      "grad_norm": 1.6157766312139887,
+      "learning_rate": 1.8310111602493193e-07,
+      "loss": 0.645,
+      "step": 4582
+    },
+    {
+      "epoch": 8.487037037037037,
+      "grad_norm": 2.2192004255214353,
+      "learning_rate": 1.8309277446004012e-07,
+      "loss": 0.6336,
+      "step": 4583
+    },
+    {
+      "epoch": 8.488888888888889,
+      "grad_norm": 1.5177882417896689,
+      "learning_rate": 1.830844310269971e-07,
+      "loss": 0.6259,
+      "step": 4584
+    },
+    {
+      "epoch": 8.49074074074074,
+      "grad_norm": 2.2876322684769095,
+      "learning_rate": 1.830760857259904e-07,
+      "loss": 0.6529,
+      "step": 4585
+    },
+    {
+      "epoch": 8.492592592592592,
+      "grad_norm": 1.6443906210680563,
+      "learning_rate": 1.830677385572077e-07,
+      "loss": 0.6066,
+      "step": 4586
+    },
+    {
+      "epoch": 8.494444444444444,
+      "grad_norm": 1.6620671081429497,
+      "learning_rate": 1.830593895208366e-07,
+      "loss": 0.605,
+      "step": 4587
+    },
+    {
+      "epoch": 8.496296296296297,
+      "grad_norm": 1.6237177762667632,
+      "learning_rate": 1.8305103861706485e-07,
+      "loss": 0.6037,
+      "step": 4588
+    },
+    {
+      "epoch": 8.498148148148148,
+      "grad_norm": 1.5424428006789714,
+      "learning_rate": 1.8304268584608017e-07,
+      "loss": 0.5897,
+      "step": 4589
+    },
+    {
+      "epoch": 8.5,
+      "grad_norm": 1.4428500929496997,
+      "learning_rate": 1.830343312080704e-07,
+      "loss": 0.6281,
+      "step": 4590
+    },
+    {
+      "epoch": 8.501851851851852,
+      "grad_norm": 1.4226147571645114,
+      "learning_rate": 1.8302597470322332e-07,
+      "loss": 0.6322,
+      "step": 4591
+    },
+    {
+      "epoch": 8.503703703703703,
+      "grad_norm": 3.66458925985423,
+      "learning_rate": 1.8301761633172684e-07,
+      "loss": 0.6583,
+      "step": 4592
+    },
+    {
+      "epoch": 8.505555555555556,
+      "grad_norm": 1.4721840927227268,
+      "learning_rate": 1.8300925609376887e-07,
+      "loss": 0.6681,
+      "step": 4593
+    },
+    {
+      "epoch": 8.507407407407408,
+      "grad_norm": 2.296984133904764,
+      "learning_rate": 1.8300089398953737e-07,
+      "loss": 0.6372,
+      "step": 4594
+    },
+    {
+      "epoch": 8.50925925925926,
+      "grad_norm": 1.6193845328258152,
+      "learning_rate": 1.8299253001922036e-07,
+      "loss": 0.6082,
+      "step": 4595
+    },
+    {
+      "epoch": 8.511111111111111,
+      "grad_norm": 1.7125968009672943,
+      "learning_rate": 1.829841641830059e-07,
+      "loss": 0.6505,
+      "step": 4596
+    },
+    {
+      "epoch": 8.512962962962963,
+      "grad_norm": 1.4994943326885157,
+      "learning_rate": 1.8297579648108198e-07,
+      "loss": 0.6387,
+      "step": 4597
+    },
+    {
+      "epoch": 8.514814814814814,
+      "grad_norm": 1.4561646082662318,
+      "learning_rate": 1.8296742691363685e-07,
+      "loss": 0.6573,
+      "step": 4598
+    },
+    {
+      "epoch": 8.516666666666667,
+      "grad_norm": 1.6034150146411446,
+      "learning_rate": 1.829590554808586e-07,
+      "loss": 0.6316,
+      "step": 4599
+    },
+    {
+      "epoch": 8.518518518518519,
+      "grad_norm": 1.6917244304149643,
+      "learning_rate": 1.8295068218293545e-07,
+      "loss": 0.6308,
+      "step": 4600
+    },
+    {
+      "epoch": 8.52037037037037,
+      "grad_norm": 1.2741846009401572,
+      "learning_rate": 1.829423070200557e-07,
+      "loss": 0.6452,
+      "step": 4601
+    },
+    {
+      "epoch": 8.522222222222222,
+      "grad_norm": 1.7116703651677392,
+      "learning_rate": 1.8293392999240765e-07,
+      "loss": 0.6435,
+      "step": 4602
+    },
+    {
+      "epoch": 8.524074074074074,
+      "grad_norm": 1.5014199062805065,
+      "learning_rate": 1.8292555110017958e-07,
+      "loss": 0.6613,
+      "step": 4603
+    },
+    {
+      "epoch": 8.525925925925925,
+      "grad_norm": 1.5460828042744386,
+      "learning_rate": 1.8291717034355988e-07,
+      "loss": 0.6739,
+      "step": 4604
+    },
+    {
+      "epoch": 8.527777777777779,
+      "grad_norm": 1.5500219760823115,
+      "learning_rate": 1.82908787722737e-07,
+      "loss": 0.6254,
+      "step": 4605
+    },
+    {
+      "epoch": 8.52962962962963,
+      "grad_norm": 1.570950037911747,
+      "learning_rate": 1.8290040323789943e-07,
+      "loss": 0.6436,
+      "step": 4606
+    },
+    {
+      "epoch": 8.531481481481482,
+      "grad_norm": 1.4662886803866015,
+      "learning_rate": 1.8289201688923562e-07,
+      "loss": 0.6453,
+      "step": 4607
+    },
+    {
+      "epoch": 8.533333333333333,
+      "grad_norm": 2.140665127568465,
+      "learning_rate": 1.8288362867693413e-07,
+      "loss": 0.6457,
+      "step": 4608
+    },
+    {
+      "epoch": 8.535185185185185,
+      "grad_norm": 1.5446819930546778,
+      "learning_rate": 1.8287523860118356e-07,
+      "loss": 0.6621,
+      "step": 4609
+    },
+    {
+      "epoch": 8.537037037037036,
+      "grad_norm": 2.2355785679658857,
+      "learning_rate": 1.8286684666217255e-07,
+      "loss": 0.6489,
+      "step": 4610
+    },
+    {
+      "epoch": 8.53888888888889,
+      "grad_norm": 1.545796072142865,
+      "learning_rate": 1.8285845286008973e-07,
+      "loss": 0.6373,
+      "step": 4611
+    },
+    {
+      "epoch": 8.540740740740741,
+      "grad_norm": 1.5205906408171752,
+      "learning_rate": 1.8285005719512388e-07,
+      "loss": 0.6166,
+      "step": 4612
+    },
+    {
+      "epoch": 8.542592592592593,
+      "grad_norm": 1.3964947578471048,
+      "learning_rate": 1.8284165966746373e-07,
+      "loss": 0.6373,
+      "step": 4613
+    },
+    {
+      "epoch": 8.544444444444444,
+      "grad_norm": 1.56279393515347,
+      "learning_rate": 1.828332602772981e-07,
+      "loss": 0.5994,
+      "step": 4614
+    },
+    {
+      "epoch": 8.546296296296296,
+      "grad_norm": 1.5192694019191217,
+      "learning_rate": 1.8282485902481575e-07,
+      "loss": 0.6423,
+      "step": 4615
+    },
+    {
+      "epoch": 8.548148148148147,
+      "grad_norm": 1.560031159362777,
+      "learning_rate": 1.8281645591020564e-07,
+      "loss": 0.6416,
+      "step": 4616
+    },
+    {
+      "epoch": 8.55,
+      "grad_norm": 1.3911895632401228,
+      "learning_rate": 1.828080509336567e-07,
+      "loss": 0.6522,
+      "step": 4617
+    },
+    {
+      "epoch": 8.551851851851852,
+      "grad_norm": 1.5042651037844108,
+      "learning_rate": 1.8279964409535783e-07,
+      "loss": 0.6435,
+      "step": 4618
+    },
+    {
+      "epoch": 8.553703703703704,
+      "grad_norm": 1.7811036677197682,
+      "learning_rate": 1.8279123539549812e-07,
+      "loss": 0.6001,
+      "step": 4619
+    },
+    {
+      "epoch": 8.555555555555555,
+      "grad_norm": 1.5563950588343298,
+      "learning_rate": 1.8278282483426657e-07,
+      "loss": 0.656,
+      "step": 4620
+    },
+    {
+      "epoch": 8.557407407407407,
+      "grad_norm": 1.6224035928703437,
+      "learning_rate": 1.8277441241185228e-07,
+      "loss": 0.6884,
+      "step": 4621
+    },
+    {
+      "epoch": 8.559259259259258,
+      "grad_norm": 3.539910385848698,
+      "learning_rate": 1.8276599812844437e-07,
+      "loss": 0.5912,
+      "step": 4622
+    },
+    {
+      "epoch": 8.561111111111112,
+      "grad_norm": 1.4926072753544442,
+      "learning_rate": 1.8275758198423205e-07,
+      "loss": 0.6165,
+      "step": 4623
+    },
+    {
+      "epoch": 8.562962962962963,
+      "grad_norm": 1.6971106085024001,
+      "learning_rate": 1.8274916397940453e-07,
+      "loss": 0.6497,
+      "step": 4624
+    },
+    {
+      "epoch": 8.564814814814815,
+      "grad_norm": 1.445082588868154,
+      "learning_rate": 1.8274074411415104e-07,
+      "loss": 0.6245,
+      "step": 4625
+    },
+    {
+      "epoch": 8.566666666666666,
+      "grad_norm": 1.6964333323163376,
+      "learning_rate": 1.827323223886609e-07,
+      "loss": 0.6466,
+      "step": 4626
+    },
+    {
+      "epoch": 8.568518518518518,
+      "grad_norm": 1.37633204202434,
+      "learning_rate": 1.827238988031235e-07,
+      "loss": 0.6411,
+      "step": 4627
+    },
+    {
+      "epoch": 8.57037037037037,
+      "grad_norm": 1.5430830872250902,
+      "learning_rate": 1.8271547335772811e-07,
+      "loss": 0.6436,
+      "step": 4628
+    },
+    {
+      "epoch": 8.572222222222223,
+      "grad_norm": 1.9230183394575824,
+      "learning_rate": 1.827070460526643e-07,
+      "loss": 0.6266,
+      "step": 4629
+    },
+    {
+      "epoch": 8.574074074074074,
+      "grad_norm": 1.4166430563025159,
+      "learning_rate": 1.826986168881214e-07,
+      "loss": 0.6583,
+      "step": 4630
+    },
+    {
+      "epoch": 8.575925925925926,
+      "grad_norm": 1.5076059966983177,
+      "learning_rate": 1.8269018586428903e-07,
+      "loss": 0.6081,
+      "step": 4631
+    },
+    {
+      "epoch": 8.577777777777778,
+      "grad_norm": 1.4557491288992066,
+      "learning_rate": 1.826817529813567e-07,
+      "loss": 0.6049,
+      "step": 4632
+    },
+    {
+      "epoch": 8.579629629629629,
+      "grad_norm": 1.4809285355161672,
+      "learning_rate": 1.8267331823951397e-07,
+      "loss": 0.6781,
+      "step": 4633
+    },
+    {
+      "epoch": 8.581481481481482,
+      "grad_norm": 1.4212208772258825,
+      "learning_rate": 1.8266488163895054e-07,
+      "loss": 0.6933,
+      "step": 4634
+    },
+    {
+      "epoch": 8.583333333333334,
+      "grad_norm": 1.5799924915930879,
+      "learning_rate": 1.8265644317985602e-07,
+      "loss": 0.646,
+      "step": 4635
+    },
+    {
+      "epoch": 8.585185185185185,
+      "grad_norm": 1.5573144804872525,
+      "learning_rate": 1.8264800286242022e-07,
+      "loss": 0.6869,
+      "step": 4636
+    },
+    {
+      "epoch": 8.587037037037037,
+      "grad_norm": 1.452921149434807,
+      "learning_rate": 1.826395606868328e-07,
+      "loss": 0.6283,
+      "step": 4637
+    },
+    {
+      "epoch": 8.588888888888889,
+      "grad_norm": 1.7743806641822972,
+      "learning_rate": 1.8263111665328362e-07,
+      "loss": 0.6298,
+      "step": 4638
+    },
+    {
+      "epoch": 8.59074074074074,
+      "grad_norm": 1.5208365051868078,
+      "learning_rate": 1.8262267076196252e-07,
+      "loss": 0.5782,
+      "step": 4639
+    },
+    {
+      "epoch": 8.592592592592592,
+      "grad_norm": 1.4851290397447496,
+      "learning_rate": 1.826142230130594e-07,
+      "loss": 0.6403,
+      "step": 4640
+    },
+    {
+      "epoch": 8.594444444444445,
+      "grad_norm": 1.409981785128526,
+      "learning_rate": 1.8260577340676413e-07,
+      "loss": 0.6583,
+      "step": 4641
+    },
+    {
+      "epoch": 8.596296296296297,
+      "grad_norm": 1.9323666449845667,
+      "learning_rate": 1.8259732194326675e-07,
+      "loss": 0.6623,
+      "step": 4642
+    },
+    {
+      "epoch": 8.598148148148148,
+      "grad_norm": 1.6448795921633081,
+      "learning_rate": 1.8258886862275725e-07,
+      "loss": 0.6429,
+      "step": 4643
+    },
+    {
+      "epoch": 8.6,
+      "grad_norm": 1.4409766365499421,
+      "learning_rate": 1.8258041344542562e-07,
+      "loss": 0.6624,
+      "step": 4644
+    },
+    {
+      "epoch": 8.601851851851851,
+      "grad_norm": 1.848940669302376,
+      "learning_rate": 1.8257195641146206e-07,
+      "loss": 0.59,
+      "step": 4645
+    },
+    {
+      "epoch": 8.603703703703705,
+      "grad_norm": 1.645895531465644,
+      "learning_rate": 1.8256349752105665e-07,
+      "loss": 0.6406,
+      "step": 4646
+    },
+    {
+      "epoch": 8.605555555555556,
+      "grad_norm": 1.5332580219964758,
+      "learning_rate": 1.825550367743996e-07,
+      "loss": 0.6083,
+      "step": 4647
+    },
+    {
+      "epoch": 8.607407407407408,
+      "grad_norm": 1.572436611417677,
+      "learning_rate": 1.8254657417168107e-07,
+      "loss": 0.6334,
+      "step": 4648
+    },
+    {
+      "epoch": 8.60925925925926,
+      "grad_norm": 1.4398395258793437,
+      "learning_rate": 1.8253810971309135e-07,
+      "loss": 0.6532,
+      "step": 4649
+    },
+    {
+      "epoch": 8.61111111111111,
+      "grad_norm": 1.642290473024365,
+      "learning_rate": 1.825296433988208e-07,
+      "loss": 0.6629,
+      "step": 4650
+    },
+    {
+      "epoch": 8.612962962962962,
+      "grad_norm": 1.3984362963954995,
+      "learning_rate": 1.8252117522905968e-07,
+      "loss": 0.6596,
+      "step": 4651
+    },
+    {
+      "epoch": 8.614814814814816,
+      "grad_norm": 1.6116596754796,
+      "learning_rate": 1.8251270520399844e-07,
+      "loss": 0.6489,
+      "step": 4652
+    },
+    {
+      "epoch": 8.616666666666667,
+      "grad_norm": 2.69604900841089,
+      "learning_rate": 1.8250423332382747e-07,
+      "loss": 0.6542,
+      "step": 4653
+    },
+    {
+      "epoch": 8.618518518518519,
+      "grad_norm": 1.4288185593949385,
+      "learning_rate": 1.8249575958873726e-07,
+      "loss": 0.6453,
+      "step": 4654
+    },
+    {
+      "epoch": 8.62037037037037,
+      "grad_norm": 1.647910173806169,
+      "learning_rate": 1.8248728399891832e-07,
+      "loss": 0.6345,
+      "step": 4655
+    },
+    {
+      "epoch": 8.622222222222222,
+      "grad_norm": 1.3281289359551343,
+      "learning_rate": 1.8247880655456122e-07,
+      "loss": 0.6835,
+      "step": 4656
+    },
+    {
+      "epoch": 8.624074074074073,
+      "grad_norm": 1.4710730760143462,
+      "learning_rate": 1.8247032725585653e-07,
+      "loss": 0.6275,
+      "step": 4657
+    },
+    {
+      "epoch": 8.625925925925927,
+      "grad_norm": 1.564139242155735,
+      "learning_rate": 1.824618461029949e-07,
+      "loss": 0.646,
+      "step": 4658
+    },
+    {
+      "epoch": 8.627777777777778,
+      "grad_norm": 1.5376627191948549,
+      "learning_rate": 1.8245336309616698e-07,
+      "loss": 0.5937,
+      "step": 4659
+    },
+    {
+      "epoch": 8.62962962962963,
+      "grad_norm": 1.4797688334607642,
+      "learning_rate": 1.8244487823556357e-07,
+      "loss": 0.6373,
+      "step": 4660
+    },
+    {
+      "epoch": 8.631481481481481,
+      "grad_norm": 1.5830035686800359,
+      "learning_rate": 1.824363915213753e-07,
+      "loss": 0.6417,
+      "step": 4661
+    },
+    {
+      "epoch": 8.633333333333333,
+      "grad_norm": 1.6981934717298532,
+      "learning_rate": 1.8242790295379312e-07,
+      "loss": 0.6655,
+      "step": 4662
+    },
+    {
+      "epoch": 8.635185185185184,
+      "grad_norm": 2.054784709026285,
+      "learning_rate": 1.8241941253300782e-07,
+      "loss": 0.6896,
+      "step": 4663
+    },
+    {
+      "epoch": 8.637037037037038,
+      "grad_norm": 1.598474932463622,
+      "learning_rate": 1.8241092025921025e-07,
+      "loss": 0.6477,
+      "step": 4664
+    },
+    {
+      "epoch": 8.63888888888889,
+      "grad_norm": 1.61900113925755,
+      "learning_rate": 1.8240242613259134e-07,
+      "loss": 0.6107,
+      "step": 4665
+    },
+    {
+      "epoch": 8.64074074074074,
+      "grad_norm": 1.6328890300869303,
+      "learning_rate": 1.8239393015334216e-07,
+      "loss": 0.667,
+      "step": 4666
+    },
+    {
+      "epoch": 8.642592592592592,
+      "grad_norm": 1.485073056746911,
+      "learning_rate": 1.8238543232165357e-07,
+      "loss": 0.6064,
+      "step": 4667
+    },
+    {
+      "epoch": 8.644444444444444,
+      "grad_norm": 1.6212773723414033,
+      "learning_rate": 1.8237693263771674e-07,
+      "loss": 0.6109,
+      "step": 4668
+    },
+    {
+      "epoch": 8.646296296296295,
+      "grad_norm": 1.554389555194729,
+      "learning_rate": 1.8236843110172276e-07,
+      "loss": 0.6528,
+      "step": 4669
+    },
+    {
+      "epoch": 8.648148148148149,
+      "grad_norm": 1.6280280529034825,
+      "learning_rate": 1.8235992771386272e-07,
+      "loss": 0.6346,
+      "step": 4670
+    },
+    {
+      "epoch": 8.65,
+      "grad_norm": 1.7458232937442923,
+      "learning_rate": 1.823514224743278e-07,
+      "loss": 0.6223,
+      "step": 4671
+    },
+    {
+      "epoch": 8.651851851851852,
+      "grad_norm": 1.585324461122069,
+      "learning_rate": 1.823429153833093e-07,
+      "loss": 0.6516,
+      "step": 4672
+    },
+    {
+      "epoch": 8.653703703703703,
+      "grad_norm": 1.8145245840940134,
+      "learning_rate": 1.8233440644099835e-07,
+      "loss": 0.6666,
+      "step": 4673
+    },
+    {
+      "epoch": 8.655555555555555,
+      "grad_norm": 1.647551951510653,
+      "learning_rate": 1.8232589564758634e-07,
+      "loss": 0.6099,
+      "step": 4674
+    },
+    {
+      "epoch": 8.657407407407408,
+      "grad_norm": 1.7238778767416179,
+      "learning_rate": 1.8231738300326461e-07,
+      "loss": 0.6231,
+      "step": 4675
+    },
+    {
+      "epoch": 8.65925925925926,
+      "grad_norm": 1.5130845337692778,
+      "learning_rate": 1.8230886850822458e-07,
+      "loss": 0.6302,
+      "step": 4676
+    },
+    {
+      "epoch": 8.661111111111111,
+      "grad_norm": 1.6029840751467057,
+      "learning_rate": 1.823003521626576e-07,
+      "loss": 0.672,
+      "step": 4677
+    },
+    {
+      "epoch": 8.662962962962963,
+      "grad_norm": 1.4008070921977813,
+      "learning_rate": 1.8229183396675518e-07,
+      "loss": 0.6595,
+      "step": 4678
+    },
+    {
+      "epoch": 8.664814814814815,
+      "grad_norm": 1.5119462980495688,
+      "learning_rate": 1.8228331392070881e-07,
+      "loss": 0.673,
+      "step": 4679
+    },
+    {
+      "epoch": 8.666666666666666,
+      "grad_norm": 1.498123257492758,
+      "learning_rate": 1.8227479202471012e-07,
+      "loss": 0.6271,
+      "step": 4680
+    },
+    {
+      "epoch": 8.668518518518518,
+      "grad_norm": 1.8742693532260817,
+      "learning_rate": 1.8226626827895063e-07,
+      "loss": 0.6118,
+      "step": 4681
+    },
+    {
+      "epoch": 8.670370370370371,
+      "grad_norm": 2.63094956801755,
+      "learning_rate": 1.8225774268362198e-07,
+      "loss": 0.6411,
+      "step": 4682
+    },
+    {
+      "epoch": 8.672222222222222,
+      "grad_norm": 1.5429727271348872,
+      "learning_rate": 1.8224921523891585e-07,
+      "loss": 0.6299,
+      "step": 4683
+    },
+    {
+      "epoch": 8.674074074074074,
+      "grad_norm": 1.8526143298357611,
+      "learning_rate": 1.82240685945024e-07,
+      "loss": 0.6999,
+      "step": 4684
+    },
+    {
+      "epoch": 8.675925925925926,
+      "grad_norm": 1.702897015000899,
+      "learning_rate": 1.8223215480213818e-07,
+      "loss": 0.6128,
+      "step": 4685
+    },
+    {
+      "epoch": 8.677777777777777,
+      "grad_norm": 1.4412285634416815,
+      "learning_rate": 1.8222362181045016e-07,
+      "loss": 0.664,
+      "step": 4686
+    },
+    {
+      "epoch": 8.67962962962963,
+      "grad_norm": 1.667561604277166,
+      "learning_rate": 1.8221508697015182e-07,
+      "loss": 0.6524,
+      "step": 4687
+    },
+    {
+      "epoch": 8.681481481481482,
+      "grad_norm": 2.2931596283003164,
+      "learning_rate": 1.8220655028143503e-07,
+      "loss": 0.6119,
+      "step": 4688
+    },
+    {
+      "epoch": 8.683333333333334,
+      "grad_norm": 1.6072472899378067,
+      "learning_rate": 1.821980117444917e-07,
+      "loss": 0.6313,
+      "step": 4689
+    },
+    {
+      "epoch": 8.685185185185185,
+      "grad_norm": 1.6181307696646436,
+      "learning_rate": 1.8218947135951385e-07,
+      "loss": 0.6181,
+      "step": 4690
+    },
+    {
+      "epoch": 8.687037037037037,
+      "grad_norm": 1.3917733867377007,
+      "learning_rate": 1.8218092912669346e-07,
+      "loss": 0.6825,
+      "step": 4691
+    },
+    {
+      "epoch": 8.688888888888888,
+      "grad_norm": 1.54247117128281,
+      "learning_rate": 1.8217238504622255e-07,
+      "loss": 0.6575,
+      "step": 4692
+    },
+    {
+      "epoch": 8.690740740740742,
+      "grad_norm": 1.5563706340018573,
+      "learning_rate": 1.8216383911829327e-07,
+      "loss": 0.6197,
+      "step": 4693
+    },
+    {
+      "epoch": 8.692592592592593,
+      "grad_norm": 1.5874580072919338,
+      "learning_rate": 1.8215529134309774e-07,
+      "loss": 0.6482,
+      "step": 4694
+    },
+    {
+      "epoch": 8.694444444444445,
+      "grad_norm": 1.7687437304554279,
+      "learning_rate": 1.8214674172082814e-07,
+      "loss": 0.6855,
+      "step": 4695
+    },
+    {
+      "epoch": 8.696296296296296,
+      "grad_norm": 1.6009761158263907,
+      "learning_rate": 1.8213819025167664e-07,
+      "loss": 0.637,
+      "step": 4696
+    },
+    {
+      "epoch": 8.698148148148148,
+      "grad_norm": 1.4795538809233624,
+      "learning_rate": 1.8212963693583557e-07,
+      "loss": 0.6721,
+      "step": 4697
+    },
+    {
+      "epoch": 8.7,
+      "grad_norm": 1.719666487858422,
+      "learning_rate": 1.821210817734972e-07,
+      "loss": 0.6213,
+      "step": 4698
+    },
+    {
+      "epoch": 8.701851851851853,
+      "grad_norm": 1.488734611142881,
+      "learning_rate": 1.8211252476485384e-07,
+      "loss": 0.6431,
+      "step": 4699
+    },
+    {
+      "epoch": 8.703703703703704,
+      "grad_norm": 1.5299407513110628,
+      "learning_rate": 1.8210396591009794e-07,
+      "loss": 0.6551,
+      "step": 4700
+    },
+    {
+      "epoch": 8.705555555555556,
+      "grad_norm": 1.8923040523516026,
+      "learning_rate": 1.8209540520942189e-07,
+      "loss": 0.6621,
+      "step": 4701
+    },
+    {
+      "epoch": 8.707407407407407,
+      "grad_norm": 1.5644709130386947,
+      "learning_rate": 1.8208684266301818e-07,
+      "loss": 0.6582,
+      "step": 4702
+    },
+    {
+      "epoch": 8.709259259259259,
+      "grad_norm": 1.5950998731019763,
+      "learning_rate": 1.8207827827107924e-07,
+      "loss": 0.6127,
+      "step": 4703
+    },
+    {
+      "epoch": 8.71111111111111,
+      "grad_norm": 1.5045411998125697,
+      "learning_rate": 1.8206971203379773e-07,
+      "loss": 0.6344,
+      "step": 4704
+    },
+    {
+      "epoch": 8.712962962962964,
+      "grad_norm": 1.7444162133434356,
+      "learning_rate": 1.8206114395136622e-07,
+      "loss": 0.6268,
+      "step": 4705
+    },
+    {
+      "epoch": 8.714814814814815,
+      "grad_norm": 1.6332213368195259,
+      "learning_rate": 1.820525740239773e-07,
+      "loss": 0.6272,
+      "step": 4706
+    },
+    {
+      "epoch": 8.716666666666667,
+      "grad_norm": 1.4707964202906618,
+      "learning_rate": 1.8204400225182363e-07,
+      "loss": 0.6771,
+      "step": 4707
+    },
+    {
+      "epoch": 8.718518518518518,
+      "grad_norm": 1.4141728248616752,
+      "learning_rate": 1.82035428635098e-07,
+      "loss": 0.6369,
+      "step": 4708
+    },
+    {
+      "epoch": 8.72037037037037,
+      "grad_norm": 1.6191809489828612,
+      "learning_rate": 1.820268531739931e-07,
+      "loss": 0.6402,
+      "step": 4709
+    },
+    {
+      "epoch": 8.722222222222221,
+      "grad_norm": 1.5002667373306158,
+      "learning_rate": 1.820182758687018e-07,
+      "loss": 0.6276,
+      "step": 4710
+    },
+    {
+      "epoch": 8.724074074074075,
+      "grad_norm": 1.6404742064137752,
+      "learning_rate": 1.8200969671941685e-07,
+      "loss": 0.6476,
+      "step": 4711
+    },
+    {
+      "epoch": 8.725925925925926,
+      "grad_norm": 1.5602978649137849,
+      "learning_rate": 1.820011157263312e-07,
+      "loss": 0.6384,
+      "step": 4712
+    },
+    {
+      "epoch": 8.727777777777778,
+      "grad_norm": 1.6241076608343443,
+      "learning_rate": 1.8199253288963778e-07,
+      "loss": 0.6191,
+      "step": 4713
+    },
+    {
+      "epoch": 8.72962962962963,
+      "grad_norm": 1.5924767719461121,
+      "learning_rate": 1.819839482095295e-07,
+      "loss": 0.622,
+      "step": 4714
+    },
+    {
+      "epoch": 8.731481481481481,
+      "grad_norm": 1.568942150358102,
+      "learning_rate": 1.8197536168619944e-07,
+      "loss": 0.6352,
+      "step": 4715
+    },
+    {
+      "epoch": 8.733333333333333,
+      "grad_norm": 1.5984664339770507,
+      "learning_rate": 1.819667733198406e-07,
+      "loss": 0.6301,
+      "step": 4716
+    },
+    {
+      "epoch": 8.735185185185186,
+      "grad_norm": 1.941038959656708,
+      "learning_rate": 1.8195818311064603e-07,
+      "loss": 0.5882,
+      "step": 4717
+    },
+    {
+      "epoch": 8.737037037037037,
+      "grad_norm": 1.549290053544421,
+      "learning_rate": 1.8194959105880895e-07,
+      "loss": 0.6326,
+      "step": 4718
+    },
+    {
+      "epoch": 8.738888888888889,
+      "grad_norm": 1.3673495020605901,
+      "learning_rate": 1.8194099716452248e-07,
+      "loss": 0.6238,
+      "step": 4719
+    },
+    {
+      "epoch": 8.74074074074074,
+      "grad_norm": 1.3991381699312113,
+      "learning_rate": 1.8193240142797985e-07,
+      "loss": 0.6163,
+      "step": 4720
+    },
+    {
+      "epoch": 8.742592592592592,
+      "grad_norm": 2.000034218585617,
+      "learning_rate": 1.8192380384937431e-07,
+      "loss": 0.6271,
+      "step": 4721
+    },
+    {
+      "epoch": 8.744444444444444,
+      "grad_norm": 1.5598029480816074,
+      "learning_rate": 1.8191520442889918e-07,
+      "loss": 0.6042,
+      "step": 4722
+    },
+    {
+      "epoch": 8.746296296296297,
+      "grad_norm": 1.478153850381044,
+      "learning_rate": 1.8190660316674773e-07,
+      "loss": 0.6362,
+      "step": 4723
+    },
+    {
+      "epoch": 8.748148148148148,
+      "grad_norm": 1.4748774850924247,
+      "learning_rate": 1.818980000631134e-07,
+      "loss": 0.6448,
+      "step": 4724
+    },
+    {
+      "epoch": 8.75,
+      "grad_norm": 1.7754842521417857,
+      "learning_rate": 1.8188939511818963e-07,
+      "loss": 0.6413,
+      "step": 4725
+    },
+    {
+      "epoch": 8.751851851851852,
+      "grad_norm": 1.4849536229583662,
+      "learning_rate": 1.8188078833216983e-07,
+      "loss": 0.6098,
+      "step": 4726
+    },
+    {
+      "epoch": 8.753703703703703,
+      "grad_norm": 1.5398261095887904,
+      "learning_rate": 1.818721797052475e-07,
+      "loss": 0.6302,
+      "step": 4727
+    },
+    {
+      "epoch": 8.755555555555556,
+      "grad_norm": 1.5000368681911562,
+      "learning_rate": 1.8186356923761626e-07,
+      "loss": 0.6418,
+      "step": 4728
+    },
+    {
+      "epoch": 8.757407407407408,
+      "grad_norm": 1.5013206321264958,
+      "learning_rate": 1.818549569294696e-07,
+      "loss": 0.6709,
+      "step": 4729
+    },
+    {
+      "epoch": 8.75925925925926,
+      "grad_norm": 2.6487533430746613,
+      "learning_rate": 1.8184634278100122e-07,
+      "loss": 0.6348,
+      "step": 4730
+    },
+    {
+      "epoch": 8.761111111111111,
+      "grad_norm": 1.4688366605716632,
+      "learning_rate": 1.8183772679240472e-07,
+      "loss": 0.6015,
+      "step": 4731
+    },
+    {
+      "epoch": 8.762962962962963,
+      "grad_norm": 1.2984630671770157,
+      "learning_rate": 1.818291089638739e-07,
+      "loss": 0.6261,
+      "step": 4732
+    },
+    {
+      "epoch": 8.764814814814814,
+      "grad_norm": 1.5969695983099004,
+      "learning_rate": 1.8182048929560242e-07,
+      "loss": 0.6475,
+      "step": 4733
+    },
+    {
+      "epoch": 8.766666666666667,
+      "grad_norm": 1.7820378959978886,
+      "learning_rate": 1.8181186778778416e-07,
+      "loss": 0.6313,
+      "step": 4734
+    },
+    {
+      "epoch": 8.768518518518519,
+      "grad_norm": 1.4622881840933173,
+      "learning_rate": 1.8180324444061292e-07,
+      "loss": 0.6233,
+      "step": 4735
+    },
+    {
+      "epoch": 8.77037037037037,
+      "grad_norm": 1.4612519905828116,
+      "learning_rate": 1.8179461925428254e-07,
+      "loss": 0.6488,
+      "step": 4736
+    },
+    {
+      "epoch": 8.772222222222222,
+      "grad_norm": 1.5163673743922017,
+      "learning_rate": 1.81785992228987e-07,
+      "loss": 0.6919,
+      "step": 4737
+    },
+    {
+      "epoch": 8.774074074074074,
+      "grad_norm": 1.6644282014852592,
+      "learning_rate": 1.817773633649202e-07,
+      "loss": 0.6741,
+      "step": 4738
+    },
+    {
+      "epoch": 8.775925925925925,
+      "grad_norm": 1.4647295128471252,
+      "learning_rate": 1.817687326622762e-07,
+      "loss": 0.648,
+      "step": 4739
+    },
+    {
+      "epoch": 8.777777777777779,
+      "grad_norm": 1.9867233597039433,
+      "learning_rate": 1.8176010012124895e-07,
+      "loss": 0.6546,
+      "step": 4740
+    },
+    {
+      "epoch": 8.77962962962963,
+      "grad_norm": 1.4490874637681297,
+      "learning_rate": 1.8175146574203264e-07,
+      "loss": 0.6658,
+      "step": 4741
+    },
+    {
+      "epoch": 8.781481481481482,
+      "grad_norm": 1.701503399174617,
+      "learning_rate": 1.8174282952482133e-07,
+      "loss": 0.6899,
+      "step": 4742
+    },
+    {
+      "epoch": 8.783333333333333,
+      "grad_norm": 1.5454750252062397,
+      "learning_rate": 1.8173419146980922e-07,
+      "loss": 0.6478,
+      "step": 4743
+    },
+    {
+      "epoch": 8.785185185185185,
+      "grad_norm": 1.5586226965585541,
+      "learning_rate": 1.817255515771905e-07,
+      "loss": 0.6438,
+      "step": 4744
+    },
+    {
+      "epoch": 8.787037037037036,
+      "grad_norm": 1.5067538618749443,
+      "learning_rate": 1.817169098471594e-07,
+      "loss": 0.6252,
+      "step": 4745
+    },
+    {
+      "epoch": 8.78888888888889,
+      "grad_norm": 1.3632548071742268,
+      "learning_rate": 1.8170826627991023e-07,
+      "loss": 0.614,
+      "step": 4746
+    },
+    {
+      "epoch": 8.790740740740741,
+      "grad_norm": 1.5590647594041502,
+      "learning_rate": 1.816996208756373e-07,
+      "loss": 0.6555,
+      "step": 4747
+    },
+    {
+      "epoch": 8.792592592592593,
+      "grad_norm": 1.5168107560316604,
+      "learning_rate": 1.8169097363453503e-07,
+      "loss": 0.6203,
+      "step": 4748
+    },
+    {
+      "epoch": 8.794444444444444,
+      "grad_norm": 1.7194309072778642,
+      "learning_rate": 1.8168232455679778e-07,
+      "loss": 0.6388,
+      "step": 4749
+    },
+    {
+      "epoch": 8.796296296296296,
+      "grad_norm": 1.5391918004481664,
+      "learning_rate": 1.8167367364262004e-07,
+      "loss": 0.6447,
+      "step": 4750
+    },
+    {
+      "epoch": 8.798148148148147,
+      "grad_norm": 1.8542247574495883,
+      "learning_rate": 1.816650208921963e-07,
+      "loss": 0.6345,
+      "step": 4751
+    },
+    {
+      "epoch": 8.8,
+      "grad_norm": 3.1963687310357507,
+      "learning_rate": 1.8165636630572108e-07,
+      "loss": 0.6004,
+      "step": 4752
+    },
+    {
+      "epoch": 8.801851851851852,
+      "grad_norm": 1.9949250538162673,
+      "learning_rate": 1.8164770988338897e-07,
+      "loss": 0.662,
+      "step": 4753
+    },
+    {
+      "epoch": 8.803703703703704,
+      "grad_norm": 1.880751629648909,
+      "learning_rate": 1.816390516253946e-07,
+      "loss": 0.6191,
+      "step": 4754
+    },
+    {
+      "epoch": 8.805555555555555,
+      "grad_norm": 1.7011827412918006,
+      "learning_rate": 1.8163039153193262e-07,
+      "loss": 0.6,
+      "step": 4755
+    },
+    {
+      "epoch": 8.807407407407407,
+      "grad_norm": 1.4842433890205666,
+      "learning_rate": 1.8162172960319774e-07,
+      "loss": 0.6321,
+      "step": 4756
+    },
+    {
+      "epoch": 8.809259259259258,
+      "grad_norm": 1.4332477888658655,
+      "learning_rate": 1.8161306583938466e-07,
+      "loss": 0.638,
+      "step": 4757
+    },
+    {
+      "epoch": 8.811111111111112,
+      "grad_norm": 1.5498554110002054,
+      "learning_rate": 1.8160440024068823e-07,
+      "loss": 0.6473,
+      "step": 4758
+    },
+    {
+      "epoch": 8.812962962962963,
+      "grad_norm": 1.537076131959102,
+      "learning_rate": 1.8159573280730328e-07,
+      "loss": 0.6513,
+      "step": 4759
+    },
+    {
+      "epoch": 8.814814814814815,
+      "grad_norm": 1.6855792368040095,
+      "learning_rate": 1.815870635394246e-07,
+      "loss": 0.6044,
+      "step": 4760
+    },
+    {
+      "epoch": 8.816666666666666,
+      "grad_norm": 1.5394131205208132,
+      "learning_rate": 1.8157839243724715e-07,
+      "loss": 0.6375,
+      "step": 4761
+    },
+    {
+      "epoch": 8.818518518518518,
+      "grad_norm": 1.5437336959923607,
+      "learning_rate": 1.815697195009659e-07,
+      "loss": 0.6287,
+      "step": 4762
+    },
+    {
+      "epoch": 8.82037037037037,
+      "grad_norm": 1.5997878235485652,
+      "learning_rate": 1.815610447307758e-07,
+      "loss": 0.6449,
+      "step": 4763
+    },
+    {
+      "epoch": 8.822222222222223,
+      "grad_norm": 2.0988768306639973,
+      "learning_rate": 1.815523681268719e-07,
+      "loss": 0.6324,
+      "step": 4764
+    },
+    {
+      "epoch": 8.824074074074074,
+      "grad_norm": 2.1354272439864985,
+      "learning_rate": 1.8154368968944925e-07,
+      "loss": 0.6275,
+      "step": 4765
+    },
+    {
+      "epoch": 8.825925925925926,
+      "grad_norm": 1.5347030507272574,
+      "learning_rate": 1.8153500941870302e-07,
+      "loss": 0.6508,
+      "step": 4766
+    },
+    {
+      "epoch": 8.827777777777778,
+      "grad_norm": 1.661853429257589,
+      "learning_rate": 1.8152632731482833e-07,
+      "loss": 0.6192,
+      "step": 4767
+    },
+    {
+      "epoch": 8.829629629629629,
+      "grad_norm": 1.3367538919049962,
+      "learning_rate": 1.8151764337802032e-07,
+      "loss": 0.6488,
+      "step": 4768
+    },
+    {
+      "epoch": 8.831481481481482,
+      "grad_norm": 1.3774919246792268,
+      "learning_rate": 1.8150895760847437e-07,
+      "loss": 0.6194,
+      "step": 4769
+    },
+    {
+      "epoch": 8.833333333333334,
+      "grad_norm": 1.630670597639851,
+      "learning_rate": 1.8150027000638563e-07,
+      "loss": 0.6709,
+      "step": 4770
+    },
+    {
+      "epoch": 8.835185185185185,
+      "grad_norm": 1.4075565066123874,
+      "learning_rate": 1.814915805719495e-07,
+      "loss": 0.6062,
+      "step": 4771
+    },
+    {
+      "epoch": 8.837037037037037,
+      "grad_norm": 1.6812371369625405,
+      "learning_rate": 1.8148288930536129e-07,
+      "loss": 0.6346,
+      "step": 4772
+    },
+    {
+      "epoch": 8.838888888888889,
+      "grad_norm": 1.8649683367785135,
+      "learning_rate": 1.8147419620681643e-07,
+      "loss": 0.6228,
+      "step": 4773
+    },
+    {
+      "epoch": 8.84074074074074,
+      "grad_norm": 1.4885054629074472,
+      "learning_rate": 1.8146550127651037e-07,
+      "loss": 0.6511,
+      "step": 4774
+    },
+    {
+      "epoch": 8.842592592592592,
+      "grad_norm": 1.6236803099433943,
+      "learning_rate": 1.8145680451463857e-07,
+      "loss": 0.6328,
+      "step": 4775
+    },
+    {
+      "epoch": 8.844444444444445,
+      "grad_norm": 1.940368928585568,
+      "learning_rate": 1.8144810592139655e-07,
+      "loss": 0.6136,
+      "step": 4776
+    },
+    {
+      "epoch": 8.846296296296297,
+      "grad_norm": 1.4758065614773443,
+      "learning_rate": 1.8143940549697993e-07,
+      "loss": 0.6577,
+      "step": 4777
+    },
+    {
+      "epoch": 8.848148148148148,
+      "grad_norm": 1.4390984157305504,
+      "learning_rate": 1.8143070324158427e-07,
+      "loss": 0.665,
+      "step": 4778
+    },
+    {
+      "epoch": 8.85,
+      "grad_norm": 1.7184382133170344,
+      "learning_rate": 1.8142199915540526e-07,
+      "loss": 0.6602,
+      "step": 4779
+    },
+    {
+      "epoch": 8.851851851851851,
+      "grad_norm": 1.6384462288972075,
+      "learning_rate": 1.8141329323863858e-07,
+      "loss": 0.661,
+      "step": 4780
+    },
+    {
+      "epoch": 8.853703703703705,
+      "grad_norm": 2.314472491481749,
+      "learning_rate": 1.814045854914799e-07,
+      "loss": 0.627,
+      "step": 4781
+    },
+    {
+      "epoch": 8.855555555555556,
+      "grad_norm": 1.5186661668525334,
+      "learning_rate": 1.813958759141251e-07,
+      "loss": 0.615,
+      "step": 4782
+    },
+    {
+      "epoch": 8.857407407407408,
+      "grad_norm": 1.4791592421566457,
+      "learning_rate": 1.813871645067699e-07,
+      "loss": 0.6223,
+      "step": 4783
+    },
+    {
+      "epoch": 8.85925925925926,
+      "grad_norm": 1.5437100243956041,
+      "learning_rate": 1.813784512696102e-07,
+      "loss": 0.6025,
+      "step": 4784
+    },
+    {
+      "epoch": 8.86111111111111,
+      "grad_norm": 1.643807220976132,
+      "learning_rate": 1.8136973620284193e-07,
+      "loss": 0.666,
+      "step": 4785
+    },
+    {
+      "epoch": 8.862962962962962,
+      "grad_norm": 1.6739861880512772,
+      "learning_rate": 1.81361019306661e-07,
+      "loss": 0.6507,
+      "step": 4786
+    },
+    {
+      "epoch": 8.864814814814816,
+      "grad_norm": 1.5952673058975069,
+      "learning_rate": 1.8135230058126333e-07,
+      "loss": 0.6434,
+      "step": 4787
+    },
+    {
+      "epoch": 8.866666666666667,
+      "grad_norm": 1.6437273328889874,
+      "learning_rate": 1.81343580026845e-07,
+      "loss": 0.6612,
+      "step": 4788
+    },
+    {
+      "epoch": 8.868518518518519,
+      "grad_norm": 1.622596914954736,
+      "learning_rate": 1.813348576436021e-07,
+      "loss": 0.6538,
+      "step": 4789
+    },
+    {
+      "epoch": 8.87037037037037,
+      "grad_norm": 1.3060592873421302,
+      "learning_rate": 1.8132613343173068e-07,
+      "loss": 0.6221,
+      "step": 4790
+    },
+    {
+      "epoch": 8.872222222222222,
+      "grad_norm": 1.5810597563113369,
+      "learning_rate": 1.8131740739142692e-07,
+      "loss": 0.6429,
+      "step": 4791
+    },
+    {
+      "epoch": 8.874074074074073,
+      "grad_norm": 1.6395330414936504,
+      "learning_rate": 1.8130867952288696e-07,
+      "loss": 0.5895,
+      "step": 4792
+    },
+    {
+      "epoch": 8.875925925925927,
+      "grad_norm": 1.5921565627648888,
+      "learning_rate": 1.8129994982630707e-07,
+      "loss": 0.6883,
+      "step": 4793
+    },
+    {
+      "epoch": 8.877777777777778,
+      "grad_norm": 1.5861378481811457,
+      "learning_rate": 1.8129121830188352e-07,
+      "loss": 0.6483,
+      "step": 4794
+    },
+    {
+      "epoch": 8.87962962962963,
+      "grad_norm": 2.0366477235989655,
+      "learning_rate": 1.8128248494981256e-07,
+      "loss": 0.6239,
+      "step": 4795
+    },
+    {
+      "epoch": 8.881481481481481,
+      "grad_norm": 1.4555855106498405,
+      "learning_rate": 1.8127374977029063e-07,
+      "loss": 0.6278,
+      "step": 4796
+    },
+    {
+      "epoch": 8.883333333333333,
+      "grad_norm": 1.428021416975343,
+      "learning_rate": 1.8126501276351404e-07,
+      "loss": 0.6176,
+      "step": 4797
+    },
+    {
+      "epoch": 8.885185185185184,
+      "grad_norm": 1.4028919477392796,
+      "learning_rate": 1.8125627392967923e-07,
+      "loss": 0.675,
+      "step": 4798
+    },
+    {
+      "epoch": 8.887037037037038,
+      "grad_norm": 1.6754221304722121,
+      "learning_rate": 1.812475332689827e-07,
+      "loss": 0.6281,
+      "step": 4799
+    },
+    {
+      "epoch": 8.88888888888889,
+      "grad_norm": 1.7550058280564609,
+      "learning_rate": 1.8123879078162097e-07,
+      "loss": 0.6626,
+      "step": 4800
+    },
+    {
+      "epoch": 8.89074074074074,
+      "grad_norm": 1.615829983650522,
+      "learning_rate": 1.812300464677906e-07,
+      "loss": 0.6121,
+      "step": 4801
+    },
+    {
+      "epoch": 8.892592592592592,
+      "grad_norm": 1.7515681734317055,
+      "learning_rate": 1.8122130032768812e-07,
+      "loss": 0.6196,
+      "step": 4802
+    },
+    {
+      "epoch": 8.894444444444444,
+      "grad_norm": 1.4298392742036237,
+      "learning_rate": 1.8121255236151023e-07,
+      "loss": 0.6614,
+      "step": 4803
+    },
+    {
+      "epoch": 8.896296296296295,
+      "grad_norm": 1.5809941412851993,
+      "learning_rate": 1.812038025694536e-07,
+      "loss": 0.6592,
+      "step": 4804
+    },
+    {
+      "epoch": 8.898148148148149,
+      "grad_norm": 1.514028225205355,
+      "learning_rate": 1.8119505095171497e-07,
+      "loss": 0.6342,
+      "step": 4805
+    },
+    {
+      "epoch": 8.9,
+      "grad_norm": 1.4341172218221372,
+      "learning_rate": 1.8118629750849104e-07,
+      "loss": 0.6172,
+      "step": 4806
+    },
+    {
+      "epoch": 8.901851851851852,
+      "grad_norm": 1.525011857098194,
+      "learning_rate": 1.8117754223997866e-07,
+      "loss": 0.6262,
+      "step": 4807
+    },
+    {
+      "epoch": 8.903703703703703,
+      "grad_norm": 1.6771318069019463,
+      "learning_rate": 1.8116878514637464e-07,
+      "loss": 0.6525,
+      "step": 4808
+    },
+    {
+      "epoch": 8.905555555555555,
+      "grad_norm": 1.7982890756177687,
+      "learning_rate": 1.811600262278759e-07,
+      "loss": 0.6123,
+      "step": 4809
+    },
+    {
+      "epoch": 8.907407407407408,
+      "grad_norm": 1.5055494196298373,
+      "learning_rate": 1.811512654846793e-07,
+      "loss": 0.6593,
+      "step": 4810
+    },
+    {
+      "epoch": 8.90925925925926,
+      "grad_norm": 1.7677444419090338,
+      "learning_rate": 1.8114250291698187e-07,
+      "loss": 0.6361,
+      "step": 4811
+    },
+    {
+      "epoch": 8.911111111111111,
+      "grad_norm": 1.3275828657322852,
+      "learning_rate": 1.811337385249806e-07,
+      "loss": 0.6298,
+      "step": 4812
+    },
+    {
+      "epoch": 8.912962962962963,
+      "grad_norm": 1.5191138290104935,
+      "learning_rate": 1.8112497230887256e-07,
+      "loss": 0.6709,
+      "step": 4813
+    },
+    {
+      "epoch": 8.914814814814815,
+      "grad_norm": 1.7926323308926118,
+      "learning_rate": 1.8111620426885476e-07,
+      "loss": 0.6325,
+      "step": 4814
+    },
+    {
+      "epoch": 8.916666666666666,
+      "grad_norm": 1.4347454040040613,
+      "learning_rate": 1.8110743440512444e-07,
+      "loss": 0.6121,
+      "step": 4815
+    },
+    {
+      "epoch": 8.918518518518518,
+      "grad_norm": 1.4312511052912815,
+      "learning_rate": 1.8109866271787865e-07,
+      "loss": 0.6374,
+      "step": 4816
+    },
+    {
+      "epoch": 8.920370370370371,
+      "grad_norm": 1.4237599959113545,
+      "learning_rate": 1.8108988920731469e-07,
+      "loss": 0.6318,
+      "step": 4817
+    },
+    {
+      "epoch": 8.922222222222222,
+      "grad_norm": 1.7623743746650833,
+      "learning_rate": 1.810811138736298e-07,
+      "loss": 0.6133,
+      "step": 4818
+    },
+    {
+      "epoch": 8.924074074074074,
+      "grad_norm": 1.7303734268128559,
+      "learning_rate": 1.8107233671702122e-07,
+      "loss": 0.6009,
+      "step": 4819
+    },
+    {
+      "epoch": 8.925925925925926,
+      "grad_norm": 2.0789529512237945,
+      "learning_rate": 1.8106355773768636e-07,
+      "loss": 0.6569,
+      "step": 4820
+    },
+    {
+      "epoch": 8.927777777777777,
+      "grad_norm": 1.4620603123364848,
+      "learning_rate": 1.8105477693582252e-07,
+      "loss": 0.6797,
+      "step": 4821
+    },
+    {
+      "epoch": 8.92962962962963,
+      "grad_norm": 1.5127297967909374,
+      "learning_rate": 1.8104599431162715e-07,
+      "loss": 0.6371,
+      "step": 4822
+    },
+    {
+      "epoch": 8.931481481481482,
+      "grad_norm": 2.3423121292169826,
+      "learning_rate": 1.8103720986529773e-07,
+      "loss": 0.6451,
+      "step": 4823
+    },
+    {
+      "epoch": 8.933333333333334,
+      "grad_norm": 1.5080718483619968,
+      "learning_rate": 1.8102842359703175e-07,
+      "loss": 0.6062,
+      "step": 4824
+    },
+    {
+      "epoch": 8.935185185185185,
+      "grad_norm": 1.560177617698358,
+      "learning_rate": 1.8101963550702674e-07,
+      "loss": 0.6392,
+      "step": 4825
+    },
+    {
+      "epoch": 8.937037037037037,
+      "grad_norm": 1.3734722786351372,
+      "learning_rate": 1.8101084559548023e-07,
+      "loss": 0.6425,
+      "step": 4826
+    },
+    {
+      "epoch": 8.938888888888888,
+      "grad_norm": 1.5953822169552891,
+      "learning_rate": 1.8100205386258995e-07,
+      "loss": 0.6275,
+      "step": 4827
+    },
+    {
+      "epoch": 8.940740740740742,
+      "grad_norm": 1.6082485311037924,
+      "learning_rate": 1.8099326030855346e-07,
+      "loss": 0.6271,
+      "step": 4828
+    },
+    {
+      "epoch": 8.942592592592593,
+      "grad_norm": 1.4954702487737823,
+      "learning_rate": 1.8098446493356852e-07,
+      "loss": 0.6376,
+      "step": 4829
+    },
+    {
+      "epoch": 8.944444444444445,
+      "grad_norm": 1.6226568298952204,
+      "learning_rate": 1.8097566773783285e-07,
+      "loss": 0.5922,
+      "step": 4830
+    },
+    {
+      "epoch": 8.946296296296296,
+      "grad_norm": 1.922238094573983,
+      "learning_rate": 1.8096686872154425e-07,
+      "loss": 0.6186,
+      "step": 4831
+    },
+    {
+      "epoch": 8.948148148148148,
+      "grad_norm": 1.5965934037450158,
+      "learning_rate": 1.8095806788490054e-07,
+      "loss": 0.6601,
+      "step": 4832
+    },
+    {
+      "epoch": 8.95,
+      "grad_norm": 2.3586461110550383,
+      "learning_rate": 1.8094926522809956e-07,
+      "loss": 0.6373,
+      "step": 4833
+    },
+    {
+      "epoch": 8.951851851851853,
+      "grad_norm": 1.503350121104173,
+      "learning_rate": 1.8094046075133927e-07,
+      "loss": 0.6279,
+      "step": 4834
+    },
+    {
+      "epoch": 8.953703703703704,
+      "grad_norm": 1.5018749230431285,
+      "learning_rate": 1.809316544548176e-07,
+      "loss": 0.6511,
+      "step": 4835
+    },
+    {
+      "epoch": 8.955555555555556,
+      "grad_norm": 1.5233438064006162,
+      "learning_rate": 1.8092284633873248e-07,
+      "loss": 0.6484,
+      "step": 4836
+    },
+    {
+      "epoch": 8.957407407407407,
+      "grad_norm": 1.475330476269923,
+      "learning_rate": 1.8091403640328205e-07,
+      "loss": 0.6559,
+      "step": 4837
+    },
+    {
+      "epoch": 8.959259259259259,
+      "grad_norm": 2.1607204789682855,
+      "learning_rate": 1.809052246486643e-07,
+      "loss": 0.6151,
+      "step": 4838
+    },
+    {
+      "epoch": 8.96111111111111,
+      "grad_norm": 1.5687888980311786,
+      "learning_rate": 1.808964110750774e-07,
+      "loss": 0.6497,
+      "step": 4839
+    },
+    {
+      "epoch": 8.962962962962964,
+      "grad_norm": 1.3948814041499955,
+      "learning_rate": 1.808875956827194e-07,
+      "loss": 0.6475,
+      "step": 4840
+    },
+    {
+      "epoch": 8.964814814814815,
+      "grad_norm": 1.519589769064513,
+      "learning_rate": 1.808787784717886e-07,
+      "loss": 0.6629,
+      "step": 4841
+    },
+    {
+      "epoch": 8.966666666666667,
+      "grad_norm": 2.110036887270788,
+      "learning_rate": 1.8086995944248317e-07,
+      "loss": 0.6949,
+      "step": 4842
+    },
+    {
+      "epoch": 8.968518518518518,
+      "grad_norm": 1.691389530965149,
+      "learning_rate": 1.8086113859500146e-07,
+      "loss": 0.6476,
+      "step": 4843
+    },
+    {
+      "epoch": 8.97037037037037,
+      "grad_norm": 1.6704512669708718,
+      "learning_rate": 1.808523159295417e-07,
+      "loss": 0.6792,
+      "step": 4844
+    },
+    {
+      "epoch": 8.972222222222221,
+      "grad_norm": 1.4432593127391118,
+      "learning_rate": 1.8084349144630228e-07,
+      "loss": 0.6008,
+      "step": 4845
+    },
+    {
+      "epoch": 8.974074074074075,
+      "grad_norm": 1.6026957206222825,
+      "learning_rate": 1.8083466514548163e-07,
+      "loss": 0.6528,
+      "step": 4846
+    },
+    {
+      "epoch": 8.975925925925926,
+      "grad_norm": 1.7981805280056289,
+      "learning_rate": 1.8082583702727817e-07,
+      "loss": 0.6343,
+      "step": 4847
+    },
+    {
+      "epoch": 8.977777777777778,
+      "grad_norm": 1.9680978528804036,
+      "learning_rate": 1.8081700709189037e-07,
+      "loss": 0.6557,
+      "step": 4848
+    },
+    {
+      "epoch": 8.97962962962963,
+      "grad_norm": 1.6198450599379675,
+      "learning_rate": 1.808081753395167e-07,
+      "loss": 0.6605,
+      "step": 4849
+    },
+    {
+      "epoch": 8.981481481481481,
+      "grad_norm": 1.8288338170075198,
+      "learning_rate": 1.8079934177035584e-07,
+      "loss": 0.6544,
+      "step": 4850
+    },
+    {
+      "epoch": 8.983333333333333,
+      "grad_norm": 1.5415661748127376,
+      "learning_rate": 1.8079050638460632e-07,
+      "loss": 0.6511,
+      "step": 4851
+    },
+    {
+      "epoch": 8.985185185185186,
+      "grad_norm": 1.429545976687254,
+      "learning_rate": 1.8078166918246678e-07,
+      "loss": 0.6494,
+      "step": 4852
+    },
+    {
+      "epoch": 8.987037037037037,
+      "grad_norm": 1.773454452391686,
+      "learning_rate": 1.8077283016413592e-07,
+      "loss": 0.6387,
+      "step": 4853
+    },
+    {
+      "epoch": 8.988888888888889,
+      "grad_norm": 1.8215141516909366,
+      "learning_rate": 1.8076398932981248e-07,
+      "loss": 0.6013,
+      "step": 4854
+    },
+    {
+      "epoch": 8.99074074074074,
+      "grad_norm": 1.6802766249658658,
+      "learning_rate": 1.807551466796952e-07,
+      "loss": 0.6298,
+      "step": 4855
+    },
+    {
+      "epoch": 8.992592592592592,
+      "grad_norm": 2.727235777158391,
+      "learning_rate": 1.8074630221398287e-07,
+      "loss": 0.6386,
+      "step": 4856
+    },
+    {
+      "epoch": 8.994444444444444,
+      "grad_norm": 1.702439565294306,
+      "learning_rate": 1.8073745593287438e-07,
+      "loss": 0.6381,
+      "step": 4857
+    },
+    {
+      "epoch": 8.996296296296297,
+      "grad_norm": 1.8383892174425474,
+      "learning_rate": 1.8072860783656858e-07,
+      "loss": 0.6769,
+      "step": 4858
+    },
+    {
+      "epoch": 8.998148148148148,
+      "grad_norm": 1.618941278882836,
+      "learning_rate": 1.8071975792526445e-07,
+      "loss": 0.6035,
+      "step": 4859
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 1.5401358476940168,
+      "learning_rate": 1.807109061991609e-07,
+      "loss": 0.6431,
+      "step": 4860
+    },
+    {
+      "epoch": 9.001851851851852,
+      "grad_norm": 1.670582670894463,
+      "learning_rate": 1.8070205265845703e-07,
+      "loss": 0.629,
+      "step": 4861
+    },
+    {
+      "epoch": 9.003703703703703,
+      "grad_norm": 1.5090899879571136,
+      "learning_rate": 1.8069319730335176e-07,
+      "loss": 0.5943,
+      "step": 4862
+    },
+    {
+      "epoch": 9.005555555555556,
+      "grad_norm": 1.6877111795180013,
+      "learning_rate": 1.806843401340443e-07,
+      "loss": 0.628,
+      "step": 4863
+    },
+    {
+      "epoch": 9.007407407407408,
+      "grad_norm": 1.5596527361672607,
+      "learning_rate": 1.8067548115073374e-07,
+      "loss": 0.6689,
+      "step": 4864
+    },
+    {
+      "epoch": 9.00925925925926,
+      "grad_norm": 2.143374943886906,
+      "learning_rate": 1.8066662035361925e-07,
+      "loss": 0.6323,
+      "step": 4865
+    },
+    {
+      "epoch": 9.011111111111111,
+      "grad_norm": 1.5731065339124362,
+      "learning_rate": 1.806577577429e-07,
+      "loss": 0.6204,
+      "step": 4866
+    },
+    {
+      "epoch": 9.012962962962963,
+      "grad_norm": 1.732070680461305,
+      "learning_rate": 1.8064889331877534e-07,
+      "loss": 0.6269,
+      "step": 4867
+    },
+    {
+      "epoch": 9.014814814814814,
+      "grad_norm": 1.577489965214457,
+      "learning_rate": 1.8064002708144452e-07,
+      "loss": 0.6388,
+      "step": 4868
+    },
+    {
+      "epoch": 9.016666666666667,
+      "grad_norm": 1.492399754931415,
+      "learning_rate": 1.8063115903110688e-07,
+      "loss": 0.6491,
+      "step": 4869
+    },
+    {
+      "epoch": 9.018518518518519,
+      "grad_norm": 1.3124732044189318,
+      "learning_rate": 1.8062228916796177e-07,
+      "loss": 0.6721,
+      "step": 4870
+    },
+    {
+      "epoch": 9.02037037037037,
+      "grad_norm": 1.7090493040093824,
+      "learning_rate": 1.8061341749220863e-07,
+      "loss": 0.6171,
+      "step": 4871
+    },
+    {
+      "epoch": 9.022222222222222,
+      "grad_norm": 1.8996995053563839,
+      "learning_rate": 1.8060454400404695e-07,
+      "loss": 0.6251,
+      "step": 4872
+    },
+    {
+      "epoch": 9.024074074074074,
+      "grad_norm": 1.4563608758878375,
+      "learning_rate": 1.8059566870367617e-07,
+      "loss": 0.6134,
+      "step": 4873
+    },
+    {
+      "epoch": 9.025925925925925,
+      "grad_norm": 1.5721457100028966,
+      "learning_rate": 1.805867915912959e-07,
+      "loss": 0.5932,
+      "step": 4874
+    },
+    {
+      "epoch": 9.027777777777779,
+      "grad_norm": 1.286492788291641,
+      "learning_rate": 1.8057791266710564e-07,
+      "loss": 0.6048,
+      "step": 4875
+    },
+    {
+      "epoch": 9.02962962962963,
+      "grad_norm": 1.8518693533491049,
+      "learning_rate": 1.805690319313051e-07,
+      "loss": 0.6124,
+      "step": 4876
+    },
+    {
+      "epoch": 9.031481481481482,
+      "grad_norm": 1.6898112425993053,
+      "learning_rate": 1.805601493840939e-07,
+      "loss": 0.6081,
+      "step": 4877
+    },
+    {
+      "epoch": 9.033333333333333,
+      "grad_norm": 1.5024791483925741,
+      "learning_rate": 1.805512650256717e-07,
+      "loss": 0.6136,
+      "step": 4878
+    },
+    {
+      "epoch": 9.035185185185185,
+      "grad_norm": 1.683033488553737,
+      "learning_rate": 1.8054237885623831e-07,
+      "loss": 0.6685,
+      "step": 4879
+    },
+    {
+      "epoch": 9.037037037037036,
+      "grad_norm": 1.7062090724179861,
+      "learning_rate": 1.805334908759935e-07,
+      "loss": 0.6563,
+      "step": 4880
+    },
+    {
+      "epoch": 9.03888888888889,
+      "grad_norm": 1.5285265723010917,
+      "learning_rate": 1.805246010851371e-07,
+      "loss": 0.6713,
+      "step": 4881
+    },
+    {
+      "epoch": 9.040740740740741,
+      "grad_norm": 1.5768796010668769,
+      "learning_rate": 1.8051570948386895e-07,
+      "loss": 0.6405,
+      "step": 4882
+    },
+    {
+      "epoch": 9.042592592592593,
+      "grad_norm": 1.6988096391799035,
+      "learning_rate": 1.8050681607238895e-07,
+      "loss": 0.6159,
+      "step": 4883
+    },
+    {
+      "epoch": 9.044444444444444,
+      "grad_norm": 1.5203602139755228,
+      "learning_rate": 1.8049792085089712e-07,
+      "loss": 0.6489,
+      "step": 4884
+    },
+    {
+      "epoch": 9.046296296296296,
+      "grad_norm": 1.6562165501659378,
+      "learning_rate": 1.804890238195934e-07,
+      "loss": 0.6007,
+      "step": 4885
+    },
+    {
+      "epoch": 9.048148148148147,
+      "grad_norm": 1.6500159216629178,
+      "learning_rate": 1.8048012497867776e-07,
+      "loss": 0.6326,
+      "step": 4886
+    },
+    {
+      "epoch": 9.05,
+      "grad_norm": 1.7431341341622981,
+      "learning_rate": 1.8047122432835038e-07,
+      "loss": 0.6403,
+      "step": 4887
+    },
+    {
+      "epoch": 9.051851851851852,
+      "grad_norm": 1.5421787446841935,
+      "learning_rate": 1.8046232186881132e-07,
+      "loss": 0.6467,
+      "step": 4888
+    },
+    {
+      "epoch": 9.053703703703704,
+      "grad_norm": 1.7387780891120124,
+      "learning_rate": 1.804534176002607e-07,
+      "loss": 0.6492,
+      "step": 4889
+    },
+    {
+      "epoch": 9.055555555555555,
+      "grad_norm": 1.421758201398186,
+      "learning_rate": 1.8044451152289877e-07,
+      "loss": 0.6306,
+      "step": 4890
+    },
+    {
+      "epoch": 9.057407407407407,
+      "grad_norm": 1.6624817408623478,
+      "learning_rate": 1.8043560363692574e-07,
+      "loss": 0.6303,
+      "step": 4891
+    },
+    {
+      "epoch": 9.059259259259258,
+      "grad_norm": 1.538891195740592,
+      "learning_rate": 1.804266939425419e-07,
+      "loss": 0.6219,
+      "step": 4892
+    },
+    {
+      "epoch": 9.061111111111112,
+      "grad_norm": 1.9771516523804664,
+      "learning_rate": 1.804177824399475e-07,
+      "loss": 0.6916,
+      "step": 4893
+    },
+    {
+      "epoch": 9.062962962962963,
+      "grad_norm": 1.6812721833699744,
+      "learning_rate": 1.8040886912934292e-07,
+      "loss": 0.6621,
+      "step": 4894
+    },
+    {
+      "epoch": 9.064814814814815,
+      "grad_norm": 1.431847059574879,
+      "learning_rate": 1.8039995401092862e-07,
+      "loss": 0.6211,
+      "step": 4895
+    },
+    {
+      "epoch": 9.066666666666666,
+      "grad_norm": 1.7768114597845408,
+      "learning_rate": 1.80391037084905e-07,
+      "loss": 0.6569,
+      "step": 4896
+    },
+    {
+      "epoch": 9.068518518518518,
+      "grad_norm": 1.6598328035460053,
+      "learning_rate": 1.803821183514725e-07,
+      "loss": 0.6667,
+      "step": 4897
+    },
+    {
+      "epoch": 9.07037037037037,
+      "grad_norm": 1.592453738061996,
+      "learning_rate": 1.8037319781083166e-07,
+      "loss": 0.6268,
+      "step": 4898
+    },
+    {
+      "epoch": 9.072222222222223,
+      "grad_norm": 1.7487959339946386,
+      "learning_rate": 1.8036427546318306e-07,
+      "loss": 0.6231,
+      "step": 4899
+    },
+    {
+      "epoch": 9.074074074074074,
+      "grad_norm": 1.6713807234040476,
+      "learning_rate": 1.803553513087273e-07,
+      "loss": 0.6492,
+      "step": 4900
+    },
+    {
+      "epoch": 9.075925925925926,
+      "grad_norm": 1.4033205965529207,
+      "learning_rate": 1.80346425347665e-07,
+      "loss": 0.659,
+      "step": 4901
+    },
+    {
+      "epoch": 9.077777777777778,
+      "grad_norm": 1.8068211291080067,
+      "learning_rate": 1.8033749758019678e-07,
+      "loss": 0.6495,
+      "step": 4902
+    },
+    {
+      "epoch": 9.079629629629629,
+      "grad_norm": 1.5516555526865257,
+      "learning_rate": 1.8032856800652344e-07,
+      "loss": 0.6396,
+      "step": 4903
+    },
+    {
+      "epoch": 9.081481481481482,
+      "grad_norm": 1.6807042891880808,
+      "learning_rate": 1.8031963662684577e-07,
+      "loss": 0.6317,
+      "step": 4904
+    },
+    {
+      "epoch": 9.083333333333334,
+      "grad_norm": 1.522403961214487,
+      "learning_rate": 1.803107034413645e-07,
+      "loss": 0.6484,
+      "step": 4905
+    },
+    {
+      "epoch": 9.085185185185185,
+      "grad_norm": 1.8348439508929184,
+      "learning_rate": 1.803017684502805e-07,
+      "loss": 0.7465,
+      "step": 4906
+    },
+    {
+      "epoch": 9.087037037037037,
+      "grad_norm": 1.8591497885796593,
+      "learning_rate": 1.8029283165379464e-07,
+      "loss": 0.6329,
+      "step": 4907
+    },
+    {
+      "epoch": 9.088888888888889,
+      "grad_norm": 1.6794104133695151,
+      "learning_rate": 1.8028389305210784e-07,
+      "loss": 0.623,
+      "step": 4908
+    },
+    {
+      "epoch": 9.09074074074074,
+      "grad_norm": 1.534546021159242,
+      "learning_rate": 1.802749526454211e-07,
+      "loss": 0.646,
+      "step": 4909
+    },
+    {
+      "epoch": 9.092592592592593,
+      "grad_norm": 2.2076579395792537,
+      "learning_rate": 1.802660104339354e-07,
+      "loss": 0.6616,
+      "step": 4910
+    },
+    {
+      "epoch": 9.094444444444445,
+      "grad_norm": 1.735066846237029,
+      "learning_rate": 1.8025706641785175e-07,
+      "loss": 0.6561,
+      "step": 4911
+    },
+    {
+      "epoch": 9.096296296296297,
+      "grad_norm": 1.7863350317501159,
+      "learning_rate": 1.802481205973713e-07,
+      "loss": 0.6028,
+      "step": 4912
+    },
+    {
+      "epoch": 9.098148148148148,
+      "grad_norm": 1.4377673643273938,
+      "learning_rate": 1.8023917297269517e-07,
+      "loss": 0.6386,
+      "step": 4913
+    },
+    {
+      "epoch": 9.1,
+      "grad_norm": 1.3457356985117823,
+      "learning_rate": 1.802302235440245e-07,
+      "loss": 0.6775,
+      "step": 4914
+    },
+    {
+      "epoch": 9.101851851851851,
+      "grad_norm": 2.089901384560452,
+      "learning_rate": 1.8022127231156047e-07,
+      "loss": 0.6159,
+      "step": 4915
+    },
+    {
+      "epoch": 9.103703703703705,
+      "grad_norm": 1.7794981059506234,
+      "learning_rate": 1.8021231927550437e-07,
+      "loss": 0.6239,
+      "step": 4916
+    },
+    {
+      "epoch": 9.105555555555556,
+      "grad_norm": 1.4997174730793863,
+      "learning_rate": 1.8020336443605747e-07,
+      "loss": 0.6546,
+      "step": 4917
+    },
+    {
+      "epoch": 9.107407407407408,
+      "grad_norm": 1.6246675824622723,
+      "learning_rate": 1.8019440779342113e-07,
+      "loss": 0.6521,
+      "step": 4918
+    },
+    {
+      "epoch": 9.10925925925926,
+      "grad_norm": 1.6028425939016664,
+      "learning_rate": 1.8018544934779671e-07,
+      "loss": 0.6563,
+      "step": 4919
+    },
+    {
+      "epoch": 9.11111111111111,
+      "grad_norm": 3.1309177845584384,
+      "learning_rate": 1.801764890993856e-07,
+      "loss": 0.6563,
+      "step": 4920
+    },
+    {
+      "epoch": 9.112962962962962,
+      "grad_norm": 1.7354720790564453,
+      "learning_rate": 1.8016752704838925e-07,
+      "loss": 0.6146,
+      "step": 4921
+    },
+    {
+      "epoch": 9.114814814814816,
+      "grad_norm": 1.5981173872314567,
+      "learning_rate": 1.8015856319500916e-07,
+      "loss": 0.6294,
+      "step": 4922
+    },
+    {
+      "epoch": 9.116666666666667,
+      "grad_norm": 1.6175541542904597,
+      "learning_rate": 1.8014959753944686e-07,
+      "loss": 0.6288,
+      "step": 4923
+    },
+    {
+      "epoch": 9.118518518518519,
+      "grad_norm": 1.466344210031735,
+      "learning_rate": 1.801406300819039e-07,
+      "loss": 0.5894,
+      "step": 4924
+    },
+    {
+      "epoch": 9.12037037037037,
+      "grad_norm": 2.490831660922919,
+      "learning_rate": 1.8013166082258196e-07,
+      "loss": 0.6833,
+      "step": 4925
+    },
+    {
+      "epoch": 9.122222222222222,
+      "grad_norm": 1.5062822291987519,
+      "learning_rate": 1.8012268976168264e-07,
+      "loss": 0.6362,
+      "step": 4926
+    },
+    {
+      "epoch": 9.124074074074073,
+      "grad_norm": 1.3442409120015848,
+      "learning_rate": 1.801137168994076e-07,
+      "loss": 0.5869,
+      "step": 4927
+    },
+    {
+      "epoch": 9.125925925925927,
+      "grad_norm": 1.5412123091510965,
+      "learning_rate": 1.8010474223595868e-07,
+      "loss": 0.6555,
+      "step": 4928
+    },
+    {
+      "epoch": 9.127777777777778,
+      "grad_norm": 1.3476337243530934,
+      "learning_rate": 1.8009576577153758e-07,
+      "loss": 0.6275,
+      "step": 4929
+    },
+    {
+      "epoch": 9.12962962962963,
+      "grad_norm": 1.840453833449504,
+      "learning_rate": 1.8008678750634614e-07,
+      "loss": 0.6202,
+      "step": 4930
+    },
+    {
+      "epoch": 9.131481481481481,
+      "grad_norm": 1.5198333023871702,
+      "learning_rate": 1.8007780744058617e-07,
+      "loss": 0.6419,
+      "step": 4931
+    },
+    {
+      "epoch": 9.133333333333333,
+      "grad_norm": 1.5909486322479607,
+      "learning_rate": 1.8006882557445962e-07,
+      "loss": 0.6524,
+      "step": 4932
+    },
+    {
+      "epoch": 9.135185185185184,
+      "grad_norm": 1.7759503810532487,
+      "learning_rate": 1.800598419081684e-07,
+      "loss": 0.6276,
+      "step": 4933
+    },
+    {
+      "epoch": 9.137037037037038,
+      "grad_norm": 1.6247030790685282,
+      "learning_rate": 1.800508564419145e-07,
+      "loss": 0.6128,
+      "step": 4934
+    },
+    {
+      "epoch": 9.13888888888889,
+      "grad_norm": 1.5383025035594717,
+      "learning_rate": 1.800418691758999e-07,
+      "loss": 0.6687,
+      "step": 4935
+    },
+    {
+      "epoch": 9.14074074074074,
+      "grad_norm": 1.8208035043933768,
+      "learning_rate": 1.8003288011032675e-07,
+      "loss": 0.645,
+      "step": 4936
+    },
+    {
+      "epoch": 9.142592592592592,
+      "grad_norm": 1.7065749632157245,
+      "learning_rate": 1.8002388924539703e-07,
+      "loss": 0.6567,
+      "step": 4937
+    },
+    {
+      "epoch": 9.144444444444444,
+      "grad_norm": 1.357175243788134,
+      "learning_rate": 1.80014896581313e-07,
+      "loss": 0.5964,
+      "step": 4938
+    },
+    {
+      "epoch": 9.146296296296295,
+      "grad_norm": 1.4591828467543584,
+      "learning_rate": 1.8000590211827674e-07,
+      "loss": 0.6301,
+      "step": 4939
+    },
+    {
+      "epoch": 9.148148148148149,
+      "grad_norm": 1.4515861833320292,
+      "learning_rate": 1.799969058564905e-07,
+      "loss": 0.6093,
+      "step": 4940
+    },
+    {
+      "epoch": 9.15,
+      "grad_norm": 1.401304633813347,
+      "learning_rate": 1.799879077961566e-07,
+      "loss": 0.5998,
+      "step": 4941
+    },
+    {
+      "epoch": 9.151851851851852,
+      "grad_norm": 1.631327827543034,
+      "learning_rate": 1.799789079374772e-07,
+      "loss": 0.646,
+      "step": 4942
+    },
+    {
+      "epoch": 9.153703703703703,
+      "grad_norm": 1.3837585331254347,
+      "learning_rate": 1.7996990628065478e-07,
+      "loss": 0.6236,
+      "step": 4943
+    },
+    {
+      "epoch": 9.155555555555555,
+      "grad_norm": 1.4647990563982192,
+      "learning_rate": 1.7996090282589166e-07,
+      "loss": 0.6437,
+      "step": 4944
+    },
+    {
+      "epoch": 9.157407407407407,
+      "grad_norm": 1.8412269214564083,
+      "learning_rate": 1.7995189757339032e-07,
+      "loss": 0.6281,
+      "step": 4945
+    },
+    {
+      "epoch": 9.15925925925926,
+      "grad_norm": 1.4354687122024636,
+      "learning_rate": 1.799428905233531e-07,
+      "loss": 0.6727,
+      "step": 4946
+    },
+    {
+      "epoch": 9.161111111111111,
+      "grad_norm": 2.645241182905206,
+      "learning_rate": 1.7993388167598266e-07,
+      "loss": 0.6025,
+      "step": 4947
+    },
+    {
+      "epoch": 9.162962962962963,
+      "grad_norm": 1.5911301601033796,
+      "learning_rate": 1.799248710314814e-07,
+      "loss": 0.6279,
+      "step": 4948
+    },
+    {
+      "epoch": 9.164814814814815,
+      "grad_norm": 1.4966637795374393,
+      "learning_rate": 1.7991585859005201e-07,
+      "loss": 0.6311,
+      "step": 4949
+    },
+    {
+      "epoch": 9.166666666666666,
+      "grad_norm": 1.5328918698937555,
+      "learning_rate": 1.7990684435189705e-07,
+      "loss": 0.6354,
+      "step": 4950
+    },
+    {
+      "epoch": 9.168518518518518,
+      "grad_norm": 1.5665955349191336,
+      "learning_rate": 1.798978283172192e-07,
+      "loss": 0.6376,
+      "step": 4951
+    },
+    {
+      "epoch": 9.170370370370371,
+      "grad_norm": 1.5761449533059166,
+      "learning_rate": 1.798888104862212e-07,
+      "loss": 0.6243,
+      "step": 4952
+    },
+    {
+      "epoch": 9.172222222222222,
+      "grad_norm": 1.6093057563831608,
+      "learning_rate": 1.798797908591057e-07,
+      "loss": 0.6309,
+      "step": 4953
+    },
+    {
+      "epoch": 9.174074074074074,
+      "grad_norm": 1.668206834831381,
+      "learning_rate": 1.798707694360756e-07,
+      "loss": 0.6858,
+      "step": 4954
+    },
+    {
+      "epoch": 9.175925925925926,
+      "grad_norm": 1.6301226639602802,
+      "learning_rate": 1.798617462173337e-07,
+      "loss": 0.632,
+      "step": 4955
+    },
+    {
+      "epoch": 9.177777777777777,
+      "grad_norm": 1.35186316364307,
+      "learning_rate": 1.798527212030828e-07,
+      "loss": 0.6525,
+      "step": 4956
+    },
+    {
+      "epoch": 9.17962962962963,
+      "grad_norm": 1.8015821596486283,
+      "learning_rate": 1.7984369439352585e-07,
+      "loss": 0.6318,
+      "step": 4957
+    },
+    {
+      "epoch": 9.181481481481482,
+      "grad_norm": 1.6094478716523994,
+      "learning_rate": 1.7983466578886582e-07,
+      "loss": 0.6479,
+      "step": 4958
+    },
+    {
+      "epoch": 9.183333333333334,
+      "grad_norm": 1.5646830130068923,
+      "learning_rate": 1.798256353893057e-07,
+      "loss": 0.6496,
+      "step": 4959
+    },
+    {
+      "epoch": 9.185185185185185,
+      "grad_norm": 1.5096535566202443,
+      "learning_rate": 1.7981660319504844e-07,
+      "loss": 0.63,
+      "step": 4960
+    },
+    {
+      "epoch": 9.187037037037037,
+      "grad_norm": 1.484454742990276,
+      "learning_rate": 1.7980756920629717e-07,
+      "loss": 0.6213,
+      "step": 4961
+    },
+    {
+      "epoch": 9.188888888888888,
+      "grad_norm": 1.9892772178648397,
+      "learning_rate": 1.79798533423255e-07,
+      "loss": 0.6571,
+      "step": 4962
+    },
+    {
+      "epoch": 9.190740740740742,
+      "grad_norm": 1.7467686499634199,
+      "learning_rate": 1.797894958461251e-07,
+      "loss": 0.6355,
+      "step": 4963
+    },
+    {
+      "epoch": 9.192592592592593,
+      "grad_norm": 1.7687282721631068,
+      "learning_rate": 1.797804564751106e-07,
+      "loss": 0.635,
+      "step": 4964
+    },
+    {
+      "epoch": 9.194444444444445,
+      "grad_norm": 1.4851070830982465,
+      "learning_rate": 1.7977141531041476e-07,
+      "loss": 0.6618,
+      "step": 4965
+    },
+    {
+      "epoch": 9.196296296296296,
+      "grad_norm": 1.615517358584911,
+      "learning_rate": 1.7976237235224084e-07,
+      "loss": 0.6363,
+      "step": 4966
+    },
+    {
+      "epoch": 9.198148148148148,
+      "grad_norm": 2.4971452816804574,
+      "learning_rate": 1.7975332760079216e-07,
+      "loss": 0.6355,
+      "step": 4967
+    },
+    {
+      "epoch": 9.2,
+      "grad_norm": 1.4223137090744706,
+      "learning_rate": 1.7974428105627206e-07,
+      "loss": 0.6648,
+      "step": 4968
+    },
+    {
+      "epoch": 9.201851851851853,
+      "grad_norm": 1.8314652308840857,
+      "learning_rate": 1.7973523271888397e-07,
+      "loss": 0.6548,
+      "step": 4969
+    },
+    {
+      "epoch": 9.203703703703704,
+      "grad_norm": 1.4900631781858213,
+      "learning_rate": 1.797261825888313e-07,
+      "loss": 0.6516,
+      "step": 4970
+    },
+    {
+      "epoch": 9.205555555555556,
+      "grad_norm": 1.699847563799028,
+      "learning_rate": 1.7971713066631745e-07,
+      "loss": 0.6544,
+      "step": 4971
+    },
+    {
+      "epoch": 9.207407407407407,
+      "grad_norm": 1.4835502298931005,
+      "learning_rate": 1.7970807695154603e-07,
+      "loss": 0.5828,
+      "step": 4972
+    },
+    {
+      "epoch": 9.209259259259259,
+      "grad_norm": 1.5407325633124669,
+      "learning_rate": 1.796990214447206e-07,
+      "loss": 0.6799,
+      "step": 4973
+    },
+    {
+      "epoch": 9.21111111111111,
+      "grad_norm": 1.4335740288818684,
+      "learning_rate": 1.7968996414604462e-07,
+      "loss": 0.6664,
+      "step": 4974
+    },
+    {
+      "epoch": 9.212962962962964,
+      "grad_norm": 2.8202926237552584,
+      "learning_rate": 1.7968090505572188e-07,
+      "loss": 0.631,
+      "step": 4975
+    },
+    {
+      "epoch": 9.214814814814815,
+      "grad_norm": 2.2964759395350054,
+      "learning_rate": 1.7967184417395596e-07,
+      "loss": 0.6752,
+      "step": 4976
+    },
+    {
+      "epoch": 9.216666666666667,
+      "grad_norm": 1.4400858026515644,
+      "learning_rate": 1.796627815009506e-07,
+      "loss": 0.6465,
+      "step": 4977
+    },
+    {
+      "epoch": 9.218518518518518,
+      "grad_norm": 1.4722258708980698,
+      "learning_rate": 1.7965371703690955e-07,
+      "loss": 0.6335,
+      "step": 4978
+    },
+    {
+      "epoch": 9.22037037037037,
+      "grad_norm": 1.657948224794598,
+      "learning_rate": 1.796446507820366e-07,
+      "loss": 0.6369,
+      "step": 4979
+    },
+    {
+      "epoch": 9.222222222222221,
+      "grad_norm": 1.3757415842306018,
+      "learning_rate": 1.796355827365356e-07,
+      "loss": 0.6434,
+      "step": 4980
+    },
+    {
+      "epoch": 9.224074074074075,
+      "grad_norm": 1.3779079858147176,
+      "learning_rate": 1.796265129006104e-07,
+      "loss": 0.6718,
+      "step": 4981
+    },
+    {
+      "epoch": 9.225925925925926,
+      "grad_norm": 1.67085509614113,
+      "learning_rate": 1.7961744127446493e-07,
+      "loss": 0.6541,
+      "step": 4982
+    },
+    {
+      "epoch": 9.227777777777778,
+      "grad_norm": 1.9355580127199132,
+      "learning_rate": 1.7960836785830315e-07,
+      "loss": 0.6151,
+      "step": 4983
+    },
+    {
+      "epoch": 9.22962962962963,
+      "grad_norm": 1.4446134482573796,
+      "learning_rate": 1.7959929265232905e-07,
+      "loss": 0.597,
+      "step": 4984
+    },
+    {
+      "epoch": 9.231481481481481,
+      "grad_norm": 2.0160220046845727,
+      "learning_rate": 1.7959021565674665e-07,
+      "loss": 0.66,
+      "step": 4985
+    },
+    {
+      "epoch": 9.233333333333333,
+      "grad_norm": 1.715321013057844,
+      "learning_rate": 1.7958113687176005e-07,
+      "loss": 0.6764,
+      "step": 4986
+    },
+    {
+      "epoch": 9.235185185185186,
+      "grad_norm": 1.8339720848478032,
+      "learning_rate": 1.7957205629757335e-07,
+      "loss": 0.6582,
+      "step": 4987
+    },
+    {
+      "epoch": 9.237037037037037,
+      "grad_norm": 1.5356065454225052,
+      "learning_rate": 1.795629739343907e-07,
+      "loss": 0.5962,
+      "step": 4988
+    },
+    {
+      "epoch": 9.238888888888889,
+      "grad_norm": 1.639472161260969,
+      "learning_rate": 1.795538897824163e-07,
+      "loss": 0.643,
+      "step": 4989
+    },
+    {
+      "epoch": 9.24074074074074,
+      "grad_norm": 1.4742147555421299,
+      "learning_rate": 1.7954480384185444e-07,
+      "loss": 0.6185,
+      "step": 4990
+    },
+    {
+      "epoch": 9.242592592592592,
+      "grad_norm": 1.6465143612137927,
+      "learning_rate": 1.795357161129093e-07,
+      "loss": 0.6757,
+      "step": 4991
+    },
+    {
+      "epoch": 9.244444444444444,
+      "grad_norm": 1.7632031636729146,
+      "learning_rate": 1.7952662659578528e-07,
+      "loss": 0.6389,
+      "step": 4992
+    },
+    {
+      "epoch": 9.246296296296297,
+      "grad_norm": 1.7586768861538635,
+      "learning_rate": 1.7951753529068666e-07,
+      "loss": 0.6253,
+      "step": 4993
+    },
+    {
+      "epoch": 9.248148148148148,
+      "grad_norm": 1.550385184948432,
+      "learning_rate": 1.7950844219781794e-07,
+      "loss": 0.6437,
+      "step": 4994
+    },
+    {
+      "epoch": 9.25,
+      "grad_norm": 1.5605681744560167,
+      "learning_rate": 1.7949934731738346e-07,
+      "loss": 0.6486,
+      "step": 4995
+    },
+    {
+      "epoch": 9.251851851851852,
+      "grad_norm": 1.7945936727718752,
+      "learning_rate": 1.7949025064958775e-07,
+      "loss": 0.6621,
+      "step": 4996
+    },
+    {
+      "epoch": 9.253703703703703,
+      "grad_norm": 1.684355890701175,
+      "learning_rate": 1.794811521946353e-07,
+      "loss": 0.6353,
+      "step": 4997
+    },
+    {
+      "epoch": 9.255555555555556,
+      "grad_norm": 1.6486710606711317,
+      "learning_rate": 1.7947205195273068e-07,
+      "loss": 0.6291,
+      "step": 4998
+    },
+    {
+      "epoch": 9.257407407407408,
+      "grad_norm": 4.526348700046569,
+      "learning_rate": 1.7946294992407848e-07,
+      "loss": 0.6539,
+      "step": 4999
+    },
+    {
+      "epoch": 9.25925925925926,
+      "grad_norm": 1.5726024947315718,
+      "learning_rate": 1.794538461088834e-07,
+      "loss": 0.6852,
+      "step": 5000
+    },
+    {
+      "epoch": 9.261111111111111,
+      "grad_norm": 1.4670976646891696,
+      "learning_rate": 1.7944474050734998e-07,
+      "loss": 0.6802,
+      "step": 5001
+    },
+    {
+      "epoch": 9.262962962962963,
+      "grad_norm": 1.4798819637549998,
+      "learning_rate": 1.794356331196831e-07,
+      "loss": 0.6725,
+      "step": 5002
+    },
+    {
+      "epoch": 9.264814814814814,
+      "grad_norm": 1.4375463696445736,
+      "learning_rate": 1.794265239460874e-07,
+      "loss": 0.6283,
+      "step": 5003
+    },
+    {
+      "epoch": 9.266666666666667,
+      "grad_norm": 1.7066915421212974,
+      "learning_rate": 1.7941741298676774e-07,
+      "loss": 0.5981,
+      "step": 5004
+    },
+    {
+      "epoch": 9.268518518518519,
+      "grad_norm": 1.5268251717681938,
+      "learning_rate": 1.7940830024192896e-07,
+      "loss": 0.6221,
+      "step": 5005
+    },
+    {
+      "epoch": 9.27037037037037,
+      "grad_norm": 1.3769626220092266,
+      "learning_rate": 1.793991857117759e-07,
+      "loss": 0.6737,
+      "step": 5006
+    },
+    {
+      "epoch": 9.272222222222222,
+      "grad_norm": 2.23314283440295,
+      "learning_rate": 1.793900693965135e-07,
+      "loss": 0.6003,
+      "step": 5007
+    },
+    {
+      "epoch": 9.274074074074074,
+      "grad_norm": 1.4265979488281466,
+      "learning_rate": 1.7938095129634672e-07,
+      "loss": 0.6762,
+      "step": 5008
+    },
+    {
+      "epoch": 9.275925925925925,
+      "grad_norm": 1.8198743096616483,
+      "learning_rate": 1.7937183141148055e-07,
+      "loss": 0.6462,
+      "step": 5009
+    },
+    {
+      "epoch": 9.277777777777779,
+      "grad_norm": 1.5188661639899825,
+      "learning_rate": 1.7936270974212007e-07,
+      "loss": 0.668,
+      "step": 5010
+    },
+    {
+      "epoch": 9.27962962962963,
+      "grad_norm": 1.5833737296242645,
+      "learning_rate": 1.793535862884703e-07,
+      "loss": 0.622,
+      "step": 5011
+    },
+    {
+      "epoch": 9.281481481481482,
+      "grad_norm": 1.5193944196457194,
+      "learning_rate": 1.793444610507364e-07,
+      "loss": 0.631,
+      "step": 5012
+    },
+    {
+      "epoch": 9.283333333333333,
+      "grad_norm": 1.672919488798573,
+      "learning_rate": 1.793353340291235e-07,
+      "loss": 0.6653,
+      "step": 5013
+    },
+    {
+      "epoch": 9.285185185185185,
+      "grad_norm": 1.7101912463159068,
+      "learning_rate": 1.7932620522383685e-07,
+      "loss": 0.6326,
+      "step": 5014
+    },
+    {
+      "epoch": 9.287037037037036,
+      "grad_norm": 1.533147391794233,
+      "learning_rate": 1.7931707463508166e-07,
+      "loss": 0.6221,
+      "step": 5015
+    },
+    {
+      "epoch": 9.28888888888889,
+      "grad_norm": 1.455199130564899,
+      "learning_rate": 1.793079422630632e-07,
+      "loss": 0.6618,
+      "step": 5016
+    },
+    {
+      "epoch": 9.290740740740741,
+      "grad_norm": 1.635680136387207,
+      "learning_rate": 1.792988081079868e-07,
+      "loss": 0.5942,
+      "step": 5017
+    },
+    {
+      "epoch": 9.292592592592593,
+      "grad_norm": 1.455392630163267,
+      "learning_rate": 1.7928967217005779e-07,
+      "loss": 0.6452,
+      "step": 5018
+    },
+    {
+      "epoch": 9.294444444444444,
+      "grad_norm": 1.584390106076646,
+      "learning_rate": 1.7928053444948167e-07,
+      "loss": 0.6431,
+      "step": 5019
+    },
+    {
+      "epoch": 9.296296296296296,
+      "grad_norm": 1.4646706362416826,
+      "learning_rate": 1.7927139494646375e-07,
+      "loss": 0.6261,
+      "step": 5020
+    },
+    {
+      "epoch": 9.298148148148147,
+      "grad_norm": 1.6056776596207563,
+      "learning_rate": 1.792622536612096e-07,
+      "loss": 0.6545,
+      "step": 5021
+    },
+    {
+      "epoch": 9.3,
+      "grad_norm": 1.6517396898267862,
+      "learning_rate": 1.792531105939247e-07,
+      "loss": 0.6213,
+      "step": 5022
+    },
+    {
+      "epoch": 9.301851851851852,
+      "grad_norm": 1.624852568444863,
+      "learning_rate": 1.7924396574481464e-07,
+      "loss": 0.6624,
+      "step": 5023
+    },
+    {
+      "epoch": 9.303703703703704,
+      "grad_norm": 1.6881408896966401,
+      "learning_rate": 1.7923481911408502e-07,
+      "loss": 0.6024,
+      "step": 5024
+    },
+    {
+      "epoch": 9.305555555555555,
+      "grad_norm": 1.545826666474559,
+      "learning_rate": 1.7922567070194146e-07,
+      "loss": 0.6391,
+      "step": 5025
+    },
+    {
+      "epoch": 9.307407407407407,
+      "grad_norm": 1.8592296379740392,
+      "learning_rate": 1.7921652050858965e-07,
+      "loss": 0.6494,
+      "step": 5026
+    },
+    {
+      "epoch": 9.309259259259258,
+      "grad_norm": 1.5960725446946296,
+      "learning_rate": 1.7920736853423528e-07,
+      "loss": 0.6654,
+      "step": 5027
+    },
+    {
+      "epoch": 9.311111111111112,
+      "grad_norm": 1.4753447139446967,
+      "learning_rate": 1.7919821477908417e-07,
+      "loss": 0.6531,
+      "step": 5028
+    },
+    {
+      "epoch": 9.312962962962963,
+      "grad_norm": 1.6716827366013367,
+      "learning_rate": 1.791890592433421e-07,
+      "loss": 0.6645,
+      "step": 5029
+    },
+    {
+      "epoch": 9.314814814814815,
+      "grad_norm": 1.984200267143179,
+      "learning_rate": 1.791799019272149e-07,
+      "loss": 0.5852,
+      "step": 5030
+    },
+    {
+      "epoch": 9.316666666666666,
+      "grad_norm": 1.5056632470097804,
+      "learning_rate": 1.7917074283090843e-07,
+      "loss": 0.6549,
+      "step": 5031
+    },
+    {
+      "epoch": 9.318518518518518,
+      "grad_norm": 1.4430046489844697,
+      "learning_rate": 1.7916158195462865e-07,
+      "loss": 0.6323,
+      "step": 5032
+    },
+    {
+      "epoch": 9.32037037037037,
+      "grad_norm": 1.7069982872656422,
+      "learning_rate": 1.7915241929858152e-07,
+      "loss": 0.593,
+      "step": 5033
+    },
+    {
+      "epoch": 9.322222222222223,
+      "grad_norm": 1.5312560951397207,
+      "learning_rate": 1.79143254862973e-07,
+      "loss": 0.6298,
+      "step": 5034
+    },
+    {
+      "epoch": 9.324074074074074,
+      "grad_norm": 1.414994024703364,
+      "learning_rate": 1.7913408864800919e-07,
+      "loss": 0.6053,
+      "step": 5035
+    },
+    {
+      "epoch": 9.325925925925926,
+      "grad_norm": 1.5320441508491534,
+      "learning_rate": 1.7912492065389613e-07,
+      "loss": 0.6364,
+      "step": 5036
+    },
+    {
+      "epoch": 9.327777777777778,
+      "grad_norm": 1.6043268445240035,
+      "learning_rate": 1.7911575088083993e-07,
+      "loss": 0.6392,
+      "step": 5037
+    },
+    {
+      "epoch": 9.329629629629629,
+      "grad_norm": 1.6424564574926412,
+      "learning_rate": 1.7910657932904683e-07,
+      "loss": 0.6371,
+      "step": 5038
+    },
+    {
+      "epoch": 9.331481481481482,
+      "grad_norm": 1.6082424307416978,
+      "learning_rate": 1.7909740599872296e-07,
+      "loss": 0.595,
+      "step": 5039
+    },
+    {
+      "epoch": 9.333333333333334,
+      "grad_norm": 1.6253009453461835,
+      "learning_rate": 1.7908823089007456e-07,
+      "loss": 0.634,
+      "step": 5040
+    },
+    {
+      "epoch": 9.335185185185185,
+      "grad_norm": 1.6991740815851821,
+      "learning_rate": 1.7907905400330795e-07,
+      "loss": 0.6424,
+      "step": 5041
+    },
+    {
+      "epoch": 9.337037037037037,
+      "grad_norm": 1.5019910570051236,
+      "learning_rate": 1.790698753386294e-07,
+      "loss": 0.6376,
+      "step": 5042
+    },
+    {
+      "epoch": 9.338888888888889,
+      "grad_norm": 1.56397856475829,
+      "learning_rate": 1.7906069489624532e-07,
+      "loss": 0.6495,
+      "step": 5043
+    },
+    {
+      "epoch": 9.34074074074074,
+      "grad_norm": 2.7932846153355073,
+      "learning_rate": 1.790515126763621e-07,
+      "loss": 0.6404,
+      "step": 5044
+    },
+    {
+      "epoch": 9.342592592592593,
+      "grad_norm": 1.583012494386562,
+      "learning_rate": 1.7904232867918614e-07,
+      "loss": 0.5802,
+      "step": 5045
+    },
+    {
+      "epoch": 9.344444444444445,
+      "grad_norm": 1.333726286358759,
+      "learning_rate": 1.79033142904924e-07,
+      "loss": 0.6004,
+      "step": 5046
+    },
+    {
+      "epoch": 9.346296296296297,
+      "grad_norm": 1.459715078531212,
+      "learning_rate": 1.7902395535378213e-07,
+      "loss": 0.6563,
+      "step": 5047
+    },
+    {
+      "epoch": 9.348148148148148,
+      "grad_norm": 1.615068153046038,
+      "learning_rate": 1.7901476602596714e-07,
+      "loss": 0.6644,
+      "step": 5048
+    },
+    {
+      "epoch": 9.35,
+      "grad_norm": 1.6605966537204446,
+      "learning_rate": 1.7900557492168559e-07,
+      "loss": 0.6561,
+      "step": 5049
+    },
+    {
+      "epoch": 9.351851851851851,
+      "grad_norm": 1.4285911454793239,
+      "learning_rate": 1.7899638204114413e-07,
+      "loss": 0.6379,
+      "step": 5050
+    },
+    {
+      "epoch": 9.353703703703705,
+      "grad_norm": 1.8774116831973644,
+      "learning_rate": 1.7898718738454947e-07,
+      "loss": 0.6315,
+      "step": 5051
+    },
+    {
+      "epoch": 9.355555555555556,
+      "grad_norm": 1.57376946696217,
+      "learning_rate": 1.789779909521083e-07,
+      "loss": 0.6354,
+      "step": 5052
+    },
+    {
+      "epoch": 9.357407407407408,
+      "grad_norm": 1.6494245917492636,
+      "learning_rate": 1.7896879274402742e-07,
+      "loss": 0.6921,
+      "step": 5053
+    },
+    {
+      "epoch": 9.35925925925926,
+      "grad_norm": 2.6664923061154884,
+      "learning_rate": 1.7895959276051355e-07,
+      "loss": 0.6535,
+      "step": 5054
+    },
+    {
+      "epoch": 9.36111111111111,
+      "grad_norm": 1.773639458781967,
+      "learning_rate": 1.7895039100177365e-07,
+      "loss": 0.6038,
+      "step": 5055
+    },
+    {
+      "epoch": 9.362962962962962,
+      "grad_norm": 1.736902489949008,
+      "learning_rate": 1.789411874680145e-07,
+      "loss": 0.6338,
+      "step": 5056
+    },
+    {
+      "epoch": 9.364814814814816,
+      "grad_norm": 1.6014590616897588,
+      "learning_rate": 1.7893198215944304e-07,
+      "loss": 0.6465,
+      "step": 5057
+    },
+    {
+      "epoch": 9.366666666666667,
+      "grad_norm": 1.5141258118210914,
+      "learning_rate": 1.7892277507626626e-07,
+      "loss": 0.6235,
+      "step": 5058
+    },
+    {
+      "epoch": 9.368518518518519,
+      "grad_norm": 1.76410411113449,
+      "learning_rate": 1.7891356621869115e-07,
+      "loss": 0.6613,
+      "step": 5059
+    },
+    {
+      "epoch": 9.37037037037037,
+      "grad_norm": 1.7630720752523528,
+      "learning_rate": 1.7890435558692474e-07,
+      "loss": 0.6303,
+      "step": 5060
+    },
+    {
+      "epoch": 9.372222222222222,
+      "grad_norm": 3.0677187889485094,
+      "learning_rate": 1.788951431811741e-07,
+      "loss": 0.629,
+      "step": 5061
+    },
+    {
+      "epoch": 9.374074074074073,
+      "grad_norm": 1.3081737218519582,
+      "learning_rate": 1.788859290016464e-07,
+      "loss": 0.6191,
+      "step": 5062
+    },
+    {
+      "epoch": 9.375925925925927,
+      "grad_norm": 1.3643717661850152,
+      "learning_rate": 1.7887671304854874e-07,
+      "loss": 0.6237,
+      "step": 5063
+    },
+    {
+      "epoch": 9.377777777777778,
+      "grad_norm": 2.0807765331599275,
+      "learning_rate": 1.7886749532208836e-07,
+      "loss": 0.68,
+      "step": 5064
+    },
+    {
+      "epoch": 9.37962962962963,
+      "grad_norm": 1.4550154309273073,
+      "learning_rate": 1.7885827582247245e-07,
+      "loss": 0.6136,
+      "step": 5065
+    },
+    {
+      "epoch": 9.381481481481481,
+      "grad_norm": 1.7066524589216818,
+      "learning_rate": 1.7884905454990837e-07,
+      "loss": 0.6129,
+      "step": 5066
+    },
+    {
+      "epoch": 9.383333333333333,
+      "grad_norm": 1.4183323629222169,
+      "learning_rate": 1.7883983150460336e-07,
+      "loss": 0.6747,
+      "step": 5067
+    },
+    {
+      "epoch": 9.385185185185184,
+      "grad_norm": 1.7227217516296243,
+      "learning_rate": 1.7883060668676483e-07,
+      "loss": 0.6271,
+      "step": 5068
+    },
+    {
+      "epoch": 9.387037037037038,
+      "grad_norm": 1.520639856957348,
+      "learning_rate": 1.7882138009660018e-07,
+      "loss": 0.5745,
+      "step": 5069
+    },
+    {
+      "epoch": 9.38888888888889,
+      "grad_norm": 1.5521383855107516,
+      "learning_rate": 1.7881215173431678e-07,
+      "loss": 0.6217,
+      "step": 5070
+    },
+    {
+      "epoch": 9.39074074074074,
+      "grad_norm": 1.364430124886182,
+      "learning_rate": 1.788029216001222e-07,
+      "loss": 0.6758,
+      "step": 5071
+    },
+    {
+      "epoch": 9.392592592592592,
+      "grad_norm": 1.824452990571026,
+      "learning_rate": 1.787936896942239e-07,
+      "loss": 0.58,
+      "step": 5072
+    },
+    {
+      "epoch": 9.394444444444444,
+      "grad_norm": 1.4294758168433916,
+      "learning_rate": 1.7878445601682947e-07,
+      "loss": 0.6548,
+      "step": 5073
+    },
+    {
+      "epoch": 9.396296296296295,
+      "grad_norm": 1.6753701967047336,
+      "learning_rate": 1.7877522056814644e-07,
+      "loss": 0.6287,
+      "step": 5074
+    },
+    {
+      "epoch": 9.398148148148149,
+      "grad_norm": 1.3616281473806655,
+      "learning_rate": 1.7876598334838255e-07,
+      "loss": 0.643,
+      "step": 5075
+    },
+    {
+      "epoch": 9.4,
+      "grad_norm": 1.6751669454515343,
+      "learning_rate": 1.7875674435774544e-07,
+      "loss": 0.6656,
+      "step": 5076
+    },
+    {
+      "epoch": 9.401851851851852,
+      "grad_norm": 1.6106451220892728,
+      "learning_rate": 1.787475035964428e-07,
+      "loss": 0.5968,
+      "step": 5077
+    },
+    {
+      "epoch": 9.403703703703703,
+      "grad_norm": 1.4867019568871895,
+      "learning_rate": 1.787382610646824e-07,
+      "loss": 0.6628,
+      "step": 5078
+    },
+    {
+      "epoch": 9.405555555555555,
+      "grad_norm": 1.9228905910102003,
+      "learning_rate": 1.7872901676267206e-07,
+      "loss": 0.6343,
+      "step": 5079
+    },
+    {
+      "epoch": 9.407407407407407,
+      "grad_norm": 1.6840921115711178,
+      "learning_rate": 1.7871977069061957e-07,
+      "loss": 0.6106,
+      "step": 5080
+    },
+    {
+      "epoch": 9.40925925925926,
+      "grad_norm": 1.7853776228622724,
+      "learning_rate": 1.7871052284873284e-07,
+      "loss": 0.6283,
+      "step": 5081
+    },
+    {
+      "epoch": 9.411111111111111,
+      "grad_norm": 2.6738343791571584,
+      "learning_rate": 1.7870127323721977e-07,
+      "loss": 0.6084,
+      "step": 5082
+    },
+    {
+      "epoch": 9.412962962962963,
+      "grad_norm": 1.5013127229298282,
+      "learning_rate": 1.7869202185628836e-07,
+      "loss": 0.5787,
+      "step": 5083
+    },
+    {
+      "epoch": 9.414814814814815,
+      "grad_norm": 1.523815238755022,
+      "learning_rate": 1.7868276870614656e-07,
+      "loss": 0.601,
+      "step": 5084
+    },
+    {
+      "epoch": 9.416666666666666,
+      "grad_norm": 1.8533380354214934,
+      "learning_rate": 1.786735137870024e-07,
+      "loss": 0.6182,
+      "step": 5085
+    },
+    {
+      "epoch": 9.418518518518518,
+      "grad_norm": 1.580346653812946,
+      "learning_rate": 1.7866425709906402e-07,
+      "loss": 0.6084,
+      "step": 5086
+    },
+    {
+      "epoch": 9.420370370370371,
+      "grad_norm": 1.6970851324684704,
+      "learning_rate": 1.7865499864253945e-07,
+      "loss": 0.5958,
+      "step": 5087
+    },
+    {
+      "epoch": 9.422222222222222,
+      "grad_norm": 1.719140709402049,
+      "learning_rate": 1.7864573841763693e-07,
+      "loss": 0.6361,
+      "step": 5088
+    },
+    {
+      "epoch": 9.424074074074074,
+      "grad_norm": 2.84831840004522,
+      "learning_rate": 1.7863647642456456e-07,
+      "loss": 0.654,
+      "step": 5089
+    },
+    {
+      "epoch": 9.425925925925926,
+      "grad_norm": 7.152127254163449,
+      "learning_rate": 1.7862721266353067e-07,
+      "loss": 0.6044,
+      "step": 5090
+    },
+    {
+      "epoch": 9.427777777777777,
+      "grad_norm": 1.4784232893295477,
+      "learning_rate": 1.7861794713474347e-07,
+      "loss": 0.6742,
+      "step": 5091
+    },
+    {
+      "epoch": 9.42962962962963,
+      "grad_norm": 1.509768925031899,
+      "learning_rate": 1.786086798384113e-07,
+      "loss": 0.618,
+      "step": 5092
+    },
+    {
+      "epoch": 9.431481481481482,
+      "grad_norm": 1.3703300928592905,
+      "learning_rate": 1.7859941077474253e-07,
+      "loss": 0.6689,
+      "step": 5093
+    },
+    {
+      "epoch": 9.433333333333334,
+      "grad_norm": 1.6585919113697565,
+      "learning_rate": 1.785901399439455e-07,
+      "loss": 0.6496,
+      "step": 5094
+    },
+    {
+      "epoch": 9.435185185185185,
+      "grad_norm": 1.4089886894937549,
+      "learning_rate": 1.7858086734622868e-07,
+      "loss": 0.6348,
+      "step": 5095
+    },
+    {
+      "epoch": 9.437037037037037,
+      "grad_norm": 1.5095432816949876,
+      "learning_rate": 1.7857159298180054e-07,
+      "loss": 0.6612,
+      "step": 5096
+    },
+    {
+      "epoch": 9.438888888888888,
+      "grad_norm": 1.7114269335120136,
+      "learning_rate": 1.785623168508696e-07,
+      "loss": 0.6198,
+      "step": 5097
+    },
+    {
+      "epoch": 9.440740740740742,
+      "grad_norm": 1.7210746541647095,
+      "learning_rate": 1.785530389536444e-07,
+      "loss": 0.6665,
+      "step": 5098
+    },
+    {
+      "epoch": 9.442592592592593,
+      "grad_norm": 1.7576859124920148,
+      "learning_rate": 1.7854375929033355e-07,
+      "loss": 0.6513,
+      "step": 5099
+    },
+    {
+      "epoch": 9.444444444444445,
+      "grad_norm": 1.5658834118143432,
+      "learning_rate": 1.785344778611457e-07,
+      "loss": 0.5689,
+      "step": 5100
+    },
+    {
+      "epoch": 9.446296296296296,
+      "grad_norm": 1.540458788306279,
+      "learning_rate": 1.7852519466628944e-07,
+      "loss": 0.5959,
+      "step": 5101
+    },
+    {
+      "epoch": 9.448148148148148,
+      "grad_norm": 1.6910202056877777,
+      "learning_rate": 1.7851590970597356e-07,
+      "loss": 0.625,
+      "step": 5102
+    },
+    {
+      "epoch": 9.45,
+      "grad_norm": 1.5519802611360196,
+      "learning_rate": 1.7850662298040676e-07,
+      "loss": 0.6103,
+      "step": 5103
+    },
+    {
+      "epoch": 9.451851851851853,
+      "grad_norm": 1.6778469011059622,
+      "learning_rate": 1.7849733448979784e-07,
+      "loss": 0.6501,
+      "step": 5104
+    },
+    {
+      "epoch": 9.453703703703704,
+      "grad_norm": 1.5762673708222676,
+      "learning_rate": 1.7848804423435568e-07,
+      "loss": 0.6202,
+      "step": 5105
+    },
+    {
+      "epoch": 9.455555555555556,
+      "grad_norm": 1.879610652802891,
+      "learning_rate": 1.784787522142891e-07,
+      "loss": 0.6902,
+      "step": 5106
+    },
+    {
+      "epoch": 9.457407407407407,
+      "grad_norm": 1.5301889028355498,
+      "learning_rate": 1.7846945842980702e-07,
+      "loss": 0.6206,
+      "step": 5107
+    },
+    {
+      "epoch": 9.459259259259259,
+      "grad_norm": 2.1021513100141274,
+      "learning_rate": 1.7846016288111843e-07,
+      "loss": 0.6385,
+      "step": 5108
+    },
+    {
+      "epoch": 9.46111111111111,
+      "grad_norm": 1.6625489146824344,
+      "learning_rate": 1.7845086556843224e-07,
+      "loss": 0.6725,
+      "step": 5109
+    },
+    {
+      "epoch": 9.462962962962964,
+      "grad_norm": 2.062150417433819,
+      "learning_rate": 1.7844156649195757e-07,
+      "loss": 0.5987,
+      "step": 5110
+    },
+    {
+      "epoch": 9.464814814814815,
+      "grad_norm": 1.6708835458354963,
+      "learning_rate": 1.7843226565190342e-07,
+      "loss": 0.6644,
+      "step": 5111
+    },
+    {
+      "epoch": 9.466666666666667,
+      "grad_norm": 1.7331393540503923,
+      "learning_rate": 1.784229630484789e-07,
+      "loss": 0.6783,
+      "step": 5112
+    },
+    {
+      "epoch": 9.468518518518518,
+      "grad_norm": 1.70479119710824,
+      "learning_rate": 1.784136586818932e-07,
+      "loss": 0.6195,
+      "step": 5113
+    },
+    {
+      "epoch": 9.47037037037037,
+      "grad_norm": 1.434663607746278,
+      "learning_rate": 1.7840435255235547e-07,
+      "loss": 0.6209,
+      "step": 5114
+    },
+    {
+      "epoch": 9.472222222222221,
+      "grad_norm": 2.277514589162537,
+      "learning_rate": 1.7839504466007497e-07,
+      "loss": 0.6388,
+      "step": 5115
+    },
+    {
+      "epoch": 9.474074074074075,
+      "grad_norm": 1.4296825968964497,
+      "learning_rate": 1.7838573500526093e-07,
+      "loss": 0.6048,
+      "step": 5116
+    },
+    {
+      "epoch": 9.475925925925926,
+      "grad_norm": 1.8716381595229283,
+      "learning_rate": 1.7837642358812267e-07,
+      "loss": 0.6581,
+      "step": 5117
+    },
+    {
+      "epoch": 9.477777777777778,
+      "grad_norm": 1.5201973085330085,
+      "learning_rate": 1.7836711040886953e-07,
+      "loss": 0.5798,
+      "step": 5118
+    },
+    {
+      "epoch": 9.47962962962963,
+      "grad_norm": 1.6050180766263866,
+      "learning_rate": 1.7835779546771096e-07,
+      "loss": 0.6443,
+      "step": 5119
+    },
+    {
+      "epoch": 9.481481481481481,
+      "grad_norm": 1.5172610186244304,
+      "learning_rate": 1.783484787648563e-07,
+      "loss": 0.6428,
+      "step": 5120
+    },
+    {
+      "epoch": 9.483333333333333,
+      "grad_norm": 1.618322725372126,
+      "learning_rate": 1.7833916030051503e-07,
+      "loss": 0.6407,
+      "step": 5121
+    },
+    {
+      "epoch": 9.485185185185186,
+      "grad_norm": 1.8773323526928989,
+      "learning_rate": 1.7832984007489668e-07,
+      "loss": 0.6215,
+      "step": 5122
+    },
+    {
+      "epoch": 9.487037037037037,
+      "grad_norm": 2.108726855722813,
+      "learning_rate": 1.783205180882108e-07,
+      "loss": 0.6491,
+      "step": 5123
+    },
+    {
+      "epoch": 9.488888888888889,
+      "grad_norm": 1.639773408382761,
+      "learning_rate": 1.7831119434066694e-07,
+      "loss": 0.6489,
+      "step": 5124
+    },
+    {
+      "epoch": 9.49074074074074,
+      "grad_norm": 2.0722175704082835,
+      "learning_rate": 1.7830186883247476e-07,
+      "loss": 0.661,
+      "step": 5125
+    },
+    {
+      "epoch": 9.492592592592592,
+      "grad_norm": 2.067283352480185,
+      "learning_rate": 1.782925415638439e-07,
+      "loss": 0.6838,
+      "step": 5126
+    },
+    {
+      "epoch": 9.494444444444444,
+      "grad_norm": 1.3923213384515385,
+      "learning_rate": 1.7828321253498403e-07,
+      "loss": 0.5992,
+      "step": 5127
+    },
+    {
+      "epoch": 9.496296296296297,
+      "grad_norm": 1.5739686797871817,
+      "learning_rate": 1.7827388174610497e-07,
+      "loss": 0.6504,
+      "step": 5128
+    },
+    {
+      "epoch": 9.498148148148148,
+      "grad_norm": 1.6540811415318923,
+      "learning_rate": 1.7826454919741645e-07,
+      "loss": 0.6271,
+      "step": 5129
+    },
+    {
+      "epoch": 9.5,
+      "grad_norm": 3.8587257914372284,
+      "learning_rate": 1.782552148891283e-07,
+      "loss": 0.6497,
+      "step": 5130
+    },
+    {
+      "epoch": 9.501851851851852,
+      "grad_norm": 1.6799831928268083,
+      "learning_rate": 1.7824587882145036e-07,
+      "loss": 0.5996,
+      "step": 5131
+    },
+    {
+      "epoch": 9.503703703703703,
+      "grad_norm": 1.5049728621744716,
+      "learning_rate": 1.7823654099459258e-07,
+      "loss": 0.6363,
+      "step": 5132
+    },
+    {
+      "epoch": 9.505555555555556,
+      "grad_norm": 3.021252466015714,
+      "learning_rate": 1.7822720140876485e-07,
+      "loss": 0.6546,
+      "step": 5133
+    },
+    {
+      "epoch": 9.507407407407408,
+      "grad_norm": 3.196935502182323,
+      "learning_rate": 1.7821786006417718e-07,
+      "loss": 0.5997,
+      "step": 5134
+    },
+    {
+      "epoch": 9.50925925925926,
+      "grad_norm": 1.5023900063816933,
+      "learning_rate": 1.7820851696103958e-07,
+      "loss": 0.6421,
+      "step": 5135
+    },
+    {
+      "epoch": 9.511111111111111,
+      "grad_norm": 1.6237543035315154,
+      "learning_rate": 1.7819917209956211e-07,
+      "loss": 0.6999,
+      "step": 5136
+    },
+    {
+      "epoch": 9.512962962962963,
+      "grad_norm": 1.5790958928468297,
+      "learning_rate": 1.781898254799549e-07,
+      "loss": 0.6622,
+      "step": 5137
+    },
+    {
+      "epoch": 9.514814814814814,
+      "grad_norm": 1.7173060074627764,
+      "learning_rate": 1.78180477102428e-07,
+      "loss": 0.6191,
+      "step": 5138
+    },
+    {
+      "epoch": 9.516666666666667,
+      "grad_norm": 1.6711628774117762,
+      "learning_rate": 1.7817112696719166e-07,
+      "loss": 0.607,
+      "step": 5139
+    },
+    {
+      "epoch": 9.518518518518519,
+      "grad_norm": 1.6399646472875993,
+      "learning_rate": 1.7816177507445606e-07,
+      "loss": 0.6494,
+      "step": 5140
+    },
+    {
+      "epoch": 9.52037037037037,
+      "grad_norm": 1.695678394425491,
+      "learning_rate": 1.7815242142443147e-07,
+      "loss": 0.6201,
+      "step": 5141
+    },
+    {
+      "epoch": 9.522222222222222,
+      "grad_norm": 1.6957190773111068,
+      "learning_rate": 1.781430660173282e-07,
+      "loss": 0.6229,
+      "step": 5142
+    },
+    {
+      "epoch": 9.524074074074074,
+      "grad_norm": 1.497449595135095,
+      "learning_rate": 1.781337088533566e-07,
+      "loss": 0.6557,
+      "step": 5143
+    },
+    {
+      "epoch": 9.525925925925925,
+      "grad_norm": 1.6345201826860112,
+      "learning_rate": 1.7812434993272702e-07,
+      "loss": 0.6589,
+      "step": 5144
+    },
+    {
+      "epoch": 9.527777777777779,
+      "grad_norm": 1.5690363683069524,
+      "learning_rate": 1.781149892556498e-07,
+      "loss": 0.5877,
+      "step": 5145
+    },
+    {
+      "epoch": 9.52962962962963,
+      "grad_norm": 2.1255811656063455,
+      "learning_rate": 1.7810562682233552e-07,
+      "loss": 0.6219,
+      "step": 5146
+    },
+    {
+      "epoch": 9.531481481481482,
+      "grad_norm": 2.534527525318062,
+      "learning_rate": 1.780962626329946e-07,
+      "loss": 0.6553,
+      "step": 5147
+    },
+    {
+      "epoch": 9.533333333333333,
+      "grad_norm": 1.64193888496852,
+      "learning_rate": 1.780868966878376e-07,
+      "loss": 0.6376,
+      "step": 5148
+    },
+    {
+      "epoch": 9.535185185185185,
+      "grad_norm": 1.725739169929844,
+      "learning_rate": 1.7807752898707507e-07,
+      "loss": 0.6249,
+      "step": 5149
+    },
+    {
+      "epoch": 9.537037037037036,
+      "grad_norm": 1.4721775872082243,
+      "learning_rate": 1.7806815953091766e-07,
+      "loss": 0.6355,
+      "step": 5150
+    },
+    {
+      "epoch": 9.53888888888889,
+      "grad_norm": 1.8340492004921911,
+      "learning_rate": 1.7805878831957597e-07,
+      "loss": 0.6165,
+      "step": 5151
+    },
+    {
+      "epoch": 9.540740740740741,
+      "grad_norm": 1.5088477590182456,
+      "learning_rate": 1.780494153532607e-07,
+      "loss": 0.6542,
+      "step": 5152
+    },
+    {
+      "epoch": 9.542592592592593,
+      "grad_norm": 1.6109125651047485,
+      "learning_rate": 1.780400406321826e-07,
+      "loss": 0.6269,
+      "step": 5153
+    },
+    {
+      "epoch": 9.544444444444444,
+      "grad_norm": 1.5492598227531769,
+      "learning_rate": 1.7803066415655245e-07,
+      "loss": 0.6053,
+      "step": 5154
+    },
+    {
+      "epoch": 9.546296296296296,
+      "grad_norm": 1.671871532019184,
+      "learning_rate": 1.7802128592658105e-07,
+      "loss": 0.6512,
+      "step": 5155
+    },
+    {
+      "epoch": 9.548148148148147,
+      "grad_norm": 1.6899592914073456,
+      "learning_rate": 1.7801190594247923e-07,
+      "loss": 0.6465,
+      "step": 5156
+    },
+    {
+      "epoch": 9.55,
+      "grad_norm": 1.456399761086655,
+      "learning_rate": 1.7800252420445787e-07,
+      "loss": 0.6107,
+      "step": 5157
+    },
+    {
+      "epoch": 9.551851851851852,
+      "grad_norm": 1.7004387622656227,
+      "learning_rate": 1.7799314071272794e-07,
+      "loss": 0.6229,
+      "step": 5158
+    },
+    {
+      "epoch": 9.553703703703704,
+      "grad_norm": 1.6615469148594542,
+      "learning_rate": 1.7798375546750038e-07,
+      "loss": 0.609,
+      "step": 5159
+    },
+    {
+      "epoch": 9.555555555555555,
+      "grad_norm": 1.4667138801761859,
+      "learning_rate": 1.7797436846898617e-07,
+      "loss": 0.5894,
+      "step": 5160
+    },
+    {
+      "epoch": 9.557407407407407,
+      "grad_norm": 1.9053834502967453,
+      "learning_rate": 1.779649797173964e-07,
+      "loss": 0.6887,
+      "step": 5161
+    },
+    {
+      "epoch": 9.559259259259258,
+      "grad_norm": 1.4768371089592753,
+      "learning_rate": 1.7795558921294212e-07,
+      "loss": 0.622,
+      "step": 5162
+    },
+    {
+      "epoch": 9.561111111111112,
+      "grad_norm": 2.6354511803852234,
+      "learning_rate": 1.779461969558345e-07,
+      "loss": 0.6399,
+      "step": 5163
+    },
+    {
+      "epoch": 9.562962962962963,
+      "grad_norm": 1.4310463367017272,
+      "learning_rate": 1.7793680294628465e-07,
+      "loss": 0.6697,
+      "step": 5164
+    },
+    {
+      "epoch": 9.564814814814815,
+      "grad_norm": 2.484626730847231,
+      "learning_rate": 1.779274071845038e-07,
+      "loss": 0.6155,
+      "step": 5165
+    },
+    {
+      "epoch": 9.566666666666666,
+      "grad_norm": 1.4971905872094466,
+      "learning_rate": 1.779180096707032e-07,
+      "loss": 0.6225,
+      "step": 5166
+    },
+    {
+      "epoch": 9.568518518518518,
+      "grad_norm": 1.5802732312685048,
+      "learning_rate": 1.779086104050941e-07,
+      "loss": 0.6346,
+      "step": 5167
+    },
+    {
+      "epoch": 9.57037037037037,
+      "grad_norm": 1.394884073892824,
+      "learning_rate": 1.7789920938788787e-07,
+      "loss": 0.6268,
+      "step": 5168
+    },
+    {
+      "epoch": 9.572222222222223,
+      "grad_norm": 1.5276213110667942,
+      "learning_rate": 1.7788980661929583e-07,
+      "loss": 0.629,
+      "step": 5169
+    },
+    {
+      "epoch": 9.574074074074074,
+      "grad_norm": 1.546255824459983,
+      "learning_rate": 1.778804020995294e-07,
+      "loss": 0.6479,
+      "step": 5170
+    },
+    {
+      "epoch": 9.575925925925926,
+      "grad_norm": 1.6529362687982083,
+      "learning_rate": 1.7787099582879998e-07,
+      "loss": 0.6371,
+      "step": 5171
+    },
+    {
+      "epoch": 9.577777777777778,
+      "grad_norm": 1.4738197643713067,
+      "learning_rate": 1.778615878073191e-07,
+      "loss": 0.6295,
+      "step": 5172
+    },
+    {
+      "epoch": 9.579629629629629,
+      "grad_norm": 1.6154524152006462,
+      "learning_rate": 1.7785217803529825e-07,
+      "loss": 0.5942,
+      "step": 5173
+    },
+    {
+      "epoch": 9.581481481481482,
+      "grad_norm": 1.6714505474308243,
+      "learning_rate": 1.77842766512949e-07,
+      "loss": 0.6211,
+      "step": 5174
+    },
+    {
+      "epoch": 9.583333333333334,
+      "grad_norm": 1.4056140393538294,
+      "learning_rate": 1.7783335324048294e-07,
+      "loss": 0.6406,
+      "step": 5175
+    },
+    {
+      "epoch": 9.585185185185185,
+      "grad_norm": 1.7289370028218976,
+      "learning_rate": 1.778239382181117e-07,
+      "loss": 0.6361,
+      "step": 5176
+    },
+    {
+      "epoch": 9.587037037037037,
+      "grad_norm": 1.4221791421804786,
+      "learning_rate": 1.77814521446047e-07,
+      "loss": 0.6235,
+      "step": 5177
+    },
+    {
+      "epoch": 9.588888888888889,
+      "grad_norm": 1.6712950723596691,
+      "learning_rate": 1.7780510292450046e-07,
+      "loss": 0.6586,
+      "step": 5178
+    },
+    {
+      "epoch": 9.59074074074074,
+      "grad_norm": 1.643774154377951,
+      "learning_rate": 1.7779568265368394e-07,
+      "loss": 0.6017,
+      "step": 5179
+    },
+    {
+      "epoch": 9.592592592592592,
+      "grad_norm": 1.4603321904356052,
+      "learning_rate": 1.7778626063380915e-07,
+      "loss": 0.6208,
+      "step": 5180
+    },
+    {
+      "epoch": 9.594444444444445,
+      "grad_norm": 1.3769500135065982,
+      "learning_rate": 1.77776836865088e-07,
+      "loss": 0.6763,
+      "step": 5181
+    },
+    {
+      "epoch": 9.596296296296297,
+      "grad_norm": 1.5225688446939925,
+      "learning_rate": 1.7776741134773227e-07,
+      "loss": 0.6339,
+      "step": 5182
+    },
+    {
+      "epoch": 9.598148148148148,
+      "grad_norm": 1.41329568822239,
+      "learning_rate": 1.7775798408195392e-07,
+      "loss": 0.6313,
+      "step": 5183
+    },
+    {
+      "epoch": 9.6,
+      "grad_norm": 2.763628993492429,
+      "learning_rate": 1.7774855506796493e-07,
+      "loss": 0.6063,
+      "step": 5184
+    },
+    {
+      "epoch": 9.601851851851851,
+      "grad_norm": 2.099629127338547,
+      "learning_rate": 1.7773912430597723e-07,
+      "loss": 0.648,
+      "step": 5185
+    },
+    {
+      "epoch": 9.603703703703705,
+      "grad_norm": 1.6274385462837548,
+      "learning_rate": 1.7772969179620293e-07,
+      "loss": 0.6653,
+      "step": 5186
+    },
+    {
+      "epoch": 9.605555555555556,
+      "grad_norm": 1.4847810567716533,
+      "learning_rate": 1.7772025753885397e-07,
+      "loss": 0.6121,
+      "step": 5187
+    },
+    {
+      "epoch": 9.607407407407408,
+      "grad_norm": 1.552257872441247,
+      "learning_rate": 1.777108215341426e-07,
+      "loss": 0.6946,
+      "step": 5188
+    },
+    {
+      "epoch": 9.60925925925926,
+      "grad_norm": 2.602667836660394,
+      "learning_rate": 1.777013837822809e-07,
+      "loss": 0.638,
+      "step": 5189
+    },
+    {
+      "epoch": 9.61111111111111,
+      "grad_norm": 1.4428224745624634,
+      "learning_rate": 1.7769194428348105e-07,
+      "loss": 0.6261,
+      "step": 5190
+    },
+    {
+      "epoch": 9.612962962962962,
+      "grad_norm": 1.385332057250731,
+      "learning_rate": 1.7768250303795527e-07,
+      "loss": 0.6089,
+      "step": 5191
+    },
+    {
+      "epoch": 9.614814814814816,
+      "grad_norm": 1.6071243225937228,
+      "learning_rate": 1.7767306004591584e-07,
+      "loss": 0.6226,
+      "step": 5192
+    },
+    {
+      "epoch": 9.616666666666667,
+      "grad_norm": 1.714399645500842,
+      "learning_rate": 1.7766361530757506e-07,
+      "loss": 0.6649,
+      "step": 5193
+    },
+    {
+      "epoch": 9.618518518518519,
+      "grad_norm": 1.8431869429813752,
+      "learning_rate": 1.776541688231453e-07,
+      "loss": 0.62,
+      "step": 5194
+    },
+    {
+      "epoch": 9.62037037037037,
+      "grad_norm": 1.5711684801866628,
+      "learning_rate": 1.776447205928389e-07,
+      "loss": 0.6078,
+      "step": 5195
+    },
+    {
+      "epoch": 9.622222222222222,
+      "grad_norm": 1.5294516711240607,
+      "learning_rate": 1.776352706168683e-07,
+      "loss": 0.6681,
+      "step": 5196
+    },
+    {
+      "epoch": 9.624074074074073,
+      "grad_norm": 1.7868353780968012,
+      "learning_rate": 1.7762581889544597e-07,
+      "loss": 0.6329,
+      "step": 5197
+    },
+    {
+      "epoch": 9.625925925925927,
+      "grad_norm": 1.3568398851367105,
+      "learning_rate": 1.776163654287844e-07,
+      "loss": 0.6291,
+      "step": 5198
+    },
+    {
+      "epoch": 9.627777777777778,
+      "grad_norm": 1.4893094509209255,
+      "learning_rate": 1.7760691021709613e-07,
+      "loss": 0.6506,
+      "step": 5199
+    },
+    {
+      "epoch": 9.62962962962963,
+      "grad_norm": 1.874732365358208,
+      "learning_rate": 1.7759745326059376e-07,
+      "loss": 0.6492,
+      "step": 5200
+    },
+    {
+      "epoch": 9.631481481481481,
+      "grad_norm": 1.9202279984534458,
+      "learning_rate": 1.7758799455948985e-07,
+      "loss": 0.642,
+      "step": 5201
+    },
+    {
+      "epoch": 9.633333333333333,
+      "grad_norm": 2.3227677664316544,
+      "learning_rate": 1.7757853411399712e-07,
+      "loss": 0.6467,
+      "step": 5202
+    },
+    {
+      "epoch": 9.635185185185184,
+      "grad_norm": 1.613038984739922,
+      "learning_rate": 1.7756907192432826e-07,
+      "loss": 0.6274,
+      "step": 5203
+    },
+    {
+      "epoch": 9.637037037037038,
+      "grad_norm": 1.4347573558348037,
+      "learning_rate": 1.7755960799069597e-07,
+      "loss": 0.6263,
+      "step": 5204
+    },
+    {
+      "epoch": 9.63888888888889,
+      "grad_norm": 1.7466899842732566,
+      "learning_rate": 1.7755014231331305e-07,
+      "loss": 0.638,
+      "step": 5205
+    },
+    {
+      "epoch": 9.64074074074074,
+      "grad_norm": 1.5846791364161328,
+      "learning_rate": 1.7754067489239227e-07,
+      "loss": 0.6778,
+      "step": 5206
+    },
+    {
+      "epoch": 9.642592592592592,
+      "grad_norm": 3.6658872901621606,
+      "learning_rate": 1.7753120572814656e-07,
+      "loss": 0.6194,
+      "step": 5207
+    },
+    {
+      "epoch": 9.644444444444444,
+      "grad_norm": 1.5164463332868299,
+      "learning_rate": 1.7752173482078878e-07,
+      "loss": 0.6609,
+      "step": 5208
+    },
+    {
+      "epoch": 9.646296296296295,
+      "grad_norm": 1.7075193381584803,
+      "learning_rate": 1.7751226217053183e-07,
+      "loss": 0.6115,
+      "step": 5209
+    },
+    {
+      "epoch": 9.648148148148149,
+      "grad_norm": 1.7424322027265615,
+      "learning_rate": 1.7750278777758873e-07,
+      "loss": 0.6028,
+      "step": 5210
+    },
+    {
+      "epoch": 9.65,
+      "grad_norm": 1.5986285043153505,
+      "learning_rate": 1.7749331164217247e-07,
+      "loss": 0.6683,
+      "step": 5211
+    },
+    {
+      "epoch": 9.651851851851852,
+      "grad_norm": 1.747458245049754,
+      "learning_rate": 1.7748383376449608e-07,
+      "loss": 0.639,
+      "step": 5212
+    },
+    {
+      "epoch": 9.653703703703703,
+      "grad_norm": 1.815650062970967,
+      "learning_rate": 1.7747435414477265e-07,
+      "loss": 0.6187,
+      "step": 5213
+    },
+    {
+      "epoch": 9.655555555555555,
+      "grad_norm": 1.4163869171878571,
+      "learning_rate": 1.7746487278321536e-07,
+      "loss": 0.6136,
+      "step": 5214
+    },
+    {
+      "epoch": 9.657407407407408,
+      "grad_norm": 1.37657137540675,
+      "learning_rate": 1.774553896800373e-07,
+      "loss": 0.6349,
+      "step": 5215
+    },
+    {
+      "epoch": 9.65925925925926,
+      "grad_norm": 1.682017507004502,
+      "learning_rate": 1.7744590483545174e-07,
+      "loss": 0.6413,
+      "step": 5216
+    },
+    {
+      "epoch": 9.661111111111111,
+      "grad_norm": 1.7080671854608027,
+      "learning_rate": 1.774364182496719e-07,
+      "loss": 0.6344,
+      "step": 5217
+    },
+    {
+      "epoch": 9.662962962962963,
+      "grad_norm": 1.74956277306685,
+      "learning_rate": 1.7742692992291103e-07,
+      "loss": 0.6696,
+      "step": 5218
+    },
+    {
+      "epoch": 9.664814814814815,
+      "grad_norm": 1.8695292464640438,
+      "learning_rate": 1.7741743985538253e-07,
+      "loss": 0.6596,
+      "step": 5219
+    },
+    {
+      "epoch": 9.666666666666666,
+      "grad_norm": 1.6476566264566908,
+      "learning_rate": 1.774079480472997e-07,
+      "loss": 0.6604,
+      "step": 5220
+    },
+    {
+      "epoch": 9.668518518518518,
+      "grad_norm": 1.5783736071658105,
+      "learning_rate": 1.7739845449887593e-07,
+      "loss": 0.6457,
+      "step": 5221
+    },
+    {
+      "epoch": 9.670370370370371,
+      "grad_norm": 1.9794647868971647,
+      "learning_rate": 1.7738895921032475e-07,
+      "loss": 0.655,
+      "step": 5222
+    },
+    {
+      "epoch": 9.672222222222222,
+      "grad_norm": 2.006583263700009,
+      "learning_rate": 1.7737946218185956e-07,
+      "loss": 0.6569,
+      "step": 5223
+    },
+    {
+      "epoch": 9.674074074074074,
+      "grad_norm": 2.2476789975727973,
+      "learning_rate": 1.773699634136939e-07,
+      "loss": 0.6326,
+      "step": 5224
+    },
+    {
+      "epoch": 9.675925925925926,
+      "grad_norm": 1.5229373863324145,
+      "learning_rate": 1.7736046290604132e-07,
+      "loss": 0.6416,
+      "step": 5225
+    },
+    {
+      "epoch": 9.677777777777777,
+      "grad_norm": 1.524412022051299,
+      "learning_rate": 1.7735096065911543e-07,
+      "loss": 0.6532,
+      "step": 5226
+    },
+    {
+      "epoch": 9.67962962962963,
+      "grad_norm": 1.7604936797410624,
+      "learning_rate": 1.7734145667312984e-07,
+      "loss": 0.6304,
+      "step": 5227
+    },
+    {
+      "epoch": 9.681481481481482,
+      "grad_norm": 1.5751223914392847,
+      "learning_rate": 1.773319509482983e-07,
+      "loss": 0.6129,
+      "step": 5228
+    },
+    {
+      "epoch": 9.683333333333334,
+      "grad_norm": 1.5848999306656517,
+      "learning_rate": 1.7732244348483443e-07,
+      "loss": 0.6085,
+      "step": 5229
+    },
+    {
+      "epoch": 9.685185185185185,
+      "grad_norm": 1.5888462830377457,
+      "learning_rate": 1.7731293428295204e-07,
+      "loss": 0.5864,
+      "step": 5230
+    },
+    {
+      "epoch": 9.687037037037037,
+      "grad_norm": 1.6421499962017303,
+      "learning_rate": 1.7730342334286492e-07,
+      "loss": 0.6146,
+      "step": 5231
+    },
+    {
+      "epoch": 9.688888888888888,
+      "grad_norm": 1.5107469942687584,
+      "learning_rate": 1.7729391066478686e-07,
+      "loss": 0.6208,
+      "step": 5232
+    },
+    {
+      "epoch": 9.690740740740742,
+      "grad_norm": 1.747309535084251,
+      "learning_rate": 1.7728439624893178e-07,
+      "loss": 0.6573,
+      "step": 5233
+    },
+    {
+      "epoch": 9.692592592592593,
+      "grad_norm": 1.9454191852625105,
+      "learning_rate": 1.772748800955136e-07,
+      "loss": 0.629,
+      "step": 5234
+    },
+    {
+      "epoch": 9.694444444444445,
+      "grad_norm": 1.6448648014087421,
+      "learning_rate": 1.772653622047462e-07,
+      "loss": 0.6271,
+      "step": 5235
+    },
+    {
+      "epoch": 9.696296296296296,
+      "grad_norm": 1.4772561174760583,
+      "learning_rate": 1.7725584257684362e-07,
+      "loss": 0.6263,
+      "step": 5236
+    },
+    {
+      "epoch": 9.698148148148148,
+      "grad_norm": 1.650343331512508,
+      "learning_rate": 1.7724632121201986e-07,
+      "loss": 0.6498,
+      "step": 5237
+    },
+    {
+      "epoch": 9.7,
+      "grad_norm": 1.8749089689357918,
+      "learning_rate": 1.77236798110489e-07,
+      "loss": 0.6204,
+      "step": 5238
+    },
+    {
+      "epoch": 9.701851851851853,
+      "grad_norm": 1.4546396285868493,
+      "learning_rate": 1.7722727327246515e-07,
+      "loss": 0.642,
+      "step": 5239
+    },
+    {
+      "epoch": 9.703703703703704,
+      "grad_norm": 1.4572608165712329,
+      "learning_rate": 1.7721774669816248e-07,
+      "loss": 0.6653,
+      "step": 5240
+    },
+    {
+      "epoch": 9.705555555555556,
+      "grad_norm": 1.5192498582193945,
+      "learning_rate": 1.7720821838779511e-07,
+      "loss": 0.67,
+      "step": 5241
+    },
+    {
+      "epoch": 9.707407407407407,
+      "grad_norm": 1.494429690589304,
+      "learning_rate": 1.7719868834157734e-07,
+      "loss": 0.6647,
+      "step": 5242
+    },
+    {
+      "epoch": 9.709259259259259,
+      "grad_norm": 1.4287165905348957,
+      "learning_rate": 1.7718915655972336e-07,
+      "loss": 0.6596,
+      "step": 5243
+    },
+    {
+      "epoch": 9.71111111111111,
+      "grad_norm": 1.5115700700826356,
+      "learning_rate": 1.7717962304244748e-07,
+      "loss": 0.6411,
+      "step": 5244
+    },
+    {
+      "epoch": 9.712962962962964,
+      "grad_norm": 1.5400094203294483,
+      "learning_rate": 1.771700877899641e-07,
+      "loss": 0.615,
+      "step": 5245
+    },
+    {
+      "epoch": 9.714814814814815,
+      "grad_norm": 1.775503954826172,
+      "learning_rate": 1.771605508024875e-07,
+      "loss": 0.6644,
+      "step": 5246
+    },
+    {
+      "epoch": 9.716666666666667,
+      "grad_norm": 1.6123763552366148,
+      "learning_rate": 1.771510120802322e-07,
+      "loss": 0.6777,
+      "step": 5247
+    },
+    {
+      "epoch": 9.718518518518518,
+      "grad_norm": 1.5934074967841956,
+      "learning_rate": 1.771414716234126e-07,
+      "loss": 0.66,
+      "step": 5248
+    },
+    {
+      "epoch": 9.72037037037037,
+      "grad_norm": 3.6390642150073647,
+      "learning_rate": 1.771319294322432e-07,
+      "loss": 0.6279,
+      "step": 5249
+    },
+    {
+      "epoch": 9.722222222222221,
+      "grad_norm": 1.4559192247007011,
+      "learning_rate": 1.7712238550693853e-07,
+      "loss": 0.6046,
+      "step": 5250
+    },
+    {
+      "epoch": 9.724074074074075,
+      "grad_norm": 1.5205214395705757,
+      "learning_rate": 1.7711283984771317e-07,
+      "loss": 0.6399,
+      "step": 5251
+    },
+    {
+      "epoch": 9.725925925925926,
+      "grad_norm": 1.3154845564893412,
+      "learning_rate": 1.7710329245478175e-07,
+      "loss": 0.6508,
+      "step": 5252
+    },
+    {
+      "epoch": 9.727777777777778,
+      "grad_norm": 1.7129914179371908,
+      "learning_rate": 1.7709374332835892e-07,
+      "loss": 0.6251,
+      "step": 5253
+    },
+    {
+      "epoch": 9.72962962962963,
+      "grad_norm": 1.4944424389958704,
+      "learning_rate": 1.7708419246865934e-07,
+      "loss": 0.6002,
+      "step": 5254
+    },
+    {
+      "epoch": 9.731481481481481,
+      "grad_norm": 1.5077483905996683,
+      "learning_rate": 1.7707463987589772e-07,
+      "loss": 0.6613,
+      "step": 5255
+    },
+    {
+      "epoch": 9.733333333333333,
+      "grad_norm": 1.630186030937466,
+      "learning_rate": 1.7706508555028893e-07,
+      "loss": 0.652,
+      "step": 5256
+    },
+    {
+      "epoch": 9.735185185185186,
+      "grad_norm": 1.5880493840753005,
+      "learning_rate": 1.7705552949204767e-07,
+      "loss": 0.6419,
+      "step": 5257
+    },
+    {
+      "epoch": 9.737037037037037,
+      "grad_norm": 1.5770507669124723,
+      "learning_rate": 1.7704597170138884e-07,
+      "loss": 0.6157,
+      "step": 5258
+    },
+    {
+      "epoch": 9.738888888888889,
+      "grad_norm": 1.4930190540130588,
+      "learning_rate": 1.770364121785273e-07,
+      "loss": 0.6493,
+      "step": 5259
+    },
+    {
+      "epoch": 9.74074074074074,
+      "grad_norm": 1.646377011549875,
+      "learning_rate": 1.77026850923678e-07,
+      "loss": 0.6613,
+      "step": 5260
+    },
+    {
+      "epoch": 9.742592592592592,
+      "grad_norm": 1.5063095099695083,
+      "learning_rate": 1.7701728793705589e-07,
+      "loss": 0.6197,
+      "step": 5261
+    },
+    {
+      "epoch": 9.744444444444444,
+      "grad_norm": 1.6553336471964604,
+      "learning_rate": 1.7700772321887595e-07,
+      "loss": 0.6186,
+      "step": 5262
+    },
+    {
+      "epoch": 9.746296296296297,
+      "grad_norm": 1.9790675332897267,
+      "learning_rate": 1.7699815676935327e-07,
+      "loss": 0.6633,
+      "step": 5263
+    },
+    {
+      "epoch": 9.748148148148148,
+      "grad_norm": 1.363777732735928,
+      "learning_rate": 1.769885885887029e-07,
+      "loss": 0.6359,
+      "step": 5264
+    },
+    {
+      "epoch": 9.75,
+      "grad_norm": 1.498978146140254,
+      "learning_rate": 1.7697901867713995e-07,
+      "loss": 0.6541,
+      "step": 5265
+    },
+    {
+      "epoch": 9.751851851851852,
+      "grad_norm": 1.6838159098378294,
+      "learning_rate": 1.7696944703487958e-07,
+      "loss": 0.6837,
+      "step": 5266
+    },
+    {
+      "epoch": 9.753703703703703,
+      "grad_norm": 1.514738306471442,
+      "learning_rate": 1.7695987366213702e-07,
+      "loss": 0.6426,
+      "step": 5267
+    },
+    {
+      "epoch": 9.755555555555556,
+      "grad_norm": 1.6546117840693506,
+      "learning_rate": 1.7695029855912743e-07,
+      "loss": 0.6833,
+      "step": 5268
+    },
+    {
+      "epoch": 9.757407407407408,
+      "grad_norm": 1.4651313483032045,
+      "learning_rate": 1.7694072172606618e-07,
+      "loss": 0.6115,
+      "step": 5269
+    },
+    {
+      "epoch": 9.75925925925926,
+      "grad_norm": 2.0002075217696738,
+      "learning_rate": 1.7693114316316854e-07,
+      "loss": 0.6583,
+      "step": 5270
+    },
+    {
+      "epoch": 9.761111111111111,
+      "grad_norm": 1.7746366776187028,
+      "learning_rate": 1.7692156287064984e-07,
+      "loss": 0.6356,
+      "step": 5271
+    },
+    {
+      "epoch": 9.762962962962963,
+      "grad_norm": 1.786737727680658,
+      "learning_rate": 1.769119808487255e-07,
+      "loss": 0.6106,
+      "step": 5272
+    },
+    {
+      "epoch": 9.764814814814814,
+      "grad_norm": 1.464309271783353,
+      "learning_rate": 1.7690239709761095e-07,
+      "loss": 0.5845,
+      "step": 5273
+    },
+    {
+      "epoch": 9.766666666666667,
+      "grad_norm": 1.5374527532805757,
+      "learning_rate": 1.7689281161752164e-07,
+      "loss": 0.5869,
+      "step": 5274
+    },
+    {
+      "epoch": 9.768518518518519,
+      "grad_norm": 1.7620044736693015,
+      "learning_rate": 1.7688322440867307e-07,
+      "loss": 0.6231,
+      "step": 5275
+    },
+    {
+      "epoch": 9.77037037037037,
+      "grad_norm": 1.4999844246670662,
+      "learning_rate": 1.768736354712808e-07,
+      "loss": 0.6649,
+      "step": 5276
+    },
+    {
+      "epoch": 9.772222222222222,
+      "grad_norm": 1.5764728308769347,
+      "learning_rate": 1.7686404480556046e-07,
+      "loss": 0.66,
+      "step": 5277
+    },
+    {
+      "epoch": 9.774074074074074,
+      "grad_norm": 1.886952361432808,
+      "learning_rate": 1.768544524117276e-07,
+      "loss": 0.6631,
+      "step": 5278
+    },
+    {
+      "epoch": 9.775925925925925,
+      "grad_norm": 1.5460929826942942,
+      "learning_rate": 1.7684485828999794e-07,
+      "loss": 0.6338,
+      "step": 5279
+    },
+    {
+      "epoch": 9.777777777777779,
+      "grad_norm": 2.708974725338093,
+      "learning_rate": 1.7683526244058716e-07,
+      "loss": 0.6425,
+      "step": 5280
+    },
+    {
+      "epoch": 9.77962962962963,
+      "grad_norm": 1.7331154549128358,
+      "learning_rate": 1.7682566486371094e-07,
+      "loss": 0.6274,
+      "step": 5281
+    },
+    {
+      "epoch": 9.781481481481482,
+      "grad_norm": 1.6884322541823826,
+      "learning_rate": 1.7681606555958518e-07,
+      "loss": 0.6599,
+      "step": 5282
+    },
+    {
+      "epoch": 9.783333333333333,
+      "grad_norm": 1.7651416967728226,
+      "learning_rate": 1.7680646452842562e-07,
+      "loss": 0.6081,
+      "step": 5283
+    },
+    {
+      "epoch": 9.785185185185185,
+      "grad_norm": 1.740057331475001,
+      "learning_rate": 1.7679686177044814e-07,
+      "loss": 0.627,
+      "step": 5284
+    },
+    {
+      "epoch": 9.787037037037036,
+      "grad_norm": 1.5136167931913198,
+      "learning_rate": 1.7678725728586863e-07,
+      "loss": 0.6463,
+      "step": 5285
+    },
+    {
+      "epoch": 9.78888888888889,
+      "grad_norm": 1.5109049800713026,
+      "learning_rate": 1.7677765107490302e-07,
+      "loss": 0.652,
+      "step": 5286
+    },
+    {
+      "epoch": 9.790740740740741,
+      "grad_norm": 1.6276218339291435,
+      "learning_rate": 1.767680431377673e-07,
+      "loss": 0.6248,
+      "step": 5287
+    },
+    {
+      "epoch": 9.792592592592593,
+      "grad_norm": 2.7170698614671567,
+      "learning_rate": 1.7675843347467746e-07,
+      "loss": 0.6184,
+      "step": 5288
+    },
+    {
+      "epoch": 9.794444444444444,
+      "grad_norm": 1.4623660631081234,
+      "learning_rate": 1.7674882208584955e-07,
+      "loss": 0.6526,
+      "step": 5289
+    },
+    {
+      "epoch": 9.796296296296296,
+      "grad_norm": 1.521351286985686,
+      "learning_rate": 1.7673920897149967e-07,
+      "loss": 0.6159,
+      "step": 5290
+    },
+    {
+      "epoch": 9.798148148148147,
+      "grad_norm": 1.4630503565355908,
+      "learning_rate": 1.76729594131844e-07,
+      "loss": 0.6392,
+      "step": 5291
+    },
+    {
+      "epoch": 9.8,
+      "grad_norm": 1.7630952834722897,
+      "learning_rate": 1.767199775670986e-07,
+      "loss": 0.6093,
+      "step": 5292
+    },
+    {
+      "epoch": 9.801851851851852,
+      "grad_norm": 1.4478057493671301,
+      "learning_rate": 1.7671035927747977e-07,
+      "loss": 0.6181,
+      "step": 5293
+    },
+    {
+      "epoch": 9.803703703703704,
+      "grad_norm": 1.5197541075930006,
+      "learning_rate": 1.767007392632037e-07,
+      "loss": 0.622,
+      "step": 5294
+    },
+    {
+      "epoch": 9.805555555555555,
+      "grad_norm": 1.8091658952086973,
+      "learning_rate": 1.7669111752448672e-07,
+      "loss": 0.6152,
+      "step": 5295
+    },
+    {
+      "epoch": 9.807407407407407,
+      "grad_norm": 1.7136907115795952,
+      "learning_rate": 1.766814940615451e-07,
+      "loss": 0.6496,
+      "step": 5296
+    },
+    {
+      "epoch": 9.809259259259258,
+      "grad_norm": 1.6788840747635472,
+      "learning_rate": 1.7667186887459527e-07,
+      "loss": 0.6679,
+      "step": 5297
+    },
+    {
+      "epoch": 9.811111111111112,
+      "grad_norm": 1.7258284364674372,
+      "learning_rate": 1.7666224196385359e-07,
+      "loss": 0.6687,
+      "step": 5298
+    },
+    {
+      "epoch": 9.812962962962963,
+      "grad_norm": 1.697097113947268,
+      "learning_rate": 1.7665261332953646e-07,
+      "loss": 0.5883,
+      "step": 5299
+    },
+    {
+      "epoch": 9.814814814814815,
+      "grad_norm": 1.4382860883142714,
+      "learning_rate": 1.766429829718604e-07,
+      "loss": 0.6054,
+      "step": 5300
+    },
+    {
+      "epoch": 9.816666666666666,
+      "grad_norm": 1.7931926426564386,
+      "learning_rate": 1.7663335089104194e-07,
+      "loss": 0.6849,
+      "step": 5301
+    },
+    {
+      "epoch": 9.818518518518518,
+      "grad_norm": 1.4821293743274302,
+      "learning_rate": 1.766237170872976e-07,
+      "loss": 0.6523,
+      "step": 5302
+    },
+    {
+      "epoch": 9.82037037037037,
+      "grad_norm": 1.7843876403444132,
+      "learning_rate": 1.7661408156084402e-07,
+      "loss": 0.5702,
+      "step": 5303
+    },
+    {
+      "epoch": 9.822222222222223,
+      "grad_norm": 1.4007354210390233,
+      "learning_rate": 1.766044443118978e-07,
+      "loss": 0.6183,
+      "step": 5304
+    },
+    {
+      "epoch": 9.824074074074074,
+      "grad_norm": 1.4357521960087263,
+      "learning_rate": 1.765948053406756e-07,
+      "loss": 0.5976,
+      "step": 5305
+    },
+    {
+      "epoch": 9.825925925925926,
+      "grad_norm": 1.5725788271690366,
+      "learning_rate": 1.7658516464739417e-07,
+      "loss": 0.5929,
+      "step": 5306
+    },
+    {
+      "epoch": 9.827777777777778,
+      "grad_norm": 4.937413932776677,
+      "learning_rate": 1.7657552223227024e-07,
+      "loss": 0.634,
+      "step": 5307
+    },
+    {
+      "epoch": 9.829629629629629,
+      "grad_norm": 1.5047120246494943,
+      "learning_rate": 1.7656587809552054e-07,
+      "loss": 0.6355,
+      "step": 5308
+    },
+    {
+      "epoch": 9.831481481481482,
+      "grad_norm": 1.5846261773078147,
+      "learning_rate": 1.7655623223736204e-07,
+      "loss": 0.6933,
+      "step": 5309
+    },
+    {
+      "epoch": 9.833333333333334,
+      "grad_norm": 1.6748626802572621,
+      "learning_rate": 1.7654658465801146e-07,
+      "loss": 0.5682,
+      "step": 5310
+    },
+    {
+      "epoch": 9.835185185185185,
+      "grad_norm": 1.5915870294780337,
+      "learning_rate": 1.7653693535768578e-07,
+      "loss": 0.5975,
+      "step": 5311
+    },
+    {
+      "epoch": 9.837037037037037,
+      "grad_norm": 1.5562064018839896,
+      "learning_rate": 1.7652728433660187e-07,
+      "loss": 0.6757,
+      "step": 5312
+    },
+    {
+      "epoch": 9.838888888888889,
+      "grad_norm": 1.6453119304021864,
+      "learning_rate": 1.765176315949768e-07,
+      "loss": 0.608,
+      "step": 5313
+    },
+    {
+      "epoch": 9.84074074074074,
+      "grad_norm": 1.4980455477625951,
+      "learning_rate": 1.7650797713302756e-07,
+      "loss": 0.6574,
+      "step": 5314
+    },
+    {
+      "epoch": 9.842592592592592,
+      "grad_norm": 1.8328229344616933,
+      "learning_rate": 1.7649832095097121e-07,
+      "loss": 0.6033,
+      "step": 5315
+    },
+    {
+      "epoch": 9.844444444444445,
+      "grad_norm": 1.3542069895684337,
+      "learning_rate": 1.7648866304902482e-07,
+      "loss": 0.6447,
+      "step": 5316
+    },
+    {
+      "epoch": 9.846296296296297,
+      "grad_norm": 1.707694633535408,
+      "learning_rate": 1.7647900342740556e-07,
+      "loss": 0.6693,
+      "step": 5317
+    },
+    {
+      "epoch": 9.848148148148148,
+      "grad_norm": 2.0291500123416286,
+      "learning_rate": 1.7646934208633056e-07,
+      "loss": 0.6256,
+      "step": 5318
+    },
+    {
+      "epoch": 9.85,
+      "grad_norm": 1.7503395391193268,
+      "learning_rate": 1.7645967902601707e-07,
+      "loss": 0.6492,
+      "step": 5319
+    },
+    {
+      "epoch": 9.851851851851851,
+      "grad_norm": 5.864940814403709,
+      "learning_rate": 1.7645001424668235e-07,
+      "loss": 0.6285,
+      "step": 5320
+    },
+    {
+      "epoch": 9.853703703703705,
+      "grad_norm": 2.0773261135265457,
+      "learning_rate": 1.764403477485437e-07,
+      "loss": 0.6434,
+      "step": 5321
+    },
+    {
+      "epoch": 9.855555555555556,
+      "grad_norm": 1.699094600802781,
+      "learning_rate": 1.7643067953181837e-07,
+      "loss": 0.6079,
+      "step": 5322
+    },
+    {
+      "epoch": 9.857407407407408,
+      "grad_norm": 1.4861082914699397,
+      "learning_rate": 1.764210095967238e-07,
+      "loss": 0.6191,
+      "step": 5323
+    },
+    {
+      "epoch": 9.85925925925926,
+      "grad_norm": 1.577006047602114,
+      "learning_rate": 1.764113379434774e-07,
+      "loss": 0.6471,
+      "step": 5324
+    },
+    {
+      "epoch": 9.86111111111111,
+      "grad_norm": 1.9236590728273362,
+      "learning_rate": 1.7640166457229656e-07,
+      "loss": 0.6571,
+      "step": 5325
+    },
+    {
+      "epoch": 9.862962962962962,
+      "grad_norm": 1.5246499120812889,
+      "learning_rate": 1.763919894833988e-07,
+      "loss": 0.6233,
+      "step": 5326
+    },
+    {
+      "epoch": 9.864814814814816,
+      "grad_norm": 1.8307438622079537,
+      "learning_rate": 1.763823126770017e-07,
+      "loss": 0.6496,
+      "step": 5327
+    },
+    {
+      "epoch": 9.866666666666667,
+      "grad_norm": 1.7739520451367077,
+      "learning_rate": 1.763726341533227e-07,
+      "loss": 0.6594,
+      "step": 5328
+    },
+    {
+      "epoch": 9.868518518518519,
+      "grad_norm": 1.3932948409685997,
+      "learning_rate": 1.7636295391257944e-07,
+      "loss": 0.6342,
+      "step": 5329
+    },
+    {
+      "epoch": 9.87037037037037,
+      "grad_norm": 1.497762520048915,
+      "learning_rate": 1.7635327195498962e-07,
+      "loss": 0.6341,
+      "step": 5330
+    },
+    {
+      "epoch": 9.872222222222222,
+      "grad_norm": 1.4184599020485498,
+      "learning_rate": 1.7634358828077088e-07,
+      "loss": 0.6494,
+      "step": 5331
+    },
+    {
+      "epoch": 9.874074074074073,
+      "grad_norm": 1.663299106333982,
+      "learning_rate": 1.763339028901409e-07,
+      "loss": 0.6449,
+      "step": 5332
+    },
+    {
+      "epoch": 9.875925925925927,
+      "grad_norm": 1.7195664615453008,
+      "learning_rate": 1.7632421578331748e-07,
+      "loss": 0.6277,
+      "step": 5333
+    },
+    {
+      "epoch": 9.877777777777778,
+      "grad_norm": 1.6271712449034368,
+      "learning_rate": 1.763145269605184e-07,
+      "loss": 0.6341,
+      "step": 5334
+    },
+    {
+      "epoch": 9.87962962962963,
+      "grad_norm": 1.4758485475156362,
+      "learning_rate": 1.7630483642196148e-07,
+      "loss": 0.6476,
+      "step": 5335
+    },
+    {
+      "epoch": 9.881481481481481,
+      "grad_norm": 1.6705914011332588,
+      "learning_rate": 1.7629514416786456e-07,
+      "loss": 0.7132,
+      "step": 5336
+    },
+    {
+      "epoch": 9.883333333333333,
+      "grad_norm": 1.6184902163359418,
+      "learning_rate": 1.7628545019844564e-07,
+      "loss": 0.6398,
+      "step": 5337
+    },
+    {
+      "epoch": 9.885185185185184,
+      "grad_norm": 1.4582106948595206,
+      "learning_rate": 1.762757545139226e-07,
+      "loss": 0.6158,
+      "step": 5338
+    },
+    {
+      "epoch": 9.887037037037038,
+      "grad_norm": 1.6877987475704237,
+      "learning_rate": 1.7626605711451342e-07,
+      "loss": 0.6593,
+      "step": 5339
+    },
+    {
+      "epoch": 9.88888888888889,
+      "grad_norm": 1.8651588210336025,
+      "learning_rate": 1.7625635800043616e-07,
+      "loss": 0.6497,
+      "step": 5340
+    },
+    {
+      "epoch": 9.89074074074074,
+      "grad_norm": 1.6340075736002309,
+      "learning_rate": 1.7624665717190883e-07,
+      "loss": 0.6583,
+      "step": 5341
+    },
+    {
+      "epoch": 9.892592592592592,
+      "grad_norm": 4.53238246778444,
+      "learning_rate": 1.762369546291496e-07,
+      "loss": 0.621,
+      "step": 5342
+    },
+    {
+      "epoch": 9.894444444444444,
+      "grad_norm": 1.6127496253864746,
+      "learning_rate": 1.7622725037237656e-07,
+      "loss": 0.6421,
+      "step": 5343
+    },
+    {
+      "epoch": 9.896296296296295,
+      "grad_norm": 1.394107095517648,
+      "learning_rate": 1.762175444018079e-07,
+      "loss": 0.5696,
+      "step": 5344
+    },
+    {
+      "epoch": 9.898148148148149,
+      "grad_norm": 1.5695701952721104,
+      "learning_rate": 1.7620783671766182e-07,
+      "loss": 0.665,
+      "step": 5345
+    },
+    {
+      "epoch": 9.9,
+      "grad_norm": 1.5916270431232358,
+      "learning_rate": 1.7619812732015663e-07,
+      "loss": 0.6309,
+      "step": 5346
+    },
+    {
+      "epoch": 9.901851851851852,
+      "grad_norm": 1.5027975880451432,
+      "learning_rate": 1.7618841620951054e-07,
+      "loss": 0.6487,
+      "step": 5347
+    },
+    {
+      "epoch": 9.903703703703703,
+      "grad_norm": 1.527701969204879,
+      "learning_rate": 1.7617870338594197e-07,
+      "loss": 0.6271,
+      "step": 5348
+    },
+    {
+      "epoch": 9.905555555555555,
+      "grad_norm": 1.4788529705550135,
+      "learning_rate": 1.761689888496692e-07,
+      "loss": 0.663,
+      "step": 5349
+    },
+    {
+      "epoch": 9.907407407407408,
+      "grad_norm": 1.724145661614413,
+      "learning_rate": 1.7615927260091076e-07,
+      "loss": 0.6544,
+      "step": 5350
+    },
+    {
+      "epoch": 9.90925925925926,
+      "grad_norm": 1.634704755671979,
+      "learning_rate": 1.76149554639885e-07,
+      "loss": 0.6182,
+      "step": 5351
+    },
+    {
+      "epoch": 9.911111111111111,
+      "grad_norm": 1.5429822541157592,
+      "learning_rate": 1.7613983496681041e-07,
+      "loss": 0.6391,
+      "step": 5352
+    },
+    {
+      "epoch": 9.912962962962963,
+      "grad_norm": 1.6321498855605314,
+      "learning_rate": 1.7613011358190554e-07,
+      "loss": 0.5995,
+      "step": 5353
+    },
+    {
+      "epoch": 9.914814814814815,
+      "grad_norm": 1.5008649767888658,
+      "learning_rate": 1.76120390485389e-07,
+      "loss": 0.6234,
+      "step": 5354
+    },
+    {
+      "epoch": 9.916666666666666,
+      "grad_norm": 3.3425832365486925,
+      "learning_rate": 1.761106656774793e-07,
+      "loss": 0.6615,
+      "step": 5355
+    },
+    {
+      "epoch": 9.918518518518518,
+      "grad_norm": 1.8997812292990122,
+      "learning_rate": 1.7610093915839512e-07,
+      "loss": 0.6791,
+      "step": 5356
+    },
+    {
+      "epoch": 9.920370370370371,
+      "grad_norm": 1.6848479657868247,
+      "learning_rate": 1.7609121092835513e-07,
+      "loss": 0.6052,
+      "step": 5357
+    },
+    {
+      "epoch": 9.922222222222222,
+      "grad_norm": 1.631792916206395,
+      "learning_rate": 1.760814809875781e-07,
+      "loss": 0.6534,
+      "step": 5358
+    },
+    {
+      "epoch": 9.924074074074074,
+      "grad_norm": 1.6183607352929394,
+      "learning_rate": 1.760717493362827e-07,
+      "loss": 0.6203,
+      "step": 5359
+    },
+    {
+      "epoch": 9.925925925925926,
+      "grad_norm": 1.6124872318896195,
+      "learning_rate": 1.760620159746878e-07,
+      "loss": 0.6402,
+      "step": 5360
+    },
+    {
+      "epoch": 9.927777777777777,
+      "grad_norm": 1.4943692515199258,
+      "learning_rate": 1.760522809030122e-07,
+      "loss": 0.6118,
+      "step": 5361
+    },
+    {
+      "epoch": 9.92962962962963,
+      "grad_norm": 1.777777824796479,
+      "learning_rate": 1.7604254412147474e-07,
+      "loss": 0.5952,
+      "step": 5362
+    },
+    {
+      "epoch": 9.931481481481482,
+      "grad_norm": 1.5167330317720857,
+      "learning_rate": 1.7603280563029437e-07,
+      "loss": 0.6375,
+      "step": 5363
+    },
+    {
+      "epoch": 9.933333333333334,
+      "grad_norm": 1.5923467687297315,
+      "learning_rate": 1.7602306542969003e-07,
+      "loss": 0.6558,
+      "step": 5364
+    },
+    {
+      "epoch": 9.935185185185185,
+      "grad_norm": 1.7759369249187156,
+      "learning_rate": 1.7601332351988072e-07,
+      "loss": 0.6209,
+      "step": 5365
+    },
+    {
+      "epoch": 9.937037037037037,
+      "grad_norm": 1.5618602035154268,
+      "learning_rate": 1.760035799010854e-07,
+      "loss": 0.5908,
+      "step": 5366
+    },
+    {
+      "epoch": 9.938888888888888,
+      "grad_norm": 1.4925859812994535,
+      "learning_rate": 1.7599383457352324e-07,
+      "loss": 0.6082,
+      "step": 5367
+    },
+    {
+      "epoch": 9.940740740740742,
+      "grad_norm": 1.4609494948842388,
+      "learning_rate": 1.7598408753741326e-07,
+      "loss": 0.5975,
+      "step": 5368
+    },
+    {
+      "epoch": 9.942592592592593,
+      "grad_norm": 1.4664494305030327,
+      "learning_rate": 1.7597433879297464e-07,
+      "loss": 0.6382,
+      "step": 5369
+    },
+    {
+      "epoch": 9.944444444444445,
+      "grad_norm": 1.7799782387798435,
+      "learning_rate": 1.759645883404265e-07,
+      "loss": 0.6191,
+      "step": 5370
+    },
+    {
+      "epoch": 9.946296296296296,
+      "grad_norm": 1.8259353449534388,
+      "learning_rate": 1.7595483617998813e-07,
+      "loss": 0.6341,
+      "step": 5371
+    },
+    {
+      "epoch": 9.948148148148148,
+      "grad_norm": 1.6952519825780845,
+      "learning_rate": 1.7594508231187873e-07,
+      "loss": 0.6485,
+      "step": 5372
+    },
+    {
+      "epoch": 9.95,
+      "grad_norm": 2.308087360435426,
+      "learning_rate": 1.7593532673631763e-07,
+      "loss": 0.6199,
+      "step": 5373
+    },
+    {
+      "epoch": 9.951851851851853,
+      "grad_norm": 1.874540168542581,
+      "learning_rate": 1.7592556945352413e-07,
+      "loss": 0.6188,
+      "step": 5374
+    },
+    {
+      "epoch": 9.953703703703704,
+      "grad_norm": 1.5275759330369276,
+      "learning_rate": 1.7591581046371763e-07,
+      "loss": 0.616,
+      "step": 5375
+    },
+    {
+      "epoch": 9.955555555555556,
+      "grad_norm": 1.616211587389306,
+      "learning_rate": 1.7590604976711754e-07,
+      "loss": 0.6447,
+      "step": 5376
+    },
+    {
+      "epoch": 9.957407407407407,
+      "grad_norm": 2.034094775213003,
+      "learning_rate": 1.7589628736394328e-07,
+      "loss": 0.5575,
+      "step": 5377
+    },
+    {
+      "epoch": 9.959259259259259,
+      "grad_norm": 1.6730283451438421,
+      "learning_rate": 1.7588652325441435e-07,
+      "loss": 0.6389,
+      "step": 5378
+    },
+    {
+      "epoch": 9.96111111111111,
+      "grad_norm": 1.6651520189680318,
+      "learning_rate": 1.7587675743875028e-07,
+      "loss": 0.6511,
+      "step": 5379
+    },
+    {
+      "epoch": 9.962962962962964,
+      "grad_norm": 1.6902678548932415,
+      "learning_rate": 1.7586698991717062e-07,
+      "loss": 0.6404,
+      "step": 5380
+    },
+    {
+      "epoch": 9.964814814814815,
+      "grad_norm": 1.9416466745535363,
+      "learning_rate": 1.7585722068989498e-07,
+      "loss": 0.6295,
+      "step": 5381
+    },
+    {
+      "epoch": 9.966666666666667,
+      "grad_norm": 1.995507662113881,
+      "learning_rate": 1.75847449757143e-07,
+      "loss": 0.6474,
+      "step": 5382
+    },
+    {
+      "epoch": 9.968518518518518,
+      "grad_norm": 1.6857697214829845,
+      "learning_rate": 1.7583767711913432e-07,
+      "loss": 0.6113,
+      "step": 5383
+    },
+    {
+      "epoch": 9.97037037037037,
+      "grad_norm": 1.7062100986463136,
+      "learning_rate": 1.7582790277608872e-07,
+      "loss": 0.6494,
+      "step": 5384
+    },
+    {
+      "epoch": 9.972222222222221,
+      "grad_norm": 1.651946582790938,
+      "learning_rate": 1.7581812672822588e-07,
+      "loss": 0.6111,
+      "step": 5385
+    },
+    {
+      "epoch": 9.974074074074075,
+      "grad_norm": 2.1863709450469253,
+      "learning_rate": 1.7580834897576568e-07,
+      "loss": 0.6293,
+      "step": 5386
+    },
+    {
+      "epoch": 9.975925925925926,
+      "grad_norm": 1.6077756891565993,
+      "learning_rate": 1.7579856951892788e-07,
+      "loss": 0.6269,
+      "step": 5387
+    },
+    {
+      "epoch": 9.977777777777778,
+      "grad_norm": 2.0861880204839744,
+      "learning_rate": 1.7578878835793237e-07,
+      "loss": 0.5877,
+      "step": 5388
+    },
+    {
+      "epoch": 9.97962962962963,
+      "grad_norm": 1.5138865055336816,
+      "learning_rate": 1.7577900549299906e-07,
+      "loss": 0.616,
+      "step": 5389
+    },
+    {
+      "epoch": 9.981481481481481,
+      "grad_norm": 1.724550504678632,
+      "learning_rate": 1.7576922092434789e-07,
+      "loss": 0.6512,
+      "step": 5390
+    },
+    {
+      "epoch": 9.983333333333333,
+      "grad_norm": 1.3473731699165248,
+      "learning_rate": 1.7575943465219884e-07,
+      "loss": 0.6763,
+      "step": 5391
+    },
+    {
+      "epoch": 9.985185185185186,
+      "grad_norm": 1.5434756588619276,
+      "learning_rate": 1.7574964667677197e-07,
+      "loss": 0.6308,
+      "step": 5392
+    },
+    {
+      "epoch": 9.987037037037037,
+      "grad_norm": 1.512171513091537,
+      "learning_rate": 1.757398569982873e-07,
+      "loss": 0.5866,
+      "step": 5393
+    },
+    {
+      "epoch": 9.988888888888889,
+      "grad_norm": 1.6069055533257244,
+      "learning_rate": 1.757300656169649e-07,
+      "loss": 0.6236,
+      "step": 5394
+    },
+    {
+      "epoch": 9.99074074074074,
+      "grad_norm": 1.5165261896035422,
+      "learning_rate": 1.75720272533025e-07,
+      "loss": 0.6161,
+      "step": 5395
+    },
+    {
+      "epoch": 9.992592592592592,
+      "grad_norm": 1.4634293951237676,
+      "learning_rate": 1.757104777466877e-07,
+      "loss": 0.6208,
+      "step": 5396
+    },
+    {
+      "epoch": 9.994444444444444,
+      "grad_norm": 1.560741807941215,
+      "learning_rate": 1.757006812581732e-07,
+      "loss": 0.6292,
+      "step": 5397
+    },
+    {
+      "epoch": 9.996296296296297,
+      "grad_norm": 1.7143380232528396,
+      "learning_rate": 1.7569088306770184e-07,
+      "loss": 0.6235,
+      "step": 5398
+    },
+    {
+      "epoch": 9.998148148148148,
+      "grad_norm": 1.5106003236497454,
+      "learning_rate": 1.7568108317549385e-07,
+      "loss": 0.6211,
+      "step": 5399
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 1.437281630754344,
+      "learning_rate": 1.7567128158176952e-07,
+      "loss": 0.6278,
+      "step": 5400
+    },
+    {
+      "epoch": 10.001851851851852,
+      "grad_norm": 1.7098573916799773,
+      "learning_rate": 1.7566147828674928e-07,
+      "loss": 0.641,
+      "step": 5401
+    },
+    {
+      "epoch": 10.003703703703703,
+      "grad_norm": 1.762501013213143,
+      "learning_rate": 1.756516732906535e-07,
+      "loss": 0.6129,
+      "step": 5402
+    },
+    {
+      "epoch": 10.005555555555556,
+      "grad_norm": 1.8107878864228215,
+      "learning_rate": 1.7564186659370267e-07,
+      "loss": 0.6388,
+      "step": 5403
+    },
+    {
+      "epoch": 10.007407407407408,
+      "grad_norm": 1.8437931273681092,
+      "learning_rate": 1.756320581961172e-07,
+      "loss": 0.6488,
+      "step": 5404
+    },
+    {
+      "epoch": 10.00925925925926,
+      "grad_norm": 1.399536817733659,
+      "learning_rate": 1.7562224809811768e-07,
+      "loss": 0.6201,
+      "step": 5405
+    },
+    {
+      "epoch": 10.011111111111111,
+      "grad_norm": 1.3749930273296522,
+      "learning_rate": 1.7561243629992464e-07,
+      "loss": 0.6225,
+      "step": 5406
+    },
+    {
+      "epoch": 10.012962962962963,
+      "grad_norm": 1.6733135468429279,
+      "learning_rate": 1.7560262280175868e-07,
+      "loss": 0.6356,
+      "step": 5407
+    },
+    {
+      "epoch": 10.014814814814814,
+      "grad_norm": 1.3853374438282582,
+      "learning_rate": 1.7559280760384038e-07,
+      "loss": 0.6259,
+      "step": 5408
+    },
+    {
+      "epoch": 10.016666666666667,
+      "grad_norm": 1.7394206880810161,
+      "learning_rate": 1.7558299070639052e-07,
+      "loss": 0.6159,
+      "step": 5409
+    },
+    {
+      "epoch": 10.018518518518519,
+      "grad_norm": 1.9959962114385614,
+      "learning_rate": 1.755731721096297e-07,
+      "loss": 0.6591,
+      "step": 5410
+    },
+    {
+      "epoch": 10.02037037037037,
+      "grad_norm": 3.1104791556069147,
+      "learning_rate": 1.7556335181377872e-07,
+      "loss": 0.6816,
+      "step": 5411
+    },
+    {
+      "epoch": 10.022222222222222,
+      "grad_norm": 1.569045766373436,
+      "learning_rate": 1.7555352981905838e-07,
+      "loss": 0.6263,
+      "step": 5412
+    },
+    {
+      "epoch": 10.024074074074074,
+      "grad_norm": 1.574548976842346,
+      "learning_rate": 1.7554370612568946e-07,
+      "loss": 0.6361,
+      "step": 5413
+    },
+    {
+      "epoch": 10.025925925925925,
+      "grad_norm": 1.5966426102196793,
+      "learning_rate": 1.755338807338929e-07,
+      "loss": 0.6626,
+      "step": 5414
+    },
+    {
+      "epoch": 10.027777777777779,
+      "grad_norm": 1.6311327368399573,
+      "learning_rate": 1.7552405364388952e-07,
+      "loss": 0.6615,
+      "step": 5415
+    },
+    {
+      "epoch": 10.02962962962963,
+      "grad_norm": 1.4758553669940013,
+      "learning_rate": 1.755142248559003e-07,
+      "loss": 0.5891,
+      "step": 5416
+    },
+    {
+      "epoch": 10.031481481481482,
+      "grad_norm": 1.6698168419392883,
+      "learning_rate": 1.7550439437014623e-07,
+      "loss": 0.6646,
+      "step": 5417
+    },
+    {
+      "epoch": 10.033333333333333,
+      "grad_norm": 1.6827280407422223,
+      "learning_rate": 1.7549456218684832e-07,
+      "loss": 0.6154,
+      "step": 5418
+    },
+    {
+      "epoch": 10.035185185185185,
+      "grad_norm": 1.7781165273835078,
+      "learning_rate": 1.7548472830622756e-07,
+      "loss": 0.6568,
+      "step": 5419
+    },
+    {
+      "epoch": 10.037037037037036,
+      "grad_norm": 1.5133939721052163,
+      "learning_rate": 1.7547489272850512e-07,
+      "loss": 0.6142,
+      "step": 5420
+    },
+    {
+      "epoch": 10.03888888888889,
+      "grad_norm": 1.806437339783129,
+      "learning_rate": 1.7546505545390212e-07,
+      "loss": 0.6297,
+      "step": 5421
+    },
+    {
+      "epoch": 10.040740740740741,
+      "grad_norm": 1.5629938681275963,
+      "learning_rate": 1.7545521648263969e-07,
+      "loss": 0.6022,
+      "step": 5422
+    },
+    {
+      "epoch": 10.042592592592593,
+      "grad_norm": 1.9452022259578916,
+      "learning_rate": 1.7544537581493906e-07,
+      "loss": 0.5931,
+      "step": 5423
+    },
+    {
+      "epoch": 10.044444444444444,
+      "grad_norm": 1.5695491944405213,
+      "learning_rate": 1.754355334510215e-07,
+      "loss": 0.629,
+      "step": 5424
+    },
+    {
+      "epoch": 10.046296296296296,
+      "grad_norm": 1.5529083569681585,
+      "learning_rate": 1.7542568939110823e-07,
+      "loss": 0.62,
+      "step": 5425
+    },
+    {
+      "epoch": 10.048148148148147,
+      "grad_norm": 1.5028566603604283,
+      "learning_rate": 1.7541584363542064e-07,
+      "loss": 0.6398,
+      "step": 5426
+    },
+    {
+      "epoch": 10.05,
+      "grad_norm": 1.7213447199997993,
+      "learning_rate": 1.7540599618418005e-07,
+      "loss": 0.6043,
+      "step": 5427
+    },
+    {
+      "epoch": 10.051851851851852,
+      "grad_norm": 1.5159099790962234,
+      "learning_rate": 1.7539614703760786e-07,
+      "loss": 0.6235,
+      "step": 5428
+    },
+    {
+      "epoch": 10.053703703703704,
+      "grad_norm": 1.528433817159488,
+      "learning_rate": 1.7538629619592552e-07,
+      "loss": 0.6684,
+      "step": 5429
+    },
+    {
+      "epoch": 10.055555555555555,
+      "grad_norm": 1.5031113085458192,
+      "learning_rate": 1.753764436593545e-07,
+      "loss": 0.6907,
+      "step": 5430
+    },
+    {
+      "epoch": 10.057407407407407,
+      "grad_norm": 1.8304059969812667,
+      "learning_rate": 1.7536658942811632e-07,
+      "loss": 0.624,
+      "step": 5431
+    },
+    {
+      "epoch": 10.059259259259258,
+      "grad_norm": 1.606070419953307,
+      "learning_rate": 1.7535673350243245e-07,
+      "loss": 0.6269,
+      "step": 5432
+    },
+    {
+      "epoch": 10.061111111111112,
+      "grad_norm": 1.4545851842975484,
+      "learning_rate": 1.7534687588252462e-07,
+      "loss": 0.6631,
+      "step": 5433
+    },
+    {
+      "epoch": 10.062962962962963,
+      "grad_norm": 1.5851564348222162,
+      "learning_rate": 1.7533701656861434e-07,
+      "loss": 0.6289,
+      "step": 5434
+    },
+    {
+      "epoch": 10.064814814814815,
+      "grad_norm": 1.8128035180214417,
+      "learning_rate": 1.7532715556092335e-07,
+      "loss": 0.6251,
+      "step": 5435
+    },
+    {
+      "epoch": 10.066666666666666,
+      "grad_norm": 1.7227149687478553,
+      "learning_rate": 1.7531729285967328e-07,
+      "loss": 0.6262,
+      "step": 5436
+    },
+    {
+      "epoch": 10.068518518518518,
+      "grad_norm": 1.6464925960931711,
+      "learning_rate": 1.7530742846508593e-07,
+      "loss": 0.6109,
+      "step": 5437
+    },
+    {
+      "epoch": 10.07037037037037,
+      "grad_norm": 1.3926493573294632,
+      "learning_rate": 1.7529756237738305e-07,
+      "loss": 0.6139,
+      "step": 5438
+    },
+    {
+      "epoch": 10.072222222222223,
+      "grad_norm": 3.0896580145107557,
+      "learning_rate": 1.7528769459678646e-07,
+      "loss": 0.591,
+      "step": 5439
+    },
+    {
+      "epoch": 10.074074074074074,
+      "grad_norm": 1.6418635059843978,
+      "learning_rate": 1.7527782512351804e-07,
+      "loss": 0.6718,
+      "step": 5440
+    },
+    {
+      "epoch": 10.075925925925926,
+      "grad_norm": 1.380776230271832,
+      "learning_rate": 1.752679539577996e-07,
+      "loss": 0.6071,
+      "step": 5441
+    },
+    {
+      "epoch": 10.077777777777778,
+      "grad_norm": 1.5018014045666626,
+      "learning_rate": 1.752580810998532e-07,
+      "loss": 0.5957,
+      "step": 5442
+    },
+    {
+      "epoch": 10.079629629629629,
+      "grad_norm": 1.5644677745854392,
+      "learning_rate": 1.752482065499007e-07,
+      "loss": 0.6367,
+      "step": 5443
+    },
+    {
+      "epoch": 10.081481481481482,
+      "grad_norm": 1.589005607459597,
+      "learning_rate": 1.7523833030816414e-07,
+      "loss": 0.6243,
+      "step": 5444
+    },
+    {
+      "epoch": 10.083333333333334,
+      "grad_norm": 1.4444195485919817,
+      "learning_rate": 1.752284523748656e-07,
+      "loss": 0.6538,
+      "step": 5445
+    },
+    {
+      "epoch": 10.085185185185185,
+      "grad_norm": 1.5359609767188398,
+      "learning_rate": 1.7521857275022712e-07,
+      "loss": 0.5972,
+      "step": 5446
+    },
+    {
+      "epoch": 10.087037037037037,
+      "grad_norm": 1.410038322340428,
+      "learning_rate": 1.7520869143447084e-07,
+      "loss": 0.6285,
+      "step": 5447
+    },
+    {
+      "epoch": 10.088888888888889,
+      "grad_norm": 1.7825376721065744,
+      "learning_rate": 1.7519880842781892e-07,
+      "loss": 0.6047,
+      "step": 5448
+    },
+    {
+      "epoch": 10.09074074074074,
+      "grad_norm": 1.7384020488052006,
+      "learning_rate": 1.7518892373049353e-07,
+      "loss": 0.644,
+      "step": 5449
+    },
+    {
+      "epoch": 10.092592592592593,
+      "grad_norm": 1.6543661230294295,
+      "learning_rate": 1.7517903734271693e-07,
+      "loss": 0.6267,
+      "step": 5450
+    },
+    {
+      "epoch": 10.094444444444445,
+      "grad_norm": 1.539215619690721,
+      "learning_rate": 1.7516914926471138e-07,
+      "loss": 0.5882,
+      "step": 5451
+    },
+    {
+      "epoch": 10.096296296296297,
+      "grad_norm": 1.649971961981206,
+      "learning_rate": 1.7515925949669922e-07,
+      "loss": 0.6527,
+      "step": 5452
+    },
+    {
+      "epoch": 10.098148148148148,
+      "grad_norm": 1.3220548537596515,
+      "learning_rate": 1.7514936803890275e-07,
+      "loss": 0.6301,
+      "step": 5453
+    },
+    {
+      "epoch": 10.1,
+      "grad_norm": 1.5495174583103943,
+      "learning_rate": 1.751394748915444e-07,
+      "loss": 0.619,
+      "step": 5454
+    },
+    {
+      "epoch": 10.101851851851851,
+      "grad_norm": 1.531033106425347,
+      "learning_rate": 1.7512958005484657e-07,
+      "loss": 0.6549,
+      "step": 5455
+    },
+    {
+      "epoch": 10.103703703703705,
+      "grad_norm": 1.559076341876545,
+      "learning_rate": 1.7511968352903175e-07,
+      "loss": 0.6354,
+      "step": 5456
+    },
+    {
+      "epoch": 10.105555555555556,
+      "grad_norm": 1.768673995635982,
+      "learning_rate": 1.751097853143224e-07,
+      "loss": 0.6283,
+      "step": 5457
+    },
+    {
+      "epoch": 10.107407407407408,
+      "grad_norm": 1.45193003555617,
+      "learning_rate": 1.7509988541094108e-07,
+      "loss": 0.6338,
+      "step": 5458
+    },
+    {
+      "epoch": 10.10925925925926,
+      "grad_norm": 1.682710732335579,
+      "learning_rate": 1.750899838191104e-07,
+      "loss": 0.6195,
+      "step": 5459
+    },
+    {
+      "epoch": 10.11111111111111,
+      "grad_norm": 1.5266955215137943,
+      "learning_rate": 1.750800805390529e-07,
+      "loss": 0.6191,
+      "step": 5460
+    },
+    {
+      "epoch": 10.112962962962962,
+      "grad_norm": 1.6048594704958947,
+      "learning_rate": 1.750701755709913e-07,
+      "loss": 0.6392,
+      "step": 5461
+    },
+    {
+      "epoch": 10.114814814814816,
+      "grad_norm": 1.6260188220193799,
+      "learning_rate": 1.7506026891514827e-07,
+      "loss": 0.6452,
+      "step": 5462
+    },
+    {
+      "epoch": 10.116666666666667,
+      "grad_norm": 1.7308424331945567,
+      "learning_rate": 1.750503605717465e-07,
+      "loss": 0.5788,
+      "step": 5463
+    },
+    {
+      "epoch": 10.118518518518519,
+      "grad_norm": 1.956225410862961,
+      "learning_rate": 1.7504045054100884e-07,
+      "loss": 0.6594,
+      "step": 5464
+    },
+    {
+      "epoch": 10.12037037037037,
+      "grad_norm": 1.5532658265287898,
+      "learning_rate": 1.7503053882315802e-07,
+      "loss": 0.6293,
+      "step": 5465
+    },
+    {
+      "epoch": 10.122222222222222,
+      "grad_norm": 3.412774904691906,
+      "learning_rate": 1.7502062541841692e-07,
+      "loss": 0.6732,
+      "step": 5466
+    },
+    {
+      "epoch": 10.124074074074073,
+      "grad_norm": 1.5537896442815855,
+      "learning_rate": 1.7501071032700838e-07,
+      "loss": 0.6623,
+      "step": 5467
+    },
+    {
+      "epoch": 10.125925925925927,
+      "grad_norm": 1.4694556242385732,
+      "learning_rate": 1.7500079354915535e-07,
+      "loss": 0.6628,
+      "step": 5468
+    },
+    {
+      "epoch": 10.127777777777778,
+      "grad_norm": 1.6229229898959951,
+      "learning_rate": 1.749908750850808e-07,
+      "loss": 0.604,
+      "step": 5469
+    },
+    {
+      "epoch": 10.12962962962963,
+      "grad_norm": 1.792271862927219,
+      "learning_rate": 1.7498095493500774e-07,
+      "loss": 0.6358,
+      "step": 5470
+    },
+    {
+      "epoch": 10.131481481481481,
+      "grad_norm": 1.766618300118669,
+      "learning_rate": 1.749710330991591e-07,
+      "loss": 0.6485,
+      "step": 5471
+    },
+    {
+      "epoch": 10.133333333333333,
+      "grad_norm": 1.978909843463822,
+      "learning_rate": 1.7496110957775808e-07,
+      "loss": 0.6566,
+      "step": 5472
+    },
+    {
+      "epoch": 10.135185185185184,
+      "grad_norm": 1.5660534328838762,
+      "learning_rate": 1.7495118437102766e-07,
+      "loss": 0.6437,
+      "step": 5473
+    },
+    {
+      "epoch": 10.137037037037038,
+      "grad_norm": 1.8229889221917583,
+      "learning_rate": 1.7494125747919109e-07,
+      "loss": 0.6508,
+      "step": 5474
+    },
+    {
+      "epoch": 10.13888888888889,
+      "grad_norm": 1.6391646748274622,
+      "learning_rate": 1.749313289024715e-07,
+      "loss": 0.6481,
+      "step": 5475
+    },
+    {
+      "epoch": 10.14074074074074,
+      "grad_norm": 2.029009025041515,
+      "learning_rate": 1.7492139864109216e-07,
+      "loss": 0.6257,
+      "step": 5476
+    },
+    {
+      "epoch": 10.142592592592592,
+      "grad_norm": 1.3708176002505783,
+      "learning_rate": 1.7491146669527624e-07,
+      "loss": 0.6122,
+      "step": 5477
+    },
+    {
+      "epoch": 10.144444444444444,
+      "grad_norm": 1.5143378257760454,
+      "learning_rate": 1.7490153306524712e-07,
+      "loss": 0.6557,
+      "step": 5478
+    },
+    {
+      "epoch": 10.146296296296295,
+      "grad_norm": 1.7993450592830094,
+      "learning_rate": 1.7489159775122812e-07,
+      "loss": 0.6053,
+      "step": 5479
+    },
+    {
+      "epoch": 10.148148148148149,
+      "grad_norm": 1.824416286209922,
+      "learning_rate": 1.7488166075344258e-07,
+      "loss": 0.6475,
+      "step": 5480
+    },
+    {
+      "epoch": 10.15,
+      "grad_norm": 1.5359348155297259,
+      "learning_rate": 1.7487172207211393e-07,
+      "loss": 0.6438,
+      "step": 5481
+    },
+    {
+      "epoch": 10.151851851851852,
+      "grad_norm": 2.181821129056764,
+      "learning_rate": 1.7486178170746563e-07,
+      "loss": 0.6386,
+      "step": 5482
+    },
+    {
+      "epoch": 10.153703703703703,
+      "grad_norm": 1.4890011471993854,
+      "learning_rate": 1.7485183965972117e-07,
+      "loss": 0.6863,
+      "step": 5483
+    },
+    {
+      "epoch": 10.155555555555555,
+      "grad_norm": 1.8035843706339572,
+      "learning_rate": 1.7484189592910402e-07,
+      "loss": 0.6343,
+      "step": 5484
+    },
+    {
+      "epoch": 10.157407407407407,
+      "grad_norm": 1.4596870801261845,
+      "learning_rate": 1.7483195051583781e-07,
+      "loss": 0.6442,
+      "step": 5485
+    },
+    {
+      "epoch": 10.15925925925926,
+      "grad_norm": 1.6045386504073424,
+      "learning_rate": 1.748220034201461e-07,
+      "loss": 0.6483,
+      "step": 5486
+    },
+    {
+      "epoch": 10.161111111111111,
+      "grad_norm": 1.443354283286236,
+      "learning_rate": 1.7481205464225254e-07,
+      "loss": 0.6324,
+      "step": 5487
+    },
+    {
+      "epoch": 10.162962962962963,
+      "grad_norm": 2.5576780325241186,
+      "learning_rate": 1.7480210418238084e-07,
+      "loss": 0.6107,
+      "step": 5488
+    },
+    {
+      "epoch": 10.164814814814815,
+      "grad_norm": 1.6782656832253358,
+      "learning_rate": 1.7479215204075466e-07,
+      "loss": 0.6011,
+      "step": 5489
+    },
+    {
+      "epoch": 10.166666666666666,
+      "grad_norm": 1.497783018464314,
+      "learning_rate": 1.7478219821759775e-07,
+      "loss": 0.5917,
+      "step": 5490
+    },
+    {
+      "epoch": 10.168518518518518,
+      "grad_norm": 1.4028401969888447,
+      "learning_rate": 1.747722427131339e-07,
+      "loss": 0.5957,
+      "step": 5491
+    },
+    {
+      "epoch": 10.170370370370371,
+      "grad_norm": 1.7500533193866945,
+      "learning_rate": 1.74762285527587e-07,
+      "loss": 0.6411,
+      "step": 5492
+    },
+    {
+      "epoch": 10.172222222222222,
+      "grad_norm": 2.536000570561538,
+      "learning_rate": 1.7475232666118086e-07,
+      "loss": 0.6671,
+      "step": 5493
+    },
+    {
+      "epoch": 10.174074074074074,
+      "grad_norm": 1.6274632337016584,
+      "learning_rate": 1.747423661141394e-07,
+      "loss": 0.6275,
+      "step": 5494
+    },
+    {
+      "epoch": 10.175925925925926,
+      "grad_norm": 1.7027974976720113,
+      "learning_rate": 1.7473240388668654e-07,
+      "loss": 0.6125,
+      "step": 5495
+    },
+    {
+      "epoch": 10.177777777777777,
+      "grad_norm": 1.6481532436486679,
+      "learning_rate": 1.7472243997904626e-07,
+      "loss": 0.6063,
+      "step": 5496
+    },
+    {
+      "epoch": 10.17962962962963,
+      "grad_norm": 1.9632694961044845,
+      "learning_rate": 1.747124743914426e-07,
+      "loss": 0.6325,
+      "step": 5497
+    },
+    {
+      "epoch": 10.181481481481482,
+      "grad_norm": 1.6392790247435027,
+      "learning_rate": 1.747025071240996e-07,
+      "loss": 0.6108,
+      "step": 5498
+    },
+    {
+      "epoch": 10.183333333333334,
+      "grad_norm": 1.8209886790195473,
+      "learning_rate": 1.7469253817724133e-07,
+      "loss": 0.6047,
+      "step": 5499
+    },
+    {
+      "epoch": 10.185185185185185,
+      "grad_norm": 1.904186901142017,
+      "learning_rate": 1.7468256755109196e-07,
+      "loss": 0.5867,
+      "step": 5500
+    },
+    {
+      "epoch": 10.187037037037037,
+      "grad_norm": 1.5541576103066044,
+      "learning_rate": 1.7467259524587564e-07,
+      "loss": 0.6415,
+      "step": 5501
+    },
+    {
+      "epoch": 10.188888888888888,
+      "grad_norm": 2.507280815088853,
+      "learning_rate": 1.7466262126181657e-07,
+      "loss": 0.5996,
+      "step": 5502
+    },
+    {
+      "epoch": 10.190740740740742,
+      "grad_norm": 1.546303851731707,
+      "learning_rate": 1.7465264559913897e-07,
+      "loss": 0.6551,
+      "step": 5503
+    },
+    {
+      "epoch": 10.192592592592593,
+      "grad_norm": 1.5705319691122943,
+      "learning_rate": 1.7464266825806717e-07,
+      "loss": 0.6469,
+      "step": 5504
+    },
+    {
+      "epoch": 10.194444444444445,
+      "grad_norm": 1.5193132845643262,
+      "learning_rate": 1.7463268923882545e-07,
+      "loss": 0.6011,
+      "step": 5505
+    },
+    {
+      "epoch": 10.196296296296296,
+      "grad_norm": 1.6649974369870668,
+      "learning_rate": 1.746227085416382e-07,
+      "loss": 0.6289,
+      "step": 5506
+    },
+    {
+      "epoch": 10.198148148148148,
+      "grad_norm": 1.6219604889928134,
+      "learning_rate": 1.7461272616672974e-07,
+      "loss": 0.5924,
+      "step": 5507
+    },
+    {
+      "epoch": 10.2,
+      "grad_norm": 1.319202947556063,
+      "learning_rate": 1.746027421143246e-07,
+      "loss": 0.6358,
+      "step": 5508
+    },
+    {
+      "epoch": 10.201851851851853,
+      "grad_norm": 1.39331139402287,
+      "learning_rate": 1.7459275638464716e-07,
+      "loss": 0.591,
+      "step": 5509
+    },
+    {
+      "epoch": 10.203703703703704,
+      "grad_norm": 1.581059424479091,
+      "learning_rate": 1.74582768977922e-07,
+      "loss": 0.5983,
+      "step": 5510
+    },
+    {
+      "epoch": 10.205555555555556,
+      "grad_norm": 1.7827362833276263,
+      "learning_rate": 1.745727798943736e-07,
+      "loss": 0.6486,
+      "step": 5511
+    },
+    {
+      "epoch": 10.207407407407407,
+      "grad_norm": 1.676344083964432,
+      "learning_rate": 1.7456278913422659e-07,
+      "loss": 0.6447,
+      "step": 5512
+    },
+    {
+      "epoch": 10.209259259259259,
+      "grad_norm": 1.7267224663675524,
+      "learning_rate": 1.7455279669770558e-07,
+      "loss": 0.6147,
+      "step": 5513
+    },
+    {
+      "epoch": 10.21111111111111,
+      "grad_norm": 1.4654553691012164,
+      "learning_rate": 1.745428025850352e-07,
+      "loss": 0.6496,
+      "step": 5514
+    },
+    {
+      "epoch": 10.212962962962964,
+      "grad_norm": 2.2067321695075512,
+      "learning_rate": 1.7453280679644018e-07,
+      "loss": 0.6167,
+      "step": 5515
+    },
+    {
+      "epoch": 10.214814814814815,
+      "grad_norm": 1.5442336004459,
+      "learning_rate": 1.7452280933214523e-07,
+      "loss": 0.6437,
+      "step": 5516
+    },
+    {
+      "epoch": 10.216666666666667,
+      "grad_norm": 1.6302117104545533,
+      "learning_rate": 1.745128101923751e-07,
+      "loss": 0.6386,
+      "step": 5517
+    },
+    {
+      "epoch": 10.218518518518518,
+      "grad_norm": 1.863647711142828,
+      "learning_rate": 1.7450280937735464e-07,
+      "loss": 0.6131,
+      "step": 5518
+    },
+    {
+      "epoch": 10.22037037037037,
+      "grad_norm": 1.6803911691655162,
+      "learning_rate": 1.7449280688730869e-07,
+      "loss": 0.6177,
+      "step": 5519
+    },
+    {
+      "epoch": 10.222222222222221,
+      "grad_norm": 1.4533503352272066,
+      "learning_rate": 1.744828027224621e-07,
+      "loss": 0.6401,
+      "step": 5520
+    },
+    {
+      "epoch": 10.224074074074075,
+      "grad_norm": 4.717111465419407,
+      "learning_rate": 1.7447279688303982e-07,
+      "loss": 0.6392,
+      "step": 5521
+    },
+    {
+      "epoch": 10.225925925925926,
+      "grad_norm": 1.4044209547715456,
+      "learning_rate": 1.7446278936926681e-07,
+      "loss": 0.6368,
+      "step": 5522
+    },
+    {
+      "epoch": 10.227777777777778,
+      "grad_norm": 1.8145224553152064,
+      "learning_rate": 1.7445278018136805e-07,
+      "loss": 0.6355,
+      "step": 5523
+    },
+    {
+      "epoch": 10.22962962962963,
+      "grad_norm": 1.421038230452516,
+      "learning_rate": 1.744427693195686e-07,
+      "loss": 0.611,
+      "step": 5524
+    },
+    {
+      "epoch": 10.231481481481481,
+      "grad_norm": 1.3321516046857718,
+      "learning_rate": 1.7443275678409348e-07,
+      "loss": 0.5833,
+      "step": 5525
+    },
+    {
+      "epoch": 10.233333333333333,
+      "grad_norm": 1.4950411109701227,
+      "learning_rate": 1.7442274257516781e-07,
+      "loss": 0.5672,
+      "step": 5526
+    },
+    {
+      "epoch": 10.235185185185186,
+      "grad_norm": 1.5837934135649152,
+      "learning_rate": 1.7441272669301684e-07,
+      "loss": 0.6305,
+      "step": 5527
+    },
+    {
+      "epoch": 10.237037037037037,
+      "grad_norm": 1.5805286526469449,
+      "learning_rate": 1.744027091378656e-07,
+      "loss": 0.6363,
+      "step": 5528
+    },
+    {
+      "epoch": 10.238888888888889,
+      "grad_norm": 2.5909401288928353,
+      "learning_rate": 1.7439268990993944e-07,
+      "loss": 0.646,
+      "step": 5529
+    },
+    {
+      "epoch": 10.24074074074074,
+      "grad_norm": 1.3892380470013994,
+      "learning_rate": 1.7438266900946353e-07,
+      "loss": 0.6475,
+      "step": 5530
+    },
+    {
+      "epoch": 10.242592592592592,
+      "grad_norm": 1.7218547567376672,
+      "learning_rate": 1.7437264643666317e-07,
+      "loss": 0.6554,
+      "step": 5531
+    },
+    {
+      "epoch": 10.244444444444444,
+      "grad_norm": 1.9632540575265303,
+      "learning_rate": 1.7436262219176378e-07,
+      "loss": 0.649,
+      "step": 5532
+    },
+    {
+      "epoch": 10.246296296296297,
+      "grad_norm": 1.862987607175111,
+      "learning_rate": 1.7435259627499064e-07,
+      "loss": 0.5927,
+      "step": 5533
+    },
+    {
+      "epoch": 10.248148148148148,
+      "grad_norm": 1.5017592493813212,
+      "learning_rate": 1.7434256868656924e-07,
+      "loss": 0.6283,
+      "step": 5534
+    },
+    {
+      "epoch": 10.25,
+      "grad_norm": 1.7186720218181681,
+      "learning_rate": 1.7433253942672495e-07,
+      "loss": 0.6077,
+      "step": 5535
+    },
+    {
+      "epoch": 10.251851851851852,
+      "grad_norm": 1.4885062582660054,
+      "learning_rate": 1.743225084956833e-07,
+      "loss": 0.6157,
+      "step": 5536
+    },
+    {
+      "epoch": 10.253703703703703,
+      "grad_norm": 1.80912790417594,
+      "learning_rate": 1.743124758936698e-07,
+      "loss": 0.6485,
+      "step": 5537
+    },
+    {
+      "epoch": 10.255555555555556,
+      "grad_norm": 1.3966421337815214,
+      "learning_rate": 1.7430244162091002e-07,
+      "loss": 0.6237,
+      "step": 5538
+    },
+    {
+      "epoch": 10.257407407407408,
+      "grad_norm": 1.5076718384621615,
+      "learning_rate": 1.7429240567762953e-07,
+      "loss": 0.6205,
+      "step": 5539
+    },
+    {
+      "epoch": 10.25925925925926,
+      "grad_norm": 1.6343518892967013,
+      "learning_rate": 1.74282368064054e-07,
+      "loss": 0.6438,
+      "step": 5540
+    },
+    {
+      "epoch": 10.261111111111111,
+      "grad_norm": 1.6166409071831247,
+      "learning_rate": 1.742723287804091e-07,
+      "loss": 0.6507,
+      "step": 5541
+    },
+    {
+      "epoch": 10.262962962962963,
+      "grad_norm": 1.5338285153498954,
+      "learning_rate": 1.742622878269205e-07,
+      "loss": 0.6277,
+      "step": 5542
+    },
+    {
+      "epoch": 10.264814814814814,
+      "grad_norm": 1.5009842496230708,
+      "learning_rate": 1.7425224520381397e-07,
+      "loss": 0.6639,
+      "step": 5543
+    },
+    {
+      "epoch": 10.266666666666667,
+      "grad_norm": 1.5348752243529402,
+      "learning_rate": 1.7424220091131535e-07,
+      "loss": 0.661,
+      "step": 5544
+    },
+    {
+      "epoch": 10.268518518518519,
+      "grad_norm": 3.1096101725094165,
+      "learning_rate": 1.7423215494965038e-07,
+      "loss": 0.6,
+      "step": 5545
+    },
+    {
+      "epoch": 10.27037037037037,
+      "grad_norm": 1.582307403119614,
+      "learning_rate": 1.7422210731904497e-07,
+      "loss": 0.6297,
+      "step": 5546
+    },
+    {
+      "epoch": 10.272222222222222,
+      "grad_norm": 1.589105049232147,
+      "learning_rate": 1.7421205801972499e-07,
+      "loss": 0.7048,
+      "step": 5547
+    },
+    {
+      "epoch": 10.274074074074074,
+      "grad_norm": 1.5981124117530174,
+      "learning_rate": 1.7420200705191638e-07,
+      "loss": 0.6651,
+      "step": 5548
+    },
+    {
+      "epoch": 10.275925925925925,
+      "grad_norm": 1.8703227544827403,
+      "learning_rate": 1.7419195441584513e-07,
+      "loss": 0.6231,
+      "step": 5549
+    },
+    {
+      "epoch": 10.277777777777779,
+      "grad_norm": 1.472798494260182,
+      "learning_rate": 1.7418190011173728e-07,
+      "loss": 0.6568,
+      "step": 5550
+    },
+    {
+      "epoch": 10.27962962962963,
+      "grad_norm": 4.887250444622802,
+      "learning_rate": 1.7417184413981877e-07,
+      "loss": 0.6312,
+      "step": 5551
+    },
+    {
+      "epoch": 10.281481481481482,
+      "grad_norm": 1.4889889626658819,
+      "learning_rate": 1.741617865003158e-07,
+      "loss": 0.64,
+      "step": 5552
+    },
+    {
+      "epoch": 10.283333333333333,
+      "grad_norm": 1.686091022171425,
+      "learning_rate": 1.7415172719345446e-07,
+      "loss": 0.6263,
+      "step": 5553
+    },
+    {
+      "epoch": 10.285185185185185,
+      "grad_norm": 1.664668886009631,
+      "learning_rate": 1.741416662194609e-07,
+      "loss": 0.6431,
+      "step": 5554
+    },
+    {
+      "epoch": 10.287037037037036,
+      "grad_norm": 1.7054550805628932,
+      "learning_rate": 1.741316035785613e-07,
+      "loss": 0.6317,
+      "step": 5555
+    },
+    {
+      "epoch": 10.28888888888889,
+      "grad_norm": 1.4083990332617717,
+      "learning_rate": 1.741215392709819e-07,
+      "loss": 0.6087,
+      "step": 5556
+    },
+    {
+      "epoch": 10.290740740740741,
+      "grad_norm": 2.4208191751102075,
+      "learning_rate": 1.7411147329694898e-07,
+      "loss": 0.6478,
+      "step": 5557
+    },
+    {
+      "epoch": 10.292592592592593,
+      "grad_norm": 1.7109823064464107,
+      "learning_rate": 1.741014056566889e-07,
+      "loss": 0.6182,
+      "step": 5558
+    },
+    {
+      "epoch": 10.294444444444444,
+      "grad_norm": 1.5950325601201951,
+      "learning_rate": 1.7409133635042795e-07,
+      "loss": 0.656,
+      "step": 5559
+    },
+    {
+      "epoch": 10.296296296296296,
+      "grad_norm": 1.4501350924335195,
+      "learning_rate": 1.740812653783925e-07,
+      "loss": 0.6322,
+      "step": 5560
+    },
+    {
+      "epoch": 10.298148148148147,
+      "grad_norm": 1.6473791712265817,
+      "learning_rate": 1.7407119274080904e-07,
+      "loss": 0.6046,
+      "step": 5561
+    },
+    {
+      "epoch": 10.3,
+      "grad_norm": 1.568230333855012,
+      "learning_rate": 1.7406111843790398e-07,
+      "loss": 0.639,
+      "step": 5562
+    },
+    {
+      "epoch": 10.301851851851852,
+      "grad_norm": 1.996711631565388,
+      "learning_rate": 1.740510424699038e-07,
+      "loss": 0.6124,
+      "step": 5563
+    },
+    {
+      "epoch": 10.303703703703704,
+      "grad_norm": 1.4386105777575093,
+      "learning_rate": 1.7404096483703512e-07,
+      "loss": 0.671,
+      "step": 5564
+    },
+    {
+      "epoch": 10.305555555555555,
+      "grad_norm": 1.506472839373819,
+      "learning_rate": 1.7403088553952442e-07,
+      "loss": 0.6233,
+      "step": 5565
+    },
+    {
+      "epoch": 10.307407407407407,
+      "grad_norm": 1.5168896943004355,
+      "learning_rate": 1.7402080457759836e-07,
+      "loss": 0.6057,
+      "step": 5566
+    },
+    {
+      "epoch": 10.309259259259258,
+      "grad_norm": 1.6397689211594215,
+      "learning_rate": 1.7401072195148357e-07,
+      "loss": 0.6669,
+      "step": 5567
+    },
+    {
+      "epoch": 10.311111111111112,
+      "grad_norm": 1.8173012835702376,
+      "learning_rate": 1.7400063766140677e-07,
+      "loss": 0.6176,
+      "step": 5568
+    },
+    {
+      "epoch": 10.312962962962963,
+      "grad_norm": 1.8232472211201778,
+      "learning_rate": 1.7399055170759462e-07,
+      "loss": 0.6303,
+      "step": 5569
+    },
+    {
+      "epoch": 10.314814814814815,
+      "grad_norm": 1.8704698502710055,
+      "learning_rate": 1.739804640902739e-07,
+      "loss": 0.6187,
+      "step": 5570
+    },
+    {
+      "epoch": 10.316666666666666,
+      "grad_norm": 1.447568043820789,
+      "learning_rate": 1.7397037480967146e-07,
+      "loss": 0.6411,
+      "step": 5571
+    },
+    {
+      "epoch": 10.318518518518518,
+      "grad_norm": 1.84325159171906,
+      "learning_rate": 1.7396028386601408e-07,
+      "loss": 0.6292,
+      "step": 5572
+    },
+    {
+      "epoch": 10.32037037037037,
+      "grad_norm": 1.8953526264542464,
+      "learning_rate": 1.7395019125952864e-07,
+      "loss": 0.6542,
+      "step": 5573
+    },
+    {
+      "epoch": 10.322222222222223,
+      "grad_norm": 1.776440025858486,
+      "learning_rate": 1.7394009699044204e-07,
+      "loss": 0.6341,
+      "step": 5574
+    },
+    {
+      "epoch": 10.324074074074074,
+      "grad_norm": 1.3759567847625795,
+      "learning_rate": 1.7393000105898126e-07,
+      "loss": 0.6077,
+      "step": 5575
+    },
+    {
+      "epoch": 10.325925925925926,
+      "grad_norm": 1.718473701447944,
+      "learning_rate": 1.7391990346537327e-07,
+      "loss": 0.6443,
+      "step": 5576
+    },
+    {
+      "epoch": 10.327777777777778,
+      "grad_norm": 3.1139820921655565,
+      "learning_rate": 1.7390980420984506e-07,
+      "loss": 0.6682,
+      "step": 5577
+    },
+    {
+      "epoch": 10.329629629629629,
+      "grad_norm": 2.1036378025468214,
+      "learning_rate": 1.7389970329262375e-07,
+      "loss": 0.6181,
+      "step": 5578
+    },
+    {
+      "epoch": 10.331481481481482,
+      "grad_norm": 1.697639162940773,
+      "learning_rate": 1.7388960071393637e-07,
+      "loss": 0.6514,
+      "step": 5579
+    },
+    {
+      "epoch": 10.333333333333334,
+      "grad_norm": 1.5413222475807067,
+      "learning_rate": 1.738794964740101e-07,
+      "loss": 0.6014,
+      "step": 5580
+    },
+    {
+      "epoch": 10.335185185185185,
+      "grad_norm": 1.4884563975506548,
+      "learning_rate": 1.738693905730721e-07,
+      "loss": 0.5975,
+      "step": 5581
+    },
+    {
+      "epoch": 10.337037037037037,
+      "grad_norm": 1.4849875083875392,
+      "learning_rate": 1.7385928301134957e-07,
+      "loss": 0.6463,
+      "step": 5582
+    },
+    {
+      "epoch": 10.338888888888889,
+      "grad_norm": 1.7043338941610162,
+      "learning_rate": 1.7384917378906977e-07,
+      "loss": 0.5953,
+      "step": 5583
+    },
+    {
+      "epoch": 10.34074074074074,
+      "grad_norm": 1.4547009229947307,
+      "learning_rate": 1.7383906290645994e-07,
+      "loss": 0.6641,
+      "step": 5584
+    },
+    {
+      "epoch": 10.342592592592593,
+      "grad_norm": 1.6012708629009933,
+      "learning_rate": 1.7382895036374746e-07,
+      "loss": 0.6118,
+      "step": 5585
+    },
+    {
+      "epoch": 10.344444444444445,
+      "grad_norm": 1.4828069656544804,
+      "learning_rate": 1.7381883616115966e-07,
+      "loss": 0.6028,
+      "step": 5586
+    },
+    {
+      "epoch": 10.346296296296297,
+      "grad_norm": 1.4452112374201271,
+      "learning_rate": 1.7380872029892392e-07,
+      "loss": 0.5754,
+      "step": 5587
+    },
+    {
+      "epoch": 10.348148148148148,
+      "grad_norm": 1.5868306913319499,
+      "learning_rate": 1.7379860277726768e-07,
+      "loss": 0.5982,
+      "step": 5588
+    },
+    {
+      "epoch": 10.35,
+      "grad_norm": 1.7519571473136228,
+      "learning_rate": 1.7378848359641845e-07,
+      "loss": 0.6333,
+      "step": 5589
+    },
+    {
+      "epoch": 10.351851851851851,
+      "grad_norm": 1.7493028509881299,
+      "learning_rate": 1.7377836275660368e-07,
+      "loss": 0.5886,
+      "step": 5590
+    },
+    {
+      "epoch": 10.353703703703705,
+      "grad_norm": 1.711461949245045,
+      "learning_rate": 1.737682402580509e-07,
+      "loss": 0.6272,
+      "step": 5591
+    },
+    {
+      "epoch": 10.355555555555556,
+      "grad_norm": 1.5582610946205975,
+      "learning_rate": 1.7375811610098778e-07,
+      "loss": 0.644,
+      "step": 5592
+    },
+    {
+      "epoch": 10.357407407407408,
+      "grad_norm": 1.8660506883667933,
+      "learning_rate": 1.737479902856419e-07,
+      "loss": 0.6436,
+      "step": 5593
+    },
+    {
+      "epoch": 10.35925925925926,
+      "grad_norm": 1.5297975489859659,
+      "learning_rate": 1.7373786281224082e-07,
+      "loss": 0.6,
+      "step": 5594
+    },
+    {
+      "epoch": 10.36111111111111,
+      "grad_norm": 1.4456178552997858,
+      "learning_rate": 1.737277336810124e-07,
+      "loss": 0.617,
+      "step": 5595
+    },
+    {
+      "epoch": 10.362962962962962,
+      "grad_norm": 1.6777621052576792,
+      "learning_rate": 1.7371760289218426e-07,
+      "loss": 0.5846,
+      "step": 5596
+    },
+    {
+      "epoch": 10.364814814814816,
+      "grad_norm": 3.0504269261398242,
+      "learning_rate": 1.7370747044598417e-07,
+      "loss": 0.6568,
+      "step": 5597
+    },
+    {
+      "epoch": 10.366666666666667,
+      "grad_norm": 1.5898361692025798,
+      "learning_rate": 1.7369733634264e-07,
+      "loss": 0.6459,
+      "step": 5598
+    },
+    {
+      "epoch": 10.368518518518519,
+      "grad_norm": 1.5635668893132577,
+      "learning_rate": 1.736872005823795e-07,
+      "loss": 0.6297,
+      "step": 5599
+    },
+    {
+      "epoch": 10.37037037037037,
+      "grad_norm": 1.5461290969141985,
+      "learning_rate": 1.736770631654306e-07,
+      "loss": 0.621,
+      "step": 5600
+    },
+    {
+      "epoch": 10.372222222222222,
+      "grad_norm": 1.4832596952264288,
+      "learning_rate": 1.7366692409202127e-07,
+      "loss": 0.6278,
+      "step": 5601
+    },
+    {
+      "epoch": 10.374074074074073,
+      "grad_norm": 1.8431309223150518,
+      "learning_rate": 1.736567833623794e-07,
+      "loss": 0.6339,
+      "step": 5602
+    },
+    {
+      "epoch": 10.375925925925927,
+      "grad_norm": 1.453667585698071,
+      "learning_rate": 1.7364664097673295e-07,
+      "loss": 0.6327,
+      "step": 5603
+    },
+    {
+      "epoch": 10.377777777777778,
+      "grad_norm": 1.4787439998930816,
+      "learning_rate": 1.7363649693530999e-07,
+      "loss": 0.5829,
+      "step": 5604
+    },
+    {
+      "epoch": 10.37962962962963,
+      "grad_norm": 1.815058250165841,
+      "learning_rate": 1.736263512383386e-07,
+      "loss": 0.6344,
+      "step": 5605
+    },
+    {
+      "epoch": 10.381481481481481,
+      "grad_norm": 1.5130607145217372,
+      "learning_rate": 1.7361620388604687e-07,
+      "loss": 0.6057,
+      "step": 5606
+    },
+    {
+      "epoch": 10.383333333333333,
+      "grad_norm": 1.5348777031692438,
+      "learning_rate": 1.7360605487866297e-07,
+      "loss": 0.6713,
+      "step": 5607
+    },
+    {
+      "epoch": 10.385185185185184,
+      "grad_norm": 1.7045118441962281,
+      "learning_rate": 1.7359590421641502e-07,
+      "loss": 0.5957,
+      "step": 5608
+    },
+    {
+      "epoch": 10.387037037037038,
+      "grad_norm": 1.6201407612969965,
+      "learning_rate": 1.7358575189953123e-07,
+      "loss": 0.6239,
+      "step": 5609
+    },
+    {
+      "epoch": 10.38888888888889,
+      "grad_norm": 1.8643119793238843,
+      "learning_rate": 1.7357559792823991e-07,
+      "loss": 0.6457,
+      "step": 5610
+    },
+    {
+      "epoch": 10.39074074074074,
+      "grad_norm": 2.1230210261776192,
+      "learning_rate": 1.7356544230276933e-07,
+      "loss": 0.6267,
+      "step": 5611
+    },
+    {
+      "epoch": 10.392592592592592,
+      "grad_norm": 1.6433140724899655,
+      "learning_rate": 1.735552850233478e-07,
+      "loss": 0.6299,
+      "step": 5612
+    },
+    {
+      "epoch": 10.394444444444444,
+      "grad_norm": 1.5505719414011734,
+      "learning_rate": 1.7354512609020365e-07,
+      "loss": 0.6614,
+      "step": 5613
+    },
+    {
+      "epoch": 10.396296296296295,
+      "grad_norm": 1.6085290625304194,
+      "learning_rate": 1.7353496550356538e-07,
+      "loss": 0.6335,
+      "step": 5614
+    },
+    {
+      "epoch": 10.398148148148149,
+      "grad_norm": 1.7753758012329388,
+      "learning_rate": 1.7352480326366133e-07,
+      "loss": 0.6409,
+      "step": 5615
+    },
+    {
+      "epoch": 10.4,
+      "grad_norm": 1.6355605303039746,
+      "learning_rate": 1.7351463937072005e-07,
+      "loss": 0.6592,
+      "step": 5616
+    },
+    {
+      "epoch": 10.401851851851852,
+      "grad_norm": 1.3998280358780555,
+      "learning_rate": 1.7350447382496997e-07,
+      "loss": 0.6363,
+      "step": 5617
+    },
+    {
+      "epoch": 10.403703703703703,
+      "grad_norm": 1.6072157041710209,
+      "learning_rate": 1.7349430662663972e-07,
+      "loss": 0.6104,
+      "step": 5618
+    },
+    {
+      "epoch": 10.405555555555555,
+      "grad_norm": 1.4306611968293306,
+      "learning_rate": 1.7348413777595782e-07,
+      "loss": 0.6098,
+      "step": 5619
+    },
+    {
+      "epoch": 10.407407407407407,
+      "grad_norm": 1.804048104237655,
+      "learning_rate": 1.7347396727315293e-07,
+      "loss": 0.6416,
+      "step": 5620
+    },
+    {
+      "epoch": 10.40925925925926,
+      "grad_norm": 1.3984142985830783,
+      "learning_rate": 1.7346379511845371e-07,
+      "loss": 0.6243,
+      "step": 5621
+    },
+    {
+      "epoch": 10.411111111111111,
+      "grad_norm": 1.4859437010383754,
+      "learning_rate": 1.7345362131208885e-07,
+      "loss": 0.6492,
+      "step": 5622
+    },
+    {
+      "epoch": 10.412962962962963,
+      "grad_norm": 1.6291128754008786,
+      "learning_rate": 1.7344344585428707e-07,
+      "loss": 0.6321,
+      "step": 5623
+    },
+    {
+      "epoch": 10.414814814814815,
+      "grad_norm": 1.8993033776711115,
+      "learning_rate": 1.7343326874527717e-07,
+      "loss": 0.6746,
+      "step": 5624
+    },
+    {
+      "epoch": 10.416666666666666,
+      "grad_norm": 1.4486317926151744,
+      "learning_rate": 1.7342308998528795e-07,
+      "loss": 0.5651,
+      "step": 5625
+    },
+    {
+      "epoch": 10.418518518518518,
+      "grad_norm": 1.3376030482538848,
+      "learning_rate": 1.7341290957454825e-07,
+      "loss": 0.6434,
+      "step": 5626
+    },
+    {
+      "epoch": 10.420370370370371,
+      "grad_norm": 1.6244110671107352,
+      "learning_rate": 1.7340272751328695e-07,
+      "loss": 0.5995,
+      "step": 5627
+    },
+    {
+      "epoch": 10.422222222222222,
+      "grad_norm": 1.6323174797127578,
+      "learning_rate": 1.73392543801733e-07,
+      "loss": 0.6639,
+      "step": 5628
+    },
+    {
+      "epoch": 10.424074074074074,
+      "grad_norm": 2.0721136553374593,
+      "learning_rate": 1.7338235844011534e-07,
+      "loss": 0.6388,
+      "step": 5629
+    },
+    {
+      "epoch": 10.425925925925926,
+      "grad_norm": 1.7090250952487094,
+      "learning_rate": 1.7337217142866292e-07,
+      "loss": 0.6143,
+      "step": 5630
+    },
+    {
+      "epoch": 10.427777777777777,
+      "grad_norm": 1.874663675977847,
+      "learning_rate": 1.7336198276760484e-07,
+      "loss": 0.6422,
+      "step": 5631
+    },
+    {
+      "epoch": 10.42962962962963,
+      "grad_norm": 1.989424696624551,
+      "learning_rate": 1.7335179245717011e-07,
+      "loss": 0.6638,
+      "step": 5632
+    },
+    {
+      "epoch": 10.431481481481482,
+      "grad_norm": 1.4182583820806682,
+      "learning_rate": 1.7334160049758789e-07,
+      "loss": 0.596,
+      "step": 5633
+    },
+    {
+      "epoch": 10.433333333333334,
+      "grad_norm": 1.7317402853825656,
+      "learning_rate": 1.7333140688908728e-07,
+      "loss": 0.6233,
+      "step": 5634
+    },
+    {
+      "epoch": 10.435185185185185,
+      "grad_norm": 1.9737827564139214,
+      "learning_rate": 1.7332121163189748e-07,
+      "loss": 0.6323,
+      "step": 5635
+    },
+    {
+      "epoch": 10.437037037037037,
+      "grad_norm": 1.6008264958018399,
+      "learning_rate": 1.7331101472624772e-07,
+      "loss": 0.6386,
+      "step": 5636
+    },
+    {
+      "epoch": 10.438888888888888,
+      "grad_norm": 1.4406854334249448,
+      "learning_rate": 1.7330081617236725e-07,
+      "loss": 0.6482,
+      "step": 5637
+    },
+    {
+      "epoch": 10.440740740740742,
+      "grad_norm": 1.403411125963815,
+      "learning_rate": 1.7329061597048532e-07,
+      "loss": 0.618,
+      "step": 5638
+    },
+    {
+      "epoch": 10.442592592592593,
+      "grad_norm": 1.8285387661774546,
+      "learning_rate": 1.732804141208313e-07,
+      "loss": 0.6322,
+      "step": 5639
+    },
+    {
+      "epoch": 10.444444444444445,
+      "grad_norm": 1.6346942011361483,
+      "learning_rate": 1.7327021062363456e-07,
+      "loss": 0.6356,
+      "step": 5640
+    },
+    {
+      "epoch": 10.446296296296296,
+      "grad_norm": 1.4195442215071123,
+      "learning_rate": 1.7326000547912447e-07,
+      "loss": 0.6489,
+      "step": 5641
+    },
+    {
+      "epoch": 10.448148148148148,
+      "grad_norm": 1.4869905493495832,
+      "learning_rate": 1.732497986875305e-07,
+      "loss": 0.5884,
+      "step": 5642
+    },
+    {
+      "epoch": 10.45,
+      "grad_norm": 1.7577182872924617,
+      "learning_rate": 1.7323959024908209e-07,
+      "loss": 0.6078,
+      "step": 5643
+    },
+    {
+      "epoch": 10.451851851851853,
+      "grad_norm": 1.4224877283071675,
+      "learning_rate": 1.732293801640088e-07,
+      "loss": 0.5932,
+      "step": 5644
+    },
+    {
+      "epoch": 10.453703703703704,
+      "grad_norm": 1.3154210498915635,
+      "learning_rate": 1.732191684325401e-07,
+      "loss": 0.6175,
+      "step": 5645
+    },
+    {
+      "epoch": 10.455555555555556,
+      "grad_norm": 1.634533795488109,
+      "learning_rate": 1.732089550549057e-07,
+      "loss": 0.6838,
+      "step": 5646
+    },
+    {
+      "epoch": 10.457407407407407,
+      "grad_norm": 1.4610794154473281,
+      "learning_rate": 1.7319874003133513e-07,
+      "loss": 0.614,
+      "step": 5647
+    },
+    {
+      "epoch": 10.459259259259259,
+      "grad_norm": 1.4141051217475848,
+      "learning_rate": 1.7318852336205806e-07,
+      "loss": 0.6489,
+      "step": 5648
+    },
+    {
+      "epoch": 10.46111111111111,
+      "grad_norm": 1.557943309573647,
+      "learning_rate": 1.7317830504730426e-07,
+      "loss": 0.6452,
+      "step": 5649
+    },
+    {
+      "epoch": 10.462962962962964,
+      "grad_norm": 1.474957456195557,
+      "learning_rate": 1.7316808508730336e-07,
+      "loss": 0.6344,
+      "step": 5650
+    },
+    {
+      "epoch": 10.464814814814815,
+      "grad_norm": 2.331331729557835,
+      "learning_rate": 1.7315786348228523e-07,
+      "loss": 0.6216,
+      "step": 5651
+    },
+    {
+      "epoch": 10.466666666666667,
+      "grad_norm": 1.4983931908397468,
+      "learning_rate": 1.731476402324796e-07,
+      "loss": 0.641,
+      "step": 5652
+    },
+    {
+      "epoch": 10.468518518518518,
+      "grad_norm": 2.9008750975051107,
+      "learning_rate": 1.7313741533811638e-07,
+      "loss": 0.6115,
+      "step": 5653
+    },
+    {
+      "epoch": 10.47037037037037,
+      "grad_norm": 1.6952039861295856,
+      "learning_rate": 1.7312718879942541e-07,
+      "loss": 0.6399,
+      "step": 5654
+    },
+    {
+      "epoch": 10.472222222222221,
+      "grad_norm": 1.489424301538423,
+      "learning_rate": 1.7311696061663664e-07,
+      "loss": 0.5911,
+      "step": 5655
+    },
+    {
+      "epoch": 10.474074074074075,
+      "grad_norm": 1.7277111543420562,
+      "learning_rate": 1.7310673078997997e-07,
+      "loss": 0.6915,
+      "step": 5656
+    },
+    {
+      "epoch": 10.475925925925926,
+      "grad_norm": 1.6628745760137444,
+      "learning_rate": 1.7309649931968547e-07,
+      "loss": 0.6315,
+      "step": 5657
+    },
+    {
+      "epoch": 10.477777777777778,
+      "grad_norm": 1.533700252751984,
+      "learning_rate": 1.7308626620598314e-07,
+      "loss": 0.6502,
+      "step": 5658
+    },
+    {
+      "epoch": 10.47962962962963,
+      "grad_norm": 1.66732880869917,
+      "learning_rate": 1.7307603144910305e-07,
+      "loss": 0.6413,
+      "step": 5659
+    },
+    {
+      "epoch": 10.481481481481481,
+      "grad_norm": 1.520279876691348,
+      "learning_rate": 1.730657950492753e-07,
+      "loss": 0.6429,
+      "step": 5660
+    },
+    {
+      "epoch": 10.483333333333333,
+      "grad_norm": 1.4445882224270583,
+      "learning_rate": 1.7305555700673001e-07,
+      "loss": 0.6364,
+      "step": 5661
+    },
+    {
+      "epoch": 10.485185185185186,
+      "grad_norm": 1.8669137043784276,
+      "learning_rate": 1.7304531732169738e-07,
+      "loss": 0.5668,
+      "step": 5662
+    },
+    {
+      "epoch": 10.487037037037037,
+      "grad_norm": 1.6118966459197406,
+      "learning_rate": 1.7303507599440765e-07,
+      "loss": 0.6213,
+      "step": 5663
+    },
+    {
+      "epoch": 10.488888888888889,
+      "grad_norm": 1.4928326356671953,
+      "learning_rate": 1.7302483302509105e-07,
+      "loss": 0.6851,
+      "step": 5664
+    },
+    {
+      "epoch": 10.49074074074074,
+      "grad_norm": 1.590635808259245,
+      "learning_rate": 1.7301458841397788e-07,
+      "loss": 0.6228,
+      "step": 5665
+    },
+    {
+      "epoch": 10.492592592592592,
+      "grad_norm": 2.1337903928787227,
+      "learning_rate": 1.7300434216129847e-07,
+      "loss": 0.6367,
+      "step": 5666
+    },
+    {
+      "epoch": 10.494444444444444,
+      "grad_norm": 1.598801441389825,
+      "learning_rate": 1.7299409426728314e-07,
+      "loss": 0.6091,
+      "step": 5667
+    },
+    {
+      "epoch": 10.496296296296297,
+      "grad_norm": 1.5868704317155193,
+      "learning_rate": 1.729838447321623e-07,
+      "loss": 0.6569,
+      "step": 5668
+    },
+    {
+      "epoch": 10.498148148148148,
+      "grad_norm": 1.4258910093065258,
+      "learning_rate": 1.7297359355616644e-07,
+      "loss": 0.6369,
+      "step": 5669
+    },
+    {
+      "epoch": 10.5,
+      "grad_norm": 1.3472595043845652,
+      "learning_rate": 1.7296334073952604e-07,
+      "loss": 0.5985,
+      "step": 5670
+    },
+    {
+      "epoch": 10.501851851851852,
+      "grad_norm": 2.3170841769906643,
+      "learning_rate": 1.729530862824715e-07,
+      "loss": 0.6263,
+      "step": 5671
+    },
+    {
+      "epoch": 10.503703703703703,
+      "grad_norm": 1.7231266880447955,
+      "learning_rate": 1.7294283018523351e-07,
+      "loss": 0.5966,
+      "step": 5672
+    },
+    {
+      "epoch": 10.505555555555556,
+      "grad_norm": 1.5144459103973558,
+      "learning_rate": 1.7293257244804256e-07,
+      "loss": 0.6486,
+      "step": 5673
+    },
+    {
+      "epoch": 10.507407407407408,
+      "grad_norm": 1.594242607675054,
+      "learning_rate": 1.729223130711293e-07,
+      "loss": 0.6169,
+      "step": 5674
+    },
+    {
+      "epoch": 10.50925925925926,
+      "grad_norm": 1.936084630040173,
+      "learning_rate": 1.729120520547244e-07,
+      "loss": 0.6232,
+      "step": 5675
+    },
+    {
+      "epoch": 10.511111111111111,
+      "grad_norm": 1.571595714329208,
+      "learning_rate": 1.729017893990585e-07,
+      "loss": 0.6038,
+      "step": 5676
+    },
+    {
+      "epoch": 10.512962962962963,
+      "grad_norm": 1.3521668243150566,
+      "learning_rate": 1.7289152510436243e-07,
+      "loss": 0.6291,
+      "step": 5677
+    },
+    {
+      "epoch": 10.514814814814814,
+      "grad_norm": 1.6619348080724383,
+      "learning_rate": 1.7288125917086686e-07,
+      "loss": 0.6381,
+      "step": 5678
+    },
+    {
+      "epoch": 10.516666666666667,
+      "grad_norm": 1.6376784420357435,
+      "learning_rate": 1.7287099159880264e-07,
+      "loss": 0.589,
+      "step": 5679
+    },
+    {
+      "epoch": 10.518518518518519,
+      "grad_norm": 1.625498230622435,
+      "learning_rate": 1.7286072238840066e-07,
+      "loss": 0.5898,
+      "step": 5680
+    },
+    {
+      "epoch": 10.52037037037037,
+      "grad_norm": 3.0222733102603594,
+      "learning_rate": 1.7285045153989172e-07,
+      "loss": 0.6327,
+      "step": 5681
+    },
+    {
+      "epoch": 10.522222222222222,
+      "grad_norm": 2.762616949259456,
+      "learning_rate": 1.7284017905350677e-07,
+      "loss": 0.5789,
+      "step": 5682
+    },
+    {
+      "epoch": 10.524074074074074,
+      "grad_norm": 1.6590225967526497,
+      "learning_rate": 1.728299049294768e-07,
+      "loss": 0.5932,
+      "step": 5683
+    },
+    {
+      "epoch": 10.525925925925925,
+      "grad_norm": 1.8747846440555545,
+      "learning_rate": 1.7281962916803274e-07,
+      "loss": 0.5865,
+      "step": 5684
+    },
+    {
+      "epoch": 10.527777777777779,
+      "grad_norm": 1.9302440237974134,
+      "learning_rate": 1.7280935176940563e-07,
+      "loss": 0.612,
+      "step": 5685
+    },
+    {
+      "epoch": 10.52962962962963,
+      "grad_norm": 1.6468961884768185,
+      "learning_rate": 1.7279907273382658e-07,
+      "loss": 0.6485,
+      "step": 5686
+    },
+    {
+      "epoch": 10.531481481481482,
+      "grad_norm": 1.639521751589659,
+      "learning_rate": 1.727887920615266e-07,
+      "loss": 0.6149,
+      "step": 5687
+    },
+    {
+      "epoch": 10.533333333333333,
+      "grad_norm": 1.6145716487603634,
+      "learning_rate": 1.7277850975273693e-07,
+      "loss": 0.6682,
+      "step": 5688
+    },
+    {
+      "epoch": 10.535185185185185,
+      "grad_norm": 1.6953885478437642,
+      "learning_rate": 1.7276822580768871e-07,
+      "loss": 0.6303,
+      "step": 5689
+    },
+    {
+      "epoch": 10.537037037037036,
+      "grad_norm": 1.6567657720319768,
+      "learning_rate": 1.727579402266131e-07,
+      "loss": 0.6471,
+      "step": 5690
+    },
+    {
+      "epoch": 10.53888888888889,
+      "grad_norm": 1.5145919834696557,
+      "learning_rate": 1.7274765300974137e-07,
+      "loss": 0.6277,
+      "step": 5691
+    },
+    {
+      "epoch": 10.540740740740741,
+      "grad_norm": 1.6714727130434472,
+      "learning_rate": 1.7273736415730486e-07,
+      "loss": 0.688,
+      "step": 5692
+    },
+    {
+      "epoch": 10.542592592592593,
+      "grad_norm": 1.4287889166568395,
+      "learning_rate": 1.7272707366953485e-07,
+      "loss": 0.5754,
+      "step": 5693
+    },
+    {
+      "epoch": 10.544444444444444,
+      "grad_norm": 1.5359363884638486,
+      "learning_rate": 1.7271678154666267e-07,
+      "loss": 0.6593,
+      "step": 5694
+    },
+    {
+      "epoch": 10.546296296296296,
+      "grad_norm": 1.6102839148142356,
+      "learning_rate": 1.7270648778891977e-07,
+      "loss": 0.6192,
+      "step": 5695
+    },
+    {
+      "epoch": 10.548148148148147,
+      "grad_norm": 1.5233105207479953,
+      "learning_rate": 1.7269619239653756e-07,
+      "loss": 0.6478,
+      "step": 5696
+    },
+    {
+      "epoch": 10.55,
+      "grad_norm": 1.467582711316881,
+      "learning_rate": 1.726858953697475e-07,
+      "loss": 0.628,
+      "step": 5697
+    },
+    {
+      "epoch": 10.551851851851852,
+      "grad_norm": 1.8767352136281532,
+      "learning_rate": 1.726755967087811e-07,
+      "loss": 0.6115,
+      "step": 5698
+    },
+    {
+      "epoch": 10.553703703703704,
+      "grad_norm": 1.4990376673647625,
+      "learning_rate": 1.7266529641386989e-07,
+      "loss": 0.6566,
+      "step": 5699
+    },
+    {
+      "epoch": 10.555555555555555,
+      "grad_norm": 1.569918657540948,
+      "learning_rate": 1.7265499448524547e-07,
+      "loss": 0.6156,
+      "step": 5700
+    },
+    {
+      "epoch": 10.557407407407407,
+      "grad_norm": 1.3450302475744602,
+      "learning_rate": 1.7264469092313945e-07,
+      "loss": 0.6125,
+      "step": 5701
+    },
+    {
+      "epoch": 10.559259259259258,
+      "grad_norm": 1.475919061188164,
+      "learning_rate": 1.7263438572778347e-07,
+      "loss": 0.5925,
+      "step": 5702
+    },
+    {
+      "epoch": 10.561111111111112,
+      "grad_norm": 1.4708315566287036,
+      "learning_rate": 1.7262407889940925e-07,
+      "loss": 0.6282,
+      "step": 5703
+    },
+    {
+      "epoch": 10.562962962962963,
+      "grad_norm": 1.6717141845646533,
+      "learning_rate": 1.7261377043824844e-07,
+      "loss": 0.6169,
+      "step": 5704
+    },
+    {
+      "epoch": 10.564814814814815,
+      "grad_norm": 1.6198375480838834,
+      "learning_rate": 1.726034603445329e-07,
+      "loss": 0.6206,
+      "step": 5705
+    },
+    {
+      "epoch": 10.566666666666666,
+      "grad_norm": 1.636936084274294,
+      "learning_rate": 1.7259314861849438e-07,
+      "loss": 0.6336,
+      "step": 5706
+    },
+    {
+      "epoch": 10.568518518518518,
+      "grad_norm": 1.487970724137033,
+      "learning_rate": 1.725828352603647e-07,
+      "loss": 0.6703,
+      "step": 5707
+    },
+    {
+      "epoch": 10.57037037037037,
+      "grad_norm": 2.3176179585290857,
+      "learning_rate": 1.7257252027037577e-07,
+      "loss": 0.6244,
+      "step": 5708
+    },
+    {
+      "epoch": 10.572222222222223,
+      "grad_norm": 1.550963673680878,
+      "learning_rate": 1.7256220364875945e-07,
+      "loss": 0.6422,
+      "step": 5709
+    },
+    {
+      "epoch": 10.574074074074074,
+      "grad_norm": 1.5167872392552564,
+      "learning_rate": 1.7255188539574774e-07,
+      "loss": 0.6428,
+      "step": 5710
+    },
+    {
+      "epoch": 10.575925925925926,
+      "grad_norm": 1.5793100816140422,
+      "learning_rate": 1.7254156551157258e-07,
+      "loss": 0.5963,
+      "step": 5711
+    },
+    {
+      "epoch": 10.577777777777778,
+      "grad_norm": 1.4774129154914482,
+      "learning_rate": 1.7253124399646603e-07,
+      "loss": 0.5963,
+      "step": 5712
+    },
+    {
+      "epoch": 10.579629629629629,
+      "grad_norm": 1.5535667891951814,
+      "learning_rate": 1.7252092085066012e-07,
+      "loss": 0.6168,
+      "step": 5713
+    },
+    {
+      "epoch": 10.581481481481482,
+      "grad_norm": 1.536798729302369,
+      "learning_rate": 1.7251059607438696e-07,
+      "loss": 0.6407,
+      "step": 5714
+    },
+    {
+      "epoch": 10.583333333333334,
+      "grad_norm": 1.8388621059463646,
+      "learning_rate": 1.7250026966787866e-07,
+      "loss": 0.616,
+      "step": 5715
+    },
+    {
+      "epoch": 10.585185185185185,
+      "grad_norm": 1.5777272704786476,
+      "learning_rate": 1.7248994163136735e-07,
+      "loss": 0.6322,
+      "step": 5716
+    },
+    {
+      "epoch": 10.587037037037037,
+      "grad_norm": 1.7491826442959557,
+      "learning_rate": 1.7247961196508532e-07,
+      "loss": 0.6216,
+      "step": 5717
+    },
+    {
+      "epoch": 10.588888888888889,
+      "grad_norm": 1.493476496291092,
+      "learning_rate": 1.7246928066926473e-07,
+      "loss": 0.5683,
+      "step": 5718
+    },
+    {
+      "epoch": 10.59074074074074,
+      "grad_norm": 1.4899634244819258,
+      "learning_rate": 1.724589477441379e-07,
+      "loss": 0.6373,
+      "step": 5719
+    },
+    {
+      "epoch": 10.592592592592592,
+      "grad_norm": 1.7165190032007915,
+      "learning_rate": 1.7244861318993712e-07,
+      "loss": 0.6147,
+      "step": 5720
+    },
+    {
+      "epoch": 10.594444444444445,
+      "grad_norm": 1.4410333142919904,
+      "learning_rate": 1.7243827700689474e-07,
+      "loss": 0.6469,
+      "step": 5721
+    },
+    {
+      "epoch": 10.596296296296297,
+      "grad_norm": 1.5160481873400484,
+      "learning_rate": 1.724279391952432e-07,
+      "loss": 0.6318,
+      "step": 5722
+    },
+    {
+      "epoch": 10.598148148148148,
+      "grad_norm": 1.8621485526873551,
+      "learning_rate": 1.7241759975521484e-07,
+      "loss": 0.6337,
+      "step": 5723
+    },
+    {
+      "epoch": 10.6,
+      "grad_norm": 1.6650670094474247,
+      "learning_rate": 1.7240725868704216e-07,
+      "loss": 0.6251,
+      "step": 5724
+    },
+    {
+      "epoch": 10.601851851851851,
+      "grad_norm": 1.482307189758289,
+      "learning_rate": 1.7239691599095762e-07,
+      "loss": 0.6694,
+      "step": 5725
+    },
+    {
+      "epoch": 10.603703703703705,
+      "grad_norm": 1.581866869248674,
+      "learning_rate": 1.7238657166719382e-07,
+      "loss": 0.6438,
+      "step": 5726
+    },
+    {
+      "epoch": 10.605555555555556,
+      "grad_norm": 1.4675099285895001,
+      "learning_rate": 1.7237622571598327e-07,
+      "loss": 0.6506,
+      "step": 5727
+    },
+    {
+      "epoch": 10.607407407407408,
+      "grad_norm": 1.7920583366206484,
+      "learning_rate": 1.7236587813755863e-07,
+      "loss": 0.6525,
+      "step": 5728
+    },
+    {
+      "epoch": 10.60925925925926,
+      "grad_norm": 1.4665089772309414,
+      "learning_rate": 1.7235552893215247e-07,
+      "loss": 0.6294,
+      "step": 5729
+    },
+    {
+      "epoch": 10.61111111111111,
+      "grad_norm": 1.466159662623476,
+      "learning_rate": 1.7234517809999752e-07,
+      "loss": 0.6321,
+      "step": 5730
+    },
+    {
+      "epoch": 10.612962962962962,
+      "grad_norm": 1.4962356532612326,
+      "learning_rate": 1.7233482564132646e-07,
+      "loss": 0.6242,
+      "step": 5731
+    },
+    {
+      "epoch": 10.614814814814816,
+      "grad_norm": 1.650078858730073,
+      "learning_rate": 1.723244715563721e-07,
+      "loss": 0.5948,
+      "step": 5732
+    },
+    {
+      "epoch": 10.616666666666667,
+      "grad_norm": 1.6468620810718793,
+      "learning_rate": 1.7231411584536714e-07,
+      "loss": 0.6371,
+      "step": 5733
+    },
+    {
+      "epoch": 10.618518518518519,
+      "grad_norm": 1.815103855647812,
+      "learning_rate": 1.723037585085445e-07,
+      "loss": 0.6111,
+      "step": 5734
+    },
+    {
+      "epoch": 10.62037037037037,
+      "grad_norm": 1.6811049186790263,
+      "learning_rate": 1.7229339954613694e-07,
+      "loss": 0.6129,
+      "step": 5735
+    },
+    {
+      "epoch": 10.622222222222222,
+      "grad_norm": 1.4564777650193013,
+      "learning_rate": 1.7228303895837745e-07,
+      "loss": 0.5777,
+      "step": 5736
+    },
+    {
+      "epoch": 10.624074074074073,
+      "grad_norm": 1.3821775978180875,
+      "learning_rate": 1.7227267674549894e-07,
+      "loss": 0.6312,
+      "step": 5737
+    },
+    {
+      "epoch": 10.625925925925927,
+      "grad_norm": 1.5338686232048901,
+      "learning_rate": 1.7226231290773435e-07,
+      "loss": 0.6593,
+      "step": 5738
+    },
+    {
+      "epoch": 10.627777777777778,
+      "grad_norm": 2.0359658791350177,
+      "learning_rate": 1.722519474453167e-07,
+      "loss": 0.6529,
+      "step": 5739
+    },
+    {
+      "epoch": 10.62962962962963,
+      "grad_norm": 1.757406146016718,
+      "learning_rate": 1.7224158035847903e-07,
+      "loss": 0.6288,
+      "step": 5740
+    },
+    {
+      "epoch": 10.631481481481481,
+      "grad_norm": 1.314803996865729,
+      "learning_rate": 1.7223121164745446e-07,
+      "loss": 0.6225,
+      "step": 5741
+    },
+    {
+      "epoch": 10.633333333333333,
+      "grad_norm": 1.606924771041962,
+      "learning_rate": 1.7222084131247606e-07,
+      "loss": 0.6266,
+      "step": 5742
+    },
+    {
+      "epoch": 10.635185185185184,
+      "grad_norm": 1.4731856113852748,
+      "learning_rate": 1.7221046935377698e-07,
+      "loss": 0.6235,
+      "step": 5743
+    },
+    {
+      "epoch": 10.637037037037038,
+      "grad_norm": 2.0233919686711443,
+      "learning_rate": 1.7220009577159047e-07,
+      "loss": 0.6523,
+      "step": 5744
+    },
+    {
+      "epoch": 10.63888888888889,
+      "grad_norm": 1.8177760915677952,
+      "learning_rate": 1.7218972056614967e-07,
+      "loss": 0.6492,
+      "step": 5745
+    },
+    {
+      "epoch": 10.64074074074074,
+      "grad_norm": 1.5758205075256775,
+      "learning_rate": 1.721793437376879e-07,
+      "loss": 0.6311,
+      "step": 5746
+    },
+    {
+      "epoch": 10.642592592592592,
+      "grad_norm": 1.576370034078635,
+      "learning_rate": 1.7216896528643847e-07,
+      "loss": 0.6539,
+      "step": 5747
+    },
+    {
+      "epoch": 10.644444444444444,
+      "grad_norm": 1.5142773170881674,
+      "learning_rate": 1.7215858521263465e-07,
+      "loss": 0.6399,
+      "step": 5748
+    },
+    {
+      "epoch": 10.646296296296295,
+      "grad_norm": 1.694950143845439,
+      "learning_rate": 1.721482035165099e-07,
+      "loss": 0.6753,
+      "step": 5749
+    },
+    {
+      "epoch": 10.648148148148149,
+      "grad_norm": 1.4014231838157987,
+      "learning_rate": 1.7213782019829754e-07,
+      "loss": 0.6303,
+      "step": 5750
+    },
+    {
+      "epoch": 10.65,
+      "grad_norm": 1.598392630736057,
+      "learning_rate": 1.721274352582311e-07,
+      "loss": 0.6241,
+      "step": 5751
+    },
+    {
+      "epoch": 10.651851851851852,
+      "grad_norm": 1.3685750005585275,
+      "learning_rate": 1.7211704869654399e-07,
+      "loss": 0.6327,
+      "step": 5752
+    },
+    {
+      "epoch": 10.653703703703703,
+      "grad_norm": 1.9655545281334443,
+      "learning_rate": 1.721066605134698e-07,
+      "loss": 0.6537,
+      "step": 5753
+    },
+    {
+      "epoch": 10.655555555555555,
+      "grad_norm": 1.5167603329062058,
+      "learning_rate": 1.7209627070924198e-07,
+      "loss": 0.6374,
+      "step": 5754
+    },
+    {
+      "epoch": 10.657407407407408,
+      "grad_norm": 1.7027541708785325,
+      "learning_rate": 1.7208587928409422e-07,
+      "loss": 0.5996,
+      "step": 5755
+    },
+    {
+      "epoch": 10.65925925925926,
+      "grad_norm": 1.6180101606610673,
+      "learning_rate": 1.7207548623826012e-07,
+      "loss": 0.6302,
+      "step": 5756
+    },
+    {
+      "epoch": 10.661111111111111,
+      "grad_norm": 1.7882501712593906,
+      "learning_rate": 1.7206509157197333e-07,
+      "loss": 0.6742,
+      "step": 5757
+    },
+    {
+      "epoch": 10.662962962962963,
+      "grad_norm": 1.6386248611687573,
+      "learning_rate": 1.7205469528546751e-07,
+      "loss": 0.6479,
+      "step": 5758
+    },
+    {
+      "epoch": 10.664814814814815,
+      "grad_norm": 1.4597843438575,
+      "learning_rate": 1.720442973789765e-07,
+      "loss": 0.6559,
+      "step": 5759
+    },
+    {
+      "epoch": 10.666666666666666,
+      "grad_norm": 1.5974565688131441,
+      "learning_rate": 1.72033897852734e-07,
+      "loss": 0.6493,
+      "step": 5760
+    },
+    {
+      "epoch": 10.668518518518518,
+      "grad_norm": 1.6548839630022452,
+      "learning_rate": 1.7202349670697383e-07,
+      "loss": 0.5894,
+      "step": 5761
+    },
+    {
+      "epoch": 10.670370370370371,
+      "grad_norm": 1.604307414003859,
+      "learning_rate": 1.7201309394192985e-07,
+      "loss": 0.5961,
+      "step": 5762
+    },
+    {
+      "epoch": 10.672222222222222,
+      "grad_norm": 1.5628062678756305,
+      "learning_rate": 1.7200268955783593e-07,
+      "loss": 0.6429,
+      "step": 5763
+    },
+    {
+      "epoch": 10.674074074074074,
+      "grad_norm": 1.8009219845733089,
+      "learning_rate": 1.7199228355492598e-07,
+      "loss": 0.5842,
+      "step": 5764
+    },
+    {
+      "epoch": 10.675925925925926,
+      "grad_norm": 1.642301145030084,
+      "learning_rate": 1.71981875933434e-07,
+      "loss": 0.6378,
+      "step": 5765
+    },
+    {
+      "epoch": 10.677777777777777,
+      "grad_norm": 1.8840836079181922,
+      "learning_rate": 1.719714666935939e-07,
+      "loss": 0.6246,
+      "step": 5766
+    },
+    {
+      "epoch": 10.67962962962963,
+      "grad_norm": 1.6072095113297304,
+      "learning_rate": 1.7196105583563977e-07,
+      "loss": 0.6515,
+      "step": 5767
+    },
+    {
+      "epoch": 10.681481481481482,
+      "grad_norm": 1.5274787248533939,
+      "learning_rate": 1.719506433598057e-07,
+      "loss": 0.6703,
+      "step": 5768
+    },
+    {
+      "epoch": 10.683333333333334,
+      "grad_norm": 1.5825837593269407,
+      "learning_rate": 1.7194022926632571e-07,
+      "loss": 0.5988,
+      "step": 5769
+    },
+    {
+      "epoch": 10.685185185185185,
+      "grad_norm": 1.6630970104806428,
+      "learning_rate": 1.7192981355543398e-07,
+      "loss": 0.6266,
+      "step": 5770
+    },
+    {
+      "epoch": 10.687037037037037,
+      "grad_norm": 1.4298503678880312,
+      "learning_rate": 1.719193962273647e-07,
+      "loss": 0.602,
+      "step": 5771
+    },
+    {
+      "epoch": 10.688888888888888,
+      "grad_norm": 1.5686657351578537,
+      "learning_rate": 1.7190897728235208e-07,
+      "loss": 0.5993,
+      "step": 5772
+    },
+    {
+      "epoch": 10.690740740740742,
+      "grad_norm": 1.5180783702596443,
+      "learning_rate": 1.7189855672063033e-07,
+      "loss": 0.6315,
+      "step": 5773
+    },
+    {
+      "epoch": 10.692592592592593,
+      "grad_norm": 2.135821158532443,
+      "learning_rate": 1.7188813454243375e-07,
+      "loss": 0.596,
+      "step": 5774
+    },
+    {
+      "epoch": 10.694444444444445,
+      "grad_norm": 1.5792821781851023,
+      "learning_rate": 1.7187771074799667e-07,
+      "loss": 0.6045,
+      "step": 5775
+    },
+    {
+      "epoch": 10.696296296296296,
+      "grad_norm": 1.487706572139368,
+      "learning_rate": 1.7186728533755344e-07,
+      "loss": 0.6608,
+      "step": 5776
+    },
+    {
+      "epoch": 10.698148148148148,
+      "grad_norm": 1.4857700394336428,
+      "learning_rate": 1.718568583113384e-07,
+      "loss": 0.6295,
+      "step": 5777
+    },
+    {
+      "epoch": 10.7,
+      "grad_norm": 1.5599470941653037,
+      "learning_rate": 1.7184642966958607e-07,
+      "loss": 0.6593,
+      "step": 5778
+    },
+    {
+      "epoch": 10.701851851851853,
+      "grad_norm": 1.771713445759415,
+      "learning_rate": 1.7183599941253085e-07,
+      "loss": 0.6233,
+      "step": 5779
+    },
+    {
+      "epoch": 10.703703703703704,
+      "grad_norm": 1.5448332302225707,
+      "learning_rate": 1.7182556754040727e-07,
+      "loss": 0.6518,
+      "step": 5780
+    },
+    {
+      "epoch": 10.705555555555556,
+      "grad_norm": 1.5666295817756888,
+      "learning_rate": 1.7181513405344986e-07,
+      "loss": 0.6013,
+      "step": 5781
+    },
+    {
+      "epoch": 10.707407407407407,
+      "grad_norm": 1.464570150520032,
+      "learning_rate": 1.7180469895189318e-07,
+      "loss": 0.6614,
+      "step": 5782
+    },
+    {
+      "epoch": 10.709259259259259,
+      "grad_norm": 1.4885006350243783,
+      "learning_rate": 1.7179426223597187e-07,
+      "loss": 0.6304,
+      "step": 5783
+    },
+    {
+      "epoch": 10.71111111111111,
+      "grad_norm": 1.751847051815446,
+      "learning_rate": 1.7178382390592055e-07,
+      "loss": 0.6198,
+      "step": 5784
+    },
+    {
+      "epoch": 10.712962962962964,
+      "grad_norm": 1.4465671912199698,
+      "learning_rate": 1.717733839619739e-07,
+      "loss": 0.6121,
+      "step": 5785
+    },
+    {
+      "epoch": 10.714814814814815,
+      "grad_norm": 1.7351913173179225,
+      "learning_rate": 1.7176294240436662e-07,
+      "loss": 0.6296,
+      "step": 5786
+    },
+    {
+      "epoch": 10.716666666666667,
+      "grad_norm": 1.6720490000121169,
+      "learning_rate": 1.7175249923333351e-07,
+      "loss": 0.6136,
+      "step": 5787
+    },
+    {
+      "epoch": 10.718518518518518,
+      "grad_norm": 1.4161435815053065,
+      "learning_rate": 1.7174205444910935e-07,
+      "loss": 0.6599,
+      "step": 5788
+    },
+    {
+      "epoch": 10.72037037037037,
+      "grad_norm": 1.547747891511945,
+      "learning_rate": 1.7173160805192892e-07,
+      "loss": 0.6271,
+      "step": 5789
+    },
+    {
+      "epoch": 10.722222222222221,
+      "grad_norm": 1.8505863824612934,
+      "learning_rate": 1.7172116004202716e-07,
+      "loss": 0.6554,
+      "step": 5790
+    },
+    {
+      "epoch": 10.724074074074075,
+      "grad_norm": 1.5591947262246384,
+      "learning_rate": 1.7171071041963892e-07,
+      "loss": 0.618,
+      "step": 5791
+    },
+    {
+      "epoch": 10.725925925925926,
+      "grad_norm": 1.4283225260216625,
+      "learning_rate": 1.7170025918499916e-07,
+      "loss": 0.6655,
+      "step": 5792
+    },
+    {
+      "epoch": 10.727777777777778,
+      "grad_norm": 1.649494296468817,
+      "learning_rate": 1.7168980633834282e-07,
+      "loss": 0.6089,
+      "step": 5793
+    },
+    {
+      "epoch": 10.72962962962963,
+      "grad_norm": 1.6474498661389734,
+      "learning_rate": 1.7167935187990493e-07,
+      "loss": 0.6465,
+      "step": 5794
+    },
+    {
+      "epoch": 10.731481481481481,
+      "grad_norm": 1.7337928549553054,
+      "learning_rate": 1.7166889580992053e-07,
+      "loss": 0.7171,
+      "step": 5795
+    },
+    {
+      "epoch": 10.733333333333333,
+      "grad_norm": 1.6481526050767683,
+      "learning_rate": 1.716584381286247e-07,
+      "loss": 0.6055,
+      "step": 5796
+    },
+    {
+      "epoch": 10.735185185185186,
+      "grad_norm": 1.5917129592174264,
+      "learning_rate": 1.7164797883625255e-07,
+      "loss": 0.6149,
+      "step": 5797
+    },
+    {
+      "epoch": 10.737037037037037,
+      "grad_norm": 1.7334651950372966,
+      "learning_rate": 1.7163751793303926e-07,
+      "loss": 0.6086,
+      "step": 5798
+    },
+    {
+      "epoch": 10.738888888888889,
+      "grad_norm": 1.5973658229373153,
+      "learning_rate": 1.7162705541922002e-07,
+      "loss": 0.6547,
+      "step": 5799
+    },
+    {
+      "epoch": 10.74074074074074,
+      "grad_norm": 1.7060440299390252,
+      "learning_rate": 1.7161659129503e-07,
+      "loss": 0.6168,
+      "step": 5800
+    },
+    {
+      "epoch": 10.742592592592592,
+      "grad_norm": 2.829532026594358,
+      "learning_rate": 1.716061255607045e-07,
+      "loss": 0.6111,
+      "step": 5801
+    },
+    {
+      "epoch": 10.744444444444444,
+      "grad_norm": 2.1300543124298885,
+      "learning_rate": 1.7159565821647886e-07,
+      "loss": 0.6051,
+      "step": 5802
+    },
+    {
+      "epoch": 10.746296296296297,
+      "grad_norm": 1.5018888552860983,
+      "learning_rate": 1.7158518926258835e-07,
+      "loss": 0.63,
+      "step": 5803
+    },
+    {
+      "epoch": 10.748148148148148,
+      "grad_norm": 1.7256997847152065,
+      "learning_rate": 1.7157471869926835e-07,
+      "loss": 0.6034,
+      "step": 5804
+    },
+    {
+      "epoch": 10.75,
+      "grad_norm": 1.7022402193355048,
+      "learning_rate": 1.7156424652675428e-07,
+      "loss": 0.6287,
+      "step": 5805
+    },
+    {
+      "epoch": 10.751851851851852,
+      "grad_norm": 1.6337867590197357,
+      "learning_rate": 1.715537727452816e-07,
+      "loss": 0.594,
+      "step": 5806
+    },
+    {
+      "epoch": 10.753703703703703,
+      "grad_norm": 1.7034862539137605,
+      "learning_rate": 1.7154329735508578e-07,
+      "loss": 0.6017,
+      "step": 5807
+    },
+    {
+      "epoch": 10.755555555555556,
+      "grad_norm": 1.5604583378884518,
+      "learning_rate": 1.7153282035640228e-07,
+      "loss": 0.6057,
+      "step": 5808
+    },
+    {
+      "epoch": 10.757407407407408,
+      "grad_norm": 1.5128060791383016,
+      "learning_rate": 1.7152234174946675e-07,
+      "loss": 0.6479,
+      "step": 5809
+    },
+    {
+      "epoch": 10.75925925925926,
+      "grad_norm": 1.7800225785340646,
+      "learning_rate": 1.7151186153451472e-07,
+      "loss": 0.6667,
+      "step": 5810
+    },
+    {
+      "epoch": 10.761111111111111,
+      "grad_norm": 1.8018205137617924,
+      "learning_rate": 1.715013797117818e-07,
+      "loss": 0.6167,
+      "step": 5811
+    },
+    {
+      "epoch": 10.762962962962963,
+      "grad_norm": 1.6405391843331543,
+      "learning_rate": 1.7149089628150367e-07,
+      "loss": 0.6408,
+      "step": 5812
+    },
+    {
+      "epoch": 10.764814814814814,
+      "grad_norm": 1.6362062112278146,
+      "learning_rate": 1.7148041124391603e-07,
+      "loss": 0.6462,
+      "step": 5813
+    },
+    {
+      "epoch": 10.766666666666667,
+      "grad_norm": 2.8559182148846984,
+      "learning_rate": 1.714699245992546e-07,
+      "loss": 0.6353,
+      "step": 5814
+    },
+    {
+      "epoch": 10.768518518518519,
+      "grad_norm": 1.5107189926914173,
+      "learning_rate": 1.7145943634775517e-07,
+      "loss": 0.6699,
+      "step": 5815
+    },
+    {
+      "epoch": 10.77037037037037,
+      "grad_norm": 1.5374575422975876,
+      "learning_rate": 1.7144894648965353e-07,
+      "loss": 0.5766,
+      "step": 5816
+    },
+    {
+      "epoch": 10.772222222222222,
+      "grad_norm": 1.4958817630047532,
+      "learning_rate": 1.714384550251855e-07,
+      "loss": 0.6388,
+      "step": 5817
+    },
+    {
+      "epoch": 10.774074074074074,
+      "grad_norm": 1.599019371328936,
+      "learning_rate": 1.71427961954587e-07,
+      "loss": 0.5986,
+      "step": 5818
+    },
+    {
+      "epoch": 10.775925925925925,
+      "grad_norm": 1.6231399426538982,
+      "learning_rate": 1.7141746727809392e-07,
+      "loss": 0.6438,
+      "step": 5819
+    },
+    {
+      "epoch": 10.777777777777779,
+      "grad_norm": 1.580626800626447,
+      "learning_rate": 1.7140697099594218e-07,
+      "loss": 0.6156,
+      "step": 5820
+    },
+    {
+      "epoch": 10.77962962962963,
+      "grad_norm": 1.4687918546796297,
+      "learning_rate": 1.713964731083678e-07,
+      "loss": 0.6253,
+      "step": 5821
+    },
+    {
+      "epoch": 10.781481481481482,
+      "grad_norm": 1.6858469050498297,
+      "learning_rate": 1.7138597361560682e-07,
+      "loss": 0.6467,
+      "step": 5822
+    },
+    {
+      "epoch": 10.783333333333333,
+      "grad_norm": 1.8416731824236705,
+      "learning_rate": 1.7137547251789527e-07,
+      "loss": 0.6079,
+      "step": 5823
+    },
+    {
+      "epoch": 10.785185185185185,
+      "grad_norm": 1.5240326735536087,
+      "learning_rate": 1.7136496981546924e-07,
+      "loss": 0.6057,
+      "step": 5824
+    },
+    {
+      "epoch": 10.787037037037036,
+      "grad_norm": 1.5893687083399342,
+      "learning_rate": 1.7135446550856486e-07,
+      "loss": 0.6172,
+      "step": 5825
+    },
+    {
+      "epoch": 10.78888888888889,
+      "grad_norm": 1.416083158743161,
+      "learning_rate": 1.7134395959741828e-07,
+      "loss": 0.6092,
+      "step": 5826
+    },
+    {
+      "epoch": 10.790740740740741,
+      "grad_norm": 1.4278803242233795,
+      "learning_rate": 1.7133345208226576e-07,
+      "loss": 0.6214,
+      "step": 5827
+    },
+    {
+      "epoch": 10.792592592592593,
+      "grad_norm": 1.4683796597877252,
+      "learning_rate": 1.7132294296334347e-07,
+      "loss": 0.6597,
+      "step": 5828
+    },
+    {
+      "epoch": 10.794444444444444,
+      "grad_norm": 1.5004471842097729,
+      "learning_rate": 1.713124322408877e-07,
+      "loss": 0.672,
+      "step": 5829
+    },
+    {
+      "epoch": 10.796296296296296,
+      "grad_norm": 1.7601918721067162,
+      "learning_rate": 1.713019199151348e-07,
+      "loss": 0.6402,
+      "step": 5830
+    },
+    {
+      "epoch": 10.798148148148147,
+      "grad_norm": 1.7234651239570857,
+      "learning_rate": 1.712914059863211e-07,
+      "loss": 0.6067,
+      "step": 5831
+    },
+    {
+      "epoch": 10.8,
+      "grad_norm": 1.3879055873838537,
+      "learning_rate": 1.7128089045468293e-07,
+      "loss": 0.6249,
+      "step": 5832
+    },
+    {
+      "epoch": 10.801851851851852,
+      "grad_norm": 1.7422974737478263,
+      "learning_rate": 1.7127037332045676e-07,
+      "loss": 0.6132,
+      "step": 5833
+    },
+    {
+      "epoch": 10.803703703703704,
+      "grad_norm": 1.5972577099884724,
+      "learning_rate": 1.7125985458387904e-07,
+      "loss": 0.6533,
+      "step": 5834
+    },
+    {
+      "epoch": 10.805555555555555,
+      "grad_norm": 1.639882097607804,
+      "learning_rate": 1.7124933424518623e-07,
+      "loss": 0.6491,
+      "step": 5835
+    },
+    {
+      "epoch": 10.807407407407407,
+      "grad_norm": 1.4821054277619221,
+      "learning_rate": 1.712388123046149e-07,
+      "loss": 0.6205,
+      "step": 5836
+    },
+    {
+      "epoch": 10.809259259259258,
+      "grad_norm": 3.6329563508240157,
+      "learning_rate": 1.7122828876240156e-07,
+      "loss": 0.615,
+      "step": 5837
+    },
+    {
+      "epoch": 10.811111111111112,
+      "grad_norm": 1.6592183703060885,
+      "learning_rate": 1.7121776361878285e-07,
+      "loss": 0.673,
+      "step": 5838
+    },
+    {
+      "epoch": 10.812962962962963,
+      "grad_norm": 1.5405493700107749,
+      "learning_rate": 1.712072368739954e-07,
+      "loss": 0.6037,
+      "step": 5839
+    },
+    {
+      "epoch": 10.814814814814815,
+      "grad_norm": 1.4304083923326163,
+      "learning_rate": 1.7119670852827586e-07,
+      "loss": 0.6308,
+      "step": 5840
+    },
+    {
+      "epoch": 10.816666666666666,
+      "grad_norm": 3.3523033067518977,
+      "learning_rate": 1.7118617858186096e-07,
+      "loss": 0.5988,
+      "step": 5841
+    },
+    {
+      "epoch": 10.818518518518518,
+      "grad_norm": 1.6273275768169826,
+      "learning_rate": 1.711756470349874e-07,
+      "loss": 0.6384,
+      "step": 5842
+    },
+    {
+      "epoch": 10.82037037037037,
+      "grad_norm": 1.9555284104965578,
+      "learning_rate": 1.7116511388789201e-07,
+      "loss": 0.5834,
+      "step": 5843
+    },
+    {
+      "epoch": 10.822222222222223,
+      "grad_norm": 1.7907661489313798,
+      "learning_rate": 1.7115457914081156e-07,
+      "loss": 0.609,
+      "step": 5844
+    },
+    {
+      "epoch": 10.824074074074074,
+      "grad_norm": 1.5144972052140446,
+      "learning_rate": 1.7114404279398294e-07,
+      "loss": 0.6217,
+      "step": 5845
+    },
+    {
+      "epoch": 10.825925925925926,
+      "grad_norm": 1.5312180497361965,
+      "learning_rate": 1.7113350484764297e-07,
+      "loss": 0.6492,
+      "step": 5846
+    },
+    {
+      "epoch": 10.827777777777778,
+      "grad_norm": 1.6680249175255166,
+      "learning_rate": 1.7112296530202865e-07,
+      "loss": 0.6299,
+      "step": 5847
+    },
+    {
+      "epoch": 10.829629629629629,
+      "grad_norm": 1.5485044703697364,
+      "learning_rate": 1.711124241573769e-07,
+      "loss": 0.6662,
+      "step": 5848
+    },
+    {
+      "epoch": 10.831481481481482,
+      "grad_norm": 1.4810048006627499,
+      "learning_rate": 1.711018814139247e-07,
+      "loss": 0.6021,
+      "step": 5849
+    },
+    {
+      "epoch": 10.833333333333334,
+      "grad_norm": 1.899062348275314,
+      "learning_rate": 1.710913370719091e-07,
+      "loss": 0.6239,
+      "step": 5850
+    },
+    {
+      "epoch": 10.835185185185185,
+      "grad_norm": 1.945187880470277,
+      "learning_rate": 1.710807911315672e-07,
+      "loss": 0.6081,
+      "step": 5851
+    },
+    {
+      "epoch": 10.837037037037037,
+      "grad_norm": 1.5811830615277291,
+      "learning_rate": 1.7107024359313603e-07,
+      "loss": 0.651,
+      "step": 5852
+    },
+    {
+      "epoch": 10.838888888888889,
+      "grad_norm": 1.5500753098618565,
+      "learning_rate": 1.7105969445685275e-07,
+      "loss": 0.6798,
+      "step": 5853
+    },
+    {
+      "epoch": 10.84074074074074,
+      "grad_norm": 1.9419497627389486,
+      "learning_rate": 1.7104914372295456e-07,
+      "loss": 0.6469,
+      "step": 5854
+    },
+    {
+      "epoch": 10.842592592592592,
+      "grad_norm": 1.462000541562202,
+      "learning_rate": 1.7103859139167867e-07,
+      "loss": 0.6413,
+      "step": 5855
+    },
+    {
+      "epoch": 10.844444444444445,
+      "grad_norm": 1.5288525604190968,
+      "learning_rate": 1.7102803746326227e-07,
+      "loss": 0.6319,
+      "step": 5856
+    },
+    {
+      "epoch": 10.846296296296297,
+      "grad_norm": 1.488358505757009,
+      "learning_rate": 1.710174819379427e-07,
+      "loss": 0.6145,
+      "step": 5857
+    },
+    {
+      "epoch": 10.848148148148148,
+      "grad_norm": 2.3875678304550707,
+      "learning_rate": 1.7100692481595727e-07,
+      "loss": 0.6432,
+      "step": 5858
+    },
+    {
+      "epoch": 10.85,
+      "grad_norm": 1.361304194274407,
+      "learning_rate": 1.7099636609754326e-07,
+      "loss": 0.6048,
+      "step": 5859
+    },
+    {
+      "epoch": 10.851851851851851,
+      "grad_norm": 1.463855705963754,
+      "learning_rate": 1.7098580578293818e-07,
+      "loss": 0.5892,
+      "step": 5860
+    },
+    {
+      "epoch": 10.853703703703705,
+      "grad_norm": 1.5644114917869907,
+      "learning_rate": 1.7097524387237936e-07,
+      "loss": 0.6075,
+      "step": 5861
+    },
+    {
+      "epoch": 10.855555555555556,
+      "grad_norm": 1.3640782096032182,
+      "learning_rate": 1.709646803661043e-07,
+      "loss": 0.6224,
+      "step": 5862
+    },
+    {
+      "epoch": 10.857407407407408,
+      "grad_norm": 1.5128254741137421,
+      "learning_rate": 1.709541152643505e-07,
+      "loss": 0.6027,
+      "step": 5863
+    },
+    {
+      "epoch": 10.85925925925926,
+      "grad_norm": 1.4619139483410486,
+      "learning_rate": 1.7094354856735548e-07,
+      "loss": 0.6242,
+      "step": 5864
+    },
+    {
+      "epoch": 10.86111111111111,
+      "grad_norm": 1.3849844022499764,
+      "learning_rate": 1.7093298027535677e-07,
+      "loss": 0.6249,
+      "step": 5865
+    },
+    {
+      "epoch": 10.862962962962962,
+      "grad_norm": 1.668214596545056,
+      "learning_rate": 1.7092241038859207e-07,
+      "loss": 0.6404,
+      "step": 5866
+    },
+    {
+      "epoch": 10.864814814814816,
+      "grad_norm": 1.344688229689299,
+      "learning_rate": 1.709118389072989e-07,
+      "loss": 0.6471,
+      "step": 5867
+    },
+    {
+      "epoch": 10.866666666666667,
+      "grad_norm": 1.4592079445921544,
+      "learning_rate": 1.7090126583171503e-07,
+      "loss": 0.6518,
+      "step": 5868
+    },
+    {
+      "epoch": 10.868518518518519,
+      "grad_norm": 1.5195929428178105,
+      "learning_rate": 1.7089069116207812e-07,
+      "loss": 0.6428,
+      "step": 5869
+    },
+    {
+      "epoch": 10.87037037037037,
+      "grad_norm": 1.7047514622055526,
+      "learning_rate": 1.7088011489862598e-07,
+      "loss": 0.6277,
+      "step": 5870
+    },
+    {
+      "epoch": 10.872222222222222,
+      "grad_norm": 1.660986182730027,
+      "learning_rate": 1.7086953704159631e-07,
+      "loss": 0.6414,
+      "step": 5871
+    },
+    {
+      "epoch": 10.874074074074073,
+      "grad_norm": 1.4967613780406208,
+      "learning_rate": 1.7085895759122697e-07,
+      "loss": 0.6457,
+      "step": 5872
+    },
+    {
+      "epoch": 10.875925925925927,
+      "grad_norm": 1.6871884626761262,
+      "learning_rate": 1.7084837654775583e-07,
+      "loss": 0.6527,
+      "step": 5873
+    },
+    {
+      "epoch": 10.877777777777778,
+      "grad_norm": 2.2725637080703462,
+      "learning_rate": 1.7083779391142077e-07,
+      "loss": 0.6361,
+      "step": 5874
+    },
+    {
+      "epoch": 10.87962962962963,
+      "grad_norm": 1.7463092073225324,
+      "learning_rate": 1.7082720968245973e-07,
+      "loss": 0.6215,
+      "step": 5875
+    },
+    {
+      "epoch": 10.881481481481481,
+      "grad_norm": 1.555899650861377,
+      "learning_rate": 1.708166238611106e-07,
+      "loss": 0.6331,
+      "step": 5876
+    },
+    {
+      "epoch": 10.883333333333333,
+      "grad_norm": 2.3952046897198165,
+      "learning_rate": 1.7080603644761146e-07,
+      "loss": 0.5985,
+      "step": 5877
+    },
+    {
+      "epoch": 10.885185185185184,
+      "grad_norm": 1.5667752679022278,
+      "learning_rate": 1.707954474422003e-07,
+      "loss": 0.6049,
+      "step": 5878
+    },
+    {
+      "epoch": 10.887037037037038,
+      "grad_norm": 1.523051244182511,
+      "learning_rate": 1.7078485684511522e-07,
+      "loss": 0.6246,
+      "step": 5879
+    },
+    {
+      "epoch": 10.88888888888889,
+      "grad_norm": 1.5159082377614266,
+      "learning_rate": 1.7077426465659431e-07,
+      "loss": 0.583,
+      "step": 5880
+    },
+    {
+      "epoch": 10.89074074074074,
+      "grad_norm": 1.3717961459429246,
+      "learning_rate": 1.707636708768757e-07,
+      "loss": 0.6163,
+      "step": 5881
+    },
+    {
+      "epoch": 10.892592592592592,
+      "grad_norm": 1.6332154906681302,
+      "learning_rate": 1.7075307550619758e-07,
+      "loss": 0.658,
+      "step": 5882
+    },
+    {
+      "epoch": 10.894444444444444,
+      "grad_norm": 1.4130658756923027,
+      "learning_rate": 1.7074247854479814e-07,
+      "loss": 0.6392,
+      "step": 5883
+    },
+    {
+      "epoch": 10.896296296296295,
+      "grad_norm": 1.5593087207293561,
+      "learning_rate": 1.7073187999291567e-07,
+      "loss": 0.6316,
+      "step": 5884
+    },
+    {
+      "epoch": 10.898148148148149,
+      "grad_norm": 1.5107680102122947,
+      "learning_rate": 1.7072127985078844e-07,
+      "loss": 0.6098,
+      "step": 5885
+    },
+    {
+      "epoch": 10.9,
+      "grad_norm": 1.5810094424258323,
+      "learning_rate": 1.7071067811865473e-07,
+      "loss": 0.6747,
+      "step": 5886
+    },
+    {
+      "epoch": 10.901851851851852,
+      "grad_norm": 1.4391330277702756,
+      "learning_rate": 1.7070007479675296e-07,
+      "loss": 0.5995,
+      "step": 5887
+    },
+    {
+      "epoch": 10.903703703703703,
+      "grad_norm": 1.58909736533067,
+      "learning_rate": 1.706894698853215e-07,
+      "loss": 0.6196,
+      "step": 5888
+    },
+    {
+      "epoch": 10.905555555555555,
+      "grad_norm": 1.6160992523099549,
+      "learning_rate": 1.7067886338459875e-07,
+      "loss": 0.6746,
+      "step": 5889
+    },
+    {
+      "epoch": 10.907407407407408,
+      "grad_norm": 1.7125014057466894,
+      "learning_rate": 1.7066825529482319e-07,
+      "loss": 0.575,
+      "step": 5890
+    },
+    {
+      "epoch": 10.90925925925926,
+      "grad_norm": 1.6978009646612635,
+      "learning_rate": 1.706576456162333e-07,
+      "loss": 0.6328,
+      "step": 5891
+    },
+    {
+      "epoch": 10.911111111111111,
+      "grad_norm": 1.6722955547612481,
+      "learning_rate": 1.7064703434906767e-07,
+      "loss": 0.6444,
+      "step": 5892
+    },
+    {
+      "epoch": 10.912962962962963,
+      "grad_norm": 1.7627915282042608,
+      "learning_rate": 1.7063642149356481e-07,
+      "loss": 0.6066,
+      "step": 5893
+    },
+    {
+      "epoch": 10.914814814814815,
+      "grad_norm": 1.7990006325096803,
+      "learning_rate": 1.7062580704996335e-07,
+      "loss": 0.6122,
+      "step": 5894
+    },
+    {
+      "epoch": 10.916666666666666,
+      "grad_norm": 2.801482688622267,
+      "learning_rate": 1.7061519101850195e-07,
+      "loss": 0.6251,
+      "step": 5895
+    },
+    {
+      "epoch": 10.918518518518518,
+      "grad_norm": 1.664448645333804,
+      "learning_rate": 1.7060457339941926e-07,
+      "loss": 0.6406,
+      "step": 5896
+    },
+    {
+      "epoch": 10.920370370370371,
+      "grad_norm": 1.4676581184883395,
+      "learning_rate": 1.7059395419295396e-07,
+      "loss": 0.6705,
+      "step": 5897
+    },
+    {
+      "epoch": 10.922222222222222,
+      "grad_norm": 1.6122419827824905,
+      "learning_rate": 1.705833333993449e-07,
+      "loss": 0.6384,
+      "step": 5898
+    },
+    {
+      "epoch": 10.924074074074074,
+      "grad_norm": 1.8826563430347905,
+      "learning_rate": 1.7057271101883077e-07,
+      "loss": 0.6345,
+      "step": 5899
+    },
+    {
+      "epoch": 10.925925925925926,
+      "grad_norm": 1.7319788554003726,
+      "learning_rate": 1.7056208705165041e-07,
+      "loss": 0.6643,
+      "step": 5900
+    },
+    {
+      "epoch": 10.927777777777777,
+      "grad_norm": 2.009172340771583,
+      "learning_rate": 1.7055146149804273e-07,
+      "loss": 0.6551,
+      "step": 5901
+    },
+    {
+      "epoch": 10.92962962962963,
+      "grad_norm": 1.7828116051587684,
+      "learning_rate": 1.7054083435824657e-07,
+      "loss": 0.5951,
+      "step": 5902
+    },
+    {
+      "epoch": 10.931481481481482,
+      "grad_norm": 1.8879528408979587,
+      "learning_rate": 1.7053020563250087e-07,
+      "loss": 0.6283,
+      "step": 5903
+    },
+    {
+      "epoch": 10.933333333333334,
+      "grad_norm": 1.4773106532183309,
+      "learning_rate": 1.7051957532104459e-07,
+      "loss": 0.615,
+      "step": 5904
+    },
+    {
+      "epoch": 10.935185185185185,
+      "grad_norm": 1.4529833705674395,
+      "learning_rate": 1.705089434241167e-07,
+      "loss": 0.5874,
+      "step": 5905
+    },
+    {
+      "epoch": 10.937037037037037,
+      "grad_norm": 1.7386534208172533,
+      "learning_rate": 1.7049830994195626e-07,
+      "loss": 0.6616,
+      "step": 5906
+    },
+    {
+      "epoch": 10.938888888888888,
+      "grad_norm": 1.480171742846781,
+      "learning_rate": 1.7048767487480239e-07,
+      "loss": 0.6581,
+      "step": 5907
+    },
+    {
+      "epoch": 10.940740740740742,
+      "grad_norm": 2.4007596250601653,
+      "learning_rate": 1.7047703822289413e-07,
+      "loss": 0.6331,
+      "step": 5908
+    },
+    {
+      "epoch": 10.942592592592593,
+      "grad_norm": 1.6496927901459952,
+      "learning_rate": 1.7046639998647062e-07,
+      "loss": 0.6502,
+      "step": 5909
+    },
+    {
+      "epoch": 10.944444444444445,
+      "grad_norm": 1.537044194589527,
+      "learning_rate": 1.7045576016577102e-07,
+      "loss": 0.6481,
+      "step": 5910
+    },
+    {
+      "epoch": 10.946296296296296,
+      "grad_norm": 1.4557301788407626,
+      "learning_rate": 1.7044511876103464e-07,
+      "loss": 0.6144,
+      "step": 5911
+    },
+    {
+      "epoch": 10.948148148148148,
+      "grad_norm": 1.6435139556027867,
+      "learning_rate": 1.7043447577250064e-07,
+      "loss": 0.6402,
+      "step": 5912
+    },
+    {
+      "epoch": 10.95,
+      "grad_norm": 1.7564321079667797,
+      "learning_rate": 1.7042383120040833e-07,
+      "loss": 0.6436,
+      "step": 5913
+    },
+    {
+      "epoch": 10.951851851851853,
+      "grad_norm": 1.5061460319688613,
+      "learning_rate": 1.7041318504499703e-07,
+      "loss": 0.6294,
+      "step": 5914
+    },
+    {
+      "epoch": 10.953703703703704,
+      "grad_norm": 2.5184878638740904,
+      "learning_rate": 1.7040253730650608e-07,
+      "loss": 0.6538,
+      "step": 5915
+    },
+    {
+      "epoch": 10.955555555555556,
+      "grad_norm": 1.5949283379518575,
+      "learning_rate": 1.703918879851749e-07,
+      "loss": 0.6158,
+      "step": 5916
+    },
+    {
+      "epoch": 10.957407407407407,
+      "grad_norm": 1.752464075535082,
+      "learning_rate": 1.7038123708124288e-07,
+      "loss": 0.6423,
+      "step": 5917
+    },
+    {
+      "epoch": 10.959259259259259,
+      "grad_norm": 1.5609427691533169,
+      "learning_rate": 1.703705845949495e-07,
+      "loss": 0.6132,
+      "step": 5918
+    },
+    {
+      "epoch": 10.96111111111111,
+      "grad_norm": 1.8231154624004224,
+      "learning_rate": 1.7035993052653427e-07,
+      "loss": 0.631,
+      "step": 5919
+    },
+    {
+      "epoch": 10.962962962962964,
+      "grad_norm": 1.5501861038383078,
+      "learning_rate": 1.7034927487623668e-07,
+      "loss": 0.6409,
+      "step": 5920
+    },
+    {
+      "epoch": 10.964814814814815,
+      "grad_norm": 1.8829032410868844,
+      "learning_rate": 1.7033861764429634e-07,
+      "loss": 0.6023,
+      "step": 5921
+    },
+    {
+      "epoch": 10.966666666666667,
+      "grad_norm": 1.9486797269862017,
+      "learning_rate": 1.7032795883095284e-07,
+      "loss": 0.6345,
+      "step": 5922
+    },
+    {
+      "epoch": 10.968518518518518,
+      "grad_norm": 2.5087948213646323,
+      "learning_rate": 1.7031729843644582e-07,
+      "loss": 0.5874,
+      "step": 5923
+    },
+    {
+      "epoch": 10.97037037037037,
+      "grad_norm": 1.4019500252654198,
+      "learning_rate": 1.7030663646101493e-07,
+      "loss": 0.6489,
+      "step": 5924
+    },
+    {
+      "epoch": 10.972222222222221,
+      "grad_norm": 2.002097245795404,
+      "learning_rate": 1.7029597290489994e-07,
+      "loss": 0.5809,
+      "step": 5925
+    },
+    {
+      "epoch": 10.974074074074075,
+      "grad_norm": 1.7688245375718317,
+      "learning_rate": 1.7028530776834056e-07,
+      "loss": 0.6196,
+      "step": 5926
+    },
+    {
+      "epoch": 10.975925925925926,
+      "grad_norm": 1.449846185082219,
+      "learning_rate": 1.7027464105157653e-07,
+      "loss": 0.608,
+      "step": 5927
+    },
+    {
+      "epoch": 10.977777777777778,
+      "grad_norm": 1.5576471207792844,
+      "learning_rate": 1.7026397275484773e-07,
+      "loss": 0.6157,
+      "step": 5928
+    },
+    {
+      "epoch": 10.97962962962963,
+      "grad_norm": 1.944666932166961,
+      "learning_rate": 1.70253302878394e-07,
+      "loss": 0.6052,
+      "step": 5929
+    },
+    {
+      "epoch": 10.981481481481481,
+      "grad_norm": 1.7634398773183664,
+      "learning_rate": 1.7024263142245515e-07,
+      "loss": 0.629,
+      "step": 5930
+    },
+    {
+      "epoch": 10.983333333333333,
+      "grad_norm": 1.698089668444641,
+      "learning_rate": 1.7023195838727124e-07,
+      "loss": 0.6443,
+      "step": 5931
+    },
+    {
+      "epoch": 10.985185185185186,
+      "grad_norm": 1.4705764205025331,
+      "learning_rate": 1.7022128377308215e-07,
+      "loss": 0.5944,
+      "step": 5932
+    },
+    {
+      "epoch": 10.987037037037037,
+      "grad_norm": 1.5898779509725238,
+      "learning_rate": 1.7021060758012785e-07,
+      "loss": 0.6544,
+      "step": 5933
+    },
+    {
+      "epoch": 10.988888888888889,
+      "grad_norm": 1.425544155866549,
+      "learning_rate": 1.7019992980864844e-07,
+      "loss": 0.6397,
+      "step": 5934
+    },
+    {
+      "epoch": 10.99074074074074,
+      "grad_norm": 1.5403476289747633,
+      "learning_rate": 1.7018925045888393e-07,
+      "loss": 0.63,
+      "step": 5935
+    },
+    {
+      "epoch": 10.992592592592592,
+      "grad_norm": 1.546030320999319,
+      "learning_rate": 1.7017856953107441e-07,
+      "loss": 0.5924,
+      "step": 5936
+    },
+    {
+      "epoch": 10.994444444444444,
+      "grad_norm": 1.7900087734556676,
+      "learning_rate": 1.701678870254601e-07,
+      "loss": 0.5829,
+      "step": 5937
+    },
+    {
+      "epoch": 10.996296296296297,
+      "grad_norm": 1.5863577571041783,
+      "learning_rate": 1.7015720294228109e-07,
+      "loss": 0.5829,
+      "step": 5938
+    },
+    {
+      "epoch": 10.998148148148148,
+      "grad_norm": 1.7033146766985157,
+      "learning_rate": 1.701465172817776e-07,
+      "loss": 0.6205,
+      "step": 5939
+    },
+    {
+      "epoch": 11.0,
+      "grad_norm": 1.4980773676279124,
+      "learning_rate": 1.7013583004418992e-07,
+      "loss": 0.6477,
+      "step": 5940
+    },
+    {
+      "epoch": 11.001851851851852,
+      "grad_norm": 1.7757209065045465,
+      "learning_rate": 1.7012514122975827e-07,
+      "loss": 0.6264,
+      "step": 5941
+    },
+    {
+      "epoch": 11.003703703703703,
+      "grad_norm": 1.5566347525346038,
+      "learning_rate": 1.7011445083872297e-07,
+      "loss": 0.6158,
+      "step": 5942
+    },
+    {
+      "epoch": 11.005555555555556,
+      "grad_norm": 1.8648723244427359,
+      "learning_rate": 1.7010375887132442e-07,
+      "loss": 0.6361,
+      "step": 5943
+    },
+    {
+      "epoch": 11.007407407407408,
+      "grad_norm": 1.6622766515932923,
+      "learning_rate": 1.7009306532780295e-07,
+      "loss": 0.6351,
+      "step": 5944
+    },
+    {
+      "epoch": 11.00925925925926,
+      "grad_norm": 1.4281526003066156,
+      "learning_rate": 1.7008237020839903e-07,
+      "loss": 0.6298,
+      "step": 5945
+    },
+    {
+      "epoch": 11.011111111111111,
+      "grad_norm": 1.615771249114317,
+      "learning_rate": 1.7007167351335307e-07,
+      "loss": 0.6221,
+      "step": 5946
+    },
+    {
+      "epoch": 11.012962962962963,
+      "grad_norm": 1.426716798274646,
+      "learning_rate": 1.7006097524290556e-07,
+      "loss": 0.6147,
+      "step": 5947
+    },
+    {
+      "epoch": 11.014814814814814,
+      "grad_norm": 2.0816540916383426,
+      "learning_rate": 1.7005027539729708e-07,
+      "loss": 0.6317,
+      "step": 5948
+    },
+    {
+      "epoch": 11.016666666666667,
+      "grad_norm": 1.5724228171986836,
+      "learning_rate": 1.7003957397676816e-07,
+      "loss": 0.646,
+      "step": 5949
+    },
+    {
+      "epoch": 11.018518518518519,
+      "grad_norm": 1.5234184039832246,
+      "learning_rate": 1.7002887098155933e-07,
+      "loss": 0.6359,
+      "step": 5950
+    },
+    {
+      "epoch": 11.02037037037037,
+      "grad_norm": 1.7935653084244785,
+      "learning_rate": 1.7001816641191133e-07,
+      "loss": 0.6182,
+      "step": 5951
+    },
+    {
+      "epoch": 11.022222222222222,
+      "grad_norm": 1.6395628034279153,
+      "learning_rate": 1.7000746026806477e-07,
+      "loss": 0.6449,
+      "step": 5952
+    },
+    {
+      "epoch": 11.024074074074074,
+      "grad_norm": 1.8087511888882657,
+      "learning_rate": 1.699967525502604e-07,
+      "loss": 0.6233,
+      "step": 5953
+    },
+    {
+      "epoch": 11.025925925925925,
+      "grad_norm": 1.5036446125841607,
+      "learning_rate": 1.699860432587389e-07,
+      "loss": 0.6061,
+      "step": 5954
+    },
+    {
+      "epoch": 11.027777777777779,
+      "grad_norm": 1.3716273776904815,
+      "learning_rate": 1.6997533239374104e-07,
+      "loss": 0.625,
+      "step": 5955
+    },
+    {
+      "epoch": 11.02962962962963,
+      "grad_norm": 1.5582109838612122,
+      "learning_rate": 1.699646199555077e-07,
+      "loss": 0.6222,
+      "step": 5956
+    },
+    {
+      "epoch": 11.031481481481482,
+      "grad_norm": 1.5441614423046037,
+      "learning_rate": 1.6995390594427965e-07,
+      "loss": 0.6588,
+      "step": 5957
+    },
+    {
+      "epoch": 11.033333333333333,
+      "grad_norm": 1.547692424104394,
+      "learning_rate": 1.6994319036029783e-07,
+      "loss": 0.6078,
+      "step": 5958
+    },
+    {
+      "epoch": 11.035185185185185,
+      "grad_norm": 1.5502695973119278,
+      "learning_rate": 1.6993247320380313e-07,
+      "loss": 0.6319,
+      "step": 5959
+    },
+    {
+      "epoch": 11.037037037037036,
+      "grad_norm": 1.5349932782159459,
+      "learning_rate": 1.6992175447503647e-07,
+      "loss": 0.6348,
+      "step": 5960
+    },
+    {
+      "epoch": 11.03888888888889,
+      "grad_norm": 1.8132430452866657,
+      "learning_rate": 1.699110341742389e-07,
+      "loss": 0.6373,
+      "step": 5961
+    },
+    {
+      "epoch": 11.040740740740741,
+      "grad_norm": 1.4455850069334868,
+      "learning_rate": 1.6990031230165135e-07,
+      "loss": 0.6484,
+      "step": 5962
+    },
+    {
+      "epoch": 11.042592592592593,
+      "grad_norm": 1.709489917083784,
+      "learning_rate": 1.69889588857515e-07,
+      "loss": 0.6083,
+      "step": 5963
+    },
+    {
+      "epoch": 11.044444444444444,
+      "grad_norm": 1.7275875346330496,
+      "learning_rate": 1.6987886384207083e-07,
+      "loss": 0.6763,
+      "step": 5964
+    },
+    {
+      "epoch": 11.046296296296296,
+      "grad_norm": 1.6342008958016723,
+      "learning_rate": 1.6986813725556002e-07,
+      "loss": 0.6267,
+      "step": 5965
+    },
+    {
+      "epoch": 11.048148148148147,
+      "grad_norm": 1.6178084650958968,
+      "learning_rate": 1.6985740909822375e-07,
+      "loss": 0.6068,
+      "step": 5966
+    },
+    {
+      "epoch": 11.05,
+      "grad_norm": 1.528443996617462,
+      "learning_rate": 1.6984667937030316e-07,
+      "loss": 0.5982,
+      "step": 5967
+    },
+    {
+      "epoch": 11.051851851851852,
+      "grad_norm": 1.729968006765182,
+      "learning_rate": 1.6983594807203953e-07,
+      "loss": 0.6391,
+      "step": 5968
+    },
+    {
+      "epoch": 11.053703703703704,
+      "grad_norm": 1.4396639839859426,
+      "learning_rate": 1.6982521520367411e-07,
+      "loss": 0.5996,
+      "step": 5969
+    },
+    {
+      "epoch": 11.055555555555555,
+      "grad_norm": 1.9726778631487216,
+      "learning_rate": 1.6981448076544825e-07,
+      "loss": 0.6212,
+      "step": 5970
+    },
+    {
+      "epoch": 11.057407407407407,
+      "grad_norm": 1.6734363757810804,
+      "learning_rate": 1.6980374475760318e-07,
+      "loss": 0.6378,
+      "step": 5971
+    },
+    {
+      "epoch": 11.059259259259258,
+      "grad_norm": 1.6628775109845524,
+      "learning_rate": 1.6979300718038042e-07,
+      "loss": 0.6211,
+      "step": 5972
+    },
+    {
+      "epoch": 11.061111111111112,
+      "grad_norm": 1.6559022609778822,
+      "learning_rate": 1.6978226803402123e-07,
+      "loss": 0.6389,
+      "step": 5973
+    },
+    {
+      "epoch": 11.062962962962963,
+      "grad_norm": 1.477957725222872,
+      "learning_rate": 1.697715273187672e-07,
+      "loss": 0.6663,
+      "step": 5974
+    },
+    {
+      "epoch": 11.064814814814815,
+      "grad_norm": 1.7470628626633204,
+      "learning_rate": 1.697607850348597e-07,
+      "loss": 0.6123,
+      "step": 5975
+    },
+    {
+      "epoch": 11.066666666666666,
+      "grad_norm": 1.5236742126645169,
+      "learning_rate": 1.6975004118254027e-07,
+      "loss": 0.6234,
+      "step": 5976
+    },
+    {
+      "epoch": 11.068518518518518,
+      "grad_norm": 1.4456330907270265,
+      "learning_rate": 1.697392957620505e-07,
+      "loss": 0.6046,
+      "step": 5977
+    },
+    {
+      "epoch": 11.07037037037037,
+      "grad_norm": 1.417079906843894,
+      "learning_rate": 1.6972854877363193e-07,
+      "loss": 0.6234,
+      "step": 5978
+    },
+    {
+      "epoch": 11.072222222222223,
+      "grad_norm": 1.5028361842938198,
+      "learning_rate": 1.6971780021752624e-07,
+      "loss": 0.6607,
+      "step": 5979
+    },
+    {
+      "epoch": 11.074074074074074,
+      "grad_norm": 1.5608358950670342,
+      "learning_rate": 1.6970705009397502e-07,
+      "loss": 0.6133,
+      "step": 5980
+    },
+    {
+      "epoch": 11.075925925925926,
+      "grad_norm": 1.727954642805163,
+      "learning_rate": 1.6969629840322e-07,
+      "loss": 0.6048,
+      "step": 5981
+    },
+    {
+      "epoch": 11.077777777777778,
+      "grad_norm": 3.716198016960567,
+      "learning_rate": 1.6968554514550291e-07,
+      "loss": 0.6014,
+      "step": 5982
+    },
+    {
+      "epoch": 11.079629629629629,
+      "grad_norm": 1.409378465235887,
+      "learning_rate": 1.6967479032106548e-07,
+      "loss": 0.6276,
+      "step": 5983
+    },
+    {
+      "epoch": 11.081481481481482,
+      "grad_norm": 1.7093019742636377,
+      "learning_rate": 1.6966403393014955e-07,
+      "loss": 0.599,
+      "step": 5984
+    },
+    {
+      "epoch": 11.083333333333334,
+      "grad_norm": 1.416583372862473,
+      "learning_rate": 1.6965327597299695e-07,
+      "loss": 0.6833,
+      "step": 5985
+    },
+    {
+      "epoch": 11.085185185185185,
+      "grad_norm": 1.5970954100280967,
+      "learning_rate": 1.6964251644984948e-07,
+      "loss": 0.6433,
+      "step": 5986
+    },
+    {
+      "epoch": 11.087037037037037,
+      "grad_norm": 1.5081580156050232,
+      "learning_rate": 1.6963175536094915e-07,
+      "loss": 0.6016,
+      "step": 5987
+    },
+    {
+      "epoch": 11.088888888888889,
+      "grad_norm": 1.4807640555911499,
+      "learning_rate": 1.696209927065378e-07,
+      "loss": 0.6372,
+      "step": 5988
+    },
+    {
+      "epoch": 11.09074074074074,
+      "grad_norm": 1.6124370048274683,
+      "learning_rate": 1.6961022848685747e-07,
+      "loss": 0.6614,
+      "step": 5989
+    },
+    {
+      "epoch": 11.092592592592593,
+      "grad_norm": 1.7178997830146034,
+      "learning_rate": 1.6959946270215012e-07,
+      "loss": 0.6196,
+      "step": 5990
+    },
+    {
+      "epoch": 11.094444444444445,
+      "grad_norm": 1.7679883986376386,
+      "learning_rate": 1.6958869535265786e-07,
+      "loss": 0.6376,
+      "step": 5991
+    },
+    {
+      "epoch": 11.096296296296297,
+      "grad_norm": 1.688111609031246,
+      "learning_rate": 1.6957792643862268e-07,
+      "loss": 0.6204,
+      "step": 5992
+    },
+    {
+      "epoch": 11.098148148148148,
+      "grad_norm": 1.5947660269786659,
+      "learning_rate": 1.6956715596028676e-07,
+      "loss": 0.6142,
+      "step": 5993
+    },
+    {
+      "epoch": 11.1,
+      "grad_norm": 1.4654212159204272,
+      "learning_rate": 1.6955638391789226e-07,
+      "loss": 0.6232,
+      "step": 5994
+    },
+    {
+      "epoch": 11.101851851851851,
+      "grad_norm": 1.956301815811382,
+      "learning_rate": 1.6954561031168127e-07,
+      "loss": 0.6134,
+      "step": 5995
+    },
+    {
+      "epoch": 11.103703703703705,
+      "grad_norm": 1.5222159748830348,
+      "learning_rate": 1.6953483514189615e-07,
+      "loss": 0.6458,
+      "step": 5996
+    },
+    {
+      "epoch": 11.105555555555556,
+      "grad_norm": 1.3575981078835275,
+      "learning_rate": 1.6952405840877906e-07,
+      "loss": 0.6493,
+      "step": 5997
+    },
+    {
+      "epoch": 11.107407407407408,
+      "grad_norm": 1.8044274184417082,
+      "learning_rate": 1.6951328011257232e-07,
+      "loss": 0.6314,
+      "step": 5998
+    },
+    {
+      "epoch": 11.10925925925926,
+      "grad_norm": 2.1737412828499827,
+      "learning_rate": 1.6950250025351825e-07,
+      "loss": 0.6257,
+      "step": 5999
+    },
+    {
+      "epoch": 11.11111111111111,
+      "grad_norm": 1.651597664799581,
+      "learning_rate": 1.6949171883185918e-07,
+      "loss": 0.5951,
+      "step": 6000
+    },
+    {
+      "epoch": 11.112962962962962,
+      "grad_norm": 2.7260874808956967,
+      "learning_rate": 1.6948093584783753e-07,
+      "loss": 0.6081,
+      "step": 6001
+    },
+    {
+      "epoch": 11.114814814814816,
+      "grad_norm": 1.610616137968889,
+      "learning_rate": 1.6947015130169576e-07,
+      "loss": 0.6384,
+      "step": 6002
+    },
+    {
+      "epoch": 11.116666666666667,
+      "grad_norm": 1.7683623770826475,
+      "learning_rate": 1.694593651936763e-07,
+      "loss": 0.6259,
+      "step": 6003
+    },
+    {
+      "epoch": 11.118518518518519,
+      "grad_norm": 1.7011868717006917,
+      "learning_rate": 1.6944857752402167e-07,
+      "loss": 0.5941,
+      "step": 6004
+    },
+    {
+      "epoch": 11.12037037037037,
+      "grad_norm": 1.5887048445925007,
+      "learning_rate": 1.694377882929744e-07,
+      "loss": 0.6372,
+      "step": 6005
+    },
+    {
+      "epoch": 11.122222222222222,
+      "grad_norm": 1.576354478728774,
+      "learning_rate": 1.6942699750077706e-07,
+      "loss": 0.6225,
+      "step": 6006
+    },
+    {
+      "epoch": 11.124074074074073,
+      "grad_norm": 1.609499867610394,
+      "learning_rate": 1.6941620514767224e-07,
+      "loss": 0.6076,
+      "step": 6007
+    },
+    {
+      "epoch": 11.125925925925927,
+      "grad_norm": 1.5540112440566092,
+      "learning_rate": 1.694054112339026e-07,
+      "loss": 0.6244,
+      "step": 6008
+    },
+    {
+      "epoch": 11.127777777777778,
+      "grad_norm": 1.6166724770567855,
+      "learning_rate": 1.693946157597108e-07,
+      "loss": 0.571,
+      "step": 6009
+    },
+    {
+      "epoch": 11.12962962962963,
+      "grad_norm": 1.6899808787818047,
+      "learning_rate": 1.6938381872533957e-07,
+      "loss": 0.6218,
+      "step": 6010
+    },
+    {
+      "epoch": 11.131481481481481,
+      "grad_norm": 2.1029401542675665,
+      "learning_rate": 1.693730201310317e-07,
+      "loss": 0.6292,
+      "step": 6011
+    },
+    {
+      "epoch": 11.133333333333333,
+      "grad_norm": 1.7258477364108313,
+      "learning_rate": 1.6936221997702988e-07,
+      "loss": 0.6695,
+      "step": 6012
+    },
+    {
+      "epoch": 11.135185185185184,
+      "grad_norm": 1.559478884145518,
+      "learning_rate": 1.6935141826357699e-07,
+      "loss": 0.6415,
+      "step": 6013
+    },
+    {
+      "epoch": 11.137037037037038,
+      "grad_norm": 1.5928205579841872,
+      "learning_rate": 1.6934061499091582e-07,
+      "loss": 0.6267,
+      "step": 6014
+    },
+    {
+      "epoch": 11.13888888888889,
+      "grad_norm": 1.4078854556125695,
+      "learning_rate": 1.6932981015928936e-07,
+      "loss": 0.6024,
+      "step": 6015
+    },
+    {
+      "epoch": 11.14074074074074,
+      "grad_norm": 1.492493794216963,
+      "learning_rate": 1.6931900376894042e-07,
+      "loss": 0.6426,
+      "step": 6016
+    },
+    {
+      "epoch": 11.142592592592592,
+      "grad_norm": 1.7437166584427275,
+      "learning_rate": 1.6930819582011203e-07,
+      "loss": 0.5964,
+      "step": 6017
+    },
+    {
+      "epoch": 11.144444444444444,
+      "grad_norm": 2.081434451286772,
+      "learning_rate": 1.6929738631304716e-07,
+      "loss": 0.6193,
+      "step": 6018
+    },
+    {
+      "epoch": 11.146296296296295,
+      "grad_norm": 1.5904257002379132,
+      "learning_rate": 1.6928657524798882e-07,
+      "loss": 0.64,
+      "step": 6019
+    },
+    {
+      "epoch": 11.148148148148149,
+      "grad_norm": 1.651310416420862,
+      "learning_rate": 1.692757626251801e-07,
+      "loss": 0.6465,
+      "step": 6020
+    },
+    {
+      "epoch": 11.15,
+      "grad_norm": 1.5319695012607,
+      "learning_rate": 1.692649484448641e-07,
+      "loss": 0.6729,
+      "step": 6021
+    },
+    {
+      "epoch": 11.151851851851852,
+      "grad_norm": 1.604386973204612,
+      "learning_rate": 1.6925413270728394e-07,
+      "loss": 0.5817,
+      "step": 6022
+    },
+    {
+      "epoch": 11.153703703703703,
+      "grad_norm": 1.623496519385197,
+      "learning_rate": 1.6924331541268274e-07,
+      "loss": 0.6162,
+      "step": 6023
+    },
+    {
+      "epoch": 11.155555555555555,
+      "grad_norm": 1.6107820446110659,
+      "learning_rate": 1.692324965613038e-07,
+      "loss": 0.6301,
+      "step": 6024
+    },
+    {
+      "epoch": 11.157407407407407,
+      "grad_norm": 1.881423819384251,
+      "learning_rate": 1.6922167615339028e-07,
+      "loss": 0.6136,
+      "step": 6025
+    },
+    {
+      "epoch": 11.15925925925926,
+      "grad_norm": 1.7450790994419552,
+      "learning_rate": 1.6921085418918546e-07,
+      "loss": 0.6168,
+      "step": 6026
+    },
+    {
+      "epoch": 11.161111111111111,
+      "grad_norm": 3.034118473800667,
+      "learning_rate": 1.6920003066893269e-07,
+      "loss": 0.6211,
+      "step": 6027
+    },
+    {
+      "epoch": 11.162962962962963,
+      "grad_norm": 1.538125211259693,
+      "learning_rate": 1.6918920559287528e-07,
+      "loss": 0.6375,
+      "step": 6028
+    },
+    {
+      "epoch": 11.164814814814815,
+      "grad_norm": 2.0464684899286696,
+      "learning_rate": 1.6917837896125661e-07,
+      "loss": 0.6308,
+      "step": 6029
+    },
+    {
+      "epoch": 11.166666666666666,
+      "grad_norm": 1.5928481151469456,
+      "learning_rate": 1.6916755077432012e-07,
+      "loss": 0.6434,
+      "step": 6030
+    },
+    {
+      "epoch": 11.168518518518518,
+      "grad_norm": 1.6956289759861047,
+      "learning_rate": 1.691567210323092e-07,
+      "loss": 0.6713,
+      "step": 6031
+    },
+    {
+      "epoch": 11.170370370370371,
+      "grad_norm": 2.1460977688328287,
+      "learning_rate": 1.6914588973546737e-07,
+      "loss": 0.6309,
+      "step": 6032
+    },
+    {
+      "epoch": 11.172222222222222,
+      "grad_norm": 1.4513230539235706,
+      "learning_rate": 1.6913505688403818e-07,
+      "loss": 0.5987,
+      "step": 6033
+    },
+    {
+      "epoch": 11.174074074074074,
+      "grad_norm": 1.6312318700342592,
+      "learning_rate": 1.6912422247826507e-07,
+      "loss": 0.6139,
+      "step": 6034
+    },
+    {
+      "epoch": 11.175925925925926,
+      "grad_norm": 1.5773566082549608,
+      "learning_rate": 1.6911338651839174e-07,
+      "loss": 0.5903,
+      "step": 6035
+    },
+    {
+      "epoch": 11.177777777777777,
+      "grad_norm": 1.6141722374712115,
+      "learning_rate": 1.691025490046618e-07,
+      "loss": 0.6339,
+      "step": 6036
+    },
+    {
+      "epoch": 11.17962962962963,
+      "grad_norm": 1.8830565967122879,
+      "learning_rate": 1.6909170993731882e-07,
+      "loss": 0.618,
+      "step": 6037
+    },
+    {
+      "epoch": 11.181481481481482,
+      "grad_norm": 1.8457149045199799,
+      "learning_rate": 1.690808693166066e-07,
+      "loss": 0.6602,
+      "step": 6038
+    },
+    {
+      "epoch": 11.183333333333334,
+      "grad_norm": 1.6331941515412243,
+      "learning_rate": 1.690700271427688e-07,
+      "loss": 0.6108,
+      "step": 6039
+    },
+    {
+      "epoch": 11.185185185185185,
+      "grad_norm": 1.5435956183842898,
+      "learning_rate": 1.690591834160492e-07,
+      "loss": 0.6495,
+      "step": 6040
+    },
+    {
+      "epoch": 11.187037037037037,
+      "grad_norm": 1.6286746877249632,
+      "learning_rate": 1.6904833813669156e-07,
+      "loss": 0.624,
+      "step": 6041
+    },
+    {
+      "epoch": 11.188888888888888,
+      "grad_norm": 1.4693250721860047,
+      "learning_rate": 1.6903749130493982e-07,
+      "loss": 0.6234,
+      "step": 6042
+    },
+    {
+      "epoch": 11.190740740740742,
+      "grad_norm": 1.7945023863999972,
+      "learning_rate": 1.6902664292103773e-07,
+      "loss": 0.6385,
+      "step": 6043
+    },
+    {
+      "epoch": 11.192592592592593,
+      "grad_norm": 1.7565048693465957,
+      "learning_rate": 1.690157929852292e-07,
+      "loss": 0.5955,
+      "step": 6044
+    },
+    {
+      "epoch": 11.194444444444445,
+      "grad_norm": 1.7556448003776346,
+      "learning_rate": 1.6900494149775827e-07,
+      "loss": 0.6122,
+      "step": 6045
+    },
+    {
+      "epoch": 11.196296296296296,
+      "grad_norm": 1.5193049872043598,
+      "learning_rate": 1.689940884588688e-07,
+      "loss": 0.6485,
+      "step": 6046
+    },
+    {
+      "epoch": 11.198148148148148,
+      "grad_norm": 1.3727336371787522,
+      "learning_rate": 1.6898323386880481e-07,
+      "loss": 0.6386,
+      "step": 6047
+    },
+    {
+      "epoch": 11.2,
+      "grad_norm": 1.325615420928115,
+      "learning_rate": 1.6897237772781044e-07,
+      "loss": 0.619,
+      "step": 6048
+    },
+    {
+      "epoch": 11.201851851851853,
+      "grad_norm": 1.7221032047844296,
+      "learning_rate": 1.6896152003612961e-07,
+      "loss": 0.6392,
+      "step": 6049
+    },
+    {
+      "epoch": 11.203703703703704,
+      "grad_norm": 2.2990644557164694,
+      "learning_rate": 1.6895066079400657e-07,
+      "loss": 0.6027,
+      "step": 6050
+    },
+    {
+      "epoch": 11.205555555555556,
+      "grad_norm": 1.7923598744216604,
+      "learning_rate": 1.6893980000168538e-07,
+      "loss": 0.5988,
+      "step": 6051
+    },
+    {
+      "epoch": 11.207407407407407,
+      "grad_norm": 1.4668453900910225,
+      "learning_rate": 1.6892893765941025e-07,
+      "loss": 0.6147,
+      "step": 6052
+    },
+    {
+      "epoch": 11.209259259259259,
+      "grad_norm": 1.4629318980876314,
+      "learning_rate": 1.689180737674254e-07,
+      "loss": 0.6128,
+      "step": 6053
+    },
+    {
+      "epoch": 11.21111111111111,
+      "grad_norm": 1.588515348146373,
+      "learning_rate": 1.689072083259751e-07,
+      "loss": 0.6025,
+      "step": 6054
+    },
+    {
+      "epoch": 11.212962962962964,
+      "grad_norm": 1.4833052016151294,
+      "learning_rate": 1.688963413353036e-07,
+      "loss": 0.5896,
+      "step": 6055
+    },
+    {
+      "epoch": 11.214814814814815,
+      "grad_norm": 2.245265316933493,
+      "learning_rate": 1.688854727956552e-07,
+      "loss": 0.5985,
+      "step": 6056
+    },
+    {
+      "epoch": 11.216666666666667,
+      "grad_norm": 1.6043734876256155,
+      "learning_rate": 1.6887460270727428e-07,
+      "loss": 0.6847,
+      "step": 6057
+    },
+    {
+      "epoch": 11.218518518518518,
+      "grad_norm": 1.6640803034910947,
+      "learning_rate": 1.6886373107040525e-07,
+      "loss": 0.6216,
+      "step": 6058
+    },
+    {
+      "epoch": 11.22037037037037,
+      "grad_norm": 1.6486378694629042,
+      "learning_rate": 1.6885285788529252e-07,
+      "loss": 0.62,
+      "step": 6059
+    },
+    {
+      "epoch": 11.222222222222221,
+      "grad_norm": 1.5718836868416153,
+      "learning_rate": 1.6884198315218052e-07,
+      "loss": 0.5949,
+      "step": 6060
+    },
+    {
+      "epoch": 11.224074074074075,
+      "grad_norm": 1.5495588270168725,
+      "learning_rate": 1.6883110687131378e-07,
+      "loss": 0.6695,
+      "step": 6061
+    },
+    {
+      "epoch": 11.225925925925926,
+      "grad_norm": 1.5390156570672184,
+      "learning_rate": 1.688202290429368e-07,
+      "loss": 0.6187,
+      "step": 6062
+    },
+    {
+      "epoch": 11.227777777777778,
+      "grad_norm": 1.876350223610275,
+      "learning_rate": 1.6880934966729415e-07,
+      "loss": 0.5815,
+      "step": 6063
+    },
+    {
+      "epoch": 11.22962962962963,
+      "grad_norm": 1.6117198714866232,
+      "learning_rate": 1.6879846874463048e-07,
+      "loss": 0.6148,
+      "step": 6064
+    },
+    {
+      "epoch": 11.231481481481481,
+      "grad_norm": 1.4746371150648672,
+      "learning_rate": 1.6878758627519036e-07,
+      "loss": 0.6279,
+      "step": 6065
+    },
+    {
+      "epoch": 11.233333333333333,
+      "grad_norm": 1.5090741366129812,
+      "learning_rate": 1.6877670225921845e-07,
+      "loss": 0.5765,
+      "step": 6066
+    },
+    {
+      "epoch": 11.235185185185186,
+      "grad_norm": 1.3697860633718788,
+      "learning_rate": 1.6876581669695953e-07,
+      "loss": 0.5757,
+      "step": 6067
+    },
+    {
+      "epoch": 11.237037037037037,
+      "grad_norm": 1.9070652412239084,
+      "learning_rate": 1.6875492958865824e-07,
+      "loss": 0.6314,
+      "step": 6068
+    },
+    {
+      "epoch": 11.238888888888889,
+      "grad_norm": 1.443061466517777,
+      "learning_rate": 1.6874404093455943e-07,
+      "loss": 0.6683,
+      "step": 6069
+    },
+    {
+      "epoch": 11.24074074074074,
+      "grad_norm": 1.5473290418442325,
+      "learning_rate": 1.6873315073490786e-07,
+      "loss": 0.6192,
+      "step": 6070
+    },
+    {
+      "epoch": 11.242592592592592,
+      "grad_norm": 1.877640418874891,
+      "learning_rate": 1.6872225898994837e-07,
+      "loss": 0.6346,
+      "step": 6071
+    },
+    {
+      "epoch": 11.244444444444444,
+      "grad_norm": 1.425461386317159,
+      "learning_rate": 1.6871136569992586e-07,
+      "loss": 0.6857,
+      "step": 6072
+    },
+    {
+      "epoch": 11.246296296296297,
+      "grad_norm": 1.5116383844670422,
+      "learning_rate": 1.6870047086508526e-07,
+      "loss": 0.6308,
+      "step": 6073
+    },
+    {
+      "epoch": 11.248148148148148,
+      "grad_norm": 1.5012634895031323,
+      "learning_rate": 1.6868957448567146e-07,
+      "loss": 0.6155,
+      "step": 6074
+    },
+    {
+      "epoch": 11.25,
+      "grad_norm": 1.9203060105502727,
+      "learning_rate": 1.6867867656192945e-07,
+      "loss": 0.5898,
+      "step": 6075
+    },
+    {
+      "epoch": 11.251851851851852,
+      "grad_norm": 1.4486571292866743,
+      "learning_rate": 1.686677770941043e-07,
+      "loss": 0.6292,
+      "step": 6076
+    },
+    {
+      "epoch": 11.253703703703703,
+      "grad_norm": 1.436259377297359,
+      "learning_rate": 1.68656876082441e-07,
+      "loss": 0.6217,
+      "step": 6077
+    },
+    {
+      "epoch": 11.255555555555556,
+      "grad_norm": 1.5332591755143443,
+      "learning_rate": 1.6864597352718467e-07,
+      "loss": 0.6712,
+      "step": 6078
+    },
+    {
+      "epoch": 11.257407407407408,
+      "grad_norm": 1.568702262379678,
+      "learning_rate": 1.686350694285804e-07,
+      "loss": 0.6143,
+      "step": 6079
+    },
+    {
+      "epoch": 11.25925925925926,
+      "grad_norm": 1.5198962337249173,
+      "learning_rate": 1.6862416378687336e-07,
+      "loss": 0.6492,
+      "step": 6080
+    },
+    {
+      "epoch": 11.261111111111111,
+      "grad_norm": 1.4832586179153602,
+      "learning_rate": 1.6861325660230872e-07,
+      "loss": 0.6537,
+      "step": 6081
+    },
+    {
+      "epoch": 11.262962962962963,
+      "grad_norm": 1.6735327668796576,
+      "learning_rate": 1.6860234787513176e-07,
+      "loss": 0.6396,
+      "step": 6082
+    },
+    {
+      "epoch": 11.264814814814814,
+      "grad_norm": 1.497741682575467,
+      "learning_rate": 1.685914376055877e-07,
+      "loss": 0.6042,
+      "step": 6083
+    },
+    {
+      "epoch": 11.266666666666667,
+      "grad_norm": 1.6488951038822184,
+      "learning_rate": 1.685805257939218e-07,
+      "loss": 0.6068,
+      "step": 6084
+    },
+    {
+      "epoch": 11.268518518518519,
+      "grad_norm": 2.1806082758097993,
+      "learning_rate": 1.6856961244037945e-07,
+      "loss": 0.6594,
+      "step": 6085
+    },
+    {
+      "epoch": 11.27037037037037,
+      "grad_norm": 2.5688896822217604,
+      "learning_rate": 1.6855869754520592e-07,
+      "loss": 0.5993,
+      "step": 6086
+    },
+    {
+      "epoch": 11.272222222222222,
+      "grad_norm": 1.4049906507190306,
+      "learning_rate": 1.6854778110864673e-07,
+      "loss": 0.6356,
+      "step": 6087
+    },
+    {
+      "epoch": 11.274074074074074,
+      "grad_norm": 1.6487246924078636,
+      "learning_rate": 1.6853686313094722e-07,
+      "loss": 0.6563,
+      "step": 6088
+    },
+    {
+      "epoch": 11.275925925925925,
+      "grad_norm": 1.694617583194286,
+      "learning_rate": 1.685259436123529e-07,
+      "loss": 0.67,
+      "step": 6089
+    },
+    {
+      "epoch": 11.277777777777779,
+      "grad_norm": 1.8596727786777985,
+      "learning_rate": 1.6851502255310923e-07,
+      "loss": 0.6749,
+      "step": 6090
+    },
+    {
+      "epoch": 11.27962962962963,
+      "grad_norm": 1.570895438672205,
+      "learning_rate": 1.685040999534618e-07,
+      "loss": 0.6384,
+      "step": 6091
+    },
+    {
+      "epoch": 11.281481481481482,
+      "grad_norm": 1.6792584951084863,
+      "learning_rate": 1.6849317581365612e-07,
+      "loss": 0.6068,
+      "step": 6092
+    },
+    {
+      "epoch": 11.283333333333333,
+      "grad_norm": 1.6733517599722272,
+      "learning_rate": 1.6848225013393785e-07,
+      "loss": 0.6563,
+      "step": 6093
+    },
+    {
+      "epoch": 11.285185185185185,
+      "grad_norm": 1.538314953048454,
+      "learning_rate": 1.6847132291455257e-07,
+      "loss": 0.6499,
+      "step": 6094
+    },
+    {
+      "epoch": 11.287037037037036,
+      "grad_norm": 1.553033337510129,
+      "learning_rate": 1.68460394155746e-07,
+      "loss": 0.6426,
+      "step": 6095
+    },
+    {
+      "epoch": 11.28888888888889,
+      "grad_norm": 1.5090041667604468,
+      "learning_rate": 1.6844946385776384e-07,
+      "loss": 0.6343,
+      "step": 6096
+    },
+    {
+      "epoch": 11.290740740740741,
+      "grad_norm": 1.917276757344583,
+      "learning_rate": 1.6843853202085182e-07,
+      "loss": 0.6182,
+      "step": 6097
+    },
+    {
+      "epoch": 11.292592592592593,
+      "grad_norm": 1.4630878373724512,
+      "learning_rate": 1.6842759864525573e-07,
+      "loss": 0.6098,
+      "step": 6098
+    },
+    {
+      "epoch": 11.294444444444444,
+      "grad_norm": 1.5734014950144677,
+      "learning_rate": 1.6841666373122138e-07,
+      "loss": 0.6259,
+      "step": 6099
+    },
+    {
+      "epoch": 11.296296296296296,
+      "grad_norm": 1.6067452972836314,
+      "learning_rate": 1.684057272789946e-07,
+      "loss": 0.6042,
+      "step": 6100
+    },
+    {
+      "epoch": 11.298148148148147,
+      "grad_norm": 1.5320340246940198,
+      "learning_rate": 1.683947892888213e-07,
+      "loss": 0.5977,
+      "step": 6101
+    },
+    {
+      "epoch": 11.3,
+      "grad_norm": 1.8003599727709685,
+      "learning_rate": 1.6838384976094736e-07,
+      "loss": 0.6112,
+      "step": 6102
+    },
+    {
+      "epoch": 11.301851851851852,
+      "grad_norm": 2.138378519944151,
+      "learning_rate": 1.6837290869561875e-07,
+      "loss": 0.6351,
+      "step": 6103
+    },
+    {
+      "epoch": 11.303703703703704,
+      "grad_norm": 1.5250346964659607,
+      "learning_rate": 1.6836196609308147e-07,
+      "loss": 0.6307,
+      "step": 6104
+    },
+    {
+      "epoch": 11.305555555555555,
+      "grad_norm": 1.5296776369039202,
+      "learning_rate": 1.683510219535815e-07,
+      "loss": 0.6403,
+      "step": 6105
+    },
+    {
+      "epoch": 11.307407407407407,
+      "grad_norm": 1.6509854862658215,
+      "learning_rate": 1.6834007627736496e-07,
+      "loss": 0.6304,
+      "step": 6106
+    },
+    {
+      "epoch": 11.309259259259258,
+      "grad_norm": 1.5471030484463637,
+      "learning_rate": 1.6832912906467786e-07,
+      "loss": 0.6282,
+      "step": 6107
+    },
+    {
+      "epoch": 11.311111111111112,
+      "grad_norm": 1.6338150097128694,
+      "learning_rate": 1.6831818031576637e-07,
+      "loss": 0.6256,
+      "step": 6108
+    },
+    {
+      "epoch": 11.312962962962963,
+      "grad_norm": 2.1878779858980777,
+      "learning_rate": 1.6830723003087664e-07,
+      "loss": 0.649,
+      "step": 6109
+    },
+    {
+      "epoch": 11.314814814814815,
+      "grad_norm": 1.4498875861507257,
+      "learning_rate": 1.6829627821025488e-07,
+      "loss": 0.6225,
+      "step": 6110
+    },
+    {
+      "epoch": 11.316666666666666,
+      "grad_norm": 2.785420159099274,
+      "learning_rate": 1.6828532485414724e-07,
+      "loss": 0.6413,
+      "step": 6111
+    },
+    {
+      "epoch": 11.318518518518518,
+      "grad_norm": 1.7407396121593037,
+      "learning_rate": 1.6827436996280007e-07,
+      "loss": 0.6234,
+      "step": 6112
+    },
+    {
+      "epoch": 11.32037037037037,
+      "grad_norm": 1.8272448599153395,
+      "learning_rate": 1.6826341353645965e-07,
+      "loss": 0.5869,
+      "step": 6113
+    },
+    {
+      "epoch": 11.322222222222223,
+      "grad_norm": 1.7073131512091284,
+      "learning_rate": 1.6825245557537228e-07,
+      "loss": 0.6266,
+      "step": 6114
+    },
+    {
+      "epoch": 11.324074074074074,
+      "grad_norm": 2.004167948563305,
+      "learning_rate": 1.6824149607978434e-07,
+      "loss": 0.6524,
+      "step": 6115
+    },
+    {
+      "epoch": 11.325925925925926,
+      "grad_norm": 1.5501093589508403,
+      "learning_rate": 1.6823053504994223e-07,
+      "loss": 0.6034,
+      "step": 6116
+    },
+    {
+      "epoch": 11.327777777777778,
+      "grad_norm": 1.7120958481742412,
+      "learning_rate": 1.6821957248609232e-07,
+      "loss": 0.6271,
+      "step": 6117
+    },
+    {
+      "epoch": 11.329629629629629,
+      "grad_norm": 1.6471374294874674,
+      "learning_rate": 1.6820860838848123e-07,
+      "loss": 0.6021,
+      "step": 6118
+    },
+    {
+      "epoch": 11.331481481481482,
+      "grad_norm": 1.6595997833019283,
+      "learning_rate": 1.6819764275735533e-07,
+      "loss": 0.6322,
+      "step": 6119
+    },
+    {
+      "epoch": 11.333333333333334,
+      "grad_norm": 1.487895686112585,
+      "learning_rate": 1.681866755929612e-07,
+      "loss": 0.6074,
+      "step": 6120
+    },
+    {
+      "epoch": 11.335185185185185,
+      "grad_norm": 1.5076138127711285,
+      "learning_rate": 1.6817570689554539e-07,
+      "loss": 0.6221,
+      "step": 6121
+    },
+    {
+      "epoch": 11.337037037037037,
+      "grad_norm": 1.3930525592568264,
+      "learning_rate": 1.6816473666535456e-07,
+      "loss": 0.6103,
+      "step": 6122
+    },
+    {
+      "epoch": 11.338888888888889,
+      "grad_norm": 1.8740855015633813,
+      "learning_rate": 1.6815376490263532e-07,
+      "loss": 0.6063,
+      "step": 6123
+    },
+    {
+      "epoch": 11.34074074074074,
+      "grad_norm": 1.4405692945863406,
+      "learning_rate": 1.6814279160763429e-07,
+      "loss": 0.6441,
+      "step": 6124
+    },
+    {
+      "epoch": 11.342592592592593,
+      "grad_norm": 1.6457619080252759,
+      "learning_rate": 1.6813181678059828e-07,
+      "loss": 0.5759,
+      "step": 6125
+    },
+    {
+      "epoch": 11.344444444444445,
+      "grad_norm": 1.7675443842494571,
+      "learning_rate": 1.68120840421774e-07,
+      "loss": 0.5969,
+      "step": 6126
+    },
+    {
+      "epoch": 11.346296296296297,
+      "grad_norm": 1.6809304572503099,
+      "learning_rate": 1.6810986253140819e-07,
+      "loss": 0.6271,
+      "step": 6127
+    },
+    {
+      "epoch": 11.348148148148148,
+      "grad_norm": 1.5544472841501098,
+      "learning_rate": 1.6809888310974768e-07,
+      "loss": 0.6307,
+      "step": 6128
+    },
+    {
+      "epoch": 11.35,
+      "grad_norm": 1.894822185973263,
+      "learning_rate": 1.6808790215703933e-07,
+      "loss": 0.6596,
+      "step": 6129
+    },
+    {
+      "epoch": 11.351851851851851,
+      "grad_norm": 3.549222274338036,
+      "learning_rate": 1.6807691967353003e-07,
+      "loss": 0.6223,
+      "step": 6130
+    },
+    {
+      "epoch": 11.353703703703705,
+      "grad_norm": 3.347767675781517,
+      "learning_rate": 1.6806593565946665e-07,
+      "loss": 0.6302,
+      "step": 6131
+    },
+    {
+      "epoch": 11.355555555555556,
+      "grad_norm": 1.7769978880340598,
+      "learning_rate": 1.6805495011509617e-07,
+      "loss": 0.6655,
+      "step": 6132
+    },
+    {
+      "epoch": 11.357407407407408,
+      "grad_norm": 1.4754736066991214,
+      "learning_rate": 1.680439630406656e-07,
+      "loss": 0.604,
+      "step": 6133
+    },
+    {
+      "epoch": 11.35925925925926,
+      "grad_norm": 1.5397310461845914,
+      "learning_rate": 1.6803297443642194e-07,
+      "loss": 0.6106,
+      "step": 6134
+    },
+    {
+      "epoch": 11.36111111111111,
+      "grad_norm": 2.111810597903459,
+      "learning_rate": 1.6802198430261222e-07,
+      "loss": 0.6363,
+      "step": 6135
+    },
+    {
+      "epoch": 11.362962962962962,
+      "grad_norm": 2.2462278560262567,
+      "learning_rate": 1.6801099263948357e-07,
+      "loss": 0.6385,
+      "step": 6136
+    },
+    {
+      "epoch": 11.364814814814816,
+      "grad_norm": 1.591875192794211,
+      "learning_rate": 1.6799999944728307e-07,
+      "loss": 0.6142,
+      "step": 6137
+    },
+    {
+      "epoch": 11.366666666666667,
+      "grad_norm": 1.4351378320019608,
+      "learning_rate": 1.679890047262579e-07,
+      "loss": 0.6516,
+      "step": 6138
+    },
+    {
+      "epoch": 11.368518518518519,
+      "grad_norm": 1.4311572403363906,
+      "learning_rate": 1.6797800847665527e-07,
+      "loss": 0.636,
+      "step": 6139
+    },
+    {
+      "epoch": 11.37037037037037,
+      "grad_norm": 1.3740052489398664,
+      "learning_rate": 1.6796701069872237e-07,
+      "loss": 0.5795,
+      "step": 6140
+    },
+    {
+      "epoch": 11.372222222222222,
+      "grad_norm": 1.5811911655716102,
+      "learning_rate": 1.6795601139270646e-07,
+      "loss": 0.5956,
+      "step": 6141
+    },
+    {
+      "epoch": 11.374074074074073,
+      "grad_norm": 1.4595634081644167,
+      "learning_rate": 1.6794501055885486e-07,
+      "loss": 0.6327,
+      "step": 6142
+    },
+    {
+      "epoch": 11.375925925925927,
+      "grad_norm": 2.65516898242403,
+      "learning_rate": 1.679340081974149e-07,
+      "loss": 0.6157,
+      "step": 6143
+    },
+    {
+      "epoch": 11.377777777777778,
+      "grad_norm": 1.6217768608074086,
+      "learning_rate": 1.6792300430863393e-07,
+      "loss": 0.5961,
+      "step": 6144
+    },
+    {
+      "epoch": 11.37962962962963,
+      "grad_norm": 1.5262965329906493,
+      "learning_rate": 1.679119988927593e-07,
+      "loss": 0.5999,
+      "step": 6145
+    },
+    {
+      "epoch": 11.381481481481481,
+      "grad_norm": 1.4136164959799644,
+      "learning_rate": 1.6790099195003854e-07,
+      "loss": 0.6184,
+      "step": 6146
+    },
+    {
+      "epoch": 11.383333333333333,
+      "grad_norm": 1.558848943983646,
+      "learning_rate": 1.6788998348071904e-07,
+      "loss": 0.7145,
+      "step": 6147
+    },
+    {
+      "epoch": 11.385185185185184,
+      "grad_norm": 1.8412967081728602,
+      "learning_rate": 1.6787897348504836e-07,
+      "loss": 0.6834,
+      "step": 6148
+    },
+    {
+      "epoch": 11.387037037037038,
+      "grad_norm": 1.5336734531670488,
+      "learning_rate": 1.6786796196327397e-07,
+      "loss": 0.6088,
+      "step": 6149
+    },
+    {
+      "epoch": 11.38888888888889,
+      "grad_norm": 4.010885856768812,
+      "learning_rate": 1.6785694891564345e-07,
+      "loss": 0.672,
+      "step": 6150
+    },
+    {
+      "epoch": 11.39074074074074,
+      "grad_norm": 1.4811792940981883,
+      "learning_rate": 1.6784593434240447e-07,
+      "loss": 0.6105,
+      "step": 6151
+    },
+    {
+      "epoch": 11.392592592592592,
+      "grad_norm": 1.9036486955386922,
+      "learning_rate": 1.6783491824380458e-07,
+      "loss": 0.5868,
+      "step": 6152
+    },
+    {
+      "epoch": 11.394444444444444,
+      "grad_norm": 1.458907816283198,
+      "learning_rate": 1.6782390062009148e-07,
+      "loss": 0.6029,
+      "step": 6153
+    },
+    {
+      "epoch": 11.396296296296295,
+      "grad_norm": 1.9278317363167718,
+      "learning_rate": 1.6781288147151295e-07,
+      "loss": 0.6459,
+      "step": 6154
+    },
+    {
+      "epoch": 11.398148148148149,
+      "grad_norm": 1.4924844805487636,
+      "learning_rate": 1.6780186079831662e-07,
+      "loss": 0.6504,
+      "step": 6155
+    },
+    {
+      "epoch": 11.4,
+      "grad_norm": 1.8339537601646378,
+      "learning_rate": 1.6779083860075033e-07,
+      "loss": 0.6086,
+      "step": 6156
+    },
+    {
+      "epoch": 11.401851851851852,
+      "grad_norm": 1.495523066543893,
+      "learning_rate": 1.6777981487906185e-07,
+      "loss": 0.5938,
+      "step": 6157
+    },
+    {
+      "epoch": 11.403703703703703,
+      "grad_norm": 1.5234592004415843,
+      "learning_rate": 1.6776878963349907e-07,
+      "loss": 0.6172,
+      "step": 6158
+    },
+    {
+      "epoch": 11.405555555555555,
+      "grad_norm": 1.525890855458275,
+      "learning_rate": 1.677577628643098e-07,
+      "loss": 0.6087,
+      "step": 6159
+    },
+    {
+      "epoch": 11.407407407407407,
+      "grad_norm": 1.5035218318707106,
+      "learning_rate": 1.6774673457174204e-07,
+      "loss": 0.6284,
+      "step": 6160
+    },
+    {
+      "epoch": 11.40925925925926,
+      "grad_norm": 2.3349176082004197,
+      "learning_rate": 1.6773570475604366e-07,
+      "loss": 0.62,
+      "step": 6161
+    },
+    {
+      "epoch": 11.411111111111111,
+      "grad_norm": 1.4882429513321542,
+      "learning_rate": 1.6772467341746266e-07,
+      "loss": 0.6622,
+      "step": 6162
+    },
+    {
+      "epoch": 11.412962962962963,
+      "grad_norm": 1.4547862053085523,
+      "learning_rate": 1.677136405562471e-07,
+      "loss": 0.599,
+      "step": 6163
+    },
+    {
+      "epoch": 11.414814814814815,
+      "grad_norm": 1.492584721341252,
+      "learning_rate": 1.6770260617264498e-07,
+      "loss": 0.6381,
+      "step": 6164
+    },
+    {
+      "epoch": 11.416666666666666,
+      "grad_norm": 1.569951233310649,
+      "learning_rate": 1.676915702669044e-07,
+      "loss": 0.6676,
+      "step": 6165
+    },
+    {
+      "epoch": 11.418518518518518,
+      "grad_norm": 1.6098508910170555,
+      "learning_rate": 1.6768053283927346e-07,
+      "loss": 0.628,
+      "step": 6166
+    },
+    {
+      "epoch": 11.420370370370371,
+      "grad_norm": 1.4088206554000426,
+      "learning_rate": 1.6766949389000034e-07,
+      "loss": 0.6037,
+      "step": 6167
+    },
+    {
+      "epoch": 11.422222222222222,
+      "grad_norm": 1.382604391730993,
+      "learning_rate": 1.6765845341933317e-07,
+      "loss": 0.5952,
+      "step": 6168
+    },
+    {
+      "epoch": 11.424074074074074,
+      "grad_norm": 1.719526958831739,
+      "learning_rate": 1.6764741142752027e-07,
+      "loss": 0.6198,
+      "step": 6169
+    },
+    {
+      "epoch": 11.425925925925926,
+      "grad_norm": 1.7198045333229666,
+      "learning_rate": 1.6763636791480978e-07,
+      "loss": 0.6039,
+      "step": 6170
+    },
+    {
+      "epoch": 11.427777777777777,
+      "grad_norm": 1.445981010339226,
+      "learning_rate": 1.676253228814501e-07,
+      "loss": 0.5835,
+      "step": 6171
+    },
+    {
+      "epoch": 11.42962962962963,
+      "grad_norm": 1.5541143313150898,
+      "learning_rate": 1.6761427632768944e-07,
+      "loss": 0.5839,
+      "step": 6172
+    },
+    {
+      "epoch": 11.431481481481482,
+      "grad_norm": 1.6137516598086628,
+      "learning_rate": 1.676032282537762e-07,
+      "loss": 0.5998,
+      "step": 6173
+    },
+    {
+      "epoch": 11.433333333333334,
+      "grad_norm": 1.559032224186864,
+      "learning_rate": 1.6759217865995883e-07,
+      "loss": 0.6313,
+      "step": 6174
+    },
+    {
+      "epoch": 11.435185185185185,
+      "grad_norm": 1.588872350946832,
+      "learning_rate": 1.675811275464857e-07,
+      "loss": 0.632,
+      "step": 6175
+    },
+    {
+      "epoch": 11.437037037037037,
+      "grad_norm": 1.5727046238120939,
+      "learning_rate": 1.6757007491360525e-07,
+      "loss": 0.5951,
+      "step": 6176
+    },
+    {
+      "epoch": 11.438888888888888,
+      "grad_norm": 1.4631903414229273,
+      "learning_rate": 1.6755902076156602e-07,
+      "loss": 0.6328,
+      "step": 6177
+    },
+    {
+      "epoch": 11.440740740740742,
+      "grad_norm": 1.8119439849567267,
+      "learning_rate": 1.675479650906165e-07,
+      "loss": 0.5837,
+      "step": 6178
+    },
+    {
+      "epoch": 11.442592592592593,
+      "grad_norm": 1.4689622619878557,
+      "learning_rate": 1.6753690790100528e-07,
+      "loss": 0.6366,
+      "step": 6179
+    },
+    {
+      "epoch": 11.444444444444445,
+      "grad_norm": 1.372096008281409,
+      "learning_rate": 1.6752584919298091e-07,
+      "loss": 0.6041,
+      "step": 6180
+    },
+    {
+      "epoch": 11.446296296296296,
+      "grad_norm": 1.3933387475011356,
+      "learning_rate": 1.675147889667921e-07,
+      "loss": 0.6458,
+      "step": 6181
+    },
+    {
+      "epoch": 11.448148148148148,
+      "grad_norm": 1.8071017325743397,
+      "learning_rate": 1.6750372722268743e-07,
+      "loss": 0.6249,
+      "step": 6182
+    },
+    {
+      "epoch": 11.45,
+      "grad_norm": 1.4771975121232566,
+      "learning_rate": 1.6749266396091568e-07,
+      "loss": 0.6295,
+      "step": 6183
+    },
+    {
+      "epoch": 11.451851851851853,
+      "grad_norm": 1.7932795890712372,
+      "learning_rate": 1.674815991817255e-07,
+      "loss": 0.6615,
+      "step": 6184
+    },
+    {
+      "epoch": 11.453703703703704,
+      "grad_norm": 1.6257021907341753,
+      "learning_rate": 1.674705328853657e-07,
+      "loss": 0.6371,
+      "step": 6185
+    },
+    {
+      "epoch": 11.455555555555556,
+      "grad_norm": 1.5260879723417615,
+      "learning_rate": 1.6745946507208508e-07,
+      "loss": 0.6589,
+      "step": 6186
+    },
+    {
+      "epoch": 11.457407407407407,
+      "grad_norm": 1.884503395385978,
+      "learning_rate": 1.6744839574213248e-07,
+      "loss": 0.6101,
+      "step": 6187
+    },
+    {
+      "epoch": 11.459259259259259,
+      "grad_norm": 1.3965438083799675,
+      "learning_rate": 1.674373248957567e-07,
+      "loss": 0.6009,
+      "step": 6188
+    },
+    {
+      "epoch": 11.46111111111111,
+      "grad_norm": 2.144178611322647,
+      "learning_rate": 1.6742625253320676e-07,
+      "loss": 0.6172,
+      "step": 6189
+    },
+    {
+      "epoch": 11.462962962962964,
+      "grad_norm": 1.4520723812939664,
+      "learning_rate": 1.674151786547315e-07,
+      "loss": 0.627,
+      "step": 6190
+    },
+    {
+      "epoch": 11.464814814814815,
+      "grad_norm": 1.6803221694476724,
+      "learning_rate": 1.6740410326057993e-07,
+      "loss": 0.6103,
+      "step": 6191
+    },
+    {
+      "epoch": 11.466666666666667,
+      "grad_norm": 1.8766644744783456,
+      "learning_rate": 1.6739302635100107e-07,
+      "loss": 0.5688,
+      "step": 6192
+    },
+    {
+      "epoch": 11.468518518518518,
+      "grad_norm": 1.6000935884917922,
+      "learning_rate": 1.6738194792624395e-07,
+      "loss": 0.594,
+      "step": 6193
+    },
+    {
+      "epoch": 11.47037037037037,
+      "grad_norm": 1.9444780893315359,
+      "learning_rate": 1.673708679865576e-07,
+      "loss": 0.6138,
+      "step": 6194
+    },
+    {
+      "epoch": 11.472222222222221,
+      "grad_norm": 1.3439450826248156,
+      "learning_rate": 1.6735978653219117e-07,
+      "loss": 0.6113,
+      "step": 6195
+    },
+    {
+      "epoch": 11.474074074074075,
+      "grad_norm": 1.8537407461103759,
+      "learning_rate": 1.6734870356339383e-07,
+      "loss": 0.6546,
+      "step": 6196
+    },
+    {
+      "epoch": 11.475925925925926,
+      "grad_norm": 1.62397838566437,
+      "learning_rate": 1.673376190804147e-07,
+      "loss": 0.5989,
+      "step": 6197
+    },
+    {
+      "epoch": 11.477777777777778,
+      "grad_norm": 1.8456980229644713,
+      "learning_rate": 1.6732653308350297e-07,
+      "loss": 0.5897,
+      "step": 6198
+    },
+    {
+      "epoch": 11.47962962962963,
+      "grad_norm": 1.5739771894128676,
+      "learning_rate": 1.6731544557290795e-07,
+      "loss": 0.5644,
+      "step": 6199
+    },
+    {
+      "epoch": 11.481481481481481,
+      "grad_norm": 1.460695117502979,
+      "learning_rate": 1.6730435654887888e-07,
+      "loss": 0.6162,
+      "step": 6200
+    },
+    {
+      "epoch": 11.483333333333333,
+      "grad_norm": 1.3832157119355897,
+      "learning_rate": 1.672932660116651e-07,
+      "loss": 0.623,
+      "step": 6201
+    },
+    {
+      "epoch": 11.485185185185186,
+      "grad_norm": 1.5460034997117744,
+      "learning_rate": 1.672821739615159e-07,
+      "loss": 0.6041,
+      "step": 6202
+    },
+    {
+      "epoch": 11.487037037037037,
+      "grad_norm": 1.698509160794228,
+      "learning_rate": 1.6727108039868072e-07,
+      "loss": 0.6197,
+      "step": 6203
+    },
+    {
+      "epoch": 11.488888888888889,
+      "grad_norm": 1.6067299769982353,
+      "learning_rate": 1.6725998532340896e-07,
+      "loss": 0.6304,
+      "step": 6204
+    },
+    {
+      "epoch": 11.49074074074074,
+      "grad_norm": 1.8519130467011935,
+      "learning_rate": 1.6724888873595e-07,
+      "loss": 0.659,
+      "step": 6205
+    },
+    {
+      "epoch": 11.492592592592592,
+      "grad_norm": 3.33231514201196,
+      "learning_rate": 1.6723779063655342e-07,
+      "loss": 0.6273,
+      "step": 6206
+    },
+    {
+      "epoch": 11.494444444444444,
+      "grad_norm": 1.522498027739761,
+      "learning_rate": 1.6722669102546868e-07,
+      "loss": 0.5818,
+      "step": 6207
+    },
+    {
+      "epoch": 11.496296296296297,
+      "grad_norm": 1.544644123030037,
+      "learning_rate": 1.6721558990294535e-07,
+      "loss": 0.5866,
+      "step": 6208
+    },
+    {
+      "epoch": 11.498148148148148,
+      "grad_norm": 1.3857098635412413,
+      "learning_rate": 1.67204487269233e-07,
+      "loss": 0.6215,
+      "step": 6209
+    },
+    {
+      "epoch": 11.5,
+      "grad_norm": 1.5846332488751016,
+      "learning_rate": 1.6719338312458123e-07,
+      "loss": 0.5937,
+      "step": 6210
+    },
+    {
+      "epoch": 11.501851851851852,
+      "grad_norm": 1.408714697606958,
+      "learning_rate": 1.6718227746923973e-07,
+      "loss": 0.5367,
+      "step": 6211
+    },
+    {
+      "epoch": 11.503703703703703,
+      "grad_norm": 1.458639915016249,
+      "learning_rate": 1.6717117030345817e-07,
+      "loss": 0.6252,
+      "step": 6212
+    },
+    {
+      "epoch": 11.505555555555556,
+      "grad_norm": 1.3651097127448846,
+      "learning_rate": 1.6716006162748627e-07,
+      "loss": 0.6101,
+      "step": 6213
+    },
+    {
+      "epoch": 11.507407407407408,
+      "grad_norm": 1.6055079286355587,
+      "learning_rate": 1.6714895144157378e-07,
+      "loss": 0.6243,
+      "step": 6214
+    },
+    {
+      "epoch": 11.50925925925926,
+      "grad_norm": 1.7737317745031491,
+      "learning_rate": 1.6713783974597048e-07,
+      "loss": 0.644,
+      "step": 6215
+    },
+    {
+      "epoch": 11.511111111111111,
+      "grad_norm": 1.6954493978407092,
+      "learning_rate": 1.6712672654092622e-07,
+      "loss": 0.6271,
+      "step": 6216
+    },
+    {
+      "epoch": 11.512962962962963,
+      "grad_norm": 1.4569796057330202,
+      "learning_rate": 1.671156118266908e-07,
+      "loss": 0.6184,
+      "step": 6217
+    },
+    {
+      "epoch": 11.514814814814814,
+      "grad_norm": 1.8402672842431333,
+      "learning_rate": 1.6710449560351413e-07,
+      "loss": 0.6382,
+      "step": 6218
+    },
+    {
+      "epoch": 11.516666666666667,
+      "grad_norm": 1.4684153407708058,
+      "learning_rate": 1.670933778716462e-07,
+      "loss": 0.595,
+      "step": 6219
+    },
+    {
+      "epoch": 11.518518518518519,
+      "grad_norm": 1.787665606412809,
+      "learning_rate": 1.670822586313369e-07,
+      "loss": 0.6503,
+      "step": 6220
+    },
+    {
+      "epoch": 11.52037037037037,
+      "grad_norm": 1.6469882618413847,
+      "learning_rate": 1.6707113788283623e-07,
+      "loss": 0.6431,
+      "step": 6221
+    },
+    {
+      "epoch": 11.522222222222222,
+      "grad_norm": 1.4877634020193584,
+      "learning_rate": 1.670600156263942e-07,
+      "loss": 0.6076,
+      "step": 6222
+    },
+    {
+      "epoch": 11.524074074074074,
+      "grad_norm": 1.6621082703447536,
+      "learning_rate": 1.670488918622609e-07,
+      "loss": 0.6271,
+      "step": 6223
+    },
+    {
+      "epoch": 11.525925925925925,
+      "grad_norm": 1.4264101397911912,
+      "learning_rate": 1.670377665906864e-07,
+      "loss": 0.6387,
+      "step": 6224
+    },
+    {
+      "epoch": 11.527777777777779,
+      "grad_norm": 1.5288064744522503,
+      "learning_rate": 1.6702663981192086e-07,
+      "loss": 0.641,
+      "step": 6225
+    },
+    {
+      "epoch": 11.52962962962963,
+      "grad_norm": 5.3506404582950715,
+      "learning_rate": 1.670155115262144e-07,
+      "loss": 0.633,
+      "step": 6226
+    },
+    {
+      "epoch": 11.531481481481482,
+      "grad_norm": 1.5853782458286365,
+      "learning_rate": 1.6700438173381718e-07,
+      "loss": 0.6663,
+      "step": 6227
+    },
+    {
+      "epoch": 11.533333333333333,
+      "grad_norm": 1.4896163997112069,
+      "learning_rate": 1.6699325043497953e-07,
+      "loss": 0.6218,
+      "step": 6228
+    },
+    {
+      "epoch": 11.535185185185185,
+      "grad_norm": 1.570512243772648,
+      "learning_rate": 1.6698211762995167e-07,
+      "loss": 0.6169,
+      "step": 6229
+    },
+    {
+      "epoch": 11.537037037037036,
+      "grad_norm": 1.6007949302657458,
+      "learning_rate": 1.6697098331898386e-07,
+      "loss": 0.6333,
+      "step": 6230
+    },
+    {
+      "epoch": 11.53888888888889,
+      "grad_norm": 1.4713092223231632,
+      "learning_rate": 1.6695984750232647e-07,
+      "loss": 0.6371,
+      "step": 6231
+    },
+    {
+      "epoch": 11.540740740740741,
+      "grad_norm": 1.636354499856578,
+      "learning_rate": 1.6694871018022982e-07,
+      "loss": 0.6696,
+      "step": 6232
+    },
+    {
+      "epoch": 11.542592592592593,
+      "grad_norm": 5.147914653041103,
+      "learning_rate": 1.6693757135294435e-07,
+      "loss": 0.6079,
+      "step": 6233
+    },
+    {
+      "epoch": 11.544444444444444,
+      "grad_norm": 2.3274418096888767,
+      "learning_rate": 1.669264310207205e-07,
+      "loss": 0.5809,
+      "step": 6234
+    },
+    {
+      "epoch": 11.546296296296296,
+      "grad_norm": 1.4832605201762499,
+      "learning_rate": 1.669152891838087e-07,
+      "loss": 0.6541,
+      "step": 6235
+    },
+    {
+      "epoch": 11.548148148148147,
+      "grad_norm": 1.5826397384281918,
+      "learning_rate": 1.669041458424594e-07,
+      "loss": 0.5781,
+      "step": 6236
+    },
+    {
+      "epoch": 11.55,
+      "grad_norm": 1.7137126589071063,
+      "learning_rate": 1.668930009969233e-07,
+      "loss": 0.6235,
+      "step": 6237
+    },
+    {
+      "epoch": 11.551851851851852,
+      "grad_norm": 1.349414839609121,
+      "learning_rate": 1.6688185464745078e-07,
+      "loss": 0.6113,
+      "step": 6238
+    },
+    {
+      "epoch": 11.553703703703704,
+      "grad_norm": 1.3085895969654802,
+      "learning_rate": 1.6687070679429252e-07,
+      "loss": 0.5878,
+      "step": 6239
+    },
+    {
+      "epoch": 11.555555555555555,
+      "grad_norm": 1.446915597307635,
+      "learning_rate": 1.668595574376992e-07,
+      "loss": 0.6184,
+      "step": 6240
+    },
+    {
+      "epoch": 11.557407407407407,
+      "grad_norm": 1.5722968347899688,
+      "learning_rate": 1.6684840657792142e-07,
+      "loss": 0.6345,
+      "step": 6241
+    },
+    {
+      "epoch": 11.559259259259258,
+      "grad_norm": 1.5380726948352605,
+      "learning_rate": 1.668372542152099e-07,
+      "loss": 0.6087,
+      "step": 6242
+    },
+    {
+      "epoch": 11.561111111111112,
+      "grad_norm": 1.5650068012898408,
+      "learning_rate": 1.668261003498154e-07,
+      "loss": 0.6339,
+      "step": 6243
+    },
+    {
+      "epoch": 11.562962962962963,
+      "grad_norm": 1.5564925554821079,
+      "learning_rate": 1.6681494498198863e-07,
+      "loss": 0.6375,
+      "step": 6244
+    },
+    {
+      "epoch": 11.564814814814815,
+      "grad_norm": 1.6524571938771602,
+      "learning_rate": 1.6680378811198044e-07,
+      "loss": 0.6117,
+      "step": 6245
+    },
+    {
+      "epoch": 11.566666666666666,
+      "grad_norm": 1.507276010440385,
+      "learning_rate": 1.6679262974004166e-07,
+      "loss": 0.6484,
+      "step": 6246
+    },
+    {
+      "epoch": 11.568518518518518,
+      "grad_norm": 1.3821805948791446,
+      "learning_rate": 1.6678146986642317e-07,
+      "loss": 0.6194,
+      "step": 6247
+    },
+    {
+      "epoch": 11.57037037037037,
+      "grad_norm": 1.4964064575704334,
+      "learning_rate": 1.6677030849137584e-07,
+      "loss": 0.6245,
+      "step": 6248
+    },
+    {
+      "epoch": 11.572222222222223,
+      "grad_norm": 1.5406692356599994,
+      "learning_rate": 1.6675914561515065e-07,
+      "loss": 0.6149,
+      "step": 6249
+    },
+    {
+      "epoch": 11.574074074074074,
+      "grad_norm": 1.5032393608953865,
+      "learning_rate": 1.6674798123799857e-07,
+      "loss": 0.6491,
+      "step": 6250
+    },
+    {
+      "epoch": 11.575925925925926,
+      "grad_norm": 1.5316813856504001,
+      "learning_rate": 1.6673681536017052e-07,
+      "loss": 0.5869,
+      "step": 6251
+    },
+    {
+      "epoch": 11.577777777777778,
+      "grad_norm": 1.5689300372691775,
+      "learning_rate": 1.6672564798191765e-07,
+      "loss": 0.5801,
+      "step": 6252
+    },
+    {
+      "epoch": 11.579629629629629,
+      "grad_norm": 1.5223383963615624,
+      "learning_rate": 1.66714479103491e-07,
+      "loss": 0.659,
+      "step": 6253
+    },
+    {
+      "epoch": 11.581481481481482,
+      "grad_norm": 1.414749129931455,
+      "learning_rate": 1.667033087251416e-07,
+      "loss": 0.648,
+      "step": 6254
+    },
+    {
+      "epoch": 11.583333333333334,
+      "grad_norm": 1.5157401555091015,
+      "learning_rate": 1.6669213684712076e-07,
+      "loss": 0.6609,
+      "step": 6255
+    },
+    {
+      "epoch": 11.585185185185185,
+      "grad_norm": 2.0440044451368022,
+      "learning_rate": 1.6668096346967948e-07,
+      "loss": 0.644,
+      "step": 6256
+    },
+    {
+      "epoch": 11.587037037037037,
+      "grad_norm": 2.0014545065171596,
+      "learning_rate": 1.6666978859306907e-07,
+      "loss": 0.6359,
+      "step": 6257
+    },
+    {
+      "epoch": 11.588888888888889,
+      "grad_norm": 1.591278352066013,
+      "learning_rate": 1.6665861221754072e-07,
+      "loss": 0.6195,
+      "step": 6258
+    },
+    {
+      "epoch": 11.59074074074074,
+      "grad_norm": 1.4294537946652528,
+      "learning_rate": 1.6664743434334571e-07,
+      "loss": 0.5972,
+      "step": 6259
+    },
+    {
+      "epoch": 11.592592592592592,
+      "grad_norm": 1.5720878263636167,
+      "learning_rate": 1.666362549707354e-07,
+      "loss": 0.6399,
+      "step": 6260
+    },
+    {
+      "epoch": 11.594444444444445,
+      "grad_norm": 1.6654548132295792,
+      "learning_rate": 1.6662507409996106e-07,
+      "loss": 0.6184,
+      "step": 6261
+    },
+    {
+      "epoch": 11.596296296296297,
+      "grad_norm": 1.6077963310588668,
+      "learning_rate": 1.6661389173127416e-07,
+      "loss": 0.6181,
+      "step": 6262
+    },
+    {
+      "epoch": 11.598148148148148,
+      "grad_norm": 2.5265451785654993,
+      "learning_rate": 1.66602707864926e-07,
+      "loss": 0.6424,
+      "step": 6263
+    },
+    {
+      "epoch": 11.6,
+      "grad_norm": 1.411964087451029,
+      "learning_rate": 1.6659152250116811e-07,
+      "loss": 0.6197,
+      "step": 6264
+    },
+    {
+      "epoch": 11.601851851851851,
+      "grad_norm": 1.4518107645800067,
+      "learning_rate": 1.6658033564025193e-07,
+      "loss": 0.6428,
+      "step": 6265
+    },
+    {
+      "epoch": 11.603703703703705,
+      "grad_norm": 1.888960815435029,
+      "learning_rate": 1.6656914728242894e-07,
+      "loss": 0.6464,
+      "step": 6266
+    },
+    {
+      "epoch": 11.605555555555556,
+      "grad_norm": 1.532147313626278,
+      "learning_rate": 1.6655795742795075e-07,
+      "loss": 0.5987,
+      "step": 6267
+    },
+    {
+      "epoch": 11.607407407407408,
+      "grad_norm": 1.6863579701247342,
+      "learning_rate": 1.6654676607706892e-07,
+      "loss": 0.6284,
+      "step": 6268
+    },
+    {
+      "epoch": 11.60925925925926,
+      "grad_norm": 1.5920634518351187,
+      "learning_rate": 1.6653557323003503e-07,
+      "loss": 0.6303,
+      "step": 6269
+    },
+    {
+      "epoch": 11.61111111111111,
+      "grad_norm": 1.4814739823819973,
+      "learning_rate": 1.6652437888710074e-07,
+      "loss": 0.6129,
+      "step": 6270
+    },
+    {
+      "epoch": 11.612962962962962,
+      "grad_norm": 1.8392935974418911,
+      "learning_rate": 1.6651318304851776e-07,
+      "loss": 0.5781,
+      "step": 6271
+    },
+    {
+      "epoch": 11.614814814814816,
+      "grad_norm": 1.558156713628762,
+      "learning_rate": 1.6650198571453777e-07,
+      "loss": 0.5946,
+      "step": 6272
+    },
+    {
+      "epoch": 11.616666666666667,
+      "grad_norm": 1.696504430959664,
+      "learning_rate": 1.6649078688541248e-07,
+      "loss": 0.6314,
+      "step": 6273
+    },
+    {
+      "epoch": 11.618518518518519,
+      "grad_norm": 1.4893108789756926,
+      "learning_rate": 1.6647958656139377e-07,
+      "loss": 0.6816,
+      "step": 6274
+    },
+    {
+      "epoch": 11.62037037037037,
+      "grad_norm": 1.492015978183013,
+      "learning_rate": 1.6646838474273338e-07,
+      "loss": 0.6302,
+      "step": 6275
+    },
+    {
+      "epoch": 11.622222222222222,
+      "grad_norm": 1.7225578348589157,
+      "learning_rate": 1.6645718142968318e-07,
+      "loss": 0.6325,
+      "step": 6276
+    },
+    {
+      "epoch": 11.624074074074073,
+      "grad_norm": 1.5960953620892344,
+      "learning_rate": 1.6644597662249505e-07,
+      "loss": 0.6014,
+      "step": 6277
+    },
+    {
+      "epoch": 11.625925925925927,
+      "grad_norm": 2.352903782285803,
+      "learning_rate": 1.6643477032142087e-07,
+      "loss": 0.6303,
+      "step": 6278
+    },
+    {
+      "epoch": 11.627777777777778,
+      "grad_norm": 1.7922186172721324,
+      "learning_rate": 1.6642356252671265e-07,
+      "loss": 0.6379,
+      "step": 6279
+    },
+    {
+      "epoch": 11.62962962962963,
+      "grad_norm": 1.4287209419383993,
+      "learning_rate": 1.6641235323862234e-07,
+      "loss": 0.609,
+      "step": 6280
+    },
+    {
+      "epoch": 11.631481481481481,
+      "grad_norm": 1.8647749058709564,
+      "learning_rate": 1.6640114245740197e-07,
+      "loss": 0.6495,
+      "step": 6281
+    },
+    {
+      "epoch": 11.633333333333333,
+      "grad_norm": 1.4511479062531873,
+      "learning_rate": 1.6638993018330357e-07,
+      "loss": 0.6148,
+      "step": 6282
+    },
+    {
+      "epoch": 11.635185185185184,
+      "grad_norm": 1.5046569050659722,
+      "learning_rate": 1.6637871641657918e-07,
+      "loss": 0.6323,
+      "step": 6283
+    },
+    {
+      "epoch": 11.637037037037038,
+      "grad_norm": 1.410894064080913,
+      "learning_rate": 1.66367501157481e-07,
+      "loss": 0.6452,
+      "step": 6284
+    },
+    {
+      "epoch": 11.63888888888889,
+      "grad_norm": 1.8972649381198574,
+      "learning_rate": 1.6635628440626115e-07,
+      "loss": 0.6288,
+      "step": 6285
+    },
+    {
+      "epoch": 11.64074074074074,
+      "grad_norm": 1.6573103161428755,
+      "learning_rate": 1.663450661631718e-07,
+      "loss": 0.6538,
+      "step": 6286
+    },
+    {
+      "epoch": 11.642592592592592,
+      "grad_norm": 1.7233344138384699,
+      "learning_rate": 1.6633384642846514e-07,
+      "loss": 0.5988,
+      "step": 6287
+    },
+    {
+      "epoch": 11.644444444444444,
+      "grad_norm": 1.6329573323226352,
+      "learning_rate": 1.6632262520239348e-07,
+      "loss": 0.6116,
+      "step": 6288
+    },
+    {
+      "epoch": 11.646296296296295,
+      "grad_norm": 1.8624530315234278,
+      "learning_rate": 1.6631140248520907e-07,
+      "loss": 0.6184,
+      "step": 6289
+    },
+    {
+      "epoch": 11.648148148148149,
+      "grad_norm": 1.6561223108206182,
+      "learning_rate": 1.6630017827716425e-07,
+      "loss": 0.6846,
+      "step": 6290
+    },
+    {
+      "epoch": 11.65,
+      "grad_norm": 1.512509148089906,
+      "learning_rate": 1.6628895257851135e-07,
+      "loss": 0.6243,
+      "step": 6291
+    },
+    {
+      "epoch": 11.651851851851852,
+      "grad_norm": 1.5409569091229534,
+      "learning_rate": 1.6627772538950274e-07,
+      "loss": 0.6318,
+      "step": 6292
+    },
+    {
+      "epoch": 11.653703703703703,
+      "grad_norm": 1.5468195934365614,
+      "learning_rate": 1.6626649671039085e-07,
+      "loss": 0.68,
+      "step": 6293
+    },
+    {
+      "epoch": 11.655555555555555,
+      "grad_norm": 1.4657078003740112,
+      "learning_rate": 1.6625526654142816e-07,
+      "loss": 0.6236,
+      "step": 6294
+    },
+    {
+      "epoch": 11.657407407407408,
+      "grad_norm": 1.574344917277673,
+      "learning_rate": 1.6624403488286713e-07,
+      "loss": 0.6011,
+      "step": 6295
+    },
+    {
+      "epoch": 11.65925925925926,
+      "grad_norm": 1.4207920157995821,
+      "learning_rate": 1.6623280173496027e-07,
+      "loss": 0.6454,
+      "step": 6296
+    },
+    {
+      "epoch": 11.661111111111111,
+      "grad_norm": 1.518694144377378,
+      "learning_rate": 1.6622156709796014e-07,
+      "loss": 0.6211,
+      "step": 6297
+    },
+    {
+      "epoch": 11.662962962962963,
+      "grad_norm": 1.6508759179831987,
+      "learning_rate": 1.6621033097211932e-07,
+      "loss": 0.6458,
+      "step": 6298
+    },
+    {
+      "epoch": 11.664814814814815,
+      "grad_norm": 1.4490673039572237,
+      "learning_rate": 1.6619909335769046e-07,
+      "loss": 0.6036,
+      "step": 6299
+    },
+    {
+      "epoch": 11.666666666666666,
+      "grad_norm": 1.694628014836841,
+      "learning_rate": 1.6618785425492616e-07,
+      "loss": 0.6582,
+      "step": 6300
+    },
+    {
+      "epoch": 11.668518518518518,
+      "grad_norm": 1.341980306918321,
+      "learning_rate": 1.6617661366407917e-07,
+      "loss": 0.6234,
+      "step": 6301
+    },
+    {
+      "epoch": 11.670370370370371,
+      "grad_norm": 1.5219227042071106,
+      "learning_rate": 1.6616537158540213e-07,
+      "loss": 0.6429,
+      "step": 6302
+    },
+    {
+      "epoch": 11.672222222222222,
+      "grad_norm": 1.5803125011634997,
+      "learning_rate": 1.6615412801914785e-07,
+      "loss": 0.6202,
+      "step": 6303
+    },
+    {
+      "epoch": 11.674074074074074,
+      "grad_norm": 1.5761684145214607,
+      "learning_rate": 1.661428829655691e-07,
+      "loss": 0.5893,
+      "step": 6304
+    },
+    {
+      "epoch": 11.675925925925926,
+      "grad_norm": 1.4905278628965921,
+      "learning_rate": 1.661316364249187e-07,
+      "loss": 0.5999,
+      "step": 6305
+    },
+    {
+      "epoch": 11.677777777777777,
+      "grad_norm": 1.3528540921139283,
+      "learning_rate": 1.6612038839744952e-07,
+      "loss": 0.6356,
+      "step": 6306
+    },
+    {
+      "epoch": 11.67962962962963,
+      "grad_norm": 3.4645169879649833,
+      "learning_rate": 1.661091388834144e-07,
+      "loss": 0.6071,
+      "step": 6307
+    },
+    {
+      "epoch": 11.681481481481482,
+      "grad_norm": 1.4887074547331722,
+      "learning_rate": 1.660978878830663e-07,
+      "loss": 0.6114,
+      "step": 6308
+    },
+    {
+      "epoch": 11.683333333333334,
+      "grad_norm": 1.3551729806318744,
+      "learning_rate": 1.6608663539665817e-07,
+      "loss": 0.6327,
+      "step": 6309
+    },
+    {
+      "epoch": 11.685185185185185,
+      "grad_norm": 1.427274671936576,
+      "learning_rate": 1.6607538142444298e-07,
+      "loss": 0.6071,
+      "step": 6310
+    },
+    {
+      "epoch": 11.687037037037037,
+      "grad_norm": 1.4034954578971253,
+      "learning_rate": 1.6606412596667376e-07,
+      "loss": 0.6769,
+      "step": 6311
+    },
+    {
+      "epoch": 11.688888888888888,
+      "grad_norm": 1.4105861666953468,
+      "learning_rate": 1.6605286902360357e-07,
+      "loss": 0.5926,
+      "step": 6312
+    },
+    {
+      "epoch": 11.690740740740742,
+      "grad_norm": 1.7708649118910318,
+      "learning_rate": 1.6604161059548548e-07,
+      "loss": 0.627,
+      "step": 6313
+    },
+    {
+      "epoch": 11.692592592592593,
+      "grad_norm": 2.2537683429355115,
+      "learning_rate": 1.6603035068257262e-07,
+      "loss": 0.6577,
+      "step": 6314
+    },
+    {
+      "epoch": 11.694444444444445,
+      "grad_norm": 1.9011910484698864,
+      "learning_rate": 1.6601908928511812e-07,
+      "loss": 0.6134,
+      "step": 6315
+    },
+    {
+      "epoch": 11.696296296296296,
+      "grad_norm": 1.8244869206006644,
+      "learning_rate": 1.6600782640337519e-07,
+      "loss": 0.6474,
+      "step": 6316
+    },
+    {
+      "epoch": 11.698148148148148,
+      "grad_norm": 1.4216886469031214,
+      "learning_rate": 1.6599656203759707e-07,
+      "loss": 0.6465,
+      "step": 6317
+    },
+    {
+      "epoch": 11.7,
+      "grad_norm": 1.508510795752423,
+      "learning_rate": 1.6598529618803698e-07,
+      "loss": 0.6472,
+      "step": 6318
+    },
+    {
+      "epoch": 11.701851851851853,
+      "grad_norm": 1.3838817303541242,
+      "learning_rate": 1.6597402885494823e-07,
+      "loss": 0.6232,
+      "step": 6319
+    },
+    {
+      "epoch": 11.703703703703704,
+      "grad_norm": 1.5044419428015603,
+      "learning_rate": 1.659627600385841e-07,
+      "loss": 0.6671,
+      "step": 6320
+    },
+    {
+      "epoch": 11.705555555555556,
+      "grad_norm": 1.6664036725195532,
+      "learning_rate": 1.6595148973919798e-07,
+      "loss": 0.6084,
+      "step": 6321
+    },
+    {
+      "epoch": 11.707407407407407,
+      "grad_norm": 1.6377579051601383,
+      "learning_rate": 1.6594021795704327e-07,
+      "loss": 0.6343,
+      "step": 6322
+    },
+    {
+      "epoch": 11.709259259259259,
+      "grad_norm": 1.7718999687809025,
+      "learning_rate": 1.6592894469237337e-07,
+      "loss": 0.5925,
+      "step": 6323
+    },
+    {
+      "epoch": 11.71111111111111,
+      "grad_norm": 1.8104779258889268,
+      "learning_rate": 1.6591766994544172e-07,
+      "loss": 0.6122,
+      "step": 6324
+    },
+    {
+      "epoch": 11.712962962962964,
+      "grad_norm": 1.667405357986239,
+      "learning_rate": 1.659063937165018e-07,
+      "loss": 0.62,
+      "step": 6325
+    },
+    {
+      "epoch": 11.714814814814815,
+      "grad_norm": 1.9286981132301937,
+      "learning_rate": 1.6589511600580722e-07,
+      "loss": 0.5899,
+      "step": 6326
+    },
+    {
+      "epoch": 11.716666666666667,
+      "grad_norm": 1.7416232824369224,
+      "learning_rate": 1.658838368136114e-07,
+      "loss": 0.5808,
+      "step": 6327
+    },
+    {
+      "epoch": 11.718518518518518,
+      "grad_norm": 1.5741337492859668,
+      "learning_rate": 1.65872556140168e-07,
+      "loss": 0.6458,
+      "step": 6328
+    },
+    {
+      "epoch": 11.72037037037037,
+      "grad_norm": 1.5139378624071405,
+      "learning_rate": 1.6586127398573064e-07,
+      "loss": 0.6071,
+      "step": 6329
+    },
+    {
+      "epoch": 11.722222222222221,
+      "grad_norm": 2.786104593113031,
+      "learning_rate": 1.6584999035055296e-07,
+      "loss": 0.6413,
+      "step": 6330
+    },
+    {
+      "epoch": 11.724074074074075,
+      "grad_norm": 1.56208165242815,
+      "learning_rate": 1.6583870523488864e-07,
+      "loss": 0.6052,
+      "step": 6331
+    },
+    {
+      "epoch": 11.725925925925926,
+      "grad_norm": 1.5084508657838658,
+      "learning_rate": 1.6582741863899147e-07,
+      "loss": 0.6292,
+      "step": 6332
+    },
+    {
+      "epoch": 11.727777777777778,
+      "grad_norm": 2.1625393344334496,
+      "learning_rate": 1.6581613056311508e-07,
+      "loss": 0.5942,
+      "step": 6333
+    },
+    {
+      "epoch": 11.72962962962963,
+      "grad_norm": 1.6970789008226272,
+      "learning_rate": 1.6580484100751334e-07,
+      "loss": 0.6188,
+      "step": 6334
+    },
+    {
+      "epoch": 11.731481481481481,
+      "grad_norm": 1.780945080553821,
+      "learning_rate": 1.6579354997244003e-07,
+      "loss": 0.6095,
+      "step": 6335
+    },
+    {
+      "epoch": 11.733333333333333,
+      "grad_norm": 1.6004377846071343,
+      "learning_rate": 1.6578225745814906e-07,
+      "loss": 0.639,
+      "step": 6336
+    },
+    {
+      "epoch": 11.735185185185186,
+      "grad_norm": 3.700453298327079,
+      "learning_rate": 1.6577096346489425e-07,
+      "loss": 0.6205,
+      "step": 6337
+    },
+    {
+      "epoch": 11.737037037037037,
+      "grad_norm": 2.0602346028089475,
+      "learning_rate": 1.657596679929296e-07,
+      "loss": 0.5841,
+      "step": 6338
+    },
+    {
+      "epoch": 11.738888888888889,
+      "grad_norm": 1.7754394071021233,
+      "learning_rate": 1.6574837104250894e-07,
+      "loss": 0.6442,
+      "step": 6339
+    },
+    {
+      "epoch": 11.74074074074074,
+      "grad_norm": 3.5535966782735025,
+      "learning_rate": 1.6573707261388638e-07,
+      "loss": 0.6182,
+      "step": 6340
+    },
+    {
+      "epoch": 11.742592592592592,
+      "grad_norm": 1.5456730744562626,
+      "learning_rate": 1.6572577270731587e-07,
+      "loss": 0.6644,
+      "step": 6341
+    },
+    {
+      "epoch": 11.744444444444444,
+      "grad_norm": 1.8826391635246373,
+      "learning_rate": 1.657144713230515e-07,
+      "loss": 0.616,
+      "step": 6342
+    },
+    {
+      "epoch": 11.746296296296297,
+      "grad_norm": 1.6971899099723748,
+      "learning_rate": 1.657031684613473e-07,
+      "loss": 0.6138,
+      "step": 6343
+    },
+    {
+      "epoch": 11.748148148148148,
+      "grad_norm": 1.7407975194083312,
+      "learning_rate": 1.656918641224575e-07,
+      "loss": 0.6241,
+      "step": 6344
+    },
+    {
+      "epoch": 11.75,
+      "grad_norm": 1.5385458587150036,
+      "learning_rate": 1.656805583066361e-07,
+      "loss": 0.5953,
+      "step": 6345
+    },
+    {
+      "epoch": 11.751851851851852,
+      "grad_norm": 1.4587167880716247,
+      "learning_rate": 1.656692510141374e-07,
+      "loss": 0.6087,
+      "step": 6346
+    },
+    {
+      "epoch": 11.753703703703703,
+      "grad_norm": 1.585098053175455,
+      "learning_rate": 1.6565794224521558e-07,
+      "loss": 0.629,
+      "step": 6347
+    },
+    {
+      "epoch": 11.755555555555556,
+      "grad_norm": 1.5673209210224417,
+      "learning_rate": 1.6564663200012487e-07,
+      "loss": 0.635,
+      "step": 6348
+    },
+    {
+      "epoch": 11.757407407407408,
+      "grad_norm": 1.4972051388538132,
+      "learning_rate": 1.656353202791196e-07,
+      "loss": 0.6328,
+      "step": 6349
+    },
+    {
+      "epoch": 11.75925925925926,
+      "grad_norm": 1.5036532459008225,
+      "learning_rate": 1.6562400708245405e-07,
+      "loss": 0.6156,
+      "step": 6350
+    },
+    {
+      "epoch": 11.761111111111111,
+      "grad_norm": 1.5707943099956645,
+      "learning_rate": 1.656126924103826e-07,
+      "loss": 0.6007,
+      "step": 6351
+    },
+    {
+      "epoch": 11.762962962962963,
+      "grad_norm": 1.4998203632117961,
+      "learning_rate": 1.6560137626315963e-07,
+      "loss": 0.6111,
+      "step": 6352
+    },
+    {
+      "epoch": 11.764814814814814,
+      "grad_norm": 1.307097115620002,
+      "learning_rate": 1.6559005864103958e-07,
+      "loss": 0.5775,
+      "step": 6353
+    },
+    {
+      "epoch": 11.766666666666667,
+      "grad_norm": 1.6128359340919014,
+      "learning_rate": 1.655787395442768e-07,
+      "loss": 0.6146,
+      "step": 6354
+    },
+    {
+      "epoch": 11.768518518518519,
+      "grad_norm": 1.6588181344785153,
+      "learning_rate": 1.655674189731259e-07,
+      "loss": 0.6638,
+      "step": 6355
+    },
+    {
+      "epoch": 11.77037037037037,
+      "grad_norm": 1.5587662598560095,
+      "learning_rate": 1.6555609692784133e-07,
+      "loss": 0.6134,
+      "step": 6356
+    },
+    {
+      "epoch": 11.772222222222222,
+      "grad_norm": 1.4800693262232518,
+      "learning_rate": 1.6554477340867766e-07,
+      "loss": 0.6701,
+      "step": 6357
+    },
+    {
+      "epoch": 11.774074074074074,
+      "grad_norm": 1.5544505860135458,
+      "learning_rate": 1.6553344841588946e-07,
+      "loss": 0.6161,
+      "step": 6358
+    },
+    {
+      "epoch": 11.775925925925925,
+      "grad_norm": 1.6498987677226291,
+      "learning_rate": 1.6552212194973133e-07,
+      "loss": 0.6489,
+      "step": 6359
+    },
+    {
+      "epoch": 11.777777777777779,
+      "grad_norm": 2.2497584476600174,
+      "learning_rate": 1.6551079401045797e-07,
+      "loss": 0.6256,
+      "step": 6360
+    },
+    {
+      "epoch": 11.77962962962963,
+      "grad_norm": 1.5613257878979847,
+      "learning_rate": 1.6549946459832402e-07,
+      "loss": 0.6436,
+      "step": 6361
+    },
+    {
+      "epoch": 11.781481481481482,
+      "grad_norm": 1.3857154439679875,
+      "learning_rate": 1.6548813371358424e-07,
+      "loss": 0.6506,
+      "step": 6362
+    },
+    {
+      "epoch": 11.783333333333333,
+      "grad_norm": 1.8815018983415994,
+      "learning_rate": 1.6547680135649334e-07,
+      "loss": 0.6418,
+      "step": 6363
+    },
+    {
+      "epoch": 11.785185185185185,
+      "grad_norm": 1.6291586631812849,
+      "learning_rate": 1.6546546752730607e-07,
+      "loss": 0.6661,
+      "step": 6364
+    },
+    {
+      "epoch": 11.787037037037036,
+      "grad_norm": 1.7625348141739272,
+      "learning_rate": 1.6545413222627733e-07,
+      "loss": 0.6041,
+      "step": 6365
+    },
+    {
+      "epoch": 11.78888888888889,
+      "grad_norm": 1.4799158452851364,
+      "learning_rate": 1.6544279545366196e-07,
+      "loss": 0.6138,
+      "step": 6366
+    },
+    {
+      "epoch": 11.790740740740741,
+      "grad_norm": 1.5219737854114233,
+      "learning_rate": 1.6543145720971475e-07,
+      "loss": 0.6465,
+      "step": 6367
+    },
+    {
+      "epoch": 11.792592592592593,
+      "grad_norm": 1.8935439148283575,
+      "learning_rate": 1.654201174946907e-07,
+      "loss": 0.6542,
+      "step": 6368
+    },
+    {
+      "epoch": 11.794444444444444,
+      "grad_norm": 1.6798835953498865,
+      "learning_rate": 1.654087763088447e-07,
+      "loss": 0.6464,
+      "step": 6369
+    },
+    {
+      "epoch": 11.796296296296296,
+      "grad_norm": 1.5619914647731885,
+      "learning_rate": 1.653974336524318e-07,
+      "loss": 0.5798,
+      "step": 6370
+    },
+    {
+      "epoch": 11.798148148148147,
+      "grad_norm": 2.2875975378425486,
+      "learning_rate": 1.6538608952570695e-07,
+      "loss": 0.6229,
+      "step": 6371
+    },
+    {
+      "epoch": 11.8,
+      "grad_norm": 1.5132196201133798,
+      "learning_rate": 1.6537474392892526e-07,
+      "loss": 0.5765,
+      "step": 6372
+    },
+    {
+      "epoch": 11.801851851851852,
+      "grad_norm": 1.5367397654286648,
+      "learning_rate": 1.6536339686234172e-07,
+      "loss": 0.6321,
+      "step": 6373
+    },
+    {
+      "epoch": 11.803703703703704,
+      "grad_norm": 1.5893518775819409,
+      "learning_rate": 1.6535204832621152e-07,
+      "loss": 0.6147,
+      "step": 6374
+    },
+    {
+      "epoch": 11.805555555555555,
+      "grad_norm": 1.6905258680258268,
+      "learning_rate": 1.653406983207898e-07,
+      "loss": 0.6269,
+      "step": 6375
+    },
+    {
+      "epoch": 11.807407407407407,
+      "grad_norm": 1.8443198418039966,
+      "learning_rate": 1.6532934684633165e-07,
+      "loss": 0.5844,
+      "step": 6376
+    },
+    {
+      "epoch": 11.809259259259258,
+      "grad_norm": 1.5548908118727325,
+      "learning_rate": 1.6531799390309242e-07,
+      "loss": 0.6114,
+      "step": 6377
+    },
+    {
+      "epoch": 11.811111111111112,
+      "grad_norm": 1.4711181296165419,
+      "learning_rate": 1.6530663949132726e-07,
+      "loss": 0.5984,
+      "step": 6378
+    },
+    {
+      "epoch": 11.812962962962963,
+      "grad_norm": 1.4625822993331419,
+      "learning_rate": 1.6529528361129147e-07,
+      "loss": 0.6172,
+      "step": 6379
+    },
+    {
+      "epoch": 11.814814814814815,
+      "grad_norm": 1.4773707975159378,
+      "learning_rate": 1.6528392626324038e-07,
+      "loss": 0.6135,
+      "step": 6380
+    },
+    {
+      "epoch": 11.816666666666666,
+      "grad_norm": 1.367464785831325,
+      "learning_rate": 1.652725674474293e-07,
+      "loss": 0.608,
+      "step": 6381
+    },
+    {
+      "epoch": 11.818518518518518,
+      "grad_norm": 1.4913928560400456,
+      "learning_rate": 1.6526120716411364e-07,
+      "loss": 0.616,
+      "step": 6382
+    },
+    {
+      "epoch": 11.82037037037037,
+      "grad_norm": 1.4295712749181289,
+      "learning_rate": 1.6524984541354882e-07,
+      "loss": 0.6174,
+      "step": 6383
+    },
+    {
+      "epoch": 11.822222222222223,
+      "grad_norm": 1.4961612361046086,
+      "learning_rate": 1.652384821959902e-07,
+      "loss": 0.5906,
+      "step": 6384
+    },
+    {
+      "epoch": 11.824074074074074,
+      "grad_norm": 1.6656333207454643,
+      "learning_rate": 1.6522711751169335e-07,
+      "loss": 0.6463,
+      "step": 6385
+    },
+    {
+      "epoch": 11.825925925925926,
+      "grad_norm": 1.6934623543651797,
+      "learning_rate": 1.6521575136091376e-07,
+      "loss": 0.6087,
+      "step": 6386
+    },
+    {
+      "epoch": 11.827777777777778,
+      "grad_norm": 1.769327561984438,
+      "learning_rate": 1.6520438374390694e-07,
+      "loss": 0.6343,
+      "step": 6387
+    },
+    {
+      "epoch": 11.829629629629629,
+      "grad_norm": 1.4703750743483683,
+      "learning_rate": 1.6519301466092847e-07,
+      "loss": 0.6376,
+      "step": 6388
+    },
+    {
+      "epoch": 11.831481481481482,
+      "grad_norm": 1.6504806922853015,
+      "learning_rate": 1.6518164411223402e-07,
+      "loss": 0.589,
+      "step": 6389
+    },
+    {
+      "epoch": 11.833333333333334,
+      "grad_norm": 1.673674541719744,
+      "learning_rate": 1.651702720980791e-07,
+      "loss": 0.6217,
+      "step": 6390
+    },
+    {
+      "epoch": 11.835185185185185,
+      "grad_norm": 1.5456181809101863,
+      "learning_rate": 1.6515889861871953e-07,
+      "loss": 0.621,
+      "step": 6391
+    },
+    {
+      "epoch": 11.837037037037037,
+      "grad_norm": 1.9772276254624102,
+      "learning_rate": 1.6514752367441092e-07,
+      "loss": 0.642,
+      "step": 6392
+    },
+    {
+      "epoch": 11.838888888888889,
+      "grad_norm": 1.7345261083362766,
+      "learning_rate": 1.6513614726540902e-07,
+      "loss": 0.6316,
+      "step": 6393
+    },
+    {
+      "epoch": 11.84074074074074,
+      "grad_norm": 1.6205672714711568,
+      "learning_rate": 1.6512476939196963e-07,
+      "loss": 0.6411,
+      "step": 6394
+    },
+    {
+      "epoch": 11.842592592592592,
+      "grad_norm": 1.556559430663859,
+      "learning_rate": 1.6511339005434856e-07,
+      "loss": 0.6244,
+      "step": 6395
+    },
+    {
+      "epoch": 11.844444444444445,
+      "grad_norm": 1.552097635584841,
+      "learning_rate": 1.6510200925280167e-07,
+      "loss": 0.6326,
+      "step": 6396
+    },
+    {
+      "epoch": 11.846296296296297,
+      "grad_norm": 1.592237644596314,
+      "learning_rate": 1.6509062698758475e-07,
+      "loss": 0.5993,
+      "step": 6397
+    },
+    {
+      "epoch": 11.848148148148148,
+      "grad_norm": 1.526142095194442,
+      "learning_rate": 1.6507924325895377e-07,
+      "loss": 0.6187,
+      "step": 6398
+    },
+    {
+      "epoch": 11.85,
+      "grad_norm": 1.7550401459096354,
+      "learning_rate": 1.6506785806716462e-07,
+      "loss": 0.603,
+      "step": 6399
+    },
+    {
+      "epoch": 11.851851851851851,
+      "grad_norm": 1.8471304883597084,
+      "learning_rate": 1.6505647141247337e-07,
+      "loss": 0.6445,
+      "step": 6400
+    },
+    {
+      "epoch": 11.853703703703705,
+      "grad_norm": 2.0399302968915523,
+      "learning_rate": 1.6504508329513591e-07,
+      "loss": 0.6329,
+      "step": 6401
+    },
+    {
+      "epoch": 11.855555555555556,
+      "grad_norm": 1.5654836953785103,
+      "learning_rate": 1.650336937154083e-07,
+      "loss": 0.6249,
+      "step": 6402
+    },
+    {
+      "epoch": 11.857407407407408,
+      "grad_norm": 1.5307159652017457,
+      "learning_rate": 1.6502230267354661e-07,
+      "loss": 0.6279,
+      "step": 6403
+    },
+    {
+      "epoch": 11.85925925925926,
+      "grad_norm": 1.499024291128072,
+      "learning_rate": 1.6501091016980704e-07,
+      "loss": 0.6263,
+      "step": 6404
+    },
+    {
+      "epoch": 11.86111111111111,
+      "grad_norm": 1.496073376617333,
+      "learning_rate": 1.6499951620444556e-07,
+      "loss": 0.6503,
+      "step": 6405
+    },
+    {
+      "epoch": 11.862962962962962,
+      "grad_norm": 1.5666873205917298,
+      "learning_rate": 1.6498812077771847e-07,
+      "loss": 0.6003,
+      "step": 6406
+    },
+    {
+      "epoch": 11.864814814814816,
+      "grad_norm": 1.5694465041331427,
+      "learning_rate": 1.6497672388988192e-07,
+      "loss": 0.6126,
+      "step": 6407
+    },
+    {
+      "epoch": 11.866666666666667,
+      "grad_norm": 1.577833565019944,
+      "learning_rate": 1.6496532554119212e-07,
+      "loss": 0.6522,
+      "step": 6408
+    },
+    {
+      "epoch": 11.868518518518519,
+      "grad_norm": 1.621700573035634,
+      "learning_rate": 1.6495392573190534e-07,
+      "loss": 0.6126,
+      "step": 6409
+    },
+    {
+      "epoch": 11.87037037037037,
+      "grad_norm": 1.5300422087108332,
+      "learning_rate": 1.649425244622779e-07,
+      "loss": 0.6288,
+      "step": 6410
+    },
+    {
+      "epoch": 11.872222222222222,
+      "grad_norm": 1.433119099736639,
+      "learning_rate": 1.6493112173256618e-07,
+      "loss": 0.6306,
+      "step": 6411
+    },
+    {
+      "epoch": 11.874074074074073,
+      "grad_norm": 1.5601472470627655,
+      "learning_rate": 1.6491971754302645e-07,
+      "loss": 0.6176,
+      "step": 6412
+    },
+    {
+      "epoch": 11.875925925925927,
+      "grad_norm": 1.6823507070755834,
+      "learning_rate": 1.649083118939152e-07,
+      "loss": 0.6166,
+      "step": 6413
+    },
+    {
+      "epoch": 11.877777777777778,
+      "grad_norm": 1.7711919611390672,
+      "learning_rate": 1.6489690478548872e-07,
+      "loss": 0.6251,
+      "step": 6414
+    },
+    {
+      "epoch": 11.87962962962963,
+      "grad_norm": 1.654190824821072,
+      "learning_rate": 1.6488549621800366e-07,
+      "loss": 0.6132,
+      "step": 6415
+    },
+    {
+      "epoch": 11.881481481481481,
+      "grad_norm": 1.3733726237488841,
+      "learning_rate": 1.6487408619171637e-07,
+      "loss": 0.6402,
+      "step": 6416
+    },
+    {
+      "epoch": 11.883333333333333,
+      "grad_norm": 1.5511284366995772,
+      "learning_rate": 1.6486267470688343e-07,
+      "loss": 0.635,
+      "step": 6417
+    },
+    {
+      "epoch": 11.885185185185184,
+      "grad_norm": 1.799343521264771,
+      "learning_rate": 1.648512617637614e-07,
+      "loss": 0.6296,
+      "step": 6418
+    },
+    {
+      "epoch": 11.887037037037038,
+      "grad_norm": 1.7945471056273337,
+      "learning_rate": 1.6483984736260688e-07,
+      "loss": 0.6819,
+      "step": 6419
+    },
+    {
+      "epoch": 11.88888888888889,
+      "grad_norm": 1.5383190163406673,
+      "learning_rate": 1.6482843150367647e-07,
+      "loss": 0.6097,
+      "step": 6420
+    },
+    {
+      "epoch": 11.89074074074074,
+      "grad_norm": 1.5212908678172758,
+      "learning_rate": 1.6481701418722685e-07,
+      "loss": 0.6487,
+      "step": 6421
+    },
+    {
+      "epoch": 11.892592592592592,
+      "grad_norm": 1.660658174900972,
+      "learning_rate": 1.6480559541351472e-07,
+      "loss": 0.6241,
+      "step": 6422
+    },
+    {
+      "epoch": 11.894444444444444,
+      "grad_norm": 1.4643803939371764,
+      "learning_rate": 1.6479417518279682e-07,
+      "loss": 0.6447,
+      "step": 6423
+    },
+    {
+      "epoch": 11.896296296296295,
+      "grad_norm": 1.462298842236054,
+      "learning_rate": 1.6478275349532984e-07,
+      "loss": 0.6325,
+      "step": 6424
+    },
+    {
+      "epoch": 11.898148148148149,
+      "grad_norm": 1.444820539070176,
+      "learning_rate": 1.6477133035137065e-07,
+      "loss": 0.6125,
+      "step": 6425
+    },
+    {
+      "epoch": 11.9,
+      "grad_norm": 1.4969345087378343,
+      "learning_rate": 1.6475990575117603e-07,
+      "loss": 0.6038,
+      "step": 6426
+    },
+    {
+      "epoch": 11.901851851851852,
+      "grad_norm": 1.6687835908810338,
+      "learning_rate": 1.6474847969500282e-07,
+      "loss": 0.5929,
+      "step": 6427
+    },
+    {
+      "epoch": 11.903703703703703,
+      "grad_norm": 1.9775388023354206,
+      "learning_rate": 1.6473705218310796e-07,
+      "loss": 0.622,
+      "step": 6428
+    },
+    {
+      "epoch": 11.905555555555555,
+      "grad_norm": 1.484128041964014,
+      "learning_rate": 1.647256232157483e-07,
+      "loss": 0.6668,
+      "step": 6429
+    },
+    {
+      "epoch": 11.907407407407408,
+      "grad_norm": 1.6213755769917255,
+      "learning_rate": 1.647141927931809e-07,
+      "loss": 0.641,
+      "step": 6430
+    },
+    {
+      "epoch": 11.90925925925926,
+      "grad_norm": 1.4009184893324613,
+      "learning_rate": 1.6470276091566265e-07,
+      "loss": 0.6553,
+      "step": 6431
+    },
+    {
+      "epoch": 11.911111111111111,
+      "grad_norm": 1.4716907526467358,
+      "learning_rate": 1.6469132758345058e-07,
+      "loss": 0.682,
+      "step": 6432
+    },
+    {
+      "epoch": 11.912962962962963,
+      "grad_norm": 1.6112091580727244,
+      "learning_rate": 1.6467989279680182e-07,
+      "loss": 0.5941,
+      "step": 6433
+    },
+    {
+      "epoch": 11.914814814814815,
+      "grad_norm": 1.562277566432166,
+      "learning_rate": 1.6466845655597336e-07,
+      "loss": 0.6399,
+      "step": 6434
+    },
+    {
+      "epoch": 11.916666666666666,
+      "grad_norm": 2.1170477161998447,
+      "learning_rate": 1.646570188612224e-07,
+      "loss": 0.6729,
+      "step": 6435
+    },
+    {
+      "epoch": 11.918518518518518,
+      "grad_norm": 1.6563289345287175,
+      "learning_rate": 1.6464557971280603e-07,
+      "loss": 0.6026,
+      "step": 6436
+    },
+    {
+      "epoch": 11.920370370370371,
+      "grad_norm": 1.389748994518346,
+      "learning_rate": 1.6463413911098145e-07,
+      "loss": 0.6546,
+      "step": 6437
+    },
+    {
+      "epoch": 11.922222222222222,
+      "grad_norm": 2.603387602889353,
+      "learning_rate": 1.646226970560059e-07,
+      "loss": 0.6188,
+      "step": 6438
+    },
+    {
+      "epoch": 11.924074074074074,
+      "grad_norm": 1.4762220057057864,
+      "learning_rate": 1.6461125354813658e-07,
+      "loss": 0.5928,
+      "step": 6439
+    },
+    {
+      "epoch": 11.925925925925926,
+      "grad_norm": 1.5134537961311547,
+      "learning_rate": 1.6459980858763077e-07,
+      "loss": 0.6077,
+      "step": 6440
+    },
+    {
+      "epoch": 11.927777777777777,
+      "grad_norm": 1.5297984512052334,
+      "learning_rate": 1.6458836217474587e-07,
+      "loss": 0.6218,
+      "step": 6441
+    },
+    {
+      "epoch": 11.92962962962963,
+      "grad_norm": 1.5873374590114824,
+      "learning_rate": 1.6457691430973915e-07,
+      "loss": 0.691,
+      "step": 6442
+    },
+    {
+      "epoch": 11.931481481481482,
+      "grad_norm": 1.3892032431167953,
+      "learning_rate": 1.6456546499286802e-07,
+      "loss": 0.6339,
+      "step": 6443
+    },
+    {
+      "epoch": 11.933333333333334,
+      "grad_norm": 1.5069824120180466,
+      "learning_rate": 1.6455401422438983e-07,
+      "loss": 0.66,
+      "step": 6444
+    },
+    {
+      "epoch": 11.935185185185185,
+      "grad_norm": 2.3065745219298672,
+      "learning_rate": 1.6454256200456214e-07,
+      "loss": 0.6281,
+      "step": 6445
+    },
+    {
+      "epoch": 11.937037037037037,
+      "grad_norm": 1.6815702486090638,
+      "learning_rate": 1.6453110833364233e-07,
+      "loss": 0.6023,
+      "step": 6446
+    },
+    {
+      "epoch": 11.938888888888888,
+      "grad_norm": 1.4346294580377503,
+      "learning_rate": 1.6451965321188794e-07,
+      "loss": 0.5717,
+      "step": 6447
+    },
+    {
+      "epoch": 11.940740740740742,
+      "grad_norm": 1.470746165912409,
+      "learning_rate": 1.6450819663955652e-07,
+      "loss": 0.6327,
+      "step": 6448
+    },
+    {
+      "epoch": 11.942592592592593,
+      "grad_norm": 1.7116026202587127,
+      "learning_rate": 1.6449673861690563e-07,
+      "loss": 0.6122,
+      "step": 6449
+    },
+    {
+      "epoch": 11.944444444444445,
+      "grad_norm": 1.7770710205614189,
+      "learning_rate": 1.6448527914419287e-07,
+      "loss": 0.6364,
+      "step": 6450
+    },
+    {
+      "epoch": 11.946296296296296,
+      "grad_norm": 2.6465345108893215,
+      "learning_rate": 1.6447381822167593e-07,
+      "loss": 0.6497,
+      "step": 6451
+    },
+    {
+      "epoch": 11.948148148148148,
+      "grad_norm": 1.7417576887351762,
+      "learning_rate": 1.6446235584961243e-07,
+      "loss": 0.5936,
+      "step": 6452
+    },
+    {
+      "epoch": 11.95,
+      "grad_norm": 1.522855779940423,
+      "learning_rate": 1.644508920282601e-07,
+      "loss": 0.6318,
+      "step": 6453
+    },
+    {
+      "epoch": 11.951851851851853,
+      "grad_norm": 1.4689981018970466,
+      "learning_rate": 1.6443942675787669e-07,
+      "loss": 0.6245,
+      "step": 6454
+    },
+    {
+      "epoch": 11.953703703703704,
+      "grad_norm": 2.1994593634690287,
+      "learning_rate": 1.644279600387199e-07,
+      "loss": 0.6404,
+      "step": 6455
+    },
+    {
+      "epoch": 11.955555555555556,
+      "grad_norm": 1.92817434977043,
+      "learning_rate": 1.644164918710476e-07,
+      "loss": 0.6261,
+      "step": 6456
+    },
+    {
+      "epoch": 11.957407407407407,
+      "grad_norm": 1.7593196941670797,
+      "learning_rate": 1.6440502225511766e-07,
+      "loss": 0.6002,
+      "step": 6457
+    },
+    {
+      "epoch": 11.959259259259259,
+      "grad_norm": 1.5181922385188862,
+      "learning_rate": 1.6439355119118785e-07,
+      "loss": 0.6328,
+      "step": 6458
+    },
+    {
+      "epoch": 11.96111111111111,
+      "grad_norm": 1.8758601171031275,
+      "learning_rate": 1.6438207867951613e-07,
+      "loss": 0.6035,
+      "step": 6459
+    },
+    {
+      "epoch": 11.962962962962964,
+      "grad_norm": 1.6212827469874653,
+      "learning_rate": 1.6437060472036044e-07,
+      "loss": 0.6333,
+      "step": 6460
+    },
+    {
+      "epoch": 11.964814814814815,
+      "grad_norm": 1.6884028365172763,
+      "learning_rate": 1.643591293139787e-07,
+      "loss": 0.6122,
+      "step": 6461
+    },
+    {
+      "epoch": 11.966666666666667,
+      "grad_norm": 1.6183811448413439,
+      "learning_rate": 1.6434765246062892e-07,
+      "loss": 0.6224,
+      "step": 6462
+    },
+    {
+      "epoch": 11.968518518518518,
+      "grad_norm": 1.5665496644766947,
+      "learning_rate": 1.643361741605692e-07,
+      "loss": 0.6304,
+      "step": 6463
+    },
+    {
+      "epoch": 11.97037037037037,
+      "grad_norm": 1.648530854681906,
+      "learning_rate": 1.643246944140575e-07,
+      "loss": 0.6428,
+      "step": 6464
+    },
+    {
+      "epoch": 11.972222222222221,
+      "grad_norm": 1.7914338123963771,
+      "learning_rate": 1.64313213221352e-07,
+      "loss": 0.5859,
+      "step": 6465
+    },
+    {
+      "epoch": 11.974074074074075,
+      "grad_norm": 1.514763172107704,
+      "learning_rate": 1.643017305827108e-07,
+      "loss": 0.6261,
+      "step": 6466
+    },
+    {
+      "epoch": 11.975925925925926,
+      "grad_norm": 1.5238718459025398,
+      "learning_rate": 1.6429024649839206e-07,
+      "loss": 0.6499,
+      "step": 6467
+    },
+    {
+      "epoch": 11.977777777777778,
+      "grad_norm": 1.6218285052342947,
+      "learning_rate": 1.6427876096865392e-07,
+      "loss": 0.5857,
+      "step": 6468
+    },
+    {
+      "epoch": 11.97962962962963,
+      "grad_norm": 1.6002142723997892,
+      "learning_rate": 1.6426727399375468e-07,
+      "loss": 0.6491,
+      "step": 6469
+    },
+    {
+      "epoch": 11.981481481481481,
+      "grad_norm": 1.409323510977805,
+      "learning_rate": 1.642557855739526e-07,
+      "loss": 0.6302,
+      "step": 6470
+    },
+    {
+      "epoch": 11.983333333333333,
+      "grad_norm": 1.7083831671987106,
+      "learning_rate": 1.642442957095059e-07,
+      "loss": 0.6142,
+      "step": 6471
+    },
+    {
+      "epoch": 11.985185185185186,
+      "grad_norm": 1.724358928249825,
+      "learning_rate": 1.64232804400673e-07,
+      "loss": 0.6108,
+      "step": 6472
+    },
+    {
+      "epoch": 11.987037037037037,
+      "grad_norm": 1.6703968721146336,
+      "learning_rate": 1.6422131164771216e-07,
+      "loss": 0.6083,
+      "step": 6473
+    },
+    {
+      "epoch": 11.988888888888889,
+      "grad_norm": 1.725340535329378,
+      "learning_rate": 1.6420981745088182e-07,
+      "loss": 0.6127,
+      "step": 6474
+    },
+    {
+      "epoch": 11.99074074074074,
+      "grad_norm": 3.6813504096107987,
+      "learning_rate": 1.641983218104404e-07,
+      "loss": 0.6151,
+      "step": 6475
+    },
+    {
+      "epoch": 11.992592592592592,
+      "grad_norm": 1.804433500069158,
+      "learning_rate": 1.641868247266464e-07,
+      "loss": 0.6267,
+      "step": 6476
+    },
+    {
+      "epoch": 11.994444444444444,
+      "grad_norm": 1.4259147593166637,
+      "learning_rate": 1.6417532619975813e-07,
+      "loss": 0.6123,
+      "step": 6477
+    },
+    {
+      "epoch": 11.996296296296297,
+      "grad_norm": 1.7361123435008383,
+      "learning_rate": 1.6416382623003431e-07,
+      "loss": 0.6257,
+      "step": 6478
+    },
+    {
+      "epoch": 11.998148148148148,
+      "grad_norm": 1.887508638725493,
+      "learning_rate": 1.641523248177334e-07,
+      "loss": 0.6214,
+      "step": 6479
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 1.784575495939203,
+      "learning_rate": 1.64140821963114e-07,
+      "loss": 0.6292,
+      "step": 6480
+    },
+    {
+      "epoch": 12.001851851851852,
+      "grad_norm": 2.2787701584976907,
+      "learning_rate": 1.641293176664347e-07,
+      "loss": 0.5842,
+      "step": 6481
+    },
+    {
+      "epoch": 12.003703703703703,
+      "grad_norm": 1.4810894806085335,
+      "learning_rate": 1.6411781192795418e-07,
+      "loss": 0.6291,
+      "step": 6482
+    },
+    {
+      "epoch": 12.005555555555556,
+      "grad_norm": 1.817293272323697,
+      "learning_rate": 1.641063047479311e-07,
+      "loss": 0.6458,
+      "step": 6483
+    },
+    {
+      "epoch": 12.007407407407408,
+      "grad_norm": 1.853385663599283,
+      "learning_rate": 1.640947961266242e-07,
+      "loss": 0.6456,
+      "step": 6484
+    },
+    {
+      "epoch": 12.00925925925926,
+      "grad_norm": 1.6402491965792916,
+      "learning_rate": 1.6408328606429216e-07,
+      "loss": 0.6472,
+      "step": 6485
+    },
+    {
+      "epoch": 12.011111111111111,
+      "grad_norm": 1.6526928363550455,
+      "learning_rate": 1.6407177456119383e-07,
+      "loss": 0.6197,
+      "step": 6486
+    },
+    {
+      "epoch": 12.012962962962963,
+      "grad_norm": 1.6501295977234771,
+      "learning_rate": 1.64060261617588e-07,
+      "loss": 0.5992,
+      "step": 6487
+    },
+    {
+      "epoch": 12.014814814814814,
+      "grad_norm": 1.4693232790439867,
+      "learning_rate": 1.6404874723373345e-07,
+      "loss": 0.6108,
+      "step": 6488
+    },
+    {
+      "epoch": 12.016666666666667,
+      "grad_norm": 1.6787383319659326,
+      "learning_rate": 1.6403723140988914e-07,
+      "loss": 0.5942,
+      "step": 6489
+    },
+    {
+      "epoch": 12.018518518518519,
+      "grad_norm": 1.6162219170357184,
+      "learning_rate": 1.64025714146314e-07,
+      "loss": 0.6001,
+      "step": 6490
+    },
+    {
+      "epoch": 12.02037037037037,
+      "grad_norm": 1.7904313042001472,
+      "learning_rate": 1.6401419544326682e-07,
+      "loss": 0.6572,
+      "step": 6491
+    },
+    {
+      "epoch": 12.022222222222222,
+      "grad_norm": 1.543355254360909,
+      "learning_rate": 1.6400267530100674e-07,
+      "loss": 0.6286,
+      "step": 6492
+    },
+    {
+      "epoch": 12.024074074074074,
+      "grad_norm": 1.6540914056004015,
+      "learning_rate": 1.6399115371979267e-07,
+      "loss": 0.6357,
+      "step": 6493
+    },
+    {
+      "epoch": 12.025925925925925,
+      "grad_norm": 1.510951231354261,
+      "learning_rate": 1.6397963069988366e-07,
+      "loss": 0.6626,
+      "step": 6494
+    },
+    {
+      "epoch": 12.027777777777779,
+      "grad_norm": 1.5335809411608767,
+      "learning_rate": 1.639681062415388e-07,
+      "loss": 0.605,
+      "step": 6495
+    },
+    {
+      "epoch": 12.02962962962963,
+      "grad_norm": 1.7204886316557308,
+      "learning_rate": 1.6395658034501717e-07,
+      "loss": 0.6482,
+      "step": 6496
+    },
+    {
+      "epoch": 12.031481481481482,
+      "grad_norm": 2.009598895747422,
+      "learning_rate": 1.639450530105779e-07,
+      "loss": 0.6036,
+      "step": 6497
+    },
+    {
+      "epoch": 12.033333333333333,
+      "grad_norm": 1.9258096937047589,
+      "learning_rate": 1.6393352423848015e-07,
+      "loss": 0.6713,
+      "step": 6498
+    },
+    {
+      "epoch": 12.035185185185185,
+      "grad_norm": 1.3965586041185545,
+      "learning_rate": 1.6392199402898317e-07,
+      "loss": 0.6065,
+      "step": 6499
+    },
+    {
+      "epoch": 12.037037037037036,
+      "grad_norm": 1.5714223010425215,
+      "learning_rate": 1.6391046238234614e-07,
+      "loss": 0.5869,
+      "step": 6500
+    },
+    {
+      "epoch": 12.03888888888889,
+      "grad_norm": 1.5101361856185822,
+      "learning_rate": 1.6389892929882834e-07,
+      "loss": 0.6497,
+      "step": 6501
+    },
+    {
+      "epoch": 12.040740740740741,
+      "grad_norm": 1.6652365874458868,
+      "learning_rate": 1.6388739477868908e-07,
+      "loss": 0.6253,
+      "step": 6502
+    },
+    {
+      "epoch": 12.042592592592593,
+      "grad_norm": 1.658238720357663,
+      "learning_rate": 1.638758588221876e-07,
+      "loss": 0.5914,
+      "step": 6503
+    },
+    {
+      "epoch": 12.044444444444444,
+      "grad_norm": 1.4565484039910486,
+      "learning_rate": 1.638643214295834e-07,
+      "loss": 0.5648,
+      "step": 6504
+    },
+    {
+      "epoch": 12.046296296296296,
+      "grad_norm": 1.664461885354162,
+      "learning_rate": 1.638527826011358e-07,
+      "loss": 0.6587,
+      "step": 6505
+    },
+    {
+      "epoch": 12.048148148148147,
+      "grad_norm": 1.4336673346446493,
+      "learning_rate": 1.6384124233710418e-07,
+      "loss": 0.6345,
+      "step": 6506
+    },
+    {
+      "epoch": 12.05,
+      "grad_norm": 1.656852925452071,
+      "learning_rate": 1.6382970063774808e-07,
+      "loss": 0.6443,
+      "step": 6507
+    },
+    {
+      "epoch": 12.051851851851852,
+      "grad_norm": 2.148405851764428,
+      "learning_rate": 1.6381815750332697e-07,
+      "loss": 0.5999,
+      "step": 6508
+    },
+    {
+      "epoch": 12.053703703703704,
+      "grad_norm": 1.3938418144482092,
+      "learning_rate": 1.6380661293410029e-07,
+      "loss": 0.6023,
+      "step": 6509
+    },
+    {
+      "epoch": 12.055555555555555,
+      "grad_norm": 1.442808686166186,
+      "learning_rate": 1.6379506693032773e-07,
+      "loss": 0.6427,
+      "step": 6510
+    },
+    {
+      "epoch": 12.057407407407407,
+      "grad_norm": 1.6405884498720538,
+      "learning_rate": 1.6378351949226874e-07,
+      "loss": 0.5915,
+      "step": 6511
+    },
+    {
+      "epoch": 12.059259259259258,
+      "grad_norm": 1.433721313892015,
+      "learning_rate": 1.63771970620183e-07,
+      "loss": 0.5985,
+      "step": 6512
+    },
+    {
+      "epoch": 12.061111111111112,
+      "grad_norm": 1.6734910927793538,
+      "learning_rate": 1.6376042031433016e-07,
+      "loss": 0.6541,
+      "step": 6513
+    },
+    {
+      "epoch": 12.062962962962963,
+      "grad_norm": 1.4287522639243595,
+      "learning_rate": 1.6374886857496995e-07,
+      "loss": 0.6426,
+      "step": 6514
+    },
+    {
+      "epoch": 12.064814814814815,
+      "grad_norm": 1.6083777544648763,
+      "learning_rate": 1.6373731540236197e-07,
+      "loss": 0.6514,
+      "step": 6515
+    },
+    {
+      "epoch": 12.066666666666666,
+      "grad_norm": 1.7425769248276788,
+      "learning_rate": 1.6372576079676606e-07,
+      "loss": 0.5949,
+      "step": 6516
+    },
+    {
+      "epoch": 12.068518518518518,
+      "grad_norm": 1.387202619465093,
+      "learning_rate": 1.6371420475844201e-07,
+      "loss": 0.5958,
+      "step": 6517
+    },
+    {
+      "epoch": 12.07037037037037,
+      "grad_norm": 2.0225468943029377,
+      "learning_rate": 1.6370264728764954e-07,
+      "loss": 0.6291,
+      "step": 6518
+    },
+    {
+      "epoch": 12.072222222222223,
+      "grad_norm": 1.4538056017265102,
+      "learning_rate": 1.6369108838464855e-07,
+      "loss": 0.6247,
+      "step": 6519
+    },
+    {
+      "epoch": 12.074074074074074,
+      "grad_norm": 1.9796249104032133,
+      "learning_rate": 1.6367952804969895e-07,
+      "loss": 0.627,
+      "step": 6520
+    },
+    {
+      "epoch": 12.075925925925926,
+      "grad_norm": 1.5088544623933808,
+      "learning_rate": 1.6366796628306056e-07,
+      "loss": 0.6069,
+      "step": 6521
+    },
+    {
+      "epoch": 12.077777777777778,
+      "grad_norm": 1.866027702548339,
+      "learning_rate": 1.636564030849934e-07,
+      "loss": 0.6123,
+      "step": 6522
+    },
+    {
+      "epoch": 12.079629629629629,
+      "grad_norm": 1.5175240916048642,
+      "learning_rate": 1.6364483845575738e-07,
+      "loss": 0.622,
+      "step": 6523
+    },
+    {
+      "epoch": 12.081481481481482,
+      "grad_norm": 1.5534456220694706,
+      "learning_rate": 1.6363327239561256e-07,
+      "loss": 0.6226,
+      "step": 6524
+    },
+    {
+      "epoch": 12.083333333333334,
+      "grad_norm": 1.8432339561778077,
+      "learning_rate": 1.6362170490481896e-07,
+      "loss": 0.5655,
+      "step": 6525
+    },
+    {
+      "epoch": 12.085185185185185,
+      "grad_norm": 1.6688563884289775,
+      "learning_rate": 1.6361013598363663e-07,
+      "loss": 0.6584,
+      "step": 6526
+    },
+    {
+      "epoch": 12.087037037037037,
+      "grad_norm": 1.7570377878114263,
+      "learning_rate": 1.6359856563232565e-07,
+      "loss": 0.6659,
+      "step": 6527
+    },
+    {
+      "epoch": 12.088888888888889,
+      "grad_norm": 1.5569023024427324,
+      "learning_rate": 1.6358699385114621e-07,
+      "loss": 0.6177,
+      "step": 6528
+    },
+    {
+      "epoch": 12.09074074074074,
+      "grad_norm": 1.5847221148183925,
+      "learning_rate": 1.6357542064035847e-07,
+      "loss": 0.6196,
+      "step": 6529
+    },
+    {
+      "epoch": 12.092592592592593,
+      "grad_norm": 1.6345527100876884,
+      "learning_rate": 1.6356384600022258e-07,
+      "loss": 0.6065,
+      "step": 6530
+    },
+    {
+      "epoch": 12.094444444444445,
+      "grad_norm": 1.6465471033362065,
+      "learning_rate": 1.635522699309988e-07,
+      "loss": 0.6248,
+      "step": 6531
+    },
+    {
+      "epoch": 12.096296296296297,
+      "grad_norm": 1.9406330249831958,
+      "learning_rate": 1.6354069243294738e-07,
+      "loss": 0.5819,
+      "step": 6532
+    },
+    {
+      "epoch": 12.098148148148148,
+      "grad_norm": 1.5492302458312346,
+      "learning_rate": 1.635291135063286e-07,
+      "loss": 0.6655,
+      "step": 6533
+    },
+    {
+      "epoch": 12.1,
+      "grad_norm": 1.8354753626822733,
+      "learning_rate": 1.6351753315140285e-07,
+      "loss": 0.6201,
+      "step": 6534
+    },
+    {
+      "epoch": 12.101851851851851,
+      "grad_norm": 1.5549770914546528,
+      "learning_rate": 1.6350595136843044e-07,
+      "loss": 0.6137,
+      "step": 6535
+    },
+    {
+      "epoch": 12.103703703703705,
+      "grad_norm": 1.5637848081548742,
+      "learning_rate": 1.6349436815767172e-07,
+      "loss": 0.6362,
+      "step": 6536
+    },
+    {
+      "epoch": 12.105555555555556,
+      "grad_norm": 1.6528884910408053,
+      "learning_rate": 1.6348278351938717e-07,
+      "loss": 0.6406,
+      "step": 6537
+    },
+    {
+      "epoch": 12.107407407407408,
+      "grad_norm": 1.9152508319253156,
+      "learning_rate": 1.6347119745383728e-07,
+      "loss": 0.6042,
+      "step": 6538
+    },
+    {
+      "epoch": 12.10925925925926,
+      "grad_norm": 1.5408813579725722,
+      "learning_rate": 1.6345960996128244e-07,
+      "loss": 0.6323,
+      "step": 6539
+    },
+    {
+      "epoch": 12.11111111111111,
+      "grad_norm": 1.5710886017162735,
+      "learning_rate": 1.6344802104198323e-07,
+      "loss": 0.6013,
+      "step": 6540
+    },
+    {
+      "epoch": 12.112962962962962,
+      "grad_norm": 1.6607233817716947,
+      "learning_rate": 1.6343643069620016e-07,
+      "loss": 0.6376,
+      "step": 6541
+    },
+    {
+      "epoch": 12.114814814814816,
+      "grad_norm": 1.4268066080879431,
+      "learning_rate": 1.6342483892419385e-07,
+      "loss": 0.6167,
+      "step": 6542
+    },
+    {
+      "epoch": 12.116666666666667,
+      "grad_norm": 2.06456398263222,
+      "learning_rate": 1.6341324572622495e-07,
+      "loss": 0.6271,
+      "step": 6543
+    },
+    {
+      "epoch": 12.118518518518519,
+      "grad_norm": 1.405837945099825,
+      "learning_rate": 1.6340165110255399e-07,
+      "loss": 0.6549,
+      "step": 6544
+    },
+    {
+      "epoch": 12.12037037037037,
+      "grad_norm": 1.492356949341976,
+      "learning_rate": 1.6339005505344174e-07,
+      "loss": 0.5785,
+      "step": 6545
+    },
+    {
+      "epoch": 12.122222222222222,
+      "grad_norm": 1.4587434473383574,
+      "learning_rate": 1.633784575791489e-07,
+      "loss": 0.6479,
+      "step": 6546
+    },
+    {
+      "epoch": 12.124074074074073,
+      "grad_norm": 2.016253170059147,
+      "learning_rate": 1.6336685867993621e-07,
+      "loss": 0.6284,
+      "step": 6547
+    },
+    {
+      "epoch": 12.125925925925927,
+      "grad_norm": 1.5193277045240394,
+      "learning_rate": 1.633552583560644e-07,
+      "loss": 0.6331,
+      "step": 6548
+    },
+    {
+      "epoch": 12.127777777777778,
+      "grad_norm": 2.521301034698399,
+      "learning_rate": 1.6334365660779432e-07,
+      "loss": 0.5935,
+      "step": 6549
+    },
+    {
+      "epoch": 12.12962962962963,
+      "grad_norm": 1.6305399538375949,
+      "learning_rate": 1.633320534353868e-07,
+      "loss": 0.6689,
+      "step": 6550
+    },
+    {
+      "epoch": 12.131481481481481,
+      "grad_norm": 2.3557694700756,
+      "learning_rate": 1.6332044883910273e-07,
+      "loss": 0.5922,
+      "step": 6551
+    },
+    {
+      "epoch": 12.133333333333333,
+      "grad_norm": 1.4017311422558687,
+      "learning_rate": 1.63308842819203e-07,
+      "loss": 0.649,
+      "step": 6552
+    },
+    {
+      "epoch": 12.135185185185184,
+      "grad_norm": 1.6734928641771962,
+      "learning_rate": 1.6329723537594852e-07,
+      "loss": 0.5723,
+      "step": 6553
+    },
+    {
+      "epoch": 12.137037037037038,
+      "grad_norm": 1.627155463241186,
+      "learning_rate": 1.6328562650960026e-07,
+      "loss": 0.6265,
+      "step": 6554
+    },
+    {
+      "epoch": 12.13888888888889,
+      "grad_norm": 1.5730465046871955,
+      "learning_rate": 1.6327401622041927e-07,
+      "loss": 0.6173,
+      "step": 6555
+    },
+    {
+      "epoch": 12.14074074074074,
+      "grad_norm": 1.563576987788256,
+      "learning_rate": 1.6326240450866653e-07,
+      "loss": 0.6453,
+      "step": 6556
+    },
+    {
+      "epoch": 12.142592592592592,
+      "grad_norm": 1.8314768806280253,
+      "learning_rate": 1.6325079137460312e-07,
+      "loss": 0.5755,
+      "step": 6557
+    },
+    {
+      "epoch": 12.144444444444444,
+      "grad_norm": 1.6788595507566144,
+      "learning_rate": 1.6323917681849013e-07,
+      "loss": 0.6029,
+      "step": 6558
+    },
+    {
+      "epoch": 12.146296296296295,
+      "grad_norm": 1.6988634075470816,
+      "learning_rate": 1.632275608405887e-07,
+      "loss": 0.6497,
+      "step": 6559
+    },
+    {
+      "epoch": 12.148148148148149,
+      "grad_norm": 1.6743549664174686,
+      "learning_rate": 1.6321594344115996e-07,
+      "loss": 0.6747,
+      "step": 6560
+    },
+    {
+      "epoch": 12.15,
+      "grad_norm": 1.61027545479233,
+      "learning_rate": 1.6320432462046513e-07,
+      "loss": 0.6127,
+      "step": 6561
+    },
+    {
+      "epoch": 12.151851851851852,
+      "grad_norm": 1.5173173449115376,
+      "learning_rate": 1.6319270437876546e-07,
+      "loss": 0.6166,
+      "step": 6562
+    },
+    {
+      "epoch": 12.153703703703703,
+      "grad_norm": 1.96836329381583,
+      "learning_rate": 1.631810827163221e-07,
+      "loss": 0.6854,
+      "step": 6563
+    },
+    {
+      "epoch": 12.155555555555555,
+      "grad_norm": 1.9295510891491456,
+      "learning_rate": 1.6316945963339644e-07,
+      "loss": 0.5745,
+      "step": 6564
+    },
+    {
+      "epoch": 12.157407407407407,
+      "grad_norm": 1.5972623403537682,
+      "learning_rate": 1.6315783513024973e-07,
+      "loss": 0.6476,
+      "step": 6565
+    },
+    {
+      "epoch": 12.15925925925926,
+      "grad_norm": 1.9266437878385532,
+      "learning_rate": 1.631462092071434e-07,
+      "loss": 0.655,
+      "step": 6566
+    },
+    {
+      "epoch": 12.161111111111111,
+      "grad_norm": 1.3488496020973324,
+      "learning_rate": 1.6313458186433878e-07,
+      "loss": 0.6698,
+      "step": 6567
+    },
+    {
+      "epoch": 12.162962962962963,
+      "grad_norm": 2.0155003994474363,
+      "learning_rate": 1.6312295310209726e-07,
+      "loss": 0.6157,
+      "step": 6568
+    },
+    {
+      "epoch": 12.164814814814815,
+      "grad_norm": 1.9602772561395967,
+      "learning_rate": 1.6311132292068033e-07,
+      "loss": 0.6049,
+      "step": 6569
+    },
+    {
+      "epoch": 12.166666666666666,
+      "grad_norm": 1.779101121190205,
+      "learning_rate": 1.6309969132034944e-07,
+      "loss": 0.57,
+      "step": 6570
+    },
+    {
+      "epoch": 12.168518518518518,
+      "grad_norm": 1.7474196550150642,
+      "learning_rate": 1.6308805830136613e-07,
+      "loss": 0.6368,
+      "step": 6571
+    },
+    {
+      "epoch": 12.170370370370371,
+      "grad_norm": 1.5636307263675404,
+      "learning_rate": 1.630764238639919e-07,
+      "loss": 0.5961,
+      "step": 6572
+    },
+    {
+      "epoch": 12.172222222222222,
+      "grad_norm": 1.3911596034294955,
+      "learning_rate": 1.6306478800848837e-07,
+      "loss": 0.6197,
+      "step": 6573
+    },
+    {
+      "epoch": 12.174074074074074,
+      "grad_norm": 1.5910468442516266,
+      "learning_rate": 1.6305315073511712e-07,
+      "loss": 0.622,
+      "step": 6574
+    },
+    {
+      "epoch": 12.175925925925926,
+      "grad_norm": 1.6640548708685794,
+      "learning_rate": 1.6304151204413978e-07,
+      "loss": 0.5823,
+      "step": 6575
+    },
+    {
+      "epoch": 12.177777777777777,
+      "grad_norm": 1.8600778641018916,
+      "learning_rate": 1.6302987193581805e-07,
+      "loss": 0.6503,
+      "step": 6576
+    },
+    {
+      "epoch": 12.17962962962963,
+      "grad_norm": 3.238108196975405,
+      "learning_rate": 1.6301823041041361e-07,
+      "loss": 0.6503,
+      "step": 6577
+    },
+    {
+      "epoch": 12.181481481481482,
+      "grad_norm": 1.5606805385472824,
+      "learning_rate": 1.6300658746818818e-07,
+      "loss": 0.656,
+      "step": 6578
+    },
+    {
+      "epoch": 12.183333333333334,
+      "grad_norm": 1.538704399617346,
+      "learning_rate": 1.6299494310940356e-07,
+      "loss": 0.6139,
+      "step": 6579
+    },
+    {
+      "epoch": 12.185185185185185,
+      "grad_norm": 1.4165390889858942,
+      "learning_rate": 1.629832973343215e-07,
+      "loss": 0.5996,
+      "step": 6580
+    },
+    {
+      "epoch": 12.187037037037037,
+      "grad_norm": 1.5930324318358666,
+      "learning_rate": 1.6297165014320388e-07,
+      "loss": 0.6355,
+      "step": 6581
+    },
+    {
+      "epoch": 12.188888888888888,
+      "grad_norm": 1.9998978511237038,
+      "learning_rate": 1.6296000153631252e-07,
+      "loss": 0.6438,
+      "step": 6582
+    },
+    {
+      "epoch": 12.190740740740742,
+      "grad_norm": 1.5156997176815612,
+      "learning_rate": 1.6294835151390936e-07,
+      "loss": 0.6188,
+      "step": 6583
+    },
+    {
+      "epoch": 12.192592592592593,
+      "grad_norm": 1.4144826874759324,
+      "learning_rate": 1.6293670007625625e-07,
+      "loss": 0.5847,
+      "step": 6584
+    },
+    {
+      "epoch": 12.194444444444445,
+      "grad_norm": 1.6889681940427033,
+      "learning_rate": 1.6292504722361524e-07,
+      "loss": 0.64,
+      "step": 6585
+    },
+    {
+      "epoch": 12.196296296296296,
+      "grad_norm": 1.5517575737154559,
+      "learning_rate": 1.6291339295624825e-07,
+      "loss": 0.62,
+      "step": 6586
+    },
+    {
+      "epoch": 12.198148148148148,
+      "grad_norm": 1.8582824280712988,
+      "learning_rate": 1.6290173727441727e-07,
+      "loss": 0.6011,
+      "step": 6587
+    },
+    {
+      "epoch": 12.2,
+      "grad_norm": 1.56630565582127,
+      "learning_rate": 1.6289008017838445e-07,
+      "loss": 0.5976,
+      "step": 6588
+    },
+    {
+      "epoch": 12.201851851851853,
+      "grad_norm": 1.500845821932329,
+      "learning_rate": 1.628784216684118e-07,
+      "loss": 0.6252,
+      "step": 6589
+    },
+    {
+      "epoch": 12.203703703703704,
+      "grad_norm": 1.5890627908192463,
+      "learning_rate": 1.6286676174476143e-07,
+      "loss": 0.6445,
+      "step": 6590
+    },
+    {
+      "epoch": 12.205555555555556,
+      "grad_norm": 1.5182906680373982,
+      "learning_rate": 1.6285510040769557e-07,
+      "loss": 0.6042,
+      "step": 6591
+    },
+    {
+      "epoch": 12.207407407407407,
+      "grad_norm": 1.7853847086911154,
+      "learning_rate": 1.628434376574763e-07,
+      "loss": 0.6038,
+      "step": 6592
+    },
+    {
+      "epoch": 12.209259259259259,
+      "grad_norm": 1.6615417061196436,
+      "learning_rate": 1.6283177349436587e-07,
+      "loss": 0.6743,
+      "step": 6593
+    },
+    {
+      "epoch": 12.21111111111111,
+      "grad_norm": 1.5307108825072282,
+      "learning_rate": 1.6282010791862654e-07,
+      "loss": 0.618,
+      "step": 6594
+    },
+    {
+      "epoch": 12.212962962962964,
+      "grad_norm": 1.6159445296531232,
+      "learning_rate": 1.6280844093052058e-07,
+      "loss": 0.5802,
+      "step": 6595
+    },
+    {
+      "epoch": 12.214814814814815,
+      "grad_norm": 1.5872601922843572,
+      "learning_rate": 1.6279677253031025e-07,
+      "loss": 0.6365,
+      "step": 6596
+    },
+    {
+      "epoch": 12.216666666666667,
+      "grad_norm": 1.7130597138637507,
+      "learning_rate": 1.6278510271825793e-07,
+      "loss": 0.6163,
+      "step": 6597
+    },
+    {
+      "epoch": 12.218518518518518,
+      "grad_norm": 1.5024369543590266,
+      "learning_rate": 1.62773431494626e-07,
+      "loss": 0.5931,
+      "step": 6598
+    },
+    {
+      "epoch": 12.22037037037037,
+      "grad_norm": 1.4031452228124768,
+      "learning_rate": 1.6276175885967683e-07,
+      "loss": 0.6454,
+      "step": 6599
+    },
+    {
+      "epoch": 12.222222222222221,
+      "grad_norm": 1.6618976941807972,
+      "learning_rate": 1.6275008481367286e-07,
+      "loss": 0.6908,
+      "step": 6600
+    },
+    {
+      "epoch": 12.224074074074075,
+      "grad_norm": 1.6554977564411542,
+      "learning_rate": 1.6273840935687654e-07,
+      "loss": 0.6406,
+      "step": 6601
+    },
+    {
+      "epoch": 12.225925925925926,
+      "grad_norm": 1.614720235985635,
+      "learning_rate": 1.627267324895504e-07,
+      "loss": 0.5723,
+      "step": 6602
+    },
+    {
+      "epoch": 12.227777777777778,
+      "grad_norm": 1.6783746757530102,
+      "learning_rate": 1.6271505421195694e-07,
+      "loss": 0.5999,
+      "step": 6603
+    },
+    {
+      "epoch": 12.22962962962963,
+      "grad_norm": 1.5762241754213928,
+      "learning_rate": 1.6270337452435874e-07,
+      "loss": 0.6318,
+      "step": 6604
+    },
+    {
+      "epoch": 12.231481481481481,
+      "grad_norm": 1.5116313846961869,
+      "learning_rate": 1.6269169342701838e-07,
+      "loss": 0.6202,
+      "step": 6605
+    },
+    {
+      "epoch": 12.233333333333333,
+      "grad_norm": 1.699175362466538,
+      "learning_rate": 1.626800109201985e-07,
+      "loss": 0.6381,
+      "step": 6606
+    },
+    {
+      "epoch": 12.235185185185186,
+      "grad_norm": 1.6300570154576297,
+      "learning_rate": 1.626683270041617e-07,
+      "loss": 0.6303,
+      "step": 6607
+    },
+    {
+      "epoch": 12.237037037037037,
+      "grad_norm": 1.6055710022349048,
+      "learning_rate": 1.6265664167917073e-07,
+      "loss": 0.6651,
+      "step": 6608
+    },
+    {
+      "epoch": 12.238888888888889,
+      "grad_norm": 1.8727886695448739,
+      "learning_rate": 1.6264495494548828e-07,
+      "loss": 0.5932,
+      "step": 6609
+    },
+    {
+      "epoch": 12.24074074074074,
+      "grad_norm": 1.4214025258412486,
+      "learning_rate": 1.626332668033771e-07,
+      "loss": 0.6026,
+      "step": 6610
+    },
+    {
+      "epoch": 12.242592592592592,
+      "grad_norm": 1.3390510790588248,
+      "learning_rate": 1.6262157725309997e-07,
+      "loss": 0.5731,
+      "step": 6611
+    },
+    {
+      "epoch": 12.244444444444444,
+      "grad_norm": 1.7581170186667185,
+      "learning_rate": 1.626098862949197e-07,
+      "loss": 0.6572,
+      "step": 6612
+    },
+    {
+      "epoch": 12.246296296296297,
+      "grad_norm": 1.6051541908675027,
+      "learning_rate": 1.6259819392909913e-07,
+      "loss": 0.6331,
+      "step": 6613
+    },
+    {
+      "epoch": 12.248148148148148,
+      "grad_norm": 1.4941806113201141,
+      "learning_rate": 1.6258650015590116e-07,
+      "loss": 0.5985,
+      "step": 6614
+    },
+    {
+      "epoch": 12.25,
+      "grad_norm": 1.435015237958376,
+      "learning_rate": 1.6257480497558871e-07,
+      "loss": 0.5999,
+      "step": 6615
+    },
+    {
+      "epoch": 12.251851851851852,
+      "grad_norm": 1.9553315189194644,
+      "learning_rate": 1.6256310838842463e-07,
+      "loss": 0.6355,
+      "step": 6616
+    },
+    {
+      "epoch": 12.253703703703703,
+      "grad_norm": 1.6553117790453324,
+      "learning_rate": 1.6255141039467203e-07,
+      "loss": 0.6082,
+      "step": 6617
+    },
+    {
+      "epoch": 12.255555555555556,
+      "grad_norm": 1.5272240377855684,
+      "learning_rate": 1.625397109945938e-07,
+      "loss": 0.6261,
+      "step": 6618
+    },
+    {
+      "epoch": 12.257407407407408,
+      "grad_norm": 1.4704925857518294,
+      "learning_rate": 1.6252801018845302e-07,
+      "loss": 0.6327,
+      "step": 6619
+    },
+    {
+      "epoch": 12.25925925925926,
+      "grad_norm": 1.6127325226807936,
+      "learning_rate": 1.6251630797651273e-07,
+      "loss": 0.5891,
+      "step": 6620
+    },
+    {
+      "epoch": 12.261111111111111,
+      "grad_norm": 1.551038443458036,
+      "learning_rate": 1.6250460435903607e-07,
+      "loss": 0.6259,
+      "step": 6621
+    },
+    {
+      "epoch": 12.262962962962963,
+      "grad_norm": 2.3189172429940585,
+      "learning_rate": 1.624928993362861e-07,
+      "loss": 0.6409,
+      "step": 6622
+    },
+    {
+      "epoch": 12.264814814814814,
+      "grad_norm": 1.3565462691550279,
+      "learning_rate": 1.6248119290852608e-07,
+      "loss": 0.6293,
+      "step": 6623
+    },
+    {
+      "epoch": 12.266666666666667,
+      "grad_norm": 1.9321640098027901,
+      "learning_rate": 1.6246948507601913e-07,
+      "loss": 0.6464,
+      "step": 6624
+    },
+    {
+      "epoch": 12.268518518518519,
+      "grad_norm": 1.8439408605762335,
+      "learning_rate": 1.6245777583902848e-07,
+      "loss": 0.613,
+      "step": 6625
+    },
+    {
+      "epoch": 12.27037037037037,
+      "grad_norm": 1.520599900625449,
+      "learning_rate": 1.624460651978174e-07,
+      "loss": 0.6197,
+      "step": 6626
+    },
+    {
+      "epoch": 12.272222222222222,
+      "grad_norm": 1.5565971027623497,
+      "learning_rate": 1.6243435315264917e-07,
+      "loss": 0.63,
+      "step": 6627
+    },
+    {
+      "epoch": 12.274074074074074,
+      "grad_norm": 1.5292274508510055,
+      "learning_rate": 1.6242263970378711e-07,
+      "loss": 0.6174,
+      "step": 6628
+    },
+    {
+      "epoch": 12.275925925925925,
+      "grad_norm": 1.5064823121259525,
+      "learning_rate": 1.6241092485149456e-07,
+      "loss": 0.6484,
+      "step": 6629
+    },
+    {
+      "epoch": 12.277777777777779,
+      "grad_norm": 1.5585272861276653,
+      "learning_rate": 1.6239920859603494e-07,
+      "loss": 0.6323,
+      "step": 6630
+    },
+    {
+      "epoch": 12.27962962962963,
+      "grad_norm": 1.4883952896329649,
+      "learning_rate": 1.6238749093767163e-07,
+      "loss": 0.6782,
+      "step": 6631
+    },
+    {
+      "epoch": 12.281481481481482,
+      "grad_norm": 1.5806561979946274,
+      "learning_rate": 1.6237577187666808e-07,
+      "loss": 0.6412,
+      "step": 6632
+    },
+    {
+      "epoch": 12.283333333333333,
+      "grad_norm": 2.7180298210675953,
+      "learning_rate": 1.6236405141328777e-07,
+      "loss": 0.6242,
+      "step": 6633
+    },
+    {
+      "epoch": 12.285185185185185,
+      "grad_norm": 1.5444227813345974,
+      "learning_rate": 1.623523295477942e-07,
+      "loss": 0.5554,
+      "step": 6634
+    },
+    {
+      "epoch": 12.287037037037036,
+      "grad_norm": 1.4465858218865122,
+      "learning_rate": 1.6234060628045093e-07,
+      "loss": 0.621,
+      "step": 6635
+    },
+    {
+      "epoch": 12.28888888888889,
+      "grad_norm": 1.425017455330324,
+      "learning_rate": 1.623288816115215e-07,
+      "loss": 0.6189,
+      "step": 6636
+    },
+    {
+      "epoch": 12.290740740740741,
+      "grad_norm": 1.512138796582942,
+      "learning_rate": 1.6231715554126955e-07,
+      "loss": 0.6125,
+      "step": 6637
+    },
+    {
+      "epoch": 12.292592592592593,
+      "grad_norm": 1.4588839274034164,
+      "learning_rate": 1.623054280699587e-07,
+      "loss": 0.6586,
+      "step": 6638
+    },
+    {
+      "epoch": 12.294444444444444,
+      "grad_norm": 1.5547541341306146,
+      "learning_rate": 1.6229369919785255e-07,
+      "loss": 0.5733,
+      "step": 6639
+    },
+    {
+      "epoch": 12.296296296296296,
+      "grad_norm": 1.5848618649918327,
+      "learning_rate": 1.622819689252149e-07,
+      "loss": 0.6168,
+      "step": 6640
+    },
+    {
+      "epoch": 12.298148148148147,
+      "grad_norm": 1.529607003296878,
+      "learning_rate": 1.622702372523094e-07,
+      "loss": 0.6325,
+      "step": 6641
+    },
+    {
+      "epoch": 12.3,
+      "grad_norm": 1.5430004695018045,
+      "learning_rate": 1.6225850417939988e-07,
+      "loss": 0.6033,
+      "step": 6642
+    },
+    {
+      "epoch": 12.301851851851852,
+      "grad_norm": 1.4390134576694458,
+      "learning_rate": 1.622467697067501e-07,
+      "loss": 0.6173,
+      "step": 6643
+    },
+    {
+      "epoch": 12.303703703703704,
+      "grad_norm": 1.9117665942689086,
+      "learning_rate": 1.6223503383462385e-07,
+      "loss": 0.6169,
+      "step": 6644
+    },
+    {
+      "epoch": 12.305555555555555,
+      "grad_norm": 1.5362139899252765,
+      "learning_rate": 1.62223296563285e-07,
+      "loss": 0.616,
+      "step": 6645
+    },
+    {
+      "epoch": 12.307407407407407,
+      "grad_norm": 1.5332623328060124,
+      "learning_rate": 1.6221155789299749e-07,
+      "loss": 0.631,
+      "step": 6646
+    },
+    {
+      "epoch": 12.309259259259258,
+      "grad_norm": 1.6702510202179586,
+      "learning_rate": 1.6219981782402516e-07,
+      "loss": 0.6605,
+      "step": 6647
+    },
+    {
+      "epoch": 12.311111111111112,
+      "grad_norm": 1.7122450359221875,
+      "learning_rate": 1.62188076356632e-07,
+      "loss": 0.6384,
+      "step": 6648
+    },
+    {
+      "epoch": 12.312962962962963,
+      "grad_norm": 1.8828853178244909,
+      "learning_rate": 1.62176333491082e-07,
+      "loss": 0.568,
+      "step": 6649
+    },
+    {
+      "epoch": 12.314814814814815,
+      "grad_norm": 1.6105207777805626,
+      "learning_rate": 1.6216458922763914e-07,
+      "loss": 0.6364,
+      "step": 6650
+    },
+    {
+      "epoch": 12.316666666666666,
+      "grad_norm": 1.5398965599915089,
+      "learning_rate": 1.621528435665675e-07,
+      "loss": 0.6296,
+      "step": 6651
+    },
+    {
+      "epoch": 12.318518518518518,
+      "grad_norm": 1.679341353015418,
+      "learning_rate": 1.6214109650813107e-07,
+      "loss": 0.5752,
+      "step": 6652
+    },
+    {
+      "epoch": 12.32037037037037,
+      "grad_norm": 1.6629218897533944,
+      "learning_rate": 1.6212934805259408e-07,
+      "loss": 0.6229,
+      "step": 6653
+    },
+    {
+      "epoch": 12.322222222222223,
+      "grad_norm": 1.5660911809385762,
+      "learning_rate": 1.6211759820022058e-07,
+      "loss": 0.6368,
+      "step": 6654
+    },
+    {
+      "epoch": 12.324074074074074,
+      "grad_norm": 1.4445883548145186,
+      "learning_rate": 1.6210584695127478e-07,
+      "loss": 0.6101,
+      "step": 6655
+    },
+    {
+      "epoch": 12.325925925925926,
+      "grad_norm": 1.4622276754859247,
+      "learning_rate": 1.6209409430602085e-07,
+      "loss": 0.6072,
+      "step": 6656
+    },
+    {
+      "epoch": 12.327777777777778,
+      "grad_norm": 1.8274624429311774,
+      "learning_rate": 1.6208234026472305e-07,
+      "loss": 0.6663,
+      "step": 6657
+    },
+    {
+      "epoch": 12.329629629629629,
+      "grad_norm": 1.596237889931007,
+      "learning_rate": 1.620705848276456e-07,
+      "loss": 0.6235,
+      "step": 6658
+    },
+    {
+      "epoch": 12.331481481481482,
+      "grad_norm": 1.4381326986273786,
+      "learning_rate": 1.6205882799505284e-07,
+      "loss": 0.6533,
+      "step": 6659
+    },
+    {
+      "epoch": 12.333333333333334,
+      "grad_norm": 1.6141048868813446,
+      "learning_rate": 1.6204706976720907e-07,
+      "loss": 0.6056,
+      "step": 6660
+    },
+    {
+      "epoch": 12.335185185185185,
+      "grad_norm": 1.5752245919685965,
+      "learning_rate": 1.6203531014437867e-07,
+      "loss": 0.5937,
+      "step": 6661
+    },
+    {
+      "epoch": 12.337037037037037,
+      "grad_norm": 2.776928552961708,
+      "learning_rate": 1.6202354912682598e-07,
+      "loss": 0.5596,
+      "step": 6662
+    },
+    {
+      "epoch": 12.338888888888889,
+      "grad_norm": 1.4192099048175186,
+      "learning_rate": 1.620117867148155e-07,
+      "loss": 0.6136,
+      "step": 6663
+    },
+    {
+      "epoch": 12.34074074074074,
+      "grad_norm": 1.372271166929822,
+      "learning_rate": 1.6200002290861158e-07,
+      "loss": 0.5909,
+      "step": 6664
+    },
+    {
+      "epoch": 12.342592592592593,
+      "grad_norm": 4.803161356529421,
+      "learning_rate": 1.619882577084788e-07,
+      "loss": 0.645,
+      "step": 6665
+    },
+    {
+      "epoch": 12.344444444444445,
+      "grad_norm": 1.6570008096125624,
+      "learning_rate": 1.6197649111468162e-07,
+      "loss": 0.6471,
+      "step": 6666
+    },
+    {
+      "epoch": 12.346296296296297,
+      "grad_norm": 1.7599154016337184,
+      "learning_rate": 1.6196472312748458e-07,
+      "loss": 0.5933,
+      "step": 6667
+    },
+    {
+      "epoch": 12.348148148148148,
+      "grad_norm": 1.526100165994934,
+      "learning_rate": 1.619529537471523e-07,
+      "loss": 0.6111,
+      "step": 6668
+    },
+    {
+      "epoch": 12.35,
+      "grad_norm": 1.5675438673126456,
+      "learning_rate": 1.6194118297394935e-07,
+      "loss": 0.5964,
+      "step": 6669
+    },
+    {
+      "epoch": 12.351851851851851,
+      "grad_norm": 1.4495462583452638,
+      "learning_rate": 1.6192941080814034e-07,
+      "loss": 0.6223,
+      "step": 6670
+    },
+    {
+      "epoch": 12.353703703703705,
+      "grad_norm": 2.1730774512058155,
+      "learning_rate": 1.6191763724998998e-07,
+      "loss": 0.6195,
+      "step": 6671
+    },
+    {
+      "epoch": 12.355555555555556,
+      "grad_norm": 3.3725143006871763,
+      "learning_rate": 1.6190586229976302e-07,
+      "loss": 0.6391,
+      "step": 6672
+    },
+    {
+      "epoch": 12.357407407407408,
+      "grad_norm": 1.7214815507989247,
+      "learning_rate": 1.6189408595772406e-07,
+      "loss": 0.5992,
+      "step": 6673
+    },
+    {
+      "epoch": 12.35925925925926,
+      "grad_norm": 1.5272373084204767,
+      "learning_rate": 1.6188230822413802e-07,
+      "loss": 0.6355,
+      "step": 6674
+    },
+    {
+      "epoch": 12.36111111111111,
+      "grad_norm": 1.627865729553433,
+      "learning_rate": 1.618705290992696e-07,
+      "loss": 0.6049,
+      "step": 6675
+    },
+    {
+      "epoch": 12.362962962962962,
+      "grad_norm": 1.9066500323363462,
+      "learning_rate": 1.6185874858338365e-07,
+      "loss": 0.6042,
+      "step": 6676
+    },
+    {
+      "epoch": 12.364814814814816,
+      "grad_norm": 2.04115358499272,
+      "learning_rate": 1.6184696667674504e-07,
+      "loss": 0.6366,
+      "step": 6677
+    },
+    {
+      "epoch": 12.366666666666667,
+      "grad_norm": 3.2010776354981854,
+      "learning_rate": 1.6183518337961862e-07,
+      "loss": 0.6333,
+      "step": 6678
+    },
+    {
+      "epoch": 12.368518518518519,
+      "grad_norm": 1.6365931904066346,
+      "learning_rate": 1.6182339869226933e-07,
+      "loss": 0.5968,
+      "step": 6679
+    },
+    {
+      "epoch": 12.37037037037037,
+      "grad_norm": 3.021436372165264,
+      "learning_rate": 1.6181161261496213e-07,
+      "loss": 0.621,
+      "step": 6680
+    },
+    {
+      "epoch": 12.372222222222222,
+      "grad_norm": 1.472734920206644,
+      "learning_rate": 1.6179982514796202e-07,
+      "loss": 0.6569,
+      "step": 6681
+    },
+    {
+      "epoch": 12.374074074074073,
+      "grad_norm": 1.5195528675956615,
+      "learning_rate": 1.61788036291534e-07,
+      "loss": 0.6035,
+      "step": 6682
+    },
+    {
+      "epoch": 12.375925925925927,
+      "grad_norm": 2.1156055599206214,
+      "learning_rate": 1.617762460459431e-07,
+      "loss": 0.6221,
+      "step": 6683
+    },
+    {
+      "epoch": 12.377777777777778,
+      "grad_norm": 2.0329831742732267,
+      "learning_rate": 1.6176445441145438e-07,
+      "loss": 0.6308,
+      "step": 6684
+    },
+    {
+      "epoch": 12.37962962962963,
+      "grad_norm": 1.6021361103864833,
+      "learning_rate": 1.6175266138833302e-07,
+      "loss": 0.6194,
+      "step": 6685
+    },
+    {
+      "epoch": 12.381481481481481,
+      "grad_norm": 1.5650015936914683,
+      "learning_rate": 1.617408669768441e-07,
+      "loss": 0.6369,
+      "step": 6686
+    },
+    {
+      "epoch": 12.383333333333333,
+      "grad_norm": 1.6215876413509036,
+      "learning_rate": 1.6172907117725279e-07,
+      "loss": 0.6228,
+      "step": 6687
+    },
+    {
+      "epoch": 12.385185185185184,
+      "grad_norm": 1.5098351993197505,
+      "learning_rate": 1.6171727398982428e-07,
+      "loss": 0.5764,
+      "step": 6688
+    },
+    {
+      "epoch": 12.387037037037038,
+      "grad_norm": 1.5395020853703976,
+      "learning_rate": 1.6170547541482388e-07,
+      "loss": 0.628,
+      "step": 6689
+    },
+    {
+      "epoch": 12.38888888888889,
+      "grad_norm": 1.5620088815710957,
+      "learning_rate": 1.616936754525168e-07,
+      "loss": 0.5611,
+      "step": 6690
+    },
+    {
+      "epoch": 12.39074074074074,
+      "grad_norm": 1.7235453802719307,
+      "learning_rate": 1.616818741031683e-07,
+      "loss": 0.6266,
+      "step": 6691
+    },
+    {
+      "epoch": 12.392592592592592,
+      "grad_norm": 1.3915572789105382,
+      "learning_rate": 1.616700713670438e-07,
+      "loss": 0.6384,
+      "step": 6692
+    },
+    {
+      "epoch": 12.394444444444444,
+      "grad_norm": 1.7370381306710112,
+      "learning_rate": 1.6165826724440854e-07,
+      "loss": 0.6192,
+      "step": 6693
+    },
+    {
+      "epoch": 12.396296296296295,
+      "grad_norm": 1.6081971050431976,
+      "learning_rate": 1.6164646173552802e-07,
+      "loss": 0.6453,
+      "step": 6694
+    },
+    {
+      "epoch": 12.398148148148149,
+      "grad_norm": 1.7344468339917467,
+      "learning_rate": 1.616346548406676e-07,
+      "loss": 0.6368,
+      "step": 6695
+    },
+    {
+      "epoch": 12.4,
+      "grad_norm": 1.4422640100939879,
+      "learning_rate": 1.6162284656009274e-07,
+      "loss": 0.6057,
+      "step": 6696
+    },
+    {
+      "epoch": 12.401851851851852,
+      "grad_norm": 1.621497058307677,
+      "learning_rate": 1.616110368940689e-07,
+      "loss": 0.5783,
+      "step": 6697
+    },
+    {
+      "epoch": 12.403703703703703,
+      "grad_norm": 1.546223165874739,
+      "learning_rate": 1.6159922584286164e-07,
+      "loss": 0.671,
+      "step": 6698
+    },
+    {
+      "epoch": 12.405555555555555,
+      "grad_norm": 1.5474537248013673,
+      "learning_rate": 1.615874134067365e-07,
+      "loss": 0.6123,
+      "step": 6699
+    },
+    {
+      "epoch": 12.407407407407407,
+      "grad_norm": 1.4787270387375124,
+      "learning_rate": 1.61575599585959e-07,
+      "loss": 0.6314,
+      "step": 6700
+    },
+    {
+      "epoch": 12.40925925925926,
+      "grad_norm": 1.7273091249486154,
+      "learning_rate": 1.615637843807948e-07,
+      "loss": 0.6376,
+      "step": 6701
+    },
+    {
+      "epoch": 12.411111111111111,
+      "grad_norm": 1.8812011119113892,
+      "learning_rate": 1.6155196779150954e-07,
+      "loss": 0.5996,
+      "step": 6702
+    },
+    {
+      "epoch": 12.412962962962963,
+      "grad_norm": 1.619848258974585,
+      "learning_rate": 1.6154014981836884e-07,
+      "loss": 0.6333,
+      "step": 6703
+    },
+    {
+      "epoch": 12.414814814814815,
+      "grad_norm": 1.3868376269066072,
+      "learning_rate": 1.6152833046163843e-07,
+      "loss": 0.6217,
+      "step": 6704
+    },
+    {
+      "epoch": 12.416666666666666,
+      "grad_norm": 1.7234605081825254,
+      "learning_rate": 1.6151650972158408e-07,
+      "loss": 0.6541,
+      "step": 6705
+    },
+    {
+      "epoch": 12.418518518518518,
+      "grad_norm": 1.732866130801652,
+      "learning_rate": 1.615046875984715e-07,
+      "loss": 0.6189,
+      "step": 6706
+    },
+    {
+      "epoch": 12.420370370370371,
+      "grad_norm": 1.39916696619329,
+      "learning_rate": 1.6149286409256647e-07,
+      "loss": 0.6182,
+      "step": 6707
+    },
+    {
+      "epoch": 12.422222222222222,
+      "grad_norm": 1.6613702358771392,
+      "learning_rate": 1.6148103920413489e-07,
+      "loss": 0.6727,
+      "step": 6708
+    },
+    {
+      "epoch": 12.424074074074074,
+      "grad_norm": 1.821863792053127,
+      "learning_rate": 1.6146921293344254e-07,
+      "loss": 0.5735,
+      "step": 6709
+    },
+    {
+      "epoch": 12.425925925925926,
+      "grad_norm": 1.3577622022312832,
+      "learning_rate": 1.6145738528075535e-07,
+      "loss": 0.6106,
+      "step": 6710
+    },
+    {
+      "epoch": 12.427777777777777,
+      "grad_norm": 1.7602412179167821,
+      "learning_rate": 1.6144555624633918e-07,
+      "loss": 0.6459,
+      "step": 6711
+    },
+    {
+      "epoch": 12.42962962962963,
+      "grad_norm": 1.5101847967225224,
+      "learning_rate": 1.6143372583046006e-07,
+      "loss": 0.6251,
+      "step": 6712
+    },
+    {
+      "epoch": 12.431481481481482,
+      "grad_norm": 1.695320221636631,
+      "learning_rate": 1.6142189403338394e-07,
+      "loss": 0.6341,
+      "step": 6713
+    },
+    {
+      "epoch": 12.433333333333334,
+      "grad_norm": 1.6764135747248183,
+      "learning_rate": 1.614100608553768e-07,
+      "loss": 0.6065,
+      "step": 6714
+    },
+    {
+      "epoch": 12.435185185185185,
+      "grad_norm": 1.5180336853503937,
+      "learning_rate": 1.613982262967047e-07,
+      "loss": 0.6297,
+      "step": 6715
+    },
+    {
+      "epoch": 12.437037037037037,
+      "grad_norm": 1.4451097703190439,
+      "learning_rate": 1.6138639035763375e-07,
+      "loss": 0.5801,
+      "step": 6716
+    },
+    {
+      "epoch": 12.438888888888888,
+      "grad_norm": 1.501209780115876,
+      "learning_rate": 1.6137455303843e-07,
+      "loss": 0.672,
+      "step": 6717
+    },
+    {
+      "epoch": 12.440740740740742,
+      "grad_norm": 1.551995743307302,
+      "learning_rate": 1.6136271433935963e-07,
+      "loss": 0.638,
+      "step": 6718
+    },
+    {
+      "epoch": 12.442592592592593,
+      "grad_norm": 1.444861987967557,
+      "learning_rate": 1.6135087426068876e-07,
+      "loss": 0.6459,
+      "step": 6719
+    },
+    {
+      "epoch": 12.444444444444445,
+      "grad_norm": 1.7419807203155202,
+      "learning_rate": 1.6133903280268362e-07,
+      "loss": 0.6045,
+      "step": 6720
+    },
+    {
+      "epoch": 12.446296296296296,
+      "grad_norm": 2.0614981942526236,
+      "learning_rate": 1.6132718996561044e-07,
+      "loss": 0.6103,
+      "step": 6721
+    },
+    {
+      "epoch": 12.448148148148148,
+      "grad_norm": 1.8342747789336824,
+      "learning_rate": 1.6131534574973543e-07,
+      "loss": 0.6207,
+      "step": 6722
+    },
+    {
+      "epoch": 12.45,
+      "grad_norm": 1.6919490628090412,
+      "learning_rate": 1.6130350015532494e-07,
+      "loss": 0.6331,
+      "step": 6723
+    },
+    {
+      "epoch": 12.451851851851853,
+      "grad_norm": 1.7157488714688314,
+      "learning_rate": 1.612916531826453e-07,
+      "loss": 0.5967,
+      "step": 6724
+    },
+    {
+      "epoch": 12.453703703703704,
+      "grad_norm": 1.6300345835805943,
+      "learning_rate": 1.6127980483196278e-07,
+      "loss": 0.6013,
+      "step": 6725
+    },
+    {
+      "epoch": 12.455555555555556,
+      "grad_norm": 1.4957333485476974,
+      "learning_rate": 1.612679551035438e-07,
+      "loss": 0.6032,
+      "step": 6726
+    },
+    {
+      "epoch": 12.457407407407407,
+      "grad_norm": 1.843506681342497,
+      "learning_rate": 1.6125610399765486e-07,
+      "loss": 0.6412,
+      "step": 6727
+    },
+    {
+      "epoch": 12.459259259259259,
+      "grad_norm": 1.6542256609222146,
+      "learning_rate": 1.6124425151456227e-07,
+      "loss": 0.6732,
+      "step": 6728
+    },
+    {
+      "epoch": 12.46111111111111,
+      "grad_norm": 1.438272911441366,
+      "learning_rate": 1.6123239765453257e-07,
+      "loss": 0.5925,
+      "step": 6729
+    },
+    {
+      "epoch": 12.462962962962964,
+      "grad_norm": 1.5756248187825583,
+      "learning_rate": 1.612205424178323e-07,
+      "loss": 0.6276,
+      "step": 6730
+    },
+    {
+      "epoch": 12.464814814814815,
+      "grad_norm": 1.630278639994648,
+      "learning_rate": 1.6120868580472796e-07,
+      "loss": 0.5784,
+      "step": 6731
+    },
+    {
+      "epoch": 12.466666666666667,
+      "grad_norm": 1.527909196796277,
+      "learning_rate": 1.6119682781548612e-07,
+      "loss": 0.5841,
+      "step": 6732
+    },
+    {
+      "epoch": 12.468518518518518,
+      "grad_norm": 1.4289853368852803,
+      "learning_rate": 1.611849684503734e-07,
+      "loss": 0.6061,
+      "step": 6733
+    },
+    {
+      "epoch": 12.47037037037037,
+      "grad_norm": 1.708498419700212,
+      "learning_rate": 1.6117310770965638e-07,
+      "loss": 0.6562,
+      "step": 6734
+    },
+    {
+      "epoch": 12.472222222222221,
+      "grad_norm": 1.8476025550069186,
+      "learning_rate": 1.6116124559360176e-07,
+      "loss": 0.6044,
+      "step": 6735
+    },
+    {
+      "epoch": 12.474074074074075,
+      "grad_norm": 1.577579629598864,
+      "learning_rate": 1.611493821024762e-07,
+      "loss": 0.6457,
+      "step": 6736
+    },
+    {
+      "epoch": 12.475925925925926,
+      "grad_norm": 1.8771734073270754,
+      "learning_rate": 1.6113751723654652e-07,
+      "loss": 0.6331,
+      "step": 6737
+    },
+    {
+      "epoch": 12.477777777777778,
+      "grad_norm": 1.3985466495227243,
+      "learning_rate": 1.6112565099607936e-07,
+      "loss": 0.6236,
+      "step": 6738
+    },
+    {
+      "epoch": 12.47962962962963,
+      "grad_norm": 1.5931850879221627,
+      "learning_rate": 1.611137833813416e-07,
+      "loss": 0.6501,
+      "step": 6739
+    },
+    {
+      "epoch": 12.481481481481481,
+      "grad_norm": 1.4911377057410506,
+      "learning_rate": 1.6110191439259994e-07,
+      "loss": 0.623,
+      "step": 6740
+    },
+    {
+      "epoch": 12.483333333333333,
+      "grad_norm": 1.493446596010548,
+      "learning_rate": 1.610900440301213e-07,
+      "loss": 0.6062,
+      "step": 6741
+    },
+    {
+      "epoch": 12.485185185185186,
+      "grad_norm": 1.8745480233916807,
+      "learning_rate": 1.610781722941726e-07,
+      "loss": 0.6358,
+      "step": 6742
+    },
+    {
+      "epoch": 12.487037037037037,
+      "grad_norm": 1.4628326994055836,
+      "learning_rate": 1.6106629918502068e-07,
+      "loss": 0.5937,
+      "step": 6743
+    },
+    {
+      "epoch": 12.488888888888889,
+      "grad_norm": 1.4370141627877862,
+      "learning_rate": 1.6105442470293248e-07,
+      "loss": 0.6698,
+      "step": 6744
+    },
+    {
+      "epoch": 12.49074074074074,
+      "grad_norm": 1.4019496182120215,
+      "learning_rate": 1.61042548848175e-07,
+      "loss": 0.6451,
+      "step": 6745
+    },
+    {
+      "epoch": 12.492592592592592,
+      "grad_norm": 1.5803883327203812,
+      "learning_rate": 1.6103067162101523e-07,
+      "loss": 0.6052,
+      "step": 6746
+    },
+    {
+      "epoch": 12.494444444444444,
+      "grad_norm": 1.4849732380091656,
+      "learning_rate": 1.6101879302172024e-07,
+      "loss": 0.6337,
+      "step": 6747
+    },
+    {
+      "epoch": 12.496296296296297,
+      "grad_norm": 1.5523537680839534,
+      "learning_rate": 1.6100691305055705e-07,
+      "loss": 0.6368,
+      "step": 6748
+    },
+    {
+      "epoch": 12.498148148148148,
+      "grad_norm": 1.4503353980944524,
+      "learning_rate": 1.6099503170779275e-07,
+      "loss": 0.6384,
+      "step": 6749
+    },
+    {
+      "epoch": 12.5,
+      "grad_norm": 1.336787061169782,
+      "learning_rate": 1.6098314899369444e-07,
+      "loss": 0.6313,
+      "step": 6750
+    },
+    {
+      "epoch": 12.501851851851852,
+      "grad_norm": 1.58137241698652,
+      "learning_rate": 1.6097126490852936e-07,
+      "loss": 0.6414,
+      "step": 6751
+    },
+    {
+      "epoch": 12.503703703703703,
+      "grad_norm": 1.4015886532799962,
+      "learning_rate": 1.6095937945256463e-07,
+      "loss": 0.6134,
+      "step": 6752
+    },
+    {
+      "epoch": 12.505555555555556,
+      "grad_norm": 3.0615025113085435,
+      "learning_rate": 1.609474926260675e-07,
+      "loss": 0.6017,
+      "step": 6753
+    },
+    {
+      "epoch": 12.507407407407408,
+      "grad_norm": 1.785876336755664,
+      "learning_rate": 1.609356044293052e-07,
+      "loss": 0.6475,
+      "step": 6754
+    },
+    {
+      "epoch": 12.50925925925926,
+      "grad_norm": 1.549268553515418,
+      "learning_rate": 1.6092371486254505e-07,
+      "loss": 0.6153,
+      "step": 6755
+    },
+    {
+      "epoch": 12.511111111111111,
+      "grad_norm": 1.5512241377844795,
+      "learning_rate": 1.6091182392605424e-07,
+      "loss": 0.6657,
+      "step": 6756
+    },
+    {
+      "epoch": 12.512962962962963,
+      "grad_norm": 1.919523385233102,
+      "learning_rate": 1.6089993162010025e-07,
+      "loss": 0.628,
+      "step": 6757
+    },
+    {
+      "epoch": 12.514814814814814,
+      "grad_norm": 1.5779930159400157,
+      "learning_rate": 1.6088803794495039e-07,
+      "loss": 0.6121,
+      "step": 6758
+    },
+    {
+      "epoch": 12.516666666666667,
+      "grad_norm": 1.453727811521583,
+      "learning_rate": 1.6087614290087206e-07,
+      "loss": 0.6021,
+      "step": 6759
+    },
+    {
+      "epoch": 12.518518518518519,
+      "grad_norm": 1.606121439166297,
+      "learning_rate": 1.608642464881327e-07,
+      "loss": 0.5676,
+      "step": 6760
+    },
+    {
+      "epoch": 12.52037037037037,
+      "grad_norm": 1.491608123734651,
+      "learning_rate": 1.608523487069998e-07,
+      "loss": 0.6244,
+      "step": 6761
+    },
+    {
+      "epoch": 12.522222222222222,
+      "grad_norm": 1.4895390869520764,
+      "learning_rate": 1.608404495577408e-07,
+      "loss": 0.64,
+      "step": 6762
+    },
+    {
+      "epoch": 12.524074074074074,
+      "grad_norm": 1.576671313126675,
+      "learning_rate": 1.6082854904062325e-07,
+      "loss": 0.6252,
+      "step": 6763
+    },
+    {
+      "epoch": 12.525925925925925,
+      "grad_norm": 1.744803875587018,
+      "learning_rate": 1.6081664715591472e-07,
+      "loss": 0.5892,
+      "step": 6764
+    },
+    {
+      "epoch": 12.527777777777779,
+      "grad_norm": 1.5210954340070728,
+      "learning_rate": 1.608047439038828e-07,
+      "loss": 0.6102,
+      "step": 6765
+    },
+    {
+      "epoch": 12.52962962962963,
+      "grad_norm": 1.6319152672684785,
+      "learning_rate": 1.6079283928479508e-07,
+      "loss": 0.5924,
+      "step": 6766
+    },
+    {
+      "epoch": 12.531481481481482,
+      "grad_norm": 2.8294307494249327,
+      "learning_rate": 1.607809332989192e-07,
+      "loss": 0.6504,
+      "step": 6767
+    },
+    {
+      "epoch": 12.533333333333333,
+      "grad_norm": 1.704731451763355,
+      "learning_rate": 1.6076902594652288e-07,
+      "loss": 0.5722,
+      "step": 6768
+    },
+    {
+      "epoch": 12.535185185185185,
+      "grad_norm": 1.7691983169778456,
+      "learning_rate": 1.6075711722787385e-07,
+      "loss": 0.6573,
+      "step": 6769
+    },
+    {
+      "epoch": 12.537037037037036,
+      "grad_norm": 2.1711031930002163,
+      "learning_rate": 1.6074520714323977e-07,
+      "loss": 0.5993,
+      "step": 6770
+    },
+    {
+      "epoch": 12.53888888888889,
+      "grad_norm": 1.6125800014290208,
+      "learning_rate": 1.6073329569288843e-07,
+      "loss": 0.6025,
+      "step": 6771
+    },
+    {
+      "epoch": 12.540740740740741,
+      "grad_norm": 1.5873208356131967,
+      "learning_rate": 1.6072138287708767e-07,
+      "loss": 0.6335,
+      "step": 6772
+    },
+    {
+      "epoch": 12.542592592592593,
+      "grad_norm": 1.6046972054598476,
+      "learning_rate": 1.6070946869610532e-07,
+      "loss": 0.5901,
+      "step": 6773
+    },
+    {
+      "epoch": 12.544444444444444,
+      "grad_norm": 1.595423085591024,
+      "learning_rate": 1.6069755315020924e-07,
+      "loss": 0.6045,
+      "step": 6774
+    },
+    {
+      "epoch": 12.546296296296296,
+      "grad_norm": 1.9371582764788189,
+      "learning_rate": 1.606856362396673e-07,
+      "loss": 0.5927,
+      "step": 6775
+    },
+    {
+      "epoch": 12.548148148148147,
+      "grad_norm": 1.5694052249405797,
+      "learning_rate": 1.6067371796474742e-07,
+      "loss": 0.6323,
+      "step": 6776
+    },
+    {
+      "epoch": 12.55,
+      "grad_norm": 1.5614783347936312,
+      "learning_rate": 1.606617983257176e-07,
+      "loss": 0.585,
+      "step": 6777
+    },
+    {
+      "epoch": 12.551851851851852,
+      "grad_norm": 4.647419220422835,
+      "learning_rate": 1.606498773228458e-07,
+      "loss": 0.5856,
+      "step": 6778
+    },
+    {
+      "epoch": 12.553703703703704,
+      "grad_norm": 1.9430957089556058,
+      "learning_rate": 1.606379549564e-07,
+      "loss": 0.6378,
+      "step": 6779
+    },
+    {
+      "epoch": 12.555555555555555,
+      "grad_norm": 1.6921605452649477,
+      "learning_rate": 1.6062603122664831e-07,
+      "loss": 0.6167,
+      "step": 6780
+    },
+    {
+      "epoch": 12.557407407407407,
+      "grad_norm": 1.729361586729261,
+      "learning_rate": 1.6061410613385879e-07,
+      "loss": 0.6161,
+      "step": 6781
+    },
+    {
+      "epoch": 12.559259259259258,
+      "grad_norm": 2.160195075360732,
+      "learning_rate": 1.6060217967829952e-07,
+      "loss": 0.6505,
+      "step": 6782
+    },
+    {
+      "epoch": 12.561111111111112,
+      "grad_norm": 1.4684027863187157,
+      "learning_rate": 1.6059025186023864e-07,
+      "loss": 0.5937,
+      "step": 6783
+    },
+    {
+      "epoch": 12.562962962962963,
+      "grad_norm": 1.746339979668932,
+      "learning_rate": 1.6057832267994437e-07,
+      "loss": 0.6088,
+      "step": 6784
+    },
+    {
+      "epoch": 12.564814814814815,
+      "grad_norm": 1.4975468709928144,
+      "learning_rate": 1.605663921376849e-07,
+      "loss": 0.5995,
+      "step": 6785
+    },
+    {
+      "epoch": 12.566666666666666,
+      "grad_norm": 1.7525048146158066,
+      "learning_rate": 1.6055446023372836e-07,
+      "loss": 0.623,
+      "step": 6786
+    },
+    {
+      "epoch": 12.568518518518518,
+      "grad_norm": 1.6800059561658405,
+      "learning_rate": 1.6054252696834314e-07,
+      "loss": 0.6074,
+      "step": 6787
+    },
+    {
+      "epoch": 12.57037037037037,
+      "grad_norm": 1.9485820795576567,
+      "learning_rate": 1.6053059234179749e-07,
+      "loss": 0.6501,
+      "step": 6788
+    },
+    {
+      "epoch": 12.572222222222223,
+      "grad_norm": 1.5573476411654357,
+      "learning_rate": 1.6051865635435973e-07,
+      "loss": 0.5981,
+      "step": 6789
+    },
+    {
+      "epoch": 12.574074074074074,
+      "grad_norm": 1.5966695575987275,
+      "learning_rate": 1.6050671900629818e-07,
+      "loss": 0.6331,
+      "step": 6790
+    },
+    {
+      "epoch": 12.575925925925926,
+      "grad_norm": 2.451439210832922,
+      "learning_rate": 1.6049478029788127e-07,
+      "loss": 0.6374,
+      "step": 6791
+    },
+    {
+      "epoch": 12.577777777777778,
+      "grad_norm": 2.399746586688835,
+      "learning_rate": 1.604828402293774e-07,
+      "loss": 0.5894,
+      "step": 6792
+    },
+    {
+      "epoch": 12.579629629629629,
+      "grad_norm": 1.7952955023936086,
+      "learning_rate": 1.60470898801055e-07,
+      "loss": 0.6445,
+      "step": 6793
+    },
+    {
+      "epoch": 12.581481481481482,
+      "grad_norm": 1.6228837070931483,
+      "learning_rate": 1.6045895601318258e-07,
+      "loss": 0.5721,
+      "step": 6794
+    },
+    {
+      "epoch": 12.583333333333334,
+      "grad_norm": 1.7018523689577818,
+      "learning_rate": 1.604470118660286e-07,
+      "loss": 0.6271,
+      "step": 6795
+    },
+    {
+      "epoch": 12.585185185185185,
+      "grad_norm": 1.379640586503175,
+      "learning_rate": 1.6043506635986162e-07,
+      "loss": 0.5684,
+      "step": 6796
+    },
+    {
+      "epoch": 12.587037037037037,
+      "grad_norm": 1.7931694900504649,
+      "learning_rate": 1.6042311949495024e-07,
+      "loss": 0.6121,
+      "step": 6797
+    },
+    {
+      "epoch": 12.588888888888889,
+      "grad_norm": 1.7854155733545496,
+      "learning_rate": 1.60411171271563e-07,
+      "loss": 0.6266,
+      "step": 6798
+    },
+    {
+      "epoch": 12.59074074074074,
+      "grad_norm": 1.717485265669741,
+      "learning_rate": 1.6039922168996857e-07,
+      "loss": 0.6164,
+      "step": 6799
+    },
+    {
+      "epoch": 12.592592592592592,
+      "grad_norm": 1.7821970231296957,
+      "learning_rate": 1.6038727075043558e-07,
+      "loss": 0.5831,
+      "step": 6800
+    },
+    {
+      "epoch": 12.594444444444445,
+      "grad_norm": 1.5811840551132714,
+      "learning_rate": 1.6037531845323274e-07,
+      "loss": 0.6138,
+      "step": 6801
+    },
+    {
+      "epoch": 12.596296296296297,
+      "grad_norm": 1.411211185081435,
+      "learning_rate": 1.603633647986288e-07,
+      "loss": 0.6105,
+      "step": 6802
+    },
+    {
+      "epoch": 12.598148148148148,
+      "grad_norm": 1.7168552057894608,
+      "learning_rate": 1.6035140978689242e-07,
+      "loss": 0.6202,
+      "step": 6803
+    },
+    {
+      "epoch": 12.6,
+      "grad_norm": 1.8197273072150357,
+      "learning_rate": 1.6033945341829247e-07,
+      "loss": 0.6073,
+      "step": 6804
+    },
+    {
+      "epoch": 12.601851851851851,
+      "grad_norm": 3.783030194013275,
+      "learning_rate": 1.603274956930977e-07,
+      "loss": 0.6039,
+      "step": 6805
+    },
+    {
+      "epoch": 12.603703703703705,
+      "grad_norm": 1.4465897650135064,
+      "learning_rate": 1.60315536611577e-07,
+      "loss": 0.5757,
+      "step": 6806
+    },
+    {
+      "epoch": 12.605555555555556,
+      "grad_norm": 1.7660923193512383,
+      "learning_rate": 1.603035761739992e-07,
+      "loss": 0.6213,
+      "step": 6807
+    },
+    {
+      "epoch": 12.607407407407408,
+      "grad_norm": 1.5832966737920553,
+      "learning_rate": 1.6029161438063325e-07,
+      "loss": 0.6123,
+      "step": 6808
+    },
+    {
+      "epoch": 12.60925925925926,
+      "grad_norm": 1.4625725710762831,
+      "learning_rate": 1.6027965123174806e-07,
+      "loss": 0.6183,
+      "step": 6809
+    },
+    {
+      "epoch": 12.61111111111111,
+      "grad_norm": 1.4750028356899958,
+      "learning_rate": 1.602676867276126e-07,
+      "loss": 0.6155,
+      "step": 6810
+    },
+    {
+      "epoch": 12.612962962962962,
+      "grad_norm": 1.4285935297020942,
+      "learning_rate": 1.6025572086849583e-07,
+      "loss": 0.6567,
+      "step": 6811
+    },
+    {
+      "epoch": 12.614814814814816,
+      "grad_norm": 1.7802253508382944,
+      "learning_rate": 1.6024375365466683e-07,
+      "loss": 0.5838,
+      "step": 6812
+    },
+    {
+      "epoch": 12.616666666666667,
+      "grad_norm": 1.563898316628401,
+      "learning_rate": 1.6023178508639462e-07,
+      "loss": 0.6276,
+      "step": 6813
+    },
+    {
+      "epoch": 12.618518518518519,
+      "grad_norm": 1.6908489757004845,
+      "learning_rate": 1.6021981516394826e-07,
+      "loss": 0.598,
+      "step": 6814
+    },
+    {
+      "epoch": 12.62037037037037,
+      "grad_norm": 1.5568319661955032,
+      "learning_rate": 1.6020784388759695e-07,
+      "loss": 0.617,
+      "step": 6815
+    },
+    {
+      "epoch": 12.622222222222222,
+      "grad_norm": 1.61159548424979,
+      "learning_rate": 1.6019587125760977e-07,
+      "loss": 0.5927,
+      "step": 6816
+    },
+    {
+      "epoch": 12.624074074074073,
+      "grad_norm": 1.4522496756525654,
+      "learning_rate": 1.6018389727425592e-07,
+      "loss": 0.6019,
+      "step": 6817
+    },
+    {
+      "epoch": 12.625925925925927,
+      "grad_norm": 1.6529772869382366,
+      "learning_rate": 1.601719219378046e-07,
+      "loss": 0.6439,
+      "step": 6818
+    },
+    {
+      "epoch": 12.627777777777778,
+      "grad_norm": 1.7132252520891806,
+      "learning_rate": 1.6015994524852504e-07,
+      "loss": 0.6447,
+      "step": 6819
+    },
+    {
+      "epoch": 12.62962962962963,
+      "grad_norm": 1.4998307154610613,
+      "learning_rate": 1.6014796720668652e-07,
+      "loss": 0.6528,
+      "step": 6820
+    },
+    {
+      "epoch": 12.631481481481481,
+      "grad_norm": 1.5692311404371204,
+      "learning_rate": 1.6013598781255831e-07,
+      "loss": 0.6305,
+      "step": 6821
+    },
+    {
+      "epoch": 12.633333333333333,
+      "grad_norm": 1.543445686295679,
+      "learning_rate": 1.6012400706640983e-07,
+      "loss": 0.6889,
+      "step": 6822
+    },
+    {
+      "epoch": 12.635185185185184,
+      "grad_norm": 2.3071664843383792,
+      "learning_rate": 1.6011202496851034e-07,
+      "loss": 0.6086,
+      "step": 6823
+    },
+    {
+      "epoch": 12.637037037037038,
+      "grad_norm": 1.7648127834545924,
+      "learning_rate": 1.6010004151912926e-07,
+      "loss": 0.6412,
+      "step": 6824
+    },
+    {
+      "epoch": 12.63888888888889,
+      "grad_norm": 1.4903289633097423,
+      "learning_rate": 1.6008805671853605e-07,
+      "loss": 0.6048,
+      "step": 6825
+    },
+    {
+      "epoch": 12.64074074074074,
+      "grad_norm": 1.5962944557466139,
+      "learning_rate": 1.600760705670001e-07,
+      "loss": 0.649,
+      "step": 6826
+    },
+    {
+      "epoch": 12.642592592592592,
+      "grad_norm": 1.581592432581292,
+      "learning_rate": 1.600640830647909e-07,
+      "loss": 0.6207,
+      "step": 6827
+    },
+    {
+      "epoch": 12.644444444444444,
+      "grad_norm": 1.8286967803900258,
+      "learning_rate": 1.6005209421217803e-07,
+      "loss": 0.6303,
+      "step": 6828
+    },
+    {
+      "epoch": 12.646296296296295,
+      "grad_norm": 1.3631116369340788,
+      "learning_rate": 1.6004010400943097e-07,
+      "loss": 0.6053,
+      "step": 6829
+    },
+    {
+      "epoch": 12.648148148148149,
+      "grad_norm": 1.4722906278412151,
+      "learning_rate": 1.6002811245681929e-07,
+      "loss": 0.6349,
+      "step": 6830
+    },
+    {
+      "epoch": 12.65,
+      "grad_norm": 1.7000184921463328,
+      "learning_rate": 1.6001611955461264e-07,
+      "loss": 0.6113,
+      "step": 6831
+    },
+    {
+      "epoch": 12.651851851851852,
+      "grad_norm": 1.557111446109822,
+      "learning_rate": 1.6000412530308055e-07,
+      "loss": 0.5934,
+      "step": 6832
+    },
+    {
+      "epoch": 12.653703703703703,
+      "grad_norm": 3.9814872391656166,
+      "learning_rate": 1.5999212970249282e-07,
+      "loss": 0.6469,
+      "step": 6833
+    },
+    {
+      "epoch": 12.655555555555555,
+      "grad_norm": 1.4576137386246166,
+      "learning_rate": 1.5998013275311905e-07,
+      "loss": 0.5807,
+      "step": 6834
+    },
+    {
+      "epoch": 12.657407407407408,
+      "grad_norm": 1.4728520668214509,
+      "learning_rate": 1.5996813445522902e-07,
+      "loss": 0.6252,
+      "step": 6835
+    },
+    {
+      "epoch": 12.65925925925926,
+      "grad_norm": 2.160072341824441,
+      "learning_rate": 1.5995613480909243e-07,
+      "loss": 0.6194,
+      "step": 6836
+    },
+    {
+      "epoch": 12.661111111111111,
+      "grad_norm": 1.7776333713749342,
+      "learning_rate": 1.5994413381497908e-07,
+      "loss": 0.6191,
+      "step": 6837
+    },
+    {
+      "epoch": 12.662962962962963,
+      "grad_norm": 1.4491885027916003,
+      "learning_rate": 1.5993213147315886e-07,
+      "loss": 0.637,
+      "step": 6838
+    },
+    {
+      "epoch": 12.664814814814815,
+      "grad_norm": 1.5962358372956102,
+      "learning_rate": 1.5992012778390147e-07,
+      "loss": 0.6044,
+      "step": 6839
+    },
+    {
+      "epoch": 12.666666666666666,
+      "grad_norm": 1.5968504474145202,
+      "learning_rate": 1.5990812274747693e-07,
+      "loss": 0.6383,
+      "step": 6840
+    },
+    {
+      "epoch": 12.668518518518518,
+      "grad_norm": 1.568961725746722,
+      "learning_rate": 1.5989611636415504e-07,
+      "loss": 0.6345,
+      "step": 6841
+    },
+    {
+      "epoch": 12.670370370370371,
+      "grad_norm": 1.6413495044092195,
+      "learning_rate": 1.5988410863420578e-07,
+      "loss": 0.6342,
+      "step": 6842
+    },
+    {
+      "epoch": 12.672222222222222,
+      "grad_norm": 1.7594545805007782,
+      "learning_rate": 1.5987209955789913e-07,
+      "loss": 0.572,
+      "step": 6843
+    },
+    {
+      "epoch": 12.674074074074074,
+      "grad_norm": 1.7350443329675036,
+      "learning_rate": 1.5986008913550508e-07,
+      "loss": 0.6188,
+      "step": 6844
+    },
+    {
+      "epoch": 12.675925925925926,
+      "grad_norm": 1.7694067535008446,
+      "learning_rate": 1.5984807736729363e-07,
+      "loss": 0.581,
+      "step": 6845
+    },
+    {
+      "epoch": 12.677777777777777,
+      "grad_norm": 1.5073123721987824,
+      "learning_rate": 1.5983606425353488e-07,
+      "loss": 0.5741,
+      "step": 6846
+    },
+    {
+      "epoch": 12.67962962962963,
+      "grad_norm": 2.026385496371346,
+      "learning_rate": 1.5982404979449889e-07,
+      "loss": 0.6204,
+      "step": 6847
+    },
+    {
+      "epoch": 12.681481481481482,
+      "grad_norm": 1.9565160971881237,
+      "learning_rate": 1.5981203399045576e-07,
+      "loss": 0.5927,
+      "step": 6848
+    },
+    {
+      "epoch": 12.683333333333334,
+      "grad_norm": 1.4061477408305016,
+      "learning_rate": 1.5980001684167567e-07,
+      "loss": 0.6142,
+      "step": 6849
+    },
+    {
+      "epoch": 12.685185185185185,
+      "grad_norm": 1.484356137606153,
+      "learning_rate": 1.597879983484288e-07,
+      "loss": 0.6054,
+      "step": 6850
+    },
+    {
+      "epoch": 12.687037037037037,
+      "grad_norm": 1.6292511337993587,
+      "learning_rate": 1.597759785109853e-07,
+      "loss": 0.6308,
+      "step": 6851
+    },
+    {
+      "epoch": 12.688888888888888,
+      "grad_norm": 1.7648019980114067,
+      "learning_rate": 1.597639573296155e-07,
+      "loss": 0.5971,
+      "step": 6852
+    },
+    {
+      "epoch": 12.690740740740742,
+      "grad_norm": 1.5122404590378937,
+      "learning_rate": 1.5975193480458962e-07,
+      "loss": 0.6165,
+      "step": 6853
+    },
+    {
+      "epoch": 12.692592592592593,
+      "grad_norm": 2.1351292203467325,
+      "learning_rate": 1.5973991093617794e-07,
+      "loss": 0.61,
+      "step": 6854
+    },
+    {
+      "epoch": 12.694444444444445,
+      "grad_norm": 1.770293579050696,
+      "learning_rate": 1.5972788572465081e-07,
+      "loss": 0.6206,
+      "step": 6855
+    },
+    {
+      "epoch": 12.696296296296296,
+      "grad_norm": 1.6241130942930582,
+      "learning_rate": 1.597158591702786e-07,
+      "loss": 0.6677,
+      "step": 6856
+    },
+    {
+      "epoch": 12.698148148148148,
+      "grad_norm": 2.2842202003581766,
+      "learning_rate": 1.597038312733317e-07,
+      "loss": 0.6571,
+      "step": 6857
+    },
+    {
+      "epoch": 12.7,
+      "grad_norm": 1.6896403378883804,
+      "learning_rate": 1.596918020340805e-07,
+      "loss": 0.6133,
+      "step": 6858
+    },
+    {
+      "epoch": 12.701851851851853,
+      "grad_norm": 1.3828107158238865,
+      "learning_rate": 1.596797714527955e-07,
+      "loss": 0.6183,
+      "step": 6859
+    },
+    {
+      "epoch": 12.703703703703704,
+      "grad_norm": 1.5329900756319352,
+      "learning_rate": 1.5966773952974711e-07,
+      "loss": 0.5826,
+      "step": 6860
+    },
+    {
+      "epoch": 12.705555555555556,
+      "grad_norm": 1.3821215581887945,
+      "learning_rate": 1.596557062652059e-07,
+      "loss": 0.6027,
+      "step": 6861
+    },
+    {
+      "epoch": 12.707407407407407,
+      "grad_norm": 1.5824028104828178,
+      "learning_rate": 1.596436716594424e-07,
+      "loss": 0.6302,
+      "step": 6862
+    },
+    {
+      "epoch": 12.709259259259259,
+      "grad_norm": 10.540412692913531,
+      "learning_rate": 1.5963163571272717e-07,
+      "loss": 0.6208,
+      "step": 6863
+    },
+    {
+      "epoch": 12.71111111111111,
+      "grad_norm": 1.2948847490202897,
+      "learning_rate": 1.596195984253308e-07,
+      "loss": 0.5975,
+      "step": 6864
+    },
+    {
+      "epoch": 12.712962962962964,
+      "grad_norm": 1.648527501671313,
+      "learning_rate": 1.5960755979752395e-07,
+      "loss": 0.5892,
+      "step": 6865
+    },
+    {
+      "epoch": 12.714814814814815,
+      "grad_norm": 1.5823797338444276,
+      "learning_rate": 1.5959551982957726e-07,
+      "loss": 0.6602,
+      "step": 6866
+    },
+    {
+      "epoch": 12.716666666666667,
+      "grad_norm": 1.5387005759209103,
+      "learning_rate": 1.595834785217614e-07,
+      "loss": 0.6281,
+      "step": 6867
+    },
+    {
+      "epoch": 12.718518518518518,
+      "grad_norm": 1.464430108480258,
+      "learning_rate": 1.5957143587434715e-07,
+      "loss": 0.6149,
+      "step": 6868
+    },
+    {
+      "epoch": 12.72037037037037,
+      "grad_norm": 1.9434408643543009,
+      "learning_rate": 1.595593918876052e-07,
+      "loss": 0.5966,
+      "step": 6869
+    },
+    {
+      "epoch": 12.722222222222221,
+      "grad_norm": 1.8048751824567402,
+      "learning_rate": 1.5954734656180638e-07,
+      "loss": 0.622,
+      "step": 6870
+    },
+    {
+      "epoch": 12.724074074074075,
+      "grad_norm": 1.521188957720398,
+      "learning_rate": 1.5953529989722148e-07,
+      "loss": 0.6098,
+      "step": 6871
+    },
+    {
+      "epoch": 12.725925925925926,
+      "grad_norm": 1.9586869029061609,
+      "learning_rate": 1.5952325189412134e-07,
+      "loss": 0.634,
+      "step": 6872
+    },
+    {
+      "epoch": 12.727777777777778,
+      "grad_norm": 1.6100028935188142,
+      "learning_rate": 1.5951120255277682e-07,
+      "loss": 0.609,
+      "step": 6873
+    },
+    {
+      "epoch": 12.72962962962963,
+      "grad_norm": 1.5161375239048271,
+      "learning_rate": 1.5949915187345885e-07,
+      "loss": 0.6724,
+      "step": 6874
+    },
+    {
+      "epoch": 12.731481481481481,
+      "grad_norm": 1.5647460359540073,
+      "learning_rate": 1.5948709985643834e-07,
+      "loss": 0.6354,
+      "step": 6875
+    },
+    {
+      "epoch": 12.733333333333333,
+      "grad_norm": 1.7624037027376658,
+      "learning_rate": 1.5947504650198627e-07,
+      "loss": 0.5979,
+      "step": 6876
+    },
+    {
+      "epoch": 12.735185185185186,
+      "grad_norm": 1.5096608238512021,
+      "learning_rate": 1.5946299181037362e-07,
+      "loss": 0.597,
+      "step": 6877
+    },
+    {
+      "epoch": 12.737037037037037,
+      "grad_norm": 1.7792576772758397,
+      "learning_rate": 1.594509357818714e-07,
+      "loss": 0.6325,
+      "step": 6878
+    },
+    {
+      "epoch": 12.738888888888889,
+      "grad_norm": 1.5436864718796397,
+      "learning_rate": 1.5943887841675073e-07,
+      "loss": 0.6263,
+      "step": 6879
+    },
+    {
+      "epoch": 12.74074074074074,
+      "grad_norm": 1.8466790091564382,
+      "learning_rate": 1.5942681971528258e-07,
+      "loss": 0.6453,
+      "step": 6880
+    },
+    {
+      "epoch": 12.742592592592592,
+      "grad_norm": 1.4496392228453736,
+      "learning_rate": 1.5941475967773814e-07,
+      "loss": 0.616,
+      "step": 6881
+    },
+    {
+      "epoch": 12.744444444444444,
+      "grad_norm": 1.5716183231308856,
+      "learning_rate": 1.5940269830438852e-07,
+      "loss": 0.6161,
+      "step": 6882
+    },
+    {
+      "epoch": 12.746296296296297,
+      "grad_norm": 1.9364924908466352,
+      "learning_rate": 1.5939063559550495e-07,
+      "loss": 0.606,
+      "step": 6883
+    },
+    {
+      "epoch": 12.748148148148148,
+      "grad_norm": 1.6666008967833845,
+      "learning_rate": 1.5937857155135857e-07,
+      "loss": 0.599,
+      "step": 6884
+    },
+    {
+      "epoch": 12.75,
+      "grad_norm": 2.188154980147139,
+      "learning_rate": 1.593665061722206e-07,
+      "loss": 0.6101,
+      "step": 6885
+    },
+    {
+      "epoch": 12.751851851851852,
+      "grad_norm": 1.711231187628782,
+      "learning_rate": 1.5935443945836237e-07,
+      "loss": 0.564,
+      "step": 6886
+    },
+    {
+      "epoch": 12.753703703703703,
+      "grad_norm": 1.9539432327794675,
+      "learning_rate": 1.593423714100551e-07,
+      "loss": 0.6371,
+      "step": 6887
+    },
+    {
+      "epoch": 12.755555555555556,
+      "grad_norm": 1.4984715345635842,
+      "learning_rate": 1.593303020275702e-07,
+      "loss": 0.5845,
+      "step": 6888
+    },
+    {
+      "epoch": 12.757407407407408,
+      "grad_norm": 1.6281732241446114,
+      "learning_rate": 1.5931823131117893e-07,
+      "loss": 0.5992,
+      "step": 6889
+    },
+    {
+      "epoch": 12.75925925925926,
+      "grad_norm": 1.5732456862049327,
+      "learning_rate": 1.5930615926115273e-07,
+      "loss": 0.5998,
+      "step": 6890
+    },
+    {
+      "epoch": 12.761111111111111,
+      "grad_norm": 1.6336412374430413,
+      "learning_rate": 1.5929408587776303e-07,
+      "loss": 0.6292,
+      "step": 6891
+    },
+    {
+      "epoch": 12.762962962962963,
+      "grad_norm": 1.6595460275293108,
+      "learning_rate": 1.5928201116128118e-07,
+      "loss": 0.626,
+      "step": 6892
+    },
+    {
+      "epoch": 12.764814814814814,
+      "grad_norm": 1.7743089326630253,
+      "learning_rate": 1.5926993511197876e-07,
+      "loss": 0.6265,
+      "step": 6893
+    },
+    {
+      "epoch": 12.766666666666667,
+      "grad_norm": 1.4690682030967814,
+      "learning_rate": 1.592578577301272e-07,
+      "loss": 0.5921,
+      "step": 6894
+    },
+    {
+      "epoch": 12.768518518518519,
+      "grad_norm": 1.9112448566114246,
+      "learning_rate": 1.5924577901599808e-07,
+      "loss": 0.6525,
+      "step": 6895
+    },
+    {
+      "epoch": 12.77037037037037,
+      "grad_norm": 1.5096988302004613,
+      "learning_rate": 1.5923369896986291e-07,
+      "loss": 0.6245,
+      "step": 6896
+    },
+    {
+      "epoch": 12.772222222222222,
+      "grad_norm": 1.4499871108322002,
+      "learning_rate": 1.5922161759199332e-07,
+      "loss": 0.6091,
+      "step": 6897
+    },
+    {
+      "epoch": 12.774074074074074,
+      "grad_norm": 1.3584437905518292,
+      "learning_rate": 1.5920953488266094e-07,
+      "loss": 0.6296,
+      "step": 6898
+    },
+    {
+      "epoch": 12.775925925925925,
+      "grad_norm": 1.4414205456580722,
+      "learning_rate": 1.5919745084213737e-07,
+      "loss": 0.5988,
+      "step": 6899
+    },
+    {
+      "epoch": 12.777777777777779,
+      "grad_norm": 1.7255951279242674,
+      "learning_rate": 1.5918536547069434e-07,
+      "loss": 0.6115,
+      "step": 6900
+    },
+    {
+      "epoch": 12.77962962962963,
+      "grad_norm": 3.674450625944859,
+      "learning_rate": 1.591732787686036e-07,
+      "loss": 0.6289,
+      "step": 6901
+    },
+    {
+      "epoch": 12.781481481481482,
+      "grad_norm": 1.708081372876104,
+      "learning_rate": 1.5916119073613675e-07,
+      "loss": 0.636,
+      "step": 6902
+    },
+    {
+      "epoch": 12.783333333333333,
+      "grad_norm": 1.7121803383373337,
+      "learning_rate": 1.591491013735657e-07,
+      "loss": 0.5837,
+      "step": 6903
+    },
+    {
+      "epoch": 12.785185185185185,
+      "grad_norm": 1.4779633527254814,
+      "learning_rate": 1.591370106811622e-07,
+      "loss": 0.5874,
+      "step": 6904
+    },
+    {
+      "epoch": 12.787037037037036,
+      "grad_norm": 1.8615589615450188,
+      "learning_rate": 1.5912491865919808e-07,
+      "loss": 0.6467,
+      "step": 6905
+    },
+    {
+      "epoch": 12.78888888888889,
+      "grad_norm": 1.7408483278296838,
+      "learning_rate": 1.5911282530794524e-07,
+      "loss": 0.6539,
+      "step": 6906
+    },
+    {
+      "epoch": 12.790740740740741,
+      "grad_norm": 1.7013220753869909,
+      "learning_rate": 1.5910073062767552e-07,
+      "loss": 0.6232,
+      "step": 6907
+    },
+    {
+      "epoch": 12.792592592592593,
+      "grad_norm": 1.7918484485255404,
+      "learning_rate": 1.5908863461866085e-07,
+      "loss": 0.6218,
+      "step": 6908
+    },
+    {
+      "epoch": 12.794444444444444,
+      "grad_norm": 1.5314227850314672,
+      "learning_rate": 1.5907653728117317e-07,
+      "loss": 0.6062,
+      "step": 6909
+    },
+    {
+      "epoch": 12.796296296296296,
+      "grad_norm": 1.7631125019601799,
+      "learning_rate": 1.5906443861548453e-07,
+      "loss": 0.5881,
+      "step": 6910
+    },
+    {
+      "epoch": 12.798148148148147,
+      "grad_norm": 1.4775292759926024,
+      "learning_rate": 1.5905233862186686e-07,
+      "loss": 0.5994,
+      "step": 6911
+    },
+    {
+      "epoch": 12.8,
+      "grad_norm": 1.863183881251582,
+      "learning_rate": 1.5904023730059226e-07,
+      "loss": 0.6068,
+      "step": 6912
+    },
+    {
+      "epoch": 12.801851851851852,
+      "grad_norm": 1.5371736650753063,
+      "learning_rate": 1.590281346519328e-07,
+      "loss": 0.6102,
+      "step": 6913
+    },
+    {
+      "epoch": 12.803703703703704,
+      "grad_norm": 1.733718031950228,
+      "learning_rate": 1.5901603067616046e-07,
+      "loss": 0.6045,
+      "step": 6914
+    },
+    {
+      "epoch": 12.805555555555555,
+      "grad_norm": 1.4975076115540258,
+      "learning_rate": 1.5900392537354754e-07,
+      "loss": 0.6265,
+      "step": 6915
+    },
+    {
+      "epoch": 12.807407407407407,
+      "grad_norm": 1.5923752420447312,
+      "learning_rate": 1.5899181874436612e-07,
+      "loss": 0.5812,
+      "step": 6916
+    },
+    {
+      "epoch": 12.809259259259258,
+      "grad_norm": 1.6211205434507365,
+      "learning_rate": 1.5897971078888836e-07,
+      "loss": 0.6326,
+      "step": 6917
+    },
+    {
+      "epoch": 12.811111111111112,
+      "grad_norm": 1.6139633455756157,
+      "learning_rate": 1.5896760150738656e-07,
+      "loss": 0.6246,
+      "step": 6918
+    },
+    {
+      "epoch": 12.812962962962963,
+      "grad_norm": 1.5812682066273467,
+      "learning_rate": 1.589554909001329e-07,
+      "loss": 0.6211,
+      "step": 6919
+    },
+    {
+      "epoch": 12.814814814814815,
+      "grad_norm": 1.4334726785478022,
+      "learning_rate": 1.5894337896739968e-07,
+      "loss": 0.6422,
+      "step": 6920
+    },
+    {
+      "epoch": 12.816666666666666,
+      "grad_norm": 1.9094142664963845,
+      "learning_rate": 1.5893126570945922e-07,
+      "loss": 0.5732,
+      "step": 6921
+    },
+    {
+      "epoch": 12.818518518518518,
+      "grad_norm": 1.8636920512448556,
+      "learning_rate": 1.5891915112658385e-07,
+      "loss": 0.6007,
+      "step": 6922
+    },
+    {
+      "epoch": 12.82037037037037,
+      "grad_norm": 1.5996726394467193,
+      "learning_rate": 1.5890703521904593e-07,
+      "loss": 0.6209,
+      "step": 6923
+    },
+    {
+      "epoch": 12.822222222222223,
+      "grad_norm": 1.5243014029250295,
+      "learning_rate": 1.5889491798711791e-07,
+      "loss": 0.6179,
+      "step": 6924
+    },
+    {
+      "epoch": 12.824074074074074,
+      "grad_norm": 1.7310053055441255,
+      "learning_rate": 1.5888279943107213e-07,
+      "loss": 0.5861,
+      "step": 6925
+    },
+    {
+      "epoch": 12.825925925925926,
+      "grad_norm": 1.6452361730581457,
+      "learning_rate": 1.5887067955118111e-07,
+      "loss": 0.5392,
+      "step": 6926
+    },
+    {
+      "epoch": 12.827777777777778,
+      "grad_norm": 1.4391954697054465,
+      "learning_rate": 1.5885855834771734e-07,
+      "loss": 0.619,
+      "step": 6927
+    },
+    {
+      "epoch": 12.829629629629629,
+      "grad_norm": 1.9432479194480319,
+      "learning_rate": 1.5884643582095331e-07,
+      "loss": 0.5908,
+      "step": 6928
+    },
+    {
+      "epoch": 12.831481481481482,
+      "grad_norm": 1.4715285715227817,
+      "learning_rate": 1.5883431197116157e-07,
+      "loss": 0.5974,
+      "step": 6929
+    },
+    {
+      "epoch": 12.833333333333334,
+      "grad_norm": 1.5714643873072307,
+      "learning_rate": 1.5882218679861472e-07,
+      "loss": 0.6449,
+      "step": 6930
+    },
+    {
+      "epoch": 12.835185185185185,
+      "grad_norm": 1.647207993509161,
+      "learning_rate": 1.5881006030358535e-07,
+      "loss": 0.6473,
+      "step": 6931
+    },
+    {
+      "epoch": 12.837037037037037,
+      "grad_norm": 1.7201150031593586,
+      "learning_rate": 1.5879793248634608e-07,
+      "loss": 0.6268,
+      "step": 6932
+    },
+    {
+      "epoch": 12.838888888888889,
+      "grad_norm": 1.4418662937024034,
+      "learning_rate": 1.587858033471696e-07,
+      "loss": 0.6755,
+      "step": 6933
+    },
+    {
+      "epoch": 12.84074074074074,
+      "grad_norm": 1.4569044943214895,
+      "learning_rate": 1.5877367288632862e-07,
+      "loss": 0.6244,
+      "step": 6934
+    },
+    {
+      "epoch": 12.842592592592592,
+      "grad_norm": 1.6720215772575935,
+      "learning_rate": 1.5876154110409585e-07,
+      "loss": 0.6174,
+      "step": 6935
+    },
+    {
+      "epoch": 12.844444444444445,
+      "grad_norm": 1.4168324603730078,
+      "learning_rate": 1.5874940800074402e-07,
+      "loss": 0.6202,
+      "step": 6936
+    },
+    {
+      "epoch": 12.846296296296297,
+      "grad_norm": 1.7694777187568598,
+      "learning_rate": 1.5873727357654593e-07,
+      "loss": 0.6351,
+      "step": 6937
+    },
+    {
+      "epoch": 12.848148148148148,
+      "grad_norm": 2.1419415089418496,
+      "learning_rate": 1.5872513783177442e-07,
+      "loss": 0.6219,
+      "step": 6938
+    },
+    {
+      "epoch": 12.85,
+      "grad_norm": 1.8717743149142723,
+      "learning_rate": 1.5871300076670233e-07,
+      "loss": 0.6287,
+      "step": 6939
+    },
+    {
+      "epoch": 12.851851851851851,
+      "grad_norm": 1.943203339902939,
+      "learning_rate": 1.587008623816025e-07,
+      "loss": 0.6563,
+      "step": 6940
+    },
+    {
+      "epoch": 12.853703703703705,
+      "grad_norm": 2.703950936589274,
+      "learning_rate": 1.5868872267674785e-07,
+      "loss": 0.605,
+      "step": 6941
+    },
+    {
+      "epoch": 12.855555555555556,
+      "grad_norm": 1.624191640151769,
+      "learning_rate": 1.5867658165241133e-07,
+      "loss": 0.6027,
+      "step": 6942
+    },
+    {
+      "epoch": 12.857407407407408,
+      "grad_norm": 1.3510287854323508,
+      "learning_rate": 1.5866443930886588e-07,
+      "loss": 0.6227,
+      "step": 6943
+    },
+    {
+      "epoch": 12.85925925925926,
+      "grad_norm": 2.6022799631669016,
+      "learning_rate": 1.5865229564638452e-07,
+      "loss": 0.5973,
+      "step": 6944
+    },
+    {
+      "epoch": 12.86111111111111,
+      "grad_norm": 1.7242501443944458,
+      "learning_rate": 1.5864015066524026e-07,
+      "loss": 0.6333,
+      "step": 6945
+    },
+    {
+      "epoch": 12.862962962962962,
+      "grad_norm": 1.5306756716384318,
+      "learning_rate": 1.5862800436570613e-07,
+      "loss": 0.6768,
+      "step": 6946
+    },
+    {
+      "epoch": 12.864814814814816,
+      "grad_norm": 1.518392861496032,
+      "learning_rate": 1.5861585674805524e-07,
+      "loss": 0.6613,
+      "step": 6947
+    },
+    {
+      "epoch": 12.866666666666667,
+      "grad_norm": 1.5098754002583945,
+      "learning_rate": 1.586037078125607e-07,
+      "loss": 0.6401,
+      "step": 6948
+    },
+    {
+      "epoch": 12.868518518518519,
+      "grad_norm": 1.4303987804425788,
+      "learning_rate": 1.5859155755949563e-07,
+      "loss": 0.6113,
+      "step": 6949
+    },
+    {
+      "epoch": 12.87037037037037,
+      "grad_norm": 1.449785974003684,
+      "learning_rate": 1.5857940598913323e-07,
+      "loss": 0.6035,
+      "step": 6950
+    },
+    {
+      "epoch": 12.872222222222222,
+      "grad_norm": 1.4559852264373372,
+      "learning_rate": 1.5856725310174667e-07,
+      "loss": 0.5877,
+      "step": 6951
+    },
+    {
+      "epoch": 12.874074074074073,
+      "grad_norm": 1.6857319457332511,
+      "learning_rate": 1.5855509889760917e-07,
+      "loss": 0.6279,
+      "step": 6952
+    },
+    {
+      "epoch": 12.875925925925927,
+      "grad_norm": 1.7141514974392464,
+      "learning_rate": 1.5854294337699406e-07,
+      "loss": 0.6412,
+      "step": 6953
+    },
+    {
+      "epoch": 12.877777777777778,
+      "grad_norm": 1.6928035016161531,
+      "learning_rate": 1.5853078654017456e-07,
+      "loss": 0.6583,
+      "step": 6954
+    },
+    {
+      "epoch": 12.87962962962963,
+      "grad_norm": 1.6064487142525163,
+      "learning_rate": 1.5851862838742397e-07,
+      "loss": 0.6488,
+      "step": 6955
+    },
+    {
+      "epoch": 12.881481481481481,
+      "grad_norm": 1.7822770556018663,
+      "learning_rate": 1.5850646891901575e-07,
+      "loss": 0.6116,
+      "step": 6956
+    },
+    {
+      "epoch": 12.883333333333333,
+      "grad_norm": 1.7613025405935119,
+      "learning_rate": 1.584943081352232e-07,
+      "loss": 0.6449,
+      "step": 6957
+    },
+    {
+      "epoch": 12.885185185185184,
+      "grad_norm": 1.580041681251821,
+      "learning_rate": 1.5848214603631972e-07,
+      "loss": 0.6504,
+      "step": 6958
+    },
+    {
+      "epoch": 12.887037037037038,
+      "grad_norm": 1.8036107019762868,
+      "learning_rate": 1.5846998262257872e-07,
+      "loss": 0.639,
+      "step": 6959
+    },
+    {
+      "epoch": 12.88888888888889,
+      "grad_norm": 1.5809629190605328,
+      "learning_rate": 1.5845781789427375e-07,
+      "loss": 0.6227,
+      "step": 6960
+    },
+    {
+      "epoch": 12.89074074074074,
+      "grad_norm": 1.5945560407753643,
+      "learning_rate": 1.5844565185167825e-07,
+      "loss": 0.6073,
+      "step": 6961
+    },
+    {
+      "epoch": 12.892592592592592,
+      "grad_norm": 1.831433857915805,
+      "learning_rate": 1.584334844950658e-07,
+      "loss": 0.6301,
+      "step": 6962
+    },
+    {
+      "epoch": 12.894444444444444,
+      "grad_norm": 1.568157789412509,
+      "learning_rate": 1.584213158247099e-07,
+      "loss": 0.5873,
+      "step": 6963
+    },
+    {
+      "epoch": 12.896296296296295,
+      "grad_norm": 1.4298543076861676,
+      "learning_rate": 1.5840914584088412e-07,
+      "loss": 0.5826,
+      "step": 6964
+    },
+    {
+      "epoch": 12.898148148148149,
+      "grad_norm": 1.6206316375965726,
+      "learning_rate": 1.5839697454386212e-07,
+      "loss": 0.6048,
+      "step": 6965
+    },
+    {
+      "epoch": 12.9,
+      "grad_norm": 1.7415934414137002,
+      "learning_rate": 1.5838480193391753e-07,
+      "loss": 0.6275,
+      "step": 6966
+    },
+    {
+      "epoch": 12.901851851851852,
+      "grad_norm": 1.5390717404474823,
+      "learning_rate": 1.58372628011324e-07,
+      "loss": 0.6191,
+      "step": 6967
+    },
+    {
+      "epoch": 12.903703703703703,
+      "grad_norm": 1.6978985594300364,
+      "learning_rate": 1.583604527763553e-07,
+      "loss": 0.5884,
+      "step": 6968
+    },
+    {
+      "epoch": 12.905555555555555,
+      "grad_norm": 1.4935896724237472,
+      "learning_rate": 1.583482762292851e-07,
+      "loss": 0.612,
+      "step": 6969
+    },
+    {
+      "epoch": 12.907407407407408,
+      "grad_norm": 1.63411268014853,
+      "learning_rate": 1.5833609837038716e-07,
+      "loss": 0.6193,
+      "step": 6970
+    },
+    {
+      "epoch": 12.90925925925926,
+      "grad_norm": 1.7637309920632391,
+      "learning_rate": 1.583239191999353e-07,
+      "loss": 0.6573,
+      "step": 6971
+    },
+    {
+      "epoch": 12.911111111111111,
+      "grad_norm": 1.4453907275111304,
+      "learning_rate": 1.5831173871820336e-07,
+      "loss": 0.6431,
+      "step": 6972
+    },
+    {
+      "epoch": 12.912962962962963,
+      "grad_norm": 1.7008633559195536,
+      "learning_rate": 1.5829955692546514e-07,
+      "loss": 0.6554,
+      "step": 6973
+    },
+    {
+      "epoch": 12.914814814814815,
+      "grad_norm": 1.512307483605181,
+      "learning_rate": 1.5828737382199452e-07,
+      "loss": 0.6205,
+      "step": 6974
+    },
+    {
+      "epoch": 12.916666666666666,
+      "grad_norm": 1.5766727888728036,
+      "learning_rate": 1.5827518940806543e-07,
+      "loss": 0.6312,
+      "step": 6975
+    },
+    {
+      "epoch": 12.918518518518518,
+      "grad_norm": 1.6919612820452719,
+      "learning_rate": 1.5826300368395184e-07,
+      "loss": 0.6676,
+      "step": 6976
+    },
+    {
+      "epoch": 12.920370370370371,
+      "grad_norm": 1.7250342390113371,
+      "learning_rate": 1.5825081664992767e-07,
+      "loss": 0.6012,
+      "step": 6977
+    },
+    {
+      "epoch": 12.922222222222222,
+      "grad_norm": 1.4318583791806843,
+      "learning_rate": 1.5823862830626694e-07,
+      "loss": 0.6358,
+      "step": 6978
+    },
+    {
+      "epoch": 12.924074074074074,
+      "grad_norm": 1.7413914272721611,
+      "learning_rate": 1.5822643865324365e-07,
+      "loss": 0.6111,
+      "step": 6979
+    },
+    {
+      "epoch": 12.925925925925926,
+      "grad_norm": 1.622968580250366,
+      "learning_rate": 1.5821424769113191e-07,
+      "loss": 0.6435,
+      "step": 6980
+    },
+    {
+      "epoch": 12.927777777777777,
+      "grad_norm": 1.5138749289499516,
+      "learning_rate": 1.5820205542020578e-07,
+      "loss": 0.5911,
+      "step": 6981
+    },
+    {
+      "epoch": 12.92962962962963,
+      "grad_norm": 1.643142531046615,
+      "learning_rate": 1.5818986184073935e-07,
+      "loss": 0.5724,
+      "step": 6982
+    },
+    {
+      "epoch": 12.931481481481482,
+      "grad_norm": 1.423707992260151,
+      "learning_rate": 1.5817766695300678e-07,
+      "loss": 0.6058,
+      "step": 6983
+    },
+    {
+      "epoch": 12.933333333333334,
+      "grad_norm": 1.3639529825988501,
+      "learning_rate": 1.5816547075728226e-07,
+      "loss": 0.6187,
+      "step": 6984
+    },
+    {
+      "epoch": 12.935185185185185,
+      "grad_norm": 1.5587202908539697,
+      "learning_rate": 1.5815327325383998e-07,
+      "loss": 0.6304,
+      "step": 6985
+    },
+    {
+      "epoch": 12.937037037037037,
+      "grad_norm": 1.4850261269029026,
+      "learning_rate": 1.5814107444295415e-07,
+      "loss": 0.6534,
+      "step": 6986
+    },
+    {
+      "epoch": 12.938888888888888,
+      "grad_norm": 1.8257773458899504,
+      "learning_rate": 1.581288743248991e-07,
+      "loss": 0.5648,
+      "step": 6987
+    },
+    {
+      "epoch": 12.940740740740742,
+      "grad_norm": 1.7394621738283145,
+      "learning_rate": 1.5811667289994907e-07,
+      "loss": 0.6233,
+      "step": 6988
+    },
+    {
+      "epoch": 12.942592592592593,
+      "grad_norm": 1.6293088979028831,
+      "learning_rate": 1.5810447016837838e-07,
+      "loss": 0.6221,
+      "step": 6989
+    },
+    {
+      "epoch": 12.944444444444445,
+      "grad_norm": 4.663484142314459,
+      "learning_rate": 1.580922661304614e-07,
+      "loss": 0.6473,
+      "step": 6990
+    },
+    {
+      "epoch": 12.946296296296296,
+      "grad_norm": 1.6690144557644095,
+      "learning_rate": 1.5808006078647248e-07,
+      "loss": 0.6027,
+      "step": 6991
+    },
+    {
+      "epoch": 12.948148148148148,
+      "grad_norm": 2.1842778211814156,
+      "learning_rate": 1.5806785413668606e-07,
+      "loss": 0.6497,
+      "step": 6992
+    },
+    {
+      "epoch": 12.95,
+      "grad_norm": 1.4722082470509708,
+      "learning_rate": 1.5805564618137658e-07,
+      "loss": 0.6214,
+      "step": 6993
+    },
+    {
+      "epoch": 12.951851851851853,
+      "grad_norm": 1.6097807020775359,
+      "learning_rate": 1.5804343692081846e-07,
+      "loss": 0.6343,
+      "step": 6994
+    },
+    {
+      "epoch": 12.953703703703704,
+      "grad_norm": 1.4633283074982837,
+      "learning_rate": 1.580312263552863e-07,
+      "loss": 0.6109,
+      "step": 6995
+    },
+    {
+      "epoch": 12.955555555555556,
+      "grad_norm": 1.5217539361475318,
+      "learning_rate": 1.5801901448505453e-07,
+      "loss": 0.59,
+      "step": 6996
+    },
+    {
+      "epoch": 12.957407407407407,
+      "grad_norm": 1.6648615011599652,
+      "learning_rate": 1.5800680131039772e-07,
+      "loss": 0.5859,
+      "step": 6997
+    },
+    {
+      "epoch": 12.959259259259259,
+      "grad_norm": 1.4857559300464391,
+      "learning_rate": 1.579945868315905e-07,
+      "loss": 0.6228,
+      "step": 6998
+    },
+    {
+      "epoch": 12.96111111111111,
+      "grad_norm": 1.6245193304820784,
+      "learning_rate": 1.5798237104890743e-07,
+      "loss": 0.6035,
+      "step": 6999
+    },
+    {
+      "epoch": 12.962962962962964,
+      "grad_norm": 1.7805554382779154,
+      "learning_rate": 1.579701539626232e-07,
+      "loss": 0.6077,
+      "step": 7000
+    },
+    {
+      "epoch": 12.964814814814815,
+      "grad_norm": 1.576147586724914,
+      "learning_rate": 1.5795793557301244e-07,
+      "loss": 0.6575,
+      "step": 7001
+    },
+    {
+      "epoch": 12.966666666666667,
+      "grad_norm": 1.6560448293355672,
+      "learning_rate": 1.579457158803499e-07,
+      "loss": 0.5856,
+      "step": 7002
+    },
+    {
+      "epoch": 12.968518518518518,
+      "grad_norm": 1.5203719283934036,
+      "learning_rate": 1.579334948849103e-07,
+      "loss": 0.5934,
+      "step": 7003
+    },
+    {
+      "epoch": 12.97037037037037,
+      "grad_norm": 1.4615112475366825,
+      "learning_rate": 1.5792127258696838e-07,
+      "loss": 0.6516,
+      "step": 7004
+    },
+    {
+      "epoch": 12.972222222222221,
+      "grad_norm": 1.346287901487352,
+      "learning_rate": 1.5790904898679892e-07,
+      "loss": 0.6213,
+      "step": 7005
+    },
+    {
+      "epoch": 12.974074074074075,
+      "grad_norm": 1.4355437594059939,
+      "learning_rate": 1.5789682408467677e-07,
+      "loss": 0.6156,
+      "step": 7006
+    },
+    {
+      "epoch": 12.975925925925926,
+      "grad_norm": 1.5389726803122858,
+      "learning_rate": 1.578845978808768e-07,
+      "loss": 0.6665,
+      "step": 7007
+    },
+    {
+      "epoch": 12.977777777777778,
+      "grad_norm": 1.6814046833548433,
+      "learning_rate": 1.5787237037567382e-07,
+      "loss": 0.6415,
+      "step": 7008
+    },
+    {
+      "epoch": 12.97962962962963,
+      "grad_norm": 1.609268038728392,
+      "learning_rate": 1.5786014156934277e-07,
+      "loss": 0.6169,
+      "step": 7009
+    },
+    {
+      "epoch": 12.981481481481481,
+      "grad_norm": 1.6100856550206866,
+      "learning_rate": 1.578479114621586e-07,
+      "loss": 0.5911,
+      "step": 7010
+    },
+    {
+      "epoch": 12.983333333333333,
+      "grad_norm": 1.5181380655697292,
+      "learning_rate": 1.5783568005439629e-07,
+      "loss": 0.6462,
+      "step": 7011
+    },
+    {
+      "epoch": 12.985185185185186,
+      "grad_norm": 1.3889241714683453,
+      "learning_rate": 1.5782344734633078e-07,
+      "loss": 0.6237,
+      "step": 7012
+    },
+    {
+      "epoch": 12.987037037037037,
+      "grad_norm": 1.3594050534025974,
+      "learning_rate": 1.5781121333823716e-07,
+      "loss": 0.6205,
+      "step": 7013
+    },
+    {
+      "epoch": 12.988888888888889,
+      "grad_norm": 1.4884411938890487,
+      "learning_rate": 1.5779897803039045e-07,
+      "loss": 0.5819,
+      "step": 7014
+    },
+    {
+      "epoch": 12.99074074074074,
+      "grad_norm": 1.4055462518083262,
+      "learning_rate": 1.5778674142306572e-07,
+      "loss": 0.6102,
+      "step": 7015
+    },
+    {
+      "epoch": 12.992592592592592,
+      "grad_norm": 1.7603573039322413,
+      "learning_rate": 1.577745035165381e-07,
+      "loss": 0.6351,
+      "step": 7016
+    },
+    {
+      "epoch": 12.994444444444444,
+      "grad_norm": 6.30460241917364,
+      "learning_rate": 1.5776226431108273e-07,
+      "loss": 0.583,
+      "step": 7017
+    },
+    {
+      "epoch": 12.996296296296297,
+      "grad_norm": 1.4766214122948749,
+      "learning_rate": 1.5775002380697474e-07,
+      "loss": 0.6199,
+      "step": 7018
+    },
+    {
+      "epoch": 12.998148148148148,
+      "grad_norm": 1.5270464215188726,
+      "learning_rate": 1.5773778200448944e-07,
+      "loss": 0.6206,
+      "step": 7019
+    },
+    {
+      "epoch": 13.0,
+      "grad_norm": 1.6216188368603912,
+      "learning_rate": 1.5772553890390196e-07,
+      "loss": 0.597,
+      "step": 7020
+    },
+    {
+      "epoch": 13.001851851851852,
+      "grad_norm": 1.4958417777720983,
+      "learning_rate": 1.5771329450548757e-07,
+      "loss": 0.6209,
+      "step": 7021
+    },
+    {
+      "epoch": 13.003703703703703,
+      "grad_norm": 1.475676342937982,
+      "learning_rate": 1.5770104880952157e-07,
+      "loss": 0.6296,
+      "step": 7022
+    },
+    {
+      "epoch": 13.005555555555556,
+      "grad_norm": 1.4018004363315908,
+      "learning_rate": 1.576888018162793e-07,
+      "loss": 0.6302,
+      "step": 7023
+    },
+    {
+      "epoch": 13.007407407407408,
+      "grad_norm": 1.9329323468873099,
+      "learning_rate": 1.5767655352603606e-07,
+      "loss": 0.6204,
+      "step": 7024
+    },
+    {
+      "epoch": 13.00925925925926,
+      "grad_norm": 1.9800664296948831,
+      "learning_rate": 1.5766430393906727e-07,
+      "loss": 0.6187,
+      "step": 7025
+    },
+    {
+      "epoch": 13.011111111111111,
+      "grad_norm": 1.382362839353423,
+      "learning_rate": 1.576520530556483e-07,
+      "loss": 0.6031,
+      "step": 7026
+    },
+    {
+      "epoch": 13.012962962962963,
+      "grad_norm": 1.608236489165492,
+      "learning_rate": 1.5763980087605456e-07,
+      "loss": 0.6426,
+      "step": 7027
+    },
+    {
+      "epoch": 13.014814814814814,
+      "grad_norm": 2.225481161842278,
+      "learning_rate": 1.5762754740056163e-07,
+      "loss": 0.6479,
+      "step": 7028
+    },
+    {
+      "epoch": 13.016666666666667,
+      "grad_norm": 1.8838160375784836,
+      "learning_rate": 1.5761529262944487e-07,
+      "loss": 0.6572,
+      "step": 7029
+    },
+    {
+      "epoch": 13.018518518518519,
+      "grad_norm": 1.8351727014629593,
+      "learning_rate": 1.5760303656297988e-07,
+      "loss": 0.5973,
+      "step": 7030
+    },
+    {
+      "epoch": 13.02037037037037,
+      "grad_norm": 1.72451616371401,
+      "learning_rate": 1.5759077920144216e-07,
+      "loss": 0.5977,
+      "step": 7031
+    },
+    {
+      "epoch": 13.022222222222222,
+      "grad_norm": 1.560677502276557,
+      "learning_rate": 1.575785205451073e-07,
+      "loss": 0.6225,
+      "step": 7032
+    },
+    {
+      "epoch": 13.024074074074074,
+      "grad_norm": 1.79634903907718,
+      "learning_rate": 1.5756626059425093e-07,
+      "loss": 0.6229,
+      "step": 7033
+    },
+    {
+      "epoch": 13.025925925925925,
+      "grad_norm": 1.717439029823767,
+      "learning_rate": 1.5755399934914868e-07,
+      "loss": 0.6283,
+      "step": 7034
+    },
+    {
+      "epoch": 13.027777777777779,
+      "grad_norm": 1.542868156702427,
+      "learning_rate": 1.575417368100762e-07,
+      "loss": 0.5781,
+      "step": 7035
+    },
+    {
+      "epoch": 13.02962962962963,
+      "grad_norm": 1.5881375922006693,
+      "learning_rate": 1.575294729773092e-07,
+      "loss": 0.5924,
+      "step": 7036
+    },
+    {
+      "epoch": 13.031481481481482,
+      "grad_norm": 1.8992551398870448,
+      "learning_rate": 1.575172078511234e-07,
+      "loss": 0.5847,
+      "step": 7037
+    },
+    {
+      "epoch": 13.033333333333333,
+      "grad_norm": 1.6845763999849395,
+      "learning_rate": 1.5750494143179453e-07,
+      "loss": 0.6345,
+      "step": 7038
+    },
+    {
+      "epoch": 13.035185185185185,
+      "grad_norm": 1.4250985818152622,
+      "learning_rate": 1.5749267371959843e-07,
+      "loss": 0.6317,
+      "step": 7039
+    },
+    {
+      "epoch": 13.037037037037036,
+      "grad_norm": 1.6941437706055298,
+      "learning_rate": 1.5748040471481088e-07,
+      "loss": 0.6216,
+      "step": 7040
+    },
+    {
+      "epoch": 13.03888888888889,
+      "grad_norm": 1.6374785398221825,
+      "learning_rate": 1.574681344177077e-07,
+      "loss": 0.6524,
+      "step": 7041
+    },
+    {
+      "epoch": 13.040740740740741,
+      "grad_norm": 1.6135624625140734,
+      "learning_rate": 1.5745586282856475e-07,
+      "loss": 0.5744,
+      "step": 7042
+    },
+    {
+      "epoch": 13.042592592592593,
+      "grad_norm": 1.8245704324093361,
+      "learning_rate": 1.57443589947658e-07,
+      "loss": 0.6129,
+      "step": 7043
+    },
+    {
+      "epoch": 13.044444444444444,
+      "grad_norm": 2.6964552210065826,
+      "learning_rate": 1.5743131577526332e-07,
+      "loss": 0.6274,
+      "step": 7044
+    },
+    {
+      "epoch": 13.046296296296296,
+      "grad_norm": 1.46257375695365,
+      "learning_rate": 1.574190403116567e-07,
+      "loss": 0.5931,
+      "step": 7045
+    },
+    {
+      "epoch": 13.048148148148147,
+      "grad_norm": 1.4312721588467456,
+      "learning_rate": 1.574067635571141e-07,
+      "loss": 0.6314,
+      "step": 7046
+    },
+    {
+      "epoch": 13.05,
+      "grad_norm": 1.5314058617642976,
+      "learning_rate": 1.573944855119115e-07,
+      "loss": 0.6102,
+      "step": 7047
+    },
+    {
+      "epoch": 13.051851851851852,
+      "grad_norm": 1.7135392632513267,
+      "learning_rate": 1.57382206176325e-07,
+      "loss": 0.6992,
+      "step": 7048
+    },
+    {
+      "epoch": 13.053703703703704,
+      "grad_norm": 1.8784351213062935,
+      "learning_rate": 1.5736992555063066e-07,
+      "loss": 0.627,
+      "step": 7049
+    },
+    {
+      "epoch": 13.055555555555555,
+      "grad_norm": 2.750254011801841,
+      "learning_rate": 1.573576436351046e-07,
+      "loss": 0.5894,
+      "step": 7050
+    },
+    {
+      "epoch": 13.057407407407407,
+      "grad_norm": 1.5886573150655008,
+      "learning_rate": 1.5734536043002292e-07,
+      "loss": 0.6178,
+      "step": 7051
+    },
+    {
+      "epoch": 13.059259259259258,
+      "grad_norm": 1.66627603545752,
+      "learning_rate": 1.573330759356618e-07,
+      "loss": 0.6044,
+      "step": 7052
+    },
+    {
+      "epoch": 13.061111111111112,
+      "grad_norm": 1.813911209527428,
+      "learning_rate": 1.573207901522974e-07,
+      "loss": 0.5918,
+      "step": 7053
+    },
+    {
+      "epoch": 13.062962962962963,
+      "grad_norm": 1.6198117244874142,
+      "learning_rate": 1.5730850308020597e-07,
+      "loss": 0.5649,
+      "step": 7054
+    },
+    {
+      "epoch": 13.064814814814815,
+      "grad_norm": 1.5884962898388135,
+      "learning_rate": 1.5729621471966375e-07,
+      "loss": 0.6669,
+      "step": 7055
+    },
+    {
+      "epoch": 13.066666666666666,
+      "grad_norm": 1.5741596935430127,
+      "learning_rate": 1.5728392507094696e-07,
+      "loss": 0.6259,
+      "step": 7056
+    },
+    {
+      "epoch": 13.068518518518518,
+      "grad_norm": 1.5108260359930092,
+      "learning_rate": 1.57271634134332e-07,
+      "loss": 0.617,
+      "step": 7057
+    },
+    {
+      "epoch": 13.07037037037037,
+      "grad_norm": 1.7346379844524948,
+      "learning_rate": 1.5725934191009514e-07,
+      "loss": 0.6281,
+      "step": 7058
+    },
+    {
+      "epoch": 13.072222222222223,
+      "grad_norm": 1.890229380921913,
+      "learning_rate": 1.5724704839851277e-07,
+      "loss": 0.6647,
+      "step": 7059
+    },
+    {
+      "epoch": 13.074074074074074,
+      "grad_norm": 1.6543594369475483,
+      "learning_rate": 1.5723475359986126e-07,
+      "loss": 0.5998,
+      "step": 7060
+    },
+    {
+      "epoch": 13.075925925925926,
+      "grad_norm": 1.620814655848228,
+      "learning_rate": 1.5722245751441708e-07,
+      "loss": 0.5585,
+      "step": 7061
+    },
+    {
+      "epoch": 13.077777777777778,
+      "grad_norm": 2.058514693021922,
+      "learning_rate": 1.5721016014245658e-07,
+      "loss": 0.5706,
+      "step": 7062
+    },
+    {
+      "epoch": 13.079629629629629,
+      "grad_norm": 2.0494715950841305,
+      "learning_rate": 1.5719786148425633e-07,
+      "loss": 0.6655,
+      "step": 7063
+    },
+    {
+      "epoch": 13.081481481481482,
+      "grad_norm": 1.705113359979874,
+      "learning_rate": 1.571855615400928e-07,
+      "loss": 0.6096,
+      "step": 7064
+    },
+    {
+      "epoch": 13.083333333333334,
+      "grad_norm": 1.4199487604869294,
+      "learning_rate": 1.5717326031024254e-07,
+      "loss": 0.6267,
+      "step": 7065
+    },
+    {
+      "epoch": 13.085185185185185,
+      "grad_norm": 3.5770625576567316,
+      "learning_rate": 1.5716095779498213e-07,
+      "loss": 0.6881,
+      "step": 7066
+    },
+    {
+      "epoch": 13.087037037037037,
+      "grad_norm": 3.338691735492136,
+      "learning_rate": 1.571486539945881e-07,
+      "loss": 0.6055,
+      "step": 7067
+    },
+    {
+      "epoch": 13.088888888888889,
+      "grad_norm": 1.4570520402825595,
+      "learning_rate": 1.5713634890933715e-07,
+      "loss": 0.6193,
+      "step": 7068
+    },
+    {
+      "epoch": 13.09074074074074,
+      "grad_norm": 1.5439779787829557,
+      "learning_rate": 1.5712404253950587e-07,
+      "loss": 0.6189,
+      "step": 7069
+    },
+    {
+      "epoch": 13.092592592592593,
+      "grad_norm": 1.4409118852189011,
+      "learning_rate": 1.57111734885371e-07,
+      "loss": 0.6166,
+      "step": 7070
+    },
+    {
+      "epoch": 13.094444444444445,
+      "grad_norm": 2.323020085454377,
+      "learning_rate": 1.5709942594720918e-07,
+      "loss": 0.5975,
+      "step": 7071
+    },
+    {
+      "epoch": 13.096296296296297,
+      "grad_norm": 1.7822081002871646,
+      "learning_rate": 1.570871157252972e-07,
+      "loss": 0.6218,
+      "step": 7072
+    },
+    {
+      "epoch": 13.098148148148148,
+      "grad_norm": 1.4554626851309314,
+      "learning_rate": 1.5707480421991183e-07,
+      "loss": 0.6137,
+      "step": 7073
+    },
+    {
+      "epoch": 13.1,
+      "grad_norm": 1.8352518797954986,
+      "learning_rate": 1.570624914313298e-07,
+      "loss": 0.6119,
+      "step": 7074
+    },
+    {
+      "epoch": 13.101851851851851,
+      "grad_norm": 1.5115228936448304,
+      "learning_rate": 1.5705017735982801e-07,
+      "loss": 0.6396,
+      "step": 7075
+    },
+    {
+      "epoch": 13.103703703703705,
+      "grad_norm": 1.5479924722574259,
+      "learning_rate": 1.570378620056833e-07,
+      "loss": 0.6371,
+      "step": 7076
+    },
+    {
+      "epoch": 13.105555555555556,
+      "grad_norm": 1.7283385592692195,
+      "learning_rate": 1.5702554536917254e-07,
+      "loss": 0.6128,
+      "step": 7077
+    },
+    {
+      "epoch": 13.107407407407408,
+      "grad_norm": 1.5073843938268008,
+      "learning_rate": 1.5701322745057262e-07,
+      "loss": 0.6133,
+      "step": 7078
+    },
+    {
+      "epoch": 13.10925925925926,
+      "grad_norm": 2.171924463991556,
+      "learning_rate": 1.5700090825016051e-07,
+      "loss": 0.6039,
+      "step": 7079
+    },
+    {
+      "epoch": 13.11111111111111,
+      "grad_norm": 1.4858684453635949,
+      "learning_rate": 1.5698858776821318e-07,
+      "loss": 0.6126,
+      "step": 7080
+    },
+    {
+      "epoch": 13.112962962962962,
+      "grad_norm": 1.5847552179941182,
+      "learning_rate": 1.569762660050076e-07,
+      "loss": 0.6034,
+      "step": 7081
+    },
+    {
+      "epoch": 13.114814814814816,
+      "grad_norm": 1.556310345714512,
+      "learning_rate": 1.5696394296082084e-07,
+      "loss": 0.5952,
+      "step": 7082
+    },
+    {
+      "epoch": 13.116666666666667,
+      "grad_norm": 1.4876415424146252,
+      "learning_rate": 1.5695161863592992e-07,
+      "loss": 0.6369,
+      "step": 7083
+    },
+    {
+      "epoch": 13.118518518518519,
+      "grad_norm": 1.8342770872925007,
+      "learning_rate": 1.5693929303061193e-07,
+      "loss": 0.6302,
+      "step": 7084
+    },
+    {
+      "epoch": 13.12037037037037,
+      "grad_norm": 1.4306705325041016,
+      "learning_rate": 1.5692696614514401e-07,
+      "loss": 0.616,
+      "step": 7085
+    },
+    {
+      "epoch": 13.122222222222222,
+      "grad_norm": 1.4972129471710074,
+      "learning_rate": 1.5691463797980325e-07,
+      "loss": 0.592,
+      "step": 7086
+    },
+    {
+      "epoch": 13.124074074074073,
+      "grad_norm": 1.714298186085463,
+      "learning_rate": 1.5690230853486686e-07,
+      "loss": 0.5892,
+      "step": 7087
+    },
+    {
+      "epoch": 13.125925925925927,
+      "grad_norm": 1.6004699226733163,
+      "learning_rate": 1.56889977810612e-07,
+      "loss": 0.6446,
+      "step": 7088
+    },
+    {
+      "epoch": 13.127777777777778,
+      "grad_norm": 1.6330351856888123,
+      "learning_rate": 1.5687764580731596e-07,
+      "loss": 0.6599,
+      "step": 7089
+    },
+    {
+      "epoch": 13.12962962962963,
+      "grad_norm": 1.650666070105923,
+      "learning_rate": 1.5686531252525597e-07,
+      "loss": 0.594,
+      "step": 7090
+    },
+    {
+      "epoch": 13.131481481481481,
+      "grad_norm": 1.8705760231052602,
+      "learning_rate": 1.5685297796470933e-07,
+      "loss": 0.6452,
+      "step": 7091
+    },
+    {
+      "epoch": 13.133333333333333,
+      "grad_norm": 1.7710449592802031,
+      "learning_rate": 1.568406421259533e-07,
+      "loss": 0.626,
+      "step": 7092
+    },
+    {
+      "epoch": 13.135185185185184,
+      "grad_norm": 1.6900422810429627,
+      "learning_rate": 1.5682830500926527e-07,
+      "loss": 0.6404,
+      "step": 7093
+    },
+    {
+      "epoch": 13.137037037037038,
+      "grad_norm": 2.1407381259962954,
+      "learning_rate": 1.568159666149226e-07,
+      "loss": 0.6206,
+      "step": 7094
+    },
+    {
+      "epoch": 13.13888888888889,
+      "grad_norm": 2.4730224236064835,
+      "learning_rate": 1.5680362694320266e-07,
+      "loss": 0.6267,
+      "step": 7095
+    },
+    {
+      "epoch": 13.14074074074074,
+      "grad_norm": 1.4709880177963799,
+      "learning_rate": 1.5679128599438295e-07,
+      "loss": 0.613,
+      "step": 7096
+    },
+    {
+      "epoch": 13.142592592592592,
+      "grad_norm": 1.4658075830738226,
+      "learning_rate": 1.5677894376874088e-07,
+      "loss": 0.5994,
+      "step": 7097
+    },
+    {
+      "epoch": 13.144444444444444,
+      "grad_norm": 1.5557579315422567,
+      "learning_rate": 1.5676660026655393e-07,
+      "loss": 0.5647,
+      "step": 7098
+    },
+    {
+      "epoch": 13.146296296296295,
+      "grad_norm": 1.4104751968510527,
+      "learning_rate": 1.5675425548809964e-07,
+      "loss": 0.5821,
+      "step": 7099
+    },
+    {
+      "epoch": 13.148148148148149,
+      "grad_norm": 2.4936061108386025,
+      "learning_rate": 1.5674190943365554e-07,
+      "loss": 0.6075,
+      "step": 7100
+    },
+    {
+      "epoch": 13.15,
+      "grad_norm": 1.5967834745204277,
+      "learning_rate": 1.5672956210349922e-07,
+      "loss": 0.6263,
+      "step": 7101
+    },
+    {
+      "epoch": 13.151851851851852,
+      "grad_norm": 1.4593158302998328,
+      "learning_rate": 1.5671721349790827e-07,
+      "loss": 0.6097,
+      "step": 7102
+    },
+    {
+      "epoch": 13.153703703703703,
+      "grad_norm": 1.9576601989944968,
+      "learning_rate": 1.567048636171603e-07,
+      "loss": 0.6016,
+      "step": 7103
+    },
+    {
+      "epoch": 13.155555555555555,
+      "grad_norm": 1.5141568945366077,
+      "learning_rate": 1.56692512461533e-07,
+      "loss": 0.5833,
+      "step": 7104
+    },
+    {
+      "epoch": 13.157407407407407,
+      "grad_norm": 1.8817099256151053,
+      "learning_rate": 1.5668016003130403e-07,
+      "loss": 0.5668,
+      "step": 7105
+    },
+    {
+      "epoch": 13.15925925925926,
+      "grad_norm": 1.5539965335364627,
+      "learning_rate": 1.5666780632675116e-07,
+      "loss": 0.6258,
+      "step": 7106
+    },
+    {
+      "epoch": 13.161111111111111,
+      "grad_norm": 1.6876452793846377,
+      "learning_rate": 1.5665545134815205e-07,
+      "loss": 0.6041,
+      "step": 7107
+    },
+    {
+      "epoch": 13.162962962962963,
+      "grad_norm": 1.94022983646296,
+      "learning_rate": 1.5664309509578456e-07,
+      "loss": 0.6058,
+      "step": 7108
+    },
+    {
+      "epoch": 13.164814814814815,
+      "grad_norm": 1.548248407237479,
+      "learning_rate": 1.5663073756992643e-07,
+      "loss": 0.6167,
+      "step": 7109
+    },
+    {
+      "epoch": 13.166666666666666,
+      "grad_norm": 1.6352882254801526,
+      "learning_rate": 1.566183787708555e-07,
+      "loss": 0.6307,
+      "step": 7110
+    },
+    {
+      "epoch": 13.168518518518518,
+      "grad_norm": 2.2855313650822566,
+      "learning_rate": 1.5660601869884967e-07,
+      "loss": 0.6218,
+      "step": 7111
+    },
+    {
+      "epoch": 13.170370370370371,
+      "grad_norm": 1.3847395082687353,
+      "learning_rate": 1.5659365735418678e-07,
+      "loss": 0.6219,
+      "step": 7112
+    },
+    {
+      "epoch": 13.172222222222222,
+      "grad_norm": 1.6927535566155099,
+      "learning_rate": 1.5658129473714477e-07,
+      "loss": 0.6325,
+      "step": 7113
+    },
+    {
+      "epoch": 13.174074074074074,
+      "grad_norm": 1.7956571728487167,
+      "learning_rate": 1.565689308480016e-07,
+      "loss": 0.5942,
+      "step": 7114
+    },
+    {
+      "epoch": 13.175925925925926,
+      "grad_norm": 1.7541327153587984,
+      "learning_rate": 1.5655656568703523e-07,
+      "loss": 0.6522,
+      "step": 7115
+    },
+    {
+      "epoch": 13.177777777777777,
+      "grad_norm": 1.4885519507175027,
+      "learning_rate": 1.565441992545236e-07,
+      "loss": 0.5946,
+      "step": 7116
+    },
+    {
+      "epoch": 13.17962962962963,
+      "grad_norm": 1.5915189221090702,
+      "learning_rate": 1.5653183155074486e-07,
+      "loss": 0.6077,
+      "step": 7117
+    },
+    {
+      "epoch": 13.181481481481482,
+      "grad_norm": 2.502650475644679,
+      "learning_rate": 1.5651946257597696e-07,
+      "loss": 0.6433,
+      "step": 7118
+    },
+    {
+      "epoch": 13.183333333333334,
+      "grad_norm": 1.376787567560288,
+      "learning_rate": 1.5650709233049807e-07,
+      "loss": 0.6266,
+      "step": 7119
+    },
+    {
+      "epoch": 13.185185185185185,
+      "grad_norm": 1.518433654511087,
+      "learning_rate": 1.5649472081458628e-07,
+      "loss": 0.632,
+      "step": 7120
+    },
+    {
+      "epoch": 13.187037037037037,
+      "grad_norm": 1.9058944420754775,
+      "learning_rate": 1.564823480285197e-07,
+      "loss": 0.5861,
+      "step": 7121
+    },
+    {
+      "epoch": 13.188888888888888,
+      "grad_norm": 1.5878063196337242,
+      "learning_rate": 1.5646997397257655e-07,
+      "loss": 0.6206,
+      "step": 7122
+    },
+    {
+      "epoch": 13.190740740740742,
+      "grad_norm": 1.403100950652496,
+      "learning_rate": 1.5645759864703503e-07,
+      "loss": 0.6282,
+      "step": 7123
+    },
+    {
+      "epoch": 13.192592592592593,
+      "grad_norm": 1.4179022264877932,
+      "learning_rate": 1.5644522205217335e-07,
+      "loss": 0.6348,
+      "step": 7124
+    },
+    {
+      "epoch": 13.194444444444445,
+      "grad_norm": 1.490665425752897,
+      "learning_rate": 1.5643284418826975e-07,
+      "loss": 0.6042,
+      "step": 7125
+    },
+    {
+      "epoch": 13.196296296296296,
+      "grad_norm": 1.5270277794993898,
+      "learning_rate": 1.5642046505560256e-07,
+      "loss": 0.5997,
+      "step": 7126
+    },
+    {
+      "epoch": 13.198148148148148,
+      "grad_norm": 1.494866735653359,
+      "learning_rate": 1.5640808465445007e-07,
+      "loss": 0.569,
+      "step": 7127
+    },
+    {
+      "epoch": 13.2,
+      "grad_norm": 1.7168862780206393,
+      "learning_rate": 1.5639570298509063e-07,
+      "loss": 0.6054,
+      "step": 7128
+    },
+    {
+      "epoch": 13.201851851851853,
+      "grad_norm": 2.8516847300923267,
+      "learning_rate": 1.5638332004780262e-07,
+      "loss": 0.6486,
+      "step": 7129
+    },
+    {
+      "epoch": 13.203703703703704,
+      "grad_norm": 1.5282436234024466,
+      "learning_rate": 1.5637093584286448e-07,
+      "loss": 0.5545,
+      "step": 7130
+    },
+    {
+      "epoch": 13.205555555555556,
+      "grad_norm": 1.4862224379288491,
+      "learning_rate": 1.5635855037055459e-07,
+      "loss": 0.588,
+      "step": 7131
+    },
+    {
+      "epoch": 13.207407407407407,
+      "grad_norm": 1.4233588105279789,
+      "learning_rate": 1.5634616363115138e-07,
+      "loss": 0.6393,
+      "step": 7132
+    },
+    {
+      "epoch": 13.209259259259259,
+      "grad_norm": 1.8378596363917523,
+      "learning_rate": 1.5633377562493342e-07,
+      "loss": 0.6219,
+      "step": 7133
+    },
+    {
+      "epoch": 13.21111111111111,
+      "grad_norm": 1.5898260549190242,
+      "learning_rate": 1.5632138635217916e-07,
+      "loss": 0.6304,
+      "step": 7134
+    },
+    {
+      "epoch": 13.212962962962964,
+      "grad_norm": 2.012889751994183,
+      "learning_rate": 1.5630899581316718e-07,
+      "loss": 0.6446,
+      "step": 7135
+    },
+    {
+      "epoch": 13.214814814814815,
+      "grad_norm": 3.2732604266456153,
+      "learning_rate": 1.5629660400817604e-07,
+      "loss": 0.5904,
+      "step": 7136
+    },
+    {
+      "epoch": 13.216666666666667,
+      "grad_norm": 1.5106587828841185,
+      "learning_rate": 1.562842109374843e-07,
+      "loss": 0.643,
+      "step": 7137
+    },
+    {
+      "epoch": 13.218518518518518,
+      "grad_norm": 2.0836126144573575,
+      "learning_rate": 1.562718166013707e-07,
+      "loss": 0.5976,
+      "step": 7138
+    },
+    {
+      "epoch": 13.22037037037037,
+      "grad_norm": 1.5558278498467972,
+      "learning_rate": 1.562594210001138e-07,
+      "loss": 0.631,
+      "step": 7139
+    },
+    {
+      "epoch": 13.222222222222221,
+      "grad_norm": 1.7223859995467241,
+      "learning_rate": 1.562470241339923e-07,
+      "loss": 0.6146,
+      "step": 7140
+    },
+    {
+      "epoch": 13.224074074074075,
+      "grad_norm": 1.5080277195184704,
+      "learning_rate": 1.5623462600328497e-07,
+      "loss": 0.5845,
+      "step": 7141
+    },
+    {
+      "epoch": 13.225925925925926,
+      "grad_norm": 1.5007770968310816,
+      "learning_rate": 1.5622222660827049e-07,
+      "loss": 0.6359,
+      "step": 7142
+    },
+    {
+      "epoch": 13.227777777777778,
+      "grad_norm": 1.4711528896239094,
+      "learning_rate": 1.5620982594922768e-07,
+      "loss": 0.6069,
+      "step": 7143
+    },
+    {
+      "epoch": 13.22962962962963,
+      "grad_norm": 1.4682487332492835,
+      "learning_rate": 1.561974240264353e-07,
+      "loss": 0.6015,
+      "step": 7144
+    },
+    {
+      "epoch": 13.231481481481481,
+      "grad_norm": 1.5438506406875174,
+      "learning_rate": 1.5618502084017218e-07,
+      "loss": 0.6155,
+      "step": 7145
+    },
+    {
+      "epoch": 13.233333333333333,
+      "grad_norm": 1.8011109053017722,
+      "learning_rate": 1.5617261639071723e-07,
+      "loss": 0.6285,
+      "step": 7146
+    },
+    {
+      "epoch": 13.235185185185186,
+      "grad_norm": 1.5024381589216425,
+      "learning_rate": 1.5616021067834928e-07,
+      "loss": 0.5624,
+      "step": 7147
+    },
+    {
+      "epoch": 13.237037037037037,
+      "grad_norm": 1.6527681664251175,
+      "learning_rate": 1.5614780370334728e-07,
+      "loss": 0.5806,
+      "step": 7148
+    },
+    {
+      "epoch": 13.238888888888889,
+      "grad_norm": 1.4644161808606149,
+      "learning_rate": 1.5613539546599014e-07,
+      "loss": 0.566,
+      "step": 7149
+    },
+    {
+      "epoch": 13.24074074074074,
+      "grad_norm": 1.9885807005798977,
+      "learning_rate": 1.561229859665569e-07,
+      "loss": 0.6165,
+      "step": 7150
+    },
+    {
+      "epoch": 13.242592592592592,
+      "grad_norm": 1.7490679004657346,
+      "learning_rate": 1.5611057520532643e-07,
+      "loss": 0.5784,
+      "step": 7151
+    },
+    {
+      "epoch": 13.244444444444444,
+      "grad_norm": 1.974566314147183,
+      "learning_rate": 1.5609816318257788e-07,
+      "loss": 0.6377,
+      "step": 7152
+    },
+    {
+      "epoch": 13.246296296296297,
+      "grad_norm": 1.844730433219054,
+      "learning_rate": 1.5608574989859026e-07,
+      "loss": 0.6257,
+      "step": 7153
+    },
+    {
+      "epoch": 13.248148148148148,
+      "grad_norm": 2.085746332903787,
+      "learning_rate": 1.5607333535364266e-07,
+      "loss": 0.6051,
+      "step": 7154
+    },
+    {
+      "epoch": 13.25,
+      "grad_norm": 1.6632466478469985,
+      "learning_rate": 1.5606091954801418e-07,
+      "loss": 0.6094,
+      "step": 7155
+    },
+    {
+      "epoch": 13.251851851851852,
+      "grad_norm": 1.4712284219704022,
+      "learning_rate": 1.5604850248198395e-07,
+      "loss": 0.6055,
+      "step": 7156
+    },
+    {
+      "epoch": 13.253703703703703,
+      "grad_norm": 1.496918293587102,
+      "learning_rate": 1.560360841558312e-07,
+      "loss": 0.6155,
+      "step": 7157
+    },
+    {
+      "epoch": 13.255555555555556,
+      "grad_norm": 1.548264825894146,
+      "learning_rate": 1.5602366456983506e-07,
+      "loss": 0.6222,
+      "step": 7158
+    },
+    {
+      "epoch": 13.257407407407408,
+      "grad_norm": 1.3363685501441196,
+      "learning_rate": 1.5601124372427478e-07,
+      "loss": 0.6055,
+      "step": 7159
+    },
+    {
+      "epoch": 13.25925925925926,
+      "grad_norm": 1.4565240443840588,
+      "learning_rate": 1.5599882161942964e-07,
+      "loss": 0.6502,
+      "step": 7160
+    },
+    {
+      "epoch": 13.261111111111111,
+      "grad_norm": 1.6176817858151131,
+      "learning_rate": 1.5598639825557887e-07,
+      "loss": 0.6142,
+      "step": 7161
+    },
+    {
+      "epoch": 13.262962962962963,
+      "grad_norm": 1.5527666367366177,
+      "learning_rate": 1.5597397363300183e-07,
+      "loss": 0.5862,
+      "step": 7162
+    },
+    {
+      "epoch": 13.264814814814814,
+      "grad_norm": 1.513823629471141,
+      "learning_rate": 1.5596154775197783e-07,
+      "loss": 0.6125,
+      "step": 7163
+    },
+    {
+      "epoch": 13.266666666666667,
+      "grad_norm": 2.3678619192155796,
+      "learning_rate": 1.5594912061278625e-07,
+      "loss": 0.6055,
+      "step": 7164
+    },
+    {
+      "epoch": 13.268518518518519,
+      "grad_norm": 1.6877413243806767,
+      "learning_rate": 1.5593669221570647e-07,
+      "loss": 0.6099,
+      "step": 7165
+    },
+    {
+      "epoch": 13.27037037037037,
+      "grad_norm": 1.5670393451826474,
+      "learning_rate": 1.5592426256101794e-07,
+      "loss": 0.5953,
+      "step": 7166
+    },
+    {
+      "epoch": 13.272222222222222,
+      "grad_norm": 1.637019713254645,
+      "learning_rate": 1.559118316490001e-07,
+      "loss": 0.6245,
+      "step": 7167
+    },
+    {
+      "epoch": 13.274074074074074,
+      "grad_norm": 1.4936109091676781,
+      "learning_rate": 1.5589939947993244e-07,
+      "loss": 0.6231,
+      "step": 7168
+    },
+    {
+      "epoch": 13.275925925925925,
+      "grad_norm": 1.516841577638341,
+      "learning_rate": 1.5588696605409444e-07,
+      "loss": 0.6186,
+      "step": 7169
+    },
+    {
+      "epoch": 13.277777777777779,
+      "grad_norm": 1.5601332465833169,
+      "learning_rate": 1.5587453137176568e-07,
+      "loss": 0.6262,
+      "step": 7170
+    },
+    {
+      "epoch": 13.27962962962963,
+      "grad_norm": 2.136025791024716,
+      "learning_rate": 1.558620954332257e-07,
+      "loss": 0.5759,
+      "step": 7171
+    },
+    {
+      "epoch": 13.281481481481482,
+      "grad_norm": 1.5916842080098519,
+      "learning_rate": 1.5584965823875406e-07,
+      "loss": 0.6032,
+      "step": 7172
+    },
+    {
+      "epoch": 13.283333333333333,
+      "grad_norm": 1.5822974119647308,
+      "learning_rate": 1.5583721978863044e-07,
+      "loss": 0.6148,
+      "step": 7173
+    },
+    {
+      "epoch": 13.285185185185185,
+      "grad_norm": 1.3879963852549086,
+      "learning_rate": 1.558247800831345e-07,
+      "loss": 0.6115,
+      "step": 7174
+    },
+    {
+      "epoch": 13.287037037037036,
+      "grad_norm": 1.7031413341957036,
+      "learning_rate": 1.5581233912254582e-07,
+      "loss": 0.6403,
+      "step": 7175
+    },
+    {
+      "epoch": 13.28888888888889,
+      "grad_norm": 1.4872108702002669,
+      "learning_rate": 1.557998969071442e-07,
+      "loss": 0.5937,
+      "step": 7176
+    },
+    {
+      "epoch": 13.290740740740741,
+      "grad_norm": 1.5827488685868027,
+      "learning_rate": 1.557874534372094e-07,
+      "loss": 0.6261,
+      "step": 7177
+    },
+    {
+      "epoch": 13.292592592592593,
+      "grad_norm": 1.4314220413011183,
+      "learning_rate": 1.5577500871302104e-07,
+      "loss": 0.5681,
+      "step": 7178
+    },
+    {
+      "epoch": 13.294444444444444,
+      "grad_norm": 1.5904591622898008,
+      "learning_rate": 1.5576256273485907e-07,
+      "loss": 0.5818,
+      "step": 7179
+    },
+    {
+      "epoch": 13.296296296296296,
+      "grad_norm": 2.2518748080039392,
+      "learning_rate": 1.557501155030032e-07,
+      "loss": 0.6015,
+      "step": 7180
+    },
+    {
+      "epoch": 13.298148148148147,
+      "grad_norm": 2.1293974731108634,
+      "learning_rate": 1.5573766701773335e-07,
+      "loss": 0.6023,
+      "step": 7181
+    },
+    {
+      "epoch": 13.3,
+      "grad_norm": 1.4576436345808617,
+      "learning_rate": 1.5572521727932936e-07,
+      "loss": 0.6093,
+      "step": 7182
+    },
+    {
+      "epoch": 13.301851851851852,
+      "grad_norm": 2.805747573234161,
+      "learning_rate": 1.5571276628807113e-07,
+      "loss": 0.615,
+      "step": 7183
+    },
+    {
+      "epoch": 13.303703703703704,
+      "grad_norm": 1.5033622270393323,
+      "learning_rate": 1.5570031404423857e-07,
+      "loss": 0.5731,
+      "step": 7184
+    },
+    {
+      "epoch": 13.305555555555555,
+      "grad_norm": 1.688349133988544,
+      "learning_rate": 1.5568786054811172e-07,
+      "loss": 0.6441,
+      "step": 7185
+    },
+    {
+      "epoch": 13.307407407407407,
+      "grad_norm": 1.476820335603725,
+      "learning_rate": 1.5567540579997048e-07,
+      "loss": 0.6229,
+      "step": 7186
+    },
+    {
+      "epoch": 13.309259259259258,
+      "grad_norm": 1.7225069094992602,
+      "learning_rate": 1.5566294980009494e-07,
+      "loss": 0.6151,
+      "step": 7187
+    },
+    {
+      "epoch": 13.311111111111112,
+      "grad_norm": 2.1389717721659394,
+      "learning_rate": 1.5565049254876507e-07,
+      "loss": 0.6297,
+      "step": 7188
+    },
+    {
+      "epoch": 13.312962962962963,
+      "grad_norm": 1.6711063435386788,
+      "learning_rate": 1.5563803404626105e-07,
+      "loss": 0.6661,
+      "step": 7189
+    },
+    {
+      "epoch": 13.314814814814815,
+      "grad_norm": 1.7705777056635927,
+      "learning_rate": 1.5562557429286284e-07,
+      "loss": 0.6033,
+      "step": 7190
+    },
+    {
+      "epoch": 13.316666666666666,
+      "grad_norm": 1.667030791659813,
+      "learning_rate": 1.556131132888507e-07,
+      "loss": 0.6139,
+      "step": 7191
+    },
+    {
+      "epoch": 13.318518518518518,
+      "grad_norm": 1.5127777248251448,
+      "learning_rate": 1.556006510345047e-07,
+      "loss": 0.6473,
+      "step": 7192
+    },
+    {
+      "epoch": 13.32037037037037,
+      "grad_norm": 1.6200711571689175,
+      "learning_rate": 1.5558818753010503e-07,
+      "loss": 0.6363,
+      "step": 7193
+    },
+    {
+      "epoch": 13.322222222222223,
+      "grad_norm": 1.8210418813816627,
+      "learning_rate": 1.5557572277593196e-07,
+      "loss": 0.5975,
+      "step": 7194
+    },
+    {
+      "epoch": 13.324074074074074,
+      "grad_norm": 1.6835784244307808,
+      "learning_rate": 1.5556325677226572e-07,
+      "loss": 0.6308,
+      "step": 7195
+    },
+    {
+      "epoch": 13.325925925925926,
+      "grad_norm": 1.6793993339238522,
+      "learning_rate": 1.555507895193865e-07,
+      "loss": 0.6399,
+      "step": 7196
+    },
+    {
+      "epoch": 13.327777777777778,
+      "grad_norm": 1.4709646184328942,
+      "learning_rate": 1.5553832101757474e-07,
+      "loss": 0.5732,
+      "step": 7197
+    },
+    {
+      "epoch": 13.329629629629629,
+      "grad_norm": 1.6320202024501513,
+      "learning_rate": 1.5552585126711065e-07,
+      "loss": 0.6282,
+      "step": 7198
+    },
+    {
+      "epoch": 13.331481481481482,
+      "grad_norm": 1.6373435069553492,
+      "learning_rate": 1.5551338026827458e-07,
+      "loss": 0.5896,
+      "step": 7199
+    },
+    {
+      "epoch": 13.333333333333334,
+      "grad_norm": 1.4650827802238808,
+      "learning_rate": 1.55500908021347e-07,
+      "loss": 0.6318,
+      "step": 7200
+    },
+    {
+      "epoch": 13.335185185185185,
+      "grad_norm": 1.5513335421585195,
+      "learning_rate": 1.5548843452660823e-07,
+      "loss": 0.6044,
+      "step": 7201
+    },
+    {
+      "epoch": 13.337037037037037,
+      "grad_norm": 1.5297668390045105,
+      "learning_rate": 1.554759597843388e-07,
+      "loss": 0.5951,
+      "step": 7202
+    },
+    {
+      "epoch": 13.338888888888889,
+      "grad_norm": 1.9096775338258725,
+      "learning_rate": 1.5546348379481906e-07,
+      "loss": 0.6609,
+      "step": 7203
+    },
+    {
+      "epoch": 13.34074074074074,
+      "grad_norm": 3.274762102893177,
+      "learning_rate": 1.554510065583296e-07,
+      "loss": 0.6278,
+      "step": 7204
+    },
+    {
+      "epoch": 13.342592592592593,
+      "grad_norm": 1.4427923949228452,
+      "learning_rate": 1.554385280751509e-07,
+      "loss": 0.626,
+      "step": 7205
+    },
+    {
+      "epoch": 13.344444444444445,
+      "grad_norm": 1.6533534661881462,
+      "learning_rate": 1.5542604834556352e-07,
+      "loss": 0.6243,
+      "step": 7206
+    },
+    {
+      "epoch": 13.346296296296297,
+      "grad_norm": 1.592967389770644,
+      "learning_rate": 1.5541356736984805e-07,
+      "loss": 0.6168,
+      "step": 7207
+    },
+    {
+      "epoch": 13.348148148148148,
+      "grad_norm": 1.5639555245975274,
+      "learning_rate": 1.5540108514828508e-07,
+      "loss": 0.6152,
+      "step": 7208
+    },
+    {
+      "epoch": 13.35,
+      "grad_norm": 1.853515960216099,
+      "learning_rate": 1.5538860168115525e-07,
+      "loss": 0.6179,
+      "step": 7209
+    },
+    {
+      "epoch": 13.351851851851851,
+      "grad_norm": 1.4444427657327827,
+      "learning_rate": 1.5537611696873922e-07,
+      "loss": 0.6189,
+      "step": 7210
+    },
+    {
+      "epoch": 13.353703703703705,
+      "grad_norm": 1.5211776580755323,
+      "learning_rate": 1.5536363101131768e-07,
+      "loss": 0.6341,
+      "step": 7211
+    },
+    {
+      "epoch": 13.355555555555556,
+      "grad_norm": 2.048897242291352,
+      "learning_rate": 1.5535114380917136e-07,
+      "loss": 0.5821,
+      "step": 7212
+    },
+    {
+      "epoch": 13.357407407407408,
+      "grad_norm": 1.8078325145819816,
+      "learning_rate": 1.5533865536258098e-07,
+      "loss": 0.5753,
+      "step": 7213
+    },
+    {
+      "epoch": 13.35925925925926,
+      "grad_norm": 4.399871436313289,
+      "learning_rate": 1.5532616567182735e-07,
+      "loss": 0.6235,
+      "step": 7214
+    },
+    {
+      "epoch": 13.36111111111111,
+      "grad_norm": 1.4027593518509,
+      "learning_rate": 1.5531367473719122e-07,
+      "loss": 0.6343,
+      "step": 7215
+    },
+    {
+      "epoch": 13.362962962962962,
+      "grad_norm": 1.5334102322901992,
+      "learning_rate": 1.553011825589535e-07,
+      "loss": 0.5927,
+      "step": 7216
+    },
+    {
+      "epoch": 13.364814814814816,
+      "grad_norm": 2.885473030185685,
+      "learning_rate": 1.5528868913739496e-07,
+      "loss": 0.6401,
+      "step": 7217
+    },
+    {
+      "epoch": 13.366666666666667,
+      "grad_norm": 1.6874424494807094,
+      "learning_rate": 1.5527619447279654e-07,
+      "loss": 0.6482,
+      "step": 7218
+    },
+    {
+      "epoch": 13.368518518518519,
+      "grad_norm": 1.676149172295434,
+      "learning_rate": 1.5526369856543912e-07,
+      "loss": 0.6299,
+      "step": 7219
+    },
+    {
+      "epoch": 13.37037037037037,
+      "grad_norm": 1.5605094292624055,
+      "learning_rate": 1.5525120141560368e-07,
+      "loss": 0.5801,
+      "step": 7220
+    },
+    {
+      "epoch": 13.372222222222222,
+      "grad_norm": 1.7229455207529392,
+      "learning_rate": 1.552387030235712e-07,
+      "loss": 0.613,
+      "step": 7221
+    },
+    {
+      "epoch": 13.374074074074073,
+      "grad_norm": 2.5048857770039175,
+      "learning_rate": 1.552262033896226e-07,
+      "loss": 0.6439,
+      "step": 7222
+    },
+    {
+      "epoch": 13.375925925925927,
+      "grad_norm": 1.5525856630085684,
+      "learning_rate": 1.5521370251403897e-07,
+      "loss": 0.6456,
+      "step": 7223
+    },
+    {
+      "epoch": 13.377777777777778,
+      "grad_norm": 1.450106358722777,
+      "learning_rate": 1.5520120039710135e-07,
+      "loss": 0.6366,
+      "step": 7224
+    },
+    {
+      "epoch": 13.37962962962963,
+      "grad_norm": 1.5866116806712873,
+      "learning_rate": 1.5518869703909085e-07,
+      "loss": 0.6531,
+      "step": 7225
+    },
+    {
+      "epoch": 13.381481481481481,
+      "grad_norm": 1.39321956578684,
+      "learning_rate": 1.5517619244028851e-07,
+      "loss": 0.6266,
+      "step": 7226
+    },
+    {
+      "epoch": 13.383333333333333,
+      "grad_norm": 1.7032694536662418,
+      "learning_rate": 1.5516368660097554e-07,
+      "loss": 0.6059,
+      "step": 7227
+    },
+    {
+      "epoch": 13.385185185185184,
+      "grad_norm": 1.719515595469275,
+      "learning_rate": 1.5515117952143306e-07,
+      "loss": 0.6041,
+      "step": 7228
+    },
+    {
+      "epoch": 13.387037037037038,
+      "grad_norm": 1.4816846150279102,
+      "learning_rate": 1.5513867120194228e-07,
+      "loss": 0.603,
+      "step": 7229
+    },
+    {
+      "epoch": 13.38888888888889,
+      "grad_norm": 1.6810960457715094,
+      "learning_rate": 1.551261616427844e-07,
+      "loss": 0.5998,
+      "step": 7230
+    },
+    {
+      "epoch": 13.39074074074074,
+      "grad_norm": 1.460274833542817,
+      "learning_rate": 1.551136508442407e-07,
+      "loss": 0.5875,
+      "step": 7231
+    },
+    {
+      "epoch": 13.392592592592592,
+      "grad_norm": 1.5976815313291926,
+      "learning_rate": 1.5510113880659245e-07,
+      "loss": 0.6369,
+      "step": 7232
+    },
+    {
+      "epoch": 13.394444444444444,
+      "grad_norm": 1.6637518168805452,
+      "learning_rate": 1.5508862553012095e-07,
+      "loss": 0.6443,
+      "step": 7233
+    },
+    {
+      "epoch": 13.396296296296295,
+      "grad_norm": 1.584093813551479,
+      "learning_rate": 1.550761110151075e-07,
+      "loss": 0.5938,
+      "step": 7234
+    },
+    {
+      "epoch": 13.398148148148149,
+      "grad_norm": 1.3902828841954127,
+      "learning_rate": 1.550635952618335e-07,
+      "loss": 0.5905,
+      "step": 7235
+    },
+    {
+      "epoch": 13.4,
+      "grad_norm": 1.6112465683726291,
+      "learning_rate": 1.5505107827058036e-07,
+      "loss": 0.611,
+      "step": 7236
+    },
+    {
+      "epoch": 13.401851851851852,
+      "grad_norm": 1.5427502555257506,
+      "learning_rate": 1.5503856004162943e-07,
+      "loss": 0.6195,
+      "step": 7237
+    },
+    {
+      "epoch": 13.403703703703703,
+      "grad_norm": 2.1613888322822903,
+      "learning_rate": 1.5502604057526218e-07,
+      "loss": 0.5967,
+      "step": 7238
+    },
+    {
+      "epoch": 13.405555555555555,
+      "grad_norm": 1.4482183321569075,
+      "learning_rate": 1.5501351987176013e-07,
+      "loss": 0.6236,
+      "step": 7239
+    },
+    {
+      "epoch": 13.407407407407407,
+      "grad_norm": 1.7071839678327017,
+      "learning_rate": 1.5500099793140472e-07,
+      "loss": 0.6317,
+      "step": 7240
+    },
+    {
+      "epoch": 13.40925925925926,
+      "grad_norm": 1.4328849031745656,
+      "learning_rate": 1.5498847475447748e-07,
+      "loss": 0.556,
+      "step": 7241
+    },
+    {
+      "epoch": 13.411111111111111,
+      "grad_norm": 2.0680950725964804,
+      "learning_rate": 1.5497595034126e-07,
+      "loss": 0.6157,
+      "step": 7242
+    },
+    {
+      "epoch": 13.412962962962963,
+      "grad_norm": 1.4340663397466076,
+      "learning_rate": 1.5496342469203382e-07,
+      "loss": 0.5666,
+      "step": 7243
+    },
+    {
+      "epoch": 13.414814814814815,
+      "grad_norm": 2.2323254468394285,
+      "learning_rate": 1.549508978070806e-07,
+      "loss": 0.5856,
+      "step": 7244
+    },
+    {
+      "epoch": 13.416666666666666,
+      "grad_norm": 1.5738682735443401,
+      "learning_rate": 1.5493836968668196e-07,
+      "loss": 0.6544,
+      "step": 7245
+    },
+    {
+      "epoch": 13.418518518518518,
+      "grad_norm": 3.4473329389301184,
+      "learning_rate": 1.5492584033111952e-07,
+      "loss": 0.5675,
+      "step": 7246
+    },
+    {
+      "epoch": 13.420370370370371,
+      "grad_norm": 1.7431074658088808,
+      "learning_rate": 1.5491330974067506e-07,
+      "loss": 0.6204,
+      "step": 7247
+    },
+    {
+      "epoch": 13.422222222222222,
+      "grad_norm": 1.7102770591626848,
+      "learning_rate": 1.549007779156302e-07,
+      "loss": 0.6341,
+      "step": 7248
+    },
+    {
+      "epoch": 13.424074074074074,
+      "grad_norm": 1.4716501837398548,
+      "learning_rate": 1.548882448562668e-07,
+      "loss": 0.5965,
+      "step": 7249
+    },
+    {
+      "epoch": 13.425925925925926,
+      "grad_norm": 1.6122510352119186,
+      "learning_rate": 1.5487571056286655e-07,
+      "loss": 0.6262,
+      "step": 7250
+    },
+    {
+      "epoch": 13.427777777777777,
+      "grad_norm": 1.5630625688884,
+      "learning_rate": 1.5486317503571128e-07,
+      "loss": 0.6029,
+      "step": 7251
+    },
+    {
+      "epoch": 13.42962962962963,
+      "grad_norm": 1.7753991485483485,
+      "learning_rate": 1.5485063827508284e-07,
+      "loss": 0.6419,
+      "step": 7252
+    },
+    {
+      "epoch": 13.431481481481482,
+      "grad_norm": 1.5111736578440107,
+      "learning_rate": 1.548381002812631e-07,
+      "loss": 0.6219,
+      "step": 7253
+    },
+    {
+      "epoch": 13.433333333333334,
+      "grad_norm": 1.4277885822816156,
+      "learning_rate": 1.548255610545339e-07,
+      "loss": 0.5958,
+      "step": 7254
+    },
+    {
+      "epoch": 13.435185185185185,
+      "grad_norm": 1.659906211316063,
+      "learning_rate": 1.548130205951772e-07,
+      "loss": 0.6223,
+      "step": 7255
+    },
+    {
+      "epoch": 13.437037037037037,
+      "grad_norm": 1.4701232226134355,
+      "learning_rate": 1.548004789034749e-07,
+      "loss": 0.6083,
+      "step": 7256
+    },
+    {
+      "epoch": 13.438888888888888,
+      "grad_norm": 2.099838260800555,
+      "learning_rate": 1.5478793597970904e-07,
+      "loss": 0.6277,
+      "step": 7257
+    },
+    {
+      "epoch": 13.440740740740742,
+      "grad_norm": 1.397297940851872,
+      "learning_rate": 1.5477539182416155e-07,
+      "loss": 0.6081,
+      "step": 7258
+    },
+    {
+      "epoch": 13.442592592592593,
+      "grad_norm": 1.6824252374009836,
+      "learning_rate": 1.547628464371145e-07,
+      "loss": 0.6071,
+      "step": 7259
+    },
+    {
+      "epoch": 13.444444444444445,
+      "grad_norm": 1.6801908021288223,
+      "learning_rate": 1.5475029981884992e-07,
+      "loss": 0.6226,
+      "step": 7260
+    },
+    {
+      "epoch": 13.446296296296296,
+      "grad_norm": 1.693111036716262,
+      "learning_rate": 1.5473775196964988e-07,
+      "loss": 0.649,
+      "step": 7261
+    },
+    {
+      "epoch": 13.448148148148148,
+      "grad_norm": 1.5285946865132818,
+      "learning_rate": 1.5472520288979654e-07,
+      "loss": 0.6693,
+      "step": 7262
+    },
+    {
+      "epoch": 13.45,
+      "grad_norm": 1.6776202118672494,
+      "learning_rate": 1.54712652579572e-07,
+      "loss": 0.5978,
+      "step": 7263
+    },
+    {
+      "epoch": 13.451851851851853,
+      "grad_norm": 1.4903361181878623,
+      "learning_rate": 1.5470010103925841e-07,
+      "loss": 0.554,
+      "step": 7264
+    },
+    {
+      "epoch": 13.453703703703704,
+      "grad_norm": 1.6270620847033141,
+      "learning_rate": 1.54687548269138e-07,
+      "loss": 0.6114,
+      "step": 7265
+    },
+    {
+      "epoch": 13.455555555555556,
+      "grad_norm": 1.450077681166889,
+      "learning_rate": 1.5467499426949301e-07,
+      "loss": 0.6432,
+      "step": 7266
+    },
+    {
+      "epoch": 13.457407407407407,
+      "grad_norm": 2.0873861628103962,
+      "learning_rate": 1.546624390406056e-07,
+      "loss": 0.6107,
+      "step": 7267
+    },
+    {
+      "epoch": 13.459259259259259,
+      "grad_norm": 2.0368006397081495,
+      "learning_rate": 1.5464988258275816e-07,
+      "loss": 0.6325,
+      "step": 7268
+    },
+    {
+      "epoch": 13.46111111111111,
+      "grad_norm": 1.5276026367458988,
+      "learning_rate": 1.546373248962329e-07,
+      "loss": 0.5599,
+      "step": 7269
+    },
+    {
+      "epoch": 13.462962962962964,
+      "grad_norm": 1.9163286826330306,
+      "learning_rate": 1.5462476598131215e-07,
+      "loss": 0.6354,
+      "step": 7270
+    },
+    {
+      "epoch": 13.464814814814815,
+      "grad_norm": 2.8722769750608483,
+      "learning_rate": 1.5461220583827834e-07,
+      "loss": 0.6556,
+      "step": 7271
+    },
+    {
+      "epoch": 13.466666666666667,
+      "grad_norm": 1.6272131929280682,
+      "learning_rate": 1.545996444674138e-07,
+      "loss": 0.6361,
+      "step": 7272
+    },
+    {
+      "epoch": 13.468518518518518,
+      "grad_norm": 3.1862731656053076,
+      "learning_rate": 1.5458708186900097e-07,
+      "loss": 0.6238,
+      "step": 7273
+    },
+    {
+      "epoch": 13.47037037037037,
+      "grad_norm": 1.5230550313486333,
+      "learning_rate": 1.545745180433223e-07,
+      "loss": 0.6548,
+      "step": 7274
+    },
+    {
+      "epoch": 13.472222222222221,
+      "grad_norm": 1.5796975329947691,
+      "learning_rate": 1.5456195299066024e-07,
+      "loss": 0.6761,
+      "step": 7275
+    },
+    {
+      "epoch": 13.474074074074075,
+      "grad_norm": 2.2177988597927794,
+      "learning_rate": 1.5454938671129725e-07,
+      "loss": 0.5899,
+      "step": 7276
+    },
+    {
+      "epoch": 13.475925925925926,
+      "grad_norm": 2.9647813476820613,
+      "learning_rate": 1.5453681920551593e-07,
+      "loss": 0.5911,
+      "step": 7277
+    },
+    {
+      "epoch": 13.477777777777778,
+      "grad_norm": 3.747270398907113,
+      "learning_rate": 1.545242504735988e-07,
+      "loss": 0.5929,
+      "step": 7278
+    },
+    {
+      "epoch": 13.47962962962963,
+      "grad_norm": 1.701062681988641,
+      "learning_rate": 1.545116805158284e-07,
+      "loss": 0.6342,
+      "step": 7279
+    },
+    {
+      "epoch": 13.481481481481481,
+      "grad_norm": 1.5296037841240349,
+      "learning_rate": 1.544991093324874e-07,
+      "loss": 0.5982,
+      "step": 7280
+    },
+    {
+      "epoch": 13.483333333333333,
+      "grad_norm": 1.5355962732251203,
+      "learning_rate": 1.544865369238584e-07,
+      "loss": 0.5723,
+      "step": 7281
+    },
+    {
+      "epoch": 13.485185185185186,
+      "grad_norm": 1.4924029441769386,
+      "learning_rate": 1.5447396329022404e-07,
+      "loss": 0.5913,
+      "step": 7282
+    },
+    {
+      "epoch": 13.487037037037037,
+      "grad_norm": 1.7686834754944267,
+      "learning_rate": 1.5446138843186708e-07,
+      "loss": 0.6184,
+      "step": 7283
+    },
+    {
+      "epoch": 13.488888888888889,
+      "grad_norm": 1.609445053178199,
+      "learning_rate": 1.5444881234907015e-07,
+      "loss": 0.5756,
+      "step": 7284
+    },
+    {
+      "epoch": 13.49074074074074,
+      "grad_norm": 1.511023451679778,
+      "learning_rate": 1.5443623504211604e-07,
+      "loss": 0.6127,
+      "step": 7285
+    },
+    {
+      "epoch": 13.492592592592592,
+      "grad_norm": 1.5958671644925277,
+      "learning_rate": 1.5442365651128757e-07,
+      "loss": 0.6159,
+      "step": 7286
+    },
+    {
+      "epoch": 13.494444444444444,
+      "grad_norm": 1.6427870429851574,
+      "learning_rate": 1.5441107675686746e-07,
+      "loss": 0.5427,
+      "step": 7287
+    },
+    {
+      "epoch": 13.496296296296297,
+      "grad_norm": 1.48078786290128,
+      "learning_rate": 1.5439849577913854e-07,
+      "loss": 0.6036,
+      "step": 7288
+    },
+    {
+      "epoch": 13.498148148148148,
+      "grad_norm": 1.536288241700882,
+      "learning_rate": 1.5438591357838373e-07,
+      "loss": 0.5611,
+      "step": 7289
+    },
+    {
+      "epoch": 13.5,
+      "grad_norm": 3.2795814027610666,
+      "learning_rate": 1.5437333015488586e-07,
+      "loss": 0.6404,
+      "step": 7290
+    },
+    {
+      "epoch": 13.501851851851852,
+      "grad_norm": 2.4877275128378393,
+      "learning_rate": 1.5436074550892782e-07,
+      "loss": 0.5842,
+      "step": 7291
+    },
+    {
+      "epoch": 13.503703703703703,
+      "grad_norm": 1.8444215746675836,
+      "learning_rate": 1.5434815964079265e-07,
+      "loss": 0.6054,
+      "step": 7292
+    },
+    {
+      "epoch": 13.505555555555556,
+      "grad_norm": 1.7101993425511546,
+      "learning_rate": 1.5433557255076316e-07,
+      "loss": 0.5957,
+      "step": 7293
+    },
+    {
+      "epoch": 13.507407407407408,
+      "grad_norm": 1.9236807851678925,
+      "learning_rate": 1.5432298423912248e-07,
+      "loss": 0.5625,
+      "step": 7294
+    },
+    {
+      "epoch": 13.50925925925926,
+      "grad_norm": 1.6001900064404895,
+      "learning_rate": 1.5431039470615356e-07,
+      "loss": 0.6268,
+      "step": 7295
+    },
+    {
+      "epoch": 13.511111111111111,
+      "grad_norm": 1.5960329218471694,
+      "learning_rate": 1.5429780395213946e-07,
+      "loss": 0.6261,
+      "step": 7296
+    },
+    {
+      "epoch": 13.512962962962963,
+      "grad_norm": 1.7449133974113504,
+      "learning_rate": 1.5428521197736327e-07,
+      "loss": 0.6013,
+      "step": 7297
+    },
+    {
+      "epoch": 13.514814814814814,
+      "grad_norm": 1.8771426929945274,
+      "learning_rate": 1.542726187821081e-07,
+      "loss": 0.6271,
+      "step": 7298
+    },
+    {
+      "epoch": 13.516666666666667,
+      "grad_norm": 1.663045731213215,
+      "learning_rate": 1.5426002436665702e-07,
+      "loss": 0.6337,
+      "step": 7299
+    },
+    {
+      "epoch": 13.518518518518519,
+      "grad_norm": 1.658408113420754,
+      "learning_rate": 1.542474287312932e-07,
+      "loss": 0.596,
+      "step": 7300
+    },
+    {
+      "epoch": 13.52037037037037,
+      "grad_norm": 1.5682888823924663,
+      "learning_rate": 1.5423483187629992e-07,
+      "loss": 0.6405,
+      "step": 7301
+    },
+    {
+      "epoch": 13.522222222222222,
+      "grad_norm": 1.765702104168939,
+      "learning_rate": 1.5422223380196027e-07,
+      "loss": 0.6186,
+      "step": 7302
+    },
+    {
+      "epoch": 13.524074074074074,
+      "grad_norm": 1.8488512607310224,
+      "learning_rate": 1.5420963450855752e-07,
+      "loss": 0.62,
+      "step": 7303
+    },
+    {
+      "epoch": 13.525925925925925,
+      "grad_norm": 1.8507928254643464,
+      "learning_rate": 1.54197033996375e-07,
+      "loss": 0.6267,
+      "step": 7304
+    },
+    {
+      "epoch": 13.527777777777779,
+      "grad_norm": 1.5225311310517957,
+      "learning_rate": 1.5418443226569595e-07,
+      "loss": 0.6084,
+      "step": 7305
+    },
+    {
+      "epoch": 13.52962962962963,
+      "grad_norm": 1.5031336554184112,
+      "learning_rate": 1.541718293168037e-07,
+      "loss": 0.5904,
+      "step": 7306
+    },
+    {
+      "epoch": 13.531481481481482,
+      "grad_norm": 1.678870570414537,
+      "learning_rate": 1.5415922514998158e-07,
+      "loss": 0.5988,
+      "step": 7307
+    },
+    {
+      "epoch": 13.533333333333333,
+      "grad_norm": 1.8478465513422078,
+      "learning_rate": 1.54146619765513e-07,
+      "loss": 0.6099,
+      "step": 7308
+    },
+    {
+      "epoch": 13.535185185185185,
+      "grad_norm": 1.789387149025911,
+      "learning_rate": 1.5413401316368132e-07,
+      "loss": 0.6125,
+      "step": 7309
+    },
+    {
+      "epoch": 13.537037037037036,
+      "grad_norm": 1.6405450700681674,
+      "learning_rate": 1.5412140534477004e-07,
+      "loss": 0.6049,
+      "step": 7310
+    },
+    {
+      "epoch": 13.53888888888889,
+      "grad_norm": 1.5386246610526035,
+      "learning_rate": 1.5410879630906257e-07,
+      "loss": 0.6097,
+      "step": 7311
+    },
+    {
+      "epoch": 13.540740740740741,
+      "grad_norm": 1.9177223826017549,
+      "learning_rate": 1.540961860568424e-07,
+      "loss": 0.6289,
+      "step": 7312
+    },
+    {
+      "epoch": 13.542592592592593,
+      "grad_norm": 1.526017841772827,
+      "learning_rate": 1.54083574588393e-07,
+      "loss": 0.6126,
+      "step": 7313
+    },
+    {
+      "epoch": 13.544444444444444,
+      "grad_norm": 1.4722972119743998,
+      "learning_rate": 1.5407096190399803e-07,
+      "loss": 0.6409,
+      "step": 7314
+    },
+    {
+      "epoch": 13.546296296296296,
+      "grad_norm": 1.6244677154763651,
+      "learning_rate": 1.540583480039409e-07,
+      "loss": 0.5764,
+      "step": 7315
+    },
+    {
+      "epoch": 13.548148148148147,
+      "grad_norm": 1.4614190575251744,
+      "learning_rate": 1.5404573288850533e-07,
+      "loss": 0.5818,
+      "step": 7316
+    },
+    {
+      "epoch": 13.55,
+      "grad_norm": 1.4984861578826052,
+      "learning_rate": 1.5403311655797492e-07,
+      "loss": 0.5604,
+      "step": 7317
+    },
+    {
+      "epoch": 13.551851851851852,
+      "grad_norm": 1.610429523590397,
+      "learning_rate": 1.5402049901263324e-07,
+      "loss": 0.5921,
+      "step": 7318
+    },
+    {
+      "epoch": 13.553703703703704,
+      "grad_norm": 1.8484256165244102,
+      "learning_rate": 1.5400788025276408e-07,
+      "loss": 0.6097,
+      "step": 7319
+    },
+    {
+      "epoch": 13.555555555555555,
+      "grad_norm": 1.5898008302380773,
+      "learning_rate": 1.5399526027865106e-07,
+      "loss": 0.6159,
+      "step": 7320
+    },
+    {
+      "epoch": 13.557407407407407,
+      "grad_norm": 1.4882874953799088,
+      "learning_rate": 1.5398263909057797e-07,
+      "loss": 0.6075,
+      "step": 7321
+    },
+    {
+      "epoch": 13.559259259259258,
+      "grad_norm": 1.5062444618629536,
+      "learning_rate": 1.5397001668882853e-07,
+      "loss": 0.6304,
+      "step": 7322
+    },
+    {
+      "epoch": 13.561111111111112,
+      "grad_norm": 1.608475389720989,
+      "learning_rate": 1.5395739307368652e-07,
+      "loss": 0.6446,
+      "step": 7323
+    },
+    {
+      "epoch": 13.562962962962963,
+      "grad_norm": 1.4871846207653443,
+      "learning_rate": 1.5394476824543574e-07,
+      "loss": 0.6002,
+      "step": 7324
+    },
+    {
+      "epoch": 13.564814814814815,
+      "grad_norm": 1.4487657844756618,
+      "learning_rate": 1.5393214220436008e-07,
+      "loss": 0.5768,
+      "step": 7325
+    },
+    {
+      "epoch": 13.566666666666666,
+      "grad_norm": 1.7145655159524333,
+      "learning_rate": 1.539195149507434e-07,
+      "loss": 0.6152,
+      "step": 7326
+    },
+    {
+      "epoch": 13.568518518518518,
+      "grad_norm": 1.4793640287813234,
+      "learning_rate": 1.5390688648486957e-07,
+      "loss": 0.5893,
+      "step": 7327
+    },
+    {
+      "epoch": 13.57037037037037,
+      "grad_norm": 1.4231671694241494,
+      "learning_rate": 1.5389425680702254e-07,
+      "loss": 0.5981,
+      "step": 7328
+    },
+    {
+      "epoch": 13.572222222222223,
+      "grad_norm": 1.644217236502398,
+      "learning_rate": 1.5388162591748623e-07,
+      "loss": 0.5845,
+      "step": 7329
+    },
+    {
+      "epoch": 13.574074074074074,
+      "grad_norm": 1.8000892300667184,
+      "learning_rate": 1.5386899381654462e-07,
+      "loss": 0.5989,
+      "step": 7330
+    },
+    {
+      "epoch": 13.575925925925926,
+      "grad_norm": 1.5449991231500702,
+      "learning_rate": 1.538563605044817e-07,
+      "loss": 0.5856,
+      "step": 7331
+    },
+    {
+      "epoch": 13.577777777777778,
+      "grad_norm": 1.5575671892729914,
+      "learning_rate": 1.5384372598158158e-07,
+      "loss": 0.5955,
+      "step": 7332
+    },
+    {
+      "epoch": 13.579629629629629,
+      "grad_norm": 1.569823369996743,
+      "learning_rate": 1.5383109024812822e-07,
+      "loss": 0.5563,
+      "step": 7333
+    },
+    {
+      "epoch": 13.581481481481482,
+      "grad_norm": 1.521514141677872,
+      "learning_rate": 1.5381845330440578e-07,
+      "loss": 0.5996,
+      "step": 7334
+    },
+    {
+      "epoch": 13.583333333333334,
+      "grad_norm": 1.6437900401362477,
+      "learning_rate": 1.5380581515069832e-07,
+      "loss": 0.6201,
+      "step": 7335
+    },
+    {
+      "epoch": 13.585185185185185,
+      "grad_norm": 2.001926645392715,
+      "learning_rate": 1.5379317578729003e-07,
+      "loss": 0.628,
+      "step": 7336
+    },
+    {
+      "epoch": 13.587037037037037,
+      "grad_norm": 1.8155289947708695,
+      "learning_rate": 1.5378053521446504e-07,
+      "loss": 0.6114,
+      "step": 7337
+    },
+    {
+      "epoch": 13.588888888888889,
+      "grad_norm": 2.8460597973518165,
+      "learning_rate": 1.537678934325075e-07,
+      "loss": 0.62,
+      "step": 7338
+    },
+    {
+      "epoch": 13.59074074074074,
+      "grad_norm": 1.4605211043987703,
+      "learning_rate": 1.537552504417017e-07,
+      "loss": 0.6138,
+      "step": 7339
+    },
+    {
+      "epoch": 13.592592592592592,
+      "grad_norm": 1.613214358650215,
+      "learning_rate": 1.5374260624233192e-07,
+      "loss": 0.6319,
+      "step": 7340
+    },
+    {
+      "epoch": 13.594444444444445,
+      "grad_norm": 1.765575642499274,
+      "learning_rate": 1.537299608346824e-07,
+      "loss": 0.5596,
+      "step": 7341
+    },
+    {
+      "epoch": 13.596296296296297,
+      "grad_norm": 1.4628720372800443,
+      "learning_rate": 1.5371731421903737e-07,
+      "loss": 0.652,
+      "step": 7342
+    },
+    {
+      "epoch": 13.598148148148148,
+      "grad_norm": 1.6429603751209803,
+      "learning_rate": 1.5370466639568127e-07,
+      "loss": 0.609,
+      "step": 7343
+    },
+    {
+      "epoch": 13.6,
+      "grad_norm": 1.9259401560372884,
+      "learning_rate": 1.5369201736489839e-07,
+      "loss": 0.6622,
+      "step": 7344
+    },
+    {
+      "epoch": 13.601851851851851,
+      "grad_norm": 1.5081132064301976,
+      "learning_rate": 1.5367936712697311e-07,
+      "loss": 0.5951,
+      "step": 7345
+    },
+    {
+      "epoch": 13.603703703703705,
+      "grad_norm": 1.9195505522029188,
+      "learning_rate": 1.536667156821899e-07,
+      "loss": 0.6105,
+      "step": 7346
+    },
+    {
+      "epoch": 13.605555555555556,
+      "grad_norm": 1.7798404802093304,
+      "learning_rate": 1.5365406303083316e-07,
+      "loss": 0.6236,
+      "step": 7347
+    },
+    {
+      "epoch": 13.607407407407408,
+      "grad_norm": 1.5916803097408845,
+      "learning_rate": 1.5364140917318734e-07,
+      "loss": 0.6189,
+      "step": 7348
+    },
+    {
+      "epoch": 13.60925925925926,
+      "grad_norm": 1.9182538034717207,
+      "learning_rate": 1.5362875410953696e-07,
+      "loss": 0.6027,
+      "step": 7349
+    },
+    {
+      "epoch": 13.61111111111111,
+      "grad_norm": 1.6311275093183528,
+      "learning_rate": 1.5361609784016655e-07,
+      "loss": 0.6236,
+      "step": 7350
+    },
+    {
+      "epoch": 13.612962962962962,
+      "grad_norm": 2.1395493644448034,
+      "learning_rate": 1.5360344036536063e-07,
+      "loss": 0.6166,
+      "step": 7351
+    },
+    {
+      "epoch": 13.614814814814816,
+      "grad_norm": 1.748014452546716,
+      "learning_rate": 1.535907816854038e-07,
+      "loss": 0.6055,
+      "step": 7352
+    },
+    {
+      "epoch": 13.616666666666667,
+      "grad_norm": 1.5338843474064465,
+      "learning_rate": 1.5357812180058064e-07,
+      "loss": 0.6107,
+      "step": 7353
+    },
+    {
+      "epoch": 13.618518518518519,
+      "grad_norm": 1.8157125463606256,
+      "learning_rate": 1.5356546071117578e-07,
+      "loss": 0.6071,
+      "step": 7354
+    },
+    {
+      "epoch": 13.62037037037037,
+      "grad_norm": 1.8375164802228954,
+      "learning_rate": 1.535527984174739e-07,
+      "loss": 0.608,
+      "step": 7355
+    },
+    {
+      "epoch": 13.622222222222222,
+      "grad_norm": 6.514712478612919,
+      "learning_rate": 1.5354013491975964e-07,
+      "loss": 0.577,
+      "step": 7356
+    },
+    {
+      "epoch": 13.624074074074073,
+      "grad_norm": 1.8417580603042927,
+      "learning_rate": 1.5352747021831775e-07,
+      "loss": 0.6347,
+      "step": 7357
+    },
+    {
+      "epoch": 13.625925925925927,
+      "grad_norm": 1.544444473697421,
+      "learning_rate": 1.5351480431343296e-07,
+      "loss": 0.6283,
+      "step": 7358
+    },
+    {
+      "epoch": 13.627777777777778,
+      "grad_norm": 1.5083167596088436,
+      "learning_rate": 1.5350213720539e-07,
+      "loss": 0.6048,
+      "step": 7359
+    },
+    {
+      "epoch": 13.62962962962963,
+      "grad_norm": 2.098424536431358,
+      "learning_rate": 1.5348946889447368e-07,
+      "loss": 0.5886,
+      "step": 7360
+    },
+    {
+      "epoch": 13.631481481481481,
+      "grad_norm": 1.31449862759255,
+      "learning_rate": 1.5347679938096887e-07,
+      "loss": 0.5811,
+      "step": 7361
+    },
+    {
+      "epoch": 13.633333333333333,
+      "grad_norm": 1.5898861162618418,
+      "learning_rate": 1.534641286651603e-07,
+      "loss": 0.5947,
+      "step": 7362
+    },
+    {
+      "epoch": 13.635185185185184,
+      "grad_norm": 6.156323347452639,
+      "learning_rate": 1.5345145674733296e-07,
+      "loss": 0.6359,
+      "step": 7363
+    },
+    {
+      "epoch": 13.637037037037038,
+      "grad_norm": 1.4335975194366528,
+      "learning_rate": 1.534387836277717e-07,
+      "loss": 0.5937,
+      "step": 7364
+    },
+    {
+      "epoch": 13.63888888888889,
+      "grad_norm": 1.4793057527138505,
+      "learning_rate": 1.5342610930676144e-07,
+      "loss": 0.6375,
+      "step": 7365
+    },
+    {
+      "epoch": 13.64074074074074,
+      "grad_norm": 1.543080566756526,
+      "learning_rate": 1.5341343378458714e-07,
+      "loss": 0.612,
+      "step": 7366
+    },
+    {
+      "epoch": 13.642592592592592,
+      "grad_norm": 1.5461909039071664,
+      "learning_rate": 1.534007570615338e-07,
+      "loss": 0.6266,
+      "step": 7367
+    },
+    {
+      "epoch": 13.644444444444444,
+      "grad_norm": 1.552960230418376,
+      "learning_rate": 1.5338807913788637e-07,
+      "loss": 0.6147,
+      "step": 7368
+    },
+    {
+      "epoch": 13.646296296296295,
+      "grad_norm": 1.5874116701323173,
+      "learning_rate": 1.533754000139299e-07,
+      "loss": 0.6015,
+      "step": 7369
+    },
+    {
+      "epoch": 13.648148148148149,
+      "grad_norm": 1.4788500930859279,
+      "learning_rate": 1.5336271968994953e-07,
+      "loss": 0.6026,
+      "step": 7370
+    },
+    {
+      "epoch": 13.65,
+      "grad_norm": 1.8610380865636993,
+      "learning_rate": 1.5335003816623025e-07,
+      "loss": 0.6267,
+      "step": 7371
+    },
+    {
+      "epoch": 13.651851851851852,
+      "grad_norm": 1.6122250355233885,
+      "learning_rate": 1.5333735544305724e-07,
+      "loss": 0.6289,
+      "step": 7372
+    },
+    {
+      "epoch": 13.653703703703703,
+      "grad_norm": 1.634396649438595,
+      "learning_rate": 1.5332467152071562e-07,
+      "loss": 0.6382,
+      "step": 7373
+    },
+    {
+      "epoch": 13.655555555555555,
+      "grad_norm": 1.4913035533512933,
+      "learning_rate": 1.5331198639949054e-07,
+      "loss": 0.5998,
+      "step": 7374
+    },
+    {
+      "epoch": 13.657407407407408,
+      "grad_norm": 1.7665538456455654,
+      "learning_rate": 1.5329930007966724e-07,
+      "loss": 0.6186,
+      "step": 7375
+    },
+    {
+      "epoch": 13.65925925925926,
+      "grad_norm": 1.4900024599502697,
+      "learning_rate": 1.532866125615309e-07,
+      "loss": 0.5899,
+      "step": 7376
+    },
+    {
+      "epoch": 13.661111111111111,
+      "grad_norm": 2.6166246127599644,
+      "learning_rate": 1.5327392384536677e-07,
+      "loss": 0.6437,
+      "step": 7377
+    },
+    {
+      "epoch": 13.662962962962963,
+      "grad_norm": 1.9815153624534798,
+      "learning_rate": 1.5326123393146015e-07,
+      "loss": 0.6009,
+      "step": 7378
+    },
+    {
+      "epoch": 13.664814814814815,
+      "grad_norm": 1.654381623895425,
+      "learning_rate": 1.5324854282009635e-07,
+      "loss": 0.6323,
+      "step": 7379
+    },
+    {
+      "epoch": 13.666666666666666,
+      "grad_norm": 1.4508373524002751,
+      "learning_rate": 1.5323585051156067e-07,
+      "loss": 0.5848,
+      "step": 7380
+    },
+    {
+      "epoch": 13.668518518518518,
+      "grad_norm": 1.4805706596039312,
+      "learning_rate": 1.532231570061385e-07,
+      "loss": 0.5719,
+      "step": 7381
+    },
+    {
+      "epoch": 13.670370370370371,
+      "grad_norm": 1.4998068656259442,
+      "learning_rate": 1.5321046230411518e-07,
+      "loss": 0.6457,
+      "step": 7382
+    },
+    {
+      "epoch": 13.672222222222222,
+      "grad_norm": 1.7749691130908583,
+      "learning_rate": 1.5319776640577615e-07,
+      "loss": 0.6141,
+      "step": 7383
+    },
+    {
+      "epoch": 13.674074074074074,
+      "grad_norm": 1.6413897860270015,
+      "learning_rate": 1.531850693114069e-07,
+      "loss": 0.5974,
+      "step": 7384
+    },
+    {
+      "epoch": 13.675925925925926,
+      "grad_norm": 1.5885803883419858,
+      "learning_rate": 1.5317237102129278e-07,
+      "loss": 0.6212,
+      "step": 7385
+    },
+    {
+      "epoch": 13.677777777777777,
+      "grad_norm": 1.7864568282209805,
+      "learning_rate": 1.5315967153571938e-07,
+      "loss": 0.6203,
+      "step": 7386
+    },
+    {
+      "epoch": 13.67962962962963,
+      "grad_norm": 1.4952710468927601,
+      "learning_rate": 1.531469708549722e-07,
+      "loss": 0.5791,
+      "step": 7387
+    },
+    {
+      "epoch": 13.681481481481482,
+      "grad_norm": 1.6540614503238513,
+      "learning_rate": 1.5313426897933676e-07,
+      "loss": 0.632,
+      "step": 7388
+    },
+    {
+      "epoch": 13.683333333333334,
+      "grad_norm": 1.530135127565535,
+      "learning_rate": 1.5312156590909862e-07,
+      "loss": 0.6766,
+      "step": 7389
+    },
+    {
+      "epoch": 13.685185185185185,
+      "grad_norm": 2.724539342665472,
+      "learning_rate": 1.5310886164454342e-07,
+      "loss": 0.6259,
+      "step": 7390
+    },
+    {
+      "epoch": 13.687037037037037,
+      "grad_norm": 1.5831600943349637,
+      "learning_rate": 1.5309615618595676e-07,
+      "loss": 0.6185,
+      "step": 7391
+    },
+    {
+      "epoch": 13.688888888888888,
+      "grad_norm": 1.7304900947870607,
+      "learning_rate": 1.530834495336243e-07,
+      "loss": 0.6375,
+      "step": 7392
+    },
+    {
+      "epoch": 13.690740740740742,
+      "grad_norm": 1.5969775695476667,
+      "learning_rate": 1.5307074168783172e-07,
+      "loss": 0.6067,
+      "step": 7393
+    },
+    {
+      "epoch": 13.692592592592593,
+      "grad_norm": 2.12885004052106,
+      "learning_rate": 1.5305803264886478e-07,
+      "loss": 0.6166,
+      "step": 7394
+    },
+    {
+      "epoch": 13.694444444444445,
+      "grad_norm": 1.479468479344743,
+      "learning_rate": 1.530453224170091e-07,
+      "loss": 0.6492,
+      "step": 7395
+    },
+    {
+      "epoch": 13.696296296296296,
+      "grad_norm": 1.415858567106282,
+      "learning_rate": 1.5303261099255051e-07,
+      "loss": 0.6009,
+      "step": 7396
+    },
+    {
+      "epoch": 13.698148148148148,
+      "grad_norm": 1.713486469796353,
+      "learning_rate": 1.5301989837577482e-07,
+      "loss": 0.62,
+      "step": 7397
+    },
+    {
+      "epoch": 13.7,
+      "grad_norm": 1.4126780018601661,
+      "learning_rate": 1.5300718456696777e-07,
+      "loss": 0.5976,
+      "step": 7398
+    },
+    {
+      "epoch": 13.701851851851853,
+      "grad_norm": 2.963230108176807,
+      "learning_rate": 1.5299446956641528e-07,
+      "loss": 0.6037,
+      "step": 7399
+    },
+    {
+      "epoch": 13.703703703703704,
+      "grad_norm": 1.6065410080413833,
+      "learning_rate": 1.529817533744032e-07,
+      "loss": 0.5605,
+      "step": 7400
+    },
+    {
+      "epoch": 13.705555555555556,
+      "grad_norm": 1.4501392385522893,
+      "learning_rate": 1.5296903599121738e-07,
+      "loss": 0.5731,
+      "step": 7401
+    },
+    {
+      "epoch": 13.707407407407407,
+      "grad_norm": 1.6026306288183525,
+      "learning_rate": 1.5295631741714376e-07,
+      "loss": 0.6419,
+      "step": 7402
+    },
+    {
+      "epoch": 13.709259259259259,
+      "grad_norm": 3.99042240217245,
+      "learning_rate": 1.5294359765246833e-07,
+      "loss": 0.5675,
+      "step": 7403
+    },
+    {
+      "epoch": 13.71111111111111,
+      "grad_norm": 1.753696972944809,
+      "learning_rate": 1.52930876697477e-07,
+      "loss": 0.648,
+      "step": 7404
+    },
+    {
+      "epoch": 13.712962962962964,
+      "grad_norm": 1.6167464511805714,
+      "learning_rate": 1.5291815455245582e-07,
+      "loss": 0.6131,
+      "step": 7405
+    },
+    {
+      "epoch": 13.714814814814815,
+      "grad_norm": 1.5039602910889693,
+      "learning_rate": 1.529054312176908e-07,
+      "loss": 0.6076,
+      "step": 7406
+    },
+    {
+      "epoch": 13.716666666666667,
+      "grad_norm": 1.6105651325131158,
+      "learning_rate": 1.5289270669346798e-07,
+      "loss": 0.5775,
+      "step": 7407
+    },
+    {
+      "epoch": 13.718518518518518,
+      "grad_norm": 2.269332215107279,
+      "learning_rate": 1.5287998098007348e-07,
+      "loss": 0.6132,
+      "step": 7408
+    },
+    {
+      "epoch": 13.72037037037037,
+      "grad_norm": 1.6379395157746772,
+      "learning_rate": 1.528672540777934e-07,
+      "loss": 0.6155,
+      "step": 7409
+    },
+    {
+      "epoch": 13.722222222222221,
+      "grad_norm": 1.696520219701363,
+      "learning_rate": 1.5285452598691385e-07,
+      "loss": 0.6205,
+      "step": 7410
+    },
+    {
+      "epoch": 13.724074074074075,
+      "grad_norm": 1.6390249447210787,
+      "learning_rate": 1.52841796707721e-07,
+      "loss": 0.5892,
+      "step": 7411
+    },
+    {
+      "epoch": 13.725925925925926,
+      "grad_norm": 1.4353779472714796,
+      "learning_rate": 1.5282906624050106e-07,
+      "loss": 0.6021,
+      "step": 7412
+    },
+    {
+      "epoch": 13.727777777777778,
+      "grad_norm": 1.665041093117239,
+      "learning_rate": 1.528163345855402e-07,
+      "loss": 0.6219,
+      "step": 7413
+    },
+    {
+      "epoch": 13.72962962962963,
+      "grad_norm": 1.8361238860153073,
+      "learning_rate": 1.528036017431247e-07,
+      "loss": 0.6302,
+      "step": 7414
+    },
+    {
+      "epoch": 13.731481481481481,
+      "grad_norm": 2.444227357715716,
+      "learning_rate": 1.5279086771354085e-07,
+      "loss": 0.6384,
+      "step": 7415
+    },
+    {
+      "epoch": 13.733333333333333,
+      "grad_norm": 1.676112674264642,
+      "learning_rate": 1.5277813249707485e-07,
+      "loss": 0.6277,
+      "step": 7416
+    },
+    {
+      "epoch": 13.735185185185186,
+      "grad_norm": 1.5411296763924103,
+      "learning_rate": 1.5276539609401313e-07,
+      "loss": 0.5846,
+      "step": 7417
+    },
+    {
+      "epoch": 13.737037037037037,
+      "grad_norm": 1.8206556729405725,
+      "learning_rate": 1.52752658504642e-07,
+      "loss": 0.6279,
+      "step": 7418
+    },
+    {
+      "epoch": 13.738888888888889,
+      "grad_norm": 1.5270173900009647,
+      "learning_rate": 1.5273991972924785e-07,
+      "loss": 0.5975,
+      "step": 7419
+    },
+    {
+      "epoch": 13.74074074074074,
+      "grad_norm": 1.4001710954295337,
+      "learning_rate": 1.5272717976811706e-07,
+      "loss": 0.5899,
+      "step": 7420
+    },
+    {
+      "epoch": 13.742592592592592,
+      "grad_norm": 1.7603814253515966,
+      "learning_rate": 1.5271443862153602e-07,
+      "loss": 0.6483,
+      "step": 7421
+    },
+    {
+      "epoch": 13.744444444444444,
+      "grad_norm": 2.231649106584098,
+      "learning_rate": 1.5270169628979125e-07,
+      "loss": 0.6435,
+      "step": 7422
+    },
+    {
+      "epoch": 13.746296296296297,
+      "grad_norm": 1.9902801010244227,
+      "learning_rate": 1.5268895277316927e-07,
+      "loss": 0.6372,
+      "step": 7423
+    },
+    {
+      "epoch": 13.748148148148148,
+      "grad_norm": 2.219839439468968,
+      "learning_rate": 1.5267620807195648e-07,
+      "loss": 0.6347,
+      "step": 7424
+    },
+    {
+      "epoch": 13.75,
+      "grad_norm": 1.6064749472489948,
+      "learning_rate": 1.5266346218643947e-07,
+      "loss": 0.6216,
+      "step": 7425
+    },
+    {
+      "epoch": 13.751851851851852,
+      "grad_norm": 2.9189469580457255,
+      "learning_rate": 1.526507151169048e-07,
+      "loss": 0.633,
+      "step": 7426
+    },
+    {
+      "epoch": 13.753703703703703,
+      "grad_norm": 1.8347096468245494,
+      "learning_rate": 1.5263796686363906e-07,
+      "loss": 0.6008,
+      "step": 7427
+    },
+    {
+      "epoch": 13.755555555555556,
+      "grad_norm": 1.7388257856616471,
+      "learning_rate": 1.5262521742692882e-07,
+      "loss": 0.6186,
+      "step": 7428
+    },
+    {
+      "epoch": 13.757407407407408,
+      "grad_norm": 1.7701552838125272,
+      "learning_rate": 1.5261246680706081e-07,
+      "loss": 0.6224,
+      "step": 7429
+    },
+    {
+      "epoch": 13.75925925925926,
+      "grad_norm": 1.4783343364839439,
+      "learning_rate": 1.5259971500432164e-07,
+      "loss": 0.5875,
+      "step": 7430
+    },
+    {
+      "epoch": 13.761111111111111,
+      "grad_norm": 1.727381410170526,
+      "learning_rate": 1.52586962018998e-07,
+      "loss": 0.5658,
+      "step": 7431
+    },
+    {
+      "epoch": 13.762962962962963,
+      "grad_norm": 1.6390493812494877,
+      "learning_rate": 1.5257420785137667e-07,
+      "loss": 0.6398,
+      "step": 7432
+    },
+    {
+      "epoch": 13.764814814814814,
+      "grad_norm": 1.4442205600902545,
+      "learning_rate": 1.5256145250174432e-07,
+      "loss": 0.596,
+      "step": 7433
+    },
+    {
+      "epoch": 13.766666666666667,
+      "grad_norm": 1.7294103959350773,
+      "learning_rate": 1.525486959703878e-07,
+      "loss": 0.6363,
+      "step": 7434
+    },
+    {
+      "epoch": 13.768518518518519,
+      "grad_norm": 1.4786264646766074,
+      "learning_rate": 1.5253593825759389e-07,
+      "loss": 0.6227,
+      "step": 7435
+    },
+    {
+      "epoch": 13.77037037037037,
+      "grad_norm": 1.7818247306411379,
+      "learning_rate": 1.5252317936364936e-07,
+      "loss": 0.6169,
+      "step": 7436
+    },
+    {
+      "epoch": 13.772222222222222,
+      "grad_norm": 1.8125349478392643,
+      "learning_rate": 1.5251041928884109e-07,
+      "loss": 0.5846,
+      "step": 7437
+    },
+    {
+      "epoch": 13.774074074074074,
+      "grad_norm": 1.9276764408965896,
+      "learning_rate": 1.52497658033456e-07,
+      "loss": 0.6637,
+      "step": 7438
+    },
+    {
+      "epoch": 13.775925925925925,
+      "grad_norm": 1.4501373463490161,
+      "learning_rate": 1.52484895597781e-07,
+      "loss": 0.6232,
+      "step": 7439
+    },
+    {
+      "epoch": 13.777777777777779,
+      "grad_norm": 13.210332842084,
+      "learning_rate": 1.5247213198210298e-07,
+      "loss": 0.5964,
+      "step": 7440
+    },
+    {
+      "epoch": 13.77962962962963,
+      "grad_norm": 1.4691993631638716,
+      "learning_rate": 1.5245936718670894e-07,
+      "loss": 0.6464,
+      "step": 7441
+    },
+    {
+      "epoch": 13.781481481481482,
+      "grad_norm": 1.7313381439660658,
+      "learning_rate": 1.5244660121188582e-07,
+      "loss": 0.5708,
+      "step": 7442
+    },
+    {
+      "epoch": 13.783333333333333,
+      "grad_norm": 1.8955162085848853,
+      "learning_rate": 1.5243383405792065e-07,
+      "loss": 0.5971,
+      "step": 7443
+    },
+    {
+      "epoch": 13.785185185185185,
+      "grad_norm": 1.3887040752912538,
+      "learning_rate": 1.5242106572510053e-07,
+      "loss": 0.5704,
+      "step": 7444
+    },
+    {
+      "epoch": 13.787037037037036,
+      "grad_norm": 1.4437963360413026,
+      "learning_rate": 1.5240829621371246e-07,
+      "loss": 0.6285,
+      "step": 7445
+    },
+    {
+      "epoch": 13.78888888888889,
+      "grad_norm": 1.6516054223999488,
+      "learning_rate": 1.5239552552404352e-07,
+      "loss": 0.5954,
+      "step": 7446
+    },
+    {
+      "epoch": 13.790740740740741,
+      "grad_norm": 1.6698616190381441,
+      "learning_rate": 1.5238275365638087e-07,
+      "loss": 0.6293,
+      "step": 7447
+    },
+    {
+      "epoch": 13.792592592592593,
+      "grad_norm": 1.5485082839958755,
+      "learning_rate": 1.523699806110117e-07,
+      "loss": 0.6373,
+      "step": 7448
+    },
+    {
+      "epoch": 13.794444444444444,
+      "grad_norm": 1.6018300173580908,
+      "learning_rate": 1.5235720638822307e-07,
+      "loss": 0.6118,
+      "step": 7449
+    },
+    {
+      "epoch": 13.796296296296296,
+      "grad_norm": 1.5638209961825218,
+      "learning_rate": 1.5234443098830227e-07,
+      "loss": 0.6082,
+      "step": 7450
+    },
+    {
+      "epoch": 13.798148148148147,
+      "grad_norm": 1.6205340752199529,
+      "learning_rate": 1.523316544115365e-07,
+      "loss": 0.6201,
+      "step": 7451
+    },
+    {
+      "epoch": 13.8,
+      "grad_norm": 2.270837949949198,
+      "learning_rate": 1.52318876658213e-07,
+      "loss": 0.6118,
+      "step": 7452
+    },
+    {
+      "epoch": 13.801851851851852,
+      "grad_norm": 1.5336222349376536,
+      "learning_rate": 1.5230609772861906e-07,
+      "loss": 0.6208,
+      "step": 7453
+    },
+    {
+      "epoch": 13.803703703703704,
+      "grad_norm": 1.6090759455501418,
+      "learning_rate": 1.5229331762304198e-07,
+      "loss": 0.676,
+      "step": 7454
+    },
+    {
+      "epoch": 13.805555555555555,
+      "grad_norm": 1.5691235604657463,
+      "learning_rate": 1.5228053634176908e-07,
+      "loss": 0.6195,
+      "step": 7455
+    },
+    {
+      "epoch": 13.807407407407407,
+      "grad_norm": 1.6087985340238542,
+      "learning_rate": 1.5226775388508776e-07,
+      "loss": 0.5408,
+      "step": 7456
+    },
+    {
+      "epoch": 13.809259259259258,
+      "grad_norm": 1.7289181264702418,
+      "learning_rate": 1.5225497025328538e-07,
+      "loss": 0.6325,
+      "step": 7457
+    },
+    {
+      "epoch": 13.811111111111112,
+      "grad_norm": 1.616759707457118,
+      "learning_rate": 1.5224218544664932e-07,
+      "loss": 0.5784,
+      "step": 7458
+    },
+    {
+      "epoch": 13.812962962962963,
+      "grad_norm": 2.358364085658103,
+      "learning_rate": 1.5222939946546705e-07,
+      "loss": 0.5743,
+      "step": 7459
+    },
+    {
+      "epoch": 13.814814814814815,
+      "grad_norm": 1.624200904225749,
+      "learning_rate": 1.5221661231002604e-07,
+      "loss": 0.6122,
+      "step": 7460
+    },
+    {
+      "epoch": 13.816666666666666,
+      "grad_norm": 2.0094410410225074,
+      "learning_rate": 1.5220382398061372e-07,
+      "loss": 0.5968,
+      "step": 7461
+    },
+    {
+      "epoch": 13.818518518518518,
+      "grad_norm": 1.7524849572883625,
+      "learning_rate": 1.521910344775177e-07,
+      "loss": 0.6084,
+      "step": 7462
+    },
+    {
+      "epoch": 13.82037037037037,
+      "grad_norm": 2.2907665283520746,
+      "learning_rate": 1.5217824380102546e-07,
+      "loss": 0.6396,
+      "step": 7463
+    },
+    {
+      "epoch": 13.822222222222223,
+      "grad_norm": 1.6997170285552388,
+      "learning_rate": 1.521654519514246e-07,
+      "loss": 0.5967,
+      "step": 7464
+    },
+    {
+      "epoch": 13.824074074074074,
+      "grad_norm": 1.4812071035840988,
+      "learning_rate": 1.5215265892900268e-07,
+      "loss": 0.6304,
+      "step": 7465
+    },
+    {
+      "epoch": 13.825925925925926,
+      "grad_norm": 1.3337623288502478,
+      "learning_rate": 1.5213986473404738e-07,
+      "loss": 0.6512,
+      "step": 7466
+    },
+    {
+      "epoch": 13.827777777777778,
+      "grad_norm": 1.774889839335562,
+      "learning_rate": 1.5212706936684625e-07,
+      "loss": 0.5953,
+      "step": 7467
+    },
+    {
+      "epoch": 13.829629629629629,
+      "grad_norm": 1.5128671129439468,
+      "learning_rate": 1.5211427282768707e-07,
+      "loss": 0.6017,
+      "step": 7468
+    },
+    {
+      "epoch": 13.831481481481482,
+      "grad_norm": 1.8758854791859465,
+      "learning_rate": 1.5210147511685748e-07,
+      "loss": 0.6377,
+      "step": 7469
+    },
+    {
+      "epoch": 13.833333333333334,
+      "grad_norm": 1.776259658725903,
+      "learning_rate": 1.5208867623464525e-07,
+      "loss": 0.6105,
+      "step": 7470
+    },
+    {
+      "epoch": 13.835185185185185,
+      "grad_norm": 1.6788360374008477,
+      "learning_rate": 1.520758761813381e-07,
+      "loss": 0.6365,
+      "step": 7471
+    },
+    {
+      "epoch": 13.837037037037037,
+      "grad_norm": 1.7986688118321117,
+      "learning_rate": 1.5206307495722376e-07,
+      "loss": 0.6342,
+      "step": 7472
+    },
+    {
+      "epoch": 13.838888888888889,
+      "grad_norm": 1.7672589686278617,
+      "learning_rate": 1.5205027256259015e-07,
+      "loss": 0.6109,
+      "step": 7473
+    },
+    {
+      "epoch": 13.84074074074074,
+      "grad_norm": 1.5984206854799181,
+      "learning_rate": 1.5203746899772506e-07,
+      "loss": 0.6271,
+      "step": 7474
+    },
+    {
+      "epoch": 13.842592592592592,
+      "grad_norm": 1.6617294882748468,
+      "learning_rate": 1.5202466426291629e-07,
+      "loss": 0.6443,
+      "step": 7475
+    },
+    {
+      "epoch": 13.844444444444445,
+      "grad_norm": 1.777347321358636,
+      "learning_rate": 1.520118583584518e-07,
+      "loss": 0.6037,
+      "step": 7476
+    },
+    {
+      "epoch": 13.846296296296297,
+      "grad_norm": 3.0112949931328545,
+      "learning_rate": 1.5199905128461946e-07,
+      "loss": 0.6764,
+      "step": 7477
+    },
+    {
+      "epoch": 13.848148148148148,
+      "grad_norm": 1.3697416622569698,
+      "learning_rate": 1.5198624304170724e-07,
+      "loss": 0.624,
+      "step": 7478
+    },
+    {
+      "epoch": 13.85,
+      "grad_norm": 1.6242529378445814,
+      "learning_rate": 1.5197343363000306e-07,
+      "loss": 0.6214,
+      "step": 7479
+    },
+    {
+      "epoch": 13.851851851851851,
+      "grad_norm": 1.8263633319167236,
+      "learning_rate": 1.5196062304979495e-07,
+      "loss": 0.6323,
+      "step": 7480
+    },
+    {
+      "epoch": 13.853703703703705,
+      "grad_norm": 1.9313327263554292,
+      "learning_rate": 1.519478113013709e-07,
+      "loss": 0.5819,
+      "step": 7481
+    },
+    {
+      "epoch": 13.855555555555556,
+      "grad_norm": 1.8538232588938657,
+      "learning_rate": 1.5193499838501894e-07,
+      "loss": 0.6124,
+      "step": 7482
+    },
+    {
+      "epoch": 13.857407407407408,
+      "grad_norm": 1.5878132919931682,
+      "learning_rate": 1.519221843010272e-07,
+      "loss": 0.5779,
+      "step": 7483
+    },
+    {
+      "epoch": 13.85925925925926,
+      "grad_norm": 1.7082702356973152,
+      "learning_rate": 1.5190936904968373e-07,
+      "loss": 0.6115,
+      "step": 7484
+    },
+    {
+      "epoch": 13.86111111111111,
+      "grad_norm": 1.44388484363116,
+      "learning_rate": 1.5189655263127665e-07,
+      "loss": 0.5855,
+      "step": 7485
+    },
+    {
+      "epoch": 13.862962962962962,
+      "grad_norm": 1.9176393634695494,
+      "learning_rate": 1.5188373504609416e-07,
+      "loss": 0.64,
+      "step": 7486
+    },
+    {
+      "epoch": 13.864814814814816,
+      "grad_norm": 1.7849555872534677,
+      "learning_rate": 1.5187091629442435e-07,
+      "loss": 0.5627,
+      "step": 7487
+    },
+    {
+      "epoch": 13.866666666666667,
+      "grad_norm": 2.241748322442699,
+      "learning_rate": 1.5185809637655547e-07,
+      "loss": 0.5627,
+      "step": 7488
+    },
+    {
+      "epoch": 13.868518518518519,
+      "grad_norm": 1.658344388672207,
+      "learning_rate": 1.5184527529277573e-07,
+      "loss": 0.5979,
+      "step": 7489
+    },
+    {
+      "epoch": 13.87037037037037,
+      "grad_norm": 1.6379350935913863,
+      "learning_rate": 1.518324530433734e-07,
+      "loss": 0.6274,
+      "step": 7490
+    },
+    {
+      "epoch": 13.872222222222222,
+      "grad_norm": 1.6469432022062998,
+      "learning_rate": 1.5181962962863675e-07,
+      "loss": 0.6199,
+      "step": 7491
+    },
+    {
+      "epoch": 13.874074074074073,
+      "grad_norm": 1.6831439133032498,
+      "learning_rate": 1.518068050488541e-07,
+      "loss": 0.6441,
+      "step": 7492
+    },
+    {
+      "epoch": 13.875925925925927,
+      "grad_norm": 2.4944938096229006,
+      "learning_rate": 1.5179397930431376e-07,
+      "loss": 0.6658,
+      "step": 7493
+    },
+    {
+      "epoch": 13.877777777777778,
+      "grad_norm": 1.6180551369726002,
+      "learning_rate": 1.5178115239530408e-07,
+      "loss": 0.6256,
+      "step": 7494
+    },
+    {
+      "epoch": 13.87962962962963,
+      "grad_norm": 1.6910039192989743,
+      "learning_rate": 1.5176832432211347e-07,
+      "loss": 0.6018,
+      "step": 7495
+    },
+    {
+      "epoch": 13.881481481481481,
+      "grad_norm": 1.90187540617632,
+      "learning_rate": 1.5175549508503035e-07,
+      "loss": 0.6609,
+      "step": 7496
+    },
+    {
+      "epoch": 13.883333333333333,
+      "grad_norm": 1.52060504381097,
+      "learning_rate": 1.517426646843431e-07,
+      "loss": 0.644,
+      "step": 7497
+    },
+    {
+      "epoch": 13.885185185185184,
+      "grad_norm": 1.4153839951882483,
+      "learning_rate": 1.5172983312034022e-07,
+      "loss": 0.5626,
+      "step": 7498
+    },
+    {
+      "epoch": 13.887037037037038,
+      "grad_norm": 1.919721454186169,
+      "learning_rate": 1.517170003933102e-07,
+      "loss": 0.61,
+      "step": 7499
+    },
+    {
+      "epoch": 13.88888888888889,
+      "grad_norm": 1.5001673423864423,
+      "learning_rate": 1.5170416650354154e-07,
+      "loss": 0.6032,
+      "step": 7500
+    },
+    {
+      "epoch": 13.89074074074074,
+      "grad_norm": 1.5437771002134788,
+      "learning_rate": 1.5169133145132283e-07,
+      "loss": 0.613,
+      "step": 7501
+    },
+    {
+      "epoch": 13.892592592592592,
+      "grad_norm": 1.547238307456619,
+      "learning_rate": 1.5167849523694258e-07,
+      "loss": 0.6299,
+      "step": 7502
+    },
+    {
+      "epoch": 13.894444444444444,
+      "grad_norm": 2.0772894368064967,
+      "learning_rate": 1.516656578606894e-07,
+      "loss": 0.6165,
+      "step": 7503
+    },
+    {
+      "epoch": 13.896296296296295,
+      "grad_norm": 1.423394506398061,
+      "learning_rate": 1.516528193228519e-07,
+      "loss": 0.6214,
+      "step": 7504
+    },
+    {
+      "epoch": 13.898148148148149,
+      "grad_norm": 1.6038144541634802,
+      "learning_rate": 1.516399796237187e-07,
+      "loss": 0.6162,
+      "step": 7505
+    },
+    {
+      "epoch": 13.9,
+      "grad_norm": 1.6410251630798371,
+      "learning_rate": 1.5162713876357859e-07,
+      "loss": 0.6452,
+      "step": 7506
+    },
+    {
+      "epoch": 13.901851851851852,
+      "grad_norm": 1.5118939799219775,
+      "learning_rate": 1.5161429674272014e-07,
+      "loss": 0.5967,
+      "step": 7507
+    },
+    {
+      "epoch": 13.903703703703703,
+      "grad_norm": 2.072781119916379,
+      "learning_rate": 1.516014535614321e-07,
+      "loss": 0.6068,
+      "step": 7508
+    },
+    {
+      "epoch": 13.905555555555555,
+      "grad_norm": 1.6248162068287648,
+      "learning_rate": 1.5158860922000326e-07,
+      "loss": 0.5914,
+      "step": 7509
+    },
+    {
+      "epoch": 13.907407407407408,
+      "grad_norm": 1.6968485150986787,
+      "learning_rate": 1.5157576371872235e-07,
+      "loss": 0.5756,
+      "step": 7510
+    },
+    {
+      "epoch": 13.90925925925926,
+      "grad_norm": 1.6217502390994523,
+      "learning_rate": 1.515629170578782e-07,
+      "loss": 0.6423,
+      "step": 7511
+    },
+    {
+      "epoch": 13.911111111111111,
+      "grad_norm": 1.761199447734665,
+      "learning_rate": 1.5155006923775965e-07,
+      "loss": 0.6122,
+      "step": 7512
+    },
+    {
+      "epoch": 13.912962962962963,
+      "grad_norm": 1.3563400575727296,
+      "learning_rate": 1.515372202586555e-07,
+      "loss": 0.6462,
+      "step": 7513
+    },
+    {
+      "epoch": 13.914814814814815,
+      "grad_norm": 1.4803800264250055,
+      "learning_rate": 1.5152437012085468e-07,
+      "loss": 0.5768,
+      "step": 7514
+    },
+    {
+      "epoch": 13.916666666666666,
+      "grad_norm": 1.537054205882389,
+      "learning_rate": 1.515115188246461e-07,
+      "loss": 0.6204,
+      "step": 7515
+    },
+    {
+      "epoch": 13.918518518518518,
+      "grad_norm": 1.435735383814482,
+      "learning_rate": 1.514986663703187e-07,
+      "loss": 0.6501,
+      "step": 7516
+    },
+    {
+      "epoch": 13.920370370370371,
+      "grad_norm": 1.5522668155834092,
+      "learning_rate": 1.5148581275816134e-07,
+      "loss": 0.6057,
+      "step": 7517
+    },
+    {
+      "epoch": 13.922222222222222,
+      "grad_norm": 1.4938712213309915,
+      "learning_rate": 1.5147295798846312e-07,
+      "loss": 0.601,
+      "step": 7518
+    },
+    {
+      "epoch": 13.924074074074074,
+      "grad_norm": 1.6221738162386912,
+      "learning_rate": 1.51460102061513e-07,
+      "loss": 0.5997,
+      "step": 7519
+    },
+    {
+      "epoch": 13.925925925925926,
+      "grad_norm": 1.6207622347607011,
+      "learning_rate": 1.514472449776e-07,
+      "loss": 0.6107,
+      "step": 7520
+    },
+    {
+      "epoch": 13.927777777777777,
+      "grad_norm": 1.7690674963714104,
+      "learning_rate": 1.5143438673701324e-07,
+      "loss": 0.5976,
+      "step": 7521
+    },
+    {
+      "epoch": 13.92962962962963,
+      "grad_norm": 1.7220497798666956,
+      "learning_rate": 1.5142152734004177e-07,
+      "loss": 0.6027,
+      "step": 7522
+    },
+    {
+      "epoch": 13.931481481481482,
+      "grad_norm": 1.3988593283737722,
+      "learning_rate": 1.514086667869747e-07,
+      "loss": 0.6133,
+      "step": 7523
+    },
+    {
+      "epoch": 13.933333333333334,
+      "grad_norm": 1.7188763586541154,
+      "learning_rate": 1.5139580507810117e-07,
+      "loss": 0.5559,
+      "step": 7524
+    },
+    {
+      "epoch": 13.935185185185185,
+      "grad_norm": 1.589229315802639,
+      "learning_rate": 1.5138294221371036e-07,
+      "loss": 0.5956,
+      "step": 7525
+    },
+    {
+      "epoch": 13.937037037037037,
+      "grad_norm": 2.445605083038169,
+      "learning_rate": 1.513700781940914e-07,
+      "loss": 0.666,
+      "step": 7526
+    },
+    {
+      "epoch": 13.938888888888888,
+      "grad_norm": 1.8101960082229966,
+      "learning_rate": 1.5135721301953367e-07,
+      "loss": 0.601,
+      "step": 7527
+    },
+    {
+      "epoch": 13.940740740740742,
+      "grad_norm": 2.3588938450912442,
+      "learning_rate": 1.5134434669032623e-07,
+      "loss": 0.6456,
+      "step": 7528
+    },
+    {
+      "epoch": 13.942592592592593,
+      "grad_norm": 1.432804026013408,
+      "learning_rate": 1.5133147920675842e-07,
+      "loss": 0.5891,
+      "step": 7529
+    },
+    {
+      "epoch": 13.944444444444445,
+      "grad_norm": 1.475675975937916,
+      "learning_rate": 1.5131861056911958e-07,
+      "loss": 0.6639,
+      "step": 7530
+    },
+    {
+      "epoch": 13.946296296296296,
+      "grad_norm": 1.7043737475500285,
+      "learning_rate": 1.5130574077769901e-07,
+      "loss": 0.6023,
+      "step": 7531
+    },
+    {
+      "epoch": 13.948148148148148,
+      "grad_norm": 1.8090890352571212,
+      "learning_rate": 1.51292869832786e-07,
+      "loss": 0.6031,
+      "step": 7532
+    },
+    {
+      "epoch": 13.95,
+      "grad_norm": 1.5429994691003408,
+      "learning_rate": 1.5127999773467e-07,
+      "loss": 0.56,
+      "step": 7533
+    },
+    {
+      "epoch": 13.951851851851853,
+      "grad_norm": 1.5946418815525665,
+      "learning_rate": 1.5126712448364035e-07,
+      "loss": 0.5918,
+      "step": 7534
+    },
+    {
+      "epoch": 13.953703703703704,
+      "grad_norm": 1.9922957525974152,
+      "learning_rate": 1.512542500799865e-07,
+      "loss": 0.5742,
+      "step": 7535
+    },
+    {
+      "epoch": 13.955555555555556,
+      "grad_norm": 1.7302611749906858,
+      "learning_rate": 1.5124137452399793e-07,
+      "loss": 0.5735,
+      "step": 7536
+    },
+    {
+      "epoch": 13.957407407407407,
+      "grad_norm": 1.6184961319236364,
+      "learning_rate": 1.512284978159641e-07,
+      "loss": 0.645,
+      "step": 7537
+    },
+    {
+      "epoch": 13.959259259259259,
+      "grad_norm": 2.7841777943945534,
+      "learning_rate": 1.5121561995617446e-07,
+      "loss": 0.5999,
+      "step": 7538
+    },
+    {
+      "epoch": 13.96111111111111,
+      "grad_norm": 1.6213083384182314,
+      "learning_rate": 1.5120274094491863e-07,
+      "loss": 0.6352,
+      "step": 7539
+    },
+    {
+      "epoch": 13.962962962962964,
+      "grad_norm": 1.7370440385827413,
+      "learning_rate": 1.511898607824861e-07,
+      "loss": 0.5674,
+      "step": 7540
+    },
+    {
+      "epoch": 13.964814814814815,
+      "grad_norm": 1.6061926243213158,
+      "learning_rate": 1.5117697946916645e-07,
+      "loss": 0.6464,
+      "step": 7541
+    },
+    {
+      "epoch": 13.966666666666667,
+      "grad_norm": 1.4386439004730254,
+      "learning_rate": 1.5116409700524933e-07,
+      "loss": 0.6309,
+      "step": 7542
+    },
+    {
+      "epoch": 13.968518518518518,
+      "grad_norm": 1.5727221414471406,
+      "learning_rate": 1.5115121339102435e-07,
+      "loss": 0.5621,
+      "step": 7543
+    },
+    {
+      "epoch": 13.97037037037037,
+      "grad_norm": 1.7831249372749607,
+      "learning_rate": 1.5113832862678113e-07,
+      "loss": 0.6112,
+      "step": 7544
+    },
+    {
+      "epoch": 13.972222222222221,
+      "grad_norm": 1.6562551890252326,
+      "learning_rate": 1.5112544271280942e-07,
+      "loss": 0.623,
+      "step": 7545
+    },
+    {
+      "epoch": 13.974074074074075,
+      "grad_norm": 1.486421417830827,
+      "learning_rate": 1.5111255564939894e-07,
+      "loss": 0.5853,
+      "step": 7546
+    },
+    {
+      "epoch": 13.975925925925926,
+      "grad_norm": 1.814560222068078,
+      "learning_rate": 1.5109966743683936e-07,
+      "loss": 0.6129,
+      "step": 7547
+    },
+    {
+      "epoch": 13.977777777777778,
+      "grad_norm": 1.3952354730022678,
+      "learning_rate": 1.5108677807542045e-07,
+      "loss": 0.5786,
+      "step": 7548
+    },
+    {
+      "epoch": 13.97962962962963,
+      "grad_norm": 1.7886332037068213,
+      "learning_rate": 1.5107388756543203e-07,
+      "loss": 0.6267,
+      "step": 7549
+    },
+    {
+      "epoch": 13.981481481481481,
+      "grad_norm": 1.672169837339197,
+      "learning_rate": 1.510609959071639e-07,
+      "loss": 0.6354,
+      "step": 7550
+    },
+    {
+      "epoch": 13.983333333333333,
+      "grad_norm": 3.231602464363265,
+      "learning_rate": 1.5104810310090592e-07,
+      "loss": 0.6085,
+      "step": 7551
+    },
+    {
+      "epoch": 13.985185185185186,
+      "grad_norm": 1.582589941045024,
+      "learning_rate": 1.5103520914694793e-07,
+      "loss": 0.6216,
+      "step": 7552
+    },
+    {
+      "epoch": 13.987037037037037,
+      "grad_norm": 1.488304524830707,
+      "learning_rate": 1.5102231404557984e-07,
+      "loss": 0.595,
+      "step": 7553
+    },
+    {
+      "epoch": 13.988888888888889,
+      "grad_norm": 1.4600270865327551,
+      "learning_rate": 1.5100941779709154e-07,
+      "loss": 0.6377,
+      "step": 7554
+    },
+    {
+      "epoch": 13.99074074074074,
+      "grad_norm": 1.5290059930363933,
+      "learning_rate": 1.50996520401773e-07,
+      "loss": 0.5955,
+      "step": 7555
+    },
+    {
+      "epoch": 13.992592592592592,
+      "grad_norm": 1.4632364788713286,
+      "learning_rate": 1.5098362185991417e-07,
+      "loss": 0.6243,
+      "step": 7556
+    },
+    {
+      "epoch": 13.994444444444444,
+      "grad_norm": 1.5916757089413898,
+      "learning_rate": 1.5097072217180508e-07,
+      "loss": 0.6132,
+      "step": 7557
+    },
+    {
+      "epoch": 13.996296296296297,
+      "grad_norm": 1.9864804342332922,
+      "learning_rate": 1.5095782133773568e-07,
+      "loss": 0.6081,
+      "step": 7558
+    },
+    {
+      "epoch": 13.998148148148148,
+      "grad_norm": 2.192119728323222,
+      "learning_rate": 1.5094491935799604e-07,
+      "loss": 0.6164,
+      "step": 7559
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 1.6209309242937744,
+      "learning_rate": 1.509320162328763e-07,
+      "loss": 0.6022,
+      "step": 7560
+    },
+    {
+      "epoch": 14.001851851851852,
+      "grad_norm": 2.604758127966735,
+      "learning_rate": 1.5091911196266648e-07,
+      "loss": 0.5719,
+      "step": 7561
+    },
+    {
+      "epoch": 14.003703703703703,
+      "grad_norm": 1.75994585102352,
+      "learning_rate": 1.5090620654765674e-07,
+      "loss": 0.5956,
+      "step": 7562
+    },
+    {
+      "epoch": 14.005555555555556,
+      "grad_norm": 2.827194029502754,
+      "learning_rate": 1.508932999881372e-07,
+      "loss": 0.6064,
+      "step": 7563
+    },
+    {
+      "epoch": 14.007407407407408,
+      "grad_norm": 1.9463303038473025,
+      "learning_rate": 1.5088039228439807e-07,
+      "loss": 0.6247,
+      "step": 7564
+    },
+    {
+      "epoch": 14.00925925925926,
+      "grad_norm": 1.50148938554801,
+      "learning_rate": 1.5086748343672953e-07,
+      "loss": 0.6056,
+      "step": 7565
+    },
+    {
+      "epoch": 14.011111111111111,
+      "grad_norm": 1.5116653951358134,
+      "learning_rate": 1.508545734454218e-07,
+      "loss": 0.639,
+      "step": 7566
+    },
+    {
+      "epoch": 14.012962962962963,
+      "grad_norm": 1.4749593468434457,
+      "learning_rate": 1.5084166231076515e-07,
+      "loss": 0.6337,
+      "step": 7567
+    },
+    {
+      "epoch": 14.014814814814814,
+      "grad_norm": 9.499893469555719,
+      "learning_rate": 1.5082875003304985e-07,
+      "loss": 0.5658,
+      "step": 7568
+    },
+    {
+      "epoch": 14.016666666666667,
+      "grad_norm": 1.5937056086985182,
+      "learning_rate": 1.5081583661256619e-07,
+      "loss": 0.6391,
+      "step": 7569
+    },
+    {
+      "epoch": 14.018518518518519,
+      "grad_norm": 1.5027774272678838,
+      "learning_rate": 1.5080292204960448e-07,
+      "loss": 0.6541,
+      "step": 7570
+    },
+    {
+      "epoch": 14.02037037037037,
+      "grad_norm": 1.5736168915432671,
+      "learning_rate": 1.5079000634445514e-07,
+      "loss": 0.5793,
+      "step": 7571
+    },
+    {
+      "epoch": 14.022222222222222,
+      "grad_norm": 2.466699452266482,
+      "learning_rate": 1.5077708949740854e-07,
+      "loss": 0.6091,
+      "step": 7572
+    },
+    {
+      "epoch": 14.024074074074074,
+      "grad_norm": 1.4343678001463427,
+      "learning_rate": 1.50764171508755e-07,
+      "loss": 0.6364,
+      "step": 7573
+    },
+    {
+      "epoch": 14.025925925925925,
+      "grad_norm": 1.9198521254907177,
+      "learning_rate": 1.5075125237878505e-07,
+      "loss": 0.5933,
+      "step": 7574
+    },
+    {
+      "epoch": 14.027777777777779,
+      "grad_norm": 1.5941251095107105,
+      "learning_rate": 1.5073833210778907e-07,
+      "loss": 0.5754,
+      "step": 7575
+    },
+    {
+      "epoch": 14.02962962962963,
+      "grad_norm": 2.0927889274354974,
+      "learning_rate": 1.5072541069605764e-07,
+      "loss": 0.6054,
+      "step": 7576
+    },
+    {
+      "epoch": 14.031481481481482,
+      "grad_norm": 2.1084662846595412,
+      "learning_rate": 1.5071248814388118e-07,
+      "loss": 0.6371,
+      "step": 7577
+    },
+    {
+      "epoch": 14.033333333333333,
+      "grad_norm": 1.598864436970099,
+      "learning_rate": 1.5069956445155025e-07,
+      "loss": 0.6429,
+      "step": 7578
+    },
+    {
+      "epoch": 14.035185185185185,
+      "grad_norm": 2.225775838516352,
+      "learning_rate": 1.506866396193554e-07,
+      "loss": 0.5802,
+      "step": 7579
+    },
+    {
+      "epoch": 14.037037037037036,
+      "grad_norm": 1.4592249300620121,
+      "learning_rate": 1.5067371364758724e-07,
+      "loss": 0.5609,
+      "step": 7580
+    },
+    {
+      "epoch": 14.03888888888889,
+      "grad_norm": 2.2113102484368277,
+      "learning_rate": 1.506607865365364e-07,
+      "loss": 0.6126,
+      "step": 7581
+    },
+    {
+      "epoch": 14.040740740740741,
+      "grad_norm": 1.5418281542151708,
+      "learning_rate": 1.506478582864934e-07,
+      "loss": 0.6794,
+      "step": 7582
+    },
+    {
+      "epoch": 14.042592592592593,
+      "grad_norm": 1.7288794564584509,
+      "learning_rate": 1.506349288977491e-07,
+      "loss": 0.6055,
+      "step": 7583
+    },
+    {
+      "epoch": 14.044444444444444,
+      "grad_norm": 1.9271581446159896,
+      "learning_rate": 1.5062199837059404e-07,
+      "loss": 0.5879,
+      "step": 7584
+    },
+    {
+      "epoch": 14.046296296296296,
+      "grad_norm": 1.6954936895958168,
+      "learning_rate": 1.5060906670531893e-07,
+      "loss": 0.6039,
+      "step": 7585
+    },
+    {
+      "epoch": 14.048148148148147,
+      "grad_norm": 1.4981864965486094,
+      "learning_rate": 1.5059613390221456e-07,
+      "loss": 0.5947,
+      "step": 7586
+    },
+    {
+      "epoch": 14.05,
+      "grad_norm": 1.5304071770474366,
+      "learning_rate": 1.505831999615717e-07,
+      "loss": 0.5875,
+      "step": 7587
+    },
+    {
+      "epoch": 14.051851851851852,
+      "grad_norm": 1.7913263733850884,
+      "learning_rate": 1.505702648836811e-07,
+      "loss": 0.6424,
+      "step": 7588
+    },
+    {
+      "epoch": 14.053703703703704,
+      "grad_norm": 1.6768149640514025,
+      "learning_rate": 1.5055732866883361e-07,
+      "loss": 0.6334,
+      "step": 7589
+    },
+    {
+      "epoch": 14.055555555555555,
+      "grad_norm": 1.441874454108305,
+      "learning_rate": 1.5054439131732008e-07,
+      "loss": 0.5887,
+      "step": 7590
+    },
+    {
+      "epoch": 14.057407407407407,
+      "grad_norm": 1.473412943784779,
+      "learning_rate": 1.5053145282943135e-07,
+      "loss": 0.6247,
+      "step": 7591
+    },
+    {
+      "epoch": 14.059259259259258,
+      "grad_norm": 1.5871441973868543,
+      "learning_rate": 1.5051851320545833e-07,
+      "loss": 0.6483,
+      "step": 7592
+    },
+    {
+      "epoch": 14.061111111111112,
+      "grad_norm": 1.5657553313386878,
+      "learning_rate": 1.505055724456919e-07,
+      "loss": 0.6223,
+      "step": 7593
+    },
+    {
+      "epoch": 14.062962962962963,
+      "grad_norm": 1.6483418326868546,
+      "learning_rate": 1.50492630550423e-07,
+      "loss": 0.5793,
+      "step": 7594
+    },
+    {
+      "epoch": 14.064814814814815,
+      "grad_norm": 6.132313332796144,
+      "learning_rate": 1.504796875199427e-07,
+      "loss": 0.6096,
+      "step": 7595
+    },
+    {
+      "epoch": 14.066666666666666,
+      "grad_norm": 1.6499612497563398,
+      "learning_rate": 1.5046674335454188e-07,
+      "loss": 0.6386,
+      "step": 7596
+    },
+    {
+      "epoch": 14.068518518518518,
+      "grad_norm": 1.5551339861687337,
+      "learning_rate": 1.504537980545116e-07,
+      "loss": 0.5956,
+      "step": 7597
+    },
+    {
+      "epoch": 14.07037037037037,
+      "grad_norm": 1.7670669510990304,
+      "learning_rate": 1.5044085162014295e-07,
+      "loss": 0.6107,
+      "step": 7598
+    },
+    {
+      "epoch": 14.072222222222223,
+      "grad_norm": 1.7605663739524313,
+      "learning_rate": 1.5042790405172695e-07,
+      "loss": 0.602,
+      "step": 7599
+    },
+    {
+      "epoch": 14.074074074074074,
+      "grad_norm": 1.6594233043733004,
+      "learning_rate": 1.5041495534955465e-07,
+      "loss": 0.6264,
+      "step": 7600
+    },
+    {
+      "epoch": 14.075925925925926,
+      "grad_norm": 1.4956276920432126,
+      "learning_rate": 1.5040200551391727e-07,
+      "loss": 0.6471,
+      "step": 7601
+    },
+    {
+      "epoch": 14.077777777777778,
+      "grad_norm": 1.4932250653635204,
+      "learning_rate": 1.5038905454510596e-07,
+      "loss": 0.599,
+      "step": 7602
+    },
+    {
+      "epoch": 14.079629629629629,
+      "grad_norm": 1.7689607643767626,
+      "learning_rate": 1.5037610244341176e-07,
+      "loss": 0.5956,
+      "step": 7603
+    },
+    {
+      "epoch": 14.081481481481482,
+      "grad_norm": 1.436908409289378,
+      "learning_rate": 1.50363149209126e-07,
+      "loss": 0.6033,
+      "step": 7604
+    },
+    {
+      "epoch": 14.083333333333334,
+      "grad_norm": 1.8453985541699853,
+      "learning_rate": 1.503501948425399e-07,
+      "loss": 0.6214,
+      "step": 7605
+    },
+    {
+      "epoch": 14.085185185185185,
+      "grad_norm": 1.4393099057386045,
+      "learning_rate": 1.5033723934394464e-07,
+      "loss": 0.5947,
+      "step": 7606
+    },
+    {
+      "epoch": 14.087037037037037,
+      "grad_norm": 1.6488557182242323,
+      "learning_rate": 1.503242827136315e-07,
+      "loss": 0.6292,
+      "step": 7607
+    },
+    {
+      "epoch": 14.088888888888889,
+      "grad_norm": 1.7530432214019205,
+      "learning_rate": 1.5031132495189185e-07,
+      "loss": 0.6218,
+      "step": 7608
+    },
+    {
+      "epoch": 14.09074074074074,
+      "grad_norm": 2.5422581918690805,
+      "learning_rate": 1.502983660590169e-07,
+      "loss": 0.6567,
+      "step": 7609
+    },
+    {
+      "epoch": 14.092592592592593,
+      "grad_norm": 1.8493467810826807,
+      "learning_rate": 1.5028540603529815e-07,
+      "loss": 0.6185,
+      "step": 7610
+    },
+    {
+      "epoch": 14.094444444444445,
+      "grad_norm": 1.4943640633340753,
+      "learning_rate": 1.5027244488102687e-07,
+      "loss": 0.6216,
+      "step": 7611
+    },
+    {
+      "epoch": 14.096296296296297,
+      "grad_norm": 1.8210036901950928,
+      "learning_rate": 1.502594825964945e-07,
+      "loss": 0.5988,
+      "step": 7612
+    },
+    {
+      "epoch": 14.098148148148148,
+      "grad_norm": 1.555931715786282,
+      "learning_rate": 1.5024651918199242e-07,
+      "loss": 0.6004,
+      "step": 7613
+    },
+    {
+      "epoch": 14.1,
+      "grad_norm": 1.7247720858698312,
+      "learning_rate": 1.502335546378122e-07,
+      "loss": 0.6045,
+      "step": 7614
+    },
+    {
+      "epoch": 14.101851851851851,
+      "grad_norm": 1.4646727458798356,
+      "learning_rate": 1.5022058896424517e-07,
+      "loss": 0.621,
+      "step": 7615
+    },
+    {
+      "epoch": 14.103703703703705,
+      "grad_norm": 1.4953265761874894,
+      "learning_rate": 1.5020762216158294e-07,
+      "loss": 0.6064,
+      "step": 7616
+    },
+    {
+      "epoch": 14.105555555555556,
+      "grad_norm": 1.5197029256162708,
+      "learning_rate": 1.5019465423011698e-07,
+      "loss": 0.6288,
+      "step": 7617
+    },
+    {
+      "epoch": 14.107407407407408,
+      "grad_norm": 1.6198709381597964,
+      "learning_rate": 1.5018168517013886e-07,
+      "loss": 0.6414,
+      "step": 7618
+    },
+    {
+      "epoch": 14.10925925925926,
+      "grad_norm": 1.5377842106867432,
+      "learning_rate": 1.5016871498194017e-07,
+      "loss": 0.6465,
+      "step": 7619
+    },
+    {
+      "epoch": 14.11111111111111,
+      "grad_norm": 2.04785401176347,
+      "learning_rate": 1.5015574366581253e-07,
+      "loss": 0.6183,
+      "step": 7620
+    },
+    {
+      "epoch": 14.112962962962962,
+      "grad_norm": 1.5187260781190668,
+      "learning_rate": 1.5014277122204756e-07,
+      "loss": 0.6024,
+      "step": 7621
+    },
+    {
+      "epoch": 14.114814814814816,
+      "grad_norm": 1.7489821246163195,
+      "learning_rate": 1.501297976509369e-07,
+      "loss": 0.6031,
+      "step": 7622
+    },
+    {
+      "epoch": 14.116666666666667,
+      "grad_norm": 2.487700463693705,
+      "learning_rate": 1.501168229527722e-07,
+      "loss": 0.6167,
+      "step": 7623
+    },
+    {
+      "epoch": 14.118518518518519,
+      "grad_norm": 1.614565649548583,
+      "learning_rate": 1.5010384712784523e-07,
+      "loss": 0.6251,
+      "step": 7624
+    },
+    {
+      "epoch": 14.12037037037037,
+      "grad_norm": 1.6249839690387244,
+      "learning_rate": 1.500908701764477e-07,
+      "loss": 0.6033,
+      "step": 7625
+    },
+    {
+      "epoch": 14.122222222222222,
+      "grad_norm": 1.9488993266066592,
+      "learning_rate": 1.5007789209887136e-07,
+      "loss": 0.6536,
+      "step": 7626
+    },
+    {
+      "epoch": 14.124074074074073,
+      "grad_norm": 1.5632818709300507,
+      "learning_rate": 1.50064912895408e-07,
+      "loss": 0.5731,
+      "step": 7627
+    },
+    {
+      "epoch": 14.125925925925927,
+      "grad_norm": 1.5443205092160375,
+      "learning_rate": 1.500519325663494e-07,
+      "loss": 0.5885,
+      "step": 7628
+    },
+    {
+      "epoch": 14.127777777777778,
+      "grad_norm": 1.9428723937062082,
+      "learning_rate": 1.500389511119875e-07,
+      "loss": 0.5547,
+      "step": 7629
+    },
+    {
+      "epoch": 14.12962962962963,
+      "grad_norm": 1.520061206649567,
+      "learning_rate": 1.5002596853261399e-07,
+      "loss": 0.5838,
+      "step": 7630
+    },
+    {
+      "epoch": 14.131481481481481,
+      "grad_norm": 1.7347340981983514,
+      "learning_rate": 1.5001298482852086e-07,
+      "loss": 0.5738,
+      "step": 7631
+    },
+    {
+      "epoch": 14.133333333333333,
+      "grad_norm": 3.684925826045263,
+      "learning_rate": 1.5e-07,
+      "loss": 0.6324,
+      "step": 7632
+    },
+    {
+      "epoch": 14.135185185185184,
+      "grad_norm": 2.434272090130225,
+      "learning_rate": 1.4998701404734332e-07,
+      "loss": 0.5837,
+      "step": 7633
+    },
+    {
+      "epoch": 14.137037037037038,
+      "grad_norm": 1.492139275235571,
+      "learning_rate": 1.4997402697084286e-07,
+      "loss": 0.5885,
+      "step": 7634
+    },
+    {
+      "epoch": 14.13888888888889,
+      "grad_norm": 1.5201663881417409,
+      "learning_rate": 1.499610387707905e-07,
+      "loss": 0.6163,
+      "step": 7635
+    },
+    {
+      "epoch": 14.14074074074074,
+      "grad_norm": 1.3930291664301588,
+      "learning_rate": 1.499480494474783e-07,
+      "loss": 0.5987,
+      "step": 7636
+    },
+    {
+      "epoch": 14.142592592592592,
+      "grad_norm": 1.7987670500985442,
+      "learning_rate": 1.4993505900119831e-07,
+      "loss": 0.6126,
+      "step": 7637
+    },
+    {
+      "epoch": 14.144444444444444,
+      "grad_norm": 1.597447549399388,
+      "learning_rate": 1.4992206743224257e-07,
+      "loss": 0.5895,
+      "step": 7638
+    },
+    {
+      "epoch": 14.146296296296295,
+      "grad_norm": 1.5927572520024187,
+      "learning_rate": 1.4990907474090313e-07,
+      "loss": 0.6414,
+      "step": 7639
+    },
+    {
+      "epoch": 14.148148148148149,
+      "grad_norm": 1.5592852943566924,
+      "learning_rate": 1.4989608092747218e-07,
+      "loss": 0.6204,
+      "step": 7640
+    },
+    {
+      "epoch": 14.15,
+      "grad_norm": 2.1378556212453197,
+      "learning_rate": 1.498830859922418e-07,
+      "loss": 0.5974,
+      "step": 7641
+    },
+    {
+      "epoch": 14.151851851851852,
+      "grad_norm": 1.9946481489879193,
+      "learning_rate": 1.4987008993550417e-07,
+      "loss": 0.6215,
+      "step": 7642
+    },
+    {
+      "epoch": 14.153703703703703,
+      "grad_norm": 1.6028189793660668,
+      "learning_rate": 1.4985709275755147e-07,
+      "loss": 0.6377,
+      "step": 7643
+    },
+    {
+      "epoch": 14.155555555555555,
+      "grad_norm": 1.4314696429646245,
+      "learning_rate": 1.4984409445867594e-07,
+      "loss": 0.6332,
+      "step": 7644
+    },
+    {
+      "epoch": 14.157407407407407,
+      "grad_norm": 1.561400966355092,
+      "learning_rate": 1.4983109503916976e-07,
+      "loss": 0.6147,
+      "step": 7645
+    },
+    {
+      "epoch": 14.15925925925926,
+      "grad_norm": 1.6507683150564743,
+      "learning_rate": 1.4981809449932523e-07,
+      "loss": 0.6264,
+      "step": 7646
+    },
+    {
+      "epoch": 14.161111111111111,
+      "grad_norm": 1.515116871958776,
+      "learning_rate": 1.4980509283943464e-07,
+      "loss": 0.6147,
+      "step": 7647
+    },
+    {
+      "epoch": 14.162962962962963,
+      "grad_norm": 1.6215723547405965,
+      "learning_rate": 1.4979209005979029e-07,
+      "loss": 0.6058,
+      "step": 7648
+    },
+    {
+      "epoch": 14.164814814814815,
+      "grad_norm": 1.733154360280619,
+      "learning_rate": 1.497790861606845e-07,
+      "loss": 0.6051,
+      "step": 7649
+    },
+    {
+      "epoch": 14.166666666666666,
+      "grad_norm": 1.4438849647600247,
+      "learning_rate": 1.4976608114240969e-07,
+      "loss": 0.6082,
+      "step": 7650
+    },
+    {
+      "epoch": 14.168518518518518,
+      "grad_norm": 1.7901110483185685,
+      "learning_rate": 1.497530750052582e-07,
+      "loss": 0.6275,
+      "step": 7651
+    },
+    {
+      "epoch": 14.170370370370371,
+      "grad_norm": 1.5358185602227856,
+      "learning_rate": 1.4974006774952246e-07,
+      "loss": 0.6168,
+      "step": 7652
+    },
+    {
+      "epoch": 14.172222222222222,
+      "grad_norm": 1.8131998431280767,
+      "learning_rate": 1.497270593754949e-07,
+      "loss": 0.6791,
+      "step": 7653
+    },
+    {
+      "epoch": 14.174074074074074,
+      "grad_norm": 1.7169095142064215,
+      "learning_rate": 1.49714049883468e-07,
+      "loss": 0.6135,
+      "step": 7654
+    },
+    {
+      "epoch": 14.175925925925926,
+      "grad_norm": 1.4135901127366672,
+      "learning_rate": 1.4970103927373422e-07,
+      "loss": 0.5671,
+      "step": 7655
+    },
+    {
+      "epoch": 14.177777777777777,
+      "grad_norm": 1.756800089014376,
+      "learning_rate": 1.4968802754658613e-07,
+      "loss": 0.6248,
+      "step": 7656
+    },
+    {
+      "epoch": 14.17962962962963,
+      "grad_norm": 1.5923061937883074,
+      "learning_rate": 1.4967501470231617e-07,
+      "loss": 0.6283,
+      "step": 7657
+    },
+    {
+      "epoch": 14.181481481481482,
+      "grad_norm": 2.833276024868875,
+      "learning_rate": 1.4966200074121697e-07,
+      "loss": 0.5987,
+      "step": 7658
+    },
+    {
+      "epoch": 14.183333333333334,
+      "grad_norm": 1.7511325232950894,
+      "learning_rate": 1.4964898566358114e-07,
+      "loss": 0.5746,
+      "step": 7659
+    },
+    {
+      "epoch": 14.185185185185185,
+      "grad_norm": 1.4061610290079498,
+      "learning_rate": 1.4963596946970126e-07,
+      "loss": 0.6166,
+      "step": 7660
+    },
+    {
+      "epoch": 14.187037037037037,
+      "grad_norm": 1.8981733129510918,
+      "learning_rate": 1.4962295215987003e-07,
+      "loss": 0.6297,
+      "step": 7661
+    },
+    {
+      "epoch": 14.188888888888888,
+      "grad_norm": 1.6293501022419705,
+      "learning_rate": 1.4960993373438e-07,
+      "loss": 0.5681,
+      "step": 7662
+    },
+    {
+      "epoch": 14.190740740740742,
+      "grad_norm": 1.8347079229215582,
+      "learning_rate": 1.495969141935239e-07,
+      "loss": 0.6091,
+      "step": 7663
+    },
+    {
+      "epoch": 14.192592592592593,
+      "grad_norm": 1.5014665760350658,
+      "learning_rate": 1.495838935375945e-07,
+      "loss": 0.6041,
+      "step": 7664
+    },
+    {
+      "epoch": 14.194444444444445,
+      "grad_norm": 1.8858033223748825,
+      "learning_rate": 1.495708717668845e-07,
+      "loss": 0.6614,
+      "step": 7665
+    },
+    {
+      "epoch": 14.196296296296296,
+      "grad_norm": 6.74372532141603,
+      "learning_rate": 1.495578488816867e-07,
+      "loss": 0.6533,
+      "step": 7666
+    },
+    {
+      "epoch": 14.198148148148148,
+      "grad_norm": 1.409799488030474,
+      "learning_rate": 1.4954482488229385e-07,
+      "loss": 0.5977,
+      "step": 7667
+    },
+    {
+      "epoch": 14.2,
+      "grad_norm": 2.085750659052171,
+      "learning_rate": 1.4953179976899877e-07,
+      "loss": 0.6156,
+      "step": 7668
+    },
+    {
+      "epoch": 14.201851851851853,
+      "grad_norm": 1.6764833276514248,
+      "learning_rate": 1.4951877354209428e-07,
+      "loss": 0.6273,
+      "step": 7669
+    },
+    {
+      "epoch": 14.203703703703704,
+      "grad_norm": 1.83934575450301,
+      "learning_rate": 1.4950574620187328e-07,
+      "loss": 0.5682,
+      "step": 7670
+    },
+    {
+      "epoch": 14.205555555555556,
+      "grad_norm": 1.4636627657960086,
+      "learning_rate": 1.4949271774862866e-07,
+      "loss": 0.6106,
+      "step": 7671
+    },
+    {
+      "epoch": 14.207407407407407,
+      "grad_norm": 1.8500699615407037,
+      "learning_rate": 1.494796881826533e-07,
+      "loss": 0.5924,
+      "step": 7672
+    },
+    {
+      "epoch": 14.209259259259259,
+      "grad_norm": 1.561324444249611,
+      "learning_rate": 1.494666575042402e-07,
+      "loss": 0.562,
+      "step": 7673
+    },
+    {
+      "epoch": 14.21111111111111,
+      "grad_norm": 1.4859167818717407,
+      "learning_rate": 1.4945362571368227e-07,
+      "loss": 0.62,
+      "step": 7674
+    },
+    {
+      "epoch": 14.212962962962964,
+      "grad_norm": 1.5950125881315613,
+      "learning_rate": 1.4944059281127254e-07,
+      "loss": 0.6099,
+      "step": 7675
+    },
+    {
+      "epoch": 14.214814814814815,
+      "grad_norm": 1.4518940386902717,
+      "learning_rate": 1.49427558797304e-07,
+      "loss": 0.6653,
+      "step": 7676
+    },
+    {
+      "epoch": 14.216666666666667,
+      "grad_norm": 1.8477784907743204,
+      "learning_rate": 1.4941452367206967e-07,
+      "loss": 0.5994,
+      "step": 7677
+    },
+    {
+      "epoch": 14.218518518518518,
+      "grad_norm": 1.4786489051890814,
+      "learning_rate": 1.4940148743586266e-07,
+      "loss": 0.6161,
+      "step": 7678
+    },
+    {
+      "epoch": 14.22037037037037,
+      "grad_norm": 1.7293487452444025,
+      "learning_rate": 1.4938845008897603e-07,
+      "loss": 0.6041,
+      "step": 7679
+    },
+    {
+      "epoch": 14.222222222222221,
+      "grad_norm": 1.5791788974353476,
+      "learning_rate": 1.493754116317029e-07,
+      "loss": 0.5896,
+      "step": 7680
+    },
+    {
+      "epoch": 14.224074074074075,
+      "grad_norm": 1.6171428519136037,
+      "learning_rate": 1.4936237206433645e-07,
+      "loss": 0.5953,
+      "step": 7681
+    },
+    {
+      "epoch": 14.225925925925926,
+      "grad_norm": 1.940828537753481,
+      "learning_rate": 1.493493313871698e-07,
+      "loss": 0.6346,
+      "step": 7682
+    },
+    {
+      "epoch": 14.227777777777778,
+      "grad_norm": 1.6527453259714904,
+      "learning_rate": 1.4933628960049612e-07,
+      "loss": 0.6111,
+      "step": 7683
+    },
+    {
+      "epoch": 14.22962962962963,
+      "grad_norm": 1.5722399279352006,
+      "learning_rate": 1.4932324670460866e-07,
+      "loss": 0.6197,
+      "step": 7684
+    },
+    {
+      "epoch": 14.231481481481481,
+      "grad_norm": 1.554248635050263,
+      "learning_rate": 1.4931020269980067e-07,
+      "loss": 0.5845,
+      "step": 7685
+    },
+    {
+      "epoch": 14.233333333333333,
+      "grad_norm": 1.5771759511127588,
+      "learning_rate": 1.492971575863654e-07,
+      "loss": 0.6477,
+      "step": 7686
+    },
+    {
+      "epoch": 14.235185185185186,
+      "grad_norm": 1.4677889257671481,
+      "learning_rate": 1.4928411136459608e-07,
+      "loss": 0.6339,
+      "step": 7687
+    },
+    {
+      "epoch": 14.237037037037037,
+      "grad_norm": 2.478884288984545,
+      "learning_rate": 1.4927106403478614e-07,
+      "loss": 0.5941,
+      "step": 7688
+    },
+    {
+      "epoch": 14.238888888888889,
+      "grad_norm": 2.1931546077170383,
+      "learning_rate": 1.4925801559722888e-07,
+      "loss": 0.6825,
+      "step": 7689
+    },
+    {
+      "epoch": 14.24074074074074,
+      "grad_norm": 1.6656675026051615,
+      "learning_rate": 1.4924496605221762e-07,
+      "loss": 0.5905,
+      "step": 7690
+    },
+    {
+      "epoch": 14.242592592592592,
+      "grad_norm": 1.4647836954121627,
+      "learning_rate": 1.4923191540004578e-07,
+      "loss": 0.6287,
+      "step": 7691
+    },
+    {
+      "epoch": 14.244444444444444,
+      "grad_norm": 1.5100096469472872,
+      "learning_rate": 1.4921886364100675e-07,
+      "loss": 0.6045,
+      "step": 7692
+    },
+    {
+      "epoch": 14.246296296296297,
+      "grad_norm": 1.7365110390741048,
+      "learning_rate": 1.4920581077539402e-07,
+      "loss": 0.6124,
+      "step": 7693
+    },
+    {
+      "epoch": 14.248148148148148,
+      "grad_norm": 1.5762062344180878,
+      "learning_rate": 1.49192756803501e-07,
+      "loss": 0.6256,
+      "step": 7694
+    },
+    {
+      "epoch": 14.25,
+      "grad_norm": 1.4857106470645267,
+      "learning_rate": 1.491797017256212e-07,
+      "loss": 0.6127,
+      "step": 7695
+    },
+    {
+      "epoch": 14.251851851851852,
+      "grad_norm": 1.5902289624051167,
+      "learning_rate": 1.4916664554204816e-07,
+      "loss": 0.5948,
+      "step": 7696
+    },
+    {
+      "epoch": 14.253703703703703,
+      "grad_norm": 2.7200043247983645,
+      "learning_rate": 1.491535882530754e-07,
+      "loss": 0.6158,
+      "step": 7697
+    },
+    {
+      "epoch": 14.255555555555556,
+      "grad_norm": 1.5649669819434802,
+      "learning_rate": 1.4914052985899645e-07,
+      "loss": 0.6054,
+      "step": 7698
+    },
+    {
+      "epoch": 14.257407407407408,
+      "grad_norm": 1.432775402247199,
+      "learning_rate": 1.4912747036010493e-07,
+      "loss": 0.6169,
+      "step": 7699
+    },
+    {
+      "epoch": 14.25925925925926,
+      "grad_norm": 1.4562983326941146,
+      "learning_rate": 1.4911440975669444e-07,
+      "loss": 0.6148,
+      "step": 7700
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 21600,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 40,
+  "save_steps": 50,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6700262394036224.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}