diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,11221 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.23539970133662894,
+  "eval_steps": 500,
+  "global_step": 8000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.865354537963867,
+      "learning_rate": 5e-06,
+      "loss": 14.1919,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.778242111206055,
+      "learning_rate": 1e-05,
+      "loss": 13.847,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.41965103149414,
+      "learning_rate": 1.5e-05,
+      "loss": 13.0987,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.182560443878174,
+      "learning_rate": 2e-05,
+      "loss": 12.1878,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.488456726074219,
+      "learning_rate": 2.5e-05,
+      "loss": 11.3594,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.141963005065918,
+      "learning_rate": 3e-05,
+      "loss": 10.709,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.098189353942871,
+      "learning_rate": 3.5e-05,
+      "loss": 10.0102,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.459712505340576,
+      "learning_rate": 4e-05,
+      "loss": 9.5047,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1383228302001953,
+      "learning_rate": 4.5e-05,
+      "loss": 9.0663,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7991732358932495,
+      "learning_rate": 5e-05,
+      "loss": 8.8198,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5260429382324219,
+      "learning_rate": 5.500000000000001e-05,
+      "loss": 8.7401,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0736186504364014,
+      "learning_rate": 6e-05,
+      "loss": 8.6149,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7068235278129578,
+      "learning_rate": 6.500000000000001e-05,
+      "loss": 8.6282,
+      "step": 65
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.85589998960495,
+      "learning_rate": 7e-05,
+      "loss": 8.6237,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.79108726978302,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 8.5789,
+      "step": 75
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9634334444999695,
+      "learning_rate": 8e-05,
+      "loss": 8.5194,
+      "step": 80
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0866438150405884,
+      "learning_rate": 8.5e-05,
+      "loss": 8.2837,
+      "step": 85
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4814759492874146,
+      "learning_rate": 9e-05,
+      "loss": 8.2037,
+      "step": 90
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5262881517410278,
+      "learning_rate": 9.5e-05,
+      "loss": 8.1493,
+      "step": 95
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.561881184577942,
+      "learning_rate": 0.0001,
+      "loss": 8.0516,
+      "step": 100
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.761247992515564,
+      "learning_rate": 9.999999462732701e-05,
+      "loss": 7.8303,
+      "step": 105
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3252863883972168,
+      "learning_rate": 9.999997850930916e-05,
+      "loss": 7.6997,
+      "step": 110
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0572723150253296,
+      "learning_rate": 9.999995164594991e-05,
+      "loss": 7.5416,
+      "step": 115
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3021457195281982,
+      "learning_rate": 9.999991403725507e-05,
+      "loss": 7.3206,
+      "step": 120
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7624614238739014,
+      "learning_rate": 9.99998656832327e-05,
+      "loss": 7.294,
+      "step": 125
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1716172695159912,
+      "learning_rate": 9.999980658389317e-05,
+      "loss": 7.3259,
+      "step": 130
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1100866794586182,
+      "learning_rate": 9.999973673924921e-05,
+      "loss": 7.1081,
+      "step": 135
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9766339659690857,
+      "learning_rate": 9.999965614931583e-05,
+      "loss": 7.0571,
+      "step": 140
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4021155834197998,
+      "learning_rate": 9.999956481411035e-05,
+      "loss": 6.9068,
+      "step": 145
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7319382429122925,
+      "learning_rate": 9.999946273365238e-05,
+      "loss": 6.9107,
+      "step": 150
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.308290481567383,
+      "learning_rate": 9.999934990796387e-05,
+      "loss": 6.9045,
+      "step": 155
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7409007549285889,
+      "learning_rate": 9.999922633706906e-05,
+      "loss": 6.8212,
+      "step": 160
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0711333751678467,
+      "learning_rate": 9.999909202099453e-05,
+      "loss": 6.609,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5754401683807373,
+      "learning_rate": 9.999894695976912e-05,
+      "loss": 6.641,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7120509147644043,
+      "learning_rate": 9.9998791153424e-05,
+      "loss": 6.5758,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5836427211761475,
+      "learning_rate": 9.999862460199267e-05,
+      "loss": 6.6846,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4431571960449219,
+      "learning_rate": 9.999844730551092e-05,
+      "loss": 6.552,
+      "step": 185
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.563740611076355,
+      "learning_rate": 9.999825926401686e-05,
+      "loss": 6.5245,
+      "step": 190
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1652429103851318,
+      "learning_rate": 9.999806047755088e-05,
+      "loss": 6.4741,
+      "step": 195
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5559320449829102,
+      "learning_rate": 9.999785094615573e-05,
+      "loss": 6.4826,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9728102684020996,
+      "learning_rate": 9.99976306698764e-05,
+      "loss": 6.4217,
+      "step": 205
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.218796730041504,
+      "learning_rate": 9.999739964876026e-05,
+      "loss": 6.335,
+      "step": 210
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1594576835632324,
+      "learning_rate": 9.999715788285695e-05,
+      "loss": 6.3192,
+      "step": 215
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.018867254257202,
+      "learning_rate": 9.999690537221844e-05,
+      "loss": 6.1887,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6369565725326538,
+      "learning_rate": 9.999664211689897e-05,
+      "loss": 6.2351,
+      "step": 225
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4256953001022339,
+      "learning_rate": 9.999636811695513e-05,
+      "loss": 6.2389,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4589308500289917,
+      "learning_rate": 9.99960833724458e-05,
+      "loss": 6.2416,
+      "step": 235
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5116280317306519,
+      "learning_rate": 9.999578788343218e-05,
+      "loss": 6.1381,
+      "step": 240
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.226797580718994,
+      "learning_rate": 9.999548164997778e-05,
+      "loss": 6.0452,
+      "step": 245
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7254366874694824,
+      "learning_rate": 9.999516467214839e-05,
+      "loss": 6.0741,
+      "step": 250
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6304724216461182,
+      "learning_rate": 9.999483695001214e-05,
+      "loss": 5.9646,
+      "step": 255
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5827312469482422,
+      "learning_rate": 9.999449848363946e-05,
+      "loss": 6.0637,
+      "step": 260
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1163673400878906,
+      "learning_rate": 9.999414927310309e-05,
+      "loss": 5.9827,
+      "step": 265
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.500626802444458,
+      "learning_rate": 9.99937893184781e-05,
+      "loss": 5.8313,
+      "step": 270
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.47915780544281,
+      "learning_rate": 9.99934186198418e-05,
+      "loss": 6.0491,
+      "step": 275
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7926409244537354,
+      "learning_rate": 9.999303717727389e-05,
+      "loss": 5.9035,
+      "step": 280
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.449931263923645,
+      "learning_rate": 9.999264499085633e-05,
+      "loss": 5.9786,
+      "step": 285
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9798824787139893,
+      "learning_rate": 9.999224206067341e-05,
+      "loss": 5.8911,
+      "step": 290
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.063843250274658,
+      "learning_rate": 9.999182838681171e-05,
+      "loss": 5.9272,
+      "step": 295
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1161229610443115,
+      "learning_rate": 9.999140396936016e-05,
+      "loss": 5.8374,
+      "step": 300
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.283904552459717,
+      "learning_rate": 9.999096880840994e-05,
+      "loss": 5.8794,
+      "step": 305
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.148697853088379,
+      "learning_rate": 9.999052290405459e-05,
+      "loss": 5.8126,
+      "step": 310
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7647688388824463,
+      "learning_rate": 9.999006625638994e-05,
+      "loss": 5.7574,
+      "step": 315
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9153484106063843,
+      "learning_rate": 9.998959886551411e-05,
+      "loss": 5.7272,
+      "step": 320
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8421961069107056,
+      "learning_rate": 9.998912073152755e-05,
+      "loss": 5.7124,
+      "step": 325
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0302577018737793,
+      "learning_rate": 9.998863185453302e-05,
+      "loss": 5.749,
+      "step": 330
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6643224954605103,
+      "learning_rate": 9.998813223463557e-05,
+      "loss": 5.7703,
+      "step": 335
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2592408657073975,
+      "learning_rate": 9.998762187194258e-05,
+      "loss": 5.6041,
+      "step": 340
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2765250205993652,
+      "learning_rate": 9.998710076656373e-05,
+      "loss": 5.6188,
+      "step": 345
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8423378467559814,
+      "learning_rate": 9.998656891861102e-05,
+      "loss": 5.7069,
+      "step": 350
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7426402568817139,
+      "learning_rate": 9.998602632819874e-05,
+      "loss": 5.6764,
+      "step": 355
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6326072216033936,
+      "learning_rate": 9.998547299544348e-05,
+      "loss": 5.6667,
+      "step": 360
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7141144275665283,
+      "learning_rate": 9.99849089204642e-05,
+      "loss": 5.575,
+      "step": 365
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5443716049194336,
+      "learning_rate": 9.998433410338206e-05,
+      "loss": 5.669,
+      "step": 370
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8691034317016602,
+      "learning_rate": 9.998374854432064e-05,
+      "loss": 5.659,
+      "step": 375
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5928624868392944,
+      "learning_rate": 9.998315224340577e-05,
+      "loss": 5.7325,
+      "step": 380
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9967966079711914,
+      "learning_rate": 9.99825452007656e-05,
+      "loss": 5.5303,
+      "step": 385
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6495732069015503,
+      "learning_rate": 9.998192741653057e-05,
+      "loss": 5.7298,
+      "step": 390
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8097381591796875,
+      "learning_rate": 9.998129889083346e-05,
+      "loss": 5.375,
+      "step": 395
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8526384830474854,
+      "learning_rate": 9.998065962380936e-05,
+      "loss": 5.4918,
+      "step": 400
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.775524616241455,
+      "learning_rate": 9.998000961559562e-05,
+      "loss": 5.4301,
+      "step": 405
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6495391130447388,
+      "learning_rate": 9.997934886633195e-05,
+      "loss": 5.4728,
+      "step": 410
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8126803636550903,
+      "learning_rate": 9.997867737616035e-05,
+      "loss": 5.534,
+      "step": 415
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7900750637054443,
+      "learning_rate": 9.997799514522513e-05,
+      "loss": 5.3895,
+      "step": 420
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9649373292922974,
+      "learning_rate": 9.99773021736729e-05,
+      "loss": 5.5427,
+      "step": 425
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.551987886428833,
+      "learning_rate": 9.99765984616526e-05,
+      "loss": 5.5865,
+      "step": 430
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5227116346359253,
+      "learning_rate": 9.997588400931542e-05,
+      "loss": 5.553,
+      "step": 435
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9641531705856323,
+      "learning_rate": 9.997515881681496e-05,
+      "loss": 5.4945,
+      "step": 440
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5408072471618652,
+      "learning_rate": 9.9974422884307e-05,
+      "loss": 5.5347,
+      "step": 445
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.405747413635254,
+      "learning_rate": 9.997367621194977e-05,
+      "loss": 5.4443,
+      "step": 450
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.597363829612732,
+      "learning_rate": 9.99729187999037e-05,
+      "loss": 5.3571,
+      "step": 455
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2353439331054688,
+      "learning_rate": 9.997215064833155e-05,
+      "loss": 5.562,
+      "step": 460
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9211082458496094,
+      "learning_rate": 9.99713717573984e-05,
+      "loss": 5.267,
+      "step": 465
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5198642015457153,
+      "learning_rate": 9.997058212727169e-05,
+      "loss": 5.3297,
+      "step": 470
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9213910102844238,
+      "learning_rate": 9.996978175812105e-05,
+      "loss": 5.3912,
+      "step": 475
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.290919065475464,
+      "learning_rate": 9.996897065011852e-05,
+      "loss": 5.3424,
+      "step": 480
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4496021270751953,
+      "learning_rate": 9.996814880343842e-05,
+      "loss": 5.4215,
+      "step": 485
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2969412803649902,
+      "learning_rate": 9.996731621825736e-05,
+      "loss": 5.4071,
+      "step": 490
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.938884973526001,
+      "learning_rate": 9.996647289475426e-05,
+      "loss": 5.4035,
+      "step": 495
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5944199562072754,
+      "learning_rate": 9.996561883311036e-05,
+      "loss": 5.3945,
+      "step": 500
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.525998592376709,
+      "learning_rate": 9.996475403350921e-05,
+      "loss": 5.232,
+      "step": 505
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5026615858078003,
+      "learning_rate": 9.996387849613667e-05,
+      "loss": 5.2223,
+      "step": 510
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4775668382644653,
+      "learning_rate": 9.996299222118086e-05,
+      "loss": 5.3493,
+      "step": 515
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.476988434791565,
+      "learning_rate": 9.99620952088323e-05,
+      "loss": 5.2578,
+      "step": 520
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4739878177642822,
+      "learning_rate": 9.996118745928373e-05,
+      "loss": 5.2386,
+      "step": 525
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5050435066223145,
+      "learning_rate": 9.996026897273024e-05,
+      "loss": 5.4264,
+      "step": 530
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8701190948486328,
+      "learning_rate": 9.995933974936922e-05,
+      "loss": 5.2351,
+      "step": 535
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3322926759719849,
+      "learning_rate": 9.995839978940036e-05,
+      "loss": 5.2161,
+      "step": 540
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7951281070709229,
+      "learning_rate": 9.995744909302567e-05,
+      "loss": 5.2909,
+      "step": 545
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.581093192100525,
+      "learning_rate": 9.995648766044947e-05,
+      "loss": 5.2707,
+      "step": 550
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.367189884185791,
+      "learning_rate": 9.995551549187836e-05,
+      "loss": 5.2156,
+      "step": 555
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.812924861907959,
+      "learning_rate": 9.995453258752127e-05,
+      "loss": 5.2069,
+      "step": 560
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8363821506500244,
+      "learning_rate": 9.995353894758945e-05,
+      "loss": 5.2574,
+      "step": 565
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6378493309020996,
+      "learning_rate": 9.995253457229643e-05,
+      "loss": 5.1155,
+      "step": 570
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7905001640319824,
+      "learning_rate": 9.995151946185804e-05,
+      "loss": 5.2042,
+      "step": 575
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.662427306175232,
+      "learning_rate": 9.995049361649247e-05,
+      "loss": 5.1488,
+      "step": 580
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8039860725402832,
+      "learning_rate": 9.994945703642013e-05,
+      "loss": 5.132,
+      "step": 585
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5987797975540161,
+      "learning_rate": 9.994840972186386e-05,
+      "loss": 5.1835,
+      "step": 590
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6902951002120972,
+      "learning_rate": 9.994735167304866e-05,
+      "loss": 5.1557,
+      "step": 595
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6339097023010254,
+      "learning_rate": 9.994628289020195e-05,
+      "loss": 5.2125,
+      "step": 600
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2942930459976196,
+      "learning_rate": 9.994520337355344e-05,
+      "loss": 5.1907,
+      "step": 605
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.281216025352478,
+      "learning_rate": 9.994411312333507e-05,
+      "loss": 5.2578,
+      "step": 610
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3871283531188965,
+      "learning_rate": 9.99430121397812e-05,
+      "loss": 5.2106,
+      "step": 615
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0334229469299316,
+      "learning_rate": 9.994190042312839e-05,
+      "loss": 5.1711,
+      "step": 620
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.686562180519104,
+      "learning_rate": 9.994077797361559e-05,
+      "loss": 5.1558,
+      "step": 625
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7922908067703247,
+      "learning_rate": 9.9939644791484e-05,
+      "loss": 5.1135,
+      "step": 630
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8174651861190796,
+      "learning_rate": 9.993850087697716e-05,
+      "loss": 5.0822,
+      "step": 635
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5060467720031738,
+      "learning_rate": 9.99373462303409e-05,
+      "loss": 5.1859,
+      "step": 640
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2765015363693237,
+      "learning_rate": 9.993618085182337e-05,
+      "loss": 5.1506,
+      "step": 645
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.809071660041809,
+      "learning_rate": 9.993500474167501e-05,
+      "loss": 5.3114,
+      "step": 650
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4875119924545288,
+      "learning_rate": 9.99338179001486e-05,
+      "loss": 5.0786,
+      "step": 655
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6791081428527832,
+      "learning_rate": 9.993262032749914e-05,
+      "loss": 5.1644,
+      "step": 660
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5744774341583252,
+      "learning_rate": 9.993141202398407e-05,
+      "loss": 4.9858,
+      "step": 665
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5015692710876465,
+      "learning_rate": 9.993019298986301e-05,
+      "loss": 5.0266,
+      "step": 670
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2777938842773438,
+      "learning_rate": 9.992896322539796e-05,
+      "loss": 5.0644,
+      "step": 675
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2695177793502808,
+      "learning_rate": 9.992772273085319e-05,
+      "loss": 5.064,
+      "step": 680
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3356993198394775,
+      "learning_rate": 9.99264715064953e-05,
+      "loss": 5.0866,
+      "step": 685
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.769729733467102,
+      "learning_rate": 9.992520955259319e-05,
+      "loss": 5.1338,
+      "step": 690
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6709634065628052,
+      "learning_rate": 9.992393686941808e-05,
+      "loss": 5.0015,
+      "step": 695
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5945712327957153,
+      "learning_rate": 9.992265345724344e-05,
+      "loss": 5.0277,
+      "step": 700
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3791865110397339,
+      "learning_rate": 9.99213593163451e-05,
+      "loss": 5.0546,
+      "step": 705
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5065523386001587,
+      "learning_rate": 9.992005444700121e-05,
+      "loss": 4.9627,
+      "step": 710
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3527815341949463,
+      "learning_rate": 9.991873884949214e-05,
+      "loss": 5.061,
+      "step": 715
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.495449185371399,
+      "learning_rate": 9.991741252410067e-05,
+      "loss": 5.0612,
+      "step": 720
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.798466682434082,
+      "learning_rate": 9.99160754711118e-05,
+      "loss": 4.8878,
+      "step": 725
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7019239664077759,
+      "learning_rate": 9.99147276908129e-05,
+      "loss": 5.0656,
+      "step": 730
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5076030492782593,
+      "learning_rate": 9.991336918349361e-05,
+      "loss": 5.0071,
+      "step": 735
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6452229022979736,
+      "learning_rate": 9.991199994944586e-05,
+      "loss": 5.0957,
+      "step": 740
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1944454908370972,
+      "learning_rate": 9.991061998896393e-05,
+      "loss": 5.0381,
+      "step": 745
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.357327938079834,
+      "learning_rate": 9.99092293023444e-05,
+      "loss": 4.9872,
+      "step": 750
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3705363273620605,
+      "learning_rate": 9.990782788988611e-05,
+      "loss": 5.048,
+      "step": 755
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4363281726837158,
+      "learning_rate": 9.990641575189022e-05,
+      "loss": 5.0795,
+      "step": 760
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4852482080459595,
+      "learning_rate": 9.990499288866024e-05,
+      "loss": 4.9506,
+      "step": 765
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.30362868309021,
+      "learning_rate": 9.990355930050194e-05,
+      "loss": 4.913,
+      "step": 770
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4449853897094727,
+      "learning_rate": 9.990211498772341e-05,
+      "loss": 4.9994,
+      "step": 775
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3332302570343018,
+      "learning_rate": 9.990065995063506e-05,
+      "loss": 5.0883,
+      "step": 780
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.280041217803955,
+      "learning_rate": 9.989919418954956e-05,
+      "loss": 4.8842,
+      "step": 785
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3570388555526733,
+      "learning_rate": 9.989771770478191e-05,
+      "loss": 4.9676,
+      "step": 790
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1844346523284912,
+      "learning_rate": 9.989623049664944e-05,
+      "loss": 4.9266,
+      "step": 795
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9854055643081665,
+      "learning_rate": 9.989473256547175e-05,
+      "loss": 4.9819,
+      "step": 800
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4344004392623901,
+      "learning_rate": 9.989322391157076e-05,
+      "loss": 5.0319,
+      "step": 805
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2796415090560913,
+      "learning_rate": 9.989170453527068e-05,
+      "loss": 4.8331,
+      "step": 810
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3244338035583496,
+      "learning_rate": 9.989017443689805e-05,
+      "loss": 4.9282,
+      "step": 815
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9755282402038574,
+      "learning_rate": 9.988863361678169e-05,
+      "loss": 5.038,
+      "step": 820
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3403288125991821,
+      "learning_rate": 9.988708207525272e-05,
+      "loss": 4.9611,
+      "step": 825
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7184112071990967,
+      "learning_rate": 9.98855198126446e-05,
+      "loss": 4.9326,
+      "step": 830
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3740100860595703,
+      "learning_rate": 9.988394682929305e-05,
+      "loss": 4.9765,
+      "step": 835
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5485782623291016,
+      "learning_rate": 9.988236312553614e-05,
+      "loss": 4.9228,
+      "step": 840
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4770698547363281,
+      "learning_rate": 9.988076870171419e-05,
+      "loss": 4.8909,
+      "step": 845
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1400439739227295,
+      "learning_rate": 9.987916355816988e-05,
+      "loss": 4.8558,
+      "step": 850
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3869225978851318,
+      "learning_rate": 9.987754769524814e-05,
+      "loss": 4.934,
+      "step": 855
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.389519214630127,
+      "learning_rate": 9.987592111329626e-05,
+      "loss": 4.8702,
+      "step": 860
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5260522365570068,
+      "learning_rate": 9.987428381266377e-05,
+      "loss": 4.9763,
+      "step": 865
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2891238927841187,
+      "learning_rate": 9.987263579370257e-05,
+      "loss": 4.8297,
+      "step": 870
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1068599224090576,
+      "learning_rate": 9.987097705676681e-05,
+      "loss": 4.9674,
+      "step": 875
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5305088758468628,
+      "learning_rate": 9.986930760221296e-05,
+      "loss": 4.8198,
+      "step": 880
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6124730110168457,
+      "learning_rate": 9.986762743039983e-05,
+      "loss": 4.8424,
+      "step": 885
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3003376722335815,
+      "learning_rate": 9.986593654168846e-05,
+      "loss": 4.8217,
+      "step": 890
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3515762090682983,
+      "learning_rate": 9.986423493644225e-05,
+      "loss": 4.8181,
+      "step": 895
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2062491178512573,
+      "learning_rate": 9.98625226150269e-05,
+      "loss": 4.8566,
+      "step": 900
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1586987972259521,
+      "learning_rate": 9.986079957781038e-05,
+      "loss": 4.8447,
+      "step": 905
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2822962999343872,
+      "learning_rate": 9.9859065825163e-05,
+      "loss": 4.774,
+      "step": 910
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2110077142715454,
+      "learning_rate": 9.985732135745733e-05,
+      "loss": 4.9038,
+      "step": 915
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2196264266967773,
+      "learning_rate": 9.985556617506828e-05,
+      "loss": 4.9109,
+      "step": 920
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2059587240219116,
+      "learning_rate": 9.985380027837308e-05,
+      "loss": 4.8844,
+      "step": 925
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.125739336013794,
+      "learning_rate": 9.98520236677512e-05,
+      "loss": 4.8504,
+      "step": 930
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0160213708877563,
+      "learning_rate": 9.985023634358443e-05,
+      "loss": 4.8104,
+      "step": 935
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.095167875289917,
+      "learning_rate": 9.984843830625691e-05,
+      "loss": 4.8427,
+      "step": 940
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.130576252937317,
+      "learning_rate": 9.984662955615507e-05,
+      "loss": 4.6818,
+      "step": 945
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2667874097824097,
+      "learning_rate": 9.984481009366758e-05,
+      "loss": 4.7782,
+      "step": 950
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2014540433883667,
+      "learning_rate": 9.984297991918548e-05,
+      "loss": 4.8951,
+      "step": 955
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.000736117362976,
+      "learning_rate": 9.984113903310206e-05,
+      "loss": 4.8708,
+      "step": 960
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.206780195236206,
+      "learning_rate": 9.983928743581296e-05,
+      "loss": 4.8771,
+      "step": 965
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3621070384979248,
+      "learning_rate": 9.983742512771611e-05,
+      "loss": 4.7518,
+      "step": 970
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5099612474441528,
+      "learning_rate": 9.983555210921171e-05,
+      "loss": 4.7208,
+      "step": 975
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2639201879501343,
+      "learning_rate": 9.983366838070231e-05,
+      "loss": 4.7603,
+      "step": 980
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2878518104553223,
+      "learning_rate": 9.983177394259271e-05,
+      "loss": 4.7245,
+      "step": 985
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2132678031921387,
+      "learning_rate": 9.982986879529006e-05,
+      "loss": 4.9052,
+      "step": 990
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0886242389678955,
+      "learning_rate": 9.982795293920377e-05,
+      "loss": 4.8411,
+      "step": 995
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1308908462524414,
+      "learning_rate": 9.982602637474558e-05,
+      "loss": 4.8148,
+      "step": 1000
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.006771445274353,
+      "learning_rate": 9.982408910232952e-05,
+      "loss": 4.7258,
+      "step": 1005
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5670878887176514,
+      "learning_rate": 9.982214112237193e-05,
+      "loss": 4.6582,
+      "step": 1010
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4064730405807495,
+      "learning_rate": 9.982018243529146e-05,
+      "loss": 4.7947,
+      "step": 1015
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2058125734329224,
+      "learning_rate": 9.981821304150901e-05,
+      "loss": 4.7599,
+      "step": 1020
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2823870182037354,
+      "learning_rate": 9.981623294144782e-05,
+      "loss": 4.7669,
+      "step": 1025
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3003207445144653,
+      "learning_rate": 9.981424213553344e-05,
+      "loss": 4.7797,
+      "step": 1030
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2630780935287476,
+      "learning_rate": 9.981224062419372e-05,
+      "loss": 4.7263,
+      "step": 1035
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3734545707702637,
+      "learning_rate": 9.981022840785877e-05,
+      "loss": 4.7885,
+      "step": 1040
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.25571608543396,
+      "learning_rate": 9.980820548696106e-05,
+      "loss": 4.697,
+      "step": 1045
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1999176740646362,
+      "learning_rate": 9.98061718619353e-05,
+      "loss": 4.7224,
+      "step": 1050
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2195305824279785,
+      "learning_rate": 9.980412753321856e-05,
+      "loss": 4.7741,
+      "step": 1055
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1094328165054321,
+      "learning_rate": 9.980207250125015e-05,
+      "loss": 4.7243,
+      "step": 1060
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3296178579330444,
+      "learning_rate": 9.980000676647174e-05,
+      "loss": 4.7365,
+      "step": 1065
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3158111572265625,
+      "learning_rate": 9.979793032932725e-05,
+      "loss": 4.7203,
+      "step": 1070
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.132800817489624,
+      "learning_rate": 9.979584319026292e-05,
+      "loss": 4.7656,
+      "step": 1075
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2001687288284302,
+      "learning_rate": 9.97937453497273e-05,
+      "loss": 4.8507,
+      "step": 1080
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1565091609954834,
+      "learning_rate": 9.979163680817124e-05,
+      "loss": 4.8588,
+      "step": 1085
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.149619460105896,
+      "learning_rate": 9.978951756604786e-05,
+      "loss": 4.8656,
+      "step": 1090
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1528968811035156,
+      "learning_rate": 9.978738762381262e-05,
+      "loss": 4.7607,
+      "step": 1095
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.009029507637024,
+      "learning_rate": 9.978524698192322e-05,
+      "loss": 4.7271,
+      "step": 1100
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9556276798248291,
+      "learning_rate": 9.978309564083975e-05,
+      "loss": 4.7656,
+      "step": 1105
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.081737756729126,
+      "learning_rate": 9.97809336010245e-05,
+      "loss": 4.6522,
+      "step": 1110
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5491801500320435,
+      "learning_rate": 9.977876086294216e-05,
+      "loss": 4.6337,
+      "step": 1115
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3452264070510864,
+      "learning_rate": 9.977657742705962e-05,
+      "loss": 4.7917,
+      "step": 1120
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2493197917938232,
+      "learning_rate": 9.977438329384614e-05,
+      "loss": 4.6985,
+      "step": 1125
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0601075887680054,
+      "learning_rate": 9.977217846377324e-05,
+      "loss": 4.7589,
+      "step": 1130
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2869666814804077,
+      "learning_rate": 9.976996293731476e-05,
+      "loss": 4.6809,
+      "step": 1135
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.186639666557312,
+      "learning_rate": 9.976773671494684e-05,
+      "loss": 4.7268,
+      "step": 1140
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.079332709312439,
+      "learning_rate": 9.976549979714791e-05,
+      "loss": 4.6859,
+      "step": 1145
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2257851362228394,
+      "learning_rate": 9.976325218439868e-05,
+      "loss": 4.6855,
+      "step": 1150
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3988691568374634,
+      "learning_rate": 9.976099387718218e-05,
+      "loss": 4.7485,
+      "step": 1155
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2305885553359985,
+      "learning_rate": 9.975872487598377e-05,
+      "loss": 4.7128,
+      "step": 1160
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0361040830612183,
+      "learning_rate": 9.975644518129104e-05,
+      "loss": 4.7135,
+      "step": 1165
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1318379640579224,
+      "learning_rate": 9.975415479359394e-05,
+      "loss": 4.7671,
+      "step": 1170
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1415613889694214,
+      "learning_rate": 9.975185371338464e-05,
+      "loss": 4.6467,
+      "step": 1175
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2138582468032837,
+      "learning_rate": 9.97495419411577e-05,
+      "loss": 4.6139,
+      "step": 1180
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1179404258728027,
+      "learning_rate": 9.974721947740995e-05,
+      "loss": 4.632,
+      "step": 1185
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0632065534591675,
+      "learning_rate": 9.974488632264046e-05,
+      "loss": 4.6807,
+      "step": 1190
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.13469660282135,
+      "learning_rate": 9.974254247735066e-05,
+      "loss": 4.5978,
+      "step": 1195
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.299612283706665,
+      "learning_rate": 9.974018794204426e-05,
+      "loss": 4.6398,
+      "step": 1200
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.02610445022583,
+      "learning_rate": 9.973782271722727e-05,
+      "loss": 4.6127,
+      "step": 1205
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1934685707092285,
+      "learning_rate": 9.973544680340799e-05,
+      "loss": 4.7208,
+      "step": 1210
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2531120777130127,
+      "learning_rate": 9.973306020109701e-05,
+      "loss": 4.6406,
+      "step": 1215
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.714058518409729,
+      "learning_rate": 9.973066291080724e-05,
+      "loss": 4.5314,
+      "step": 1220
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2945247888565063,
+      "learning_rate": 9.972825493305387e-05,
+      "loss": 4.6712,
+      "step": 1225
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1765650510787964,
+      "learning_rate": 9.97258362683544e-05,
+      "loss": 4.5582,
+      "step": 1230
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0680540800094604,
+      "learning_rate": 9.972340691722858e-05,
+      "loss": 4.6679,
+      "step": 1235
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1062068939208984,
+      "learning_rate": 9.972096688019856e-05,
+      "loss": 4.6321,
+      "step": 1240
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.103111743927002,
+      "learning_rate": 9.971851615778868e-05,
+      "loss": 4.5958,
+      "step": 1245
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2542630434036255,
+      "learning_rate": 9.971605475052561e-05,
+      "loss": 4.6106,
+      "step": 1250
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2209641933441162,
+      "learning_rate": 9.971358265893834e-05,
+      "loss": 4.6131,
+      "step": 1255
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.09658944606781,
+      "learning_rate": 9.971109988355814e-05,
+      "loss": 4.6178,
+      "step": 1260
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0809062719345093,
+      "learning_rate": 9.970860642491854e-05,
+      "loss": 4.6051,
+      "step": 1265
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9511008858680725,
+      "learning_rate": 9.970610228355548e-05,
+      "loss": 4.7189,
+      "step": 1270
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0091472864151,
+      "learning_rate": 9.970358746000704e-05,
+      "loss": 4.6165,
+      "step": 1275
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9107927680015564,
+      "learning_rate": 9.97010619548137e-05,
+      "loss": 4.5909,
+      "step": 1280
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9552775621414185,
+      "learning_rate": 9.969852576851822e-05,
+      "loss": 4.5503,
+      "step": 1285
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0112459659576416,
+      "learning_rate": 9.969597890166565e-05,
+      "loss": 4.6597,
+      "step": 1290
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1482925415039062,
+      "learning_rate": 9.969342135480328e-05,
+      "loss": 4.5875,
+      "step": 1295
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0411022901535034,
+      "learning_rate": 9.969085312848081e-05,
+      "loss": 4.5594,
+      "step": 1300
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1362754106521606,
+      "learning_rate": 9.968827422325012e-05,
+      "loss": 4.5262,
+      "step": 1305
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2394375801086426,
+      "learning_rate": 9.968568463966545e-05,
+      "loss": 4.6496,
+      "step": 1310
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2385950088500977,
+      "learning_rate": 9.968308437828333e-05,
+      "loss": 4.6223,
+      "step": 1315
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.557721495628357,
+      "learning_rate": 9.968047343966256e-05,
+      "loss": 4.485,
+      "step": 1320
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1618167161941528,
+      "learning_rate": 9.967785182436426e-05,
+      "loss": 4.4463,
+      "step": 1325
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0881215333938599,
+      "learning_rate": 9.967521953295185e-05,
+      "loss": 4.4637,
+      "step": 1330
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.930715024471283,
+      "learning_rate": 9.967257656599099e-05,
+      "loss": 4.5096,
+      "step": 1335
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9364334940910339,
+      "learning_rate": 9.966992292404967e-05,
+      "loss": 4.499,
+      "step": 1340
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.918265163898468,
+      "learning_rate": 9.966725860769821e-05,
+      "loss": 4.535,
+      "step": 1345
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9972043633460999,
+      "learning_rate": 9.966458361750916e-05,
+      "loss": 4.5577,
+      "step": 1350
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.12447190284729,
+      "learning_rate": 9.966189795405743e-05,
+      "loss": 4.614,
+      "step": 1355
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0205003023147583,
+      "learning_rate": 9.965920161792016e-05,
+      "loss": 4.5432,
+      "step": 1360
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0749421119689941,
+      "learning_rate": 9.96564946096768e-05,
+      "loss": 4.5448,
+      "step": 1365
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9702705144882202,
+      "learning_rate": 9.965377692990914e-05,
+      "loss": 4.6212,
+      "step": 1370
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.318506121635437,
+      "learning_rate": 9.965104857920121e-05,
+      "loss": 4.5245,
+      "step": 1375
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.112592101097107,
+      "learning_rate": 9.964830955813935e-05,
+      "loss": 4.5737,
+      "step": 1380
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.923689067363739,
+      "learning_rate": 9.964555986731219e-05,
+      "loss": 4.4915,
+      "step": 1385
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0956313610076904,
+      "learning_rate": 9.964279950731066e-05,
+      "loss": 4.5648,
+      "step": 1390
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1847922801971436,
+      "learning_rate": 9.9640028478728e-05,
+      "loss": 4.4671,
+      "step": 1395
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1669468879699707,
+      "learning_rate": 9.963724678215968e-05,
+      "loss": 4.538,
+      "step": 1400
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2183971405029297,
+      "learning_rate": 9.963445441820356e-05,
+      "loss": 4.4448,
+      "step": 1405
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9898837208747864,
+      "learning_rate": 9.963165138745969e-05,
+      "loss": 4.4904,
+      "step": 1410
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0325148105621338,
+      "learning_rate": 9.962883769053051e-05,
+      "loss": 4.4781,
+      "step": 1415
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0582112073898315,
+      "learning_rate": 9.962601332802065e-05,
+      "loss": 4.4382,
+      "step": 1420
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8887158632278442,
+      "learning_rate": 9.962317830053711e-05,
+      "loss": 4.622,
+      "step": 1425
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0774013996124268,
+      "learning_rate": 9.962033260868916e-05,
+      "loss": 4.6193,
+      "step": 1430
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0495344400405884,
+      "learning_rate": 9.961747625308838e-05,
+      "loss": 4.4636,
+      "step": 1435
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1240602731704712,
+      "learning_rate": 9.961460923434858e-05,
+      "loss": 4.5973,
+      "step": 1440
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0496684312820435,
+      "learning_rate": 9.961173155308591e-05,
+      "loss": 4.4211,
+      "step": 1445
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2532814741134644,
+      "learning_rate": 9.960884320991881e-05,
+      "loss": 4.4486,
+      "step": 1450
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2570927143096924,
+      "learning_rate": 9.960594420546802e-05,
+      "loss": 4.4567,
+      "step": 1455
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0263696908950806,
+      "learning_rate": 9.960303454035654e-05,
+      "loss": 4.5603,
+      "step": 1460
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0136160850524902,
+      "learning_rate": 9.960011421520969e-05,
+      "loss": 4.4888,
+      "step": 1465
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8563377261161804,
+      "learning_rate": 9.959718323065503e-05,
+      "loss": 4.455,
+      "step": 1470
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0197609663009644,
+      "learning_rate": 9.959424158732252e-05,
+      "loss": 4.4619,
+      "step": 1475
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.040071964263916,
+      "learning_rate": 9.959128928584426e-05,
+      "loss": 4.5302,
+      "step": 1480
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1960361003875732,
+      "learning_rate": 9.958832632685478e-05,
+      "loss": 4.596,
+      "step": 1485
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0609952211380005,
+      "learning_rate": 9.95853527109908e-05,
+      "loss": 4.4877,
+      "step": 1490
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9883006811141968,
+      "learning_rate": 9.95823684388914e-05,
+      "loss": 4.5208,
+      "step": 1495
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.15782630443573,
+      "learning_rate": 9.95793735111979e-05,
+      "loss": 4.5272,
+      "step": 1500
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.001089096069336,
+      "learning_rate": 9.957636792855394e-05,
+      "loss": 4.5009,
+      "step": 1505
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9634086489677429,
+      "learning_rate": 9.957335169160545e-05,
+      "loss": 4.4945,
+      "step": 1510
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9154871106147766,
+      "learning_rate": 9.957032480100061e-05,
+      "loss": 4.4185,
+      "step": 1515
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.987657368183136,
+      "learning_rate": 9.956728725738996e-05,
+      "loss": 4.4406,
+      "step": 1520
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1099871397018433,
+      "learning_rate": 9.956423906142626e-05,
+      "loss": 4.4614,
+      "step": 1525
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0751529932022095,
+      "learning_rate": 9.956118021376461e-05,
+      "loss": 4.4089,
+      "step": 1530
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9815667271614075,
+      "learning_rate": 9.955811071506236e-05,
+      "loss": 4.5563,
+      "step": 1535
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9260372519493103,
+      "learning_rate": 9.955503056597917e-05,
+      "loss": 4.3594,
+      "step": 1540
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0457096099853516,
+      "learning_rate": 9.955193976717698e-05,
+      "loss": 4.5076,
+      "step": 1545
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0650675296783447,
+      "learning_rate": 9.954883831932003e-05,
+      "loss": 4.5292,
+      "step": 1550
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9985018968582153,
+      "learning_rate": 9.954572622307485e-05,
+      "loss": 4.4933,
+      "step": 1555
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9503152370452881,
+      "learning_rate": 9.954260347911025e-05,
+      "loss": 4.4746,
+      "step": 1560
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.061326026916504,
+      "learning_rate": 9.953947008809731e-05,
+      "loss": 4.4767,
+      "step": 1565
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8622316122055054,
+      "learning_rate": 9.953632605070944e-05,
+      "loss": 4.4553,
+      "step": 1570
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9125860333442688,
+      "learning_rate": 9.95331713676223e-05,
+      "loss": 4.4784,
+      "step": 1575
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0776078701019287,
+      "learning_rate": 9.953000603951386e-05,
+      "loss": 4.3819,
+      "step": 1580
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1412805318832397,
+      "learning_rate": 9.952683006706438e-05,
+      "loss": 4.3795,
+      "step": 1585
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0255082845687866,
+      "learning_rate": 9.952364345095638e-05,
+      "loss": 4.4874,
+      "step": 1590
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.98856121301651,
+      "learning_rate": 9.95204461918747e-05,
+      "loss": 4.4316,
+      "step": 1595
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9140748381614685,
+      "learning_rate": 9.951723829050646e-05,
+      "loss": 4.4147,
+      "step": 1600
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0750501155853271,
+      "learning_rate": 9.951401974754101e-05,
+      "loss": 4.3923,
+      "step": 1605
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.15281343460083,
+      "learning_rate": 9.951079056367013e-05,
+      "loss": 4.4133,
+      "step": 1610
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9914402961730957,
+      "learning_rate": 9.95075507395877e-05,
+      "loss": 4.291,
+      "step": 1615
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9306018352508545,
+      "learning_rate": 9.950430027599003e-05,
+      "loss": 4.4024,
+      "step": 1620
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0503841638565063,
+      "learning_rate": 9.950103917357566e-05,
+      "loss": 4.4192,
+      "step": 1625
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8728650212287903,
+      "learning_rate": 9.949776743304542e-05,
+      "loss": 4.3868,
+      "step": 1630
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.179078221321106,
+      "learning_rate": 9.949448505510243e-05,
+      "loss": 4.3624,
+      "step": 1635
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9564917683601379,
+      "learning_rate": 9.94911920404521e-05,
+      "loss": 4.4932,
+      "step": 1640
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8807275891304016,
+      "learning_rate": 9.94878883898021e-05,
+      "loss": 4.3196,
+      "step": 1645
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.178863286972046,
+      "learning_rate": 9.948457410386244e-05,
+      "loss": 4.3857,
+      "step": 1650
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8887953162193298,
+      "learning_rate": 9.948124918334538e-05,
+      "loss": 4.4577,
+      "step": 1655
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9354264140129089,
+      "learning_rate": 9.947791362896543e-05,
+      "loss": 4.4074,
+      "step": 1660
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9689270853996277,
+      "learning_rate": 9.947456744143947e-05,
+      "loss": 4.4014,
+      "step": 1665
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9946916699409485,
+      "learning_rate": 9.947121062148658e-05,
+      "loss": 4.4285,
+      "step": 1670
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8885537981987,
+      "learning_rate": 9.94678431698282e-05,
+      "loss": 4.393,
+      "step": 1675
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8793570399284363,
+      "learning_rate": 9.946446508718801e-05,
+      "loss": 4.3626,
+      "step": 1680
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5086390972137451,
+      "learning_rate": 9.946107637429196e-05,
+      "loss": 4.3653,
+      "step": 1685
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8934321403503418,
+      "learning_rate": 9.945767703186832e-05,
+      "loss": 4.3879,
+      "step": 1690
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8884941935539246,
+      "learning_rate": 9.945426706064764e-05,
+      "loss": 4.453,
+      "step": 1695
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8590004444122314,
+      "learning_rate": 9.945084646136275e-05,
+      "loss": 4.3489,
+      "step": 1700
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9937050342559814,
+      "learning_rate": 9.944741523474874e-05,
+      "loss": 4.2102,
+      "step": 1705
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.979364275932312,
+      "learning_rate": 9.944397338154301e-05,
+      "loss": 4.3404,
+      "step": 1710
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.978954017162323,
+      "learning_rate": 9.944052090248526e-05,
+      "loss": 4.3546,
+      "step": 1715
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8627170324325562,
+      "learning_rate": 9.943705779831742e-05,
+      "loss": 4.3576,
+      "step": 1720
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9601417779922485,
+      "learning_rate": 9.943358406978377e-05,
+      "loss": 4.4126,
+      "step": 1725
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.966124415397644,
+      "learning_rate": 9.943009971763082e-05,
+      "loss": 4.3485,
+      "step": 1730
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0642715692520142,
+      "learning_rate": 9.942660474260737e-05,
+      "loss": 4.3064,
+      "step": 1735
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7946628332138062,
+      "learning_rate": 9.942309914546453e-05,
+      "loss": 4.3194,
+      "step": 1740
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9242978096008301,
+      "learning_rate": 9.941958292695565e-05,
+      "loss": 4.291,
+      "step": 1745
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9501816034317017,
+      "learning_rate": 9.941605608783644e-05,
+      "loss": 4.337,
+      "step": 1750
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9336662888526917,
+      "learning_rate": 9.94125186288648e-05,
+      "loss": 4.3623,
+      "step": 1755
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.080331563949585,
+      "learning_rate": 9.940897055080098e-05,
+      "loss": 4.4114,
+      "step": 1760
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8232134580612183,
+      "learning_rate": 9.940541185440746e-05,
+      "loss": 4.4185,
+      "step": 1765
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9650618433952332,
+      "learning_rate": 9.940184254044904e-05,
+      "loss": 4.3134,
+      "step": 1770
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8358234167098999,
+      "learning_rate": 9.939826260969282e-05,
+      "loss": 4.511,
+      "step": 1775
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9228266477584839,
+      "learning_rate": 9.93946720629081e-05,
+      "loss": 4.3879,
+      "step": 1780
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8005467057228088,
+      "learning_rate": 9.939107090086654e-05,
+      "loss": 4.3049,
+      "step": 1785
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8405125737190247,
+      "learning_rate": 9.938745912434208e-05,
+      "loss": 4.3667,
+      "step": 1790
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8827022314071655,
+      "learning_rate": 9.938383673411087e-05,
+      "loss": 4.265,
+      "step": 1795
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9424577355384827,
+      "learning_rate": 9.938020373095141e-05,
+      "loss": 4.2377,
+      "step": 1800
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.091784954071045,
+      "learning_rate": 9.937656011564445e-05,
+      "loss": 4.4214,
+      "step": 1805
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9332832098007202,
+      "learning_rate": 9.937290588897302e-05,
+      "loss": 4.3694,
+      "step": 1810
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9324964880943298,
+      "learning_rate": 9.936924105172248e-05,
+      "loss": 4.2452,
+      "step": 1815
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.017594337463379,
+      "learning_rate": 9.936556560468037e-05,
+      "loss": 4.3914,
+      "step": 1820
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0270326137542725,
+      "learning_rate": 9.93618795486366e-05,
+      "loss": 4.3716,
+      "step": 1825
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0003628730773926,
+      "learning_rate": 9.935818288438334e-05,
+      "loss": 4.3349,
+      "step": 1830
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9341906309127808,
+      "learning_rate": 9.935447561271501e-05,
+      "loss": 4.3086,
+      "step": 1835
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8961882591247559,
+      "learning_rate": 9.935075773442835e-05,
+      "loss": 4.3964,
+      "step": 1840
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9122427105903625,
+      "learning_rate": 9.934702925032232e-05,
+      "loss": 4.2922,
+      "step": 1845
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9396681785583496,
+      "learning_rate": 9.934329016119824e-05,
+      "loss": 4.3527,
+      "step": 1850
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9001750946044922,
+      "learning_rate": 9.933954046785963e-05,
+      "loss": 4.2763,
+      "step": 1855
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9995968341827393,
+      "learning_rate": 9.933578017111233e-05,
+      "loss": 4.3357,
+      "step": 1860
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2028285264968872,
+      "learning_rate": 9.933200927176449e-05,
+      "loss": 4.3968,
+      "step": 1865
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9057603478431702,
+      "learning_rate": 9.932822777062646e-05,
+      "loss": 4.3064,
+      "step": 1870
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9613719582557678,
+      "learning_rate": 9.932443566851092e-05,
+      "loss": 4.314,
+      "step": 1875
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0476748943328857,
+      "learning_rate": 9.932063296623284e-05,
+      "loss": 4.2914,
+      "step": 1880
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.001121163368225,
+      "learning_rate": 9.931681966460944e-05,
+      "loss": 4.4617,
+      "step": 1885
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9707443118095398,
+      "learning_rate": 9.93129957644602e-05,
+      "loss": 4.3538,
+      "step": 1890
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9188675880432129,
+      "learning_rate": 9.930916126660691e-05,
+      "loss": 4.4077,
+      "step": 1895
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8299434781074524,
+      "learning_rate": 9.930531617187367e-05,
+      "loss": 4.3521,
+      "step": 1900
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.959729790687561,
+      "learning_rate": 9.930146048108678e-05,
+      "loss": 4.3208,
+      "step": 1905
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9663805365562439,
+      "learning_rate": 9.929759419507487e-05,
+      "loss": 4.424,
+      "step": 1910
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8181895613670349,
+      "learning_rate": 9.92937173146688e-05,
+      "loss": 4.2491,
+      "step": 1915
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9376658797264099,
+      "learning_rate": 9.928982984070179e-05,
+      "loss": 4.3097,
+      "step": 1920
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8866323828697205,
+      "learning_rate": 9.928593177400925e-05,
+      "loss": 4.3246,
+      "step": 1925
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2720496654510498,
+      "learning_rate": 9.928202311542889e-05,
+      "loss": 4.3374,
+      "step": 1930
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9235008955001831,
+      "learning_rate": 9.927810386580075e-05,
+      "loss": 4.31,
+      "step": 1935
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9263464212417603,
+      "learning_rate": 9.927417402596707e-05,
+      "loss": 4.3573,
+      "step": 1940
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8662964105606079,
+      "learning_rate": 9.927023359677241e-05,
+      "loss": 4.2037,
+      "step": 1945
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8622426390647888,
+      "learning_rate": 9.92662825790636e-05,
+      "loss": 4.2122,
+      "step": 1950
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8145320415496826,
+      "learning_rate": 9.926232097368974e-05,
+      "loss": 4.3486,
+      "step": 1955
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9132172465324402,
+      "learning_rate": 9.925834878150219e-05,
+      "loss": 4.3809,
+      "step": 1960
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9075316786766052,
+      "learning_rate": 9.925436600335464e-05,
+      "loss": 4.2651,
+      "step": 1965
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8904309868812561,
+      "learning_rate": 9.925037264010298e-05,
+      "loss": 4.3309,
+      "step": 1970
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9027050137519836,
+      "learning_rate": 9.924636869260542e-05,
+      "loss": 4.3054,
+      "step": 1975
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9754094481468201,
+      "learning_rate": 9.924235416172243e-05,
+      "loss": 4.2863,
+      "step": 1980
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8844654560089111,
+      "learning_rate": 9.923832904831679e-05,
+      "loss": 4.2493,
+      "step": 1985
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9261740446090698,
+      "learning_rate": 9.923429335325349e-05,
+      "loss": 4.3031,
+      "step": 1990
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8355966806411743,
+      "learning_rate": 9.923024707739985e-05,
+      "loss": 4.3505,
+      "step": 1995
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9094793200492859,
+      "learning_rate": 9.922619022162545e-05,
+      "loss": 4.2701,
+      "step": 2000
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9037348031997681,
+      "learning_rate": 9.92221227868021e-05,
+      "loss": 4.3397,
+      "step": 2005
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8133493661880493,
+      "learning_rate": 9.921804477380394e-05,
+      "loss": 4.2946,
+      "step": 2010
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9184492230415344,
+      "learning_rate": 9.921395618350739e-05,
+      "loss": 4.2659,
+      "step": 2015
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8432704210281372,
+      "learning_rate": 9.920985701679106e-05,
+      "loss": 4.3791,
+      "step": 2020
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.966478705406189,
+      "learning_rate": 9.920574727453595e-05,
+      "loss": 4.426,
+      "step": 2025
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8585464358329773,
+      "learning_rate": 9.920162695762522e-05,
+      "loss": 4.3192,
+      "step": 2030
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8447548747062683,
+      "learning_rate": 9.919749606694439e-05,
+      "loss": 4.2775,
+      "step": 2035
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.019842505455017,
+      "learning_rate": 9.919335460338122e-05,
+      "loss": 4.1512,
+      "step": 2040
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8868736028671265,
+      "learning_rate": 9.918920256782572e-05,
+      "loss": 4.1925,
+      "step": 2045
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0602483749389648,
+      "learning_rate": 9.918503996117016e-05,
+      "loss": 4.3372,
+      "step": 2050
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9156107902526855,
+      "learning_rate": 9.918086678430917e-05,
+      "loss": 4.2662,
+      "step": 2055
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9607118964195251,
+      "learning_rate": 9.917668303813959e-05,
+      "loss": 4.1842,
+      "step": 2060
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8450588583946228,
+      "learning_rate": 9.917248872356051e-05,
+      "loss": 4.2657,
+      "step": 2065
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7863975763320923,
+      "learning_rate": 9.916828384147331e-05,
+      "loss": 4.2764,
+      "step": 2070
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.796524167060852,
+      "learning_rate": 9.916406839278167e-05,
+      "loss": 4.2972,
+      "step": 2075
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8690253496170044,
+      "learning_rate": 9.915984237839152e-05,
+      "loss": 4.2397,
+      "step": 2080
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8445404171943665,
+      "learning_rate": 9.915560579921106e-05,
+      "loss": 4.1933,
+      "step": 2085
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9288851618766785,
+      "learning_rate": 9.915135865615074e-05,
+      "loss": 4.1628,
+      "step": 2090
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0005863904953003,
+      "learning_rate": 9.914710095012332e-05,
+      "loss": 4.3106,
+      "step": 2095
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.913024365901947,
+      "learning_rate": 9.91428326820438e-05,
+      "loss": 4.2873,
+      "step": 2100
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8917052745819092,
+      "learning_rate": 9.913855385282947e-05,
+      "loss": 4.25,
+      "step": 2105
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0002299547195435,
+      "learning_rate": 9.913426446339986e-05,
+      "loss": 4.2872,
+      "step": 2110
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9522709250450134,
+      "learning_rate": 9.912996451467682e-05,
+      "loss": 4.1066,
+      "step": 2115
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9926539659500122,
+      "learning_rate": 9.912565400758443e-05,
+      "loss": 4.2711,
+      "step": 2120
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8360921144485474,
+      "learning_rate": 9.912133294304901e-05,
+      "loss": 4.3839,
+      "step": 2125
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8529282808303833,
+      "learning_rate": 9.911700132199924e-05,
+      "loss": 4.2085,
+      "step": 2130
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8825228810310364,
+      "learning_rate": 9.911265914536598e-05,
+      "loss": 4.2741,
+      "step": 2135
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0875592231750488,
+      "learning_rate": 9.910830641408242e-05,
+      "loss": 4.365,
+      "step": 2140
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8486242294311523,
+      "learning_rate": 9.910394312908397e-05,
+      "loss": 4.3242,
+      "step": 2145
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9227307438850403,
+      "learning_rate": 9.909956929130834e-05,
+      "loss": 4.214,
+      "step": 2150
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9528930187225342,
+      "learning_rate": 9.90951849016955e-05,
+      "loss": 4.2377,
+      "step": 2155
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9374765753746033,
+      "learning_rate": 9.909078996118768e-05,
+      "loss": 4.1935,
+      "step": 2160
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8720608949661255,
+      "learning_rate": 9.908638447072939e-05,
+      "loss": 4.3211,
+      "step": 2165
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0150630474090576,
+      "learning_rate": 9.90819684312674e-05,
+      "loss": 4.2422,
+      "step": 2170
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8510294556617737,
+      "learning_rate": 9.907754184375074e-05,
+      "loss": 4.2749,
+      "step": 2175
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8451830148696899,
+      "learning_rate": 9.90731047091307e-05,
+      "loss": 4.0637,
+      "step": 2180
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.813502848148346,
+      "learning_rate": 9.906865702836088e-05,
+      "loss": 4.283,
+      "step": 2185
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0244249105453491,
+      "learning_rate": 9.90641988023971e-05,
+      "loss": 4.3037,
+      "step": 2190
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0313026905059814,
+      "learning_rate": 9.905973003219748e-05,
+      "loss": 4.2405,
+      "step": 2195
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8593780994415283,
+      "learning_rate": 9.905525071872237e-05,
+      "loss": 4.1206,
+      "step": 2200
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9260755777359009,
+      "learning_rate": 9.905076086293443e-05,
+      "loss": 4.2254,
+      "step": 2205
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8005271553993225,
+      "learning_rate": 9.904626046579852e-05,
+      "loss": 4.1627,
+      "step": 2210
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8273778557777405,
+      "learning_rate": 9.904174952828186e-05,
+      "loss": 4.1319,
+      "step": 2215
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8583614826202393,
+      "learning_rate": 9.903722805135385e-05,
+      "loss": 4.2576,
+      "step": 2220
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8739277720451355,
+      "learning_rate": 9.903269603598617e-05,
+      "loss": 4.1802,
+      "step": 2225
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9247838854789734,
+      "learning_rate": 9.902815348315282e-05,
+      "loss": 4.2962,
+      "step": 2230
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8835291266441345,
+      "learning_rate": 9.902360039383e-05,
+      "loss": 4.1853,
+      "step": 2235
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9064680933952332,
+      "learning_rate": 9.901903676899621e-05,
+      "loss": 4.2153,
+      "step": 2240
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0199729204177856,
+      "learning_rate": 9.901446260963221e-05,
+      "loss": 4.1733,
+      "step": 2245
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8242191672325134,
+      "learning_rate": 9.9009877916721e-05,
+      "loss": 4.1924,
+      "step": 2250
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8620455265045166,
+      "learning_rate": 9.900528269124787e-05,
+      "loss": 4.1423,
+      "step": 2255
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8975424766540527,
+      "learning_rate": 9.900067693420039e-05,
+      "loss": 4.2253,
+      "step": 2260
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.03963303565979,
+      "learning_rate": 9.899606064656832e-05,
+      "loss": 4.1118,
+      "step": 2265
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8586543798446655,
+      "learning_rate": 9.899143382934379e-05,
+      "loss": 4.1933,
+      "step": 2270
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8085629343986511,
+      "learning_rate": 9.898679648352108e-05,
+      "loss": 4.1172,
+      "step": 2275
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9052501916885376,
+      "learning_rate": 9.898214861009683e-05,
+      "loss": 4.1274,
+      "step": 2280
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.848138689994812,
+      "learning_rate": 9.897749021006988e-05,
+      "loss": 4.2026,
+      "step": 2285
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8831045031547546,
+      "learning_rate": 9.897282128444135e-05,
+      "loss": 4.2566,
+      "step": 2290
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8589482307434082,
+      "learning_rate": 9.896814183421464e-05,
+      "loss": 4.2467,
+      "step": 2295
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0212466716766357,
+      "learning_rate": 9.896345186039538e-05,
+      "loss": 4.1567,
+      "step": 2300
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9021971225738525,
+      "learning_rate": 9.895875136399149e-05,
+      "loss": 4.2211,
+      "step": 2305
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1750144958496094,
+      "learning_rate": 9.895404034601313e-05,
+      "loss": 4.2009,
+      "step": 2310
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0178542137145996,
+      "learning_rate": 9.894931880747273e-05,
+      "loss": 4.2864,
+      "step": 2315
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.811051607131958,
+      "learning_rate": 9.894458674938499e-05,
+      "loss": 4.2829,
+      "step": 2320
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9225900173187256,
+      "learning_rate": 9.893984417276686e-05,
+      "loss": 4.291,
+      "step": 2325
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8347369432449341,
+      "learning_rate": 9.893509107863754e-05,
+      "loss": 4.1355,
+      "step": 2330
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8535067439079285,
+      "learning_rate": 9.893032746801854e-05,
+      "loss": 4.0674,
+      "step": 2335
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8876878619194031,
+      "learning_rate": 9.892555334193353e-05,
+      "loss": 4.0559,
+      "step": 2340
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.774366021156311,
+      "learning_rate": 9.892076870140857e-05,
+      "loss": 4.2742,
+      "step": 2345
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9003043174743652,
+      "learning_rate": 9.891597354747186e-05,
+      "loss": 4.2662,
+      "step": 2350
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9063124060630798,
+      "learning_rate": 9.891116788115393e-05,
+      "loss": 4.2017,
+      "step": 2355
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.863969624042511,
+      "learning_rate": 9.890635170348756e-05,
+      "loss": 4.2214,
+      "step": 2360
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8075172305107117,
+      "learning_rate": 9.890152501550777e-05,
+      "loss": 4.2056,
+      "step": 2365
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.843422532081604,
+      "learning_rate": 9.889668781825185e-05,
+      "loss": 4.2103,
+      "step": 2370
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8354036808013916,
+      "learning_rate": 9.889184011275936e-05,
+      "loss": 4.0438,
+      "step": 2375
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.792033851146698,
+      "learning_rate": 9.888698190007209e-05,
+      "loss": 4.0262,
+      "step": 2380
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8295325040817261,
+      "learning_rate": 9.88821131812341e-05,
+      "loss": 4.1299,
+      "step": 2385
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7669634819030762,
+      "learning_rate": 9.887723395729174e-05,
+      "loss": 4.2374,
+      "step": 2390
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8457567691802979,
+      "learning_rate": 9.887234422929355e-05,
+      "loss": 4.1924,
+      "step": 2395
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9156321883201599,
+      "learning_rate": 9.886744399829039e-05,
+      "loss": 4.1544,
+      "step": 2400
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8035711050033569,
+      "learning_rate": 9.886253326533535e-05,
+      "loss": 4.2035,
+      "step": 2405
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.767970860004425,
+      "learning_rate": 9.885761203148378e-05,
+      "loss": 4.1786,
+      "step": 2410
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7946574687957764,
+      "learning_rate": 9.885268029779327e-05,
+      "loss": 4.2939,
+      "step": 2415
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7956034541130066,
+      "learning_rate": 9.884773806532373e-05,
+      "loss": 4.1482,
+      "step": 2420
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8520569205284119,
+      "learning_rate": 9.884278533513723e-05,
+      "loss": 4.1314,
+      "step": 2425
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8857873678207397,
+      "learning_rate": 9.883782210829816e-05,
+      "loss": 4.1667,
+      "step": 2430
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7695557475090027,
+      "learning_rate": 9.883284838587317e-05,
+      "loss": 4.1789,
+      "step": 2435
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7941457629203796,
+      "learning_rate": 9.882786416893112e-05,
+      "loss": 4.1519,
+      "step": 2440
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8108274340629578,
+      "learning_rate": 9.882286945854319e-05,
+      "loss": 4.1915,
+      "step": 2445
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8673930168151855,
+      "learning_rate": 9.881786425578274e-05,
+      "loss": 4.0961,
+      "step": 2450
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.783336877822876,
+      "learning_rate": 9.881284856172545e-05,
+      "loss": 4.1509,
+      "step": 2455
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.856675386428833,
+      "learning_rate": 9.880782237744921e-05,
+      "loss": 4.1732,
+      "step": 2460
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8168655633926392,
+      "learning_rate": 9.880278570403419e-05,
+      "loss": 4.0057,
+      "step": 2465
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8240490555763245,
+      "learning_rate": 9.87977385425628e-05,
+      "loss": 4.1802,
+      "step": 2470
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8018542528152466,
+      "learning_rate": 9.87926808941197e-05,
+      "loss": 4.2309,
+      "step": 2475
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7763277888298035,
+      "learning_rate": 9.878761275979184e-05,
+      "loss": 4.2137,
+      "step": 2480
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7453907132148743,
+      "learning_rate": 9.878253414066839e-05,
+      "loss": 4.0526,
+      "step": 2485
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9517676830291748,
+      "learning_rate": 9.877744503784076e-05,
+      "loss": 4.1851,
+      "step": 2490
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9704002737998962,
+      "learning_rate": 9.877234545240265e-05,
+      "loss": 4.1618,
+      "step": 2495
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8855175375938416,
+      "learning_rate": 9.876723538545001e-05,
+      "loss": 4.1823,
+      "step": 2500
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8798913359642029,
+      "learning_rate": 9.8762114838081e-05,
+      "loss": 4.2168,
+      "step": 2505
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9619790315628052,
+      "learning_rate": 9.875698381139607e-05,
+      "loss": 4.1052,
+      "step": 2510
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8987843990325928,
+      "learning_rate": 9.875184230649792e-05,
+      "loss": 4.3012,
+      "step": 2515
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8634008765220642,
+      "learning_rate": 9.874669032449149e-05,
+      "loss": 4.0108,
+      "step": 2520
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8625091910362244,
+      "learning_rate": 9.874152786648399e-05,
+      "loss": 4.106,
+      "step": 2525
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7973573803901672,
+      "learning_rate": 9.873635493358484e-05,
+      "loss": 4.219,
+      "step": 2530
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8727997541427612,
+      "learning_rate": 9.873117152690576e-05,
+      "loss": 4.2618,
+      "step": 2535
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8032433390617371,
+      "learning_rate": 9.872597764756071e-05,
+      "loss": 4.2342,
+      "step": 2540
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7865430116653442,
+      "learning_rate": 9.872077329666587e-05,
+      "loss": 4.1499,
+      "step": 2545
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8177130222320557,
+      "learning_rate": 9.871555847533968e-05,
+      "loss": 4.2918,
+      "step": 2550
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8597103953361511,
+      "learning_rate": 9.871033318470289e-05,
+      "loss": 4.13,
+      "step": 2555
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7673056125640869,
+      "learning_rate": 9.870509742587838e-05,
+      "loss": 4.1584,
+      "step": 2560
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9154881238937378,
+      "learning_rate": 9.869985119999143e-05,
+      "loss": 4.1633,
+      "step": 2565
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.859653115272522,
+      "learning_rate": 9.869459450816942e-05,
+      "loss": 4.0417,
+      "step": 2570
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8476735353469849,
+      "learning_rate": 9.86893273515421e-05,
+      "loss": 4.0928,
+      "step": 2575
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.827778160572052,
+      "learning_rate": 9.868404973124139e-05,
+      "loss": 4.1314,
+      "step": 2580
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9780545234680176,
+      "learning_rate": 9.86787616484015e-05,
+      "loss": 4.1972,
+      "step": 2585
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7707496881484985,
+      "learning_rate": 9.867346310415886e-05,
+      "loss": 4.1353,
+      "step": 2590
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8499646782875061,
+      "learning_rate": 9.866815409965219e-05,
+      "loss": 4.1573,
+      "step": 2595
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3792710304260254,
+      "learning_rate": 9.86628346360224e-05,
+      "loss": 4.1416,
+      "step": 2600
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7860751748085022,
+      "learning_rate": 9.865750471441271e-05,
+      "loss": 4.1358,
+      "step": 2605
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8300714492797852,
+      "learning_rate": 9.865216433596853e-05,
+      "loss": 4.0823,
+      "step": 2610
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9836693406105042,
+      "learning_rate": 9.864681350183758e-05,
+      "loss": 4.1331,
+      "step": 2615
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9571417570114136,
+      "learning_rate": 9.864145221316975e-05,
+      "loss": 4.177,
+      "step": 2620
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7734148502349854,
+      "learning_rate": 9.863608047111725e-05,
+      "loss": 4.2281,
+      "step": 2625
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.989045262336731,
+      "learning_rate": 9.863069827683446e-05,
+      "loss": 4.077,
+      "step": 2630
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9619313478469849,
+      "learning_rate": 9.862530563147812e-05,
+      "loss": 4.1661,
+      "step": 2635
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8962774872779846,
+      "learning_rate": 9.861990253620709e-05,
+      "loss": 4.175,
+      "step": 2640
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8202120661735535,
+      "learning_rate": 9.861448899218255e-05,
+      "loss": 4.1376,
+      "step": 2645
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8472998142242432,
+      "learning_rate": 9.860906500056792e-05,
+      "loss": 4.1785,
+      "step": 2650
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8804469108581543,
+      "learning_rate": 9.860363056252883e-05,
+      "loss": 4.1965,
+      "step": 2655
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8699422478675842,
+      "learning_rate": 9.859818567923321e-05,
+      "loss": 4.1674,
+      "step": 2660
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8098098039627075,
+      "learning_rate": 9.859273035185118e-05,
+      "loss": 4.0318,
+      "step": 2665
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8125141859054565,
+      "learning_rate": 9.858726458155512e-05,
+      "loss": 4.2092,
+      "step": 2670
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7835256457328796,
+      "learning_rate": 9.858178836951967e-05,
+      "loss": 4.1416,
+      "step": 2675
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8320144414901733,
+      "learning_rate": 9.857630171692174e-05,
+      "loss": 4.0987,
+      "step": 2680
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7360141277313232,
+      "learning_rate": 9.857080462494039e-05,
+      "loss": 3.9873,
+      "step": 2685
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7672869563102722,
+      "learning_rate": 9.856529709475703e-05,
+      "loss": 4.0046,
+      "step": 2690
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9761074781417847,
+      "learning_rate": 9.855977912755524e-05,
+      "loss": 4.0818,
+      "step": 2695
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.739989697933197,
+      "learning_rate": 9.855425072452086e-05,
+      "loss": 4.0855,
+      "step": 2700
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.900478720664978,
+      "learning_rate": 9.854871188684204e-05,
+      "loss": 4.1451,
+      "step": 2705
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8419812917709351,
+      "learning_rate": 9.854316261570905e-05,
+      "loss": 4.1191,
+      "step": 2710
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9926355481147766,
+      "learning_rate": 9.85376029123145e-05,
+      "loss": 4.0149,
+      "step": 2715
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8116398453712463,
+      "learning_rate": 9.853203277785318e-05,
+      "loss": 4.088,
+      "step": 2720
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9778822660446167,
+      "learning_rate": 9.85264522135222e-05,
+      "loss": 4.1498,
+      "step": 2725
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8457492589950562,
+      "learning_rate": 9.852086122052083e-05,
+      "loss": 4.1111,
+      "step": 2730
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8294225931167603,
+      "learning_rate": 9.85152598000506e-05,
+      "loss": 4.0539,
+      "step": 2735
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7900792956352234,
+      "learning_rate": 9.850964795331531e-05,
+      "loss": 4.0914,
+      "step": 2740
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8610666990280151,
+      "learning_rate": 9.8504025681521e-05,
+      "loss": 4.0341,
+      "step": 2745
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7812069058418274,
+      "learning_rate": 9.84983929858759e-05,
+      "loss": 4.0626,
+      "step": 2750
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8641030192375183,
+      "learning_rate": 9.849274986759054e-05,
+      "loss": 4.1734,
+      "step": 2755
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7941977381706238,
+      "learning_rate": 9.848709632787768e-05,
+      "loss": 4.1126,
+      "step": 2760
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1504554748535156,
+      "learning_rate": 9.848143236795225e-05,
+      "loss": 4.1171,
+      "step": 2765
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7732850313186646,
+      "learning_rate": 9.847575798903152e-05,
+      "loss": 4.0961,
+      "step": 2770
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8347551822662354,
+      "learning_rate": 9.847007319233496e-05,
+      "loss": 4.0172,
+      "step": 2775
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8266686201095581,
+      "learning_rate": 9.846437797908424e-05,
+      "loss": 4.065,
+      "step": 2780
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9070795774459839,
+      "learning_rate": 9.845867235050333e-05,
+      "loss": 4.1795,
+      "step": 2785
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8572375774383545,
+      "learning_rate": 9.845295630781837e-05,
+      "loss": 4.1443,
+      "step": 2790
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7847888469696045,
+      "learning_rate": 9.844722985225783e-05,
+      "loss": 4.1202,
+      "step": 2795
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7725611329078674,
+      "learning_rate": 9.844149298505232e-05,
+      "loss": 4.1284,
+      "step": 2800
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7498067617416382,
+      "learning_rate": 9.843574570743474e-05,
+      "loss": 4.0639,
+      "step": 2805
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7982293367385864,
+      "learning_rate": 9.842998802064025e-05,
+      "loss": 4.0614,
+      "step": 2810
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7394840121269226,
+      "learning_rate": 9.842421992590619e-05,
+      "loss": 4.0831,
+      "step": 2815
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7713798880577087,
+      "learning_rate": 9.841844142447215e-05,
+      "loss": 4.0212,
+      "step": 2820
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7344151735305786,
+      "learning_rate": 9.841265251758e-05,
+      "loss": 4.0928,
+      "step": 2825
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7637064456939697,
+      "learning_rate": 9.840685320647381e-05,
+      "loss": 4.0547,
+      "step": 2830
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7163331508636475,
+      "learning_rate": 9.840104349239988e-05,
+      "loss": 4.1047,
+      "step": 2835
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7959428429603577,
+      "learning_rate": 9.839522337660677e-05,
+      "loss": 4.0451,
+      "step": 2840
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9641393423080444,
+      "learning_rate": 9.838939286034526e-05,
+      "loss": 3.9946,
+      "step": 2845
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.771903395652771,
+      "learning_rate": 9.838355194486836e-05,
+      "loss": 4.0716,
+      "step": 2850
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7979621291160583,
+      "learning_rate": 9.837770063143133e-05,
+      "loss": 4.1688,
+      "step": 2855
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8572778105735779,
+      "learning_rate": 9.837183892129164e-05,
+      "loss": 3.9741,
+      "step": 2860
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8652186393737793,
+      "learning_rate": 9.836596681570906e-05,
+      "loss": 3.8919,
+      "step": 2865
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1240743398666382,
+      "learning_rate": 9.83600843159455e-05,
+      "loss": 4.0348,
+      "step": 2870
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7439741492271423,
+      "learning_rate": 9.835419142326517e-05,
+      "loss": 4.151,
+      "step": 2875
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7896543741226196,
+      "learning_rate": 9.834828813893448e-05,
+      "loss": 4.0417,
+      "step": 2880
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8672937154769897,
+      "learning_rate": 9.83423744642221e-05,
+      "loss": 4.178,
+      "step": 2885
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7366281151771545,
+      "learning_rate": 9.833645040039892e-05,
+      "loss": 3.9895,
+      "step": 2890
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7278005480766296,
+      "learning_rate": 9.833051594873805e-05,
+      "loss": 4.0237,
+      "step": 2895
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8357091546058655,
+      "learning_rate": 9.832457111051485e-05,
+      "loss": 4.1676,
+      "step": 2900
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8501677513122559,
+      "learning_rate": 9.831861588700692e-05,
+      "loss": 4.0208,
+      "step": 2905
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7765496969223022,
+      "learning_rate": 9.831265027949406e-05,
+      "loss": 4.1306,
+      "step": 2910
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.940582811832428,
+      "learning_rate": 9.830667428925833e-05,
+      "loss": 4.1398,
+      "step": 2915
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.770330548286438,
+      "learning_rate": 9.8300687917584e-05,
+      "loss": 4.1057,
+      "step": 2920
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7469985485076904,
+      "learning_rate": 9.82946911657576e-05,
+      "loss": 4.1488,
+      "step": 2925
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.843786358833313,
+      "learning_rate": 9.828868403506788e-05,
+      "loss": 3.9817,
+      "step": 2930
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7778830528259277,
+      "learning_rate": 9.828266652680577e-05,
+      "loss": 3.9885,
+      "step": 2935
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8030687570571899,
+      "learning_rate": 9.827663864226453e-05,
+      "loss": 4.0707,
+      "step": 2940
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7301581501960754,
+      "learning_rate": 9.827060038273956e-05,
+      "loss": 4.0123,
+      "step": 2945
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8521193265914917,
+      "learning_rate": 9.826455174952852e-05,
+      "loss": 4.082,
+      "step": 2950
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7837696671485901,
+      "learning_rate": 9.825849274393133e-05,
+      "loss": 3.9087,
+      "step": 2955
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7679029703140259,
+      "learning_rate": 9.825242336725009e-05,
+      "loss": 4.0479,
+      "step": 2960
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8081327676773071,
+      "learning_rate": 9.824634362078915e-05,
+      "loss": 3.9772,
+      "step": 2965
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7829834818840027,
+      "learning_rate": 9.824025350585511e-05,
+      "loss": 3.9618,
+      "step": 2970
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8335613012313843,
+      "learning_rate": 9.823415302375676e-05,
+      "loss": 4.1615,
+      "step": 2975
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7775614857673645,
+      "learning_rate": 9.822804217580515e-05,
+      "loss": 4.0562,
+      "step": 2980
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7770524024963379,
+      "learning_rate": 9.822192096331351e-05,
+      "loss": 4.06,
+      "step": 2985
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.79961758852005,
+      "learning_rate": 9.82157893875974e-05,
+      "loss": 4.0399,
+      "step": 2990
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7700688242912292,
+      "learning_rate": 9.820964744997445e-05,
+      "loss": 4.1084,
+      "step": 2995
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8124660849571228,
+      "learning_rate": 9.820349515176466e-05,
+      "loss": 3.9924,
+      "step": 3000
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8372924327850342,
+      "learning_rate": 9.819733249429019e-05,
+      "loss": 4.0425,
+      "step": 3005
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8001456260681152,
+      "learning_rate": 9.819115947887545e-05,
+      "loss": 4.0908,
+      "step": 3010
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0193424224853516,
+      "learning_rate": 9.818497610684705e-05,
+      "loss": 4.1331,
+      "step": 3015
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7650743126869202,
+      "learning_rate": 9.817878237953383e-05,
+      "loss": 3.9738,
+      "step": 3020
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9089260697364807,
+      "learning_rate": 9.817257829826689e-05,
+      "loss": 4.0622,
+      "step": 3025
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7920193076133728,
+      "learning_rate": 9.81663638643795e-05,
+      "loss": 3.9885,
+      "step": 3030
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1940184831619263,
+      "learning_rate": 9.81601390792072e-05,
+      "loss": 4.0653,
+      "step": 3035
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.952750563621521,
+      "learning_rate": 9.815390394408776e-05,
+      "loss": 4.0396,
+      "step": 3040
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9346968531608582,
+      "learning_rate": 9.814765846036111e-05,
+      "loss": 4.1286,
+      "step": 3045
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8506262898445129,
+      "learning_rate": 9.81414026293695e-05,
+      "loss": 4.0425,
+      "step": 3050
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7965868711471558,
+      "learning_rate": 9.813513645245729e-05,
+      "loss": 4.0632,
+      "step": 3055
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7659491300582886,
+      "learning_rate": 9.812885993097118e-05,
+      "loss": 4.0396,
+      "step": 3060
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7796339988708496,
+      "learning_rate": 9.812257306626002e-05,
+      "loss": 3.9624,
+      "step": 3065
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7833777666091919,
+      "learning_rate": 9.811627585967487e-05,
+      "loss": 3.9802,
+      "step": 3070
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8041028380393982,
+      "learning_rate": 9.810996831256909e-05,
+      "loss": 3.9706,
+      "step": 3075
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8143548965454102,
+      "learning_rate": 9.81036504262982e-05,
+      "loss": 4.1615,
+      "step": 3080
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7737205624580383,
+      "learning_rate": 9.809732220221995e-05,
+      "loss": 4.0999,
+      "step": 3085
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8043818473815918,
+      "learning_rate": 9.80909836416943e-05,
+      "loss": 4.0044,
+      "step": 3090
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7989277243614197,
+      "learning_rate": 9.80846347460835e-05,
+      "loss": 4.1803,
+      "step": 3095
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.869310200214386,
+      "learning_rate": 9.807827551675192e-05,
+      "loss": 3.9496,
+      "step": 3100
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.100605845451355,
+      "learning_rate": 9.807190595506623e-05,
+      "loss": 4.0185,
+      "step": 3105
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8150125741958618,
+      "learning_rate": 9.80655260623953e-05,
+      "loss": 3.9793,
+      "step": 3110
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8614518046379089,
+      "learning_rate": 9.805913584011019e-05,
+      "loss": 4.0579,
+      "step": 3115
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7724331617355347,
+      "learning_rate": 9.805273528958423e-05,
+      "loss": 4.0474,
+      "step": 3120
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8028234839439392,
+      "learning_rate": 9.804632441219293e-05,
+      "loss": 4.0914,
+      "step": 3125
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8180481791496277,
+      "learning_rate": 9.803990320931402e-05,
+      "loss": 3.9523,
+      "step": 3130
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7312827110290527,
+      "learning_rate": 9.803347168232746e-05,
+      "loss": 4.0979,
+      "step": 3135
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8292685747146606,
+      "learning_rate": 9.802702983261545e-05,
+      "loss": 3.907,
+      "step": 3140
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7670102119445801,
+      "learning_rate": 9.802057766156238e-05,
+      "loss": 4.0154,
+      "step": 3145
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7940794825553894,
+      "learning_rate": 9.801411517055488e-05,
+      "loss": 4.0159,
+      "step": 3150
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8141523003578186,
+      "learning_rate": 9.800764236098175e-05,
+      "loss": 4.0251,
+      "step": 3155
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7281124591827393,
+      "learning_rate": 9.800115923423405e-05,
+      "loss": 3.9239,
+      "step": 3160
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7306647896766663,
+      "learning_rate": 9.799466579170508e-05,
+      "loss": 4.0453,
+      "step": 3165
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7612237930297852,
+      "learning_rate": 9.798816203479029e-05,
+      "loss": 3.9924,
+      "step": 3170
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8770204186439514,
+      "learning_rate": 9.79816479648874e-05,
+      "loss": 4.0778,
+      "step": 3175
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7630795836448669,
+      "learning_rate": 9.797512358339633e-05,
+      "loss": 4.0702,
+      "step": 3180
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7618923783302307,
+      "learning_rate": 9.796858889171921e-05,
+      "loss": 4.0583,
+      "step": 3185
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.818384051322937,
+      "learning_rate": 9.79620438912604e-05,
+      "loss": 4.0408,
+      "step": 3190
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7824782729148865,
+      "learning_rate": 9.795548858342645e-05,
+      "loss": 4.0149,
+      "step": 3195
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8857114315032959,
+      "learning_rate": 9.794892296962615e-05,
+      "loss": 4.0374,
+      "step": 3200
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8463344573974609,
+      "learning_rate": 9.794234705127047e-05,
+      "loss": 3.9549,
+      "step": 3205
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8507333993911743,
+      "learning_rate": 9.793576082977267e-05,
+      "loss": 3.9373,
+      "step": 3210
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0891833305358887,
+      "learning_rate": 9.792916430654815e-05,
+      "loss": 3.8881,
+      "step": 3215
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8575323820114136,
+      "learning_rate": 9.792255748301453e-05,
+      "loss": 4.0221,
+      "step": 3220
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7708762288093567,
+      "learning_rate": 9.791594036059168e-05,
+      "loss": 3.9076,
+      "step": 3225
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.83617103099823,
+      "learning_rate": 9.790931294070169e-05,
+      "loss": 4.0132,
+      "step": 3230
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7307329773902893,
+      "learning_rate": 9.79026752247688e-05,
+      "loss": 3.9073,
+      "step": 3235
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7178450226783752,
+      "learning_rate": 9.78960272142195e-05,
+      "loss": 4.0073,
+      "step": 3240
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7541383504867554,
+      "learning_rate": 9.788936891048251e-05,
+      "loss": 3.9971,
+      "step": 3245
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8955694437026978,
+      "learning_rate": 9.788270031498876e-05,
+      "loss": 4.0143,
+      "step": 3250
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7889512777328491,
+      "learning_rate": 9.787602142917135e-05,
+      "loss": 3.9717,
+      "step": 3255
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7644156217575073,
+      "learning_rate": 9.786933225446563e-05,
+      "loss": 3.93,
+      "step": 3260
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8339189291000366,
+      "learning_rate": 9.786263279230915e-05,
+      "loss": 4.0888,
+      "step": 3265
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7627922892570496,
+      "learning_rate": 9.785592304414167e-05,
+      "loss": 4.023,
+      "step": 3270
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7756912112236023,
+      "learning_rate": 9.784920301140516e-05,
+      "loss": 3.9603,
+      "step": 3275
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7174762487411499,
+      "learning_rate": 9.78424726955438e-05,
+      "loss": 3.9913,
+      "step": 3280
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7982318997383118,
+      "learning_rate": 9.783573209800398e-05,
+      "loss": 3.9703,
+      "step": 3285
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7398289442062378,
+      "learning_rate": 9.782898122023432e-05,
+      "loss": 4.055,
+      "step": 3290
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7623131275177002,
+      "learning_rate": 9.78222200636856e-05,
+      "loss": 4.135,
+      "step": 3295
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.74330735206604,
+      "learning_rate": 9.781544862981087e-05,
+      "loss": 3.9564,
+      "step": 3300
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7418188452720642,
+      "learning_rate": 9.780866692006534e-05,
+      "loss": 3.9709,
+      "step": 3305
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9443297386169434,
+      "learning_rate": 9.780187493590645e-05,
+      "loss": 3.911,
+      "step": 3310
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7579126358032227,
+      "learning_rate": 9.779507267879385e-05,
+      "loss": 3.9914,
+      "step": 3315
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.904556930065155,
+      "learning_rate": 9.778826015018936e-05,
+      "loss": 3.8799,
+      "step": 3320
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7957940101623535,
+      "learning_rate": 9.778143735155709e-05,
+      "loss": 3.9986,
+      "step": 3325
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7432476878166199,
+      "learning_rate": 9.777460428436327e-05,
+      "loss": 4.0692,
+      "step": 3330
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.000910997390747,
+      "learning_rate": 9.776776095007638e-05,
+      "loss": 4.0815,
+      "step": 3335
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7548974752426147,
+      "learning_rate": 9.776090735016711e-05,
+      "loss": 4.0288,
+      "step": 3340
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7548381090164185,
+      "learning_rate": 9.775404348610835e-05,
+      "loss": 4.0369,
+      "step": 3345
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7925072312355042,
+      "learning_rate": 9.774716935937518e-05,
+      "loss": 4.0104,
+      "step": 3350
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7259548902511597,
+      "learning_rate": 9.77402849714449e-05,
+      "loss": 3.8567,
+      "step": 3355
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7419453859329224,
+      "learning_rate": 9.773339032379701e-05,
+      "loss": 4.0048,
+      "step": 3360
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0442363023757935,
+      "learning_rate": 9.772648541791323e-05,
+      "loss": 3.891,
+      "step": 3365
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7861118316650391,
+      "learning_rate": 9.771957025527745e-05,
+      "loss": 3.9593,
+      "step": 3370
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7535933256149292,
+      "learning_rate": 9.771264483737581e-05,
+      "loss": 3.9414,
+      "step": 3375
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7460413575172424,
+      "learning_rate": 9.770570916569661e-05,
+      "loss": 3.9403,
+      "step": 3380
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7163098454475403,
+      "learning_rate": 9.76987632417304e-05,
+      "loss": 3.9364,
+      "step": 3385
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7522199749946594,
+      "learning_rate": 9.769180706696988e-05,
+      "loss": 3.9593,
+      "step": 3390
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8179682493209839,
+      "learning_rate": 9.768484064290997e-05,
+      "loss": 3.9763,
+      "step": 3395
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7668630480766296,
+      "learning_rate": 9.767786397104786e-05,
+      "loss": 4.0449,
+      "step": 3400
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8921141028404236,
+      "learning_rate": 9.767087705288283e-05,
+      "loss": 3.9112,
+      "step": 3405
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7747445106506348,
+      "learning_rate": 9.766387988991642e-05,
+      "loss": 4.0566,
+      "step": 3410
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7571500539779663,
+      "learning_rate": 9.765687248365241e-05,
+      "loss": 4.0412,
+      "step": 3415
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1251405477523804,
+      "learning_rate": 9.76498548355967e-05,
+      "loss": 3.8834,
+      "step": 3420
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.772466242313385,
+      "learning_rate": 9.764282694725744e-05,
+      "loss": 3.8681,
+      "step": 3425
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9793023467063904,
+      "learning_rate": 9.7635788820145e-05,
+      "loss": 4.0077,
+      "step": 3430
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8201249241828918,
+      "learning_rate": 9.762874045577189e-05,
+      "loss": 3.965,
+      "step": 3435
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7866344451904297,
+      "learning_rate": 9.762168185565284e-05,
+      "loss": 3.9254,
+      "step": 3440
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7204021215438843,
+      "learning_rate": 9.761461302130484e-05,
+      "loss": 3.9863,
+      "step": 3445
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7343882918357849,
+      "learning_rate": 9.760753395424699e-05,
+      "loss": 3.9642,
+      "step": 3450
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7686299681663513,
+      "learning_rate": 9.760044465600065e-05,
+      "loss": 3.9262,
+      "step": 3455
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6907262802124023,
+      "learning_rate": 9.759334512808935e-05,
+      "loss": 4.0311,
+      "step": 3460
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7561254501342773,
+      "learning_rate": 9.758623537203883e-05,
+      "loss": 3.8309,
+      "step": 3465
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7108284831047058,
+      "learning_rate": 9.757911538937706e-05,
+      "loss": 3.8814,
+      "step": 3470
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7558962106704712,
+      "learning_rate": 9.757198518163411e-05,
+      "loss": 3.8804,
+      "step": 3475
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7611292004585266,
+      "learning_rate": 9.756484475034235e-05,
+      "loss": 4.0067,
+      "step": 3480
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7121755480766296,
+      "learning_rate": 9.75576940970363e-05,
+      "loss": 3.9441,
+      "step": 3485
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7752575278282166,
+      "learning_rate": 9.755053322325269e-05,
+      "loss": 3.9217,
+      "step": 3490
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7309322953224182,
+      "learning_rate": 9.754336213053043e-05,
+      "loss": 3.9791,
+      "step": 3495
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7653898000717163,
+      "learning_rate": 9.753618082041066e-05,
+      "loss": 4.0277,
+      "step": 3500
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.862130880355835,
+      "learning_rate": 9.752898929443666e-05,
+      "loss": 3.8495,
+      "step": 3505
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7566332221031189,
+      "learning_rate": 9.752178755415396e-05,
+      "loss": 3.8289,
+      "step": 3510
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7841774821281433,
+      "learning_rate": 9.751457560111026e-05,
+      "loss": 3.896,
+      "step": 3515
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7289752960205078,
+      "learning_rate": 9.750735343685547e-05,
+      "loss": 4.1058,
+      "step": 3520
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7284510135650635,
+      "learning_rate": 9.750012106294167e-05,
+      "loss": 4.0212,
+      "step": 3525
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7350246906280518,
+      "learning_rate": 9.749287848092315e-05,
+      "loss": 3.9869,
+      "step": 3530
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7356334924697876,
+      "learning_rate": 9.748562569235637e-05,
+      "loss": 3.9254,
+      "step": 3535
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.744530200958252,
+      "learning_rate": 9.747836269880003e-05,
+      "loss": 4.0003,
+      "step": 3540
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7327910661697388,
+      "learning_rate": 9.7471089501815e-05,
+      "loss": 3.8927,
+      "step": 3545
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7226892113685608,
+      "learning_rate": 9.746380610296435e-05,
+      "loss": 4.0003,
+      "step": 3550
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8997710347175598,
+      "learning_rate": 9.74565125038133e-05,
+      "loss": 3.8772,
+      "step": 3555
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7877281308174133,
+      "learning_rate": 9.744920870592931e-05,
+      "loss": 3.8677,
+      "step": 3560
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7937436103820801,
+      "learning_rate": 9.7441894710882e-05,
+      "loss": 3.8988,
+      "step": 3565
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7360289692878723,
+      "learning_rate": 9.743457052024323e-05,
+      "loss": 4.0414,
+      "step": 3570
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7351023554801941,
+      "learning_rate": 9.742723613558702e-05,
+      "loss": 4.0398,
+      "step": 3575
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7748407125473022,
+      "learning_rate": 9.741989155848953e-05,
+      "loss": 3.9266,
+      "step": 3580
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7629684805870056,
+      "learning_rate": 9.741253679052923e-05,
+      "loss": 3.9804,
+      "step": 3585
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8270654678344727,
+      "learning_rate": 9.740517183328666e-05,
+      "loss": 3.9386,
+      "step": 3590
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7086227536201477,
+      "learning_rate": 9.739779668834461e-05,
+      "loss": 3.9894,
+      "step": 3595
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7253994345664978,
+      "learning_rate": 9.739041135728806e-05,
+      "loss": 3.7815,
+      "step": 3600
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7592572569847107,
+      "learning_rate": 9.738301584170417e-05,
+      "loss": 4.0258,
+      "step": 3605
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7342005372047424,
+      "learning_rate": 9.737561014318228e-05,
+      "loss": 3.9935,
+      "step": 3610
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7062914967536926,
+      "learning_rate": 9.736819426331392e-05,
+      "loss": 3.9862,
+      "step": 3615
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7178187966346741,
+      "learning_rate": 9.736076820369283e-05,
+      "loss": 3.9346,
+      "step": 3620
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7413998246192932,
+      "learning_rate": 9.735333196591492e-05,
+      "loss": 3.9815,
+      "step": 3625
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7442604899406433,
+      "learning_rate": 9.734588555157826e-05,
+      "loss": 4.0414,
+      "step": 3630
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.726449191570282,
+      "learning_rate": 9.733842896228318e-05,
+      "loss": 3.8553,
+      "step": 3635
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7738696336746216,
+      "learning_rate": 9.733096219963211e-05,
+      "loss": 3.9669,
+      "step": 3640
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7260298728942871,
+      "learning_rate": 9.732348526522974e-05,
+      "loss": 3.9546,
+      "step": 3645
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7806318998336792,
+      "learning_rate": 9.73159981606829e-05,
+      "loss": 3.9133,
+      "step": 3650
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7431362271308899,
+      "learning_rate": 9.730850088760064e-05,
+      "loss": 3.8909,
+      "step": 3655
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7421375513076782,
+      "learning_rate": 9.730099344759413e-05,
+      "loss": 4.0284,
+      "step": 3660
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7068578004837036,
+      "learning_rate": 9.729347584227684e-05,
+      "loss": 3.9447,
+      "step": 3665
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7487486004829407,
+      "learning_rate": 9.72859480732643e-05,
+      "loss": 3.8956,
+      "step": 3670
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7758656144142151,
+      "learning_rate": 9.727841014217429e-05,
+      "loss": 3.9505,
+      "step": 3675
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7883812785148621,
+      "learning_rate": 9.727086205062676e-05,
+      "loss": 3.9176,
+      "step": 3680
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7411519289016724,
+      "learning_rate": 9.72633038002439e-05,
+      "loss": 3.9335,
+      "step": 3685
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7278258204460144,
+      "learning_rate": 9.725573539264994e-05,
+      "loss": 3.8928,
+      "step": 3690
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7816063165664673,
+      "learning_rate": 9.724815682947146e-05,
+      "loss": 4.0303,
+      "step": 3695
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7474303245544434,
+      "learning_rate": 9.72405681123371e-05,
+      "loss": 3.9508,
+      "step": 3700
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.691038966178894,
+      "learning_rate": 9.723296924287775e-05,
+      "loss": 4.0336,
+      "step": 3705
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7350514531135559,
+      "learning_rate": 9.722536022272646e-05,
+      "loss": 3.9605,
+      "step": 3710
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8015035390853882,
+      "learning_rate": 9.721774105351845e-05,
+      "loss": 3.9234,
+      "step": 3715
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7018195986747742,
+      "learning_rate": 9.721011173689114e-05,
+      "loss": 3.9862,
+      "step": 3720
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.431609869003296,
+      "learning_rate": 9.720247227448411e-05,
+      "loss": 3.928,
+      "step": 3725
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7782214879989624,
+      "learning_rate": 9.719482266793916e-05,
+      "loss": 3.8607,
+      "step": 3730
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7407627105712891,
+      "learning_rate": 9.718716291890022e-05,
+      "loss": 3.9214,
+      "step": 3735
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7137983441352844,
+      "learning_rate": 9.717949302901343e-05,
+      "loss": 3.919,
+      "step": 3740
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7712181210517883,
+      "learning_rate": 9.71718129999271e-05,
+      "loss": 3.9464,
+      "step": 3745
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7289685606956482,
+      "learning_rate": 9.716412283329172e-05,
+      "loss": 3.9661,
+      "step": 3750
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7536170482635498,
+      "learning_rate": 9.715642253075998e-05,
+      "loss": 3.9354,
+      "step": 3755
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7835754752159119,
+      "learning_rate": 9.714871209398671e-05,
+      "loss": 3.825,
+      "step": 3760
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7063088417053223,
+      "learning_rate": 9.714099152462892e-05,
+      "loss": 3.997,
+      "step": 3765
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8679980635643005,
+      "learning_rate": 9.713326082434584e-05,
+      "loss": 3.956,
+      "step": 3770
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.742300271987915,
+      "learning_rate": 9.712551999479884e-05,
+      "loss": 4.0409,
+      "step": 3775
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.706699788570404,
+      "learning_rate": 9.71177690376515e-05,
+      "loss": 3.9196,
+      "step": 3780
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8656459450721741,
+      "learning_rate": 9.711000795456951e-05,
+      "loss": 3.8748,
+      "step": 3785
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7414349913597107,
+      "learning_rate": 9.710223674722083e-05,
+      "loss": 3.8152,
+      "step": 3790
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.78056800365448,
+      "learning_rate": 9.709445541727551e-05,
+      "loss": 3.8669,
+      "step": 3795
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7553170919418335,
+      "learning_rate": 9.708666396640582e-05,
+      "loss": 3.8977,
+      "step": 3800
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9867407083511353,
+      "learning_rate": 9.707886239628621e-05,
+      "loss": 3.9045,
+      "step": 3805
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8233649134635925,
+      "learning_rate": 9.707105070859326e-05,
+      "loss": 3.9659,
+      "step": 3810
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8038836121559143,
+      "learning_rate": 9.706322890500579e-05,
+      "loss": 3.8256,
+      "step": 3815
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8156123757362366,
+      "learning_rate": 9.705539698720476e-05,
+      "loss": 3.8803,
+      "step": 3820
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7836065888404846,
+      "learning_rate": 9.704755495687328e-05,
+      "loss": 3.8258,
+      "step": 3825
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6912264823913574,
+      "learning_rate": 9.703970281569668e-05,
+      "loss": 3.8727,
+      "step": 3830
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7666847705841064,
+      "learning_rate": 9.70318405653624e-05,
+      "loss": 3.8726,
+      "step": 3835
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.708160400390625,
+      "learning_rate": 9.702396820756014e-05,
+      "loss": 3.8491,
+      "step": 3840
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7295490503311157,
+      "learning_rate": 9.701608574398169e-05,
+      "loss": 3.8304,
+      "step": 3845
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8284698128700256,
+      "learning_rate": 9.700819317632107e-05,
+      "loss": 3.9517,
+      "step": 3850
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7878208756446838,
+      "learning_rate": 9.700029050627443e-05,
+      "loss": 3.8495,
+      "step": 3855
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7629729509353638,
+      "learning_rate": 9.699237773554012e-05,
+      "loss": 3.92,
+      "step": 3860
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7698236703872681,
+      "learning_rate": 9.698445486581864e-05,
+      "loss": 3.8632,
+      "step": 3865
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7303275465965271,
+      "learning_rate": 9.697652189881266e-05,
+      "loss": 3.7871,
+      "step": 3870
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6592829823493958,
+      "learning_rate": 9.696857883622706e-05,
+      "loss": 3.9166,
+      "step": 3875
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7273129224777222,
+      "learning_rate": 9.696062567976885e-05,
+      "loss": 3.9,
+      "step": 3880
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7478437423706055,
+      "learning_rate": 9.695266243114718e-05,
+      "loss": 3.8547,
+      "step": 3885
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7080117464065552,
+      "learning_rate": 9.694468909207345e-05,
+      "loss": 3.8857,
+      "step": 3890
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.768325686454773,
+      "learning_rate": 9.693670566426119e-05,
+      "loss": 3.8496,
+      "step": 3895
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7026990652084351,
+      "learning_rate": 9.692871214942608e-05,
+      "loss": 4.0713,
+      "step": 3900
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8750876784324646,
+      "learning_rate": 9.692070854928596e-05,
+      "loss": 3.8697,
+      "step": 3905
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6634063720703125,
+      "learning_rate": 9.691269486556088e-05,
+      "loss": 3.9056,
+      "step": 3910
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8170890212059021,
+      "learning_rate": 9.690467109997303e-05,
+      "loss": 3.9145,
+      "step": 3915
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7423869967460632,
+      "learning_rate": 9.68966372542468e-05,
+      "loss": 3.8719,
+      "step": 3920
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7168678641319275,
+      "learning_rate": 9.688859333010866e-05,
+      "loss": 3.8551,
+      "step": 3925
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7893345952033997,
+      "learning_rate": 9.688053932928737e-05,
+      "loss": 3.8256,
+      "step": 3930
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7576113343238831,
+      "learning_rate": 9.687247525351375e-05,
+      "loss": 3.9768,
+      "step": 3935
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7799345850944519,
+      "learning_rate": 9.686440110452084e-05,
+      "loss": 3.9122,
+      "step": 3940
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7348023056983948,
+      "learning_rate": 9.685631688404384e-05,
+      "loss": 3.9425,
+      "step": 3945
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7079779505729675,
+      "learning_rate": 9.684822259382008e-05,
+      "loss": 3.8484,
+      "step": 3950
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7347009181976318,
+      "learning_rate": 9.684011823558909e-05,
+      "loss": 3.8527,
+      "step": 3955
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6870574951171875,
+      "learning_rate": 9.683200381109254e-05,
+      "loss": 3.9316,
+      "step": 3960
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7763539552688599,
+      "learning_rate": 9.68238793220743e-05,
+      "loss": 3.8645,
+      "step": 3965
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7124796509742737,
+      "learning_rate": 9.681574477028039e-05,
+      "loss": 3.8034,
+      "step": 3970
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6972880363464355,
+      "learning_rate": 9.680760015745894e-05,
+      "loss": 3.8912,
+      "step": 3975
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7260239124298096,
+      "learning_rate": 9.67994454853603e-05,
+      "loss": 3.9123,
+      "step": 3980
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7084234952926636,
+      "learning_rate": 9.679128075573698e-05,
+      "loss": 3.841,
+      "step": 3985
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7248141169548035,
+      "learning_rate": 9.678310597034364e-05,
+      "loss": 3.747,
+      "step": 3990
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6935074925422668,
+      "learning_rate": 9.677492113093706e-05,
+      "loss": 3.9803,
+      "step": 3995
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7236133813858032,
+      "learning_rate": 9.676672623927625e-05,
+      "loss": 3.8767,
+      "step": 4000
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.675078809261322,
+      "learning_rate": 9.675852129712234e-05,
+      "loss": 3.908,
+      "step": 4005
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.674927830696106,
+      "learning_rate": 9.675030630623864e-05,
+      "loss": 3.9077,
+      "step": 4010
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8050296306610107,
+      "learning_rate": 9.67420812683906e-05,
+      "loss": 3.8721,
+      "step": 4015
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7453052401542664,
+      "learning_rate": 9.673384618534582e-05,
+      "loss": 3.829,
+      "step": 4020
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8475931882858276,
+      "learning_rate": 9.672560105887409e-05,
+      "loss": 3.8638,
+      "step": 4025
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7031151652336121,
+      "learning_rate": 9.671734589074736e-05,
+      "loss": 3.8616,
+      "step": 4030
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7344084978103638,
+      "learning_rate": 9.670908068273972e-05,
+      "loss": 3.8902,
+      "step": 4035
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8092424869537354,
+      "learning_rate": 9.67008054366274e-05,
+      "loss": 3.859,
+      "step": 4040
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9113495349884033,
+      "learning_rate": 9.669252015418882e-05,
+      "loss": 3.998,
+      "step": 4045
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7960372567176819,
+      "learning_rate": 9.668422483720456e-05,
+      "loss": 3.9039,
+      "step": 4050
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.745573103427887,
+      "learning_rate": 9.66759194874573e-05,
+      "loss": 3.862,
+      "step": 4055
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7504746913909912,
+      "learning_rate": 9.666760410673195e-05,
+      "loss": 3.9103,
+      "step": 4060
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7168198823928833,
+      "learning_rate": 9.665927869681553e-05,
+      "loss": 3.8404,
+      "step": 4065
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8342441916465759,
+      "learning_rate": 9.665094325949723e-05,
+      "loss": 3.8108,
+      "step": 4070
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.733777642250061,
+      "learning_rate": 9.664259779656842e-05,
+      "loss": 3.739,
+      "step": 4075
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.769227921962738,
+      "learning_rate": 9.663424230982254e-05,
+      "loss": 3.9371,
+      "step": 4080
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.758838951587677,
+      "learning_rate": 9.66258768010553e-05,
+      "loss": 3.9342,
+      "step": 4085
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6924288272857666,
+      "learning_rate": 9.661750127206445e-05,
+      "loss": 3.828,
+      "step": 4090
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7193908095359802,
+      "learning_rate": 9.660911572465e-05,
+      "loss": 3.8929,
+      "step": 4095
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.685756266117096,
+      "learning_rate": 9.660072016061406e-05,
+      "loss": 3.8938,
+      "step": 4100
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7149713039398193,
+      "learning_rate": 9.659231458176085e-05,
+      "loss": 3.9669,
+      "step": 4105
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6959943771362305,
+      "learning_rate": 9.658389898989684e-05,
+      "loss": 3.9272,
+      "step": 4110
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6774330735206604,
+      "learning_rate": 9.657547338683055e-05,
+      "loss": 3.8376,
+      "step": 4115
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6747409105300903,
+      "learning_rate": 9.656703777437273e-05,
+      "loss": 3.8382,
+      "step": 4120
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7991278767585754,
+      "learning_rate": 9.655859215433625e-05,
+      "loss": 3.8439,
+      "step": 4125
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7032038569450378,
+      "learning_rate": 9.655013652853612e-05,
+      "loss": 3.7907,
+      "step": 4130
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6994839310646057,
+      "learning_rate": 9.654167089878952e-05,
+      "loss": 3.9135,
+      "step": 4135
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6752082109451294,
+      "learning_rate": 9.653319526691577e-05,
+      "loss": 3.841,
+      "step": 4140
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8327478766441345,
+      "learning_rate": 9.652470963473635e-05,
+      "loss": 3.9614,
+      "step": 4145
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.775684654712677,
+      "learning_rate": 9.651621400407487e-05,
+      "loss": 3.7919,
+      "step": 4150
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7279064059257507,
+      "learning_rate": 9.650770837675712e-05,
+      "loss": 3.8597,
+      "step": 4155
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.4536027908325195,
+      "learning_rate": 9.649919275461099e-05,
+      "loss": 3.8139,
+      "step": 4160
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6802389621734619,
+      "learning_rate": 9.649066713946655e-05,
+      "loss": 3.8368,
+      "step": 4165
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7550429105758667,
+      "learning_rate": 9.648213153315603e-05,
+      "loss": 3.8408,
+      "step": 4170
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6647107601165771,
+      "learning_rate": 9.647358593751378e-05,
+      "loss": 3.8765,
+      "step": 4175
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6905397772789001,
+      "learning_rate": 9.646503035437633e-05,
+      "loss": 3.7738,
+      "step": 4180
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7378332018852234,
+      "learning_rate": 9.64564647855823e-05,
+      "loss": 3.8695,
+      "step": 4185
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7089004516601562,
+      "learning_rate": 9.644788923297253e-05,
+      "loss": 3.8001,
+      "step": 4190
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6580518484115601,
+      "learning_rate": 9.64393036983899e-05,
+      "loss": 3.8816,
+      "step": 4195
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7614989280700684,
+      "learning_rate": 9.643070818367957e-05,
+      "loss": 3.9159,
+      "step": 4200
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7648181915283203,
+      "learning_rate": 9.642210269068874e-05,
+      "loss": 3.8602,
+      "step": 4205
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7083746790885925,
+      "learning_rate": 9.641348722126679e-05,
+      "loss": 3.9299,
+      "step": 4210
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8207822442054749,
+      "learning_rate": 9.640486177726527e-05,
+      "loss": 3.9205,
+      "step": 4215
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8256579637527466,
+      "learning_rate": 9.639622636053781e-05,
+      "loss": 3.9224,
+      "step": 4220
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6737625598907471,
+      "learning_rate": 9.638758097294024e-05,
+      "loss": 3.7932,
+      "step": 4225
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7057338953018188,
+      "learning_rate": 9.637892561633054e-05,
+      "loss": 3.8497,
+      "step": 4230
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6843653917312622,
+      "learning_rate": 9.637026029256874e-05,
+      "loss": 3.8726,
+      "step": 4235
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.724190890789032,
+      "learning_rate": 9.636158500351713e-05,
+      "loss": 3.7906,
+      "step": 4240
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7710651755332947,
+      "learning_rate": 9.635289975104009e-05,
+      "loss": 4.0024,
+      "step": 4245
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.728609561920166,
+      "learning_rate": 9.634420453700413e-05,
+      "loss": 3.8899,
+      "step": 4250
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7018215656280518,
+      "learning_rate": 9.633549936327789e-05,
+      "loss": 3.9028,
+      "step": 4255
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7313151955604553,
+      "learning_rate": 9.63267842317322e-05,
+      "loss": 3.9632,
+      "step": 4260
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8182834982872009,
+      "learning_rate": 9.631805914424e-05,
+      "loss": 3.8936,
+      "step": 4265
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7878748178482056,
+      "learning_rate": 9.630932410267637e-05,
+      "loss": 3.9133,
+      "step": 4270
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8007236123085022,
+      "learning_rate": 9.630057910891852e-05,
+      "loss": 3.807,
+      "step": 4275
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0376439094543457,
+      "learning_rate": 9.629182416484582e-05,
+      "loss": 3.7731,
+      "step": 4280
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6923347115516663,
+      "learning_rate": 9.628305927233975e-05,
+      "loss": 3.9031,
+      "step": 4285
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7408250570297241,
+      "learning_rate": 9.627428443328397e-05,
+      "loss": 3.9418,
+      "step": 4290
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7062495946884155,
+      "learning_rate": 9.626549964956424e-05,
+      "loss": 3.779,
+      "step": 4295
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7415155172348022,
+      "learning_rate": 9.625670492306848e-05,
+      "loss": 3.9259,
+      "step": 4300
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8561193346977234,
+      "learning_rate": 9.624790025568673e-05,
+      "loss": 3.7927,
+      "step": 4305
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7086501717567444,
+      "learning_rate": 9.623908564931117e-05,
+      "loss": 3.8821,
+      "step": 4310
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.716729462146759,
+      "learning_rate": 9.623026110583612e-05,
+      "loss": 3.8511,
+      "step": 4315
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9607390761375427,
+      "learning_rate": 9.622142662715808e-05,
+      "loss": 3.8236,
+      "step": 4320
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7795659899711609,
+      "learning_rate": 9.621258221517556e-05,
+      "loss": 3.912,
+      "step": 4325
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6799745559692383,
+      "learning_rate": 9.620372787178934e-05,
+      "loss": 3.8052,
+      "step": 4330
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7117394208908081,
+      "learning_rate": 9.619486359890228e-05,
+      "loss": 3.6378,
+      "step": 4335
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7361997961997986,
+      "learning_rate": 9.618598939841934e-05,
+      "loss": 3.9453,
+      "step": 4340
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6551253795623779,
+      "learning_rate": 9.617710527224768e-05,
+      "loss": 3.7064,
+      "step": 4345
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7140074372291565,
+      "learning_rate": 9.616821122229656e-05,
+      "loss": 3.8003,
+      "step": 4350
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6917895078659058,
+      "learning_rate": 9.615930725047732e-05,
+      "loss": 3.98,
+      "step": 4355
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6637527346611023,
+      "learning_rate": 9.615039335870357e-05,
+      "loss": 3.9137,
+      "step": 4360
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7833606600761414,
+      "learning_rate": 9.614146954889091e-05,
+      "loss": 3.7354,
+      "step": 4365
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7077828049659729,
+      "learning_rate": 9.613253582295714e-05,
+      "loss": 3.7939,
+      "step": 4370
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7515588998794556,
+      "learning_rate": 9.612359218282216e-05,
+      "loss": 3.9507,
+      "step": 4375
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6966031789779663,
+      "learning_rate": 9.611463863040807e-05,
+      "loss": 3.814,
+      "step": 4380
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.752215564250946,
+      "learning_rate": 9.610567516763902e-05,
+      "loss": 3.7884,
+      "step": 4385
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7231920957565308,
+      "learning_rate": 9.60967017964413e-05,
+      "loss": 3.8404,
+      "step": 4390
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6622960567474365,
+      "learning_rate": 9.608771851874338e-05,
+      "loss": 3.7215,
+      "step": 4395
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.760969340801239,
+      "learning_rate": 9.607872533647584e-05,
+      "loss": 3.8959,
+      "step": 4400
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6574525237083435,
+      "learning_rate": 9.606972225157134e-05,
+      "loss": 3.7066,
+      "step": 4405
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7396848201751709,
+      "learning_rate": 9.606070926596471e-05,
+      "loss": 3.8061,
+      "step": 4410
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.78935307264328,
+      "learning_rate": 9.605168638159294e-05,
+      "loss": 3.7823,
+      "step": 4415
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.715259313583374,
+      "learning_rate": 9.604265360039507e-05,
+      "loss": 3.8308,
+      "step": 4420
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7350407242774963,
+      "learning_rate": 9.603361092431234e-05,
+      "loss": 3.8608,
+      "step": 4425
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7217309474945068,
+      "learning_rate": 9.602455835528804e-05,
+      "loss": 3.7426,
+      "step": 4430
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7091699838638306,
+      "learning_rate": 9.601549589526767e-05,
+      "loss": 3.8374,
+      "step": 4435
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7744762301445007,
+      "learning_rate": 9.60064235461988e-05,
+      "loss": 3.8806,
+      "step": 4440
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7336158752441406,
+      "learning_rate": 9.599734131003114e-05,
+      "loss": 3.7784,
+      "step": 4445
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7141445875167847,
+      "learning_rate": 9.598824918871654e-05,
+      "loss": 3.8514,
+      "step": 4450
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7358189821243286,
+      "learning_rate": 9.597914718420892e-05,
+      "loss": 3.8428,
+      "step": 4455
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7002937197685242,
+      "learning_rate": 9.597003529846442e-05,
+      "loss": 3.849,
+      "step": 4460
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6829643249511719,
+      "learning_rate": 9.59609135334412e-05,
+      "loss": 3.7578,
+      "step": 4465
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6991536617279053,
+      "learning_rate": 9.595178189109963e-05,
+      "loss": 3.815,
+      "step": 4470
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7061537504196167,
+      "learning_rate": 9.594264037340212e-05,
+      "loss": 3.8289,
+      "step": 4475
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7507505416870117,
+      "learning_rate": 9.593348898231326e-05,
+      "loss": 3.8543,
+      "step": 4480
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7783147692680359,
+      "learning_rate": 9.592432771979978e-05,
+      "loss": 3.8181,
+      "step": 4485
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7291918396949768,
+      "learning_rate": 9.591515658783046e-05,
+      "loss": 3.7758,
+      "step": 4490
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7302814722061157,
+      "learning_rate": 9.590597558837626e-05,
+      "loss": 3.9115,
+      "step": 4495
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6841729283332825,
+      "learning_rate": 9.589678472341023e-05,
+      "loss": 3.8074,
+      "step": 4500
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.761920154094696,
+      "learning_rate": 9.588758399490755e-05,
+      "loss": 3.8112,
+      "step": 4505
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7996997833251953,
+      "learning_rate": 9.587837340484553e-05,
+      "loss": 3.8731,
+      "step": 4510
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.793246865272522,
+      "learning_rate": 9.586915295520358e-05,
+      "loss": 3.7586,
+      "step": 4515
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7111709713935852,
+      "learning_rate": 9.585992264796324e-05,
+      "loss": 3.8043,
+      "step": 4520
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7443453669548035,
+      "learning_rate": 9.585068248510817e-05,
+      "loss": 3.9087,
+      "step": 4525
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7052918672561646,
+      "learning_rate": 9.584143246862414e-05,
+      "loss": 3.9084,
+      "step": 4530
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7016591429710388,
+      "learning_rate": 9.583217260049907e-05,
+      "loss": 3.7758,
+      "step": 4535
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8275299072265625,
+      "learning_rate": 9.582290288272291e-05,
+      "loss": 3.8121,
+      "step": 4540
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7940973043441772,
+      "learning_rate": 9.581362331728784e-05,
+      "loss": 3.8508,
+      "step": 4545
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8400365114212036,
+      "learning_rate": 9.580433390618808e-05,
+      "loss": 3.8205,
+      "step": 4550
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.733667254447937,
+      "learning_rate": 9.579503465142e-05,
+      "loss": 3.8629,
+      "step": 4555
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6928481459617615,
+      "learning_rate": 9.578572555498207e-05,
+      "loss": 3.719,
+      "step": 4560
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8230397701263428,
+      "learning_rate": 9.577640661887488e-05,
+      "loss": 3.8193,
+      "step": 4565
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7690085768699646,
+      "learning_rate": 9.576707784510113e-05,
+      "loss": 3.8977,
+      "step": 4570
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6714286804199219,
+      "learning_rate": 9.575773923566563e-05,
+      "loss": 3.9309,
+      "step": 4575
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.754267156124115,
+      "learning_rate": 9.574839079257532e-05,
+      "loss": 3.7256,
+      "step": 4580
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6660600304603577,
+      "learning_rate": 9.573903251783925e-05,
+      "loss": 3.7103,
+      "step": 4585
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6558008790016174,
+      "learning_rate": 9.572966441346858e-05,
+      "loss": 3.7771,
+      "step": 4590
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6764190196990967,
+      "learning_rate": 9.572028648147657e-05,
+      "loss": 3.8102,
+      "step": 4595
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6661336421966553,
+      "learning_rate": 9.571089872387862e-05,
+      "loss": 3.7945,
+      "step": 4600
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7078689932823181,
+      "learning_rate": 9.57015011426922e-05,
+      "loss": 3.899,
+      "step": 4605
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.682138204574585,
+      "learning_rate": 9.569209373993693e-05,
+      "loss": 3.7664,
+      "step": 4610
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7368147969245911,
+      "learning_rate": 9.568267651763451e-05,
+      "loss": 3.7355,
+      "step": 4615
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6613869667053223,
+      "learning_rate": 9.56732494778088e-05,
+      "loss": 3.796,
+      "step": 4620
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6816684603691101,
+      "learning_rate": 9.56638126224857e-05,
+      "loss": 3.9283,
+      "step": 4625
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7001058459281921,
+      "learning_rate": 9.565436595369328e-05,
+      "loss": 3.9135,
+      "step": 4630
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7152508497238159,
+      "learning_rate": 9.564490947346168e-05,
+      "loss": 3.8025,
+      "step": 4635
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7125628590583801,
+      "learning_rate": 9.563544318382316e-05,
+      "loss": 3.8678,
+      "step": 4640
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7804508805274963,
+      "learning_rate": 9.56259670868121e-05,
+      "loss": 3.8406,
+      "step": 4645
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7251895666122437,
+      "learning_rate": 9.5616481184465e-05,
+      "loss": 3.7333,
+      "step": 4650
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6744594573974609,
+      "learning_rate": 9.560698547882041e-05,
+      "loss": 3.785,
+      "step": 4655
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7523850202560425,
+      "learning_rate": 9.559747997191903e-05,
+      "loss": 3.7934,
+      "step": 4660
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7060679197311401,
+      "learning_rate": 9.558796466580366e-05,
+      "loss": 3.8355,
+      "step": 4665
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7760443687438965,
+      "learning_rate": 9.557843956251924e-05,
+      "loss": 3.7943,
+      "step": 4670
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7123631238937378,
+      "learning_rate": 9.556890466411272e-05,
+      "loss": 3.8073,
+      "step": 4675
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6738618612289429,
+      "learning_rate": 9.555935997263328e-05,
+      "loss": 3.849,
+      "step": 4680
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6865014433860779,
+      "learning_rate": 9.554980549013209e-05,
+      "loss": 3.9213,
+      "step": 4685
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6967694163322449,
+      "learning_rate": 9.554024121866249e-05,
+      "loss": 3.8677,
+      "step": 4690
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7120850682258606,
+      "learning_rate": 9.55306671602799e-05,
+      "loss": 3.7283,
+      "step": 4695
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7586105465888977,
+      "learning_rate": 9.552108331704189e-05,
+      "loss": 3.8766,
+      "step": 4700
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.72507244348526,
+      "learning_rate": 9.551148969100804e-05,
+      "loss": 3.8401,
+      "step": 4705
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7157936692237854,
+      "learning_rate": 9.550188628424012e-05,
+      "loss": 3.7902,
+      "step": 4710
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6995603442192078,
+      "learning_rate": 9.549227309880197e-05,
+      "loss": 3.7569,
+      "step": 4715
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6488493084907532,
+      "learning_rate": 9.548265013675951e-05,
+      "loss": 3.7815,
+      "step": 4720
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7236054539680481,
+      "learning_rate": 9.54730174001808e-05,
+      "loss": 3.8263,
+      "step": 4725
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6475001573562622,
+      "learning_rate": 9.546337489113597e-05,
+      "loss": 3.7561,
+      "step": 4730
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6639286279678345,
+      "learning_rate": 9.545372261169725e-05,
+      "loss": 3.6599,
+      "step": 4735
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6827556490898132,
+      "learning_rate": 9.544406056393903e-05,
+      "loss": 3.7491,
+      "step": 4740
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6949857473373413,
+      "learning_rate": 9.54343887499377e-05,
+      "loss": 3.8477,
+      "step": 4745
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7333765029907227,
+      "learning_rate": 9.542470717177182e-05,
+      "loss": 3.718,
+      "step": 4750
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6635848879814148,
+      "learning_rate": 9.541501583152204e-05,
+      "loss": 3.7922,
+      "step": 4755
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.722284734249115,
+      "learning_rate": 9.540531473127106e-05,
+      "loss": 3.8281,
+      "step": 4760
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6738249659538269,
+      "learning_rate": 9.539560387310375e-05,
+      "loss": 3.7978,
+      "step": 4765
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6769446134567261,
+      "learning_rate": 9.538588325910702e-05,
+      "loss": 3.8587,
+      "step": 4770
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0422422885894775,
+      "learning_rate": 9.53761528913699e-05,
+      "loss": 3.7942,
+      "step": 4775
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7257570028305054,
+      "learning_rate": 9.536641277198353e-05,
+      "loss": 3.7185,
+      "step": 4780
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.779447615146637,
+      "learning_rate": 9.535666290304111e-05,
+      "loss": 3.852,
+      "step": 4785
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7202276587486267,
+      "learning_rate": 9.534690328663796e-05,
+      "loss": 3.9197,
+      "step": 4790
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9228225946426392,
+      "learning_rate": 9.533713392487149e-05,
+      "loss": 3.7899,
+      "step": 4795
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9983485341072083,
+      "learning_rate": 9.532735481984121e-05,
+      "loss": 3.8182,
+      "step": 4800
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7186928391456604,
+      "learning_rate": 9.53175659736487e-05,
+      "loss": 3.8605,
+      "step": 4805
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.76997971534729,
+      "learning_rate": 9.530776738839766e-05,
+      "loss": 3.8844,
+      "step": 4810
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.745994508266449,
+      "learning_rate": 9.529795906619387e-05,
+      "loss": 3.8006,
+      "step": 4815
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7213228940963745,
+      "learning_rate": 9.528814100914522e-05,
+      "loss": 3.7319,
+      "step": 4820
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7529492974281311,
+      "learning_rate": 9.52783132193617e-05,
+      "loss": 3.8047,
+      "step": 4825
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7520737648010254,
+      "learning_rate": 9.52684756989553e-05,
+      "loss": 3.7435,
+      "step": 4830
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7191484570503235,
+      "learning_rate": 9.525862845004022e-05,
+      "loss": 3.7928,
+      "step": 4835
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6883547902107239,
+      "learning_rate": 9.52487714747327e-05,
+      "loss": 3.7359,
+      "step": 4840
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.804142415523529,
+      "learning_rate": 9.523890477515105e-05,
+      "loss": 3.9449,
+      "step": 4845
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.03458571434021,
+      "learning_rate": 9.522902835341573e-05,
+      "loss": 3.8129,
+      "step": 4850
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6640298366546631,
+      "learning_rate": 9.521914221164922e-05,
+      "loss": 3.7506,
+      "step": 4855
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6912886500358582,
+      "learning_rate": 9.520924635197612e-05,
+      "loss": 3.7763,
+      "step": 4860
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6663967370986938,
+      "learning_rate": 9.519934077652312e-05,
+      "loss": 3.7707,
+      "step": 4865
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6560251116752625,
+      "learning_rate": 9.518942548741902e-05,
+      "loss": 3.7992,
+      "step": 4870
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7256091833114624,
+      "learning_rate": 9.517950048679467e-05,
+      "loss": 3.7044,
+      "step": 4875
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7679279446601868,
+      "learning_rate": 9.516956577678302e-05,
+      "loss": 3.7959,
+      "step": 4880
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6698256134986877,
+      "learning_rate": 9.51596213595191e-05,
+      "loss": 3.8403,
+      "step": 4885
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8171987533569336,
+      "learning_rate": 9.514966723714004e-05,
+      "loss": 3.7708,
+      "step": 4890
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7046585083007812,
+      "learning_rate": 9.513970341178506e-05,
+      "loss": 3.7214,
+      "step": 4895
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6659003496170044,
+      "learning_rate": 9.512972988559545e-05,
+      "loss": 3.8401,
+      "step": 4900
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6867298483848572,
+      "learning_rate": 9.511974666071458e-05,
+      "loss": 3.8051,
+      "step": 4905
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6445996165275574,
+      "learning_rate": 9.510975373928792e-05,
+      "loss": 3.7971,
+      "step": 4910
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6604128479957581,
+      "learning_rate": 9.509975112346303e-05,
+      "loss": 3.6672,
+      "step": 4915
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.703145444393158,
+      "learning_rate": 9.508973881538952e-05,
+      "loss": 3.6926,
+      "step": 4920
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6997295618057251,
+      "learning_rate": 9.507971681721911e-05,
+      "loss": 3.8285,
+      "step": 4925
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6642308235168457,
+      "learning_rate": 9.50696851311056e-05,
+      "loss": 3.7823,
+      "step": 4930
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6974987387657166,
+      "learning_rate": 9.505964375920489e-05,
+      "loss": 3.7897,
+      "step": 4935
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6724978089332581,
+      "learning_rate": 9.50495927036749e-05,
+      "loss": 3.7802,
+      "step": 4940
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6299294829368591,
+      "learning_rate": 9.503953196667569e-05,
+      "loss": 3.7741,
+      "step": 4945
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6976083517074585,
+      "learning_rate": 9.502946155036938e-05,
+      "loss": 3.7183,
+      "step": 4950
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7229387164115906,
+      "learning_rate": 9.50193814569202e-05,
+      "loss": 3.7343,
+      "step": 4955
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7873827219009399,
+      "learning_rate": 9.500929168849439e-05,
+      "loss": 3.7238,
+      "step": 4960
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6839877963066101,
+      "learning_rate": 9.499919224726032e-05,
+      "loss": 3.7285,
+      "step": 4965
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6798027157783508,
+      "learning_rate": 9.498908313538846e-05,
+      "loss": 3.9365,
+      "step": 4970
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7304661273956299,
+      "learning_rate": 9.49789643550513e-05,
+      "loss": 3.9386,
+      "step": 4975
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6882640719413757,
+      "learning_rate": 9.496883590842343e-05,
+      "loss": 3.7294,
+      "step": 4980
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6708075404167175,
+      "learning_rate": 9.495869779768154e-05,
+      "loss": 3.9011,
+      "step": 4985
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.653045117855072,
+      "learning_rate": 9.494855002500438e-05,
+      "loss": 3.7957,
+      "step": 4990
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7089905738830566,
+      "learning_rate": 9.493839259257278e-05,
+      "loss": 3.7687,
+      "step": 4995
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8092687726020813,
+      "learning_rate": 9.492822550256962e-05,
+      "loss": 3.7198,
+      "step": 5000
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7344354391098022,
+      "learning_rate": 9.49180487571799e-05,
+      "loss": 3.7549,
+      "step": 5005
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7380156517028809,
+      "learning_rate": 9.490786235859066e-05,
+      "loss": 3.7617,
+      "step": 5010
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6712434887886047,
+      "learning_rate": 9.489766630899104e-05,
+      "loss": 3.8334,
+      "step": 5015
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6558536887168884,
+      "learning_rate": 9.488746061057223e-05,
+      "loss": 3.7293,
+      "step": 5020
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7083954811096191,
+      "learning_rate": 9.487724526552753e-05,
+      "loss": 3.7604,
+      "step": 5025
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7317386269569397,
+      "learning_rate": 9.486702027605224e-05,
+      "loss": 3.698,
+      "step": 5030
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6877592206001282,
+      "learning_rate": 9.485678564434381e-05,
+      "loss": 3.7478,
+      "step": 5035
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7440779209136963,
+      "learning_rate": 9.484654137260176e-05,
+      "loss": 3.8109,
+      "step": 5040
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.789886474609375,
+      "learning_rate": 9.483628746302761e-05,
+      "loss": 3.8197,
+      "step": 5045
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7372972965240479,
+      "learning_rate": 9.482602391782501e-05,
+      "loss": 3.742,
+      "step": 5050
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7077298164367676,
+      "learning_rate": 9.481575073919968e-05,
+      "loss": 3.7199,
+      "step": 5055
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6764664053916931,
+      "learning_rate": 9.480546792935939e-05,
+      "loss": 3.6733,
+      "step": 5060
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6339630484580994,
+      "learning_rate": 9.479517549051398e-05,
+      "loss": 3.6975,
+      "step": 5065
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7570455074310303,
+      "learning_rate": 9.478487342487536e-05,
+      "loss": 3.8122,
+      "step": 5070
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.663226306438446,
+      "learning_rate": 9.477456173465754e-05,
+      "loss": 3.7568,
+      "step": 5075
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7077811360359192,
+      "learning_rate": 9.476424042207655e-05,
+      "loss": 3.8852,
+      "step": 5080
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6552476286888123,
+      "learning_rate": 9.475390948935054e-05,
+      "loss": 3.7769,
+      "step": 5085
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.68292236328125,
+      "learning_rate": 9.474356893869967e-05,
+      "loss": 3.6778,
+      "step": 5090
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6603934168815613,
+      "learning_rate": 9.47332187723462e-05,
+      "loss": 3.7477,
+      "step": 5095
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6220848560333252,
+      "learning_rate": 9.472285899251445e-05,
+      "loss": 3.6978,
+      "step": 5100
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6640552878379822,
+      "learning_rate": 9.471248960143085e-05,
+      "loss": 3.7654,
+      "step": 5105
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6455395817756653,
+      "learning_rate": 9.47021106013238e-05,
+      "loss": 3.6803,
+      "step": 5110
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6592692136764526,
+      "learning_rate": 9.469172199442383e-05,
+      "loss": 3.8701,
+      "step": 5115
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8167845010757446,
+      "learning_rate": 9.468132378296355e-05,
+      "loss": 3.8234,
+      "step": 5120
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6505124568939209,
+      "learning_rate": 9.467091596917758e-05,
+      "loss": 3.743,
+      "step": 5125
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6753003597259521,
+      "learning_rate": 9.466049855530265e-05,
+      "loss": 3.7967,
+      "step": 5130
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6740750074386597,
+      "learning_rate": 9.465007154357752e-05,
+      "loss": 3.826,
+      "step": 5135
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6580204963684082,
+      "learning_rate": 9.463963493624304e-05,
+      "loss": 3.7344,
+      "step": 5140
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6611657738685608,
+      "learning_rate": 9.462918873554209e-05,
+      "loss": 3.7983,
+      "step": 5145
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7158042192459106,
+      "learning_rate": 9.461873294371965e-05,
+      "loss": 3.8223,
+      "step": 5150
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7062550783157349,
+      "learning_rate": 9.460826756302275e-05,
+      "loss": 3.7927,
+      "step": 5155
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.700351357460022,
+      "learning_rate": 9.459779259570045e-05,
+      "loss": 3.7337,
+      "step": 5160
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7420654296875,
+      "learning_rate": 9.458730804400389e-05,
+      "loss": 3.6803,
+      "step": 5165
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6943119764328003,
+      "learning_rate": 9.457681391018629e-05,
+      "loss": 3.7614,
+      "step": 5170
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7723524570465088,
+      "learning_rate": 9.45663101965029e-05,
+      "loss": 3.7148,
+      "step": 5175
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6643696427345276,
+      "learning_rate": 9.455579690521106e-05,
+      "loss": 3.8314,
+      "step": 5180
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6814690232276917,
+      "learning_rate": 9.454527403857013e-05,
+      "loss": 3.6093,
+      "step": 5185
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7119035124778748,
+      "learning_rate": 9.453474159884155e-05,
+      "loss": 3.7551,
+      "step": 5190
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6340065002441406,
+      "learning_rate": 9.452419958828883e-05,
+      "loss": 3.7059,
+      "step": 5195
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6689919829368591,
+      "learning_rate": 9.45136480091775e-05,
+      "loss": 3.6952,
+      "step": 5200
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6336925625801086,
+      "learning_rate": 9.450308686377516e-05,
+      "loss": 3.6254,
+      "step": 5205
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6840919852256775,
+      "learning_rate": 9.449251615435151e-05,
+      "loss": 3.7864,
+      "step": 5210
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7399453520774841,
+      "learning_rate": 9.448193588317822e-05,
+      "loss": 3.7458,
+      "step": 5215
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6756210327148438,
+      "learning_rate": 9.447134605252912e-05,
+      "loss": 3.7613,
+      "step": 5220
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6753889918327332,
+      "learning_rate": 9.446074666468e-05,
+      "loss": 3.7607,
+      "step": 5225
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.731918454170227,
+      "learning_rate": 9.445013772190874e-05,
+      "loss": 3.6653,
+      "step": 5230
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6321896910667419,
+      "learning_rate": 9.443951922649529e-05,
+      "loss": 3.737,
+      "step": 5235
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7150954008102417,
+      "learning_rate": 9.442889118072162e-05,
+      "loss": 3.8176,
+      "step": 5240
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6617087721824646,
+      "learning_rate": 9.441825358687181e-05,
+      "loss": 3.6793,
+      "step": 5245
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6665310263633728,
+      "learning_rate": 9.440760644723192e-05,
+      "loss": 3.746,
+      "step": 5250
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6749745011329651,
+      "learning_rate": 9.439694976409008e-05,
+      "loss": 3.669,
+      "step": 5255
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7269194722175598,
+      "learning_rate": 9.438628353973653e-05,
+      "loss": 3.6424,
+      "step": 5260
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.060640573501587,
+      "learning_rate": 9.437560777646348e-05,
+      "loss": 3.699,
+      "step": 5265
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6673661470413208,
+      "learning_rate": 9.436492247656523e-05,
+      "loss": 3.6786,
+      "step": 5270
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.781356930732727,
+      "learning_rate": 9.435422764233814e-05,
+      "loss": 3.6595,
+      "step": 5275
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6430137753486633,
+      "learning_rate": 9.43435232760806e-05,
+      "loss": 3.6186,
+      "step": 5280
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7284837365150452,
+      "learning_rate": 9.433280938009304e-05,
+      "loss": 3.7037,
+      "step": 5285
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7333751916885376,
+      "learning_rate": 9.432208595667797e-05,
+      "loss": 3.8359,
+      "step": 5290
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6739276051521301,
+      "learning_rate": 9.431135300813989e-05,
+      "loss": 3.6563,
+      "step": 5295
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6965942978858948,
+      "learning_rate": 9.430061053678542e-05,
+      "loss": 3.92,
+      "step": 5300
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7025936841964722,
+      "learning_rate": 9.42898585449232e-05,
+      "loss": 3.6049,
+      "step": 5305
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7284041047096252,
+      "learning_rate": 9.427909703486388e-05,
+      "loss": 3.7023,
+      "step": 5310
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7397528886795044,
+      "learning_rate": 9.426832600892019e-05,
+      "loss": 3.7618,
+      "step": 5315
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7659059762954712,
+      "learning_rate": 9.425754546940691e-05,
+      "loss": 3.7052,
+      "step": 5320
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7439866065979004,
+      "learning_rate": 9.424675541864081e-05,
+      "loss": 3.6557,
+      "step": 5325
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6896258592605591,
+      "learning_rate": 9.423595585894082e-05,
+      "loss": 3.6258,
+      "step": 5330
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.680923581123352,
+      "learning_rate": 9.422514679262778e-05,
+      "loss": 3.8763,
+      "step": 5335
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6358628273010254,
+      "learning_rate": 9.421432822202465e-05,
+      "loss": 3.6749,
+      "step": 5340
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6697905659675598,
+      "learning_rate": 9.420350014945643e-05,
+      "loss": 3.7416,
+      "step": 5345
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.657863974571228,
+      "learning_rate": 9.419266257725013e-05,
+      "loss": 3.7824,
+      "step": 5350
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.721548318862915,
+      "learning_rate": 9.418181550773483e-05,
+      "loss": 3.6817,
+      "step": 5355
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6709536910057068,
+      "learning_rate": 9.417095894324163e-05,
+      "loss": 3.7236,
+      "step": 5360
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6940340995788574,
+      "learning_rate": 9.41600928861037e-05,
+      "loss": 3.691,
+      "step": 5365
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6462624669075012,
+      "learning_rate": 9.41492173386562e-05,
+      "loss": 3.7399,
+      "step": 5370
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8651660084724426,
+      "learning_rate": 9.413833230323639e-05,
+      "loss": 3.6692,
+      "step": 5375
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7003412842750549,
+      "learning_rate": 9.412743778218351e-05,
+      "loss": 3.7993,
+      "step": 5380
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6482155323028564,
+      "learning_rate": 9.41165337778389e-05,
+      "loss": 3.8455,
+      "step": 5385
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.647760272026062,
+      "learning_rate": 9.410562029254589e-05,
+      "loss": 3.7214,
+      "step": 5390
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7166098952293396,
+      "learning_rate": 9.409469732864986e-05,
+      "loss": 3.8174,
+      "step": 5395
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6635168790817261,
+      "learning_rate": 9.408376488849824e-05,
+      "loss": 3.6971,
+      "step": 5400
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6379615664482117,
+      "learning_rate": 9.407282297444048e-05,
+      "loss": 3.6306,
+      "step": 5405
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7536595463752747,
+      "learning_rate": 9.406187158882808e-05,
+      "loss": 3.7222,
+      "step": 5410
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7054377198219299,
+      "learning_rate": 9.405091073401458e-05,
+      "loss": 3.6837,
+      "step": 5415
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7000157833099365,
+      "learning_rate": 9.40399404123555e-05,
+      "loss": 3.6044,
+      "step": 5420
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6436945796012878,
+      "learning_rate": 9.402896062620848e-05,
+      "loss": 3.8125,
+      "step": 5425
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.677832841873169,
+      "learning_rate": 9.401797137793313e-05,
+      "loss": 3.7887,
+      "step": 5430
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7575086951255798,
+      "learning_rate": 9.400697266989112e-05,
+      "loss": 3.7774,
+      "step": 5435
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6609094738960266,
+      "learning_rate": 9.399596450444617e-05,
+      "loss": 3.6961,
+      "step": 5440
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6928625702857971,
+      "learning_rate": 9.398494688396398e-05,
+      "loss": 3.7726,
+      "step": 5445
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.656806230545044,
+      "learning_rate": 9.397391981081232e-05,
+      "loss": 3.7878,
+      "step": 5450
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6309849619865417,
+      "learning_rate": 9.3962883287361e-05,
+      "loss": 3.8181,
+      "step": 5455
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6577248573303223,
+      "learning_rate": 9.395183731598183e-05,
+      "loss": 3.7469,
+      "step": 5460
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7005400061607361,
+      "learning_rate": 9.394078189904867e-05,
+      "loss": 3.6001,
+      "step": 5465
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8009021878242493,
+      "learning_rate": 9.392971703893741e-05,
+      "loss": 3.7326,
+      "step": 5470
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9077003598213196,
+      "learning_rate": 9.391864273802595e-05,
+      "loss": 3.7162,
+      "step": 5475
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6839208006858826,
+      "learning_rate": 9.390755899869425e-05,
+      "loss": 3.6893,
+      "step": 5480
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6394956707954407,
+      "learning_rate": 9.38964658233243e-05,
+      "loss": 3.7081,
+      "step": 5485
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6393072605133057,
+      "learning_rate": 9.388536321430005e-05,
+      "loss": 3.7715,
+      "step": 5490
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6483945250511169,
+      "learning_rate": 9.387425117400757e-05,
+      "loss": 3.7239,
+      "step": 5495
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7142335772514343,
+      "learning_rate": 9.386312970483489e-05,
+      "loss": 3.6885,
+      "step": 5500
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6611399054527283,
+      "learning_rate": 9.385199880917211e-05,
+      "loss": 3.5961,
+      "step": 5505
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6243033409118652,
+      "learning_rate": 9.384085848941132e-05,
+      "loss": 3.6586,
+      "step": 5510
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6203659176826477,
+      "learning_rate": 9.382970874794666e-05,
+      "loss": 3.7132,
+      "step": 5515
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6120381355285645,
+      "learning_rate": 9.381854958717429e-05,
+      "loss": 3.8567,
+      "step": 5520
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6671572923660278,
+      "learning_rate": 9.380738100949238e-05,
+      "loss": 3.6895,
+      "step": 5525
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6535508036613464,
+      "learning_rate": 9.379620301730114e-05,
+      "loss": 3.7317,
+      "step": 5530
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6301713585853577,
+      "learning_rate": 9.378501561300282e-05,
+      "loss": 3.684,
+      "step": 5535
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6375166773796082,
+      "learning_rate": 9.377381879900162e-05,
+      "loss": 3.8229,
+      "step": 5540
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6607962846755981,
+      "learning_rate": 9.376261257770384e-05,
+      "loss": 3.7,
+      "step": 5545
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6922563314437866,
+      "learning_rate": 9.37513969515178e-05,
+      "loss": 3.7506,
+      "step": 5550
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.708430826663971,
+      "learning_rate": 9.374017192285378e-05,
+      "loss": 3.7007,
+      "step": 5555
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6569916605949402,
+      "learning_rate": 9.372893749412413e-05,
+      "loss": 3.674,
+      "step": 5560
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.661085844039917,
+      "learning_rate": 9.371769366774321e-05,
+      "loss": 3.8167,
+      "step": 5565
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6912073493003845,
+      "learning_rate": 9.370644044612738e-05,
+      "loss": 3.7473,
+      "step": 5570
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6510387063026428,
+      "learning_rate": 9.369517783169505e-05,
+      "loss": 3.6655,
+      "step": 5575
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6248131990432739,
+      "learning_rate": 9.368390582686663e-05,
+      "loss": 3.692,
+      "step": 5580
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7090110778808594,
+      "learning_rate": 9.367262443406456e-05,
+      "loss": 3.7887,
+      "step": 5585
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.757709801197052,
+      "learning_rate": 9.366133365571328e-05,
+      "loss": 3.805,
+      "step": 5590
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.650056779384613,
+      "learning_rate": 9.365003349423924e-05,
+      "loss": 3.7135,
+      "step": 5595
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6473081707954407,
+      "learning_rate": 9.363872395207097e-05,
+      "loss": 3.5655,
+      "step": 5600
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6282681822776794,
+      "learning_rate": 9.362740503163891e-05,
+      "loss": 3.7753,
+      "step": 5605
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.650528073310852,
+      "learning_rate": 9.361607673537563e-05,
+      "loss": 3.7316,
+      "step": 5610
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8883610963821411,
+      "learning_rate": 9.36047390657156e-05,
+      "loss": 3.5922,
+      "step": 5615
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.690132200717926,
+      "learning_rate": 9.359339202509543e-05,
+      "loss": 3.7161,
+      "step": 5620
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.902965784072876,
+      "learning_rate": 9.358203561595363e-05,
+      "loss": 3.7133,
+      "step": 5625
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8393006324768066,
+      "learning_rate": 9.357066984073079e-05,
+      "loss": 3.6961,
+      "step": 5630
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.756287157535553,
+      "learning_rate": 9.355929470186948e-05,
+      "loss": 3.7679,
+      "step": 5635
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6115703582763672,
+      "learning_rate": 9.354791020181431e-05,
+      "loss": 3.7173,
+      "step": 5640
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.608020544052124,
+      "learning_rate": 9.353651634301188e-05,
+      "loss": 3.7507,
+      "step": 5645
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.661238431930542,
+      "learning_rate": 9.352511312791083e-05,
+      "loss": 3.699,
+      "step": 5650
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6859521269798279,
+      "learning_rate": 9.351370055896175e-05,
+      "loss": 3.7729,
+      "step": 5655
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6613646149635315,
+      "learning_rate": 9.350227863861732e-05,
+      "loss": 3.7005,
+      "step": 5660
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.684928834438324,
+      "learning_rate": 9.349084736933215e-05,
+      "loss": 3.6484,
+      "step": 5665
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6706328988075256,
+      "learning_rate": 9.347940675356292e-05,
+      "loss": 3.7413,
+      "step": 5670
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6811142563819885,
+      "learning_rate": 9.346795679376832e-05,
+      "loss": 3.7307,
+      "step": 5675
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7273861169815063,
+      "learning_rate": 9.345649749240898e-05,
+      "loss": 3.8394,
+      "step": 5680
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7066673040390015,
+      "learning_rate": 9.344502885194763e-05,
+      "loss": 3.6649,
+      "step": 5685
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6519045233726501,
+      "learning_rate": 9.343355087484894e-05,
+      "loss": 3.6792,
+      "step": 5690
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6553910374641418,
+      "learning_rate": 9.342206356357956e-05,
+      "loss": 3.6079,
+      "step": 5695
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.689484715461731,
+      "learning_rate": 9.341056692060828e-05,
+      "loss": 3.6204,
+      "step": 5700
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6506314277648926,
+      "learning_rate": 9.339906094840575e-05,
+      "loss": 3.6783,
+      "step": 5705
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1032545566558838,
+      "learning_rate": 9.33875456494447e-05,
+      "loss": 3.7866,
+      "step": 5710
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6673218607902527,
+      "learning_rate": 9.337602102619985e-05,
+      "loss": 3.6253,
+      "step": 5715
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6747058629989624,
+      "learning_rate": 9.336448708114789e-05,
+      "loss": 3.7598,
+      "step": 5720
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.633236289024353,
+      "learning_rate": 9.33529438167676e-05,
+      "loss": 3.589,
+      "step": 5725
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6544418334960938,
+      "learning_rate": 9.334139123553967e-05,
+      "loss": 3.7826,
+      "step": 5730
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.702997088432312,
+      "learning_rate": 9.332982933994683e-05,
+      "loss": 3.6459,
+      "step": 5735
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6619644165039062,
+      "learning_rate": 9.331825813247381e-05,
+      "loss": 3.6986,
+      "step": 5740
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6559497714042664,
+      "learning_rate": 9.330667761560739e-05,
+      "loss": 3.7714,
+      "step": 5745
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.651055634021759,
+      "learning_rate": 9.329508779183623e-05,
+      "loss": 3.7178,
+      "step": 5750
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6792038083076477,
+      "learning_rate": 9.328348866365111e-05,
+      "loss": 3.6472,
+      "step": 5755
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.818943500518799,
+      "learning_rate": 9.327188023354474e-05,
+      "loss": 3.6964,
+      "step": 5760
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6860701441764832,
+      "learning_rate": 9.326026250401188e-05,
+      "loss": 3.7492,
+      "step": 5765
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7152941823005676,
+      "learning_rate": 9.324863547754924e-05,
+      "loss": 3.6547,
+      "step": 5770
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6729847192764282,
+      "learning_rate": 9.323699915665555e-05,
+      "loss": 3.7743,
+      "step": 5775
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6548579931259155,
+      "learning_rate": 9.322535354383153e-05,
+      "loss": 3.6845,
+      "step": 5780
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6511702537536621,
+      "learning_rate": 9.321369864157992e-05,
+      "loss": 3.6767,
+      "step": 5785
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6451547741889954,
+      "learning_rate": 9.320203445240543e-05,
+      "loss": 3.5779,
+      "step": 5790
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6702418923377991,
+      "learning_rate": 9.319036097881477e-05,
+      "loss": 3.7368,
+      "step": 5795
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6477304697036743,
+      "learning_rate": 9.317867822331667e-05,
+      "loss": 3.6687,
+      "step": 5800
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7858580946922302,
+      "learning_rate": 9.31669861884218e-05,
+      "loss": 3.6699,
+      "step": 5805
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7417999505996704,
+      "learning_rate": 9.315528487664289e-05,
+      "loss": 3.6177,
+      "step": 5810
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.635173499584198,
+      "learning_rate": 9.314357429049462e-05,
+      "loss": 3.7526,
+      "step": 5815
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.640451967716217,
+      "learning_rate": 9.313185443249367e-05,
+      "loss": 3.7464,
+      "step": 5820
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6347212195396423,
+      "learning_rate": 9.312012530515875e-05,
+      "loss": 3.6533,
+      "step": 5825
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6002013683319092,
+      "learning_rate": 9.310838691101048e-05,
+      "loss": 3.7033,
+      "step": 5830
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.580357015132904,
+      "learning_rate": 9.309663925257156e-05,
+      "loss": 3.6432,
+      "step": 5835
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.665326714515686,
+      "learning_rate": 9.308488233236663e-05,
+      "loss": 3.6762,
+      "step": 5840
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6348349452018738,
+      "learning_rate": 9.307311615292234e-05,
+      "loss": 3.6412,
+      "step": 5845
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6799684762954712,
+      "learning_rate": 9.306134071676734e-05,
+      "loss": 3.6944,
+      "step": 5850
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6309449076652527,
+      "learning_rate": 9.30495560264322e-05,
+      "loss": 3.6342,
+      "step": 5855
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6344949007034302,
+      "learning_rate": 9.303776208444958e-05,
+      "loss": 3.6005,
+      "step": 5860
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6141017079353333,
+      "learning_rate": 9.302595889335405e-05,
+      "loss": 3.6942,
+      "step": 5865
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6967468857765198,
+      "learning_rate": 9.301414645568223e-05,
+      "loss": 3.5144,
+      "step": 5870
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6360249519348145,
+      "learning_rate": 9.300232477397266e-05,
+      "loss": 3.7915,
+      "step": 5875
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6376981139183044,
+      "learning_rate": 9.299049385076592e-05,
+      "loss": 3.6842,
+      "step": 5880
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7109939455986023,
+      "learning_rate": 9.297865368860456e-05,
+      "loss": 3.5676,
+      "step": 5885
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6608805656433105,
+      "learning_rate": 9.296680429003311e-05,
+      "loss": 3.6543,
+      "step": 5890
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6582044363021851,
+      "learning_rate": 9.295494565759808e-05,
+      "loss": 3.6879,
+      "step": 5895
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6704488396644592,
+      "learning_rate": 9.294307779384796e-05,
+      "loss": 3.7306,
+      "step": 5900
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7448645830154419,
+      "learning_rate": 9.293120070133327e-05,
+      "loss": 3.6384,
+      "step": 5905
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6338740587234497,
+      "learning_rate": 9.291931438260646e-05,
+      "loss": 3.5693,
+      "step": 5910
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6971124410629272,
+      "learning_rate": 9.290741884022198e-05,
+      "loss": 3.6973,
+      "step": 5915
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6637726426124573,
+      "learning_rate": 9.28955140767363e-05,
+      "loss": 3.5763,
+      "step": 5920
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6214706301689148,
+      "learning_rate": 9.288360009470776e-05,
+      "loss": 3.8267,
+      "step": 5925
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.691270112991333,
+      "learning_rate": 9.287167689669683e-05,
+      "loss": 3.7389,
+      "step": 5930
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8174451589584351,
+      "learning_rate": 9.285974448526585e-05,
+      "loss": 3.7612,
+      "step": 5935
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8745969533920288,
+      "learning_rate": 9.284780286297918e-05,
+      "loss": 3.6769,
+      "step": 5940
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7884221076965332,
+      "learning_rate": 9.283585203240317e-05,
+      "loss": 3.767,
+      "step": 5945
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6723948121070862,
+      "learning_rate": 9.282389199610613e-05,
+      "loss": 3.6327,
+      "step": 5950
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6624341607093811,
+      "learning_rate": 9.281192275665836e-05,
+      "loss": 3.6951,
+      "step": 5955
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6747727990150452,
+      "learning_rate": 9.279994431663212e-05,
+      "loss": 3.7265,
+      "step": 5960
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.614228367805481,
+      "learning_rate": 9.278795667860167e-05,
+      "loss": 3.6438,
+      "step": 5965
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.656100869178772,
+      "learning_rate": 9.277595984514325e-05,
+      "loss": 3.6529,
+      "step": 5970
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6369115710258484,
+      "learning_rate": 9.276395381883502e-05,
+      "loss": 3.6812,
+      "step": 5975
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.643017053604126,
+      "learning_rate": 9.275193860225722e-05,
+      "loss": 3.6786,
+      "step": 5980
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0630841255187988,
+      "learning_rate": 9.273991419799193e-05,
+      "loss": 3.6383,
+      "step": 5985
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6776756644248962,
+      "learning_rate": 9.272788060862334e-05,
+      "loss": 3.7242,
+      "step": 5990
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6619054079055786,
+      "learning_rate": 9.271583783673751e-05,
+      "loss": 3.6294,
+      "step": 5995
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6388033628463745,
+      "learning_rate": 9.270378588492255e-05,
+      "loss": 3.6657,
+      "step": 6000
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6708295941352844,
+      "learning_rate": 9.269172475576849e-05,
+      "loss": 3.717,
+      "step": 6005
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7127484083175659,
+      "learning_rate": 9.267965445186733e-05,
+      "loss": 3.7559,
+      "step": 6010
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6613791584968567,
+      "learning_rate": 9.266757497581308e-05,
+      "loss": 3.7012,
+      "step": 6015
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6745616793632507,
+      "learning_rate": 9.265548633020174e-05,
+      "loss": 3.6724,
+      "step": 6020
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6218417286872864,
+      "learning_rate": 9.264338851763117e-05,
+      "loss": 3.7013,
+      "step": 6025
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6837544441223145,
+      "learning_rate": 9.263128154070133e-05,
+      "loss": 3.688,
+      "step": 6030
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.646815299987793,
+      "learning_rate": 9.261916540201407e-05,
+      "loss": 3.666,
+      "step": 6035
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6446861028671265,
+      "learning_rate": 9.260704010417325e-05,
+      "loss": 3.688,
+      "step": 6040
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6488479375839233,
+      "learning_rate": 9.259490564978466e-05,
+      "loss": 3.703,
+      "step": 6045
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6030767560005188,
+      "learning_rate": 9.258276204145609e-05,
+      "loss": 3.561,
+      "step": 6050
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6617406010627747,
+      "learning_rate": 9.257060928179726e-05,
+      "loss": 3.6222,
+      "step": 6055
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7117607593536377,
+      "learning_rate": 9.255844737341992e-05,
+      "loss": 3.6855,
+      "step": 6060
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6948263645172119,
+      "learning_rate": 9.254627631893773e-05,
+      "loss": 3.5813,
+      "step": 6065
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6472216248512268,
+      "learning_rate": 9.253409612096635e-05,
+      "loss": 3.7588,
+      "step": 6070
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6088003516197205,
+      "learning_rate": 9.252190678212335e-05,
+      "loss": 3.696,
+      "step": 6075
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6327986717224121,
+      "learning_rate": 9.250970830502834e-05,
+      "loss": 3.7343,
+      "step": 6080
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6973710656166077,
+      "learning_rate": 9.249750069230285e-05,
+      "loss": 3.6496,
+      "step": 6085
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6609290838241577,
+      "learning_rate": 9.248528394657036e-05,
+      "loss": 3.6184,
+      "step": 6090
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4919941425323486,
+      "learning_rate": 9.247305807045637e-05,
+      "loss": 3.6502,
+      "step": 6095
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6744839549064636,
+      "learning_rate": 9.246082306658827e-05,
+      "loss": 3.6309,
+      "step": 6100
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.672412633895874,
+      "learning_rate": 9.244857893759545e-05,
+      "loss": 3.7089,
+      "step": 6105
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7039036750793457,
+      "learning_rate": 9.243632568610929e-05,
+      "loss": 3.6505,
+      "step": 6110
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7028497457504272,
+      "learning_rate": 9.242406331476307e-05,
+      "loss": 3.773,
+      "step": 6115
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6230226755142212,
+      "learning_rate": 9.241179182619206e-05,
+      "loss": 3.6678,
+      "step": 6120
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6302070617675781,
+      "learning_rate": 9.239951122303349e-05,
+      "loss": 3.6145,
+      "step": 6125
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6892881989479065,
+      "learning_rate": 9.238722150792658e-05,
+      "loss": 3.6781,
+      "step": 6130
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6155871152877808,
+      "learning_rate": 9.237492268351242e-05,
+      "loss": 3.6989,
+      "step": 6135
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.642099916934967,
+      "learning_rate": 9.236261475243415e-05,
+      "loss": 3.6694,
+      "step": 6140
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6773453950881958,
+      "learning_rate": 9.23502977173368e-05,
+      "loss": 3.6706,
+      "step": 6145
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6855302453041077,
+      "learning_rate": 9.233797158086742e-05,
+      "loss": 3.6322,
+      "step": 6150
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8033294081687927,
+      "learning_rate": 9.232563634567495e-05,
+      "loss": 3.5964,
+      "step": 6155
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8495514392852783,
+      "learning_rate": 9.231329201441034e-05,
+      "loss": 3.6235,
+      "step": 6160
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6426656246185303,
+      "learning_rate": 9.230093858972646e-05,
+      "loss": 3.5409,
+      "step": 6165
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7213931083679199,
+      "learning_rate": 9.228857607427816e-05,
+      "loss": 3.6349,
+      "step": 6170
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6703625917434692,
+      "learning_rate": 9.227620447072221e-05,
+      "loss": 3.6018,
+      "step": 6175
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6350034475326538,
+      "learning_rate": 9.226382378171738e-05,
+      "loss": 3.7629,
+      "step": 6180
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6193051338195801,
+      "learning_rate": 9.225143400992435e-05,
+      "loss": 3.5726,
+      "step": 6185
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6390960216522217,
+      "learning_rate": 9.223903515800576e-05,
+      "loss": 3.6624,
+      "step": 6190
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6209862232208252,
+      "learning_rate": 9.222662722862622e-05,
+      "loss": 3.6287,
+      "step": 6195
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6479684114456177,
+      "learning_rate": 9.221421022445228e-05,
+      "loss": 3.685,
+      "step": 6200
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0272868871688843,
+      "learning_rate": 9.220178414815242e-05,
+      "loss": 3.7486,
+      "step": 6205
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7484703063964844,
+      "learning_rate": 9.218934900239713e-05,
+      "loss": 3.6016,
+      "step": 6210
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6864889860153198,
+      "learning_rate": 9.217690478985878e-05,
+      "loss": 3.6361,
+      "step": 6215
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6312556266784668,
+      "learning_rate": 9.216445151321171e-05,
+      "loss": 3.6617,
+      "step": 6220
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6576535105705261,
+      "learning_rate": 9.215198917513223e-05,
+      "loss": 3.7002,
+      "step": 6225
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6815969944000244,
+      "learning_rate": 9.213951777829861e-05,
+      "loss": 3.6127,
+      "step": 6230
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6535267233848572,
+      "learning_rate": 9.212703732539098e-05,
+      "loss": 3.6648,
+      "step": 6235
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7077336311340332,
+      "learning_rate": 9.211454781909153e-05,
+      "loss": 3.563,
+      "step": 6240
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6654105186462402,
+      "learning_rate": 9.21020492620843e-05,
+      "loss": 3.7769,
+      "step": 6245
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7664467692375183,
+      "learning_rate": 9.208954165705534e-05,
+      "loss": 3.6503,
+      "step": 6250
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6360939145088196,
+      "learning_rate": 9.207702500669262e-05,
+      "loss": 3.69,
+      "step": 6255
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6217556595802307,
+      "learning_rate": 9.206449931368604e-05,
+      "loss": 3.6539,
+      "step": 6260
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7074015736579895,
+      "learning_rate": 9.205196458072747e-05,
+      "loss": 3.5584,
+      "step": 6265
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6666274666786194,
+      "learning_rate": 9.203942081051071e-05,
+      "loss": 3.6604,
+      "step": 6270
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6795749068260193,
+      "learning_rate": 9.20268680057315e-05,
+      "loss": 3.8339,
+      "step": 6275
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.631287693977356,
+      "learning_rate": 9.201430616908752e-05,
+      "loss": 3.6053,
+      "step": 6280
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1515945196151733,
+      "learning_rate": 9.200173530327842e-05,
+      "loss": 3.609,
+      "step": 6285
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8341479301452637,
+      "learning_rate": 9.198915541100575e-05,
+      "loss": 3.6393,
+      "step": 6290
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7924326062202454,
+      "learning_rate": 9.1976566494973e-05,
+      "loss": 3.5724,
+      "step": 6295
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6725490093231201,
+      "learning_rate": 9.196396855788563e-05,
+      "loss": 3.6039,
+      "step": 6300
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6777785420417786,
+      "learning_rate": 9.195136160245104e-05,
+      "loss": 3.6515,
+      "step": 6305
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6182354688644409,
+      "learning_rate": 9.193874563137853e-05,
+      "loss": 3.6425,
+      "step": 6310
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6612922549247742,
+      "learning_rate": 9.192612064737934e-05,
+      "loss": 3.5554,
+      "step": 6315
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6831209063529968,
+      "learning_rate": 9.19134866531667e-05,
+      "loss": 3.6914,
+      "step": 6320
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6318866014480591,
+      "learning_rate": 9.190084365145575e-05,
+      "loss": 3.6225,
+      "step": 6325
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6270996928215027,
+      "learning_rate": 9.188819164496354e-05,
+      "loss": 3.6078,
+      "step": 6330
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6741576194763184,
+      "learning_rate": 9.187553063640906e-05,
+      "loss": 3.5466,
+      "step": 6335
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.658142626285553,
+      "learning_rate": 9.186286062851327e-05,
+      "loss": 3.6382,
+      "step": 6340
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6297696232795715,
+      "learning_rate": 9.185018162399905e-05,
+      "loss": 3.6156,
+      "step": 6345
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6201626658439636,
+      "learning_rate": 9.183749362559118e-05,
+      "loss": 3.6959,
+      "step": 6350
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6419195532798767,
+      "learning_rate": 9.182479663601642e-05,
+      "loss": 3.6545,
+      "step": 6355
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6703521013259888,
+      "learning_rate": 9.181209065800342e-05,
+      "loss": 3.5909,
+      "step": 6360
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6766203045845032,
+      "learning_rate": 9.17993756942828e-05,
+      "loss": 3.7279,
+      "step": 6365
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6403589248657227,
+      "learning_rate": 9.17866517475871e-05,
+      "loss": 3.5355,
+      "step": 6370
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6516534686088562,
+      "learning_rate": 9.177391882065077e-05,
+      "loss": 3.7593,
+      "step": 6375
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6633095741271973,
+      "learning_rate": 9.176117691621017e-05,
+      "loss": 3.7391,
+      "step": 6380
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6580092310905457,
+      "learning_rate": 9.17484260370037e-05,
+      "loss": 3.6513,
+      "step": 6385
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6619282960891724,
+      "learning_rate": 9.173566618577156e-05,
+      "loss": 3.596,
+      "step": 6390
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7399194240570068,
+      "learning_rate": 9.172289736525594e-05,
+      "loss": 3.5827,
+      "step": 6395
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6927294135093689,
+      "learning_rate": 9.171011957820097e-05,
+      "loss": 3.6634,
+      "step": 6400
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.735564112663269,
+      "learning_rate": 9.169733282735265e-05,
+      "loss": 3.7298,
+      "step": 6405
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7396203875541687,
+      "learning_rate": 9.168453711545895e-05,
+      "loss": 3.6398,
+      "step": 6410
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6419672966003418,
+      "learning_rate": 9.167173244526977e-05,
+      "loss": 3.6318,
+      "step": 6415
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7140073776245117,
+      "learning_rate": 9.165891881953692e-05,
+      "loss": 3.5822,
+      "step": 6420
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6780020594596863,
+      "learning_rate": 9.164609624101414e-05,
+      "loss": 3.6585,
+      "step": 6425
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6852356791496277,
+      "learning_rate": 9.163326471245707e-05,
+      "loss": 3.6802,
+      "step": 6430
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7554282546043396,
+      "learning_rate": 9.162042423662331e-05,
+      "loss": 3.5746,
+      "step": 6435
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6545906662940979,
+      "learning_rate": 9.160757481627237e-05,
+      "loss": 3.6957,
+      "step": 6440
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6377950310707092,
+      "learning_rate": 9.159471645416567e-05,
+      "loss": 3.6351,
+      "step": 6445
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.625011682510376,
+      "learning_rate": 9.158184915306655e-05,
+      "loss": 3.6626,
+      "step": 6450
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6501367688179016,
+      "learning_rate": 9.156897291574034e-05,
+      "loss": 3.7196,
+      "step": 6455
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6951277852058411,
+      "learning_rate": 9.155608774495415e-05,
+      "loss": 3.5991,
+      "step": 6460
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6517708897590637,
+      "learning_rate": 9.154319364347715e-05,
+      "loss": 3.6315,
+      "step": 6465
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6652078032493591,
+      "learning_rate": 9.153029061408037e-05,
+      "loss": 3.7044,
+      "step": 6470
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6485198140144348,
+      "learning_rate": 9.151737865953671e-05,
+      "loss": 3.632,
+      "step": 6475
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6219106912612915,
+      "learning_rate": 9.150445778262109e-05,
+      "loss": 3.6548,
+      "step": 6480
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6472190618515015,
+      "learning_rate": 9.149152798611027e-05,
+      "loss": 3.6765,
+      "step": 6485
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7456123232841492,
+      "learning_rate": 9.147858927278296e-05,
+      "loss": 3.7046,
+      "step": 6490
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8964935541152954,
+      "learning_rate": 9.14656416454198e-05,
+      "loss": 3.6613,
+      "step": 6495
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5985890626907349,
+      "learning_rate": 9.14526851068033e-05,
+      "loss": 3.7074,
+      "step": 6500
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7010729312896729,
+      "learning_rate": 9.143971965971789e-05,
+      "loss": 3.5827,
+      "step": 6505
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.605023205280304,
+      "learning_rate": 9.142674530694999e-05,
+      "loss": 3.6382,
+      "step": 6510
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6686131358146667,
+      "learning_rate": 9.141376205128784e-05,
+      "loss": 3.7312,
+      "step": 6515
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6271719336509705,
+      "learning_rate": 9.140076989552162e-05,
+      "loss": 3.5349,
+      "step": 6520
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6345179677009583,
+      "learning_rate": 9.138776884244347e-05,
+      "loss": 3.669,
+      "step": 6525
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6318054795265198,
+      "learning_rate": 9.137475889484739e-05,
+      "loss": 3.7473,
+      "step": 6530
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6742165684700012,
+      "learning_rate": 9.13617400555293e-05,
+      "loss": 3.5855,
+      "step": 6535
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5957058072090149,
+      "learning_rate": 9.134871232728706e-05,
+      "loss": 3.68,
+      "step": 6540
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6261151432991028,
+      "learning_rate": 9.133567571292038e-05,
+      "loss": 3.7109,
+      "step": 6545
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6768052577972412,
+      "learning_rate": 9.132263021523096e-05,
+      "loss": 3.6603,
+      "step": 6550
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7402241230010986,
+      "learning_rate": 9.130957583702236e-05,
+      "loss": 3.6404,
+      "step": 6555
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6815353631973267,
+      "learning_rate": 9.129651258110003e-05,
+      "loss": 3.61,
+      "step": 6560
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7246091365814209,
+      "learning_rate": 9.128344045027137e-05,
+      "loss": 3.6155,
+      "step": 6565
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6492617130279541,
+      "learning_rate": 9.127035944734568e-05,
+      "loss": 3.6303,
+      "step": 6570
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6493353843688965,
+      "learning_rate": 9.125726957513414e-05,
+      "loss": 3.6089,
+      "step": 6575
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6899619698524475,
+      "learning_rate": 9.124417083644988e-05,
+      "loss": 3.6144,
+      "step": 6580
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6046091914176941,
+      "learning_rate": 9.12310632341079e-05,
+      "loss": 3.5522,
+      "step": 6585
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6009009480476379,
+      "learning_rate": 9.121794677092509e-05,
+      "loss": 3.5431,
+      "step": 6590
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6662131547927856,
+      "learning_rate": 9.120482144972029e-05,
+      "loss": 3.5625,
+      "step": 6595
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6702367067337036,
+      "learning_rate": 9.119168727331424e-05,
+      "loss": 3.6052,
+      "step": 6600
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6265493631362915,
+      "learning_rate": 9.117854424452952e-05,
+      "loss": 3.5123,
+      "step": 6605
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6714627742767334,
+      "learning_rate": 9.116539236619071e-05,
+      "loss": 3.4254,
+      "step": 6610
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9042956233024597,
+      "learning_rate": 9.115223164112418e-05,
+      "loss": 3.7119,
+      "step": 6615
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6815637350082397,
+      "learning_rate": 9.113906207215832e-05,
+      "loss": 3.6331,
+      "step": 6620
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6754025816917419,
+      "learning_rate": 9.112588366212333e-05,
+      "loss": 3.7334,
+      "step": 6625
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.693716824054718,
+      "learning_rate": 9.111269641385134e-05,
+      "loss": 3.6804,
+      "step": 6630
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6276625394821167,
+      "learning_rate": 9.10995003301764e-05,
+      "loss": 3.663,
+      "step": 6635
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6902511119842529,
+      "learning_rate": 9.10862954139344e-05,
+      "loss": 3.556,
+      "step": 6640
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6989098787307739,
+      "learning_rate": 9.107308166796322e-05,
+      "loss": 3.8107,
+      "step": 6645
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6390373110771179,
+      "learning_rate": 9.105985909510255e-05,
+      "loss": 3.6238,
+      "step": 6650
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.646295964717865,
+      "learning_rate": 9.104662769819404e-05,
+      "loss": 3.5797,
+      "step": 6655
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6221117973327637,
+      "learning_rate": 9.103338748008118e-05,
+      "loss": 3.661,
+      "step": 6660
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6219208836555481,
+      "learning_rate": 9.102013844360939e-05,
+      "loss": 3.6751,
+      "step": 6665
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6670545339584351,
+      "learning_rate": 9.100688059162599e-05,
+      "loss": 3.5458,
+      "step": 6670
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6524112224578857,
+      "learning_rate": 9.099361392698018e-05,
+      "loss": 3.6215,
+      "step": 6675
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6660169363021851,
+      "learning_rate": 9.098033845252307e-05,
+      "loss": 3.7306,
+      "step": 6680
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6620325446128845,
+      "learning_rate": 9.096705417110762e-05,
+      "loss": 3.6542,
+      "step": 6685
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.649667501449585,
+      "learning_rate": 9.095376108558875e-05,
+      "loss": 3.695,
+      "step": 6690
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6704279184341431,
+      "learning_rate": 9.094045919882322e-05,
+      "loss": 3.6827,
+      "step": 6695
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6440548896789551,
+      "learning_rate": 9.092714851366968e-05,
+      "loss": 3.6942,
+      "step": 6700
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6055727601051331,
+      "learning_rate": 9.091382903298872e-05,
+      "loss": 3.6551,
+      "step": 6705
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.63572096824646,
+      "learning_rate": 9.090050075964276e-05,
+      "loss": 3.6482,
+      "step": 6710
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7293950915336609,
+      "learning_rate": 9.088716369649616e-05,
+      "loss": 3.6384,
+      "step": 6715
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6475096344947815,
+      "learning_rate": 9.087381784641515e-05,
+      "loss": 3.5661,
+      "step": 6720
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6106477379798889,
+      "learning_rate": 9.086046321226783e-05,
+      "loss": 3.6906,
+      "step": 6725
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.709546685218811,
+      "learning_rate": 9.08470997969242e-05,
+      "loss": 3.6843,
+      "step": 6730
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6067266464233398,
+      "learning_rate": 9.083372760325617e-05,
+      "loss": 3.5679,
+      "step": 6735
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6003871560096741,
+      "learning_rate": 9.08203466341375e-05,
+      "loss": 3.6779,
+      "step": 6740
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6335563063621521,
+      "learning_rate": 9.080695689244385e-05,
+      "loss": 3.4717,
+      "step": 6745
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6089677214622498,
+      "learning_rate": 9.079355838105279e-05,
+      "loss": 3.6444,
+      "step": 6750
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6419903039932251,
+      "learning_rate": 9.078015110284373e-05,
+      "loss": 3.661,
+      "step": 6755
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6160089373588562,
+      "learning_rate": 9.076673506069801e-05,
+      "loss": 3.5884,
+      "step": 6760
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6147283315658569,
+      "learning_rate": 9.07533102574988e-05,
+      "loss": 3.6029,
+      "step": 6765
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6442582011222839,
+      "learning_rate": 9.07398766961312e-05,
+      "loss": 3.5697,
+      "step": 6770
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6847180724143982,
+      "learning_rate": 9.072643437948218e-05,
+      "loss": 3.5782,
+      "step": 6775
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.609647810459137,
+      "learning_rate": 9.07129833104406e-05,
+      "loss": 3.6538,
+      "step": 6780
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6177313923835754,
+      "learning_rate": 9.069952349189714e-05,
+      "loss": 3.5657,
+      "step": 6785
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6500287055969238,
+      "learning_rate": 9.068605492674446e-05,
+      "loss": 3.6273,
+      "step": 6790
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6147716045379639,
+      "learning_rate": 9.0672577617877e-05,
+      "loss": 3.5549,
+      "step": 6795
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6033639311790466,
+      "learning_rate": 9.065909156819118e-05,
+      "loss": 3.6442,
+      "step": 6800
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6429266929626465,
+      "learning_rate": 9.06455967805852e-05,
+      "loss": 3.6242,
+      "step": 6805
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6387028098106384,
+      "learning_rate": 9.063209325795921e-05,
+      "loss": 3.5959,
+      "step": 6810
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6550421714782715,
+      "learning_rate": 9.06185810032152e-05,
+      "loss": 3.5943,
+      "step": 6815
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6509239673614502,
+      "learning_rate": 9.060506001925707e-05,
+      "loss": 3.5083,
+      "step": 6820
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6172776818275452,
+      "learning_rate": 9.059153030899051e-05,
+      "loss": 3.6895,
+      "step": 6825
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6380317211151123,
+      "learning_rate": 9.05779918753232e-05,
+      "loss": 3.6573,
+      "step": 6830
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6028115749359131,
+      "learning_rate": 9.056444472116465e-05,
+      "loss": 3.5863,
+      "step": 6835
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6052219867706299,
+      "learning_rate": 9.055088884942622e-05,
+      "loss": 3.5531,
+      "step": 6840
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6171693205833435,
+      "learning_rate": 9.053732426302116e-05,
+      "loss": 3.5917,
+      "step": 6845
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6217378377914429,
+      "learning_rate": 9.052375096486458e-05,
+      "loss": 3.6174,
+      "step": 6850
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6109480857849121,
+      "learning_rate": 9.051016895787352e-05,
+      "loss": 3.5944,
+      "step": 6855
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6807569861412048,
+      "learning_rate": 9.049657824496678e-05,
+      "loss": 3.5382,
+      "step": 6860
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5830181241035461,
+      "learning_rate": 9.048297882906516e-05,
+      "loss": 3.528,
+      "step": 6865
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6308459639549255,
+      "learning_rate": 9.046937071309124e-05,
+      "loss": 3.773,
+      "step": 6870
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6233350038528442,
+      "learning_rate": 9.04557538999695e-05,
+      "loss": 3.5791,
+      "step": 6875
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6124426126480103,
+      "learning_rate": 9.04421283926263e-05,
+      "loss": 3.626,
+      "step": 6880
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6691504716873169,
+      "learning_rate": 9.042849419398982e-05,
+      "loss": 3.5537,
+      "step": 6885
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6202459931373596,
+      "learning_rate": 9.041485130699018e-05,
+      "loss": 3.707,
+      "step": 6890
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6235212087631226,
+      "learning_rate": 9.040119973455934e-05,
+      "loss": 3.6184,
+      "step": 6895
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6367961168289185,
+      "learning_rate": 9.038753947963108e-05,
+      "loss": 3.761,
+      "step": 6900
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6065836548805237,
+      "learning_rate": 9.03738705451411e-05,
+      "loss": 3.63,
+      "step": 6905
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6393280625343323,
+      "learning_rate": 9.036019293402697e-05,
+      "loss": 3.6714,
+      "step": 6910
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9464275240898132,
+      "learning_rate": 9.034650664922807e-05,
+      "loss": 3.5718,
+      "step": 6915
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.587316632270813,
+      "learning_rate": 9.033281169368568e-05,
+      "loss": 3.6928,
+      "step": 6920
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6801005601882935,
+      "learning_rate": 9.031910807034296e-05,
+      "loss": 3.7189,
+      "step": 6925
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6121535897254944,
+      "learning_rate": 9.03053957821449e-05,
+      "loss": 3.5936,
+      "step": 6930
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5981286764144897,
+      "learning_rate": 9.029167483203837e-05,
+      "loss": 3.6564,
+      "step": 6935
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6156281232833862,
+      "learning_rate": 9.027794522297211e-05,
+      "loss": 3.7379,
+      "step": 6940
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6822390556335449,
+      "learning_rate": 9.026420695789668e-05,
+      "loss": 3.5559,
+      "step": 6945
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6664403676986694,
+      "learning_rate": 9.025046003976457e-05,
+      "loss": 3.7128,
+      "step": 6950
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6222808361053467,
+      "learning_rate": 9.023670447153003e-05,
+      "loss": 3.604,
+      "step": 6955
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.619886040687561,
+      "learning_rate": 9.022294025614926e-05,
+      "loss": 3.6674,
+      "step": 6960
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6248182058334351,
+      "learning_rate": 9.02091673965803e-05,
+      "loss": 3.5409,
+      "step": 6965
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6373429894447327,
+      "learning_rate": 9.0195385895783e-05,
+      "loss": 3.7041,
+      "step": 6970
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6263178586959839,
+      "learning_rate": 9.018159575671914e-05,
+      "loss": 3.6546,
+      "step": 6975
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6629490852355957,
+      "learning_rate": 9.016779698235227e-05,
+      "loss": 3.5952,
+      "step": 6980
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9641255140304565,
+      "learning_rate": 9.015398957564787e-05,
+      "loss": 3.6525,
+      "step": 6985
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6224432587623596,
+      "learning_rate": 9.014017353957324e-05,
+      "loss": 3.5963,
+      "step": 6990
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6418037414550781,
+      "learning_rate": 9.012634887709754e-05,
+      "loss": 3.625,
+      "step": 6995
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6747340559959412,
+      "learning_rate": 9.01125155911918e-05,
+      "loss": 3.643,
+      "step": 7000
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6656273007392883,
+      "learning_rate": 9.009867368482887e-05,
+      "loss": 3.4906,
+      "step": 7005
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6343880295753479,
+      "learning_rate": 9.008482316098348e-05,
+      "loss": 3.6144,
+      "step": 7010
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6222692728042603,
+      "learning_rate": 9.007096402263221e-05,
+      "loss": 3.596,
+      "step": 7015
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6203016638755798,
+      "learning_rate": 9.005709627275348e-05,
+      "loss": 3.705,
+      "step": 7020
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6504935622215271,
+      "learning_rate": 9.004321991432753e-05,
+      "loss": 3.6636,
+      "step": 7025
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.618148684501648,
+      "learning_rate": 9.002933495033657e-05,
+      "loss": 3.6457,
+      "step": 7030
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6605909466743469,
+      "learning_rate": 9.001544138376447e-05,
+      "loss": 3.6223,
+      "step": 7035
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6305635571479797,
+      "learning_rate": 9.000153921759714e-05,
+      "loss": 3.5525,
+      "step": 7040
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6906110644340515,
+      "learning_rate": 8.998762845482223e-05,
+      "loss": 3.5922,
+      "step": 7045
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6018814444541931,
+      "learning_rate": 8.997370909842921e-05,
+      "loss": 3.6781,
+      "step": 7050
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6599273085594177,
+      "learning_rate": 8.99597811514095e-05,
+      "loss": 3.6171,
+      "step": 7055
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6832717061042786,
+      "learning_rate": 8.99458446167563e-05,
+      "loss": 3.5798,
+      "step": 7060
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6442378163337708,
+      "learning_rate": 8.993189949746468e-05,
+      "loss": 3.6803,
+      "step": 7065
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6228021383285522,
+      "learning_rate": 8.99179457965315e-05,
+      "loss": 3.6071,
+      "step": 7070
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.750568151473999,
+      "learning_rate": 8.990398351695555e-05,
+      "loss": 3.6375,
+      "step": 7075
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6378205418586731,
+      "learning_rate": 8.989001266173738e-05,
+      "loss": 3.51,
+      "step": 7080
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6649860739707947,
+      "learning_rate": 8.987603323387948e-05,
+      "loss": 3.5074,
+      "step": 7085
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.63639235496521,
+      "learning_rate": 8.986204523638606e-05,
+      "loss": 3.7007,
+      "step": 7090
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6309627890586853,
+      "learning_rate": 8.984804867226328e-05,
+      "loss": 3.592,
+      "step": 7095
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6214967370033264,
+      "learning_rate": 8.98340435445191e-05,
+      "loss": 3.6596,
+      "step": 7100
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6079266667366028,
+      "learning_rate": 8.982002985616331e-05,
+      "loss": 3.6071,
+      "step": 7105
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6420413851737976,
+      "learning_rate": 8.980600761020754e-05,
+      "loss": 3.5596,
+      "step": 7110
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6389975547790527,
+      "learning_rate": 8.979197680966527e-05,
+      "loss": 3.66,
+      "step": 7115
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6986696124076843,
+      "learning_rate": 8.977793745755182e-05,
+      "loss": 3.6178,
+      "step": 7120
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5919976830482483,
+      "learning_rate": 8.976388955688434e-05,
+      "loss": 3.6425,
+      "step": 7125
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6332110166549683,
+      "learning_rate": 8.974983311068184e-05,
+      "loss": 3.6862,
+      "step": 7130
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6868057250976562,
+      "learning_rate": 8.973576812196513e-05,
+      "loss": 3.7558,
+      "step": 7135
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5996155142784119,
+      "learning_rate": 8.972169459375686e-05,
+      "loss": 3.5383,
+      "step": 7140
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6755701303482056,
+      "learning_rate": 8.970761252908156e-05,
+      "loss": 3.6204,
+      "step": 7145
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6613823175430298,
+      "learning_rate": 8.969352193096554e-05,
+      "loss": 3.5389,
+      "step": 7150
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6834219694137573,
+      "learning_rate": 8.967942280243699e-05,
+      "loss": 3.6719,
+      "step": 7155
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7740730047225952,
+      "learning_rate": 8.966531514652588e-05,
+      "loss": 3.5922,
+      "step": 7160
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7059200406074524,
+      "learning_rate": 8.965119896626406e-05,
+      "loss": 3.5966,
+      "step": 7165
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6294935345649719,
+      "learning_rate": 8.963707426468519e-05,
+      "loss": 3.6139,
+      "step": 7170
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6977207660675049,
+      "learning_rate": 8.962294104482478e-05,
+      "loss": 3.5911,
+      "step": 7175
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6249411702156067,
+      "learning_rate": 8.960879930972014e-05,
+      "loss": 3.5235,
+      "step": 7180
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7075315117835999,
+      "learning_rate": 8.959464906241044e-05,
+      "loss": 3.6006,
+      "step": 7185
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6995972394943237,
+      "learning_rate": 8.958049030593665e-05,
+      "loss": 3.6003,
+      "step": 7190
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6933475136756897,
+      "learning_rate": 8.956632304334158e-05,
+      "loss": 3.5816,
+      "step": 7195
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.658586323261261,
+      "learning_rate": 8.955214727766989e-05,
+      "loss": 3.6394,
+      "step": 7200
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6238296627998352,
+      "learning_rate": 8.953796301196807e-05,
+      "loss": 3.5896,
+      "step": 7205
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7330600023269653,
+      "learning_rate": 8.952377024928437e-05,
+      "loss": 3.6199,
+      "step": 7210
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5955808162689209,
+      "learning_rate": 8.950956899266895e-05,
+      "loss": 3.5327,
+      "step": 7215
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6118154525756836,
+      "learning_rate": 8.949535924517374e-05,
+      "loss": 3.692,
+      "step": 7220
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6050345301628113,
+      "learning_rate": 8.948114100985252e-05,
+      "loss": 3.5663,
+      "step": 7225
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6501654982566833,
+      "learning_rate": 8.946691428976088e-05,
+      "loss": 3.5274,
+      "step": 7230
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6236889958381653,
+      "learning_rate": 8.945267908795625e-05,
+      "loss": 3.4946,
+      "step": 7235
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6258043646812439,
+      "learning_rate": 8.943843540749785e-05,
+      "loss": 3.6215,
+      "step": 7240
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6433592438697815,
+      "learning_rate": 8.942418325144679e-05,
+      "loss": 3.5583,
+      "step": 7245
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6381149888038635,
+      "learning_rate": 8.940992262286592e-05,
+      "loss": 3.5345,
+      "step": 7250
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6156508922576904,
+      "learning_rate": 8.939565352481996e-05,
+      "loss": 3.6987,
+      "step": 7255
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6194127202033997,
+      "learning_rate": 8.938137596037544e-05,
+      "loss": 3.585,
+      "step": 7260
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6027131080627441,
+      "learning_rate": 8.93670899326007e-05,
+      "loss": 3.641,
+      "step": 7265
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5821008086204529,
+      "learning_rate": 8.93527954445659e-05,
+      "loss": 3.5172,
+      "step": 7270
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6400082111358643,
+      "learning_rate": 8.933849249934306e-05,
+      "loss": 3.6571,
+      "step": 7275
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6383677124977112,
+      "learning_rate": 8.932418110000594e-05,
+      "loss": 3.6384,
+      "step": 7280
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6943330764770508,
+      "learning_rate": 8.930986124963016e-05,
+      "loss": 3.7469,
+      "step": 7285
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6976456642150879,
+      "learning_rate": 8.92955329512932e-05,
+      "loss": 3.5218,
+      "step": 7290
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.653202474117279,
+      "learning_rate": 8.928119620807426e-05,
+      "loss": 3.6649,
+      "step": 7295
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6405032873153687,
+      "learning_rate": 8.926685102305443e-05,
+      "loss": 3.6129,
+      "step": 7300
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6589599251747131,
+      "learning_rate": 8.925249739931659e-05,
+      "loss": 3.7171,
+      "step": 7305
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6682443618774414,
+      "learning_rate": 8.923813533994543e-05,
+      "loss": 3.5462,
+      "step": 7310
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6117516160011292,
+      "learning_rate": 8.922376484802745e-05,
+      "loss": 3.6511,
+      "step": 7315
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6581892371177673,
+      "learning_rate": 8.920938592665098e-05,
+      "loss": 3.5194,
+      "step": 7320
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.694196879863739,
+      "learning_rate": 8.919499857890613e-05,
+      "loss": 3.523,
+      "step": 7325
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6860597729682922,
+      "learning_rate": 8.918060280788486e-05,
+      "loss": 3.57,
+      "step": 7330
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6871790885925293,
+      "learning_rate": 8.916619861668092e-05,
+      "loss": 3.6612,
+      "step": 7335
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6024455428123474,
+      "learning_rate": 8.915178600838985e-05,
+      "loss": 3.5378,
+      "step": 7340
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6459681987762451,
+      "learning_rate": 8.913736498610906e-05,
+      "loss": 3.5946,
+      "step": 7345
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6452612280845642,
+      "learning_rate": 8.912293555293767e-05,
+      "loss": 3.6069,
+      "step": 7350
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6114992499351501,
+      "learning_rate": 8.910849771197671e-05,
+      "loss": 3.6417,
+      "step": 7355
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6363154053688049,
+      "learning_rate": 8.909405146632896e-05,
+      "loss": 3.4792,
+      "step": 7360
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6284523606300354,
+      "learning_rate": 8.907959681909902e-05,
+      "loss": 3.5592,
+      "step": 7365
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6296967267990112,
+      "learning_rate": 8.906513377339327e-05,
+      "loss": 3.487,
+      "step": 7370
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6165747046470642,
+      "learning_rate": 8.905066233231995e-05,
+      "loss": 3.6827,
+      "step": 7375
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6058087944984436,
+      "learning_rate": 8.903618249898906e-05,
+      "loss": 3.5689,
+      "step": 7380
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6135293245315552,
+      "learning_rate": 8.902169427651242e-05,
+      "loss": 3.6444,
+      "step": 7385
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6486176252365112,
+      "learning_rate": 8.900719766800365e-05,
+      "loss": 3.6375,
+      "step": 7390
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6227015256881714,
+      "learning_rate": 8.899269267657817e-05,
+      "loss": 3.5991,
+      "step": 7395
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6042312979698181,
+      "learning_rate": 8.89781793053532e-05,
+      "loss": 3.6329,
+      "step": 7400
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5910986661911011,
+      "learning_rate": 8.896365755744775e-05,
+      "loss": 3.5594,
+      "step": 7405
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6082831025123596,
+      "learning_rate": 8.894912743598268e-05,
+      "loss": 3.5511,
+      "step": 7410
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6199930310249329,
+      "learning_rate": 8.893458894408058e-05,
+      "loss": 3.5941,
+      "step": 7415
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5959880948066711,
+      "learning_rate": 8.89200420848659e-05,
+      "loss": 3.6117,
+      "step": 7420
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6170229911804199,
+      "learning_rate": 8.890548686146483e-05,
+      "loss": 3.5293,
+      "step": 7425
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6392413377761841,
+      "learning_rate": 8.889092327700541e-05,
+      "loss": 3.5689,
+      "step": 7430
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7164409160614014,
+      "learning_rate": 8.887635133461745e-05,
+      "loss": 3.617,
+      "step": 7435
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5981169939041138,
+      "learning_rate": 8.886177103743257e-05,
+      "loss": 3.5323,
+      "step": 7440
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6519257426261902,
+      "learning_rate": 8.884718238858416e-05,
+      "loss": 3.5603,
+      "step": 7445
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6216931939125061,
+      "learning_rate": 8.883258539120742e-05,
+      "loss": 3.5346,
+      "step": 7450
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5975487232208252,
+      "learning_rate": 8.881798004843938e-05,
+      "loss": 3.5814,
+      "step": 7455
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6166713237762451,
+      "learning_rate": 8.880336636341878e-05,
+      "loss": 3.5707,
+      "step": 7460
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6165528297424316,
+      "learning_rate": 8.878874433928623e-05,
+      "loss": 3.5084,
+      "step": 7465
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6397777795791626,
+      "learning_rate": 8.877411397918411e-05,
+      "loss": 3.5327,
+      "step": 7470
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5709030032157898,
+      "learning_rate": 8.875947528625658e-05,
+      "loss": 3.5783,
+      "step": 7475
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6828928589820862,
+      "learning_rate": 8.874482826364958e-05,
+      "loss": 3.5514,
+      "step": 7480
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6735356450080872,
+      "learning_rate": 8.873017291451086e-05,
+      "loss": 3.5694,
+      "step": 7485
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5955107808113098,
+      "learning_rate": 8.871550924198999e-05,
+      "loss": 3.7195,
+      "step": 7490
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6493826508522034,
+      "learning_rate": 8.870083724923826e-05,
+      "loss": 3.6604,
+      "step": 7495
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7016242742538452,
+      "learning_rate": 8.868615693940879e-05,
+      "loss": 3.5872,
+      "step": 7500
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6070040464401245,
+      "learning_rate": 8.86714683156565e-05,
+      "loss": 3.6087,
+      "step": 7505
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6062846183776855,
+      "learning_rate": 8.865677138113803e-05,
+      "loss": 3.6094,
+      "step": 7510
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5861921906471252,
+      "learning_rate": 8.864206613901189e-05,
+      "loss": 3.6076,
+      "step": 7515
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.63309645652771,
+      "learning_rate": 8.862735259243832e-05,
+      "loss": 3.5824,
+      "step": 7520
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6046217679977417,
+      "learning_rate": 8.861263074457941e-05,
+      "loss": 3.5087,
+      "step": 7525
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6083025932312012,
+      "learning_rate": 8.859790059859893e-05,
+      "loss": 3.6194,
+      "step": 7530
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6252034902572632,
+      "learning_rate": 8.858316215766252e-05,
+      "loss": 3.5869,
+      "step": 7535
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5797540545463562,
+      "learning_rate": 8.856841542493754e-05,
+      "loss": 3.5189,
+      "step": 7540
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6712521314620972,
+      "learning_rate": 8.855366040359321e-05,
+      "loss": 3.5287,
+      "step": 7545
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6107014417648315,
+      "learning_rate": 8.853889709680045e-05,
+      "loss": 3.5669,
+      "step": 7550
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5952281355857849,
+      "learning_rate": 8.852412550773202e-05,
+      "loss": 3.6078,
+      "step": 7555
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6131510734558105,
+      "learning_rate": 8.850934563956244e-05,
+      "loss": 3.5107,
+      "step": 7560
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6593649387359619,
+      "learning_rate": 8.849455749546797e-05,
+      "loss": 3.64,
+      "step": 7565
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6309821009635925,
+      "learning_rate": 8.847976107862673e-05,
+      "loss": 3.6651,
+      "step": 7570
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6182429790496826,
+      "learning_rate": 8.846495639221854e-05,
+      "loss": 3.6732,
+      "step": 7575
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6071635484695435,
+      "learning_rate": 8.845014343942505e-05,
+      "loss": 3.596,
+      "step": 7580
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7131217122077942,
+      "learning_rate": 8.843532222342965e-05,
+      "loss": 3.5648,
+      "step": 7585
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1229063272476196,
+      "learning_rate": 8.842049274741753e-05,
+      "loss": 3.5761,
+      "step": 7590
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6005117297172546,
+      "learning_rate": 8.840565501457565e-05,
+      "loss": 3.5917,
+      "step": 7595
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7124444246292114,
+      "learning_rate": 8.839080902809274e-05,
+      "loss": 3.5306,
+      "step": 7600
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8110289573669434,
+      "learning_rate": 8.83759547911593e-05,
+      "loss": 3.551,
+      "step": 7605
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6084639430046082,
+      "learning_rate": 8.83610923069676e-05,
+      "loss": 3.5729,
+      "step": 7610
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6201856732368469,
+      "learning_rate": 8.83462215787117e-05,
+      "loss": 3.4529,
+      "step": 7615
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.614809513092041,
+      "learning_rate": 8.833134260958745e-05,
+      "loss": 3.4328,
+      "step": 7620
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6645103693008423,
+      "learning_rate": 8.831645540279238e-05,
+      "loss": 3.5735,
+      "step": 7625
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6269055008888245,
+      "learning_rate": 8.830155996152591e-05,
+      "loss": 3.5334,
+      "step": 7630
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6366388201713562,
+      "learning_rate": 8.828665628898918e-05,
+      "loss": 3.7404,
+      "step": 7635
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6442884206771851,
+      "learning_rate": 8.827174438838503e-05,
+      "loss": 3.5471,
+      "step": 7640
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6180920600891113,
+      "learning_rate": 8.825682426291819e-05,
+      "loss": 3.5981,
+      "step": 7645
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.578211784362793,
+      "learning_rate": 8.824189591579505e-05,
+      "loss": 3.6139,
+      "step": 7650
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6585428714752197,
+      "learning_rate": 8.822695935022387e-05,
+      "loss": 3.5208,
+      "step": 7655
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6824387907981873,
+      "learning_rate": 8.821201456941458e-05,
+      "loss": 3.568,
+      "step": 7660
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5872287750244141,
+      "learning_rate": 8.819706157657893e-05,
+      "loss": 3.5758,
+      "step": 7665
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6620607376098633,
+      "learning_rate": 8.818210037493042e-05,
+      "loss": 3.5385,
+      "step": 7670
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5877419710159302,
+      "learning_rate": 8.816713096768431e-05,
+      "loss": 3.5289,
+      "step": 7675
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6390382051467896,
+      "learning_rate": 8.815215335805765e-05,
+      "loss": 3.6195,
+      "step": 7680
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6211628317832947,
+      "learning_rate": 8.81371675492692e-05,
+      "loss": 3.5081,
+      "step": 7685
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.645075798034668,
+      "learning_rate": 8.812217354453956e-05,
+      "loss": 3.522,
+      "step": 7690
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5981563925743103,
+      "learning_rate": 8.810717134709099e-05,
+      "loss": 3.4559,
+      "step": 7695
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6034389138221741,
+      "learning_rate": 8.809216096014759e-05,
+      "loss": 3.5619,
+      "step": 7700
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7003628015518188,
+      "learning_rate": 8.807714238693521e-05,
+      "loss": 3.3979,
+      "step": 7705
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6121801137924194,
+      "learning_rate": 8.806211563068145e-05,
+      "loss": 3.5409,
+      "step": 7710
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6418849229812622,
+      "learning_rate": 8.804708069461563e-05,
+      "loss": 3.4039,
+      "step": 7715
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6116931438446045,
+      "learning_rate": 8.803203758196887e-05,
+      "loss": 3.6759,
+      "step": 7720
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6935160160064697,
+      "learning_rate": 8.801698629597407e-05,
+      "loss": 3.6147,
+      "step": 7725
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6301100254058838,
+      "learning_rate": 8.800192683986583e-05,
+      "loss": 3.4998,
+      "step": 7730
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6241505146026611,
+      "learning_rate": 8.798685921688053e-05,
+      "loss": 3.5505,
+      "step": 7735
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6059820652008057,
+      "learning_rate": 8.797178343025631e-05,
+      "loss": 3.5795,
+      "step": 7740
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6309882402420044,
+      "learning_rate": 8.795669948323308e-05,
+      "loss": 3.7013,
+      "step": 7745
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6364321112632751,
+      "learning_rate": 8.794160737905245e-05,
+      "loss": 3.5318,
+      "step": 7750
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6866723895072937,
+      "learning_rate": 8.792650712095784e-05,
+      "loss": 3.5518,
+      "step": 7755
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6202917098999023,
+      "learning_rate": 8.791139871219441e-05,
+      "loss": 3.4917,
+      "step": 7760
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6145130395889282,
+      "learning_rate": 8.789628215600902e-05,
+      "loss": 3.567,
+      "step": 7765
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6332247853279114,
+      "learning_rate": 8.788115745565036e-05,
+      "loss": 3.5201,
+      "step": 7770
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5983673334121704,
+      "learning_rate": 8.786602461436884e-05,
+      "loss": 3.4794,
+      "step": 7775
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.600166916847229,
+      "learning_rate": 8.785088363541657e-05,
+      "loss": 3.6497,
+      "step": 7780
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6262897253036499,
+      "learning_rate": 8.78357345220475e-05,
+      "loss": 3.528,
+      "step": 7785
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.603503942489624,
+      "learning_rate": 8.782057727751722e-05,
+      "loss": 3.6122,
+      "step": 7790
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6356224417686462,
+      "learning_rate": 8.78054119050832e-05,
+      "loss": 3.5875,
+      "step": 7795
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6476446390151978,
+      "learning_rate": 8.779023840800452e-05,
+      "loss": 3.5356,
+      "step": 7800
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.620455265045166,
+      "learning_rate": 8.777505678954211e-05,
+      "loss": 3.5682,
+      "step": 7805
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6363736987113953,
+      "learning_rate": 8.775986705295858e-05,
+      "loss": 3.4729,
+      "step": 7810
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6706913113594055,
+      "learning_rate": 8.774466920151833e-05,
+      "loss": 3.585,
+      "step": 7815
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5840045809745789,
+      "learning_rate": 8.772946323848746e-05,
+      "loss": 3.5445,
+      "step": 7820
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5783376097679138,
+      "learning_rate": 8.771424916713386e-05,
+      "loss": 3.5587,
+      "step": 7825
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.625701904296875,
+      "learning_rate": 8.769902699072712e-05,
+      "loss": 3.5002,
+      "step": 7830
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5966210961341858,
+      "learning_rate": 8.768379671253859e-05,
+      "loss": 3.5735,
+      "step": 7835
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6389672160148621,
+      "learning_rate": 8.76685583358414e-05,
+      "loss": 3.6013,
+      "step": 7840
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5734913349151611,
+      "learning_rate": 8.765331186391033e-05,
+      "loss": 3.6311,
+      "step": 7845
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6379283666610718,
+      "learning_rate": 8.763805730002199e-05,
+      "loss": 3.5382,
+      "step": 7850
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5836450457572937,
+      "learning_rate": 8.762279464745466e-05,
+      "loss": 3.6118,
+      "step": 7855
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8633871674537659,
+      "learning_rate": 8.760752390948843e-05,
+      "loss": 3.4767,
+      "step": 7860
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6150323748588562,
+      "learning_rate": 8.759224508940505e-05,
+      "loss": 3.5932,
+      "step": 7865
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6048701405525208,
+      "learning_rate": 8.757695819048808e-05,
+      "loss": 3.5043,
+      "step": 7870
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6121411323547363,
+      "learning_rate": 8.756166321602273e-05,
+      "loss": 3.5673,
+      "step": 7875
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6299585103988647,
+      "learning_rate": 8.754636016929604e-05,
+      "loss": 3.6326,
+      "step": 7880
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6321864128112793,
+      "learning_rate": 8.753104905359674e-05,
+      "loss": 3.5809,
+      "step": 7885
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1350053548812866,
+      "learning_rate": 8.751572987221526e-05,
+      "loss": 3.5039,
+      "step": 7890
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5985552072525024,
+      "learning_rate": 8.750040262844384e-05,
+      "loss": 3.4951,
+      "step": 7895
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6465233564376831,
+      "learning_rate": 8.748506732557637e-05,
+      "loss": 3.5941,
+      "step": 7900
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9732844829559326,
+      "learning_rate": 8.746972396690853e-05,
+      "loss": 3.5643,
+      "step": 7905
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.631447434425354,
+      "learning_rate": 8.745437255573773e-05,
+      "loss": 3.6104,
+      "step": 7910
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6856198906898499,
+      "learning_rate": 8.743901309536307e-05,
+      "loss": 3.5252,
+      "step": 7915
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5918905735015869,
+      "learning_rate": 8.742364558908543e-05,
+      "loss": 3.4826,
+      "step": 7920
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6145909428596497,
+      "learning_rate": 8.740827004020735e-05,
+      "loss": 3.608,
+      "step": 7925
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5764964818954468,
+      "learning_rate": 8.739288645203319e-05,
+      "loss": 3.5595,
+      "step": 7930
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6890547871589661,
+      "learning_rate": 8.737749482786897e-05,
+      "loss": 3.6249,
+      "step": 7935
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5963539481163025,
+      "learning_rate": 8.736209517102247e-05,
+      "loss": 3.4614,
+      "step": 7940
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6819955110549927,
+      "learning_rate": 8.734668748480315e-05,
+      "loss": 3.5262,
+      "step": 7945
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6976768374443054,
+      "learning_rate": 8.733127177252225e-05,
+      "loss": 3.475,
+      "step": 7950
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.596578061580658,
+      "learning_rate": 8.731584803749271e-05,
+      "loss": 3.4844,
+      "step": 7955
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6109974980354309,
+      "learning_rate": 8.73004162830292e-05,
+      "loss": 3.5886,
+      "step": 7960
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6416093707084656,
+      "learning_rate": 8.728497651244812e-05,
+      "loss": 3.4466,
+      "step": 7965
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6429569721221924,
+      "learning_rate": 8.726952872906755e-05,
+      "loss": 3.5273,
+      "step": 7970
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5707230567932129,
+      "learning_rate": 8.725407293620737e-05,
+      "loss": 3.4936,
+      "step": 7975
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6317285299301147,
+      "learning_rate": 8.72386091371891e-05,
+      "loss": 3.5684,
+      "step": 7980
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.662795901298523,
+      "learning_rate": 8.722313733533603e-05,
+      "loss": 3.4832,
+      "step": 7985
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6150239706039429,
+      "learning_rate": 8.720765753397317e-05,
+      "loss": 3.5294,
+      "step": 7990
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6103739738464355,
+      "learning_rate": 8.719216973642723e-05,
+      "loss": 3.5568,
+      "step": 7995
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.578384518623352,
+      "learning_rate": 8.717667394602664e-05,
+      "loss": 3.5777,
+      "step": 8000
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 33984,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "total_flos": 2.208473472093389e+19,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}