{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 3125,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0032,
      "grad_norm": 3.097726066927376,
      "learning_rate": 2.875399361022364e-07,
      "loss": 0.4182,
      "step": 10
    },
    {
      "epoch": 0.0064,
      "grad_norm": 2.6698471287095344,
      "learning_rate": 6.070287539936103e-07,
      "loss": 0.3969,
      "step": 20
    },
    {
      "epoch": 0.0096,
      "grad_norm": 2.3652285610261403,
      "learning_rate": 9.265175718849841e-07,
      "loss": 0.3746,
      "step": 30
    },
    {
      "epoch": 0.0128,
      "grad_norm": 1.4765456643161126,
      "learning_rate": 1.2460063897763578e-06,
      "loss": 0.3695,
      "step": 40
    },
    {
      "epoch": 0.016,
      "grad_norm": 2.191519526172484,
      "learning_rate": 1.565495207667732e-06,
      "loss": 0.3449,
      "step": 50
    },
    {
      "epoch": 0.0192,
      "grad_norm": 1.6285826522074962,
      "learning_rate": 1.8849840255591056e-06,
      "loss": 0.3181,
      "step": 60
    },
    {
      "epoch": 0.0224,
      "grad_norm": 1.8683549636521075,
      "learning_rate": 2.2044728434504793e-06,
      "loss": 0.322,
      "step": 70
    },
    {
      "epoch": 0.0256,
      "grad_norm": 1.5301019294828755,
      "learning_rate": 2.5239616613418532e-06,
      "loss": 0.3009,
      "step": 80
    },
    {
      "epoch": 0.0288,
      "grad_norm": 1.538896137107791,
      "learning_rate": 2.8434504792332267e-06,
      "loss": 0.293,
      "step": 90
    },
    {
      "epoch": 0.032,
      "grad_norm": 1.6984302868653567,
      "learning_rate": 3.162939297124601e-06,
      "loss": 0.2999,
      "step": 100
    },
    {
      "epoch": 0.0352,
      "grad_norm": 1.7119603328381132,
      "learning_rate": 3.482428115015975e-06,
      "loss": 0.2945,
      "step": 110
    },
    {
      "epoch": 0.0384,
      "grad_norm": 1.6069777286559,
      "learning_rate": 3.8019169329073485e-06,
      "loss": 0.2895,
      "step": 120
    },
    {
      "epoch": 0.0416,
      "grad_norm": 1.7566779999753588,
      "learning_rate": 4.121405750798722e-06,
      "loss": 0.2923,
      "step": 130
    },
    {
      "epoch": 0.0448,
      "grad_norm": 1.7352248180581427,
      "learning_rate": 4.440894568690096e-06,
      "loss": 0.3638,
      "step": 140
    },
    {
      "epoch": 0.048,
      "grad_norm": 1.8483046207545708,
      "learning_rate": 4.76038338658147e-06,
      "loss": 0.2861,
      "step": 150
    },
    {
      "epoch": 0.0512,
      "grad_norm": 1.6829534856608535,
      "learning_rate": 5.079872204472844e-06,
      "loss": 0.5296,
      "step": 160
    },
    {
      "epoch": 0.0544,
      "grad_norm": 1.7746050883459261,
      "learning_rate": 5.399361022364218e-06,
      "loss": 0.2934,
      "step": 170
    },
    {
      "epoch": 0.0576,
      "grad_norm": 1.7179977616371749,
      "learning_rate": 5.718849840255591e-06,
      "loss": 0.2713,
      "step": 180
    },
    {
      "epoch": 0.0608,
      "grad_norm": 1.5706330739618684,
      "learning_rate": 6.038338658146965e-06,
      "loss": 0.2779,
      "step": 190
    },
    {
      "epoch": 0.064,
      "grad_norm": 1.8232979139526881,
      "learning_rate": 6.35782747603834e-06,
      "loss": 0.2835,
      "step": 200
    },
    {
      "epoch": 0.0672,
      "grad_norm": 1.755606637900451,
      "learning_rate": 6.677316293929713e-06,
      "loss": 0.3767,
      "step": 210
    },
    {
      "epoch": 0.0704,
      "grad_norm": 1.7142425655315976,
      "learning_rate": 6.996805111821087e-06,
      "loss": 0.2864,
      "step": 220
    },
    {
      "epoch": 0.0736,
      "grad_norm": 1.6051059574805915,
      "learning_rate": 7.316293929712461e-06,
      "loss": 0.2767,
      "step": 230
    },
    {
      "epoch": 0.0768,
      "grad_norm": 1.49334442261346,
      "learning_rate": 7.635782747603835e-06,
      "loss": 0.2689,
      "step": 240
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.6270294243342234,
      "learning_rate": 7.955271565495208e-06,
      "loss": 0.2857,
      "step": 250
    },
    {
      "epoch": 0.0832,
      "grad_norm": 1.4820640434767234,
      "learning_rate": 8.274760383386582e-06,
      "loss": 0.2838,
      "step": 260
    },
    {
      "epoch": 0.0864,
      "grad_norm": 1.4755509754091296,
      "learning_rate": 8.594249201277956e-06,
      "loss": 0.2829,
      "step": 270
    },
    {
      "epoch": 0.0896,
      "grad_norm": 1.5750758252112724,
      "learning_rate": 8.91373801916933e-06,
      "loss": 0.2716,
      "step": 280
    },
    {
      "epoch": 0.0928,
      "grad_norm": 1.8211062452729152,
      "learning_rate": 9.233226837060704e-06,
      "loss": 0.2985,
      "step": 290
    },
    {
      "epoch": 0.096,
      "grad_norm": 1.470802373475764,
      "learning_rate": 9.552715654952077e-06,
      "loss": 0.2722,
      "step": 300
    },
    {
      "epoch": 0.0992,
      "grad_norm": 1.5951743640227587,
      "learning_rate": 9.87220447284345e-06,
      "loss": 0.2834,
      "step": 310
    },
    {
      "epoch": 0.1024,
      "grad_norm": 1.5085331338575516,
      "learning_rate": 9.999887666317538e-06,
      "loss": 0.2866,
      "step": 320
    },
    {
      "epoch": 0.1056,
      "grad_norm": 1.5890905665931712,
      "learning_rate": 9.999201200981566e-06,
      "loss": 0.2774,
      "step": 330
    },
    {
      "epoch": 0.1088,
      "grad_norm": 1.4824327210922925,
      "learning_rate": 9.997890763487869e-06,
      "loss": 0.2865,
      "step": 340
    },
    {
      "epoch": 0.112,
      "grad_norm": 1.4179211935288807,
      "learning_rate": 9.995956517397884e-06,
      "loss": 0.3445,
      "step": 350
    },
    {
      "epoch": 0.1152,
      "grad_norm": 1.6311477640409462,
      "learning_rate": 9.993398704133318e-06,
      "loss": 0.2961,
      "step": 360
    },
    {
      "epoch": 0.1184,
      "grad_norm": 1.3493949424853904,
      "learning_rate": 9.990217642946028e-06,
      "loss": 0.2693,
      "step": 370
    },
    {
      "epoch": 0.1216,
      "grad_norm": 1.5074472178009761,
      "learning_rate": 9.986413730878168e-06,
      "loss": 0.2908,
      "step": 380
    },
    {
      "epoch": 0.1248,
      "grad_norm": 1.5552421320809011,
      "learning_rate": 9.981987442712634e-06,
      "loss": 0.3047,
      "step": 390
    },
    {
      "epoch": 0.128,
      "grad_norm": 1.5085252753416616,
      "learning_rate": 9.976939330913801e-06,
      "loss": 0.2761,
      "step": 400
    },
    {
      "epoch": 0.1312,
      "grad_norm": 1.3894985533924997,
      "learning_rate": 9.971270025558576e-06,
      "loss": 0.2827,
      "step": 410
    },
    {
      "epoch": 0.1344,
      "grad_norm": 1.376517568828387,
      "learning_rate": 9.96498023425774e-06,
      "loss": 0.282,
      "step": 420
    },
    {
      "epoch": 0.1376,
      "grad_norm": 1.293372373569424,
      "learning_rate": 9.958070742067649e-06,
      "loss": 0.2864,
      "step": 430
    },
    {
      "epoch": 0.1408,
      "grad_norm": 1.3933504307268871,
      "learning_rate": 9.95054241139223e-06,
      "loss": 0.2886,
      "step": 440
    },
    {
      "epoch": 0.144,
      "grad_norm": 1.5512528120422318,
      "learning_rate": 9.942396181875342e-06,
      "loss": 0.2792,
      "step": 450
    },
    {
      "epoch": 0.1472,
      "grad_norm": 1.3955106828693524,
      "learning_rate": 9.933633070283512e-06,
      "loss": 0.2823,
      "step": 460
    },
    {
      "epoch": 0.1504,
      "grad_norm": 1.6468424852669425,
      "learning_rate": 9.924254170379007e-06,
      "loss": 0.2984,
      "step": 470
    },
    {
      "epoch": 0.1536,
      "grad_norm": 1.5159031541701018,
      "learning_rate": 9.914260652783323e-06,
      "loss": 0.271,
      "step": 480
    },
    {
      "epoch": 0.1568,
      "grad_norm": 1.5805212725674178,
      "learning_rate": 9.903653764831088e-06,
      "loss": 0.2957,
      "step": 490
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.1956547216731315,
      "learning_rate": 9.892434830414354e-06,
      "loss": 0.2684,
      "step": 500
    },
    {
      "epoch": 0.1632,
      "grad_norm": 1.4268809698190317,
      "learning_rate": 9.880605249817377e-06,
      "loss": 0.3657,
      "step": 510
    },
    {
      "epoch": 0.1664,
      "grad_norm": 1.4179964964215248,
      "learning_rate": 9.868166499541824e-06,
      "loss": 0.2751,
      "step": 520
    },
    {
      "epoch": 0.1696,
      "grad_norm": 1.2892232060188915,
      "learning_rate": 9.855120132122503e-06,
      "loss": 0.2729,
      "step": 530
    },
    {
      "epoch": 0.1728,
      "grad_norm": 1.261114838277135,
      "learning_rate": 9.841467775933566e-06,
      "loss": 0.268,
      "step": 540
    },
    {
      "epoch": 0.176,
      "grad_norm": 1.430271043321,
      "learning_rate": 9.827211134985273e-06,
      "loss": 0.4323,
      "step": 550
    },
    {
      "epoch": 0.1792,
      "grad_norm": 1.2954706652401755,
      "learning_rate": 9.812351988711312e-06,
      "loss": 0.2622,
      "step": 560
    },
    {
      "epoch": 0.1824,
      "grad_norm": 1.7636694848560834,
      "learning_rate": 9.79689219174669e-06,
      "loss": 0.3947,
      "step": 570
    },
    {
      "epoch": 0.1856,
      "grad_norm": 1.5362869932266252,
      "learning_rate": 9.780833673696255e-06,
      "loss": 0.261,
      "step": 580
    },
    {
      "epoch": 0.1888,
      "grad_norm": 1.2516741612873126,
      "learning_rate": 9.76417843889385e-06,
      "loss": 0.2632,
      "step": 590
    },
    {
      "epoch": 0.192,
      "grad_norm": 1.330313742879521,
      "learning_rate": 9.746928566152148e-06,
      "loss": 0.2612,
      "step": 600
    },
    {
      "epoch": 0.1952,
      "grad_norm": 1.369062918644532,
      "learning_rate": 9.729086208503174e-06,
      "loss": 0.2672,
      "step": 610
    },
    {
      "epoch": 0.1984,
      "grad_norm": 1.4183586038757179,
      "learning_rate": 9.710653592929595e-06,
      "loss": 0.2738,
      "step": 620
    },
    {
      "epoch": 0.2016,
      "grad_norm": 1.52682494954233,
      "learning_rate": 9.691633020086745e-06,
      "loss": 0.3808,
      "step": 630
    },
    {
      "epoch": 0.2048,
      "grad_norm": 1.4376998703858688,
      "learning_rate": 9.672026864015476e-06,
      "loss": 0.2643,
      "step": 640
    },
    {
      "epoch": 0.208,
      "grad_norm": 1.583565617875829,
      "learning_rate": 9.651837571845842e-06,
      "loss": 0.2772,
      "step": 650
    },
    {
      "epoch": 0.2112,
      "grad_norm": 1.4806395536482786,
      "learning_rate": 9.631067663491663e-06,
      "loss": 0.3377,
      "step": 660
    },
    {
      "epoch": 0.2144,
      "grad_norm": 1.2496121580498412,
      "learning_rate": 9.609719731336005e-06,
      "loss": 0.2752,
      "step": 670
    },
    {
      "epoch": 0.2176,
      "grad_norm": 1.3439013535339106,
      "learning_rate": 9.587796439907609e-06,
      "loss": 0.2685,
      "step": 680
    },
    {
      "epoch": 0.2208,
      "grad_norm": 1.4305791278136326,
      "learning_rate": 9.565300525548327e-06,
      "loss": 0.2754,
      "step": 690
    },
    {
      "epoch": 0.224,
      "grad_norm": 1.4621614330412347,
      "learning_rate": 9.542234796071577e-06,
      "loss": 0.2627,
      "step": 700
    },
    {
      "epoch": 0.2272,
      "grad_norm": 1.3498296779237806,
      "learning_rate": 9.518602130411894e-06,
      "loss": 0.2861,
      "step": 710
    },
    {
      "epoch": 0.2304,
      "grad_norm": 1.3530290691335103,
      "learning_rate": 9.4944054782656e-06,
      "loss": 0.2668,
      "step": 720
    },
    {
      "epoch": 0.2336,
      "grad_norm": 1.5377603964582789,
      "learning_rate": 9.469647859722634e-06,
      "loss": 0.2679,
      "step": 730
    },
    {
      "epoch": 0.2368,
      "grad_norm": 1.225090999497597,
      "learning_rate": 9.444332364889603e-06,
      "loss": 0.2491,
      "step": 740
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.2724375938152441,
      "learning_rate": 9.41846215350409e-06,
      "loss": 0.2756,
      "step": 750
    },
    {
      "epoch": 0.2432,
      "grad_norm": 1.3649187680293995,
      "learning_rate": 9.392040454540284e-06,
      "loss": 0.2685,
      "step": 760
    },
    {
      "epoch": 0.2464,
      "grad_norm": 1.506689299970985,
      "learning_rate": 9.365070565805941e-06,
      "loss": 0.2727,
      "step": 770
    },
    {
      "epoch": 0.2496,
      "grad_norm": 1.5365021125318565,
      "learning_rate": 9.337555853530785e-06,
      "loss": 0.2695,
      "step": 780
    },
    {
      "epoch": 0.2528,
      "grad_norm": 1.4793118284689457,
      "learning_rate": 9.309499751946345e-06,
      "loss": 0.2712,
      "step": 790
    },
    {
      "epoch": 0.256,
      "grad_norm": 1.3520775224079904,
      "learning_rate": 9.280905762857315e-06,
      "loss": 0.2658,
      "step": 800
    },
    {
      "epoch": 0.2592,
      "grad_norm": 1.0943480269165575,
      "learning_rate": 9.251777455204485e-06,
      "loss": 0.2699,
      "step": 810
    },
    {
      "epoch": 0.2624,
      "grad_norm": 1.4272492824528293,
      "learning_rate": 9.222118464619278e-06,
      "loss": 0.2552,
      "step": 820
    },
    {
      "epoch": 0.2656,
      "grad_norm": 2.210970619104712,
      "learning_rate": 9.191932492969972e-06,
      "loss": 0.2609,
      "step": 830
    },
    {
      "epoch": 0.2688,
      "grad_norm": 1.3860115638657557,
      "learning_rate": 9.161223307899659e-06,
      "loss": 0.2614,
      "step": 840
    },
    {
      "epoch": 0.272,
      "grad_norm": 1.3097054200264115,
      "learning_rate": 9.129994742355985e-06,
      "loss": 0.2558,
      "step": 850
    },
    {
      "epoch": 0.2752,
      "grad_norm": 1.2863046042758517,
      "learning_rate": 9.09825069411274e-06,
      "loss": 0.3506,
      "step": 860
    },
    {
      "epoch": 0.2784,
      "grad_norm": 1.4154177255509635,
      "learning_rate": 9.065995125283367e-06,
      "loss": 0.3483,
      "step": 870
    },
    {
      "epoch": 0.2816,
      "grad_norm": 1.2614915111427545,
      "learning_rate": 9.033232061826428e-06,
      "loss": 0.3094,
      "step": 880
    },
    {
      "epoch": 0.2848,
      "grad_norm": 1.392215585602565,
      "learning_rate": 8.999965593043113e-06,
      "loss": 0.2552,
      "step": 890
    },
    {
      "epoch": 0.288,
      "grad_norm": 1.2787168993445066,
      "learning_rate": 8.96619987106682e-06,
      "loss": 0.2635,
      "step": 900
    },
    {
      "epoch": 0.2912,
      "grad_norm": 1.1874290348203185,
      "learning_rate": 8.931939110344935e-06,
      "loss": 0.3925,
      "step": 910
    },
    {
      "epoch": 0.2944,
      "grad_norm": 1.2892860612183799,
      "learning_rate": 8.897187587112783e-06,
      "loss": 0.472,
      "step": 920
    },
    {
      "epoch": 0.2976,
      "grad_norm": 1.1180916115878632,
      "learning_rate": 8.861949638859908e-06,
      "loss": 0.4072,
      "step": 930
    },
    {
      "epoch": 0.3008,
      "grad_norm": 1.311746880897253,
      "learning_rate": 8.826229663788688e-06,
      "loss": 0.2632,
      "step": 940
    },
    {
      "epoch": 0.304,
      "grad_norm": 1.2164327029930522,
      "learning_rate": 8.790032120265373e-06,
      "loss": 0.2532,
      "step": 950
    },
    {
      "epoch": 0.3072,
      "grad_norm": 1.436604570629404,
      "learning_rate": 8.753361526263622e-06,
      "loss": 0.2591,
      "step": 960
    },
    {
      "epoch": 0.3104,
      "grad_norm": 1.2626583454725855,
      "learning_rate": 8.716222458800591e-06,
      "loss": 0.2617,
      "step": 970
    },
    {
      "epoch": 0.3136,
      "grad_norm": 1.5056105724631803,
      "learning_rate": 8.67861955336566e-06,
      "loss": 0.2622,
      "step": 980
    },
    {
      "epoch": 0.3168,
      "grad_norm": 1.4940608590847797,
      "learning_rate": 8.640557503341843e-06,
      "loss": 0.2689,
      "step": 990
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.4135983918557036,
      "learning_rate": 8.602041059420017e-06,
      "loss": 0.2599,
      "step": 1000
    },
    {
      "epoch": 0.3232,
      "grad_norm": 1.2654674588600259,
      "learning_rate": 8.563075029005924e-06,
      "loss": 0.2598,
      "step": 1010
    },
    {
      "epoch": 0.3264,
      "grad_norm": 1.5394425180175606,
      "learning_rate": 8.523664275620185e-06,
      "loss": 0.284,
      "step": 1020
    },
    {
      "epoch": 0.3296,
      "grad_norm": 1.2453557042164298,
      "learning_rate": 8.483813718291223e-06,
      "loss": 0.2475,
      "step": 1030
    },
    {
      "epoch": 0.3328,
      "grad_norm": 1.3932335036293408,
      "learning_rate": 8.443528330941322e-06,
      "loss": 0.2733,
      "step": 1040
    },
    {
      "epoch": 0.336,
      "grad_norm": 1.3326668118526488,
      "learning_rate": 8.402813141765796e-06,
      "loss": 0.2706,
      "step": 1050
    },
    {
      "epoch": 0.3392,
      "grad_norm": 1.4498463244679263,
      "learning_rate": 8.361673232605408e-06,
      "loss": 0.2599,
      "step": 1060
    },
    {
      "epoch": 0.3424,
      "grad_norm": 1.5505084458466836,
      "learning_rate": 8.320113738312081e-06,
      "loss": 0.2582,
      "step": 1070
    },
    {
      "epoch": 0.3456,
      "grad_norm": 1.3933287663129275,
      "learning_rate": 8.27813984610799e-06,
      "loss": 0.251,
      "step": 1080
    },
    {
      "epoch": 0.3488,
      "grad_norm": 1.3411841017130814,
      "learning_rate": 8.235756794938123e-06,
      "loss": 0.2545,
      "step": 1090
    },
    {
      "epoch": 0.352,
      "grad_norm": 1.2762415468326629,
      "learning_rate": 8.19296987481639e-06,
      "loss": 0.2511,
      "step": 1100
    },
    {
      "epoch": 0.3552,
      "grad_norm": 1.3756624147468053,
      "learning_rate": 8.149784426165351e-06,
      "loss": 0.2471,
      "step": 1110
    },
    {
      "epoch": 0.3584,
      "grad_norm": 1.2367975395806792,
      "learning_rate": 8.106205839149653e-06,
      "loss": 0.2398,
      "step": 1120
    },
    {
      "epoch": 0.3616,
      "grad_norm": 1.2795711630605267,
      "learning_rate": 8.06223955300326e-06,
      "loss": 0.25,
      "step": 1130
    },
    {
      "epoch": 0.3648,
      "grad_norm": 1.181366173079793,
      "learning_rate": 8.017891055350563e-06,
      "loss": 0.3787,
      "step": 1140
    },
    {
      "epoch": 0.368,
      "grad_norm": 1.235985789921985,
      "learning_rate": 7.973165881521435e-06,
      "loss": 0.2489,
      "step": 1150
    },
    {
      "epoch": 0.3712,
      "grad_norm": 1.3716107886978977,
      "learning_rate": 7.928069613860357e-06,
      "loss": 0.2638,
      "step": 1160
    },
    {
      "epoch": 0.3744,
      "grad_norm": 1.3314537160679019,
      "learning_rate": 7.882607881029652e-06,
      "loss": 0.2666,
      "step": 1170
    },
    {
      "epoch": 0.3776,
      "grad_norm": 1.47532226068613,
      "learning_rate": 7.836786357306943e-06,
      "loss": 0.2469,
      "step": 1180
    },
    {
      "epoch": 0.3808,
      "grad_norm": 1.2910123474743094,
      "learning_rate": 7.790610761876936e-06,
      "loss": 0.245,
      "step": 1190
    },
    {
      "epoch": 0.384,
      "grad_norm": 1.219675861544086,
      "learning_rate": 7.744086858117565e-06,
      "loss": 0.2532,
      "step": 1200
    },
    {
      "epoch": 0.3872,
      "grad_norm": 1.2663481766467186,
      "learning_rate": 7.69722045288066e-06,
      "loss": 0.2648,
      "step": 1210
    },
    {
      "epoch": 0.3904,
      "grad_norm": 1.2601430683107997,
      "learning_rate": 7.650017395767149e-06,
      "loss": 0.2428,
      "step": 1220
    },
    {
      "epoch": 0.3936,
      "grad_norm": 1.306765832939207,
      "learning_rate": 7.602483578396955e-06,
      "loss": 0.244,
      "step": 1230
    },
    {
      "epoch": 0.3968,
      "grad_norm": 1.3723024837409241,
      "learning_rate": 7.554624933673638e-06,
      "loss": 0.3634,
      "step": 1240
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.5507546450379777,
      "learning_rate": 7.5064474350438755e-06,
      "loss": 0.2618,
      "step": 1250
    },
    {
      "epoch": 0.4032,
      "grad_norm": 1.366012809594849,
      "learning_rate": 7.457957095751896e-06,
      "loss": 0.2487,
      "step": 1260
    },
    {
      "epoch": 0.4064,
      "grad_norm": 1.1873848911132183,
      "learning_rate": 7.4091599680889425e-06,
      "loss": 0.2553,
      "step": 1270
    },
    {
      "epoch": 0.4096,
      "grad_norm": 1.4014484490952397,
      "learning_rate": 7.3600621426378515e-06,
      "loss": 0.2402,
      "step": 1280
    },
    {
      "epoch": 0.4128,
      "grad_norm": 1.1617991472531028,
      "learning_rate": 7.3106697475128655e-06,
      "loss": 0.2569,
      "step": 1290
    },
    {
      "epoch": 0.416,
      "grad_norm": 1.3751891734766228,
      "learning_rate": 7.260988947594759e-06,
      "loss": 0.2488,
      "step": 1300
    },
    {
      "epoch": 0.4192,
      "grad_norm": 1.237845283871077,
      "learning_rate": 7.211025943761367e-06,
      "loss": 0.2454,
      "step": 1310
    },
    {
      "epoch": 0.4224,
      "grad_norm": 1.1628031806651409,
      "learning_rate": 7.160786972113627e-06,
      "loss": 0.2554,
      "step": 1320
    },
    {
      "epoch": 0.4256,
      "grad_norm": 1.3404586293984995,
      "learning_rate": 7.1102783031972326e-06,
      "loss": 0.2551,
      "step": 1330
    },
    {
      "epoch": 0.4288,
      "grad_norm": 1.4670462801897188,
      "learning_rate": 7.059506241219964e-06,
      "loss": 0.2502,
      "step": 1340
    },
    {
      "epoch": 0.432,
      "grad_norm": 1.0302147200635003,
      "learning_rate": 7.008477123264849e-06,
      "loss": 0.2295,
      "step": 1350
    },
    {
      "epoch": 0.4352,
      "grad_norm": 1.2824619758646878,
      "learning_rate": 6.957197318499187e-06,
      "loss": 0.2539,
      "step": 1360
    },
    {
      "epoch": 0.4384,
      "grad_norm": 1.3608888568790518,
      "learning_rate": 6.905673227379606e-06,
      "loss": 0.342,
      "step": 1370
    },
    {
      "epoch": 0.4416,
      "grad_norm": 1.2538554464958942,
      "learning_rate": 6.853911280853168e-06,
      "loss": 0.2405,
      "step": 1380
    },
    {
      "epoch": 0.4448,
      "grad_norm": 1.377080653341555,
      "learning_rate": 6.801917939554721e-06,
      "loss": 0.246,
      "step": 1390
    },
    {
      "epoch": 0.448,
      "grad_norm": 1.295752065257539,
      "learning_rate": 6.749699693000495e-06,
      "loss": 0.238,
      "step": 1400
    },
    {
      "epoch": 0.4512,
      "grad_norm": 1.1623556510041717,
      "learning_rate": 6.6972630587781385e-06,
      "loss": 0.3522,
      "step": 1410
    },
    {
      "epoch": 0.4544,
      "grad_norm": 1.0514412139989395,
      "learning_rate": 6.6446145817332105e-06,
      "loss": 0.2448,
      "step": 1420
    },
    {
      "epoch": 0.4576,
      "grad_norm": 1.3006684707343785,
      "learning_rate": 6.591760833152306e-06,
      "loss": 0.2388,
      "step": 1430
    },
    {
      "epoch": 0.4608,
      "grad_norm": 1.3688639995228327,
      "learning_rate": 6.538708409942854e-06,
      "loss": 0.2484,
      "step": 1440
    },
    {
      "epoch": 0.464,
      "grad_norm": 1.182581571604436,
      "learning_rate": 6.48546393380973e-06,
      "loss": 0.2411,
      "step": 1450
    },
    {
      "epoch": 0.4672,
      "grad_norm": 1.2229123644507025,
      "learning_rate": 6.4320340504287825e-06,
      "loss": 0.2913,
      "step": 1460
    },
    {
      "epoch": 0.4704,
      "grad_norm": 1.114177546269806,
      "learning_rate": 6.378425428617343e-06,
      "loss": 0.5038,
      "step": 1470
    },
    {
      "epoch": 0.4736,
      "grad_norm": 1.400649143081864,
      "learning_rate": 6.324644759501869e-06,
      "loss": 0.2514,
      "step": 1480
    },
    {
      "epoch": 0.4768,
      "grad_norm": 1.2414663469351401,
      "learning_rate": 6.270698755682792e-06,
      "loss": 0.2434,
      "step": 1490
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.3094558123042377,
      "learning_rate": 6.2165941503966995e-06,
      "loss": 0.2486,
      "step": 1500
    },
    {
      "epoch": 0.4832,
      "grad_norm": 1.1308739268431762,
      "learning_rate": 6.162337696675909e-06,
      "loss": 0.224,
      "step": 1510
    },
    {
      "epoch": 0.4864,
      "grad_norm": 1.2921131535571986,
      "learning_rate": 6.107936166505615e-06,
      "loss": 0.25,
      "step": 1520
    },
    {
      "epoch": 0.4896,
      "grad_norm": 1.277856644085611,
      "learning_rate": 6.053396349978632e-06,
      "loss": 0.2418,
      "step": 1530
    },
    {
      "epoch": 0.4928,
      "grad_norm": 1.207819320049489,
      "learning_rate": 5.998725054447904e-06,
      "loss": 0.2429,
      "step": 1540
    },
    {
      "epoch": 0.496,
      "grad_norm": 1.3964622577997952,
      "learning_rate": 5.943929103676839e-06,
      "loss": 0.2339,
      "step": 1550
    },
    {
      "epoch": 0.4992,
      "grad_norm": 1.0444429321557864,
      "learning_rate": 5.889015336987614e-06,
      "loss": 0.2296,
      "step": 1560
    },
    {
      "epoch": 0.5024,
      "grad_norm": 1.000405925226846,
      "learning_rate": 5.833990608407525e-06,
      "loss": 0.2227,
      "step": 1570
    },
    {
      "epoch": 0.5056,
      "grad_norm": 1.243996070580087,
      "learning_rate": 5.778861785813508e-06,
      "loss": 0.228,
      "step": 1580
    },
    {
      "epoch": 0.5088,
      "grad_norm": 1.4163588613795817,
      "learning_rate": 5.723635750074924e-06,
      "loss": 0.262,
      "step": 1590
    },
    {
      "epoch": 0.512,
      "grad_norm": 1.2497990764938063,
      "learning_rate": 5.6683193941947365e-06,
      "loss": 0.2354,
      "step": 1600
    },
    {
      "epoch": 0.5152,
      "grad_norm": 1.0918427100928676,
      "learning_rate": 5.61291962244916e-06,
      "loss": 0.2343,
      "step": 1610
    },
    {
      "epoch": 0.5184,
      "grad_norm": 1.2313539946902519,
      "learning_rate": 5.5574433495259015e-06,
      "loss": 0.2224,
      "step": 1620
    },
    {
      "epoch": 0.5216,
      "grad_norm": 1.3110693967381668,
      "learning_rate": 5.501897499661123e-06,
      "loss": 0.2353,
      "step": 1630
    },
    {
      "epoch": 0.5248,
      "grad_norm": 1.3211565707814268,
      "learning_rate": 5.446289005775185e-06,
      "loss": 0.2448,
      "step": 1640
    },
    {
      "epoch": 0.528,
      "grad_norm": 1.336496780231579,
      "learning_rate": 5.390624808607321e-06,
      "loss": 0.2353,
      "step": 1650
    },
    {
      "epoch": 0.5312,
      "grad_norm": 1.4349961842702135,
      "learning_rate": 5.334911855849334e-06,
      "loss": 0.2344,
      "step": 1660
    },
    {
      "epoch": 0.5344,
      "grad_norm": 1.300276553470609,
      "learning_rate": 5.279157101278433e-06,
      "loss": 0.2488,
      "step": 1670
    },
    {
      "epoch": 0.5376,
      "grad_norm": 1.3938747275358123,
      "learning_rate": 5.2233675038892815e-06,
      "loss": 0.235,
      "step": 1680
    },
    {
      "epoch": 0.5408,
      "grad_norm": 1.0695710070831115,
      "learning_rate": 5.1675500270254385e-06,
      "loss": 0.2358,
      "step": 1690
    },
    {
      "epoch": 0.544,
      "grad_norm": 1.3464266574587438,
      "learning_rate": 5.111711637510216e-06,
      "loss": 0.262,
      "step": 1700
    },
    {
      "epoch": 0.5472,
      "grad_norm": 1.2054969480183788,
      "learning_rate": 5.055859304777127e-06,
      "loss": 0.2346,
      "step": 1710
    },
    {
      "epoch": 0.5504,
      "grad_norm": 1.592945506341502,
      "learning_rate": 5e-06,
      "loss": 0.2491,
      "step": 1720
    },
    {
      "epoch": 0.5536,
      "grad_norm": 1.0841849724930561,
      "learning_rate": 4.944140695222874e-06,
      "loss": 0.2373,
      "step": 1730
    },
    {
      "epoch": 0.5568,
      "grad_norm": 1.1462434966725683,
      "learning_rate": 4.888288362489786e-06,
      "loss": 0.2245,
      "step": 1740
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.2006687090364805,
      "learning_rate": 4.832449972974564e-06,
      "loss": 0.2281,
      "step": 1750
    },
    {
      "epoch": 0.5632,
      "grad_norm": 1.31046822802538,
      "learning_rate": 4.776632496110721e-06,
      "loss": 0.2343,
      "step": 1760
    },
    {
      "epoch": 0.5664,
      "grad_norm": 1.3130856213140962,
      "learning_rate": 4.720842898721569e-06,
      "loss": 0.3716,
      "step": 1770
    },
    {
      "epoch": 0.5696,
      "grad_norm": 1.3909345006491758,
      "learning_rate": 4.665088144150666e-06,
      "loss": 0.2344,
      "step": 1780
    },
    {
      "epoch": 0.5728,
      "grad_norm": 1.103425988971943,
      "learning_rate": 4.60937519139268e-06,
      "loss": 0.2207,
      "step": 1790
    },
    {
      "epoch": 0.576,
      "grad_norm": 1.3110124719599046,
      "learning_rate": 4.553710994224816e-06,
      "loss": 0.2584,
      "step": 1800
    },
    {
      "epoch": 0.5792,
      "grad_norm": 1.1358436006113255,
      "learning_rate": 4.498102500338879e-06,
      "loss": 0.2408,
      "step": 1810
    },
    {
      "epoch": 0.5824,
      "grad_norm": 1.1712875527581932,
      "learning_rate": 4.442556650474099e-06,
      "loss": 0.2338,
      "step": 1820
    },
    {
      "epoch": 0.5856,
      "grad_norm": 1.4200454612891542,
      "learning_rate": 4.387080377550843e-06,
      "loss": 0.3351,
      "step": 1830
    },
    {
      "epoch": 0.5888,
      "grad_norm": 1.1792392310210245,
      "learning_rate": 4.331680605805264e-06,
      "loss": 0.2284,
      "step": 1840
    },
    {
      "epoch": 0.592,
      "grad_norm": 1.4269025040790715,
      "learning_rate": 4.2763642499250765e-06,
      "loss": 0.2831,
      "step": 1850
    },
    {
      "epoch": 0.5952,
      "grad_norm": 1.2178218832805392,
      "learning_rate": 4.221138214186493e-06,
      "loss": 0.4656,
      "step": 1860
    },
    {
      "epoch": 0.5984,
      "grad_norm": 1.381880289226324,
      "learning_rate": 4.166009391592476e-06,
      "loss": 0.2226,
      "step": 1870
    },
    {
      "epoch": 0.6016,
      "grad_norm": 1.2861868856506182,
      "learning_rate": 4.110984663012388e-06,
      "loss": 0.2436,
      "step": 1880
    },
    {
      "epoch": 0.6048,
      "grad_norm": 1.460603893087498,
      "learning_rate": 4.056070896323163e-06,
      "loss": 0.2348,
      "step": 1890
    },
    {
      "epoch": 0.608,
      "grad_norm": 1.3411148998390934,
      "learning_rate": 4.001274945552098e-06,
      "loss": 0.2294,
      "step": 1900
    },
    {
      "epoch": 0.6112,
      "grad_norm": 1.2503467379311224,
      "learning_rate": 3.94660365002137e-06,
      "loss": 0.2179,
      "step": 1910
    },
    {
      "epoch": 0.6144,
      "grad_norm": 0.8844502114106245,
      "learning_rate": 3.892063833494387e-06,
      "loss": 0.2293,
      "step": 1920
    },
    {
      "epoch": 0.6176,
      "grad_norm": 1.323943188387191,
      "learning_rate": 3.837662303324093e-06,
      "loss": 0.2395,
      "step": 1930
    },
    {
      "epoch": 0.6208,
      "grad_norm": 1.3150182927928105,
      "learning_rate": 3.783405849603302e-06,
      "loss": 0.2265,
      "step": 1940
    },
    {
      "epoch": 0.624,
      "grad_norm": 1.5251447223837293,
      "learning_rate": 3.729301244317208e-06,
      "loss": 0.2445,
      "step": 1950
    },
    {
      "epoch": 0.6272,
      "grad_norm": 1.2518632060745116,
      "learning_rate": 3.675355240498133e-06,
      "loss": 0.2229,
      "step": 1960
    },
    {
      "epoch": 0.6304,
      "grad_norm": 1.4410331274554835,
      "learning_rate": 3.6215745713826585e-06,
      "loss": 0.2269,
      "step": 1970
    },
    {
      "epoch": 0.6336,
      "grad_norm": 1.2681938269526776,
      "learning_rate": 3.567965949571219e-06,
      "loss": 0.2216,
      "step": 1980
    },
    {
      "epoch": 0.6368,
      "grad_norm": 1.2119772204720247,
      "learning_rate": 3.5145360661902717e-06,
      "loss": 0.2029,
      "step": 1990
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.2769167413279376,
      "learning_rate": 3.4612915900571493e-06,
      "loss": 0.2234,
      "step": 2000
    },
    {
      "epoch": 0.6432,
      "grad_norm": 1.1935986503511073,
      "learning_rate": 3.408239166847696e-06,
      "loss": 0.2313,
      "step": 2010
    },
    {
      "epoch": 0.6464,
      "grad_norm": 1.0781337815097756,
      "learning_rate": 3.355385418266792e-06,
      "loss": 0.2292,
      "step": 2020
    },
    {
      "epoch": 0.6496,
      "grad_norm": 1.2668633413118764,
      "learning_rate": 3.3027369412218623e-06,
      "loss": 0.2231,
      "step": 2030
    },
    {
      "epoch": 0.6528,
      "grad_norm": 1.2243691251251998,
      "learning_rate": 3.2503003069995057e-06,
      "loss": 0.2127,
      "step": 2040
    },
    {
      "epoch": 0.656,
      "grad_norm": 1.1678450454685523,
      "learning_rate": 3.198082060445281e-06,
      "loss": 0.2286,
      "step": 2050
    },
    {
      "epoch": 0.6592,
      "grad_norm": 1.3654151111307413,
      "learning_rate": 3.1460887191468324e-06,
      "loss": 0.2262,
      "step": 2060
    },
    {
      "epoch": 0.6624,
      "grad_norm": 1.1453122239206175,
      "learning_rate": 3.0943267726203965e-06,
      "loss": 0.2251,
      "step": 2070
    },
    {
      "epoch": 0.6656,
      "grad_norm": 1.4758563467786459,
      "learning_rate": 3.042802681500814e-06,
      "loss": 0.2293,
      "step": 2080
    },
    {
      "epoch": 0.6688,
      "grad_norm": 1.1793224562603977,
      "learning_rate": 2.991522876735154e-06,
      "loss": 0.2205,
      "step": 2090
    },
    {
      "epoch": 0.672,
      "grad_norm": 1.2381651466593462,
      "learning_rate": 2.9404937587800374e-06,
      "loss": 0.2189,
      "step": 2100
    },
    {
      "epoch": 0.6752,
      "grad_norm": 19.66766001236314,
      "learning_rate": 2.889721696802768e-06,
      "loss": 0.2491,
      "step": 2110
    },
    {
      "epoch": 0.6784,
      "grad_norm": 1.0721784074518568,
      "learning_rate": 2.839213027886373e-06,
      "loss": 0.2266,
      "step": 2120
    },
    {
      "epoch": 0.6816,
      "grad_norm": 1.0952730571552138,
      "learning_rate": 2.7889740562386357e-06,
      "loss": 0.2156,
      "step": 2130
    },
    {
      "epoch": 0.6848,
      "grad_norm": 1.0572399244670028,
      "learning_rate": 2.7390110524052415e-06,
      "loss": 0.2355,
      "step": 2140
    },
    {
      "epoch": 0.688,
      "grad_norm": 1.0010243056654815,
      "learning_rate": 2.6893302524871357e-06,
      "loss": 0.2339,
      "step": 2150
    },
    {
      "epoch": 0.6912,
      "grad_norm": 1.1936155230445462,
      "learning_rate": 2.6399378573621493e-06,
      "loss": 0.2057,
      "step": 2160
    },
    {
      "epoch": 0.6944,
      "grad_norm": 1.4177935849928514,
      "learning_rate": 2.5908400319110588e-06,
      "loss": 0.2845,
      "step": 2170
    },
    {
      "epoch": 0.6976,
      "grad_norm": 1.3791416946344361,
      "learning_rate": 2.5420429042481054e-06,
      "loss": 0.287,
      "step": 2180
    },
    {
      "epoch": 0.7008,
      "grad_norm": 1.4748810133302899,
      "learning_rate": 2.493552564956126e-06,
      "loss": 0.2317,
      "step": 2190
    },
    {
      "epoch": 0.704,
      "grad_norm": 1.1836491485069627,
      "learning_rate": 2.445375066326362e-06,
      "loss": 0.2172,
      "step": 2200
    },
    {
      "epoch": 0.7072,
      "grad_norm": 1.2813045848755424,
      "learning_rate": 2.3975164216030456e-06,
      "loss": 0.2232,
      "step": 2210
    },
    {
      "epoch": 0.7104,
      "grad_norm": 1.3145112714821583,
      "learning_rate": 2.349982604232851e-06,
      "loss": 0.2343,
      "step": 2220
    },
    {
      "epoch": 0.7136,
      "grad_norm": 1.1808875801974772,
      "learning_rate": 2.3027795471193404e-06,
      "loss": 0.2465,
      "step": 2230
    },
    {
      "epoch": 0.7168,
      "grad_norm": 1.0981462474429815,
      "learning_rate": 2.255913141882436e-06,
      "loss": 0.3201,
      "step": 2240
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.230734760390504,
      "learning_rate": 2.209389238123066e-06,
      "loss": 0.2201,
      "step": 2250
    },
    {
      "epoch": 0.7232,
      "grad_norm": 1.2070256506474353,
      "learning_rate": 2.163213642693059e-06,
      "loss": 0.2218,
      "step": 2260
    },
    {
      "epoch": 0.7264,
      "grad_norm": 1.2704909154384314,
      "learning_rate": 2.1173921189703523e-06,
      "loss": 0.3296,
      "step": 2270
    },
    {
      "epoch": 0.7296,
      "grad_norm": 1.2731910215448914,
      "learning_rate": 2.0719303861396435e-06,
      "loss": 0.2509,
      "step": 2280
    },
    {
      "epoch": 0.7328,
      "grad_norm": 1.2952080019044347,
      "learning_rate": 2.0268341184785674e-06,
      "loss": 0.2296,
      "step": 2290
    },
    {
      "epoch": 0.736,
      "grad_norm": 1.1752545729106638,
      "learning_rate": 1.982108944649441e-06,
      "loss": 0.2502,
      "step": 2300
    },
    {
      "epoch": 0.7392,
      "grad_norm": 1.3810746184648486,
      "learning_rate": 1.937760446996741e-06,
      "loss": 0.2234,
      "step": 2310
    },
    {
      "epoch": 0.7424,
      "grad_norm": 1.24815248782901,
      "learning_rate": 1.8937941608503484e-06,
      "loss": 0.3464,
      "step": 2320
    },
    {
      "epoch": 0.7456,
      "grad_norm": 1.1328500128827523,
      "learning_rate": 1.8502155738346488e-06,
      "loss": 0.2319,
      "step": 2330
    },
    {
      "epoch": 0.7488,
      "grad_norm": 1.1922344333046615,
      "learning_rate": 1.8070301251836108e-06,
      "loss": 0.2231,
      "step": 2340
    },
    {
      "epoch": 0.752,
      "grad_norm": 1.2519090435571332,
      "learning_rate": 1.764243205061879e-06,
      "loss": 0.2213,
      "step": 2350
    },
    {
      "epoch": 0.7552,
      "grad_norm": 1.2296446772094094,
      "learning_rate": 1.721860153892011e-06,
      "loss": 0.2334,
      "step": 2360
    },
    {
      "epoch": 0.7584,
      "grad_norm": 1.1405936822992957,
      "learning_rate": 1.6798862616879185e-06,
      "loss": 0.2139,
      "step": 2370
    },
    {
      "epoch": 0.7616,
      "grad_norm": 1.3260125376610297,
      "learning_rate": 1.6383267673945925e-06,
      "loss": 0.222,
      "step": 2380
    },
    {
      "epoch": 0.7648,
      "grad_norm": 1.1316011229474554,
      "learning_rate": 1.5971868582342047e-06,
      "loss": 0.2026,
      "step": 2390
    },
    {
      "epoch": 0.768,
      "grad_norm": 1.1138427441279388,
      "learning_rate": 1.55647166905868e-06,
      "loss": 0.2136,
      "step": 2400
    },
    {
      "epoch": 0.7712,
      "grad_norm": 1.3571505881357013,
      "learning_rate": 1.516186281708778e-06,
      "loss": 0.2216,
      "step": 2410
    },
    {
      "epoch": 0.7744,
      "grad_norm": 1.2375006700475994,
      "learning_rate": 1.4763357243798154e-06,
      "loss": 0.2139,
      "step": 2420
    },
    {
      "epoch": 0.7776,
      "grad_norm": 1.1212490843977576,
      "learning_rate": 1.4369249709940759e-06,
      "loss": 0.2067,
      "step": 2430
    },
    {
      "epoch": 0.7808,
      "grad_norm": 1.3356196332358288,
      "learning_rate": 1.3979589405799865e-06,
      "loss": 0.207,
      "step": 2440
    },
    {
      "epoch": 0.784,
      "grad_norm": 1.28648624602483,
      "learning_rate": 1.3594424966581555e-06,
      "loss": 0.2186,
      "step": 2450
    },
    {
      "epoch": 0.7872,
      "grad_norm": 1.1736305743675295,
      "learning_rate": 1.321380446634342e-06,
      "loss": 0.2116,
      "step": 2460
    },
    {
      "epoch": 0.7904,
      "grad_norm": 1.2598458372212618,
      "learning_rate": 1.2837775411994092e-06,
      "loss": 0.2228,
      "step": 2470
    },
    {
      "epoch": 0.7936,
      "grad_norm": 1.0979929469461585,
      "learning_rate": 1.246638473736378e-06,
      "loss": 0.2067,
      "step": 2480
    },
    {
      "epoch": 0.7968,
      "grad_norm": 1.6135875508628044,
      "learning_rate": 1.2099678797346282e-06,
      "loss": 0.3298,
      "step": 2490
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.230081395796401,
      "learning_rate": 1.1737703362113134e-06,
      "loss": 0.2149,
      "step": 2500
    },
    {
      "epoch": 0.8032,
      "grad_norm": 1.191155602222997,
      "learning_rate": 1.1380503611400933e-06,
      "loss": 0.2114,
      "step": 2510
    },
    {
      "epoch": 0.8064,
      "grad_norm": 1.0941583722626504,
      "learning_rate": 1.1028124128872191e-06,
      "loss": 0.2197,
      "step": 2520
    },
    {
      "epoch": 0.8096,
      "grad_norm": 1.3743010322287874,
      "learning_rate": 1.068060889655066e-06,
      "loss": 0.2201,
      "step": 2530
    },
    {
      "epoch": 0.8128,
      "grad_norm": 1.367609495826361,
      "learning_rate": 1.033800128933179e-06,
      "loss": 0.2263,
      "step": 2540
    },
    {
      "epoch": 0.816,
      "grad_norm": 1.2841195283902214,
      "learning_rate": 1.0000344069568885e-06,
      "loss": 0.2108,
      "step": 2550
    },
    {
      "epoch": 0.8192,
      "grad_norm": 1.262506622679509,
      "learning_rate": 9.667679381735706e-07,
      "loss": 0.2099,
      "step": 2560
    },
    {
      "epoch": 0.8224,
      "grad_norm": 1.3708525231737212,
      "learning_rate": 9.340048747166341e-07,
      "loss": 0.2123,
      "step": 2570
    },
    {
      "epoch": 0.8256,
      "grad_norm": 1.0907835378818949,
      "learning_rate": 9.017493058872623e-07,
      "loss": 0.2124,
      "step": 2580
    },
    {
      "epoch": 0.8288,
      "grad_norm": 1.1324488490089442,
      "learning_rate": 8.700052576440166e-07,
      "loss": 0.2152,
      "step": 2590
    },
    {
      "epoch": 0.832,
      "grad_norm": 1.1888142582084373,
      "learning_rate": 8.387766921003427e-07,
      "loss": 0.2218,
      "step": 2600
    },
    {
      "epoch": 0.8352,
      "grad_norm": 1.230386423249033,
      "learning_rate": 8.080675070300303e-07,
      "loss": 0.4094,
      "step": 2610
    },
    {
      "epoch": 0.8384,
      "grad_norm": 1.2059240836852692,
      "learning_rate": 7.77881535380724e-07,
      "loss": 0.2182,
      "step": 2620
    },
    {
      "epoch": 0.8416,
      "grad_norm": 1.2730243315669996,
      "learning_rate": 7.482225447955155e-07,
      "loss": 0.2242,
      "step": 2630
    },
    {
      "epoch": 0.8448,
      "grad_norm": 1.3097925731697693,
      "learning_rate": 7.190942371426862e-07,
      "loss": 0.2334,
      "step": 2640
    },
    {
      "epoch": 0.848,
      "grad_norm": 1.3089263483412454,
      "learning_rate": 6.905002480536565e-07,
      "loss": 0.2129,
      "step": 2650
    },
    {
      "epoch": 0.8512,
      "grad_norm": 1.2487331361219,
      "learning_rate": 6.624441464692161e-07,
      "loss": 0.2059,
      "step": 2660
    },
    {
      "epoch": 0.8544,
      "grad_norm": 1.1900169075542317,
      "learning_rate": 6.349294341940593e-07,
      "loss": 0.2119,
      "step": 2670
    },
    {
      "epoch": 0.8576,
      "grad_norm": 1.4499451387635829,
      "learning_rate": 6.07959545459717e-07,
      "loss": 0.2881,
      "step": 2680
    },
    {
      "epoch": 0.8608,
      "grad_norm": 1.2365542029204284,
      "learning_rate": 5.815378464959109e-07,
      "loss": 0.21,
      "step": 2690
    },
    {
      "epoch": 0.864,
      "grad_norm": 1.169403617807277,
      "learning_rate": 5.55667635110399e-07,
      "loss": 0.2131,
      "step": 2700
    },
    {
      "epoch": 0.8672,
      "grad_norm": 1.1917798187061128,
      "learning_rate": 5.303521402773665e-07,
      "loss": 0.2171,
      "step": 2710
    },
    {
      "epoch": 0.8704,
      "grad_norm": 1.3227237208234792,
      "learning_rate": 5.055945217344004e-07,
      "loss": 0.2232,
      "step": 2720
    },
    {
      "epoch": 0.8736,
      "grad_norm": 1.244493433459287,
      "learning_rate": 4.81397869588106e-07,
      "loss": 0.3424,
      "step": 2730
    },
    {
      "epoch": 0.8768,
      "grad_norm": 1.2918840219931813,
      "learning_rate": 4.5776520392842473e-07,
      "loss": 0.2041,
      "step": 2740
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.161291630108919,
      "learning_rate": 4.346994744516747e-07,
      "loss": 0.2062,
      "step": 2750
    },
    {
      "epoch": 0.8832,
      "grad_norm": 1.188169889946088,
      "learning_rate": 4.122035600923913e-07,
      "loss": 0.2138,
      "step": 2760
    },
    {
      "epoch": 0.8864,
      "grad_norm": 1.2759791115877892,
      "learning_rate": 3.902802686639967e-07,
      "loss": 0.2076,
      "step": 2770
    },
    {
      "epoch": 0.8896,
      "grad_norm": 1.2428248160576683,
      "learning_rate": 3.6893233650833916e-07,
      "loss": 0.2231,
      "step": 2780
    },
    {
      "epoch": 0.8928,
      "grad_norm": 1.2872717018149586,
      "learning_rate": 3.4816242815416014e-07,
      "loss": 0.2167,
      "step": 2790
    },
    {
      "epoch": 0.896,
      "grad_norm": 1.120924586598326,
      "learning_rate": 3.2797313598452506e-07,
      "loss": 0.2125,
      "step": 2800
    },
    {
      "epoch": 0.8992,
      "grad_norm": 1.235367729092687,
      "learning_rate": 3.0836697991325547e-07,
      "loss": 0.2211,
      "step": 2810
    },
    {
      "epoch": 0.9024,
      "grad_norm": 1.1425245188644768,
      "learning_rate": 2.893464070704055e-07,
      "loss": 0.2059,
      "step": 2820
    },
    {
      "epoch": 0.9056,
      "grad_norm": 1.240438287494313,
      "learning_rate": 2.7091379149682683e-07,
      "loss": 0.2038,
      "step": 2830
    },
    {
      "epoch": 0.9088,
      "grad_norm": 1.2119903261308935,
      "learning_rate": 2.53071433847854e-07,
      "loss": 0.208,
      "step": 2840
    },
    {
      "epoch": 0.912,
      "grad_norm": 1.3047012967784055,
      "learning_rate": 2.3582156110614985e-07,
      "loss": 0.2065,
      "step": 2850
    },
    {
      "epoch": 0.9152,
      "grad_norm": 1.2159745001331477,
      "learning_rate": 2.1916632630374579e-07,
      "loss": 0.2092,
      "step": 2860
    },
    {
      "epoch": 0.9184,
      "grad_norm": 1.4115746282687291,
      "learning_rate": 2.0310780825331056e-07,
      "loss": 0.3065,
      "step": 2870
    },
    {
      "epoch": 0.9216,
      "grad_norm": 1.1260909387649622,
      "learning_rate": 1.876480112886886e-07,
      "loss": 0.2149,
      "step": 2880
    },
    {
      "epoch": 0.9248,
      "grad_norm": 1.321400120727542,
      "learning_rate": 1.7278886501472804e-07,
      "loss": 0.2151,
      "step": 2890
    },
    {
      "epoch": 0.928,
      "grad_norm": 1.2811002757998282,
      "learning_rate": 1.5853222406643555e-07,
      "loss": 0.2124,
      "step": 2900
    },
    {
      "epoch": 0.9312,
      "grad_norm": 1.3188098070211447,
      "learning_rate": 1.4487986787749763e-07,
      "loss": 0.2103,
      "step": 2910
    },
    {
      "epoch": 0.9344,
      "grad_norm": 1.1879475265636688,
      "learning_rate": 1.318335004581761e-07,
      "loss": 0.2114,
      "step": 2920
    },
    {
      "epoch": 0.9376,
      "grad_norm": 1.119348294167629,
      "learning_rate": 1.1939475018262481e-07,
      "loss": 0.2144,
      "step": 2930
    },
    {
      "epoch": 0.9408,
      "grad_norm": 1.160073841048408,
      "learning_rate": 1.0756516958564667e-07,
      "loss": 0.2119,
      "step": 2940
    },
    {
      "epoch": 0.944,
      "grad_norm": 1.3050171867050437,
      "learning_rate": 9.634623516891372e-08,
      "loss": 0.2077,
      "step": 2950
    },
    {
      "epoch": 0.9472,
      "grad_norm": 1.102635547442788,
      "learning_rate": 8.573934721667731e-08,
      "loss": 0.2107,
      "step": 2960
    },
    {
      "epoch": 0.9504,
      "grad_norm": 1.1976820144956357,
      "learning_rate": 7.574582962099508e-08,
      "loss": 0.3234,
      "step": 2970
    },
    {
      "epoch": 0.9536,
      "grad_norm": 1.0289812852531994,
      "learning_rate": 6.636692971648873e-08,
      "loss": 0.2946,
      "step": 2980
    },
    {
      "epoch": 0.9568,
      "grad_norm": 1.1192107741554636,
      "learning_rate": 5.7603818124657984e-08,
      "loss": 0.2146,
      "step": 2990
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.319177806968747,
      "learning_rate": 4.9457588607772497e-08,
      "loss": 0.2113,
      "step": 3000
    },
    {
      "epoch": 0.9632,
      "grad_norm": 1.0857028410476908,
      "learning_rate": 4.192925793235159e-08,
      "loss": 0.2097,
      "step": 3010
    },
    {
      "epoch": 0.9664,
      "grad_norm": 1.0588162046549727,
      "learning_rate": 3.501976574226018e-08,
      "loss": 0.2138,
      "step": 3020
    },
    {
      "epoch": 0.9696,
      "grad_norm": 1.3776741815700904,
      "learning_rate": 2.8729974441426557e-08,
      "loss": 0.1994,
      "step": 3030
    },
    {
      "epoch": 0.9728,
      "grad_norm": 1.403580554324849,
      "learning_rate": 2.3060669086199526e-08,
      "loss": 0.2166,
      "step": 3040
    },
    {
      "epoch": 0.976,
      "grad_norm": 1.2568358655842473,
      "learning_rate": 1.8012557287367394e-08,
      "loss": 0.2843,
      "step": 3050
    },
    {
      "epoch": 0.9792,
      "grad_norm": 1.3656578926442577,
      "learning_rate": 1.3586269121833028e-08,
      "loss": 0.2138,
      "step": 3060
    },
    {
      "epoch": 0.9824,
      "grad_norm": 1.178254860310462,
      "learning_rate": 9.782357053972902e-09,
      "loss": 0.2169,
      "step": 3070
    },
    {
      "epoch": 0.9856,
      "grad_norm": 1.2773015970417392,
      "learning_rate": 6.6012958666827886e-09,
      "loss": 0.2163,
      "step": 3080
    },
    {
      "epoch": 0.9888,
      "grad_norm": 1.2949804935819031,
      "learning_rate": 4.043482602116844e-09,
      "loss": 0.2152,
      "step": 3090
    },
    {
      "epoch": 0.992,
      "grad_norm": 1.129218160141717,
      "learning_rate": 2.1092365121305745e-09,
      "loss": 0.2566,
      "step": 3100
    },
    {
      "epoch": 0.9952,
      "grad_norm": 1.3984530229483965,
      "learning_rate": 7.987990184354921e-10,
      "loss": 0.3123,
      "step": 3110
    },
    {
      "epoch": 0.9984,
      "grad_norm": 1.3456687728542012,
      "learning_rate": 1.1233368246321708e-10,
      "loss": 0.2285,
      "step": 3120
    }
  ],
  "logging_steps": 10,
  "max_steps": 3125,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 68659404668928.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}