{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 88,
  "global_step": 700,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0,
      "eval_loss": 2.872528314590454,
      "eval_runtime": 166.5845,
      "eval_samples_per_second": 1.909,
      "eval_steps_per_second": 0.318,
      "memory/device_mem_reserved(gib)": 42.76,
      "memory/max_mem_active(gib)": 42.42,
      "memory/max_mem_allocated(gib)": 40.92,
      "step": 0
    },
    {
      "epoch": 0.002857142857142857,
      "grad_norm": 6.58348657731275,
      "learning_rate": 0.0,
      "loss": 2.7506,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 45.24,
      "memory/max_mem_allocated(gib)": 45.11,
      "step": 1
    },
    {
      "epoch": 0.005714285714285714,
      "grad_norm": 6.813951873498306,
      "learning_rate": 1.1428571428571427e-08,
      "loss": 2.7698,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.43,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 2
    },
    {
      "epoch": 0.008571428571428572,
      "grad_norm": 6.666348098712079,
      "learning_rate": 2.2857142857142854e-08,
      "loss": 2.7508,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.59,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 3
    },
    {
      "epoch": 0.011428571428571429,
      "grad_norm": 6.478993209664458,
      "learning_rate": 3.4285714285714286e-08,
      "loss": 2.8363,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.59,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 4
    },
    {
      "epoch": 0.014285714285714285,
      "grad_norm": 6.303993913620244,
      "learning_rate": 4.571428571428571e-08,
      "loss": 2.7726,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.59,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 5
    },
    {
      "epoch": 0.017142857142857144,
      "grad_norm": 6.3921203080148095,
      "learning_rate": 5.714285714285714e-08,
      "loss": 2.7048,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.59,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 6
    },
    {
      "epoch": 0.02,
      "grad_norm": 6.602738486446332,
      "learning_rate": 6.857142857142857e-08,
      "loss": 2.7637,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.59,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 7
    },
    {
      "epoch": 0.022857142857142857,
      "grad_norm": 6.719550313122406,
      "learning_rate": 8e-08,
      "loss": 2.7754,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.59,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 8
    },
    {
      "epoch": 0.025714285714285714,
      "grad_norm": 6.4971954494628115,
      "learning_rate": 9.142857142857142e-08,
      "loss": 2.7538,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.59,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 9
    },
    {
      "epoch": 0.02857142857142857,
      "grad_norm": 6.726473136739958,
      "learning_rate": 1.0285714285714285e-07,
      "loss": 2.8036,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.59,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 10
    },
    {
      "epoch": 0.03142857142857143,
      "grad_norm": 6.568188044376355,
      "learning_rate": 1.1428571428571427e-07,
      "loss": 2.7332,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.59,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 11
    },
    {
      "epoch": 0.03428571428571429,
      "grad_norm": 6.405225352287575,
      "learning_rate": 1.2571428571428572e-07,
      "loss": 2.724,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.59,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 12
    },
    {
      "epoch": 0.037142857142857144,
      "grad_norm": 6.424971153625209,
      "learning_rate": 1.3714285714285715e-07,
      "loss": 2.7921,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.59,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 13
    },
    {
      "epoch": 0.04,
      "grad_norm": 6.425588279476742,
      "learning_rate": 1.4857142857142857e-07,
      "loss": 2.7596,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.59,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 14
    },
    {
      "epoch": 0.04285714285714286,
      "grad_norm": 6.400458418721808,
      "learning_rate": 1.6e-07,
      "loss": 2.7168,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.59,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 15
    },
    {
      "epoch": 0.045714285714285714,
      "grad_norm": 6.6612260694408105,
      "learning_rate": 1.714285714285714e-07,
      "loss": 2.7651,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 16
    },
    {
      "epoch": 0.04857142857142857,
      "grad_norm": 6.73638332890746,
      "learning_rate": 1.8285714285714283e-07,
      "loss": 2.8347,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 17
    },
    {
      "epoch": 0.05142857142857143,
      "grad_norm": 6.559839480834488,
      "learning_rate": 1.9428571428571426e-07,
      "loss": 2.7817,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 18
    },
    {
      "epoch": 0.054285714285714284,
      "grad_norm": 6.567012577867464,
      "learning_rate": 2.057142857142857e-07,
      "loss": 2.7938,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 19
    },
    {
      "epoch": 0.05714285714285714,
      "grad_norm": 6.518612302699156,
      "learning_rate": 2.1714285714285712e-07,
      "loss": 2.7971,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 20
    },
    {
      "epoch": 0.06,
      "grad_norm": 6.343130744558327,
      "learning_rate": 2.2857142857142855e-07,
      "loss": 2.755,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 21
    },
    {
      "epoch": 0.06285714285714286,
      "grad_norm": 6.448256185177304,
      "learning_rate": 2.4e-07,
      "loss": 2.7483,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 22
    },
    {
      "epoch": 0.06571428571428571,
      "grad_norm": 6.425682215291841,
      "learning_rate": 2.5142857142857143e-07,
      "loss": 2.7504,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 23
    },
    {
      "epoch": 0.06857142857142857,
      "grad_norm": 6.318707120246484,
      "learning_rate": 2.6285714285714284e-07,
      "loss": 2.823,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 24
    },
    {
      "epoch": 0.07142857142857142,
      "grad_norm": 6.400793804988853,
      "learning_rate": 2.742857142857143e-07,
      "loss": 2.7952,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 25
    },
    {
      "epoch": 0.07428571428571429,
      "grad_norm": 6.345763318012626,
      "learning_rate": 2.857142857142857e-07,
      "loss": 2.7693,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 26
    },
    {
      "epoch": 0.07714285714285714,
      "grad_norm": 6.220793403292321,
      "learning_rate": 2.9714285714285715e-07,
      "loss": 2.7459,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 27
    },
    {
      "epoch": 0.08,
      "grad_norm": 6.030925598055583,
      "learning_rate": 3.0857142857142855e-07,
      "loss": 2.7339,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 28
    },
    {
      "epoch": 0.08285714285714285,
      "grad_norm": 5.760615240130594,
      "learning_rate": 3.2e-07,
      "loss": 2.7072,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 29
    },
    {
      "epoch": 0.08571428571428572,
      "grad_norm": 5.66496205841438,
      "learning_rate": 3.314285714285714e-07,
      "loss": 2.7539,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 30
    },
    {
      "epoch": 0.08857142857142856,
      "grad_norm": 5.8084456388036285,
      "learning_rate": 3.428571428571428e-07,
      "loss": 2.7033,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 31
    },
    {
      "epoch": 0.09142857142857143,
      "grad_norm": 5.958937944903284,
      "learning_rate": 3.5428571428571427e-07,
      "loss": 2.7354,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 32
    },
    {
      "epoch": 0.09428571428571429,
      "grad_norm": 5.764877868755503,
      "learning_rate": 3.6571428571428567e-07,
      "loss": 2.7711,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 33
    },
    {
      "epoch": 0.09714285714285714,
      "grad_norm": 5.711557024303864,
      "learning_rate": 3.771428571428571e-07,
      "loss": 2.7797,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 34
    },
    {
      "epoch": 0.1,
      "grad_norm": 5.601709705929177,
      "learning_rate": 3.8857142857142853e-07,
      "loss": 2.7564,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 35
    },
    {
      "epoch": 0.10285714285714286,
      "grad_norm": 5.518114977001241,
      "learning_rate": 4e-07,
      "loss": 2.7957,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 36
    },
    {
      "epoch": 0.10571428571428572,
      "grad_norm": 5.642264875436979,
      "learning_rate": 4.114285714285714e-07,
      "loss": 2.7587,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 37
    },
    {
      "epoch": 0.10857142857142857,
      "grad_norm": 5.212441194691044,
      "learning_rate": 4.2285714285714284e-07,
      "loss": 2.742,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 38
    },
    {
      "epoch": 0.11142857142857143,
      "grad_norm": 4.666748450442973,
      "learning_rate": 4.3428571428571424e-07,
      "loss": 2.6992,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 39
    },
    {
      "epoch": 0.11428571428571428,
      "grad_norm": 4.54434857149615,
      "learning_rate": 4.457142857142857e-07,
      "loss": 2.7528,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 40
    },
    {
      "epoch": 0.11714285714285715,
      "grad_norm": 4.358923272400012,
      "learning_rate": 4.571428571428571e-07,
      "loss": 2.6074,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 41
    },
    {
      "epoch": 0.12,
      "grad_norm": 4.3843081603995016,
      "learning_rate": 4.6857142857142855e-07,
      "loss": 2.6699,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 42
    },
    {
      "epoch": 0.12285714285714286,
      "grad_norm": 4.448911106079794,
      "learning_rate": 4.8e-07,
      "loss": 2.777,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 43
    },
    {
      "epoch": 0.12571428571428572,
      "grad_norm": 4.403856293157942,
      "learning_rate": 4.914285714285714e-07,
      "loss": 2.6892,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 44
    },
    {
      "epoch": 0.12857142857142856,
      "grad_norm": 4.195353461893667,
      "learning_rate": 5.028571428571429e-07,
      "loss": 2.6949,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 45
    },
    {
      "epoch": 0.13142857142857142,
      "grad_norm": 4.174553702515705,
      "learning_rate": 5.142857142857143e-07,
      "loss": 2.7634,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 46
    },
    {
      "epoch": 0.13428571428571429,
      "grad_norm": 3.95408142590073,
      "learning_rate": 5.257142857142857e-07,
      "loss": 2.6915,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 47
    },
    {
      "epoch": 0.13714285714285715,
      "grad_norm": 4.019386952161042,
      "learning_rate": 5.371428571428571e-07,
      "loss": 2.7148,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 48
    },
    {
      "epoch": 0.14,
      "grad_norm": 3.825252112024877,
      "learning_rate": 5.485714285714286e-07,
      "loss": 2.7291,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 49
    },
    {
      "epoch": 0.14285714285714285,
      "grad_norm": 3.733904265878752,
      "learning_rate": 5.599999999999999e-07,
      "loss": 2.5917,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 50
    },
    {
      "epoch": 0.1457142857142857,
      "grad_norm": 3.757282847646802,
      "learning_rate": 5.714285714285714e-07,
      "loss": 2.719,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 51
    },
    {
      "epoch": 0.14857142857142858,
      "grad_norm": 3.6009044552554244,
      "learning_rate": 5.828571428571428e-07,
      "loss": 2.6519,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 52
    },
    {
      "epoch": 0.15142857142857144,
      "grad_norm": 3.174032663481999,
      "learning_rate": 5.942857142857143e-07,
      "loss": 2.6718,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 53
    },
    {
      "epoch": 0.15428571428571428,
      "grad_norm": 2.7458739161032963,
      "learning_rate": 6.057142857142856e-07,
      "loss": 2.6289,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 54
    },
    {
      "epoch": 0.15714285714285714,
      "grad_norm": 2.2417357930798345,
      "learning_rate": 6.171428571428571e-07,
      "loss": 2.6509,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 55
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.9967768113859634,
      "learning_rate": 6.285714285714286e-07,
      "loss": 2.7027,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 56
    },
    {
      "epoch": 0.16285714285714287,
      "grad_norm": 1.7818062253264118,
      "learning_rate": 6.4e-07,
      "loss": 2.5927,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 57
    },
    {
      "epoch": 0.1657142857142857,
      "grad_norm": 1.8101165146863603,
      "learning_rate": 6.514285714285714e-07,
      "loss": 2.6014,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 58
    },
    {
      "epoch": 0.16857142857142857,
      "grad_norm": 1.7199494503180122,
      "learning_rate": 6.628571428571428e-07,
      "loss": 2.6479,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 59
    },
    {
      "epoch": 0.17142857142857143,
      "grad_norm": 1.7027786074487685,
      "learning_rate": 6.742857142857143e-07,
      "loss": 2.5656,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 60
    },
    {
      "epoch": 0.1742857142857143,
      "grad_norm": 1.6041478784950989,
      "learning_rate": 6.857142857142856e-07,
      "loss": 2.623,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 61
    },
    {
      "epoch": 0.17714285714285713,
      "grad_norm": 1.5179428021816292,
      "learning_rate": 6.971428571428571e-07,
      "loss": 2.6156,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 62
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.5541460903370177,
      "learning_rate": 7.085714285714285e-07,
      "loss": 2.664,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 63
    },
    {
      "epoch": 0.18285714285714286,
      "grad_norm": 1.537144027084823,
      "learning_rate": 7.2e-07,
      "loss": 2.5862,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 64
    },
    {
      "epoch": 0.18571428571428572,
      "grad_norm": 1.4523439240374998,
      "learning_rate": 7.314285714285713e-07,
      "loss": 2.5683,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 65
    },
    {
      "epoch": 0.18857142857142858,
      "grad_norm": 1.4185513332988053,
      "learning_rate": 7.428571428571429e-07,
      "loss": 2.6369,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 66
    },
    {
      "epoch": 0.19142857142857142,
      "grad_norm": 1.3361429266518214,
      "learning_rate": 7.542857142857142e-07,
      "loss": 2.6076,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 67
    },
    {
      "epoch": 0.19428571428571428,
      "grad_norm": 1.29197647623745,
      "learning_rate": 7.657142857142857e-07,
      "loss": 2.5964,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 68
    },
    {
      "epoch": 0.19714285714285715,
      "grad_norm": 1.2497598589474845,
      "learning_rate": 7.771428571428571e-07,
      "loss": 2.5873,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 69
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.1570834784726216,
      "learning_rate": 7.885714285714286e-07,
      "loss": 2.6002,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 70
    },
    {
      "epoch": 0.20285714285714285,
      "grad_norm": 1.1016979784273444,
      "learning_rate": 8e-07,
      "loss": 2.5841,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 71
    },
    {
      "epoch": 0.2057142857142857,
      "grad_norm": 0.950663056332938,
      "learning_rate": 7.999950266646766e-07,
      "loss": 2.5934,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 72
    },
    {
      "epoch": 0.20857142857142857,
      "grad_norm": 0.8909462148253868,
      "learning_rate": 7.999801067823771e-07,
      "loss": 2.594,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 73
    },
    {
      "epoch": 0.21142857142857144,
      "grad_norm": 0.8886356233732262,
      "learning_rate": 7.999552407241093e-07,
      "loss": 2.5218,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 74
    },
    {
      "epoch": 0.21428571428571427,
      "grad_norm": 0.8394903734907394,
      "learning_rate": 7.999204291082094e-07,
      "loss": 2.5083,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 75
    },
    {
      "epoch": 0.21714285714285714,
      "grad_norm": 0.8094558233882342,
      "learning_rate": 7.998756728003265e-07,
      "loss": 2.4894,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 76
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.7910385839315002,
      "learning_rate": 7.998209729134013e-07,
      "loss": 2.5891,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 77
    },
    {
      "epoch": 0.22285714285714286,
      "grad_norm": 0.7737634514349553,
      "learning_rate": 7.997563308076383e-07,
      "loss": 2.5191,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 78
    },
    {
      "epoch": 0.2257142857142857,
      "grad_norm": 0.7516151950536224,
      "learning_rate": 7.996817480904717e-07,
      "loss": 2.5655,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 79
    },
    {
      "epoch": 0.22857142857142856,
      "grad_norm": 0.7059846983575885,
      "learning_rate": 7.995972266165259e-07,
      "loss": 2.5462,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 80
    },
    {
      "epoch": 0.23142857142857143,
      "grad_norm": 0.768223010395181,
      "learning_rate": 7.995027684875689e-07,
      "loss": 2.6042,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 81
    },
    {
      "epoch": 0.2342857142857143,
      "grad_norm": 0.7148069881300689,
      "learning_rate": 7.993983760524606e-07,
      "loss": 2.5948,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 82
    },
    {
      "epoch": 0.23714285714285716,
      "grad_norm": 0.7092585586288884,
      "learning_rate": 7.99284051907094e-07,
      "loss": 2.5794,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 83
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.6676231552245798,
      "learning_rate": 7.991597988943305e-07,
      "loss": 2.5542,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 84
    },
    {
      "epoch": 0.24285714285714285,
      "grad_norm": 0.5959637634275424,
      "learning_rate": 7.990256201039296e-07,
      "loss": 2.4888,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 85
    },
    {
      "epoch": 0.24571428571428572,
      "grad_norm": 0.6141665164176573,
      "learning_rate": 7.98881518872472e-07,
      "loss": 2.5658,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 86
    },
    {
      "epoch": 0.24857142857142858,
      "grad_norm": 0.6275332188572059,
      "learning_rate": 7.987274987832763e-07,
      "loss": 2.5957,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 87
    },
    {
      "epoch": 0.25142857142857145,
      "grad_norm": 0.6267994007916001,
      "learning_rate": 7.985635636663103e-07,
      "loss": 2.532,
      "memory/device_mem_reserved(gib)": 69.56,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 88
    },
    {
      "epoch": 0.25142857142857145,
      "eval_loss": 2.6638412475585938,
      "eval_runtime": 166.4779,
      "eval_samples_per_second": 1.91,
      "eval_steps_per_second": 0.318,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 88
    },
    {
      "epoch": 0.2542857142857143,
      "grad_norm": 0.5627177962566883,
      "learning_rate": 7.983897175980956e-07,
      "loss": 2.5373,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 89
    },
    {
      "epoch": 0.2571428571428571,
      "grad_norm": 0.5738675014033745,
      "learning_rate": 7.98205964901606e-07,
      "loss": 2.5749,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 90
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.5875092171023301,
      "learning_rate": 7.980123101461606e-07,
      "loss": 2.553,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 91
    },
    {
      "epoch": 0.26285714285714284,
      "grad_norm": 0.575532656707433,
      "learning_rate": 7.978087581473093e-07,
      "loss": 2.4559,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 92
    },
    {
      "epoch": 0.26571428571428574,
      "grad_norm": 0.5614727897839459,
      "learning_rate": 7.97595313966714e-07,
      "loss": 2.5354,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 93
    },
    {
      "epoch": 0.26857142857142857,
      "grad_norm": 0.5350744027637109,
      "learning_rate": 7.973719829120222e-07,
      "loss": 2.5417,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 94
    },
    {
      "epoch": 0.2714285714285714,
      "grad_norm": 0.5673844795217607,
      "learning_rate": 7.971387705367349e-07,
      "loss": 2.505,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 95
    },
    {
      "epoch": 0.2742857142857143,
      "grad_norm": 0.5858985498367595,
      "learning_rate": 7.968956826400689e-07,
      "loss": 2.5808,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 96
    },
    {
      "epoch": 0.27714285714285714,
      "grad_norm": 0.5363260908481103,
      "learning_rate": 7.96642725266812e-07,
      "loss": 2.5521,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 97
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.4910312827977321,
      "learning_rate": 7.963799047071739e-07,
      "loss": 2.5247,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 98
    },
    {
      "epoch": 0.28285714285714286,
      "grad_norm": 0.5335036189849376,
      "learning_rate": 7.961072274966281e-07,
      "loss": 2.5422,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 99
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 0.5277251997314476,
      "learning_rate": 7.958247004157508e-07,
      "loss": 2.4883,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 100
    },
    {
      "epoch": 0.2885714285714286,
      "grad_norm": 0.5122194401989285,
      "learning_rate": 7.955323304900513e-07,
      "loss": 2.5195,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 101
    },
    {
      "epoch": 0.2914285714285714,
      "grad_norm": 0.5211540137102477,
      "learning_rate": 7.952301249897984e-07,
      "loss": 2.5336,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 102
    },
    {
      "epoch": 0.29428571428571426,
      "grad_norm": 0.5105209709604847,
      "learning_rate": 7.949180914298383e-07,
      "loss": 2.5249,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 103
    },
    {
      "epoch": 0.29714285714285715,
      "grad_norm": 0.4911237234034178,
      "learning_rate": 7.945962375694085e-07,
      "loss": 2.5195,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 104
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.46777080055237363,
      "learning_rate": 7.942645714119451e-07,
      "loss": 2.5202,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 105
    },
    {
      "epoch": 0.3028571428571429,
      "grad_norm": 0.47437550719038307,
      "learning_rate": 7.939231012048832e-07,
      "loss": 2.53,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 106
    },
    {
      "epoch": 0.3057142857142857,
      "grad_norm": 0.46101643620766136,
      "learning_rate": 7.935718354394518e-07,
      "loss": 2.5093,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 107
    },
    {
      "epoch": 0.30857142857142855,
      "grad_norm": 0.49214900544846946,
      "learning_rate": 7.932107828504632e-07,
      "loss": 2.4851,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 108
    },
    {
      "epoch": 0.31142857142857144,
      "grad_norm": 0.4819794369269858,
      "learning_rate": 7.928399524160954e-07,
      "loss": 2.5211,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 109
    },
    {
      "epoch": 0.3142857142857143,
      "grad_norm": 0.46587583985249315,
      "learning_rate": 7.92459353357669e-07,
      "loss": 2.5571,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 110
    },
    {
      "epoch": 0.3171428571428571,
      "grad_norm": 0.45340861566299884,
      "learning_rate": 7.920689951394175e-07,
      "loss": 2.4902,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 111
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.44962117010551356,
      "learning_rate": 7.916688874682526e-07,
      "loss": 2.5304,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 112
    },
    {
      "epoch": 0.32285714285714284,
      "grad_norm": 0.44619942524862266,
      "learning_rate": 7.912590402935222e-07,
      "loss": 2.5763,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 113
    },
    {
      "epoch": 0.32571428571428573,
      "grad_norm": 0.4807037475627005,
      "learning_rate": 7.908394638067637e-07,
      "loss": 2.4933,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 114
    },
    {
      "epoch": 0.32857142857142857,
      "grad_norm": 0.45846080823182794,
      "learning_rate": 7.904101684414497e-07,
      "loss": 2.5518,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 115
    },
    {
      "epoch": 0.3314285714285714,
      "grad_norm": 0.4383864367384291,
      "learning_rate": 7.899711648727293e-07,
      "loss": 2.5315,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 116
    },
    {
      "epoch": 0.3342857142857143,
      "grad_norm": 0.4289151901194594,
      "learning_rate": 7.895224640171625e-07,
      "loss": 2.5271,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 117
    },
    {
      "epoch": 0.33714285714285713,
      "grad_norm": 0.42803383731471234,
      "learning_rate": 7.89064077032448e-07,
      "loss": 2.6065,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 118
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.4269876103905192,
      "learning_rate": 7.885960153171469e-07,
      "loss": 2.5424,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 119
    },
    {
      "epoch": 0.34285714285714286,
      "grad_norm": 0.4368457798444619,
      "learning_rate": 7.881182905103985e-07,
      "loss": 2.4526,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 120
    },
    {
      "epoch": 0.3457142857142857,
      "grad_norm": 0.4630287318961494,
      "learning_rate": 7.876309144916312e-07,
      "loss": 2.4772,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 121
    },
    {
      "epoch": 0.3485714285714286,
      "grad_norm": 0.4423142229104516,
      "learning_rate": 7.871338993802666e-07,
      "loss": 2.4977,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 122
    },
    {
      "epoch": 0.3514285714285714,
      "grad_norm": 0.4352017550526162,
      "learning_rate": 7.866272575354188e-07,
      "loss": 2.5323,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 123
    },
    {
      "epoch": 0.35428571428571426,
      "grad_norm": 0.438946189149841,
      "learning_rate": 7.861110015555869e-07,
      "loss": 2.5601,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 124
    },
    {
      "epoch": 0.35714285714285715,
      "grad_norm": 0.44069886946626124,
      "learning_rate": 7.855851442783412e-07,
      "loss": 2.5248,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 125
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.42755455576746176,
      "learning_rate": 7.850496987800048e-07,
      "loss": 2.5078,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 126
    },
    {
      "epoch": 0.3628571428571429,
      "grad_norm": 0.597709861760904,
      "learning_rate": 7.845046783753275e-07,
      "loss": 2.5526,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 127
    },
    {
      "epoch": 0.3657142857142857,
      "grad_norm": 0.42078323653882893,
      "learning_rate": 7.839500966171556e-07,
      "loss": 2.4814,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 128
    },
    {
      "epoch": 0.36857142857142855,
      "grad_norm": 0.4701429648662771,
      "learning_rate": 7.833859672960942e-07,
      "loss": 2.5523,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 129
    },
    {
      "epoch": 0.37142857142857144,
      "grad_norm": 0.4229409246907706,
      "learning_rate": 7.828123044401649e-07,
      "loss": 2.4875,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 130
    },
    {
      "epoch": 0.3742857142857143,
      "grad_norm": 0.43457415950466366,
      "learning_rate": 7.822291223144563e-07,
      "loss": 2.5383,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 131
    },
    {
      "epoch": 0.37714285714285717,
      "grad_norm": 0.49201226919688135,
      "learning_rate": 7.816364354207696e-07,
      "loss": 2.4684,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 132
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.4142877503478154,
      "learning_rate": 7.810342584972584e-07,
      "loss": 2.5038,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 133
    },
    {
      "epoch": 0.38285714285714284,
      "grad_norm": 0.3993259108611048,
      "learning_rate": 7.804226065180615e-07,
      "loss": 2.4903,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 134
    },
    {
      "epoch": 0.38571428571428573,
      "grad_norm": 0.40865564645290375,
      "learning_rate": 7.798014946929305e-07,
      "loss": 2.5307,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 135
    },
    {
      "epoch": 0.38857142857142857,
      "grad_norm": 0.4261461152082098,
      "learning_rate": 7.791709384668527e-07,
      "loss": 2.5687,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 136
    },
    {
      "epoch": 0.3914285714285714,
      "grad_norm": 0.418038471120841,
      "learning_rate": 7.785309535196656e-07,
      "loss": 2.5427,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 137
    },
    {
      "epoch": 0.3942857142857143,
      "grad_norm": 0.40943036179603,
      "learning_rate": 7.778815557656681e-07,
      "loss": 2.5602,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 138
    },
    {
      "epoch": 0.39714285714285713,
      "grad_norm": 0.4272805299552179,
      "learning_rate": 7.772227613532242e-07,
      "loss": 2.4869,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 139
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.43006490216654675,
      "learning_rate": 7.765545866643613e-07,
      "loss": 2.4804,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 140
    },
    {
      "epoch": 0.40285714285714286,
      "grad_norm": 0.39125905565576724,
      "learning_rate": 7.758770483143634e-07,
      "loss": 2.5186,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 141
    },
    {
      "epoch": 0.4057142857142857,
      "grad_norm": 0.3968057506061393,
      "learning_rate": 7.751901631513575e-07,
      "loss": 2.5011,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 142
    },
    {
      "epoch": 0.4085714285714286,
      "grad_norm": 0.4223526322439004,
      "learning_rate": 7.744939482558948e-07,
      "loss": 2.4832,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 143
    },
    {
      "epoch": 0.4114285714285714,
      "grad_norm": 0.41552832703930476,
      "learning_rate": 7.737884209405261e-07,
      "loss": 2.4487,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 144
    },
    {
      "epoch": 0.4142857142857143,
      "grad_norm": 0.402866507162364,
      "learning_rate": 7.730735987493711e-07,
      "loss": 2.498,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 145
    },
    {
      "epoch": 0.41714285714285715,
      "grad_norm": 0.3948363752989273,
      "learning_rate": 7.723494994576816e-07,
      "loss": 2.4952,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 146
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.39286427623641407,
      "learning_rate": 7.71616141071401e-07,
      "loss": 2.4962,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 147
    },
    {
      "epoch": 0.4228571428571429,
      "grad_norm": 0.400561181786353,
      "learning_rate": 7.708735418267149e-07,
      "loss": 2.4952,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 148
    },
    {
      "epoch": 0.4257142857142857,
      "grad_norm": 0.39888323030727985,
      "learning_rate": 7.701217201895986e-07,
      "loss": 2.504,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 149
    },
    {
      "epoch": 0.42857142857142855,
      "grad_norm": 0.39255654087748887,
      "learning_rate": 7.693606948553577e-07,
      "loss": 2.4889,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 150
    },
    {
      "epoch": 0.43142857142857144,
      "grad_norm": 0.3840938629496403,
      "learning_rate": 7.685904847481631e-07,
      "loss": 2.5343,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 151
    },
    {
      "epoch": 0.4342857142857143,
      "grad_norm": 0.3765304120378203,
      "learning_rate": 7.678111090205803e-07,
      "loss": 2.5038,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 152
    },
    {
      "epoch": 0.43714285714285717,
      "grad_norm": 0.3953066814694025,
      "learning_rate": 7.670225870530934e-07,
      "loss": 2.4764,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 153
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.39031006206050495,
      "learning_rate": 7.662249384536236e-07,
      "loss": 2.5059,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 154
    },
    {
      "epoch": 0.44285714285714284,
      "grad_norm": 0.4000590915344439,
      "learning_rate": 7.654181830570403e-07,
      "loss": 2.4905,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 155
    },
    {
      "epoch": 0.44571428571428573,
      "grad_norm": 0.3967220458237404,
      "learning_rate": 7.646023409246693e-07,
      "loss": 2.4971,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 156
    },
    {
      "epoch": 0.44857142857142857,
      "grad_norm": 0.40092559714917914,
      "learning_rate": 7.637774323437928e-07,
      "loss": 2.4906,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 157
    },
    {
      "epoch": 0.4514285714285714,
      "grad_norm": 0.43172640483892405,
      "learning_rate": 7.629434778271459e-07,
      "loss": 2.5429,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 158
    },
    {
      "epoch": 0.4542857142857143,
      "grad_norm": 0.38549146140624163,
      "learning_rate": 7.621004981124059e-07,
      "loss": 2.5144,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 159
    },
    {
      "epoch": 0.45714285714285713,
      "grad_norm": 0.37099072572148417,
      "learning_rate": 7.612485141616766e-07,
      "loss": 2.4858,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 160
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.4229052153862539,
      "learning_rate": 7.603875471609676e-07,
      "loss": 2.467,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 161
    },
    {
      "epoch": 0.46285714285714286,
      "grad_norm": 0.41508842006882457,
      "learning_rate": 7.595176185196667e-07,
      "loss": 2.4726,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 162
    },
    {
      "epoch": 0.4657142857142857,
      "grad_norm": 0.3800593705594027,
      "learning_rate": 7.586387498700083e-07,
      "loss": 2.48,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 163
    },
    {
      "epoch": 0.4685714285714286,
      "grad_norm": 0.3882101503179876,
      "learning_rate": 7.577509630665347e-07,
      "loss": 2.4934,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 164
    },
    {
      "epoch": 0.4714285714285714,
      "grad_norm": 0.3790970007535719,
      "learning_rate": 7.568542801855533e-07,
      "loss": 2.4724,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 165
    },
    {
      "epoch": 0.4742857142857143,
      "grad_norm": 0.3874478584401711,
      "learning_rate": 7.559487235245875e-07,
      "loss": 2.4418,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 166
    },
    {
      "epoch": 0.47714285714285715,
      "grad_norm": 0.4072890336635849,
      "learning_rate": 7.550343156018216e-07,
      "loss": 2.5073,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 167
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.39208822644979924,
      "learning_rate": 7.54111079155542e-07,
      "loss": 2.4916,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 168
    },
    {
      "epoch": 0.4828571428571429,
      "grad_norm": 0.4150286468043869,
      "learning_rate": 7.531790371435707e-07,
      "loss": 2.5122,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 169
    },
    {
      "epoch": 0.4857142857142857,
      "grad_norm": 0.43892680418291813,
      "learning_rate": 7.522382127426952e-07,
      "loss": 2.4611,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 170
    },
    {
      "epoch": 0.48857142857142855,
      "grad_norm": 0.3866484457996398,
      "learning_rate": 7.512886293480914e-07,
      "loss": 2.5292,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 171
    },
    {
      "epoch": 0.49142857142857144,
      "grad_norm": 0.374270487411839,
      "learning_rate": 7.503303105727426e-07,
      "loss": 2.5349,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 172
    },
    {
      "epoch": 0.4942857142857143,
      "grad_norm": 0.3870344898458331,
      "learning_rate": 7.493632802468518e-07,
      "loss": 2.409,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 173
    },
    {
      "epoch": 0.49714285714285716,
      "grad_norm": 0.3828087659941473,
      "learning_rate": 7.483875624172496e-07,
      "loss": 2.4868,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 174
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.39195156916142243,
      "learning_rate": 7.474031813467955e-07,
      "loss": 2.4981,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 175
    },
    {
      "epoch": 0.5028571428571429,
      "grad_norm": 0.3762305036491876,
      "learning_rate": 7.464101615137755e-07,
      "loss": 2.4817,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 176
    },
    {
      "epoch": 0.5028571428571429,
      "eval_loss": 2.5930516719818115,
      "eval_runtime": 166.3488,
      "eval_samples_per_second": 1.912,
      "eval_steps_per_second": 0.319,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 176
    },
    {
      "epoch": 0.5057142857142857,
      "grad_norm": 0.37519392017744374,
      "learning_rate": 7.454085276112924e-07,
      "loss": 2.5042,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 177
    },
    {
      "epoch": 0.5085714285714286,
      "grad_norm": 0.37385205421645823,
      "learning_rate": 7.443983045466525e-07,
      "loss": 2.5327,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 178
    },
    {
      "epoch": 0.5114285714285715,
      "grad_norm": 0.38429324818547256,
      "learning_rate": 7.433795174407464e-07,
      "loss": 2.5141,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 179
    },
    {
      "epoch": 0.5142857142857142,
      "grad_norm": 0.37169737236975714,
      "learning_rate": 7.423521916274234e-07,
      "loss": 2.4033,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 180
    },
    {
      "epoch": 0.5171428571428571,
      "grad_norm": 0.3718458557397947,
      "learning_rate": 7.413163526528622e-07,
      "loss": 2.4936,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 181
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.39027962316950426,
      "learning_rate": 7.402720262749359e-07,
      "loss": 2.5185,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 182
    },
    {
      "epoch": 0.5228571428571429,
      "grad_norm": 0.38568588797346665,
      "learning_rate": 7.392192384625703e-07,
      "loss": 2.5115,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 183
    },
    {
      "epoch": 0.5257142857142857,
      "grad_norm": 0.36629863462398143,
      "learning_rate": 7.381580153951e-07,
      "loss": 2.5349,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 184
    },
    {
      "epoch": 0.5285714285714286,
      "grad_norm": 0.38697254308789736,
      "learning_rate": 7.370883834616156e-07,
      "loss": 2.4925,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 185
    },
    {
      "epoch": 0.5314285714285715,
      "grad_norm": 0.3670681252764195,
      "learning_rate": 7.360103692603086e-07,
      "loss": 2.527,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 186
    },
    {
      "epoch": 0.5342857142857143,
      "grad_norm": 0.370260215594253,
      "learning_rate": 7.349239995978094e-07,
      "loss": 2.4639,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 187
    },
    {
      "epoch": 0.5371428571428571,
      "grad_norm": 0.4238961283062229,
      "learning_rate": 7.33829301488521e-07,
      "loss": 2.4649,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 188
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.35922294875975963,
      "learning_rate": 7.327263021539476e-07,
      "loss": 2.499,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 189
    },
    {
      "epoch": 0.5428571428571428,
      "grad_norm": 0.3668321573240473,
      "learning_rate": 7.316150290220166e-07,
      "loss": 2.5089,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 190
    },
    {
      "epoch": 0.5457142857142857,
      "grad_norm": 0.3656484396425331,
      "learning_rate": 7.304955097263979e-07,
      "loss": 2.4967,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 191
    },
    {
      "epoch": 0.5485714285714286,
      "grad_norm": 0.3572495999236213,
      "learning_rate": 7.293677721058156e-07,
      "loss": 2.4714,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 192
    },
    {
      "epoch": 0.5514285714285714,
      "grad_norm": 0.3653390381041467,
      "learning_rate": 7.282318442033566e-07,
      "loss": 2.5329,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 193
    },
    {
      "epoch": 0.5542857142857143,
      "grad_norm": 0.37176497769920275,
      "learning_rate": 7.270877542657725e-07,
      "loss": 2.472,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 194
    },
    {
      "epoch": 0.5571428571428572,
      "grad_norm": 0.37548393631079474,
      "learning_rate": 7.25935530742778e-07,
      "loss": 2.4136,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 195
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.36600893819205693,
      "learning_rate": 7.247752022863426e-07,
      "loss": 2.466,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 196
    },
    {
      "epoch": 0.5628571428571428,
      "grad_norm": 0.3786638899387433,
      "learning_rate": 7.23606797749979e-07,
      "loss": 2.5049,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 197
    },
    {
      "epoch": 0.5657142857142857,
      "grad_norm": 0.3573567232595577,
      "learning_rate": 7.224303461880248e-07,
      "loss": 2.5074,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 198
    },
    {
      "epoch": 0.5685714285714286,
      "grad_norm": 0.35820433976274557,
      "learning_rate": 7.212458768549207e-07,
      "loss": 2.5282,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 199
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 0.37434826849729697,
      "learning_rate": 7.200534192044825e-07,
      "loss": 2.5049,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 200
    },
    {
      "epoch": 0.5742857142857143,
      "grad_norm": 0.376513282394864,
      "learning_rate": 7.18853002889169e-07,
      "loss": 2.5523,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 201
    },
    {
      "epoch": 0.5771428571428572,
      "grad_norm": 0.3694477316722928,
      "learning_rate": 7.176446577593444e-07,
      "loss": 2.4656,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 202
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.36718646520606185,
      "learning_rate": 7.164284138625366e-07,
      "loss": 2.5154,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 203
    },
    {
      "epoch": 0.5828571428571429,
      "grad_norm": 0.3872658815719157,
      "learning_rate": 7.152043014426887e-07,
      "loss": 2.4765,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 204
    },
    {
      "epoch": 0.5857142857142857,
      "grad_norm": 0.3561649216531024,
      "learning_rate": 7.139723509394089e-07,
      "loss": 2.4781,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 205
    },
    {
      "epoch": 0.5885714285714285,
      "grad_norm": 0.38255434129156507,
      "learning_rate": 7.12732592987212e-07,
      "loss": 2.4943,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 206
    },
    {
      "epoch": 0.5914285714285714,
      "grad_norm": 0.3755870761342919,
      "learning_rate": 7.114850584147576e-07,
      "loss": 2.4948,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 207
    },
    {
      "epoch": 0.5942857142857143,
      "grad_norm": 0.37469846735313367,
      "learning_rate": 7.102297782440848e-07,
      "loss": 2.5028,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 208
    },
    {
      "epoch": 0.5971428571428572,
      "grad_norm": 0.348032646562249,
      "learning_rate": 7.089667836898397e-07,
      "loss": 2.4474,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 209
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.3562480926853211,
      "learning_rate": 7.076961061584995e-07,
      "loss": 2.4841,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 210
    },
    {
      "epoch": 0.6028571428571429,
      "grad_norm": 0.35444730365280064,
      "learning_rate": 7.064177772475911e-07,
      "loss": 2.4572,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 211
    },
    {
      "epoch": 0.6057142857142858,
      "grad_norm": 0.35078366951080886,
      "learning_rate": 7.051318287449066e-07,
      "loss": 2.5588,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 212
    },
    {
      "epoch": 0.6085714285714285,
      "grad_norm": 0.3705344934520105,
      "learning_rate": 7.038382926277112e-07,
      "loss": 2.5239,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 213
    },
    {
      "epoch": 0.6114285714285714,
      "grad_norm": 0.3718368713458524,
      "learning_rate": 7.025372010619493e-07,
      "loss": 2.5013,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 214
    },
    {
      "epoch": 0.6142857142857143,
      "grad_norm": 0.3654665915105373,
      "learning_rate": 7.012285864014444e-07,
      "loss": 2.4875,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 215
    },
    {
      "epoch": 0.6171428571428571,
      "grad_norm": 0.3546240195117714,
      "learning_rate": 6.999124811870936e-07,
      "loss": 2.5023,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 216
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.42051092189800676,
      "learning_rate": 6.9858891814606e-07,
      "loss": 2.5003,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 217
    },
    {
      "epoch": 0.6228571428571429,
      "grad_norm": 0.3491245054666944,
      "learning_rate": 6.972579301909577e-07,
      "loss": 2.461,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 218
    },
    {
      "epoch": 0.6257142857142857,
      "grad_norm": 0.35658395759457545,
      "learning_rate": 6.959195504190336e-07,
      "loss": 2.5134,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 219
    },
    {
      "epoch": 0.6285714285714286,
      "grad_norm": 0.36361102512046456,
      "learning_rate": 6.945738121113448e-07,
      "loss": 2.5126,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 220
    },
    {
      "epoch": 0.6314285714285715,
      "grad_norm": 0.34356912968361036,
      "learning_rate": 6.932207487319305e-07,
      "loss": 2.4981,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 221
    },
    {
      "epoch": 0.6342857142857142,
      "grad_norm": 0.36129225439181545,
      "learning_rate": 6.918603939269803e-07,
      "loss": 2.4806,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 222
    },
    {
      "epoch": 0.6371428571428571,
      "grad_norm": 0.3838496084297271,
      "learning_rate": 6.904927815239972e-07,
      "loss": 2.4465,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 223
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.3731089194966591,
      "learning_rate": 6.891179455309565e-07,
      "loss": 2.5182,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 224
    },
    {
      "epoch": 0.6428571428571429,
      "grad_norm": 0.3586114444632,
      "learning_rate": 6.877359201354604e-07,
      "loss": 2.4027,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 225
    },
    {
      "epoch": 0.6457142857142857,
      "grad_norm": 0.3772140684266388,
      "learning_rate": 6.863467397038874e-07,
      "loss": 2.4736,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 226
    },
    {
      "epoch": 0.6485714285714286,
      "grad_norm": 0.3751799521678175,
      "learning_rate": 6.849504387805379e-07,
      "loss": 2.4965,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 227
    },
    {
      "epoch": 0.6514285714285715,
      "grad_norm": 0.36426930615242076,
      "learning_rate": 6.835470520867757e-07,
      "loss": 2.4426,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 228
    },
    {
      "epoch": 0.6542857142857142,
      "grad_norm": 0.3619221552357011,
      "learning_rate": 6.821366145201635e-07,
      "loss": 2.5004,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 229
    },
    {
      "epoch": 0.6571428571428571,
      "grad_norm": 0.3441559013579583,
      "learning_rate": 6.807191611535966e-07,
      "loss": 2.4278,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 230
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.3525102144925795,
      "learning_rate": 6.792947272344291e-07,
      "loss": 2.4986,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 231
    },
    {
      "epoch": 0.6628571428571428,
      "grad_norm": 0.36803179296456207,
      "learning_rate": 6.778633481835989e-07,
      "loss": 2.4796,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 232
    },
    {
      "epoch": 0.6657142857142857,
      "grad_norm": 0.3401902245889775,
      "learning_rate": 6.764250595947458e-07,
      "loss": 2.3971,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 233
    },
    {
      "epoch": 0.6685714285714286,
      "grad_norm": 0.34844003283488245,
      "learning_rate": 6.749798972333271e-07,
      "loss": 2.509,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 234
    },
    {
      "epoch": 0.6714285714285714,
      "grad_norm": 0.3623544280284417,
      "learning_rate": 6.735278970357279e-07,
      "loss": 2.5387,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 235
    },
    {
      "epoch": 0.6742857142857143,
      "grad_norm": 0.3602058774995079,
      "learning_rate": 6.720690951083677e-07,
      "loss": 2.4656,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 236
    },
    {
      "epoch": 0.6771428571428572,
      "grad_norm": 0.3589282701413255,
      "learning_rate": 6.706035277268021e-07,
      "loss": 2.4562,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 237
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.3438222001156716,
      "learning_rate": 6.691312313348211e-07,
      "loss": 2.4953,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 238
    },
    {
      "epoch": 0.6828571428571428,
      "grad_norm": 0.37953839811743645,
      "learning_rate": 6.676522425435432e-07,
      "loss": 2.5069,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 239
    },
    {
      "epoch": 0.6857142857142857,
      "grad_norm": 0.4025431860338764,
      "learning_rate": 6.661665981305044e-07,
      "loss": 2.5095,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 240
    },
    {
      "epoch": 0.6885714285714286,
      "grad_norm": 0.36980627031248103,
      "learning_rate": 6.646743350387437e-07,
      "loss": 2.4929,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 241
    },
    {
      "epoch": 0.6914285714285714,
      "grad_norm": 0.3458583250356646,
      "learning_rate": 6.63175490375885e-07,
      "loss": 2.4784,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 242
    },
    {
      "epoch": 0.6942857142857143,
      "grad_norm": 0.3570435105604994,
      "learning_rate": 6.616701014132138e-07,
      "loss": 2.4906,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 243
    },
    {
      "epoch": 0.6971428571428572,
      "grad_norm": 0.35463642906397286,
      "learning_rate": 6.601582055847504e-07,
      "loss": 2.4485,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 244
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.36840330286900297,
      "learning_rate": 6.586398404863198e-07,
      "loss": 2.4521,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 245
    },
    {
      "epoch": 0.7028571428571428,
      "grad_norm": 0.34466271455495584,
      "learning_rate": 6.571150438746157e-07,
      "loss": 2.4799,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 246
    },
    {
      "epoch": 0.7057142857142857,
      "grad_norm": 0.3421913203365403,
      "learning_rate": 6.555838536662624e-07,
      "loss": 2.4996,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 247
    },
    {
      "epoch": 0.7085714285714285,
      "grad_norm": 0.3377100744517225,
      "learning_rate": 6.540463079368714e-07,
      "loss": 2.5031,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 248
    },
    {
      "epoch": 0.7114285714285714,
      "grad_norm": 0.35394202496673854,
      "learning_rate": 6.525024449200955e-07,
      "loss": 2.4462,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 249
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 0.3569603019955735,
      "learning_rate": 6.50952303006677e-07,
      "loss": 2.4558,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 250
    },
    {
      "epoch": 0.7171428571428572,
      "grad_norm": 0.35487770612468417,
      "learning_rate": 6.493959207434934e-07,
      "loss": 2.4587,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 251
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.3822919254629326,
      "learning_rate": 6.478333368325993e-07,
      "loss": 2.4925,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 252
    },
    {
      "epoch": 0.7228571428571429,
      "grad_norm": 0.3405814925489275,
      "learning_rate": 6.462645901302632e-07,
      "loss": 2.5055,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 253
    },
    {
      "epoch": 0.7257142857142858,
      "grad_norm": 0.37745990434258037,
      "learning_rate": 6.446897196460025e-07,
      "loss": 2.5104,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 254
    },
    {
      "epoch": 0.7285714285714285,
      "grad_norm": 0.3516630571600152,
      "learning_rate": 6.43108764541612e-07,
      "loss": 2.5216,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 255
    },
    {
      "epoch": 0.7314285714285714,
      "grad_norm": 0.4332727052093402,
      "learning_rate": 6.41521764130191e-07,
      "loss": 2.4266,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 256
    },
    {
      "epoch": 0.7342857142857143,
      "grad_norm": 0.35301862780660254,
      "learning_rate": 6.399287578751655e-07,
      "loss": 2.5304,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 257
    },
    {
      "epoch": 0.7371428571428571,
      "grad_norm": 0.3456685078132819,
      "learning_rate": 6.383297853893069e-07,
      "loss": 2.4224,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 258
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.3636772884814302,
      "learning_rate": 6.367248864337471e-07,
      "loss": 2.3889,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 259
    },
    {
      "epoch": 0.7428571428571429,
      "grad_norm": 0.3608518088641067,
      "learning_rate": 6.351141009169893e-07,
      "loss": 2.4369,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 260
    },
    {
      "epoch": 0.7457142857142857,
      "grad_norm": 0.34810272871258524,
      "learning_rate": 6.33497468893916e-07,
      "loss": 2.4061,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 261
    },
    {
      "epoch": 0.7485714285714286,
      "grad_norm": 0.3421689292003609,
      "learning_rate": 6.31875030564793e-07,
      "loss": 2.511,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 262
    },
    {
      "epoch": 0.7514285714285714,
      "grad_norm": 0.34911833160977096,
      "learning_rate": 6.302468262742694e-07,
      "loss": 2.442,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 263
    },
    {
      "epoch": 0.7542857142857143,
      "grad_norm": 0.39355376033973877,
      "learning_rate": 6.286128965103749e-07,
      "loss": 2.4687,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 264
    },
    {
      "epoch": 0.7542857142857143,
      "eval_loss": 2.565204620361328,
      "eval_runtime": 167.6749,
      "eval_samples_per_second": 1.897,
      "eval_steps_per_second": 0.316,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 264
    },
    {
      "epoch": 0.7571428571428571,
      "grad_norm": 0.39457207771925396,
      "learning_rate": 6.269732819035127e-07,
      "loss": 2.4698,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 265
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.3334084012424826,
      "learning_rate": 6.253280232254487e-07,
      "loss": 2.4585,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 266
    },
    {
      "epoch": 0.7628571428571429,
      "grad_norm": 0.34318029118007076,
      "learning_rate": 6.236771613882987e-07,
      "loss": 2.4876,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 267
    },
    {
      "epoch": 0.7657142857142857,
      "grad_norm": 0.3486014500652036,
      "learning_rate": 6.2202073744351e-07,
      "loss": 2.4419,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 268
    },
    {
      "epoch": 0.7685714285714286,
      "grad_norm": 0.34709699915754927,
      "learning_rate": 6.20358792580841e-07,
      "loss": 2.5366,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 269
    },
    {
      "epoch": 0.7714285714285715,
      "grad_norm": 0.35321437117932697,
      "learning_rate": 6.186913681273374e-07,
      "loss": 2.4664,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 270
    },
    {
      "epoch": 0.7742857142857142,
      "grad_norm": 0.3666149829104433,
      "learning_rate": 6.170185055463037e-07,
      "loss": 2.4437,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 271
    },
    {
      "epoch": 0.7771428571428571,
      "grad_norm": 0.3564294569104391,
      "learning_rate": 6.153402464362729e-07,
      "loss": 2.4571,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 272
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.351816369317707,
      "learning_rate": 6.136566325299714e-07,
      "loss": 2.4257,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 273
    },
    {
      "epoch": 0.7828571428571428,
      "grad_norm": 0.34530603321181286,
      "learning_rate": 6.119677056932819e-07,
      "loss": 2.4608,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 274
    },
    {
      "epoch": 0.7857142857142857,
      "grad_norm": 0.34434546456869247,
      "learning_rate": 6.102735079242018e-07,
      "loss": 2.4469,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 275
    },
    {
      "epoch": 0.7885714285714286,
      "grad_norm": 0.3432147331263689,
      "learning_rate": 6.085740813517991e-07,
      "loss": 2.4452,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 276
    },
    {
      "epoch": 0.7914285714285715,
      "grad_norm": 0.3591286629845744,
      "learning_rate": 6.06869468235165e-07,
      "loss": 2.4868,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 277
    },
    {
      "epoch": 0.7942857142857143,
      "grad_norm": 0.3378444324938211,
      "learning_rate": 6.051597109623624e-07,
      "loss": 2.481,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 278
    },
    {
      "epoch": 0.7971428571428572,
      "grad_norm": 0.4122524038732501,
      "learning_rate": 6.034448520493727e-07,
      "loss": 2.473,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 279
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.363566312056061,
      "learning_rate": 6.017249341390376e-07,
      "loss": 2.4669,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 280
    },
    {
      "epoch": 0.8028571428571428,
      "grad_norm": 0.3672200412080197,
      "learning_rate": 6e-07,
      "loss": 2.4796,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 281
    },
    {
      "epoch": 0.8057142857142857,
      "grad_norm": 0.3875340509826031,
      "learning_rate": 5.982700925256391e-07,
      "loss": 2.4532,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 282
    },
    {
      "epoch": 0.8085714285714286,
      "grad_norm": 0.33943192402270767,
      "learning_rate": 5.965352547330045e-07,
      "loss": 2.4828,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 283
    },
    {
      "epoch": 0.8114285714285714,
      "grad_norm": 0.3182851807778329,
      "learning_rate": 5.947955297617469e-07,
      "loss": 2.5062,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 284
    },
    {
      "epoch": 0.8142857142857143,
      "grad_norm": 0.3351183005878844,
      "learning_rate": 5.930509608730443e-07,
      "loss": 2.4893,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 285
    },
    {
      "epoch": 0.8171428571428572,
      "grad_norm": 0.33759896567086656,
      "learning_rate": 5.913015914485272e-07,
      "loss": 2.4442,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 286
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.3399416489048959,
      "learning_rate": 5.895474649891994e-07,
      "loss": 2.4514,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 287
    },
    {
      "epoch": 0.8228571428571428,
      "grad_norm": 0.3545440479908436,
      "learning_rate": 5.877886251143563e-07,
      "loss": 2.4718,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 288
    },
    {
      "epoch": 0.8257142857142857,
      "grad_norm": 0.335178601825211,
      "learning_rate": 5.860251155605003e-07,
      "loss": 2.4825,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 289
    },
    {
      "epoch": 0.8285714285714286,
      "grad_norm": 0.35774830391005497,
      "learning_rate": 5.842569801802529e-07,
      "loss": 2.4472,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 290
    },
    {
      "epoch": 0.8314285714285714,
      "grad_norm": 0.36084872276367186,
      "learning_rate": 5.824842629412652e-07,
      "loss": 2.445,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 291
    },
    {
      "epoch": 0.8342857142857143,
      "grad_norm": 0.35242565851295765,
      "learning_rate": 5.807070079251234e-07,
      "loss": 2.4886,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 292
    },
    {
      "epoch": 0.8371428571428572,
      "grad_norm": 0.3495350057811423,
      "learning_rate": 5.78925259326253e-07,
      "loss": 2.471,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 293
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.3806230860115398,
      "learning_rate": 5.771390614508207e-07,
      "loss": 2.4581,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 294
    },
    {
      "epoch": 0.8428571428571429,
      "grad_norm": 0.38113919527634504,
      "learning_rate": 5.75348458715631e-07,
      "loss": 2.5104,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 295
    },
    {
      "epoch": 0.8457142857142858,
      "grad_norm": 0.349648018876053,
      "learning_rate": 5.735534956470232e-07,
      "loss": 2.5099,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 296
    },
    {
      "epoch": 0.8485714285714285,
      "grad_norm": 0.34479267434882577,
      "learning_rate": 5.717542168797637e-07,
      "loss": 2.469,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 297
    },
    {
      "epoch": 0.8514285714285714,
      "grad_norm": 0.3663200349720271,
      "learning_rate": 5.699506671559354e-07,
      "loss": 2.4824,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 298
    },
    {
      "epoch": 0.8542857142857143,
      "grad_norm": 0.366066016946436,
      "learning_rate": 5.681428913238262e-07,
      "loss": 2.4725,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 299
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 0.3807073720670669,
      "learning_rate": 5.663309343368132e-07,
      "loss": 2.5099,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 300
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.3515126047720806,
      "learning_rate": 5.645148412522447e-07,
      "loss": 2.5591,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 301
    },
    {
      "epoch": 0.8628571428571429,
      "grad_norm": 0.5357579807606813,
      "learning_rate": 5.626946572303201e-07,
      "loss": 2.5283,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 302
    },
    {
      "epoch": 0.8657142857142858,
      "grad_norm": 0.3416675874952649,
      "learning_rate": 5.60870427532967e-07,
      "loss": 2.4697,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 303
    },
    {
      "epoch": 0.8685714285714285,
      "grad_norm": 0.3635810289310478,
      "learning_rate": 5.590421975227153e-07,
      "loss": 2.4615,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 304
    },
    {
      "epoch": 0.8714285714285714,
      "grad_norm": 0.358882250757311,
      "learning_rate": 5.572100126615694e-07,
      "loss": 2.4288,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 305
    },
    {
      "epoch": 0.8742857142857143,
      "grad_norm": 0.37834950353981683,
      "learning_rate": 5.553739185098779e-07,
      "loss": 2.4728,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 306
    },
    {
      "epoch": 0.8771428571428571,
      "grad_norm": 0.3387084046110639,
      "learning_rate": 5.535339607252002e-07,
      "loss": 2.4579,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 307
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.3436579538730971,
      "learning_rate": 5.516901850611714e-07,
      "loss": 2.476,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 308
    },
    {
      "epoch": 0.8828571428571429,
      "grad_norm": 0.3396777904636505,
      "learning_rate": 5.498426373663648e-07,
      "loss": 2.4724,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 309
    },
    {
      "epoch": 0.8857142857142857,
      "grad_norm": 0.37046893914260753,
      "learning_rate": 5.479913635831516e-07,
      "loss": 2.39,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 310
    },
    {
      "epoch": 0.8885714285714286,
      "grad_norm": 0.3635286693484049,
      "learning_rate": 5.461364097465581e-07,
      "loss": 2.5156,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 311
    },
    {
      "epoch": 0.8914285714285715,
      "grad_norm": 0.33577475696395825,
      "learning_rate": 5.442778219831215e-07,
      "loss": 2.4255,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 312
    },
    {
      "epoch": 0.8942857142857142,
      "grad_norm": 0.34473312763568614,
      "learning_rate": 5.424156465097427e-07,
      "loss": 2.4546,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 313
    },
    {
      "epoch": 0.8971428571428571,
      "grad_norm": 0.3798074503322467,
      "learning_rate": 5.405499296325371e-07,
      "loss": 2.4886,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 314
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.3728293506945305,
      "learning_rate": 5.386807177456829e-07,
      "loss": 2.4671,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 315
    },
    {
      "epoch": 0.9028571428571428,
      "grad_norm": 0.3391677878009453,
      "learning_rate": 5.368080573302676e-07,
      "loss": 2.5042,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 316
    },
    {
      "epoch": 0.9057142857142857,
      "grad_norm": 0.33238222520907174,
      "learning_rate": 5.34931994953132e-07,
      "loss": 2.4796,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 317
    },
    {
      "epoch": 0.9085714285714286,
      "grad_norm": 0.3604834906195999,
      "learning_rate": 5.330525772657128e-07,
      "loss": 2.4535,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 318
    },
    {
      "epoch": 0.9114285714285715,
      "grad_norm": 0.3420935222719714,
      "learning_rate": 5.311698510028818e-07,
      "loss": 2.4873,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 319
    },
    {
      "epoch": 0.9142857142857143,
      "grad_norm": 0.345113659125592,
      "learning_rate": 5.29283862981784e-07,
      "loss": 2.4684,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 320
    },
    {
      "epoch": 0.9171428571428571,
      "grad_norm": 0.3349235797984999,
      "learning_rate": 5.273946601006738e-07,
      "loss": 2.4981,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 321
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.3467727945903166,
      "learning_rate": 5.255022893377483e-07,
      "loss": 2.515,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 322
    },
    {
      "epoch": 0.9228571428571428,
      "grad_norm": 0.3371310270277657,
      "learning_rate": 5.23606797749979e-07,
      "loss": 2.4812,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 323
    },
    {
      "epoch": 0.9257142857142857,
      "grad_norm": 0.34310966718111924,
      "learning_rate": 5.217082324719425e-07,
      "loss": 2.4542,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 324
    },
    {
      "epoch": 0.9285714285714286,
      "grad_norm": 0.35562479535135727,
      "learning_rate": 5.198066407146477e-07,
      "loss": 2.4605,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 325
    },
    {
      "epoch": 0.9314285714285714,
      "grad_norm": 0.3812077570186057,
      "learning_rate": 5.179020697643617e-07,
      "loss": 2.4506,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 326
    },
    {
      "epoch": 0.9342857142857143,
      "grad_norm": 0.3508570582420508,
      "learning_rate": 5.159945669814344e-07,
      "loss": 2.4514,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 327
    },
    {
      "epoch": 0.9371428571428572,
      "grad_norm": 0.39699693296370153,
      "learning_rate": 5.140841797991209e-07,
      "loss": 2.5083,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 328
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.33743971851632737,
      "learning_rate": 5.121709557224011e-07,
      "loss": 2.5216,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 329
    },
    {
      "epoch": 0.9428571428571428,
      "grad_norm": 0.33393298334172467,
      "learning_rate": 5.102549423267996e-07,
      "loss": 2.5103,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 330
    },
    {
      "epoch": 0.9457142857142857,
      "grad_norm": 0.33910952276555056,
      "learning_rate": 5.08336187257202e-07,
      "loss": 2.4786,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 331
    },
    {
      "epoch": 0.9485714285714286,
      "grad_norm": 0.3561519068878474,
      "learning_rate": 5.0641473822667e-07,
      "loss": 2.4609,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 332
    },
    {
      "epoch": 0.9514285714285714,
      "grad_norm": 0.34159371173439246,
      "learning_rate": 5.044906430152553e-07,
      "loss": 2.4168,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 333
    },
    {
      "epoch": 0.9542857142857143,
      "grad_norm": 0.3393671508648039,
      "learning_rate": 5.025639494688113e-07,
      "loss": 2.4547,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 334
    },
    {
      "epoch": 0.9571428571428572,
      "grad_norm": 0.3506954165274522,
      "learning_rate": 5.006347054978034e-07,
      "loss": 2.4408,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 335
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.3385523672512557,
      "learning_rate": 4.987029590761174e-07,
      "loss": 2.4604,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 336
    },
    {
      "epoch": 0.9628571428571429,
      "grad_norm": 0.353159692195445,
      "learning_rate": 4.96768758239867e-07,
      "loss": 2.4716,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 337
    },
    {
      "epoch": 0.9657142857142857,
      "grad_norm": 0.3475130299316105,
      "learning_rate": 4.94832151086199e-07,
      "loss": 2.4077,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 338
    },
    {
      "epoch": 0.9685714285714285,
      "grad_norm": 0.3586068561700244,
      "learning_rate": 4.92893185772097e-07,
      "loss": 2.4094,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 339
    },
    {
      "epoch": 0.9714285714285714,
      "grad_norm": 0.36365819067701627,
      "learning_rate": 4.909519105131845e-07,
      "loss": 2.4414,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 340
    },
    {
      "epoch": 0.9742857142857143,
      "grad_norm": 0.3562742387071953,
      "learning_rate": 4.890083735825257e-07,
      "loss": 2.4208,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 341
    },
    {
      "epoch": 0.9771428571428571,
      "grad_norm": 0.33248196034594213,
      "learning_rate": 4.87062623309425e-07,
      "loss": 2.472,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 342
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.338789431522657,
      "learning_rate": 4.851147080782249e-07,
      "loss": 2.4743,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 343
    },
    {
      "epoch": 0.9828571428571429,
      "grad_norm": 0.3403725818218743,
      "learning_rate": 4.831646763271037e-07,
      "loss": 2.4962,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 344
    },
    {
      "epoch": 0.9857142857142858,
      "grad_norm": 0.334286704524842,
      "learning_rate": 4.812125765468704e-07,
      "loss": 2.4826,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 345
    },
    {
      "epoch": 0.9885714285714285,
      "grad_norm": 0.3510406468570529,
      "learning_rate": 4.79258457279759e-07,
      "loss": 2.5145,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 346
    },
    {
      "epoch": 0.9914285714285714,
      "grad_norm": 0.3517494367287261,
      "learning_rate": 4.773023671182213e-07,
      "loss": 2.4645,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 347
    },
    {
      "epoch": 0.9942857142857143,
      "grad_norm": 0.3348806749973408,
      "learning_rate": 4.7534435470371877e-07,
      "loss": 2.4781,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 348
    },
    {
      "epoch": 0.9971428571428571,
      "grad_norm": 0.34530897795031784,
      "learning_rate": 4.733844687255129e-07,
      "loss": 2.4746,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 349
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.36909691649461324,
      "learning_rate": 4.7142275791945465e-07,
      "loss": 2.5425,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 350
    },
    {
      "epoch": 1.002857142857143,
      "grad_norm": 0.3578530170999241,
      "learning_rate": 4.694592710667722e-07,
      "loss": 2.4482,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 351
    },
    {
      "epoch": 1.0057142857142858,
      "grad_norm": 0.36348392212072117,
      "learning_rate": 4.6749405699285794e-07,
      "loss": 2.4625,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 352
    },
    {
      "epoch": 1.0057142857142858,
      "eval_loss": 2.5527780055999756,
      "eval_runtime": 166.4811,
      "eval_samples_per_second": 1.91,
      "eval_steps_per_second": 0.318,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 352
    },
    {
      "epoch": 1.0085714285714287,
      "grad_norm": 0.3577329043966054,
      "learning_rate": 4.6552716456605507e-07,
      "loss": 2.4481,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 353
    },
    {
      "epoch": 1.0114285714285713,
      "grad_norm": 0.352880101184416,
      "learning_rate": 4.635586426964413e-07,
      "loss": 2.5377,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 354
    },
    {
      "epoch": 1.0142857142857142,
      "grad_norm": 0.42889431086007757,
      "learning_rate": 4.6158854033461337e-07,
      "loss": 2.4847,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 355
    },
    {
      "epoch": 1.0171428571428571,
      "grad_norm": 0.3381662796929482,
      "learning_rate": 4.596169064704697e-07,
      "loss": 2.4133,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 356
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.350548022193557,
      "learning_rate": 4.5764379013199203e-07,
      "loss": 2.4632,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 357
    },
    {
      "epoch": 1.022857142857143,
      "grad_norm": 0.3478085448863657,
      "learning_rate": 4.5566924038402616e-07,
      "loss": 2.4731,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 358
    },
    {
      "epoch": 1.0257142857142858,
      "grad_norm": 0.34385631708589265,
      "learning_rate": 4.536933063270622e-07,
      "loss": 2.4597,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 359
    },
    {
      "epoch": 1.0285714285714285,
      "grad_norm": 0.36510630405161154,
      "learning_rate": 4.517160370960133e-07,
      "loss": 2.4969,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 360
    },
    {
      "epoch": 1.0314285714285714,
      "grad_norm": 0.3455317005759876,
      "learning_rate": 4.4973748185899413e-07,
      "loss": 2.4323,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 361
    },
    {
      "epoch": 1.0342857142857143,
      "grad_norm": 0.35113457340719884,
      "learning_rate": 4.477576898160976e-07,
      "loss": 2.4361,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 362
    },
    {
      "epoch": 1.0371428571428571,
      "grad_norm": 0.38263449758609114,
      "learning_rate": 4.457767101981727e-07,
      "loss": 2.4954,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 363
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.36341836552851947,
      "learning_rate": 4.437945922655988e-07,
      "loss": 2.4656,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 364
    },
    {
      "epoch": 1.042857142857143,
      "grad_norm": 0.3915156770516886,
      "learning_rate": 4.418113853070614e-07,
      "loss": 2.415,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 365
    },
    {
      "epoch": 1.0457142857142858,
      "grad_norm": 0.33195674839892325,
      "learning_rate": 4.3982713863832666e-07,
      "loss": 2.4604,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 366
    },
    {
      "epoch": 1.0485714285714285,
      "grad_norm": 0.33623308524494583,
      "learning_rate": 4.3784190160101484e-07,
      "loss": 2.5282,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 367
    },
    {
      "epoch": 1.0514285714285714,
      "grad_norm": 0.34274894731728034,
      "learning_rate": 4.358557235613734e-07,
      "loss": 2.4833,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 368
    },
    {
      "epoch": 1.0542857142857143,
      "grad_norm": 0.3443588545705623,
      "learning_rate": 4.338686539090493e-07,
      "loss": 2.4915,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 369
    },
    {
      "epoch": 1.0571428571428572,
      "grad_norm": 0.33689255831949766,
      "learning_rate": 4.3188074205586105e-07,
      "loss": 2.4955,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 370
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.35974276935560007,
      "learning_rate": 4.298920374345697e-07,
      "loss": 2.4612,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 371
    },
    {
      "epoch": 1.062857142857143,
      "grad_norm": 0.36308621230901933,
      "learning_rate": 4.2790258949765005e-07,
      "loss": 2.4531,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 372
    },
    {
      "epoch": 1.0657142857142856,
      "grad_norm": 0.3316452780139297,
      "learning_rate": 4.2591244771606065e-07,
      "loss": 2.4572,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 373
    },
    {
      "epoch": 1.0685714285714285,
      "grad_norm": 0.4357272850215446,
      "learning_rate": 4.239216615780137e-07,
      "loss": 2.5267,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 374
    },
    {
      "epoch": 1.0714285714285714,
      "grad_norm": 0.3400549968340237,
      "learning_rate": 4.21930280587744e-07,
      "loss": 2.4962,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 375
    },
    {
      "epoch": 1.0742857142857143,
      "grad_norm": 0.3857487670450144,
      "learning_rate": 4.1993835426427883e-07,
      "loss": 2.4742,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 376
    },
    {
      "epoch": 1.0771428571428572,
      "grad_norm": 0.3283783363348819,
      "learning_rate": 4.17945932140206e-07,
      "loss": 2.4549,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 377
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.3518027148209554,
      "learning_rate": 4.15953063760442e-07,
      "loss": 2.4465,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 378
    },
    {
      "epoch": 1.0828571428571427,
      "grad_norm": 0.3726941488206686,
      "learning_rate": 4.1395979868100045e-07,
      "loss": 2.4325,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 379
    },
    {
      "epoch": 1.0857142857142856,
      "grad_norm": 0.3336066212925924,
      "learning_rate": 4.119661864677594e-07,
      "loss": 2.4815,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 380
    },
    {
      "epoch": 1.0885714285714285,
      "grad_norm": 0.33868704810395195,
      "learning_rate": 4.0997227669522914e-07,
      "loss": 2.4242,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 381
    },
    {
      "epoch": 1.0914285714285714,
      "grad_norm": 0.341965769286426,
      "learning_rate": 4.0797811894531924e-07,
      "loss": 2.4474,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 382
    },
    {
      "epoch": 1.0942857142857143,
      "grad_norm": 0.3346430420174228,
      "learning_rate": 4.0598376280610545e-07,
      "loss": 2.4926,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 383
    },
    {
      "epoch": 1.0971428571428572,
      "grad_norm": 0.36277934596341493,
      "learning_rate": 4.03989257870597e-07,
      "loss": 2.5039,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 384
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.3604456268974496,
      "learning_rate": 4.019946537355032e-07,
      "loss": 2.4841,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 385
    },
    {
      "epoch": 1.1028571428571428,
      "grad_norm": 0.3482087648645755,
      "learning_rate": 4e-07,
      "loss": 2.5283,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 386
    },
    {
      "epoch": 1.1057142857142856,
      "grad_norm": 0.3547337148632499,
      "learning_rate": 3.980053462644968e-07,
      "loss": 2.4847,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 387
    },
    {
      "epoch": 1.1085714285714285,
      "grad_norm": 0.36292593219541636,
      "learning_rate": 3.9601074212940297e-07,
      "loss": 2.4835,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 388
    },
    {
      "epoch": 1.1114285714285714,
      "grad_norm": 0.3453897469743714,
      "learning_rate": 3.940162371938946e-07,
      "loss": 2.4567,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 389
    },
    {
      "epoch": 1.1142857142857143,
      "grad_norm": 0.35139175655598165,
      "learning_rate": 3.9202188105468083e-07,
      "loss": 2.5112,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 390
    },
    {
      "epoch": 1.1171428571428572,
      "grad_norm": 0.3533033977421551,
      "learning_rate": 3.900277233047709e-07,
      "loss": 2.3753,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 391
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.33957884295880786,
      "learning_rate": 3.880338135322407e-07,
      "loss": 2.4334,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 392
    },
    {
      "epoch": 1.1228571428571428,
      "grad_norm": 0.36081617151412526,
      "learning_rate": 3.860402013189997e-07,
      "loss": 2.5385,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 393
    },
    {
      "epoch": 1.1257142857142857,
      "grad_norm": 0.3327290678763158,
      "learning_rate": 3.8404693623955813e-07,
      "loss": 2.4517,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 394
    },
    {
      "epoch": 1.1285714285714286,
      "grad_norm": 0.34818532351495185,
      "learning_rate": 3.8205406785979413e-07,
      "loss": 2.4664,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 395
    },
    {
      "epoch": 1.1314285714285715,
      "grad_norm": 0.3415725648684764,
      "learning_rate": 3.800616457357211e-07,
      "loss": 2.5378,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 396
    },
    {
      "epoch": 1.1342857142857143,
      "grad_norm": 0.35676653198878794,
      "learning_rate": 3.780697194122559e-07,
      "loss": 2.4732,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 397
    },
    {
      "epoch": 1.1371428571428572,
      "grad_norm": 0.3476180456340453,
      "learning_rate": 3.760783384219863e-07,
      "loss": 2.4922,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 398
    },
    {
      "epoch": 1.1400000000000001,
      "grad_norm": 0.33920096065661004,
      "learning_rate": 3.740875522839392e-07,
      "loss": 2.5146,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 399
    },
    {
      "epoch": 1.1428571428571428,
      "grad_norm": 0.3531091814802872,
      "learning_rate": 3.7209741050234986e-07,
      "loss": 2.3837,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 400
    },
    {
      "epoch": 1.1457142857142857,
      "grad_norm": 0.33247904592495287,
      "learning_rate": 3.7010796256543027e-07,
      "loss": 2.5067,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 401
    },
    {
      "epoch": 1.1485714285714286,
      "grad_norm": 0.4399124106113077,
      "learning_rate": 3.68119257944139e-07,
      "loss": 2.448,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 402
    },
    {
      "epoch": 1.1514285714285715,
      "grad_norm": 0.36179117750921613,
      "learning_rate": 3.661313460909507e-07,
      "loss": 2.4843,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 403
    },
    {
      "epoch": 1.1542857142857144,
      "grad_norm": 0.3361677234737866,
      "learning_rate": 3.641442764386266e-07,
      "loss": 2.4532,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 404
    },
    {
      "epoch": 1.157142857142857,
      "grad_norm": 0.3927827286282074,
      "learning_rate": 3.621580983989851e-07,
      "loss": 2.4851,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 405
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.34005483930461033,
      "learning_rate": 3.6017286136167336e-07,
      "loss": 2.5411,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 406
    },
    {
      "epoch": 1.1628571428571428,
      "grad_norm": 0.3353254472552839,
      "learning_rate": 3.5818861469293866e-07,
      "loss": 2.4415,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 407
    },
    {
      "epoch": 1.1657142857142857,
      "grad_norm": 0.3335659225170607,
      "learning_rate": 3.562054077344013e-07,
      "loss": 2.4474,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 408
    },
    {
      "epoch": 1.1685714285714286,
      "grad_norm": 0.33701336028632173,
      "learning_rate": 3.542232898018273e-07,
      "loss": 2.4963,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 409
    },
    {
      "epoch": 1.1714285714285715,
      "grad_norm": 0.3527686112559015,
      "learning_rate": 3.5224231018390234e-07,
      "loss": 2.4164,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 410
    },
    {
      "epoch": 1.1742857142857144,
      "grad_norm": 0.3362320736535409,
      "learning_rate": 3.5026251814100604e-07,
      "loss": 2.4785,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 411
    },
    {
      "epoch": 1.177142857142857,
      "grad_norm": 0.33055501008306276,
      "learning_rate": 3.4828396290398677e-07,
      "loss": 2.4777,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 412
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.34922552471666174,
      "learning_rate": 3.4630669367293793e-07,
      "loss": 2.522,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 413
    },
    {
      "epoch": 1.1828571428571428,
      "grad_norm": 0.3367320180750741,
      "learning_rate": 3.443307596159739e-07,
      "loss": 2.4472,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 414
    },
    {
      "epoch": 1.1857142857142857,
      "grad_norm": 0.3507157389252014,
      "learning_rate": 3.4235620986800804e-07,
      "loss": 2.4304,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 415
    },
    {
      "epoch": 1.1885714285714286,
      "grad_norm": 0.3251151853680131,
      "learning_rate": 3.4038309352953014e-07,
      "loss": 2.5007,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 416
    },
    {
      "epoch": 1.1914285714285715,
      "grad_norm": 0.3446907607816989,
      "learning_rate": 3.384114596653865e-07,
      "loss": 2.4741,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 417
    },
    {
      "epoch": 1.1942857142857144,
      "grad_norm": 0.34743920694914115,
      "learning_rate": 3.364413573035587e-07,
      "loss": 2.4654,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 418
    },
    {
      "epoch": 1.197142857142857,
      "grad_norm": 0.33975327745810197,
      "learning_rate": 3.344728354339449e-07,
      "loss": 2.4647,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 419
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.354959363885649,
      "learning_rate": 3.3250594300714203e-07,
      "loss": 2.4779,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 420
    },
    {
      "epoch": 1.2028571428571428,
      "grad_norm": 0.3752907049307172,
      "learning_rate": 3.3054072893322785e-07,
      "loss": 2.463,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 421
    },
    {
      "epoch": 1.2057142857142857,
      "grad_norm": 0.3425092134767236,
      "learning_rate": 3.2857724208054536e-07,
      "loss": 2.4799,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 422
    },
    {
      "epoch": 1.2085714285714286,
      "grad_norm": 0.3588274090576282,
      "learning_rate": 3.2661553127448705e-07,
      "loss": 2.4862,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 423
    },
    {
      "epoch": 1.2114285714285715,
      "grad_norm": 0.3504117789982341,
      "learning_rate": 3.246556452962813e-07,
      "loss": 2.4142,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 424
    },
    {
      "epoch": 1.2142857142857142,
      "grad_norm": 0.3512702389020557,
      "learning_rate": 3.2269763288177875e-07,
      "loss": 2.4092,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 425
    },
    {
      "epoch": 1.217142857142857,
      "grad_norm": 0.3361544242039997,
      "learning_rate": 3.2074154272024104e-07,
      "loss": 2.3932,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 426
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.33399403939706285,
      "learning_rate": 3.1878742345312957e-07,
      "loss": 2.4882,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 427
    },
    {
      "epoch": 1.2228571428571429,
      "grad_norm": 0.3457039747884681,
      "learning_rate": 3.168353236728963e-07,
      "loss": 2.4226,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 428
    },
    {
      "epoch": 1.2257142857142858,
      "grad_norm": 0.3532537531657652,
      "learning_rate": 3.1488529192177524e-07,
      "loss": 2.4732,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 429
    },
    {
      "epoch": 1.2285714285714286,
      "grad_norm": 0.35045021563166595,
      "learning_rate": 3.129373766905751e-07,
      "loss": 2.4555,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 430
    },
    {
      "epoch": 1.2314285714285713,
      "grad_norm": 0.33520391682990736,
      "learning_rate": 3.1099162641747425e-07,
      "loss": 2.5139,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 431
    },
    {
      "epoch": 1.2342857142857142,
      "grad_norm": 0.35681155455156227,
      "learning_rate": 3.0904808948681544e-07,
      "loss": 2.5038,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 432
    },
    {
      "epoch": 1.237142857142857,
      "grad_norm": 0.3509944640046914,
      "learning_rate": 3.07106814227903e-07,
      "loss": 2.4906,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 433
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.3458934267039996,
      "learning_rate": 3.05167848913801e-07,
      "loss": 2.4671,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 434
    },
    {
      "epoch": 1.2428571428571429,
      "grad_norm": 0.33055940743220014,
      "learning_rate": 3.032312417601329e-07,
      "loss": 2.4087,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 435
    },
    {
      "epoch": 1.2457142857142858,
      "grad_norm": 0.3749997526291379,
      "learning_rate": 3.012970409238825e-07,
      "loss": 2.4837,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 436
    },
    {
      "epoch": 1.2485714285714287,
      "grad_norm": 0.39144911082992606,
      "learning_rate": 2.993652945021966e-07,
      "loss": 2.5109,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 437
    },
    {
      "epoch": 1.2514285714285713,
      "grad_norm": 0.3321496973119456,
      "learning_rate": 2.974360505311886e-07,
      "loss": 2.4476,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 438
    },
    {
      "epoch": 1.2542857142857142,
      "grad_norm": 0.34574723458836265,
      "learning_rate": 2.9550935698474465e-07,
      "loss": 2.4604,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 439
    },
    {
      "epoch": 1.2571428571428571,
      "grad_norm": 0.3523656764858021,
      "learning_rate": 2.9358526177332995e-07,
      "loss": 2.495,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 440
    },
    {
      "epoch": 1.2571428571428571,
      "eval_loss": 2.546485424041748,
      "eval_runtime": 166.4628,
      "eval_samples_per_second": 1.91,
      "eval_steps_per_second": 0.318,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 440
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.3321939537503061,
      "learning_rate": 2.91663812742798e-07,
      "loss": 2.4735,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 441
    },
    {
      "epoch": 1.262857142857143,
      "grad_norm": 0.34338688396251343,
      "learning_rate": 2.897450576732003e-07,
      "loss": 2.3807,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 442
    },
    {
      "epoch": 1.2657142857142858,
      "grad_norm": 0.33978554764711166,
      "learning_rate": 2.8782904427759893e-07,
      "loss": 2.4602,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 443
    },
    {
      "epoch": 1.2685714285714287,
      "grad_norm": 0.3420346371356678,
      "learning_rate": 2.8591582020087914e-07,
      "loss": 2.4696,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 444
    },
    {
      "epoch": 1.2714285714285714,
      "grad_norm": 0.34476523924518143,
      "learning_rate": 2.840054330185655e-07,
      "loss": 2.4325,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 445
    },
    {
      "epoch": 1.2742857142857142,
      "grad_norm": 0.3582287165372729,
      "learning_rate": 2.820979302356383e-07,
      "loss": 2.5059,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 446
    },
    {
      "epoch": 1.2771428571428571,
      "grad_norm": 0.34540242091181167,
      "learning_rate": 2.801933592853523e-07,
      "loss": 2.4805,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 447
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.33052181090145644,
      "learning_rate": 2.782917675280575e-07,
      "loss": 2.4579,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 448
    },
    {
      "epoch": 1.282857142857143,
      "grad_norm": 0.34247949407246303,
      "learning_rate": 2.7639320225002107e-07,
      "loss": 2.472,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 449
    },
    {
      "epoch": 1.2857142857142856,
      "grad_norm": 0.3361093994431006,
      "learning_rate": 2.7449771066225184e-07,
      "loss": 2.4187,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 450
    },
    {
      "epoch": 1.2885714285714287,
      "grad_norm": 0.34538987944656585,
      "learning_rate": 2.7260533989932626e-07,
      "loss": 2.4521,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 451
    },
    {
      "epoch": 1.2914285714285714,
      "grad_norm": 0.3426770292361572,
      "learning_rate": 2.7071613701821607e-07,
      "loss": 2.4656,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 452
    },
    {
      "epoch": 1.2942857142857143,
      "grad_norm": 0.32968657725231604,
      "learning_rate": 2.6883014899711825e-07,
      "loss": 2.4574,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 453
    },
    {
      "epoch": 1.2971428571428572,
      "grad_norm": 0.35044251708152785,
      "learning_rate": 2.6694742273428726e-07,
      "loss": 2.4551,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 454
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.34959363182065906,
      "learning_rate": 2.65068005046868e-07,
      "loss": 2.4565,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 455
    },
    {
      "epoch": 1.302857142857143,
      "grad_norm": 0.3422238545143212,
      "learning_rate": 2.631919426697325e-07,
      "loss": 2.4678,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 456
    },
    {
      "epoch": 1.3057142857142856,
      "grad_norm": 0.37913659988492787,
      "learning_rate": 2.613192822543171e-07,
      "loss": 2.4486,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 457
    },
    {
      "epoch": 1.3085714285714285,
      "grad_norm": 0.3637767051786548,
      "learning_rate": 2.5945007036746285e-07,
      "loss": 2.4226,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 458
    },
    {
      "epoch": 1.3114285714285714,
      "grad_norm": 0.3599015328306578,
      "learning_rate": 2.575843534902573e-07,
      "loss": 2.4585,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 459
    },
    {
      "epoch": 1.3142857142857143,
      "grad_norm": 0.3445640398996027,
      "learning_rate": 2.5572217801687854e-07,
      "loss": 2.4987,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 460
    },
    {
      "epoch": 1.3171428571428572,
      "grad_norm": 0.33762572143996233,
      "learning_rate": 2.5386359025344197e-07,
      "loss": 2.4297,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 461
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.33563054816394783,
      "learning_rate": 2.5200863641684844e-07,
      "loss": 2.4738,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 462
    },
    {
      "epoch": 1.322857142857143,
      "grad_norm": 0.3338562701602384,
      "learning_rate": 2.5015736263363514e-07,
      "loss": 2.5161,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 463
    },
    {
      "epoch": 1.3257142857142856,
      "grad_norm": 0.35645963873295,
      "learning_rate": 2.4830981493882865e-07,
      "loss": 2.4373,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 464
    },
    {
      "epoch": 1.3285714285714285,
      "grad_norm": 0.3471161778657677,
      "learning_rate": 2.4646603927479987e-07,
      "loss": 2.4936,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 465
    },
    {
      "epoch": 1.3314285714285714,
      "grad_norm": 0.39265332133401115,
      "learning_rate": 2.4462608149012213e-07,
      "loss": 2.4768,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 466
    },
    {
      "epoch": 1.3342857142857143,
      "grad_norm": 0.3416863257272438,
      "learning_rate": 2.4278998733843056e-07,
      "loss": 2.4706,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 467
    },
    {
      "epoch": 1.3371428571428572,
      "grad_norm": 0.3347809774717219,
      "learning_rate": 2.409578024772847e-07,
      "loss": 2.5501,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 468
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.35228187875665185,
      "learning_rate": 2.3912957246703304e-07,
      "loss": 2.488,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 469
    },
    {
      "epoch": 1.342857142857143,
      "grad_norm": 0.34462129889305404,
      "learning_rate": 2.3730534276967998e-07,
      "loss": 2.4003,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 470
    },
    {
      "epoch": 1.3457142857142856,
      "grad_norm": 0.33489254911062727,
      "learning_rate": 2.3548515874775546e-07,
      "loss": 2.4259,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 471
    },
    {
      "epoch": 1.3485714285714285,
      "grad_norm": 0.35294976460362765,
      "learning_rate": 2.3366906566318696e-07,
      "loss": 2.443,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 472
    },
    {
      "epoch": 1.3514285714285714,
      "grad_norm": 0.3546250076602948,
      "learning_rate": 2.3185710867617386e-07,
      "loss": 2.4791,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 473
    },
    {
      "epoch": 1.3542857142857143,
      "grad_norm": 0.35593850896876045,
      "learning_rate": 2.3004933284406464e-07,
      "loss": 2.5077,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 474
    },
    {
      "epoch": 1.3571428571428572,
      "grad_norm": 0.3536112472762103,
      "learning_rate": 2.2824578312023627e-07,
      "loss": 2.4725,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 475
    },
    {
      "epoch": 1.3599999999999999,
      "grad_norm": 0.3314781883544676,
      "learning_rate": 2.2644650435297675e-07,
      "loss": 2.4557,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 476
    },
    {
      "epoch": 1.362857142857143,
      "grad_norm": 0.3329765349642462,
      "learning_rate": 2.24651541284369e-07,
      "loss": 2.5021,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 477
    },
    {
      "epoch": 1.3657142857142857,
      "grad_norm": 0.3460759613981604,
      "learning_rate": 2.2286093854917944e-07,
      "loss": 2.4306,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 478
    },
    {
      "epoch": 1.3685714285714285,
      "grad_norm": 0.3764768090910693,
      "learning_rate": 2.210747406737469e-07,
      "loss": 2.5008,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 479
    },
    {
      "epoch": 1.3714285714285714,
      "grad_norm": 0.3537812228756639,
      "learning_rate": 2.1929299207487673e-07,
      "loss": 2.4376,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 480
    },
    {
      "epoch": 1.3742857142857143,
      "grad_norm": 0.36263357692551546,
      "learning_rate": 2.1751573705873477e-07,
      "loss": 2.4912,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 481
    },
    {
      "epoch": 1.3771428571428572,
      "grad_norm": 0.3560473316572667,
      "learning_rate": 2.1574301981974715e-07,
      "loss": 2.4212,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 482
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.36786110701048486,
      "learning_rate": 2.1397488443949984e-07,
      "loss": 2.4563,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 483
    },
    {
      "epoch": 1.3828571428571428,
      "grad_norm": 0.33147890245274236,
      "learning_rate": 2.1221137488564378e-07,
      "loss": 2.4437,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 484
    },
    {
      "epoch": 1.3857142857142857,
      "grad_norm": 0.34355371115176303,
      "learning_rate": 2.1045253501080054e-07,
      "loss": 2.4841,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 485
    },
    {
      "epoch": 1.3885714285714286,
      "grad_norm": 0.3443268121727172,
      "learning_rate": 2.0869840855147282e-07,
      "loss": 2.5203,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 486
    },
    {
      "epoch": 1.3914285714285715,
      "grad_norm": 0.33559733712481915,
      "learning_rate": 2.0694903912695573e-07,
      "loss": 2.4933,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 487
    },
    {
      "epoch": 1.3942857142857144,
      "grad_norm": 0.35275493381686746,
      "learning_rate": 2.0520447023825306e-07,
      "loss": 2.5177,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 488
    },
    {
      "epoch": 1.3971428571428572,
      "grad_norm": 0.3767561332152279,
      "learning_rate": 2.0346474526699548e-07,
      "loss": 2.4418,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 489
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.3727743955756971,
      "learning_rate": 2.0172990747436095e-07,
      "loss": 2.4352,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 490
    },
    {
      "epoch": 1.4028571428571428,
      "grad_norm": 0.3294630205343925,
      "learning_rate": 2.0000000000000007e-07,
      "loss": 2.4745,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 491
    },
    {
      "epoch": 1.4057142857142857,
      "grad_norm": 0.34195274878253873,
      "learning_rate": 1.9827506586096234e-07,
      "loss": 2.457,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 492
    },
    {
      "epoch": 1.4085714285714286,
      "grad_norm": 0.3545046180786858,
      "learning_rate": 1.965551479506274e-07,
      "loss": 2.4404,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 493
    },
    {
      "epoch": 1.4114285714285715,
      "grad_norm": 0.3468571594608192,
      "learning_rate": 1.948402890376376e-07,
      "loss": 2.4052,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 494
    },
    {
      "epoch": 1.4142857142857144,
      "grad_norm": 0.3457291274126277,
      "learning_rate": 1.931305317648349e-07,
      "loss": 2.4569,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 495
    },
    {
      "epoch": 1.4171428571428573,
      "grad_norm": 0.32596382958789394,
      "learning_rate": 1.9142591864820078e-07,
      "loss": 2.4516,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 496
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.3447284238795954,
      "learning_rate": 1.8972649207579807e-07,
      "loss": 2.4542,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 497
    },
    {
      "epoch": 1.4228571428571428,
      "grad_norm": 0.41448386682174737,
      "learning_rate": 1.8803229430671808e-07,
      "loss": 2.4533,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 498
    },
    {
      "epoch": 1.4257142857142857,
      "grad_norm": 0.3436711917548033,
      "learning_rate": 1.863433674700285e-07,
      "loss": 2.4636,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 499
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 0.3538855192926404,
      "learning_rate": 1.8465975356372713e-07,
      "loss": 2.4469,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 500
    },
    {
      "epoch": 1.4314285714285715,
      "grad_norm": 0.3428259931666106,
      "learning_rate": 1.8298149445369627e-07,
      "loss": 2.4932,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 501
    },
    {
      "epoch": 1.4342857142857142,
      "grad_norm": 0.3158299400013455,
      "learning_rate": 1.8130863187266256e-07,
      "loss": 2.4642,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 502
    },
    {
      "epoch": 1.4371428571428573,
      "grad_norm": 0.3435423822044267,
      "learning_rate": 1.7964120741915902e-07,
      "loss": 2.4346,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 503
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.33776650396363245,
      "learning_rate": 1.7797926255649004e-07,
      "loss": 2.4662,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 504
    },
    {
      "epoch": 1.4428571428571428,
      "grad_norm": 0.3533020697371498,
      "learning_rate": 1.763228386117013e-07,
      "loss": 2.4511,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 505
    },
    {
      "epoch": 1.4457142857142857,
      "grad_norm": 0.33503337168391256,
      "learning_rate": 1.7467197677455116e-07,
      "loss": 2.4564,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 506
    },
    {
      "epoch": 1.4485714285714286,
      "grad_norm": 0.3482818087228521,
      "learning_rate": 1.730267180964873e-07,
      "loss": 2.4511,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 507
    },
    {
      "epoch": 1.4514285714285715,
      "grad_norm": 0.33950893100833385,
      "learning_rate": 1.71387103489625e-07,
      "loss": 2.5049,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 508
    },
    {
      "epoch": 1.4542857142857142,
      "grad_norm": 0.3312630848391548,
      "learning_rate": 1.6975317372573065e-07,
      "loss": 2.4756,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 509
    },
    {
      "epoch": 1.457142857142857,
      "grad_norm": 0.3265227265102468,
      "learning_rate": 1.6812496943520711e-07,
      "loss": 2.4495,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 510
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.6796096615724381,
      "learning_rate": 1.6650253110608414e-07,
      "loss": 2.4267,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 511
    },
    {
      "epoch": 1.4628571428571429,
      "grad_norm": 0.3514346729388687,
      "learning_rate": 1.6488589908301078e-07,
      "loss": 2.4328,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 512
    },
    {
      "epoch": 1.4657142857142857,
      "grad_norm": 0.3494024100149988,
      "learning_rate": 1.6327511356625298e-07,
      "loss": 2.4432,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 513
    },
    {
      "epoch": 1.4685714285714286,
      "grad_norm": 0.3485481500048259,
      "learning_rate": 1.6167021461069291e-07,
      "loss": 2.4573,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 514
    },
    {
      "epoch": 1.4714285714285715,
      "grad_norm": 0.3427374828571882,
      "learning_rate": 1.600712421248345e-07,
      "loss": 2.4359,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 515
    },
    {
      "epoch": 1.4742857142857142,
      "grad_norm": 0.3351780182395755,
      "learning_rate": 1.5847823586980895e-07,
      "loss": 2.4051,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 516
    },
    {
      "epoch": 1.477142857142857,
      "grad_norm": 0.34787082390928886,
      "learning_rate": 1.56891235458388e-07,
      "loss": 2.4698,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 517
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.3778661555436346,
      "learning_rate": 1.553102803539974e-07,
      "loss": 2.4545,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 518
    },
    {
      "epoch": 1.4828571428571429,
      "grad_norm": 0.36086432555218123,
      "learning_rate": 1.5373540986973668e-07,
      "loss": 2.4747,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 519
    },
    {
      "epoch": 1.4857142857142858,
      "grad_norm": 0.35826527392310303,
      "learning_rate": 1.521666631674007e-07,
      "loss": 2.425,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 520
    },
    {
      "epoch": 1.4885714285714284,
      "grad_norm": 0.33158918917604324,
      "learning_rate": 1.506040792565066e-07,
      "loss": 2.4932,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 521
    },
    {
      "epoch": 1.4914285714285715,
      "grad_norm": 0.3503624113286146,
      "learning_rate": 1.4904769699332298e-07,
      "loss": 2.498,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 522
    },
    {
      "epoch": 1.4942857142857142,
      "grad_norm": 0.3613906555193044,
      "learning_rate": 1.4749755507990447e-07,
      "loss": 2.3736,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 523
    },
    {
      "epoch": 1.497142857142857,
      "grad_norm": 0.34788577168590956,
      "learning_rate": 1.459536920631285e-07,
      "loss": 2.4519,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 524
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.34120807550278204,
      "learning_rate": 1.4441614633373772e-07,
      "loss": 2.4631,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 525
    },
    {
      "epoch": 1.502857142857143,
      "grad_norm": 0.45587011530768606,
      "learning_rate": 1.4288495612538425e-07,
      "loss": 2.4473,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 526
    },
    {
      "epoch": 1.5057142857142858,
      "grad_norm": 0.3381720357608468,
      "learning_rate": 1.4136015951368018e-07,
      "loss": 2.4703,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 527
    },
    {
      "epoch": 1.5085714285714285,
      "grad_norm": 0.3592158835023677,
      "learning_rate": 1.398417944152495e-07,
      "loss": 2.4983,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 528
    },
    {
      "epoch": 1.5085714285714285,
      "eval_loss": 2.54315185546875,
      "eval_runtime": 165.2145,
      "eval_samples_per_second": 1.925,
      "eval_steps_per_second": 0.321,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 528
    },
    {
      "epoch": 1.5114285714285716,
      "grad_norm": 0.33866858639539815,
      "learning_rate": 1.3832989858678628e-07,
      "loss": 2.4799,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 529
    },
    {
      "epoch": 1.5142857142857142,
      "grad_norm": 0.34198154791452534,
      "learning_rate": 1.3682450962411495e-07,
      "loss": 2.3718,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 530
    },
    {
      "epoch": 1.5171428571428571,
      "grad_norm": 0.33960692793152314,
      "learning_rate": 1.3532566496125632e-07,
      "loss": 2.4604,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 531
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.3440527496348003,
      "learning_rate": 1.3383340186949563e-07,
      "loss": 2.4843,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 532
    },
    {
      "epoch": 1.522857142857143,
      "grad_norm": 0.3487866940189364,
      "learning_rate": 1.3234775745645682e-07,
      "loss": 2.4777,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 533
    },
    {
      "epoch": 1.5257142857142858,
      "grad_norm": 0.35231528530366213,
      "learning_rate": 1.3086876866517874e-07,
      "loss": 2.5022,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 534
    },
    {
      "epoch": 1.5285714285714285,
      "grad_norm": 0.3496058235427625,
      "learning_rate": 1.293964722731979e-07,
      "loss": 2.4603,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 535
    },
    {
      "epoch": 1.5314285714285716,
      "grad_norm": 0.33132002173934666,
      "learning_rate": 1.2793090489163218e-07,
      "loss": 2.4956,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 536
    },
    {
      "epoch": 1.5342857142857143,
      "grad_norm": 0.35003216379209373,
      "learning_rate": 1.2647210296427196e-07,
      "loss": 2.4323,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 537
    },
    {
      "epoch": 1.5371428571428571,
      "grad_norm": 0.35235677145199595,
      "learning_rate": 1.2502010276667272e-07,
      "loss": 2.4329,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 538
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.3443206018717129,
      "learning_rate": 1.2357494040525414e-07,
      "loss": 2.4675,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 539
    },
    {
      "epoch": 1.5428571428571427,
      "grad_norm": 0.34005074014541814,
      "learning_rate": 1.2213665181640104e-07,
      "loss": 2.4766,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 540
    },
    {
      "epoch": 1.5457142857142858,
      "grad_norm": 0.3398976370068044,
      "learning_rate": 1.207052727655709e-07,
      "loss": 2.4666,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 541
    },
    {
      "epoch": 1.5485714285714285,
      "grad_norm": 0.3276315155809813,
      "learning_rate": 1.192808388464034e-07,
      "loss": 2.4412,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 542
    },
    {
      "epoch": 1.5514285714285714,
      "grad_norm": 0.3322318853927801,
      "learning_rate": 1.178633854798365e-07,
      "loss": 2.5029,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 543
    },
    {
      "epoch": 1.5542857142857143,
      "grad_norm": 0.3376674113912205,
      "learning_rate": 1.164529479132244e-07,
      "loss": 2.441,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 544
    },
    {
      "epoch": 1.5571428571428572,
      "grad_norm": 0.3619028417182483,
      "learning_rate": 1.1504956121946214e-07,
      "loss": 2.3838,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 545
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.3598935869345813,
      "learning_rate": 1.1365326029611262e-07,
      "loss": 2.4365,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 546
    },
    {
      "epoch": 1.5628571428571427,
      "grad_norm": 0.3431896201199915,
      "learning_rate": 1.1226407986453961e-07,
      "loss": 2.4743,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 547
    },
    {
      "epoch": 1.5657142857142858,
      "grad_norm": 0.37504529828084726,
      "learning_rate": 1.1088205446904342e-07,
      "loss": 2.4799,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 548
    },
    {
      "epoch": 1.5685714285714285,
      "grad_norm": 0.3299028739276816,
      "learning_rate": 1.0950721847600282e-07,
      "loss": 2.4987,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 549
    },
    {
      "epoch": 1.5714285714285714,
      "grad_norm": 0.34601658476157854,
      "learning_rate": 1.0813960607301967e-07,
      "loss": 2.4755,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 550
    },
    {
      "epoch": 1.5742857142857143,
      "grad_norm": 0.3425736543654905,
      "learning_rate": 1.0677925126806955e-07,
      "loss": 2.5236,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 551
    },
    {
      "epoch": 1.5771428571428572,
      "grad_norm": 0.3580970601232447,
      "learning_rate": 1.0542618788865527e-07,
      "loss": 2.4377,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 552
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.34578459885119206,
      "learning_rate": 1.040804495809665e-07,
      "loss": 2.4874,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 553
    },
    {
      "epoch": 1.5828571428571427,
      "grad_norm": 0.36805552622979093,
      "learning_rate": 1.0274206980904226e-07,
      "loss": 2.4487,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 554
    },
    {
      "epoch": 1.5857142857142859,
      "grad_norm": 0.3385696371208761,
      "learning_rate": 1.0141108185393994e-07,
      "loss": 2.451,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 555
    },
    {
      "epoch": 1.5885714285714285,
      "grad_norm": 0.3410112915103823,
      "learning_rate": 1.0008751881290627e-07,
      "loss": 2.4671,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 556
    },
    {
      "epoch": 1.5914285714285714,
      "grad_norm": 0.3679373473613422,
      "learning_rate": 9.877141359855565e-08,
      "loss": 2.4679,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 557
    },
    {
      "epoch": 1.5942857142857143,
      "grad_norm": 0.3486766166612844,
      "learning_rate": 9.746279893805054e-08,
      "loss": 2.4753,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 558
    },
    {
      "epoch": 1.5971428571428572,
      "grad_norm": 0.3363798438033145,
      "learning_rate": 9.616170737228881e-08,
      "loss": 2.4203,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 559
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.34491030261226907,
      "learning_rate": 9.486817125509339e-08,
      "loss": 2.4564,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 560
    },
    {
      "epoch": 1.6028571428571428,
      "grad_norm": 0.3411179736870957,
      "learning_rate": 9.358222275240884e-08,
      "loss": 2.4308,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 561
    },
    {
      "epoch": 1.6057142857142859,
      "grad_norm": 0.3308270758493747,
      "learning_rate": 9.230389384150052e-08,
      "loss": 2.5318,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 562
    },
    {
      "epoch": 1.6085714285714285,
      "grad_norm": 0.33892946459895457,
      "learning_rate": 9.103321631016023e-08,
      "loss": 2.4965,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 563
    },
    {
      "epoch": 1.6114285714285714,
      "grad_norm": 0.35720218731249365,
      "learning_rate": 8.977022175591508e-08,
      "loss": 2.4754,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 564
    },
    {
      "epoch": 1.6142857142857143,
      "grad_norm": 0.3418917859369647,
      "learning_rate": 8.851494158524243e-08,
      "loss": 2.4592,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 565
    },
    {
      "epoch": 1.617142857142857,
      "grad_norm": 0.3270997851350058,
      "learning_rate": 8.726740701278807e-08,
      "loss": 2.4755,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 566
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.3753352090842295,
      "learning_rate": 8.602764906059108e-08,
      "loss": 2.4722,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 567
    },
    {
      "epoch": 1.6228571428571428,
      "grad_norm": 0.33717529134357965,
      "learning_rate": 8.479569855731123e-08,
      "loss": 2.4364,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 568
    },
    {
      "epoch": 1.6257142857142857,
      "grad_norm": 0.3456096010067795,
      "learning_rate": 8.35715861374636e-08,
      "loss": 2.4875,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 569
    },
    {
      "epoch": 1.6285714285714286,
      "grad_norm": 0.33806299518339794,
      "learning_rate": 8.23553422406556e-08,
      "loss": 2.4876,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 570
    },
    {
      "epoch": 1.6314285714285715,
      "grad_norm": 0.33251020500317413,
      "learning_rate": 8.114699711083113e-08,
      "loss": 2.4732,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 571
    },
    {
      "epoch": 1.6342857142857143,
      "grad_norm": 0.3423176460981214,
      "learning_rate": 7.994658079551752e-08,
      "loss": 2.4558,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 572
    },
    {
      "epoch": 1.637142857142857,
      "grad_norm": 0.3357660970811657,
      "learning_rate": 7.875412314507941e-08,
      "loss": 2.4216,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 573
    },
    {
      "epoch": 1.6400000000000001,
      "grad_norm": 0.34506110835890963,
      "learning_rate": 7.756965381197514e-08,
      "loss": 2.493,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 574
    },
    {
      "epoch": 1.6428571428571428,
      "grad_norm": 0.34006361927838363,
      "learning_rate": 7.639320225002106e-08,
      "loss": 2.3794,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 575
    },
    {
      "epoch": 1.6457142857142857,
      "grad_norm": 0.3518300061897772,
      "learning_rate": 7.52247977136574e-08,
      "loss": 2.4493,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 576
    },
    {
      "epoch": 1.6485714285714286,
      "grad_norm": 0.3489606887135397,
      "learning_rate": 7.40644692572221e-08,
      "loss": 2.4719,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 577
    },
    {
      "epoch": 1.6514285714285715,
      "grad_norm": 0.6140151597686516,
      "learning_rate": 7.291224573422746e-08,
      "loss": 2.4184,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 578
    },
    {
      "epoch": 1.6542857142857144,
      "grad_norm": 0.35428082303827063,
      "learning_rate": 7.176815579664341e-08,
      "loss": 2.4762,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 579
    },
    {
      "epoch": 1.657142857142857,
      "grad_norm": 0.3248512929628147,
      "learning_rate": 7.06322278941843e-08,
      "loss": 2.4042,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 580
    },
    {
      "epoch": 1.6600000000000001,
      "grad_norm": 0.3423595895004747,
      "learning_rate": 6.950449027360213e-08,
      "loss": 2.4749,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 581
    },
    {
      "epoch": 1.6628571428571428,
      "grad_norm": 0.34251702878817736,
      "learning_rate": 6.838497097798335e-08,
      "loss": 2.455,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 582
    },
    {
      "epoch": 1.6657142857142857,
      "grad_norm": 0.33067229035889634,
      "learning_rate": 6.72736978460525e-08,
      "loss": 2.375,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 583
    },
    {
      "epoch": 1.6685714285714286,
      "grad_norm": 0.34822146736584114,
      "learning_rate": 6.617069851147893e-08,
      "loss": 2.4862,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 584
    },
    {
      "epoch": 1.6714285714285713,
      "grad_norm": 0.34117473081678046,
      "learning_rate": 6.507600040219072e-08,
      "loss": 2.5152,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 585
    },
    {
      "epoch": 1.6742857142857144,
      "grad_norm": 0.35842329536939344,
      "learning_rate": 6.398963073969144e-08,
      "loss": 2.4421,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 586
    },
    {
      "epoch": 1.677142857142857,
      "grad_norm": 0.34360806724384274,
      "learning_rate": 6.291161653838433e-08,
      "loss": 2.4335,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 587
    },
    {
      "epoch": 1.6800000000000002,
      "grad_norm": 0.36235009155434456,
      "learning_rate": 6.18419846049e-08,
      "loss": 2.4728,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 588
    },
    {
      "epoch": 1.6828571428571428,
      "grad_norm": 0.34823564669279866,
      "learning_rate": 6.078076153742962e-08,
      "loss": 2.4839,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 589
    },
    {
      "epoch": 1.6857142857142857,
      "grad_norm": 0.3623716471686509,
      "learning_rate": 5.972797372506421e-08,
      "loss": 2.4866,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 590
    },
    {
      "epoch": 1.6885714285714286,
      "grad_norm": 0.39698076411904776,
      "learning_rate": 5.8683647347137754e-08,
      "loss": 2.4701,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 591
    },
    {
      "epoch": 1.6914285714285713,
      "grad_norm": 0.3862977473719727,
      "learning_rate": 5.764780837257666e-08,
      "loss": 2.4566,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 592
    },
    {
      "epoch": 1.6942857142857144,
      "grad_norm": 0.33065248574326134,
      "learning_rate": 5.662048255925356e-08,
      "loss": 2.4687,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 593
    },
    {
      "epoch": 1.697142857142857,
      "grad_norm": 0.3525466540373024,
      "learning_rate": 5.560169545334736e-08,
      "loss": 2.4272,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 594
    },
    {
      "epoch": 1.7,
      "grad_norm": 0.3695933454080633,
      "learning_rate": 5.459147238870767e-08,
      "loss": 2.4302,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 595
    },
    {
      "epoch": 1.7028571428571428,
      "grad_norm": 0.3295134539615001,
      "learning_rate": 5.358983848622451e-08,
      "loss": 2.4588,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 596
    },
    {
      "epoch": 1.7057142857142857,
      "grad_norm": 0.3316179041606098,
      "learning_rate": 5.2596818653204465e-08,
      "loss": 2.4786,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 597
    },
    {
      "epoch": 1.7085714285714286,
      "grad_norm": 0.3364049728077785,
      "learning_rate": 5.161243758275038e-08,
      "loss": 2.4828,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 598
    },
    {
      "epoch": 1.7114285714285713,
      "grad_norm": 0.3480909264892675,
      "learning_rate": 5.063671975314814e-08,
      "loss": 2.4241,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 599
    },
    {
      "epoch": 1.7142857142857144,
      "grad_norm": 0.34828750822116666,
      "learning_rate": 4.966968942725738e-08,
      "loss": 2.4349,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 600
    },
    {
      "epoch": 1.717142857142857,
      "grad_norm": 0.3535474690259664,
      "learning_rate": 4.8711370651908536e-08,
      "loss": 2.4373,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 601
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.3712733582093995,
      "learning_rate": 4.7761787257304796e-08,
      "loss": 2.4714,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 602
    },
    {
      "epoch": 1.7228571428571429,
      "grad_norm": 0.3317448854896828,
      "learning_rate": 4.6820962856429203e-08,
      "loss": 2.4847,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 603
    },
    {
      "epoch": 1.7257142857142858,
      "grad_norm": 0.3554639179356067,
      "learning_rate": 4.588892084445799e-08,
      "loss": 2.4895,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 604
    },
    {
      "epoch": 1.7285714285714286,
      "grad_norm": 0.33678710806563894,
      "learning_rate": 4.496568439817836e-08,
      "loss": 2.5013,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 605
    },
    {
      "epoch": 1.7314285714285713,
      "grad_norm": 0.3547113719729211,
      "learning_rate": 4.405127647541258e-08,
      "loss": 2.4055,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 606
    },
    {
      "epoch": 1.7342857142857144,
      "grad_norm": 0.33855444869836243,
      "learning_rate": 4.314571981444666e-08,
      "loss": 2.5109,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 607
    },
    {
      "epoch": 1.737142857142857,
      "grad_norm": 0.33480434582398616,
      "learning_rate": 4.224903693346533e-08,
      "loss": 2.4026,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 608
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.34343999614717974,
      "learning_rate": 4.1361250129991674e-08,
      "loss": 2.3694,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 609
    },
    {
      "epoch": 1.7428571428571429,
      "grad_norm": 0.3440939194811736,
      "learning_rate": 4.048238148033327e-08,
      "loss": 2.4162,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 610
    },
    {
      "epoch": 1.7457142857142856,
      "grad_norm": 0.3362937944885254,
      "learning_rate": 3.9612452839032386e-08,
      "loss": 2.3872,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 611
    },
    {
      "epoch": 1.7485714285714287,
      "grad_norm": 0.3299530512838386,
      "learning_rate": 3.875148583832337e-08,
      "loss": 2.4917,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 612
    },
    {
      "epoch": 1.7514285714285713,
      "grad_norm": 0.33410972169208125,
      "learning_rate": 3.78995018875941e-08,
      "loss": 2.422,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 613
    },
    {
      "epoch": 1.7542857142857144,
      "grad_norm": 0.3795916016462024,
      "learning_rate": 3.7056522172854045e-08,
      "loss": 2.4494,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 614
    },
    {
      "epoch": 1.7571428571428571,
      "grad_norm": 0.32851385127566174,
      "learning_rate": 3.622256765620713e-08,
      "loss": 2.451,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 615
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.332716966525936,
      "learning_rate": 3.539765907533074e-08,
      "loss": 2.4396,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 616
    },
    {
      "epoch": 1.76,
      "eval_loss": 2.542086362838745,
      "eval_runtime": 166.6377,
      "eval_samples_per_second": 1.908,
      "eval_steps_per_second": 0.318,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 616
    },
    {
      "epoch": 1.762857142857143,
      "grad_norm": 0.3359566302567548,
      "learning_rate": 3.4581816942959606e-08,
      "loss": 2.4689,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 617
    },
    {
      "epoch": 1.7657142857142856,
      "grad_norm": 0.3381348552776297,
      "learning_rate": 3.377506154637637e-08,
      "loss": 2.4233,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 618
    },
    {
      "epoch": 1.7685714285714287,
      "grad_norm": 0.34002272406776,
      "learning_rate": 3.2977412946906434e-08,
      "loss": 2.5185,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 619
    },
    {
      "epoch": 1.7714285714285714,
      "grad_norm": 0.34258151620949384,
      "learning_rate": 3.218889097941977e-08,
      "loss": 2.4475,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 620
    },
    {
      "epoch": 1.7742857142857142,
      "grad_norm": 0.3442996094842595,
      "learning_rate": 3.1409515251836907e-08,
      "loss": 2.4248,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 621
    },
    {
      "epoch": 1.7771428571428571,
      "grad_norm": 0.3419161852567631,
      "learning_rate": 3.063930514464226e-08,
      "loss": 2.4388,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 622
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.3413613680041689,
      "learning_rate": 2.987827981040132e-08,
      "loss": 2.4063,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 623
    },
    {
      "epoch": 1.782857142857143,
      "grad_norm": 0.34257965871856,
      "learning_rate": 2.9126458173285073e-08,
      "loss": 2.4426,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 624
    },
    {
      "epoch": 1.7857142857142856,
      "grad_norm": 0.33234333480481903,
      "learning_rate": 2.8383858928598957e-08,
      "loss": 2.429,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 625
    },
    {
      "epoch": 1.7885714285714287,
      "grad_norm": 0.34995155384015036,
      "learning_rate": 2.7650500542318344e-08,
      "loss": 2.427,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 626
    },
    {
      "epoch": 1.7914285714285714,
      "grad_norm": 0.3488146239094762,
      "learning_rate": 2.692640125062895e-08,
      "loss": 2.4678,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 627
    },
    {
      "epoch": 1.7942857142857143,
      "grad_norm": 0.33155562975311703,
      "learning_rate": 2.621157905947382e-08,
      "loss": 2.4639,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 628
    },
    {
      "epoch": 1.7971428571428572,
      "grad_norm": 0.343308402552018,
      "learning_rate": 2.5506051744105117e-08,
      "loss": 2.4559,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 629
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.3663010303118854,
      "learning_rate": 2.480983684864255e-08,
      "loss": 2.4496,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 630
    },
    {
      "epoch": 1.802857142857143,
      "grad_norm": 0.3496087155407632,
      "learning_rate": 2.4122951685636674e-08,
      "loss": 2.4618,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 631
    },
    {
      "epoch": 1.8057142857142856,
      "grad_norm": 0.3709143157008778,
      "learning_rate": 2.3445413335638763e-08,
      "loss": 2.4336,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 632
    },
    {
      "epoch": 1.8085714285714287,
      "grad_norm": 0.334780955068954,
      "learning_rate": 2.2777238646775766e-08,
      "loss": 2.4661,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 633
    },
    {
      "epoch": 1.8114285714285714,
      "grad_norm": 0.3151809215226177,
      "learning_rate": 2.21184442343318e-08,
      "loss": 2.4896,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 634
    },
    {
      "epoch": 1.8142857142857143,
      "grad_norm": 0.32729671161010176,
      "learning_rate": 2.1469046480334296e-08,
      "loss": 2.4719,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 635
    },
    {
      "epoch": 1.8171428571428572,
      "grad_norm": 0.3336729669396178,
      "learning_rate": 2.082906153314732e-08,
      "loss": 2.4275,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 636
    },
    {
      "epoch": 1.8199999999999998,
      "grad_norm": 0.330415951237473,
      "learning_rate": 2.0198505307069457e-08,
      "loss": 2.4348,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 637
    },
    {
      "epoch": 1.822857142857143,
      "grad_norm": 0.3391893951193496,
      "learning_rate": 1.9577393481938586e-08,
      "loss": 2.4558,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 638
    },
    {
      "epoch": 1.8257142857142856,
      "grad_norm": 0.33196410030214013,
      "learning_rate": 1.8965741502741506e-08,
      "loss": 2.4663,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 639
    },
    {
      "epoch": 1.8285714285714287,
      "grad_norm": 0.3553964914137974,
      "learning_rate": 1.8363564579230294e-08,
      "loss": 2.4312,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 640
    },
    {
      "epoch": 1.8314285714285714,
      "grad_norm": 0.36070852715175145,
      "learning_rate": 1.7770877685543685e-08,
      "loss": 2.4286,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 641
    },
    {
      "epoch": 1.8342857142857143,
      "grad_norm": 0.34450829450639314,
      "learning_rate": 1.718769555983508e-08,
      "loss": 2.4728,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 642
    },
    {
      "epoch": 1.8371428571428572,
      "grad_norm": 0.337811651965076,
      "learning_rate": 1.6614032703905713e-08,
      "loss": 2.4548,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 643
    },
    {
      "epoch": 1.8399999999999999,
      "grad_norm": 0.3379976920167282,
      "learning_rate": 1.604990338284442e-08,
      "loss": 2.442,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 644
    },
    {
      "epoch": 1.842857142857143,
      "grad_norm": 0.3297931279922318,
      "learning_rate": 1.5495321624672443e-08,
      "loss": 2.4945,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 645
    },
    {
      "epoch": 1.8457142857142856,
      "grad_norm": 0.34058231197562855,
      "learning_rate": 1.4950301219995187e-08,
      "loss": 2.4943,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 646
    },
    {
      "epoch": 1.8485714285714285,
      "grad_norm": 0.3387246448133857,
      "learning_rate": 1.4414855721658703e-08,
      "loss": 2.4537,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 647
    },
    {
      "epoch": 1.8514285714285714,
      "grad_norm": 0.3686759506494177,
      "learning_rate": 1.3888998444413091e-08,
      "loss": 2.4667,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 648
    },
    {
      "epoch": 1.8542857142857143,
      "grad_norm": 0.3796237992053229,
      "learning_rate": 1.3372742464581133e-08,
      "loss": 2.4568,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 649
    },
    {
      "epoch": 1.8571428571428572,
      "grad_norm": 0.3391532747183515,
      "learning_rate": 1.2866100619733434e-08,
      "loss": 2.494,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 650
    },
    {
      "epoch": 1.8599999999999999,
      "grad_norm": 0.4002833628577297,
      "learning_rate": 1.236908550836886e-08,
      "loss": 2.5439,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 651
    },
    {
      "epoch": 1.862857142857143,
      "grad_norm": 0.33307535479066336,
      "learning_rate": 1.188170948960141e-08,
      "loss": 2.5131,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 652
    },
    {
      "epoch": 1.8657142857142857,
      "grad_norm": 0.3429562508098538,
      "learning_rate": 1.1403984682852997e-08,
      "loss": 2.4546,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 653
    },
    {
      "epoch": 1.8685714285714285,
      "grad_norm": 0.3489322483707893,
      "learning_rate": 1.0935922967551948e-08,
      "loss": 2.4452,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 654
    },
    {
      "epoch": 1.8714285714285714,
      "grad_norm": 0.3430712066250772,
      "learning_rate": 1.047753598283747e-08,
      "loss": 2.4145,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 655
    },
    {
      "epoch": 1.8742857142857143,
      "grad_norm": 0.3928865435281231,
      "learning_rate": 1.0028835127270552e-08,
      "loss": 2.458,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 656
    },
    {
      "epoch": 1.8771428571428572,
      "grad_norm": 0.3407058992896231,
      "learning_rate": 9.58983155855022e-09,
      "loss": 2.4436,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 657
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.479982896520525,
      "learning_rate": 9.160536193236312e-09,
      "loss": 2.4619,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 658
    },
    {
      "epoch": 1.882857142857143,
      "grad_norm": 0.3580787356797423,
      "learning_rate": 8.740959706477725e-09,
      "loss": 2.458,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 659
    },
    {
      "epoch": 1.8857142857142857,
      "grad_norm": 0.3541061658868137,
      "learning_rate": 8.33111253174743e-09,
      "loss": 2.3754,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 660
    },
    {
      "epoch": 1.8885714285714286,
      "grad_norm": 0.35829791900460084,
      "learning_rate": 7.931004860582468e-09,
      "loss": 2.5012,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 661
    },
    {
      "epoch": 1.8914285714285715,
      "grad_norm": 0.33458331075284653,
      "learning_rate": 7.540646642330984e-09,
      "loss": 2.4107,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 662
    },
    {
      "epoch": 1.8942857142857141,
      "grad_norm": 0.34331138367538366,
      "learning_rate": 7.160047583904472e-09,
      "loss": 2.4403,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 663
    },
    {
      "epoch": 1.8971428571428572,
      "grad_norm": 0.36548742681426033,
      "learning_rate": 6.78921714953673e-09,
      "loss": 2.4743,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 664
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.3497969291535889,
      "learning_rate": 6.428164560548133e-09,
      "loss": 2.4529,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 665
    },
    {
      "epoch": 1.9028571428571428,
      "grad_norm": 0.3355422135767946,
      "learning_rate": 6.076898795116792e-09,
      "loss": 2.4909,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 666
    },
    {
      "epoch": 1.9057142857142857,
      "grad_norm": 0.3232248349935825,
      "learning_rate": 5.7354285880548244e-09,
      "loss": 2.4664,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 667
    },
    {
      "epoch": 1.9085714285714286,
      "grad_norm": 0.33841007572995646,
      "learning_rate": 5.403762430591463e-09,
      "loss": 2.4401,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 668
    },
    {
      "epoch": 1.9114285714285715,
      "grad_norm": 0.3406640187096029,
      "learning_rate": 5.081908570161753e-09,
      "loss": 2.4738,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 669
    },
    {
      "epoch": 1.9142857142857141,
      "grad_norm": 0.33874180232547246,
      "learning_rate": 4.7698750102016075e-09,
      "loss": 2.4551,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 670
    },
    {
      "epoch": 1.9171428571428573,
      "grad_norm": 0.3291228076697962,
      "learning_rate": 4.467669509948591e-09,
      "loss": 2.4846,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 671
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.3626765176554161,
      "learning_rate": 4.175299584249225e-09,
      "loss": 2.5021,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 672
    },
    {
      "epoch": 1.9228571428571428,
      "grad_norm": 0.3347019454342824,
      "learning_rate": 3.892772503371855e-09,
      "loss": 2.4682,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 673
    },
    {
      "epoch": 1.9257142857142857,
      "grad_norm": 0.33761362577097176,
      "learning_rate": 3.62009529282612e-09,
      "loss": 2.4407,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 674
    },
    {
      "epoch": 1.9285714285714286,
      "grad_norm": 0.33760169922994093,
      "learning_rate": 3.357274733187898e-09,
      "loss": 2.4484,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 675
    },
    {
      "epoch": 1.9314285714285715,
      "grad_norm": 0.3476129490779765,
      "learning_rate": 3.1043173599311746e-09,
      "loss": 2.4375,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 676
    },
    {
      "epoch": 1.9342857142857142,
      "grad_norm": 0.34028937592893826,
      "learning_rate": 2.8612294632650578e-09,
      "loss": 2.4387,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 677
    },
    {
      "epoch": 1.9371428571428573,
      "grad_norm": 0.34519206369560324,
      "learning_rate": 2.628017087977774e-09,
      "loss": 2.4953,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 678
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.33291875611950034,
      "learning_rate": 2.4046860332858966e-09,
      "loss": 2.5094,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 679
    },
    {
      "epoch": 1.9428571428571428,
      "grad_norm": 0.32799649676351506,
      "learning_rate": 2.191241852690684e-09,
      "loss": 2.4977,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 680
    },
    {
      "epoch": 1.9457142857142857,
      "grad_norm": 0.33629426510324656,
      "learning_rate": 1.987689853839436e-09,
      "loss": 2.4661,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 681
    },
    {
      "epoch": 1.9485714285714286,
      "grad_norm": 0.34263632647856,
      "learning_rate": 1.7940350983939534e-09,
      "loss": 2.4489,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 682
    },
    {
      "epoch": 1.9514285714285715,
      "grad_norm": 0.3452579801170354,
      "learning_rate": 1.6102824019043725e-09,
      "loss": 2.4044,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 683
    },
    {
      "epoch": 1.9542857142857142,
      "grad_norm": 0.3340176449177557,
      "learning_rate": 1.4364363336896613e-09,
      "loss": 2.4426,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 684
    },
    {
      "epoch": 1.9571428571428573,
      "grad_norm": 0.33283106211620805,
      "learning_rate": 1.2725012167236205e-09,
      "loss": 2.4289,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 685
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.34599321837249125,
      "learning_rate": 1.1184811275279481e-09,
      "loss": 2.4486,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 686
    },
    {
      "epoch": 1.9628571428571429,
      "grad_norm": 0.3561494883021287,
      "learning_rate": 9.74379896070321e-10,
      "loss": 2.4589,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 687
    },
    {
      "epoch": 1.9657142857142857,
      "grad_norm": 0.38387327989422493,
      "learning_rate": 8.402011056694469e-10,
      "loss": 2.3951,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 688
    },
    {
      "epoch": 1.9685714285714284,
      "grad_norm": 0.35645074299253215,
      "learning_rate": 7.159480929059381e-10,
      "loss": 2.3971,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 689
    },
    {
      "epoch": 1.9714285714285715,
      "grad_norm": 0.439617615471271,
      "learning_rate": 6.016239475393093e-10,
      "loss": 2.4295,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 690
    },
    {
      "epoch": 1.9742857142857142,
      "grad_norm": 0.3391147248670222,
      "learning_rate": 4.972315124310622e-10,
      "loss": 2.4092,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 691
    },
    {
      "epoch": 1.977142857142857,
      "grad_norm": 0.34516127593718166,
      "learning_rate": 4.027733834741642e-10,
      "loss": 2.4605,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 692
    },
    {
      "epoch": 1.98,
      "grad_norm": 0.3332109775061957,
      "learning_rate": 3.1825190952829983e-10,
      "loss": 2.4632,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 693
    },
    {
      "epoch": 1.9828571428571429,
      "grad_norm": 0.33664125214104,
      "learning_rate": 2.4366919236169516e-10,
      "loss": 2.4848,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 694
    },
    {
      "epoch": 1.9857142857142858,
      "grad_norm": 0.33614383368258566,
      "learning_rate": 1.7902708659867095e-10,
      "loss": 2.4717,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 695
    },
    {
      "epoch": 1.9885714285714284,
      "grad_norm": 0.341797069919402,
      "learning_rate": 1.243271996735018e-10,
      "loss": 2.5035,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 696
    },
    {
      "epoch": 1.9914285714285715,
      "grad_norm": 0.3583032368307794,
      "learning_rate": 7.95708917905813e-11,
      "loss": 2.4531,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 697
    },
    {
      "epoch": 1.9942857142857142,
      "grad_norm": 0.3489704945574538,
      "learning_rate": 4.475927589062678e-11,
      "loss": 2.4672,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 698
    },
    {
      "epoch": 1.997142857142857,
      "grad_norm": 0.3302867938790412,
      "learning_rate": 1.9893217622790614e-11,
      "loss": 2.4633,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 699
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.37895609396079216,
      "learning_rate": 4.9733353232550565e-12,
      "loss": 2.5314,
      "memory/device_mem_reserved(gib)": 73.92,
      "memory/max_mem_active(gib)": 49.62,
      "memory/max_mem_allocated(gib)": 49.43,
      "step": 700
    }
  ],
  "logging_steps": 1,
  "max_steps": 700,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 175,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.5689826883057222e+19,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}