{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.08533333333333333,
  "eval_steps": 500,
  "global_step": 48,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0017777777777777779,
      "grad_norm": 2.875,
      "learning_rate": 0.0,
      "loss": 1.09912109375,
      "memory/device_reserved (GiB)": 40.7,
      "memory/max_active (GiB)": 34.07,
      "memory/max_allocated (GiB)": 34.07,
      "ppl": 3.00153,
      "step": 1,
      "tokens/total": 2097152,
      "tokens/train_per_sec_per_gpu": 631.15,
      "tokens/trainable": 1319329
    },
    {
      "epoch": 0.0035555555555555557,
      "grad_norm": 2.890625,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.0859375,
      "memory/device_reserved (GiB)": 47.77,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.96222,
      "step": 2,
      "tokens/total": 4194304,
      "tokens/train_per_sec_per_gpu": 616.0,
      "tokens/trainable": 2628278
    },
    {
      "epoch": 0.005333333333333333,
      "grad_norm": 2.578125,
      "learning_rate": 6.666666666666667e-06,
      "loss": 1.0758056640625,
      "memory/device_reserved (GiB)": 47.77,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.93235,
      "step": 3,
      "tokens/total": 6291456,
      "tokens/train_per_sec_per_gpu": 655.05,
      "tokens/trainable": 3948806
    },
    {
      "epoch": 0.0071111111111111115,
      "grad_norm": 1.71875,
      "learning_rate": 1e-05,
      "loss": 1.0162353515625,
      "memory/device_reserved (GiB)": 47.77,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.76277,
      "step": 4,
      "tokens/total": 8388608,
      "tokens/train_per_sec_per_gpu": 598.59,
      "tokens/trainable": 5229996
    },
    {
      "epoch": 0.008888888888888889,
      "grad_norm": 2.203125,
      "learning_rate": 9.98903822616921e-06,
      "loss": 1.031494140625,
      "memory/device_reserved (GiB)": 47.8,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.80525,
      "step": 5,
      "tokens/total": 10485760,
      "tokens/train_per_sec_per_gpu": 684.59,
      "tokens/trainable": 6541073
    },
    {
      "epoch": 0.010666666666666666,
      "grad_norm": 1.84375,
      "learning_rate": 9.956206309337067e-06,
      "loss": 1.009765625,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.74496,
      "step": 6,
      "tokens/total": 12582912,
      "tokens/train_per_sec_per_gpu": 634.42,
      "tokens/trainable": 7870413
    },
    {
      "epoch": 0.012444444444444444,
      "grad_norm": 1.234375,
      "learning_rate": 9.901664203302126e-06,
      "loss": 0.948486328125,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.5818,
      "step": 7,
      "tokens/total": 14680064,
      "tokens/train_per_sec_per_gpu": 634.71,
      "tokens/trainable": 9189777
    },
    {
      "epoch": 0.014222222222222223,
      "grad_norm": 1.0,
      "learning_rate": 9.825677631722436e-06,
      "loss": 0.9532470703125,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.59412,
      "step": 8,
      "tokens/total": 16777216,
      "tokens/train_per_sec_per_gpu": 619.22,
      "tokens/trainable": 10479490
    },
    {
      "epoch": 0.016,
      "grad_norm": 0.8359375,
      "learning_rate": 9.728616793536588e-06,
      "loss": 0.944091796875,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.57048,
      "step": 9,
      "tokens/total": 18874368,
      "tokens/train_per_sec_per_gpu": 652.49,
      "tokens/trainable": 11784999
    },
    {
      "epoch": 0.017777777777777778,
      "grad_norm": 0.74609375,
      "learning_rate": 9.610954559391704e-06,
      "loss": 0.956298828125,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.60205,
      "step": 10,
      "tokens/total": 20971520,
      "tokens/train_per_sec_per_gpu": 600.6,
      "tokens/trainable": 13094236
    },
    {
      "epoch": 0.019555555555555555,
      "grad_norm": 0.7890625,
      "learning_rate": 9.473264167865172e-06,
      "loss": 0.9410400390625,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.56265,
      "step": 11,
      "tokens/total": 23068672,
      "tokens/train_per_sec_per_gpu": 610.15,
      "tokens/trainable": 14404718
    },
    {
      "epoch": 0.021333333333333333,
      "grad_norm": 0.58203125,
      "learning_rate": 9.316216432703918e-06,
      "loss": 0.8973388671875,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.45307,
      "step": 12,
      "tokens/total": 25165824,
      "tokens/train_per_sec_per_gpu": 667.59,
      "tokens/trainable": 15735257
    },
    {
      "epoch": 0.02311111111111111,
      "grad_norm": 0.6953125,
      "learning_rate": 9.140576474687263e-06,
      "loss": 0.9390869140625,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.55765,
      "step": 13,
      "tokens/total": 27262976,
      "tokens/train_per_sec_per_gpu": 634.35,
      "tokens/trainable": 17038230
    },
    {
      "epoch": 0.024888888888888887,
      "grad_norm": 0.55859375,
      "learning_rate": 8.947199994035402e-06,
      "loss": 0.9205322265625,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.51063,
      "step": 14,
      "tokens/total": 29360128,
      "tokens/train_per_sec_per_gpu": 705.07,
      "tokens/trainable": 18338536
    },
    {
      "epoch": 0.02666666666666667,
      "grad_norm": 0.54296875,
      "learning_rate": 8.737029101523931e-06,
      "loss": 0.9287109375,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.53124,
      "step": 15,
      "tokens/total": 31457280,
      "tokens/train_per_sec_per_gpu": 587.61,
      "tokens/trainable": 19638784
    },
    {
      "epoch": 0.028444444444444446,
      "grad_norm": 0.51171875,
      "learning_rate": 8.511087728614863e-06,
      "loss": 0.9176025390625,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.50328,
      "step": 16,
      "tokens/total": 33554432,
      "tokens/train_per_sec_per_gpu": 625.82,
      "tokens/trainable": 20962100
    },
    {
      "epoch": 0.030222222222222223,
      "grad_norm": 0.6875,
      "learning_rate": 8.270476638965463e-06,
      "loss": 0.9228515625,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.51646,
      "step": 17,
      "tokens/total": 35651584,
      "tokens/train_per_sec_per_gpu": 647.29,
      "tokens/trainable": 22244548
    },
    {
      "epoch": 0.032,
      "grad_norm": 0.47265625,
      "learning_rate": 8.016368065618361e-06,
      "loss": 0.9234619140625,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.51799,
      "step": 18,
      "tokens/total": 37748736,
      "tokens/train_per_sec_per_gpu": 633.36,
      "tokens/trainable": 23550208
    },
    {
      "epoch": 0.033777777777777775,
      "grad_norm": 0.455078125,
      "learning_rate": 7.75e-06,
      "loss": 0.9197998046875,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.50879,
      "step": 19,
      "tokens/total": 39845888,
      "tokens/train_per_sec_per_gpu": 655.15,
      "tokens/trainable": 24856052
    },
    {
      "epoch": 0.035555555555555556,
      "grad_norm": 0.443359375,
      "learning_rate": 7.472670160550849e-06,
      "loss": 0.933349609375,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.54301,
      "step": 20,
      "tokens/total": 41943040,
      "tokens/train_per_sec_per_gpu": 614.04,
      "tokens/trainable": 26140298
    },
    {
      "epoch": 0.037333333333333336,
      "grad_norm": 0.455078125,
      "learning_rate": 7.185729670371605e-06,
      "loss": 0.9356689453125,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.54892,
      "step": 21,
      "tokens/total": 44040192,
      "tokens/train_per_sec_per_gpu": 627.31,
      "tokens/trainable": 27452072
    },
    {
      "epoch": 0.03911111111111111,
      "grad_norm": 0.46484375,
      "learning_rate": 6.890576474687264e-06,
      "loss": 0.937255859375,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.55297,
      "step": 22,
      "tokens/total": 46137344,
      "tokens/train_per_sec_per_gpu": 640.26,
      "tokens/trainable": 28743186
    },
    {
      "epoch": 0.04088888888888889,
      "grad_norm": 0.4140625,
      "learning_rate": 6.588648530198505e-06,
      "loss": 0.900146484375,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.45996,
      "step": 23,
      "tokens/total": 48234496,
      "tokens/train_per_sec_per_gpu": 595.44,
      "tokens/trainable": 30066808
    },
    {
      "epoch": 0.042666666666666665,
      "grad_norm": 0.4921875,
      "learning_rate": 6.281416799501188e-06,
      "loss": 0.9246826171875,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.52107,
      "step": 24,
      "tokens/total": 50331648,
      "tokens/train_per_sec_per_gpu": 595.01,
      "tokens/trainable": 31368300
    },
    {
      "epoch": 0.044444444444444446,
      "grad_norm": 0.3828125,
      "learning_rate": 5.970378084704441e-06,
      "loss": 0.9300537109375,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.53465,
      "step": 25,
      "tokens/total": 52428800,
      "tokens/train_per_sec_per_gpu": 621.98,
      "tokens/trainable": 32679264
    },
    {
      "epoch": 0.04622222222222222,
      "grad_norm": 0.3671875,
      "learning_rate": 5.657047735161256e-06,
      "loss": 0.934326171875,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.5455,
      "step": 26,
      "tokens/total": 54525952,
      "tokens/train_per_sec_per_gpu": 600.67,
      "tokens/trainable": 33988396
    },
    {
      "epoch": 0.048,
      "grad_norm": 0.55859375,
      "learning_rate": 5.342952264838748e-06,
      "loss": 0.9188232421875,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.50634,
      "step": 27,
      "tokens/total": 56623104,
      "tokens/train_per_sec_per_gpu": 633.67,
      "tokens/trainable": 35298496
    },
    {
      "epoch": 0.049777777777777775,
      "grad_norm": 0.416015625,
      "learning_rate": 5.02962191529556e-06,
      "loss": 0.9195556640625,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.50818,
      "step": 28,
      "tokens/total": 58720256,
      "tokens/train_per_sec_per_gpu": 635.47,
      "tokens/trainable": 36597120
    },
    {
      "epoch": 0.051555555555555556,
      "grad_norm": 0.353515625,
      "learning_rate": 4.718583200498814e-06,
      "loss": 0.9129638671875,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.4917,
      "step": 29,
      "tokens/total": 60817408,
      "tokens/train_per_sec_per_gpu": 660.96,
      "tokens/trainable": 37907664
    },
    {
      "epoch": 0.05333333333333334,
      "grad_norm": 0.44140625,
      "learning_rate": 4.4113514698014955e-06,
      "loss": 0.905517578125,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.47321,
      "step": 30,
      "tokens/total": 62889984,
      "tokens/train_per_sec_per_gpu": 611.89,
      "tokens/trainable": 39186144
    },
    {
      "epoch": 0.05511111111111111,
      "grad_norm": 0.38671875,
      "learning_rate": 4.109423525312738e-06,
      "loss": 0.919189453125,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.50726,
      "step": 31,
      "tokens/total": 64987136,
      "tokens/train_per_sec_per_gpu": 646.87,
      "tokens/trainable": 40499084
    },
    {
      "epoch": 0.05688888888888889,
      "grad_norm": 0.369140625,
      "learning_rate": 3.8142703296283954e-06,
      "loss": 0.9398193359375,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.55952,
      "step": 32,
      "tokens/total": 67084288,
      "tokens/train_per_sec_per_gpu": 578.65,
      "tokens/trainable": 41754008
    },
    {
      "epoch": 0.058666666666666666,
      "grad_norm": 0.57421875,
      "learning_rate": 3.527329839449152e-06,
      "loss": 0.919921875,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.50909,
      "step": 33,
      "tokens/total": 69181440,
      "tokens/train_per_sec_per_gpu": 644.77,
      "tokens/trainable": 43059700
    },
    {
      "epoch": 0.060444444444444446,
      "grad_norm": 0.3515625,
      "learning_rate": 3.2500000000000015e-06,
      "loss": 0.88623046875,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.42597,
      "step": 34,
      "tokens/total": 71278592,
      "tokens/train_per_sec_per_gpu": 546.47,
      "tokens/trainable": 44340384
    },
    {
      "epoch": 0.06222222222222222,
      "grad_norm": 0.44140625,
      "learning_rate": 2.98363193438164e-06,
      "loss": 0.918701171875,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.50603,
      "step": 35,
      "tokens/total": 73375744,
      "tokens/train_per_sec_per_gpu": 633.05,
      "tokens/trainable": 45649076
    },
    {
      "epoch": 0.064,
      "grad_norm": 0.4296875,
      "learning_rate": 2.7295233610345384e-06,
      "loss": 0.9283447265625,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.53032,
      "step": 36,
      "tokens/total": 75472896,
      "tokens/train_per_sec_per_gpu": 601.0,
      "tokens/trainable": 46969288
    },
    {
      "epoch": 0.06577777777777778,
      "grad_norm": 0.34765625,
      "learning_rate": 2.4889122713851397e-06,
      "loss": 0.9442138671875,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.57079,
      "step": 37,
      "tokens/total": 77570048,
      "tokens/train_per_sec_per_gpu": 660.13,
      "tokens/trainable": 48301204
    },
    {
      "epoch": 0.06755555555555555,
      "grad_norm": 0.32421875,
      "learning_rate": 2.262970898476071e-06,
      "loss": 0.9041748046875,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.46989,
      "step": 38,
      "tokens/total": 79667200,
      "tokens/train_per_sec_per_gpu": 600.34,
      "tokens/trainable": 49615048
    },
    {
      "epoch": 0.06933333333333333,
      "grad_norm": 0.4609375,
      "learning_rate": 2.0528000059646e-06,
      "loss": 0.8941650390625,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.44529,
      "step": 39,
      "tokens/total": 81764352,
      "tokens/train_per_sec_per_gpu": 643.27,
      "tokens/trainable": 50907048
    },
    {
      "epoch": 0.07111111111111111,
      "grad_norm": 0.34375,
      "learning_rate": 1.8594235253127373e-06,
      "loss": 0.9310302734375,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.53712,
      "step": 40,
      "tokens/total": 83861504,
      "tokens/train_per_sec_per_gpu": 603.96,
      "tokens/trainable": 52206344
    },
    {
      "epoch": 0.07288888888888889,
      "grad_norm": 0.421875,
      "learning_rate": 1.6837835672960834e-06,
      "loss": 0.9283447265625,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.53032,
      "step": 41,
      "tokens/total": 85958656,
      "tokens/train_per_sec_per_gpu": 606.25,
      "tokens/trainable": 53511636
    },
    {
      "epoch": 0.07466666666666667,
      "grad_norm": 0.34765625,
      "learning_rate": 1.526735832134829e-06,
      "loss": 0.9222412109375,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.51492,
      "step": 42,
      "tokens/total": 88055808,
      "tokens/train_per_sec_per_gpu": 640.97,
      "tokens/trainable": 54832244
    },
    {
      "epoch": 0.07644444444444444,
      "grad_norm": 0.345703125,
      "learning_rate": 1.389045440608296e-06,
      "loss": 0.9063720703125,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.47533,
      "step": 43,
      "tokens/total": 90152960,
      "tokens/train_per_sec_per_gpu": 599.17,
      "tokens/trainable": 56163436
    },
    {
      "epoch": 0.07822222222222222,
      "grad_norm": 0.412109375,
      "learning_rate": 1.2713832064634127e-06,
      "loss": 0.94482421875,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.57236,
      "step": 44,
      "tokens/total": 92250112,
      "tokens/train_per_sec_per_gpu": 640.52,
      "tokens/trainable": 57456500
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.353515625,
      "learning_rate": 1.174322368277565e-06,
      "loss": 0.914794921875,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.49626,
      "step": 45,
      "tokens/total": 94347264,
      "tokens/train_per_sec_per_gpu": 616.68,
      "tokens/trainable": 58768064
    },
    {
      "epoch": 0.08177777777777778,
      "grad_norm": 0.388671875,
      "learning_rate": 1.0983357966978747e-06,
      "loss": 0.914306640625,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.49504,
      "step": 46,
      "tokens/total": 96444416,
      "tokens/train_per_sec_per_gpu": 652.42,
      "tokens/trainable": 60042444
    },
    {
      "epoch": 0.08355555555555555,
      "grad_norm": 0.380859375,
      "learning_rate": 1.0437936906629336e-06,
      "loss": 0.9169921875,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.50175,
      "step": 47,
      "tokens/total": 98541568,
      "tokens/train_per_sec_per_gpu": 605.58,
      "tokens/trainable": 61333816
    },
    {
      "epoch": 0.08533333333333333,
      "grad_norm": 0.34375,
      "learning_rate": 1.0109617738307914e-06,
      "loss": 0.9178466796875,
      "memory/device_reserved (GiB)": 47.82,
      "memory/max_active (GiB)": 41.3,
      "memory/max_allocated (GiB)": 41.3,
      "ppl": 2.50389,
      "step": 48,
      "tokens/total": 100638720,
      "tokens/train_per_sec_per_gpu": 616.89,
      "tokens/trainable": 62647352
    }
  ],
  "logging_steps": 1,
  "max_steps": 48,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.1883537665147535e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}