{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 416,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.024067388688327317,
      "grad_norm": 4.96875,
      "learning_rate": 3.6000000000000003e-06,
      "loss": 2.9962,
      "memory/device_reserved (GiB)": 32.53,
      "memory/max_active (GiB)": 31.83,
      "memory/max_allocated (GiB)": 31.83,
      "step": 10,
      "tokens_per_second_per_gpu": 2526.79
    },
    {
      "epoch": 0.048134777376654635,
      "grad_norm": 4.375,
      "learning_rate": 7.600000000000001e-06,
      "loss": 3.0077,
      "memory/device_reserved (GiB)": 32.54,
      "memory/max_active (GiB)": 31.81,
      "memory/max_allocated (GiB)": 31.81,
      "step": 20,
      "tokens_per_second_per_gpu": 2746.26
    },
    {
      "epoch": 0.07220216606498195,
      "grad_norm": 4.21875,
      "learning_rate": 1.16e-05,
      "loss": 2.8836,
      "memory/device_reserved (GiB)": 32.54,
      "memory/max_active (GiB)": 31.81,
      "memory/max_allocated (GiB)": 31.81,
      "step": 30,
      "tokens_per_second_per_gpu": 2704.02
    },
    {
      "epoch": 0.09626955475330927,
      "grad_norm": 4.15625,
      "learning_rate": 1.5600000000000003e-05,
      "loss": 2.7347,
      "memory/device_reserved (GiB)": 32.54,
      "memory/max_active (GiB)": 31.81,
      "memory/max_allocated (GiB)": 31.81,
      "step": 40,
      "tokens_per_second_per_gpu": 2751.94
    },
    {
      "epoch": 0.12033694344163658,
      "grad_norm": 13.8125,
      "learning_rate": 1.9600000000000002e-05,
      "loss": 2.6617,
      "memory/device_reserved (GiB)": 32.55,
      "memory/max_active (GiB)": 31.8,
      "memory/max_allocated (GiB)": 31.8,
      "step": 50,
      "tokens_per_second_per_gpu": 12.16
    },
    {
      "epoch": 0.1444043321299639,
      "grad_norm": 2.953125,
      "learning_rate": 1.9970175264485268e-05,
      "loss": 2.6474,
      "memory/device_reserved (GiB)": 32.9,
      "memory/max_active (GiB)": 32.02,
      "memory/max_allocated (GiB)": 32.02,
      "step": 60,
      "tokens_per_second_per_gpu": 3420.31
    },
    {
      "epoch": 0.1684717208182912,
      "grad_norm": 3.359375,
      "learning_rate": 1.9867305793119814e-05,
      "loss": 2.5291,
      "memory/device_reserved (GiB)": 32.9,
      "memory/max_active (GiB)": 31.81,
      "memory/max_allocated (GiB)": 31.81,
      "step": 70,
      "tokens_per_second_per_gpu": 2842.63
    },
    {
      "epoch": 0.19253910950661854,
      "grad_norm": 3.53125,
      "learning_rate": 1.9691780654392538e-05,
      "loss": 2.5769,
      "memory/device_reserved (GiB)": 32.9,
      "memory/max_active (GiB)": 31.81,
      "memory/max_allocated (GiB)": 31.81,
      "step": 80,
      "tokens_per_second_per_gpu": 2713.62
    },
    {
      "epoch": 0.21660649819494585,
      "grad_norm": 4.0,
      "learning_rate": 1.9444892287836614e-05,
      "loss": 2.509,
      "memory/device_reserved (GiB)": 32.9,
      "memory/max_active (GiB)": 31.81,
      "memory/max_allocated (GiB)": 31.81,
      "step": 90,
      "tokens_per_second_per_gpu": 2652.2
    },
    {
      "epoch": 0.24067388688327315,
      "grad_norm": 9.6875,
      "learning_rate": 1.9128458599921358e-05,
      "loss": 2.3728,
      "memory/device_reserved (GiB)": 32.9,
      "memory/max_active (GiB)": 31.81,
      "memory/max_allocated (GiB)": 31.81,
      "step": 100,
      "tokens_per_second_per_gpu": 15.95
    },
    {
      "epoch": 0.2647412755716005,
      "grad_norm": 3.25,
      "learning_rate": 1.8744809578310398e-05,
      "loss": 2.5744,
      "memory/device_reserved (GiB)": 32.76,
      "memory/max_active (GiB)": 31.92,
      "memory/max_allocated (GiB)": 31.92,
      "step": 110,
      "tokens_per_second_per_gpu": 3061.35
    },
    {
      "epoch": 0.2888086642599278,
      "grad_norm": 3.203125,
      "learning_rate": 1.829677013552619e-05,
      "loss": 2.4197,
      "memory/device_reserved (GiB)": 32.76,
      "memory/max_active (GiB)": 31.81,
      "memory/max_allocated (GiB)": 31.81,
      "step": 120,
      "tokens_per_second_per_gpu": 2742.16
    },
    {
      "epoch": 0.3128760529482551,
      "grad_norm": 3.6875,
      "learning_rate": 1.778763930834761e-05,
      "loss": 2.3961,
      "memory/device_reserved (GiB)": 32.76,
      "memory/max_active (GiB)": 31.81,
      "memory/max_allocated (GiB)": 31.81,
      "step": 130,
      "tokens_per_second_per_gpu": 2790.74
    },
    {
      "epoch": 0.3369434416365824,
      "grad_norm": 4.15625,
      "learning_rate": 1.7221165966101163e-05,
      "loss": 2.4885,
      "memory/device_reserved (GiB)": 32.76,
      "memory/max_active (GiB)": 31.81,
      "memory/max_allocated (GiB)": 31.81,
      "step": 140,
      "tokens_per_second_per_gpu": 2710.33
    },
    {
      "epoch": 0.36101083032490977,
      "grad_norm": 8.5,
      "learning_rate": 1.660152120671232e-05,
      "loss": 2.3189,
      "memory/device_reserved (GiB)": 32.76,
      "memory/max_active (GiB)": 31.8,
      "memory/max_allocated (GiB)": 31.8,
      "step": 150,
      "tokens_per_second_per_gpu": 13.9
    },
    {
      "epoch": 0.3850782190132371,
      "grad_norm": 3.109375,
      "learning_rate": 1.593326764377232e-05,
      "loss": 2.4858,
      "memory/device_reserved (GiB)": 32.82,
      "memory/max_active (GiB)": 31.96,
      "memory/max_allocated (GiB)": 31.96,
      "step": 160,
      "tokens_per_second_per_gpu": 3034.65
    },
    {
      "epoch": 0.4091456077015644,
      "grad_norm": 3.296875,
      "learning_rate": 1.5221325810768251e-05,
      "loss": 2.3984,
      "memory/device_reserved (GiB)": 32.82,
      "memory/max_active (GiB)": 31.81,
      "memory/max_allocated (GiB)": 31.81,
      "step": 170,
      "tokens_per_second_per_gpu": 2794.98
    },
    {
      "epoch": 0.4332129963898917,
      "grad_norm": 3.765625,
      "learning_rate": 1.4470937929851142e-05,
      "loss": 2.4386,
      "memory/device_reserved (GiB)": 32.82,
      "memory/max_active (GiB)": 31.81,
      "memory/max_allocated (GiB)": 31.81,
      "step": 180,
      "tokens_per_second_per_gpu": 2640.62
    },
    {
      "epoch": 0.457280385078219,
      "grad_norm": 3.9375,
      "learning_rate": 1.3687629311922604e-05,
      "loss": 2.3784,
      "memory/device_reserved (GiB)": 32.82,
      "memory/max_active (GiB)": 31.81,
      "memory/max_allocated (GiB)": 31.81,
      "step": 190,
      "tokens_per_second_per_gpu": 2688.2
    },
    {
      "epoch": 0.4813477737665463,
      "grad_norm": 6.375,
      "learning_rate": 1.287716767226167e-05,
      "loss": 2.254,
      "memory/device_reserved (GiB)": 32.82,
      "memory/max_active (GiB)": 31.8,
      "memory/max_allocated (GiB)": 31.8,
      "step": 200,
      "tokens_per_second_per_gpu": 16.53
    },
    {
      "epoch": 0.5054151624548736,
      "grad_norm": 3.125,
      "learning_rate": 1.2045520661262011e-05,
      "loss": 2.5127,
      "memory/device_reserved (GiB)": 32.94,
      "memory/max_active (GiB)": 31.96,
      "memory/max_allocated (GiB)": 31.96,
      "step": 210,
      "tokens_per_second_per_gpu": 3069.32
    },
    {
      "epoch": 0.529482551143201,
      "grad_norm": 3.328125,
      "learning_rate": 1.1198811922992274e-05,
      "loss": 2.4208,
      "memory/device_reserved (GiB)": 32.94,
      "memory/max_active (GiB)": 31.81,
      "memory/max_allocated (GiB)": 31.81,
      "step": 220,
      "tokens_per_second_per_gpu": 2758.53
    },
    {
      "epoch": 0.5535499398315282,
      "grad_norm": 3.390625,
      "learning_rate": 1.0343276005132436e-05,
      "loss": 2.3724,
      "memory/device_reserved (GiB)": 32.94,
      "memory/max_active (GiB)": 31.81,
      "memory/max_allocated (GiB)": 31.81,
      "step": 230,
      "tokens_per_second_per_gpu": 2799.61
    },
    {
      "epoch": 0.5776173285198556,
      "grad_norm": 3.96875,
      "learning_rate": 9.485212452296535e-06,
      "loss": 2.3658,
      "memory/device_reserved (GiB)": 32.94,
      "memory/max_active (GiB)": 31.81,
      "memory/max_allocated (GiB)": 31.81,
      "step": 240,
      "tokens_per_second_per_gpu": 2669.67
    },
    {
      "epoch": 0.601684717208183,
      "grad_norm": 7.875,
      "learning_rate": 8.630939420765247e-06,
      "loss": 2.2695,
      "memory/device_reserved (GiB)": 32.94,
      "memory/max_active (GiB)": 31.81,
      "memory/max_allocated (GiB)": 31.81,
      "step": 250,
      "tokens_per_second_per_gpu": 16.97
    },
    {
      "epoch": 0.6257521058965102,
      "grad_norm": 3.0625,
      "learning_rate": 7.786747156175675e-06,
      "loss": 2.4657,
      "memory/device_reserved (GiB)": 32.64,
      "memory/max_active (GiB)": 31.99,
      "memory/max_allocated (GiB)": 31.99,
      "step": 260,
      "tokens_per_second_per_gpu": 3091.19
    },
    {
      "epoch": 0.6498194945848376,
      "grad_norm": 3.234375,
      "learning_rate": 6.958851676724823e-06,
      "loss": 2.4131,
      "memory/device_reserved (GiB)": 32.64,
      "memory/max_active (GiB)": 31.81,
      "memory/max_allocated (GiB)": 31.81,
      "step": 270,
      "tokens_per_second_per_gpu": 2802.37
    },
    {
      "epoch": 0.6738868832731648,
      "grad_norm": 3.421875,
      "learning_rate": 6.153349002929988e-06,
      "loss": 2.3716,
      "memory/device_reserved (GiB)": 32.64,
      "memory/max_active (GiB)": 31.81,
      "memory/max_allocated (GiB)": 31.81,
      "step": 280,
      "tokens_per_second_per_gpu": 2645.5
    },
    {
      "epoch": 0.6979542719614922,
      "grad_norm": 4.125,
      "learning_rate": 5.3761702709648555e-06,
      "loss": 2.3805,
      "memory/device_reserved (GiB)": 32.64,
      "memory/max_active (GiB)": 31.81,
      "memory/max_allocated (GiB)": 31.81,
      "step": 290,
      "tokens_per_second_per_gpu": 2761.33
    },
    {
      "epoch": 0.7220216606498195,
      "grad_norm": 11.75,
      "learning_rate": 4.633038060083996e-06,
      "loss": 2.1915,
      "memory/device_reserved (GiB)": 32.65,
      "memory/max_active (GiB)": 31.8,
      "memory/max_allocated (GiB)": 31.8,
      "step": 300,
      "tokens_per_second_per_gpu": 14.43
    },
    {
      "epoch": 0.7460890493381468,
      "grad_norm": 3.078125,
      "learning_rate": 3.929424255708999e-06,
      "loss": 2.4542,
      "memory/device_reserved (GiB)": 33.33,
      "memory/max_active (GiB)": 31.99,
      "memory/max_allocated (GiB)": 31.99,
      "step": 310,
      "tokens_per_second_per_gpu": 3503.84
    },
    {
      "epoch": 0.7701564380264742,
      "grad_norm": 3.28125,
      "learning_rate": 3.2705097584416712e-06,
      "loss": 2.4018,
      "memory/device_reserved (GiB)": 33.33,
      "memory/max_active (GiB)": 31.81,
      "memory/max_allocated (GiB)": 31.81,
      "step": 320,
      "tokens_per_second_per_gpu": 2889.9
    },
    {
      "epoch": 0.7942238267148014,
      "grad_norm": 3.53125,
      "learning_rate": 2.66114633567801e-06,
      "loss": 2.4133,
      "memory/device_reserved (GiB)": 33.33,
      "memory/max_active (GiB)": 31.81,
      "memory/max_allocated (GiB)": 31.81,
      "step": 330,
      "tokens_per_second_per_gpu": 2775.32
    },
    {
      "epoch": 0.8182912154031288,
      "grad_norm": 4.09375,
      "learning_rate": 2.1058208967198046e-06,
      "loss": 2.3389,
      "memory/device_reserved (GiB)": 33.33,
      "memory/max_active (GiB)": 31.81,
      "memory/max_allocated (GiB)": 31.81,
      "step": 340,
      "tokens_per_second_per_gpu": 2850.91
    },
    {
      "epoch": 0.8423586040914561,
      "grad_norm": 12.9375,
      "learning_rate": 1.6086224544360617e-06,
      "loss": 2.2469,
      "memory/device_reserved (GiB)": 33.33,
      "memory/max_active (GiB)": 31.8,
      "memory/max_allocated (GiB)": 31.8,
      "step": 350,
      "tokens_per_second_per_gpu": 13.75
    },
    {
      "epoch": 0.8664259927797834,
      "grad_norm": 3.015625,
      "learning_rate": 1.1732120167445248e-06,
      "loss": 2.4261,
      "memory/device_reserved (GiB)": 32.94,
      "memory/max_active (GiB)": 31.92,
      "memory/max_allocated (GiB)": 31.92,
      "step": 360,
      "tokens_per_second_per_gpu": 3009.01
    },
    {
      "epoch": 0.8904933814681107,
      "grad_norm": 3.28125,
      "learning_rate": 8.027956296105355e-07,
      "loss": 2.411,
      "memory/device_reserved (GiB)": 32.94,
      "memory/max_active (GiB)": 31.81,
      "memory/max_allocated (GiB)": 31.81,
      "step": 370,
      "tokens_per_second_per_gpu": 2749.41
    },
    {
      "epoch": 0.914560770156438,
      "grad_norm": 3.453125,
      "learning_rate": 5.001007700549898e-07,
      "loss": 2.3689,
      "memory/device_reserved (GiB)": 32.94,
      "memory/max_active (GiB)": 31.81,
      "memory/max_allocated (GiB)": 31.81,
      "step": 380,
      "tokens_per_second_per_gpu": 2769.43
    },
    {
      "epoch": 0.9386281588447654,
      "grad_norm": 3.84375,
      "learning_rate": 2.6735626299617456e-07,
      "loss": 2.3496,
      "memory/device_reserved (GiB)": 32.94,
      "memory/max_active (GiB)": 31.81,
      "memory/max_allocated (GiB)": 31.81,
      "step": 390,
      "tokens_per_second_per_gpu": 2799.6
    },
    {
      "epoch": 0.9626955475330926,
      "grad_norm": 10.625,
      "learning_rate": 1.0627586980317073e-07,
      "loss": 2.1982,
      "memory/device_reserved (GiB)": 32.94,
      "memory/max_active (GiB)": 31.8,
      "memory/max_allocated (GiB)": 31.8,
      "step": 400,
      "tokens_per_second_per_gpu": 14.1
    },
    {
      "epoch": 0.98676293622142,
      "grad_norm": 3.75,
      "learning_rate": 1.8045669402859678e-08,
      "loss": 2.473,
      "memory/device_reserved (GiB)": 32.76,
      "memory/max_active (GiB)": 31.86,
      "memory/max_allocated (GiB)": 31.86,
      "step": 410,
      "tokens_per_second_per_gpu": 3564.04
    }
  ],
  "logging_steps": 10,
  "max_steps": 416,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.6853339156578304e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}