{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 9.923076923076923,
  "eval_steps": 500,
  "global_step": 39,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.3076923076923077,
      "grad_norm": 429.5760192871094,
      "learning_rate": 5e-06,
      "loss": 1.9642,
      "memory/device_mem_reserved(gib)": 51.52,
      "memory/max_mem_active(gib)": 46.79,
      "memory/max_mem_allocated(gib)": 45.85,
      "step": 1
    },
    {
      "epoch": 0.6153846153846154,
      "grad_norm": 88.243408203125,
      "learning_rate": 1e-05,
      "loss": 1.43,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 2
    },
    {
      "epoch": 0.9230769230769231,
      "grad_norm": 188.16270446777344,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 1.1989,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 3
    },
    {
      "epoch": 1.0,
      "grad_norm": 238.33164978027344,
      "learning_rate": 2e-05,
      "loss": 1.4779,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 4
    },
    {
      "epoch": 1.3076923076923077,
      "grad_norm": 60.53327178955078,
      "learning_rate": 1.9953596287703015e-05,
      "loss": 1.4517,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 5
    },
    {
      "epoch": 1.6153846153846154,
      "grad_norm": 89.0710678100586,
      "learning_rate": 1.9905213270142184e-05,
      "loss": 1.1788,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 6
    },
    {
      "epoch": 1.9230769230769231,
      "grad_norm": 74.55708312988281,
      "learning_rate": 1.9854721549636805e-05,
      "loss": 1.3326,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 7
    },
    {
      "epoch": 2.0,
      "grad_norm": 46.6354866027832,
      "learning_rate": 1.9801980198019806e-05,
      "loss": 1.1162,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 8
    },
    {
      "epoch": 2.3076923076923075,
      "grad_norm": 11.567599296569824,
      "learning_rate": 1.974683544303798e-05,
      "loss": 1.0117,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 9
    },
    {
      "epoch": 2.6153846153846154,
      "grad_norm": 15.884381294250488,
      "learning_rate": 1.9689119170984456e-05,
      "loss": 0.9432,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 10
    },
    {
      "epoch": 2.9230769230769234,
      "grad_norm": 8.479954719543457,
      "learning_rate": 1.9628647214854114e-05,
      "loss": 0.9247,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 11
    },
    {
      "epoch": 3.0,
      "grad_norm": 17.43779754638672,
      "learning_rate": 1.956521739130435e-05,
      "loss": 1.0189,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 12
    },
    {
      "epoch": 3.3076923076923075,
      "grad_norm": 8.543362617492676,
      "learning_rate": 1.9498607242339832e-05,
      "loss": 0.9282,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 13
    },
    {
      "epoch": 3.6153846153846154,
      "grad_norm": 5.215980052947998,
      "learning_rate": 1.942857142857143e-05,
      "loss": 0.8694,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 14
    },
    {
      "epoch": 3.9230769230769234,
      "grad_norm": 4.073164463043213,
      "learning_rate": 1.935483870967742e-05,
      "loss": 0.8029,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 15
    },
    {
      "epoch": 4.0,
      "grad_norm": 5.382778167724609,
      "learning_rate": 1.9277108433734944e-05,
      "loss": 0.7638,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 16
    },
    {
      "epoch": 4.3076923076923075,
      "grad_norm": 4.377191066741943,
      "learning_rate": 1.9195046439628485e-05,
      "loss": 0.7751,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 17
    },
    {
      "epoch": 4.615384615384615,
      "grad_norm": 3.5090882778167725,
      "learning_rate": 1.9108280254777068e-05,
      "loss": 0.731,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 18
    },
    {
      "epoch": 4.923076923076923,
      "grad_norm": 4.811877727508545,
      "learning_rate": 1.9016393442622952e-05,
      "loss": 0.7118,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 19
    },
    {
      "epoch": 5.0,
      "grad_norm": 4.822802543640137,
      "learning_rate": 1.891891891891892e-05,
      "loss": 0.6149,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 20
    },
    {
      "epoch": 5.3076923076923075,
      "grad_norm": 4.164008140563965,
      "learning_rate": 1.8815331010452963e-05,
      "loss": 0.6936,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 21
    },
    {
      "epoch": 5.615384615384615,
      "grad_norm": 3.9381167888641357,
      "learning_rate": 1.8705035971223024e-05,
      "loss": 0.6502,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 22
    },
    {
      "epoch": 5.923076923076923,
      "grad_norm": 3.9260995388031006,
      "learning_rate": 1.858736059479554e-05,
      "loss": 0.633,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 23
    },
    {
      "epoch": 6.0,
      "grad_norm": 5.373617649078369,
      "learning_rate": 1.846153846153846e-05,
      "loss": 0.4986,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 24
    },
    {
      "epoch": 6.3076923076923075,
      "grad_norm": 4.031383514404297,
      "learning_rate": 1.8326693227091633e-05,
      "loss": 0.6118,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 25
    },
    {
      "epoch": 6.615384615384615,
      "grad_norm": 4.3576436042785645,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 0.5786,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 26
    },
    {
      "epoch": 6.923076923076923,
      "grad_norm": 3.389698028564453,
      "learning_rate": 1.8025751072961374e-05,
      "loss": 0.569,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 27
    },
    {
      "epoch": 7.0,
      "grad_norm": 3.289379596710205,
      "learning_rate": 1.785714285714286e-05,
      "loss": 0.3683,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 28
    },
    {
      "epoch": 7.3076923076923075,
      "grad_norm": 3.358076333999634,
      "learning_rate": 1.7674418604651163e-05,
      "loss": 0.5345,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 29
    },
    {
      "epoch": 7.615384615384615,
      "grad_norm": 3.419854164123535,
      "learning_rate": 1.7475728155339808e-05,
      "loss": 0.5051,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 30
    },
    {
      "epoch": 7.923076923076923,
      "grad_norm": 3.624562978744507,
      "learning_rate": 1.7258883248730966e-05,
      "loss": 0.4952,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 31
    },
    {
      "epoch": 8.0,
      "grad_norm": 3.478980541229248,
      "learning_rate": 1.7021276595744686e-05,
      "loss": 0.2747,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 32
    },
    {
      "epoch": 8.307692307692308,
      "grad_norm": 3.6138954162597656,
      "learning_rate": 1.675977653631285e-05,
      "loss": 0.4662,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 33
    },
    {
      "epoch": 8.615384615384615,
      "grad_norm": 3.260951042175293,
      "learning_rate": 1.647058823529412e-05,
      "loss": 0.4351,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 34
    },
    {
      "epoch": 8.923076923076923,
      "grad_norm": 4.216782093048096,
      "learning_rate": 1.6149068322981367e-05,
      "loss": 0.4366,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 35
    },
    {
      "epoch": 9.0,
      "grad_norm": 3.045259475708008,
      "learning_rate": 1.5789473684210526e-05,
      "loss": 0.2067,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 36
    },
    {
      "epoch": 9.307692307692308,
      "grad_norm": 3.721900463104248,
      "learning_rate": 1.5384615384615384e-05,
      "loss": 0.3979,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 37
    },
    {
      "epoch": 9.615384615384615,
      "grad_norm": 4.223248481750488,
      "learning_rate": 1.4925373134328359e-05,
      "loss": 0.3831,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 38
    },
    {
      "epoch": 9.923076923076923,
      "grad_norm": 4.570195198059082,
      "learning_rate": 1.4400000000000001e-05,
      "loss": 0.3919,
      "memory/device_mem_reserved(gib)": 51.53,
      "memory/max_mem_active(gib)": 46.82,
      "memory/max_mem_allocated(gib)": 45.88,
      "step": 39
    }
  ],
  "logging_steps": 1,
  "max_steps": 48,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 12,
  "save_steps": 3,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.482969594438615e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}