{
  "best_metric": 0.8724454641342163,
  "best_model_checkpoint": "./results/checkpoint-2700",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 18900,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.18518518518518517,
      "grad_norm": 2.5104269981384277,
      "learning_rate": 4.94212962962963e-05,
      "loss": 0.9092,
      "step": 500
    },
    {
      "epoch": 0.37037037037037035,
      "grad_norm": 1.030969262123108,
      "learning_rate": 4.8842592592592595e-05,
      "loss": 0.869,
      "step": 1000
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 0.49540138244628906,
      "learning_rate": 4.8263888888888895e-05,
      "loss": 0.8595,
      "step": 1500
    },
    {
      "epoch": 0.7407407407407407,
      "grad_norm": 0.6758684515953064,
      "learning_rate": 4.768518518518519e-05,
      "loss": 0.8446,
      "step": 2000
    },
    {
      "epoch": 0.9259259259259259,
      "grad_norm": 1.0114092826843262,
      "learning_rate": 4.710648148148149e-05,
      "loss": 0.8403,
      "step": 2500
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.8724454641342163,
      "eval_runtime": 30.4341,
      "eval_samples_per_second": 39.43,
      "eval_steps_per_second": 9.857,
      "step": 2700
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 0.8358873128890991,
      "learning_rate": 4.652777777777778e-05,
      "loss": 0.8109,
      "step": 3000
    },
    {
      "epoch": 1.2962962962962963,
      "grad_norm": 0.8434183597564697,
      "learning_rate": 4.594907407407408e-05,
      "loss": 0.7912,
      "step": 3500
    },
    {
      "epoch": 1.4814814814814814,
      "grad_norm": 1.2122215032577515,
      "learning_rate": 4.5370370370370374e-05,
      "loss": 0.7891,
      "step": 4000
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.7747897505760193,
      "learning_rate": 4.4791666666666673e-05,
      "loss": 0.8132,
      "step": 4500
    },
    {
      "epoch": 1.8518518518518519,
      "grad_norm": 0.9554975032806396,
      "learning_rate": 4.4212962962962966e-05,
      "loss": 0.8047,
      "step": 5000
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.8918996453285217,
      "eval_runtime": 30.4698,
      "eval_samples_per_second": 39.383,
      "eval_steps_per_second": 9.846,
      "step": 5400
    },
    {
      "epoch": 2.037037037037037,
      "grad_norm": 0.6646651029586792,
      "learning_rate": 4.3634259259259266e-05,
      "loss": 0.8091,
      "step": 5500
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 4.084255218505859,
      "learning_rate": 4.305555555555556e-05,
      "loss": 0.7432,
      "step": 6000
    },
    {
      "epoch": 2.4074074074074074,
      "grad_norm": 2.1203970909118652,
      "learning_rate": 4.247685185185186e-05,
      "loss": 0.7355,
      "step": 6500
    },
    {
      "epoch": 2.5925925925925926,
      "grad_norm": 4.367093086242676,
      "learning_rate": 4.1898148148148145e-05,
      "loss": 0.751,
      "step": 7000
    },
    {
      "epoch": 2.7777777777777777,
      "grad_norm": 1.1563506126403809,
      "learning_rate": 4.1319444444444445e-05,
      "loss": 0.7632,
      "step": 7500
    },
    {
      "epoch": 2.962962962962963,
      "grad_norm": 0.9595785140991211,
      "learning_rate": 4.074074074074074e-05,
      "loss": 0.7714,
      "step": 8000
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.9504669308662415,
      "eval_runtime": 30.5135,
      "eval_samples_per_second": 39.327,
      "eval_steps_per_second": 9.832,
      "step": 8100
    },
    {
      "epoch": 3.148148148148148,
      "grad_norm": 0.6189069747924805,
      "learning_rate": 4.016203703703704e-05,
      "loss": 0.7277,
      "step": 8500
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.6079156994819641,
      "learning_rate": 3.958333333333333e-05,
      "loss": 0.7373,
      "step": 9000
    },
    {
      "epoch": 3.5185185185185186,
      "grad_norm": 0.4996514320373535,
      "learning_rate": 3.900462962962963e-05,
      "loss": 0.7206,
      "step": 9500
    },
    {
      "epoch": 3.7037037037037037,
      "grad_norm": 0.9442146420478821,
      "learning_rate": 3.8425925925925924e-05,
      "loss": 0.7215,
      "step": 10000
    },
    {
      "epoch": 3.888888888888889,
      "grad_norm": 0.46321621537208557,
      "learning_rate": 3.7847222222222224e-05,
      "loss": 0.7238,
      "step": 10500
    },
    {
      "epoch": 4.0,
      "eval_loss": 1.0164023637771606,
      "eval_runtime": 30.4613,
      "eval_samples_per_second": 39.394,
      "eval_steps_per_second": 9.849,
      "step": 10800
    },
    {
      "epoch": 4.074074074074074,
      "grad_norm": 0.808857798576355,
      "learning_rate": 3.726851851851852e-05,
      "loss": 0.7143,
      "step": 11000
    },
    {
      "epoch": 4.2592592592592595,
      "grad_norm": 0.2266809195280075,
      "learning_rate": 3.6689814814814816e-05,
      "loss": 0.7017,
      "step": 11500
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 0.8129966259002686,
      "learning_rate": 3.611111111111111e-05,
      "loss": 0.7165,
      "step": 12000
    },
    {
      "epoch": 4.62962962962963,
      "grad_norm": 0.7154943943023682,
      "learning_rate": 3.553240740740741e-05,
      "loss": 0.6892,
      "step": 12500
    },
    {
      "epoch": 4.814814814814815,
      "grad_norm": 0.823897659778595,
      "learning_rate": 3.49537037037037e-05,
      "loss": 0.7026,
      "step": 13000
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.7548332810401917,
      "learning_rate": 3.4375e-05,
      "loss": 0.7019,
      "step": 13500
    },
    {
      "epoch": 5.0,
      "eval_loss": 1.056677222251892,
      "eval_runtime": 30.3702,
      "eval_samples_per_second": 39.512,
      "eval_steps_per_second": 9.878,
      "step": 13500
    },
    {
      "epoch": 5.185185185185185,
      "grad_norm": 0.6250707507133484,
      "learning_rate": 3.3796296296296295e-05,
      "loss": 0.7107,
      "step": 14000
    },
    {
      "epoch": 5.37037037037037,
      "grad_norm": 0.7014070749282837,
      "learning_rate": 3.3217592592592595e-05,
      "loss": 0.696,
      "step": 14500
    },
    {
      "epoch": 5.555555555555555,
      "grad_norm": 0.8305183053016663,
      "learning_rate": 3.263888888888889e-05,
      "loss": 0.6858,
      "step": 15000
    },
    {
      "epoch": 5.7407407407407405,
      "grad_norm": 0.5459818840026855,
      "learning_rate": 3.206018518518519e-05,
      "loss": 0.6828,
      "step": 15500
    },
    {
      "epoch": 5.925925925925926,
      "grad_norm": 0.40176087617874146,
      "learning_rate": 3.148148148148148e-05,
      "loss": 0.6841,
      "step": 16000
    },
    {
      "epoch": 6.0,
      "eval_loss": 1.0900229215621948,
      "eval_runtime": 30.4123,
      "eval_samples_per_second": 39.458,
      "eval_steps_per_second": 9.864,
      "step": 16200
    },
    {
      "epoch": 6.111111111111111,
      "grad_norm": 0.4161689281463623,
      "learning_rate": 3.090277777777778e-05,
      "loss": 0.6615,
      "step": 16500
    },
    {
      "epoch": 6.296296296296296,
      "grad_norm": 0.7132428288459778,
      "learning_rate": 3.0324074074074077e-05,
      "loss": 0.683,
      "step": 17000
    },
    {
      "epoch": 6.481481481481482,
      "grad_norm": 0.6823524236679077,
      "learning_rate": 2.9745370370370373e-05,
      "loss": 0.6692,
      "step": 17500
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.3051627278327942,
      "learning_rate": 2.916666666666667e-05,
      "loss": 0.6771,
      "step": 18000
    },
    {
      "epoch": 6.851851851851852,
      "grad_norm": 0.5912793278694153,
      "learning_rate": 2.8587962962962966e-05,
      "loss": 0.6959,
      "step": 18500
    },
    {
      "epoch": 7.0,
      "eval_loss": 1.12686288356781,
      "eval_runtime": 30.4381,
      "eval_samples_per_second": 39.424,
      "eval_steps_per_second": 9.856,
      "step": 18900
    }
  ],
  "logging_steps": 500,
  "max_steps": 43200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 16,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.97536776192e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}