{
  "best_global_step": 140,
  "best_metric": 0.22836367785930634,
  "best_model_checkpoint": "/content/models/gemma_jigsaw_lmh/checkpoint-140",
  "epoch": 2.7450980392156863,
  "eval_steps": 20,
  "global_step": 140,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 2.560735213756561,
      "epoch": 0.39215686274509803,
      "grad_norm": 6.186038494110107,
      "learning_rate": 8.758169934640524e-06,
      "loss": 0.5541,
      "mean_token_accuracy": 0.7453125,
      "num_tokens": 67412.0,
      "step": 20
    },
    {
      "epoch": 0.39215686274509803,
      "eval_entropy": 2.491304580981915,
      "eval_loss": 0.3785918951034546,
      "eval_mean_token_accuracy": 0.7578671345343957,
      "eval_num_tokens": 67412.0,
      "eval_runtime": 2.6932,
      "eval_samples_per_second": 75.376,
      "eval_steps_per_second": 4.827,
      "step": 20
    },
    {
      "entropy": 2.419952464103699,
      "epoch": 0.7843137254901961,
      "grad_norm": 2.4226012229919434,
      "learning_rate": 7.450980392156863e-06,
      "loss": 0.3293,
      "mean_token_accuracy": 0.81640625,
      "num_tokens": 134808.0,
      "step": 40
    },
    {
      "epoch": 0.7843137254901961,
      "eval_entropy": 2.4435020043299747,
      "eval_loss": 0.2946617007255554,
      "eval_mean_token_accuracy": 0.8513986009817857,
      "eval_num_tokens": 134808.0,
      "eval_runtime": 2.6642,
      "eval_samples_per_second": 76.195,
      "eval_steps_per_second": 4.879,
      "step": 40
    },
    {
      "entropy": 2.423317462205887,
      "epoch": 1.1764705882352942,
      "grad_norm": 10.875091552734375,
      "learning_rate": 6.143790849673204e-06,
      "loss": 0.2871,
      "mean_token_accuracy": 0.8487723216414451,
      "num_tokens": 201046.0,
      "step": 60
    },
    {
      "epoch": 1.1764705882352942,
      "eval_entropy": 2.448130937723013,
      "eval_loss": 0.2594100534915924,
      "eval_mean_token_accuracy": 0.8621066441902747,
      "eval_num_tokens": 201046.0,
      "eval_runtime": 2.6633,
      "eval_samples_per_second": 76.221,
      "eval_steps_per_second": 4.881,
      "step": 60
    },
    {
      "entropy": 2.3511528968811035,
      "epoch": 1.5686274509803921,
      "grad_norm": 5.017323970794678,
      "learning_rate": 4.836601307189543e-06,
      "loss": 0.2503,
      "mean_token_accuracy": 0.87109375,
      "num_tokens": 269334.0,
      "step": 80
    },
    {
      "epoch": 1.5686274509803921,
      "eval_entropy": 2.3352334682758036,
      "eval_loss": 0.2594275176525116,
      "eval_mean_token_accuracy": 0.8824300720141485,
      "eval_num_tokens": 269334.0,
      "eval_runtime": 2.6613,
      "eval_samples_per_second": 76.278,
      "eval_steps_per_second": 4.885,
      "step": 80
    },
    {
      "entropy": 2.3079891920089723,
      "epoch": 1.9607843137254903,
      "grad_norm": 7.2991743087768555,
      "learning_rate": 3.529411764705883e-06,
      "loss": 0.247,
      "mean_token_accuracy": 0.87890625,
      "num_tokens": 337424.0,
      "step": 100
    },
    {
      "epoch": 1.9607843137254903,
      "eval_entropy": 2.3540270145122824,
      "eval_loss": 0.23124322295188904,
      "eval_mean_token_accuracy": 0.8861451057287363,
      "eval_num_tokens": 337424.0,
      "eval_runtime": 2.7035,
      "eval_samples_per_second": 75.087,
      "eval_steps_per_second": 4.809,
      "step": 100
    },
    {
      "entropy": 2.3495707869529725,
      "epoch": 2.3529411764705883,
      "grad_norm": 5.988176345825195,
      "learning_rate": 2.222222222222222e-06,
      "loss": 0.2009,
      "mean_token_accuracy": 0.9171875,
      "num_tokens": 404955.0,
      "step": 120
    },
    {
      "epoch": 2.3529411764705883,
      "eval_entropy": 2.4085900966937723,
      "eval_loss": 0.23444519937038422,
      "eval_mean_token_accuracy": 0.8957604903441209,
      "eval_num_tokens": 404955.0,
      "eval_runtime": 2.6365,
      "eval_samples_per_second": 76.995,
      "eval_steps_per_second": 4.931,
      "step": 120
    },
    {
      "entropy": 2.3811947822570803,
      "epoch": 2.7450980392156863,
      "grad_norm": 9.767471313476562,
      "learning_rate": 9.150326797385621e-07,
      "loss": 0.1973,
      "mean_token_accuracy": 0.91328125,
      "num_tokens": 472246.0,
      "step": 140
    },
    {
      "epoch": 2.7450980392156863,
      "eval_entropy": 2.4008009983943057,
      "eval_loss": 0.22836367785930634,
      "eval_mean_token_accuracy": 0.9040646873987638,
      "eval_num_tokens": 472246.0,
      "eval_runtime": 2.645,
      "eval_samples_per_second": 76.748,
      "eval_steps_per_second": 4.915,
      "step": 140
    }
  ],
  "logging_steps": 20,
  "max_steps": 153,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 20,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.0222031863807488e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}