checkpoint-111/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 100,
  "global_step": 111,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 1.418435550481081,
      "epoch": 0.27303754266211605,
      "grad_norm": 0.056396484375,
      "learning_rate": 0.00019959742939952392,
      "loss": 1.6161483764648437,
      "mean_token_accuracy": 0.6638647213578224,
      "num_tokens": 151209.0,
      "step": 10
    },
    {
      "entropy": 1.097722884826362,
      "epoch": 0.5460750853242321,
      "grad_norm": 0.033203125,
      "learning_rate": 0.00019253043004739968,
      "loss": 1.1661317825317383,
      "mean_token_accuracy": 0.7458787495270371,
      "num_tokens": 320046.0,
      "step": 20
    },
    {
      "entropy": 1.063631435856223,
      "epoch": 0.8191126279863481,
      "grad_norm": 0.0247802734375,
      "learning_rate": 0.00017724169592245995,
      "loss": 1.118991756439209,
      "mean_token_accuracy": 0.7535936305299401,
      "num_tokens": 492438.0,
      "step": 30
    },
    {
      "entropy": 0.9953024551852957,
      "epoch": 1.0819112627986347,
      "grad_norm": 0.0255126953125,
      "learning_rate": 0.00015508969814521025,
      "loss": 1.0578977584838867,
      "mean_token_accuracy": 0.7685656903626082,
      "num_tokens": 642939.0,
      "step": 40
    },
    {
      "entropy": 0.9422929083928466,
      "epoch": 1.3549488054607508,
      "grad_norm": 0.025390625,
      "learning_rate": 0.00012804273893060028,
      "loss": 1.0265376091003418,
      "mean_token_accuracy": 0.7775206623598934,
      "num_tokens": 817432.0,
      "step": 50
    },
    {
      "entropy": 0.9282504981383681,
      "epoch": 1.627986348122867,
      "grad_norm": 0.025390625,
      "learning_rate": 9.850405929847366e-05,
      "loss": 0.9812464714050293,
      "mean_token_accuracy": 0.7824548855423927,
      "num_tokens": 971956.0,
      "step": 60
    },
    {
      "entropy": 0.9165813697502017,
      "epoch": 1.901023890784983,
      "grad_norm": 0.028076171875,
      "learning_rate": 6.909830056250527e-05,
      "loss": 0.9443504333496093,
      "mean_token_accuracy": 0.7839378425851464,
      "num_tokens": 1134082.0,
      "step": 70
    },
    {
      "entropy": 0.946953180354911,
      "epoch": 2.1638225255972694,
      "grad_norm": 0.02880859375,
      "learning_rate": 4.2438293431432665e-05,
      "loss": 0.968726921081543,
      "mean_token_accuracy": 0.7822197033213331,
      "num_tokens": 1290177.0,
      "step": 80
    },
    {
      "entropy": 0.8450189002789557,
      "epoch": 2.4368600682593855,
      "grad_norm": 0.031982421875,
      "learning_rate": 2.0892896534365904e-05,
      "loss": 0.9016420364379882,
      "mean_token_accuracy": 0.7995872467756271,
      "num_tokens": 1461672.0,
      "step": 90
    },
    {
      "entropy": 0.8309150729328394,
      "epoch": 2.7098976109215016,
      "grad_norm": 0.037353515625,
      "learning_rate": 6.37651293602628e-06,
      "loss": 0.8841998100280761,
      "mean_token_accuracy": 0.8054205430671573,
      "num_tokens": 1624257.0,
      "step": 100
    },
    {
      "epoch": 2.7098976109215016,
      "eval_entropy": 0.9007232843926458,
      "eval_loss": 0.9238418340682983,
      "eval_mean_token_accuracy": 0.7843361054406022,
      "eval_num_tokens": 1624257.0,
      "eval_runtime": 42.5945,
      "eval_samples_per_second": 1.549,
      "eval_steps_per_second": 1.549,
      "step": 100
    },
    {
      "entropy": 0.8297194179147482,
      "epoch": 2.9829351535836177,
      "grad_norm": 0.0322265625,
      "learning_rate": 1.7898702322648453e-07,
      "loss": 0.8666117668151856,
      "mean_token_accuracy": 0.8052121920511126,
      "num_tokens": 1778979.0,
      "step": 110
    }
  ],
  "logging_steps": 10,
  "max_steps": 111,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7.754556700156723e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}