trainer_state.json

{
  "best_global_step": 100,
  "best_metric": 0.09561321139335632,
  "best_model_checkpoint": "/content/models/gemma_qlora_lmh/checkpoint-100",
  "epoch": 1.7008547008547008,
  "eval_steps": 20,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 2.458172196149826,
      "epoch": 0.3418803418803419,
      "grad_norm": 6.135525226593018,
      "learning_rate": 8.389830508474577e-06,
      "loss": 0.3827,
      "mean_token_accuracy": 0.8760345175862312,
      "num_tokens": 113164.0,
      "step": 20
    },
    {
      "epoch": 0.3418803418803419,
      "eval_entropy": 2.299590313536489,
      "eval_loss": 0.14134672284126282,
      "eval_mean_token_accuracy": 0.9523653464439588,
      "eval_num_tokens": 113164.0,
      "eval_runtime": 46.8429,
      "eval_samples_per_second": 39.707,
      "eval_steps_per_second": 2.498,
      "step": 20
    },
    {
      "entropy": 2.298478972911835,
      "epoch": 0.6837606837606838,
      "grad_norm": 2.0751421451568604,
      "learning_rate": 6.694915254237288e-06,
      "loss": 0.1357,
      "mean_token_accuracy": 0.9575570523738861,
      "num_tokens": 225335.0,
      "step": 40
    },
    {
      "epoch": 0.6837606837606838,
      "eval_entropy": 2.2715458065016656,
      "eval_loss": 0.11509539932012558,
      "eval_mean_token_accuracy": 0.9629310033260248,
      "eval_num_tokens": 225335.0,
      "eval_runtime": 45.696,
      "eval_samples_per_second": 40.704,
      "eval_steps_per_second": 2.56,
      "step": 40
    },
    {
      "entropy": 2.295832566725902,
      "epoch": 1.017094017094017,
      "grad_norm": 2.188286542892456,
      "learning_rate": 5e-06,
      "loss": 0.113,
      "mean_token_accuracy": 0.9653458717541817,
      "num_tokens": 330390.0,
      "step": 60
    },
    {
      "epoch": 1.017094017094017,
      "eval_entropy": 2.2908162163873005,
      "eval_loss": 0.10838180035352707,
      "eval_mean_token_accuracy": 0.9647057086993487,
      "eval_num_tokens": 330390.0,
      "eval_runtime": 46.2535,
      "eval_samples_per_second": 40.213,
      "eval_steps_per_second": 2.53,
      "step": 60
    },
    {
      "entropy": 2.271016186475754,
      "epoch": 1.358974358974359,
      "grad_norm": 2.2554891109466553,
      "learning_rate": 3.305084745762712e-06,
      "loss": 0.0848,
      "mean_token_accuracy": 0.9718978926539421,
      "num_tokens": 440357.0,
      "step": 80
    },
    {
      "epoch": 1.358974358974359,
      "eval_entropy": 2.254208923405052,
      "eval_loss": 0.10406262427568436,
      "eval_mean_token_accuracy": 0.9654262356269054,
      "eval_num_tokens": 440357.0,
      "eval_runtime": 46.3191,
      "eval_samples_per_second": 40.156,
      "eval_steps_per_second": 2.526,
      "step": 80
    },
    {
      "entropy": 2.2658998131752015,
      "epoch": 1.7008547008547008,
      "grad_norm": 1.6946748495101929,
      "learning_rate": 1.6101694915254237e-06,
      "loss": 0.0716,
      "mean_token_accuracy": 0.9734383270144462,
      "num_tokens": 552807.0,
      "step": 100
    },
    {
      "epoch": 1.7008547008547008,
      "eval_entropy": 2.2408512260159874,
      "eval_loss": 0.09561321139335632,
      "eval_mean_token_accuracy": 0.9683785734013615,
      "eval_num_tokens": 552807.0,
      "eval_runtime": 47.0074,
      "eval_samples_per_second": 39.568,
      "eval_steps_per_second": 2.489,
      "step": 100
    }
  ],
  "logging_steps": 20,
  "max_steps": 118,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "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": 1.614440163723264e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}