trainer_state.json

{
  "best_global_step": 100,
  "best_metric": 0.09543673694133759,
  "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": 2.4621351063251495,
      "epoch": 0.3418803418803419,
      "grad_norm": 6.045567035675049,
      "learning_rate": 8.389830508474577e-06,
      "loss": 0.3829,
      "mean_token_accuracy": 0.8746399849653244,
      "num_tokens": 113164.0,
      "step": 20
    },
    {
      "epoch": 0.3418803418803419,
      "eval_entropy": 2.310356734145401,
      "eval_loss": 0.14163336157798767,
      "eval_mean_token_accuracy": 0.9525125485200149,
      "eval_num_tokens": 113164.0,
      "eval_runtime": 46.7154,
      "eval_samples_per_second": 39.816,
      "eval_steps_per_second": 2.505,
      "step": 20
    },
    {
      "entropy": 2.3071940779685973,
      "epoch": 0.6837606837606838,
      "grad_norm": 2.0642659664154053,
      "learning_rate": 6.694915254237288e-06,
      "loss": 0.1358,
      "mean_token_accuracy": 0.9574040338397026,
      "num_tokens": 225335.0,
      "step": 40
    },
    {
      "epoch": 0.6837606837606838,
      "eval_entropy": 2.279157841307485,
      "eval_loss": 0.11512647569179535,
      "eval_mean_token_accuracy": 0.9622599701596122,
      "eval_num_tokens": 225335.0,
      "eval_runtime": 45.5615,
      "eval_samples_per_second": 40.824,
      "eval_steps_per_second": 2.568,
      "step": 40
    },
    {
      "entropy": 2.3005140133393116,
      "epoch": 1.017094017094017,
      "grad_norm": 2.1764779090881348,
      "learning_rate": 5e-06,
      "loss": 0.113,
      "mean_token_accuracy": 0.9657797859265254,
      "num_tokens": 330390.0,
      "step": 60
    },
    {
      "epoch": 1.017094017094017,
      "eval_entropy": 2.293168124989567,
      "eval_loss": 0.10884715616703033,
      "eval_mean_token_accuracy": 0.9645043224351019,
      "eval_num_tokens": 330390.0,
      "eval_runtime": 46.0807,
      "eval_samples_per_second": 40.364,
      "eval_steps_per_second": 2.539,
      "step": 60
    },
    {
      "entropy": 2.27779643535614,
      "epoch": 1.358974358974359,
      "grad_norm": 2.2281429767608643,
      "learning_rate": 3.305084745762712e-06,
      "loss": 0.0849,
      "mean_token_accuracy": 0.9723166093230248,
      "num_tokens": 440357.0,
      "step": 80
    },
    {
      "epoch": 1.358974358974359,
      "eval_entropy": 2.2607989514994826,
      "eval_loss": 0.10449662059545517,
      "eval_mean_token_accuracy": 0.9651819151690882,
      "eval_num_tokens": 440357.0,
      "eval_runtime": 46.3813,
      "eval_samples_per_second": 40.102,
      "eval_steps_per_second": 2.523,
      "step": 80
    },
    {
      "entropy": 2.2726868212223055,
      "epoch": 1.7008547008547008,
      "grad_norm": 1.7332485914230347,
      "learning_rate": 1.6101694915254237e-06,
      "loss": 0.0719,
      "mean_token_accuracy": 0.9728455126285553,
      "num_tokens": 552807.0,
      "step": 100
    },
    {
      "epoch": 1.7008547008547008,
      "eval_entropy": 2.2467452420128717,
      "eval_loss": 0.09543673694133759,
      "eval_mean_token_accuracy": 0.9682549915762029,
      "eval_num_tokens": 552807.0,
      "eval_runtime": 46.2095,
      "eval_samples_per_second": 40.251,
      "eval_steps_per_second": 2.532,
      "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
}