Mistral-7B-v0.3_pct_ortho_r32/checkpoint-8/trainer_state.json

{
  "best_metric": 1.967491626739502,
  "best_model_checkpoint": "/home/datta0/models/lora_final/Mistral-7B-v0.3_pct_ortho_r32/checkpoint-8",
  "epoch": 0.020632681845657867,
  "eval_steps": 8,
  "global_step": 8,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0025790852307072333,
      "grad_norm": 7.501750946044922,
      "learning_rate": 1.25e-05,
      "loss": 2.2239,
      "step": 1
    },
    {
      "epoch": 0.010316340922828933,
      "grad_norm": 4.365421295166016,
      "learning_rate": 5e-05,
      "loss": 2.1072,
      "step": 4
    },
    {
      "epoch": 0.020632681845657867,
      "grad_norm": 3.2742128372192383,
      "learning_rate": 0.0001,
      "loss": 1.9673,
      "step": 8
    },
    {
      "epoch": 0.020632681845657867,
      "eval_loss": 1.967491626739502,
      "eval_runtime": 94.9785,
      "eval_samples_per_second": 2.59,
      "eval_steps_per_second": 2.59,
      "step": 8
    }
  ],
  "logging_steps": 4,
  "max_steps": 387,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 8,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 9743832228888576.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}