checkpoint-175/trainer_state.json

{
  "best_metric": 2.602226495742798,
  "best_model_checkpoint": "./Qwen1-5-4B-Chat-hindi-sft/checkpoint-175",
  "epoch": 0.025250703412452204,
  "eval_steps": 25,
  "global_step": 175,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 4.076544761657715,
      "learning_rate": 4.807692307692308e-06,
      "loss": 6.0722,
      "step": 25
    },
    {
      "epoch": 0.0,
      "eval_loss": 6.239687919616699,
      "eval_runtime": 258.5805,
      "eval_samples_per_second": 1.087,
      "eval_steps_per_second": 1.087,
      "step": 25
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.1901891231536865,
      "learning_rate": 9.615384615384616e-06,
      "loss": 6.1105,
      "step": 50
    },
    {
      "epoch": 0.01,
      "eval_loss": 5.878845691680908,
      "eval_runtime": 259.0078,
      "eval_samples_per_second": 1.085,
      "eval_steps_per_second": 1.085,
      "step": 50
    },
    {
      "epoch": 0.01,
      "grad_norm": 7.468161582946777,
      "learning_rate": 1.4423076923076923e-05,
      "loss": 5.1608,
      "step": 75
    },
    {
      "epoch": 0.01,
      "eval_loss": 4.886234283447266,
      "eval_runtime": 259.2966,
      "eval_samples_per_second": 1.084,
      "eval_steps_per_second": 1.084,
      "step": 75
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.9563682079315186,
      "learning_rate": 1.923076923076923e-05,
      "loss": 4.5709,
      "step": 100
    },
    {
      "epoch": 0.01,
      "eval_loss": 3.820255994796753,
      "eval_runtime": 258.4491,
      "eval_samples_per_second": 1.087,
      "eval_steps_per_second": 1.087,
      "step": 100
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.1275033950805664,
      "learning_rate": 2.4038461538461542e-05,
      "loss": 3.7854,
      "step": 125
    },
    {
      "epoch": 0.02,
      "eval_loss": 3.2866950035095215,
      "eval_runtime": 258.8408,
      "eval_samples_per_second": 1.086,
      "eval_steps_per_second": 1.086,
      "step": 125
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.8067069053649902,
      "learning_rate": 2.8846153846153845e-05,
      "loss": 2.8926,
      "step": 150
    },
    {
      "epoch": 0.02,
      "eval_loss": 2.8484082221984863,
      "eval_runtime": 259.0414,
      "eval_samples_per_second": 1.085,
      "eval_steps_per_second": 1.085,
      "step": 150
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.6189056038856506,
      "learning_rate": 3.365384615384616e-05,
      "loss": 2.6005,
      "step": 175
    },
    {
      "epoch": 0.03,
      "eval_loss": 2.602226495742798,
      "eval_runtime": 258.7432,
      "eval_samples_per_second": 1.086,
      "eval_steps_per_second": 1.086,
      "step": 175
    }
  ],
  "logging_steps": 25,
  "max_steps": 34650,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 25,
  "total_flos": 4324033139834880.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}