checkpoint-244/trainer_state.json

{
  "best_metric": 1.7466135025024414,
  "best_model_checkpoint": "outputs/checkpoint-244",
  "epoch": 6.996415770609319,
  "eval_steps": 500,
  "global_step": 244,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.5734767025089605,
      "grad_norm": 0.8502072691917419,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 2.925,
      "step": 20
    },
    {
      "epoch": 0.974910394265233,
      "eval_loss": 2.817981243133545,
      "eval_runtime": 11.5812,
      "eval_samples_per_second": 32.121,
      "eval_steps_per_second": 4.058,
      "step": 34
    },
    {
      "epoch": 1.146953405017921,
      "grad_norm": 0.916117787361145,
      "learning_rate": 4.000000000000001e-06,
      "loss": 2.8937,
      "step": 40
    },
    {
      "epoch": 1.7204301075268817,
      "grad_norm": 1.073569416999817,
      "learning_rate": 6e-06,
      "loss": 2.8277,
      "step": 60
    },
    {
      "epoch": 1.978494623655914,
      "eval_loss": 2.588735342025757,
      "eval_runtime": 11.5877,
      "eval_samples_per_second": 32.103,
      "eval_steps_per_second": 4.056,
      "step": 69
    },
    {
      "epoch": 2.293906810035842,
      "grad_norm": 1.3921942710876465,
      "learning_rate": 8.000000000000001e-06,
      "loss": 2.6548,
      "step": 80
    },
    {
      "epoch": 2.867383512544803,
      "grad_norm": 1.2179991006851196,
      "learning_rate": 1e-05,
      "loss": 2.2871,
      "step": 100
    },
    {
      "epoch": 2.982078853046595,
      "eval_loss": 1.9897719621658325,
      "eval_runtime": 11.5904,
      "eval_samples_per_second": 32.096,
      "eval_steps_per_second": 4.055,
      "step": 104
    },
    {
      "epoch": 3.4408602150537635,
      "grad_norm": 0.5500979423522949,
      "learning_rate": 9.829629131445342e-06,
      "loss": 2.0042,
      "step": 120
    },
    {
      "epoch": 3.985663082437276,
      "eval_loss": 1.8679059743881226,
      "eval_runtime": 11.5852,
      "eval_samples_per_second": 32.11,
      "eval_steps_per_second": 4.057,
      "step": 139
    },
    {
      "epoch": 4.014336917562724,
      "grad_norm": 0.46987271308898926,
      "learning_rate": 9.330127018922195e-06,
      "loss": 1.9362,
      "step": 140
    },
    {
      "epoch": 4.587813620071684,
      "grad_norm": 0.44649583101272583,
      "learning_rate": 8.535533905932739e-06,
      "loss": 1.891,
      "step": 160
    },
    {
      "epoch": 4.989247311827957,
      "eval_loss": 1.8080483675003052,
      "eval_runtime": 11.5897,
      "eval_samples_per_second": 32.097,
      "eval_steps_per_second": 4.055,
      "step": 174
    },
    {
      "epoch": 5.161290322580645,
      "grad_norm": 0.4501400589942932,
      "learning_rate": 7.500000000000001e-06,
      "loss": 1.8352,
      "step": 180
    },
    {
      "epoch": 5.734767025089606,
      "grad_norm": 0.44334864616394043,
      "learning_rate": 6.294095225512604e-06,
      "loss": 1.8157,
      "step": 200
    },
    {
      "epoch": 5.992831541218638,
      "eval_loss": 1.7702627182006836,
      "eval_runtime": 11.5903,
      "eval_samples_per_second": 32.096,
      "eval_steps_per_second": 4.055,
      "step": 209
    },
    {
      "epoch": 6.308243727598566,
      "grad_norm": 0.4709676504135132,
      "learning_rate": 5e-06,
      "loss": 1.769,
      "step": 220
    },
    {
      "epoch": 6.881720430107527,
      "grad_norm": 0.45132553577423096,
      "learning_rate": 3.705904774487396e-06,
      "loss": 1.7697,
      "step": 240
    },
    {
      "epoch": 6.996415770609319,
      "eval_loss": 1.7466135025024414,
      "eval_runtime": 11.5882,
      "eval_samples_per_second": 32.102,
      "eval_steps_per_second": 4.056,
      "step": 244
    }
  ],
  "logging_steps": 20,
  "max_steps": 340,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 1.2904027978850304e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}