last-checkpoint/trainer_state.json

{
  "best_metric": 0.3573495149612427,
  "best_model_checkpoint": "../KevinKibe/nllb-200-distilled-1.3B-finetuned-finetuned-finetuned-finetuned/checkpoint-400",
  "epoch": 2000.0,
  "eval_steps": 200,
  "global_step": 2000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 200.0,
      "grad_norm": 0.0014342718059197068,
      "learning_rate": 4.878851863841287e-05,
      "loss": 0.0032,
      "step": 200
    },
    {
      "epoch": 200.0,
      "eval_gen_len": 22.0,
      "eval_loss": 0.4362771511077881,
      "eval_rouge": 0.2835,
      "eval_runtime": 1.8779,
      "eval_samples_per_second": 1.065,
      "eval_steps_per_second": 0.533,
      "step": 200
    },
    {
      "epoch": 400.0,
      "grad_norm": 0.0009787885937839746,
      "learning_rate": 4.524848217064997e-05,
      "loss": 0.0003,
      "step": 400
    },
    {
      "epoch": 400.0,
      "eval_gen_len": 22.5,
      "eval_loss": 0.3573495149612427,
      "eval_rouge": 0.2713,
      "eval_runtime": 1.0507,
      "eval_samples_per_second": 1.903,
      "eval_steps_per_second": 0.952,
      "step": 400
    },
    {
      "epoch": 600.0,
      "grad_norm": 0.00037960291956551373,
      "learning_rate": 3.972638318855291e-05,
      "loss": 0.0,
      "step": 600
    },
    {
      "epoch": 600.0,
      "eval_gen_len": 19.5,
      "eval_loss": 0.37564775347709656,
      "eval_rouge": 0.2762,
      "eval_runtime": 0.9,
      "eval_samples_per_second": 2.222,
      "eval_steps_per_second": 1.111,
      "step": 600
    },
    {
      "epoch": 800.0,
      "grad_norm": 0.00027565439813770354,
      "learning_rate": 3.2762763215215306e-05,
      "loss": 0.0,
      "step": 800
    },
    {
      "epoch": 800.0,
      "eval_gen_len": 19.5,
      "eval_loss": 0.3804549276828766,
      "eval_rouge": 0.2762,
      "eval_runtime": 0.8354,
      "eval_samples_per_second": 2.394,
      "eval_steps_per_second": 1.197,
      "step": 800
    },
    {
      "epoch": 1000.0,
      "grad_norm": 0.00021920226572547108,
      "learning_rate": 2.5039269892020772e-05,
      "loss": 0.0,
      "step": 1000
    },
    {
      "epoch": 1000.0,
      "eval_gen_len": 19.5,
      "eval_loss": 0.38089361786842346,
      "eval_rouge": 0.2762,
      "eval_runtime": 0.9583,
      "eval_samples_per_second": 2.087,
      "eval_steps_per_second": 1.044,
      "step": 1000
    },
    {
      "epoch": 1200.0,
      "grad_norm": 0.00018822183483280241,
      "learning_rate": 1.731193255818582e-05,
      "loss": 0.0,
      "step": 1200
    },
    {
      "epoch": 1200.0,
      "eval_gen_len": 19.5,
      "eval_loss": 0.38475799560546875,
      "eval_rouge": 0.2762,
      "eval_runtime": 0.8748,
      "eval_samples_per_second": 2.286,
      "eval_steps_per_second": 1.143,
      "step": 1200
    },
    {
      "epoch": 1400.0,
      "grad_norm": 0.00018423172878101468,
      "learning_rate": 1.0337156831471246e-05,
      "loss": 0.0,
      "step": 1400
    },
    {
      "epoch": 1400.0,
      "eval_gen_len": 22.5,
      "eval_loss": 0.3964395225048065,
      "eval_rouge": 0.2713,
      "eval_runtime": 0.8684,
      "eval_samples_per_second": 2.303,
      "eval_steps_per_second": 1.152,
      "step": 1400
    },
    {
      "epoch": 1600.0,
      "grad_norm": 0.00017020950326696038,
      "learning_rate": 4.7976823561278865e-06,
      "loss": 0.0,
      "step": 1600
    },
    {
      "epoch": 1600.0,
      "eval_gen_len": 22.5,
      "eval_loss": 0.39614760875701904,
      "eval_rouge": 0.2713,
      "eval_runtime": 1.0804,
      "eval_samples_per_second": 1.851,
      "eval_steps_per_second": 0.926,
      "step": 1600
    },
    {
      "epoch": 1800.0,
      "grad_norm": 0.00016995037731248885,
      "learning_rate": 1.2357514895905003e-06,
      "loss": 0.0,
      "step": 1800
    },
    {
      "epoch": 1800.0,
      "eval_gen_len": 22.5,
      "eval_loss": 0.3964083194732666,
      "eval_rouge": 0.2713,
      "eval_runtime": 0.9109,
      "eval_samples_per_second": 2.196,
      "eval_steps_per_second": 1.098,
      "step": 1800
    },
    {
      "epoch": 2000.0,
      "grad_norm": 0.00016411281831096858,
      "learning_rate": 3.0842507411921185e-11,
      "loss": 0.0,
      "step": 2000
    },
    {
      "epoch": 2000.0,
      "eval_gen_len": 22.5,
      "eval_loss": 0.3965863287448883,
      "eval_rouge": 0.2713,
      "eval_runtime": 0.8715,
      "eval_samples_per_second": 2.295,
      "eval_steps_per_second": 1.147,
      "step": 2000
    }
  ],
  "logging_steps": 200,
  "max_steps": 2000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2000,
  "save_steps": 200,
  "total_flos": 5000574664704000.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}