trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 9.51086956521739,
  "global_step": 3500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_bleu": 62.8839,
      "eval_gen_len": 18.6587,
      "eval_loss": 0.6685489416122437,
      "eval_runtime": 219.9201,
      "eval_samples_per_second": 6.689,
      "eval_steps_per_second": 0.418,
      "step": 368
    },
    {
      "epoch": 1.36,
      "learning_rate": 3.456521739130435e-05,
      "loss": 0.9723,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_bleu": 70.5409,
      "eval_gen_len": 18.0659,
      "eval_loss": 0.5173786878585815,
      "eval_runtime": 185.8072,
      "eval_samples_per_second": 7.917,
      "eval_steps_per_second": 0.495,
      "step": 736
    },
    {
      "epoch": 2.72,
      "learning_rate": 2.91304347826087e-05,
      "loss": 0.4138,
      "step": 1000
    },
    {
      "epoch": 3.0,
      "eval_bleu": 70.6822,
      "eval_gen_len": 18.2522,
      "eval_loss": 0.4692835211753845,
      "eval_runtime": 305.3472,
      "eval_samples_per_second": 4.817,
      "eval_steps_per_second": 0.301,
      "step": 1104
    },
    {
      "epoch": 4.0,
      "eval_bleu": 76.1884,
      "eval_gen_len": 18.2624,
      "eval_loss": 0.43825143575668335,
      "eval_runtime": 180.0844,
      "eval_samples_per_second": 8.168,
      "eval_steps_per_second": 0.511,
      "step": 1472
    },
    {
      "epoch": 4.08,
      "learning_rate": 2.3695652173913045e-05,
      "loss": 0.2544,
      "step": 1500
    },
    {
      "epoch": 5.0,
      "eval_bleu": 76.4649,
      "eval_gen_len": 17.9884,
      "eval_loss": 0.42623549699783325,
      "eval_runtime": 265.6449,
      "eval_samples_per_second": 5.537,
      "eval_steps_per_second": 0.346,
      "step": 1840
    },
    {
      "epoch": 5.43,
      "learning_rate": 1.8260869565217393e-05,
      "loss": 0.1709,
      "step": 2000
    },
    {
      "epoch": 6.0,
      "eval_bleu": 77.8959,
      "eval_gen_len": 17.9952,
      "eval_loss": 0.41967156529426575,
      "eval_runtime": 222.1851,
      "eval_samples_per_second": 6.621,
      "eval_steps_per_second": 0.414,
      "step": 2208
    },
    {
      "epoch": 6.79,
      "learning_rate": 1.282608695652174e-05,
      "loss": 0.1287,
      "step": 2500
    },
    {
      "epoch": 7.0,
      "eval_bleu": 78.8909,
      "eval_gen_len": 18.0598,
      "eval_loss": 0.4178922474384308,
      "eval_runtime": 226.0293,
      "eval_samples_per_second": 6.508,
      "eval_steps_per_second": 0.407,
      "step": 2576
    },
    {
      "epoch": 8.0,
      "eval_bleu": 78.7098,
      "eval_gen_len": 18.0034,
      "eval_loss": 0.4161533713340759,
      "eval_runtime": 218.1192,
      "eval_samples_per_second": 6.744,
      "eval_steps_per_second": 0.422,
      "step": 2944
    },
    {
      "epoch": 8.15,
      "learning_rate": 7.391304347826087e-06,
      "loss": 0.1028,
      "step": 3000
    },
    {
      "epoch": 9.0,
      "eval_bleu": 79.2546,
      "eval_gen_len": 18.0659,
      "eval_loss": 0.4158268868923187,
      "eval_runtime": 170.3307,
      "eval_samples_per_second": 8.636,
      "eval_steps_per_second": 0.54,
      "step": 3312
    },
    {
      "epoch": 9.51,
      "learning_rate": 1.956521739130435e-06,
      "loss": 0.084,
      "step": 3500
    }
  ],
  "max_steps": 3680,
  "num_train_epochs": 10,
  "total_flos": 2284115790790656.0,
  "trial_name": null,
  "trial_params": null
}