trainer_state.json

{
  "best_metric": 2.227576732635498,
  "best_model_checkpoint": "tiny_bert_km_50_v1_stsb/checkpoint-23",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 138,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 4.95888090133667,
      "learning_rate": 4.9e-05,
      "loss": 3.8515,
      "step": 23
    },
    {
      "epoch": 1.0,
      "eval_combined_score": 0.18734264198444261,
      "eval_loss": 2.227576732635498,
      "eval_pearson": 0.1923505309545669,
      "eval_runtime": 0.5509,
      "eval_samples_per_second": 2722.84,
      "eval_spearmanr": 0.18233475301431837,
      "eval_steps_per_second": 10.891,
      "step": 23
    },
    {
      "epoch": 2.0,
      "grad_norm": 8.362377166748047,
      "learning_rate": 4.8e-05,
      "loss": 2.1315,
      "step": 46
    },
    {
      "epoch": 2.0,
      "eval_combined_score": 0.2089354886210551,
      "eval_loss": 2.2513928413391113,
      "eval_pearson": 0.21651479454360775,
      "eval_runtime": 0.5991,
      "eval_samples_per_second": 2503.728,
      "eval_spearmanr": 0.20135618269850244,
      "eval_steps_per_second": 10.015,
      "step": 46
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.1979072093963623,
      "learning_rate": 4.7e-05,
      "loss": 1.9864,
      "step": 69
    },
    {
      "epoch": 3.0,
      "eval_combined_score": 0.19251859064923532,
      "eval_loss": 2.2710776329040527,
      "eval_pearson": 0.1966087155320278,
      "eval_runtime": 0.6964,
      "eval_samples_per_second": 2154.04,
      "eval_spearmanr": 0.18842846576644287,
      "eval_steps_per_second": 8.616,
      "step": 69
    },
    {
      "epoch": 4.0,
      "grad_norm": 4.702291488647461,
      "learning_rate": 4.600000000000001e-05,
      "loss": 1.8498,
      "step": 92
    },
    {
      "epoch": 4.0,
      "eval_combined_score": 0.21904044995109873,
      "eval_loss": 2.3629276752471924,
      "eval_pearson": 0.22169527715203102,
      "eval_runtime": 0.5734,
      "eval_samples_per_second": 2616.115,
      "eval_spearmanr": 0.2163856227501664,
      "eval_steps_per_second": 10.464,
      "step": 92
    },
    {
      "epoch": 5.0,
      "grad_norm": 6.9626946449279785,
      "learning_rate": 4.5e-05,
      "loss": 1.6182,
      "step": 115
    },
    {
      "epoch": 5.0,
      "eval_combined_score": 0.24536057902618041,
      "eval_loss": 2.371917724609375,
      "eval_pearson": 0.24478048351146486,
      "eval_runtime": 0.6115,
      "eval_samples_per_second": 2453.097,
      "eval_spearmanr": 0.24594067454089594,
      "eval_steps_per_second": 9.812,
      "step": 115
    },
    {
      "epoch": 6.0,
      "grad_norm": 7.2307891845703125,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 1.3745,
      "step": 138
    },
    {
      "epoch": 6.0,
      "eval_combined_score": 0.24094916196629926,
      "eval_loss": 2.4488790035247803,
      "eval_pearson": 0.2402419127861277,
      "eval_runtime": 0.5723,
      "eval_samples_per_second": 2621.166,
      "eval_spearmanr": 0.24165641114647082,
      "eval_steps_per_second": 10.485,
      "step": 138
    },
    {
      "epoch": 6.0,
      "step": 138,
      "total_flos": 904528923227136.0,
      "train_loss": 2.1353097998577617,
      "train_runtime": 31.7448,
      "train_samples_per_second": 9055.015,
      "train_steps_per_second": 36.226
    }
  ],
  "logging_steps": 1,
  "max_steps": 1150,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 500,
  "total_flos": 904528923227136.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}