trainer_state.json

{
  "best_metric": 2.2016942501068115,
  "best_model_checkpoint": "tiny_bert_km_100_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": 2.6959726810455322,
      "learning_rate": 4.9e-05,
      "loss": 3.7897,
      "step": 23
    },
    {
      "epoch": 1.0,
      "eval_combined_score": 0.18695206741134152,
      "eval_loss": 2.2016942501068115,
      "eval_pearson": 0.19832934084442527,
      "eval_runtime": 0.5834,
      "eval_samples_per_second": 2570.957,
      "eval_spearmanr": 0.17557479397825773,
      "eval_steps_per_second": 10.284,
      "step": 23
    },
    {
      "epoch": 2.0,
      "grad_norm": 8.435463905334473,
      "learning_rate": 4.8e-05,
      "loss": 2.1293,
      "step": 46
    },
    {
      "epoch": 2.0,
      "eval_combined_score": 0.195336433686907,
      "eval_loss": 2.2599353790283203,
      "eval_pearson": 0.20618687628879384,
      "eval_runtime": 0.5695,
      "eval_samples_per_second": 2633.833,
      "eval_spearmanr": 0.18448599108502017,
      "eval_steps_per_second": 10.535,
      "step": 46
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.052290439605713,
      "learning_rate": 4.7e-05,
      "loss": 1.9869,
      "step": 69
    },
    {
      "epoch": 3.0,
      "eval_combined_score": 0.17145564110767253,
      "eval_loss": 2.3073794841766357,
      "eval_pearson": 0.1839128375346377,
      "eval_runtime": 0.6249,
      "eval_samples_per_second": 2400.342,
      "eval_spearmanr": 0.15899844468070734,
      "eval_steps_per_second": 9.601,
      "step": 69
    },
    {
      "epoch": 4.0,
      "grad_norm": 7.113165378570557,
      "learning_rate": 4.600000000000001e-05,
      "loss": 1.8494,
      "step": 92
    },
    {
      "epoch": 4.0,
      "eval_combined_score": 0.20344559502937293,
      "eval_loss": 2.3755438327789307,
      "eval_pearson": 0.21031045738602444,
      "eval_runtime": 0.574,
      "eval_samples_per_second": 2613.391,
      "eval_spearmanr": 0.19658073267272141,
      "eval_steps_per_second": 10.454,
      "step": 92
    },
    {
      "epoch": 5.0,
      "grad_norm": 6.30102014541626,
      "learning_rate": 4.5e-05,
      "loss": 1.62,
      "step": 115
    },
    {
      "epoch": 5.0,
      "eval_combined_score": 0.25598026887111425,
      "eval_loss": 2.345820903778076,
      "eval_pearson": 0.25888621802633294,
      "eval_runtime": 0.6092,
      "eval_samples_per_second": 2462.335,
      "eval_spearmanr": 0.2530743197158956,
      "eval_steps_per_second": 9.849,
      "step": 115
    },
    {
      "epoch": 6.0,
      "grad_norm": 10.555892944335938,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 1.4033,
      "step": 138
    },
    {
      "epoch": 6.0,
      "eval_combined_score": 0.2698149868649785,
      "eval_loss": 2.323427438735962,
      "eval_pearson": 0.2691045935746744,
      "eval_runtime": 0.5851,
      "eval_samples_per_second": 2563.76,
      "eval_spearmanr": 0.27052538015528255,
      "eval_steps_per_second": 10.255,
      "step": 138
    },
    {
      "epoch": 6.0,
      "step": 138,
      "total_flos": 904528923227136.0,
      "train_loss": 2.1297607421875,
      "train_runtime": 32.0911,
      "train_samples_per_second": 8957.301,
      "train_steps_per_second": 35.835
    }
  ],
  "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
}