{
  "best_metric": 0.694903552532196,
  "best_model_checkpoint": "tiny_bert_km_100_v1_rte/checkpoint-30",
  "epoch": 8.0,
  "eval_steps": 500,
  "global_step": 80,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.6791782975196838,
      "learning_rate": 4.9e-05,
      "loss": 0.699,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5270758122743683,
      "eval_loss": 0.6959471106529236,
      "eval_runtime": 0.1358,
      "eval_samples_per_second": 2040.346,
      "eval_steps_per_second": 14.732,
      "step": 10
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.7037239074707031,
      "learning_rate": 4.8e-05,
      "loss": 0.6852,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.4981949458483754,
      "eval_loss": 0.7013058662414551,
      "eval_runtime": 0.1226,
      "eval_samples_per_second": 2259.809,
      "eval_steps_per_second": 16.316,
      "step": 20
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.6154415011405945,
      "learning_rate": 4.7e-05,
      "loss": 0.6713,
      "step": 30
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5270758122743683,
      "eval_loss": 0.694903552532196,
      "eval_runtime": 0.1148,
      "eval_samples_per_second": 2412.236,
      "eval_steps_per_second": 17.417,
      "step": 30
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.636819064617157,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.6531,
      "step": 40
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5451263537906137,
      "eval_loss": 0.7031038403511047,
      "eval_runtime": 0.1179,
      "eval_samples_per_second": 2348.643,
      "eval_steps_per_second": 16.958,
      "step": 40
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.8013365268707275,
      "learning_rate": 4.5e-05,
      "loss": 0.6265,
      "step": 50
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5234657039711191,
      "eval_loss": 0.708032488822937,
      "eval_runtime": 0.1174,
      "eval_samples_per_second": 2359.336,
      "eval_steps_per_second": 17.035,
      "step": 50
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.9887241125106812,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.5887,
      "step": 60
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.516245487364621,
      "eval_loss": 0.7479658126831055,
      "eval_runtime": 0.1163,
      "eval_samples_per_second": 2380.993,
      "eval_steps_per_second": 17.191,
      "step": 60
    },
    {
      "epoch": 7.0,
      "grad_norm": 1.28850519657135,
      "learning_rate": 4.3e-05,
      "loss": 0.5347,
      "step": 70
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.5342960288808665,
      "eval_loss": 0.7832542657852173,
      "eval_runtime": 0.125,
      "eval_samples_per_second": 2216.571,
      "eval_steps_per_second": 16.004,
      "step": 70
    },
    {
      "epoch": 8.0,
      "grad_norm": 1.8739584684371948,
      "learning_rate": 4.2e-05,
      "loss": 0.457,
      "step": 80
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.5306859205776173,
      "eval_loss": 0.8416646718978882,
      "eval_runtime": 0.1169,
      "eval_samples_per_second": 2369.547,
      "eval_steps_per_second": 17.109,
      "step": 80
    },
    {
      "epoch": 8.0,
      "step": 80,
      "total_flos": 522373675991040.0,
      "train_loss": 0.6144436895847321,
      "train_runtime": 21.1487,
      "train_samples_per_second": 5886.889,
      "train_steps_per_second": 23.642
    }
  ],
  "logging_steps": 1,
  "max_steps": 500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 500,
  "total_flos": 522373675991040.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}