{
  "best_metric": 1.05314302444458,
  "best_model_checkpoint": "th_cl_33epochs_lora_pos_neg/checkpoint-32",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 160,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.3125,
      "grad_norm": 8.306446075439453,
      "learning_rate": 9.375e-05,
      "loss": 0.164,
      "step": 10
    },
    {
      "epoch": 0.625,
      "grad_norm": 26.88968849182129,
      "learning_rate": 8.75e-05,
      "loss": 0.3455,
      "step": 20
    },
    {
      "epoch": 0.9375,
      "grad_norm": 10.799789428710938,
      "learning_rate": 8.125000000000001e-05,
      "loss": 0.3999,
      "step": 30
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4888888888888889,
      "eval_balanced_accuracy": 0.4939271255060729,
      "eval_loss": 1.05314302444458,
      "eval_runtime": 88.4283,
      "eval_samples_per_second": 0.509,
      "eval_steps_per_second": 0.068,
      "step": 32
    },
    {
      "epoch": 1.25,
      "grad_norm": 20.668262481689453,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.1668,
      "step": 40
    },
    {
      "epoch": 1.5625,
      "grad_norm": 19.973417282104492,
      "learning_rate": 6.875e-05,
      "loss": 0.1598,
      "step": 50
    },
    {
      "epoch": 1.875,
      "grad_norm": 21.18987464904785,
      "learning_rate": 6.25e-05,
      "loss": 0.2178,
      "step": 60
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.4666666666666667,
      "eval_balanced_accuracy": 0.4722222222222222,
      "eval_loss": 1.058489203453064,
      "eval_runtime": 88.3706,
      "eval_samples_per_second": 0.509,
      "eval_steps_per_second": 0.068,
      "step": 64
    },
    {
      "epoch": 2.1875,
      "grad_norm": 8.074424743652344,
      "learning_rate": 5.6250000000000005e-05,
      "loss": 0.1336,
      "step": 70
    },
    {
      "epoch": 2.5,
      "grad_norm": 7.036731243133545,
      "learning_rate": 5e-05,
      "loss": 0.13,
      "step": 80
    },
    {
      "epoch": 2.8125,
      "grad_norm": 3.6135730743408203,
      "learning_rate": 4.375e-05,
      "loss": 0.0794,
      "step": 90
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5555555555555556,
      "eval_balanced_accuracy": 0.55,
      "eval_loss": 1.0634422302246094,
      "eval_runtime": 88.5007,
      "eval_samples_per_second": 0.508,
      "eval_steps_per_second": 0.068,
      "step": 96
    },
    {
      "epoch": 3.125,
      "grad_norm": 1.4257570505142212,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.0825,
      "step": 100
    },
    {
      "epoch": 3.4375,
      "grad_norm": 23.794193267822266,
      "learning_rate": 3.125e-05,
      "loss": 0.0622,
      "step": 110
    },
    {
      "epoch": 3.75,
      "grad_norm": 7.325411796569824,
      "learning_rate": 2.5e-05,
      "loss": 0.0271,
      "step": 120
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5333333333333333,
      "eval_balanced_accuracy": 0.53,
      "eval_loss": 1.208985686302185,
      "eval_runtime": 88.4567,
      "eval_samples_per_second": 0.509,
      "eval_steps_per_second": 0.068,
      "step": 128
    },
    {
      "epoch": 4.0625,
      "grad_norm": 0.5730175971984863,
      "learning_rate": 1.8750000000000002e-05,
      "loss": 0.0257,
      "step": 130
    },
    {
      "epoch": 4.375,
      "grad_norm": 1.101705551147461,
      "learning_rate": 1.25e-05,
      "loss": 0.0126,
      "step": 140
    },
    {
      "epoch": 4.6875,
      "grad_norm": 2.9783823490142822,
      "learning_rate": 6.25e-06,
      "loss": 0.0166,
      "step": 150
    },
    {
      "epoch": 5.0,
      "grad_norm": 9.750654220581055,
      "learning_rate": 0.0,
      "loss": 0.0412,
      "step": 160
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.4666666666666667,
      "eval_balanced_accuracy": 0.47,
      "eval_loss": 1.2303385734558105,
      "eval_runtime": 88.5783,
      "eval_samples_per_second": 0.508,
      "eval_steps_per_second": 0.068,
      "step": 160
    },
    {
      "epoch": 5.0,
      "step": 160,
      "total_flos": 2.696144222158848e+16,
      "train_loss": 0.12904777154326438,
      "train_runtime": 8156.5384,
      "train_samples_per_second": 0.154,
      "train_steps_per_second": 0.02
    }
  ],
  "logging_steps": 10,
  "max_steps": 160,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 2.696144222158848e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}