{
  "best_metric": 1.6402901411056519,
  "best_model_checkpoint": "../../experiments_checkpoints/MAdAiLab/Qwen/Qwen1.5_1.8B_scotus/checkpoint-300",
  "epoch": 1.910828025477707,
  "eval_steps": 50,
  "global_step": 300,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.06,
      "grad_norm": 206.6490020751953,
      "learning_rate": 4.893842887473461e-06,
      "loss": 5.8906,
      "step": 10
    },
    {
      "epoch": 0.13,
      "grad_norm": 142.15260314941406,
      "learning_rate": 4.787685774946922e-06,
      "loss": 4.4914,
      "step": 20
    },
    {
      "epoch": 0.19,
      "grad_norm": 47.5128173828125,
      "learning_rate": 4.6815286624203824e-06,
      "loss": 3.1742,
      "step": 30
    },
    {
      "epoch": 0.25,
      "grad_norm": 90.23841094970703,
      "learning_rate": 4.575371549893844e-06,
      "loss": 2.3301,
      "step": 40
    },
    {
      "epoch": 0.32,
      "grad_norm": 106.74323272705078,
      "learning_rate": 4.469214437367304e-06,
      "loss": 2.1746,
      "step": 50
    },
    {
      "epoch": 0.32,
      "eval_accuracy": 0.2692857142857143,
      "eval_f1_macro": 0.11482650807382964,
      "eval_f1_micro": 0.2692857142857143,
      "eval_loss": 2.2791965007781982,
      "eval_runtime": 3.6276,
      "eval_samples_per_second": 385.929,
      "eval_steps_per_second": 12.129,
      "step": 50
    },
    {
      "epoch": 0.38,
      "grad_norm": 87.4228515625,
      "learning_rate": 4.3630573248407645e-06,
      "loss": 1.9594,
      "step": 60
    },
    {
      "epoch": 0.45,
      "grad_norm": 61.52133560180664,
      "learning_rate": 4.256900212314226e-06,
      "loss": 1.8109,
      "step": 70
    },
    {
      "epoch": 0.51,
      "grad_norm": 55.21061706542969,
      "learning_rate": 4.150743099787686e-06,
      "loss": 1.7418,
      "step": 80
    },
    {
      "epoch": 0.57,
      "grad_norm": 71.5464096069336,
      "learning_rate": 4.044585987261147e-06,
      "loss": 1.6996,
      "step": 90
    },
    {
      "epoch": 0.64,
      "grad_norm": 50.51081848144531,
      "learning_rate": 3.938428874734608e-06,
      "loss": 1.5758,
      "step": 100
    },
    {
      "epoch": 0.64,
      "eval_accuracy": 0.41,
      "eval_f1_macro": 0.22906680413417876,
      "eval_f1_micro": 0.41,
      "eval_loss": 1.930892825126648,
      "eval_runtime": 3.6465,
      "eval_samples_per_second": 383.927,
      "eval_steps_per_second": 12.066,
      "step": 100
    },
    {
      "epoch": 0.7,
      "grad_norm": 108.21426391601562,
      "learning_rate": 3.832271762208068e-06,
      "loss": 1.6871,
      "step": 110
    },
    {
      "epoch": 0.76,
      "grad_norm": 62.29310607910156,
      "learning_rate": 3.7261146496815285e-06,
      "loss": 1.6055,
      "step": 120
    },
    {
      "epoch": 0.83,
      "grad_norm": 79.33343505859375,
      "learning_rate": 3.6199575371549893e-06,
      "loss": 1.6043,
      "step": 130
    },
    {
      "epoch": 0.89,
      "grad_norm": 77.69229888916016,
      "learning_rate": 3.51380042462845e-06,
      "loss": 1.5793,
      "step": 140
    },
    {
      "epoch": 0.96,
      "grad_norm": 67.4636001586914,
      "learning_rate": 3.407643312101911e-06,
      "loss": 1.7312,
      "step": 150
    },
    {
      "epoch": 0.96,
      "eval_accuracy": 0.4714285714285714,
      "eval_f1_macro": 0.2546511715546434,
      "eval_f1_micro": 0.4714285714285714,
      "eval_loss": 1.7475446462631226,
      "eval_runtime": 3.6415,
      "eval_samples_per_second": 384.454,
      "eval_steps_per_second": 12.083,
      "step": 150
    },
    {
      "epoch": 1.02,
      "grad_norm": 72.03321838378906,
      "learning_rate": 3.3014861995753718e-06,
      "loss": 1.5695,
      "step": 160
    },
    {
      "epoch": 1.08,
      "grad_norm": 59.265869140625,
      "learning_rate": 3.195329087048832e-06,
      "loss": 1.2879,
      "step": 170
    },
    {
      "epoch": 1.15,
      "grad_norm": 78.44482421875,
      "learning_rate": 3.089171974522293e-06,
      "loss": 1.3846,
      "step": 180
    },
    {
      "epoch": 1.21,
      "grad_norm": 46.30598068237305,
      "learning_rate": 2.983014861995754e-06,
      "loss": 1.365,
      "step": 190
    },
    {
      "epoch": 1.27,
      "grad_norm": 62.2653923034668,
      "learning_rate": 2.8768577494692146e-06,
      "loss": 1.2859,
      "step": 200
    },
    {
      "epoch": 1.27,
      "eval_accuracy": 0.4685714285714286,
      "eval_f1_macro": 0.30187587586686543,
      "eval_f1_micro": 0.4685714285714286,
      "eval_loss": 1.793928623199463,
      "eval_runtime": 3.6406,
      "eval_samples_per_second": 384.55,
      "eval_steps_per_second": 12.086,
      "step": 200
    },
    {
      "epoch": 1.34,
      "grad_norm": 38.229000091552734,
      "learning_rate": 2.7707006369426754e-06,
      "loss": 1.3562,
      "step": 210
    },
    {
      "epoch": 1.4,
      "grad_norm": 64.51058197021484,
      "learning_rate": 2.6645435244161363e-06,
      "loss": 1.2486,
      "step": 220
    },
    {
      "epoch": 1.46,
      "grad_norm": 45.04010009765625,
      "learning_rate": 2.5583864118895966e-06,
      "loss": 1.3775,
      "step": 230
    },
    {
      "epoch": 1.53,
      "grad_norm": 36.0023307800293,
      "learning_rate": 2.4522292993630575e-06,
      "loss": 1.2434,
      "step": 240
    },
    {
      "epoch": 1.59,
      "grad_norm": 42.964847564697266,
      "learning_rate": 2.3460721868365183e-06,
      "loss": 1.1971,
      "step": 250
    },
    {
      "epoch": 1.59,
      "eval_accuracy": 0.4957142857142857,
      "eval_f1_macro": 0.33073042661457275,
      "eval_f1_micro": 0.4957142857142857,
      "eval_loss": 1.7145981788635254,
      "eval_runtime": 3.8826,
      "eval_samples_per_second": 360.584,
      "eval_steps_per_second": 11.333,
      "step": 250
    },
    {
      "epoch": 1.66,
      "grad_norm": 43.22111511230469,
      "learning_rate": 2.239915074309979e-06,
      "loss": 1.2471,
      "step": 260
    },
    {
      "epoch": 1.72,
      "grad_norm": 49.360374450683594,
      "learning_rate": 2.13375796178344e-06,
      "loss": 1.2869,
      "step": 270
    },
    {
      "epoch": 1.78,
      "grad_norm": 54.94723129272461,
      "learning_rate": 2.0276008492569003e-06,
      "loss": 1.1953,
      "step": 280
    },
    {
      "epoch": 1.85,
      "grad_norm": 58.562225341796875,
      "learning_rate": 1.921443736730361e-06,
      "loss": 1.1973,
      "step": 290
    },
    {
      "epoch": 1.91,
      "grad_norm": 47.98759460449219,
      "learning_rate": 1.8152866242038217e-06,
      "loss": 1.1271,
      "step": 300
    },
    {
      "epoch": 1.91,
      "eval_accuracy": 0.515,
      "eval_f1_macro": 0.35688242427831446,
      "eval_f1_micro": 0.515,
      "eval_loss": 1.6402901411056519,
      "eval_runtime": 3.6453,
      "eval_samples_per_second": 384.061,
      "eval_steps_per_second": 12.071,
      "step": 300
    }
  ],
  "logging_steps": 10,
  "max_steps": 471,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 50,
  "total_flos": 8954453164032000.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}