{
  "best_metric": 0.605766236782074,
  "best_model_checkpoint": "./vit-base-cats/checkpoint-500",
  "epoch": 2.857142857142857,
  "eval_steps": 100,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.06,
      "grad_norm": 104665.1796875,
      "learning_rate": 0.00019714285714285716,
      "loss": 2.8888,
      "step": 10
    },
    {
      "epoch": 0.11,
      "grad_norm": 102043.53125,
      "learning_rate": 0.0001942857142857143,
      "loss": 2.6146,
      "step": 20
    },
    {
      "epoch": 0.17,
      "grad_norm": 97863.03125,
      "learning_rate": 0.00019142857142857145,
      "loss": 2.3775,
      "step": 30
    },
    {
      "epoch": 0.23,
      "grad_norm": 98961.4453125,
      "learning_rate": 0.00018857142857142857,
      "loss": 2.0395,
      "step": 40
    },
    {
      "epoch": 0.29,
      "grad_norm": 142313.515625,
      "learning_rate": 0.00018571428571428572,
      "loss": 1.8599,
      "step": 50
    },
    {
      "epoch": 0.34,
      "grad_norm": 112630.1328125,
      "learning_rate": 0.00018285714285714286,
      "loss": 1.7573,
      "step": 60
    },
    {
      "epoch": 0.4,
      "grad_norm": 180249.9375,
      "learning_rate": 0.00018,
      "loss": 1.5562,
      "step": 70
    },
    {
      "epoch": 0.46,
      "grad_norm": 124194.484375,
      "learning_rate": 0.00017714285714285713,
      "loss": 1.4203,
      "step": 80
    },
    {
      "epoch": 0.51,
      "grad_norm": 187150.671875,
      "learning_rate": 0.0001742857142857143,
      "loss": 1.2605,
      "step": 90
    },
    {
      "epoch": 0.57,
      "grad_norm": 151804.65625,
      "learning_rate": 0.00017142857142857143,
      "loss": 1.2422,
      "step": 100
    },
    {
      "epoch": 0.57,
      "eval_accuracy": 0.7314285714285714,
      "eval_loss": 1.1657222509384155,
      "eval_runtime": 15.2094,
      "eval_samples_per_second": 46.024,
      "eval_steps_per_second": 2.893,
      "step": 100
    },
    {
      "epoch": 0.63,
      "grad_norm": 134052.609375,
      "learning_rate": 0.00016857142857142857,
      "loss": 1.1438,
      "step": 110
    },
    {
      "epoch": 0.69,
      "grad_norm": 163657.515625,
      "learning_rate": 0.00016571428571428575,
      "loss": 1.2158,
      "step": 120
    },
    {
      "epoch": 0.74,
      "grad_norm": 159906.9375,
      "learning_rate": 0.00016285714285714287,
      "loss": 1.0575,
      "step": 130
    },
    {
      "epoch": 0.8,
      "grad_norm": 171710.640625,
      "learning_rate": 0.00016,
      "loss": 1.0049,
      "step": 140
    },
    {
      "epoch": 0.86,
      "grad_norm": 171094.046875,
      "learning_rate": 0.00015714285714285716,
      "loss": 1.0786,
      "step": 150
    },
    {
      "epoch": 0.91,
      "grad_norm": 140285.921875,
      "learning_rate": 0.0001542857142857143,
      "loss": 0.9912,
      "step": 160
    },
    {
      "epoch": 0.97,
      "grad_norm": 170517.640625,
      "learning_rate": 0.00015142857142857143,
      "loss": 0.9448,
      "step": 170
    },
    {
      "epoch": 1.03,
      "grad_norm": 125416.1484375,
      "learning_rate": 0.00014857142857142857,
      "loss": 0.826,
      "step": 180
    },
    {
      "epoch": 1.09,
      "grad_norm": 205800.3125,
      "learning_rate": 0.00014571428571428572,
      "loss": 0.7471,
      "step": 190
    },
    {
      "epoch": 1.14,
      "grad_norm": 74866.34375,
      "learning_rate": 0.00014285714285714287,
      "loss": 0.5948,
      "step": 200
    },
    {
      "epoch": 1.14,
      "eval_accuracy": 0.8,
      "eval_loss": 0.7744152545928955,
      "eval_runtime": 11.8882,
      "eval_samples_per_second": 58.882,
      "eval_steps_per_second": 3.701,
      "step": 200
    },
    {
      "epoch": 1.2,
      "grad_norm": 191153.328125,
      "learning_rate": 0.00014,
      "loss": 0.5425,
      "step": 210
    },
    {
      "epoch": 1.26,
      "grad_norm": 181985.578125,
      "learning_rate": 0.00013714285714285716,
      "loss": 0.6606,
      "step": 220
    },
    {
      "epoch": 1.31,
      "grad_norm": 193704.78125,
      "learning_rate": 0.00013428571428571428,
      "loss": 0.5681,
      "step": 230
    },
    {
      "epoch": 1.37,
      "grad_norm": 135817.890625,
      "learning_rate": 0.00013142857142857143,
      "loss": 0.5497,
      "step": 240
    },
    {
      "epoch": 1.43,
      "grad_norm": 154213.90625,
      "learning_rate": 0.00012857142857142858,
      "loss": 0.562,
      "step": 250
    },
    {
      "epoch": 1.49,
      "grad_norm": 132152.8125,
      "learning_rate": 0.00012571428571428572,
      "loss": 0.5728,
      "step": 260
    },
    {
      "epoch": 1.54,
      "grad_norm": 207587.0625,
      "learning_rate": 0.00012285714285714287,
      "loss": 0.6254,
      "step": 270
    },
    {
      "epoch": 1.6,
      "grad_norm": 98314.2109375,
      "learning_rate": 0.00012,
      "loss": 0.4033,
      "step": 280
    },
    {
      "epoch": 1.66,
      "grad_norm": 107417.6328125,
      "learning_rate": 0.00011714285714285715,
      "loss": 0.4892,
      "step": 290
    },
    {
      "epoch": 1.71,
      "grad_norm": 250248.421875,
      "learning_rate": 0.00011428571428571428,
      "loss": 0.5324,
      "step": 300
    },
    {
      "epoch": 1.71,
      "eval_accuracy": 0.8128571428571428,
      "eval_loss": 0.6736084818840027,
      "eval_runtime": 13.3903,
      "eval_samples_per_second": 52.277,
      "eval_steps_per_second": 3.286,
      "step": 300
    },
    {
      "epoch": 1.77,
      "grad_norm": 165674.15625,
      "learning_rate": 0.00011142857142857144,
      "loss": 0.4641,
      "step": 310
    },
    {
      "epoch": 1.83,
      "grad_norm": 194581.40625,
      "learning_rate": 0.00010857142857142856,
      "loss": 0.4818,
      "step": 320
    },
    {
      "epoch": 1.89,
      "grad_norm": 138774.1875,
      "learning_rate": 0.00010571428571428572,
      "loss": 0.4677,
      "step": 330
    },
    {
      "epoch": 1.94,
      "grad_norm": 203262.421875,
      "learning_rate": 0.00010285714285714286,
      "loss": 0.4034,
      "step": 340
    },
    {
      "epoch": 2.0,
      "grad_norm": 128663.1875,
      "learning_rate": 0.0001,
      "loss": 0.3978,
      "step": 350
    },
    {
      "epoch": 2.06,
      "grad_norm": 35406.80078125,
      "learning_rate": 9.714285714285715e-05,
      "loss": 0.2415,
      "step": 360
    },
    {
      "epoch": 2.11,
      "grad_norm": 128107.3046875,
      "learning_rate": 9.428571428571429e-05,
      "loss": 0.2798,
      "step": 370
    },
    {
      "epoch": 2.17,
      "grad_norm": 52566.515625,
      "learning_rate": 9.142857142857143e-05,
      "loss": 0.2742,
      "step": 380
    },
    {
      "epoch": 2.23,
      "grad_norm": 49925.12890625,
      "learning_rate": 8.857142857142857e-05,
      "loss": 0.2293,
      "step": 390
    },
    {
      "epoch": 2.29,
      "grad_norm": 175336.984375,
      "learning_rate": 8.571428571428571e-05,
      "loss": 0.212,
      "step": 400
    },
    {
      "epoch": 2.29,
      "eval_accuracy": 0.8085714285714286,
      "eval_loss": 0.7056272625923157,
      "eval_runtime": 11.5847,
      "eval_samples_per_second": 60.424,
      "eval_steps_per_second": 3.798,
      "step": 400
    },
    {
      "epoch": 2.34,
      "grad_norm": 24206.984375,
      "learning_rate": 8.285714285714287e-05,
      "loss": 0.2311,
      "step": 410
    },
    {
      "epoch": 2.4,
      "grad_norm": 124311.4140625,
      "learning_rate": 8e-05,
      "loss": 0.2342,
      "step": 420
    },
    {
      "epoch": 2.46,
      "grad_norm": 68467.265625,
      "learning_rate": 7.714285714285715e-05,
      "loss": 0.2055,
      "step": 430
    },
    {
      "epoch": 2.51,
      "grad_norm": 104893.53125,
      "learning_rate": 7.428571428571429e-05,
      "loss": 0.1814,
      "step": 440
    },
    {
      "epoch": 2.57,
      "grad_norm": 76138.921875,
      "learning_rate": 7.142857142857143e-05,
      "loss": 0.2127,
      "step": 450
    },
    {
      "epoch": 2.63,
      "grad_norm": 138337.28125,
      "learning_rate": 6.857142857142858e-05,
      "loss": 0.1863,
      "step": 460
    },
    {
      "epoch": 2.69,
      "grad_norm": 47311.3203125,
      "learning_rate": 6.571428571428571e-05,
      "loss": 0.2281,
      "step": 470
    },
    {
      "epoch": 2.74,
      "grad_norm": 107393.203125,
      "learning_rate": 6.285714285714286e-05,
      "loss": 0.1875,
      "step": 480
    },
    {
      "epoch": 2.8,
      "grad_norm": 167001.0,
      "learning_rate": 6e-05,
      "loss": 0.1896,
      "step": 490
    },
    {
      "epoch": 2.86,
      "grad_norm": 90490.375,
      "learning_rate": 5.714285714285714e-05,
      "loss": 0.2158,
      "step": 500
    },
    {
      "epoch": 2.86,
      "eval_accuracy": 0.82,
      "eval_loss": 0.605766236782074,
      "eval_runtime": 11.55,
      "eval_samples_per_second": 60.606,
      "eval_steps_per_second": 3.81,
      "step": 500
    }
  ],
  "logging_steps": 10,
  "max_steps": 700,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 100,
  "total_flos": 1.240071860256768e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}