{
  "best_metric": 0.7207300066947937,
  "best_model_checkpoint": "/raildefectfft2/checkpoint-30",
  "epoch": 30.0,
  "global_step": 450,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.67,
      "learning_rate": 0.00019555555555555556,
      "loss": 1.3922,
      "step": 10
    },
    {
      "epoch": 0.67,
      "eval_accuracy": 0.6114285714285714,
      "eval_loss": 1.1690133810043335,
      "eval_runtime": 217.3963,
      "eval_samples_per_second": 1.61,
      "eval_steps_per_second": 0.202,
      "step": 10
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.00019111111111111114,
      "loss": 0.8518,
      "step": 20
    },
    {
      "epoch": 1.33,
      "eval_accuracy": 0.6828571428571428,
      "eval_loss": 0.8874489068984985,
      "eval_runtime": 217.0024,
      "eval_samples_per_second": 1.613,
      "eval_steps_per_second": 0.203,
      "step": 20
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0001866666666666667,
      "loss": 0.5386,
      "step": 30
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7542857142857143,
      "eval_loss": 0.7207300066947937,
      "eval_runtime": 221.2829,
      "eval_samples_per_second": 1.582,
      "eval_steps_per_second": 0.199,
      "step": 30
    },
    {
      "epoch": 2.67,
      "learning_rate": 0.00018222222222222224,
      "loss": 0.3125,
      "step": 40
    },
    {
      "epoch": 2.67,
      "eval_accuracy": 0.7285714285714285,
      "eval_loss": 0.8382583260536194,
      "eval_runtime": 217.2612,
      "eval_samples_per_second": 1.611,
      "eval_steps_per_second": 0.203,
      "step": 40
    },
    {
      "epoch": 3.33,
      "learning_rate": 0.00017777777777777779,
      "loss": 0.2264,
      "step": 50
    },
    {
      "epoch": 3.33,
      "eval_accuracy": 0.7428571428571429,
      "eval_loss": 0.8440293669700623,
      "eval_runtime": 215.3358,
      "eval_samples_per_second": 1.625,
      "eval_steps_per_second": 0.204,
      "step": 50
    },
    {
      "epoch": 4.0,
      "learning_rate": 0.00017333333333333334,
      "loss": 0.1613,
      "step": 60
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7457142857142857,
      "eval_loss": 0.8516280055046082,
      "eval_runtime": 221.3905,
      "eval_samples_per_second": 1.581,
      "eval_steps_per_second": 0.199,
      "step": 60
    },
    {
      "epoch": 4.67,
      "learning_rate": 0.00016888888888888889,
      "loss": 0.119,
      "step": 70
    },
    {
      "epoch": 4.67,
      "eval_accuracy": 0.6,
      "eval_loss": 1.3625210523605347,
      "eval_runtime": 210.1415,
      "eval_samples_per_second": 1.666,
      "eval_steps_per_second": 0.209,
      "step": 70
    },
    {
      "epoch": 5.33,
      "learning_rate": 0.00016444444444444444,
      "loss": 0.0972,
      "step": 80
    },
    {
      "epoch": 5.33,
      "eval_accuracy": 0.7428571428571429,
      "eval_loss": 0.9109606146812439,
      "eval_runtime": 208.8901,
      "eval_samples_per_second": 1.676,
      "eval_steps_per_second": 0.211,
      "step": 80
    },
    {
      "epoch": 6.0,
      "learning_rate": 0.00016,
      "loss": 0.0844,
      "step": 90
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.78,
      "eval_loss": 0.8271930813789368,
      "eval_runtime": 214.6994,
      "eval_samples_per_second": 1.63,
      "eval_steps_per_second": 0.205,
      "step": 90
    },
    {
      "epoch": 6.67,
      "learning_rate": 0.00015555555555555556,
      "loss": 0.0725,
      "step": 100
    },
    {
      "epoch": 6.67,
      "eval_accuracy": 0.74,
      "eval_loss": 0.8958377242088318,
      "eval_runtime": 209.0619,
      "eval_samples_per_second": 1.674,
      "eval_steps_per_second": 0.21,
      "step": 100
    },
    {
      "epoch": 7.33,
      "learning_rate": 0.0001511111111111111,
      "loss": 0.0708,
      "step": 110
    },
    {
      "epoch": 7.33,
      "eval_accuracy": 0.7371428571428571,
      "eval_loss": 1.0972360372543335,
      "eval_runtime": 210.819,
      "eval_samples_per_second": 1.66,
      "eval_steps_per_second": 0.209,
      "step": 110
    },
    {
      "epoch": 8.0,
      "learning_rate": 0.00014666666666666666,
      "loss": 0.041,
      "step": 120
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.7628571428571429,
      "eval_loss": 1.0088900327682495,
      "eval_runtime": 209.2911,
      "eval_samples_per_second": 1.672,
      "eval_steps_per_second": 0.21,
      "step": 120
    },
    {
      "epoch": 8.67,
      "learning_rate": 0.00014222222222222224,
      "loss": 0.0312,
      "step": 130
    },
    {
      "epoch": 8.67,
      "eval_accuracy": 0.7628571428571429,
      "eval_loss": 1.03481125831604,
      "eval_runtime": 215.9471,
      "eval_samples_per_second": 1.621,
      "eval_steps_per_second": 0.204,
      "step": 130
    },
    {
      "epoch": 9.33,
      "learning_rate": 0.0001377777777777778,
      "loss": 0.0401,
      "step": 140
    },
    {
      "epoch": 9.33,
      "eval_accuracy": 0.7257142857142858,
      "eval_loss": 1.2426719665527344,
      "eval_runtime": 211.184,
      "eval_samples_per_second": 1.657,
      "eval_steps_per_second": 0.208,
      "step": 140
    },
    {
      "epoch": 10.0,
      "learning_rate": 0.00013333333333333334,
      "loss": 0.0271,
      "step": 150
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.7542857142857143,
      "eval_loss": 1.0153539180755615,
      "eval_runtime": 208.7196,
      "eval_samples_per_second": 1.677,
      "eval_steps_per_second": 0.211,
      "step": 150
    },
    {
      "epoch": 10.67,
      "learning_rate": 0.00012888888888888892,
      "loss": 0.0328,
      "step": 160
    },
    {
      "epoch": 10.67,
      "eval_accuracy": 0.7714285714285715,
      "eval_loss": 1.0373210906982422,
      "eval_runtime": 216.803,
      "eval_samples_per_second": 1.614,
      "eval_steps_per_second": 0.203,
      "step": 160
    },
    {
      "epoch": 11.33,
      "learning_rate": 0.00012444444444444444,
      "loss": 0.023,
      "step": 170
    },
    {
      "epoch": 11.33,
      "eval_accuracy": 0.7685714285714286,
      "eval_loss": 1.005110502243042,
      "eval_runtime": 209.183,
      "eval_samples_per_second": 1.673,
      "eval_steps_per_second": 0.21,
      "step": 170
    },
    {
      "epoch": 12.0,
      "learning_rate": 0.00012,
      "loss": 0.0199,
      "step": 180
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.7657142857142857,
      "eval_loss": 0.9775477647781372,
      "eval_runtime": 208.6639,
      "eval_samples_per_second": 1.677,
      "eval_steps_per_second": 0.211,
      "step": 180
    },
    {
      "epoch": 12.67,
      "learning_rate": 0.00011555555555555555,
      "loss": 0.0189,
      "step": 190
    },
    {
      "epoch": 12.67,
      "eval_accuracy": 0.7657142857142857,
      "eval_loss": 1.008815050125122,
      "eval_runtime": 216.3653,
      "eval_samples_per_second": 1.618,
      "eval_steps_per_second": 0.203,
      "step": 190
    },
    {
      "epoch": 13.33,
      "learning_rate": 0.00011111111111111112,
      "loss": 0.0188,
      "step": 200
    },
    {
      "epoch": 13.33,
      "eval_accuracy": 0.7342857142857143,
      "eval_loss": 1.1904319524765015,
      "eval_runtime": 209.8186,
      "eval_samples_per_second": 1.668,
      "eval_steps_per_second": 0.21,
      "step": 200
    },
    {
      "epoch": 14.0,
      "learning_rate": 0.00010666666666666667,
      "loss": 0.0167,
      "step": 210
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.7285714285714285,
      "eval_loss": 1.2999135255813599,
      "eval_runtime": 209.7587,
      "eval_samples_per_second": 1.669,
      "eval_steps_per_second": 0.21,
      "step": 210
    },
    {
      "epoch": 14.67,
      "learning_rate": 0.00010222222222222222,
      "loss": 0.0159,
      "step": 220
    },
    {
      "epoch": 14.67,
      "eval_accuracy": 0.7514285714285714,
      "eval_loss": 1.1326370239257812,
      "eval_runtime": 215.8574,
      "eval_samples_per_second": 1.621,
      "eval_steps_per_second": 0.204,
      "step": 220
    },
    {
      "epoch": 15.33,
      "learning_rate": 9.777777777777778e-05,
      "loss": 0.0145,
      "step": 230
    },
    {
      "epoch": 15.33,
      "eval_accuracy": 0.7542857142857143,
      "eval_loss": 1.1385791301727295,
      "eval_runtime": 209.3052,
      "eval_samples_per_second": 1.672,
      "eval_steps_per_second": 0.21,
      "step": 230
    },
    {
      "epoch": 16.0,
      "learning_rate": 9.333333333333334e-05,
      "loss": 0.015,
      "step": 240
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.7542857142857143,
      "eval_loss": 1.1441489458084106,
      "eval_runtime": 209.1403,
      "eval_samples_per_second": 1.674,
      "eval_steps_per_second": 0.21,
      "step": 240
    },
    {
      "epoch": 16.67,
      "learning_rate": 8.888888888888889e-05,
      "loss": 0.0133,
      "step": 250
    },
    {
      "epoch": 16.67,
      "eval_accuracy": 0.7514285714285714,
      "eval_loss": 1.154445767402649,
      "eval_runtime": 212.9184,
      "eval_samples_per_second": 1.644,
      "eval_steps_per_second": 0.207,
      "step": 250
    },
    {
      "epoch": 17.33,
      "learning_rate": 8.444444444444444e-05,
      "loss": 0.0132,
      "step": 260
    },
    {
      "epoch": 17.33,
      "eval_accuracy": 0.7514285714285714,
      "eval_loss": 1.1629431247711182,
      "eval_runtime": 209.529,
      "eval_samples_per_second": 1.67,
      "eval_steps_per_second": 0.21,
      "step": 260
    },
    {
      "epoch": 18.0,
      "learning_rate": 8e-05,
      "loss": 0.0121,
      "step": 270
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.7514285714285714,
      "eval_loss": 1.1707779169082642,
      "eval_runtime": 209.7822,
      "eval_samples_per_second": 1.668,
      "eval_steps_per_second": 0.21,
      "step": 270
    },
    {
      "epoch": 18.67,
      "learning_rate": 7.555555555555556e-05,
      "loss": 0.0121,
      "step": 280
    },
    {
      "epoch": 18.67,
      "eval_accuracy": 0.7514285714285714,
      "eval_loss": 1.1773394346237183,
      "eval_runtime": 220.9849,
      "eval_samples_per_second": 1.584,
      "eval_steps_per_second": 0.199,
      "step": 280
    },
    {
      "epoch": 19.33,
      "learning_rate": 7.111111111111112e-05,
      "loss": 0.0114,
      "step": 290
    },
    {
      "epoch": 19.33,
      "eval_accuracy": 0.7514285714285714,
      "eval_loss": 1.1831494569778442,
      "eval_runtime": 216.1426,
      "eval_samples_per_second": 1.619,
      "eval_steps_per_second": 0.204,
      "step": 290
    },
    {
      "epoch": 20.0,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.0111,
      "step": 300
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.7514285714285714,
      "eval_loss": 1.1883198022842407,
      "eval_runtime": 210.1269,
      "eval_samples_per_second": 1.666,
      "eval_steps_per_second": 0.209,
      "step": 300
    },
    {
      "epoch": 20.67,
      "learning_rate": 6.222222222222222e-05,
      "loss": 0.011,
      "step": 310
    },
    {
      "epoch": 20.67,
      "eval_accuracy": 0.7514285714285714,
      "eval_loss": 1.1936721801757812,
      "eval_runtime": 210.5377,
      "eval_samples_per_second": 1.662,
      "eval_steps_per_second": 0.209,
      "step": 310
    },
    {
      "epoch": 21.33,
      "learning_rate": 5.7777777777777776e-05,
      "loss": 0.0103,
      "step": 320
    },
    {
      "epoch": 21.33,
      "eval_accuracy": 0.7514285714285714,
      "eval_loss": 1.1992676258087158,
      "eval_runtime": 214.4415,
      "eval_samples_per_second": 1.632,
      "eval_steps_per_second": 0.205,
      "step": 320
    },
    {
      "epoch": 22.0,
      "learning_rate": 5.333333333333333e-05,
      "loss": 0.0103,
      "step": 330
    },
    {
      "epoch": 22.0,
      "eval_accuracy": 0.7514285714285714,
      "eval_loss": 1.2045563459396362,
      "eval_runtime": 211.851,
      "eval_samples_per_second": 1.652,
      "eval_steps_per_second": 0.208,
      "step": 330
    },
    {
      "epoch": 22.67,
      "learning_rate": 4.888888888888889e-05,
      "loss": 0.0103,
      "step": 340
    },
    {
      "epoch": 22.67,
      "eval_accuracy": 0.7514285714285714,
      "eval_loss": 1.208925724029541,
      "eval_runtime": 209.6112,
      "eval_samples_per_second": 1.67,
      "eval_steps_per_second": 0.21,
      "step": 340
    },
    {
      "epoch": 23.33,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 0.0096,
      "step": 350
    },
    {
      "epoch": 23.33,
      "eval_accuracy": 0.7514285714285714,
      "eval_loss": 1.2132576704025269,
      "eval_runtime": 219.2216,
      "eval_samples_per_second": 1.597,
      "eval_steps_per_second": 0.201,
      "step": 350
    },
    {
      "epoch": 24.0,
      "learning_rate": 4e-05,
      "loss": 0.0095,
      "step": 360
    },
    {
      "epoch": 24.0,
      "eval_accuracy": 0.7514285714285714,
      "eval_loss": 1.2171136140823364,
      "eval_runtime": 208.9362,
      "eval_samples_per_second": 1.675,
      "eval_steps_per_second": 0.211,
      "step": 360
    },
    {
      "epoch": 24.67,
      "learning_rate": 3.555555555555556e-05,
      "loss": 0.0096,
      "step": 370
    },
    {
      "epoch": 24.67,
      "eval_accuracy": 0.7514285714285714,
      "eval_loss": 1.2204023599624634,
      "eval_runtime": 209.9857,
      "eval_samples_per_second": 1.667,
      "eval_steps_per_second": 0.21,
      "step": 370
    },
    {
      "epoch": 25.33,
      "learning_rate": 3.111111111111111e-05,
      "loss": 0.0093,
      "step": 380
    },
    {
      "epoch": 25.33,
      "eval_accuracy": 0.7485714285714286,
      "eval_loss": 1.2234961986541748,
      "eval_runtime": 216.2311,
      "eval_samples_per_second": 1.619,
      "eval_steps_per_second": 0.203,
      "step": 380
    },
    {
      "epoch": 26.0,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 0.0091,
      "step": 390
    },
    {
      "epoch": 26.0,
      "eval_accuracy": 0.7485714285714286,
      "eval_loss": 1.2261521816253662,
      "eval_runtime": 210.1553,
      "eval_samples_per_second": 1.665,
      "eval_steps_per_second": 0.209,
      "step": 390
    },
    {
      "epoch": 26.67,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 0.0092,
      "step": 400
    },
    {
      "epoch": 26.67,
      "eval_accuracy": 0.7514285714285714,
      "eval_loss": 1.227960228919983,
      "eval_runtime": 211.3289,
      "eval_samples_per_second": 1.656,
      "eval_steps_per_second": 0.208,
      "step": 400
    },
    {
      "epoch": 27.33,
      "learning_rate": 1.777777777777778e-05,
      "loss": 0.0089,
      "step": 410
    },
    {
      "epoch": 27.33,
      "eval_accuracy": 0.7514285714285714,
      "eval_loss": 1.2296239137649536,
      "eval_runtime": 213.6977,
      "eval_samples_per_second": 1.638,
      "eval_steps_per_second": 0.206,
      "step": 410
    },
    {
      "epoch": 28.0,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.0092,
      "step": 420
    },
    {
      "epoch": 28.0,
      "eval_accuracy": 0.7514285714285714,
      "eval_loss": 1.2309640645980835,
      "eval_runtime": 210.9587,
      "eval_samples_per_second": 1.659,
      "eval_steps_per_second": 0.209,
      "step": 420
    },
    {
      "epoch": 28.67,
      "learning_rate": 8.88888888888889e-06,
      "loss": 0.0089,
      "step": 430
    },
    {
      "epoch": 28.67,
      "eval_accuracy": 0.7485714285714286,
      "eval_loss": 1.2319449186325073,
      "eval_runtime": 210.158,
      "eval_samples_per_second": 1.665,
      "eval_steps_per_second": 0.209,
      "step": 430
    },
    {
      "epoch": 29.33,
      "learning_rate": 4.444444444444445e-06,
      "loss": 0.0089,
      "step": 440
    },
    {
      "epoch": 29.33,
      "eval_accuracy": 0.7485714285714286,
      "eval_loss": 1.23252272605896,
      "eval_runtime": 216.1534,
      "eval_samples_per_second": 1.619,
      "eval_steps_per_second": 0.204,
      "step": 440
    },
    {
      "epoch": 30.0,
      "learning_rate": 0.0,
      "loss": 0.0088,
      "step": 450
    },
    {
      "epoch": 30.0,
      "eval_accuracy": 0.7485714285714286,
      "eval_loss": 1.2327271699905396,
      "eval_runtime": 216.0649,
      "eval_samples_per_second": 1.62,
      "eval_steps_per_second": 0.204,
      "step": 450
    },
    {
      "epoch": 30.0,
      "step": 450,
      "total_flos": 9.8805828893184e+17,
      "train_loss": 0.09906705205639203,
      "train_runtime": 36217.9644,
      "train_samples_per_second": 0.352,
      "train_steps_per_second": 0.012
    }
  ],
  "max_steps": 450,
  "num_train_epochs": 30,
  "total_flos": 9.8805828893184e+17,
  "trial_name": null,
  "trial_params": null
}