{
  "best_metric": 3.962663412094116,
  "best_model_checkpoint": "output_hemo_aug_4/checkpoint-990",
  "epoch": 200.0,
  "eval_steps": 500,
  "global_step": 1200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "learning_rate": 9.950000000000001e-06,
      "loss": 9.4301,
      "step": 6
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.024437927663734114,
      "eval_loss": 8.684534072875977,
      "eval_runtime": 3.4336,
      "eval_samples_per_second": 0.291,
      "eval_steps_per_second": 0.291,
      "step": 6
    },
    {
      "epoch": 2.0,
      "learning_rate": 9.9e-06,
      "loss": 8.4172,
      "step": 12
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.08113391984359726,
      "eval_loss": 7.8148698806762695,
      "eval_runtime": 3.5386,
      "eval_samples_per_second": 0.283,
      "eval_steps_per_second": 0.283,
      "step": 12
    },
    {
      "epoch": 3.0,
      "learning_rate": 9.85e-06,
      "loss": 7.6869,
      "step": 18
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.15542521994134897,
      "eval_loss": 7.201221942901611,
      "eval_runtime": 3.2749,
      "eval_samples_per_second": 0.305,
      "eval_steps_per_second": 0.305,
      "step": 18
    },
    {
      "epoch": 4.0,
      "learning_rate": 9.800000000000001e-06,
      "loss": 7.1731,
      "step": 24
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.1827956989247312,
      "eval_loss": 6.913944721221924,
      "eval_runtime": 3.2424,
      "eval_samples_per_second": 0.308,
      "eval_steps_per_second": 0.308,
      "step": 24
    },
    {
      "epoch": 5.0,
      "learning_rate": 9.75e-06,
      "loss": 6.8807,
      "step": 30
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.19550342130987292,
      "eval_loss": 6.623848915100098,
      "eval_runtime": 3.4613,
      "eval_samples_per_second": 0.289,
      "eval_steps_per_second": 0.289,
      "step": 30
    },
    {
      "epoch": 6.0,
      "learning_rate": 9.7e-06,
      "loss": 6.6009,
      "step": 36
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.1935483870967742,
      "eval_loss": 6.384663105010986,
      "eval_runtime": 4.2768,
      "eval_samples_per_second": 0.234,
      "eval_steps_per_second": 0.234,
      "step": 36
    },
    {
      "epoch": 7.0,
      "learning_rate": 9.65e-06,
      "loss": 6.4347,
      "step": 42
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.20625610948191594,
      "eval_loss": 6.234148979187012,
      "eval_runtime": 3.3994,
      "eval_samples_per_second": 0.294,
      "eval_steps_per_second": 0.294,
      "step": 42
    },
    {
      "epoch": 8.0,
      "learning_rate": 9.600000000000001e-06,
      "loss": 6.2831,
      "step": 48
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.21407624633431085,
      "eval_loss": 6.096428394317627,
      "eval_runtime": 3.1809,
      "eval_samples_per_second": 0.314,
      "eval_steps_per_second": 0.314,
      "step": 48
    },
    {
      "epoch": 9.0,
      "learning_rate": 9.55e-06,
      "loss": 6.1728,
      "step": 54
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.2209188660801564,
      "eval_loss": 5.98640775680542,
      "eval_runtime": 3.4302,
      "eval_samples_per_second": 0.292,
      "eval_steps_per_second": 0.292,
      "step": 54
    },
    {
      "epoch": 10.0,
      "learning_rate": 9.5e-06,
      "loss": 6.0805,
      "step": 60
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.2316715542521994,
      "eval_loss": 5.893611907958984,
      "eval_runtime": 3.6923,
      "eval_samples_per_second": 0.271,
      "eval_steps_per_second": 0.271,
      "step": 60
    },
    {
      "epoch": 11.0,
      "learning_rate": 9.450000000000001e-06,
      "loss": 5.9959,
      "step": 66
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.2404692082111437,
      "eval_loss": 5.81611967086792,
      "eval_runtime": 3.2748,
      "eval_samples_per_second": 0.305,
      "eval_steps_per_second": 0.305,
      "step": 66
    },
    {
      "epoch": 12.0,
      "learning_rate": 9.4e-06,
      "loss": 5.925,
      "step": 72
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.23851417399804498,
      "eval_loss": 5.745607852935791,
      "eval_runtime": 3.4474,
      "eval_samples_per_second": 0.29,
      "eval_steps_per_second": 0.29,
      "step": 72
    },
    {
      "epoch": 13.0,
      "learning_rate": 9.350000000000002e-06,
      "loss": 5.8787,
      "step": 78
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.2482893450635386,
      "eval_loss": 5.664583683013916,
      "eval_runtime": 3.5432,
      "eval_samples_per_second": 0.282,
      "eval_steps_per_second": 0.282,
      "step": 78
    },
    {
      "epoch": 14.0,
      "learning_rate": 9.3e-06,
      "loss": 5.7996,
      "step": 84
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.24926686217008798,
      "eval_loss": 5.590141296386719,
      "eval_runtime": 3.2321,
      "eval_samples_per_second": 0.309,
      "eval_steps_per_second": 0.309,
      "step": 84
    },
    {
      "epoch": 15.0,
      "learning_rate": 9.250000000000001e-06,
      "loss": 5.7312,
      "step": 90
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.25317693059628543,
      "eval_loss": 5.521559238433838,
      "eval_runtime": 3.5703,
      "eval_samples_per_second": 0.28,
      "eval_steps_per_second": 0.28,
      "step": 90
    },
    {
      "epoch": 16.0,
      "learning_rate": 9.200000000000002e-06,
      "loss": 5.6751,
      "step": 96
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.2590420332355816,
      "eval_loss": 5.469524383544922,
      "eval_runtime": 3.5655,
      "eval_samples_per_second": 0.28,
      "eval_steps_per_second": 0.28,
      "step": 96
    },
    {
      "epoch": 17.0,
      "learning_rate": 9.15e-06,
      "loss": 5.6076,
      "step": 102
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.2619745845552297,
      "eval_loss": 5.421563148498535,
      "eval_runtime": 3.4299,
      "eval_samples_per_second": 0.292,
      "eval_steps_per_second": 0.292,
      "step": 102
    },
    {
      "epoch": 18.0,
      "learning_rate": 9.100000000000001e-06,
      "loss": 5.569,
      "step": 108
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.2619745845552297,
      "eval_loss": 5.373498916625977,
      "eval_runtime": 3.4799,
      "eval_samples_per_second": 0.287,
      "eval_steps_per_second": 0.287,
      "step": 108
    },
    {
      "epoch": 19.0,
      "learning_rate": 9.050000000000001e-06,
      "loss": 5.5037,
      "step": 114
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.26295210166177907,
      "eval_loss": 5.327227592468262,
      "eval_runtime": 3.5293,
      "eval_samples_per_second": 0.283,
      "eval_steps_per_second": 0.283,
      "step": 114
    },
    {
      "epoch": 20.0,
      "learning_rate": 9e-06,
      "loss": 5.4681,
      "step": 120
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.2697947214076246,
      "eval_loss": 5.285594463348389,
      "eval_runtime": 3.6414,
      "eval_samples_per_second": 0.275,
      "eval_steps_per_second": 0.275,
      "step": 120
    },
    {
      "epoch": 21.0,
      "learning_rate": 8.95e-06,
      "loss": 5.4225,
      "step": 126
    },
    {
      "epoch": 21.0,
      "eval_accuracy": 0.27174975562072334,
      "eval_loss": 5.255927085876465,
      "eval_runtime": 3.6672,
      "eval_samples_per_second": 0.273,
      "eval_steps_per_second": 0.273,
      "step": 126
    },
    {
      "epoch": 22.0,
      "learning_rate": 8.900000000000001e-06,
      "loss": 5.3805,
      "step": 132
    },
    {
      "epoch": 22.0,
      "eval_accuracy": 0.27663734115347016,
      "eval_loss": 5.212594509124756,
      "eval_runtime": 3.547,
      "eval_samples_per_second": 0.282,
      "eval_steps_per_second": 0.282,
      "step": 132
    },
    {
      "epoch": 23.0,
      "learning_rate": 8.85e-06,
      "loss": 5.3527,
      "step": 138
    },
    {
      "epoch": 23.0,
      "eval_accuracy": 0.2756598240469208,
      "eval_loss": 5.1883745193481445,
      "eval_runtime": 3.3405,
      "eval_samples_per_second": 0.299,
      "eval_steps_per_second": 0.299,
      "step": 138
    },
    {
      "epoch": 24.0,
      "learning_rate": 8.8e-06,
      "loss": 5.3033,
      "step": 144
    },
    {
      "epoch": 24.0,
      "eval_accuracy": 0.27956989247311825,
      "eval_loss": 5.153918266296387,
      "eval_runtime": 3.6205,
      "eval_samples_per_second": 0.276,
      "eval_steps_per_second": 0.276,
      "step": 144
    },
    {
      "epoch": 25.0,
      "learning_rate": 8.750000000000001e-06,
      "loss": 5.2635,
      "step": 150
    },
    {
      "epoch": 25.0,
      "eval_accuracy": 0.2854349951124145,
      "eval_loss": 5.111028671264648,
      "eval_runtime": 3.658,
      "eval_samples_per_second": 0.273,
      "eval_steps_per_second": 0.273,
      "step": 150
    },
    {
      "epoch": 26.0,
      "learning_rate": 8.700000000000001e-06,
      "loss": 5.2411,
      "step": 156
    },
    {
      "epoch": 26.0,
      "eval_accuracy": 0.2854349951124145,
      "eval_loss": 5.0881805419921875,
      "eval_runtime": 3.5731,
      "eval_samples_per_second": 0.28,
      "eval_steps_per_second": 0.28,
      "step": 156
    },
    {
      "epoch": 27.0,
      "learning_rate": 8.65e-06,
      "loss": 5.1972,
      "step": 162
    },
    {
      "epoch": 27.0,
      "eval_accuracy": 0.2903225806451613,
      "eval_loss": 5.057516098022461,
      "eval_runtime": 3.6003,
      "eval_samples_per_second": 0.278,
      "eval_steps_per_second": 0.278,
      "step": 162
    },
    {
      "epoch": 28.0,
      "learning_rate": 8.6e-06,
      "loss": 5.163,
      "step": 168
    },
    {
      "epoch": 28.0,
      "eval_accuracy": 0.2913000977517107,
      "eval_loss": 5.029298782348633,
      "eval_runtime": 3.5678,
      "eval_samples_per_second": 0.28,
      "eval_steps_per_second": 0.28,
      "step": 168
    },
    {
      "epoch": 29.0,
      "learning_rate": 8.550000000000001e-06,
      "loss": 5.1273,
      "step": 174
    },
    {
      "epoch": 29.0,
      "eval_accuracy": 0.2903225806451613,
      "eval_loss": 5.00468111038208,
      "eval_runtime": 3.5535,
      "eval_samples_per_second": 0.281,
      "eval_steps_per_second": 0.281,
      "step": 174
    },
    {
      "epoch": 30.0,
      "learning_rate": 8.5e-06,
      "loss": 5.1032,
      "step": 180
    },
    {
      "epoch": 30.0,
      "eval_accuracy": 0.29521016617790813,
      "eval_loss": 4.981687545776367,
      "eval_runtime": 3.5476,
      "eval_samples_per_second": 0.282,
      "eval_steps_per_second": 0.282,
      "step": 180
    },
    {
      "epoch": 31.0,
      "learning_rate": 8.45e-06,
      "loss": 5.0726,
      "step": 186
    },
    {
      "epoch": 31.0,
      "eval_accuracy": 0.29521016617790813,
      "eval_loss": 4.958263874053955,
      "eval_runtime": 3.574,
      "eval_samples_per_second": 0.28,
      "eval_steps_per_second": 0.28,
      "step": 186
    },
    {
      "epoch": 32.0,
      "learning_rate": 8.400000000000001e-06,
      "loss": 5.0405,
      "step": 192
    },
    {
      "epoch": 32.0,
      "eval_accuracy": 0.29521016617790813,
      "eval_loss": 4.9354777336120605,
      "eval_runtime": 3.2861,
      "eval_samples_per_second": 0.304,
      "eval_steps_per_second": 0.304,
      "step": 192
    },
    {
      "epoch": 33.0,
      "learning_rate": 8.35e-06,
      "loss": 5.007,
      "step": 198
    },
    {
      "epoch": 33.0,
      "eval_accuracy": 0.29521016617790813,
      "eval_loss": 4.918440818786621,
      "eval_runtime": 3.5379,
      "eval_samples_per_second": 0.283,
      "eval_steps_per_second": 0.283,
      "step": 198
    },
    {
      "epoch": 34.0,
      "learning_rate": 8.3e-06,
      "loss": 4.9897,
      "step": 204
    },
    {
      "epoch": 34.0,
      "eval_accuracy": 0.29716520039100686,
      "eval_loss": 4.891113758087158,
      "eval_runtime": 3.2282,
      "eval_samples_per_second": 0.31,
      "eval_steps_per_second": 0.31,
      "step": 204
    },
    {
      "epoch": 35.0,
      "learning_rate": 8.25e-06,
      "loss": 4.9416,
      "step": 210
    },
    {
      "epoch": 35.0,
      "eval_accuracy": 0.29716520039100686,
      "eval_loss": 4.862751483917236,
      "eval_runtime": 3.5473,
      "eval_samples_per_second": 0.282,
      "eval_steps_per_second": 0.282,
      "step": 210
    },
    {
      "epoch": 36.0,
      "learning_rate": 8.2e-06,
      "loss": 4.9245,
      "step": 216
    },
    {
      "epoch": 36.0,
      "eval_accuracy": 0.2981427174975562,
      "eval_loss": 4.849916934967041,
      "eval_runtime": 3.5627,
      "eval_samples_per_second": 0.281,
      "eval_steps_per_second": 0.281,
      "step": 216
    },
    {
      "epoch": 37.0,
      "learning_rate": 8.15e-06,
      "loss": 4.901,
      "step": 222
    },
    {
      "epoch": 37.0,
      "eval_accuracy": 0.30303030303030304,
      "eval_loss": 4.826337814331055,
      "eval_runtime": 3.475,
      "eval_samples_per_second": 0.288,
      "eval_steps_per_second": 0.288,
      "step": 222
    },
    {
      "epoch": 38.0,
      "learning_rate": 8.1e-06,
      "loss": 4.8713,
      "step": 228
    },
    {
      "epoch": 38.0,
      "eval_accuracy": 0.30303030303030304,
      "eval_loss": 4.803491115570068,
      "eval_runtime": 3.4779,
      "eval_samples_per_second": 0.288,
      "eval_steps_per_second": 0.288,
      "step": 228
    },
    {
      "epoch": 39.0,
      "learning_rate": 8.050000000000001e-06,
      "loss": 4.845,
      "step": 234
    },
    {
      "epoch": 39.0,
      "eval_accuracy": 0.30596285434995113,
      "eval_loss": 4.7873663902282715,
      "eval_runtime": 3.3398,
      "eval_samples_per_second": 0.299,
      "eval_steps_per_second": 0.299,
      "step": 234
    },
    {
      "epoch": 40.0,
      "learning_rate": 8.000000000000001e-06,
      "loss": 4.8052,
      "step": 240
    },
    {
      "epoch": 40.0,
      "eval_accuracy": 0.3040078201368524,
      "eval_loss": 4.753478527069092,
      "eval_runtime": 3.1758,
      "eval_samples_per_second": 0.315,
      "eval_steps_per_second": 0.315,
      "step": 240
    },
    {
      "epoch": 41.0,
      "learning_rate": 7.950000000000002e-06,
      "loss": 4.7786,
      "step": 246
    },
    {
      "epoch": 41.0,
      "eval_accuracy": 0.30596285434995113,
      "eval_loss": 4.731250762939453,
      "eval_runtime": 3.2392,
      "eval_samples_per_second": 0.309,
      "eval_steps_per_second": 0.309,
      "step": 246
    },
    {
      "epoch": 42.0,
      "learning_rate": 7.9e-06,
      "loss": 4.7501,
      "step": 252
    },
    {
      "epoch": 42.0,
      "eval_accuracy": 0.3088954056695992,
      "eval_loss": 4.717497825622559,
      "eval_runtime": 3.5586,
      "eval_samples_per_second": 0.281,
      "eval_steps_per_second": 0.281,
      "step": 252
    },
    {
      "epoch": 43.0,
      "learning_rate": 7.850000000000001e-06,
      "loss": 4.7221,
      "step": 258
    },
    {
      "epoch": 43.0,
      "eval_accuracy": 0.3118279569892473,
      "eval_loss": 4.697778701782227,
      "eval_runtime": 3.4899,
      "eval_samples_per_second": 0.287,
      "eval_steps_per_second": 0.287,
      "step": 258
    },
    {
      "epoch": 44.0,
      "learning_rate": 7.800000000000002e-06,
      "loss": 4.7038,
      "step": 264
    },
    {
      "epoch": 44.0,
      "eval_accuracy": 0.31085043988269795,
      "eval_loss": 4.678452014923096,
      "eval_runtime": 3.2849,
      "eval_samples_per_second": 0.304,
      "eval_steps_per_second": 0.304,
      "step": 264
    },
    {
      "epoch": 45.0,
      "learning_rate": 7.75e-06,
      "loss": 4.681,
      "step": 270
    },
    {
      "epoch": 45.0,
      "eval_accuracy": 0.3128054740957967,
      "eval_loss": 4.666133403778076,
      "eval_runtime": 3.4251,
      "eval_samples_per_second": 0.292,
      "eval_steps_per_second": 0.292,
      "step": 270
    },
    {
      "epoch": 46.0,
      "learning_rate": 7.7e-06,
      "loss": 4.6566,
      "step": 276
    },
    {
      "epoch": 46.0,
      "eval_accuracy": 0.31573802541544477,
      "eval_loss": 4.653237342834473,
      "eval_runtime": 3.1995,
      "eval_samples_per_second": 0.313,
      "eval_steps_per_second": 0.313,
      "step": 276
    },
    {
      "epoch": 47.0,
      "learning_rate": 7.650000000000001e-06,
      "loss": 4.632,
      "step": 282
    },
    {
      "epoch": 47.0,
      "eval_accuracy": 0.31573802541544477,
      "eval_loss": 4.636072635650635,
      "eval_runtime": 3.5333,
      "eval_samples_per_second": 0.283,
      "eval_steps_per_second": 0.283,
      "step": 282
    },
    {
      "epoch": 48.0,
      "learning_rate": 7.600000000000001e-06,
      "loss": 4.618,
      "step": 288
    },
    {
      "epoch": 48.0,
      "eval_accuracy": 0.3196480938416422,
      "eval_loss": 4.6162004470825195,
      "eval_runtime": 3.2635,
      "eval_samples_per_second": 0.306,
      "eval_steps_per_second": 0.306,
      "step": 288
    },
    {
      "epoch": 49.0,
      "learning_rate": 7.5500000000000006e-06,
      "loss": 4.5928,
      "step": 294
    },
    {
      "epoch": 49.0,
      "eval_accuracy": 0.32453567937438904,
      "eval_loss": 4.598696708679199,
      "eval_runtime": 3.5544,
      "eval_samples_per_second": 0.281,
      "eval_steps_per_second": 0.281,
      "step": 294
    },
    {
      "epoch": 50.0,
      "learning_rate": 7.500000000000001e-06,
      "loss": 4.5716,
      "step": 300
    },
    {
      "epoch": 50.0,
      "eval_accuracy": 0.32160312805474095,
      "eval_loss": 4.584750175476074,
      "eval_runtime": 3.2866,
      "eval_samples_per_second": 0.304,
      "eval_steps_per_second": 0.304,
      "step": 300
    },
    {
      "epoch": 51.0,
      "learning_rate": 7.450000000000001e-06,
      "loss": 4.5485,
      "step": 306
    },
    {
      "epoch": 51.0,
      "eval_accuracy": 0.32453567937438904,
      "eval_loss": 4.572133541107178,
      "eval_runtime": 4.8161,
      "eval_samples_per_second": 0.208,
      "eval_steps_per_second": 0.208,
      "step": 306
    },
    {
      "epoch": 52.0,
      "learning_rate": 7.4e-06,
      "loss": 4.5324,
      "step": 312
    },
    {
      "epoch": 52.0,
      "eval_accuracy": 0.3196480938416422,
      "eval_loss": 4.557907581329346,
      "eval_runtime": 3.1695,
      "eval_samples_per_second": 0.316,
      "eval_steps_per_second": 0.316,
      "step": 312
    },
    {
      "epoch": 53.0,
      "learning_rate": 7.350000000000001e-06,
      "loss": 4.5038,
      "step": 318
    },
    {
      "epoch": 53.0,
      "eval_accuracy": 0.3196480938416422,
      "eval_loss": 4.542334079742432,
      "eval_runtime": 3.3713,
      "eval_samples_per_second": 0.297,
      "eval_steps_per_second": 0.297,
      "step": 318
    },
    {
      "epoch": 54.0,
      "learning_rate": 7.3e-06,
      "loss": 4.4831,
      "step": 324
    },
    {
      "epoch": 54.0,
      "eval_accuracy": 0.32649071358748777,
      "eval_loss": 4.524044036865234,
      "eval_runtime": 3.1022,
      "eval_samples_per_second": 0.322,
      "eval_steps_per_second": 0.322,
      "step": 324
    },
    {
      "epoch": 55.0,
      "learning_rate": 7.25e-06,
      "loss": 4.4347,
      "step": 330
    },
    {
      "epoch": 55.0,
      "eval_accuracy": 0.3255131964809384,
      "eval_loss": 4.508722305297852,
      "eval_runtime": 3.0991,
      "eval_samples_per_second": 0.323,
      "eval_steps_per_second": 0.323,
      "step": 330
    },
    {
      "epoch": 56.0,
      "learning_rate": 7.2000000000000005e-06,
      "loss": 4.4218,
      "step": 336
    },
    {
      "epoch": 56.0,
      "eval_accuracy": 0.3255131964809384,
      "eval_loss": 4.484982013702393,
      "eval_runtime": 3.3635,
      "eval_samples_per_second": 0.297,
      "eval_steps_per_second": 0.297,
      "step": 336
    },
    {
      "epoch": 57.0,
      "learning_rate": 7.15e-06,
      "loss": 4.3939,
      "step": 342
    },
    {
      "epoch": 57.0,
      "eval_accuracy": 0.32746823069403713,
      "eval_loss": 4.479069709777832,
      "eval_runtime": 3.1386,
      "eval_samples_per_second": 0.319,
      "eval_steps_per_second": 0.319,
      "step": 342
    },
    {
      "epoch": 58.0,
      "learning_rate": 7.100000000000001e-06,
      "loss": 4.3766,
      "step": 348
    },
    {
      "epoch": 58.0,
      "eval_accuracy": 0.32649071358748777,
      "eval_loss": 4.463999271392822,
      "eval_runtime": 3.2574,
      "eval_samples_per_second": 0.307,
      "eval_steps_per_second": 0.307,
      "step": 348
    },
    {
      "epoch": 59.0,
      "learning_rate": 7.05e-06,
      "loss": 4.3472,
      "step": 354
    },
    {
      "epoch": 59.0,
      "eval_accuracy": 0.32746823069403713,
      "eval_loss": 4.4470648765563965,
      "eval_runtime": 3.1788,
      "eval_samples_per_second": 0.315,
      "eval_steps_per_second": 0.315,
      "step": 354
    },
    {
      "epoch": 60.0,
      "learning_rate": 7e-06,
      "loss": 4.3241,
      "step": 360
    },
    {
      "epoch": 60.0,
      "eval_accuracy": 0.32746823069403713,
      "eval_loss": 4.433444499969482,
      "eval_runtime": 3.1305,
      "eval_samples_per_second": 0.319,
      "eval_steps_per_second": 0.319,
      "step": 360
    },
    {
      "epoch": 61.0,
      "learning_rate": 6.95e-06,
      "loss": 4.2919,
      "step": 366
    },
    {
      "epoch": 61.0,
      "eval_accuracy": 0.3304007820136852,
      "eval_loss": 4.42958927154541,
      "eval_runtime": 3.1351,
      "eval_samples_per_second": 0.319,
      "eval_steps_per_second": 0.319,
      "step": 366
    },
    {
      "epoch": 62.0,
      "learning_rate": 6.9e-06,
      "loss": 4.2678,
      "step": 372
    },
    {
      "epoch": 62.0,
      "eval_accuracy": 0.3343108504398827,
      "eval_loss": 4.428121566772461,
      "eval_runtime": 3.0787,
      "eval_samples_per_second": 0.325,
      "eval_steps_per_second": 0.325,
      "step": 372
    },
    {
      "epoch": 63.0,
      "learning_rate": 6.850000000000001e-06,
      "loss": 4.2515,
      "step": 378
    },
    {
      "epoch": 63.0,
      "eval_accuracy": 0.33724340175953077,
      "eval_loss": 4.412038803100586,
      "eval_runtime": 3.3246,
      "eval_samples_per_second": 0.301,
      "eval_steps_per_second": 0.301,
      "step": 378
    },
    {
      "epoch": 64.0,
      "learning_rate": 6.800000000000001e-06,
      "loss": 4.2244,
      "step": 384
    },
    {
      "epoch": 64.0,
      "eval_accuracy": 0.3343108504398827,
      "eval_loss": 4.403759002685547,
      "eval_runtime": 3.1858,
      "eval_samples_per_second": 0.314,
      "eval_steps_per_second": 0.314,
      "step": 384
    },
    {
      "epoch": 65.0,
      "learning_rate": 6.750000000000001e-06,
      "loss": 4.2129,
      "step": 390
    },
    {
      "epoch": 65.0,
      "eval_accuracy": 0.3391984359726295,
      "eval_loss": 4.382594585418701,
      "eval_runtime": 3.1111,
      "eval_samples_per_second": 0.321,
      "eval_steps_per_second": 0.321,
      "step": 390
    },
    {
      "epoch": 66.0,
      "learning_rate": 6.700000000000001e-06,
      "loss": 4.1882,
      "step": 396
    },
    {
      "epoch": 66.0,
      "eval_accuracy": 0.33724340175953077,
      "eval_loss": 4.383427143096924,
      "eval_runtime": 3.2114,
      "eval_samples_per_second": 0.311,
      "eval_steps_per_second": 0.311,
      "step": 396
    },
    {
      "epoch": 67.0,
      "learning_rate": 6.650000000000001e-06,
      "loss": 4.1503,
      "step": 402
    },
    {
      "epoch": 67.0,
      "eval_accuracy": 0.33724340175953077,
      "eval_loss": 4.37380838394165,
      "eval_runtime": 3.1592,
      "eval_samples_per_second": 0.317,
      "eval_steps_per_second": 0.317,
      "step": 402
    },
    {
      "epoch": 68.0,
      "learning_rate": 6.600000000000001e-06,
      "loss": 4.1398,
      "step": 408
    },
    {
      "epoch": 68.0,
      "eval_accuracy": 0.33724340175953077,
      "eval_loss": 4.359556674957275,
      "eval_runtime": 3.2732,
      "eval_samples_per_second": 0.306,
      "eval_steps_per_second": 0.306,
      "step": 408
    },
    {
      "epoch": 69.0,
      "learning_rate": 6.550000000000001e-06,
      "loss": 4.115,
      "step": 414
    },
    {
      "epoch": 69.0,
      "eval_accuracy": 0.3411534701857282,
      "eval_loss": 4.337615966796875,
      "eval_runtime": 3.5266,
      "eval_samples_per_second": 0.284,
      "eval_steps_per_second": 0.284,
      "step": 414
    },
    {
      "epoch": 70.0,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 4.1052,
      "step": 420
    },
    {
      "epoch": 70.0,
      "eval_accuracy": 0.3411534701857282,
      "eval_loss": 4.333048343658447,
      "eval_runtime": 3.1824,
      "eval_samples_per_second": 0.314,
      "eval_steps_per_second": 0.314,
      "step": 420
    },
    {
      "epoch": 71.0,
      "learning_rate": 6.450000000000001e-06,
      "loss": 4.0932,
      "step": 426
    },
    {
      "epoch": 71.0,
      "eval_accuracy": 0.3411534701857282,
      "eval_loss": 4.329537391662598,
      "eval_runtime": 3.1403,
      "eval_samples_per_second": 0.318,
      "eval_steps_per_second": 0.318,
      "step": 426
    },
    {
      "epoch": 72.0,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 4.0573,
      "step": 432
    },
    {
      "epoch": 72.0,
      "eval_accuracy": 0.3411534701857282,
      "eval_loss": 4.311087131500244,
      "eval_runtime": 3.3734,
      "eval_samples_per_second": 0.296,
      "eval_steps_per_second": 0.296,
      "step": 432
    },
    {
      "epoch": 73.0,
      "learning_rate": 6.35e-06,
      "loss": 4.0449,
      "step": 438
    },
    {
      "epoch": 73.0,
      "eval_accuracy": 0.34408602150537637,
      "eval_loss": 4.304787635803223,
      "eval_runtime": 3.3551,
      "eval_samples_per_second": 0.298,
      "eval_steps_per_second": 0.298,
      "step": 438
    },
    {
      "epoch": 74.0,
      "learning_rate": 6.300000000000001e-06,
      "loss": 4.0165,
      "step": 444
    },
    {
      "epoch": 74.0,
      "eval_accuracy": 0.3460410557184751,
      "eval_loss": 4.293551445007324,
      "eval_runtime": 3.1253,
      "eval_samples_per_second": 0.32,
      "eval_steps_per_second": 0.32,
      "step": 444
    },
    {
      "epoch": 75.0,
      "learning_rate": 6.25e-06,
      "loss": 3.9936,
      "step": 450
    },
    {
      "epoch": 75.0,
      "eval_accuracy": 0.3509286412512219,
      "eval_loss": 4.281482696533203,
      "eval_runtime": 3.2326,
      "eval_samples_per_second": 0.309,
      "eval_steps_per_second": 0.309,
      "step": 450
    },
    {
      "epoch": 76.0,
      "learning_rate": 6.200000000000001e-06,
      "loss": 3.967,
      "step": 456
    },
    {
      "epoch": 76.0,
      "eval_accuracy": 0.35386119257087,
      "eval_loss": 4.268586158752441,
      "eval_runtime": 3.1926,
      "eval_samples_per_second": 0.313,
      "eval_steps_per_second": 0.313,
      "step": 456
    },
    {
      "epoch": 77.0,
      "learning_rate": 6.15e-06,
      "loss": 3.9524,
      "step": 462
    },
    {
      "epoch": 77.0,
      "eval_accuracy": 0.3509286412512219,
      "eval_loss": 4.269726276397705,
      "eval_runtime": 3.0654,
      "eval_samples_per_second": 0.326,
      "eval_steps_per_second": 0.326,
      "step": 462
    },
    {
      "epoch": 78.0,
      "learning_rate": 6.1e-06,
      "loss": 3.9287,
      "step": 468
    },
    {
      "epoch": 78.0,
      "eval_accuracy": 0.35288367546432065,
      "eval_loss": 4.254611015319824,
      "eval_runtime": 3.2483,
      "eval_samples_per_second": 0.308,
      "eval_steps_per_second": 0.308,
      "step": 468
    },
    {
      "epoch": 79.0,
      "learning_rate": 6.0500000000000005e-06,
      "loss": 3.9092,
      "step": 474
    },
    {
      "epoch": 79.0,
      "eval_accuracy": 0.35386119257087,
      "eval_loss": 4.248389720916748,
      "eval_runtime": 3.279,
      "eval_samples_per_second": 0.305,
      "eval_steps_per_second": 0.305,
      "step": 474
    },
    {
      "epoch": 80.0,
      "learning_rate": 6e-06,
      "loss": 3.8907,
      "step": 480
    },
    {
      "epoch": 80.0,
      "eval_accuracy": 0.35386119257087,
      "eval_loss": 4.242007255554199,
      "eval_runtime": 3.1903,
      "eval_samples_per_second": 0.313,
      "eval_steps_per_second": 0.313,
      "step": 480
    },
    {
      "epoch": 81.0,
      "learning_rate": 5.950000000000001e-06,
      "loss": 3.8704,
      "step": 486
    },
    {
      "epoch": 81.0,
      "eval_accuracy": 0.35288367546432065,
      "eval_loss": 4.241766929626465,
      "eval_runtime": 3.0667,
      "eval_samples_per_second": 0.326,
      "eval_steps_per_second": 0.326,
      "step": 486
    },
    {
      "epoch": 82.0,
      "learning_rate": 5.9e-06,
      "loss": 3.8499,
      "step": 492
    },
    {
      "epoch": 82.0,
      "eval_accuracy": 0.3548387096774194,
      "eval_loss": 4.226541519165039,
      "eval_runtime": 3.1032,
      "eval_samples_per_second": 0.322,
      "eval_steps_per_second": 0.322,
      "step": 492
    },
    {
      "epoch": 83.0,
      "learning_rate": 5.85e-06,
      "loss": 3.8325,
      "step": 498
    },
    {
      "epoch": 83.0,
      "eval_accuracy": 0.3548387096774194,
      "eval_loss": 4.208946704864502,
      "eval_runtime": 3.0764,
      "eval_samples_per_second": 0.325,
      "eval_steps_per_second": 0.325,
      "step": 498
    },
    {
      "epoch": 84.0,
      "learning_rate": 5.8e-06,
      "loss": 3.8024,
      "step": 504
    },
    {
      "epoch": 84.0,
      "eval_accuracy": 0.3567937438905181,
      "eval_loss": 4.205758094787598,
      "eval_runtime": 3.0957,
      "eval_samples_per_second": 0.323,
      "eval_steps_per_second": 0.323,
      "step": 504
    },
    {
      "epoch": 85.0,
      "learning_rate": 5.75e-06,
      "loss": 3.8058,
      "step": 510
    },
    {
      "epoch": 85.0,
      "eval_accuracy": 0.35581622678396874,
      "eval_loss": 4.203883647918701,
      "eval_runtime": 3.1095,
      "eval_samples_per_second": 0.322,
      "eval_steps_per_second": 0.322,
      "step": 510
    },
    {
      "epoch": 86.0,
      "learning_rate": 5.7e-06,
      "loss": 3.7888,
      "step": 516
    },
    {
      "epoch": 86.0,
      "eval_accuracy": 0.35777126099706746,
      "eval_loss": 4.190591335296631,
      "eval_runtime": 3.1,
      "eval_samples_per_second": 0.323,
      "eval_steps_per_second": 0.323,
      "step": 516
    },
    {
      "epoch": 87.0,
      "learning_rate": 5.65e-06,
      "loss": 3.7622,
      "step": 522
    },
    {
      "epoch": 87.0,
      "eval_accuracy": 0.3616813294232649,
      "eval_loss": 4.179159164428711,
      "eval_runtime": 3.1107,
      "eval_samples_per_second": 0.321,
      "eval_steps_per_second": 0.321,
      "step": 522
    },
    {
      "epoch": 88.0,
      "learning_rate": 5.600000000000001e-06,
      "loss": 3.746,
      "step": 528
    },
    {
      "epoch": 88.0,
      "eval_accuracy": 0.35777126099706746,
      "eval_loss": 4.18191385269165,
      "eval_runtime": 3.1176,
      "eval_samples_per_second": 0.321,
      "eval_steps_per_second": 0.321,
      "step": 528
    },
    {
      "epoch": 89.0,
      "learning_rate": 5.550000000000001e-06,
      "loss": 3.7196,
      "step": 534
    },
    {
      "epoch": 89.0,
      "eval_accuracy": 0.3597262952101662,
      "eval_loss": 4.178895473480225,
      "eval_runtime": 3.3763,
      "eval_samples_per_second": 0.296,
      "eval_steps_per_second": 0.296,
      "step": 534
    },
    {
      "epoch": 90.0,
      "learning_rate": 5.500000000000001e-06,
      "loss": 3.7046,
      "step": 540
    },
    {
      "epoch": 90.0,
      "eval_accuracy": 0.36070381231671556,
      "eval_loss": 4.160989284515381,
      "eval_runtime": 3.3055,
      "eval_samples_per_second": 0.303,
      "eval_steps_per_second": 0.303,
      "step": 540
    },
    {
      "epoch": 91.0,
      "learning_rate": 5.450000000000001e-06,
      "loss": 3.7078,
      "step": 546
    },
    {
      "epoch": 91.0,
      "eval_accuracy": 0.36070381231671556,
      "eval_loss": 4.151462078094482,
      "eval_runtime": 3.0842,
      "eval_samples_per_second": 0.324,
      "eval_steps_per_second": 0.324,
      "step": 546
    },
    {
      "epoch": 92.0,
      "learning_rate": 5.400000000000001e-06,
      "loss": 3.6687,
      "step": 552
    },
    {
      "epoch": 92.0,
      "eval_accuracy": 0.36070381231671556,
      "eval_loss": 4.17518424987793,
      "eval_runtime": 3.097,
      "eval_samples_per_second": 0.323,
      "eval_steps_per_second": 0.323,
      "step": 552
    },
    {
      "epoch": 93.0,
      "learning_rate": 5.3500000000000004e-06,
      "loss": 3.6559,
      "step": 558
    },
    {
      "epoch": 93.0,
      "eval_accuracy": 0.36363636363636365,
      "eval_loss": 4.128724575042725,
      "eval_runtime": 3.2229,
      "eval_samples_per_second": 0.31,
      "eval_steps_per_second": 0.31,
      "step": 558
    },
    {
      "epoch": 94.0,
      "learning_rate": 5.300000000000001e-06,
      "loss": 3.6401,
      "step": 564
    },
    {
      "epoch": 94.0,
      "eval_accuracy": 0.364613880742913,
      "eval_loss": 4.156914234161377,
      "eval_runtime": 3.0553,
      "eval_samples_per_second": 0.327,
      "eval_steps_per_second": 0.327,
      "step": 564
    },
    {
      "epoch": 95.0,
      "learning_rate": 5.2500000000000006e-06,
      "loss": 3.6281,
      "step": 570
    },
    {
      "epoch": 95.0,
      "eval_accuracy": 0.3626588465298143,
      "eval_loss": 4.123438358306885,
      "eval_runtime": 3.127,
      "eval_samples_per_second": 0.32,
      "eval_steps_per_second": 0.32,
      "step": 570
    },
    {
      "epoch": 96.0,
      "learning_rate": 5.2e-06,
      "loss": 3.5978,
      "step": 576
    },
    {
      "epoch": 96.0,
      "eval_accuracy": 0.36950146627565983,
      "eval_loss": 4.1269941329956055,
      "eval_runtime": 3.0893,
      "eval_samples_per_second": 0.324,
      "eval_steps_per_second": 0.324,
      "step": 576
    },
    {
      "epoch": 97.0,
      "learning_rate": 5.150000000000001e-06,
      "loss": 3.5951,
      "step": 582
    },
    {
      "epoch": 97.0,
      "eval_accuracy": 0.364613880742913,
      "eval_loss": 4.118751049041748,
      "eval_runtime": 3.148,
      "eval_samples_per_second": 0.318,
      "eval_steps_per_second": 0.318,
      "step": 582
    },
    {
      "epoch": 98.0,
      "learning_rate": 5.1e-06,
      "loss": 3.5679,
      "step": 588
    },
    {
      "epoch": 98.0,
      "eval_accuracy": 0.36852394916911047,
      "eval_loss": 4.128164291381836,
      "eval_runtime": 3.0607,
      "eval_samples_per_second": 0.327,
      "eval_steps_per_second": 0.327,
      "step": 588
    },
    {
      "epoch": 99.0,
      "learning_rate": 5.050000000000001e-06,
      "loss": 3.5618,
      "step": 594
    },
    {
      "epoch": 99.0,
      "eval_accuracy": 0.364613880742913,
      "eval_loss": 4.1089253425598145,
      "eval_runtime": 3.0857,
      "eval_samples_per_second": 0.324,
      "eval_steps_per_second": 0.324,
      "step": 594
    },
    {
      "epoch": 100.0,
      "learning_rate": 5e-06,
      "loss": 3.5404,
      "step": 600
    },
    {
      "epoch": 100.0,
      "eval_accuracy": 0.36950146627565983,
      "eval_loss": 4.1090216636657715,
      "eval_runtime": 3.1735,
      "eval_samples_per_second": 0.315,
      "eval_steps_per_second": 0.315,
      "step": 600
    },
    {
      "epoch": 101.0,
      "learning_rate": 4.95e-06,
      "loss": 3.5255,
      "step": 606
    },
    {
      "epoch": 101.0,
      "eval_accuracy": 0.364613880742913,
      "eval_loss": 4.103869915008545,
      "eval_runtime": 3.1307,
      "eval_samples_per_second": 0.319,
      "eval_steps_per_second": 0.319,
      "step": 606
    },
    {
      "epoch": 102.0,
      "learning_rate": 4.9000000000000005e-06,
      "loss": 3.5111,
      "step": 612
    },
    {
      "epoch": 102.0,
      "eval_accuracy": 0.36950146627565983,
      "eval_loss": 4.100970268249512,
      "eval_runtime": 3.116,
      "eval_samples_per_second": 0.321,
      "eval_steps_per_second": 0.321,
      "step": 612
    },
    {
      "epoch": 103.0,
      "learning_rate": 4.85e-06,
      "loss": 3.5015,
      "step": 618
    },
    {
      "epoch": 103.0,
      "eval_accuracy": 0.3704789833822092,
      "eval_loss": 4.088879585266113,
      "eval_runtime": 3.3403,
      "eval_samples_per_second": 0.299,
      "eval_steps_per_second": 0.299,
      "step": 618
    },
    {
      "epoch": 104.0,
      "learning_rate": 4.800000000000001e-06,
      "loss": 3.493,
      "step": 624
    },
    {
      "epoch": 104.0,
      "eval_accuracy": 0.3704789833822092,
      "eval_loss": 4.082551956176758,
      "eval_runtime": 3.3467,
      "eval_samples_per_second": 0.299,
      "eval_steps_per_second": 0.299,
      "step": 624
    },
    {
      "epoch": 105.0,
      "learning_rate": 4.75e-06,
      "loss": 3.5643,
      "step": 630
    },
    {
      "epoch": 105.0,
      "eval_accuracy": 0.375366568914956,
      "eval_loss": 4.091529846191406,
      "eval_runtime": 3.5246,
      "eval_samples_per_second": 0.284,
      "eval_steps_per_second": 0.284,
      "step": 630
    },
    {
      "epoch": 106.0,
      "learning_rate": 4.7e-06,
      "loss": 3.4543,
      "step": 636
    },
    {
      "epoch": 106.0,
      "eval_accuracy": 0.3724340175953079,
      "eval_loss": 4.091198921203613,
      "eval_runtime": 3.1421,
      "eval_samples_per_second": 0.318,
      "eval_steps_per_second": 0.318,
      "step": 636
    },
    {
      "epoch": 107.0,
      "learning_rate": 4.65e-06,
      "loss": 3.4517,
      "step": 642
    },
    {
      "epoch": 107.0,
      "eval_accuracy": 0.375366568914956,
      "eval_loss": 4.084360599517822,
      "eval_runtime": 3.5262,
      "eval_samples_per_second": 0.284,
      "eval_steps_per_second": 0.284,
      "step": 642
    },
    {
      "epoch": 108.0,
      "learning_rate": 4.600000000000001e-06,
      "loss": 3.4387,
      "step": 648
    },
    {
      "epoch": 108.0,
      "eval_accuracy": 0.375366568914956,
      "eval_loss": 4.066359996795654,
      "eval_runtime": 3.1772,
      "eval_samples_per_second": 0.315,
      "eval_steps_per_second": 0.315,
      "step": 648
    },
    {
      "epoch": 109.0,
      "learning_rate": 4.5500000000000005e-06,
      "loss": 3.4274,
      "step": 654
    },
    {
      "epoch": 109.0,
      "eval_accuracy": 0.3763440860215054,
      "eval_loss": 4.088470458984375,
      "eval_runtime": 3.106,
      "eval_samples_per_second": 0.322,
      "eval_steps_per_second": 0.322,
      "step": 654
    },
    {
      "epoch": 110.0,
      "learning_rate": 4.5e-06,
      "loss": 3.4241,
      "step": 660
    },
    {
      "epoch": 110.0,
      "eval_accuracy": 0.37927663734115347,
      "eval_loss": 4.058298587799072,
      "eval_runtime": 3.0985,
      "eval_samples_per_second": 0.323,
      "eval_steps_per_second": 0.323,
      "step": 660
    },
    {
      "epoch": 111.0,
      "learning_rate": 4.450000000000001e-06,
      "loss": 3.4016,
      "step": 666
    },
    {
      "epoch": 111.0,
      "eval_accuracy": 0.38025415444770283,
      "eval_loss": 4.0626630783081055,
      "eval_runtime": 3.5144,
      "eval_samples_per_second": 0.285,
      "eval_steps_per_second": 0.285,
      "step": 666
    },
    {
      "epoch": 112.0,
      "learning_rate": 4.4e-06,
      "loss": 3.383,
      "step": 672
    },
    {
      "epoch": 112.0,
      "eval_accuracy": 0.3812316715542522,
      "eval_loss": 4.062616348266602,
      "eval_runtime": 3.2502,
      "eval_samples_per_second": 0.308,
      "eval_steps_per_second": 0.308,
      "step": 672
    },
    {
      "epoch": 113.0,
      "learning_rate": 4.350000000000001e-06,
      "loss": 3.3709,
      "step": 678
    },
    {
      "epoch": 113.0,
      "eval_accuracy": 0.3870967741935484,
      "eval_loss": 4.0414323806762695,
      "eval_runtime": 3.4468,
      "eval_samples_per_second": 0.29,
      "eval_steps_per_second": 0.29,
      "step": 678
    },
    {
      "epoch": 114.0,
      "learning_rate": 4.3e-06,
      "loss": 3.3646,
      "step": 684
    },
    {
      "epoch": 114.0,
      "eval_accuracy": 0.38220918866080156,
      "eval_loss": 4.0561604499816895,
      "eval_runtime": 3.0962,
      "eval_samples_per_second": 0.323,
      "eval_steps_per_second": 0.323,
      "step": 684
    },
    {
      "epoch": 115.0,
      "learning_rate": 4.25e-06,
      "loss": 3.3456,
      "step": 690
    },
    {
      "epoch": 115.0,
      "eval_accuracy": 0.386119257086999,
      "eval_loss": 4.036128044128418,
      "eval_runtime": 3.3227,
      "eval_samples_per_second": 0.301,
      "eval_steps_per_second": 0.301,
      "step": 690
    },
    {
      "epoch": 116.0,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 3.3369,
      "step": 696
    },
    {
      "epoch": 116.0,
      "eval_accuracy": 0.38514173998044965,
      "eval_loss": 4.052359104156494,
      "eval_runtime": 3.0856,
      "eval_samples_per_second": 0.324,
      "eval_steps_per_second": 0.324,
      "step": 696
    },
    {
      "epoch": 117.0,
      "learning_rate": 4.15e-06,
      "loss": 3.3136,
      "step": 702
    },
    {
      "epoch": 117.0,
      "eval_accuracy": 0.3841642228739003,
      "eval_loss": 4.042443752288818,
      "eval_runtime": 3.358,
      "eval_samples_per_second": 0.298,
      "eval_steps_per_second": 0.298,
      "step": 702
    },
    {
      "epoch": 118.0,
      "learning_rate": 4.1e-06,
      "loss": 3.307,
      "step": 708
    },
    {
      "epoch": 118.0,
      "eval_accuracy": 0.386119257086999,
      "eval_loss": 4.047730922698975,
      "eval_runtime": 3.2939,
      "eval_samples_per_second": 0.304,
      "eval_steps_per_second": 0.304,
      "step": 708
    },
    {
      "epoch": 119.0,
      "learning_rate": 4.05e-06,
      "loss": 3.2954,
      "step": 714
    },
    {
      "epoch": 119.0,
      "eval_accuracy": 0.38514173998044965,
      "eval_loss": 4.028741359710693,
      "eval_runtime": 3.1849,
      "eval_samples_per_second": 0.314,
      "eval_steps_per_second": 0.314,
      "step": 714
    },
    {
      "epoch": 120.0,
      "learning_rate": 4.000000000000001e-06,
      "loss": 3.2887,
      "step": 720
    },
    {
      "epoch": 120.0,
      "eval_accuracy": 0.39002932551319647,
      "eval_loss": 4.0391716957092285,
      "eval_runtime": 3.373,
      "eval_samples_per_second": 0.296,
      "eval_steps_per_second": 0.296,
      "step": 720
    },
    {
      "epoch": 121.0,
      "learning_rate": 3.95e-06,
      "loss": 3.2776,
      "step": 726
    },
    {
      "epoch": 121.0,
      "eval_accuracy": 0.39100684261974583,
      "eval_loss": 4.019059658050537,
      "eval_runtime": 3.0671,
      "eval_samples_per_second": 0.326,
      "eval_steps_per_second": 0.326,
      "step": 726
    },
    {
      "epoch": 122.0,
      "learning_rate": 3.900000000000001e-06,
      "loss": 3.2527,
      "step": 732
    },
    {
      "epoch": 122.0,
      "eval_accuracy": 0.39100684261974583,
      "eval_loss": 4.03394889831543,
      "eval_runtime": 3.608,
      "eval_samples_per_second": 0.277,
      "eval_steps_per_second": 0.277,
      "step": 732
    },
    {
      "epoch": 123.0,
      "learning_rate": 3.85e-06,
      "loss": 3.259,
      "step": 738
    },
    {
      "epoch": 123.0,
      "eval_accuracy": 0.39296187683284456,
      "eval_loss": 4.006428241729736,
      "eval_runtime": 3.37,
      "eval_samples_per_second": 0.297,
      "eval_steps_per_second": 0.297,
      "step": 738
    },
    {
      "epoch": 124.0,
      "learning_rate": 3.8000000000000005e-06,
      "loss": 3.2559,
      "step": 744
    },
    {
      "epoch": 124.0,
      "eval_accuracy": 0.38807429130009774,
      "eval_loss": 4.028494834899902,
      "eval_runtime": 3.4043,
      "eval_samples_per_second": 0.294,
      "eval_steps_per_second": 0.294,
      "step": 744
    },
    {
      "epoch": 125.0,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 3.2335,
      "step": 750
    },
    {
      "epoch": 125.0,
      "eval_accuracy": 0.39296187683284456,
      "eval_loss": 4.015052795410156,
      "eval_runtime": 3.1425,
      "eval_samples_per_second": 0.318,
      "eval_steps_per_second": 0.318,
      "step": 750
    },
    {
      "epoch": 126.0,
      "learning_rate": 3.7e-06,
      "loss": 3.2318,
      "step": 756
    },
    {
      "epoch": 126.0,
      "eval_accuracy": 0.39002932551319647,
      "eval_loss": 4.027667999267578,
      "eval_runtime": 3.3897,
      "eval_samples_per_second": 0.295,
      "eval_steps_per_second": 0.295,
      "step": 756
    },
    {
      "epoch": 127.0,
      "learning_rate": 3.65e-06,
      "loss": 3.2266,
      "step": 762
    },
    {
      "epoch": 127.0,
      "eval_accuracy": 0.3978494623655914,
      "eval_loss": 3.992929697036743,
      "eval_runtime": 3.1289,
      "eval_samples_per_second": 0.32,
      "eval_steps_per_second": 0.32,
      "step": 762
    },
    {
      "epoch": 128.0,
      "learning_rate": 3.6000000000000003e-06,
      "loss": 3.2051,
      "step": 768
    },
    {
      "epoch": 128.0,
      "eval_accuracy": 0.3978494623655914,
      "eval_loss": 3.9944605827331543,
      "eval_runtime": 3.5639,
      "eval_samples_per_second": 0.281,
      "eval_steps_per_second": 0.281,
      "step": 768
    },
    {
      "epoch": 129.0,
      "learning_rate": 3.5500000000000003e-06,
      "loss": 3.2009,
      "step": 774
    },
    {
      "epoch": 129.0,
      "eval_accuracy": 0.39296187683284456,
      "eval_loss": 4.02908182144165,
      "eval_runtime": 3.1278,
      "eval_samples_per_second": 0.32,
      "eval_steps_per_second": 0.32,
      "step": 774
    },
    {
      "epoch": 130.0,
      "learning_rate": 3.5e-06,
      "loss": 3.1791,
      "step": 780
    },
    {
      "epoch": 130.0,
      "eval_accuracy": 0.39296187683284456,
      "eval_loss": 3.9955568313598633,
      "eval_runtime": 3.1553,
      "eval_samples_per_second": 0.317,
      "eval_steps_per_second": 0.317,
      "step": 780
    },
    {
      "epoch": 131.0,
      "learning_rate": 3.45e-06,
      "loss": 3.1759,
      "step": 786
    },
    {
      "epoch": 131.0,
      "eval_accuracy": 0.396871945259042,
      "eval_loss": 4.001156806945801,
      "eval_runtime": 3.1012,
      "eval_samples_per_second": 0.322,
      "eval_steps_per_second": 0.322,
      "step": 786
    },
    {
      "epoch": 132.0,
      "learning_rate": 3.4000000000000005e-06,
      "loss": 3.1622,
      "step": 792
    },
    {
      "epoch": 132.0,
      "eval_accuracy": 0.3949169110459433,
      "eval_loss": 4.010651111602783,
      "eval_runtime": 3.2622,
      "eval_samples_per_second": 0.307,
      "eval_steps_per_second": 0.307,
      "step": 792
    },
    {
      "epoch": 133.0,
      "learning_rate": 3.3500000000000005e-06,
      "loss": 3.1559,
      "step": 798
    },
    {
      "epoch": 133.0,
      "eval_accuracy": 0.3939393939393939,
      "eval_loss": 4.009001731872559,
      "eval_runtime": 3.147,
      "eval_samples_per_second": 0.318,
      "eval_steps_per_second": 0.318,
      "step": 798
    },
    {
      "epoch": 134.0,
      "learning_rate": 3.3000000000000006e-06,
      "loss": 3.1521,
      "step": 804
    },
    {
      "epoch": 134.0,
      "eval_accuracy": 0.39100684261974583,
      "eval_loss": 4.00282621383667,
      "eval_runtime": 3.2649,
      "eval_samples_per_second": 0.306,
      "eval_steps_per_second": 0.306,
      "step": 804
    },
    {
      "epoch": 135.0,
      "learning_rate": 3.2500000000000002e-06,
      "loss": 3.1353,
      "step": 810
    },
    {
      "epoch": 135.0,
      "eval_accuracy": 0.3939393939393939,
      "eval_loss": 4.0033183097839355,
      "eval_runtime": 3.0829,
      "eval_samples_per_second": 0.324,
      "eval_steps_per_second": 0.324,
      "step": 810
    },
    {
      "epoch": 136.0,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 3.1427,
      "step": 816
    },
    {
      "epoch": 136.0,
      "eval_accuracy": 0.3939393939393939,
      "eval_loss": 3.999516248703003,
      "eval_runtime": 3.3725,
      "eval_samples_per_second": 0.297,
      "eval_steps_per_second": 0.297,
      "step": 816
    },
    {
      "epoch": 137.0,
      "learning_rate": 3.1500000000000003e-06,
      "loss": 3.1276,
      "step": 822
    },
    {
      "epoch": 137.0,
      "eval_accuracy": 0.3919843597262952,
      "eval_loss": 3.9962964057922363,
      "eval_runtime": 3.3881,
      "eval_samples_per_second": 0.295,
      "eval_steps_per_second": 0.295,
      "step": 822
    },
    {
      "epoch": 138.0,
      "learning_rate": 3.1000000000000004e-06,
      "loss": 3.1228,
      "step": 828
    },
    {
      "epoch": 138.0,
      "eval_accuracy": 0.3978494623655914,
      "eval_loss": 3.9996395111083984,
      "eval_runtime": 3.1889,
      "eval_samples_per_second": 0.314,
      "eval_steps_per_second": 0.314,
      "step": 828
    },
    {
      "epoch": 139.0,
      "learning_rate": 3.05e-06,
      "loss": 3.1039,
      "step": 834
    },
    {
      "epoch": 139.0,
      "eval_accuracy": 0.39882697947214074,
      "eval_loss": 3.992779493331909,
      "eval_runtime": 3.1905,
      "eval_samples_per_second": 0.313,
      "eval_steps_per_second": 0.313,
      "step": 834
    },
    {
      "epoch": 140.0,
      "learning_rate": 3e-06,
      "loss": 3.097,
      "step": 840
    },
    {
      "epoch": 140.0,
      "eval_accuracy": 0.396871945259042,
      "eval_loss": 3.9969444274902344,
      "eval_runtime": 3.2194,
      "eval_samples_per_second": 0.311,
      "eval_steps_per_second": 0.311,
      "step": 840
    },
    {
      "epoch": 141.0,
      "learning_rate": 2.95e-06,
      "loss": 3.083,
      "step": 846
    },
    {
      "epoch": 141.0,
      "eval_accuracy": 0.3949169110459433,
      "eval_loss": 3.991848945617676,
      "eval_runtime": 3.2252,
      "eval_samples_per_second": 0.31,
      "eval_steps_per_second": 0.31,
      "step": 846
    },
    {
      "epoch": 142.0,
      "learning_rate": 2.9e-06,
      "loss": 3.0844,
      "step": 852
    },
    {
      "epoch": 142.0,
      "eval_accuracy": 0.396871945259042,
      "eval_loss": 3.9899985790252686,
      "eval_runtime": 3.3224,
      "eval_samples_per_second": 0.301,
      "eval_steps_per_second": 0.301,
      "step": 852
    },
    {
      "epoch": 143.0,
      "learning_rate": 2.85e-06,
      "loss": 3.077,
      "step": 858
    },
    {
      "epoch": 143.0,
      "eval_accuracy": 0.39589442815249265,
      "eval_loss": 3.981245756149292,
      "eval_runtime": 3.3063,
      "eval_samples_per_second": 0.302,
      "eval_steps_per_second": 0.302,
      "step": 858
    },
    {
      "epoch": 144.0,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 3.0601,
      "step": 864
    },
    {
      "epoch": 144.0,
      "eval_accuracy": 0.39589442815249265,
      "eval_loss": 3.9948182106018066,
      "eval_runtime": 3.5418,
      "eval_samples_per_second": 0.282,
      "eval_steps_per_second": 0.282,
      "step": 864
    },
    {
      "epoch": 145.0,
      "learning_rate": 2.7500000000000004e-06,
      "loss": 3.0669,
      "step": 870
    },
    {
      "epoch": 145.0,
      "eval_accuracy": 0.39589442815249265,
      "eval_loss": 3.9938085079193115,
      "eval_runtime": 3.3604,
      "eval_samples_per_second": 0.298,
      "eval_steps_per_second": 0.298,
      "step": 870
    },
    {
      "epoch": 146.0,
      "learning_rate": 2.7000000000000004e-06,
      "loss": 3.0515,
      "step": 876
    },
    {
      "epoch": 146.0,
      "eval_accuracy": 0.3978494623655914,
      "eval_loss": 3.9895379543304443,
      "eval_runtime": 3.1997,
      "eval_samples_per_second": 0.313,
      "eval_steps_per_second": 0.313,
      "step": 876
    },
    {
      "epoch": 147.0,
      "learning_rate": 2.6500000000000005e-06,
      "loss": 3.0405,
      "step": 882
    },
    {
      "epoch": 147.0,
      "eval_accuracy": 0.39882697947214074,
      "eval_loss": 3.9802987575531006,
      "eval_runtime": 3.2702,
      "eval_samples_per_second": 0.306,
      "eval_steps_per_second": 0.306,
      "step": 882
    },
    {
      "epoch": 148.0,
      "learning_rate": 2.6e-06,
      "loss": 3.029,
      "step": 888
    },
    {
      "epoch": 148.0,
      "eval_accuracy": 0.396871945259042,
      "eval_loss": 3.985629081726074,
      "eval_runtime": 3.064,
      "eval_samples_per_second": 0.326,
      "eval_steps_per_second": 0.326,
      "step": 888
    },
    {
      "epoch": 149.0,
      "learning_rate": 2.55e-06,
      "loss": 3.0342,
      "step": 894
    },
    {
      "epoch": 149.0,
      "eval_accuracy": 0.396871945259042,
      "eval_loss": 3.982790231704712,
      "eval_runtime": 3.2872,
      "eval_samples_per_second": 0.304,
      "eval_steps_per_second": 0.304,
      "step": 894
    },
    {
      "epoch": 150.0,
      "learning_rate": 2.5e-06,
      "loss": 3.0137,
      "step": 900
    },
    {
      "epoch": 150.0,
      "eval_accuracy": 0.3978494623655914,
      "eval_loss": 3.997673273086548,
      "eval_runtime": 3.3166,
      "eval_samples_per_second": 0.302,
      "eval_steps_per_second": 0.302,
      "step": 900
    },
    {
      "epoch": 151.0,
      "learning_rate": 2.4500000000000003e-06,
      "loss": 3.0277,
      "step": 906
    },
    {
      "epoch": 151.0,
      "eval_accuracy": 0.3998044965786901,
      "eval_loss": 3.9793360233306885,
      "eval_runtime": 3.1051,
      "eval_samples_per_second": 0.322,
      "eval_steps_per_second": 0.322,
      "step": 906
    },
    {
      "epoch": 152.0,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 3.0005,
      "step": 912
    },
    {
      "epoch": 152.0,
      "eval_accuracy": 0.3998044965786901,
      "eval_loss": 3.9779205322265625,
      "eval_runtime": 3.1637,
      "eval_samples_per_second": 0.316,
      "eval_steps_per_second": 0.316,
      "step": 912
    },
    {
      "epoch": 153.0,
      "learning_rate": 2.35e-06,
      "loss": 3.0027,
      "step": 918
    },
    {
      "epoch": 153.0,
      "eval_accuracy": 0.39882697947214074,
      "eval_loss": 3.989084482192993,
      "eval_runtime": 3.2887,
      "eval_samples_per_second": 0.304,
      "eval_steps_per_second": 0.304,
      "step": 918
    },
    {
      "epoch": 154.0,
      "learning_rate": 2.3000000000000004e-06,
      "loss": 3.0034,
      "step": 924
    },
    {
      "epoch": 154.0,
      "eval_accuracy": 0.40078201368523947,
      "eval_loss": 3.9687039852142334,
      "eval_runtime": 3.2813,
      "eval_samples_per_second": 0.305,
      "eval_steps_per_second": 0.305,
      "step": 924
    },
    {
      "epoch": 155.0,
      "learning_rate": 2.25e-06,
      "loss": 2.9853,
      "step": 930
    },
    {
      "epoch": 155.0,
      "eval_accuracy": 0.3978494623655914,
      "eval_loss": 3.988722801208496,
      "eval_runtime": 3.4968,
      "eval_samples_per_second": 0.286,
      "eval_steps_per_second": 0.286,
      "step": 930
    },
    {
      "epoch": 156.0,
      "learning_rate": 2.2e-06,
      "loss": 2.9947,
      "step": 936
    },
    {
      "epoch": 156.0,
      "eval_accuracy": 0.4027370478983382,
      "eval_loss": 3.9860475063323975,
      "eval_runtime": 3.29,
      "eval_samples_per_second": 0.304,
      "eval_steps_per_second": 0.304,
      "step": 936
    },
    {
      "epoch": 157.0,
      "learning_rate": 2.15e-06,
      "loss": 2.9768,
      "step": 942
    },
    {
      "epoch": 157.0,
      "eval_accuracy": 0.4027370478983382,
      "eval_loss": 3.989997148513794,
      "eval_runtime": 3.3737,
      "eval_samples_per_second": 0.296,
      "eval_steps_per_second": 0.296,
      "step": 942
    },
    {
      "epoch": 158.0,
      "learning_rate": 2.1000000000000002e-06,
      "loss": 2.9752,
      "step": 948
    },
    {
      "epoch": 158.0,
      "eval_accuracy": 0.39882697947214074,
      "eval_loss": 3.9992799758911133,
      "eval_runtime": 3.3931,
      "eval_samples_per_second": 0.295,
      "eval_steps_per_second": 0.295,
      "step": 948
    },
    {
      "epoch": 159.0,
      "learning_rate": 2.05e-06,
      "loss": 2.9773,
      "step": 954
    },
    {
      "epoch": 159.0,
      "eval_accuracy": 0.40175953079178883,
      "eval_loss": 3.9693987369537354,
      "eval_runtime": 3.1141,
      "eval_samples_per_second": 0.321,
      "eval_steps_per_second": 0.321,
      "step": 954
    },
    {
      "epoch": 160.0,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 2.9662,
      "step": 960
    },
    {
      "epoch": 160.0,
      "eval_accuracy": 0.3998044965786901,
      "eval_loss": 3.9923715591430664,
      "eval_runtime": 3.2181,
      "eval_samples_per_second": 0.311,
      "eval_steps_per_second": 0.311,
      "step": 960
    },
    {
      "epoch": 161.0,
      "learning_rate": 1.9500000000000004e-06,
      "loss": 2.9661,
      "step": 966
    },
    {
      "epoch": 161.0,
      "eval_accuracy": 0.39882697947214074,
      "eval_loss": 4.008890151977539,
      "eval_runtime": 3.1472,
      "eval_samples_per_second": 0.318,
      "eval_steps_per_second": 0.318,
      "step": 966
    },
    {
      "epoch": 162.0,
      "learning_rate": 1.9000000000000002e-06,
      "loss": 2.9488,
      "step": 972
    },
    {
      "epoch": 162.0,
      "eval_accuracy": 0.39882697947214074,
      "eval_loss": 3.974891424179077,
      "eval_runtime": 3.2045,
      "eval_samples_per_second": 0.312,
      "eval_steps_per_second": 0.312,
      "step": 972
    },
    {
      "epoch": 163.0,
      "learning_rate": 1.85e-06,
      "loss": 2.9487,
      "step": 978
    },
    {
      "epoch": 163.0,
      "eval_accuracy": 0.3978494623655914,
      "eval_loss": 3.9931938648223877,
      "eval_runtime": 2.9879,
      "eval_samples_per_second": 0.335,
      "eval_steps_per_second": 0.335,
      "step": 978
    },
    {
      "epoch": 164.0,
      "learning_rate": 1.8000000000000001e-06,
      "loss": 2.9482,
      "step": 984
    },
    {
      "epoch": 164.0,
      "eval_accuracy": 0.39882697947214074,
      "eval_loss": 3.998748540878296,
      "eval_runtime": 3.0761,
      "eval_samples_per_second": 0.325,
      "eval_steps_per_second": 0.325,
      "step": 984
    },
    {
      "epoch": 165.0,
      "learning_rate": 1.75e-06,
      "loss": 2.9624,
      "step": 990
    },
    {
      "epoch": 165.0,
      "eval_accuracy": 0.3978494623655914,
      "eval_loss": 3.962663412094116,
      "eval_runtime": 3.2497,
      "eval_samples_per_second": 0.308,
      "eval_steps_per_second": 0.308,
      "step": 990
    },
    {
      "epoch": 166.0,
      "learning_rate": 1.7000000000000002e-06,
      "loss": 2.9524,
      "step": 996
    },
    {
      "epoch": 166.0,
      "eval_accuracy": 0.40078201368523947,
      "eval_loss": 3.979146957397461,
      "eval_runtime": 3.2339,
      "eval_samples_per_second": 0.309,
      "eval_steps_per_second": 0.309,
      "step": 996
    },
    {
      "epoch": 167.0,
      "learning_rate": 1.6500000000000003e-06,
      "loss": 2.9357,
      "step": 1002
    },
    {
      "epoch": 167.0,
      "eval_accuracy": 0.3998044965786901,
      "eval_loss": 3.996922731399536,
      "eval_runtime": 3.0918,
      "eval_samples_per_second": 0.323,
      "eval_steps_per_second": 0.323,
      "step": 1002
    },
    {
      "epoch": 168.0,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 2.9323,
      "step": 1008
    },
    {
      "epoch": 168.0,
      "eval_accuracy": 0.40078201368523947,
      "eval_loss": 3.9853515625,
      "eval_runtime": 3.2542,
      "eval_samples_per_second": 0.307,
      "eval_steps_per_second": 0.307,
      "step": 1008
    },
    {
      "epoch": 169.0,
      "learning_rate": 1.5500000000000002e-06,
      "loss": 2.9334,
      "step": 1014
    },
    {
      "epoch": 169.0,
      "eval_accuracy": 0.40078201368523947,
      "eval_loss": 3.977756977081299,
      "eval_runtime": 3.2318,
      "eval_samples_per_second": 0.309,
      "eval_steps_per_second": 0.309,
      "step": 1014
    },
    {
      "epoch": 170.0,
      "learning_rate": 1.5e-06,
      "loss": 2.9228,
      "step": 1020
    },
    {
      "epoch": 170.0,
      "eval_accuracy": 0.4027370478983382,
      "eval_loss": 3.9858651161193848,
      "eval_runtime": 3.1222,
      "eval_samples_per_second": 0.32,
      "eval_steps_per_second": 0.32,
      "step": 1020
    },
    {
      "epoch": 171.0,
      "learning_rate": 1.45e-06,
      "loss": 2.9305,
      "step": 1026
    },
    {
      "epoch": 171.0,
      "eval_accuracy": 0.40371456500488756,
      "eval_loss": 3.9820897579193115,
      "eval_runtime": 3.1725,
      "eval_samples_per_second": 0.315,
      "eval_steps_per_second": 0.315,
      "step": 1026
    },
    {
      "epoch": 172.0,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 2.9239,
      "step": 1032
    },
    {
      "epoch": 172.0,
      "eval_accuracy": 0.4066471163245357,
      "eval_loss": 3.987579107284546,
      "eval_runtime": 3.1039,
      "eval_samples_per_second": 0.322,
      "eval_steps_per_second": 0.322,
      "step": 1032
    },
    {
      "epoch": 173.0,
      "learning_rate": 1.3500000000000002e-06,
      "loss": 2.9181,
      "step": 1038
    },
    {
      "epoch": 173.0,
      "eval_accuracy": 0.4056695992179863,
      "eval_loss": 3.9791972637176514,
      "eval_runtime": 3.503,
      "eval_samples_per_second": 0.285,
      "eval_steps_per_second": 0.285,
      "step": 1038
    },
    {
      "epoch": 174.0,
      "learning_rate": 1.3e-06,
      "loss": 2.9162,
      "step": 1044
    },
    {
      "epoch": 174.0,
      "eval_accuracy": 0.40371456500488756,
      "eval_loss": 3.9731061458587646,
      "eval_runtime": 3.0863,
      "eval_samples_per_second": 0.324,
      "eval_steps_per_second": 0.324,
      "step": 1044
    },
    {
      "epoch": 175.0,
      "learning_rate": 1.25e-06,
      "loss": 2.9171,
      "step": 1050
    },
    {
      "epoch": 175.0,
      "eval_accuracy": 0.4066471163245357,
      "eval_loss": 3.9795916080474854,
      "eval_runtime": 3.0684,
      "eval_samples_per_second": 0.326,
      "eval_steps_per_second": 0.326,
      "step": 1050
    },
    {
      "epoch": 176.0,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 2.9132,
      "step": 1056
    },
    {
      "epoch": 176.0,
      "eval_accuracy": 0.4046920821114369,
      "eval_loss": 3.9914140701293945,
      "eval_runtime": 3.5253,
      "eval_samples_per_second": 0.284,
      "eval_steps_per_second": 0.284,
      "step": 1056
    },
    {
      "epoch": 177.0,
      "learning_rate": 1.1500000000000002e-06,
      "loss": 2.9168,
      "step": 1062
    },
    {
      "epoch": 177.0,
      "eval_accuracy": 0.4046920821114369,
      "eval_loss": 3.9826488494873047,
      "eval_runtime": 3.0836,
      "eval_samples_per_second": 0.324,
      "eval_steps_per_second": 0.324,
      "step": 1062
    },
    {
      "epoch": 178.0,
      "learning_rate": 1.1e-06,
      "loss": 2.8974,
      "step": 1068
    },
    {
      "epoch": 178.0,
      "eval_accuracy": 0.4056695992179863,
      "eval_loss": 3.9753177165985107,
      "eval_runtime": 3.165,
      "eval_samples_per_second": 0.316,
      "eval_steps_per_second": 0.316,
      "step": 1068
    },
    {
      "epoch": 179.0,
      "learning_rate": 1.0500000000000001e-06,
      "loss": 2.8954,
      "step": 1074
    },
    {
      "epoch": 179.0,
      "eval_accuracy": 0.4056695992179863,
      "eval_loss": 3.976564884185791,
      "eval_runtime": 3.123,
      "eval_samples_per_second": 0.32,
      "eval_steps_per_second": 0.32,
      "step": 1074
    },
    {
      "epoch": 180.0,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 2.9003,
      "step": 1080
    },
    {
      "epoch": 180.0,
      "eval_accuracy": 0.4027370478983382,
      "eval_loss": 3.986520290374756,
      "eval_runtime": 3.106,
      "eval_samples_per_second": 0.322,
      "eval_steps_per_second": 0.322,
      "step": 1080
    },
    {
      "epoch": 181.0,
      "learning_rate": 9.500000000000001e-07,
      "loss": 2.9012,
      "step": 1086
    },
    {
      "epoch": 181.0,
      "eval_accuracy": 0.4046920821114369,
      "eval_loss": 3.983490228652954,
      "eval_runtime": 3.1013,
      "eval_samples_per_second": 0.322,
      "eval_steps_per_second": 0.322,
      "step": 1086
    },
    {
      "epoch": 182.0,
      "learning_rate": 9.000000000000001e-07,
      "loss": 2.8994,
      "step": 1092
    },
    {
      "epoch": 182.0,
      "eval_accuracy": 0.4046920821114369,
      "eval_loss": 3.980245590209961,
      "eval_runtime": 3.2714,
      "eval_samples_per_second": 0.306,
      "eval_steps_per_second": 0.306,
      "step": 1092
    },
    {
      "epoch": 183.0,
      "learning_rate": 8.500000000000001e-07,
      "loss": 2.8918,
      "step": 1098
    },
    {
      "epoch": 183.0,
      "eval_accuracy": 0.4066471163245357,
      "eval_loss": 3.9810588359832764,
      "eval_runtime": 3.2463,
      "eval_samples_per_second": 0.308,
      "eval_steps_per_second": 0.308,
      "step": 1098
    },
    {
      "epoch": 184.0,
      "learning_rate": 8.000000000000001e-07,
      "loss": 2.8893,
      "step": 1104
    },
    {
      "epoch": 184.0,
      "eval_accuracy": 0.4056695992179863,
      "eval_loss": 3.981043815612793,
      "eval_runtime": 3.5146,
      "eval_samples_per_second": 0.285,
      "eval_steps_per_second": 0.285,
      "step": 1104
    },
    {
      "epoch": 185.0,
      "learning_rate": 7.5e-07,
      "loss": 2.8865,
      "step": 1110
    },
    {
      "epoch": 185.0,
      "eval_accuracy": 0.40762463343108507,
      "eval_loss": 3.9851670265197754,
      "eval_runtime": 3.121,
      "eval_samples_per_second": 0.32,
      "eval_steps_per_second": 0.32,
      "step": 1110
    },
    {
      "epoch": 186.0,
      "learning_rate": 7.000000000000001e-07,
      "loss": 2.8784,
      "step": 1116
    },
    {
      "epoch": 186.0,
      "eval_accuracy": 0.4056695992179863,
      "eval_loss": 3.9805047512054443,
      "eval_runtime": 3.1275,
      "eval_samples_per_second": 0.32,
      "eval_steps_per_second": 0.32,
      "step": 1116
    },
    {
      "epoch": 187.0,
      "learning_rate": 6.5e-07,
      "loss": 2.8875,
      "step": 1122
    },
    {
      "epoch": 187.0,
      "eval_accuracy": 0.4066471163245357,
      "eval_loss": 3.978147029876709,
      "eval_runtime": 3.3509,
      "eval_samples_per_second": 0.298,
      "eval_steps_per_second": 0.298,
      "step": 1122
    },
    {
      "epoch": 188.0,
      "learning_rate": 6.000000000000001e-07,
      "loss": 2.8948,
      "step": 1128
    },
    {
      "epoch": 188.0,
      "eval_accuracy": 0.4056695992179863,
      "eval_loss": 3.9830515384674072,
      "eval_runtime": 3.3453,
      "eval_samples_per_second": 0.299,
      "eval_steps_per_second": 0.299,
      "step": 1128
    },
    {
      "epoch": 189.0,
      "learning_rate": 5.5e-07,
      "loss": 2.8927,
      "step": 1134
    },
    {
      "epoch": 189.0,
      "eval_accuracy": 0.4066471163245357,
      "eval_loss": 3.9836947917938232,
      "eval_runtime": 3.1414,
      "eval_samples_per_second": 0.318,
      "eval_steps_per_second": 0.318,
      "step": 1134
    },
    {
      "epoch": 190.0,
      "learning_rate": 5.000000000000001e-07,
      "loss": 2.8739,
      "step": 1140
    },
    {
      "epoch": 190.0,
      "eval_accuracy": 0.4056695992179863,
      "eval_loss": 3.9822254180908203,
      "eval_runtime": 3.3757,
      "eval_samples_per_second": 0.296,
      "eval_steps_per_second": 0.296,
      "step": 1140
    },
    {
      "epoch": 191.0,
      "learning_rate": 4.5000000000000003e-07,
      "loss": 2.8919,
      "step": 1146
    },
    {
      "epoch": 191.0,
      "eval_accuracy": 0.4066471163245357,
      "eval_loss": 3.9791526794433594,
      "eval_runtime": 3.5323,
      "eval_samples_per_second": 0.283,
      "eval_steps_per_second": 0.283,
      "step": 1146
    },
    {
      "epoch": 192.0,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 2.8713,
      "step": 1152
    },
    {
      "epoch": 192.0,
      "eval_accuracy": 0.4056695992179863,
      "eval_loss": 3.9800055027008057,
      "eval_runtime": 3.4653,
      "eval_samples_per_second": 0.289,
      "eval_steps_per_second": 0.289,
      "step": 1152
    },
    {
      "epoch": 193.0,
      "learning_rate": 3.5000000000000004e-07,
      "loss": 2.8798,
      "step": 1158
    },
    {
      "epoch": 193.0,
      "eval_accuracy": 0.4046920821114369,
      "eval_loss": 3.985433578491211,
      "eval_runtime": 3.0972,
      "eval_samples_per_second": 0.323,
      "eval_steps_per_second": 0.323,
      "step": 1158
    },
    {
      "epoch": 194.0,
      "learning_rate": 3.0000000000000004e-07,
      "loss": 2.8835,
      "step": 1164
    },
    {
      "epoch": 194.0,
      "eval_accuracy": 0.4056695992179863,
      "eval_loss": 3.984498977661133,
      "eval_runtime": 4.5165,
      "eval_samples_per_second": 0.221,
      "eval_steps_per_second": 0.221,
      "step": 1164
    },
    {
      "epoch": 195.0,
      "learning_rate": 2.5000000000000004e-07,
      "loss": 2.878,
      "step": 1170
    },
    {
      "epoch": 195.0,
      "eval_accuracy": 0.4056695992179863,
      "eval_loss": 3.981985330581665,
      "eval_runtime": 3.1036,
      "eval_samples_per_second": 0.322,
      "eval_steps_per_second": 0.322,
      "step": 1170
    },
    {
      "epoch": 196.0,
      "learning_rate": 2.0000000000000002e-07,
      "loss": 2.8931,
      "step": 1176
    },
    {
      "epoch": 196.0,
      "eval_accuracy": 0.4056695992179863,
      "eval_loss": 3.9816386699676514,
      "eval_runtime": 3.2982,
      "eval_samples_per_second": 0.303,
      "eval_steps_per_second": 0.303,
      "step": 1176
    },
    {
      "epoch": 197.0,
      "learning_rate": 1.5000000000000002e-07,
      "loss": 2.8662,
      "step": 1182
    },
    {
      "epoch": 197.0,
      "eval_accuracy": 0.4056695992179863,
      "eval_loss": 3.982978105545044,
      "eval_runtime": 4.8102,
      "eval_samples_per_second": 0.208,
      "eval_steps_per_second": 0.208,
      "step": 1182
    },
    {
      "epoch": 198.0,
      "learning_rate": 1.0000000000000001e-07,
      "loss": 2.8734,
      "step": 1188
    },
    {
      "epoch": 198.0,
      "eval_accuracy": 0.4056695992179863,
      "eval_loss": 3.984077215194702,
      "eval_runtime": 3.1139,
      "eval_samples_per_second": 0.321,
      "eval_steps_per_second": 0.321,
      "step": 1188
    },
    {
      "epoch": 199.0,
      "learning_rate": 5.0000000000000004e-08,
      "loss": 2.8825,
      "step": 1194
    },
    {
      "epoch": 199.0,
      "eval_accuracy": 0.4056695992179863,
      "eval_loss": 3.98300838470459,
      "eval_runtime": 3.644,
      "eval_samples_per_second": 0.274,
      "eval_steps_per_second": 0.274,
      "step": 1194
    },
    {
      "epoch": 200.0,
      "learning_rate": 0.0,
      "loss": 2.8825,
      "step": 1200
    },
    {
      "epoch": 200.0,
      "eval_accuracy": 0.4056695992179863,
      "eval_loss": 3.982684373855591,
      "eval_runtime": 3.4062,
      "eval_samples_per_second": 0.294,
      "eval_steps_per_second": 0.294,
      "step": 1200
    },
    {
      "epoch": 200.0,
      "step": 1200,
      "total_flos": 1.044564148224e+16,
      "train_loss": 3.9290491278966266,
      "train_runtime": 32687.0411,
      "train_samples_per_second": 0.073,
      "train_steps_per_second": 0.037
    }
  ],
  "logging_steps": 500,
  "max_steps": 1200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 200,
  "save_steps": 500,
  "total_flos": 1.044564148224e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}