| { | |
| "best_metric": 0.8037383177570093, | |
| "best_model_checkpoint": "beit-base-patch16-224-RH\\checkpoint-232", | |
| "epoch": 40.0, | |
| "eval_steps": 500, | |
| "global_step": 320, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.5887850467289719, | |
| "eval_loss": 0.7926577925682068, | |
| "eval_runtime": 2.0185, | |
| "eval_samples_per_second": 53.011, | |
| "eval_steps_per_second": 3.468, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "learning_rate": 1.5625e-05, | |
| "loss": 0.8183, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5887850467289719, | |
| "eval_loss": 0.7411813735961914, | |
| "eval_runtime": 1.969, | |
| "eval_samples_per_second": 54.343, | |
| "eval_steps_per_second": 3.555, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "learning_rate": 3.125e-05, | |
| "loss": 0.7414, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.5887850467289719, | |
| "eval_loss": 0.6850691437721252, | |
| "eval_runtime": 2.004, | |
| "eval_samples_per_second": 53.394, | |
| "eval_steps_per_second": 3.493, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "learning_rate": 4.6875e-05, | |
| "loss": 0.6837, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.5887850467289719, | |
| "eval_loss": 0.6638379693031311, | |
| "eval_runtime": 2.1135, | |
| "eval_samples_per_second": 50.627, | |
| "eval_steps_per_second": 3.312, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 4.8611111111111115e-05, | |
| "loss": 0.6621, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.5981308411214953, | |
| "eval_loss": 0.6619369387626648, | |
| "eval_runtime": 1.9694, | |
| "eval_samples_per_second": 54.331, | |
| "eval_steps_per_second": 3.554, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.6261682242990654, | |
| "eval_loss": 0.6446465253829956, | |
| "eval_runtime": 2.0965, | |
| "eval_samples_per_second": 51.037, | |
| "eval_steps_per_second": 3.339, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 6.25, | |
| "learning_rate": 4.6875e-05, | |
| "loss": 0.6538, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.6728971962616822, | |
| "eval_loss": 0.6369854807853699, | |
| "eval_runtime": 1.9695, | |
| "eval_samples_per_second": 54.33, | |
| "eval_steps_per_second": 3.554, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "learning_rate": 4.5138888888888894e-05, | |
| "loss": 0.641, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.6635514018691588, | |
| "eval_loss": 0.6484658122062683, | |
| "eval_runtime": 2.082, | |
| "eval_samples_per_second": 51.394, | |
| "eval_steps_per_second": 3.362, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 8.75, | |
| "learning_rate": 4.340277777777778e-05, | |
| "loss": 0.628, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.6448598130841121, | |
| "eval_loss": 0.6392771601676941, | |
| "eval_runtime": 1.9785, | |
| "eval_samples_per_second": 54.083, | |
| "eval_steps_per_second": 3.538, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 4.166666666666667e-05, | |
| "loss": 0.6187, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.5794392523364486, | |
| "eval_loss": 0.640946626663208, | |
| "eval_runtime": 1.972, | |
| "eval_samples_per_second": 54.261, | |
| "eval_steps_per_second": 3.55, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.5794392523364486, | |
| "eval_loss": 0.6359996795654297, | |
| "eval_runtime": 1.9835, | |
| "eval_samples_per_second": 53.946, | |
| "eval_steps_per_second": 3.529, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 11.25, | |
| "learning_rate": 3.993055555555556e-05, | |
| "loss": 0.6075, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.6355140186915887, | |
| "eval_loss": 0.6208956837654114, | |
| "eval_runtime": 1.9903, | |
| "eval_samples_per_second": 53.761, | |
| "eval_steps_per_second": 3.517, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "learning_rate": 3.8194444444444444e-05, | |
| "loss": 0.6081, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.6448598130841121, | |
| "eval_loss": 0.6377462148666382, | |
| "eval_runtime": 1.992, | |
| "eval_samples_per_second": 53.716, | |
| "eval_steps_per_second": 3.514, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 13.75, | |
| "learning_rate": 3.6458333333333336e-05, | |
| "loss": 0.5886, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.6728971962616822, | |
| "eval_loss": 0.5931380987167358, | |
| "eval_runtime": 1.9725, | |
| "eval_samples_per_second": 54.247, | |
| "eval_steps_per_second": 3.549, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "learning_rate": 3.472222222222222e-05, | |
| "loss": 0.5945, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.6635514018691588, | |
| "eval_loss": 0.6107914447784424, | |
| "eval_runtime": 1.983, | |
| "eval_samples_per_second": 53.959, | |
| "eval_steps_per_second": 3.53, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.7009345794392523, | |
| "eval_loss": 0.584554135799408, | |
| "eval_runtime": 1.992, | |
| "eval_samples_per_second": 53.716, | |
| "eval_steps_per_second": 3.514, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 16.25, | |
| "learning_rate": 3.2986111111111115e-05, | |
| "loss": 0.5808, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.6822429906542056, | |
| "eval_loss": 0.5945072770118713, | |
| "eval_runtime": 1.98, | |
| "eval_samples_per_second": 54.041, | |
| "eval_steps_per_second": 3.535, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 17.5, | |
| "learning_rate": 3.125e-05, | |
| "loss": 0.5636, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.6635514018691588, | |
| "eval_loss": 0.7401838302612305, | |
| "eval_runtime": 1.9965, | |
| "eval_samples_per_second": 53.594, | |
| "eval_steps_per_second": 3.506, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 18.75, | |
| "learning_rate": 2.951388888888889e-05, | |
| "loss": 0.5839, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.6915887850467289, | |
| "eval_loss": 0.5661436319351196, | |
| "eval_runtime": 2.014, | |
| "eval_samples_per_second": 53.129, | |
| "eval_steps_per_second": 3.476, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 0.5166, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.6635514018691588, | |
| "eval_loss": 0.5359623432159424, | |
| "eval_runtime": 1.98, | |
| "eval_samples_per_second": 54.041, | |
| "eval_steps_per_second": 3.535, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "eval_accuracy": 0.6728971962616822, | |
| "eval_loss": 0.5621436238288879, | |
| "eval_runtime": 2.0255, | |
| "eval_samples_per_second": 52.827, | |
| "eval_steps_per_second": 3.456, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 21.25, | |
| "learning_rate": 2.604166666666667e-05, | |
| "loss": 0.5165, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "eval_accuracy": 0.719626168224299, | |
| "eval_loss": 0.5508888363838196, | |
| "eval_runtime": 1.9745, | |
| "eval_samples_per_second": 54.192, | |
| "eval_steps_per_second": 3.545, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 22.5, | |
| "learning_rate": 2.4305555555555558e-05, | |
| "loss": 0.5308, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "eval_accuracy": 0.7570093457943925, | |
| "eval_loss": 0.5601618885993958, | |
| "eval_runtime": 1.9885, | |
| "eval_samples_per_second": 53.81, | |
| "eval_steps_per_second": 3.52, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 23.75, | |
| "learning_rate": 2.2569444444444447e-05, | |
| "loss": 0.4595, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_accuracy": 0.7850467289719626, | |
| "eval_loss": 0.47350895404815674, | |
| "eval_runtime": 1.969, | |
| "eval_samples_per_second": 54.343, | |
| "eval_steps_per_second": 3.555, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "learning_rate": 2.0833333333333336e-05, | |
| "loss": 0.4553, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "eval_accuracy": 0.7663551401869159, | |
| "eval_loss": 0.4696148931980133, | |
| "eval_runtime": 1.974, | |
| "eval_samples_per_second": 54.206, | |
| "eval_steps_per_second": 3.546, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "eval_accuracy": 0.7850467289719626, | |
| "eval_loss": 0.5305755734443665, | |
| "eval_runtime": 1.97, | |
| "eval_samples_per_second": 54.316, | |
| "eval_steps_per_second": 3.553, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 26.25, | |
| "learning_rate": 1.9097222222222222e-05, | |
| "loss": 0.4004, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "eval_accuracy": 0.794392523364486, | |
| "eval_loss": 0.48186224699020386, | |
| "eval_runtime": 1.972, | |
| "eval_samples_per_second": 54.26, | |
| "eval_steps_per_second": 3.55, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 27.5, | |
| "learning_rate": 1.736111111111111e-05, | |
| "loss": 0.3954, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "eval_accuracy": 0.794392523364486, | |
| "eval_loss": 0.483099102973938, | |
| "eval_runtime": 2.0135, | |
| "eval_samples_per_second": 53.142, | |
| "eval_steps_per_second": 3.477, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 28.75, | |
| "learning_rate": 1.5625e-05, | |
| "loss": 0.3521, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 29.0, | |
| "eval_accuracy": 0.8037383177570093, | |
| "eval_loss": 0.43397656083106995, | |
| "eval_runtime": 1.9725, | |
| "eval_samples_per_second": 54.246, | |
| "eval_steps_per_second": 3.549, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "learning_rate": 1.388888888888889e-05, | |
| "loss": 0.3436, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "eval_accuracy": 0.7757009345794392, | |
| "eval_loss": 0.4789830446243286, | |
| "eval_runtime": 2.015, | |
| "eval_samples_per_second": 53.102, | |
| "eval_steps_per_second": 3.474, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 31.0, | |
| "eval_accuracy": 0.7757009345794392, | |
| "eval_loss": 0.4719656705856323, | |
| "eval_runtime": 1.9775, | |
| "eval_samples_per_second": 54.11, | |
| "eval_steps_per_second": 3.54, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 31.25, | |
| "learning_rate": 1.2152777777777779e-05, | |
| "loss": 0.34, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "eval_accuracy": 0.7850467289719626, | |
| "eval_loss": 0.5282993912696838, | |
| "eval_runtime": 1.992, | |
| "eval_samples_per_second": 53.716, | |
| "eval_steps_per_second": 3.514, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 32.5, | |
| "learning_rate": 1.0416666666666668e-05, | |
| "loss": 0.2995, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 33.0, | |
| "eval_accuracy": 0.794392523364486, | |
| "eval_loss": 0.43829986453056335, | |
| "eval_runtime": 1.968, | |
| "eval_samples_per_second": 54.371, | |
| "eval_steps_per_second": 3.557, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 33.75, | |
| "learning_rate": 8.680555555555556e-06, | |
| "loss": 0.2951, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 34.0, | |
| "eval_accuracy": 0.794392523364486, | |
| "eval_loss": 0.4739982783794403, | |
| "eval_runtime": 1.9943, | |
| "eval_samples_per_second": 53.653, | |
| "eval_steps_per_second": 3.51, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 35.0, | |
| "learning_rate": 6.944444444444445e-06, | |
| "loss": 0.3094, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 35.0, | |
| "eval_accuracy": 0.7663551401869159, | |
| "eval_loss": 0.5863447785377502, | |
| "eval_runtime": 1.9985, | |
| "eval_samples_per_second": 53.541, | |
| "eval_steps_per_second": 3.503, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 36.0, | |
| "eval_accuracy": 0.7850467289719626, | |
| "eval_loss": 0.4483156204223633, | |
| "eval_runtime": 1.98, | |
| "eval_samples_per_second": 54.041, | |
| "eval_steps_per_second": 3.535, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 36.25, | |
| "learning_rate": 5.208333333333334e-06, | |
| "loss": 0.2963, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 37.0, | |
| "eval_accuracy": 0.794392523364486, | |
| "eval_loss": 0.4758929908275604, | |
| "eval_runtime": 1.989, | |
| "eval_samples_per_second": 53.797, | |
| "eval_steps_per_second": 3.519, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 37.5, | |
| "learning_rate": 3.4722222222222224e-06, | |
| "loss": 0.3045, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 38.0, | |
| "eval_accuracy": 0.794392523364486, | |
| "eval_loss": 0.446876585483551, | |
| "eval_runtime": 2.006, | |
| "eval_samples_per_second": 53.341, | |
| "eval_steps_per_second": 3.49, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 38.75, | |
| "learning_rate": 1.7361111111111112e-06, | |
| "loss": 0.2739, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 39.0, | |
| "eval_accuracy": 0.7850467289719626, | |
| "eval_loss": 0.4517090618610382, | |
| "eval_runtime": 1.971, | |
| "eval_samples_per_second": 54.288, | |
| "eval_steps_per_second": 3.552, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "learning_rate": 0.0, | |
| "loss": 0.2717, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "eval_accuracy": 0.794392523364486, | |
| "eval_loss": 0.4654483497142792, | |
| "eval_runtime": 2.08, | |
| "eval_samples_per_second": 51.442, | |
| "eval_steps_per_second": 3.365, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "step": 320, | |
| "total_flos": 1.5429806632629043e+18, | |
| "train_loss": 0.4979598492383957, | |
| "train_runtime": 690.9933, | |
| "train_samples_per_second": 28.828, | |
| "train_steps_per_second": 0.463 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 320, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 40, | |
| "save_steps": 500, | |
| "total_flos": 1.5429806632629043e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |