| { | |
| "best_metric": 0.8317757009345794, | |
| "best_model_checkpoint": "beit-base-patch16-224-OT\\checkpoint-248", | |
| "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": 0.6887365579605103, | |
| "eval_runtime": 2.8498, | |
| "eval_samples_per_second": 37.546, | |
| "eval_steps_per_second": 2.456, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "learning_rate": 1.5625e-05, | |
| "loss": 0.692, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5887850467289719, | |
| "eval_loss": 0.678210437297821, | |
| "eval_runtime": 1.9785, | |
| "eval_samples_per_second": 54.082, | |
| "eval_steps_per_second": 3.538, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "learning_rate": 3.125e-05, | |
| "loss": 0.6801, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.5887850467289719, | |
| "eval_loss": 0.6668981313705444, | |
| "eval_runtime": 2.0656, | |
| "eval_samples_per_second": 51.802, | |
| "eval_steps_per_second": 3.389, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "learning_rate": 4.6875e-05, | |
| "loss": 0.6696, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.5887850467289719, | |
| "eval_loss": 0.6644209623336792, | |
| "eval_runtime": 2.1267, | |
| "eval_samples_per_second": 50.313, | |
| "eval_steps_per_second": 3.291, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 4.8611111111111115e-05, | |
| "loss": 0.6607, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.6635514018691588, | |
| "eval_loss": 0.6661449074745178, | |
| "eval_runtime": 1.995, | |
| "eval_samples_per_second": 53.635, | |
| "eval_steps_per_second": 3.509, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.6542056074766355, | |
| "eval_loss": 0.6241438388824463, | |
| "eval_runtime": 2.0189, | |
| "eval_samples_per_second": 52.999, | |
| "eval_steps_per_second": 3.467, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 6.25, | |
| "learning_rate": 4.6875e-05, | |
| "loss": 0.6341, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.6542056074766355, | |
| "eval_loss": 0.6234968900680542, | |
| "eval_runtime": 1.9955, | |
| "eval_samples_per_second": 53.622, | |
| "eval_steps_per_second": 3.508, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "learning_rate": 4.5138888888888894e-05, | |
| "loss": 0.6089, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.6915887850467289, | |
| "eval_loss": 0.6088296175003052, | |
| "eval_runtime": 2.06, | |
| "eval_samples_per_second": 51.942, | |
| "eval_steps_per_second": 3.398, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 8.75, | |
| "learning_rate": 4.340277777777778e-05, | |
| "loss": 0.6095, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.6915887850467289, | |
| "eval_loss": 0.5911644101142883, | |
| "eval_runtime": 2.01, | |
| "eval_samples_per_second": 53.235, | |
| "eval_steps_per_second": 3.483, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 4.166666666666667e-05, | |
| "loss": 0.5632, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.6355140186915887, | |
| "eval_loss": 0.660692036151886, | |
| "eval_runtime": 2.126, | |
| "eval_samples_per_second": 50.329, | |
| "eval_steps_per_second": 3.293, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.7009345794392523, | |
| "eval_loss": 0.5792553424835205, | |
| "eval_runtime": 1.972, | |
| "eval_samples_per_second": 54.261, | |
| "eval_steps_per_second": 3.55, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 11.25, | |
| "learning_rate": 3.993055555555556e-05, | |
| "loss": 0.5418, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.6822429906542056, | |
| "eval_loss": 0.5953279733657837, | |
| "eval_runtime": 2.2228, | |
| "eval_samples_per_second": 48.137, | |
| "eval_steps_per_second": 3.149, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "learning_rate": 3.8194444444444444e-05, | |
| "loss": 0.5336, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.7102803738317757, | |
| "eval_loss": 0.5792534947395325, | |
| "eval_runtime": 1.991, | |
| "eval_samples_per_second": 53.742, | |
| "eval_steps_per_second": 3.516, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 13.75, | |
| "learning_rate": 3.6458333333333336e-05, | |
| "loss": 0.5102, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.719626168224299, | |
| "eval_loss": 0.5291872024536133, | |
| "eval_runtime": 2.0585, | |
| "eval_samples_per_second": 51.98, | |
| "eval_steps_per_second": 3.401, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "learning_rate": 3.472222222222222e-05, | |
| "loss": 0.4762, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.7009345794392523, | |
| "eval_loss": 0.6557727456092834, | |
| "eval_runtime": 1.9985, | |
| "eval_samples_per_second": 53.541, | |
| "eval_steps_per_second": 3.503, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.7102803738317757, | |
| "eval_loss": 0.5371208786964417, | |
| "eval_runtime": 1.971, | |
| "eval_samples_per_second": 54.288, | |
| "eval_steps_per_second": 3.552, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 16.25, | |
| "learning_rate": 3.2986111111111115e-05, | |
| "loss": 0.544, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.7570093457943925, | |
| "eval_loss": 0.5400705337524414, | |
| "eval_runtime": 1.981, | |
| "eval_samples_per_second": 54.014, | |
| "eval_steps_per_second": 3.534, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 17.5, | |
| "learning_rate": 3.125e-05, | |
| "loss": 0.4256, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.794392523364486, | |
| "eval_loss": 0.49267861247062683, | |
| "eval_runtime": 1.9622, | |
| "eval_samples_per_second": 54.531, | |
| "eval_steps_per_second": 3.567, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 18.75, | |
| "learning_rate": 2.951388888888889e-05, | |
| "loss": 0.4082, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.7383177570093458, | |
| "eval_loss": 0.5800967216491699, | |
| "eval_runtime": 1.9825, | |
| "eval_samples_per_second": 53.973, | |
| "eval_steps_per_second": 3.531, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 0.4014, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.7383177570093458, | |
| "eval_loss": 0.5822688937187195, | |
| "eval_runtime": 1.985, | |
| "eval_samples_per_second": 53.905, | |
| "eval_steps_per_second": 3.526, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "eval_accuracy": 0.7757009345794392, | |
| "eval_loss": 0.5392723083496094, | |
| "eval_runtime": 2.117, | |
| "eval_samples_per_second": 50.542, | |
| "eval_steps_per_second": 3.306, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 21.25, | |
| "learning_rate": 2.604166666666667e-05, | |
| "loss": 0.3483, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "eval_accuracy": 0.7102803738317757, | |
| "eval_loss": 0.5940819382667542, | |
| "eval_runtime": 1.9765, | |
| "eval_samples_per_second": 54.137, | |
| "eval_steps_per_second": 3.542, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 22.5, | |
| "learning_rate": 2.4305555555555558e-05, | |
| "loss": 0.3121, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "eval_accuracy": 0.7383177570093458, | |
| "eval_loss": 0.5568514466285706, | |
| "eval_runtime": 2.1005, | |
| "eval_samples_per_second": 50.94, | |
| "eval_steps_per_second": 3.333, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 23.75, | |
| "learning_rate": 2.2569444444444447e-05, | |
| "loss": 0.3484, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_accuracy": 0.7663551401869159, | |
| "eval_loss": 0.5975044369697571, | |
| "eval_runtime": 1.97, | |
| "eval_samples_per_second": 54.316, | |
| "eval_steps_per_second": 3.553, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "learning_rate": 2.0833333333333336e-05, | |
| "loss": 0.263, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "eval_accuracy": 0.7570093457943925, | |
| "eval_loss": 0.6544022560119629, | |
| "eval_runtime": 2.3716, | |
| "eval_samples_per_second": 45.118, | |
| "eval_steps_per_second": 2.952, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "eval_accuracy": 0.7757009345794392, | |
| "eval_loss": 0.5743973851203918, | |
| "eval_runtime": 2.0105, | |
| "eval_samples_per_second": 53.221, | |
| "eval_steps_per_second": 3.482, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 26.25, | |
| "learning_rate": 1.9097222222222222e-05, | |
| "loss": 0.2633, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "eval_accuracy": 0.7663551401869159, | |
| "eval_loss": 0.6095036864280701, | |
| "eval_runtime": 1.9725, | |
| "eval_samples_per_second": 54.247, | |
| "eval_steps_per_second": 3.549, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 27.5, | |
| "learning_rate": 1.736111111111111e-05, | |
| "loss": 0.2935, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "eval_accuracy": 0.7663551401869159, | |
| "eval_loss": 0.528620719909668, | |
| "eval_runtime": 2.0605, | |
| "eval_samples_per_second": 51.929, | |
| "eval_steps_per_second": 3.397, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 28.75, | |
| "learning_rate": 1.5625e-05, | |
| "loss": 0.2332, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 29.0, | |
| "eval_accuracy": 0.7850467289719626, | |
| "eval_loss": 0.6027860045433044, | |
| "eval_runtime": 1.9665, | |
| "eval_samples_per_second": 54.412, | |
| "eval_steps_per_second": 3.56, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "learning_rate": 1.388888888888889e-05, | |
| "loss": 0.2314, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "eval_accuracy": 0.794392523364486, | |
| "eval_loss": 0.5935384631156921, | |
| "eval_runtime": 1.962, | |
| "eval_samples_per_second": 54.537, | |
| "eval_steps_per_second": 3.568, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 31.0, | |
| "eval_accuracy": 0.8317757009345794, | |
| "eval_loss": 0.5392867922782898, | |
| "eval_runtime": 1.969, | |
| "eval_samples_per_second": 54.343, | |
| "eval_steps_per_second": 3.555, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 31.25, | |
| "learning_rate": 1.2152777777777779e-05, | |
| "loss": 0.202, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "eval_accuracy": 0.822429906542056, | |
| "eval_loss": 0.5556337833404541, | |
| "eval_runtime": 2.012, | |
| "eval_samples_per_second": 53.182, | |
| "eval_steps_per_second": 3.479, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 32.5, | |
| "learning_rate": 1.0416666666666668e-05, | |
| "loss": 0.2127, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 33.0, | |
| "eval_accuracy": 0.8037383177570093, | |
| "eval_loss": 0.5912833213806152, | |
| "eval_runtime": 1.98, | |
| "eval_samples_per_second": 54.041, | |
| "eval_steps_per_second": 3.535, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 33.75, | |
| "learning_rate": 8.680555555555556e-06, | |
| "loss": 0.2035, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 34.0, | |
| "eval_accuracy": 0.8037383177570093, | |
| "eval_loss": 0.5337203741073608, | |
| "eval_runtime": 1.9834, | |
| "eval_samples_per_second": 53.947, | |
| "eval_steps_per_second": 3.529, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 35.0, | |
| "learning_rate": 6.944444444444445e-06, | |
| "loss": 0.2618, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 35.0, | |
| "eval_accuracy": 0.8037383177570093, | |
| "eval_loss": 0.622107982635498, | |
| "eval_runtime": 2.0025, | |
| "eval_samples_per_second": 53.434, | |
| "eval_steps_per_second": 3.496, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 36.0, | |
| "eval_accuracy": 0.8317757009345794, | |
| "eval_loss": 0.5089600682258606, | |
| "eval_runtime": 1.9755, | |
| "eval_samples_per_second": 54.164, | |
| "eval_steps_per_second": 3.543, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 36.25, | |
| "learning_rate": 5.208333333333334e-06, | |
| "loss": 0.217, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 37.0, | |
| "eval_accuracy": 0.822429906542056, | |
| "eval_loss": 0.5649047493934631, | |
| "eval_runtime": 1.991, | |
| "eval_samples_per_second": 53.743, | |
| "eval_steps_per_second": 3.516, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 37.5, | |
| "learning_rate": 3.4722222222222224e-06, | |
| "loss": 0.2111, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 38.0, | |
| "eval_accuracy": 0.8130841121495327, | |
| "eval_loss": 0.568317174911499, | |
| "eval_runtime": 2.0365, | |
| "eval_samples_per_second": 52.542, | |
| "eval_steps_per_second": 3.437, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 38.75, | |
| "learning_rate": 1.7361111111111112e-06, | |
| "loss": 0.2085, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 39.0, | |
| "eval_accuracy": 0.822429906542056, | |
| "eval_loss": 0.5397838950157166, | |
| "eval_runtime": 1.9935, | |
| "eval_samples_per_second": 53.675, | |
| "eval_steps_per_second": 3.511, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "learning_rate": 0.0, | |
| "loss": 0.1912, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "eval_accuracy": 0.822429906542056, | |
| "eval_loss": 0.5548034310340881, | |
| "eval_runtime": 1.978, | |
| "eval_samples_per_second": 54.096, | |
| "eval_steps_per_second": 3.539, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "step": 320, | |
| "total_flos": 1.5429806632629043e+18, | |
| "train_loss": 0.4049976162612438, | |
| "train_runtime": 701.3243, | |
| "train_samples_per_second": 28.403, | |
| "train_steps_per_second": 0.456 | |
| } | |
| ], | |
| "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 | |
| } | |