| { | |
| "best_metric": 0.8032710280373832, | |
| "best_model_checkpoint": "resnet-152-finetuned-cassava-leaf-disease/checkpoint-703", | |
| "epoch": 29.690721649484537, | |
| "eval_steps": 500, | |
| "global_step": 720, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 6.944444444444445e-06, | |
| "loss": 7.3336, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 1.388888888888889e-05, | |
| "loss": 7.2158, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "eval_accuracy": 0.0014018691588785046, | |
| "eval_loss": 6.815145015716553, | |
| "eval_runtime": 8.158, | |
| "eval_samples_per_second": 262.318, | |
| "eval_steps_per_second": 1.348, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "learning_rate": 2.0833333333333336e-05, | |
| "loss": 6.9635, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 6.6136, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "eval_accuracy": 0.046261682242990654, | |
| "eval_loss": 5.699448108673096, | |
| "eval_runtime": 8.3972, | |
| "eval_samples_per_second": 254.846, | |
| "eval_steps_per_second": 1.31, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "learning_rate": 3.472222222222222e-05, | |
| "loss": 6.1712, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "learning_rate": 4.166666666666667e-05, | |
| "loss": 5.5789, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "learning_rate": 4.8611111111111115e-05, | |
| "loss": 4.7064, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "eval_accuracy": 0.45560747663551404, | |
| "eval_loss": 3.352895975112915, | |
| "eval_runtime": 7.9592, | |
| "eval_samples_per_second": 268.872, | |
| "eval_steps_per_second": 1.382, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "learning_rate": 4.938271604938271e-05, | |
| "loss": 3.8038, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "learning_rate": 4.8611111111111115e-05, | |
| "loss": 3.0888, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.5920560747663551, | |
| "eval_loss": 2.1475517749786377, | |
| "eval_runtime": 8.0136, | |
| "eval_samples_per_second": 267.048, | |
| "eval_steps_per_second": 1.373, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 4.12, | |
| "learning_rate": 4.783950617283951e-05, | |
| "loss": 2.5659, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 4.54, | |
| "learning_rate": 4.70679012345679e-05, | |
| "loss": 2.1887, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.95, | |
| "learning_rate": 4.62962962962963e-05, | |
| "loss": 1.8458, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.99, | |
| "eval_accuracy": 0.6285046728971962, | |
| "eval_loss": 1.588365077972412, | |
| "eval_runtime": 8.4217, | |
| "eval_samples_per_second": 254.105, | |
| "eval_steps_per_second": 1.306, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 5.36, | |
| "learning_rate": 4.5524691358024696e-05, | |
| "loss": 1.6147, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 5.77, | |
| "learning_rate": 4.4753086419753084e-05, | |
| "loss": 1.4168, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 5.98, | |
| "eval_accuracy": 0.6588785046728972, | |
| "eval_loss": 1.2460048198699951, | |
| "eval_runtime": 7.9049, | |
| "eval_samples_per_second": 270.718, | |
| "eval_steps_per_second": 1.392, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 6.19, | |
| "learning_rate": 4.3981481481481486e-05, | |
| "loss": 1.271, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 6.6, | |
| "learning_rate": 4.3209876543209875e-05, | |
| "loss": 1.1951, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 6.97, | |
| "eval_accuracy": 0.6869158878504673, | |
| "eval_loss": 1.0658469200134277, | |
| "eval_runtime": 7.9886, | |
| "eval_samples_per_second": 267.881, | |
| "eval_steps_per_second": 1.377, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 7.01, | |
| "learning_rate": 4.243827160493827e-05, | |
| "loss": 1.1222, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 7.42, | |
| "learning_rate": 4.166666666666667e-05, | |
| "loss": 1.0719, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 7.84, | |
| "learning_rate": 4.089506172839506e-05, | |
| "loss": 0.9871, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.7037383177570093, | |
| "eval_loss": 0.9436482191085815, | |
| "eval_runtime": 8.7539, | |
| "eval_samples_per_second": 244.461, | |
| "eval_steps_per_second": 1.257, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 8.25, | |
| "learning_rate": 4.012345679012346e-05, | |
| "loss": 0.9461, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 8.66, | |
| "learning_rate": 3.935185185185186e-05, | |
| "loss": 0.9247, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 8.99, | |
| "eval_accuracy": 0.7238317757009346, | |
| "eval_loss": 0.8639922738075256, | |
| "eval_runtime": 8.532, | |
| "eval_samples_per_second": 250.821, | |
| "eval_steps_per_second": 1.289, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 9.07, | |
| "learning_rate": 3.8580246913580246e-05, | |
| "loss": 0.8806, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 9.48, | |
| "learning_rate": 3.780864197530865e-05, | |
| "loss": 0.8467, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 9.9, | |
| "learning_rate": 3.7037037037037037e-05, | |
| "loss": 0.826, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 9.98, | |
| "eval_accuracy": 0.7317757009345794, | |
| "eval_loss": 0.7976669073104858, | |
| "eval_runtime": 7.7765, | |
| "eval_samples_per_second": 275.189, | |
| "eval_steps_per_second": 1.415, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 10.31, | |
| "learning_rate": 3.626543209876543e-05, | |
| "loss": 0.79, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 10.72, | |
| "learning_rate": 3.5493827160493834e-05, | |
| "loss": 0.7964, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 10.97, | |
| "eval_accuracy": 0.7509345794392523, | |
| "eval_loss": 0.7539175152778625, | |
| "eval_runtime": 7.6095, | |
| "eval_samples_per_second": 281.226, | |
| "eval_steps_per_second": 1.446, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 11.13, | |
| "learning_rate": 3.472222222222222e-05, | |
| "loss": 0.7405, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 11.55, | |
| "learning_rate": 3.395061728395062e-05, | |
| "loss": 0.7471, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 11.96, | |
| "learning_rate": 3.317901234567901e-05, | |
| "loss": 0.7168, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.7514018691588785, | |
| "eval_loss": 0.7179247140884399, | |
| "eval_runtime": 8.7034, | |
| "eval_samples_per_second": 245.882, | |
| "eval_steps_per_second": 1.264, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 12.37, | |
| "learning_rate": 3.240740740740741e-05, | |
| "loss": 0.7166, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 12.78, | |
| "learning_rate": 3.16358024691358e-05, | |
| "loss": 0.6893, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 12.99, | |
| "eval_accuracy": 0.7630841121495328, | |
| "eval_loss": 0.6952459812164307, | |
| "eval_runtime": 8.8433, | |
| "eval_samples_per_second": 241.991, | |
| "eval_steps_per_second": 1.244, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 13.2, | |
| "learning_rate": 3.08641975308642e-05, | |
| "loss": 0.6942, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 13.61, | |
| "learning_rate": 3.0092592592592593e-05, | |
| "loss": 0.687, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 13.98, | |
| "eval_accuracy": 0.764018691588785, | |
| "eval_loss": 0.6812998652458191, | |
| "eval_runtime": 8.1933, | |
| "eval_samples_per_second": 261.19, | |
| "eval_steps_per_second": 1.343, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 14.02, | |
| "learning_rate": 2.9320987654320992e-05, | |
| "loss": 0.6621, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 14.43, | |
| "learning_rate": 2.8549382716049384e-05, | |
| "loss": 0.6721, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 14.85, | |
| "learning_rate": 2.777777777777778e-05, | |
| "loss": 0.6496, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 14.97, | |
| "eval_accuracy": 0.7691588785046729, | |
| "eval_loss": 0.6577351689338684, | |
| "eval_runtime": 7.8611, | |
| "eval_samples_per_second": 272.227, | |
| "eval_steps_per_second": 1.399, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 15.26, | |
| "learning_rate": 2.700617283950617e-05, | |
| "loss": 0.6403, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 15.67, | |
| "learning_rate": 2.623456790123457e-05, | |
| "loss": 0.6477, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.7813084112149533, | |
| "eval_loss": 0.6433616280555725, | |
| "eval_runtime": 7.6827, | |
| "eval_samples_per_second": 278.548, | |
| "eval_steps_per_second": 1.432, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 16.08, | |
| "learning_rate": 2.5462962962962965e-05, | |
| "loss": 0.6452, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 16.49, | |
| "learning_rate": 2.4691358024691357e-05, | |
| "loss": 0.6371, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 16.91, | |
| "learning_rate": 2.3919753086419755e-05, | |
| "loss": 0.5992, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 16.99, | |
| "eval_accuracy": 0.7771028037383177, | |
| "eval_loss": 0.6388670802116394, | |
| "eval_runtime": 8.24, | |
| "eval_samples_per_second": 259.708, | |
| "eval_steps_per_second": 1.335, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 17.32, | |
| "learning_rate": 2.314814814814815e-05, | |
| "loss": 0.6016, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 17.73, | |
| "learning_rate": 2.2376543209876542e-05, | |
| "loss": 0.6059, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 17.98, | |
| "eval_accuracy": 0.7906542056074767, | |
| "eval_loss": 0.623566746711731, | |
| "eval_runtime": 8.516, | |
| "eval_samples_per_second": 251.293, | |
| "eval_steps_per_second": 1.292, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 18.14, | |
| "learning_rate": 2.1604938271604937e-05, | |
| "loss": 0.6031, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 18.56, | |
| "learning_rate": 2.0833333333333336e-05, | |
| "loss": 0.6126, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 18.97, | |
| "learning_rate": 2.006172839506173e-05, | |
| "loss": 0.5993, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 18.97, | |
| "eval_accuracy": 0.7934579439252336, | |
| "eval_loss": 0.6181445717811584, | |
| "eval_runtime": 7.8023, | |
| "eval_samples_per_second": 274.277, | |
| "eval_steps_per_second": 1.41, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 19.38, | |
| "learning_rate": 1.9290123456790123e-05, | |
| "loss": 0.5872, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 19.79, | |
| "learning_rate": 1.8518518518518518e-05, | |
| "loss": 0.5802, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.7911214953271029, | |
| "eval_loss": 0.611458957195282, | |
| "eval_runtime": 8.4234, | |
| "eval_samples_per_second": 254.054, | |
| "eval_steps_per_second": 1.306, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 20.21, | |
| "learning_rate": 1.7746913580246917e-05, | |
| "loss": 0.6008, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 20.62, | |
| "learning_rate": 1.697530864197531e-05, | |
| "loss": 0.5905, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 20.99, | |
| "eval_accuracy": 0.7929906542056074, | |
| "eval_loss": 0.6005003452301025, | |
| "eval_runtime": 7.7149, | |
| "eval_samples_per_second": 277.386, | |
| "eval_steps_per_second": 1.426, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 21.03, | |
| "learning_rate": 1.6203703703703704e-05, | |
| "loss": 0.5861, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 21.44, | |
| "learning_rate": 1.54320987654321e-05, | |
| "loss": 0.5865, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 21.86, | |
| "learning_rate": 1.4660493827160496e-05, | |
| "loss": 0.5825, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 21.98, | |
| "eval_accuracy": 0.7953271028037383, | |
| "eval_loss": 0.5945016145706177, | |
| "eval_runtime": 7.5452, | |
| "eval_samples_per_second": 283.623, | |
| "eval_steps_per_second": 1.458, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 22.27, | |
| "learning_rate": 1.388888888888889e-05, | |
| "loss": 0.5732, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 22.68, | |
| "learning_rate": 1.3117283950617285e-05, | |
| "loss": 0.5837, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 22.97, | |
| "eval_accuracy": 0.7981308411214953, | |
| "eval_loss": 0.5882120132446289, | |
| "eval_runtime": 8.1333, | |
| "eval_samples_per_second": 263.116, | |
| "eval_steps_per_second": 1.352, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 23.09, | |
| "learning_rate": 1.2345679012345678e-05, | |
| "loss": 0.5791, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 23.51, | |
| "learning_rate": 1.1574074074074075e-05, | |
| "loss": 0.5783, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 23.92, | |
| "learning_rate": 1.0802469135802469e-05, | |
| "loss": 0.5501, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_accuracy": 0.8023364485981308, | |
| "eval_loss": 0.5847517251968384, | |
| "eval_runtime": 8.0927, | |
| "eval_samples_per_second": 264.435, | |
| "eval_steps_per_second": 1.359, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 24.33, | |
| "learning_rate": 1.0030864197530866e-05, | |
| "loss": 0.5841, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 24.74, | |
| "learning_rate": 9.259259259259259e-06, | |
| "loss": 0.5593, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 24.99, | |
| "eval_accuracy": 0.7962616822429907, | |
| "eval_loss": 0.5861061811447144, | |
| "eval_runtime": 7.5867, | |
| "eval_samples_per_second": 282.074, | |
| "eval_steps_per_second": 1.45, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 25.15, | |
| "learning_rate": 8.487654320987654e-06, | |
| "loss": 0.5526, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 25.57, | |
| "learning_rate": 7.71604938271605e-06, | |
| "loss": 0.5792, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 25.98, | |
| "learning_rate": 6.944444444444445e-06, | |
| "loss": 0.5454, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 25.98, | |
| "eval_accuracy": 0.8014018691588785, | |
| "eval_loss": 0.5776079893112183, | |
| "eval_runtime": 7.665, | |
| "eval_samples_per_second": 279.191, | |
| "eval_steps_per_second": 1.435, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 26.39, | |
| "learning_rate": 6.172839506172839e-06, | |
| "loss": 0.5477, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 26.8, | |
| "learning_rate": 5.401234567901234e-06, | |
| "loss": 0.5656, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 26.97, | |
| "eval_accuracy": 0.7976635514018692, | |
| "eval_loss": 0.579258143901825, | |
| "eval_runtime": 7.9989, | |
| "eval_samples_per_second": 267.536, | |
| "eval_steps_per_second": 1.375, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 27.22, | |
| "learning_rate": 4.6296296296296296e-06, | |
| "loss": 0.5546, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 27.63, | |
| "learning_rate": 3.858024691358025e-06, | |
| "loss": 0.5708, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "eval_accuracy": 0.8004672897196262, | |
| "eval_loss": 0.5814595818519592, | |
| "eval_runtime": 8.2466, | |
| "eval_samples_per_second": 259.502, | |
| "eval_steps_per_second": 1.334, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 28.04, | |
| "learning_rate": 3.0864197530864196e-06, | |
| "loss": 0.5505, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 28.45, | |
| "learning_rate": 2.3148148148148148e-06, | |
| "loss": 0.5565, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 28.87, | |
| "learning_rate": 1.5432098765432098e-06, | |
| "loss": 0.5541, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 28.99, | |
| "eval_accuracy": 0.8032710280373832, | |
| "eval_loss": 0.5749428868293762, | |
| "eval_runtime": 8.685, | |
| "eval_samples_per_second": 246.402, | |
| "eval_steps_per_second": 1.267, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 29.28, | |
| "learning_rate": 7.716049382716049e-07, | |
| "loss": 0.5511, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 29.69, | |
| "learning_rate": 0.0, | |
| "loss": 0.5439, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 29.69, | |
| "eval_accuracy": 0.7995327102803739, | |
| "eval_loss": 0.5757396221160889, | |
| "eval_runtime": 7.6093, | |
| "eval_samples_per_second": 281.236, | |
| "eval_steps_per_second": 1.446, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 29.69, | |
| "step": 720, | |
| "total_flos": 3.1088341164957254e+19, | |
| "train_loss": 1.3989773021803962, | |
| "train_runtime": 2686.7823, | |
| "train_samples_per_second": 215.019, | |
| "train_steps_per_second": 0.268 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 720, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 30, | |
| "save_steps": 500, | |
| "total_flos": 3.1088341164957254e+19, | |
| "train_batch_size": 200, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |