| { | |
| "best_metric": 0.8854679802955665, | |
| "best_model_checkpoint": "vit-msn-small-wbc-classifier-0316-cropped-cleaned-dataset-10/checkpoint-357", | |
| "epoch": 25.0, | |
| "eval_steps": 500, | |
| "global_step": 425, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.5882352941176471, | |
| "grad_norm": 12.153236389160156, | |
| "learning_rate": 1.1627906976744187e-05, | |
| "loss": 1.3709, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.805008210180624, | |
| "eval_loss": 0.6976552605628967, | |
| "eval_runtime": 7.6934, | |
| "eval_samples_per_second": 316.633, | |
| "eval_steps_per_second": 5.069, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 1.1764705882352942, | |
| "grad_norm": 14.669098854064941, | |
| "learning_rate": 2.3255813953488374e-05, | |
| "loss": 0.7911, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.7647058823529411, | |
| "grad_norm": 5.835587024688721, | |
| "learning_rate": 3.488372093023256e-05, | |
| "loss": 0.5673, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8099343185550082, | |
| "eval_loss": 0.5948964357376099, | |
| "eval_runtime": 7.6055, | |
| "eval_samples_per_second": 320.293, | |
| "eval_steps_per_second": 5.128, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 2.3529411764705883, | |
| "grad_norm": 10.460927963256836, | |
| "learning_rate": 4.651162790697675e-05, | |
| "loss": 0.5463, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 2.9411764705882355, | |
| "grad_norm": 9.529006004333496, | |
| "learning_rate": 4.9083769633507855e-05, | |
| "loss": 0.5227, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7931034482758621, | |
| "eval_loss": 0.6151543259620667, | |
| "eval_runtime": 7.5675, | |
| "eval_samples_per_second": 321.904, | |
| "eval_steps_per_second": 5.154, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 3.5294117647058822, | |
| "grad_norm": 6.369431495666504, | |
| "learning_rate": 4.7774869109947645e-05, | |
| "loss": 0.4958, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8435960591133005, | |
| "eval_loss": 0.4351496994495392, | |
| "eval_runtime": 7.5745, | |
| "eval_samples_per_second": 321.605, | |
| "eval_steps_per_second": 5.149, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 4.117647058823529, | |
| "grad_norm": 10.749176979064941, | |
| "learning_rate": 4.6465968586387436e-05, | |
| "loss": 0.483, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 4.705882352941177, | |
| "grad_norm": 11.76589584350586, | |
| "learning_rate": 4.5157068062827226e-05, | |
| "loss": 0.4402, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8579638752052545, | |
| "eval_loss": 0.3776942789554596, | |
| "eval_runtime": 7.5784, | |
| "eval_samples_per_second": 321.442, | |
| "eval_steps_per_second": 5.146, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 5.294117647058823, | |
| "grad_norm": 4.830715656280518, | |
| "learning_rate": 4.384816753926702e-05, | |
| "loss": 0.418, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 5.882352941176471, | |
| "grad_norm": 7.1909027099609375, | |
| "learning_rate": 4.253926701570681e-05, | |
| "loss": 0.3878, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8698686371100164, | |
| "eval_loss": 0.39695626497268677, | |
| "eval_runtime": 7.4679, | |
| "eval_samples_per_second": 326.194, | |
| "eval_steps_per_second": 5.222, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 6.470588235294118, | |
| "grad_norm": 4.031527042388916, | |
| "learning_rate": 4.12303664921466e-05, | |
| "loss": 0.3646, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8641215106732348, | |
| "eval_loss": 0.3792504668235779, | |
| "eval_runtime": 7.7676, | |
| "eval_samples_per_second": 313.611, | |
| "eval_steps_per_second": 5.021, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 7.0588235294117645, | |
| "grad_norm": 5.431394100189209, | |
| "learning_rate": 3.992146596858639e-05, | |
| "loss": 0.3965, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 7.647058823529412, | |
| "grad_norm": 4.352228164672852, | |
| "learning_rate": 3.861256544502618e-05, | |
| "loss": 0.3452, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8805418719211823, | |
| "eval_loss": 0.3549965023994446, | |
| "eval_runtime": 7.4981, | |
| "eval_samples_per_second": 324.881, | |
| "eval_steps_per_second": 5.201, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 8.235294117647058, | |
| "grad_norm": 13.60920524597168, | |
| "learning_rate": 3.730366492146597e-05, | |
| "loss": 0.3754, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 8.823529411764707, | |
| "grad_norm": 7.518331050872803, | |
| "learning_rate": 3.599476439790576e-05, | |
| "loss": 0.344, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.8735632183908046, | |
| "eval_loss": 0.40034234523773193, | |
| "eval_runtime": 7.6227, | |
| "eval_samples_per_second": 319.572, | |
| "eval_steps_per_second": 5.116, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 9.411764705882353, | |
| "grad_norm": 7.200953960418701, | |
| "learning_rate": 3.468586387434556e-05, | |
| "loss": 0.3418, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 7.663539886474609, | |
| "learning_rate": 3.337696335078534e-05, | |
| "loss": 0.3365, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.8830049261083743, | |
| "eval_loss": 0.36542433500289917, | |
| "eval_runtime": 7.6247, | |
| "eval_samples_per_second": 319.489, | |
| "eval_steps_per_second": 5.115, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 10.588235294117647, | |
| "grad_norm": 4.720971584320068, | |
| "learning_rate": 3.206806282722513e-05, | |
| "loss": 0.3223, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.8764367816091954, | |
| "eval_loss": 0.3570525646209717, | |
| "eval_runtime": 7.6607, | |
| "eval_samples_per_second": 317.988, | |
| "eval_steps_per_second": 5.091, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 11.176470588235293, | |
| "grad_norm": 4.508999824523926, | |
| "learning_rate": 3.075916230366492e-05, | |
| "loss": 0.3084, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 11.764705882352942, | |
| "grad_norm": 7.280440807342529, | |
| "learning_rate": 2.9450261780104715e-05, | |
| "loss": 0.2819, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.8788998357963875, | |
| "eval_loss": 0.3665020167827606, | |
| "eval_runtime": 7.4797, | |
| "eval_samples_per_second": 325.683, | |
| "eval_steps_per_second": 5.214, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 12.352941176470589, | |
| "grad_norm": 5.721529483795166, | |
| "learning_rate": 2.8141361256544502e-05, | |
| "loss": 0.3067, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 12.941176470588236, | |
| "grad_norm": 5.043443202972412, | |
| "learning_rate": 2.6832460732984293e-05, | |
| "loss": 0.2998, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.8838259441707718, | |
| "eval_loss": 0.36087653040885925, | |
| "eval_runtime": 7.7018, | |
| "eval_samples_per_second": 316.291, | |
| "eval_steps_per_second": 5.064, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 13.529411764705882, | |
| "grad_norm": 9.1295804977417, | |
| "learning_rate": 2.5523560209424086e-05, | |
| "loss": 0.2959, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.8719211822660099, | |
| "eval_loss": 0.4335402846336365, | |
| "eval_runtime": 7.5815, | |
| "eval_samples_per_second": 321.307, | |
| "eval_steps_per_second": 5.144, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 14.117647058823529, | |
| "grad_norm": 6.499159812927246, | |
| "learning_rate": 2.4214659685863873e-05, | |
| "loss": 0.2732, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 14.705882352941176, | |
| "grad_norm": 5.268309116363525, | |
| "learning_rate": 2.2905759162303667e-05, | |
| "loss": 0.2662, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.8784893267651889, | |
| "eval_loss": 0.42450225353240967, | |
| "eval_runtime": 7.7843, | |
| "eval_samples_per_second": 312.939, | |
| "eval_steps_per_second": 5.01, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 15.294117647058824, | |
| "grad_norm": 3.793020725250244, | |
| "learning_rate": 2.1596858638743454e-05, | |
| "loss": 0.2693, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 15.882352941176471, | |
| "grad_norm": 8.22383975982666, | |
| "learning_rate": 2.0287958115183248e-05, | |
| "loss": 0.2668, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.8846469622331691, | |
| "eval_loss": 0.37603434920310974, | |
| "eval_runtime": 7.5418, | |
| "eval_samples_per_second": 322.999, | |
| "eval_steps_per_second": 5.171, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 16.470588235294116, | |
| "grad_norm": 4.452105522155762, | |
| "learning_rate": 1.8979057591623035e-05, | |
| "loss": 0.2576, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.8830049261083743, | |
| "eval_loss": 0.3727973699569702, | |
| "eval_runtime": 7.5894, | |
| "eval_samples_per_second": 320.973, | |
| "eval_steps_per_second": 5.139, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 17.058823529411764, | |
| "grad_norm": 5.180832386016846, | |
| "learning_rate": 1.767015706806283e-05, | |
| "loss": 0.2578, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 17.647058823529413, | |
| "grad_norm": 6.889744758605957, | |
| "learning_rate": 1.636125654450262e-05, | |
| "loss": 0.2398, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.8813628899835796, | |
| "eval_loss": 0.41920730471611023, | |
| "eval_runtime": 7.5395, | |
| "eval_samples_per_second": 323.1, | |
| "eval_steps_per_second": 5.173, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 18.235294117647058, | |
| "grad_norm": 5.117687225341797, | |
| "learning_rate": 1.505235602094241e-05, | |
| "loss": 0.2353, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 18.823529411764707, | |
| "grad_norm": 5.613910675048828, | |
| "learning_rate": 1.3743455497382199e-05, | |
| "loss": 0.2278, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.8805418719211823, | |
| "eval_loss": 0.41564103960990906, | |
| "eval_runtime": 7.6556, | |
| "eval_samples_per_second": 318.197, | |
| "eval_steps_per_second": 5.094, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 19.41176470588235, | |
| "grad_norm": 5.283733367919922, | |
| "learning_rate": 1.243455497382199e-05, | |
| "loss": 0.2297, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 5.314838886260986, | |
| "learning_rate": 1.1125654450261781e-05, | |
| "loss": 0.2033, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.8850574712643678, | |
| "eval_loss": 0.41589659452438354, | |
| "eval_runtime": 7.4417, | |
| "eval_samples_per_second": 327.342, | |
| "eval_steps_per_second": 5.241, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 20.58823529411765, | |
| "grad_norm": 5.559089183807373, | |
| "learning_rate": 9.816753926701572e-06, | |
| "loss": 0.2037, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "eval_accuracy": 0.8854679802955665, | |
| "eval_loss": 0.3985845744609833, | |
| "eval_runtime": 7.5913, | |
| "eval_samples_per_second": 320.893, | |
| "eval_steps_per_second": 5.137, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 21.176470588235293, | |
| "grad_norm": 4.5341596603393555, | |
| "learning_rate": 8.507853403141362e-06, | |
| "loss": 0.2106, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 21.764705882352942, | |
| "grad_norm": 4.3524580001831055, | |
| "learning_rate": 7.1989528795811526e-06, | |
| "loss": 0.1934, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "eval_accuracy": 0.882183908045977, | |
| "eval_loss": 0.42198774218559265, | |
| "eval_runtime": 7.5092, | |
| "eval_samples_per_second": 324.401, | |
| "eval_steps_per_second": 5.194, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 22.352941176470587, | |
| "grad_norm": 6.747894763946533, | |
| "learning_rate": 5.890052356020943e-06, | |
| "loss": 0.2043, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 22.941176470588236, | |
| "grad_norm": 5.149715900421143, | |
| "learning_rate": 4.5811518324607335e-06, | |
| "loss": 0.1983, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "eval_accuracy": 0.8854679802955665, | |
| "eval_loss": 0.4159087836742401, | |
| "eval_runtime": 7.4544, | |
| "eval_samples_per_second": 326.786, | |
| "eval_steps_per_second": 5.232, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 23.529411764705884, | |
| "grad_norm": 4.871130466461182, | |
| "learning_rate": 3.272251308900524e-06, | |
| "loss": 0.1746, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_accuracy": 0.8854679802955665, | |
| "eval_loss": 0.41789308190345764, | |
| "eval_runtime": 7.7744, | |
| "eval_samples_per_second": 313.335, | |
| "eval_steps_per_second": 5.016, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 24.11764705882353, | |
| "grad_norm": 4.728577136993408, | |
| "learning_rate": 1.9633507853403143e-06, | |
| "loss": 0.1925, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 24.705882352941178, | |
| "grad_norm": 5.179035663604736, | |
| "learning_rate": 6.544502617801048e-07, | |
| "loss": 0.1776, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "eval_accuracy": 0.8834154351395731, | |
| "eval_loss": 0.42465320229530334, | |
| "eval_runtime": 7.6371, | |
| "eval_samples_per_second": 318.97, | |
| "eval_steps_per_second": 5.107, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "step": 425, | |
| "total_flos": 2.113465532645376e+18, | |
| "train_loss": 0.34626551782383636, | |
| "train_runtime": 1013.337, | |
| "train_samples_per_second": 106.579, | |
| "train_steps_per_second": 0.419 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 425, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 25, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.113465532645376e+18, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |