| { | |
| "best_global_step": 500, | |
| "best_metric": 0.9950738019660021, | |
| "best_model_checkpoint": "hiera-finetuned-stroke-binary-ultrasound/checkpoint-500", | |
| "epoch": 12.0, | |
| "eval_steps": 100, | |
| "global_step": 1380, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.437636761487965, | |
| "grad_norm": 4.08018684387207, | |
| "learning_rate": 6.811594202898551e-06, | |
| "loss": 0.0505, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.87527352297593, | |
| "grad_norm": 0.09228578954935074, | |
| "learning_rate": 1.4057971014492755e-05, | |
| "loss": 0.038, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.87527352297593, | |
| "eval_accuracy": 0.9938423645320197, | |
| "eval_f1": 0.9938419442238937, | |
| "eval_loss": 0.018425090238451958, | |
| "eval_precision": 0.9939169121769106, | |
| "eval_recall": 0.9938423645320197, | |
| "eval_runtime": 4.7821, | |
| "eval_samples_per_second": 169.799, | |
| "eval_steps_per_second": 21.329, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.3063457330415755, | |
| "grad_norm": 13.409342765808105, | |
| "learning_rate": 1.9997952651793994e-05, | |
| "loss": 0.0537, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.7439824945295404, | |
| "grad_norm": 4.2889485359191895, | |
| "learning_rate": 1.9892575474365977e-05, | |
| "loss": 0.0468, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.7439824945295404, | |
| "eval_accuracy": 0.9926108374384236, | |
| "eval_f1": 0.9926101649260403, | |
| "eval_loss": 0.02063354291021824, | |
| "eval_precision": 0.9927179267509103, | |
| "eval_recall": 0.9926108374384236, | |
| "eval_runtime": 4.2841, | |
| "eval_samples_per_second": 189.54, | |
| "eval_steps_per_second": 23.809, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.175054704595186, | |
| "grad_norm": 27.501617431640625, | |
| "learning_rate": 1.9629172873477995e-05, | |
| "loss": 0.055, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.612691466083151, | |
| "grad_norm": 34.34234619140625, | |
| "learning_rate": 1.9211952480161382e-05, | |
| "loss": 0.0445, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.612691466083151, | |
| "eval_accuracy": 0.9901477832512315, | |
| "eval_f1": 0.9901478430214063, | |
| "eval_loss": 0.02250731736421585, | |
| "eval_precision": 0.9901599165230894, | |
| "eval_recall": 0.9901477832512315, | |
| "eval_runtime": 4.3458, | |
| "eval_samples_per_second": 186.849, | |
| "eval_steps_per_second": 23.471, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.0437636761487963, | |
| "grad_norm": 1.1840153932571411, | |
| "learning_rate": 1.866025403784439e-05, | |
| "loss": 0.0309, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 3.4814004376367613, | |
| "grad_norm": 0.7110581994056702, | |
| "learning_rate": 1.7960402776569358e-05, | |
| "loss": 0.0415, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.4814004376367613, | |
| "eval_accuracy": 0.9889162561576355, | |
| "eval_f1": 0.9889161721031502, | |
| "eval_loss": 0.018740125000476837, | |
| "eval_precision": 0.9889190879070968, | |
| "eval_recall": 0.9889162561576355, | |
| "eval_runtime": 4.4065, | |
| "eval_samples_per_second": 184.274, | |
| "eval_steps_per_second": 23.148, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.9190371991247264, | |
| "grad_norm": 44.404239654541016, | |
| "learning_rate": 1.7133390896410106e-05, | |
| "loss": 0.034, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 4.350109409190372, | |
| "grad_norm": 0.5970667004585266, | |
| "learning_rate": 1.6192429204088022e-05, | |
| "loss": 0.0465, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.350109409190372, | |
| "eval_accuracy": 0.9950738916256158, | |
| "eval_f1": 0.9950738019660021, | |
| "eval_loss": 0.009755603037774563, | |
| "eval_precision": 0.9950857869728612, | |
| "eval_recall": 0.9950738916256158, | |
| "eval_runtime": 5.2229, | |
| "eval_samples_per_second": 155.47, | |
| "eval_steps_per_second": 19.53, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.787746170678337, | |
| "grad_norm": 0.04525822028517723, | |
| "learning_rate": 1.5152548757013183e-05, | |
| "loss": 0.0284, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 5.218818380743983, | |
| "grad_norm": 6.67152214050293, | |
| "learning_rate": 1.4030360755011423e-05, | |
| "loss": 0.0397, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 5.218818380743983, | |
| "eval_accuracy": 0.9901477832512315, | |
| "eval_f1": 0.9901463484623846, | |
| "eval_loss": 0.028631530702114105, | |
| "eval_precision": 0.9903372489579386, | |
| "eval_recall": 0.9901477832512315, | |
| "eval_runtime": 5.1408, | |
| "eval_samples_per_second": 157.953, | |
| "eval_steps_per_second": 19.841, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 5.656455142231947, | |
| "grad_norm": 2.804370641708374, | |
| "learning_rate": 1.284379119062912e-05, | |
| "loss": 0.0334, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 6.087527352297593, | |
| "grad_norm": 0.5940968990325928, | |
| "learning_rate": 1.1611794496750019e-05, | |
| "loss": 0.0257, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 6.087527352297593, | |
| "eval_accuracy": 0.9926108374384236, | |
| "eval_f1": 0.9926104787847378, | |
| "eval_loss": 0.01877174898982048, | |
| "eval_precision": 0.992658305036109, | |
| "eval_recall": 0.9926108374384236, | |
| "eval_runtime": 4.3633, | |
| "eval_samples_per_second": 186.097, | |
| "eval_steps_per_second": 23.377, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 6.525164113785558, | |
| "grad_norm": 6.283336162567139, | |
| "learning_rate": 1.0354050765758148e-05, | |
| "loss": 0.0226, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 6.962800875273523, | |
| "grad_norm": 0.4039391279220581, | |
| "learning_rate": 9.090651376911532e-06, | |
| "loss": 0.0434, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 6.962800875273523, | |
| "eval_accuracy": 0.9938423645320197, | |
| "eval_f1": 0.9938419442238937, | |
| "eval_loss": 0.02092530019581318, | |
| "eval_precision": 0.9939169121769106, | |
| "eval_recall": 0.9938423645320197, | |
| "eval_runtime": 5.4149, | |
| "eval_samples_per_second": 149.956, | |
| "eval_steps_per_second": 18.837, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 7.393873085339169, | |
| "grad_norm": 26.558956146240234, | |
| "learning_rate": 7.841778053760212e-06, | |
| "loss": 0.0256, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 7.831509846827133, | |
| "grad_norm": 1.8528072834014893, | |
| "learning_rate": 6.627380478391543e-06, | |
| "loss": 0.0261, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 7.831509846827133, | |
| "eval_accuracy": 0.9926108374384236, | |
| "eval_f1": 0.9926107029490034, | |
| "eval_loss": 0.015434592962265015, | |
| "eval_precision": 0.9926226132344405, | |
| "eval_recall": 0.9926108374384236, | |
| "eval_runtime": 4.4967, | |
| "eval_samples_per_second": 180.577, | |
| "eval_steps_per_second": 22.683, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 8.262582056892779, | |
| "grad_norm": 0.028279444202780724, | |
| "learning_rate": 5.466857612339229e-06, | |
| "loss": 0.0132, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 8.700218818380744, | |
| "grad_norm": 10.328960418701172, | |
| "learning_rate": 4.378747814781629e-06, | |
| "loss": 0.0198, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 8.700218818380744, | |
| "eval_accuracy": 0.9950738916256158, | |
| "eval_f1": 0.9950736525231585, | |
| "eval_loss": 0.009430773556232452, | |
| "eval_precision": 0.9951217179205126, | |
| "eval_recall": 0.9950738916256158, | |
| "eval_runtime": 5.241, | |
| "eval_samples_per_second": 154.932, | |
| "eval_steps_per_second": 19.462, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 9.13129102844639, | |
| "grad_norm": 5.253521919250488, | |
| "learning_rate": 3.3804327081259304e-06, | |
| "loss": 0.0244, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 9.568927789934355, | |
| "grad_norm": 0.006149813067167997, | |
| "learning_rate": 2.4878595214718236e-06, | |
| "loss": 0.0207, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 9.568927789934355, | |
| "eval_accuracy": 0.9938423645320197, | |
| "eval_f1": 0.9938421683977477, | |
| "eval_loss": 0.012234929017722607, | |
| "eval_precision": 0.9938691081078003, | |
| "eval_recall": 0.9938423645320197, | |
| "eval_runtime": 4.3976, | |
| "eval_samples_per_second": 184.648, | |
| "eval_steps_per_second": 23.195, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.003324387362226844, | |
| "learning_rate": 1.7152863472802195e-06, | |
| "loss": 0.0212, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 10.437636761487965, | |
| "grad_norm": 5.904305934906006, | |
| "learning_rate": 1.075054380553552e-06, | |
| "loss": 0.0157, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 10.437636761487965, | |
| "eval_accuracy": 0.9950738916256158, | |
| "eval_f1": 0.9950736525231585, | |
| "eval_loss": 0.01014081109315157, | |
| "eval_precision": 0.9951217179205126, | |
| "eval_recall": 0.9950738916256158, | |
| "eval_runtime": 5.4913, | |
| "eval_samples_per_second": 147.869, | |
| "eval_steps_per_second": 18.575, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 10.87527352297593, | |
| "grad_norm": 0.6864893436431885, | |
| "learning_rate": 5.77390778811796e-07, | |
| "loss": 0.0143, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 11.306345733041576, | |
| "grad_norm": 39.91190719604492, | |
| "learning_rate": 2.3024529200728952e-07, | |
| "loss": 0.0188, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 11.306345733041576, | |
| "eval_accuracy": 0.9950738916256158, | |
| "eval_f1": 0.9950736525231585, | |
| "eval_loss": 0.010448944754898548, | |
| "eval_precision": 0.9951217179205126, | |
| "eval_recall": 0.9950738916256158, | |
| "eval_runtime": 5.3991, | |
| "eval_samples_per_second": 150.395, | |
| "eval_steps_per_second": 18.892, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 11.74398249452954, | |
| "grad_norm": 40.59695816040039, | |
| "learning_rate": 3.916327207559967e-08, | |
| "loss": 0.0198, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "step": 1380, | |
| "total_flos": 4.018008276329693e+18, | |
| "train_loss": 0.031628868545311085, | |
| "train_runtime": 1401.2705, | |
| "train_samples_per_second": 62.549, | |
| "train_steps_per_second": 0.985 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 1380, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 12, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.018008276329693e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |