| { | |
| "best_metric": 0.9141630901287554, | |
| "best_model_checkpoint": "beit-base-patch16-224-pt22k-ft22k-finetuned-lora-medmnistv2/checkpoint-63", | |
| "epoch": 9.882352941176471, | |
| "eval_steps": 500, | |
| "global_step": 630, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 2.7611124515533447, | |
| "learning_rate": 0.004920634920634921, | |
| "loss": 0.6696, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 3.1516690254211426, | |
| "learning_rate": 0.004841269841269842, | |
| "loss": 0.5037, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 1.6632912158966064, | |
| "learning_rate": 0.0047619047619047615, | |
| "loss": 0.5343, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 15.660941123962402, | |
| "learning_rate": 0.004682539682539683, | |
| "loss": 0.7623, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 1.5720008611679077, | |
| "learning_rate": 0.004603174603174603, | |
| "loss": 0.5904, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 3.288884401321411, | |
| "learning_rate": 0.004523809523809524, | |
| "loss": 0.4775, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "eval_accuracy": 0.9141630901287554, | |
| "eval_f1": 0.8903401432261271, | |
| "eval_loss": 0.22644232213497162, | |
| "eval_precision": 0.884993651288677, | |
| "eval_recall": 0.8961752704933064, | |
| "eval_runtime": 12.2128, | |
| "eval_samples_per_second": 95.391, | |
| "eval_steps_per_second": 5.977, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 2.6683709621429443, | |
| "learning_rate": 0.0044603174603174605, | |
| "loss": 0.4832, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 1.2820926904678345, | |
| "learning_rate": 0.004380952380952381, | |
| "loss": 0.5064, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 5.157268047332764, | |
| "learning_rate": 0.004301587301587302, | |
| "loss": 0.4561, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 0.9834429025650024, | |
| "learning_rate": 0.004222222222222223, | |
| "loss": 0.3981, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 7.78388786315918, | |
| "learning_rate": 0.0041428571428571434, | |
| "loss": 0.5016, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 12.203431129455566, | |
| "learning_rate": 0.004063492063492063, | |
| "loss": 0.7117, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "eval_accuracy": 0.7390557939914163, | |
| "eval_f1": 0.4249753208292202, | |
| "eval_loss": 0.4007842242717743, | |
| "eval_precision": 0.36952789699570815, | |
| "eval_recall": 0.5, | |
| "eval_runtime": 12.6185, | |
| "eval_samples_per_second": 92.325, | |
| "eval_steps_per_second": 5.785, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "grad_norm": 9.85489273071289, | |
| "learning_rate": 0.003992063492063492, | |
| "loss": 0.4575, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 2.926109552383423, | |
| "learning_rate": 0.003912698412698413, | |
| "loss": 0.4562, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 0.5027350187301636, | |
| "learning_rate": 0.0038333333333333336, | |
| "loss": 0.417, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "grad_norm": 0.8259086608886719, | |
| "learning_rate": 0.003753968253968254, | |
| "loss": 0.3953, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "grad_norm": 2.8119688034057617, | |
| "learning_rate": 0.003674603174603175, | |
| "loss": 0.4175, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "grad_norm": 2.306551218032837, | |
| "learning_rate": 0.0035952380952380954, | |
| "loss": 0.4226, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.98, | |
| "grad_norm": 6.00565242767334, | |
| "learning_rate": 0.003515873015873016, | |
| "loss": 0.4115, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8154506437768241, | |
| "eval_f1": 0.7957275475643482, | |
| "eval_loss": 0.43578293919563293, | |
| "eval_precision": 0.7870816708492059, | |
| "eval_recall": 0.8645050125313283, | |
| "eval_runtime": 12.1326, | |
| "eval_samples_per_second": 96.022, | |
| "eval_steps_per_second": 6.017, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "grad_norm": 1.8762693405151367, | |
| "learning_rate": 0.0034365079365079364, | |
| "loss": 0.3876, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.29, | |
| "grad_norm": 1.1961586475372314, | |
| "learning_rate": 0.003357142857142857, | |
| "loss": 0.3232, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 3.45, | |
| "grad_norm": 0.5015142560005188, | |
| "learning_rate": 0.003277777777777778, | |
| "loss": 0.4061, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 3.61, | |
| "grad_norm": 2.750140428543091, | |
| "learning_rate": 0.003198412698412698, | |
| "loss": 0.372, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 3.76, | |
| "grad_norm": 2.029989242553711, | |
| "learning_rate": 0.003119047619047619, | |
| "loss": 0.3498, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.92, | |
| "grad_norm": 2.366288661956787, | |
| "learning_rate": 0.0030396825396825397, | |
| "loss": 0.3631, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8798283261802575, | |
| "eval_f1": 0.8517509925684618, | |
| "eval_loss": 0.30910125374794006, | |
| "eval_precision": 0.8380756825290672, | |
| "eval_recall": 0.8708184332783178, | |
| "eval_runtime": 12.051, | |
| "eval_samples_per_second": 96.673, | |
| "eval_steps_per_second": 6.058, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 4.08, | |
| "grad_norm": 0.5986310839653015, | |
| "learning_rate": 0.0029603174603174604, | |
| "loss": 0.3413, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 4.24, | |
| "grad_norm": 4.78090763092041, | |
| "learning_rate": 0.0028809523809523807, | |
| "loss": 0.384, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 4.39, | |
| "grad_norm": 3.392075300216675, | |
| "learning_rate": 0.0028015873015873015, | |
| "loss": 0.3579, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 4.55, | |
| "grad_norm": 1.982884407043457, | |
| "learning_rate": 0.0027222222222222222, | |
| "loss": 0.3397, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 4.71, | |
| "grad_norm": 4.884500980377197, | |
| "learning_rate": 0.002642857142857143, | |
| "loss": 0.3609, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 4.86, | |
| "grad_norm": 0.9579557776451111, | |
| "learning_rate": 0.0025634920634920633, | |
| "loss": 0.3794, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 4.99, | |
| "eval_accuracy": 0.8798283261802575, | |
| "eval_f1": 0.8494965377744312, | |
| "eval_loss": 0.2801721394062042, | |
| "eval_precision": 0.8392582443135881, | |
| "eval_recall": 0.8623062992847974, | |
| "eval_runtime": 12.197, | |
| "eval_samples_per_second": 95.516, | |
| "eval_steps_per_second": 5.985, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 5.02, | |
| "grad_norm": 1.1442433595657349, | |
| "learning_rate": 0.002484126984126984, | |
| "loss": 0.3566, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 5.18, | |
| "grad_norm": 4.931128978729248, | |
| "learning_rate": 0.0024047619047619048, | |
| "loss": 0.4874, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 5.33, | |
| "grad_norm": 2.474273443222046, | |
| "learning_rate": 0.0023253968253968255, | |
| "loss": 0.375, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 5.49, | |
| "grad_norm": 2.144062042236328, | |
| "learning_rate": 0.0022460317460317463, | |
| "loss": 0.368, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 5.65, | |
| "grad_norm": 0.6386366486549377, | |
| "learning_rate": 0.002166666666666667, | |
| "loss": 0.3755, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 5.8, | |
| "grad_norm": 3.0398051738739014, | |
| "learning_rate": 0.0020873015873015873, | |
| "loss": 0.3248, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 5.96, | |
| "grad_norm": 2.5777676105499268, | |
| "learning_rate": 0.002007936507936508, | |
| "loss": 0.3713, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 5.99, | |
| "eval_accuracy": 0.8772532188841202, | |
| "eval_f1": 0.8449126373928156, | |
| "eval_loss": 0.2805176377296448, | |
| "eval_precision": 0.8370533441922019, | |
| "eval_recall": 0.8541800385109115, | |
| "eval_runtime": 11.9639, | |
| "eval_samples_per_second": 97.376, | |
| "eval_steps_per_second": 6.102, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 6.12, | |
| "grad_norm": 0.9707914590835571, | |
| "learning_rate": 0.0019285714285714288, | |
| "loss": 0.3227, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 6.27, | |
| "grad_norm": 1.1869500875473022, | |
| "learning_rate": 0.0018492063492063493, | |
| "loss": 0.3271, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 6.43, | |
| "grad_norm": 0.6983945369720459, | |
| "learning_rate": 0.00176984126984127, | |
| "loss": 0.3709, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 6.59, | |
| "grad_norm": 1.19561767578125, | |
| "learning_rate": 0.0016904761904761906, | |
| "loss": 0.3629, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 6.75, | |
| "grad_norm": 1.7224721908569336, | |
| "learning_rate": 0.0016111111111111113, | |
| "loss": 0.3182, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 6.9, | |
| "grad_norm": 2.228806972503662, | |
| "learning_rate": 0.0015317460317460319, | |
| "loss": 0.3953, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8583690987124464, | |
| "eval_f1": 0.836744521418762, | |
| "eval_loss": 0.3396996855735779, | |
| "eval_precision": 0.8185304398119043, | |
| "eval_recall": 0.8871569166819487, | |
| "eval_runtime": 12.0715, | |
| "eval_samples_per_second": 96.508, | |
| "eval_steps_per_second": 6.047, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 7.06, | |
| "grad_norm": 3.1751720905303955, | |
| "learning_rate": 0.0014523809523809526, | |
| "loss": 0.3753, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 7.22, | |
| "grad_norm": 2.4551353454589844, | |
| "learning_rate": 0.0013730158730158731, | |
| "loss": 0.3061, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 7.37, | |
| "grad_norm": 1.707070231437683, | |
| "learning_rate": 0.0012936507936507939, | |
| "loss": 0.3525, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 7.53, | |
| "grad_norm": 0.43897438049316406, | |
| "learning_rate": 0.0012142857142857144, | |
| "loss": 0.3032, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 7.69, | |
| "grad_norm": 2.149365186691284, | |
| "learning_rate": 0.001134920634920635, | |
| "loss": 0.3031, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 7.84, | |
| "grad_norm": 1.5541889667510986, | |
| "learning_rate": 0.0010555555555555557, | |
| "loss": 0.3199, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.5194265842437744, | |
| "learning_rate": 0.0009761904761904762, | |
| "loss": 0.3218, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8669527896995708, | |
| "eval_f1": 0.8448168439144914, | |
| "eval_loss": 0.3072466254234314, | |
| "eval_precision": 0.8257109927719459, | |
| "eval_recall": 0.8897720673635308, | |
| "eval_runtime": 12.0816, | |
| "eval_samples_per_second": 96.428, | |
| "eval_steps_per_second": 6.042, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 8.16, | |
| "grad_norm": 0.4812917709350586, | |
| "learning_rate": 0.0008968253968253968, | |
| "loss": 0.3105, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 8.31, | |
| "grad_norm": 3.871387004852295, | |
| "learning_rate": 0.0008174603174603175, | |
| "loss": 0.3398, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 8.47, | |
| "grad_norm": 0.3649653196334839, | |
| "learning_rate": 0.000746031746031746, | |
| "loss": 0.3169, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 8.63, | |
| "grad_norm": 0.4373646676540375, | |
| "learning_rate": 0.0006666666666666666, | |
| "loss": 0.3019, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 8.78, | |
| "grad_norm": 0.21746236085891724, | |
| "learning_rate": 0.0005873015873015873, | |
| "loss": 0.3463, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 8.94, | |
| "grad_norm": 1.6411595344543457, | |
| "learning_rate": 0.0005079365079365079, | |
| "loss": 0.3219, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 8.99, | |
| "eval_accuracy": 0.896137339055794, | |
| "eval_f1": 0.870789179160911, | |
| "eval_loss": 0.263265997171402, | |
| "eval_precision": 0.8582184517497349, | |
| "eval_recall": 0.887172198789657, | |
| "eval_runtime": 12.1817, | |
| "eval_samples_per_second": 95.635, | |
| "eval_steps_per_second": 5.993, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 9.1, | |
| "grad_norm": 1.573140025138855, | |
| "learning_rate": 0.0004285714285714286, | |
| "loss": 0.3078, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 9.25, | |
| "grad_norm": 0.6549800634384155, | |
| "learning_rate": 0.00034920634920634924, | |
| "loss": 0.2872, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 9.41, | |
| "grad_norm": 0.4417751133441925, | |
| "learning_rate": 0.0002698412698412699, | |
| "loss": 0.3256, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 9.57, | |
| "grad_norm": 3.714104413986206, | |
| "learning_rate": 0.0001904761904761905, | |
| "loss": 0.361, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 9.73, | |
| "grad_norm": 0.5447297096252441, | |
| "learning_rate": 0.00011111111111111112, | |
| "loss": 0.3773, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 9.88, | |
| "grad_norm": 1.347410798072815, | |
| "learning_rate": 3.1746031746031745e-05, | |
| "loss": 0.3049, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 9.88, | |
| "eval_accuracy": 0.8927038626609443, | |
| "eval_f1": 0.8684775000564475, | |
| "eval_loss": 0.27391761541366577, | |
| "eval_precision": 0.8528169402296, | |
| "eval_recall": 0.8912334189131366, | |
| "eval_runtime": 12.284, | |
| "eval_samples_per_second": 94.839, | |
| "eval_steps_per_second": 5.943, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 9.88, | |
| "step": 630, | |
| "total_flos": 3.142570654487126e+18, | |
| "train_loss": 0.3948629246817695, | |
| "train_runtime": 760.2155, | |
| "train_samples_per_second": 53.63, | |
| "train_steps_per_second": 0.829 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 630, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 3.142570654487126e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |