| { | |
| "best_metric": 0.6332945285215367, | |
| "best_model_checkpoint": "dinov2-base-ODIR-5K/checkpoint-263", | |
| "epoch": 5.914691943127962, | |
| "eval_steps": 500, | |
| "global_step": 312, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1895734597156398, | |
| "grad_norm": 151.11000061035156, | |
| "learning_rate": 1.5625e-05, | |
| "loss": 0.7609, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3791469194312796, | |
| "grad_norm": 34.450782775878906, | |
| "learning_rate": 3.125e-05, | |
| "loss": 0.6601, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.5687203791469194, | |
| "grad_norm": 58.26628112792969, | |
| "learning_rate": 4.6875e-05, | |
| "loss": 0.6456, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.7582938388625592, | |
| "grad_norm": 36.494163513183594, | |
| "learning_rate": 4.8571428571428576e-05, | |
| "loss": 0.7091, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.9478672985781991, | |
| "grad_norm": 71.59378051757812, | |
| "learning_rate": 4.678571428571429e-05, | |
| "loss": 0.6374, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.985781990521327, | |
| "eval_accuracy": 0.6778224007139669, | |
| "eval_f1": 0.20309050772626933, | |
| "eval_loss": 0.6185599565505981, | |
| "eval_runtime": 48.0839, | |
| "eval_samples_per_second": 46.606, | |
| "eval_steps_per_second": 1.477, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.1374407582938388, | |
| "grad_norm": 30.247587203979492, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.5721, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.3270142180094786, | |
| "grad_norm": 39.30122756958008, | |
| "learning_rate": 4.3214285714285716e-05, | |
| "loss": 0.6093, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.5165876777251186, | |
| "grad_norm": 32.839881896972656, | |
| "learning_rate": 4.1428571428571437e-05, | |
| "loss": 0.5805, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.7061611374407581, | |
| "grad_norm": 15.105513572692871, | |
| "learning_rate": 3.964285714285714e-05, | |
| "loss": 0.5933, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.8957345971563981, | |
| "grad_norm": 12.636411666870117, | |
| "learning_rate": 3.785714285714286e-05, | |
| "loss": 0.5789, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.9905213270142181, | |
| "eval_accuracy": 0.715305667112896, | |
| "eval_f1": 0.3793774319066148, | |
| "eval_loss": 0.5660730600357056, | |
| "eval_runtime": 48.0436, | |
| "eval_samples_per_second": 46.645, | |
| "eval_steps_per_second": 1.478, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 2.085308056872038, | |
| "grad_norm": 10.535428047180176, | |
| "learning_rate": 3.607142857142857e-05, | |
| "loss": 0.5118, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.2748815165876777, | |
| "grad_norm": 20.61358642578125, | |
| "learning_rate": 3.428571428571429e-05, | |
| "loss": 0.5222, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.4644549763033177, | |
| "grad_norm": 16.988859176635742, | |
| "learning_rate": 3.2500000000000004e-05, | |
| "loss": 0.552, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.654028436018957, | |
| "grad_norm": 27.578903198242188, | |
| "learning_rate": 3.071428571428572e-05, | |
| "loss": 0.5025, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.843601895734597, | |
| "grad_norm": 7.29171895980835, | |
| "learning_rate": 2.8928571428571434e-05, | |
| "loss": 0.5368, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.995260663507109, | |
| "eval_accuracy": 0.7407407407407407, | |
| "eval_f1": 0.5756026296566837, | |
| "eval_loss": 0.5333759188652039, | |
| "eval_runtime": 48.0715, | |
| "eval_samples_per_second": 46.618, | |
| "eval_steps_per_second": 1.477, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 3.0331753554502368, | |
| "grad_norm": 10.034445762634277, | |
| "learning_rate": 2.714285714285714e-05, | |
| "loss": 0.4669, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 3.2227488151658767, | |
| "grad_norm": 10.450629234313965, | |
| "learning_rate": 2.5357142857142858e-05, | |
| "loss": 0.4631, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 3.4123222748815167, | |
| "grad_norm": 17.410293579101562, | |
| "learning_rate": 2.357142857142857e-05, | |
| "loss": 0.4565, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.6018957345971563, | |
| "grad_norm": 11.578069686889648, | |
| "learning_rate": 2.1785714285714288e-05, | |
| "loss": 0.4292, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.7914691943127963, | |
| "grad_norm": 12.56495189666748, | |
| "learning_rate": 2e-05, | |
| "loss": 0.4289, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.9810426540284363, | |
| "grad_norm": 11.08157730102539, | |
| "learning_rate": 1.8214285714285715e-05, | |
| "loss": 0.4162, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.6983489513609995, | |
| "eval_f1": 0.6197975253093364, | |
| "eval_loss": 0.574705183506012, | |
| "eval_runtime": 48.2487, | |
| "eval_samples_per_second": 46.447, | |
| "eval_steps_per_second": 1.472, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 4.170616113744076, | |
| "grad_norm": 16.88544273376465, | |
| "learning_rate": 1.642857142857143e-05, | |
| "loss": 0.3839, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 4.360189573459715, | |
| "grad_norm": 10.79478931427002, | |
| "learning_rate": 1.4642857142857144e-05, | |
| "loss": 0.3881, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 4.549763033175355, | |
| "grad_norm": 19.362905502319336, | |
| "learning_rate": 1.2857142857142857e-05, | |
| "loss": 0.3377, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 4.739336492890995, | |
| "grad_norm": 14.351540565490723, | |
| "learning_rate": 1.107142857142857e-05, | |
| "loss": 0.348, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.928909952606635, | |
| "grad_norm": 12.308573722839355, | |
| "learning_rate": 9.285714285714286e-06, | |
| "loss": 0.3679, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 4.9857819905213265, | |
| "eval_accuracy": 0.7188755020080321, | |
| "eval_f1": 0.6332945285215367, | |
| "eval_loss": 0.5700241923332214, | |
| "eval_runtime": 48.6206, | |
| "eval_samples_per_second": 46.092, | |
| "eval_steps_per_second": 1.46, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 5.118483412322275, | |
| "grad_norm": 6.65657377243042, | |
| "learning_rate": 7.5e-06, | |
| "loss": 0.3325, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 5.308056872037914, | |
| "grad_norm": 14.3682222366333, | |
| "learning_rate": 5.7142857142857145e-06, | |
| "loss": 0.2559, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 5.497630331753554, | |
| "grad_norm": 14.902679443359375, | |
| "learning_rate": 3.928571428571429e-06, | |
| "loss": 0.238, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 5.687203791469194, | |
| "grad_norm": 12.136739730834961, | |
| "learning_rate": 2.142857142857143e-06, | |
| "loss": 0.2202, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 5.876777251184834, | |
| "grad_norm": 14.554306030273438, | |
| "learning_rate": 3.5714285714285716e-07, | |
| "loss": 0.2431, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 5.914691943127962, | |
| "eval_accuracy": 0.7563587684069611, | |
| "eval_f1": 0.6330645161290323, | |
| "eval_loss": 0.6110782027244568, | |
| "eval_runtime": 48.257, | |
| "eval_samples_per_second": 46.439, | |
| "eval_steps_per_second": 1.471, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 5.914691943127962, | |
| "step": 312, | |
| "total_flos": 3.110854122492383e+18, | |
| "train_loss": 0.474308407268463, | |
| "train_runtime": 2029.1858, | |
| "train_samples_per_second": 19.873, | |
| "train_steps_per_second": 0.154 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 312, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 6, | |
| "save_steps": 500, | |
| "total_flos": 3.110854122492383e+18, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |