| { | |
| "best_metric": 0.5404761904761904, | |
| "best_model_checkpoint": "videomae-base-finetuned-ElderReact-anger-balanced-hp/checkpoint-400", | |
| "epoch": 19.010416666666668, | |
| "eval_steps": 500, | |
| "global_step": 480, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.783530235290527, | |
| "learning_rate": 0.00020833333333333335, | |
| "loss": 0.7372, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 3.4003899097442627, | |
| "learning_rate": 0.0004166666666666667, | |
| "loss": 0.7532, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "eval_accuracy": 0.5238095238095238, | |
| "eval_loss": 0.7077798247337341, | |
| "eval_runtime": 171.2381, | |
| "eval_samples_per_second": 2.453, | |
| "eval_steps_per_second": 0.158, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 4.148990631103516, | |
| "learning_rate": 0.000625, | |
| "loss": 0.8004, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 2.1322150230407715, | |
| "learning_rate": 0.0008333333333333334, | |
| "loss": 0.7903, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 13.590558052062988, | |
| "learning_rate": 0.0009953703703703704, | |
| "loss": 0.7571, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "eval_accuracy": 0.47619047619047616, | |
| "eval_loss": 0.7033982276916504, | |
| "eval_runtime": 171.2144, | |
| "eval_samples_per_second": 2.453, | |
| "eval_steps_per_second": 0.158, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 3.682979106903076, | |
| "learning_rate": 0.0009722222222222222, | |
| "loss": 0.723, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "grad_norm": 1.8970884084701538, | |
| "learning_rate": 0.0009490740740740741, | |
| "loss": 0.7357, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "eval_accuracy": 0.44285714285714284, | |
| "eval_loss": 0.7080289125442505, | |
| "eval_runtime": 124.8954, | |
| "eval_samples_per_second": 3.363, | |
| "eval_steps_per_second": 0.216, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "grad_norm": 4.713415145874023, | |
| "learning_rate": 0.000925925925925926, | |
| "loss": 0.7389, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "grad_norm": 2.461824893951416, | |
| "learning_rate": 0.0009027777777777778, | |
| "loss": 0.6913, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "grad_norm": 9.549880981445312, | |
| "learning_rate": 0.0008796296296296296, | |
| "loss": 0.6976, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "eval_accuracy": 0.5238095238095238, | |
| "eval_loss": 0.716019332408905, | |
| "eval_runtime": 123.304, | |
| "eval_samples_per_second": 3.406, | |
| "eval_steps_per_second": 0.219, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "grad_norm": 2.853116273880005, | |
| "learning_rate": 0.0008564814814814815, | |
| "loss": 0.7026, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.04, | |
| "grad_norm": 5.189712047576904, | |
| "learning_rate": 0.0008333333333333334, | |
| "loss": 0.7131, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.05, | |
| "eval_accuracy": 0.4714285714285714, | |
| "eval_loss": 0.6892691254615784, | |
| "eval_runtime": 124.5641, | |
| "eval_samples_per_second": 3.372, | |
| "eval_steps_per_second": 0.217, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 5.01, | |
| "grad_norm": 9.913081169128418, | |
| "learning_rate": 0.0008101851851851853, | |
| "loss": 0.7349, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 5.03, | |
| "grad_norm": 5.214167594909668, | |
| "learning_rate": 0.0007870370370370372, | |
| "loss": 0.7542, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 5.05, | |
| "grad_norm": 7.558966636657715, | |
| "learning_rate": 0.0007638888888888888, | |
| "loss": 0.7275, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 5.05, | |
| "eval_accuracy": 0.4928571428571429, | |
| "eval_loss": 0.8349756002426147, | |
| "eval_runtime": 173.5792, | |
| "eval_samples_per_second": 2.42, | |
| "eval_steps_per_second": 0.156, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 6.02, | |
| "grad_norm": 0.9638558030128479, | |
| "learning_rate": 0.0007407407407407407, | |
| "loss": 0.8133, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 6.04, | |
| "grad_norm": 2.0914902687072754, | |
| "learning_rate": 0.0007175925925925926, | |
| "loss": 0.7334, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 6.05, | |
| "eval_accuracy": 0.4738095238095238, | |
| "eval_loss": 0.7126601338386536, | |
| "eval_runtime": 163.6881, | |
| "eval_samples_per_second": 2.566, | |
| "eval_steps_per_second": 0.165, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 7.01, | |
| "grad_norm": 1.6074931621551514, | |
| "learning_rate": 0.0006944444444444445, | |
| "loss": 0.6938, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 7.03, | |
| "grad_norm": 1.0195335149765015, | |
| "learning_rate": 0.0006712962962962962, | |
| "loss": 0.6986, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 7.05, | |
| "grad_norm": 10.589719772338867, | |
| "learning_rate": 0.0006481481481481481, | |
| "loss": 0.7274, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 7.05, | |
| "eval_accuracy": 0.5047619047619047, | |
| "eval_loss": 0.7088494896888733, | |
| "eval_runtime": 163.5307, | |
| "eval_samples_per_second": 2.568, | |
| "eval_steps_per_second": 0.165, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 8.02, | |
| "grad_norm": 1.649876356124878, | |
| "learning_rate": 0.000625, | |
| "loss": 0.6904, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 8.04, | |
| "grad_norm": 6.6575798988342285, | |
| "learning_rate": 0.0006018518518518519, | |
| "loss": 0.697, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 8.05, | |
| "eval_accuracy": 0.5190476190476191, | |
| "eval_loss": 0.6910755038261414, | |
| "eval_runtime": 125.391, | |
| "eval_samples_per_second": 3.35, | |
| "eval_steps_per_second": 0.215, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 9.01, | |
| "grad_norm": 3.279046058654785, | |
| "learning_rate": 0.0005787037037037038, | |
| "loss": 0.7072, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 9.03, | |
| "grad_norm": 1.2495840787887573, | |
| "learning_rate": 0.0005555555555555556, | |
| "loss": 0.694, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 9.05, | |
| "grad_norm": 11.012711524963379, | |
| "learning_rate": 0.0005324074074074074, | |
| "loss": 0.7605, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 9.05, | |
| "eval_accuracy": 0.4976190476190476, | |
| "eval_loss": 0.7295739054679871, | |
| "eval_runtime": 127.4402, | |
| "eval_samples_per_second": 3.296, | |
| "eval_steps_per_second": 0.212, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 10.02, | |
| "grad_norm": 1.9475804567337036, | |
| "learning_rate": 0.0005092592592592593, | |
| "loss": 0.7224, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 10.04, | |
| "grad_norm": 2.378416061401367, | |
| "learning_rate": 0.0004861111111111111, | |
| "loss": 0.7105, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 10.05, | |
| "eval_accuracy": 0.48333333333333334, | |
| "eval_loss": 0.710009753704071, | |
| "eval_runtime": 122.4996, | |
| "eval_samples_per_second": 3.429, | |
| "eval_steps_per_second": 0.22, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 11.01, | |
| "grad_norm": 0.702026903629303, | |
| "learning_rate": 0.000462962962962963, | |
| "loss": 0.695, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 11.03, | |
| "grad_norm": 2.676254987716675, | |
| "learning_rate": 0.0004398148148148148, | |
| "loss": 0.6865, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 11.05, | |
| "grad_norm": 7.562650680541992, | |
| "learning_rate": 0.0004166666666666667, | |
| "loss": 0.6745, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 11.05, | |
| "eval_accuracy": 0.45476190476190476, | |
| "eval_loss": 0.7270693778991699, | |
| "eval_runtime": 175.5484, | |
| "eval_samples_per_second": 2.393, | |
| "eval_steps_per_second": 0.154, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 12.02, | |
| "grad_norm": 4.304387092590332, | |
| "learning_rate": 0.0003935185185185186, | |
| "loss": 0.6889, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 12.04, | |
| "grad_norm": 0.35395267605781555, | |
| "learning_rate": 0.00037037037037037035, | |
| "loss": 0.7166, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 12.05, | |
| "eval_accuracy": 0.5285714285714286, | |
| "eval_loss": 0.6954912543296814, | |
| "eval_runtime": 177.5444, | |
| "eval_samples_per_second": 2.366, | |
| "eval_steps_per_second": 0.152, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 13.01, | |
| "grad_norm": 0.8164138793945312, | |
| "learning_rate": 0.00034722222222222224, | |
| "loss": 0.7363, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 13.03, | |
| "grad_norm": 1.0167158842086792, | |
| "learning_rate": 0.00032407407407407406, | |
| "loss": 0.6993, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 13.05, | |
| "grad_norm": 7.524409294128418, | |
| "learning_rate": 0.00030092592592592595, | |
| "loss": 0.6849, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 13.05, | |
| "eval_accuracy": 0.4976190476190476, | |
| "eval_loss": 0.698103666305542, | |
| "eval_runtime": 161.7269, | |
| "eval_samples_per_second": 2.597, | |
| "eval_steps_per_second": 0.167, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 14.02, | |
| "grad_norm": 1.1957156658172607, | |
| "learning_rate": 0.0002777777777777778, | |
| "loss": 0.6888, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 14.04, | |
| "grad_norm": 1.1630483865737915, | |
| "learning_rate": 0.00025462962962962966, | |
| "loss": 0.6978, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 14.05, | |
| "eval_accuracy": 0.49523809523809526, | |
| "eval_loss": 0.697565495967865, | |
| "eval_runtime": 162.8303, | |
| "eval_samples_per_second": 2.579, | |
| "eval_steps_per_second": 0.166, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 15.01, | |
| "grad_norm": 2.8283958435058594, | |
| "learning_rate": 0.0002314814814814815, | |
| "loss": 0.6931, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 15.03, | |
| "grad_norm": 1.4520641565322876, | |
| "learning_rate": 0.00020833333333333335, | |
| "loss": 0.6966, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 15.05, | |
| "grad_norm": 9.017864227294922, | |
| "learning_rate": 0.00018518518518518518, | |
| "loss": 0.6928, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 15.05, | |
| "eval_accuracy": 0.5404761904761904, | |
| "eval_loss": 0.6940857768058777, | |
| "eval_runtime": 165.3836, | |
| "eval_samples_per_second": 2.54, | |
| "eval_steps_per_second": 0.163, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 16.02, | |
| "grad_norm": 2.541412591934204, | |
| "learning_rate": 0.00016203703703703703, | |
| "loss": 0.6795, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 16.04, | |
| "grad_norm": 0.8520733118057251, | |
| "learning_rate": 0.0001388888888888889, | |
| "loss": 0.7057, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 16.05, | |
| "eval_accuracy": 0.5, | |
| "eval_loss": 0.7022379636764526, | |
| "eval_runtime": 164.8925, | |
| "eval_samples_per_second": 2.547, | |
| "eval_steps_per_second": 0.164, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 17.01, | |
| "grad_norm": 2.4603936672210693, | |
| "learning_rate": 0.00011574074074074075, | |
| "loss": 0.733, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 17.03, | |
| "grad_norm": 2.6278584003448486, | |
| "learning_rate": 9.259259259259259e-05, | |
| "loss": 0.6902, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 17.05, | |
| "grad_norm": 8.034465789794922, | |
| "learning_rate": 6.944444444444444e-05, | |
| "loss": 0.6842, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 17.05, | |
| "eval_accuracy": 0.4738095238095238, | |
| "eval_loss": 0.6942616701126099, | |
| "eval_runtime": 172.9688, | |
| "eval_samples_per_second": 2.428, | |
| "eval_steps_per_second": 0.156, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 18.02, | |
| "grad_norm": 1.0102304220199585, | |
| "learning_rate": 4.6296296296296294e-05, | |
| "loss": 0.6921, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 18.04, | |
| "grad_norm": 1.547524094581604, | |
| "learning_rate": 2.3148148148148147e-05, | |
| "loss": 0.6824, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 18.05, | |
| "eval_accuracy": 0.5166666666666667, | |
| "eval_loss": 0.694518506526947, | |
| "eval_runtime": 164.6584, | |
| "eval_samples_per_second": 2.551, | |
| "eval_steps_per_second": 0.164, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 19.01, | |
| "grad_norm": 0.8265247941017151, | |
| "learning_rate": 0.0, | |
| "loss": 0.7065, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 19.01, | |
| "eval_accuracy": 0.5142857142857142, | |
| "eval_loss": 0.6948422193527222, | |
| "eval_runtime": 157.477, | |
| "eval_samples_per_second": 2.667, | |
| "eval_steps_per_second": 0.171, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 19.01, | |
| "step": 480, | |
| "total_flos": 9.214647639143547e+18, | |
| "train_loss": 0.7152137100696564, | |
| "train_runtime": 6018.7779, | |
| "train_samples_per_second": 1.276, | |
| "train_steps_per_second": 0.08 | |
| }, | |
| { | |
| "epoch": 19.01, | |
| "eval_accuracy": 0.4671916010498688, | |
| "eval_loss": 0.6937770247459412, | |
| "eval_runtime": 144.3495, | |
| "eval_samples_per_second": 2.639, | |
| "eval_steps_per_second": 0.166, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 19.01, | |
| "eval_accuracy": 0.4671916010498688, | |
| "eval_loss": 0.6937723755836487, | |
| "eval_runtime": 152.0578, | |
| "eval_samples_per_second": 2.506, | |
| "eval_steps_per_second": 0.158, | |
| "step": 480 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 480, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9223372036854775807, | |
| "save_steps": 500, | |
| "total_flos": 9.214647639143547e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |