| { | |
| "best_metric": 0.5935187339782715, | |
| "best_model_checkpoint": "facial_emotions_image_detection_rafdb_google_vit/checkpoint-14616", | |
| "epoch": 17.0, | |
| "eval_steps": 500, | |
| "global_step": 17748, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 4.908149719238281, | |
| "learning_rate": 2.9351896303408547e-06, | |
| "loss": 1.6796, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 6.714761257171631, | |
| "learning_rate": 2.8631781084973597e-06, | |
| "loss": 1.1697, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6166883963494133, | |
| "eval_loss": 1.1752560138702393, | |
| "eval_runtime": 43.7506, | |
| "eval_samples_per_second": 70.125, | |
| "eval_steps_per_second": 4.389, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 5.284860134124756, | |
| "learning_rate": 2.7911665866538647e-06, | |
| "loss": 0.9317, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 6.737391471862793, | |
| "learning_rate": 2.7191550648103698e-06, | |
| "loss": 0.7959, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6988265971316818, | |
| "eval_loss": 0.9186690449714661, | |
| "eval_runtime": 27.1747, | |
| "eval_samples_per_second": 112.899, | |
| "eval_steps_per_second": 7.065, | |
| "step": 2088 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "grad_norm": 7.392563819885254, | |
| "learning_rate": 2.6471435429668748e-06, | |
| "loss": 0.6865, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "grad_norm": 6.488108158111572, | |
| "learning_rate": 2.57513202112338e-06, | |
| "loss": 0.6157, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7314211212516297, | |
| "eval_loss": 0.798936128616333, | |
| "eval_runtime": 26.6756, | |
| "eval_samples_per_second": 115.012, | |
| "eval_steps_per_second": 7.198, | |
| "step": 3132 | |
| }, | |
| { | |
| "epoch": 3.35, | |
| "grad_norm": 6.2023773193359375, | |
| "learning_rate": 2.503120499279885e-06, | |
| "loss": 0.5444, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "grad_norm": 7.885190010070801, | |
| "learning_rate": 2.43110897743639e-06, | |
| "loss": 0.4788, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7496740547588006, | |
| "eval_loss": 0.7347991466522217, | |
| "eval_runtime": 26.7532, | |
| "eval_samples_per_second": 114.678, | |
| "eval_steps_per_second": 7.177, | |
| "step": 4176 | |
| }, | |
| { | |
| "epoch": 4.31, | |
| "grad_norm": 6.879584312438965, | |
| "learning_rate": 2.359097455592895e-06, | |
| "loss": 0.4213, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 4.79, | |
| "grad_norm": 5.687563896179199, | |
| "learning_rate": 2.2870859337494e-06, | |
| "loss": 0.3859, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7601043024771839, | |
| "eval_loss": 0.6939254999160767, | |
| "eval_runtime": 34.3214, | |
| "eval_samples_per_second": 89.39, | |
| "eval_steps_per_second": 5.594, | |
| "step": 5220 | |
| }, | |
| { | |
| "epoch": 5.27, | |
| "grad_norm": 8.583109855651855, | |
| "learning_rate": 2.215074411905905e-06, | |
| "loss": 0.3529, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 5.75, | |
| "grad_norm": 8.925189971923828, | |
| "learning_rate": 2.14306289006241e-06, | |
| "loss": 0.319, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.7728161668839635, | |
| "eval_loss": 0.6545543074607849, | |
| "eval_runtime": 26.3981, | |
| "eval_samples_per_second": 116.22, | |
| "eval_steps_per_second": 7.273, | |
| "step": 6264 | |
| }, | |
| { | |
| "epoch": 6.23, | |
| "grad_norm": 6.763670921325684, | |
| "learning_rate": 2.071051368218915e-06, | |
| "loss": 0.2914, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 6.7, | |
| "grad_norm": 5.260948657989502, | |
| "learning_rate": 1.9990398463754204e-06, | |
| "loss": 0.2635, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.7692307692307693, | |
| "eval_loss": 0.6396156549453735, | |
| "eval_runtime": 29.4081, | |
| "eval_samples_per_second": 104.325, | |
| "eval_steps_per_second": 6.529, | |
| "step": 7308 | |
| }, | |
| { | |
| "epoch": 7.18, | |
| "grad_norm": 5.159913539886475, | |
| "learning_rate": 1.9270283245319254e-06, | |
| "loss": 0.2463, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 7.66, | |
| "grad_norm": 6.131282806396484, | |
| "learning_rate": 1.8550168026884303e-06, | |
| "loss": 0.2254, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.7829204693611473, | |
| "eval_loss": 0.6089679598808289, | |
| "eval_runtime": 26.3772, | |
| "eval_samples_per_second": 116.313, | |
| "eval_steps_per_second": 7.279, | |
| "step": 8352 | |
| }, | |
| { | |
| "epoch": 8.14, | |
| "grad_norm": 9.69652271270752, | |
| "learning_rate": 1.7830052808449353e-06, | |
| "loss": 0.206, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 8.62, | |
| "grad_norm": 19.338401794433594, | |
| "learning_rate": 1.71099375900144e-06, | |
| "loss": 0.189, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.7842242503259452, | |
| "eval_loss": 0.6003961563110352, | |
| "eval_runtime": 26.7103, | |
| "eval_samples_per_second": 114.862, | |
| "eval_steps_per_second": 7.188, | |
| "step": 9396 | |
| }, | |
| { | |
| "epoch": 9.1, | |
| "grad_norm": 7.628934383392334, | |
| "learning_rate": 1.6389822371579451e-06, | |
| "loss": 0.1867, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 9.58, | |
| "grad_norm": 7.943005084991455, | |
| "learning_rate": 1.5669707153144506e-06, | |
| "loss": 0.1675, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.7832464146023468, | |
| "eval_loss": 0.6076902747154236, | |
| "eval_runtime": 28.2446, | |
| "eval_samples_per_second": 108.623, | |
| "eval_steps_per_second": 6.798, | |
| "step": 10440 | |
| }, | |
| { | |
| "epoch": 10.06, | |
| "grad_norm": 5.741328239440918, | |
| "learning_rate": 1.4949591934709554e-06, | |
| "loss": 0.154, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 10.54, | |
| "grad_norm": 12.926642417907715, | |
| "learning_rate": 1.4229476716274604e-06, | |
| "loss": 0.1515, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.7845501955671447, | |
| "eval_loss": 0.5965496897697449, | |
| "eval_runtime": 26.4509, | |
| "eval_samples_per_second": 115.989, | |
| "eval_steps_per_second": 7.259, | |
| "step": 11484 | |
| }, | |
| { | |
| "epoch": 11.02, | |
| "grad_norm": 7.206885814666748, | |
| "learning_rate": 1.3509361497839654e-06, | |
| "loss": 0.1346, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 11.49, | |
| "grad_norm": 5.869232177734375, | |
| "learning_rate": 1.2789246279404704e-06, | |
| "loss": 0.1292, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 11.97, | |
| "grad_norm": 5.3356828689575195, | |
| "learning_rate": 1.2069131060969755e-06, | |
| "loss": 0.1212, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.7868318122555411, | |
| "eval_loss": 0.5986641645431519, | |
| "eval_runtime": 28.927, | |
| "eval_samples_per_second": 106.06, | |
| "eval_steps_per_second": 6.637, | |
| "step": 12528 | |
| }, | |
| { | |
| "epoch": 12.45, | |
| "grad_norm": 4.9229350090026855, | |
| "learning_rate": 1.1349015842534805e-06, | |
| "loss": 0.1127, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 12.93, | |
| "grad_norm": 2.7797281742095947, | |
| "learning_rate": 1.0628900624099857e-06, | |
| "loss": 0.1147, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.7891134289439374, | |
| "eval_loss": 0.6010680794715881, | |
| "eval_runtime": 26.5314, | |
| "eval_samples_per_second": 115.637, | |
| "eval_steps_per_second": 7.237, | |
| "step": 13572 | |
| }, | |
| { | |
| "epoch": 13.41, | |
| "grad_norm": 6.798295497894287, | |
| "learning_rate": 9.908785405664905e-07, | |
| "loss": 0.1059, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 13.89, | |
| "grad_norm": 7.220839500427246, | |
| "learning_rate": 9.188670187229958e-07, | |
| "loss": 0.0989, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.795632333767927, | |
| "eval_loss": 0.5935187339782715, | |
| "eval_runtime": 26.5413, | |
| "eval_samples_per_second": 115.593, | |
| "eval_steps_per_second": 7.234, | |
| "step": 14616 | |
| }, | |
| { | |
| "epoch": 14.37, | |
| "grad_norm": 7.945505619049072, | |
| "learning_rate": 8.468554968795008e-07, | |
| "loss": 0.0972, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 14.85, | |
| "grad_norm": 2.5897045135498047, | |
| "learning_rate": 7.748439750360057e-07, | |
| "loss": 0.0915, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.7940026075619296, | |
| "eval_loss": 0.5987181663513184, | |
| "eval_runtime": 26.6784, | |
| "eval_samples_per_second": 114.999, | |
| "eval_steps_per_second": 7.197, | |
| "step": 15660 | |
| }, | |
| { | |
| "epoch": 15.33, | |
| "grad_norm": 5.564732074737549, | |
| "learning_rate": 7.028324531925108e-07, | |
| "loss": 0.0898, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 15.8, | |
| "grad_norm": 2.869757890701294, | |
| "learning_rate": 6.308209313490159e-07, | |
| "loss": 0.0887, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.7966101694915254, | |
| "eval_loss": 0.6017541885375977, | |
| "eval_runtime": 29.5882, | |
| "eval_samples_per_second": 103.69, | |
| "eval_steps_per_second": 6.489, | |
| "step": 16704 | |
| }, | |
| { | |
| "epoch": 16.28, | |
| "grad_norm": 10.52423095703125, | |
| "learning_rate": 5.588094095055209e-07, | |
| "loss": 0.0812, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 16.76, | |
| "grad_norm": 2.270158052444458, | |
| "learning_rate": 4.867978876620259e-07, | |
| "loss": 0.0857, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.7985658409387223, | |
| "eval_loss": 0.6071637272834778, | |
| "eval_runtime": 32.54, | |
| "eval_samples_per_second": 94.284, | |
| "eval_steps_per_second": 5.9, | |
| "step": 17748 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 20880, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 4.400719318285052e+19, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |