| { | |
| "best_metric": 0.87021494370522, | |
| "best_model_checkpoint": "./outputs/finetuning/mnli_ChcE/checkpoint-20000", | |
| "epoch": 5.0, | |
| "global_step": 59960, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 1.933288859239493e-05, | |
| "loss": 0.3102, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "eval_acc": 0.8544524053224155, | |
| "eval_loss": 0.41351282596588135, | |
| "eval_runtime": 16.7417, | |
| "eval_samples_per_second": 583.571, | |
| "eval_steps_per_second": 18.278, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 1.866577718478986e-05, | |
| "loss": 0.3046, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "eval_acc": 0.8644831115660184, | |
| "eval_loss": 0.40244850516319275, | |
| "eval_runtime": 16.6646, | |
| "eval_samples_per_second": 586.274, | |
| "eval_steps_per_second": 18.362, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 1.7998665777184793e-05, | |
| "loss": 0.3038, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_acc": 0.8668372569089048, | |
| "eval_loss": 0.39355912804603577, | |
| "eval_runtime": 16.6429, | |
| "eval_samples_per_second": 587.039, | |
| "eval_steps_per_second": 18.386, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.7331554369579722e-05, | |
| "loss": 0.3012, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "eval_acc": 0.8625383828045036, | |
| "eval_loss": 0.40068650245666504, | |
| "eval_runtime": 16.6669, | |
| "eval_samples_per_second": 586.192, | |
| "eval_steps_per_second": 18.36, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 1.666444296197465e-05, | |
| "loss": 0.2979, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "eval_acc": 0.8620266120777892, | |
| "eval_loss": 0.42349421977996826, | |
| "eval_runtime": 16.6784, | |
| "eval_samples_per_second": 585.787, | |
| "eval_steps_per_second": 18.347, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 1.599733155436958e-05, | |
| "loss": 0.2997, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_acc": 0.8643807574206755, | |
| "eval_loss": 0.403084933757782, | |
| "eval_runtime": 16.6652, | |
| "eval_samples_per_second": 586.251, | |
| "eval_steps_per_second": 18.362, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "learning_rate": 1.533022014676451e-05, | |
| "loss": 0.2099, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "eval_acc": 0.8632548618219038, | |
| "eval_loss": 0.43926700949668884, | |
| "eval_runtime": 16.6492, | |
| "eval_samples_per_second": 586.816, | |
| "eval_steps_per_second": 18.379, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "learning_rate": 1.4663108739159441e-05, | |
| "loss": 0.2114, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "eval_acc": 0.8628454452405322, | |
| "eval_loss": 0.46620267629623413, | |
| "eval_runtime": 16.6551, | |
| "eval_samples_per_second": 586.607, | |
| "eval_steps_per_second": 18.373, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "learning_rate": 1.3995997331554372e-05, | |
| "loss": 0.2147, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "eval_acc": 0.864790174002047, | |
| "eval_loss": 0.4331408143043518, | |
| "eval_runtime": 16.6453, | |
| "eval_samples_per_second": 586.954, | |
| "eval_steps_per_second": 18.384, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "learning_rate": 1.33288859239493e-05, | |
| "loss": 0.2122, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "eval_acc": 0.87021494370522, | |
| "eval_loss": 0.4166066646575928, | |
| "eval_runtime": 16.67, | |
| "eval_samples_per_second": 586.082, | |
| "eval_steps_per_second": 18.356, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 1.2661774516344229e-05, | |
| "loss": 0.2156, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "eval_acc": 0.8632548618219038, | |
| "eval_loss": 0.4462782144546509, | |
| "eval_runtime": 16.6855, | |
| "eval_samples_per_second": 585.539, | |
| "eval_steps_per_second": 18.339, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 1.199466310873916e-05, | |
| "loss": 0.2117, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_acc": 0.8679631525076765, | |
| "eval_loss": 0.46366986632347107, | |
| "eval_runtime": 16.6496, | |
| "eval_samples_per_second": 586.801, | |
| "eval_steps_per_second": 18.379, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "learning_rate": 1.132755170113409e-05, | |
| "loss": 0.1469, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "eval_acc": 0.8680655066530194, | |
| "eval_loss": 0.5210850834846497, | |
| "eval_runtime": 16.6494, | |
| "eval_samples_per_second": 586.808, | |
| "eval_steps_per_second": 18.379, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "learning_rate": 1.066044029352902e-05, | |
| "loss": 0.1526, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "eval_acc": 0.8620266120777892, | |
| "eval_loss": 0.5206254720687866, | |
| "eval_runtime": 16.6442, | |
| "eval_samples_per_second": 586.99, | |
| "eval_steps_per_second": 18.385, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "learning_rate": 9.99332888592395e-06, | |
| "loss": 0.1494, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "eval_acc": 0.8664278403275333, | |
| "eval_loss": 0.5167897939682007, | |
| "eval_runtime": 16.6489, | |
| "eval_samples_per_second": 586.825, | |
| "eval_steps_per_second": 18.38, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "learning_rate": 9.326217478318879e-06, | |
| "loss": 0.1519, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "eval_acc": 0.8700102354145343, | |
| "eval_loss": 0.48301637172698975, | |
| "eval_runtime": 16.6933, | |
| "eval_samples_per_second": 585.265, | |
| "eval_steps_per_second": 18.331, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "learning_rate": 8.65910607071381e-06, | |
| "loss": 0.152, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "eval_acc": 0.8635619242579324, | |
| "eval_loss": 0.5464909672737122, | |
| "eval_runtime": 16.6533, | |
| "eval_samples_per_second": 586.671, | |
| "eval_steps_per_second": 18.375, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 7.99199466310874e-06, | |
| "loss": 0.1498, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_acc": 0.8679631525076765, | |
| "eval_loss": 0.5550450682640076, | |
| "eval_runtime": 16.6712, | |
| "eval_samples_per_second": 586.039, | |
| "eval_steps_per_second": 18.355, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "learning_rate": 7.324883255503669e-06, | |
| "loss": 0.1131, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "eval_acc": 0.8601842374616172, | |
| "eval_loss": 0.6764107346534729, | |
| "eval_runtime": 16.6576, | |
| "eval_samples_per_second": 586.52, | |
| "eval_steps_per_second": 18.37, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 3.34, | |
| "learning_rate": 6.6577718478985995e-06, | |
| "loss": 0.1135, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 3.34, | |
| "eval_acc": 0.865711361310133, | |
| "eval_loss": 0.6199905276298523, | |
| "eval_runtime": 16.6534, | |
| "eval_samples_per_second": 586.668, | |
| "eval_steps_per_second": 18.375, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "learning_rate": 5.99066044029353e-06, | |
| "loss": 0.1175, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "eval_acc": 0.8671443193449335, | |
| "eval_loss": 0.588898241519928, | |
| "eval_runtime": 16.664, | |
| "eval_samples_per_second": 586.293, | |
| "eval_steps_per_second": 18.363, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 3.67, | |
| "learning_rate": 5.32354903268846e-06, | |
| "loss": 0.1156, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 3.67, | |
| "eval_acc": 0.8663254861821904, | |
| "eval_loss": 0.6299933195114136, | |
| "eval_runtime": 16.6491, | |
| "eval_samples_per_second": 586.818, | |
| "eval_steps_per_second": 18.379, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "learning_rate": 4.656437625083389e-06, | |
| "loss": 0.1104, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "eval_acc": 0.8689866939611054, | |
| "eval_loss": 0.6044776439666748, | |
| "eval_runtime": 16.6309, | |
| "eval_samples_per_second": 587.46, | |
| "eval_steps_per_second": 18.399, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 3.989326217478319e-06, | |
| "loss": 0.1111, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_acc": 0.869396110542477, | |
| "eval_loss": 0.6412716507911682, | |
| "eval_runtime": 16.6717, | |
| "eval_samples_per_second": 586.023, | |
| "eval_steps_per_second": 18.354, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "learning_rate": 3.3222148098732494e-06, | |
| "loss": 0.086, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "eval_acc": 0.8658137154554759, | |
| "eval_loss": 0.7271037697792053, | |
| "eval_runtime": 16.6567, | |
| "eval_samples_per_second": 586.55, | |
| "eval_steps_per_second": 18.371, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 4.34, | |
| "learning_rate": 2.6551034022681787e-06, | |
| "loss": 0.0895, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 4.34, | |
| "eval_acc": 0.8682702149437053, | |
| "eval_loss": 0.7273563146591187, | |
| "eval_runtime": 16.678, | |
| "eval_samples_per_second": 585.801, | |
| "eval_steps_per_second": 18.348, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "learning_rate": 1.987991994663109e-06, | |
| "loss": 0.0867, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "eval_acc": 0.8658137154554759, | |
| "eval_loss": 0.7226472496986389, | |
| "eval_runtime": 16.6651, | |
| "eval_samples_per_second": 586.253, | |
| "eval_steps_per_second": 18.362, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "learning_rate": 1.3208805870580388e-06, | |
| "loss": 0.0886, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "eval_acc": 0.8690890481064483, | |
| "eval_loss": 0.7181665897369385, | |
| "eval_runtime": 16.6571, | |
| "eval_samples_per_second": 586.536, | |
| "eval_steps_per_second": 18.371, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 4.84, | |
| "learning_rate": 6.537691794529687e-07, | |
| "loss": 0.0849, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 4.84, | |
| "eval_acc": 0.8698055271238485, | |
| "eval_loss": 0.7094203233718872, | |
| "eval_runtime": 16.6693, | |
| "eval_samples_per_second": 586.108, | |
| "eval_steps_per_second": 18.357, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 59960, | |
| "total_flos": 1.2620695529804544e+17, | |
| "train_loss": 0.17320066710326415, | |
| "train_runtime": 11406.3483, | |
| "train_samples_per_second": 168.211, | |
| "train_steps_per_second": 5.257 | |
| } | |
| ], | |
| "max_steps": 59960, | |
| "num_train_epochs": 5, | |
| "total_flos": 1.2620695529804544e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |