| { | |
| "best_metric": 0.8503582395087002, | |
| "best_model_checkpoint": "./outputs/finetuning/mnli_CollSgE/checkpoint-48000", | |
| "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": 0.4123, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "eval_acc": 0.8331627430910952, | |
| "eval_loss": 0.46929866075515747, | |
| "eval_runtime": 16.6562, | |
| "eval_samples_per_second": 586.569, | |
| "eval_steps_per_second": 18.372, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 1.866577718478986e-05, | |
| "loss": 0.4028, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "eval_acc": 0.8337768679631525, | |
| "eval_loss": 0.46244016289711, | |
| "eval_runtime": 16.6313, | |
| "eval_samples_per_second": 587.445, | |
| "eval_steps_per_second": 18.399, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 1.7998665777184793e-05, | |
| "loss": 0.3888, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_acc": 0.8374616171954964, | |
| "eval_loss": 0.45001623034477234, | |
| "eval_runtime": 16.6286, | |
| "eval_samples_per_second": 587.542, | |
| "eval_steps_per_second": 18.402, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.7331554369579722e-05, | |
| "loss": 0.3841, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "eval_acc": 0.8415557830092119, | |
| "eval_loss": 0.4280681312084198, | |
| "eval_runtime": 16.6312, | |
| "eval_samples_per_second": 587.451, | |
| "eval_steps_per_second": 18.399, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 1.666444296197465e-05, | |
| "loss": 0.3783, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "eval_acc": 0.8365404298874104, | |
| "eval_loss": 0.4433709681034088, | |
| "eval_runtime": 16.6487, | |
| "eval_samples_per_second": 586.832, | |
| "eval_steps_per_second": 18.38, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 1.599733155436958e-05, | |
| "loss": 0.3759, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_acc": 0.8417604912998976, | |
| "eval_loss": 0.4399653375148773, | |
| "eval_runtime": 16.6479, | |
| "eval_samples_per_second": 586.862, | |
| "eval_steps_per_second": 18.381, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "learning_rate": 1.533022014676451e-05, | |
| "loss": 0.2721, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "eval_acc": 0.8426816786079836, | |
| "eval_loss": 0.5021729469299316, | |
| "eval_runtime": 16.6186, | |
| "eval_samples_per_second": 587.897, | |
| "eval_steps_per_second": 18.413, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "learning_rate": 1.4663108739159441e-05, | |
| "loss": 0.2736, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "eval_acc": 0.8430910951893552, | |
| "eval_loss": 0.5251772403717041, | |
| "eval_runtime": 16.6517, | |
| "eval_samples_per_second": 586.728, | |
| "eval_steps_per_second": 18.377, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "learning_rate": 1.3995997331554372e-05, | |
| "loss": 0.2821, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "eval_acc": 0.8409416581371546, | |
| "eval_loss": 0.48868492245674133, | |
| "eval_runtime": 16.6338, | |
| "eval_samples_per_second": 587.358, | |
| "eval_steps_per_second": 18.396, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "learning_rate": 1.33288859239493e-05, | |
| "loss": 0.2802, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "eval_acc": 0.8457523029682702, | |
| "eval_loss": 0.4757882356643677, | |
| "eval_runtime": 16.6205, | |
| "eval_samples_per_second": 587.829, | |
| "eval_steps_per_second": 18.411, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 1.2661774516344229e-05, | |
| "loss": 0.2794, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "eval_acc": 0.8457523029682702, | |
| "eval_loss": 0.4611198604106903, | |
| "eval_runtime": 16.6377, | |
| "eval_samples_per_second": 587.22, | |
| "eval_steps_per_second": 18.392, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 1.199466310873916e-05, | |
| "loss": 0.2797, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_acc": 0.8456499488229273, | |
| "eval_loss": 0.49363580346107483, | |
| "eval_runtime": 16.6189, | |
| "eval_samples_per_second": 587.883, | |
| "eval_steps_per_second": 18.413, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "learning_rate": 1.132755170113409e-05, | |
| "loss": 0.1915, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "eval_acc": 0.8461617195496418, | |
| "eval_loss": 0.5545384883880615, | |
| "eval_runtime": 16.6314, | |
| "eval_samples_per_second": 587.443, | |
| "eval_steps_per_second": 18.399, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "learning_rate": 1.066044029352902e-05, | |
| "loss": 0.1946, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "eval_acc": 0.8443193449334698, | |
| "eval_loss": 0.5731399059295654, | |
| "eval_runtime": 16.6331, | |
| "eval_samples_per_second": 587.383, | |
| "eval_steps_per_second": 18.397, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "learning_rate": 9.99332888592395e-06, | |
| "loss": 0.2007, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "eval_acc": 0.8427840327533265, | |
| "eval_loss": 0.5507048964500427, | |
| "eval_runtime": 16.6762, | |
| "eval_samples_per_second": 585.864, | |
| "eval_steps_per_second": 18.349, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "learning_rate": 9.326217478318879e-06, | |
| "loss": 0.2008, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "eval_acc": 0.8454452405322416, | |
| "eval_loss": 0.5498657822608948, | |
| "eval_runtime": 16.6352, | |
| "eval_samples_per_second": 587.307, | |
| "eval_steps_per_second": 18.395, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "learning_rate": 8.65910607071381e-06, | |
| "loss": 0.1971, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "eval_acc": 0.8483111566018424, | |
| "eval_loss": 0.5273796916007996, | |
| "eval_runtime": 16.6214, | |
| "eval_samples_per_second": 587.797, | |
| "eval_steps_per_second": 18.41, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 7.99199466310874e-06, | |
| "loss": 0.2054, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_acc": 0.8475946775844422, | |
| "eval_loss": 0.5454184412956238, | |
| "eval_runtime": 16.6713, | |
| "eval_samples_per_second": 586.039, | |
| "eval_steps_per_second": 18.355, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "learning_rate": 7.324883255503669e-06, | |
| "loss": 0.1436, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "eval_acc": 0.844216990788127, | |
| "eval_loss": 0.6787278652191162, | |
| "eval_runtime": 16.636, | |
| "eval_samples_per_second": 587.281, | |
| "eval_steps_per_second": 18.394, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 3.34, | |
| "learning_rate": 6.6577718478985995e-06, | |
| "loss": 0.1426, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 3.34, | |
| "eval_acc": 0.8420675537359263, | |
| "eval_loss": 0.6932513117790222, | |
| "eval_runtime": 16.6446, | |
| "eval_samples_per_second": 586.978, | |
| "eval_steps_per_second": 18.384, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "learning_rate": 5.99066044029353e-06, | |
| "loss": 0.1463, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "eval_acc": 0.8455475946775844, | |
| "eval_loss": 0.65469890832901, | |
| "eval_runtime": 16.6534, | |
| "eval_samples_per_second": 586.666, | |
| "eval_steps_per_second": 18.375, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 3.67, | |
| "learning_rate": 5.32354903268846e-06, | |
| "loss": 0.1447, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 3.67, | |
| "eval_acc": 0.8438075742067553, | |
| "eval_loss": 0.646902859210968, | |
| "eval_runtime": 16.6236, | |
| "eval_samples_per_second": 587.717, | |
| "eval_steps_per_second": 18.408, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "learning_rate": 4.656437625083389e-06, | |
| "loss": 0.1445, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "eval_acc": 0.8471852610030707, | |
| "eval_loss": 0.6626366376876831, | |
| "eval_runtime": 16.6568, | |
| "eval_samples_per_second": 586.548, | |
| "eval_steps_per_second": 18.371, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 3.989326217478319e-06, | |
| "loss": 0.1457, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_acc": 0.8503582395087002, | |
| "eval_loss": 0.6494254469871521, | |
| "eval_runtime": 16.658, | |
| "eval_samples_per_second": 586.503, | |
| "eval_steps_per_second": 18.37, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "learning_rate": 3.3222148098732494e-06, | |
| "loss": 0.1133, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "eval_acc": 0.8458546571136131, | |
| "eval_loss": 0.7663929462432861, | |
| "eval_runtime": 16.6347, | |
| "eval_samples_per_second": 587.326, | |
| "eval_steps_per_second": 18.395, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 4.34, | |
| "learning_rate": 2.6551034022681787e-06, | |
| "loss": 0.1138, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 4.34, | |
| "eval_acc": 0.8452405322415558, | |
| "eval_loss": 0.7856881022453308, | |
| "eval_runtime": 16.65, | |
| "eval_samples_per_second": 586.787, | |
| "eval_steps_per_second": 18.378, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "learning_rate": 1.987991994663109e-06, | |
| "loss": 0.1154, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "eval_acc": 0.848618219037871, | |
| "eval_loss": 0.7623312473297119, | |
| "eval_runtime": 16.6261, | |
| "eval_samples_per_second": 587.63, | |
| "eval_steps_per_second": 18.405, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "learning_rate": 1.3208805870580388e-06, | |
| "loss": 0.1102, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "eval_acc": 0.845957011258956, | |
| "eval_loss": 0.7739743590354919, | |
| "eval_runtime": 16.636, | |
| "eval_samples_per_second": 587.28, | |
| "eval_steps_per_second": 18.394, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 4.84, | |
| "learning_rate": 6.537691794529687e-07, | |
| "loss": 0.1143, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 4.84, | |
| "eval_acc": 0.8445240532241556, | |
| "eval_loss": 0.7610092759132385, | |
| "eval_runtime": 16.6493, | |
| "eval_samples_per_second": 586.812, | |
| "eval_steps_per_second": 18.379, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 59960, | |
| "total_flos": 1.2620432416386816e+17, | |
| "train_loss": 0.22490199587200385, | |
| "train_runtime": 11405.3413, | |
| "train_samples_per_second": 168.222, | |
| "train_steps_per_second": 5.257 | |
| } | |
| ], | |
| "max_steps": 59960, | |
| "num_train_epochs": 5, | |
| "total_flos": 1.2620432416386816e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |