| { | |
| "best_metric": 0.4779281382373973, | |
| "best_model_checkpoint": "distilbert-base-uncased-finetuned-cola/run-2/checkpoint-8552", | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 10690, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 4.28505277633667, | |
| "learning_rate": 3.0702893894484785e-06, | |
| "loss": 0.6069, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 9.482794761657715, | |
| "learning_rate": 2.9196373094951675e-06, | |
| "loss": 0.5628, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 22.521339416503906, | |
| "learning_rate": 2.7689852295418565e-06, | |
| "loss": 0.5565, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 26.7753849029541, | |
| "learning_rate": 2.6183331495885454e-06, | |
| "loss": 0.5184, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.5730993747711182, | |
| "eval_matthews_correlation": 0.3853198145814999, | |
| "eval_runtime": 0.7612, | |
| "eval_samples_per_second": 1370.225, | |
| "eval_steps_per_second": 86.706, | |
| "step": 2138 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 17.77669334411621, | |
| "learning_rate": 2.4676810696352344e-06, | |
| "loss": 0.4619, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 37.4239387512207, | |
| "learning_rate": 2.3170289896819234e-06, | |
| "loss": 0.5014, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "grad_norm": 46.75569534301758, | |
| "learning_rate": 2.1663769097286124e-06, | |
| "loss": 0.492, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 66.9134750366211, | |
| "learning_rate": 2.0157248297753013e-06, | |
| "loss": 0.4809, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.6646500825881958, | |
| "eval_matthews_correlation": 0.4691032179514943, | |
| "eval_runtime": 0.8224, | |
| "eval_samples_per_second": 1268.193, | |
| "eval_steps_per_second": 80.25, | |
| "step": 4276 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 11.169896125793457, | |
| "learning_rate": 1.8650727498219905e-06, | |
| "loss": 0.4934, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "grad_norm": 64.01177215576172, | |
| "learning_rate": 1.7144206698686793e-06, | |
| "loss": 0.4653, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "grad_norm": 0.48781171441078186, | |
| "learning_rate": 1.5637685899153684e-06, | |
| "loss": 0.4639, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "grad_norm": 0.21301893889904022, | |
| "learning_rate": 1.4131165099620574e-06, | |
| "loss": 0.514, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.8122562766075134, | |
| "eval_matthews_correlation": 0.44860917123689154, | |
| "eval_runtime": 0.7584, | |
| "eval_samples_per_second": 1375.339, | |
| "eval_steps_per_second": 87.03, | |
| "step": 6414 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "grad_norm": 76.10978698730469, | |
| "learning_rate": 1.2624644300087464e-06, | |
| "loss": 0.5175, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 3.27, | |
| "grad_norm": 13.475150108337402, | |
| "learning_rate": 1.1118123500554353e-06, | |
| "loss": 0.4448, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.51, | |
| "grad_norm": 0.116688072681427, | |
| "learning_rate": 9.611602701021243e-07, | |
| "loss": 0.4159, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.74, | |
| "grad_norm": 0.9807508587837219, | |
| "learning_rate": 8.105081901488134e-07, | |
| "loss": 0.4664, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.98, | |
| "grad_norm": 0.25140833854675293, | |
| "learning_rate": 6.598561101955022e-07, | |
| "loss": 0.4243, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.8493317365646362, | |
| "eval_matthews_correlation": 0.4779281382373973, | |
| "eval_runtime": 0.8106, | |
| "eval_samples_per_second": 1286.694, | |
| "eval_steps_per_second": 81.421, | |
| "step": 8552 | |
| }, | |
| { | |
| "epoch": 4.21, | |
| "grad_norm": 149.03529357910156, | |
| "learning_rate": 5.092040302421912e-07, | |
| "loss": 0.4192, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 4.44, | |
| "grad_norm": 0.7424508929252625, | |
| "learning_rate": 3.585519502888802e-07, | |
| "loss": 0.4218, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.68, | |
| "grad_norm": 11.759597778320312, | |
| "learning_rate": 2.078998703355692e-07, | |
| "loss": 0.469, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.91, | |
| "grad_norm": 7.273904800415039, | |
| "learning_rate": 5.7247790382258184e-08, | |
| "loss": 0.4144, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 0.8652405738830566, | |
| "eval_matthews_correlation": 0.4755357129573294, | |
| "eval_runtime": 0.7744, | |
| "eval_samples_per_second": 1346.924, | |
| "eval_steps_per_second": 85.232, | |
| "step": 10690 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 10690, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 171279829011432.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": { | |
| "learning_rate": 3.2209414694017896e-06, | |
| "num_train_epochs": 5, | |
| "per_device_train_batch_size": 4, | |
| "seed": 16 | |
| } | |
| } | |