| { | |
| "best_metric": 0.7885178676039836, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-13/checkpoint-3210", | |
| "epoch": 15.0, | |
| "eval_steps": 500, | |
| "global_step": 3210, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 8.763052940368652, | |
| "learning_rate": 0.00036729100192842157, | |
| "loss": 0.5966, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7211482132396017, | |
| "eval_f1": 0.5132924335378324, | |
| "eval_loss": 0.5418887734413147, | |
| "eval_mcc": 0.33384466816371045, | |
| "eval_precision": 0.6136919315403423, | |
| "eval_recall": 0.44112478031634444, | |
| "eval_runtime": 3.1521, | |
| "eval_samples_per_second": 541.55, | |
| "eval_steps_per_second": 17.132, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 3.084683895111084, | |
| "learning_rate": 0.00034688594626573147, | |
| "loss": 0.5299, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7404803749267721, | |
| "eval_f1": 0.5179542981501633, | |
| "eval_loss": 0.5285571813583374, | |
| "eval_mcc": 0.37347577729616044, | |
| "eval_precision": 0.68, | |
| "eval_recall": 0.4182776801405975, | |
| "eval_runtime": 3.8839, | |
| "eval_samples_per_second": 439.512, | |
| "eval_steps_per_second": 13.904, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.8376232385635376, | |
| "learning_rate": 0.0003264808906030414, | |
| "loss": 0.5041, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7475102519039251, | |
| "eval_f1": 0.4800965018094088, | |
| "eval_loss": 0.5334233045578003, | |
| "eval_mcc": 0.3885026303183633, | |
| "eval_precision": 0.7653846153846153, | |
| "eval_recall": 0.34973637961335674, | |
| "eval_runtime": 3.1807, | |
| "eval_samples_per_second": 536.674, | |
| "eval_steps_per_second": 16.977, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 2.807384490966797, | |
| "learning_rate": 0.00030607583494035134, | |
| "loss": 0.4955, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7398945518453427, | |
| "eval_f1": 0.4422110552763819, | |
| "eval_loss": 0.5175681710243225, | |
| "eval_mcc": 0.36720426049390004, | |
| "eval_precision": 0.775330396475771, | |
| "eval_recall": 0.3093145869947276, | |
| "eval_runtime": 3.9525, | |
| "eval_samples_per_second": 431.878, | |
| "eval_steps_per_second": 13.662, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 7.096841335296631, | |
| "learning_rate": 0.0002856707792776612, | |
| "loss": 0.4841, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7516110134739309, | |
| "eval_f1": 0.6052141527001862, | |
| "eval_loss": 0.49192342162132263, | |
| "eval_mcc": 0.4265647238499904, | |
| "eval_precision": 0.6435643564356436, | |
| "eval_recall": 0.5711775043936731, | |
| "eval_runtime": 3.1288, | |
| "eval_samples_per_second": 545.582, | |
| "eval_steps_per_second": 17.259, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 4.960063457489014, | |
| "learning_rate": 0.00026526572361497115, | |
| "loss": 0.4738, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.7621558289396603, | |
| "eval_f1": 0.621268656716418, | |
| "eval_loss": 0.4805011451244354, | |
| "eval_mcc": 0.4506812006793125, | |
| "eval_precision": 0.6620278330019881, | |
| "eval_recall": 0.5852372583479789, | |
| "eval_runtime": 3.2097, | |
| "eval_samples_per_second": 531.827, | |
| "eval_steps_per_second": 16.824, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 6.75669002532959, | |
| "learning_rate": 0.000244860667952281, | |
| "loss": 0.4601, | |
| "step": 1498 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.7586408904510837, | |
| "eval_f1": 0.5264367816091953, | |
| "eval_loss": 0.4859382212162018, | |
| "eval_mcc": 0.41956224318934715, | |
| "eval_precision": 0.760797342192691, | |
| "eval_recall": 0.4024604569420035, | |
| "eval_runtime": 3.1647, | |
| "eval_samples_per_second": 539.388, | |
| "eval_steps_per_second": 17.063, | |
| "step": 1498 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 5.769745349884033, | |
| "learning_rate": 0.00022445561228959097, | |
| "loss": 0.4483, | |
| "step": 1712 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.7785588752196837, | |
| "eval_f1": 0.6337209302325582, | |
| "eval_loss": 0.466266006231308, | |
| "eval_mcc": 0.4826297358701252, | |
| "eval_precision": 0.7062634989200864, | |
| "eval_recall": 0.5746924428822495, | |
| "eval_runtime": 3.8302, | |
| "eval_samples_per_second": 445.67, | |
| "eval_steps_per_second": 14.099, | |
| "step": 1712 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 7.356119632720947, | |
| "learning_rate": 0.00020405055662690085, | |
| "loss": 0.4408, | |
| "step": 1926 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.7809021675454013, | |
| "eval_f1": 0.6191446028513238, | |
| "eval_loss": 0.46509596705436707, | |
| "eval_mcc": 0.48266230635542773, | |
| "eval_precision": 0.7360774818401937, | |
| "eval_recall": 0.5342706502636204, | |
| "eval_runtime": 3.1727, | |
| "eval_samples_per_second": 538.023, | |
| "eval_steps_per_second": 17.02, | |
| "step": 1926 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 3.752133369445801, | |
| "learning_rate": 0.00018364550096421078, | |
| "loss": 0.4322, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.7809021675454013, | |
| "eval_f1": 0.661231884057971, | |
| "eval_loss": 0.46195492148399353, | |
| "eval_mcc": 0.5000717628957632, | |
| "eval_precision": 0.6822429906542056, | |
| "eval_recall": 0.6414762741652021, | |
| "eval_runtime": 3.329, | |
| "eval_samples_per_second": 512.771, | |
| "eval_steps_per_second": 16.221, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 2.258147954940796, | |
| "learning_rate": 0.0001632404453015207, | |
| "loss": 0.4251, | |
| "step": 2354 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.7756297598125366, | |
| "eval_f1": 0.6135216952573159, | |
| "eval_loss": 0.46607503294944763, | |
| "eval_mcc": 0.4705148045618678, | |
| "eval_precision": 0.7203791469194313, | |
| "eval_recall": 0.5342706502636204, | |
| "eval_runtime": 3.1529, | |
| "eval_samples_per_second": 541.404, | |
| "eval_steps_per_second": 17.127, | |
| "step": 2354 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 3.4573206901550293, | |
| "learning_rate": 0.0001428353896388306, | |
| "loss": 0.417, | |
| "step": 2568 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.7832454598711189, | |
| "eval_f1": 0.6777003484320557, | |
| "eval_loss": 0.4633403718471527, | |
| "eval_mcc": 0.5144799999787437, | |
| "eval_precision": 0.6718480138169257, | |
| "eval_recall": 0.6836555360281195, | |
| "eval_runtime": 3.2478, | |
| "eval_samples_per_second": 525.584, | |
| "eval_steps_per_second": 16.627, | |
| "step": 2568 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 16.82988166809082, | |
| "learning_rate": 0.0001224303339761405, | |
| "loss": 0.4122, | |
| "step": 2782 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.7855887521968365, | |
| "eval_f1": 0.6623616236162362, | |
| "eval_loss": 0.47581344842910767, | |
| "eval_mcc": 0.5072003861458727, | |
| "eval_precision": 0.6970873786407767, | |
| "eval_recall": 0.6309314586994728, | |
| "eval_runtime": 3.1465, | |
| "eval_samples_per_second": 542.51, | |
| "eval_steps_per_second": 17.162, | |
| "step": 2782 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 14.468791961669922, | |
| "learning_rate": 0.00010202527831345042, | |
| "loss": 0.4053, | |
| "step": 2996 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.7814879906268307, | |
| "eval_f1": 0.6642664266426643, | |
| "eval_loss": 0.464487224817276, | |
| "eval_mcc": 0.5027717230279147, | |
| "eval_precision": 0.6808118081180812, | |
| "eval_recall": 0.648506151142355, | |
| "eval_runtime": 3.3781, | |
| "eval_samples_per_second": 505.31, | |
| "eval_steps_per_second": 15.985, | |
| "step": 2996 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 3.5750694274902344, | |
| "learning_rate": 8.162022265076034e-05, | |
| "loss": 0.3972, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.7885178676039836, | |
| "eval_f1": 0.6762331838565022, | |
| "eval_loss": 0.4647216796875, | |
| "eval_mcc": 0.5195513014296814, | |
| "eval_precision": 0.6904761904761905, | |
| "eval_recall": 0.6625659050966608, | |
| "eval_runtime": 3.1634, | |
| "eval_samples_per_second": 539.617, | |
| "eval_steps_per_second": 17.07, | |
| "step": 3210 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 4066, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 19, | |
| "save_steps": 500, | |
| "total_flos": 7871634966600.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.9979252549744665, | |
| "learning_rate": 0.00038769605759111166, | |
| "num_train_epochs": 19, | |
| "temperature": 10 | |
| } | |
| } | |