| { | |
| "best_metric": 0.7709431751611013, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-10/checkpoint-1926", | |
| "epoch": 9.0, | |
| "eval_steps": 500, | |
| "global_step": 1926, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 8.383261680603027, | |
| "learning_rate": 0.00019166625457051745, | |
| "loss": 0.5952, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7076742823667252, | |
| "eval_f1": 0.3907203907203907, | |
| "eval_loss": 0.5565764307975769, | |
| "eval_mcc": 0.2694716353370074, | |
| "eval_precision": 0.64, | |
| "eval_recall": 0.281195079086116, | |
| "eval_runtime": 3.1354, | |
| "eval_samples_per_second": 544.427, | |
| "eval_steps_per_second": 17.223, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 2.881709575653076, | |
| "learning_rate": 0.00017037000406268217, | |
| "loss": 0.531, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7246631517281781, | |
| "eval_f1": 0.520408163265306, | |
| "eval_loss": 0.5205469727516174, | |
| "eval_mcc": 0.3429768856541182, | |
| "eval_precision": 0.6204379562043796, | |
| "eval_recall": 0.44815465729349735, | |
| "eval_runtime": 3.9179, | |
| "eval_samples_per_second": 435.687, | |
| "eval_steps_per_second": 13.783, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 2.652766227722168, | |
| "learning_rate": 0.00014907375355484689, | |
| "loss": 0.5039, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7492677211482133, | |
| "eval_f1": 0.5169300225733634, | |
| "eval_loss": 0.5036008358001709, | |
| "eval_mcc": 0.3941392583108005, | |
| "eval_precision": 0.722397476340694, | |
| "eval_recall": 0.4024604569420035, | |
| "eval_runtime": 3.1479, | |
| "eval_samples_per_second": 542.261, | |
| "eval_steps_per_second": 17.154, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 5.058462619781494, | |
| "learning_rate": 0.0001277775030470116, | |
| "loss": 0.4923, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7521968365553603, | |
| "eval_f1": 0.5209513023782559, | |
| "eval_loss": 0.49394065141677856, | |
| "eval_mcc": 0.40200593502267246, | |
| "eval_precision": 0.732484076433121, | |
| "eval_recall": 0.40421792618629176, | |
| "eval_runtime": 3.2248, | |
| "eval_samples_per_second": 529.338, | |
| "eval_steps_per_second": 16.745, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 5.336209297180176, | |
| "learning_rate": 0.00010648125253917635, | |
| "loss": 0.4809, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.760398359695372, | |
| "eval_f1": 0.5490628445424476, | |
| "eval_loss": 0.48668524622917175, | |
| "eval_mcc": 0.42515592515592515, | |
| "eval_precision": 0.7366863905325444, | |
| "eval_recall": 0.437609841827768, | |
| "eval_runtime": 3.1454, | |
| "eval_samples_per_second": 542.701, | |
| "eval_steps_per_second": 17.168, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 5.197123050689697, | |
| "learning_rate": 8.518500203134108e-05, | |
| "loss": 0.4741, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.7598125366139425, | |
| "eval_f1": 0.5883534136546184, | |
| "eval_loss": 0.48185786604881287, | |
| "eval_mcc": 0.4323191909966307, | |
| "eval_precision": 0.6861826697892272, | |
| "eval_recall": 0.5149384885764499, | |
| "eval_runtime": 3.2476, | |
| "eval_samples_per_second": 525.615, | |
| "eval_steps_per_second": 16.628, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 5.824602127075195, | |
| "learning_rate": 6.38887515235058e-05, | |
| "loss": 0.4642, | |
| "step": 1498 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.7697715289982425, | |
| "eval_f1": 0.5814696485623003, | |
| "eval_loss": 0.47905731201171875, | |
| "eval_mcc": 0.45140334971229035, | |
| "eval_precision": 0.7378378378378379, | |
| "eval_recall": 0.4797891036906854, | |
| "eval_runtime": 3.1541, | |
| "eval_samples_per_second": 541.195, | |
| "eval_steps_per_second": 17.12, | |
| "step": 1498 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 6.0699849128723145, | |
| "learning_rate": 4.259250101567054e-05, | |
| "loss": 0.4534, | |
| "step": 1712 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.7639132981839485, | |
| "eval_f1": 0.5780104712041885, | |
| "eval_loss": 0.4776459038257599, | |
| "eval_mcc": 0.4376855515817711, | |
| "eval_precision": 0.7150259067357513, | |
| "eval_recall": 0.4850615114235501, | |
| "eval_runtime": 3.3925, | |
| "eval_samples_per_second": 503.169, | |
| "eval_steps_per_second": 15.917, | |
| "step": 1712 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 4.322149753570557, | |
| "learning_rate": 2.129625050783527e-05, | |
| "loss": 0.454, | |
| "step": 1926 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.7709431751611013, | |
| "eval_f1": 0.606243705941591, | |
| "eval_loss": 0.4749068319797516, | |
| "eval_mcc": 0.4592236637293635, | |
| "eval_precision": 0.7099056603773585, | |
| "eval_recall": 0.5289982425307557, | |
| "eval_runtime": 3.1452, | |
| "eval_samples_per_second": 542.727, | |
| "eval_steps_per_second": 17.169, | |
| "step": 1926 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2140, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 4722980979960.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.9841007089206046, | |
| "learning_rate": 0.0002129625050783527, | |
| "num_train_epochs": 10, | |
| "temperature": 10 | |
| } | |
| } | |