| { | |
| "best_metric": 0.9720466022505865, | |
| "best_model_checkpoint": "/scratch/mrahma45/pixel/finetuned_models/bert/bert-base-finetuned-pos-ud-English-EWT/checkpoint-3500", | |
| "epoch": 15.306122448979592, | |
| "global_step": 6000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.9500000000000004e-05, | |
| "loss": 1.4315, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 4.966778523489933e-05, | |
| "loss": 0.1414, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 4.933221476510068e-05, | |
| "loss": 0.1147, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "learning_rate": 4.8996644295302016e-05, | |
| "loss": 0.0869, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "learning_rate": 4.8664429530201344e-05, | |
| "loss": 0.0599, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "eval_accuracy": 0.9675931448566544, | |
| "eval_loss": 0.12289105355739594, | |
| "eval_runtime": 10.3063, | |
| "eval_samples_per_second": 194.153, | |
| "eval_steps_per_second": 24.354, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "learning_rate": 4.8328859060402684e-05, | |
| "loss": 0.0573, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "learning_rate": 4.799328859060403e-05, | |
| "loss": 0.0606, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "learning_rate": 4.7657718120805376e-05, | |
| "loss": 0.0503, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "learning_rate": 4.7322147651006715e-05, | |
| "loss": 0.0279, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "learning_rate": 4.698657718120806e-05, | |
| "loss": 0.0285, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "eval_accuracy": 0.9700584516282954, | |
| "eval_loss": 0.12456455826759338, | |
| "eval_runtime": 10.3011, | |
| "eval_samples_per_second": 194.252, | |
| "eval_steps_per_second": 24.366, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "learning_rate": 4.66510067114094e-05, | |
| "loss": 0.0309, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.06, | |
| "learning_rate": 4.631543624161074e-05, | |
| "loss": 0.0263, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.32, | |
| "learning_rate": 4.597986577181208e-05, | |
| "loss": 0.0162, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "learning_rate": 4.5644295302013425e-05, | |
| "loss": 0.0175, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "learning_rate": 4.5308724832214764e-05, | |
| "loss": 0.0165, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "eval_accuracy": 0.9707741858523202, | |
| "eval_loss": 0.14595353603363037, | |
| "eval_runtime": 10.2794, | |
| "eval_samples_per_second": 194.66, | |
| "eval_steps_per_second": 24.418, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.08, | |
| "learning_rate": 4.497315436241611e-05, | |
| "loss": 0.016, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 4.34, | |
| "learning_rate": 4.463758389261745e-05, | |
| "loss": 0.0107, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 4.59, | |
| "learning_rate": 4.4302013422818796e-05, | |
| "loss": 0.0102, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 4.85, | |
| "learning_rate": 4.3966442953020135e-05, | |
| "loss": 0.0137, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "learning_rate": 4.363087248322148e-05, | |
| "loss": 0.0084, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "eval_accuracy": 0.970575370790091, | |
| "eval_loss": 0.16072939336299896, | |
| "eval_runtime": 10.2179, | |
| "eval_samples_per_second": 195.833, | |
| "eval_steps_per_second": 24.565, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 5.36, | |
| "learning_rate": 4.329530201342282e-05, | |
| "loss": 0.0061, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 5.61, | |
| "learning_rate": 4.2959731543624166e-05, | |
| "loss": 0.0076, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 5.87, | |
| "learning_rate": 4.2624161073825505e-05, | |
| "loss": 0.009, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 6.12, | |
| "learning_rate": 4.2288590604026845e-05, | |
| "loss": 0.0076, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 6.38, | |
| "learning_rate": 4.195302013422819e-05, | |
| "loss": 0.005, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 6.38, | |
| "eval_accuracy": 0.9695812954789454, | |
| "eval_loss": 0.18822818994522095, | |
| "eval_runtime": 10.2296, | |
| "eval_samples_per_second": 195.609, | |
| "eval_steps_per_second": 24.537, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 6.63, | |
| "learning_rate": 4.161744966442953e-05, | |
| "loss": 0.0065, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 6.89, | |
| "learning_rate": 4.1281879194630876e-05, | |
| "loss": 0.0056, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 7.14, | |
| "learning_rate": 4.0946308724832215e-05, | |
| "loss": 0.0058, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 7.4, | |
| "learning_rate": 4.061073825503356e-05, | |
| "loss": 0.0045, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 7.65, | |
| "learning_rate": 4.02751677852349e-05, | |
| "loss": 0.0054, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 7.65, | |
| "eval_accuracy": 0.9698993995785121, | |
| "eval_loss": 0.17441630363464355, | |
| "eval_runtime": 10.2421, | |
| "eval_samples_per_second": 195.37, | |
| "eval_steps_per_second": 24.507, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 7.91, | |
| "learning_rate": 3.993959731543625e-05, | |
| "loss": 0.0046, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 8.16, | |
| "learning_rate": 3.9604026845637586e-05, | |
| "loss": 0.0056, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 8.42, | |
| "learning_rate": 3.926845637583893e-05, | |
| "loss": 0.0039, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 8.67, | |
| "learning_rate": 3.893288590604027e-05, | |
| "loss": 0.004, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 8.93, | |
| "learning_rate": 3.859731543624161e-05, | |
| "loss": 0.0058, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 8.93, | |
| "eval_accuracy": 0.9720466022505865, | |
| "eval_loss": 0.16418756544589996, | |
| "eval_runtime": 10.248, | |
| "eval_samples_per_second": 195.258, | |
| "eval_steps_per_second": 24.493, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 9.18, | |
| "learning_rate": 3.826174496644295e-05, | |
| "loss": 0.0039, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 9.44, | |
| "learning_rate": 3.7926174496644296e-05, | |
| "loss": 0.0037, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 9.69, | |
| "learning_rate": 3.7590604026845635e-05, | |
| "loss": 0.0026, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 9.95, | |
| "learning_rate": 3.725503355704698e-05, | |
| "loss": 0.0034, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 10.2, | |
| "learning_rate": 3.691946308724833e-05, | |
| "loss": 0.0034, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 10.2, | |
| "eval_accuracy": 0.9707344228398743, | |
| "eval_loss": 0.1923176646232605, | |
| "eval_runtime": 10.2481, | |
| "eval_samples_per_second": 195.256, | |
| "eval_steps_per_second": 24.492, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 10.46, | |
| "learning_rate": 3.6583892617449666e-05, | |
| "loss": 0.0033, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 10.71, | |
| "learning_rate": 3.624832214765101e-05, | |
| "loss": 0.0029, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 10.97, | |
| "learning_rate": 3.591275167785235e-05, | |
| "loss": 0.0046, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 11.22, | |
| "learning_rate": 3.55771812080537e-05, | |
| "loss": 0.0037, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 11.48, | |
| "learning_rate": 3.524161073825504e-05, | |
| "loss": 0.0032, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 11.48, | |
| "eval_accuracy": 0.9701777406656328, | |
| "eval_loss": 0.19770818948745728, | |
| "eval_runtime": 10.2544, | |
| "eval_samples_per_second": 195.135, | |
| "eval_steps_per_second": 24.477, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 11.73, | |
| "learning_rate": 3.4906040268456376e-05, | |
| "loss": 0.0034, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 11.99, | |
| "learning_rate": 3.4570469798657716e-05, | |
| "loss": 0.0038, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 12.24, | |
| "learning_rate": 3.423489932885906e-05, | |
| "loss": 0.0027, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "learning_rate": 3.38993288590604e-05, | |
| "loss": 0.0031, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 12.76, | |
| "learning_rate": 3.356375838926175e-05, | |
| "loss": 0.0023, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 12.76, | |
| "eval_accuracy": 0.9706548968149827, | |
| "eval_loss": 0.1978425234556198, | |
| "eval_runtime": 10.2317, | |
| "eval_samples_per_second": 195.569, | |
| "eval_steps_per_second": 24.532, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 13.01, | |
| "learning_rate": 3.3228187919463086e-05, | |
| "loss": 0.0034, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 13.27, | |
| "learning_rate": 3.289261744966443e-05, | |
| "loss": 0.0027, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 13.52, | |
| "learning_rate": 3.255704697986577e-05, | |
| "loss": 0.0036, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 13.78, | |
| "learning_rate": 3.222147651006712e-05, | |
| "loss": 0.0034, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 14.03, | |
| "learning_rate": 3.188590604026846e-05, | |
| "loss": 0.0034, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 14.03, | |
| "eval_accuracy": 0.9698596365660662, | |
| "eval_loss": 0.2152303010225296, | |
| "eval_runtime": 10.2495, | |
| "eval_samples_per_second": 195.228, | |
| "eval_steps_per_second": 24.489, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 14.29, | |
| "learning_rate": 3.1553691275167785e-05, | |
| "loss": 0.0029, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 14.54, | |
| "learning_rate": 3.1218120805369124e-05, | |
| "loss": 0.0023, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 14.8, | |
| "learning_rate": 3.088255033557047e-05, | |
| "loss": 0.0029, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 15.05, | |
| "learning_rate": 3.0546979865771816e-05, | |
| "loss": 0.0033, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 15.31, | |
| "learning_rate": 3.0211409395973156e-05, | |
| "loss": 0.0028, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 15.31, | |
| "eval_accuracy": 0.9705356077776452, | |
| "eval_loss": 0.20026059448719025, | |
| "eval_runtime": 10.2479, | |
| "eval_samples_per_second": 195.26, | |
| "eval_steps_per_second": 24.493, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 15.31, | |
| "step": 6000, | |
| "total_flos": 2.508593046513408e+16, | |
| "train_loss": 0.03989025120437145, | |
| "train_runtime": 1640.9678, | |
| "train_samples_per_second": 292.51, | |
| "train_steps_per_second": 9.141 | |
| } | |
| ], | |
| "max_steps": 15000, | |
| "num_train_epochs": 39, | |
| "total_flos": 2.508593046513408e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |