| { | |
| "best_metric": 0.8568872987477638, | |
| "best_model_checkpoint": "/scratch/mrahma45/pixel/finetuned_models/bert/bert-base-finetuned-pos-ud-Hindi-HDTB/checkpoint-3000", | |
| "epoch": 13.221153846153847, | |
| "global_step": 5500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 4.9500000000000004e-05, | |
| "loss": 1.5512, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 4.966778523489933e-05, | |
| "loss": 0.7227, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 4.933221476510068e-05, | |
| "loss": 0.6046, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 4.8996644295302016e-05, | |
| "loss": 0.5466, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 4.8661073825503355e-05, | |
| "loss": 0.5069, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "eval_accuracy": 0.8260783144504075, | |
| "eval_loss": 0.4958227276802063, | |
| "eval_runtime": 8.4268, | |
| "eval_samples_per_second": 196.872, | |
| "eval_steps_per_second": 24.683, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "learning_rate": 4.83255033557047e-05, | |
| "loss": 0.4882, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "learning_rate": 4.798993288590604e-05, | |
| "loss": 0.4728, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "learning_rate": 4.765436241610739e-05, | |
| "loss": 0.4576, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "learning_rate": 4.7318791946308726e-05, | |
| "loss": 0.4339, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 4.698322147651007e-05, | |
| "loss": 0.4185, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "eval_accuracy": 0.8392821648635602, | |
| "eval_loss": 0.46815404295921326, | |
| "eval_runtime": 8.5929, | |
| "eval_samples_per_second": 193.067, | |
| "eval_steps_per_second": 24.206, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "learning_rate": 4.664765100671141e-05, | |
| "loss": 0.4086, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "learning_rate": 4.631208053691276e-05, | |
| "loss": 0.4081, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "learning_rate": 4.5976510067114097e-05, | |
| "loss": 0.3903, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 3.37, | |
| "learning_rate": 4.564093959731544e-05, | |
| "loss": 0.3658, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 3.61, | |
| "learning_rate": 4.5305369127516775e-05, | |
| "loss": 0.3701, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.61, | |
| "eval_accuracy": 0.8504699434931993, | |
| "eval_loss": 0.4506780803203583, | |
| "eval_runtime": 8.4658, | |
| "eval_samples_per_second": 195.965, | |
| "eval_steps_per_second": 24.569, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "learning_rate": 4.496979865771812e-05, | |
| "loss": 0.3701, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 4.09, | |
| "learning_rate": 4.463422818791946e-05, | |
| "loss": 0.3607, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 4.33, | |
| "learning_rate": 4.4298657718120806e-05, | |
| "loss": 0.3278, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 4.57, | |
| "learning_rate": 4.3963087248322146e-05, | |
| "loss": 0.3374, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 4.81, | |
| "learning_rate": 4.362751677852349e-05, | |
| "loss": 0.3336, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.81, | |
| "eval_accuracy": 0.8520316892409916, | |
| "eval_loss": 0.44800490140914917, | |
| "eval_runtime": 8.4418, | |
| "eval_samples_per_second": 196.522, | |
| "eval_steps_per_second": 24.639, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 5.05, | |
| "learning_rate": 4.329194630872484e-05, | |
| "loss": 0.3316, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 5.29, | |
| "learning_rate": 4.295637583892618e-05, | |
| "loss": 0.2929, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 5.53, | |
| "learning_rate": 4.262080536912752e-05, | |
| "loss": 0.3, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 5.77, | |
| "learning_rate": 4.228523489932886e-05, | |
| "loss": 0.3029, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 6.01, | |
| "learning_rate": 4.194966442953021e-05, | |
| "loss": 0.3045, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 6.01, | |
| "eval_accuracy": 0.854814436209785, | |
| "eval_loss": 0.4508860409259796, | |
| "eval_runtime": 8.4218, | |
| "eval_samples_per_second": 196.99, | |
| "eval_steps_per_second": 24.698, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 6.25, | |
| "learning_rate": 4.161409395973155e-05, | |
| "loss": 0.265, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 6.49, | |
| "learning_rate": 4.127852348993289e-05, | |
| "loss": 0.2695, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 6.73, | |
| "learning_rate": 4.0942953020134226e-05, | |
| "loss": 0.2749, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 6.97, | |
| "learning_rate": 4.060738255033557e-05, | |
| "loss": 0.2735, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 7.21, | |
| "learning_rate": 4.027181208053691e-05, | |
| "loss": 0.2419, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 7.21, | |
| "eval_accuracy": 0.8568872987477638, | |
| "eval_loss": 0.4852503538131714, | |
| "eval_runtime": 8.4429, | |
| "eval_samples_per_second": 196.497, | |
| "eval_steps_per_second": 24.636, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 7.45, | |
| "learning_rate": 3.993624161073826e-05, | |
| "loss": 0.2386, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 7.69, | |
| "learning_rate": 3.96006711409396e-05, | |
| "loss": 0.241, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 7.93, | |
| "learning_rate": 3.926510067114094e-05, | |
| "loss": 0.2447, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 8.17, | |
| "learning_rate": 3.892953020134228e-05, | |
| "loss": 0.2193, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 8.41, | |
| "learning_rate": 3.859395973154363e-05, | |
| "loss": 0.2148, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 8.41, | |
| "eval_accuracy": 0.855439134508902, | |
| "eval_loss": 0.5063510537147522, | |
| "eval_runtime": 8.6104, | |
| "eval_samples_per_second": 192.673, | |
| "eval_steps_per_second": 24.157, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 8.65, | |
| "learning_rate": 3.825838926174497e-05, | |
| "loss": 0.2085, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 8.89, | |
| "learning_rate": 3.7922818791946313e-05, | |
| "loss": 0.2195, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 9.13, | |
| "learning_rate": 3.758724832214765e-05, | |
| "loss": 0.2027, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 9.38, | |
| "learning_rate": 3.725167785234899e-05, | |
| "loss": 0.183, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 9.62, | |
| "learning_rate": 3.691610738255034e-05, | |
| "loss": 0.1883, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 9.62, | |
| "eval_accuracy": 0.8551835761138087, | |
| "eval_loss": 0.5312172174453735, | |
| "eval_runtime": 8.4553, | |
| "eval_samples_per_second": 196.208, | |
| "eval_steps_per_second": 24.6, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 9.86, | |
| "learning_rate": 3.658053691275168e-05, | |
| "loss": 0.1903, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 10.1, | |
| "learning_rate": 3.624496644295302e-05, | |
| "loss": 0.1764, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 10.34, | |
| "learning_rate": 3.590939597315436e-05, | |
| "loss": 0.158, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 10.58, | |
| "learning_rate": 3.557382550335571e-05, | |
| "loss": 0.1643, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 10.82, | |
| "learning_rate": 3.523825503355705e-05, | |
| "loss": 0.1642, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 10.82, | |
| "eval_accuracy": 0.852003293863759, | |
| "eval_loss": 0.5592557787895203, | |
| "eval_runtime": 8.4747, | |
| "eval_samples_per_second": 195.759, | |
| "eval_steps_per_second": 24.544, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 11.06, | |
| "learning_rate": 3.4902684563758394e-05, | |
| "loss": 0.1645, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 11.3, | |
| "learning_rate": 3.456711409395973e-05, | |
| "loss": 0.1353, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 11.54, | |
| "learning_rate": 3.423154362416108e-05, | |
| "loss": 0.1462, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 11.78, | |
| "learning_rate": 3.389597315436242e-05, | |
| "loss": 0.1451, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 12.02, | |
| "learning_rate": 3.356040268456376e-05, | |
| "loss": 0.1488, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 12.02, | |
| "eval_accuracy": 0.8523156430133174, | |
| "eval_loss": 0.587023913860321, | |
| "eval_runtime": 8.4932, | |
| "eval_samples_per_second": 195.333, | |
| "eval_steps_per_second": 24.49, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 12.26, | |
| "learning_rate": 3.32248322147651e-05, | |
| "loss": 0.1163, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "learning_rate": 3.288926174496644e-05, | |
| "loss": 0.1224, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 12.74, | |
| "learning_rate": 3.255369127516779e-05, | |
| "loss": 0.1281, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 12.98, | |
| "learning_rate": 3.221812080536913e-05, | |
| "loss": 0.1305, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 13.22, | |
| "learning_rate": 3.1882550335570474e-05, | |
| "loss": 0.1027, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 13.22, | |
| "eval_accuracy": 0.8522588522588522, | |
| "eval_loss": 0.6439880132675171, | |
| "eval_runtime": 8.5081, | |
| "eval_samples_per_second": 194.991, | |
| "eval_steps_per_second": 24.447, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 13.22, | |
| "step": 5500, | |
| "total_flos": 2.298364364453069e+16, | |
| "train_loss": 0.310650390625, | |
| "train_runtime": 1477.8296, | |
| "train_samples_per_second": 324.801, | |
| "train_steps_per_second": 10.15 | |
| } | |
| ], | |
| "max_steps": 15000, | |
| "num_train_epochs": 37, | |
| "total_flos": 2.298364364453069e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |