| { | |
| "best_metric": 0.984617888825588, | |
| "best_model_checkpoint": "/scratch/mrahma45/pixel/finetuned_models/mbert/mbert-base-finetuned-pos-ud-Japanese-GSD/checkpoint-5500", | |
| "epoch": 36.199095022624434, | |
| "global_step": 8000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 7.840000000000001e-05, | |
| "loss": 0.901, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 7.947382550335571e-05, | |
| "loss": 0.1239, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "learning_rate": 7.893691275167786e-05, | |
| "loss": 0.0812, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "learning_rate": 7.840000000000001e-05, | |
| "loss": 0.069, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "learning_rate": 7.786308724832216e-05, | |
| "loss": 0.0517, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "eval_accuracy": 0.9805485472450558, | |
| "eval_loss": 0.07757739722728729, | |
| "eval_runtime": 2.5773, | |
| "eval_samples_per_second": 196.718, | |
| "eval_steps_per_second": 24.832, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "learning_rate": 7.73261744966443e-05, | |
| "loss": 0.0422, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "learning_rate": 7.678926174496645e-05, | |
| "loss": 0.0332, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.62, | |
| "learning_rate": 7.62523489932886e-05, | |
| "loss": 0.0268, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 4.07, | |
| "learning_rate": 7.571543624161075e-05, | |
| "loss": 0.0264, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 4.52, | |
| "learning_rate": 7.51785234899329e-05, | |
| "loss": 0.0194, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.52, | |
| "eval_accuracy": 0.9803043867502238, | |
| "eval_loss": 0.08086052536964417, | |
| "eval_runtime": 2.5816, | |
| "eval_samples_per_second": 196.393, | |
| "eval_steps_per_second": 24.791, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.98, | |
| "learning_rate": 7.464161073825505e-05, | |
| "loss": 0.0179, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 5.43, | |
| "learning_rate": 7.410469798657718e-05, | |
| "loss": 0.0133, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 5.88, | |
| "learning_rate": 7.356778523489933e-05, | |
| "loss": 0.0157, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 6.33, | |
| "learning_rate": 7.303087248322148e-05, | |
| "loss": 0.0119, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 6.79, | |
| "learning_rate": 7.249395973154363e-05, | |
| "loss": 0.0124, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 6.79, | |
| "eval_accuracy": 0.9816879628876047, | |
| "eval_loss": 0.09609220921993256, | |
| "eval_runtime": 2.603, | |
| "eval_samples_per_second": 194.776, | |
| "eval_steps_per_second": 24.587, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 7.24, | |
| "learning_rate": 7.195704697986577e-05, | |
| "loss": 0.0123, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 7.69, | |
| "learning_rate": 7.142013422818792e-05, | |
| "loss": 0.0093, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 8.14, | |
| "learning_rate": 7.088322147651007e-05, | |
| "loss": 0.0099, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 8.6, | |
| "learning_rate": 7.034630872483222e-05, | |
| "loss": 0.009, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 9.05, | |
| "learning_rate": 6.980939597315437e-05, | |
| "loss": 0.0105, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 9.05, | |
| "eval_accuracy": 0.9804671604134451, | |
| "eval_loss": 0.0997396856546402, | |
| "eval_runtime": 2.5962, | |
| "eval_samples_per_second": 195.282, | |
| "eval_steps_per_second": 24.651, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 9.5, | |
| "learning_rate": 6.927248322147651e-05, | |
| "loss": 0.008, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 9.95, | |
| "learning_rate": 6.873557046979866e-05, | |
| "loss": 0.0082, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 10.41, | |
| "learning_rate": 6.819865771812081e-05, | |
| "loss": 0.0082, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 10.86, | |
| "learning_rate": 6.766174496644296e-05, | |
| "loss": 0.0078, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 11.31, | |
| "learning_rate": 6.712483221476511e-05, | |
| "loss": 0.0082, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 11.31, | |
| "eval_accuracy": 0.9815251892243835, | |
| "eval_loss": 0.09456193447113037, | |
| "eval_runtime": 2.6055, | |
| "eval_samples_per_second": 194.586, | |
| "eval_steps_per_second": 24.563, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 11.76, | |
| "learning_rate": 6.658791946308726e-05, | |
| "loss": 0.0065, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 12.22, | |
| "learning_rate": 6.60510067114094e-05, | |
| "loss": 0.0055, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 12.67, | |
| "learning_rate": 6.551409395973155e-05, | |
| "loss": 0.0055, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 13.12, | |
| "learning_rate": 6.49771812080537e-05, | |
| "loss": 0.0058, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 13.57, | |
| "learning_rate": 6.444026845637585e-05, | |
| "loss": 0.0052, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 13.57, | |
| "eval_accuracy": 0.9798974525921705, | |
| "eval_loss": 0.10926211625337601, | |
| "eval_runtime": 2.6042, | |
| "eval_samples_per_second": 194.688, | |
| "eval_steps_per_second": 24.576, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 14.03, | |
| "learning_rate": 6.3903355704698e-05, | |
| "loss": 0.0059, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 14.48, | |
| "learning_rate": 6.336644295302015e-05, | |
| "loss": 0.0065, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 14.93, | |
| "learning_rate": 6.28295302013423e-05, | |
| "loss": 0.0063, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 15.38, | |
| "learning_rate": 6.229261744966444e-05, | |
| "loss": 0.0056, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 15.84, | |
| "learning_rate": 6.176107382550336e-05, | |
| "loss": 0.0059, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 15.84, | |
| "eval_accuracy": 0.982094897045658, | |
| "eval_loss": 0.1131383404135704, | |
| "eval_runtime": 2.6028, | |
| "eval_samples_per_second": 194.787, | |
| "eval_steps_per_second": 24.588, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 16.29, | |
| "learning_rate": 6.122416107382551e-05, | |
| "loss": 0.0056, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 16.74, | |
| "learning_rate": 6.068724832214766e-05, | |
| "loss": 0.0044, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 17.19, | |
| "learning_rate": 6.015033557046981e-05, | |
| "loss": 0.0056, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 17.65, | |
| "learning_rate": 5.9613422818791955e-05, | |
| "loss": 0.004, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 18.1, | |
| "learning_rate": 5.90765100671141e-05, | |
| "loss": 0.0031, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 18.1, | |
| "eval_accuracy": 0.9822576707088793, | |
| "eval_loss": 0.11973382532596588, | |
| "eval_runtime": 2.6076, | |
| "eval_samples_per_second": 194.43, | |
| "eval_steps_per_second": 24.543, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 18.55, | |
| "learning_rate": 5.853959731543625e-05, | |
| "loss": 0.0053, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "learning_rate": 5.80026845637584e-05, | |
| "loss": 0.0054, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 19.46, | |
| "learning_rate": 5.7465771812080534e-05, | |
| "loss": 0.0054, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 19.91, | |
| "learning_rate": 5.692885906040268e-05, | |
| "loss": 0.0048, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 20.36, | |
| "learning_rate": 5.639194630872483e-05, | |
| "loss": 0.0046, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 20.36, | |
| "eval_accuracy": 0.9817693497192155, | |
| "eval_loss": 0.11302904784679413, | |
| "eval_runtime": 2.605, | |
| "eval_samples_per_second": 194.626, | |
| "eval_steps_per_second": 24.568, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 20.81, | |
| "learning_rate": 5.5855033557046986e-05, | |
| "loss": 0.0046, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 21.27, | |
| "learning_rate": 5.5318120805369134e-05, | |
| "loss": 0.0041, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 21.72, | |
| "learning_rate": 5.478120805369128e-05, | |
| "loss": 0.003, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 22.17, | |
| "learning_rate": 5.424429530201343e-05, | |
| "loss": 0.0042, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 22.62, | |
| "learning_rate": 5.370738255033558e-05, | |
| "loss": 0.0034, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 22.62, | |
| "eval_accuracy": 0.9838854073410922, | |
| "eval_loss": 0.11620958149433136, | |
| "eval_runtime": 2.6086, | |
| "eval_samples_per_second": 194.354, | |
| "eval_steps_per_second": 24.534, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 23.08, | |
| "learning_rate": 5.317046979865773e-05, | |
| "loss": 0.0032, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 23.53, | |
| "learning_rate": 5.2633557046979875e-05, | |
| "loss": 0.0033, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 23.98, | |
| "learning_rate": 5.2096644295302024e-05, | |
| "loss": 0.0033, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 24.43, | |
| "learning_rate": 5.155973154362417e-05, | |
| "loss": 0.0024, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 24.89, | |
| "learning_rate": 5.1022818791946307e-05, | |
| "loss": 0.0027, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 24.89, | |
| "eval_accuracy": 0.984617888825588, | |
| "eval_loss": 0.11322484910488129, | |
| "eval_runtime": 2.6066, | |
| "eval_samples_per_second": 194.503, | |
| "eval_steps_per_second": 24.553, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 25.34, | |
| "learning_rate": 5.0485906040268455e-05, | |
| "loss": 0.0033, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 25.79, | |
| "learning_rate": 4.99489932885906e-05, | |
| "loss": 0.0037, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 26.24, | |
| "learning_rate": 4.941208053691275e-05, | |
| "loss": 0.0028, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 26.7, | |
| "learning_rate": 4.88751677852349e-05, | |
| "loss": 0.0023, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 27.15, | |
| "learning_rate": 4.833825503355705e-05, | |
| "loss": 0.0021, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 27.15, | |
| "eval_accuracy": 0.9831529258565964, | |
| "eval_loss": 0.12333222478628159, | |
| "eval_runtime": 2.6126, | |
| "eval_samples_per_second": 194.063, | |
| "eval_steps_per_second": 24.497, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 27.6, | |
| "learning_rate": 4.7801342281879196e-05, | |
| "loss": 0.0026, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 28.05, | |
| "learning_rate": 4.726442953020135e-05, | |
| "loss": 0.0031, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 28.51, | |
| "learning_rate": 4.67275167785235e-05, | |
| "loss": 0.0027, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 28.96, | |
| "learning_rate": 4.619060402684565e-05, | |
| "loss": 0.003, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 29.41, | |
| "learning_rate": 4.5653691275167796e-05, | |
| "loss": 0.0023, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 29.41, | |
| "eval_accuracy": 0.9822576707088793, | |
| "eval_loss": 0.12343005836009979, | |
| "eval_runtime": 2.6249, | |
| "eval_samples_per_second": 193.151, | |
| "eval_steps_per_second": 24.382, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 29.86, | |
| "learning_rate": 4.511677852348993e-05, | |
| "loss": 0.0028, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 30.32, | |
| "learning_rate": 4.457986577181208e-05, | |
| "loss": 0.0022, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 30.77, | |
| "learning_rate": 4.404295302013423e-05, | |
| "loss": 0.0021, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 31.22, | |
| "learning_rate": 4.3506040268456375e-05, | |
| "loss": 0.0022, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 31.67, | |
| "learning_rate": 4.2969127516778523e-05, | |
| "loss": 0.0022, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 31.67, | |
| "eval_accuracy": 0.9826646048669325, | |
| "eval_loss": 0.11375758796930313, | |
| "eval_runtime": 2.6259, | |
| "eval_samples_per_second": 193.08, | |
| "eval_steps_per_second": 24.373, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 32.13, | |
| "learning_rate": 4.243221476510067e-05, | |
| "loss": 0.0017, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 32.58, | |
| "learning_rate": 4.189530201342282e-05, | |
| "loss": 0.0016, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 33.03, | |
| "learning_rate": 4.135838926174497e-05, | |
| "loss": 0.002, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 33.48, | |
| "learning_rate": 4.0821476510067116e-05, | |
| "loss": 0.0018, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 33.94, | |
| "learning_rate": 4.0284563758389265e-05, | |
| "loss": 0.0018, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 33.94, | |
| "eval_accuracy": 0.9799788394237813, | |
| "eval_loss": 0.13140234351158142, | |
| "eval_runtime": 2.6023, | |
| "eval_samples_per_second": 194.831, | |
| "eval_steps_per_second": 24.594, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 34.39, | |
| "learning_rate": 3.974765100671141e-05, | |
| "loss": 0.0024, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 34.84, | |
| "learning_rate": 3.921073825503356e-05, | |
| "loss": 0.0018, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 35.29, | |
| "learning_rate": 3.867382550335571e-05, | |
| "loss": 0.0021, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 35.75, | |
| "learning_rate": 3.813691275167786e-05, | |
| "loss": 0.0013, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 36.2, | |
| "learning_rate": 3.76e-05, | |
| "loss": 0.0025, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 36.2, | |
| "eval_accuracy": 0.9829087653617644, | |
| "eval_loss": 0.11779461801052094, | |
| "eval_runtime": 2.6153, | |
| "eval_samples_per_second": 193.859, | |
| "eval_steps_per_second": 24.471, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 36.2, | |
| "step": 8000, | |
| "total_flos": 3.3347033060634624e+16, | |
| "train_loss": 0.021707101341336965, | |
| "train_runtime": 2208.5517, | |
| "train_samples_per_second": 217.337, | |
| "train_steps_per_second": 6.792 | |
| } | |
| ], | |
| "max_steps": 15000, | |
| "num_train_epochs": 68, | |
| "total_flos": 3.3347033060634624e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |