| { | |
| "best_metric": 0.9729108101201125, | |
| "best_model_checkpoint": "/scratch/mrahma45/pixel/finetuned_models/mbert/mbert-base-finetuned-pos-ud-Hindi-HDTB/checkpoint-4000", | |
| "epoch": 15.625, | |
| "global_step": 6500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 7.92e-05, | |
| "loss": 0.9473, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 7.946845637583894e-05, | |
| "loss": 0.1783, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 7.893154362416109e-05, | |
| "loss": 0.1398, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 7.840000000000001e-05, | |
| "loss": 0.1279, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 7.786308724832216e-05, | |
| "loss": 0.0987, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "eval_accuracy": 0.964704546099895, | |
| "eval_loss": 0.1141478568315506, | |
| "eval_runtime": 8.4275, | |
| "eval_samples_per_second": 196.854, | |
| "eval_steps_per_second": 24.681, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "learning_rate": 7.73261744966443e-05, | |
| "loss": 0.0919, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "learning_rate": 7.678926174496645e-05, | |
| "loss": 0.0895, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "learning_rate": 7.62523489932886e-05, | |
| "loss": 0.0863, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "learning_rate": 7.571543624161075e-05, | |
| "loss": 0.0665, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 7.51785234899329e-05, | |
| "loss": 0.0605, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "eval_accuracy": 0.9700996677740864, | |
| "eval_loss": 0.10947530716657639, | |
| "eval_runtime": 8.4196, | |
| "eval_samples_per_second": 197.04, | |
| "eval_steps_per_second": 24.704, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "learning_rate": 7.464161073825505e-05, | |
| "loss": 0.0602, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "learning_rate": 7.410469798657718e-05, | |
| "loss": 0.0603, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "learning_rate": 7.356778523489933e-05, | |
| "loss": 0.0499, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 3.37, | |
| "learning_rate": 7.303087248322148e-05, | |
| "loss": 0.0439, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 3.61, | |
| "learning_rate": 7.249395973154363e-05, | |
| "loss": 0.045, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.61, | |
| "eval_accuracy": 0.9701848539057841, | |
| "eval_loss": 0.10890347510576248, | |
| "eval_runtime": 8.429, | |
| "eval_samples_per_second": 196.82, | |
| "eval_steps_per_second": 24.677, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "learning_rate": 7.195704697986577e-05, | |
| "loss": 0.0486, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 4.09, | |
| "learning_rate": 7.142013422818792e-05, | |
| "loss": 0.0414, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 4.33, | |
| "learning_rate": 7.088322147651007e-05, | |
| "loss": 0.0348, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 4.57, | |
| "learning_rate": 7.034630872483222e-05, | |
| "loss": 0.0359, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 4.81, | |
| "learning_rate": 6.980939597315437e-05, | |
| "loss": 0.038, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.81, | |
| "eval_accuracy": 0.9715194366357157, | |
| "eval_loss": 0.11456557363271713, | |
| "eval_runtime": 8.4277, | |
| "eval_samples_per_second": 196.85, | |
| "eval_steps_per_second": 24.68, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 5.05, | |
| "learning_rate": 6.927248322147651e-05, | |
| "loss": 0.0326, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 5.29, | |
| "learning_rate": 6.873557046979866e-05, | |
| "loss": 0.0269, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 5.53, | |
| "learning_rate": 6.819865771812081e-05, | |
| "loss": 0.0285, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 5.77, | |
| "learning_rate": 6.766174496644296e-05, | |
| "loss": 0.0294, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 6.01, | |
| "learning_rate": 6.712483221476511e-05, | |
| "loss": 0.0308, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 6.01, | |
| "eval_accuracy": 0.9711219013544595, | |
| "eval_loss": 0.11870458722114563, | |
| "eval_runtime": 8.4327, | |
| "eval_samples_per_second": 196.733, | |
| "eval_steps_per_second": 24.666, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 6.25, | |
| "learning_rate": 6.658791946308726e-05, | |
| "loss": 0.0213, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 6.49, | |
| "learning_rate": 6.60510067114094e-05, | |
| "loss": 0.0229, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 6.73, | |
| "learning_rate": 6.551409395973155e-05, | |
| "loss": 0.0261, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 6.97, | |
| "learning_rate": 6.49771812080537e-05, | |
| "loss": 0.0242, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 7.21, | |
| "learning_rate": 6.444026845637585e-05, | |
| "loss": 0.02, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 7.21, | |
| "eval_accuracy": 0.9710935059772269, | |
| "eval_loss": 0.1298820823431015, | |
| "eval_runtime": 8.4732, | |
| "eval_samples_per_second": 195.793, | |
| "eval_steps_per_second": 24.548, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 7.45, | |
| "learning_rate": 6.3903355704698e-05, | |
| "loss": 0.0186, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 7.69, | |
| "learning_rate": 6.336644295302015e-05, | |
| "loss": 0.0191, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 7.93, | |
| "learning_rate": 6.28295302013423e-05, | |
| "loss": 0.0202, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 8.17, | |
| "learning_rate": 6.229261744966444e-05, | |
| "loss": 0.0163, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 8.41, | |
| "learning_rate": 6.175570469798658e-05, | |
| "loss": 0.0159, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 8.41, | |
| "eval_accuracy": 0.9705539938098078, | |
| "eval_loss": 0.15600977838039398, | |
| "eval_runtime": 8.4103, | |
| "eval_samples_per_second": 197.259, | |
| "eval_steps_per_second": 24.732, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 8.65, | |
| "learning_rate": 6.121879194630873e-05, | |
| "loss": 0.0181, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 8.89, | |
| "learning_rate": 6.068187919463087e-05, | |
| "loss": 0.0175, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 9.13, | |
| "learning_rate": 6.014496644295302e-05, | |
| "loss": 0.0147, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 9.38, | |
| "learning_rate": 5.960805369127517e-05, | |
| "loss": 0.0134, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 9.62, | |
| "learning_rate": 5.907114093959732e-05, | |
| "loss": 0.0142, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 9.62, | |
| "eval_accuracy": 0.9729108101201125, | |
| "eval_loss": 0.13532687723636627, | |
| "eval_runtime": 8.4032, | |
| "eval_samples_per_second": 197.426, | |
| "eval_steps_per_second": 24.753, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 9.86, | |
| "learning_rate": 5.8534228187919466e-05, | |
| "loss": 0.0152, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 10.1, | |
| "learning_rate": 5.7997315436241614e-05, | |
| "loss": 0.0138, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 10.34, | |
| "learning_rate": 5.746040268456376e-05, | |
| "loss": 0.0109, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 10.58, | |
| "learning_rate": 5.692348993288591e-05, | |
| "loss": 0.0099, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 10.82, | |
| "learning_rate": 5.638657718120806e-05, | |
| "loss": 0.0124, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 10.82, | |
| "eval_accuracy": 0.9713206689950876, | |
| "eval_loss": 0.14761441946029663, | |
| "eval_runtime": 8.4464, | |
| "eval_samples_per_second": 196.415, | |
| "eval_steps_per_second": 24.626, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 11.06, | |
| "learning_rate": 5.58496644295302e-05, | |
| "loss": 0.0136, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 11.3, | |
| "learning_rate": 5.531275167785235e-05, | |
| "loss": 0.0091, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 11.54, | |
| "learning_rate": 5.47758389261745e-05, | |
| "loss": 0.0104, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 11.78, | |
| "learning_rate": 5.4238926174496645e-05, | |
| "loss": 0.011, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 12.02, | |
| "learning_rate": 5.3702013422818794e-05, | |
| "loss": 0.0112, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 12.02, | |
| "eval_accuracy": 0.9716614135218786, | |
| "eval_loss": 0.1558678150177002, | |
| "eval_runtime": 8.4483, | |
| "eval_samples_per_second": 196.371, | |
| "eval_steps_per_second": 24.62, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 12.26, | |
| "learning_rate": 5.316510067114094e-05, | |
| "loss": 0.0079, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "learning_rate": 5.262818791946309e-05, | |
| "loss": 0.0089, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 12.74, | |
| "learning_rate": 5.209127516778524e-05, | |
| "loss": 0.009, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 12.98, | |
| "learning_rate": 5.155436241610739e-05, | |
| "loss": 0.0098, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 13.22, | |
| "learning_rate": 5.1017449664429535e-05, | |
| "loss": 0.0081, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 13.22, | |
| "eval_accuracy": 0.9694465740977369, | |
| "eval_loss": 0.17652873694896698, | |
| "eval_runtime": 8.4104, | |
| "eval_samples_per_second": 197.255, | |
| "eval_steps_per_second": 24.731, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 13.46, | |
| "learning_rate": 5.048053691275168e-05, | |
| "loss": 0.0078, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 13.7, | |
| "learning_rate": 4.994362416107383e-05, | |
| "loss": 0.0074, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 13.94, | |
| "learning_rate": 4.940671140939597e-05, | |
| "loss": 0.008, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 14.18, | |
| "learning_rate": 4.886979865771812e-05, | |
| "loss": 0.0068, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 14.42, | |
| "learning_rate": 4.833288590604027e-05, | |
| "loss": 0.0057, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 14.42, | |
| "eval_accuracy": 0.9713490643723202, | |
| "eval_loss": 0.1719847023487091, | |
| "eval_runtime": 8.4464, | |
| "eval_samples_per_second": 196.415, | |
| "eval_steps_per_second": 24.626, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 14.66, | |
| "learning_rate": 4.779597315436242e-05, | |
| "loss": 0.0081, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 14.9, | |
| "learning_rate": 4.7259060402684566e-05, | |
| "loss": 0.0078, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 15.14, | |
| "learning_rate": 4.6722147651006714e-05, | |
| "loss": 0.0069, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 15.38, | |
| "learning_rate": 4.619060402684565e-05, | |
| "loss": 0.0056, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 15.62, | |
| "learning_rate": 4.5653691275167796e-05, | |
| "loss": 0.0066, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 15.62, | |
| "eval_accuracy": 0.9701564585285516, | |
| "eval_loss": 0.18945461511611938, | |
| "eval_runtime": 8.4725, | |
| "eval_samples_per_second": 195.811, | |
| "eval_steps_per_second": 24.55, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 15.62, | |
| "step": 6500, | |
| "total_flos": 2.716286806308864e+16, | |
| "train_loss": 0.046262450704207786, | |
| "train_runtime": 1858.7918, | |
| "train_samples_per_second": 258.232, | |
| "train_steps_per_second": 8.07 | |
| } | |
| ], | |
| "max_steps": 15000, | |
| "num_train_epochs": 37, | |
| "total_flos": 2.716286806308864e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |