| { | |
| "best_metric": 0.9091821374811841, | |
| "best_model_checkpoint": "/scratch/mrahma45/pixel/finetuned_models/canine/canine-base-finetuned-pos-ud-Korean-GSD/checkpoint-5000", | |
| "epoch": 54.34782608695652, | |
| "global_step": 7500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 7.840000000000001e-05, | |
| "loss": 1.4677, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "learning_rate": 7.947382550335571e-05, | |
| "loss": 0.5809, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "learning_rate": 7.893691275167786e-05, | |
| "loss": 0.4184, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "learning_rate": 7.840000000000001e-05, | |
| "loss": 0.339, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.62, | |
| "learning_rate": 7.786308724832216e-05, | |
| "loss": 0.2624, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.62, | |
| "eval_accuracy": 0.8917042983776551, | |
| "eval_loss": 0.3615211546421051, | |
| "eval_runtime": 3.8662, | |
| "eval_samples_per_second": 245.722, | |
| "eval_steps_per_second": 30.78, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.35, | |
| "learning_rate": 7.73261744966443e-05, | |
| "loss": 0.2237, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 5.07, | |
| "learning_rate": 7.678926174496645e-05, | |
| "loss": 0.1908, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 5.8, | |
| "learning_rate": 7.62523489932886e-05, | |
| "loss": 0.1433, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 6.52, | |
| "learning_rate": 7.571543624161075e-05, | |
| "loss": 0.1169, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 7.25, | |
| "learning_rate": 7.51785234899329e-05, | |
| "loss": 0.1012, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 7.25, | |
| "eval_accuracy": 0.8995651446730223, | |
| "eval_loss": 0.4672793745994568, | |
| "eval_runtime": 3.904, | |
| "eval_samples_per_second": 243.342, | |
| "eval_steps_per_second": 30.482, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 7.97, | |
| "learning_rate": 7.464161073825505e-05, | |
| "loss": 0.0865, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 8.7, | |
| "learning_rate": 7.410469798657718e-05, | |
| "loss": 0.0661, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 9.42, | |
| "learning_rate": 7.356778523489933e-05, | |
| "loss": 0.0596, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 10.14, | |
| "learning_rate": 7.303087248322148e-05, | |
| "loss": 0.0566, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 10.87, | |
| "learning_rate": 7.249395973154363e-05, | |
| "loss": 0.0461, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 10.87, | |
| "eval_accuracy": 0.9004014049172102, | |
| "eval_loss": 0.5751976370811462, | |
| "eval_runtime": 3.877, | |
| "eval_samples_per_second": 245.034, | |
| "eval_steps_per_second": 30.694, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 11.59, | |
| "learning_rate": 7.195704697986577e-05, | |
| "loss": 0.0401, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 12.32, | |
| "learning_rate": 7.142013422818792e-05, | |
| "loss": 0.0363, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 13.04, | |
| "learning_rate": 7.088322147651007e-05, | |
| "loss": 0.0335, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 13.77, | |
| "learning_rate": 7.034630872483222e-05, | |
| "loss": 0.0307, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 14.49, | |
| "learning_rate": 6.980939597315437e-05, | |
| "loss": 0.0305, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 14.49, | |
| "eval_accuracy": 0.9002341528683726, | |
| "eval_loss": 0.628495991230011, | |
| "eval_runtime": 3.8938, | |
| "eval_samples_per_second": 243.98, | |
| "eval_steps_per_second": 30.562, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 15.22, | |
| "learning_rate": 6.927248322147651e-05, | |
| "loss": 0.0256, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 15.94, | |
| "learning_rate": 6.873557046979866e-05, | |
| "loss": 0.0243, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 16.67, | |
| "learning_rate": 6.819865771812081e-05, | |
| "loss": 0.0199, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 17.39, | |
| "learning_rate": 6.766174496644296e-05, | |
| "loss": 0.022, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 18.12, | |
| "learning_rate": 6.712483221476511e-05, | |
| "loss": 0.0189, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 18.12, | |
| "eval_accuracy": 0.9037464458939622, | |
| "eval_loss": 0.6671296954154968, | |
| "eval_runtime": 3.8657, | |
| "eval_samples_per_second": 245.753, | |
| "eval_steps_per_second": 30.784, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 18.84, | |
| "learning_rate": 6.658791946308726e-05, | |
| "loss": 0.0185, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 19.57, | |
| "learning_rate": 6.60510067114094e-05, | |
| "loss": 0.0173, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 20.29, | |
| "learning_rate": 6.551409395973155e-05, | |
| "loss": 0.0186, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 21.01, | |
| "learning_rate": 6.49771812080537e-05, | |
| "loss": 0.0179, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 21.74, | |
| "learning_rate": 6.444026845637585e-05, | |
| "loss": 0.0137, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 21.74, | |
| "eval_accuracy": 0.9047499581869878, | |
| "eval_loss": 0.6996186375617981, | |
| "eval_runtime": 3.8813, | |
| "eval_samples_per_second": 244.763, | |
| "eval_steps_per_second": 30.66, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 22.46, | |
| "learning_rate": 6.3903355704698e-05, | |
| "loss": 0.0131, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 23.19, | |
| "learning_rate": 6.336644295302015e-05, | |
| "loss": 0.0135, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 23.91, | |
| "learning_rate": 6.28295302013423e-05, | |
| "loss": 0.0138, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 24.64, | |
| "learning_rate": 6.229261744966444e-05, | |
| "loss": 0.0134, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 25.36, | |
| "learning_rate": 6.175570469798658e-05, | |
| "loss": 0.0114, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 25.36, | |
| "eval_accuracy": 0.9012376651613982, | |
| "eval_loss": 0.7291152477264404, | |
| "eval_runtime": 3.8811, | |
| "eval_samples_per_second": 244.777, | |
| "eval_steps_per_second": 30.662, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 26.09, | |
| "learning_rate": 6.121879194630873e-05, | |
| "loss": 0.0114, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 26.81, | |
| "learning_rate": 6.068187919463087e-05, | |
| "loss": 0.0124, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 27.54, | |
| "learning_rate": 6.014496644295302e-05, | |
| "loss": 0.0114, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 28.26, | |
| "learning_rate": 5.960805369127517e-05, | |
| "loss": 0.0095, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 28.99, | |
| "learning_rate": 5.907114093959732e-05, | |
| "loss": 0.0096, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 28.99, | |
| "eval_accuracy": 0.9044990801137314, | |
| "eval_loss": 0.755575954914093, | |
| "eval_runtime": 3.856, | |
| "eval_samples_per_second": 246.369, | |
| "eval_steps_per_second": 30.861, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 29.71, | |
| "learning_rate": 5.8534228187919466e-05, | |
| "loss": 0.0086, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 30.43, | |
| "learning_rate": 5.7997315436241614e-05, | |
| "loss": 0.0103, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 31.16, | |
| "learning_rate": 5.746040268456376e-05, | |
| "loss": 0.0085, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 31.88, | |
| "learning_rate": 5.692348993288591e-05, | |
| "loss": 0.0091, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 32.61, | |
| "learning_rate": 5.638657718120806e-05, | |
| "loss": 0.0078, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 32.61, | |
| "eval_accuracy": 0.9049172102358254, | |
| "eval_loss": 0.7404294013977051, | |
| "eval_runtime": 3.8744, | |
| "eval_samples_per_second": 245.198, | |
| "eval_steps_per_second": 30.714, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 33.33, | |
| "learning_rate": 5.58496644295302e-05, | |
| "loss": 0.0072, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 34.06, | |
| "learning_rate": 5.531275167785235e-05, | |
| "loss": 0.0086, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 34.78, | |
| "learning_rate": 5.47758389261745e-05, | |
| "loss": 0.0075, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 35.51, | |
| "learning_rate": 5.4238926174496645e-05, | |
| "loss": 0.0079, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 36.23, | |
| "learning_rate": 5.3702013422818794e-05, | |
| "loss": 0.0071, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 36.23, | |
| "eval_accuracy": 0.9091821374811841, | |
| "eval_loss": 0.7556995153427124, | |
| "eval_runtime": 3.9089, | |
| "eval_samples_per_second": 243.038, | |
| "eval_steps_per_second": 30.444, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 36.96, | |
| "learning_rate": 5.316510067114094e-05, | |
| "loss": 0.0068, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 37.68, | |
| "learning_rate": 5.262818791946309e-05, | |
| "loss": 0.0063, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 38.41, | |
| "learning_rate": 5.209127516778524e-05, | |
| "loss": 0.0082, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 39.13, | |
| "learning_rate": 5.155436241610739e-05, | |
| "loss": 0.0064, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 39.86, | |
| "learning_rate": 5.1017449664429535e-05, | |
| "loss": 0.0065, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 39.86, | |
| "eval_accuracy": 0.9068406087974578, | |
| "eval_loss": 0.7957345843315125, | |
| "eval_runtime": 3.8473, | |
| "eval_samples_per_second": 246.929, | |
| "eval_steps_per_second": 30.931, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 40.58, | |
| "learning_rate": 5.048053691275168e-05, | |
| "loss": 0.0072, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 41.3, | |
| "learning_rate": 4.994362416107383e-05, | |
| "loss": 0.0061, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 42.03, | |
| "learning_rate": 4.940671140939597e-05, | |
| "loss": 0.006, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 42.75, | |
| "learning_rate": 4.886979865771812e-05, | |
| "loss": 0.0053, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 43.48, | |
| "learning_rate": 4.833288590604027e-05, | |
| "loss": 0.0063, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 43.48, | |
| "eval_accuracy": 0.907175112895133, | |
| "eval_loss": 0.7736948728561401, | |
| "eval_runtime": 3.8584, | |
| "eval_samples_per_second": 246.219, | |
| "eval_steps_per_second": 30.842, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 44.2, | |
| "learning_rate": 4.779597315436242e-05, | |
| "loss": 0.0059, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 44.93, | |
| "learning_rate": 4.7259060402684566e-05, | |
| "loss": 0.0062, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 45.65, | |
| "learning_rate": 4.6722147651006714e-05, | |
| "loss": 0.0056, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 46.38, | |
| "learning_rate": 4.618523489932886e-05, | |
| "loss": 0.0052, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 47.1, | |
| "learning_rate": 4.564832214765101e-05, | |
| "loss": 0.0037, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 47.1, | |
| "eval_accuracy": 0.9078441210904834, | |
| "eval_loss": 0.8221492171287537, | |
| "eval_runtime": 3.8863, | |
| "eval_samples_per_second": 244.45, | |
| "eval_steps_per_second": 30.621, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 47.83, | |
| "learning_rate": 4.511140939597316e-05, | |
| "loss": 0.004, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 48.55, | |
| "learning_rate": 4.457449664429531e-05, | |
| "loss": 0.0039, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 49.28, | |
| "learning_rate": 4.4037583892617455e-05, | |
| "loss": 0.0053, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "learning_rate": 4.35006711409396e-05, | |
| "loss": 0.0053, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 50.72, | |
| "learning_rate": 4.2963758389261745e-05, | |
| "loss": 0.004, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 50.72, | |
| "eval_accuracy": 0.9079277471149022, | |
| "eval_loss": 0.8471272587776184, | |
| "eval_runtime": 3.8324, | |
| "eval_samples_per_second": 247.888, | |
| "eval_steps_per_second": 31.051, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 51.45, | |
| "learning_rate": 4.242684563758389e-05, | |
| "loss": 0.0045, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 52.17, | |
| "learning_rate": 4.188993288590604e-05, | |
| "loss": 0.0031, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 52.9, | |
| "learning_rate": 4.135302013422819e-05, | |
| "loss": 0.0047, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 53.62, | |
| "learning_rate": 4.081610738255034e-05, | |
| "loss": 0.0052, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 54.35, | |
| "learning_rate": 4.0279194630872486e-05, | |
| "loss": 0.0039, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 54.35, | |
| "eval_accuracy": 0.9078441210904834, | |
| "eval_loss": 0.8163352012634277, | |
| "eval_runtime": 3.8952, | |
| "eval_samples_per_second": 243.889, | |
| "eval_steps_per_second": 30.55, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 54.35, | |
| "step": 7500, | |
| "total_flos": 3.927232531980288e+16, | |
| "train_loss": 0.06486416501204173, | |
| "train_runtime": 1116.591, | |
| "train_samples_per_second": 429.88, | |
| "train_steps_per_second": 13.434 | |
| } | |
| ], | |
| "max_steps": 15000, | |
| "num_train_epochs": 109, | |
| "total_flos": 3.927232531980288e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |