| { | |
| "best_metric": 0.511895477771759, | |
| "best_model_checkpoint": "./vit-beta2-0.995/checkpoint-2889", | |
| "epoch": 19.0, | |
| "eval_steps": 500, | |
| "global_step": 6099, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 30.705322265625, | |
| "learning_rate": 1.8234275822273514e-05, | |
| "loss": 1.7709, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7038834951456311, | |
| "eval_f1": 0.6317206539589807, | |
| "eval_loss": 0.9409339427947998, | |
| "eval_precision": 0.6600572679415466, | |
| "eval_recall": 0.7038834951456311, | |
| "eval_runtime": 22.6099, | |
| "eval_samples_per_second": 127.555, | |
| "eval_steps_per_second": 15.966, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 38.1316032409668, | |
| "learning_rate": 3.675706866705136e-05, | |
| "loss": 1.1633, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7371705963938974, | |
| "eval_f1": 0.6969962032458188, | |
| "eval_loss": 0.7317262887954712, | |
| "eval_precision": 0.7193055769194622, | |
| "eval_recall": 0.7371705963938974, | |
| "eval_runtime": 22.4171, | |
| "eval_samples_per_second": 128.652, | |
| "eval_steps_per_second": 16.104, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 10.181236267089844, | |
| "learning_rate": 5.52798615118292e-05, | |
| "loss": 1.0429, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7624826629680999, | |
| "eval_f1": 0.7239854360368981, | |
| "eval_loss": 0.6350400447845459, | |
| "eval_precision": 0.7357005373068777, | |
| "eval_recall": 0.7624826629680999, | |
| "eval_runtime": 22.621, | |
| "eval_samples_per_second": 127.492, | |
| "eval_steps_per_second": 15.959, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 36.46014404296875, | |
| "learning_rate": 7.380265435660705e-05, | |
| "loss": 0.9649, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7694174757281553, | |
| "eval_f1": 0.78081186335, | |
| "eval_loss": 0.5759614706039429, | |
| "eval_precision": 0.803793715931661, | |
| "eval_recall": 0.7694174757281553, | |
| "eval_runtime": 22.7229, | |
| "eval_samples_per_second": 126.92, | |
| "eval_steps_per_second": 15.887, | |
| "step": 1284 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 8.231016159057617, | |
| "learning_rate": 9.232544720138489e-05, | |
| "loss": 0.9051, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7669902912621359, | |
| "eval_f1": 0.77317554088632, | |
| "eval_loss": 0.6440545320510864, | |
| "eval_precision": 0.7941009873640001, | |
| "eval_recall": 0.7669902912621359, | |
| "eval_runtime": 22.7099, | |
| "eval_samples_per_second": 126.993, | |
| "eval_steps_per_second": 15.896, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 7.757928848266602, | |
| "learning_rate": 9.984876489938473e-05, | |
| "loss": 0.9826, | |
| "step": 1926 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.7850208044382802, | |
| "eval_f1": 0.7892025324025785, | |
| "eval_loss": 0.5661589503288269, | |
| "eval_precision": 0.7956162433005612, | |
| "eval_recall": 0.7850208044382802, | |
| "eval_runtime": 22.6328, | |
| "eval_samples_per_second": 127.425, | |
| "eval_steps_per_second": 15.95, | |
| "step": 1926 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 11.347550392150879, | |
| "learning_rate": 9.889494151200358e-05, | |
| "loss": 0.8855, | |
| "step": 2247 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.7264216366158114, | |
| "eval_f1": 0.7458569130340702, | |
| "eval_loss": 0.6881958246231079, | |
| "eval_precision": 0.7937071962372872, | |
| "eval_recall": 0.7264216366158114, | |
| "eval_runtime": 22.7206, | |
| "eval_samples_per_second": 126.933, | |
| "eval_steps_per_second": 15.889, | |
| "step": 2247 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 2.9686567783355713, | |
| "learning_rate": 9.707962612088379e-05, | |
| "loss": 0.789, | |
| "step": 2568 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.7364771151178918, | |
| "eval_f1": 0.7563577225180272, | |
| "eval_loss": 0.6491453051567078, | |
| "eval_precision": 0.8089248672404397, | |
| "eval_recall": 0.7364771151178918, | |
| "eval_runtime": 22.3088, | |
| "eval_samples_per_second": 129.277, | |
| "eval_steps_per_second": 16.182, | |
| "step": 2568 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 6.504317760467529, | |
| "learning_rate": 9.443480321450928e-05, | |
| "loss": 0.7192, | |
| "step": 2889 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.8075589459084604, | |
| "eval_f1": 0.8098127959154342, | |
| "eval_loss": 0.511895477771759, | |
| "eval_precision": 0.8207082348032128, | |
| "eval_recall": 0.8075589459084604, | |
| "eval_runtime": 22.6645, | |
| "eval_samples_per_second": 127.247, | |
| "eval_steps_per_second": 15.928, | |
| "step": 2889 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 12.737798690795898, | |
| "learning_rate": 9.100707257835249e-05, | |
| "loss": 0.7012, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.79750346740638, | |
| "eval_f1": 0.807690598198484, | |
| "eval_loss": 0.5413523316383362, | |
| "eval_precision": 0.8340905013524049, | |
| "eval_recall": 0.79750346740638, | |
| "eval_runtime": 22.8925, | |
| "eval_samples_per_second": 125.98, | |
| "eval_steps_per_second": 15.769, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 33.7946891784668, | |
| "learning_rate": 8.685682824178951e-05, | |
| "loss": 0.6376, | |
| "step": 3531 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.7947295423023578, | |
| "eval_f1": 0.8066140477733247, | |
| "eval_loss": 0.5712208151817322, | |
| "eval_precision": 0.8331612477677219, | |
| "eval_recall": 0.7947295423023578, | |
| "eval_runtime": 22.3907, | |
| "eval_samples_per_second": 128.804, | |
| "eval_steps_per_second": 16.123, | |
| "step": 3531 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 3.9330718517303467, | |
| "learning_rate": 8.205719438083829e-05, | |
| "loss": 0.5412, | |
| "step": 3852 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.8058252427184466, | |
| "eval_f1": 0.8145494176055466, | |
| "eval_loss": 0.5660970211029053, | |
| "eval_precision": 0.8327663750154051, | |
| "eval_recall": 0.8058252427184466, | |
| "eval_runtime": 23.2238, | |
| "eval_samples_per_second": 124.183, | |
| "eval_steps_per_second": 15.544, | |
| "step": 3852 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 15.180066108703613, | |
| "learning_rate": 7.669273692531118e-05, | |
| "loss": 0.4667, | |
| "step": 4173 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.819001386962552, | |
| "eval_f1": 0.8179067223293143, | |
| "eval_loss": 0.6374972462654114, | |
| "eval_precision": 0.8409999751324982, | |
| "eval_recall": 0.819001386962552, | |
| "eval_runtime": 22.8695, | |
| "eval_samples_per_second": 126.107, | |
| "eval_steps_per_second": 15.785, | |
| "step": 4173 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 2.8769447803497314, | |
| "learning_rate": 7.085797357089247e-05, | |
| "loss": 0.4766, | |
| "step": 4494 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.8252427184466019, | |
| "eval_f1": 0.83131443421793, | |
| "eval_loss": 0.5736179351806641, | |
| "eval_precision": 0.850771343089076, | |
| "eval_recall": 0.8252427184466019, | |
| "eval_runtime": 22.5348, | |
| "eval_samples_per_second": 127.98, | |
| "eval_steps_per_second": 16.02, | |
| "step": 4494 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 3.6918177604675293, | |
| "learning_rate": 6.46557084486047e-05, | |
| "loss": 0.384, | |
| "step": 4815 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.8356449375866851, | |
| "eval_f1": 0.8371244710047189, | |
| "eval_loss": 0.5305333137512207, | |
| "eval_precision": 0.8414602896223483, | |
| "eval_recall": 0.8356449375866851, | |
| "eval_runtime": 22.1675, | |
| "eval_samples_per_second": 130.1, | |
| "eval_steps_per_second": 16.285, | |
| "step": 4815 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 7.413918495178223, | |
| "learning_rate": 5.8195220793532045e-05, | |
| "loss": 0.37, | |
| "step": 5136 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.8314840499306518, | |
| "eval_f1": 0.8379252493969135, | |
| "eval_loss": 0.5530928373336792, | |
| "eval_precision": 0.8498567684865045, | |
| "eval_recall": 0.8314840499306518, | |
| "eval_runtime": 22.2316, | |
| "eval_samples_per_second": 129.725, | |
| "eval_steps_per_second": 16.238, | |
| "step": 5136 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 8.914341926574707, | |
| "learning_rate": 5.15903395270923e-05, | |
| "loss": 0.2809, | |
| "step": 5457 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.8637309292649098, | |
| "eval_f1": 0.860814136748839, | |
| "eval_loss": 0.5173911452293396, | |
| "eval_precision": 0.8629791065379673, | |
| "eval_recall": 0.8637309292649098, | |
| "eval_runtime": 22.2757, | |
| "eval_samples_per_second": 129.468, | |
| "eval_steps_per_second": 16.206, | |
| "step": 5457 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 18.052043914794922, | |
| "learning_rate": 4.495743767726598e-05, | |
| "loss": 0.2681, | |
| "step": 5778 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.8477808599167822, | |
| "eval_f1": 0.8504360089943235, | |
| "eval_loss": 0.5555988550186157, | |
| "eval_precision": 0.8554685811338774, | |
| "eval_recall": 0.8477808599167822, | |
| "eval_runtime": 22.3572, | |
| "eval_samples_per_second": 128.997, | |
| "eval_steps_per_second": 16.147, | |
| "step": 5778 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 0.04609627276659012, | |
| "learning_rate": 3.841338197358591e-05, | |
| "loss": 0.2139, | |
| "step": 6099 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.8255894590846047, | |
| "eval_f1": 0.8335016397329735, | |
| "eval_loss": 0.6290740966796875, | |
| "eval_precision": 0.8517909910468258, | |
| "eval_recall": 0.8255894590846047, | |
| "eval_runtime": 22.1886, | |
| "eval_samples_per_second": 129.977, | |
| "eval_steps_per_second": 16.27, | |
| "step": 6099 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "step": 6099, | |
| "total_flos": 7.550537882380222e+18, | |
| "train_loss": 0.713874793126947, | |
| "train_runtime": 2273.3603, | |
| "train_samples_per_second": 225.569, | |
| "train_steps_per_second": 14.12 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 32100, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 100, | |
| "save_steps": 500, | |
| "total_flos": 7.550537882380222e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |