| { | |
| "best_metric": 0.29478973150253296, | |
| "best_model_checkpoint": "./convnext-nano-3e-4-augment/checkpoint-2750", | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 2750, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 17.436264038085938, | |
| "learning_rate": 0.00029902226030228247, | |
| "loss": 1.4532, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 8.700443267822266, | |
| "learning_rate": 0.00029610178754135, | |
| "loss": 0.8119, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.827037773359841, | |
| "eval_loss": 0.5636539459228516, | |
| "eval_runtime": 61.1609, | |
| "eval_samples_per_second": 41.121, | |
| "eval_steps_per_second": 0.654, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 19.5541934967041, | |
| "learning_rate": 0.0002912766545459287, | |
| "loss": 0.7009, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 5.992691516876221, | |
| "learning_rate": 0.0002846097643037037, | |
| "loss": 0.5634, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 10.371554374694824, | |
| "learning_rate": 0.0002761880299246772, | |
| "loss": 0.5829, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8497017892644135, | |
| "eval_loss": 0.5015920400619507, | |
| "eval_runtime": 60.5306, | |
| "eval_samples_per_second": 41.549, | |
| "eval_steps_per_second": 0.661, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "grad_norm": 6.42030143737793, | |
| "learning_rate": 0.00026612124159586237, | |
| "loss": 0.5118, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "grad_norm": 5.376116752624512, | |
| "learning_rate": 0.00025454063529829405, | |
| "loss": 0.4771, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.91, | |
| "grad_norm": 9.462407112121582, | |
| "learning_rate": 0.00024159718194531572, | |
| "loss": 0.4623, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8755467196819086, | |
| "eval_loss": 0.44941774010658264, | |
| "eval_runtime": 60.9271, | |
| "eval_samples_per_second": 41.279, | |
| "eval_steps_per_second": 0.657, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 3.27, | |
| "grad_norm": 6.973038196563721, | |
| "learning_rate": 0.00022745961924584428, | |
| "loss": 0.3929, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.64, | |
| "grad_norm": 6.476861953735352, | |
| "learning_rate": 0.00021231225195028297, | |
| "loss": 0.3711, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 6.22141170501709, | |
| "learning_rate": 0.0001963525491562421, | |
| "loss": 0.359, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8886679920477137, | |
| "eval_loss": 0.3809404671192169, | |
| "eval_runtime": 61.1213, | |
| "eval_samples_per_second": 41.148, | |
| "eval_steps_per_second": 0.654, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.36, | |
| "grad_norm": 7.65138578414917, | |
| "learning_rate": 0.0001797885699968618, | |
| "loss": 0.3315, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 4.73, | |
| "grad_norm": 4.203142166137695, | |
| "learning_rate": 0.00016283625127182596, | |
| "loss": 0.2881, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8998011928429424, | |
| "eval_loss": 0.3741509020328522, | |
| "eval_runtime": 61.4609, | |
| "eval_samples_per_second": 40.92, | |
| "eval_steps_per_second": 0.651, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 5.09, | |
| "grad_norm": 6.004030227661133, | |
| "learning_rate": 0.00014571659238094556, | |
| "loss": 0.286, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 5.45, | |
| "grad_norm": 6.018474578857422, | |
| "learning_rate": 0.00012865277425900724, | |
| "loss": 0.2292, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 5.82, | |
| "grad_norm": 4.614959716796875, | |
| "learning_rate": 0.00011186724987097698, | |
| "loss": 0.2302, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9113320079522863, | |
| "eval_loss": 0.3401913046836853, | |
| "eval_runtime": 61.1216, | |
| "eval_samples_per_second": 41.147, | |
| "eval_steps_per_second": 0.654, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 6.18, | |
| "grad_norm": 4.227890968322754, | |
| "learning_rate": 9.557884419740386e-05, | |
| "loss": 0.2225, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 6.55, | |
| "grad_norm": 4.184410572052002, | |
| "learning_rate": 7.999990151614894e-05, | |
| "loss": 0.1851, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 6.91, | |
| "grad_norm": 6.824435234069824, | |
| "learning_rate": 6.533351716998465e-05, | |
| "loss": 0.1827, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9121272365805169, | |
| "eval_loss": 0.31500810384750366, | |
| "eval_runtime": 60.793, | |
| "eval_samples_per_second": 41.37, | |
| "eval_steps_per_second": 0.658, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 7.27, | |
| "grad_norm": 3.374882221221924, | |
| "learning_rate": 5.1770889908207245e-05, | |
| "loss": 0.1551, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 7.64, | |
| "grad_norm": 4.39635705947876, | |
| "learning_rate": 3.948882931853924e-05, | |
| "loss": 0.1564, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 2.6348471641540527, | |
| "learning_rate": 2.8647450843757897e-05, | |
| "loss": 0.1466, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9228628230616303, | |
| "eval_loss": 0.30119746923446655, | |
| "eval_runtime": 61.3308, | |
| "eval_samples_per_second": 41.007, | |
| "eval_steps_per_second": 0.652, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 8.36, | |
| "grad_norm": 5.662447929382324, | |
| "learning_rate": 1.9388088432033443e-05, | |
| "loss": 0.1236, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 8.73, | |
| "grad_norm": 0.9912707209587097, | |
| "learning_rate": 1.1831452032772498e-05, | |
| "loss": 0.1223, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.9248508946322067, | |
| "eval_loss": 0.2995615005493164, | |
| "eval_runtime": 60.6772, | |
| "eval_samples_per_second": 41.449, | |
| "eval_steps_per_second": 0.659, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 9.09, | |
| "grad_norm": 2.5069386959075928, | |
| "learning_rate": 6.076053957825411e-06, | |
| "loss": 0.1283, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 9.45, | |
| "grad_norm": 7.067609786987305, | |
| "learning_rate": 2.1969246228460523e-06, | |
| "loss": 0.1142, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 9.82, | |
| "grad_norm": 4.694984436035156, | |
| "learning_rate": 2.4463441107965276e-07, | |
| "loss": 0.1332, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.9248508946322067, | |
| "eval_loss": 0.29478973150253296, | |
| "eval_runtime": 61.4481, | |
| "eval_samples_per_second": 40.929, | |
| "eval_steps_per_second": 0.651, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 2750, | |
| "total_flos": 7.000491898906214e+18, | |
| "train_loss": 0.35536065743186257, | |
| "train_runtime": 6476.8205, | |
| "train_samples_per_second": 27.144, | |
| "train_steps_per_second": 0.425 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 2750, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 7.000491898906214e+18, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |