| { | |
| "best_metric": 0.74, | |
| "best_model_checkpoint": "distilhubert-finetuned-not-a-word2/run-8/checkpoint-168", | |
| "epoch": 9.0, | |
| "eval_steps": 500, | |
| "global_step": 216, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 1.6496813297271729, | |
| "learning_rate": 2.9441483262927863e-06, | |
| "loss": 0.6992, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.620004415512085, | |
| "learning_rate": 5.888296652585573e-06, | |
| "loss": 0.6939, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.5726878046989441, | |
| "learning_rate": 8.832444978878358e-06, | |
| "loss": 0.6835, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 1.4239176511764526, | |
| "learning_rate": 1.1776593305171145e-05, | |
| "loss": 0.6673, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_f1": 0.7326732673267327, | |
| "eval_loss": 0.6805419921875, | |
| "eval_runtime": 1.3716, | |
| "eval_samples_per_second": 46.661, | |
| "eval_steps_per_second": 5.833, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 0.9195191264152527, | |
| "learning_rate": 1.406648644784331e-05, | |
| "loss": 0.6673, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 1.8524231910705566, | |
| "learning_rate": 1.3739358856033002e-05, | |
| "loss": 0.6155, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 1.8213531970977783, | |
| "learning_rate": 1.3412231264222692e-05, | |
| "loss": 0.5895, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 0.4818130433559418, | |
| "learning_rate": 1.3085103672412383e-05, | |
| "loss": 0.6468, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 0.6597484946250916, | |
| "learning_rate": 1.2757976080602073e-05, | |
| "loss": 0.6173, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_f1": 0.7326732673267327, | |
| "eval_loss": 0.6970634460449219, | |
| "eval_runtime": 1.3632, | |
| "eval_samples_per_second": 46.949, | |
| "eval_steps_per_second": 5.869, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 0.48903289437294006, | |
| "learning_rate": 1.2430848488791764e-05, | |
| "loss": 0.6302, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "grad_norm": 0.6064260601997375, | |
| "learning_rate": 1.2103720896981454e-05, | |
| "loss": 0.5867, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.6802453398704529, | |
| "learning_rate": 1.1776593305171145e-05, | |
| "loss": 0.6321, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "grad_norm": 1.2592875957489014, | |
| "learning_rate": 1.1449465713360835e-05, | |
| "loss": 0.6223, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 1.1591824293136597, | |
| "learning_rate": 1.1122338121550526e-05, | |
| "loss": 0.4922, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_f1": 0.7326732673267327, | |
| "eval_loss": 0.7079887390136719, | |
| "eval_runtime": 1.3669, | |
| "eval_samples_per_second": 46.821, | |
| "eval_steps_per_second": 5.853, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "grad_norm": 0.5332023501396179, | |
| "learning_rate": 1.0795210529740214e-05, | |
| "loss": 0.5989, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "grad_norm": 0.5555600523948669, | |
| "learning_rate": 1.0468082937929906e-05, | |
| "loss": 0.6104, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.54, | |
| "grad_norm": 1.2928024530410767, | |
| "learning_rate": 1.0140955346119596e-05, | |
| "loss": 0.4936, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 1.1424989700317383, | |
| "learning_rate": 9.813827754309287e-06, | |
| "loss": 0.6191, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.96, | |
| "grad_norm": 1.119732141494751, | |
| "learning_rate": 9.486700162498977e-06, | |
| "loss": 0.6004, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_f1": 0.7326732673267327, | |
| "eval_loss": 0.7053489685058594, | |
| "eval_runtime": 1.3556, | |
| "eval_samples_per_second": 47.211, | |
| "eval_steps_per_second": 5.901, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "grad_norm": 0.8135461211204529, | |
| "learning_rate": 9.159572570688668e-06, | |
| "loss": 0.5154, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 4.38, | |
| "grad_norm": 1.8034342527389526, | |
| "learning_rate": 8.832444978878358e-06, | |
| "loss": 0.6067, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 4.58, | |
| "grad_norm": 0.8029685020446777, | |
| "learning_rate": 8.505317387068049e-06, | |
| "loss": 0.5499, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.79, | |
| "grad_norm": 1.019626259803772, | |
| "learning_rate": 8.178189795257739e-06, | |
| "loss": 0.5542, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.861674427986145, | |
| "learning_rate": 7.85106220344743e-06, | |
| "loss": 0.5545, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_f1": 0.7326732673267327, | |
| "eval_loss": 0.6832504272460938, | |
| "eval_runtime": 1.3811, | |
| "eval_samples_per_second": 46.341, | |
| "eval_steps_per_second": 5.793, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 5.21, | |
| "grad_norm": 1.5949212312698364, | |
| "learning_rate": 7.523934611637121e-06, | |
| "loss": 0.4806, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 5.42, | |
| "grad_norm": 3.002861738204956, | |
| "learning_rate": 7.196807019826811e-06, | |
| "loss": 0.5832, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 5.62, | |
| "grad_norm": 1.4606820344924927, | |
| "learning_rate": 6.9351049463785626e-06, | |
| "loss": 0.5481, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 5.83, | |
| "grad_norm": 1.6088628768920898, | |
| "learning_rate": 6.607977354568253e-06, | |
| "loss": 0.5333, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_f1": 0.7326732673267327, | |
| "eval_loss": 0.6599597930908203, | |
| "eval_runtime": 1.3747, | |
| "eval_samples_per_second": 46.555, | |
| "eval_steps_per_second": 5.819, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 6.04, | |
| "grad_norm": 2.2051286697387695, | |
| "learning_rate": 6.280849762757943e-06, | |
| "loss": 0.4805, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 6.25, | |
| "grad_norm": 1.6964988708496094, | |
| "learning_rate": 5.953722170947633e-06, | |
| "loss": 0.4573, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 6.46, | |
| "grad_norm": 2.1374056339263916, | |
| "learning_rate": 5.626594579137324e-06, | |
| "loss": 0.59, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 6.67, | |
| "grad_norm": 1.8037084341049194, | |
| "learning_rate": 5.299466987327015e-06, | |
| "loss": 0.494, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 6.88, | |
| "grad_norm": 1.7916295528411865, | |
| "learning_rate": 4.972339395516706e-06, | |
| "loss": 0.4997, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_f1": 0.74, | |
| "eval_loss": 0.6653976440429688, | |
| "eval_runtime": 1.3593, | |
| "eval_samples_per_second": 47.083, | |
| "eval_steps_per_second": 5.885, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 7.08, | |
| "grad_norm": 2.2235190868377686, | |
| "learning_rate": 4.645211803706396e-06, | |
| "loss": 0.3887, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 7.29, | |
| "grad_norm": 1.239268183708191, | |
| "learning_rate": 4.318084211896087e-06, | |
| "loss": 0.4519, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 1.8677798509597778, | |
| "learning_rate": 3.990956620085777e-06, | |
| "loss": 0.5146, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 7.71, | |
| "grad_norm": 3.8495407104492188, | |
| "learning_rate": 3.6638290282754668e-06, | |
| "loss": 0.4237, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 7.92, | |
| "grad_norm": 1.8828785419464111, | |
| "learning_rate": 3.3367014364651573e-06, | |
| "loss": 0.5033, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_f1": 0.74, | |
| "eval_loss": 0.6842975616455078, | |
| "eval_runtime": 1.3685, | |
| "eval_samples_per_second": 46.767, | |
| "eval_steps_per_second": 5.846, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 8.12, | |
| "grad_norm": 4.3374714851379395, | |
| "learning_rate": 3.009573844654848e-06, | |
| "loss": 0.4588, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 8.33, | |
| "grad_norm": 2.9799509048461914, | |
| "learning_rate": 2.6824462528445384e-06, | |
| "loss": 0.3783, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 8.54, | |
| "grad_norm": 3.3768601417541504, | |
| "learning_rate": 2.3553186610342286e-06, | |
| "loss": 0.4366, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 8.75, | |
| "grad_norm": 2.2495288848876953, | |
| "learning_rate": 2.028191069223919e-06, | |
| "loss": 0.4545, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 8.96, | |
| "grad_norm": 2.078002691268921, | |
| "learning_rate": 1.7010634774136097e-06, | |
| "loss": 0.6012, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_f1": 0.74, | |
| "eval_loss": 0.6836881637573242, | |
| "eval_runtime": 1.3545, | |
| "eval_samples_per_second": 47.25, | |
| "eval_steps_per_second": 5.906, | |
| "step": 216 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 240, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 7545814381042464.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": { | |
| "learning_rate": 1.4131911966205373e-05, | |
| "per_device_train_batch_size": 8 | |
| } | |
| } | |