| { | |
| "best_metric": 0.7816455696202531, | |
| "best_model_checkpoint": "tsec_vit_model/checkpoint-276", | |
| "epoch": 9.873417721518987, | |
| "eval_steps": 500, | |
| "global_step": 390, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.25316455696202533, | |
| "grad_norm": 1.1729109287261963, | |
| "learning_rate": 1.282051282051282e-05, | |
| "loss": 0.5143, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.5063291139240507, | |
| "grad_norm": 1.1752610206604004, | |
| "learning_rate": 2.564102564102564e-05, | |
| "loss": 0.559, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.759493670886076, | |
| "grad_norm": 1.3603609800338745, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 0.4963, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.9873417721518988, | |
| "eval_accuracy": 0.6724683544303798, | |
| "eval_loss": 0.5699456334114075, | |
| "eval_runtime": 4.299, | |
| "eval_samples_per_second": 147.012, | |
| "eval_steps_per_second": 9.305, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.0126582278481013, | |
| "grad_norm": 2.1996357440948486, | |
| "learning_rate": 4.985754985754986e-05, | |
| "loss": 0.5522, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.2658227848101267, | |
| "grad_norm": 1.0542365312576294, | |
| "learning_rate": 4.8433048433048433e-05, | |
| "loss": 0.4911, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.518987341772152, | |
| "grad_norm": 1.4736473560333252, | |
| "learning_rate": 4.700854700854701e-05, | |
| "loss": 0.5448, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.7721518987341773, | |
| "grad_norm": 1.3524911403656006, | |
| "learning_rate": 4.558404558404559e-05, | |
| "loss": 0.4959, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7151898734177216, | |
| "eval_loss": 0.5484605431556702, | |
| "eval_runtime": 4.1719, | |
| "eval_samples_per_second": 151.489, | |
| "eval_steps_per_second": 9.588, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 2.0253164556962027, | |
| "grad_norm": 1.6714565753936768, | |
| "learning_rate": 4.415954415954416e-05, | |
| "loss": 0.4969, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.278481012658228, | |
| "grad_norm": 1.8887039422988892, | |
| "learning_rate": 4.2735042735042735e-05, | |
| "loss": 0.4863, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 2.5316455696202533, | |
| "grad_norm": 2.640259027481079, | |
| "learning_rate": 4.131054131054131e-05, | |
| "loss": 0.5075, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.7848101265822782, | |
| "grad_norm": 1.167457103729248, | |
| "learning_rate": 3.988603988603989e-05, | |
| "loss": 0.4879, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.9873417721518987, | |
| "eval_accuracy": 0.7689873417721519, | |
| "eval_loss": 0.48862743377685547, | |
| "eval_runtime": 4.3228, | |
| "eval_samples_per_second": 146.201, | |
| "eval_steps_per_second": 9.253, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 3.037974683544304, | |
| "grad_norm": 1.0525858402252197, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 0.4946, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 3.291139240506329, | |
| "grad_norm": 2.056792736053467, | |
| "learning_rate": 3.7037037037037037e-05, | |
| "loss": 0.4884, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 3.5443037974683547, | |
| "grad_norm": 1.7059816122055054, | |
| "learning_rate": 3.561253561253561e-05, | |
| "loss": 0.4926, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 3.7974683544303796, | |
| "grad_norm": 1.2786283493041992, | |
| "learning_rate": 3.418803418803419e-05, | |
| "loss": 0.5243, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7468354430379747, | |
| "eval_loss": 0.5133278965950012, | |
| "eval_runtime": 4.3787, | |
| "eval_samples_per_second": 144.335, | |
| "eval_steps_per_second": 9.135, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 4.050632911392405, | |
| "grad_norm": 2.1460752487182617, | |
| "learning_rate": 3.2763532763532764e-05, | |
| "loss": 0.4983, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 4.30379746835443, | |
| "grad_norm": 1.596248984336853, | |
| "learning_rate": 3.133903133903134e-05, | |
| "loss": 0.4936, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 4.556962025316456, | |
| "grad_norm": 1.6535227298736572, | |
| "learning_rate": 2.9914529914529915e-05, | |
| "loss": 0.5049, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 4.810126582278481, | |
| "grad_norm": 1.3324358463287354, | |
| "learning_rate": 2.8490028490028492e-05, | |
| "loss": 0.4654, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 4.987341772151899, | |
| "eval_accuracy": 0.7515822784810127, | |
| "eval_loss": 0.49274376034736633, | |
| "eval_runtime": 4.3657, | |
| "eval_samples_per_second": 144.763, | |
| "eval_steps_per_second": 9.162, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 5.063291139240507, | |
| "grad_norm": 1.212802529335022, | |
| "learning_rate": 2.706552706552707e-05, | |
| "loss": 0.4828, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 5.3164556962025316, | |
| "grad_norm": 1.189599633216858, | |
| "learning_rate": 2.564102564102564e-05, | |
| "loss": 0.4816, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 5.569620253164557, | |
| "grad_norm": 1.3671377897262573, | |
| "learning_rate": 2.4216524216524217e-05, | |
| "loss": 0.4709, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 5.822784810126582, | |
| "grad_norm": 1.5688464641571045, | |
| "learning_rate": 2.2792022792022794e-05, | |
| "loss": 0.4776, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.7642405063291139, | |
| "eval_loss": 0.4901277422904968, | |
| "eval_runtime": 4.2593, | |
| "eval_samples_per_second": 148.382, | |
| "eval_steps_per_second": 9.391, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 6.075949367088608, | |
| "grad_norm": 1.6727008819580078, | |
| "learning_rate": 2.1367521367521368e-05, | |
| "loss": 0.4878, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 6.329113924050633, | |
| "grad_norm": 1.2271509170532227, | |
| "learning_rate": 1.9943019943019945e-05, | |
| "loss": 0.4725, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 6.582278481012658, | |
| "grad_norm": 1.5308541059494019, | |
| "learning_rate": 1.8518518518518518e-05, | |
| "loss": 0.444, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 6.8354430379746836, | |
| "grad_norm": 1.6207118034362793, | |
| "learning_rate": 1.7094017094017095e-05, | |
| "loss": 0.4767, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 6.987341772151899, | |
| "eval_accuracy": 0.7816455696202531, | |
| "eval_loss": 0.46520036458969116, | |
| "eval_runtime": 4.1881, | |
| "eval_samples_per_second": 150.905, | |
| "eval_steps_per_second": 9.551, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 7.0886075949367084, | |
| "grad_norm": 1.3883824348449707, | |
| "learning_rate": 1.566951566951567e-05, | |
| "loss": 0.4649, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 7.341772151898734, | |
| "grad_norm": 2.2651216983795166, | |
| "learning_rate": 1.4245014245014246e-05, | |
| "loss": 0.4644, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 7.594936708860759, | |
| "grad_norm": 1.2660713195800781, | |
| "learning_rate": 1.282051282051282e-05, | |
| "loss": 0.4448, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 7.848101265822785, | |
| "grad_norm": 1.8906290531158447, | |
| "learning_rate": 1.1396011396011397e-05, | |
| "loss": 0.4465, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.7642405063291139, | |
| "eval_loss": 0.4795072674751282, | |
| "eval_runtime": 4.3585, | |
| "eval_samples_per_second": 145.004, | |
| "eval_steps_per_second": 9.177, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 8.10126582278481, | |
| "grad_norm": 1.7846542596817017, | |
| "learning_rate": 9.971509971509972e-06, | |
| "loss": 0.4552, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 8.354430379746836, | |
| "grad_norm": 1.2796516418457031, | |
| "learning_rate": 8.547008547008548e-06, | |
| "loss": 0.4284, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 8.60759493670886, | |
| "grad_norm": 1.181015133857727, | |
| "learning_rate": 7.122507122507123e-06, | |
| "loss": 0.3985, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 8.860759493670885, | |
| "grad_norm": 1.352704644203186, | |
| "learning_rate": 5.6980056980056985e-06, | |
| "loss": 0.467, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 8.987341772151899, | |
| "eval_accuracy": 0.7484177215189873, | |
| "eval_loss": 0.4690525233745575, | |
| "eval_runtime": 4.2658, | |
| "eval_samples_per_second": 148.154, | |
| "eval_steps_per_second": 9.377, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 9.113924050632912, | |
| "grad_norm": 1.3692480325698853, | |
| "learning_rate": 4.273504273504274e-06, | |
| "loss": 0.437, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 9.367088607594937, | |
| "grad_norm": 2.6086838245391846, | |
| "learning_rate": 2.8490028490028492e-06, | |
| "loss": 0.4062, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 9.620253164556962, | |
| "grad_norm": 1.352748155593872, | |
| "learning_rate": 1.4245014245014246e-06, | |
| "loss": 0.4365, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 9.873417721518987, | |
| "grad_norm": 1.9576495885849, | |
| "learning_rate": 0.0, | |
| "loss": 0.4121, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 9.873417721518987, | |
| "eval_accuracy": 0.7689873417721519, | |
| "eval_loss": 0.482129842042923, | |
| "eval_runtime": 4.3754, | |
| "eval_samples_per_second": 144.443, | |
| "eval_steps_per_second": 9.142, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 9.873417721518987, | |
| "step": 390, | |
| "total_flos": 1.9321077770606223e+18, | |
| "train_loss": 0.4781974340096498, | |
| "train_runtime": 382.4209, | |
| "train_samples_per_second": 66.027, | |
| "train_steps_per_second": 1.02 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 390, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.9321077770606223e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |