| { | |
| "best_metric": 0.9030581039755352, | |
| "best_model_checkpoint": "/n/holyscratch01/hlakkaraju_lab/Lab/aaronli/w2s_models/wts/boolq/llama3-8b-instruct/gpt2-xl/002/checkpoint-1000", | |
| "epoch": 3.903680916516389, | |
| "eval_steps": 100, | |
| "global_step": 2300, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 3.265306122448979e-07, | |
| "loss": 1.1307, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "eval_accuracy": 0.6275229357798165, | |
| "eval_loss": 0.7803781032562256, | |
| "eval_runtime": 330.8684, | |
| "eval_samples_per_second": 9.883, | |
| "eval_steps_per_second": 9.883, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 6.666666666666666e-07, | |
| "loss": 0.5918, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "eval_accuracy": 0.8012232415902141, | |
| "eval_loss": 0.48456913232803345, | |
| "eval_runtime": 330.4617, | |
| "eval_samples_per_second": 9.895, | |
| "eval_steps_per_second": 9.895, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 1e-06, | |
| "loss": 0.3622, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "eval_accuracy": 0.8660550458715597, | |
| "eval_loss": 0.3509620428085327, | |
| "eval_runtime": 330.4751, | |
| "eval_samples_per_second": 9.895, | |
| "eval_steps_per_second": 9.895, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 9.622783855149e-07, | |
| "loss": 0.3308, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "eval_accuracy": 0.8902140672782874, | |
| "eval_loss": 0.2842748761177063, | |
| "eval_runtime": 330.8369, | |
| "eval_samples_per_second": 9.884, | |
| "eval_steps_per_second": 9.884, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 9.245567710298e-07, | |
| "loss": 0.3175, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "eval_accuracy": 0.8889908256880734, | |
| "eval_loss": 0.27379733324050903, | |
| "eval_runtime": 330.6584, | |
| "eval_samples_per_second": 9.889, | |
| "eval_steps_per_second": 9.889, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "learning_rate": 8.868351565447001e-07, | |
| "loss": 0.2717, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "eval_accuracy": 0.8902140672782874, | |
| "eval_loss": 0.38391104340553284, | |
| "eval_runtime": 330.8255, | |
| "eval_samples_per_second": 9.884, | |
| "eval_steps_per_second": 9.884, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "learning_rate": 8.491135420596e-07, | |
| "loss": 0.1692, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "eval_accuracy": 0.8902140672782874, | |
| "eval_loss": 0.4322011470794678, | |
| "eval_runtime": 330.9477, | |
| "eval_samples_per_second": 9.881, | |
| "eval_steps_per_second": 9.881, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "learning_rate": 8.113919275745001e-07, | |
| "loss": 0.2002, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "eval_accuracy": 0.8963302752293578, | |
| "eval_loss": 0.5128087997436523, | |
| "eval_runtime": 330.6728, | |
| "eval_samples_per_second": 9.889, | |
| "eval_steps_per_second": 9.889, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "learning_rate": 7.736703130894002e-07, | |
| "loss": 0.2214, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "eval_accuracy": 0.9, | |
| "eval_loss": 0.3137321174144745, | |
| "eval_runtime": 330.5236, | |
| "eval_samples_per_second": 9.893, | |
| "eval_steps_per_second": 9.893, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "learning_rate": 7.359486986043002e-07, | |
| "loss": 0.1993, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "eval_accuracy": 0.9030581039755352, | |
| "eval_loss": 0.4319615960121155, | |
| "eval_runtime": 330.874, | |
| "eval_samples_per_second": 9.883, | |
| "eval_steps_per_second": 9.883, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "learning_rate": 6.982270841192002e-07, | |
| "loss": 0.1967, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "eval_accuracy": 0.901223241590214, | |
| "eval_loss": 0.43224194645881653, | |
| "eval_runtime": 330.5713, | |
| "eval_samples_per_second": 9.892, | |
| "eval_steps_per_second": 9.892, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "learning_rate": 6.605054696341003e-07, | |
| "loss": 0.1723, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "eval_accuracy": 0.9030581039755352, | |
| "eval_loss": 0.56269770860672, | |
| "eval_runtime": 330.8803, | |
| "eval_samples_per_second": 9.883, | |
| "eval_steps_per_second": 9.883, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "learning_rate": 6.227838551490004e-07, | |
| "loss": 0.0888, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "eval_accuracy": 0.9027522935779817, | |
| "eval_loss": 0.5143199563026428, | |
| "eval_runtime": 331.0535, | |
| "eval_samples_per_second": 9.878, | |
| "eval_steps_per_second": 9.878, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "learning_rate": 5.850622406639005e-07, | |
| "loss": 0.057, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "eval_accuracy": 0.9018348623853211, | |
| "eval_loss": 0.6684749722480774, | |
| "eval_runtime": 330.8995, | |
| "eval_samples_per_second": 9.882, | |
| "eval_steps_per_second": 9.882, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "learning_rate": 5.473406261788004e-07, | |
| "loss": 0.0556, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "eval_accuracy": 0.9021406727828746, | |
| "eval_loss": 0.690205454826355, | |
| "eval_runtime": 330.6813, | |
| "eval_samples_per_second": 9.889, | |
| "eval_steps_per_second": 9.889, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "learning_rate": 5.096190116937005e-07, | |
| "loss": 0.036, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "eval_accuracy": 0.9030581039755352, | |
| "eval_loss": 0.708600640296936, | |
| "eval_runtime": 330.8665, | |
| "eval_samples_per_second": 9.883, | |
| "eval_steps_per_second": 9.883, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "learning_rate": 4.718973972086005e-07, | |
| "loss": 0.0529, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "eval_accuracy": 0.901223241590214, | |
| "eval_loss": 0.6835083365440369, | |
| "eval_runtime": 330.509, | |
| "eval_samples_per_second": 9.894, | |
| "eval_steps_per_second": 9.894, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 3.06, | |
| "learning_rate": 4.3417578272350054e-07, | |
| "loss": 0.0265, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.06, | |
| "eval_accuracy": 0.9024464831804281, | |
| "eval_loss": 0.7818667888641357, | |
| "eval_runtime": 330.6981, | |
| "eval_samples_per_second": 9.888, | |
| "eval_steps_per_second": 9.888, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "learning_rate": 3.964541682384006e-07, | |
| "loss": 0.0204, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "eval_accuracy": 0.901223241590214, | |
| "eval_loss": 0.773235023021698, | |
| "eval_runtime": 330.8228, | |
| "eval_samples_per_second": 9.884, | |
| "eval_steps_per_second": 9.884, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 3.39, | |
| "learning_rate": 3.5873255375330063e-07, | |
| "loss": 0.0085, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.39, | |
| "eval_accuracy": 0.901223241590214, | |
| "eval_loss": 0.7827399373054504, | |
| "eval_runtime": 330.5421, | |
| "eval_samples_per_second": 9.893, | |
| "eval_steps_per_second": 9.893, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.56, | |
| "learning_rate": 3.2101093926820065e-07, | |
| "loss": 0.0098, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.56, | |
| "eval_accuracy": 0.900611620795107, | |
| "eval_loss": 0.8394690752029419, | |
| "eval_runtime": 330.8326, | |
| "eval_samples_per_second": 9.884, | |
| "eval_steps_per_second": 9.884, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.73, | |
| "learning_rate": 2.832893247831007e-07, | |
| "loss": 0.0059, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.73, | |
| "eval_accuracy": 0.9018348623853211, | |
| "eval_loss": 0.8700089454650879, | |
| "eval_runtime": 330.8106, | |
| "eval_samples_per_second": 9.885, | |
| "eval_steps_per_second": 9.885, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "learning_rate": 2.4556771029800075e-07, | |
| "loss": 0.002, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "eval_accuracy": 0.9015290519877676, | |
| "eval_loss": 0.9172555208206177, | |
| "eval_runtime": 330.8933, | |
| "eval_samples_per_second": 9.882, | |
| "eval_steps_per_second": 9.882, | |
| "step": 2300 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 2945, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 100, | |
| "total_flos": 1.024320316529664e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |