| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.7692307692307693, | |
| "eval_steps": 10, | |
| "global_step": 250, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.006153846153846154, | |
| "eval_loss": 1.845949411392212, | |
| "eval_runtime": 1.9761, | |
| "eval_samples_per_second": 57.69, | |
| "eval_steps_per_second": 3.036, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03076923076923077, | |
| "grad_norm": 29.46457290649414, | |
| "learning_rate": 7.692307692307694e-07, | |
| "loss": 1.944, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03076923076923077, | |
| "eval_loss": 1.7965577840805054, | |
| "eval_runtime": 1.9506, | |
| "eval_samples_per_second": 58.444, | |
| "eval_steps_per_second": 3.076, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06153846153846154, | |
| "grad_norm": 14.49207878112793, | |
| "learning_rate": 1.5384615384615387e-06, | |
| "loss": 1.6999, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06153846153846154, | |
| "eval_loss": 1.5032916069030762, | |
| "eval_runtime": 1.9599, | |
| "eval_samples_per_second": 58.168, | |
| "eval_steps_per_second": 3.061, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.09230769230769231, | |
| "grad_norm": 11.355466842651367, | |
| "learning_rate": 2.307692307692308e-06, | |
| "loss": 1.1499, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.09230769230769231, | |
| "eval_loss": 0.8391174674034119, | |
| "eval_runtime": 1.9437, | |
| "eval_samples_per_second": 58.65, | |
| "eval_steps_per_second": 3.087, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.12307692307692308, | |
| "grad_norm": 5.648468017578125, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 0.7194, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.12307692307692308, | |
| "eval_loss": 0.6716201305389404, | |
| "eval_runtime": 1.9517, | |
| "eval_samples_per_second": 58.411, | |
| "eval_steps_per_second": 3.074, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 5.752841949462891, | |
| "learning_rate": 3.846153846153847e-06, | |
| "loss": 0.6811, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.15384615384615385, | |
| "eval_loss": 0.6289324164390564, | |
| "eval_runtime": 1.952, | |
| "eval_samples_per_second": 58.401, | |
| "eval_steps_per_second": 3.074, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.18461538461538463, | |
| "grad_norm": 3.414722204208374, | |
| "learning_rate": 4.615384615384616e-06, | |
| "loss": 0.5924, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.18461538461538463, | |
| "eval_loss": 0.6140013337135315, | |
| "eval_runtime": 1.94, | |
| "eval_samples_per_second": 58.762, | |
| "eval_steps_per_second": 3.093, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.2153846153846154, | |
| "grad_norm": 5.556036949157715, | |
| "learning_rate": 5.384615384615385e-06, | |
| "loss": 0.6135, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.2153846153846154, | |
| "eval_loss": 0.6054026484489441, | |
| "eval_runtime": 1.9659, | |
| "eval_samples_per_second": 57.988, | |
| "eval_steps_per_second": 3.052, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.24615384615384617, | |
| "grad_norm": 4.436710834503174, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.5952, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.24615384615384617, | |
| "eval_loss": 0.5986860990524292, | |
| "eval_runtime": 1.9428, | |
| "eval_samples_per_second": 58.677, | |
| "eval_steps_per_second": 3.088, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.27692307692307694, | |
| "grad_norm": 3.496018409729004, | |
| "learning_rate": 6.923076923076923e-06, | |
| "loss": 0.5887, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.27692307692307694, | |
| "eval_loss": 0.594973623752594, | |
| "eval_runtime": 1.9468, | |
| "eval_samples_per_second": 58.558, | |
| "eval_steps_per_second": 3.082, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 2.3540539741516113, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.5634, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "eval_loss": 0.5954164266586304, | |
| "eval_runtime": 1.948, | |
| "eval_samples_per_second": 58.521, | |
| "eval_steps_per_second": 3.08, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3384615384615385, | |
| "grad_norm": 2.9880635738372803, | |
| "learning_rate": 8.461538461538462e-06, | |
| "loss": 0.5827, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.3384615384615385, | |
| "eval_loss": 0.5955133438110352, | |
| "eval_runtime": 1.9432, | |
| "eval_samples_per_second": 58.667, | |
| "eval_steps_per_second": 3.088, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.36923076923076925, | |
| "grad_norm": 1.9405996799468994, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.5726, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.36923076923076925, | |
| "eval_loss": 0.5951128602027893, | |
| "eval_runtime": 1.9715, | |
| "eval_samples_per_second": 57.824, | |
| "eval_steps_per_second": 3.043, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 1.9948021173477173, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6488, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "eval_loss": 0.5964463949203491, | |
| "eval_runtime": 2.1032, | |
| "eval_samples_per_second": 54.203, | |
| "eval_steps_per_second": 2.853, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.4307692307692308, | |
| "grad_norm": 1.95350980758667, | |
| "learning_rate": 1.076923076923077e-05, | |
| "loss": 0.6218, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.4307692307692308, | |
| "eval_loss": 0.5996471047401428, | |
| "eval_runtime": 1.9539, | |
| "eval_samples_per_second": 58.345, | |
| "eval_steps_per_second": 3.071, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 3.840015411376953, | |
| "learning_rate": 1.1538461538461538e-05, | |
| "loss": 0.5585, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "eval_loss": 0.6000372767448425, | |
| "eval_runtime": 1.9457, | |
| "eval_samples_per_second": 58.592, | |
| "eval_steps_per_second": 3.084, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.49230769230769234, | |
| "grad_norm": 2.345364809036255, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.5314, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.49230769230769234, | |
| "eval_loss": 0.6025042533874512, | |
| "eval_runtime": 1.9398, | |
| "eval_samples_per_second": 58.768, | |
| "eval_steps_per_second": 3.093, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.5230769230769231, | |
| "grad_norm": 3.162071704864502, | |
| "learning_rate": 1.3076923076923078e-05, | |
| "loss": 0.6901, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5230769230769231, | |
| "eval_loss": 0.598171055316925, | |
| "eval_runtime": 1.9419, | |
| "eval_samples_per_second": 58.705, | |
| "eval_steps_per_second": 3.09, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5538461538461539, | |
| "grad_norm": 3.9432108402252197, | |
| "learning_rate": 1.3846153846153847e-05, | |
| "loss": 0.6604, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.5538461538461539, | |
| "eval_loss": 0.5974885821342468, | |
| "eval_runtime": 1.9664, | |
| "eval_samples_per_second": 57.975, | |
| "eval_steps_per_second": 3.051, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.5846153846153846, | |
| "grad_norm": 1.904718279838562, | |
| "learning_rate": 1.4615384615384615e-05, | |
| "loss": 0.5806, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.5846153846153846, | |
| "eval_loss": 0.6012160778045654, | |
| "eval_runtime": 2.082, | |
| "eval_samples_per_second": 54.755, | |
| "eval_steps_per_second": 2.882, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 2.2823173999786377, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.588, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "eval_loss": 0.6052933931350708, | |
| "eval_runtime": 1.9613, | |
| "eval_samples_per_second": 58.124, | |
| "eval_steps_per_second": 3.059, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6461538461538462, | |
| "grad_norm": 1.8020161390304565, | |
| "learning_rate": 1.6153846153846154e-05, | |
| "loss": 0.5826, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.6461538461538462, | |
| "eval_loss": 0.6047356724739075, | |
| "eval_runtime": 1.9487, | |
| "eval_samples_per_second": 58.5, | |
| "eval_steps_per_second": 3.079, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.676923076923077, | |
| "grad_norm": 1.8297995328903198, | |
| "learning_rate": 1.6923076923076924e-05, | |
| "loss": 0.5442, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.676923076923077, | |
| "eval_loss": 0.6105689406394958, | |
| "eval_runtime": 2.0567, | |
| "eval_samples_per_second": 55.429, | |
| "eval_steps_per_second": 2.917, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.7076923076923077, | |
| "grad_norm": 1.8141131401062012, | |
| "learning_rate": 1.7692307692307694e-05, | |
| "loss": 0.602, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.7076923076923077, | |
| "eval_loss": 0.6149886250495911, | |
| "eval_runtime": 2.1951, | |
| "eval_samples_per_second": 51.933, | |
| "eval_steps_per_second": 2.733, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.7384615384615385, | |
| "grad_norm": 2.093683958053589, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.6103, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.7384615384615385, | |
| "eval_loss": 0.621900200843811, | |
| "eval_runtime": 1.9465, | |
| "eval_samples_per_second": 58.567, | |
| "eval_steps_per_second": 3.082, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 1.840038776397705, | |
| "learning_rate": 1.923076923076923e-05, | |
| "loss": 0.5775, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "eval_loss": 0.6232128739356995, | |
| "eval_runtime": 1.9883, | |
| "eval_samples_per_second": 57.334, | |
| "eval_steps_per_second": 3.018, | |
| "step": 250 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 2600, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 8, | |
| "save_steps": 50, | |
| "total_flos": 2.170153002126541e+16, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |