| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 250, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 3.27207350730896, | |
| "learning_rate": 4.980286753286195e-05, | |
| "loss": 1.3192, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 2.096735954284668, | |
| "learning_rate": 4.9214579028215776e-05, | |
| "loss": 1.2281, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 2.3264734745025635, | |
| "learning_rate": 4.8244412147206284e-05, | |
| "loss": 1.1693, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 2.9271645545959473, | |
| "learning_rate": 4.690766700109659e-05, | |
| "loss": 1.1378, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 2.3592920303344727, | |
| "learning_rate": 4.522542485937369e-05, | |
| "loss": 1.1255, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 2.8506274223327637, | |
| "learning_rate": 4.3224215685535294e-05, | |
| "loss": 1.077, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 3.32490873336792, | |
| "learning_rate": 4.093559974371725e-05, | |
| "loss": 1.1089, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 2.3261430263519287, | |
| "learning_rate": 3.8395669874474915e-05, | |
| "loss": 1.0634, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 2.7234225273132324, | |
| "learning_rate": 3.564448228912682e-05, | |
| "loss": 1.0618, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 2.270724296569824, | |
| "learning_rate": 3.302359024518024e-05, | |
| "loss": 1.0099, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 2.833041191101074, | |
| "learning_rate": 2.9992749512860173e-05, | |
| "loss": 1.0095, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 4.331037998199463, | |
| "learning_rate": 2.6883170138198323e-05, | |
| "loss": 1.0568, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 2.0229177474975586, | |
| "learning_rate": 2.3743892045505764e-05, | |
| "loss": 0.939, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 2.161649227142334, | |
| "learning_rate": 2.0624423525618098e-05, | |
| "loss": 0.8439, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 2.71807861328125, | |
| "learning_rate": 1.7573960460574133e-05, | |
| "loss": 0.8464, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 2.480152130126953, | |
| "learning_rate": 1.4640610475167898e-05, | |
| "loss": 0.8446, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 2.507754325866699, | |
| "learning_rate": 1.1870634250967605e-05, | |
| "loss": 0.8545, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 2.9544825553894043, | |
| "learning_rate": 9.553509672741645e-06, | |
| "loss": 0.8518, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 2.6840856075286865, | |
| "learning_rate": 7.211608069767867e-06, | |
| "loss": 0.8015, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 2.7335338592529297, | |
| "learning_rate": 5.150240033804116e-06, | |
| "loss": 0.8429, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 2.8081326484680176, | |
| "learning_rate": 3.4019145701791184e-06, | |
| "loss": 0.8786, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 2.5944387912750244, | |
| "learning_rate": 1.9942038158532407e-06, | |
| "loss": 0.8469, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 2.7935092449188232, | |
| "learning_rate": 9.493082103478517e-07, | |
| "loss": 0.8525, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 3.1116323471069336, | |
| "learning_rate": 2.8370638155215123e-07, | |
| "loss": 0.8663, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 2.7518155574798584, | |
| "learning_rate": 7.895267917501504e-09, | |
| "loss": 0.8084, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 250, | |
| "total_flos": 1.752671732105216e+17, | |
| "train_loss": 0.9777891387939454, | |
| "train_runtime": 5812.3304, | |
| "train_samples_per_second": 0.688, | |
| "train_steps_per_second": 0.043 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 250, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 1000, | |
| "total_flos": 1.752671732105216e+17, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |