| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.96, | |
| "eval_steps": 500, | |
| "global_step": 310, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 17.163537984134056, | |
| "learning_rate": 1.6129032258064517e-05, | |
| "loss": 3.247, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 5.018646756033027, | |
| "learning_rate": 3.2258064516129034e-05, | |
| "loss": 0.9917, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 3.102932183151889, | |
| "learning_rate": 4.8387096774193554e-05, | |
| "loss": 0.5627, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 3.1713603687439784, | |
| "learning_rate": 4.8387096774193554e-05, | |
| "loss": 0.5618, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 2.4859823149760207, | |
| "learning_rate": 4.659498207885305e-05, | |
| "loss": 0.5205, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 2.272313000462373, | |
| "learning_rate": 4.4802867383512545e-05, | |
| "loss": 0.5124, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 2.5273618623771386, | |
| "learning_rate": 4.301075268817205e-05, | |
| "loss": 0.4395, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 1.90386763992575, | |
| "learning_rate": 4.121863799283154e-05, | |
| "loss": 0.4075, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 2.1818059655411046, | |
| "learning_rate": 3.9426523297491045e-05, | |
| "loss": 0.4218, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 1.8586027320750607, | |
| "learning_rate": 3.763440860215054e-05, | |
| "loss": 0.4045, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 2.367612894945561, | |
| "learning_rate": 3.5842293906810036e-05, | |
| "loss": 0.4047, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 1.7552435519555452, | |
| "learning_rate": 3.405017921146954e-05, | |
| "loss": 0.4006, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 1.6187500706970905, | |
| "learning_rate": 3.2258064516129034e-05, | |
| "loss": 0.3502, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 1.621006239401913, | |
| "learning_rate": 3.046594982078853e-05, | |
| "loss": 0.2912, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 1.8302472177918652, | |
| "learning_rate": 2.8673835125448028e-05, | |
| "loss": 0.2914, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 2.137720741340305, | |
| "learning_rate": 2.6881720430107527e-05, | |
| "loss": 0.2976, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "grad_norm": 1.6840762975810266, | |
| "learning_rate": 2.5089605734767026e-05, | |
| "loss": 0.306, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 1.623859665642121, | |
| "learning_rate": 2.3297491039426525e-05, | |
| "loss": 0.3036, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "grad_norm": 1.4462761218350422, | |
| "learning_rate": 2.1505376344086024e-05, | |
| "loss": 0.2824, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 1.5715993286842083, | |
| "learning_rate": 1.9713261648745522e-05, | |
| "loss": 0.1915, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.36, | |
| "grad_norm": 1.40640376627855, | |
| "learning_rate": 1.7921146953405018e-05, | |
| "loss": 0.197, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 1.5607010212609245, | |
| "learning_rate": 1.6129032258064517e-05, | |
| "loss": 0.1906, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 3.68, | |
| "grad_norm": 1.4815811366542657, | |
| "learning_rate": 1.4336917562724014e-05, | |
| "loss": 0.1974, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 1.5236794304094388, | |
| "learning_rate": 1.2544802867383513e-05, | |
| "loss": 0.1973, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.3798039353727727, | |
| "learning_rate": 1.0752688172043012e-05, | |
| "loss": 0.1867, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 1.3649861078638328, | |
| "learning_rate": 8.960573476702509e-06, | |
| "loss": 0.11, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 4.32, | |
| "grad_norm": 1.254492221709335, | |
| "learning_rate": 7.168458781362007e-06, | |
| "loss": 0.1094, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 1.189476576700767, | |
| "learning_rate": 5.376344086021506e-06, | |
| "loss": 0.1111, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 4.64, | |
| "grad_norm": 1.1598377753968125, | |
| "learning_rate": 3.5842293906810035e-06, | |
| "loss": 0.1026, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 1.0113867726860153, | |
| "learning_rate": 1.7921146953405017e-06, | |
| "loss": 0.1001, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 4.96, | |
| "grad_norm": 0.9129278851969916, | |
| "learning_rate": 0.0, | |
| "loss": 0.102, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 4.96, | |
| "step": 310, | |
| "total_flos": 2.2549700766859264e+16, | |
| "train_loss": 0.41267242623913675, | |
| "train_runtime": 2160.1701, | |
| "train_samples_per_second": 4.629, | |
| "train_steps_per_second": 0.144 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 310, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 2.2549700766859264e+16, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |