| { | |
| "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": 34.25452966739399, | |
| "learning_rate": 1.6129032258064517e-05, | |
| "loss": 3.7761, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 8.906774244439243, | |
| "learning_rate": 3.2258064516129034e-05, | |
| "loss": 0.9603, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 10.512197040633026, | |
| "learning_rate": 4.8387096774193554e-05, | |
| "loss": 0.6829, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 5.659258132140747, | |
| "learning_rate": 4.8387096774193554e-05, | |
| "loss": 0.6871, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 5.4454466530955, | |
| "learning_rate": 4.659498207885305e-05, | |
| "loss": 0.7293, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 2.078978399137911, | |
| "learning_rate": 4.4802867383512545e-05, | |
| "loss": 0.7045, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 2.1592322472119347, | |
| "learning_rate": 4.301075268817205e-05, | |
| "loss": 0.5735, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 1.7428085520266803, | |
| "learning_rate": 4.121863799283154e-05, | |
| "loss": 0.5189, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 1.1670263564721854, | |
| "learning_rate": 3.9426523297491045e-05, | |
| "loss": 0.5237, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 1.5260706910724648, | |
| "learning_rate": 3.763440860215054e-05, | |
| "loss": 0.4715, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 1.3126339162900238, | |
| "learning_rate": 3.5842293906810036e-05, | |
| "loss": 0.4616, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 1.0545831699338555, | |
| "learning_rate": 3.405017921146954e-05, | |
| "loss": 0.4508, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 0.8889763949621304, | |
| "learning_rate": 3.2258064516129034e-05, | |
| "loss": 0.4217, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 2.578136837094038, | |
| "learning_rate": 3.046594982078853e-05, | |
| "loss": 0.3953, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 1.0640758160669577, | |
| "learning_rate": 2.8673835125448028e-05, | |
| "loss": 0.3842, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.821638760863385, | |
| "learning_rate": 2.6881720430107527e-05, | |
| "loss": 0.38, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "grad_norm": 0.8187593197474141, | |
| "learning_rate": 2.5089605734767026e-05, | |
| "loss": 0.3925, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.8203780375140627, | |
| "learning_rate": 2.3297491039426525e-05, | |
| "loss": 0.383, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "grad_norm": 0.7124541736388171, | |
| "learning_rate": 2.1505376344086024e-05, | |
| "loss": 0.3815, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.7120761765250042, | |
| "learning_rate": 1.9713261648745522e-05, | |
| "loss": 0.3396, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.36, | |
| "grad_norm": 0.8983318764339185, | |
| "learning_rate": 1.7921146953405018e-05, | |
| "loss": 0.339, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 1.1337114647621995, | |
| "learning_rate": 1.6129032258064517e-05, | |
| "loss": 0.3278, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 3.68, | |
| "grad_norm": 0.9162119185377995, | |
| "learning_rate": 1.4336917562724014e-05, | |
| "loss": 0.3301, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.7552382013838203, | |
| "learning_rate": 1.2544802867383513e-05, | |
| "loss": 0.3274, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.6437764227175861, | |
| "learning_rate": 1.0752688172043012e-05, | |
| "loss": 0.3227, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 1.2354058415848357, | |
| "learning_rate": 8.960573476702509e-06, | |
| "loss": 0.287, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 4.32, | |
| "grad_norm": 0.7223142484779251, | |
| "learning_rate": 7.168458781362007e-06, | |
| "loss": 0.2795, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 0.9339238305532896, | |
| "learning_rate": 5.376344086021506e-06, | |
| "loss": 0.2927, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 4.64, | |
| "grad_norm": 0.8495922569216602, | |
| "learning_rate": 3.5842293906810035e-06, | |
| "loss": 0.2904, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.7820132151533432, | |
| "learning_rate": 1.7921146953405017e-06, | |
| "loss": 0.2897, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 4.96, | |
| "grad_norm": 0.8762178800956718, | |
| "learning_rate": 0.0, | |
| "loss": 0.2859, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 4.96, | |
| "step": 310, | |
| "total_flos": 3.956670103289856e+16, | |
| "train_loss": 0.5480726165156211, | |
| "train_runtime": 3327.5315, | |
| "train_samples_per_second": 3.005, | |
| "train_steps_per_second": 0.093 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 310, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 3.956670103289856e+16, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |