| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.976, | |
| "eval_steps": 500, | |
| "global_step": 186, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.6117611690953515, | |
| "learning_rate": 1.0526315789473683e-08, | |
| "loss": 0.4933, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 4.485227520015997, | |
| "learning_rate": 5.2631578947368416e-08, | |
| "loss": 0.4996, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 4.765524959241675, | |
| "learning_rate": 1.0526315789473683e-07, | |
| "loss": 0.5018, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 4.710513305333378, | |
| "learning_rate": 1.5789473684210525e-07, | |
| "loss": 0.4844, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 4.537921702304328, | |
| "learning_rate": 1.9998230608242914e-07, | |
| "loss": 0.4739, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 3.917599997580791, | |
| "learning_rate": 1.993636761762317e-07, | |
| "loss": 0.4759, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 3.988201318412564, | |
| "learning_rate": 1.9786660193313148e-07, | |
| "loss": 0.4606, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 2.843430355228346, | |
| "learning_rate": 1.9550431853565574e-07, | |
| "loss": 0.4171, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 2.9412621389290257, | |
| "learning_rate": 1.9229771021990092e-07, | |
| "loss": 0.4075, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 2.5891798071506584, | |
| "learning_rate": 1.8827512564429073e-07, | |
| "loss": 0.3895, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 2.311260337839049, | |
| "learning_rate": 1.8347212726762435e-07, | |
| "loss": 0.3826, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 1.7315001394623175, | |
| "learning_rate": 1.7793117695208826e-07, | |
| "loss": 0.3429, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 1.3441016711237108, | |
| "learning_rate": 1.717012605707203e-07, | |
| "loss": 0.3348, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "eval_loss": 0.2483493983745575, | |
| "eval_runtime": 131.0647, | |
| "eval_samples_per_second": 0.275, | |
| "eval_steps_per_second": 0.069, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 1.2392330789677268, | |
| "learning_rate": 1.6483745493805867e-07, | |
| "loss": 0.3216, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 1.219453051753039, | |
| "learning_rate": 1.5740044089261134e-07, | |
| "loss": 0.3185, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 1.1222350719240004, | |
| "learning_rate": 1.4945596683583752e-07, | |
| "loss": 0.3083, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 1.1144412166765716, | |
| "learning_rate": 1.4107426747033074e-07, | |
| "loss": 0.3047, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 1.1344036025375182, | |
| "learning_rate": 1.323294428759652e-07, | |
| "loss": 0.3083, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 1.114097744195408, | |
| "learning_rate": 1.232988034134053e-07, | |
| "loss": 0.2968, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 1.04860266630711, | |
| "learning_rate": 1.1406218624648985e-07, | |
| "loss": 0.2964, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 1.0412871936488755, | |
| "learning_rate": 1.0470124952590976e-07, | |
| "loss": 0.2933, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 1.0589673455582587, | |
| "learning_rate": 9.529875047409026e-08, | |
| "loss": 0.2908, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 1.0322727028980907, | |
| "learning_rate": 8.59378137535102e-08, | |
| "loss": 0.2756, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 1.0299988465119796, | |
| "learning_rate": 7.670119658659468e-08, | |
| "loss": 0.276, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 1.0150896602116855, | |
| "learning_rate": 6.76705571240348e-08, | |
| "loss": 0.2666, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.0283555629278627, | |
| "learning_rate": 5.892573252966926e-08, | |
| "loss": 0.2659, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.18834702670574188, | |
| "eval_runtime": 131.135, | |
| "eval_samples_per_second": 0.275, | |
| "eval_steps_per_second": 0.069, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 1.0093549041505132, | |
| "learning_rate": 5.054403316416247e-08, | |
| "loss": 0.2524, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 1.018811573391022, | |
| "learning_rate": 4.2599559107388636e-08, | |
| "loss": 0.2506, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 1.039650875543289, | |
| "learning_rate": 3.516254506194133e-08, | |
| "loss": 0.2578, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 1.035580954839146, | |
| "learning_rate": 2.8298739429279706e-08, | |
| "loss": 0.2522, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 1.0360132857828035, | |
| "learning_rate": 2.2068823047911754e-08, | |
| "loss": 0.2517, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 1.0267979523571857, | |
| "learning_rate": 1.6527872732375646e-08, | |
| "loss": 0.2386, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 1.0903062573289466, | |
| "learning_rate": 1.1724874355709258e-08, | |
| "loss": 0.2412, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 1.049473662481498, | |
| "learning_rate": 7.70228978009907e-09, | |
| "loss": 0.2397, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "grad_norm": 1.0294155234866598, | |
| "learning_rate": 4.4956814643442585e-09, | |
| "loss": 0.2353, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 1.0562623811700738, | |
| "learning_rate": 2.1333980668685415e-09, | |
| "loss": 0.2418, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 1.0446606479668537, | |
| "learning_rate": 6.363238237683033e-10, | |
| "loss": 0.2383, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 1.0364893408749978, | |
| "learning_rate": 1.7693917570837934e-11, | |
| "loss": 0.2447, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.98, | |
| "eval_loss": 0.1706668883562088, | |
| "eval_runtime": 129.4132, | |
| "eval_samples_per_second": 0.278, | |
| "eval_steps_per_second": 0.07, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.98, | |
| "step": 186, | |
| "total_flos": 30250327080960.0, | |
| "train_loss": 0.3222779505034929, | |
| "train_runtime": 18917.8994, | |
| "train_samples_per_second": 0.079, | |
| "train_steps_per_second": 0.01 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 186, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 1, | |
| "total_flos": 30250327080960.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |