| { | |
| "best_metric": 0.17973944544792175, | |
| "best_model_checkpoint": "./mistral/01-03-24-Weni-ZeroShot-3.3.18-Mistral-7b-Multilanguage-3.2.0_Zeroshot-2_max_steps-800_batch_256_2024-03-01_ppid_6/checkpoint-400", | |
| "epoch": 3.9603960396039604, | |
| "eval_steps": 100, | |
| "global_step": 400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 2.1325972080230713, | |
| "learning_rate": 4.5e-05, | |
| "loss": 1.424, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.3059396743774414, | |
| "learning_rate": 9.5e-05, | |
| "loss": 0.5357, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.20484383404254913, | |
| "learning_rate": 0.000145, | |
| "loss": 0.4403, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.3068823516368866, | |
| "learning_rate": 0.000195, | |
| "loss": 0.4135, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.1791614145040512, | |
| "learning_rate": 0.0001996917333733128, | |
| "loss": 0.3859, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "eval_loss": 0.3723442554473877, | |
| "eval_runtime": 93.8197, | |
| "eval_samples_per_second": 30.559, | |
| "eval_steps_per_second": 0.959, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 0.18006961047649384, | |
| "learning_rate": 0.00019862856015372317, | |
| "loss": 0.3567, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 0.18006713688373566, | |
| "learning_rate": 0.0001968147640378108, | |
| "loss": 0.3436, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 0.20077605545520782, | |
| "learning_rate": 0.00019426414910921787, | |
| "loss": 0.3322, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 0.2087867707014084, | |
| "learning_rate": 0.00019099612708765434, | |
| "loss": 0.3143, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 0.21649543941020966, | |
| "learning_rate": 0.00018703556959398998, | |
| "loss": 0.3011, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "eval_loss": 0.2929498553276062, | |
| "eval_runtime": 93.8381, | |
| "eval_samples_per_second": 30.553, | |
| "eval_steps_per_second": 0.959, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "grad_norm": 0.2589591145515442, | |
| "learning_rate": 0.00018241261886220154, | |
| "loss": 0.275, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "grad_norm": 0.22108183801174164, | |
| "learning_rate": 0.00017716245833877201, | |
| "loss": 0.2632, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "grad_norm": 0.233080193400383, | |
| "learning_rate": 0.00017132504491541818, | |
| "loss": 0.2515, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "grad_norm": 0.23667199909687042, | |
| "learning_rate": 0.00016494480483301836, | |
| "loss": 0.2407, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "grad_norm": 0.27916812896728516, | |
| "learning_rate": 0.00015807029557109398, | |
| "loss": 0.2293, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "eval_loss": 0.2265055775642395, | |
| "eval_runtime": 93.8776, | |
| "eval_samples_per_second": 30.54, | |
| "eval_steps_per_second": 0.959, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "grad_norm": 0.25347772240638733, | |
| "learning_rate": 0.00015075383629607042, | |
| "loss": 0.2072, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 3.37, | |
| "grad_norm": 0.2635527551174164, | |
| "learning_rate": 0.00014305110968082952, | |
| "loss": 0.1977, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 3.56, | |
| "grad_norm": 0.2663807272911072, | |
| "learning_rate": 0.00013502073812594675, | |
| "loss": 0.188, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 3.76, | |
| "grad_norm": 0.2633569836616516, | |
| "learning_rate": 0.00012672383760782568, | |
| "loss": 0.1842, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 3.96, | |
| "grad_norm": 0.2857445478439331, | |
| "learning_rate": 0.00011822355254921478, | |
| "loss": 0.1728, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.96, | |
| "eval_loss": 0.17973944544792175, | |
| "eval_runtime": 93.8682, | |
| "eval_samples_per_second": 30.543, | |
| "eval_steps_per_second": 0.959, | |
| "step": 400 | |
| } | |
| ], | |
| "logging_steps": 20, | |
| "max_steps": 800, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 8, | |
| "save_steps": 100, | |
| "total_flos": 3.644899239558906e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |