| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 3.4482758620689653, |
| "eval_steps": 500, |
| "global_step": 100, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.3448275862068966, |
| "grad_norm": 3.9107234477996826, |
| "learning_rate": 8.620689655172415e-07, |
| "logits/chosen": -2.363854169845581, |
| "logits/rejected": -2.338671922683716, |
| "logps/chosen": -76.8819351196289, |
| "logps/rejected": -75.86869049072266, |
| "loss": 0.6942, |
| "rewards/accuracies": 0.375, |
| "rewards/chosen": -0.003996999468654394, |
| "rewards/margins": -0.004145228303968906, |
| "rewards/rejected": 0.00014822949015069753, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.6896551724137931, |
| "grad_norm": 4.328212738037109, |
| "learning_rate": 1.724137931034483e-06, |
| "logits/chosen": -2.3617663383483887, |
| "logits/rejected": -2.3477540016174316, |
| "logps/chosen": -93.76543426513672, |
| "logps/rejected": -75.16656494140625, |
| "loss": 0.6931, |
| "rewards/accuracies": 0.48750001192092896, |
| "rewards/chosen": 0.002653368515893817, |
| "rewards/margins": 0.006352785974740982, |
| "rewards/rejected": -0.003699416993185878, |
| "step": 20 |
| }, |
| { |
| "epoch": 1.0344827586206897, |
| "grad_norm": 3.473827838897705, |
| "learning_rate": 2.5862068965517246e-06, |
| "logits/chosen": -2.3354058265686035, |
| "logits/rejected": -2.3330740928649902, |
| "logps/chosen": -138.9960479736328, |
| "logps/rejected": -67.00438690185547, |
| "loss": 0.6927, |
| "rewards/accuracies": 0.44999998807907104, |
| "rewards/chosen": 0.01745142787694931, |
| "rewards/margins": 0.013760591857135296, |
| "rewards/rejected": 0.0036908381152898073, |
| "step": 30 |
| }, |
| { |
| "epoch": 1.3793103448275863, |
| "grad_norm": 3.9458820819854736, |
| "learning_rate": 3.448275862068966e-06, |
| "logits/chosen": -2.3478751182556152, |
| "logits/rejected": -2.3700289726257324, |
| "logps/chosen": -75.59983825683594, |
| "logps/rejected": -67.20357513427734, |
| "loss": 0.69, |
| "rewards/accuracies": 0.5, |
| "rewards/chosen": 0.00024204826331697404, |
| "rewards/margins": 0.005866119172424078, |
| "rewards/rejected": -0.005624071694910526, |
| "step": 40 |
| }, |
| { |
| "epoch": 1.7241379310344827, |
| "grad_norm": 4.299890041351318, |
| "learning_rate": 4.310344827586207e-06, |
| "logits/chosen": -2.338815212249756, |
| "logits/rejected": -2.3532769680023193, |
| "logps/chosen": -72.74342346191406, |
| "logps/rejected": -87.63409423828125, |
| "loss": 0.6899, |
| "rewards/accuracies": 0.574999988079071, |
| "rewards/chosen": -0.003249581903219223, |
| "rewards/margins": 0.006888723932206631, |
| "rewards/rejected": -0.010138307698071003, |
| "step": 50 |
| }, |
| { |
| "epoch": 2.0689655172413794, |
| "grad_norm": 4.957555294036865, |
| "learning_rate": 4.999818897894192e-06, |
| "logits/chosen": -2.344572067260742, |
| "logits/rejected": -2.345888614654541, |
| "logps/chosen": -80.16046142578125, |
| "logps/rejected": -73.05986022949219, |
| "loss": 0.6802, |
| "rewards/accuracies": 0.637499988079071, |
| "rewards/chosen": -0.013288321904838085, |
| "rewards/margins": 0.020968889817595482, |
| "rewards/rejected": -0.03425721079111099, |
| "step": 60 |
| }, |
| { |
| "epoch": 2.413793103448276, |
| "grad_norm": 4.212078094482422, |
| "learning_rate": 4.9934830787948756e-06, |
| "logits/chosen": -2.3330090045928955, |
| "logits/rejected": -2.328575372695923, |
| "logps/chosen": -73.63365173339844, |
| "logps/rejected": -74.45356750488281, |
| "loss": 0.6693, |
| "rewards/accuracies": 0.7124999761581421, |
| "rewards/chosen": -0.009874681942164898, |
| "rewards/margins": 0.04785541445016861, |
| "rewards/rejected": -0.05773010104894638, |
| "step": 70 |
| }, |
| { |
| "epoch": 2.7586206896551726, |
| "grad_norm": 4.446742534637451, |
| "learning_rate": 4.978118375700895e-06, |
| "logits/chosen": -2.3472495079040527, |
| "logits/rejected": -2.3697409629821777, |
| "logps/chosen": -73.44490814208984, |
| "logps/rejected": -89.69837951660156, |
| "loss": 0.6553, |
| "rewards/accuracies": 0.8500000238418579, |
| "rewards/chosen": -0.022817375138401985, |
| "rewards/margins": 0.09045850485563278, |
| "rewards/rejected": -0.11327588558197021, |
| "step": 80 |
| }, |
| { |
| "epoch": 3.103448275862069, |
| "grad_norm": 4.435715198516846, |
| "learning_rate": 4.953780424089803e-06, |
| "logits/chosen": -2.346717596054077, |
| "logits/rejected": -2.3626608848571777, |
| "logps/chosen": -85.64796447753906, |
| "logps/rejected": -80.06268310546875, |
| "loss": 0.6432, |
| "rewards/accuracies": 0.737500011920929, |
| "rewards/chosen": -0.0543200746178627, |
| "rewards/margins": 0.1085677295923233, |
| "rewards/rejected": -0.1628878116607666, |
| "step": 90 |
| }, |
| { |
| "epoch": 3.4482758620689653, |
| "grad_norm": 4.5200910568237305, |
| "learning_rate": 4.920557351506409e-06, |
| "logits/chosen": -2.358900547027588, |
| "logits/rejected": -2.3591980934143066, |
| "logps/chosen": -82.83473205566406, |
| "logps/rejected": -80.66204071044922, |
| "loss": 0.5996, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": -0.046993009746074677, |
| "rewards/margins": 0.20401433110237122, |
| "rewards/rejected": -0.2510073184967041, |
| "step": 100 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 580, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 20, |
| "save_steps": 100, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 3.6446446725523046e+17, |
| "train_batch_size": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
