| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9182209469153515, | |
| "eval_steps": 500, | |
| "global_step": 10, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.09182209469153516, | |
| "grad_norm": 0.5866674780845642, | |
| "learning_rate": 5e-07, | |
| "logits/chosen": 0.11353569477796555, | |
| "logits/rejected": 0.26297590136528015, | |
| "logps/chosen": -131.12338256835938, | |
| "logps/ref_chosen": -127.41575622558594, | |
| "logps/ref_rejected": -163.67575073242188, | |
| "logps/rejected": -167.3235626220703, | |
| "loss": 0.6935, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": -0.037076327949762344, | |
| "rewards/margins": -0.0005980312125757337, | |
| "rewards/rejected": -0.036478299647569656, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.1836441893830703, | |
| "grad_norm": 0.41607558727264404, | |
| "learning_rate": 4.849231551964771e-07, | |
| "logits/chosen": 0.08786322176456451, | |
| "logits/rejected": 0.31397393345832825, | |
| "logps/chosen": -128.80751037597656, | |
| "logps/ref_chosen": -125.11693572998047, | |
| "logps/ref_rejected": -166.12916564941406, | |
| "logps/rejected": -169.7572784423828, | |
| "loss": 0.6935, | |
| "rewards/accuracies": 0.4609375, | |
| "rewards/chosen": -0.03690561652183533, | |
| "rewards/margins": -0.0006243533571250737, | |
| "rewards/rejected": -0.03628126531839371, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.27546628407460544, | |
| "grad_norm": 0.2935572862625122, | |
| "learning_rate": 4.415111107797445e-07, | |
| "logits/chosen": 0.16991552710533142, | |
| "logits/rejected": 0.2849712371826172, | |
| "logps/chosen": -132.75396728515625, | |
| "logps/ref_chosen": -129.03077697753906, | |
| "logps/ref_rejected": -160.53233337402344, | |
| "logps/rejected": -164.21923828125, | |
| "loss": 0.6933, | |
| "rewards/accuracies": 0.4765625, | |
| "rewards/chosen": -0.03723179176449776, | |
| "rewards/margins": -0.0003625184472184628, | |
| "rewards/rejected": -0.036869268864393234, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.3672883787661406, | |
| "grad_norm": 0.37583044171333313, | |
| "learning_rate": 3.75e-07, | |
| "logits/chosen": 0.05217592790722847, | |
| "logits/rejected": 0.25611257553100586, | |
| "logps/chosen": -133.43521118164062, | |
| "logps/ref_chosen": -129.90614318847656, | |
| "logps/ref_rejected": -176.55445861816406, | |
| "logps/rejected": -179.9996795654297, | |
| "loss": 0.6936, | |
| "rewards/accuracies": 0.484375, | |
| "rewards/chosen": -0.0352906659245491, | |
| "rewards/margins": -0.0008385752444155514, | |
| "rewards/rejected": -0.03445209190249443, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.45911047345767575, | |
| "grad_norm": 0.3961547017097473, | |
| "learning_rate": 2.934120444167326e-07, | |
| "logits/chosen": 0.06961281597614288, | |
| "logits/rejected": 0.2353174239397049, | |
| "logps/chosen": -135.94540405273438, | |
| "logps/ref_chosen": -132.38174438476562, | |
| "logps/ref_rejected": -172.46087646484375, | |
| "logps/rejected": -175.92135620117188, | |
| "loss": 0.6937, | |
| "rewards/accuracies": 0.4296875, | |
| "rewards/chosen": -0.03563670068979263, | |
| "rewards/margins": -0.0010319410357624292, | |
| "rewards/rejected": -0.034604761749506, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.5509325681492109, | |
| "grad_norm": 0.38980162143707275, | |
| "learning_rate": 2.065879555832674e-07, | |
| "logits/chosen": 0.12377427518367767, | |
| "logits/rejected": 0.2602069675922394, | |
| "logps/chosen": -133.6548614501953, | |
| "logps/ref_chosen": -130.05429077148438, | |
| "logps/ref_rejected": -162.71273803710938, | |
| "logps/rejected": -166.23013305664062, | |
| "loss": 0.6936, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": -0.036005713045597076, | |
| "rewards/margins": -0.0008315728046000004, | |
| "rewards/rejected": -0.035174138844013214, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.642754662840746, | |
| "grad_norm": 0.3176744282245636, | |
| "learning_rate": 1.2500000000000005e-07, | |
| "logits/chosen": 0.06976456940174103, | |
| "logits/rejected": 0.22669002413749695, | |
| "logps/chosen": -129.18087768554688, | |
| "logps/ref_chosen": -125.5029296875, | |
| "logps/ref_rejected": -164.9804229736328, | |
| "logps/rejected": -168.5306396484375, | |
| "loss": 0.6938, | |
| "rewards/accuracies": 0.3828125, | |
| "rewards/chosen": -0.036779459565877914, | |
| "rewards/margins": -0.0012771700276061893, | |
| "rewards/rejected": -0.03550228849053383, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.7345767575322812, | |
| "grad_norm": 0.32249295711517334, | |
| "learning_rate": 5.848888922025552e-08, | |
| "logits/chosen": 0.21500277519226074, | |
| "logits/rejected": 0.3311355412006378, | |
| "logps/chosen": -136.3115997314453, | |
| "logps/ref_chosen": -132.56710815429688, | |
| "logps/ref_rejected": -161.65798950195312, | |
| "logps/rejected": -165.33477783203125, | |
| "loss": 0.6935, | |
| "rewards/accuracies": 0.4765625, | |
| "rewards/chosen": -0.037444986402988434, | |
| "rewards/margins": -0.0006772614433430135, | |
| "rewards/rejected": -0.03676772862672806, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.8263988522238164, | |
| "grad_norm": 0.40930500626564026, | |
| "learning_rate": 1.507684480352292e-08, | |
| "logits/chosen": 0.12441365420818329, | |
| "logits/rejected": 0.2915050983428955, | |
| "logps/chosen": -133.42391967773438, | |
| "logps/ref_chosen": -129.88262939453125, | |
| "logps/ref_rejected": -163.88711547851562, | |
| "logps/rejected": -167.4432830810547, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.03541266918182373, | |
| "rewards/margins": 0.0001489966962253675, | |
| "rewards/rejected": -0.03556166589260101, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.9182209469153515, | |
| "grad_norm": 0.3752247095108032, | |
| "learning_rate": 0.0, | |
| "logits/chosen": 0.0854257270693779, | |
| "logits/rejected": 0.24119172990322113, | |
| "logps/chosen": -133.70213317871094, | |
| "logps/ref_chosen": -130.187744140625, | |
| "logps/ref_rejected": -169.77410888671875, | |
| "logps/rejected": -173.26141357421875, | |
| "loss": 0.6933, | |
| "rewards/accuracies": 0.4921875, | |
| "rewards/chosen": -0.03514372929930687, | |
| "rewards/margins": -0.00027063535526394844, | |
| "rewards/rejected": -0.03487309440970421, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.9182209469153515, | |
| "step": 10, | |
| "total_flos": 0.0, | |
| "train_loss": 0.6934699952602387, | |
| "train_runtime": 619.8127, | |
| "train_samples_per_second": 2.247, | |
| "train_steps_per_second": 0.016 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 10, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |