| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.15808974886784685, | |
| "eval_steps": 500, | |
| "global_step": 12, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.013174145738987238, | |
| "grad_norm": 0.5080101490020752, | |
| "learning_rate": 1.25e-07, | |
| "logits/chosen": 10.32492733001709, | |
| "logits/rejected": 10.282785415649414, | |
| "logps/chosen": -1.0583709478378296, | |
| "logps/rejected": -1.1253927946090698, | |
| "loss": 1.3887, | |
| "rewards/accuracies": 0.5390625, | |
| "rewards/chosen": -2.116741895675659, | |
| "rewards/margins": 0.1340436041355133, | |
| "rewards/rejected": -2.2507855892181396, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.026348291477974475, | |
| "grad_norm": 0.5300341844558716, | |
| "learning_rate": 2.5e-07, | |
| "logits/chosen": 10.530074119567871, | |
| "logits/rejected": 10.672085762023926, | |
| "logps/chosen": -1.1093882322311401, | |
| "logps/rejected": -1.2041704654693604, | |
| "loss": 1.3086, | |
| "rewards/accuracies": 0.6171875, | |
| "rewards/chosen": -2.2187764644622803, | |
| "rewards/margins": 0.18956437706947327, | |
| "rewards/rejected": -2.4083409309387207, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03952243721696171, | |
| "grad_norm": 1.1035377979278564, | |
| "learning_rate": 3.75e-07, | |
| "logits/chosen": 10.240287780761719, | |
| "logits/rejected": 10.40180492401123, | |
| "logps/chosen": -1.1902998685836792, | |
| "logps/rejected": -1.311903715133667, | |
| "loss": 1.3245, | |
| "rewards/accuracies": 0.6484375, | |
| "rewards/chosen": -2.3805997371673584, | |
| "rewards/margins": 0.24320749938488007, | |
| "rewards/rejected": -2.623807430267334, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.05269658295594895, | |
| "grad_norm": 1.8411834239959717, | |
| "learning_rate": 5e-07, | |
| "logits/chosen": 10.358713150024414, | |
| "logits/rejected": 10.405998229980469, | |
| "logps/chosen": -1.0821034908294678, | |
| "logps/rejected": -1.1429979801177979, | |
| "loss": 1.3768, | |
| "rewards/accuracies": 0.578125, | |
| "rewards/chosen": -2.1642069816589355, | |
| "rewards/margins": 0.12178920209407806, | |
| "rewards/rejected": -2.2859959602355957, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.06587072869493618, | |
| "grad_norm": 0.5591928362846375, | |
| "learning_rate": 6.249999999999999e-07, | |
| "logits/chosen": 10.748247146606445, | |
| "logits/rejected": 10.783531188964844, | |
| "logps/chosen": -1.1161627769470215, | |
| "logps/rejected": -1.1165733337402344, | |
| "loss": 1.4573, | |
| "rewards/accuracies": 0.5859375, | |
| "rewards/chosen": -2.232325553894043, | |
| "rewards/margins": 0.0008210651576519012, | |
| "rewards/rejected": -2.2331466674804688, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.07904487443392343, | |
| "grad_norm": 1.0748578310012817, | |
| "learning_rate": 7.5e-07, | |
| "logits/chosen": 10.895101547241211, | |
| "logits/rejected": 10.935002326965332, | |
| "logps/chosen": -1.051721453666687, | |
| "logps/rejected": -1.1617565155029297, | |
| "loss": 1.3341, | |
| "rewards/accuracies": 0.609375, | |
| "rewards/chosen": -2.103442907333374, | |
| "rewards/margins": 0.22007031738758087, | |
| "rewards/rejected": -2.3235130310058594, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.09221902017291066, | |
| "grad_norm": 1.2474150657653809, | |
| "learning_rate": 8.75e-07, | |
| "logits/chosen": 10.060023307800293, | |
| "logits/rejected": 10.135562896728516, | |
| "logps/chosen": -1.0141205787658691, | |
| "logps/rejected": -1.1586568355560303, | |
| "loss": 1.3246, | |
| "rewards/accuracies": 0.6328125, | |
| "rewards/chosen": -2.0282411575317383, | |
| "rewards/margins": 0.28907278180122375, | |
| "rewards/rejected": -2.3173136711120605, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1053931659118979, | |
| "grad_norm": 1.7001811265945435, | |
| "learning_rate": 1e-06, | |
| "logits/chosen": 10.837862014770508, | |
| "logits/rejected": 10.750317573547363, | |
| "logps/chosen": -1.0156888961791992, | |
| "logps/rejected": -1.1156163215637207, | |
| "loss": 1.314, | |
| "rewards/accuracies": 0.671875, | |
| "rewards/chosen": -2.0313777923583984, | |
| "rewards/margins": 0.19985482096672058, | |
| "rewards/rejected": -2.2312326431274414, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.11856731165088513, | |
| "grad_norm": 1.0888780355453491, | |
| "learning_rate": 9.994504457428556e-07, | |
| "logits/chosen": 10.434420585632324, | |
| "logits/rejected": 10.604791641235352, | |
| "logps/chosen": -1.1994280815124512, | |
| "logps/rejected": -1.1771385669708252, | |
| "loss": 1.4926, | |
| "rewards/accuracies": 0.5234375, | |
| "rewards/chosen": -2.3988561630249023, | |
| "rewards/margins": -0.04457877576351166, | |
| "rewards/rejected": -2.3542771339416504, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.13174145738987236, | |
| "grad_norm": 0.9426828622817993, | |
| "learning_rate": 9.97802991010949e-07, | |
| "logits/chosen": 10.330013275146484, | |
| "logits/rejected": 10.423635482788086, | |
| "logps/chosen": -1.1276183128356934, | |
| "logps/rejected": -1.1535909175872803, | |
| "loss": 1.4741, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -2.2552366256713867, | |
| "rewards/margins": 0.05194506794214249, | |
| "rewards/rejected": -2.3071818351745605, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.14491560312885962, | |
| "grad_norm": 1.174481987953186, | |
| "learning_rate": 9.950612572673255e-07, | |
| "logits/chosen": 10.557825088500977, | |
| "logits/rejected": 10.500927925109863, | |
| "logps/chosen": -1.2195600271224976, | |
| "logps/rejected": -1.3650047779083252, | |
| "loss": 1.3192, | |
| "rewards/accuracies": 0.640625, | |
| "rewards/chosen": -2.439120054244995, | |
| "rewards/margins": 0.2908894717693329, | |
| "rewards/rejected": -2.7300095558166504, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.15808974886784685, | |
| "grad_norm": 0.9938077330589294, | |
| "learning_rate": 9.912312714377879e-07, | |
| "logits/chosen": 10.221735000610352, | |
| "logits/rejected": 10.339948654174805, | |
| "logps/chosen": -1.072077989578247, | |
| "logps/rejected": -1.114700198173523, | |
| "loss": 1.3846, | |
| "rewards/accuracies": 0.6015625, | |
| "rewards/chosen": -2.144155979156494, | |
| "rewards/margins": 0.08524461090564728, | |
| "rewards/rejected": -2.229400396347046, | |
| "step": 12 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 75, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 12, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |