| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9929622520793346, | |
| "eval_steps": 500, | |
| "global_step": 97, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05118362124120281, | |
| "grad_norm": 142.04534883137924, | |
| "learning_rate": 2.5e-07, | |
| "logits/chosen": -3.121891498565674, | |
| "logits/rejected": -3.0059380531311035, | |
| "logps/chosen": -1.436647653579712, | |
| "logps/rejected": -2.13486909866333, | |
| "loss": 0.263, | |
| "rewards/accuracies": 0.9781249761581421, | |
| "rewards/chosen": -3.5916194915771484, | |
| "rewards/margins": 1.7455530166625977, | |
| "rewards/rejected": -5.337172508239746, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.10236724248240563, | |
| "grad_norm": 8.928774478177914, | |
| "learning_rate": 5e-07, | |
| "logits/chosen": -3.1206612586975098, | |
| "logits/rejected": -3.020460844039917, | |
| "logps/chosen": -1.482598066329956, | |
| "logps/rejected": -3.0258779525756836, | |
| "loss": 0.0557, | |
| "rewards/accuracies": 0.9984375238418579, | |
| "rewards/chosen": -3.7064952850341797, | |
| "rewards/margins": 3.8581995964050293, | |
| "rewards/rejected": -7.564694881439209, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.15355086372360843, | |
| "grad_norm": 0.011898667770509092, | |
| "learning_rate": 4.959362109830006e-07, | |
| "logits/chosen": -3.10644793510437, | |
| "logits/rejected": -3.03446102142334, | |
| "logps/chosen": -1.575461983680725, | |
| "logps/rejected": -5.494174003601074, | |
| "loss": 0.0004, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -3.938655138015747, | |
| "rewards/margins": 9.796780586242676, | |
| "rewards/rejected": -13.735437393188477, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.20473448496481125, | |
| "grad_norm": 0.02098359966143572, | |
| "learning_rate": 4.838769589814002e-07, | |
| "logits/chosen": -3.0926451683044434, | |
| "logits/rejected": -3.0563385486602783, | |
| "logps/chosen": -1.6248165369033813, | |
| "logps/rejected": -6.373618125915527, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -4.062041759490967, | |
| "rewards/margins": 11.872003555297852, | |
| "rewards/rejected": -15.934043884277344, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2559181062060141, | |
| "grad_norm": 0.0019588953751018155, | |
| "learning_rate": 4.642142940418973e-07, | |
| "logits/chosen": -3.0899291038513184, | |
| "logits/rejected": -3.0629467964172363, | |
| "logps/chosen": -1.6836541891098022, | |
| "logps/rejected": -7.282121181488037, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -4.209136009216309, | |
| "rewards/margins": 13.996167182922363, | |
| "rewards/rejected": -18.205303192138672, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.30710172744721687, | |
| "grad_norm": 0.0011161418830464636, | |
| "learning_rate": 4.375874555391006e-07, | |
| "logits/chosen": -3.0745973587036133, | |
| "logits/rejected": -3.052220106124878, | |
| "logps/chosen": -1.7839086055755615, | |
| "logps/rejected": -7.939328670501709, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -4.459771633148193, | |
| "rewards/margins": 15.3885498046875, | |
| "rewards/rejected": -19.84832191467285, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3582853486884197, | |
| "grad_norm": 0.0009076724860371347, | |
| "learning_rate": 4.0486209030393077e-07, | |
| "logits/chosen": -3.0616514682769775, | |
| "logits/rejected": -3.033237934112549, | |
| "logps/chosen": -1.8522485494613647, | |
| "logps/rejected": -8.748382568359375, | |
| "loss": 0.0003, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -4.630621433258057, | |
| "rewards/margins": 17.24033546447754, | |
| "rewards/rejected": -21.87095832824707, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4094689699296225, | |
| "grad_norm": 0.00013148852405563184, | |
| "learning_rate": 3.6710211017494754e-07, | |
| "logits/chosen": -3.0289058685302734, | |
| "logits/rejected": -3.002310276031494, | |
| "logps/chosen": -1.935601830482483, | |
| "logps/rejected": -9.641812324523926, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -4.8390045166015625, | |
| "rewards/margins": 19.26552391052246, | |
| "rewards/rejected": -24.104528427124023, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.46065259117082535, | |
| "grad_norm": 7.439987461568028e-05, | |
| "learning_rate": 3.2553510389239316e-07, | |
| "logits/chosen": -3.0215375423431396, | |
| "logits/rejected": -2.998112201690674, | |
| "logps/chosen": -1.993088960647583, | |
| "logps/rejected": -10.117258071899414, | |
| "loss": 0.0003, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -4.982722282409668, | |
| "rewards/margins": 20.310420989990234, | |
| "rewards/rejected": -25.29314613342285, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5118362124120281, | |
| "grad_norm": 0.0002479185108431476, | |
| "learning_rate": 2.8151242780507236e-07, | |
| "logits/chosen": -3.010772466659546, | |
| "logits/rejected": -2.990293264389038, | |
| "logps/chosen": -2.0182385444641113, | |
| "logps/rejected": -10.393533706665039, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -5.045596122741699, | |
| "rewards/margins": 20.938236236572266, | |
| "rewards/rejected": -25.983835220336914, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5630198336532309, | |
| "grad_norm": 5.564430896058611e-05, | |
| "learning_rate": 2.3646527285364563e-07, | |
| "logits/chosen": -3.001908779144287, | |
| "logits/rejected": -2.984178066253662, | |
| "logps/chosen": -2.0269970893859863, | |
| "logps/rejected": -10.485576629638672, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -5.0674920082092285, | |
| "rewards/margins": 21.14645004272461, | |
| "rewards/rejected": -26.213939666748047, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.6142034548944337, | |
| "grad_norm": 2.3732297250695814e-05, | |
| "learning_rate": 1.9185813610642245e-07, | |
| "logits/chosen": -2.9939801692962646, | |
| "logits/rejected": -2.9763240814208984, | |
| "logps/chosen": -2.1204419136047363, | |
| "logps/rejected": -10.583490371704102, | |
| "loss": 0.0003, | |
| "rewards/accuracies": 0.9984375238418579, | |
| "rewards/chosen": -5.301104545593262, | |
| "rewards/margins": 21.157623291015625, | |
| "rewards/rejected": -26.458728790283203, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.6653870761356366, | |
| "grad_norm": 3.5153653165787e-05, | |
| "learning_rate": 1.491412095025479e-07, | |
| "logits/chosen": -3.0078186988830566, | |
| "logits/rejected": -2.990968704223633, | |
| "logps/chosen": -2.0652410984039307, | |
| "logps/rejected": -10.685462951660156, | |
| "loss": 0.0003, | |
| "rewards/accuracies": 0.9984375238418579, | |
| "rewards/chosen": -5.163102626800537, | |
| "rewards/margins": 21.550552368164062, | |
| "rewards/rejected": -26.71365737915039, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.7165706973768394, | |
| "grad_norm": 4.722931910809097e-05, | |
| "learning_rate": 1.0970323365940443e-07, | |
| "logits/chosen": -2.9940237998962402, | |
| "logits/rejected": -2.9749436378479004, | |
| "logps/chosen": -2.1152796745300293, | |
| "logps/rejected": -10.595209121704102, | |
| "loss": 0.0004, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -5.288198947906494, | |
| "rewards/margins": 21.1998233795166, | |
| "rewards/rejected": -26.488025665283203, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.7677543186180422, | |
| "grad_norm": 0.00043712476558793484, | |
| "learning_rate": 7.482634948166442e-08, | |
| "logits/chosen": -3.0042853355407715, | |
| "logits/rejected": -2.9820871353149414, | |
| "logps/chosen": -2.102086067199707, | |
| "logps/rejected": -10.324960708618164, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -5.255215644836426, | |
| "rewards/margins": 20.557186126708984, | |
| "rewards/rejected": -25.812402725219727, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.818937939859245, | |
| "grad_norm": 0.0013776814183359047, | |
| "learning_rate": 4.564441536027705e-08, | |
| "logits/chosen": -3.00022292137146, | |
| "logits/rejected": -2.9755115509033203, | |
| "logps/chosen": -2.1440300941467285, | |
| "logps/rejected": -10.0432767868042, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -5.360074520111084, | |
| "rewards/margins": 19.748117446899414, | |
| "rewards/rejected": -25.10819435119629, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.8701215611004478, | |
| "grad_norm": 0.0005507801821536249, | |
| "learning_rate": 2.3106145082260774e-08, | |
| "logits/chosen": -3.0011277198791504, | |
| "logits/rejected": -2.974482297897339, | |
| "logps/chosen": -2.1504433155059814, | |
| "logps/rejected": -10.013360977172852, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -5.376108646392822, | |
| "rewards/margins": 19.65729331970215, | |
| "rewards/rejected": -25.03339958190918, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.9213051823416507, | |
| "grad_norm": 0.00025207548670523616, | |
| "learning_rate": 7.944264849359172e-09, | |
| "logits/chosen": -3.0121123790740967, | |
| "logits/rejected": -2.981370449066162, | |
| "logps/chosen": -2.1106839179992676, | |
| "logps/rejected": -10.074134826660156, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -5.27670955657959, | |
| "rewards/margins": 19.90863037109375, | |
| "rewards/rejected": -25.185338973999023, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.9724888035828535, | |
| "grad_norm": 0.0001811770725552188, | |
| "learning_rate": 6.516921205125537e-10, | |
| "logits/chosen": -2.9971859455108643, | |
| "logits/rejected": -2.97003173828125, | |
| "logps/chosen": -2.143008232116699, | |
| "logps/rejected": -10.052621841430664, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -5.357520580291748, | |
| "rewards/margins": 19.774032592773438, | |
| "rewards/rejected": -25.131555557250977, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.9929622520793346, | |
| "step": 97, | |
| "total_flos": 0.0, | |
| "train_loss": 0.016540080016594853, | |
| "train_runtime": 11118.1076, | |
| "train_samples_per_second": 4.497, | |
| "train_steps_per_second": 0.009 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 97, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |