| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9879931389365352, | |
| "eval_steps": 500, | |
| "global_step": 72, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0137221269296741, | |
| "grad_norm": 23.312597274780273, | |
| "learning_rate": 0.0, | |
| "logits/chosen": -3.0826492309570312, | |
| "logits/rejected": -3.1289501190185547, | |
| "logps/chosen": -224.05235290527344, | |
| "logps/rejected": -239.40707397460938, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0686106346483705, | |
| "grad_norm": 27.118257522583008, | |
| "learning_rate": 2.5e-07, | |
| "logits/chosen": -2.9749486446380615, | |
| "logits/rejected": -3.077843427658081, | |
| "logps/chosen": -187.49241638183594, | |
| "logps/rejected": -283.5939025878906, | |
| "loss": 0.6834, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.0007624387508258224, | |
| "rewards/margins": 0.020107731223106384, | |
| "rewards/rejected": -0.02087016962468624, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.137221269296741, | |
| "grad_norm": 21.591798782348633, | |
| "learning_rate": 4.996988640512931e-07, | |
| "logits/chosen": -2.9997000694274902, | |
| "logits/rejected": -3.1022396087646484, | |
| "logps/chosen": -200.90689086914062, | |
| "logps/rejected": -292.61444091796875, | |
| "loss": 0.5854, | |
| "rewards/accuracies": 0.940625011920929, | |
| "rewards/chosen": -0.014221695251762867, | |
| "rewards/margins": 0.24389581382274628, | |
| "rewards/rejected": -0.2581174969673157, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.2058319039451115, | |
| "grad_norm": 12.920186996459961, | |
| "learning_rate": 4.892350839330522e-07, | |
| "logits/chosen": -2.883420944213867, | |
| "logits/rejected": -2.9923837184906006, | |
| "logps/chosen": -175.34622192382812, | |
| "logps/rejected": -294.7455749511719, | |
| "loss": 0.4103, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": -0.10586054623126984, | |
| "rewards/margins": 0.8374508619308472, | |
| "rewards/rejected": -0.943311333656311, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.274442538593482, | |
| "grad_norm": 14.314050674438477, | |
| "learning_rate": 4.64432152500068e-07, | |
| "logits/chosen": -2.740008592605591, | |
| "logits/rejected": -2.9438371658325195, | |
| "logps/chosen": -192.83287048339844, | |
| "logps/rejected": -330.67755126953125, | |
| "loss": 0.2578, | |
| "rewards/accuracies": 0.9593750238418579, | |
| "rewards/chosen": -0.345708966255188, | |
| "rewards/margins": 1.699308156967163, | |
| "rewards/rejected": -2.0450172424316406, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.34305317324185247, | |
| "grad_norm": 10.350852012634277, | |
| "learning_rate": 4.2677669529663686e-07, | |
| "logits/chosen": -2.7626330852508545, | |
| "logits/rejected": -2.9129269123077393, | |
| "logps/chosen": -187.16873168945312, | |
| "logps/rejected": -338.9030456542969, | |
| "loss": 0.2415, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -0.5644805431365967, | |
| "rewards/margins": 1.97623610496521, | |
| "rewards/rejected": -2.5407166481018066, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.411663807890223, | |
| "grad_norm": 12.666106224060059, | |
| "learning_rate": 3.7852568604830535e-07, | |
| "logits/chosen": -2.6965413093566895, | |
| "logits/rejected": -2.8506977558135986, | |
| "logps/chosen": -190.6026611328125, | |
| "logps/rejected": -337.8497009277344, | |
| "loss": 0.2079, | |
| "rewards/accuracies": 0.953125, | |
| "rewards/chosen": -0.7271584272384644, | |
| "rewards/margins": 2.3706743717193604, | |
| "rewards/rejected": -3.0978329181671143, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.48027444253859347, | |
| "grad_norm": 7.95162296295166, | |
| "learning_rate": 3.2257116931361555e-07, | |
| "logits/chosen": -2.6175200939178467, | |
| "logits/rejected": -2.8301939964294434, | |
| "logps/chosen": -177.00990295410156, | |
| "logps/rejected": -325.2242736816406, | |
| "loss": 0.1944, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -0.6376256346702576, | |
| "rewards/margins": 2.4539847373962402, | |
| "rewards/rejected": -3.0916104316711426, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.548885077186964, | |
| "grad_norm": 12.024945259094238, | |
| "learning_rate": 2.6226691858185454e-07, | |
| "logits/chosen": -2.6264350414276123, | |
| "logits/rejected": -2.80271315574646, | |
| "logps/chosen": -157.8688507080078, | |
| "logps/rejected": -315.71124267578125, | |
| "loss": 0.1893, | |
| "rewards/accuracies": 0.953125, | |
| "rewards/chosen": -0.7045091986656189, | |
| "rewards/margins": 2.5911941528320312, | |
| "rewards/rejected": -3.295703411102295, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.6174957118353345, | |
| "grad_norm": 12.804183006286621, | |
| "learning_rate": 2.0122741949596793e-07, | |
| "logits/chosen": -2.5655877590179443, | |
| "logits/rejected": -2.7770166397094727, | |
| "logps/chosen": -186.1898956298828, | |
| "logps/rejected": -350.7716979980469, | |
| "loss": 0.1893, | |
| "rewards/accuracies": 0.953125, | |
| "rewards/chosen": -0.8316699266433716, | |
| "rewards/margins": 2.6264994144439697, | |
| "rewards/rejected": -3.4581692218780518, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.6861063464837049, | |
| "grad_norm": 8.851630210876465, | |
| "learning_rate": 1.4311122664242953e-07, | |
| "logits/chosen": -2.564091205596924, | |
| "logits/rejected": -2.7729170322418213, | |
| "logps/chosen": -173.6750030517578, | |
| "logps/rejected": -336.9339294433594, | |
| "loss": 0.1599, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.8224509358406067, | |
| "rewards/margins": 2.8290677070617676, | |
| "rewards/rejected": -3.6515185832977295, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.7547169811320755, | |
| "grad_norm": 11.309371948242188, | |
| "learning_rate": 9.140167895908865e-08, | |
| "logits/chosen": -2.5938456058502197, | |
| "logits/rejected": -2.785677194595337, | |
| "logps/chosen": -195.90493774414062, | |
| "logps/rejected": -358.4869384765625, | |
| "loss": 0.176, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -0.9772733449935913, | |
| "rewards/margins": 2.8666346073150635, | |
| "rewards/rejected": -3.8439078330993652, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.823327615780446, | |
| "grad_norm": 11.698320388793945, | |
| "learning_rate": 4.919811712983879e-08, | |
| "logits/chosen": -2.6009421348571777, | |
| "logits/rejected": -2.8167078495025635, | |
| "logps/chosen": -196.6665802001953, | |
| "logps/rejected": -381.7016296386719, | |
| "loss": 0.163, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -0.9723421931266785, | |
| "rewards/margins": 3.029435873031616, | |
| "rewards/rejected": -4.001777648925781, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.8919382504288165, | |
| "grad_norm": 10.755302429199219, | |
| "learning_rate": 1.9030116872178314e-08, | |
| "logits/chosen": -2.5659337043762207, | |
| "logits/rejected": -2.7652087211608887, | |
| "logps/chosen": -163.1660614013672, | |
| "logps/rejected": -347.26629638671875, | |
| "loss": 0.1707, | |
| "rewards/accuracies": 0.971875011920929, | |
| "rewards/chosen": -0.8446305990219116, | |
| "rewards/margins": 2.903198003768921, | |
| "rewards/rejected": -3.747828722000122, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.9605488850771869, | |
| "grad_norm": 11.380764961242676, | |
| "learning_rate": 2.7058725088047464e-09, | |
| "logits/chosen": -2.54191255569458, | |
| "logits/rejected": -2.7878177165985107, | |
| "logps/chosen": -196.59622192382812, | |
| "logps/rejected": -356.73626708984375, | |
| "loss": 0.1788, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -0.9953180551528931, | |
| "rewards/margins": 2.8245952129364014, | |
| "rewards/rejected": -3.819913148880005, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9879931389365352, | |
| "step": 72, | |
| "total_flos": 0.0, | |
| "train_loss": 0.26905302993125385, | |
| "train_runtime": 954.9092, | |
| "train_samples_per_second": 4.882, | |
| "train_steps_per_second": 0.075 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 72, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "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": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |