| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9966329966329966, | |
| "eval_steps": 500, | |
| "global_step": 148, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 3.3333333333333334e-08, | |
| "logits/chosen": -2.023646593093872, | |
| "logits/rejected": -1.861999750137329, | |
| "logps/chosen": -160.15196228027344, | |
| "logps/rejected": -164.30947875976562, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 3.333333333333333e-07, | |
| "logits/chosen": -1.8246960639953613, | |
| "logits/rejected": -1.874166488647461, | |
| "logps/chosen": -186.62855529785156, | |
| "logps/rejected": -191.06869506835938, | |
| "loss": 0.693, | |
| "rewards/accuracies": 0.4513888955116272, | |
| "rewards/chosen": -0.0028373675886541605, | |
| "rewards/margins": -9.477811545366421e-06, | |
| "rewards/rejected": -0.002827889285981655, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.98258427321406e-07, | |
| "logits/chosen": -1.5834639072418213, | |
| "logits/rejected": -1.6468555927276611, | |
| "logps/chosen": -191.36276245117188, | |
| "logps/rejected": -201.68360900878906, | |
| "loss": 0.6936, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.09236270189285278, | |
| "rewards/margins": 0.01027429848909378, | |
| "rewards/rejected": -0.10263700783252716, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.844710954430464e-07, | |
| "logits/chosen": -1.572912335395813, | |
| "logits/rejected": -1.6341121196746826, | |
| "logps/chosen": -189.92250061035156, | |
| "logps/rejected": -188.98448181152344, | |
| "loss": 0.6947, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": -0.10429297387599945, | |
| "rewards/margins": -0.007742973975837231, | |
| "rewards/rejected": -0.09655000269412994, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.576621278295557e-07, | |
| "logits/chosen": -1.6369476318359375, | |
| "logits/rejected": -1.5702846050262451, | |
| "logps/chosen": -203.3102264404297, | |
| "logps/rejected": -203.18093872070312, | |
| "loss": 0.6897, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.03835677355527878, | |
| "rewards/margins": 0.012528707273304462, | |
| "rewards/rejected": -0.05088547617197037, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 4.193203929064353e-07, | |
| "logits/chosen": -1.3983080387115479, | |
| "logits/rejected": -1.405611276626587, | |
| "logps/chosen": -200.58035278320312, | |
| "logps/rejected": -205.0247802734375, | |
| "loss": 0.6879, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -0.2730618119239807, | |
| "rewards/margins": 0.03439151123166084, | |
| "rewards/rejected": -0.30745333433151245, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 3.715752452735703e-07, | |
| "logits/chosen": -1.2504160404205322, | |
| "logits/rejected": -1.3270930051803589, | |
| "logps/chosen": -236.2527313232422, | |
| "logps/rejected": -239.9370574951172, | |
| "loss": 0.6856, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": -0.5804754495620728, | |
| "rewards/margins": 0.015670539811253548, | |
| "rewards/rejected": -0.5961459279060364, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 3.170782694233712e-07, | |
| "logits/chosen": -1.1783835887908936, | |
| "logits/rejected": -1.141601800918579, | |
| "logps/chosen": -263.32073974609375, | |
| "logps/rejected": -272.93450927734375, | |
| "loss": 0.6956, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.9808699488639832, | |
| "rewards/margins": 0.05901496857404709, | |
| "rewards/rejected": -1.0398849248886108, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.588560207905135e-07, | |
| "logits/chosen": -1.2210887670516968, | |
| "logits/rejected": -1.3275476694107056, | |
| "logps/chosen": -277.7283020019531, | |
| "logps/rejected": -284.2450256347656, | |
| "loss": 0.6953, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": -0.8450358510017395, | |
| "rewards/margins": -0.01733619160950184, | |
| "rewards/rejected": -0.8276995420455933, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 2.001419423371019e-07, | |
| "logits/chosen": -1.2033240795135498, | |
| "logits/rejected": -1.2934813499450684, | |
| "logps/chosen": -233.0209197998047, | |
| "logps/rejected": -232.87161254882812, | |
| "loss": 0.6984, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -0.7631824612617493, | |
| "rewards/margins": -0.014957061037421227, | |
| "rewards/rejected": -0.7482253313064575, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.4419679138889375e-07, | |
| "logits/chosen": -1.5676350593566895, | |
| "logits/rejected": -1.5352352857589722, | |
| "logps/chosen": -243.58340454101562, | |
| "logps/rejected": -249.83670043945312, | |
| "loss": 0.6839, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.431363046169281, | |
| "rewards/margins": 0.026667874306440353, | |
| "rewards/rejected": -0.45803093910217285, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 9.412754953531663e-08, | |
| "logits/chosen": -1.4186036586761475, | |
| "logits/rejected": -1.4321409463882446, | |
| "logps/chosen": -207.4648895263672, | |
| "logps/rejected": -220.63540649414062, | |
| "loss": 0.6783, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.3499959409236908, | |
| "rewards/margins": 0.08647538721561432, | |
| "rewards/rejected": -0.4364713132381439, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 5.271487265090163e-08, | |
| "logits/chosen": -1.3899421691894531, | |
| "logits/rejected": -1.4946931600570679, | |
| "logps/chosen": -199.7452392578125, | |
| "logps/rejected": -206.2049102783203, | |
| "loss": 0.6857, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.31818634271621704, | |
| "rewards/margins": 0.03093295730650425, | |
| "rewards/rejected": -0.34911927580833435, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 2.2258663809784888e-08, | |
| "logits/chosen": -1.5326082706451416, | |
| "logits/rejected": -1.4976835250854492, | |
| "logps/chosen": -203.8050079345703, | |
| "logps/rejected": -208.89431762695312, | |
| "loss": 0.6865, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.27058011293411255, | |
| "rewards/margins": 0.03228816017508507, | |
| "rewards/rejected": -0.3028682768344879, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 4.45034538815614e-09, | |
| "logits/chosen": -1.4072165489196777, | |
| "logits/rejected": -1.4480218887329102, | |
| "logps/chosen": -210.232666015625, | |
| "logps/rejected": -218.79763793945312, | |
| "loss": 0.6827, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.29495373368263245, | |
| "rewards/margins": 0.03974684700369835, | |
| "rewards/rejected": -0.3347005844116211, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 148, | |
| "total_flos": 0.0, | |
| "train_loss": 0.6894009854342487, | |
| "train_runtime": 9025.6123, | |
| "train_samples_per_second": 2.105, | |
| "train_steps_per_second": 0.016 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 148, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "total_flos": 0.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |