| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.3161794977356937, | |
| "eval_steps": 500, | |
| "global_step": 24, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.013174145738987238, | |
| "grad_norm": 0.3937181234359741, | |
| "learning_rate": 6.25e-08, | |
| "logits/chosen": 10.038352012634277, | |
| "logits/rejected": 10.592904090881348, | |
| "logps/chosen": -132.56948852539062, | |
| "logps/ref_chosen": -132.56948852539062, | |
| "logps/ref_rejected": -148.778076171875, | |
| "logps/rejected": -148.778076171875, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.026348291477974475, | |
| "grad_norm": 0.7554967999458313, | |
| "learning_rate": 1.25e-07, | |
| "logits/chosen": 10.320584297180176, | |
| "logits/rejected": 10.721946716308594, | |
| "logps/chosen": -132.46385192871094, | |
| "logps/ref_chosen": -132.46385192871094, | |
| "logps/ref_rejected": -142.69049072265625, | |
| "logps/rejected": -142.69049072265625, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03952243721696171, | |
| "grad_norm": 0.5059730410575867, | |
| "learning_rate": 1.875e-07, | |
| "logits/chosen": 9.891550064086914, | |
| "logits/rejected": 10.499095916748047, | |
| "logps/chosen": -136.52415466308594, | |
| "logps/ref_chosen": -135.79869079589844, | |
| "logps/ref_rejected": -145.69479370117188, | |
| "logps/rejected": -146.24124145507812, | |
| "loss": 0.6941, | |
| "rewards/accuracies": 0.3984375, | |
| "rewards/chosen": -0.007254539057612419, | |
| "rewards/margins": -0.0017902727704495192, | |
| "rewards/rejected": -0.005464266519993544, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.05269658295594895, | |
| "grad_norm": 0.6780113577842712, | |
| "learning_rate": 2.5e-07, | |
| "logits/chosen": 10.122194290161133, | |
| "logits/rejected": 10.64792251586914, | |
| "logps/chosen": -137.41903686523438, | |
| "logps/ref_chosen": -136.85159301757812, | |
| "logps/ref_rejected": -142.64901733398438, | |
| "logps/rejected": -142.94912719726562, | |
| "loss": 0.6945, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": -0.00567453121766448, | |
| "rewards/margins": -0.0026732482947409153, | |
| "rewards/rejected": -0.0030012831557542086, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.06587072869493618, | |
| "grad_norm": 0.5103796124458313, | |
| "learning_rate": 3.1249999999999997e-07, | |
| "logits/chosen": 10.327558517456055, | |
| "logits/rejected": 11.079102516174316, | |
| "logps/chosen": -130.6144561767578, | |
| "logps/ref_chosen": -130.2961883544922, | |
| "logps/ref_rejected": -144.1498565673828, | |
| "logps/rejected": -144.40802001953125, | |
| "loss": 0.6935, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": -0.0031829699873924255, | |
| "rewards/margins": -0.000601244973950088, | |
| "rewards/rejected": -0.0025817248970270157, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.07904487443392343, | |
| "grad_norm": 0.7068921327590942, | |
| "learning_rate": 3.75e-07, | |
| "logits/chosen": 10.794892311096191, | |
| "logits/rejected": 11.326902389526367, | |
| "logps/chosen": -138.65377807617188, | |
| "logps/ref_chosen": -138.59861755371094, | |
| "logps/ref_rejected": -151.63775634765625, | |
| "logps/rejected": -151.48056030273438, | |
| "loss": 0.6942, | |
| "rewards/accuracies": 0.4765625, | |
| "rewards/chosen": -0.0005516067612916231, | |
| "rewards/margins": -0.002123636193573475, | |
| "rewards/rejected": 0.0015720291994512081, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.09221902017291066, | |
| "grad_norm": 0.4639431834220886, | |
| "learning_rate": 4.375e-07, | |
| "logits/chosen": 9.949363708496094, | |
| "logits/rejected": 10.44249153137207, | |
| "logps/chosen": -130.56231689453125, | |
| "logps/ref_chosen": -129.5895538330078, | |
| "logps/ref_rejected": -139.67816162109375, | |
| "logps/rejected": -140.58245849609375, | |
| "loss": 0.6935, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.009727789089083672, | |
| "rewards/margins": -0.000684772094246, | |
| "rewards/rejected": -0.009043017402291298, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1053931659118979, | |
| "grad_norm": 0.5582533478736877, | |
| "learning_rate": 5e-07, | |
| "logits/chosen": 10.655470848083496, | |
| "logits/rejected": 11.168412208557129, | |
| "logps/chosen": -134.7186737060547, | |
| "logps/ref_chosen": -134.24497985839844, | |
| "logps/ref_rejected": -147.19009399414062, | |
| "logps/rejected": -147.70596313476562, | |
| "loss": 0.693, | |
| "rewards/accuracies": 0.5078125, | |
| "rewards/chosen": -0.0047368621453642845, | |
| "rewards/margins": 0.0004215884255245328, | |
| "rewards/rejected": -0.0051584504544734955, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.11856731165088513, | |
| "grad_norm": 0.559306263923645, | |
| "learning_rate": 4.997252228714278e-07, | |
| "logits/chosen": 10.525918006896973, | |
| "logits/rejected": 10.821675300598145, | |
| "logps/chosen": -142.31239318847656, | |
| "logps/ref_chosen": -141.59860229492188, | |
| "logps/ref_rejected": -149.33889770507812, | |
| "logps/rejected": -149.89822387695312, | |
| "loss": 0.6939, | |
| "rewards/accuracies": 0.4140625, | |
| "rewards/chosen": -0.007138015236705542, | |
| "rewards/margins": -0.001544784870930016, | |
| "rewards/rejected": -0.005593230482190847, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.13174145738987236, | |
| "grad_norm": 0.6123551726341248, | |
| "learning_rate": 4.989014955054745e-07, | |
| "logits/chosen": 10.101277351379395, | |
| "logits/rejected": 10.55870246887207, | |
| "logps/chosen": -133.06532287597656, | |
| "logps/ref_chosen": -132.13449096679688, | |
| "logps/ref_rejected": -142.75973510742188, | |
| "logps/rejected": -143.90049743652344, | |
| "loss": 0.6921, | |
| "rewards/accuracies": 0.5390625, | |
| "rewards/chosen": -0.00930829904973507, | |
| "rewards/margins": 0.002099357545375824, | |
| "rewards/rejected": -0.011407657526433468, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.14491560312885962, | |
| "grad_norm": 0.6729086637496948, | |
| "learning_rate": 4.975306286336627e-07, | |
| "logits/chosen": 10.170294761657715, | |
| "logits/rejected": 10.829442024230957, | |
| "logps/chosen": -137.3282012939453, | |
| "logps/ref_chosen": -135.85043334960938, | |
| "logps/ref_rejected": -152.14999389648438, | |
| "logps/rejected": -153.4501953125, | |
| "loss": 0.6941, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": -0.014777721837162971, | |
| "rewards/margins": -0.0017757893074303865, | |
| "rewards/rejected": -0.013001934625208378, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.15808974886784685, | |
| "grad_norm": 0.3486435115337372, | |
| "learning_rate": 4.956156357188939e-07, | |
| "logits/chosen": 10.261972427368164, | |
| "logits/rejected": 10.63994312286377, | |
| "logps/chosen": -141.65684509277344, | |
| "logps/ref_chosen": -139.63204956054688, | |
| "logps/ref_rejected": -148.03170776367188, | |
| "logps/rejected": -150.0493621826172, | |
| "loss": 0.6932, | |
| "rewards/accuracies": 0.484375, | |
| "rewards/chosen": -0.02024805173277855, | |
| "rewards/margins": -7.15737696737051e-05, | |
| "rewards/rejected": -0.0201764777302742, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.17126389460683408, | |
| "grad_norm": 0.5037537813186646, | |
| "learning_rate": 4.931607263312032e-07, | |
| "logits/chosen": 9.811729431152344, | |
| "logits/rejected": 10.610340118408203, | |
| "logps/chosen": -131.98635864257812, | |
| "logps/ref_chosen": -129.56983947753906, | |
| "logps/ref_rejected": -143.01284790039062, | |
| "logps/rejected": -145.4197998046875, | |
| "loss": 0.6932, | |
| "rewards/accuracies": 0.5390625, | |
| "rewards/chosen": -0.024165067821741104, | |
| "rewards/margins": -9.549467358738184e-05, | |
| "rewards/rejected": -0.024069571867585182, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1844380403458213, | |
| "grad_norm": 1.0056744813919067, | |
| "learning_rate": 4.9017129689421e-07, | |
| "logits/chosen": 10.571147918701172, | |
| "logits/rejected": 11.290738105773926, | |
| "logps/chosen": -132.96786499023438, | |
| "logps/ref_chosen": -130.44677734375, | |
| "logps/ref_rejected": -142.0145721435547, | |
| "logps/rejected": -144.57803344726562, | |
| "loss": 0.693, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.025210827589035034, | |
| "rewards/margins": 0.0004238423425704241, | |
| "rewards/rejected": -0.02563466690480709, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.19761218608480857, | |
| "grad_norm": 0.44761258363723755, | |
| "learning_rate": 4.866539188226085e-07, | |
| "logits/chosen": 9.930328369140625, | |
| "logits/rejected": 10.632352828979492, | |
| "logps/chosen": -137.48495483398438, | |
| "logps/ref_chosen": -133.87620544433594, | |
| "logps/ref_rejected": -146.4284210205078, | |
| "logps/rejected": -150.177490234375, | |
| "loss": 0.6925, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.036087535321712494, | |
| "rewards/margins": 0.0014031402533873916, | |
| "rewards/rejected": -0.03749067336320877, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2107863318237958, | |
| "grad_norm": 0.7480736970901489, | |
| "learning_rate": 4.826163240767716e-07, | |
| "logits/chosen": 10.873997688293457, | |
| "logits/rejected": 11.219457626342773, | |
| "logps/chosen": -138.66485595703125, | |
| "logps/ref_chosen": -135.57931518554688, | |
| "logps/ref_rejected": -144.543701171875, | |
| "logps/rejected": -147.69003295898438, | |
| "loss": 0.6929, | |
| "rewards/accuracies": 0.4765625, | |
| "rewards/chosen": -0.0308553334325552, | |
| "rewards/margins": 0.0006078272126615047, | |
| "rewards/rejected": -0.03146316111087799, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.22396047756278303, | |
| "grad_norm": 0.5350177884101868, | |
| "learning_rate": 4.780673881662242e-07, | |
| "logits/chosen": 10.29094409942627, | |
| "logits/rejected": 10.750887870788574, | |
| "logps/chosen": -130.87939453125, | |
| "logps/ref_chosen": -127.53980255126953, | |
| "logps/ref_rejected": -141.44097900390625, | |
| "logps/rejected": -144.80685424804688, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.5078125, | |
| "rewards/chosen": -0.03339584171772003, | |
| "rewards/margins": 0.0002629577647894621, | |
| "rewards/rejected": -0.03365879878401756, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.23713462330177026, | |
| "grad_norm": 0.6819839477539062, | |
| "learning_rate": 4.730171106393466e-07, | |
| "logits/chosen": 10.341330528259277, | |
| "logits/rejected": 10.997025489807129, | |
| "logps/chosen": -135.26243591308594, | |
| "logps/ref_chosen": -130.62847900390625, | |
| "logps/ref_rejected": -141.88360595703125, | |
| "logps/rejected": -146.40476989746094, | |
| "loss": 0.6938, | |
| "rewards/accuracies": 0.484375, | |
| "rewards/chosen": -0.046339523047208786, | |
| "rewards/margins": -0.001128039206378162, | |
| "rewards/rejected": -0.045211486518383026, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2503087690407575, | |
| "grad_norm": 0.5013086199760437, | |
| "learning_rate": 4.6747659310219757e-07, | |
| "logits/chosen": 10.433052062988281, | |
| "logits/rejected": 10.780852317810059, | |
| "logps/chosen": -140.90821838378906, | |
| "logps/ref_chosen": -136.1767120361328, | |
| "logps/ref_rejected": -148.38580322265625, | |
| "logps/rejected": -153.79835510253906, | |
| "loss": 0.6898, | |
| "rewards/accuracies": 0.578125, | |
| "rewards/chosen": -0.0473150797188282, | |
| "rewards/margins": 0.006810260470956564, | |
| "rewards/rejected": -0.0541253462433815, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.2634829147797447, | |
| "grad_norm": 0.5322529673576355, | |
| "learning_rate": 4.6145801481477433e-07, | |
| "logits/chosen": 10.486635208129883, | |
| "logits/rejected": 11.192235946655273, | |
| "logps/chosen": -140.9699249267578, | |
| "logps/ref_chosen": -134.8175048828125, | |
| "logps/ref_rejected": -145.92430114746094, | |
| "logps/rejected": -151.58935546875, | |
| "loss": 0.6957, | |
| "rewards/accuracies": 0.40625, | |
| "rewards/chosen": -0.06152408942580223, | |
| "rewards/margins": -0.004873568192124367, | |
| "rewards/rejected": -0.056650519371032715, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.276657060518732, | |
| "grad_norm": 0.426982581615448, | |
| "learning_rate": 4.549746059183561e-07, | |
| "logits/chosen": 9.691137313842773, | |
| "logits/rejected": 10.519243240356445, | |
| "logps/chosen": -137.68182373046875, | |
| "logps/ref_chosen": -131.53555297851562, | |
| "logps/ref_rejected": -147.749267578125, | |
| "logps/rejected": -154.18560791015625, | |
| "loss": 0.6918, | |
| "rewards/accuracies": 0.5078125, | |
| "rewards/chosen": -0.061462655663490295, | |
| "rewards/margins": 0.002900664694607258, | |
| "rewards/rejected": -0.06436331570148468, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.28983120625771924, | |
| "grad_norm": 0.39799365401268005, | |
| "learning_rate": 4.480406183527823e-07, | |
| "logits/chosen": 10.019172668457031, | |
| "logits/rejected": 10.592238426208496, | |
| "logps/chosen": -137.14170837402344, | |
| "logps/ref_chosen": -129.1536865234375, | |
| "logps/ref_rejected": -138.03895568847656, | |
| "logps/rejected": -145.4950714111328, | |
| "loss": 0.6959, | |
| "rewards/accuracies": 0.4609375, | |
| "rewards/chosen": -0.07987993955612183, | |
| "rewards/margins": -0.005318798590451479, | |
| "rewards/rejected": -0.07456114143133163, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.3030053519967065, | |
| "grad_norm": 0.6805318593978882, | |
| "learning_rate": 4.4067129452759546e-07, | |
| "logits/chosen": 9.99167251586914, | |
| "logits/rejected": 10.793407440185547, | |
| "logps/chosen": -139.6283416748047, | |
| "logps/ref_chosen": -131.8021240234375, | |
| "logps/ref_rejected": -144.97862243652344, | |
| "logps/rejected": -152.63169860839844, | |
| "loss": 0.6942, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.07826196402311325, | |
| "rewards/margins": -0.0017311733681708574, | |
| "rewards/rejected": -0.07653079926967621, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3161794977356937, | |
| "grad_norm": 0.5804150104522705, | |
| "learning_rate": 4.3288283381591725e-07, | |
| "logits/chosen": 10.220123291015625, | |
| "logits/rejected": 10.670228958129883, | |
| "logps/chosen": -141.5391387939453, | |
| "logps/ref_chosen": -134.2378692626953, | |
| "logps/ref_rejected": -146.93955993652344, | |
| "logps/rejected": -154.45404052734375, | |
| "loss": 0.6922, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": -0.07301273196935654, | |
| "rewards/margins": 0.0021319000516086817, | |
| "rewards/rejected": -0.07514464110136032, | |
| "step": 24 | |
| } | |
| ], | |
| "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 | |
| } | |