| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 1274, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0007849293563579278, | |
| "grad_norm": 0.03268135007860773, | |
| "learning_rate": 3.90625e-09, | |
| "logits/chosen": 5914.52099609375, | |
| "logits/rejected": 2785.021484375, | |
| "logps/chosen": -0.6567006707191467, | |
| "logps/rejected": -0.778023362159729, | |
| "loss": 1.3863, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.007849293563579277, | |
| "grad_norm": 0.030196865520322073, | |
| "learning_rate": 3.9062499999999997e-08, | |
| "logits/chosen": 4973.7216796875, | |
| "logits/rejected": 4328.291015625, | |
| "logps/chosen": -0.7547220587730408, | |
| "logps/rejected": -0.8142147064208984, | |
| "loss": 1.3863, | |
| "rewards/accuracies": 0.4351852238178253, | |
| "rewards/chosen": 0.00014649056538473815, | |
| "rewards/margins": 0.00016151672753039747, | |
| "rewards/rejected": -1.5026165783638135e-05, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.015698587127158554, | |
| "grad_norm": 0.03515484842658268, | |
| "learning_rate": 7.812499999999999e-08, | |
| "logits/chosen": 6084.4482421875, | |
| "logits/rejected": 4834.47021484375, | |
| "logps/chosen": -0.6600744128227234, | |
| "logps/rejected": -0.7692333459854126, | |
| "loss": 1.3863, | |
| "rewards/accuracies": 0.4666667580604553, | |
| "rewards/chosen": 6.742794357705861e-05, | |
| "rewards/margins": 1.602613519935403e-05, | |
| "rewards/rejected": 5.140181019669399e-05, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.023547880690737835, | |
| "grad_norm": 0.031197105041082272, | |
| "learning_rate": 1.1718749999999999e-07, | |
| "logits/chosen": 6085.9990234375, | |
| "logits/rejected": 5106.2900390625, | |
| "logps/chosen": -0.7124911546707153, | |
| "logps/rejected": -0.7892431020736694, | |
| "loss": 1.3863, | |
| "rewards/accuracies": 0.6333333253860474, | |
| "rewards/chosen": 9.94208239717409e-05, | |
| "rewards/margins": 0.0004724114260170609, | |
| "rewards/rejected": -0.00037299064570106566, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03139717425431711, | |
| "grad_norm": 0.02997164386237047, | |
| "learning_rate": 1.5624999999999999e-07, | |
| "logits/chosen": 5314.44580078125, | |
| "logits/rejected": 4349.19580078125, | |
| "logps/chosen": -0.7033883929252625, | |
| "logps/rejected": -0.7918646931648254, | |
| "loss": 1.3863, | |
| "rewards/accuracies": 0.6416667103767395, | |
| "rewards/chosen": 0.001075960579328239, | |
| "rewards/margins": 0.0015331318136304617, | |
| "rewards/rejected": -0.0004571711760945618, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.03924646781789639, | |
| "grad_norm": 0.03082303956618271, | |
| "learning_rate": 1.9531249999999998e-07, | |
| "logits/chosen": 6431.498046875, | |
| "logits/rejected": 5048.09912109375, | |
| "logps/chosen": -0.7465437650680542, | |
| "logps/rejected": -0.8175977468490601, | |
| "loss": 1.3863, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.0019261162960901856, | |
| "rewards/margins": 0.004134190734475851, | |
| "rewards/rejected": -0.0022080740891397, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.04709576138147567, | |
| "grad_norm": 0.029348201524489196, | |
| "learning_rate": 2.3437499999999998e-07, | |
| "logits/chosen": 5498.6728515625, | |
| "logits/rejected": 4572.8447265625, | |
| "logps/chosen": -0.684872567653656, | |
| "logps/rejected": -0.8479803800582886, | |
| "loss": 1.3863, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.002722438657656312, | |
| "rewards/margins": 0.006867048796266317, | |
| "rewards/rejected": -0.004144609440118074, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.054945054945054944, | |
| "grad_norm": 0.029933985172415266, | |
| "learning_rate": 2.734375e-07, | |
| "logits/chosen": 5221.7255859375, | |
| "logits/rejected": 4945.2509765625, | |
| "logps/chosen": -0.6324489116668701, | |
| "logps/rejected": -0.7256419658660889, | |
| "loss": 1.3863, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.003583890153095126, | |
| "rewards/margins": 0.01209492888301611, | |
| "rewards/rejected": -0.008511039428412914, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.06279434850863422, | |
| "grad_norm": 0.0421303895483361, | |
| "learning_rate": 3.1249999999999997e-07, | |
| "logits/chosen": 5823.1435546875, | |
| "logits/rejected": 5312.83642578125, | |
| "logps/chosen": -0.6347643733024597, | |
| "logps/rejected": -0.7057794332504272, | |
| "loss": 1.3862, | |
| "rewards/accuracies": 0.6166667342185974, | |
| "rewards/chosen": -0.005841313861310482, | |
| "rewards/margins": 0.016323894262313843, | |
| "rewards/rejected": -0.02216520719230175, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.0706436420722135, | |
| "grad_norm": 0.052609413699091494, | |
| "learning_rate": 3.5156249999999997e-07, | |
| "logits/chosen": 6123.3955078125, | |
| "logits/rejected": 5257.037109375, | |
| "logps/chosen": -0.6579065322875977, | |
| "logps/rejected": -0.7939826846122742, | |
| "loss": 1.3862, | |
| "rewards/accuracies": 0.7583333253860474, | |
| "rewards/chosen": -0.019795957952737808, | |
| "rewards/margins": 0.05933200195431709, | |
| "rewards/rejected": -0.0791279524564743, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.07849293563579278, | |
| "grad_norm": 0.06474882431499306, | |
| "learning_rate": 3.9062499999999997e-07, | |
| "logits/chosen": 6086.119140625, | |
| "logits/rejected": 5854.5458984375, | |
| "logps/chosen": -0.7721215486526489, | |
| "logps/rejected": -0.9403657913208008, | |
| "loss": 1.386, | |
| "rewards/accuracies": 0.6500000953674316, | |
| "rewards/chosen": -0.11592841148376465, | |
| "rewards/margins": 0.09634516388177872, | |
| "rewards/rejected": -0.21227355301380157, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.08634222919937205, | |
| "grad_norm": 0.2514412569708036, | |
| "learning_rate": 4.2968749999999996e-07, | |
| "logits/chosen": 6355.8857421875, | |
| "logits/rejected": 5154.1103515625, | |
| "logps/chosen": -1.2314940690994263, | |
| "logps/rejected": -1.771580696105957, | |
| "loss": 1.3853, | |
| "rewards/accuracies": 0.7083333730697632, | |
| "rewards/chosen": -0.6742404699325562, | |
| "rewards/margins": 0.3862757086753845, | |
| "rewards/rejected": -1.0605162382125854, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.09419152276295134, | |
| "grad_norm": 0.4065982432993292, | |
| "learning_rate": 4.6874999999999996e-07, | |
| "logits/chosen": 5716.19189453125, | |
| "logits/rejected": 4568.13134765625, | |
| "logps/chosen": -3.0005619525909424, | |
| "logps/rejected": -4.118376731872559, | |
| "loss": 1.3839, | |
| "rewards/accuracies": 0.6333333253860474, | |
| "rewards/chosen": -2.4058868885040283, | |
| "rewards/margins": 0.977187991142273, | |
| "rewards/rejected": -3.3830745220184326, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.10204081632653061, | |
| "grad_norm": 0.6207662485091842, | |
| "learning_rate": 4.999962424962166e-07, | |
| "logits/chosen": 6246.6572265625, | |
| "logits/rejected": 5490.064453125, | |
| "logps/chosen": -4.607759475708008, | |
| "logps/rejected": -6.540151119232178, | |
| "loss": 1.3829, | |
| "rewards/accuracies": 0.6916666626930237, | |
| "rewards/chosen": -4.007367134094238, | |
| "rewards/margins": 1.8176984786987305, | |
| "rewards/rejected": -5.825065612792969, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.10989010989010989, | |
| "grad_norm": 1.6726707215948637, | |
| "learning_rate": 4.998647417232375e-07, | |
| "logits/chosen": 6259.673828125, | |
| "logits/rejected": 5075.4140625, | |
| "logps/chosen": -6.643527984619141, | |
| "logps/rejected": -9.528350830078125, | |
| "loss": 1.3806, | |
| "rewards/accuracies": 0.6416667103767395, | |
| "rewards/chosen": -6.045531272888184, | |
| "rewards/margins": 2.7727363109588623, | |
| "rewards/rejected": -8.818267822265625, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.11773940345368916, | |
| "grad_norm": 1.8074740604257467, | |
| "learning_rate": 4.995454786965036e-07, | |
| "logits/chosen": 6712.15234375, | |
| "logits/rejected": 5173.9384765625, | |
| "logps/chosen": -5.438518047332764, | |
| "logps/rejected": -10.787275314331055, | |
| "loss": 1.3787, | |
| "rewards/accuracies": 0.6583333015441895, | |
| "rewards/chosen": -4.854126453399658, | |
| "rewards/margins": 5.265545845031738, | |
| "rewards/rejected": -10.119672775268555, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.12558869701726844, | |
| "grad_norm": 3.1308973590524034, | |
| "learning_rate": 4.990386933279972e-07, | |
| "logits/chosen": 5863.80517578125, | |
| "logits/rejected": 4384.7138671875, | |
| "logps/chosen": -9.97611141204834, | |
| "logps/rejected": -14.926114082336426, | |
| "loss": 1.3759, | |
| "rewards/accuracies": 0.6666666865348816, | |
| "rewards/chosen": -9.367152214050293, | |
| "rewards/margins": 4.869722843170166, | |
| "rewards/rejected": -14.236875534057617, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.13343799058084774, | |
| "grad_norm": 4.1177149352808, | |
| "learning_rate": 4.983447664444096e-07, | |
| "logits/chosen": 6553.7626953125, | |
| "logits/rejected": 5106.4052734375, | |
| "logps/chosen": -12.286310195922852, | |
| "logps/rejected": -20.12063980102539, | |
| "loss": 1.3774, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -11.659884452819824, | |
| "rewards/margins": 7.752722263336182, | |
| "rewards/rejected": -19.412609100341797, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.141287284144427, | |
| "grad_norm": 3.8730928565663554, | |
| "learning_rate": 4.97464219500968e-07, | |
| "logits/chosen": 5846.3642578125, | |
| "logits/rejected": 4395.67578125, | |
| "logps/chosen": -6.325826644897461, | |
| "logps/rejected": -14.793339729309082, | |
| "loss": 1.3709, | |
| "rewards/accuracies": 0.6666666269302368, | |
| "rewards/chosen": -5.696617603302002, | |
| "rewards/margins": 8.336694717407227, | |
| "rewards/rejected": -14.03331184387207, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.14913657770800628, | |
| "grad_norm": 6.602624193673466, | |
| "learning_rate": 4.963977141895843e-07, | |
| "logits/chosen": 5749.099609375, | |
| "logits/rejected": 4122.6748046875, | |
| "logps/chosen": -13.3460111618042, | |
| "logps/rejected": -30.625717163085938, | |
| "loss": 1.3673, | |
| "rewards/accuracies": 0.8083332777023315, | |
| "rewards/chosen": -12.694738388061523, | |
| "rewards/margins": 17.16573143005371, | |
| "rewards/rejected": -29.860469818115234, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.15698587127158556, | |
| "grad_norm": 5.197411313874128, | |
| "learning_rate": 4.951460519416227e-07, | |
| "logits/chosen": 6093.1865234375, | |
| "logits/rejected": 4876.7451171875, | |
| "logps/chosen": -9.102602005004883, | |
| "logps/rejected": -26.314044952392578, | |
| "loss": 1.3625, | |
| "rewards/accuracies": 0.7583333849906921, | |
| "rewards/chosen": -8.488224983215332, | |
| "rewards/margins": 17.09214210510254, | |
| "rewards/rejected": -25.580366134643555, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.16483516483516483, | |
| "grad_norm": 9.618320681230124, | |
| "learning_rate": 4.937101733256606e-07, | |
| "logits/chosen": 5406.921875, | |
| "logits/rejected": 4192.8603515625, | |
| "logps/chosen": -14.340167045593262, | |
| "logps/rejected": -33.15205383300781, | |
| "loss": 1.357, | |
| "rewards/accuracies": 0.6333333253860474, | |
| "rewards/chosen": -13.77330207824707, | |
| "rewards/margins": 18.675561904907227, | |
| "rewards/rejected": -32.44886016845703, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.1726844583987441, | |
| "grad_norm": 7.375553630744227, | |
| "learning_rate": 4.920911573406924e-07, | |
| "logits/chosen": 6773.37255859375, | |
| "logits/rejected": 5361.48046875, | |
| "logps/chosen": -11.12817668914795, | |
| "logps/rejected": -28.294403076171875, | |
| "loss": 1.3615, | |
| "rewards/accuracies": 0.7166666388511658, | |
| "rewards/chosen": -10.548368453979492, | |
| "rewards/margins": 17.09095573425293, | |
| "rewards/rejected": -27.639324188232422, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.18053375196232338, | |
| "grad_norm": 8.615349191502917, | |
| "learning_rate": 4.902902206053098e-07, | |
| "logits/chosen": 5996.9853515625, | |
| "logits/rejected": 4782.43798828125, | |
| "logps/chosen": -18.20733070373535, | |
| "logps/rejected": -41.948974609375, | |
| "loss": 1.3626, | |
| "rewards/accuracies": 0.7166666984558105, | |
| "rewards/chosen": -17.62150001525879, | |
| "rewards/margins": 23.599395751953125, | |
| "rewards/rejected": -41.22089385986328, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.18838304552590268, | |
| "grad_norm": 6.3641054961132175, | |
| "learning_rate": 4.883087164434672e-07, | |
| "logits/chosen": 5568.8056640625, | |
| "logits/rejected": 3895.72412109375, | |
| "logps/chosen": -7.421477317810059, | |
| "logps/rejected": -30.459941864013672, | |
| "loss": 1.3562, | |
| "rewards/accuracies": 0.7583333253860474, | |
| "rewards/chosen": -6.827574729919434, | |
| "rewards/margins": 22.909481048583984, | |
| "rewards/rejected": -29.7370548248291, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.19623233908948196, | |
| "grad_norm": 25.511536310895185, | |
| "learning_rate": 4.861481338675183e-07, | |
| "logits/chosen": 6335.380859375, | |
| "logits/rejected": 4804.142578125, | |
| "logps/chosen": -14.736673355102539, | |
| "logps/rejected": -40.67548370361328, | |
| "loss": 1.3621, | |
| "rewards/accuracies": 0.6666666269302368, | |
| "rewards/chosen": -14.16309642791748, | |
| "rewards/margins": 25.803043365478516, | |
| "rewards/rejected": -39.96613693237305, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.20408163265306123, | |
| "grad_norm": 8.643358672198618, | |
| "learning_rate": 4.838100964592904e-07, | |
| "logits/chosen": 6742.24365234375, | |
| "logits/rejected": 5024.3515625, | |
| "logps/chosen": -8.986857414245605, | |
| "logps/rejected": -24.885251998901367, | |
| "loss": 1.3711, | |
| "rewards/accuracies": 0.7083333730697632, | |
| "rewards/chosen": -8.32063102722168, | |
| "rewards/margins": 15.831449508666992, | |
| "rewards/rejected": -24.152080535888672, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.2119309262166405, | |
| "grad_norm": 6.0870160275581044, | |
| "learning_rate": 4.812963611500339e-07, | |
| "logits/chosen": 6522.9189453125, | |
| "logits/rejected": 5658.826171875, | |
| "logps/chosen": -11.388198852539062, | |
| "logps/rejected": -31.324010848999023, | |
| "loss": 1.3595, | |
| "rewards/accuracies": 0.7083333134651184, | |
| "rewards/chosen": -10.788042068481445, | |
| "rewards/margins": 19.862857818603516, | |
| "rewards/rejected": -30.650897979736328, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.21978021978021978, | |
| "grad_norm": 8.557241921021467, | |
| "learning_rate": 4.786088169001671e-07, | |
| "logits/chosen": 5661.5009765625, | |
| "logits/rejected": 4401.69482421875, | |
| "logps/chosen": -8.053993225097656, | |
| "logps/rejected": -32.422542572021484, | |
| "loss": 1.3579, | |
| "rewards/accuracies": 0.7166666984558105, | |
| "rewards/chosen": -7.461718559265137, | |
| "rewards/margins": 24.201175689697266, | |
| "rewards/rejected": -31.662893295288086, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.22762951334379905, | |
| "grad_norm": 9.481317515747005, | |
| "learning_rate": 4.7574948327980567e-07, | |
| "logits/chosen": 7745.7158203125, | |
| "logits/rejected": 4650.02880859375, | |
| "logps/chosen": -12.464855194091797, | |
| "logps/rejected": -40.86915969848633, | |
| "loss": 1.3604, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -11.852567672729492, | |
| "rewards/margins": 28.29081153869629, | |
| "rewards/rejected": -40.14337921142578, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.23547880690737832, | |
| "grad_norm": 8.835805617844823, | |
| "learning_rate": 4.727205089511466e-07, | |
| "logits/chosen": 5531.8173828125, | |
| "logits/rejected": 4774.9970703125, | |
| "logps/chosen": -10.429524421691895, | |
| "logps/rejected": -26.804691314697266, | |
| "loss": 1.3651, | |
| "rewards/accuracies": 0.6916666626930237, | |
| "rewards/chosen": -9.819427490234375, | |
| "rewards/margins": 16.315807342529297, | |
| "rewards/rejected": -26.13523292541504, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.24332810047095763, | |
| "grad_norm": 15.454604782065125, | |
| "learning_rate": 4.6952417005384247e-07, | |
| "logits/chosen": 6384.7763671875, | |
| "logits/rejected": 5284.7490234375, | |
| "logps/chosen": -7.295445442199707, | |
| "logps/rejected": -19.249813079833984, | |
| "loss": 1.3664, | |
| "rewards/accuracies": 0.7083333730697632, | |
| "rewards/chosen": -6.720705509185791, | |
| "rewards/margins": 11.84199333190918, | |
| "rewards/rejected": -18.56269645690918, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.25117739403453687, | |
| "grad_norm": 11.201402261813897, | |
| "learning_rate": 4.661628684945851e-07, | |
| "logits/chosen": 6445.9931640625, | |
| "logits/rejected": 4742.51123046875, | |
| "logps/chosen": -10.060206413269043, | |
| "logps/rejected": -35.663307189941406, | |
| "loss": 1.3585, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -9.450899124145508, | |
| "rewards/margins": 25.444076538085938, | |
| "rewards/rejected": -34.89497375488281, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.25902668759811615, | |
| "grad_norm": 14.088756180907868, | |
| "learning_rate": 4.626391301421782e-07, | |
| "logits/chosen": 6144.3173828125, | |
| "logits/rejected": 5099.60986328125, | |
| "logps/chosen": -11.769998550415039, | |
| "logps/rejected": -26.48116683959961, | |
| "loss": 1.3667, | |
| "rewards/accuracies": 0.6833333373069763, | |
| "rewards/chosen": -11.127023696899414, | |
| "rewards/margins": 14.672645568847656, | |
| "rewards/rejected": -25.799671173095703, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.2668759811616955, | |
| "grad_norm": 7.090474377137257, | |
| "learning_rate": 4.5895560292945996e-07, | |
| "logits/chosen": 6591.3095703125, | |
| "logits/rejected": 6281.44970703125, | |
| "logps/chosen": -8.332866668701172, | |
| "logps/rejected": -23.061569213867188, | |
| "loss": 1.3574, | |
| "rewards/accuracies": 0.7083333730697632, | |
| "rewards/chosen": -7.750649929046631, | |
| "rewards/margins": 14.612638473510742, | |
| "rewards/rejected": -22.3632869720459, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.27472527472527475, | |
| "grad_norm": 11.8517335767745, | |
| "learning_rate": 4.5511505486349865e-07, | |
| "logits/chosen": 6811.2880859375, | |
| "logits/rejected": 5612.2724609375, | |
| "logps/chosen": -12.734210014343262, | |
| "logps/rejected": -46.445518493652344, | |
| "loss": 1.3473, | |
| "rewards/accuracies": 0.7249999642372131, | |
| "rewards/chosen": -12.119455337524414, | |
| "rewards/margins": 33.59000015258789, | |
| "rewards/rejected": -45.70945358276367, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.282574568288854, | |
| "grad_norm": 17.10556818380673, | |
| "learning_rate": 4.5112037194555876e-07, | |
| "logits/chosen": 6228.7158203125, | |
| "logits/rejected": 5359.6708984375, | |
| "logps/chosen": -16.61972999572754, | |
| "logps/rejected": -51.56354522705078, | |
| "loss": 1.3582, | |
| "rewards/accuracies": 0.7500000596046448, | |
| "rewards/chosen": -16.05486488342285, | |
| "rewards/margins": 34.75627517700195, | |
| "rewards/rejected": -50.81113815307617, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.2904238618524333, | |
| "grad_norm": 12.277205892634896, | |
| "learning_rate": 4.4697455600239863e-07, | |
| "logits/chosen": 5767.1728515625, | |
| "logits/rejected": 5032.4482421875, | |
| "logps/chosen": -8.446660041809082, | |
| "logps/rejected": -22.16537857055664, | |
| "loss": 1.3678, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -7.765728950500488, | |
| "rewards/margins": 13.689413070678711, | |
| "rewards/rejected": -21.455141067504883, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.29827315541601257, | |
| "grad_norm": 18.240138504142248, | |
| "learning_rate": 4.426807224305315e-07, | |
| "logits/chosen": 7020.58203125, | |
| "logits/rejected": 4838.7958984375, | |
| "logps/chosen": -8.1710205078125, | |
| "logps/rejected": -42.699180603027344, | |
| "loss": 1.36, | |
| "rewards/accuracies": 0.7666667103767395, | |
| "rewards/chosen": -7.526429653167725, | |
| "rewards/margins": 34.454978942871094, | |
| "rewards/rejected": -41.98141098022461, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.30612244897959184, | |
| "grad_norm": 9.573222554347216, | |
| "learning_rate": 4.3824209785514326e-07, | |
| "logits/chosen": 6994.20703125, | |
| "logits/rejected": 4667.099609375, | |
| "logps/chosen": -10.061126708984375, | |
| "logps/rejected": -38.28179168701172, | |
| "loss": 1.354, | |
| "rewards/accuracies": 0.6916667222976685, | |
| "rewards/chosen": -9.435959815979004, | |
| "rewards/margins": 28.09332847595215, | |
| "rewards/rejected": -37.52928924560547, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.3139717425431711, | |
| "grad_norm": 6.6100886548986475, | |
| "learning_rate": 4.3366201770542687e-07, | |
| "logits/chosen": 5962.6591796875, | |
| "logits/rejected": 5447.56884765625, | |
| "logps/chosen": -10.098733901977539, | |
| "logps/rejected": -31.517467498779297, | |
| "loss": 1.3655, | |
| "rewards/accuracies": 0.7249999642372131, | |
| "rewards/chosen": -9.449263572692871, | |
| "rewards/margins": 21.30449867248535, | |
| "rewards/rejected": -30.753765106201172, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.3218210361067504, | |
| "grad_norm": 30.548099506436554, | |
| "learning_rate": 4.2894392370815567e-07, | |
| "logits/chosen": 6273.9775390625, | |
| "logits/rejected": 4478.13330078125, | |
| "logps/chosen": -14.873064041137695, | |
| "logps/rejected": -59.72370147705078, | |
| "loss": 1.3447, | |
| "rewards/accuracies": 0.7583333253860474, | |
| "rewards/chosen": -14.206558227539062, | |
| "rewards/margins": 44.70840072631836, | |
| "rewards/rejected": -58.91496658325195, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.32967032967032966, | |
| "grad_norm": 11.996768365493867, | |
| "learning_rate": 4.2409136130137845e-07, | |
| "logits/chosen": 5747.73681640625, | |
| "logits/rejected": 4370.01708984375, | |
| "logps/chosen": -21.314987182617188, | |
| "logps/rejected": -54.006263732910156, | |
| "loss": 1.3593, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -20.65334701538086, | |
| "rewards/margins": 32.607086181640625, | |
| "rewards/rejected": -53.26043701171875, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.33751962323390894, | |
| "grad_norm": 10.590905912907617, | |
| "learning_rate": 4.1910797697018017e-07, | |
| "logits/chosen": 5875.6357421875, | |
| "logits/rejected": 4011.819091796875, | |
| "logps/chosen": -11.717180252075195, | |
| "logps/rejected": -46.7780647277832, | |
| "loss": 1.3438, | |
| "rewards/accuracies": 0.7916666865348816, | |
| "rewards/chosen": -11.10181999206543, | |
| "rewards/margins": 34.92017364501953, | |
| "rewards/rejected": -46.02199172973633, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.3453689167974882, | |
| "grad_norm": 14.453344590210275, | |
| "learning_rate": 4.1399751550651084e-07, | |
| "logits/chosen": 6271.8125, | |
| "logits/rejected": 5575.626953125, | |
| "logps/chosen": -13.031234741210938, | |
| "logps/rejected": -36.1110725402832, | |
| "loss": 1.3582, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -12.423306465148926, | |
| "rewards/margins": 22.96180534362793, | |
| "rewards/rejected": -35.385108947753906, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.3532182103610675, | |
| "grad_norm": 11.608919477063516, | |
| "learning_rate": 4.087638171951401e-07, | |
| "logits/chosen": 7476.86572265625, | |
| "logits/rejected": 4706.1162109375, | |
| "logps/chosen": -6.023161888122559, | |
| "logps/rejected": -37.36185836791992, | |
| "loss": 1.3562, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -5.43270206451416, | |
| "rewards/margins": 31.197275161743164, | |
| "rewards/rejected": -36.629974365234375, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.36106750392464676, | |
| "grad_norm": 26.28345054484464, | |
| "learning_rate": 4.034108149278543e-07, | |
| "logits/chosen": 7570.7451171875, | |
| "logits/rejected": 5338.115234375, | |
| "logps/chosen": -12.235418319702148, | |
| "logps/rejected": -36.413673400878906, | |
| "loss": 1.3609, | |
| "rewards/accuracies": 0.7916667461395264, | |
| "rewards/chosen": -11.572993278503418, | |
| "rewards/margins": 24.130626678466797, | |
| "rewards/rejected": -35.70362091064453, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.36891679748822603, | |
| "grad_norm": 9.755625997424268, | |
| "learning_rate": 3.979425312480629e-07, | |
| "logits/chosen": 6442.072265625, | |
| "logits/rejected": 5090.69287109375, | |
| "logps/chosen": -11.879281997680664, | |
| "logps/rejected": -36.57582092285156, | |
| "loss": 1.3581, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -11.22594165802002, | |
| "rewards/margins": 24.602800369262695, | |
| "rewards/rejected": -35.82874298095703, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.37676609105180536, | |
| "grad_norm": 13.488032723757092, | |
| "learning_rate": 3.923630753280357e-07, | |
| "logits/chosen": 7082.0048828125, | |
| "logits/rejected": 5317.03369140625, | |
| "logps/chosen": -10.345157623291016, | |
| "logps/rejected": -40.309444427490234, | |
| "loss": 1.3518, | |
| "rewards/accuracies": 0.7916667461395264, | |
| "rewards/chosen": -9.730131149291992, | |
| "rewards/margins": 29.90970230102539, | |
| "rewards/rejected": -39.63983154296875, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "grad_norm": 10.81857496627396, | |
| "learning_rate": 3.866766398810424e-07, | |
| "logits/chosen": 6680.68359375, | |
| "logits/rejected": 5265.3974609375, | |
| "logps/chosen": -9.079381942749023, | |
| "logps/rejected": -47.32402801513672, | |
| "loss": 1.3498, | |
| "rewards/accuracies": 0.7666667103767395, | |
| "rewards/chosen": -8.535429954528809, | |
| "rewards/margins": 38.05257034301758, | |
| "rewards/rejected": -46.58799743652344, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.3924646781789639, | |
| "grad_norm": 8.225268327717496, | |
| "learning_rate": 3.8088749801071496e-07, | |
| "logits/chosen": 6926.06640625, | |
| "logits/rejected": 4467.6005859375, | |
| "logps/chosen": -18.076974868774414, | |
| "logps/rejected": -52.8994255065918, | |
| "loss": 1.3522, | |
| "rewards/accuracies": 0.6916666626930237, | |
| "rewards/chosen": -17.428617477416992, | |
| "rewards/margins": 34.659934997558594, | |
| "rewards/rejected": -52.08855056762695, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.4003139717425432, | |
| "grad_norm": 8.403224575968139, | |
| "learning_rate": 3.75e-07, | |
| "logits/chosen": 5711.43994140625, | |
| "logits/rejected": 4712.4541015625, | |
| "logps/chosen": -10.59355640411377, | |
| "logps/rejected": -28.433752059936523, | |
| "loss": 1.357, | |
| "rewards/accuracies": 0.7416666746139526, | |
| "rewards/chosen": -9.983890533447266, | |
| "rewards/margins": 17.735435485839844, | |
| "rewards/rejected": -27.719324111938477, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.40816326530612246, | |
| "grad_norm": 7.859662724135296, | |
| "learning_rate": 3.6901857004211443e-07, | |
| "logits/chosen": 6058.2236328125, | |
| "logits/rejected": 5172.513671875, | |
| "logps/chosen": -9.832149505615234, | |
| "logps/rejected": -27.763540267944336, | |
| "loss": 1.3672, | |
| "rewards/accuracies": 0.7083333730697632, | |
| "rewards/chosen": -9.129549980163574, | |
| "rewards/margins": 17.810230255126953, | |
| "rewards/rejected": -26.93977928161621, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.41601255886970173, | |
| "grad_norm": 13.018474920832832, | |
| "learning_rate": 3.6294770291596076e-07, | |
| "logits/chosen": 6869.9970703125, | |
| "logits/rejected": 5111.6474609375, | |
| "logps/chosen": -6.732565879821777, | |
| "logps/rejected": -25.002674102783203, | |
| "loss": 1.3553, | |
| "rewards/accuracies": 0.6833333373069763, | |
| "rewards/chosen": -6.134089469909668, | |
| "rewards/margins": 18.133365631103516, | |
| "rewards/rejected": -24.2674560546875, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.423861852433281, | |
| "grad_norm": 17.461930895857275, | |
| "learning_rate": 3.5679196060850034e-07, | |
| "logits/chosen": 6209.1474609375, | |
| "logits/rejected": 4489.95556640625, | |
| "logps/chosen": -17.81626319885254, | |
| "logps/rejected": -52.312652587890625, | |
| "loss": 1.3616, | |
| "rewards/accuracies": 0.7416666746139526, | |
| "rewards/chosen": -17.198108673095703, | |
| "rewards/margins": 34.41394805908203, | |
| "rewards/rejected": -51.61205291748047, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.4317111459968603, | |
| "grad_norm": 14.537248039469688, | |
| "learning_rate": 3.505559688866229e-07, | |
| "logits/chosen": 6301.17724609375, | |
| "logits/rejected": 4744.8837890625, | |
| "logps/chosen": -10.278491973876953, | |
| "logps/rejected": -44.104400634765625, | |
| "loss": 1.3482, | |
| "rewards/accuracies": 0.7750000357627869, | |
| "rewards/chosen": -9.643937110900879, | |
| "rewards/margins": 33.664276123046875, | |
| "rewards/rejected": -43.30820846557617, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.43956043956043955, | |
| "grad_norm": 12.121639720486284, | |
| "learning_rate": 3.4424441382108826e-07, | |
| "logits/chosen": 5940.2646484375, | |
| "logits/rejected": 5123.6513671875, | |
| "logps/chosen": -16.82316017150879, | |
| "logps/rejected": -34.89044952392578, | |
| "loss": 1.3679, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -16.180652618408203, | |
| "rewards/margins": 17.98758316040039, | |
| "rewards/rejected": -34.168235778808594, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.4474097331240188, | |
| "grad_norm": 9.643273866165796, | |
| "learning_rate": 3.378620382651523e-07, | |
| "logits/chosen": 6564.9365234375, | |
| "logits/rejected": 5344.296875, | |
| "logps/chosen": -11.590880393981934, | |
| "logps/rejected": -38.3111572265625, | |
| "loss": 1.3495, | |
| "rewards/accuracies": 0.716666579246521, | |
| "rewards/chosen": -10.887467384338379, | |
| "rewards/margins": 26.603397369384766, | |
| "rewards/rejected": -37.49085998535156, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.4552590266875981, | |
| "grad_norm": 9.384348098391655, | |
| "learning_rate": 3.314136382905234e-07, | |
| "logits/chosen": 6624.49072265625, | |
| "logits/rejected": 4907.68212890625, | |
| "logps/chosen": -9.338412284851074, | |
| "logps/rejected": -45.03810119628906, | |
| "loss": 1.3524, | |
| "rewards/accuracies": 0.7916666269302368, | |
| "rewards/chosen": -8.67926025390625, | |
| "rewards/margins": 35.5770263671875, | |
| "rewards/rejected": -44.25628662109375, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.4631083202511774, | |
| "grad_norm": 11.149480684566118, | |
| "learning_rate": 3.249040595833274e-07, | |
| "logits/chosen": 7035.4873046875, | |
| "logits/rejected": 5026.8193359375, | |
| "logps/chosen": -11.245471954345703, | |
| "logps/rejected": -44.87081527709961, | |
| "loss": 1.3472, | |
| "rewards/accuracies": 0.7916666865348816, | |
| "rewards/chosen": -10.6029691696167, | |
| "rewards/margins": 33.520263671875, | |
| "rewards/rejected": -44.123233795166016, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.47095761381475665, | |
| "grad_norm": 19.526133215761668, | |
| "learning_rate": 3.1833819380279023e-07, | |
| "logits/chosen": 6805.5771484375, | |
| "logits/rejected": 5161.3935546875, | |
| "logps/chosen": -8.982612609863281, | |
| "logps/rejected": -35.16088104248047, | |
| "loss": 1.3532, | |
| "rewards/accuracies": 0.7750000357627869, | |
| "rewards/chosen": -8.421594619750977, | |
| "rewards/margins": 26.0313777923584, | |
| "rewards/rejected": -34.45297622680664, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.478806907378336, | |
| "grad_norm": 11.018750446737162, | |
| "learning_rate": 3.11720974905373e-07, | |
| "logits/chosen": 6474.22900390625, | |
| "logits/rejected": 5111.5498046875, | |
| "logps/chosen": -11.592761993408203, | |
| "logps/rejected": -42.404930114746094, | |
| "loss": 1.3414, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -10.948739051818848, | |
| "rewards/margins": 30.717721939086914, | |
| "rewards/rejected": -41.666465759277344, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.48665620094191525, | |
| "grad_norm": 8.775291118452293, | |
| "learning_rate": 3.0505737543712275e-07, | |
| "logits/chosen": 5456.4990234375, | |
| "logits/rejected": 3857.27294921875, | |
| "logps/chosen": -15.486307144165039, | |
| "logps/rejected": -48.56859588623047, | |
| "loss": 1.3521, | |
| "rewards/accuracies": 0.7083333134651184, | |
| "rewards/chosen": -14.838460922241211, | |
| "rewards/margins": 32.94873809814453, | |
| "rewards/rejected": -47.787193298339844, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.4945054945054945, | |
| "grad_norm": 9.749208955925335, | |
| "learning_rate": 2.9835240279702513e-07, | |
| "logits/chosen": 7186.5859375, | |
| "logits/rejected": 5223.5263671875, | |
| "logps/chosen": -8.555147171020508, | |
| "logps/rejected": -48.16273880004883, | |
| "loss": 1.3394, | |
| "rewards/accuracies": 0.7583333253860474, | |
| "rewards/chosen": -7.907660007476807, | |
| "rewards/margins": 39.55339813232422, | |
| "rewards/rejected": -47.461055755615234, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.5023547880690737, | |
| "grad_norm": 10.872491616524146, | |
| "learning_rate": 2.9161109547416667e-07, | |
| "logits/chosen": 6674.87646484375, | |
| "logits/rejected": 5399.53564453125, | |
| "logps/chosen": -18.24646759033203, | |
| "logps/rejected": -34.43614196777344, | |
| "loss": 1.3488, | |
| "rewards/accuracies": 0.6666666865348816, | |
| "rewards/chosen": -17.58979606628418, | |
| "rewards/margins": 16.084903717041016, | |
| "rewards/rejected": -33.67470169067383, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.5102040816326531, | |
| "grad_norm": 10.769277377138156, | |
| "learning_rate": 2.848385192615339e-07, | |
| "logits/chosen": 5799.63671875, | |
| "logits/rejected": 4268.919921875, | |
| "logps/chosen": -12.065497398376465, | |
| "logps/rejected": -39.761295318603516, | |
| "loss": 1.3529, | |
| "rewards/accuracies": 0.6750000715255737, | |
| "rewards/chosen": -11.4081449508667, | |
| "rewards/margins": 27.570276260375977, | |
| "rewards/rejected": -38.97842788696289, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.5180533751962323, | |
| "grad_norm": 12.661156373398873, | |
| "learning_rate": 2.780397634492949e-07, | |
| "logits/chosen": 6555.94775390625, | |
| "logits/rejected": 4259.31787109375, | |
| "logps/chosen": -10.676517486572266, | |
| "logps/rejected": -47.2622184753418, | |
| "loss": 1.3509, | |
| "rewards/accuracies": 0.7416666746139526, | |
| "rewards/chosen": -10.025793075561523, | |
| "rewards/margins": 36.494449615478516, | |
| "rewards/rejected": -46.520240783691406, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.5259026687598116, | |
| "grad_norm": 10.079297684272941, | |
| "learning_rate": 2.71219937000424e-07, | |
| "logits/chosen": 6487.3076171875, | |
| "logits/rejected": 4708.78515625, | |
| "logps/chosen": -9.443751335144043, | |
| "logps/rejected": -38.85082244873047, | |
| "loss": 1.3585, | |
| "rewards/accuracies": 0.7333333492279053, | |
| "rewards/chosen": -8.82435131072998, | |
| "rewards/margins": 29.266210556030273, | |
| "rewards/rejected": -38.09056854248047, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.533751962323391, | |
| "grad_norm": 9.015834795504192, | |
| "learning_rate": 2.6438416471154273e-07, | |
| "logits/chosen": 6213.49267578125, | |
| "logits/rejected": 4702.8037109375, | |
| "logps/chosen": -9.60643196105957, | |
| "logps/rejected": -42.58051300048828, | |
| "loss": 1.3547, | |
| "rewards/accuracies": 0.7750000357627869, | |
| "rewards/chosen": -8.9835205078125, | |
| "rewards/margins": 32.88523864746094, | |
| "rewards/rejected": -41.8687629699707, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.5416012558869702, | |
| "grad_norm": 11.295069759751865, | |
| "learning_rate": 2.5753758336186326e-07, | |
| "logits/chosen": 6236.7900390625, | |
| "logits/rejected": 4860.75439453125, | |
| "logps/chosen": -9.005716323852539, | |
| "logps/rejected": -47.84248352050781, | |
| "loss": 1.3411, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -8.415175437927246, | |
| "rewards/margins": 38.613529205322266, | |
| "rewards/rejected": -47.02870178222656, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.5494505494505495, | |
| "grad_norm": 12.115147537268316, | |
| "learning_rate": 2.5068533785312666e-07, | |
| "logits/chosen": 5841.62890625, | |
| "logits/rejected": 4997.1611328125, | |
| "logps/chosen": -15.859621047973633, | |
| "logps/rejected": -44.61968994140625, | |
| "loss": 1.3585, | |
| "rewards/accuracies": 0.6416667103767395, | |
| "rewards/chosen": -15.210749626159668, | |
| "rewards/margins": 28.713302612304688, | |
| "rewards/rejected": -43.92405319213867, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.5572998430141287, | |
| "grad_norm": 9.83299305993898, | |
| "learning_rate": 2.4383257734343794e-07, | |
| "logits/chosen": 5929.123046875, | |
| "logits/rejected": 5199.5263671875, | |
| "logps/chosen": -11.007040977478027, | |
| "logps/rejected": -41.53852081298828, | |
| "loss": 1.3444, | |
| "rewards/accuracies": 0.7416666746139526, | |
| "rewards/chosen": -10.386730194091797, | |
| "rewards/margins": 30.414636611938477, | |
| "rewards/rejected": -40.80137252807617, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.565149136577708, | |
| "grad_norm": 11.047220141739768, | |
| "learning_rate": 2.3698445137790258e-07, | |
| "logits/chosen": 6393.74462890625, | |
| "logits/rejected": 4948.32470703125, | |
| "logps/chosen": -7.117844581604004, | |
| "logps/rejected": -32.42450714111328, | |
| "loss": 1.3619, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -6.463726997375488, | |
| "rewards/margins": 25.183258056640625, | |
| "rewards/rejected": -31.646984100341797, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.5729984301412873, | |
| "grad_norm": 11.23974480867259, | |
| "learning_rate": 2.3014610601897157e-07, | |
| "logits/chosen": 6870.53369140625, | |
| "logits/rejected": 4605.583984375, | |
| "logps/chosen": -9.47913932800293, | |
| "logps/rejected": -41.51166534423828, | |
| "loss": 1.3438, | |
| "rewards/accuracies": 0.7416666746139526, | |
| "rewards/chosen": -8.836474418640137, | |
| "rewards/margins": 31.91958236694336, | |
| "rewards/rejected": -40.75605773925781, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.5808477237048666, | |
| "grad_norm": 15.559189693273993, | |
| "learning_rate": 2.2332267997940513e-07, | |
| "logits/chosen": 5558.8349609375, | |
| "logits/rejected": 4223.8994140625, | |
| "logps/chosen": -15.262466430664062, | |
| "logps/rejected": -46.9307746887207, | |
| "loss": 1.3578, | |
| "rewards/accuracies": 0.7583333253860474, | |
| "rewards/chosen": -14.636202812194824, | |
| "rewards/margins": 31.5787296295166, | |
| "rewards/rejected": -46.214927673339844, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.5886970172684458, | |
| "grad_norm": 14.272382010281738, | |
| "learning_rate": 2.1651930076075723e-07, | |
| "logits/chosen": 5937.4658203125, | |
| "logits/rejected": 4661.57470703125, | |
| "logps/chosen": -17.500411987304688, | |
| "logps/rejected": -43.45275115966797, | |
| "loss": 1.3582, | |
| "rewards/accuracies": 0.6916666030883789, | |
| "rewards/chosen": -16.904193878173828, | |
| "rewards/margins": 25.871822357177734, | |
| "rewards/rejected": -42.776023864746094, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.5965463108320251, | |
| "grad_norm": 8.754492592596351, | |
| "learning_rate": 2.0974108080028692e-07, | |
| "logits/chosen": 6596.6220703125, | |
| "logits/rejected": 4611.462890625, | |
| "logps/chosen": -7.955502510070801, | |
| "logps/rejected": -37.630226135253906, | |
| "loss": 1.3555, | |
| "rewards/accuracies": 0.7166666388511658, | |
| "rewards/chosen": -7.332449913024902, | |
| "rewards/margins": 29.615585327148438, | |
| "rewards/rejected": -36.948036193847656, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.6043956043956044, | |
| "grad_norm": 10.912459081597383, | |
| "learning_rate": 2.0299311362918773e-07, | |
| "logits/chosen": 6754.17578125, | |
| "logits/rejected": 5320.26806640625, | |
| "logps/chosen": -13.05921745300293, | |
| "logps/rejected": -38.955936431884766, | |
| "loss": 1.3558, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -12.377306938171387, | |
| "rewards/margins": 25.745901107788086, | |
| "rewards/rejected": -38.123207092285156, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.6122448979591837, | |
| "grad_norm": 11.267028059252095, | |
| "learning_rate": 1.962804700450265e-07, | |
| "logits/chosen": 6594.95068359375, | |
| "logits/rejected": 5792.111328125, | |
| "logps/chosen": -11.285120010375977, | |
| "logps/rejected": -36.00288009643555, | |
| "loss": 1.3521, | |
| "rewards/accuracies": 0.7083333730697632, | |
| "rewards/chosen": -10.641828536987305, | |
| "rewards/margins": 24.54639434814453, | |
| "rewards/rejected": -35.1882209777832, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.6200941915227629, | |
| "grad_norm": 16.889300889632764, | |
| "learning_rate": 1.8960819430126334e-07, | |
| "logits/chosen": 5998.2666015625, | |
| "logits/rejected": 4746.22119140625, | |
| "logps/chosen": -18.8221378326416, | |
| "logps/rejected": -49.53624725341797, | |
| "loss": 1.3518, | |
| "rewards/accuracies": 0.7416666746139526, | |
| "rewards/chosen": -18.138534545898438, | |
| "rewards/margins": 30.649255752563477, | |
| "rewards/rejected": -48.78778839111328, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.6279434850863422, | |
| "grad_norm": 10.993691880850108, | |
| "learning_rate": 1.8298130031671972e-07, | |
| "logits/chosen": 6008.68994140625, | |
| "logits/rejected": 4658.36669921875, | |
| "logps/chosen": -15.523478507995605, | |
| "logps/rejected": -42.00802230834961, | |
| "loss": 1.3645, | |
| "rewards/accuracies": 0.7500000596046448, | |
| "rewards/chosen": -14.784734725952148, | |
| "rewards/margins": 26.388931274414062, | |
| "rewards/rejected": -41.173667907714844, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.6357927786499215, | |
| "grad_norm": 9.806014319167142, | |
| "learning_rate": 1.7640476790784075e-07, | |
| "logits/chosen": 5616.755859375, | |
| "logits/rejected": 4236.2626953125, | |
| "logps/chosen": -10.850046157836914, | |
| "logps/rejected": -40.791954040527344, | |
| "loss": 1.3535, | |
| "rewards/accuracies": 0.7083333730697632, | |
| "rewards/chosen": -10.169736862182617, | |
| "rewards/margins": 29.735652923583984, | |
| "rewards/rejected": -39.905391693115234, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.6436420722135008, | |
| "grad_norm": 18.37377243229568, | |
| "learning_rate": 1.6988353904658492e-07, | |
| "logits/chosen": 6082.55322265625, | |
| "logits/rejected": 4100.5595703125, | |
| "logps/chosen": -12.024580001831055, | |
| "logps/rejected": -40.69676208496094, | |
| "loss": 1.3524, | |
| "rewards/accuracies": 0.8166667222976685, | |
| "rewards/chosen": -11.352849006652832, | |
| "rewards/margins": 28.577457427978516, | |
| "rewards/rejected": -39.93030548095703, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.6514913657770801, | |
| "grad_norm": 16.297609330074813, | |
| "learning_rate": 1.634225141467513e-07, | |
| "logits/chosen": 5812.82275390625, | |
| "logits/rejected": 4317.2626953125, | |
| "logps/chosen": -17.369579315185547, | |
| "logps/rejected": -48.95631408691406, | |
| "loss": 1.3462, | |
| "rewards/accuracies": 0.6750000715255737, | |
| "rewards/chosen": -16.705692291259766, | |
| "rewards/margins": 31.48634910583496, | |
| "rewards/rejected": -48.19204330444336, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.6593406593406593, | |
| "grad_norm": 9.301450718129091, | |
| "learning_rate": 1.570265483815364e-07, | |
| "logits/chosen": 6524.0634765625, | |
| "logits/rejected": 4608.86181640625, | |
| "logps/chosen": -14.519497871398926, | |
| "logps/rejected": -40.709632873535156, | |
| "loss": 1.3465, | |
| "rewards/accuracies": 0.6833333969116211, | |
| "rewards/chosen": -13.860455513000488, | |
| "rewards/margins": 26.043548583984375, | |
| "rewards/rejected": -39.90399932861328, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.6671899529042387, | |
| "grad_norm": 9.160256595515714, | |
| "learning_rate": 1.5070044803508691e-07, | |
| "logits/chosen": 6107.046875, | |
| "logits/rejected": 4694.4599609375, | |
| "logps/chosen": -8.90070915222168, | |
| "logps/rejected": -44.73848342895508, | |
| "loss": 1.3408, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -8.21817398071289, | |
| "rewards/margins": 35.75190353393555, | |
| "rewards/rejected": -43.97007751464844, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.6750392464678179, | |
| "grad_norm": 27.489286245121285, | |
| "learning_rate": 1.444489668907914e-07, | |
| "logits/chosen": 6378.09033203125, | |
| "logits/rejected": 4609.8115234375, | |
| "logps/chosen": -12.870359420776367, | |
| "logps/rejected": -47.970619201660156, | |
| "loss": 1.3578, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -12.144631385803223, | |
| "rewards/margins": 35.02986526489258, | |
| "rewards/rejected": -47.174495697021484, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.6828885400313972, | |
| "grad_norm": 9.941621288803526, | |
| "learning_rate": 1.3827680265902232e-07, | |
| "logits/chosen": 6576.61083984375, | |
| "logits/rejected": 4636.0556640625, | |
| "logps/chosen": -10.919339179992676, | |
| "logps/rejected": -39.724056243896484, | |
| "loss": 1.3491, | |
| "rewards/accuracies": 0.7666666507720947, | |
| "rewards/chosen": -10.253345489501953, | |
| "rewards/margins": 28.734447479248047, | |
| "rewards/rejected": -38.98779296875, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.6907378335949764, | |
| "grad_norm": 24.038360075630116, | |
| "learning_rate": 1.3218859344701632e-07, | |
| "logits/chosen": 5705.89697265625, | |
| "logits/rejected": 4821.3330078125, | |
| "logps/chosen": -12.567468643188477, | |
| "logps/rejected": -35.90740966796875, | |
| "loss": 1.3557, | |
| "rewards/accuracies": 0.7500000596046448, | |
| "rewards/chosen": -11.867032051086426, | |
| "rewards/margins": 23.217498779296875, | |
| "rewards/rejected": -35.084529876708984, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.6985871271585558, | |
| "grad_norm": 14.799566924171229, | |
| "learning_rate": 1.2618891427354172e-07, | |
| "logits/chosen": 6839.7822265625, | |
| "logits/rejected": 4843.2822265625, | |
| "logps/chosen": -10.830427169799805, | |
| "logps/rejected": -46.047332763671875, | |
| "loss": 1.3521, | |
| "rewards/accuracies": 0.7750000357627869, | |
| "rewards/chosen": -10.071599960327148, | |
| "rewards/margins": 35.167415618896484, | |
| "rewards/rejected": -45.239017486572266, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.706436420722135, | |
| "grad_norm": 12.683033226611311, | |
| "learning_rate": 1.202822736309758e-07, | |
| "logits/chosen": 5786.37744140625, | |
| "logits/rejected": 4521.42529296875, | |
| "logps/chosen": -10.186563491821289, | |
| "logps/rejected": -42.66360092163086, | |
| "loss": 1.3532, | |
| "rewards/accuracies": 0.7333333492279053, | |
| "rewards/chosen": -9.53927993774414, | |
| "rewards/margins": 32.29823303222656, | |
| "rewards/rejected": -41.83751678466797, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.7142857142857143, | |
| "grad_norm": 7.8112818904757555, | |
| "learning_rate": 1.1447311009737299e-07, | |
| "logits/chosen": 5534.1630859375, | |
| "logits/rejected": 4551.8251953125, | |
| "logps/chosen": -13.114629745483398, | |
| "logps/rejected": -46.98776626586914, | |
| "loss": 1.3477, | |
| "rewards/accuracies": 0.7416666746139526, | |
| "rewards/chosen": -12.424124717712402, | |
| "rewards/margins": 33.744571685791016, | |
| "rewards/rejected": -46.1686897277832, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.7221350078492935, | |
| "grad_norm": 13.48022639064254, | |
| "learning_rate": 1.0876578900107053e-07, | |
| "logits/chosen": 6257.46142578125, | |
| "logits/rejected": 4657.412109375, | |
| "logps/chosen": -9.951969146728516, | |
| "logps/rejected": -37.00797653198242, | |
| "loss": 1.3473, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -9.255838394165039, | |
| "rewards/margins": 26.91226577758789, | |
| "rewards/rejected": -36.1681022644043, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.7299843014128728, | |
| "grad_norm": 9.203666814948612, | |
| "learning_rate": 1.0316459914033793e-07, | |
| "logits/chosen": 6127.88720703125, | |
| "logits/rejected": 3711.983642578125, | |
| "logps/chosen": -14.189971923828125, | |
| "logps/rejected": -48.925025939941406, | |
| "loss": 1.3439, | |
| "rewards/accuracies": 0.7583333253860474, | |
| "rewards/chosen": -13.459379196166992, | |
| "rewards/margins": 34.6566047668457, | |
| "rewards/rejected": -48.1159782409668, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.7378335949764521, | |
| "grad_norm": 10.697737816990417, | |
| "learning_rate": 9.767374956053584e-08, | |
| "logits/chosen": 5963.5029296875, | |
| "logits/rejected": 4586.4345703125, | |
| "logps/chosen": -11.35425090789795, | |
| "logps/rejected": -51.100425720214844, | |
| "loss": 1.3422, | |
| "rewards/accuracies": 0.6916666626930237, | |
| "rewards/chosen": -10.636934280395508, | |
| "rewards/margins": 39.55208206176758, | |
| "rewards/rejected": -50.18900680541992, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.7456828885400314, | |
| "grad_norm": 11.783318339337354, | |
| "learning_rate": 9.229736639120561e-08, | |
| "logits/chosen": 6086.76025390625, | |
| "logits/rejected": 5080.07421875, | |
| "logps/chosen": -12.311005592346191, | |
| "logps/rejected": -28.69287109375, | |
| "loss": 1.3568, | |
| "rewards/accuracies": 0.6833332777023315, | |
| "rewards/chosen": -11.60954475402832, | |
| "rewards/margins": 16.351043701171875, | |
| "rewards/rejected": -27.960586547851562, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.7535321821036107, | |
| "grad_norm": 10.637109859687323, | |
| "learning_rate": 8.70394897454659e-08, | |
| "logits/chosen": 6191.0126953125, | |
| "logits/rejected": 4793.7890625, | |
| "logps/chosen": -6.652519226074219, | |
| "logps/rejected": -37.2434196472168, | |
| "loss": 1.339, | |
| "rewards/accuracies": 0.7166666984558105, | |
| "rewards/chosen": -5.941922187805176, | |
| "rewards/margins": 30.51517677307129, | |
| "rewards/rejected": -36.457096099853516, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.7613814756671899, | |
| "grad_norm": 21.916390858965105, | |
| "learning_rate": 8.19040706840472e-08, | |
| "logits/chosen": 6167.099609375, | |
| "logits/rejected": 4380.296875, | |
| "logps/chosen": -9.31376838684082, | |
| "logps/rejected": -49.45098114013672, | |
| "loss": 1.3427, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -8.63332748413086, | |
| "rewards/margins": 39.97867965698242, | |
| "rewards/rejected": -48.61200714111328, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 15.489524399549873, | |
| "learning_rate": 7.689496824624525e-08, | |
| "logits/chosen": 5892.37158203125, | |
| "logits/rejected": 3848.999267578125, | |
| "logps/chosen": -10.760286331176758, | |
| "logps/rejected": -56.49827194213867, | |
| "loss": 1.3421, | |
| "rewards/accuracies": 0.7916666269302368, | |
| "rewards/chosen": -10.0794677734375, | |
| "rewards/margins": 45.537147521972656, | |
| "rewards/rejected": -55.616615295410156, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.7770800627943485, | |
| "grad_norm": 11.83089798223724, | |
| "learning_rate": 7.201594655002458e-08, | |
| "logits/chosen": 6142.3076171875, | |
| "logits/rejected": 4396.3681640625, | |
| "logps/chosen": -11.698305130004883, | |
| "logps/rejected": -52.2495002746582, | |
| "loss": 1.3418, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -11.031845092773438, | |
| "rewards/margins": 40.371498107910156, | |
| "rewards/rejected": -51.403343200683594, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.7849293563579278, | |
| "grad_norm": 12.8555278251267, | |
| "learning_rate": 6.727067196345099e-08, | |
| "logits/chosen": 5779.787109375, | |
| "logits/rejected": 4448.8349609375, | |
| "logps/chosen": -12.997045516967773, | |
| "logps/rejected": -34.62881851196289, | |
| "loss": 1.3515, | |
| "rewards/accuracies": 0.6833333373069763, | |
| "rewards/chosen": -12.312833786010742, | |
| "rewards/margins": 21.579345703125, | |
| "rewards/rejected": -33.89217758178711, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.792778649921507, | |
| "grad_norm": 12.537447175753579, | |
| "learning_rate": 6.26627103495786e-08, | |
| "logits/chosen": 6107.9267578125, | |
| "logits/rejected": 4406.4130859375, | |
| "logps/chosen": -11.072089195251465, | |
| "logps/rejected": -43.155887603759766, | |
| "loss": 1.3516, | |
| "rewards/accuracies": 0.7083333134651184, | |
| "rewards/chosen": -10.391607284545898, | |
| "rewards/margins": 32.014122009277344, | |
| "rewards/rejected": -42.40572738647461, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.8006279434850864, | |
| "grad_norm": 6.910015837764057, | |
| "learning_rate": 5.8195524386862374e-08, | |
| "logits/chosen": 6251.3759765625, | |
| "logits/rejected": 4906.587890625, | |
| "logps/chosen": -8.422673225402832, | |
| "logps/rejected": -39.31324768066406, | |
| "loss": 1.3533, | |
| "rewards/accuracies": 0.7916667461395264, | |
| "rewards/chosen": -7.672184944152832, | |
| "rewards/margins": 30.805267333984375, | |
| "rewards/rejected": -38.477455139160156, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.8084772370486656, | |
| "grad_norm": 10.822080025305892, | |
| "learning_rate": 5.38724709671092e-08, | |
| "logits/chosen": 6651.43115234375, | |
| "logits/rejected": 5678.1181640625, | |
| "logps/chosen": -9.612025260925293, | |
| "logps/rejected": -36.233829498291016, | |
| "loss": 1.3491, | |
| "rewards/accuracies": 0.7500001192092896, | |
| "rewards/chosen": -8.941255569458008, | |
| "rewards/margins": 26.474712371826172, | |
| "rewards/rejected": -35.41596603393555, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.8163265306122449, | |
| "grad_norm": 8.94597165114862, | |
| "learning_rate": 4.969679867292276e-08, | |
| "logits/chosen": 5803.84375, | |
| "logits/rejected": 4507.89404296875, | |
| "logps/chosen": -11.683372497558594, | |
| "logps/rejected": -50.33116149902344, | |
| "loss": 1.3396, | |
| "rewards/accuracies": 0.7000000476837158, | |
| "rewards/chosen": -10.970834732055664, | |
| "rewards/margins": 38.524723052978516, | |
| "rewards/rejected": -49.49555587768555, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.8241758241758241, | |
| "grad_norm": 15.522644034350902, | |
| "learning_rate": 4.5671645336537416e-08, | |
| "logits/chosen": 5976.689453125, | |
| "logits/rejected": 4534.595703125, | |
| "logps/chosen": -11.027806282043457, | |
| "logps/rejected": -47.25383758544922, | |
| "loss": 1.3436, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -10.277358055114746, | |
| "rewards/margins": 36.20246505737305, | |
| "rewards/rejected": -46.479827880859375, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.8320251177394035, | |
| "grad_norm": 19.227519235685435, | |
| "learning_rate": 4.180003568187776e-08, | |
| "logits/chosen": 7209.6474609375, | |
| "logits/rejected": 5028.30029296875, | |
| "logps/chosen": -14.763595581054688, | |
| "logps/rejected": -42.880470275878906, | |
| "loss": 1.3622, | |
| "rewards/accuracies": 0.6583333611488342, | |
| "rewards/chosen": -14.041421890258789, | |
| "rewards/margins": 28.038188934326172, | |
| "rewards/rejected": -42.079612731933594, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.8398744113029827, | |
| "grad_norm": 11.191674738465105, | |
| "learning_rate": 3.8084879051612144e-08, | |
| "logits/chosen": 6095.42529296875, | |
| "logits/rejected": 4858.88671875, | |
| "logps/chosen": -8.86856460571289, | |
| "logps/rejected": -46.186073303222656, | |
| "loss": 1.3464, | |
| "rewards/accuracies": 0.6833333969116211, | |
| "rewards/chosen": -8.155261993408203, | |
| "rewards/margins": 37.2761116027832, | |
| "rewards/rejected": -45.431373596191406, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.847723704866562, | |
| "grad_norm": 7.0206188838060735, | |
| "learning_rate": 3.452896722091128e-08, | |
| "logits/chosen": 6701.546875, | |
| "logits/rejected": 4320.4267578125, | |
| "logps/chosen": -9.577341079711914, | |
| "logps/rejected": -46.688411712646484, | |
| "loss": 1.3479, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -8.867132186889648, | |
| "rewards/margins": 37.0556640625, | |
| "rewards/rejected": -45.922794342041016, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.8555729984301413, | |
| "grad_norm": 12.648028891337141, | |
| "learning_rate": 3.11349722995527e-08, | |
| "logits/chosen": 6607.6318359375, | |
| "logits/rejected": 4307.4677734375, | |
| "logps/chosen": -11.983565330505371, | |
| "logps/rejected": -44.108150482177734, | |
| "loss": 1.3466, | |
| "rewards/accuracies": 0.7416667342185974, | |
| "rewards/chosen": -11.318174362182617, | |
| "rewards/margins": 31.94852638244629, | |
| "rewards/rejected": -43.26670455932617, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.8634222919937206, | |
| "grad_norm": 16.16520843547699, | |
| "learning_rate": 2.7905444723949762e-08, | |
| "logits/chosen": 6445.07177734375, | |
| "logits/rejected": 4464.13427734375, | |
| "logps/chosen": -14.572372436523438, | |
| "logps/rejected": -57.9854850769043, | |
| "loss": 1.3444, | |
| "rewards/accuracies": 0.7750000357627869, | |
| "rewards/chosen": -13.854543685913086, | |
| "rewards/margins": 43.35486602783203, | |
| "rewards/rejected": -57.20940399169922, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.8712715855572999, | |
| "grad_norm": 12.082780970830518, | |
| "learning_rate": 2.484281134061142e-08, | |
| "logits/chosen": 6635.34619140625, | |
| "logits/rejected": 4545.72314453125, | |
| "logps/chosen": -15.10521411895752, | |
| "logps/rejected": -54.311370849609375, | |
| "loss": 1.3485, | |
| "rewards/accuracies": 0.7416666746139526, | |
| "rewards/chosen": -14.40710735321045, | |
| "rewards/margins": 39.0588264465332, | |
| "rewards/rejected": -53.4659309387207, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.8791208791208791, | |
| "grad_norm": 15.429026565586026, | |
| "learning_rate": 2.194937358247506e-08, | |
| "logits/chosen": 6658.29052734375, | |
| "logits/rejected": 4560.0166015625, | |
| "logps/chosen": -14.397059440612793, | |
| "logps/rejected": -51.39775466918945, | |
| "loss": 1.3472, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -13.709543228149414, | |
| "rewards/margins": 36.796878814697266, | |
| "rewards/rejected": -50.50641632080078, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.8869701726844584, | |
| "grad_norm": 19.175151198780238, | |
| "learning_rate": 1.9227305739481612e-08, | |
| "logits/chosen": 6183.2001953125, | |
| "logits/rejected": 3959.198486328125, | |
| "logps/chosen": -8.73318099975586, | |
| "logps/rejected": -47.819435119628906, | |
| "loss": 1.342, | |
| "rewards/accuracies": 0.7583333849906921, | |
| "rewards/chosen": -8.031963348388672, | |
| "rewards/margins": 38.94813919067383, | |
| "rewards/rejected": -46.9801025390625, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.8948194662480377, | |
| "grad_norm": 10.620817934538785, | |
| "learning_rate": 1.6678653324693787e-08, | |
| "logits/chosen": 6747.44287109375, | |
| "logits/rejected": 4794.2822265625, | |
| "logps/chosen": -10.389518737792969, | |
| "logps/rejected": -41.45861053466797, | |
| "loss": 1.3374, | |
| "rewards/accuracies": 0.7416666746139526, | |
| "rewards/chosen": -9.669957160949707, | |
| "rewards/margins": 30.935802459716797, | |
| "rewards/rejected": -40.60576629638672, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.902668759811617, | |
| "grad_norm": 10.754709061690225, | |
| "learning_rate": 1.4305331537183384e-08, | |
| "logits/chosen": 5942.2294921875, | |
| "logits/rejected": 4849.82177734375, | |
| "logps/chosen": -12.92119312286377, | |
| "logps/rejected": -41.3829345703125, | |
| "loss": 1.3461, | |
| "rewards/accuracies": 0.7083333134651184, | |
| "rewards/chosen": -12.190919876098633, | |
| "rewards/margins": 28.414255142211914, | |
| "rewards/rejected": -40.60517120361328, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.9105180533751962, | |
| "grad_norm": 13.557533071090745, | |
| "learning_rate": 1.2109123822844653e-08, | |
| "logits/chosen": 6006.47021484375, | |
| "logits/rejected": 4266.6396484375, | |
| "logps/chosen": -14.5009765625, | |
| "logps/rejected": -38.874107360839844, | |
| "loss": 1.3503, | |
| "rewards/accuracies": 0.7166666984558105, | |
| "rewards/chosen": -13.78251838684082, | |
| "rewards/margins": 24.258342742919922, | |
| "rewards/rejected": -38.040863037109375, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.9183673469387755, | |
| "grad_norm": 18.275601434548747, | |
| "learning_rate": 1.0091680534213387e-08, | |
| "logits/chosen": 6534.689453125, | |
| "logits/rejected": 5600.6259765625, | |
| "logps/chosen": -14.446474075317383, | |
| "logps/rejected": -39.4738655090332, | |
| "loss": 1.3455, | |
| "rewards/accuracies": 0.6333333253860474, | |
| "rewards/chosen": -13.775007247924805, | |
| "rewards/margins": 24.918426513671875, | |
| "rewards/rejected": -38.69343566894531, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.9262166405023547, | |
| "grad_norm": 18.235577733959495, | |
| "learning_rate": 8.254517690300944e-09, | |
| "logits/chosen": 5951.74609375, | |
| "logits/rejected": 4470.59814453125, | |
| "logps/chosen": -12.159133911132812, | |
| "logps/rejected": -45.7835807800293, | |
| "loss": 1.3548, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -11.42202091217041, | |
| "rewards/margins": 33.510128021240234, | |
| "rewards/rejected": -44.932151794433594, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.9340659340659341, | |
| "grad_norm": 10.49720602308573, | |
| "learning_rate": 6.599015837372907e-09, | |
| "logits/chosen": 6310.583984375, | |
| "logits/rejected": 4920.32861328125, | |
| "logps/chosen": -15.357757568359375, | |
| "logps/rejected": -46.599708557128906, | |
| "loss": 1.3385, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -14.61291217803955, | |
| "rewards/margins": 31.154321670532227, | |
| "rewards/rejected": -45.767234802246094, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.9419152276295133, | |
| "grad_norm": 11.862843010865799, | |
| "learning_rate": 5.126419011529992e-09, | |
| "logits/chosen": 6784.61865234375, | |
| "logits/rejected": 4886.77001953125, | |
| "logps/chosen": -10.652966499328613, | |
| "logps/rejected": -48.49315643310547, | |
| "loss": 1.3506, | |
| "rewards/accuracies": 0.7833333015441895, | |
| "rewards/chosen": -9.923705101013184, | |
| "rewards/margins": 37.731605529785156, | |
| "rewards/rejected": -47.65531539916992, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.9497645211930926, | |
| "grad_norm": 16.13028833676148, | |
| "learning_rate": 3.837833803870177e-09, | |
| "logits/chosen": 6215.4248046875, | |
| "logits/rejected": 4725.1298828125, | |
| "logps/chosen": -13.33996868133545, | |
| "logps/rejected": -44.714012145996094, | |
| "loss": 1.3548, | |
| "rewards/accuracies": 0.7583333253860474, | |
| "rewards/chosen": -12.622026443481445, | |
| "rewards/margins": 31.248937606811523, | |
| "rewards/rejected": -43.8709602355957, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.957613814756672, | |
| "grad_norm": 13.798152810624883, | |
| "learning_rate": 2.734228528934679e-09, | |
| "logits/chosen": 7684.4091796875, | |
| "logits/rejected": 4981.8173828125, | |
| "logps/chosen": -13.918993949890137, | |
| "logps/rejected": -47.229942321777344, | |
| "loss": 1.3582, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -13.145196914672852, | |
| "rewards/margins": 33.178897857666016, | |
| "rewards/rejected": -46.324092864990234, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.9654631083202512, | |
| "grad_norm": 10.813489628564165, | |
| "learning_rate": 1.8164324970625645e-09, | |
| "logits/chosen": 6866.75, | |
| "logits/rejected": 4787.5693359375, | |
| "logps/chosen": -11.430197715759277, | |
| "logps/rejected": -42.63923645019531, | |
| "loss": 1.3546, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -10.682231903076172, | |
| "rewards/margins": 31.16469383239746, | |
| "rewards/rejected": -41.846927642822266, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.9733124018838305, | |
| "grad_norm": 11.901517103335008, | |
| "learning_rate": 1.0851353912008642e-09, | |
| "logits/chosen": 5887.49462890625, | |
| "logits/rejected": 4629.6220703125, | |
| "logps/chosen": -13.619874954223633, | |
| "logps/rejected": -50.765167236328125, | |
| "loss": 1.34, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -12.862905502319336, | |
| "rewards/margins": 36.99613571166992, | |
| "rewards/rejected": -49.85904312133789, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.9811616954474097, | |
| "grad_norm": 11.112356440700552, | |
| "learning_rate": 5.408867486384471e-10, | |
| "logits/chosen": 6076.89453125, | |
| "logits/rejected": 4519.1474609375, | |
| "logps/chosen": -8.847349166870117, | |
| "logps/rejected": -37.14400863647461, | |
| "loss": 1.3532, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -8.169310569763184, | |
| "rewards/margins": 28.190332412719727, | |
| "rewards/rejected": -36.359649658203125, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.989010989010989, | |
| "grad_norm": 10.283870046812526, | |
| "learning_rate": 1.840955480532924e-10, | |
| "logits/chosen": 5829.6083984375, | |
| "logits/rejected": 4635.3076171875, | |
| "logps/chosen": -11.450944900512695, | |
| "logps/rejected": -42.468692779541016, | |
| "loss": 1.3448, | |
| "rewards/accuracies": 0.7666666507720947, | |
| "rewards/chosen": -10.698575019836426, | |
| "rewards/margins": 30.944873809814453, | |
| "rewards/rejected": -41.64345169067383, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.9968602825745683, | |
| "grad_norm": 14.458122344024726, | |
| "learning_rate": 1.502990218302247e-11, | |
| "logits/chosen": 5938.97802734375, | |
| "logits/rejected": 4195.7705078125, | |
| "logps/chosen": -12.192282676696777, | |
| "logps/rejected": -46.034873962402344, | |
| "loss": 1.3539, | |
| "rewards/accuracies": 0.6833333373069763, | |
| "rewards/chosen": -11.495050430297852, | |
| "rewards/margins": 33.76050567626953, | |
| "rewards/rejected": -45.255550384521484, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 1274, | |
| "total_flos": 0.0, | |
| "train_loss": 1.3571216397607346, | |
| "train_runtime": 13426.8072, | |
| "train_samples_per_second": 4.553, | |
| "train_steps_per_second": 0.095 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1274, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 2000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 6, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |