| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9997382884061764, | |
| "eval_steps": 500, | |
| "global_step": 955, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 5.208333333333333e-08, | |
| "logits/chosen": 0.654504120349884, | |
| "logits/rejected": 0.5926104187965393, | |
| "logps/chosen": -305.37469482421875, | |
| "logps/rejected": -235.55426025390625, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 5.208333333333334e-07, | |
| "logits/chosen": 0.7002869248390198, | |
| "logits/rejected": 0.7474254965782166, | |
| "logps/chosen": -312.4855041503906, | |
| "logps/rejected": -270.3090515136719, | |
| "loss": 0.6928, | |
| "rewards/accuracies": 0.4722222089767456, | |
| "rewards/chosen": 0.0012019119458273053, | |
| "rewards/margins": 0.0008854579646140337, | |
| "rewards/rejected": 0.00031645383569411933, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 1.0416666666666667e-06, | |
| "logits/chosen": 0.6109224557876587, | |
| "logits/rejected": 0.7636935114860535, | |
| "logps/chosen": -229.7316436767578, | |
| "logps/rejected": -217.67245483398438, | |
| "loss": 0.6928, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.0013416050933301449, | |
| "rewards/margins": 0.0012069231597706676, | |
| "rewards/rejected": 0.00013468178804032505, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.5625e-06, | |
| "logits/chosen": 0.6081576347351074, | |
| "logits/rejected": 0.65868079662323, | |
| "logps/chosen": -286.92681884765625, | |
| "logps/rejected": -243.32778930664062, | |
| "loss": 0.6929, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.00032840127823874354, | |
| "rewards/margins": 0.0011824481189250946, | |
| "rewards/rejected": -0.0008540467242710292, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 2.0833333333333334e-06, | |
| "logits/chosen": 0.616665244102478, | |
| "logits/rejected": 0.735335648059845, | |
| "logps/chosen": -291.98284912109375, | |
| "logps/rejected": -265.10198974609375, | |
| "loss": 0.6926, | |
| "rewards/accuracies": 0.4625000059604645, | |
| "rewards/chosen": -0.0013727883342653513, | |
| "rewards/margins": 0.0006782080745324492, | |
| "rewards/rejected": -0.0020509962923824787, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2.604166666666667e-06, | |
| "logits/chosen": 0.6475738286972046, | |
| "logits/rejected": 0.660605788230896, | |
| "logps/chosen": -276.0094299316406, | |
| "logps/rejected": -254.7918701171875, | |
| "loss": 0.6923, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -1.2021581824228633e-05, | |
| "rewards/margins": 0.0021027675829827785, | |
| "rewards/rejected": -0.002114789094775915, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 3.125e-06, | |
| "logits/chosen": 0.6682679057121277, | |
| "logits/rejected": 0.7323259115219116, | |
| "logps/chosen": -308.230224609375, | |
| "logps/rejected": -253.301025390625, | |
| "loss": 0.6917, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.0026237894780933857, | |
| "rewards/margins": 0.003060306655243039, | |
| "rewards/rejected": -0.005684095900505781, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 3.6458333333333333e-06, | |
| "logits/chosen": 0.6251315474510193, | |
| "logits/rejected": 0.6695118546485901, | |
| "logps/chosen": -296.9325866699219, | |
| "logps/rejected": -265.0625915527344, | |
| "loss": 0.6898, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.003425169037654996, | |
| "rewards/margins": 0.0062147388234734535, | |
| "rewards/rejected": -0.009639907628297806, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 4.166666666666667e-06, | |
| "logits/chosen": 0.6303352117538452, | |
| "logits/rejected": 0.6176969408988953, | |
| "logps/chosen": -296.875244140625, | |
| "logps/rejected": -259.2256774902344, | |
| "loss": 0.6875, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.011222435161471367, | |
| "rewards/margins": 0.009960432536900043, | |
| "rewards/rejected": -0.021182868629693985, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 4.6875000000000004e-06, | |
| "logits/chosen": 0.6014409065246582, | |
| "logits/rejected": 0.6135669946670532, | |
| "logps/chosen": -271.5030212402344, | |
| "logps/rejected": -246.84310913085938, | |
| "loss": 0.6859, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.019970973953604698, | |
| "rewards/margins": 0.015956712886691093, | |
| "rewards/rejected": -0.03592768684029579, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.9997324926814375e-06, | |
| "logits/chosen": 0.5811746716499329, | |
| "logits/rejected": 0.582471489906311, | |
| "logps/chosen": -257.50079345703125, | |
| "logps/rejected": -245.5070343017578, | |
| "loss": 0.6848, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.025636669248342514, | |
| "rewards/margins": 0.01930178329348564, | |
| "rewards/rejected": -0.044938452541828156, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 4.996723692767927e-06, | |
| "logits/chosen": 0.6834483742713928, | |
| "logits/rejected": 0.7622531056404114, | |
| "logps/chosen": -267.14837646484375, | |
| "logps/rejected": -244.15005493164062, | |
| "loss": 0.6816, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.042797841131687164, | |
| "rewards/margins": 0.023977002128958702, | |
| "rewards/rejected": -0.06677485257387161, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.9903757462135984e-06, | |
| "logits/chosen": 0.5502853393554688, | |
| "logits/rejected": 0.6681184768676758, | |
| "logps/chosen": -267.02532958984375, | |
| "logps/rejected": -238.77914428710938, | |
| "loss": 0.6792, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.05984887480735779, | |
| "rewards/margins": 0.038902174681425095, | |
| "rewards/rejected": -0.09875104576349258, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.980697142834315e-06, | |
| "logits/chosen": 0.5389119982719421, | |
| "logits/rejected": 0.6768661737442017, | |
| "logps/chosen": -308.8938903808594, | |
| "logps/rejected": -267.13861083984375, | |
| "loss": 0.6768, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.08925649523735046, | |
| "rewards/margins": 0.032300472259521484, | |
| "rewards/rejected": -0.12155697494745255, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 4.967700826904229e-06, | |
| "logits/chosen": 0.5784076452255249, | |
| "logits/rejected": 0.6090500950813293, | |
| "logps/chosen": -268.8511657714844, | |
| "logps/rejected": -252.6071319580078, | |
| "loss": 0.6733, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.12314645200967789, | |
| "rewards/margins": 0.0378914549946785, | |
| "rewards/rejected": -0.1610378921031952, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 4.951404179843963e-06, | |
| "logits/chosen": 0.6251226663589478, | |
| "logits/rejected": 0.6390718817710876, | |
| "logps/chosen": -277.3585510253906, | |
| "logps/rejected": -255.3101043701172, | |
| "loss": 0.6781, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -0.14656075835227966, | |
| "rewards/margins": 0.036974769085645676, | |
| "rewards/rejected": -0.18353553116321564, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 4.931828996974498e-06, | |
| "logits/chosen": 0.6863027811050415, | |
| "logits/rejected": 0.5657976865768433, | |
| "logps/chosen": -279.9803161621094, | |
| "logps/rejected": -273.37109375, | |
| "loss": 0.6634, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.22472751140594482, | |
| "rewards/margins": 0.05435310676693916, | |
| "rewards/rejected": -0.2790805995464325, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.909001458367867e-06, | |
| "logits/chosen": 0.6299469470977783, | |
| "logits/rejected": 0.656846284866333, | |
| "logps/chosen": -332.79541015625, | |
| "logps/rejected": -292.52850341796875, | |
| "loss": 0.6598, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.3016909658908844, | |
| "rewards/margins": 0.08058367669582367, | |
| "rewards/rejected": -0.38227468729019165, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.882952093833628e-06, | |
| "logits/chosen": 0.5225847363471985, | |
| "logits/rejected": 0.6088179349899292, | |
| "logps/chosen": -289.1798095703125, | |
| "logps/rejected": -311.8951110839844, | |
| "loss": 0.652, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.43765193223953247, | |
| "rewards/margins": 0.09067411720752716, | |
| "rewards/rejected": -0.5283260345458984, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.853715742087947e-06, | |
| "logits/chosen": 0.5354307293891907, | |
| "logits/rejected": 0.652572512626648, | |
| "logps/chosen": -359.98126220703125, | |
| "logps/rejected": -309.4668273925781, | |
| "loss": 0.6449, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.5696258544921875, | |
| "rewards/margins": 0.17413343489170074, | |
| "rewards/rejected": -0.7437592148780823, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.821331504159906e-06, | |
| "logits/chosen": 0.5104254484176636, | |
| "logits/rejected": 0.49786561727523804, | |
| "logps/chosen": -336.1935729980469, | |
| "logps/rejected": -350.4376220703125, | |
| "loss": 0.6426, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.7382306456565857, | |
| "rewards/margins": 0.15078046917915344, | |
| "rewards/rejected": -0.8890112042427063, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.7858426910973435e-06, | |
| "logits/chosen": 0.4257640838623047, | |
| "logits/rejected": 0.47642454504966736, | |
| "logps/chosen": -387.05242919921875, | |
| "logps/rejected": -345.6732177734375, | |
| "loss": 0.6381, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.8003188967704773, | |
| "rewards/margins": 0.19792547821998596, | |
| "rewards/rejected": -0.9982444047927856, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.747296766042161e-06, | |
| "logits/chosen": 0.4104679226875305, | |
| "logits/rejected": 0.4869573712348938, | |
| "logps/chosen": -348.92156982421875, | |
| "logps/rejected": -350.780029296875, | |
| "loss": 0.637, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.8022894859313965, | |
| "rewards/margins": 0.19004520773887634, | |
| "rewards/rejected": -0.9923346638679504, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 4.705745280752586e-06, | |
| "logits/chosen": 0.3983922600746155, | |
| "logits/rejected": 0.42386817932128906, | |
| "logps/chosen": -388.9286804199219, | |
| "logps/rejected": -352.74444580078125, | |
| "loss": 0.6239, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.8833669424057007, | |
| "rewards/margins": 0.20530752837657928, | |
| "rewards/rejected": -1.088674545288086, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 4.661243806657256e-06, | |
| "logits/chosen": 0.38030871748924255, | |
| "logits/rejected": 0.3976936936378479, | |
| "logps/chosen": -378.4253234863281, | |
| "logps/rejected": -352.67388916015625, | |
| "loss": 0.6149, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.9681804776191711, | |
| "rewards/margins": 0.27472421526908875, | |
| "rewards/rejected": -1.2429049015045166, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.613851860533367e-06, | |
| "logits/chosen": 0.29627329111099243, | |
| "logits/rejected": 0.30808278918266296, | |
| "logps/chosen": -370.7823791503906, | |
| "logps/rejected": -365.70843505859375, | |
| "loss": 0.6213, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.9412585496902466, | |
| "rewards/margins": 0.25268539786338806, | |
| "rewards/rejected": -1.193943977355957, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.563632824908252e-06, | |
| "logits/chosen": 0.22838203608989716, | |
| "logits/rejected": 0.2758544087409973, | |
| "logps/chosen": -370.5462341308594, | |
| "logps/rejected": -384.4304504394531, | |
| "loss": 0.6205, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -1.1226102113723755, | |
| "rewards/margins": 0.22477027773857117, | |
| "rewards/rejected": -1.347380518913269, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 4.510653863290871e-06, | |
| "logits/chosen": 0.16573426127433777, | |
| "logits/rejected": 0.24319115281105042, | |
| "logps/chosen": -432.40496826171875, | |
| "logps/rejected": -434.70465087890625, | |
| "loss": 0.6065, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -1.3656351566314697, | |
| "rewards/margins": 0.3202405571937561, | |
| "rewards/rejected": -1.6858758926391602, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.454985830346574e-06, | |
| "logits/chosen": 0.15168949961662292, | |
| "logits/rejected": 0.2517026662826538, | |
| "logps/chosen": -496.7152404785156, | |
| "logps/rejected": -481.1207580566406, | |
| "loss": 0.6244, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.8755683898925781, | |
| "rewards/margins": 0.3297354280948639, | |
| "rewards/rejected": -2.20530366897583, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 4.396703177135262e-06, | |
| "logits/chosen": 0.1856023520231247, | |
| "logits/rejected": 0.21857495605945587, | |
| "logps/chosen": -468.2583923339844, | |
| "logps/rejected": -422.20477294921875, | |
| "loss": 0.6089, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -1.88265061378479, | |
| "rewards/margins": 0.23467198014259338, | |
| "rewards/rejected": -2.1173224449157715, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 4.335883851539693e-06, | |
| "logits/chosen": 0.10853853076696396, | |
| "logits/rejected": 0.2359129637479782, | |
| "logps/chosen": -437.02569580078125, | |
| "logps/rejected": -435.40606689453125, | |
| "loss": 0.5821, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.7075822353363037, | |
| "rewards/margins": 0.393388569355011, | |
| "rewards/rejected": -2.10097074508667, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 4.2726091940171055e-06, | |
| "logits/chosen": 0.12634363770484924, | |
| "logits/rejected": 0.224656343460083, | |
| "logps/chosen": -436.81683349609375, | |
| "logps/rejected": -466.3431091308594, | |
| "loss": 0.5697, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.6430351734161377, | |
| "rewards/margins": 0.4527590870857239, | |
| "rewards/rejected": -2.095794200897217, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 4.206963828813555e-06, | |
| "logits/chosen": 0.1670975536108017, | |
| "logits/rejected": 0.20784397423267365, | |
| "logps/chosen": -454.04400634765625, | |
| "logps/rejected": -471.0421447753906, | |
| "loss": 0.5959, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.7891740798950195, | |
| "rewards/margins": 0.4047388434410095, | |
| "rewards/rejected": -2.193913221359253, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 4.139035550786495e-06, | |
| "logits/chosen": 0.13126584887504578, | |
| "logits/rejected": 0.1421881467103958, | |
| "logps/chosen": -436.185546875, | |
| "logps/rejected": -425.6346740722656, | |
| "loss": 0.621, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -1.7343782186508179, | |
| "rewards/margins": 0.29571670293807983, | |
| "rewards/rejected": -2.030094861984253, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 4.068915207986931e-06, | |
| "logits/chosen": 0.18622621893882751, | |
| "logits/rejected": 0.2519204020500183, | |
| "logps/chosen": -453.08856201171875, | |
| "logps/rejected": -439.3407287597656, | |
| "loss": 0.63, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -1.8599331378936768, | |
| "rewards/margins": 0.25747692584991455, | |
| "rewards/rejected": -2.117410182952881, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 3.996696580158211e-06, | |
| "logits/chosen": 0.15227162837982178, | |
| "logits/rejected": 0.23813900351524353, | |
| "logps/chosen": -502.65252685546875, | |
| "logps/rejected": -521.7779541015625, | |
| "loss": 0.5979, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.993268609046936, | |
| "rewards/margins": 0.44951528310775757, | |
| "rewards/rejected": -2.442783832550049, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.922476253313921e-06, | |
| "logits/chosen": 0.07589183002710342, | |
| "logits/rejected": 0.1402803361415863, | |
| "logps/chosen": -492.88824462890625, | |
| "logps/rejected": -508.8631286621094, | |
| "loss": 0.5971, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -2.206943988800049, | |
| "rewards/margins": 0.2886379361152649, | |
| "rewards/rejected": -2.495581865310669, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 3.846353490562664e-06, | |
| "logits/chosen": 0.20907124876976013, | |
| "logits/rejected": 0.17944853007793427, | |
| "logps/chosen": -460.7980041503906, | |
| "logps/rejected": -506.36151123046875, | |
| "loss": 0.5709, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.9685118198394775, | |
| "rewards/margins": 0.47328275442123413, | |
| "rewards/rejected": -2.4417946338653564, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 3.768430099352445e-06, | |
| "logits/chosen": 0.16458635032176971, | |
| "logits/rejected": 0.22938723862171173, | |
| "logps/chosen": -461.98712158203125, | |
| "logps/rejected": -475.66046142578125, | |
| "loss": 0.5929, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.9674451351165771, | |
| "rewards/margins": 0.38506731390953064, | |
| "rewards/rejected": -2.3525123596191406, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 3.6888102953122307e-06, | |
| "logits/chosen": 0.06095918267965317, | |
| "logits/rejected": 0.1491270363330841, | |
| "logps/chosen": -450.43798828125, | |
| "logps/rejected": -455.1705017089844, | |
| "loss": 0.6117, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.9474515914916992, | |
| "rewards/margins": 0.4293443560600281, | |
| "rewards/rejected": -2.376796007156372, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.607600562872785e-06, | |
| "logits/chosen": 0.09890508651733398, | |
| "logits/rejected": 0.1348509043455124, | |
| "logps/chosen": -522.7869873046875, | |
| "logps/rejected": -504.12469482421875, | |
| "loss": 0.5751, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -2.1772594451904297, | |
| "rewards/margins": 0.4235813617706299, | |
| "rewards/rejected": -2.6008408069610596, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 3.5249095128531863e-06, | |
| "logits/chosen": 0.15617763996124268, | |
| "logits/rejected": 0.2346019446849823, | |
| "logps/chosen": -525.8595581054688, | |
| "logps/rejected": -515.14599609375, | |
| "loss": 0.6036, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -2.276054859161377, | |
| "rewards/margins": 0.37463003396987915, | |
| "rewards/rejected": -2.6506848335266113, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 3.4408477372034743e-06, | |
| "logits/chosen": 0.057582952082157135, | |
| "logits/rejected": 0.13709139823913574, | |
| "logps/chosen": -447.7400817871094, | |
| "logps/rejected": -488.5953063964844, | |
| "loss": 0.5871, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -2.0702292919158936, | |
| "rewards/margins": 0.3465236723423004, | |
| "rewards/rejected": -2.4167532920837402, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 3.355527661097728e-06, | |
| "logits/chosen": 0.10504597425460815, | |
| "logits/rejected": 0.18081213533878326, | |
| "logps/chosen": -476.1048278808594, | |
| "logps/rejected": -492.5294494628906, | |
| "loss": 0.6166, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -2.1964662075042725, | |
| "rewards/margins": 0.29873451590538025, | |
| "rewards/rejected": -2.4952008724212646, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 3.269063392575352e-06, | |
| "logits/chosen": 0.11929740011692047, | |
| "logits/rejected": 0.15031102299690247, | |
| "logps/chosen": -505.28363037109375, | |
| "logps/rejected": -516.3770751953125, | |
| "loss": 0.6019, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -2.265862226486206, | |
| "rewards/margins": 0.365284264087677, | |
| "rewards/rejected": -2.6311464309692383, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 3.181570569931697e-06, | |
| "logits/chosen": 0.10959301143884659, | |
| "logits/rejected": 0.15382876992225647, | |
| "logps/chosen": -448.1092224121094, | |
| "logps/rejected": -466.357421875, | |
| "loss": 0.6033, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -2.1950180530548096, | |
| "rewards/margins": 0.3157437741756439, | |
| "rewards/rejected": -2.5107619762420654, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 3.09316620706208e-06, | |
| "logits/chosen": 0.0773458257317543, | |
| "logits/rejected": 0.18486995995044708, | |
| "logps/chosen": -508.8724670410156, | |
| "logps/rejected": -497.22882080078125, | |
| "loss": 0.585, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -2.2086341381073, | |
| "rewards/margins": 0.32577189803123474, | |
| "rewards/rejected": -2.5344061851501465, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 3.0039685369660785e-06, | |
| "logits/chosen": 0.09328463673591614, | |
| "logits/rejected": 0.20468135178089142, | |
| "logps/chosen": -522.9307861328125, | |
| "logps/rejected": -512.7855834960938, | |
| "loss": 0.5807, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -2.516700506210327, | |
| "rewards/margins": 0.32140296697616577, | |
| "rewards/rejected": -2.8381035327911377, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.91409685362137e-06, | |
| "logits/chosen": 0.1576906144618988, | |
| "logits/rejected": 0.10724158585071564, | |
| "logps/chosen": -456.68359375, | |
| "logps/rejected": -451.1465759277344, | |
| "loss": 0.6305, | |
| "rewards/accuracies": 0.44999998807907104, | |
| "rewards/chosen": -2.26971697807312, | |
| "rewards/margins": 0.18658225238323212, | |
| "rewards/rejected": -2.456299304962158, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 2.8236713524386085e-06, | |
| "logits/chosen": 0.1447945386171341, | |
| "logits/rejected": 0.1760699599981308, | |
| "logps/chosen": -467.67889404296875, | |
| "logps/rejected": -484.5015563964844, | |
| "loss": 0.5847, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -2.1700217723846436, | |
| "rewards/margins": 0.4066441059112549, | |
| "rewards/rejected": -2.5766658782958984, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 2.7328129695107205e-06, | |
| "logits/chosen": 0.09771154820919037, | |
| "logits/rejected": 0.21410468220710754, | |
| "logps/chosen": -505.1236877441406, | |
| "logps/rejected": -521.1968994140625, | |
| "loss": 0.5852, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -2.445769786834717, | |
| "rewards/margins": 0.5605564117431641, | |
| "rewards/rejected": -3.00632643699646, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 2.641643219871597e-06, | |
| "logits/chosen": 0.0740189403295517, | |
| "logits/rejected": 0.10937446355819702, | |
| "logps/chosen": -575.5813598632812, | |
| "logps/rejected": -578.539306640625, | |
| "loss": 0.5878, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -3.042142152786255, | |
| "rewards/margins": 0.41739901900291443, | |
| "rewards/rejected": -3.4595413208007812, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.5502840349805074e-06, | |
| "logits/chosen": 0.13737405836582184, | |
| "logits/rejected": 0.1366177350282669, | |
| "logps/chosen": -577.433837890625, | |
| "logps/rejected": -586.6515502929688, | |
| "loss": 0.6018, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -2.9127590656280518, | |
| "rewards/margins": 0.4386274814605713, | |
| "rewards/rejected": -3.351386308670044, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 2.4588575996495797e-06, | |
| "logits/chosen": 0.009027734398841858, | |
| "logits/rejected": 0.15047816932201385, | |
| "logps/chosen": -578.1231689453125, | |
| "logps/rejected": -593.139404296875, | |
| "loss": 0.584, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -2.788166046142578, | |
| "rewards/margins": 0.4355267882347107, | |
| "rewards/rejected": -3.2236924171447754, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 2.367486188632446e-06, | |
| "logits/chosen": 0.12746676802635193, | |
| "logits/rejected": 0.15015694499015808, | |
| "logps/chosen": -549.587890625, | |
| "logps/rejected": -567.0057373046875, | |
| "loss": 0.5919, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -2.6341264247894287, | |
| "rewards/margins": 0.4422709047794342, | |
| "rewards/rejected": -3.07639741897583, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2.276292003092593e-06, | |
| "logits/chosen": 0.06688917428255081, | |
| "logits/rejected": 0.1280103623867035, | |
| "logps/chosen": -534.5496215820312, | |
| "logps/rejected": -526.1148071289062, | |
| "loss": 0.6045, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -2.5599446296691895, | |
| "rewards/margins": 0.3815520107746124, | |
| "rewards/rejected": -2.9414966106414795, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 2.1853970071701415e-06, | |
| "logits/chosen": 0.05697401612997055, | |
| "logits/rejected": 0.15568815171718597, | |
| "logps/chosen": -523.0081787109375, | |
| "logps/rejected": -505.5555114746094, | |
| "loss": 0.5914, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -2.705841541290283, | |
| "rewards/margins": 0.403696209192276, | |
| "rewards/rejected": -3.1095376014709473, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 2.0949227648656194e-06, | |
| "logits/chosen": 0.08012814819812775, | |
| "logits/rejected": 0.1830509603023529, | |
| "logps/chosen": -562.3424682617188, | |
| "logps/rejected": -572.2830200195312, | |
| "loss": 0.6085, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -2.9859776496887207, | |
| "rewards/margins": 0.37897008657455444, | |
| "rewards/rejected": -3.364948272705078, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 2.00499027745888e-06, | |
| "logits/chosen": 0.13961026072502136, | |
| "logits/rejected": 0.18276450037956238, | |
| "logps/chosen": -562.8177490234375, | |
| "logps/rejected": -579.6517333984375, | |
| "loss": 0.6114, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -2.9555232524871826, | |
| "rewards/margins": 0.380531370639801, | |
| "rewards/rejected": -3.336055040359497, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 1.915719821680624e-06, | |
| "logits/chosen": 0.10515383630990982, | |
| "logits/rejected": 0.12738659977912903, | |
| "logps/chosen": -508.9317932128906, | |
| "logps/rejected": -502.7018127441406, | |
| "loss": 0.5984, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -2.51983904838562, | |
| "rewards/margins": 0.23314905166625977, | |
| "rewards/rejected": -2.752988338470459, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 1.8272307888529276e-06, | |
| "logits/chosen": 0.1120181530714035, | |
| "logits/rejected": 0.18100281059741974, | |
| "logps/chosen": -511.30645751953125, | |
| "logps/rejected": -542.5186767578125, | |
| "loss": 0.5837, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -2.2869346141815186, | |
| "rewards/margins": 0.4697488248348236, | |
| "rewards/rejected": -2.756683111190796, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 1.739641525213929e-06, | |
| "logits/chosen": 0.13541504740715027, | |
| "logits/rejected": 0.1992477923631668, | |
| "logps/chosen": -526.412841796875, | |
| "logps/rejected": -498.51318359375, | |
| "loss": 0.5803, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -2.441995143890381, | |
| "rewards/margins": 0.4095155596733093, | |
| "rewards/rejected": -2.851510524749756, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.6530691736402317e-06, | |
| "logits/chosen": 0.11898718029260635, | |
| "logits/rejected": 0.11196433007717133, | |
| "logps/chosen": -530.7737426757812, | |
| "logps/rejected": -540.7352294921875, | |
| "loss": 0.5947, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -2.539415121078491, | |
| "rewards/margins": 0.4597281813621521, | |
| "rewards/rejected": -2.999143362045288, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 1.5676295169786864e-06, | |
| "logits/chosen": 0.12535694241523743, | |
| "logits/rejected": 0.2773950397968292, | |
| "logps/chosen": -558.296630859375, | |
| "logps/rejected": -556.3736572265625, | |
| "loss": 0.5937, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -2.703909158706665, | |
| "rewards/margins": 0.4821166396141052, | |
| "rewards/rejected": -3.186025857925415, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.4834368231970922e-06, | |
| "logits/chosen": 0.13473114371299744, | |
| "logits/rejected": 0.18143931031227112, | |
| "logps/chosen": -497.7254333496094, | |
| "logps/rejected": -533.7047119140625, | |
| "loss": 0.5819, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -2.562814235687256, | |
| "rewards/margins": 0.5369300842285156, | |
| "rewards/rejected": -3.0997443199157715, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 1.4006036925609245e-06, | |
| "logits/chosen": 0.051880162209272385, | |
| "logits/rejected": 0.20864169299602509, | |
| "logps/chosen": -551.8402099609375, | |
| "logps/rejected": -567.7971801757812, | |
| "loss": 0.5719, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -2.570906162261963, | |
| "rewards/margins": 0.5435839295387268, | |
| "rewards/rejected": -3.114490270614624, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.3192409070404582e-06, | |
| "logits/chosen": 0.12426358461380005, | |
| "logits/rejected": 0.08786197006702423, | |
| "logps/chosen": -543.6322021484375, | |
| "logps/rejected": -535.1307373046875, | |
| "loss": 0.596, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -2.673713445663452, | |
| "rewards/margins": 0.34852334856987, | |
| "rewards/rejected": -3.0222370624542236, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.2394572821496953e-06, | |
| "logits/chosen": 0.07642360031604767, | |
| "logits/rejected": 0.1105254516005516, | |
| "logps/chosen": -521.8109130859375, | |
| "logps/rejected": -533.8194580078125, | |
| "loss": 0.5713, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -2.5358338356018066, | |
| "rewards/margins": 0.5434619188308716, | |
| "rewards/rejected": -3.0792956352233887, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 1.1613595214152713e-06, | |
| "logits/chosen": 0.07207540422677994, | |
| "logits/rejected": 0.1519361436367035, | |
| "logps/chosen": -538.2649536132812, | |
| "logps/rejected": -527.5178833007812, | |
| "loss": 0.6009, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -2.5930490493774414, | |
| "rewards/margins": 0.24494171142578125, | |
| "rewards/rejected": -2.8379909992218018, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 1.0850520736699362e-06, | |
| "logits/chosen": 0.09551219642162323, | |
| "logits/rejected": 0.20561787486076355, | |
| "logps/chosen": -522.4210815429688, | |
| "logps/rejected": -531.2017822265625, | |
| "loss": 0.5918, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -2.5857510566711426, | |
| "rewards/margins": 0.47004079818725586, | |
| "rewards/rejected": -3.0557916164398193, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 1.0106369933615043e-06, | |
| "logits/chosen": 0.0834534615278244, | |
| "logits/rejected": 0.27686116099357605, | |
| "logps/chosen": -509.4612731933594, | |
| "logps/rejected": -507.83258056640625, | |
| "loss": 0.5759, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -2.489625930786133, | |
| "rewards/margins": 0.5027724504470825, | |
| "rewards/rejected": -2.992398262023926, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 9.382138040640714e-07, | |
| "logits/chosen": 0.08181905746459961, | |
| "logits/rejected": 0.10011023283004761, | |
| "logps/chosen": -566.5830078125, | |
| "logps/rejected": -552.9130249023438, | |
| "loss": 0.5783, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -2.718501091003418, | |
| "rewards/margins": 0.40024662017822266, | |
| "rewards/rejected": -3.1187477111816406, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 8.678793653740633e-07, | |
| "logits/chosen": 0.0735187977552414, | |
| "logits/rejected": 0.13505037128925323, | |
| "logps/chosen": -513.2781372070312, | |
| "logps/rejected": -523.3275146484375, | |
| "loss": 0.5936, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -2.5609078407287598, | |
| "rewards/margins": 0.3891633450984955, | |
| "rewards/rejected": -2.9500718116760254, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 7.997277433690984e-07, | |
| "logits/chosen": 0.06112724542617798, | |
| "logits/rejected": 0.0807592123746872, | |
| "logps/chosen": -558.6268310546875, | |
| "logps/rejected": -531.7501220703125, | |
| "loss": 0.5879, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -2.575448989868164, | |
| "rewards/margins": 0.3756105601787567, | |
| "rewards/rejected": -2.951059341430664, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 7.338500848029603e-07, | |
| "logits/chosen": 0.005916979629546404, | |
| "logits/rejected": 0.17464521527290344, | |
| "logps/chosen": -569.9360961914062, | |
| "logps/rejected": -577.1893310546875, | |
| "loss": 0.5867, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -2.632016181945801, | |
| "rewards/margins": 0.4428483545780182, | |
| "rewards/rejected": -3.074864387512207, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 6.70334495204884e-07, | |
| "logits/chosen": 0.1314985156059265, | |
| "logits/rejected": 0.16822925209999084, | |
| "logps/chosen": -500.46636962890625, | |
| "logps/rejected": -552.1305541992188, | |
| "loss": 0.5901, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -2.497035503387451, | |
| "rewards/margins": 0.5148124694824219, | |
| "rewards/rejected": -3.011847972869873, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 6.092659210462232e-07, | |
| "logits/chosen": 0.052326373755931854, | |
| "logits/rejected": 0.16787569224834442, | |
| "logps/chosen": -557.7818603515625, | |
| "logps/rejected": -567.0888061523438, | |
| "loss": 0.5689, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -2.623483657836914, | |
| "rewards/margins": 0.46536707878112793, | |
| "rewards/rejected": -3.088850736618042, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 5.507260361320738e-07, | |
| "logits/chosen": 0.08203858882188797, | |
| "logits/rejected": 0.14716270565986633, | |
| "logps/chosen": -574.7208862304688, | |
| "logps/rejected": -586.5050659179688, | |
| "loss": 0.5999, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -2.7186100482940674, | |
| "rewards/margins": 0.42362919449806213, | |
| "rewards/rejected": -3.1422390937805176, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 4.947931323697983e-07, | |
| "logits/chosen": 0.10785499960184097, | |
| "logits/rejected": 0.18369507789611816, | |
| "logps/chosen": -542.8053588867188, | |
| "logps/rejected": -549.775146484375, | |
| "loss": 0.5772, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -2.621422529220581, | |
| "rewards/margins": 0.4870278239250183, | |
| "rewards/rejected": -3.108450412750244, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 4.4154201506053985e-07, | |
| "logits/chosen": 0.11180669069290161, | |
| "logits/rejected": 0.14754420518875122, | |
| "logps/chosen": -532.26123046875, | |
| "logps/rejected": -550.2486572265625, | |
| "loss": 0.6016, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -2.573822021484375, | |
| "rewards/margins": 0.3509921431541443, | |
| "rewards/rejected": -2.924813985824585, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 3.910439028537638e-07, | |
| "logits/chosen": 0.10717664659023285, | |
| "logits/rejected": 0.24534320831298828, | |
| "logps/chosen": -510.2649841308594, | |
| "logps/rejected": -519.7151489257812, | |
| "loss": 0.562, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -2.3766260147094727, | |
| "rewards/margins": 0.4747850000858307, | |
| "rewards/rejected": -2.8514111042022705, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 3.4336633249862084e-07, | |
| "logits/chosen": 0.07645310461521149, | |
| "logits/rejected": 0.17721445858478546, | |
| "logps/chosen": -552.8776245117188, | |
| "logps/rejected": -514.6980590820312, | |
| "loss": 0.5707, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -2.620849847793579, | |
| "rewards/margins": 0.33325469493865967, | |
| "rewards/rejected": -2.95410418510437, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 2.98573068519539e-07, | |
| "logits/chosen": 0.04426421970129013, | |
| "logits/rejected": 0.21165044605731964, | |
| "logps/chosen": -541.8455810546875, | |
| "logps/rejected": -530.4864501953125, | |
| "loss": 0.5871, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -2.5347847938537598, | |
| "rewards/margins": 0.558411180973053, | |
| "rewards/rejected": -3.093196153640747, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 2.5672401793681854e-07, | |
| "logits/chosen": 0.11617372184991837, | |
| "logits/rejected": 0.16498759388923645, | |
| "logps/chosen": -526.9368286132812, | |
| "logps/rejected": -577.3372802734375, | |
| "loss": 0.568, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -2.5850746631622314, | |
| "rewards/margins": 0.4706156253814697, | |
| "rewards/rejected": -3.055690050125122, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 2.178751501463036e-07, | |
| "logits/chosen": 0.06383179128170013, | |
| "logits/rejected": 0.056464362889528275, | |
| "logps/chosen": -508.4523010253906, | |
| "logps/rejected": -521.2445068359375, | |
| "loss": 0.6203, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -2.663949728012085, | |
| "rewards/margins": 0.27986857295036316, | |
| "rewards/rejected": -2.9438185691833496, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 1.820784220652766e-07, | |
| "logits/chosen": 0.09964722394943237, | |
| "logits/rejected": 0.19454631209373474, | |
| "logps/chosen": -553.30517578125, | |
| "logps/rejected": -551.3551635742188, | |
| "loss": 0.5899, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -2.6559367179870605, | |
| "rewards/margins": 0.4735372066497803, | |
| "rewards/rejected": -3.12947416305542, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.4938170864468636e-07, | |
| "logits/chosen": 0.05580083280801773, | |
| "logits/rejected": 0.1297743022441864, | |
| "logps/chosen": -535.3272705078125, | |
| "logps/rejected": -541.6774291992188, | |
| "loss": 0.6136, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -2.696950912475586, | |
| "rewards/margins": 0.3545316159725189, | |
| "rewards/rejected": -3.0514824390411377, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 1.1982873884064466e-07, | |
| "logits/chosen": 0.04851340129971504, | |
| "logits/rejected": 0.13393761217594147, | |
| "logps/chosen": -480.47125244140625, | |
| "logps/rejected": -523.5387573242188, | |
| "loss": 0.592, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -2.486987590789795, | |
| "rewards/margins": 0.4138285219669342, | |
| "rewards/rejected": -2.900815963745117, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 9.345903713082305e-08, | |
| "logits/chosen": 0.04031291604042053, | |
| "logits/rejected": 0.15236200392246246, | |
| "logps/chosen": -546.6204833984375, | |
| "logps/rejected": -551.7069702148438, | |
| "loss": 0.6008, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -2.5852303504943848, | |
| "rewards/margins": 0.3579770624637604, | |
| "rewards/rejected": -2.943207263946533, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 7.030787065396866e-08, | |
| "logits/chosen": 0.13291895389556885, | |
| "logits/rejected": 0.15924690663814545, | |
| "logps/chosen": -508.44775390625, | |
| "logps/rejected": -528.5171508789062, | |
| "loss": 0.6118, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -2.3992390632629395, | |
| "rewards/margins": 0.3830375373363495, | |
| "rewards/rejected": -2.7822766304016113, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 5.0406202043228604e-08, | |
| "logits/chosen": 0.09626724570989609, | |
| "logits/rejected": 0.17059974372386932, | |
| "logps/chosen": -527.4398193359375, | |
| "logps/rejected": -531.8056640625, | |
| "loss": 0.6108, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -2.479921817779541, | |
| "rewards/margins": 0.4598274827003479, | |
| "rewards/rejected": -2.9397494792938232, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 3.378064801637687e-08, | |
| "logits/chosen": 0.02929285541176796, | |
| "logits/rejected": 0.18183095753192902, | |
| "logps/chosen": -556.4517822265625, | |
| "logps/rejected": -552.9277954101562, | |
| "loss": 0.5773, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -2.6010327339172363, | |
| "rewards/margins": 0.4194592535495758, | |
| "rewards/rejected": -3.0204920768737793, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2.0453443778310766e-08, | |
| "logits/chosen": 0.1071668416261673, | |
| "logits/rejected": 0.20206472277641296, | |
| "logps/chosen": -557.05419921875, | |
| "logps/rejected": -545.5590209960938, | |
| "loss": 0.5713, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -2.5374794006347656, | |
| "rewards/margins": 0.5199281573295593, | |
| "rewards/rejected": -3.057407855987549, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 1.0442413283435759e-08, | |
| "logits/chosen": 0.07009036839008331, | |
| "logits/rejected": 0.16552026569843292, | |
| "logps/chosen": -491.12237548828125, | |
| "logps/rejected": -496.7190856933594, | |
| "loss": 0.569, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -2.3404624462127686, | |
| "rewards/margins": 0.5322117209434509, | |
| "rewards/rejected": -2.8726742267608643, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 3.760945397705828e-09, | |
| "logits/chosen": 0.007874694652855396, | |
| "logits/rejected": 0.0997551903128624, | |
| "logps/chosen": -552.2734985351562, | |
| "logps/rejected": -549.2908935546875, | |
| "loss": 0.5796, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -2.5978238582611084, | |
| "rewards/margins": 0.399721622467041, | |
| "rewards/rejected": -2.9975457191467285, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 4.1797599220405605e-10, | |
| "logits/chosen": 0.1123199313879013, | |
| "logits/rejected": 0.08204374462366104, | |
| "logps/chosen": -510.1936950683594, | |
| "logps/rejected": -531.1867065429688, | |
| "loss": 0.5859, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -2.560084819793701, | |
| "rewards/margins": 0.4064672589302063, | |
| "rewards/rejected": -2.9665520191192627, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 955, | |
| "total_flos": 0.0, | |
| "train_loss": 0.0333975282639109, | |
| "train_runtime": 873.8284, | |
| "train_samples_per_second": 69.962, | |
| "train_steps_per_second": 1.093 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 955, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "total_flos": 0.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |