| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 625, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 7.936507936507937e-08, | |
| "logits/chosen": 0.1349913775920868, | |
| "logits/rejected": 0.1167084276676178, | |
| "logps/chosen": -428.1781005859375, | |
| "logps/rejected": -484.54522705078125, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 7.936507936507937e-07, | |
| "logits/chosen": 0.2796071171760559, | |
| "logits/rejected": 0.24151457846164703, | |
| "logps/chosen": -499.6032409667969, | |
| "logps/rejected": -473.9646301269531, | |
| "loss": 0.6932, | |
| "rewards/accuracies": 0.4027777910232544, | |
| "rewards/chosen": 2.6125486328965053e-05, | |
| "rewards/margins": 5.101554052089341e-05, | |
| "rewards/rejected": -2.489005055394955e-05, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.5873015873015873e-06, | |
| "logits/chosen": 0.2518273890018463, | |
| "logits/rejected": 0.32998543977737427, | |
| "logps/chosen": -489.4015197753906, | |
| "logps/rejected": -479.5149841308594, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.4437499940395355, | |
| "rewards/chosen": -0.0016528337728232145, | |
| "rewards/margins": 0.00016125899855978787, | |
| "rewards/rejected": -0.001814092742279172, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2.380952380952381e-06, | |
| "logits/chosen": 0.18698568642139435, | |
| "logits/rejected": 0.21475109457969666, | |
| "logps/chosen": -501.27581787109375, | |
| "logps/rejected": -522.3826293945312, | |
| "loss": 0.6927, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.008201382122933865, | |
| "rewards/margins": 0.0008340775966644287, | |
| "rewards/rejected": -0.009035459719598293, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 3.1746031746031746e-06, | |
| "logits/chosen": 0.20552173256874084, | |
| "logits/rejected": 0.18397466838359833, | |
| "logps/chosen": -533.6796875, | |
| "logps/rejected": -549.6390380859375, | |
| "loss": 0.6917, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.024371124804019928, | |
| "rewards/margins": 0.0025360379368066788, | |
| "rewards/rejected": -0.026907164603471756, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 3.968253968253968e-06, | |
| "logits/chosen": 0.2108609676361084, | |
| "logits/rejected": 0.2548374831676483, | |
| "logps/chosen": -554.6732177734375, | |
| "logps/rejected": -545.2399291992188, | |
| "loss": 0.6896, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.0501377210021019, | |
| "rewards/margins": 0.006772263441234827, | |
| "rewards/rejected": -0.05690998584032059, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.761904761904762e-06, | |
| "logits/chosen": 0.19314441084861755, | |
| "logits/rejected": 0.27108246088027954, | |
| "logps/chosen": -606.186767578125, | |
| "logps/rejected": -608.7874145507812, | |
| "loss": 0.6876, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.0944763794541359, | |
| "rewards/margins": 0.008413041010499, | |
| "rewards/rejected": -0.10288941860198975, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 4.998086282661188e-06, | |
| "logits/chosen": 0.16735275089740753, | |
| "logits/rejected": 0.17189382016658783, | |
| "logps/chosen": -552.279296875, | |
| "logps/rejected": -591.57421875, | |
| "loss": 0.6852, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": -0.11828555911779404, | |
| "rewards/margins": 0.01398100983351469, | |
| "rewards/rejected": -0.1322665959596634, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.988720025682995e-06, | |
| "logits/chosen": 0.10864056646823883, | |
| "logits/rejected": 0.14530186355113983, | |
| "logps/chosen": -644.5869140625, | |
| "logps/rejected": -674.5743408203125, | |
| "loss": 0.6828, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.13946327567100525, | |
| "rewards/margins": 0.019074594601988792, | |
| "rewards/rejected": -0.1585378646850586, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.9715789537359126e-06, | |
| "logits/chosen": 0.07046853750944138, | |
| "logits/rejected": 0.1207747831940651, | |
| "logps/chosen": -684.7418212890625, | |
| "logps/rejected": -712.235107421875, | |
| "loss": 0.6792, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.1731593906879425, | |
| "rewards/margins": 0.031467147171497345, | |
| "rewards/rejected": -0.20462656021118164, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 4.946716615897932e-06, | |
| "logits/chosen": 0.0422443263232708, | |
| "logits/rejected": 0.07933792471885681, | |
| "logps/chosen": -741.9517211914062, | |
| "logps/rejected": -768.1805419921875, | |
| "loss": 0.6793, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.2586411237716675, | |
| "rewards/margins": 0.030019978061318398, | |
| "rewards/rejected": -0.28866106271743774, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.9142106826480114e-06, | |
| "logits/chosen": 0.01217989344149828, | |
| "logits/rejected": 0.036614928394556046, | |
| "logps/chosen": -704.1077880859375, | |
| "logps/rejected": -766.5753173828125, | |
| "loss": 0.6816, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.24351103603839874, | |
| "rewards/margins": 0.025916840881109238, | |
| "rewards/rejected": -0.2694278359413147, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.874162703221823e-06, | |
| "logits/chosen": 0.04041505232453346, | |
| "logits/rejected": 0.0663461834192276, | |
| "logps/chosen": -756.8778686523438, | |
| "logps/rejected": -816.00244140625, | |
| "loss": 0.6726, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.27453333139419556, | |
| "rewards/margins": 0.04802945256233215, | |
| "rewards/rejected": -0.3225627541542053, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.826697788369752e-06, | |
| "logits/chosen": -0.03461530804634094, | |
| "logits/rejected": 0.08036826550960541, | |
| "logps/chosen": -926.38720703125, | |
| "logps/rejected": -1025.3724365234375, | |
| "loss": 0.6617, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.422262579202652, | |
| "rewards/margins": 0.0845782682299614, | |
| "rewards/rejected": -0.5068408250808716, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.7719642195082224e-06, | |
| "logits/chosen": 0.032650940120220184, | |
| "logits/rejected": 0.02779592201113701, | |
| "logps/chosen": -852.9603271484375, | |
| "logps/rejected": -985.2545166015625, | |
| "loss": 0.6576, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.36906754970550537, | |
| "rewards/margins": 0.09497459977865219, | |
| "rewards/rejected": -0.46404212713241577, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 4.710132985485355e-06, | |
| "logits/chosen": 0.05756285786628723, | |
| "logits/rejected": 0.1693125069141388, | |
| "logps/chosen": -1006.541015625, | |
| "logps/rejected": -1214.77587890625, | |
| "loss": 0.6513, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.5491369962692261, | |
| "rewards/margins": 0.171253964304924, | |
| "rewards/rejected": -0.7203909158706665, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.641397248408122e-06, | |
| "logits/chosen": 0.25534456968307495, | |
| "logits/rejected": 0.15446144342422485, | |
| "logps/chosen": -1380.378173828125, | |
| "logps/rejected": -1509.271728515625, | |
| "loss": 0.6491, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.8604512214660645, | |
| "rewards/margins": 0.11827781051397324, | |
| "rewards/rejected": -0.9787289500236511, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.5659717401997655e-06, | |
| "logits/chosen": 0.11406876146793365, | |
| "logits/rejected": 0.2429320365190506, | |
| "logps/chosen": -1162.3896484375, | |
| "logps/rejected": -1300.2310791015625, | |
| "loss": 0.6593, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.6630619168281555, | |
| "rewards/margins": 0.1277274191379547, | |
| "rewards/rejected": -0.7907893061637878, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.4840920917726425e-06, | |
| "logits/chosen": 0.1443721204996109, | |
| "logits/rejected": 0.24939250946044922, | |
| "logps/chosen": -1037.646240234375, | |
| "logps/rejected": -1174.6041259765625, | |
| "loss": 0.6464, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.5330273509025574, | |
| "rewards/margins": 0.12969453632831573, | |
| "rewards/rejected": -0.6627219915390015, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 4.396014096912182e-06, | |
| "logits/chosen": 0.1525859534740448, | |
| "logits/rejected": 0.20368242263793945, | |
| "logps/chosen": -1222.385498046875, | |
| "logps/rejected": -1366.5484619140625, | |
| "loss": 0.6629, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.7235668897628784, | |
| "rewards/margins": 0.11871180683374405, | |
| "rewards/rejected": -0.8422788381576538, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 4.302012913171584e-06, | |
| "logits/chosen": 0.20741760730743408, | |
| "logits/rejected": 0.23669476807117462, | |
| "logps/chosen": -1129.2685546875, | |
| "logps/rejected": -1311.441650390625, | |
| "loss": 0.6541, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.6570686101913452, | |
| "rewards/margins": 0.16556376218795776, | |
| "rewards/rejected": -0.8226323127746582, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 4.202382202273702e-06, | |
| "logits/chosen": 0.20800559222698212, | |
| "logits/rejected": 0.2900869846343994, | |
| "logps/chosen": -1247.2933349609375, | |
| "logps/rejected": -1375.472900390625, | |
| "loss": 0.6643, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.6747683882713318, | |
| "rewards/margins": 0.1478392481803894, | |
| "rewards/rejected": -0.8226076364517212, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 4.097433212705492e-06, | |
| "logits/chosen": 0.1988043189048767, | |
| "logits/rejected": 0.2728360593318939, | |
| "logps/chosen": -1089.7906494140625, | |
| "logps/rejected": -1184.677490234375, | |
| "loss": 0.6588, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.5955246686935425, | |
| "rewards/margins": 0.07477297633886337, | |
| "rewards/rejected": -0.6702976226806641, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 3.987493807371033e-06, | |
| "logits/chosen": 0.2314763069152832, | |
| "logits/rejected": 0.34903794527053833, | |
| "logps/chosen": -1126.2254638671875, | |
| "logps/rejected": -1295.7572021484375, | |
| "loss": 0.6408, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.6263796091079712, | |
| "rewards/margins": 0.1536393165588379, | |
| "rewards/rejected": -0.7800189852714539, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.872907439340758e-06, | |
| "logits/chosen": 0.24221527576446533, | |
| "logits/rejected": 0.3353860080242157, | |
| "logps/chosen": -1068.5975341796875, | |
| "logps/rejected": -1235.1514892578125, | |
| "loss": 0.6518, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.6160915493965149, | |
| "rewards/margins": 0.1515791118144989, | |
| "rewards/rejected": -0.7676706910133362, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 3.75403207889666e-06, | |
| "logits/chosen": 0.2971830368041992, | |
| "logits/rejected": 0.34285473823547363, | |
| "logps/chosen": -1286.203369140625, | |
| "logps/rejected": -1462.2490234375, | |
| "loss": 0.647, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.7703440189361572, | |
| "rewards/margins": 0.16007237136363983, | |
| "rewards/rejected": -0.9304164052009583, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.631239095225417e-06, | |
| "logits/chosen": 0.22825905680656433, | |
| "logits/rejected": 0.32187604904174805, | |
| "logps/chosen": -1112.8629150390625, | |
| "logps/rejected": -1291.74072265625, | |
| "loss": 0.64, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.6394481658935547, | |
| "rewards/margins": 0.16354107856750488, | |
| "rewards/rejected": -0.8029893040657043, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 3.5049120962530608e-06, | |
| "logits/chosen": 0.32051190733909607, | |
| "logits/rejected": 0.2698783576488495, | |
| "logps/chosen": -1120.5859375, | |
| "logps/rejected": -1305.715087890625, | |
| "loss": 0.6356, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.6476856470108032, | |
| "rewards/margins": 0.15380369126796722, | |
| "rewards/rejected": -0.8014893531799316, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 3.3754457302455464e-06, | |
| "logits/chosen": 0.23477685451507568, | |
| "logits/rejected": 0.43997740745544434, | |
| "logps/chosen": -1300.92724609375, | |
| "logps/rejected": -1495.2685546875, | |
| "loss": 0.6537, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -0.8141768574714661, | |
| "rewards/margins": 0.17344731092453003, | |
| "rewards/rejected": -0.9876241683959961, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 3.2432444529190714e-06, | |
| "logits/chosen": 0.38444215059280396, | |
| "logits/rejected": 0.43280649185180664, | |
| "logps/chosen": -1254.042724609375, | |
| "logps/rejected": -1498.3414306640625, | |
| "loss": 0.6399, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.7942158579826355, | |
| "rewards/margins": 0.20459000766277313, | |
| "rewards/rejected": -0.9988058805465698, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 3.1087212639117057e-06, | |
| "logits/chosen": 0.35266411304473877, | |
| "logits/rejected": 0.5088330507278442, | |
| "logps/chosen": -1342.4091796875, | |
| "logps/rejected": -1464.3389892578125, | |
| "loss": 0.6541, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.8787577748298645, | |
| "rewards/margins": 0.11491602659225464, | |
| "rewards/rejected": -0.9936738014221191, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.9722964165636263e-06, | |
| "logits/chosen": 0.3081873953342438, | |
| "logits/rejected": 0.4749964773654938, | |
| "logps/chosen": -1241.6292724609375, | |
| "logps/rejected": -1437.963134765625, | |
| "loss": 0.6523, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.7512220144271851, | |
| "rewards/margins": 0.16275373101234436, | |
| "rewards/rejected": -0.913975715637207, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 2.8343961050366275e-06, | |
| "logits/chosen": 0.3727472424507141, | |
| "logits/rejected": 0.41282615065574646, | |
| "logps/chosen": -1198.703857421875, | |
| "logps/rejected": -1467.189208984375, | |
| "loss": 0.6399, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.719472348690033, | |
| "rewards/margins": 0.23017852008342743, | |
| "rewards/rejected": -0.9496508836746216, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 2.695451132874385e-06, | |
| "logits/chosen": 0.3753197193145752, | |
| "logits/rejected": 0.4588402211666107, | |
| "logps/chosen": -1306.0467529296875, | |
| "logps/rejected": -1423.9765625, | |
| "loss": 0.6339, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.821855902671814, | |
| "rewards/margins": 0.15302011370658875, | |
| "rewards/rejected": -0.9748760461807251, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.5558955671628964e-06, | |
| "logits/chosen": 0.32516294717788696, | |
| "logits/rejected": 0.49244314432144165, | |
| "logps/chosen": -1465.662353515625, | |
| "logps/rejected": -1645.747314453125, | |
| "loss": 0.6415, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.933737576007843, | |
| "rewards/margins": 0.19395960867404938, | |
| "rewards/rejected": -1.127697229385376, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 2.4161653824955654e-06, | |
| "logits/chosen": 0.37509265542030334, | |
| "logits/rejected": 0.5179189443588257, | |
| "logps/chosen": -1365.8070068359375, | |
| "logps/rejected": -1507.181396484375, | |
| "loss": 0.639, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.868498682975769, | |
| "rewards/margins": 0.15674063563346863, | |
| "rewards/rejected": -1.0252392292022705, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2.2766970989791697e-06, | |
| "logits/chosen": 0.3914629817008972, | |
| "logits/rejected": 0.39266401529312134, | |
| "logps/chosen": -1348.240966796875, | |
| "logps/rejected": -1593.037353515625, | |
| "loss": 0.6485, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.8051446080207825, | |
| "rewards/margins": 0.2146397829055786, | |
| "rewards/rejected": -1.0197844505310059, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 2.1379264185356545e-06, | |
| "logits/chosen": 0.29912859201431274, | |
| "logits/rejected": 0.4203578531742096, | |
| "logps/chosen": -1122.2073974609375, | |
| "logps/rejected": -1355.9158935546875, | |
| "loss": 0.6317, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.7034318447113037, | |
| "rewards/margins": 0.18991844356060028, | |
| "rewards/rejected": -0.8933502435684204, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 2.000286863759934e-06, | |
| "logits/chosen": 0.37296754121780396, | |
| "logits/rejected": 0.569365382194519, | |
| "logps/chosen": -1219.0467529296875, | |
| "logps/rejected": -1441.828857421875, | |
| "loss": 0.6332, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.7400010228157043, | |
| "rewards/margins": 0.20997154712677002, | |
| "rewards/rejected": -0.9499725103378296, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 1.8642084235859764e-06, | |
| "logits/chosen": 0.39869484305381775, | |
| "logits/rejected": 0.49166345596313477, | |
| "logps/chosen": -1271.091064453125, | |
| "logps/rejected": -1447.0933837890625, | |
| "loss": 0.6447, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.7757032513618469, | |
| "rewards/margins": 0.18741454184055328, | |
| "rewards/rejected": -0.9631179571151733, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 1.7301162099921013e-06, | |
| "logits/chosen": 0.48826533555984497, | |
| "logits/rejected": 0.5182665586471558, | |
| "logps/chosen": -1342.3040771484375, | |
| "logps/rejected": -1515.572021484375, | |
| "loss": 0.6527, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.8140773773193359, | |
| "rewards/margins": 0.18610136210918427, | |
| "rewards/rejected": -1.0001786947250366, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 1.5984291299420117e-06, | |
| "logits/chosen": 0.4482479989528656, | |
| "logits/rejected": 0.494678258895874, | |
| "logps/chosen": -1258.864501953125, | |
| "logps/rejected": -1553.4383544921875, | |
| "loss": 0.6073, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.7780212163925171, | |
| "rewards/margins": 0.25596219301223755, | |
| "rewards/rejected": -1.0339832305908203, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.4695585767104092e-06, | |
| "logits/chosen": 0.4765148162841797, | |
| "logits/rejected": 0.44907188415527344, | |
| "logps/chosen": -1387.763427734375, | |
| "logps/rejected": -1522.737548828125, | |
| "loss": 0.6573, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.9270384907722473, | |
| "rewards/margins": 0.10358323156833649, | |
| "rewards/rejected": -1.0306217670440674, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.3439071446815452e-06, | |
| "logits/chosen": 0.4347519278526306, | |
| "logits/rejected": 0.5749093294143677, | |
| "logps/chosen": -1321.0262451171875, | |
| "logps/rejected": -1519.8450927734375, | |
| "loss": 0.6397, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.8617701530456543, | |
| "rewards/margins": 0.19662390649318695, | |
| "rewards/rejected": -1.0583940744400024, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.2218673716356919e-06, | |
| "logits/chosen": 0.4529833197593689, | |
| "logits/rejected": 0.5917688608169556, | |
| "logps/chosen": -1364.6546630859375, | |
| "logps/rejected": -1504.9599609375, | |
| "loss": 0.6405, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.8427094221115112, | |
| "rewards/margins": 0.1536230742931366, | |
| "rewards/rejected": -0.9963325262069702, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 1.103820512452661e-06, | |
| "logits/chosen": 0.3574792444705963, | |
| "logits/rejected": 0.4730028510093689, | |
| "logps/chosen": -1273.257568359375, | |
| "logps/rejected": -1469.2452392578125, | |
| "loss": 0.616, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.7781795859336853, | |
| "rewards/margins": 0.17071351408958435, | |
| "rewards/rejected": -0.9488931894302368, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 9.901353480633468e-07, | |
| "logits/chosen": 0.43064388632774353, | |
| "logits/rejected": 0.6706501245498657, | |
| "logps/chosen": -1389.450439453125, | |
| "logps/rejected": -1536.8935546875, | |
| "loss": 0.6465, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.8454073071479797, | |
| "rewards/margins": 0.16833043098449707, | |
| "rewards/rejected": -1.0137377977371216, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 8.811670333701544e-07, | |
| "logits/chosen": 0.46944761276245117, | |
| "logits/rejected": 0.5561131238937378, | |
| "logps/chosen": -1286.3804931640625, | |
| "logps/rejected": -1454.6707763671875, | |
| "loss": 0.6409, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -0.8223176002502441, | |
| "rewards/margins": 0.14870892465114594, | |
| "rewards/rejected": -0.971026599407196, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 7.772559877354341e-07, | |
| "logits/chosen": 0.44347596168518066, | |
| "logits/rejected": 0.5607708692550659, | |
| "logps/chosen": -1329.2608642578125, | |
| "logps/rejected": -1553.534912109375, | |
| "loss": 0.6401, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.8247483968734741, | |
| "rewards/margins": 0.23265239596366882, | |
| "rewards/rejected": -1.0574008226394653, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 6.787268315040604e-07, | |
| "logits/chosen": 0.45571979880332947, | |
| "logits/rejected": 0.47239336371421814, | |
| "logps/chosen": -1211.712646484375, | |
| "logps/rejected": -1386.2025146484375, | |
| "loss": 0.6304, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.7387667298316956, | |
| "rewards/margins": 0.1702551543712616, | |
| "rewards/rejected": -0.9090218544006348, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 5.858873718824829e-07, | |
| "logits/chosen": 0.3857249319553375, | |
| "logits/rejected": 0.5553338527679443, | |
| "logps/chosen": -1262.7962646484375, | |
| "logps/rejected": -1481.9622802734375, | |
| "loss": 0.6333, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.7554186582565308, | |
| "rewards/margins": 0.20985445380210876, | |
| "rewards/rejected": -0.9652730822563171, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 4.990276413423817e-07, | |
| "logits/chosen": 0.4313405454158783, | |
| "logits/rejected": 0.590368390083313, | |
| "logps/chosen": -1306.13427734375, | |
| "logps/rejected": -1447.4407958984375, | |
| "loss": 0.6303, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.7738032341003418, | |
| "rewards/margins": 0.16366735100746155, | |
| "rewards/rejected": -0.9374706149101257, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 4.184189915529796e-07, | |
| "logits/chosen": 0.4014933109283447, | |
| "logits/rejected": 0.5485681891441345, | |
| "logps/chosen": -1255.651123046875, | |
| "logps/rejected": -1364.13232421875, | |
| "loss": 0.6527, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -0.7963151931762695, | |
| "rewards/margins": 0.10994601249694824, | |
| "rewards/rejected": -0.9062612652778625, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 3.4431324567258176e-07, | |
| "logits/chosen": 0.48007622361183167, | |
| "logits/rejected": 0.5644385814666748, | |
| "logps/chosen": -1217.1119384765625, | |
| "logps/rejected": -1428.557861328125, | |
| "loss": 0.6382, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.7597641944885254, | |
| "rewards/margins": 0.2099626362323761, | |
| "rewards/rejected": -0.9697268605232239, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 2.769419116476052e-07, | |
| "logits/chosen": 0.41416773200035095, | |
| "logits/rejected": 0.5963398218154907, | |
| "logps/chosen": -1281.758544921875, | |
| "logps/rejected": -1478.4769287109375, | |
| "loss": 0.6332, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.7958284020423889, | |
| "rewards/margins": 0.19340744614601135, | |
| "rewards/rejected": -0.9892358779907227, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 2.1651545897676512e-07, | |
| "logits/chosen": 0.4229779839515686, | |
| "logits/rejected": 0.6195346117019653, | |
| "logps/chosen": -1226.513427734375, | |
| "logps/rejected": -1382.2999267578125, | |
| "loss": 0.6559, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.76485675573349, | |
| "rewards/margins": 0.13517045974731445, | |
| "rewards/rejected": -0.9000272750854492, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.6322266119983222e-07, | |
| "logits/chosen": 0.46644434332847595, | |
| "logits/rejected": 0.5570663213729858, | |
| "logps/chosen": -1331.478271484375, | |
| "logps/rejected": -1517.951416015625, | |
| "loss": 0.6477, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.8168514966964722, | |
| "rewards/margins": 0.17577320337295532, | |
| "rewards/rejected": -0.9926246404647827, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 1.1723000616502167e-07, | |
| "logits/chosen": 0.414206326007843, | |
| "logits/rejected": 0.6192079782485962, | |
| "logps/chosen": -1307.5006103515625, | |
| "logps/rejected": -1502.477783203125, | |
| "loss": 0.6329, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.8042691349983215, | |
| "rewards/margins": 0.1876949965953827, | |
| "rewards/rejected": -0.9919641613960266, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 7.868117591737585e-08, | |
| "logits/chosen": 0.3812820017337799, | |
| "logits/rejected": 0.6579981446266174, | |
| "logps/chosen": -1247.164794921875, | |
| "logps/rejected": -1437.7857666015625, | |
| "loss": 0.643, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.7581318616867065, | |
| "rewards/margins": 0.20403996109962463, | |
| "rewards/rejected": -0.9621719121932983, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 4.769659783295383e-08, | |
| "logits/chosen": 0.38649970293045044, | |
| "logits/rejected": 0.5585275888442993, | |
| "logps/chosen": -1279.88818359375, | |
| "logps/rejected": -1423.707763671875, | |
| "loss": 0.6325, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.7771328687667847, | |
| "rewards/margins": 0.1578211635351181, | |
| "rewards/rejected": -0.9349539875984192, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2.4373068401120358e-08, | |
| "logits/chosen": 0.5281786918640137, | |
| "logits/rejected": 0.5024128556251526, | |
| "logps/chosen": -1351.9422607421875, | |
| "logps/rejected": -1575.468505859375, | |
| "loss": 0.6309, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.7931488752365112, | |
| "rewards/margins": 0.2438853532075882, | |
| "rewards/rejected": -1.037034273147583, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 8.78345083022425e-09, | |
| "logits/chosen": 0.42136329412460327, | |
| "logits/rejected": 0.5088121891021729, | |
| "logps/chosen": -1306.164794921875, | |
| "logps/rejected": -1567.685546875, | |
| "loss": 0.6363, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.8058547973632812, | |
| "rewards/margins": 0.24808284640312195, | |
| "rewards/rejected": -1.0539376735687256, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 9.764474213677654e-10, | |
| "logits/chosen": 0.45875900983810425, | |
| "logits/rejected": 0.5520389080047607, | |
| "logps/chosen": -1280.700439453125, | |
| "logps/rejected": -1439.426025390625, | |
| "loss": 0.6465, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.7951013445854187, | |
| "rewards/margins": 0.15834487974643707, | |
| "rewards/rejected": -0.9534462094306946, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 625, | |
| "total_flos": 0.0, | |
| "train_loss": 0.3321857452392578, | |
| "train_runtime": 4002.829, | |
| "train_samples_per_second": 4.996, | |
| "train_steps_per_second": 0.156 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 625, | |
| "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 | |
| } | |