| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 1743, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 2.857142857142857e-09, | |
| "logits/chosen": -2.8170976638793945, | |
| "logits/rejected": -2.583425998687744, | |
| "logps/chosen": -267.58819580078125, | |
| "logps/rejected": -55.24143981933594, | |
| "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": 2.857142857142857e-08, | |
| "logits/chosen": -2.728940486907959, | |
| "logits/rejected": -2.597886800765991, | |
| "logps/chosen": -287.7469482421875, | |
| "logps/rejected": -66.66583251953125, | |
| "loss": 0.6912, | |
| "rewards/accuracies": 0.4583333432674408, | |
| "rewards/chosen": -0.003922327421605587, | |
| "rewards/margins": -0.0009250044822692871, | |
| "rewards/rejected": -0.0029973229393363, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 5.714285714285714e-08, | |
| "logits/chosen": -2.6918883323669434, | |
| "logits/rejected": -2.6110966205596924, | |
| "logps/chosen": -280.5570068359375, | |
| "logps/rejected": -69.74317169189453, | |
| "loss": 0.6445, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": 0.10821928083896637, | |
| "rewards/margins": 0.11207082122564316, | |
| "rewards/rejected": -0.003851559478789568, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 8.571428571428572e-08, | |
| "logits/chosen": -2.6756694316864014, | |
| "logits/rejected": -2.5497803688049316, | |
| "logps/chosen": -334.42138671875, | |
| "logps/rejected": -74.35704040527344, | |
| "loss": 0.5256, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": 0.3521518111228943, | |
| "rewards/margins": 0.4118792414665222, | |
| "rewards/rejected": -0.05972742289304733, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 1.1428571428571427e-07, | |
| "logits/chosen": -2.585674524307251, | |
| "logits/rejected": -2.4876582622528076, | |
| "logps/chosen": -259.1927185058594, | |
| "logps/rejected": -70.0163345336914, | |
| "loss": 0.3967, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 0.5247544050216675, | |
| "rewards/margins": 0.7782431244850159, | |
| "rewards/rejected": -0.2534886598587036, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.4285714285714285e-07, | |
| "logits/chosen": -2.556175947189331, | |
| "logits/rejected": -2.4422554969787598, | |
| "logps/chosen": -293.9497375488281, | |
| "logps/rejected": -78.49222564697266, | |
| "loss": 0.3084, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": 0.4779079556465149, | |
| "rewards/margins": 1.3577072620391846, | |
| "rewards/rejected": -0.8797992467880249, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.7142857142857143e-07, | |
| "logits/chosen": -2.5218682289123535, | |
| "logits/rejected": -2.413239002227783, | |
| "logps/chosen": -348.8528747558594, | |
| "logps/rejected": -90.35295867919922, | |
| "loss": 0.2281, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": 0.8038774728775024, | |
| "rewards/margins": 2.1713979244232178, | |
| "rewards/rejected": -1.3675200939178467, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 2e-07, | |
| "logits/chosen": -2.5286450386047363, | |
| "logits/rejected": -2.4316279888153076, | |
| "logps/chosen": -277.53424072265625, | |
| "logps/rejected": -77.01128387451172, | |
| "loss": 0.2105, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": 0.6769354343414307, | |
| "rewards/margins": 2.166170358657837, | |
| "rewards/rejected": -1.4892350435256958, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2.2857142857142855e-07, | |
| "logits/chosen": -2.498818874359131, | |
| "logits/rejected": -2.377124071121216, | |
| "logps/chosen": -307.00531005859375, | |
| "logps/rejected": -99.22511291503906, | |
| "loss": 0.1428, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 0.6349704265594482, | |
| "rewards/margins": 2.8562676906585693, | |
| "rewards/rejected": -2.2212975025177, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2.571428571428571e-07, | |
| "logits/chosen": -2.5420005321502686, | |
| "logits/rejected": -2.4419803619384766, | |
| "logps/chosen": -297.6865539550781, | |
| "logps/rejected": -105.5224609375, | |
| "loss": 0.1443, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": 0.5000184774398804, | |
| "rewards/margins": 3.1510608196258545, | |
| "rewards/rejected": -2.6510424613952637, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 2.857142857142857e-07, | |
| "logits/chosen": -2.4766311645507812, | |
| "logits/rejected": -2.364316701889038, | |
| "logps/chosen": -269.93719482421875, | |
| "logps/rejected": -100.0908432006836, | |
| "loss": 0.1393, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": 0.4442061483860016, | |
| "rewards/margins": 3.083071231842041, | |
| "rewards/rejected": -2.638864517211914, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 3.142857142857143e-07, | |
| "logits/chosen": -2.388690233230591, | |
| "logits/rejected": -2.3019070625305176, | |
| "logps/chosen": -259.3259582519531, | |
| "logps/rejected": -92.60478210449219, | |
| "loss": 0.1255, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": 0.3840440511703491, | |
| "rewards/margins": 3.357806444168091, | |
| "rewards/rejected": -2.973762035369873, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 3.4285714285714286e-07, | |
| "logits/chosen": -2.413675546646118, | |
| "logits/rejected": -2.366042375564575, | |
| "logps/chosen": -291.9371643066406, | |
| "logps/rejected": -122.68412017822266, | |
| "loss": 0.1103, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": 0.07346852123737335, | |
| "rewards/margins": 3.9470927715301514, | |
| "rewards/rejected": -3.873624324798584, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 3.7142857142857145e-07, | |
| "logits/chosen": -2.2968249320983887, | |
| "logits/rejected": -2.225172519683838, | |
| "logps/chosen": -244.48300170898438, | |
| "logps/rejected": -106.67143249511719, | |
| "loss": 0.1152, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -0.45761212706565857, | |
| "rewards/margins": 3.7605834007263184, | |
| "rewards/rejected": -4.21819543838501, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 4e-07, | |
| "logits/chosen": -2.4523983001708984, | |
| "logits/rejected": -2.301741123199463, | |
| "logps/chosen": -299.51251220703125, | |
| "logps/rejected": -111.51058197021484, | |
| "loss": 0.0923, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": 0.3345158100128174, | |
| "rewards/margins": 4.5269455909729, | |
| "rewards/rejected": -4.192429542541504, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 4.285714285714285e-07, | |
| "logits/chosen": -2.2728230953216553, | |
| "logits/rejected": -2.1703383922576904, | |
| "logps/chosen": -217.3053741455078, | |
| "logps/rejected": -100.21586608886719, | |
| "loss": 0.0937, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": 0.14594963192939758, | |
| "rewards/margins": 3.915192127227783, | |
| "rewards/rejected": -3.76924204826355, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 4.571428571428571e-07, | |
| "logits/chosen": -2.311068058013916, | |
| "logits/rejected": -2.1951611042022705, | |
| "logps/chosen": -224.5479736328125, | |
| "logps/rejected": -97.22464752197266, | |
| "loss": 0.081, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": 0.23994378745555878, | |
| "rewards/margins": 4.240386009216309, | |
| "rewards/rejected": -4.000441551208496, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.857142857142857e-07, | |
| "logits/chosen": -2.4215264320373535, | |
| "logits/rejected": -2.3117451667785645, | |
| "logps/chosen": -338.24591064453125, | |
| "logps/rejected": -135.39236450195312, | |
| "loss": 0.0497, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": 0.741433322429657, | |
| "rewards/margins": 6.07324743270874, | |
| "rewards/rejected": -5.331814765930176, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.984056122448979e-07, | |
| "logits/chosen": -2.376217842102051, | |
| "logits/rejected": -2.2611336708068848, | |
| "logps/chosen": -308.8148498535156, | |
| "logps/rejected": -132.30233764648438, | |
| "loss": 0.1005, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": 0.16114525496959686, | |
| "rewards/margins": 5.889215469360352, | |
| "rewards/rejected": -5.72806978225708, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 4.952168367346938e-07, | |
| "logits/chosen": -2.3446619510650635, | |
| "logits/rejected": -2.239445209503174, | |
| "logps/chosen": -316.2629699707031, | |
| "logps/rejected": -133.4289093017578, | |
| "loss": 0.0733, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.2028081864118576, | |
| "rewards/margins": 5.608849048614502, | |
| "rewards/rejected": -5.811657905578613, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 4.920280612244897e-07, | |
| "logits/chosen": -2.263739585876465, | |
| "logits/rejected": -2.1694753170013428, | |
| "logps/chosen": -233.5961456298828, | |
| "logps/rejected": -109.9480209350586, | |
| "loss": 0.0772, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": 0.2176302969455719, | |
| "rewards/margins": 5.1773786544799805, | |
| "rewards/rejected": -4.959748268127441, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 4.888392857142857e-07, | |
| "logits/chosen": -2.3166961669921875, | |
| "logits/rejected": -2.2644801139831543, | |
| "logps/chosen": -266.83099365234375, | |
| "logps/rejected": -127.63531494140625, | |
| "loss": 0.0682, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": 0.30769938230514526, | |
| "rewards/margins": 5.930529594421387, | |
| "rewards/rejected": -5.622830390930176, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.856505102040816e-07, | |
| "logits/chosen": -2.289324998855591, | |
| "logits/rejected": -2.145235538482666, | |
| "logps/chosen": -287.94903564453125, | |
| "logps/rejected": -135.3939666748047, | |
| "loss": 0.0824, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": 0.18774142861366272, | |
| "rewards/margins": 6.2231669425964355, | |
| "rewards/rejected": -6.035426139831543, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.824617346938776e-07, | |
| "logits/chosen": -2.216264009475708, | |
| "logits/rejected": -2.047607183456421, | |
| "logps/chosen": -271.09869384765625, | |
| "logps/rejected": -129.23095703125, | |
| "loss": 0.1054, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": 0.05019025877118111, | |
| "rewards/margins": 6.176311016082764, | |
| "rewards/rejected": -6.126120567321777, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.792729591836734e-07, | |
| "logits/chosen": -2.2174153327941895, | |
| "logits/rejected": -2.094043254852295, | |
| "logps/chosen": -295.2961120605469, | |
| "logps/rejected": -132.24790954589844, | |
| "loss": 0.0579, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": 0.24180030822753906, | |
| "rewards/margins": 6.183707237243652, | |
| "rewards/rejected": -5.941906929016113, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.760841836734694e-07, | |
| "logits/chosen": -2.3349456787109375, | |
| "logits/rejected": -2.2391867637634277, | |
| "logps/chosen": -301.3727722167969, | |
| "logps/rejected": -140.08639526367188, | |
| "loss": 0.0578, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.2325439751148224, | |
| "rewards/margins": 6.470588684082031, | |
| "rewards/rejected": -6.703132629394531, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 4.728954081632653e-07, | |
| "logits/chosen": -2.2977206707000732, | |
| "logits/rejected": -2.1363391876220703, | |
| "logps/chosen": -305.45477294921875, | |
| "logps/rejected": -140.94717407226562, | |
| "loss": 0.0599, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -0.0075416178442537785, | |
| "rewards/margins": 6.882362365722656, | |
| "rewards/rejected": -6.8899030685424805, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 4.697066326530612e-07, | |
| "logits/chosen": -2.2209079265594482, | |
| "logits/rejected": -2.1426846981048584, | |
| "logps/chosen": -269.04010009765625, | |
| "logps/rejected": -142.6284942626953, | |
| "loss": 0.056, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.23809853196144104, | |
| "rewards/margins": 6.732804298400879, | |
| "rewards/rejected": -6.970902919769287, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 4.665178571428571e-07, | |
| "logits/chosen": -2.3430752754211426, | |
| "logits/rejected": -2.2075252532958984, | |
| "logps/chosen": -314.6163024902344, | |
| "logps/rejected": -146.22244262695312, | |
| "loss": 0.0529, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.08103661239147186, | |
| "rewards/margins": 7.245053768157959, | |
| "rewards/rejected": -7.3260908126831055, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 4.6332908163265305e-07, | |
| "logits/chosen": -2.2507033348083496, | |
| "logits/rejected": -2.094573736190796, | |
| "logps/chosen": -262.8319396972656, | |
| "logps/rejected": -141.3706817626953, | |
| "loss": 0.0589, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.06328751146793365, | |
| "rewards/margins": 6.6208062171936035, | |
| "rewards/rejected": -6.684094429016113, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 4.60140306122449e-07, | |
| "logits/chosen": -2.2667717933654785, | |
| "logits/rejected": -2.16914701461792, | |
| "logps/chosen": -282.3851623535156, | |
| "logps/rejected": -144.63890075683594, | |
| "loss": 0.0585, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -0.35556235909461975, | |
| "rewards/margins": 6.836700439453125, | |
| "rewards/rejected": -7.192262172698975, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.569515306122449e-07, | |
| "logits/chosen": -2.2799079418182373, | |
| "logits/rejected": -2.1546857357025146, | |
| "logps/chosen": -307.00457763671875, | |
| "logps/rejected": -144.7294921875, | |
| "loss": 0.0401, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": 0.4573780596256256, | |
| "rewards/margins": 7.150294303894043, | |
| "rewards/rejected": -6.6929168701171875, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.5376275510204084e-07, | |
| "logits/chosen": -2.2312560081481934, | |
| "logits/rejected": -2.0914812088012695, | |
| "logps/chosen": -261.8872985839844, | |
| "logps/rejected": -139.28701782226562, | |
| "loss": 0.0619, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.15933240950107574, | |
| "rewards/margins": 6.717618465423584, | |
| "rewards/rejected": -6.876950740814209, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.505739795918367e-07, | |
| "logits/chosen": -2.2225937843322754, | |
| "logits/rejected": -2.0411500930786133, | |
| "logps/chosen": -304.5137634277344, | |
| "logps/rejected": -132.83473205566406, | |
| "loss": 0.065, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": 0.07894410938024521, | |
| "rewards/margins": 6.419223785400391, | |
| "rewards/rejected": -6.340278625488281, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.473852040816326e-07, | |
| "logits/chosen": -2.304166793823242, | |
| "logits/rejected": -2.144609212875366, | |
| "logps/chosen": -274.3529968261719, | |
| "logps/rejected": -131.06246948242188, | |
| "loss": 0.0907, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.192825049161911, | |
| "rewards/margins": 6.38867712020874, | |
| "rewards/rejected": -6.5815019607543945, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.441964285714285e-07, | |
| "logits/chosen": -2.305800676345825, | |
| "logits/rejected": -2.111426830291748, | |
| "logps/chosen": -340.38702392578125, | |
| "logps/rejected": -148.1328887939453, | |
| "loss": 0.0496, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.030660927295684814, | |
| "rewards/margins": 7.517220973968506, | |
| "rewards/rejected": -7.547882080078125, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.4100765306122445e-07, | |
| "logits/chosen": -2.249474287033081, | |
| "logits/rejected": -2.114851474761963, | |
| "logps/chosen": -276.80303955078125, | |
| "logps/rejected": -139.46585083007812, | |
| "loss": 0.06, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -0.2887513041496277, | |
| "rewards/margins": 7.209240913391113, | |
| "rewards/rejected": -7.497992038726807, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.378188775510204e-07, | |
| "logits/chosen": -2.227515935897827, | |
| "logits/rejected": -2.0700523853302, | |
| "logps/chosen": -279.75494384765625, | |
| "logps/rejected": -127.5013198852539, | |
| "loss": 0.0517, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": 0.12594351172447205, | |
| "rewards/margins": 6.841870307922363, | |
| "rewards/rejected": -6.715925693511963, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.346301020408163e-07, | |
| "logits/chosen": -2.196239948272705, | |
| "logits/rejected": -2.0718905925750732, | |
| "logps/chosen": -291.7792663574219, | |
| "logps/rejected": -154.97567749023438, | |
| "loss": 0.0363, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.4449889063835144, | |
| "rewards/margins": 6.8718414306640625, | |
| "rewards/rejected": -7.316830635070801, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.314413265306122e-07, | |
| "logits/chosen": -2.2047321796417236, | |
| "logits/rejected": -2.04673171043396, | |
| "logps/chosen": -278.8887023925781, | |
| "logps/rejected": -148.29612731933594, | |
| "loss": 0.0549, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": 0.8236015439033508, | |
| "rewards/margins": 8.402433395385742, | |
| "rewards/rejected": -7.578830718994141, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.282525510204081e-07, | |
| "logits/chosen": -2.2842113971710205, | |
| "logits/rejected": -2.1249568462371826, | |
| "logps/chosen": -252.1570587158203, | |
| "logps/rejected": -131.91351318359375, | |
| "loss": 0.0658, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -0.4603995382785797, | |
| "rewards/margins": 6.588088035583496, | |
| "rewards/rejected": -7.048487186431885, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 4.2506377551020405e-07, | |
| "logits/chosen": -2.2204508781433105, | |
| "logits/rejected": -2.0328991413116455, | |
| "logps/chosen": -277.8280334472656, | |
| "logps/rejected": -131.99388122558594, | |
| "loss": 0.0548, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.6896305084228516, | |
| "rewards/margins": 6.525734901428223, | |
| "rewards/rejected": -7.215365409851074, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 4.21875e-07, | |
| "logits/chosen": -2.281752586364746, | |
| "logits/rejected": -2.1253433227539062, | |
| "logps/chosen": -295.5191345214844, | |
| "logps/rejected": -153.25, | |
| "loss": 0.0481, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -0.4880848824977875, | |
| "rewards/margins": 7.4977922439575195, | |
| "rewards/rejected": -7.98587703704834, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 4.186862244897959e-07, | |
| "logits/chosen": -2.2700979709625244, | |
| "logits/rejected": -2.0792222023010254, | |
| "logps/chosen": -246.54293823242188, | |
| "logps/rejected": -140.80882263183594, | |
| "loss": 0.047, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -0.25689664483070374, | |
| "rewards/margins": 7.780982971191406, | |
| "rewards/rejected": -8.037880897521973, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 4.1549744897959185e-07, | |
| "logits/chosen": -2.2732179164886475, | |
| "logits/rejected": -2.10884690284729, | |
| "logps/chosen": -301.9032897949219, | |
| "logps/rejected": -160.38320922851562, | |
| "loss": 0.0668, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.4491173326969147, | |
| "rewards/margins": 8.146552085876465, | |
| "rewards/rejected": -8.595669746398926, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.123086734693877e-07, | |
| "logits/chosen": -2.259704113006592, | |
| "logits/rejected": -2.1263084411621094, | |
| "logps/chosen": -248.52328491210938, | |
| "logps/rejected": -142.90264892578125, | |
| "loss": 0.037, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.04825396090745926, | |
| "rewards/margins": 7.279501914978027, | |
| "rewards/rejected": -7.327755928039551, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.0911989795918365e-07, | |
| "logits/chosen": -2.1421685218811035, | |
| "logits/rejected": -2.0395822525024414, | |
| "logps/chosen": -253.42160034179688, | |
| "logps/rejected": -134.23333740234375, | |
| "loss": 0.0699, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.702133059501648, | |
| "rewards/margins": 6.782039642333984, | |
| "rewards/rejected": -7.484173774719238, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.059311224489796e-07, | |
| "logits/chosen": -2.2776074409484863, | |
| "logits/rejected": -2.075753688812256, | |
| "logps/chosen": -261.2068176269531, | |
| "logps/rejected": -140.44773864746094, | |
| "loss": 0.0891, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.07892769575119019, | |
| "rewards/margins": 7.501675605773926, | |
| "rewards/rejected": -7.580602169036865, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 4.027423469387755e-07, | |
| "logits/chosen": -2.144608974456787, | |
| "logits/rejected": -2.0202698707580566, | |
| "logps/chosen": -230.21023559570312, | |
| "logps/rejected": -146.96124267578125, | |
| "loss": 0.1102, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -1.1200422048568726, | |
| "rewards/margins": 7.595203399658203, | |
| "rewards/rejected": -8.715245246887207, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 3.9955357142857144e-07, | |
| "logits/chosen": -2.1214406490325928, | |
| "logits/rejected": -1.9482629299163818, | |
| "logps/chosen": -271.87554931640625, | |
| "logps/rejected": -156.08773803710938, | |
| "loss": 0.0559, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -1.0995118618011475, | |
| "rewards/margins": 7.6725311279296875, | |
| "rewards/rejected": -8.772043228149414, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 3.963647959183674e-07, | |
| "logits/chosen": -2.222761392593384, | |
| "logits/rejected": -2.065107822418213, | |
| "logps/chosen": -331.9025573730469, | |
| "logps/rejected": -174.4883575439453, | |
| "loss": 0.0811, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.9387734532356262, | |
| "rewards/margins": 8.301498413085938, | |
| "rewards/rejected": -9.240272521972656, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "eval_logits/chosen": -2.1856682300567627, | |
| "eval_logits/rejected": -2.003657579421997, | |
| "eval_logps/chosen": -304.0705261230469, | |
| "eval_logps/rejected": -161.6812744140625, | |
| "eval_loss": 0.05427403748035431, | |
| "eval_rewards/accuracies": 0.9719827771186829, | |
| "eval_rewards/chosen": -0.516727089881897, | |
| "eval_rewards/margins": 8.396077156066895, | |
| "eval_rewards/rejected": -8.91280460357666, | |
| "eval_runtime": 515.5452, | |
| "eval_samples_per_second": 3.575, | |
| "eval_steps_per_second": 0.113, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 3.931760204081632e-07, | |
| "logits/chosen": -2.1678292751312256, | |
| "logits/rejected": -2.0178639888763428, | |
| "logps/chosen": -281.6822509765625, | |
| "logps/rejected": -152.28237915039062, | |
| "loss": 0.0387, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.07496733963489532, | |
| "rewards/margins": 8.705911636352539, | |
| "rewards/rejected": -8.63094425201416, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 3.8998724489795913e-07, | |
| "logits/chosen": -2.124497652053833, | |
| "logits/rejected": -1.9633057117462158, | |
| "logps/chosen": -272.04278564453125, | |
| "logps/rejected": -154.85665893554688, | |
| "loss": 0.0747, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -0.6167508363723755, | |
| "rewards/margins": 7.811673164367676, | |
| "rewards/rejected": -8.428422927856445, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 3.8679846938775506e-07, | |
| "logits/chosen": -2.0578837394714355, | |
| "logits/rejected": -1.9298282861709595, | |
| "logps/chosen": -290.99639892578125, | |
| "logps/rejected": -168.87124633789062, | |
| "loss": 0.0378, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.20361948013305664, | |
| "rewards/margins": 9.225129127502441, | |
| "rewards/rejected": -9.428749084472656, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 3.83609693877551e-07, | |
| "logits/chosen": -2.1337828636169434, | |
| "logits/rejected": -1.9600626230239868, | |
| "logps/chosen": -297.354248046875, | |
| "logps/rejected": -156.31643676757812, | |
| "loss": 0.0215, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.2819109261035919, | |
| "rewards/margins": 8.338506698608398, | |
| "rewards/rejected": -8.056594848632812, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 3.804209183673469e-07, | |
| "logits/chosen": -2.1427624225616455, | |
| "logits/rejected": -1.9869788885116577, | |
| "logps/chosen": -314.75238037109375, | |
| "logps/rejected": -156.91665649414062, | |
| "loss": 0.0719, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.07266904413700104, | |
| "rewards/margins": 8.285478591918945, | |
| "rewards/rejected": -8.358146667480469, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 3.7723214285714285e-07, | |
| "logits/chosen": -2.193801164627075, | |
| "logits/rejected": -2.023911476135254, | |
| "logps/chosen": -280.3459777832031, | |
| "logps/rejected": -159.136474609375, | |
| "loss": 0.043, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.6412056684494019, | |
| "rewards/margins": 8.589024543762207, | |
| "rewards/rejected": -9.230230331420898, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 3.7404336734693873e-07, | |
| "logits/chosen": -2.1136326789855957, | |
| "logits/rejected": -1.970041275024414, | |
| "logps/chosen": -288.44378662109375, | |
| "logps/rejected": -163.71575927734375, | |
| "loss": 0.0628, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -0.7930206656455994, | |
| "rewards/margins": 8.54659652709961, | |
| "rewards/rejected": -9.339616775512695, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 3.7085459183673466e-07, | |
| "logits/chosen": -2.181980848312378, | |
| "logits/rejected": -2.0017356872558594, | |
| "logps/chosen": -269.84637451171875, | |
| "logps/rejected": -161.8075714111328, | |
| "loss": 0.0811, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -1.1092716455459595, | |
| "rewards/margins": 8.634279251098633, | |
| "rewards/rejected": -9.743551254272461, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 3.676658163265306e-07, | |
| "logits/chosen": -2.1273303031921387, | |
| "logits/rejected": -1.9688310623168945, | |
| "logps/chosen": -321.7066650390625, | |
| "logps/rejected": -174.16650390625, | |
| "loss": 0.0425, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.19933632016181946, | |
| "rewards/margins": 9.354225158691406, | |
| "rewards/rejected": -9.553561210632324, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 3.644770408163265e-07, | |
| "logits/chosen": -2.167726993560791, | |
| "logits/rejected": -1.9377820491790771, | |
| "logps/chosen": -290.7398376464844, | |
| "logps/rejected": -156.63296508789062, | |
| "loss": 0.0404, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.34938400983810425, | |
| "rewards/margins": 8.953282356262207, | |
| "rewards/rejected": -9.302667617797852, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 3.6128826530612245e-07, | |
| "logits/chosen": -2.037559747695923, | |
| "logits/rejected": -1.890639305114746, | |
| "logps/chosen": -251.7969970703125, | |
| "logps/rejected": -145.12217712402344, | |
| "loss": 0.0411, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -0.910443902015686, | |
| "rewards/margins": 7.705912113189697, | |
| "rewards/rejected": -8.61635684967041, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 3.580994897959184e-07, | |
| "logits/chosen": -2.1508290767669678, | |
| "logits/rejected": -1.962224006652832, | |
| "logps/chosen": -297.48358154296875, | |
| "logps/rejected": -163.69259643554688, | |
| "loss": 0.0546, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.33932703733444214, | |
| "rewards/margins": 9.108827590942383, | |
| "rewards/rejected": -9.448153495788574, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 3.5491071428571426e-07, | |
| "logits/chosen": -2.179138422012329, | |
| "logits/rejected": -2.000100612640381, | |
| "logps/chosen": -311.79119873046875, | |
| "logps/rejected": -169.6217041015625, | |
| "loss": 0.044, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.6057596206665039, | |
| "rewards/margins": 8.855595588684082, | |
| "rewards/rejected": -9.461355209350586, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 3.517219387755102e-07, | |
| "logits/chosen": -2.2315735816955566, | |
| "logits/rejected": -2.1057209968566895, | |
| "logps/chosen": -287.6869812011719, | |
| "logps/rejected": -158.8007354736328, | |
| "loss": 0.0373, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.49593955278396606, | |
| "rewards/margins": 7.963850975036621, | |
| "rewards/rejected": -8.459790229797363, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 3.485331632653061e-07, | |
| "logits/chosen": -2.150575876235962, | |
| "logits/rejected": -2.0577378273010254, | |
| "logps/chosen": -261.49163818359375, | |
| "logps/rejected": -165.69863891601562, | |
| "loss": 0.0429, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -0.6350365281105042, | |
| "rewards/margins": 8.936600685119629, | |
| "rewards/rejected": -9.571638107299805, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.4534438775510205e-07, | |
| "logits/chosen": -2.080155849456787, | |
| "logits/rejected": -1.9665069580078125, | |
| "logps/chosen": -288.9784240722656, | |
| "logps/rejected": -173.96405029296875, | |
| "loss": 0.0592, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -0.8485971689224243, | |
| "rewards/margins": 9.569430351257324, | |
| "rewards/rejected": -10.418027877807617, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.4215561224489793e-07, | |
| "logits/chosen": -2.1020286083221436, | |
| "logits/rejected": -2.015242338180542, | |
| "logps/chosen": -286.51593017578125, | |
| "logps/rejected": -197.00831604003906, | |
| "loss": 0.0479, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.8589476346969604, | |
| "rewards/margins": 11.049829483032227, | |
| "rewards/rejected": -11.90877628326416, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 3.3896683673469386e-07, | |
| "logits/chosen": -2.207352876663208, | |
| "logits/rejected": -2.026625394821167, | |
| "logps/chosen": -283.9382019042969, | |
| "logps/rejected": -170.56997680664062, | |
| "loss": 0.0686, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -1.428898811340332, | |
| "rewards/margins": 8.341761589050293, | |
| "rewards/rejected": -9.770662307739258, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 3.3577806122448974e-07, | |
| "logits/chosen": -2.226600408554077, | |
| "logits/rejected": -1.9801476001739502, | |
| "logps/chosen": -318.254150390625, | |
| "logps/rejected": -169.10043334960938, | |
| "loss": 0.0859, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.8905173540115356, | |
| "rewards/margins": 8.872238159179688, | |
| "rewards/rejected": -9.762755393981934, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 3.3258928571428567e-07, | |
| "logits/chosen": -2.1024184226989746, | |
| "logits/rejected": -1.9208921194076538, | |
| "logps/chosen": -336.98724365234375, | |
| "logps/rejected": -168.90625, | |
| "loss": 0.0523, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.09593725204467773, | |
| "rewards/margins": 10.029386520385742, | |
| "rewards/rejected": -10.125323295593262, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 3.294005102040816e-07, | |
| "logits/chosen": -2.137101650238037, | |
| "logits/rejected": -1.9699983596801758, | |
| "logps/chosen": -261.7487487792969, | |
| "logps/rejected": -156.60157775878906, | |
| "loss": 0.0614, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.6025804281234741, | |
| "rewards/margins": 8.209660530090332, | |
| "rewards/rejected": -8.812241554260254, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 3.2621173469387753e-07, | |
| "logits/chosen": -2.2646114826202393, | |
| "logits/rejected": -2.1247591972351074, | |
| "logps/chosen": -329.1217041015625, | |
| "logps/rejected": -185.3964385986328, | |
| "loss": 0.0578, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -1.2518744468688965, | |
| "rewards/margins": 9.136481285095215, | |
| "rewards/rejected": -10.388354301452637, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.2302295918367346e-07, | |
| "logits/chosen": -2.3025715351104736, | |
| "logits/rejected": -2.1256372928619385, | |
| "logps/chosen": -310.5058288574219, | |
| "logps/rejected": -170.10592651367188, | |
| "loss": 0.029, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.1804181337356567, | |
| "rewards/margins": 8.121767044067383, | |
| "rewards/rejected": -9.302184104919434, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.198341836734694e-07, | |
| "logits/chosen": -2.2341511249542236, | |
| "logits/rejected": -2.04396915435791, | |
| "logps/chosen": -328.27667236328125, | |
| "logps/rejected": -192.09335327148438, | |
| "loss": 0.0565, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.02377413585782051, | |
| "rewards/margins": 11.305741310119629, | |
| "rewards/rejected": -11.32951545715332, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 3.1664540816326527e-07, | |
| "logits/chosen": -2.1826584339141846, | |
| "logits/rejected": -2.023871660232544, | |
| "logps/chosen": -276.5951843261719, | |
| "logps/rejected": -160.23538208007812, | |
| "loss": 0.0376, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -0.3585100769996643, | |
| "rewards/margins": 9.415512084960938, | |
| "rewards/rejected": -9.77402114868164, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 3.134566326530612e-07, | |
| "logits/chosen": -2.227428436279297, | |
| "logits/rejected": -2.070985794067383, | |
| "logps/chosen": -294.4335021972656, | |
| "logps/rejected": -169.70211791992188, | |
| "loss": 0.0343, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.31169813871383667, | |
| "rewards/margins": 9.067781448364258, | |
| "rewards/rejected": -9.379480361938477, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 3.1026785714285713e-07, | |
| "logits/chosen": -2.1913247108459473, | |
| "logits/rejected": -1.9936134815216064, | |
| "logps/chosen": -328.67083740234375, | |
| "logps/rejected": -175.294921875, | |
| "loss": 0.0432, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.3559989333152771, | |
| "rewards/margins": 9.043487548828125, | |
| "rewards/rejected": -9.399484634399414, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 3.0707908163265306e-07, | |
| "logits/chosen": -2.1028852462768555, | |
| "logits/rejected": -1.8983999490737915, | |
| "logps/chosen": -273.2193908691406, | |
| "logps/rejected": -158.0215301513672, | |
| "loss": 0.0428, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.963762104511261, | |
| "rewards/margins": 8.092933654785156, | |
| "rewards/rejected": -9.056695938110352, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 3.03890306122449e-07, | |
| "logits/chosen": -2.1897473335266113, | |
| "logits/rejected": -1.9860061407089233, | |
| "logps/chosen": -317.2612609863281, | |
| "logps/rejected": -167.63690185546875, | |
| "loss": 0.0308, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.3040831685066223, | |
| "rewards/margins": 8.782797813415527, | |
| "rewards/rejected": -9.086880683898926, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 3.007015306122449e-07, | |
| "logits/chosen": -2.2554848194122314, | |
| "logits/rejected": -2.0466835498809814, | |
| "logps/chosen": -359.5973815917969, | |
| "logps/rejected": -168.97488403320312, | |
| "loss": 0.0485, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": 0.31903451681137085, | |
| "rewards/margins": 9.262194633483887, | |
| "rewards/rejected": -8.943160057067871, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 2.975127551020408e-07, | |
| "logits/chosen": -2.0375688076019287, | |
| "logits/rejected": -1.8554328680038452, | |
| "logps/chosen": -321.64581298828125, | |
| "logps/rejected": -150.6351776123047, | |
| "loss": 0.0425, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.05941831320524216, | |
| "rewards/margins": 8.272252082824707, | |
| "rewards/rejected": -8.331669807434082, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 2.9432397959183673e-07, | |
| "logits/chosen": -2.2465949058532715, | |
| "logits/rejected": -2.003244638442993, | |
| "logps/chosen": -339.7053527832031, | |
| "logps/rejected": -173.89718627929688, | |
| "loss": 0.0453, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -1.0983786582946777, | |
| "rewards/margins": 8.536051750183105, | |
| "rewards/rejected": -9.634429931640625, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 2.911352040816326e-07, | |
| "logits/chosen": -2.157315254211426, | |
| "logits/rejected": -2.0112011432647705, | |
| "logps/chosen": -285.68023681640625, | |
| "logps/rejected": -163.58163452148438, | |
| "loss": 0.0364, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -0.5791069269180298, | |
| "rewards/margins": 8.840615272521973, | |
| "rewards/rejected": -9.419723510742188, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 2.8794642857142854e-07, | |
| "logits/chosen": -2.1603357791900635, | |
| "logits/rejected": -1.983139991760254, | |
| "logps/chosen": -275.4844665527344, | |
| "logps/rejected": -165.48094177246094, | |
| "loss": 0.036, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.849933922290802, | |
| "rewards/margins": 8.592988967895508, | |
| "rewards/rejected": -9.442923545837402, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 2.8475765306122447e-07, | |
| "logits/chosen": -2.1025400161743164, | |
| "logits/rejected": -1.902698278427124, | |
| "logps/chosen": -294.97662353515625, | |
| "logps/rejected": -164.88980102539062, | |
| "loss": 0.0255, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": 0.18203288316726685, | |
| "rewards/margins": 9.839866638183594, | |
| "rewards/rejected": -9.657832145690918, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 2.815688775510204e-07, | |
| "logits/chosen": -2.1174845695495605, | |
| "logits/rejected": -1.9745981693267822, | |
| "logps/chosen": -289.689208984375, | |
| "logps/rejected": -225.04946899414062, | |
| "loss": 0.0306, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.11574029922485352, | |
| "rewards/margins": 14.800374031066895, | |
| "rewards/rejected": -14.916116714477539, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.783801020408163e-07, | |
| "logits/chosen": -2.099398136138916, | |
| "logits/rejected": -1.9381110668182373, | |
| "logps/chosen": -280.7049255371094, | |
| "logps/rejected": -162.89515686035156, | |
| "loss": 0.0412, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.6393367052078247, | |
| "rewards/margins": 9.100909233093262, | |
| "rewards/rejected": -9.740245819091797, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.751913265306122e-07, | |
| "logits/chosen": -2.130143404006958, | |
| "logits/rejected": -1.8786506652832031, | |
| "logps/chosen": -254.42440795898438, | |
| "logps/rejected": -148.86289978027344, | |
| "loss": 0.0416, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.5891353487968445, | |
| "rewards/margins": 8.205097198486328, | |
| "rewards/rejected": -8.794232368469238, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 2.7200255102040814e-07, | |
| "logits/chosen": -2.0689034461975098, | |
| "logits/rejected": -2.004838466644287, | |
| "logps/chosen": -253.21847534179688, | |
| "logps/rejected": -173.26658630371094, | |
| "loss": 0.0437, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.9242547154426575, | |
| "rewards/margins": 8.997157096862793, | |
| "rewards/rejected": -9.921411514282227, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 2.6881377551020407e-07, | |
| "logits/chosen": -2.24395751953125, | |
| "logits/rejected": -2.0530552864074707, | |
| "logps/chosen": -276.29327392578125, | |
| "logps/rejected": -168.1978759765625, | |
| "loss": 0.0338, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.6147713661193848, | |
| "rewards/margins": 9.07136058807373, | |
| "rewards/rejected": -9.686132431030273, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 2.65625e-07, | |
| "logits/chosen": -2.1872873306274414, | |
| "logits/rejected": -1.9537794589996338, | |
| "logps/chosen": -313.37164306640625, | |
| "logps/rejected": -169.32046508789062, | |
| "loss": 0.0408, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.2682819664478302, | |
| "rewards/margins": 9.430694580078125, | |
| "rewards/rejected": -9.69897747039795, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 2.6243622448979593e-07, | |
| "logits/chosen": -2.152787208557129, | |
| "logits/rejected": -1.9356346130371094, | |
| "logps/chosen": -280.98541259765625, | |
| "logps/rejected": -152.12704467773438, | |
| "loss": 0.0839, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.25039607286453247, | |
| "rewards/margins": 8.020809173583984, | |
| "rewards/rejected": -8.271204948425293, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 2.592474489795918e-07, | |
| "logits/chosen": -2.0773465633392334, | |
| "logits/rejected": -1.9478585720062256, | |
| "logps/chosen": -294.1889953613281, | |
| "logps/rejected": -162.88656616210938, | |
| "loss": 0.0256, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.10252566635608673, | |
| "rewards/margins": 9.275411605834961, | |
| "rewards/rejected": -9.377937316894531, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.5605867346938774e-07, | |
| "logits/chosen": -2.1434168815612793, | |
| "logits/rejected": -2.0151402950286865, | |
| "logps/chosen": -255.9226531982422, | |
| "logps/rejected": -164.32913208007812, | |
| "loss": 0.033, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.7857500314712524, | |
| "rewards/margins": 9.118104934692383, | |
| "rewards/rejected": -9.903855323791504, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 2.5286989795918367e-07, | |
| "logits/chosen": -2.177273750305176, | |
| "logits/rejected": -2.035215139389038, | |
| "logps/chosen": -275.280517578125, | |
| "logps/rejected": -174.3971405029297, | |
| "loss": 0.0274, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.5861722826957703, | |
| "rewards/margins": 9.521924018859863, | |
| "rewards/rejected": -10.1080961227417, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 2.4968112244897954e-07, | |
| "logits/chosen": -2.123955011367798, | |
| "logits/rejected": -1.9494431018829346, | |
| "logps/chosen": -271.4372863769531, | |
| "logps/rejected": -159.25759887695312, | |
| "loss": 0.0183, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.16222012042999268, | |
| "rewards/margins": 8.81199836730957, | |
| "rewards/rejected": -8.974218368530273, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 2.464923469387755e-07, | |
| "logits/chosen": -2.1805062294006348, | |
| "logits/rejected": -1.9929234981536865, | |
| "logps/chosen": -312.1669006347656, | |
| "logps/rejected": -173.0575714111328, | |
| "loss": 0.0542, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.16012373566627502, | |
| "rewards/margins": 10.075935363769531, | |
| "rewards/rejected": -9.915812492370605, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 2.433035714285714e-07, | |
| "logits/chosen": -2.1818442344665527, | |
| "logits/rejected": -1.9997507333755493, | |
| "logps/chosen": -302.0064392089844, | |
| "logps/rejected": -165.9043731689453, | |
| "loss": 0.0448, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.39435887336730957, | |
| "rewards/margins": 8.824365615844727, | |
| "rewards/rejected": -9.21872329711914, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 2.4011479591836734e-07, | |
| "logits/chosen": -2.17393159866333, | |
| "logits/rejected": -2.0282928943634033, | |
| "logps/chosen": -297.4441833496094, | |
| "logps/rejected": -168.88804626464844, | |
| "loss": 0.0296, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.8467293977737427, | |
| "rewards/margins": 8.895944595336914, | |
| "rewards/rejected": -9.742671966552734, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 2.3692602040816327e-07, | |
| "logits/chosen": -2.210867166519165, | |
| "logits/rejected": -1.9793678522109985, | |
| "logps/chosen": -287.63763427734375, | |
| "logps/rejected": -174.19747924804688, | |
| "loss": 0.0362, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.5712664723396301, | |
| "rewards/margins": 10.053550720214844, | |
| "rewards/rejected": -10.62481689453125, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "eval_logits/chosen": -2.203035593032837, | |
| "eval_logits/rejected": -2.0113065242767334, | |
| "eval_logps/chosen": -303.8834228515625, | |
| "eval_logps/rejected": -168.3771209716797, | |
| "eval_loss": 0.04832219332456589, | |
| "eval_rewards/accuracies": 0.9719827771186829, | |
| "eval_rewards/chosen": -0.4980185925960541, | |
| "eval_rewards/margins": 9.084371566772461, | |
| "eval_rewards/rejected": -9.582389831542969, | |
| "eval_runtime": 501.7343, | |
| "eval_samples_per_second": 3.673, | |
| "eval_steps_per_second": 0.116, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2.3373724489795917e-07, | |
| "logits/chosen": -2.2325656414031982, | |
| "logits/rejected": -2.099182605743408, | |
| "logps/chosen": -298.896240234375, | |
| "logps/rejected": -178.80360412597656, | |
| "loss": 0.0381, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -1.1191880702972412, | |
| "rewards/margins": 8.851495742797852, | |
| "rewards/rejected": -9.970685005187988, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 2.305484693877551e-07, | |
| "logits/chosen": -2.1347708702087402, | |
| "logits/rejected": -1.9578421115875244, | |
| "logps/chosen": -310.70367431640625, | |
| "logps/rejected": -173.8622283935547, | |
| "loss": 0.022, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.24546487629413605, | |
| "rewards/margins": 9.734888076782227, | |
| "rewards/rejected": -9.980352401733398, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 2.2735969387755103e-07, | |
| "logits/chosen": -2.114215135574341, | |
| "logits/rejected": -1.9470380544662476, | |
| "logps/chosen": -259.62371826171875, | |
| "logps/rejected": -150.71444702148438, | |
| "loss": 0.0263, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.3662714660167694, | |
| "rewards/margins": 8.222164154052734, | |
| "rewards/rejected": -8.588435173034668, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 2.241709183673469e-07, | |
| "logits/chosen": -2.0746610164642334, | |
| "logits/rejected": -1.9087474346160889, | |
| "logps/chosen": -308.8531799316406, | |
| "logps/rejected": -173.00241088867188, | |
| "loss": 0.0478, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.7547154426574707, | |
| "rewards/margins": 9.060162544250488, | |
| "rewards/rejected": -9.814878463745117, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 2.2098214285714284e-07, | |
| "logits/chosen": -2.1720070838928223, | |
| "logits/rejected": -1.9714542627334595, | |
| "logps/chosen": -282.46026611328125, | |
| "logps/rejected": -165.5819549560547, | |
| "loss": 0.0484, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.8778654336929321, | |
| "rewards/margins": 8.815220832824707, | |
| "rewards/rejected": -9.693086624145508, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 2.1779336734693877e-07, | |
| "logits/chosen": -2.160717010498047, | |
| "logits/rejected": -2.0338234901428223, | |
| "logps/chosen": -293.3428649902344, | |
| "logps/rejected": -180.75949096679688, | |
| "loss": 0.0232, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.8398637771606445, | |
| "rewards/margins": 9.461033821105957, | |
| "rewards/rejected": -10.300897598266602, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 2.1460459183673467e-07, | |
| "logits/chosen": -2.0859107971191406, | |
| "logits/rejected": -1.9392955303192139, | |
| "logps/chosen": -269.2712707519531, | |
| "logps/rejected": -160.63858032226562, | |
| "loss": 0.048, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -0.8251531720161438, | |
| "rewards/margins": 8.415184020996094, | |
| "rewards/rejected": -9.240337371826172, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 2.114158163265306e-07, | |
| "logits/chosen": -2.114478349685669, | |
| "logits/rejected": -2.0172407627105713, | |
| "logps/chosen": -313.1733093261719, | |
| "logps/rejected": -183.88912963867188, | |
| "loss": 0.02, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.37718772888183594, | |
| "rewards/margins": 10.156351089477539, | |
| "rewards/rejected": -10.533537864685059, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 2.0822704081632654e-07, | |
| "logits/chosen": -2.151115894317627, | |
| "logits/rejected": -1.9639852046966553, | |
| "logps/chosen": -304.9864196777344, | |
| "logps/rejected": -179.6061248779297, | |
| "loss": 0.0374, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.8430510759353638, | |
| "rewards/margins": 9.594018936157227, | |
| "rewards/rejected": -10.4370698928833, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 2.0503826530612244e-07, | |
| "logits/chosen": -2.07926082611084, | |
| "logits/rejected": -1.8545608520507812, | |
| "logps/chosen": -287.17510986328125, | |
| "logps/rejected": -175.82423400878906, | |
| "loss": 0.025, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.6869755387306213, | |
| "rewards/margins": 9.874262809753418, | |
| "rewards/rejected": -10.561239242553711, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 2.0184948979591837e-07, | |
| "logits/chosen": -2.0271568298339844, | |
| "logits/rejected": -1.9322795867919922, | |
| "logps/chosen": -254.9891815185547, | |
| "logps/rejected": -171.43470764160156, | |
| "loss": 0.0285, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.40473413467407227, | |
| "rewards/margins": 10.018217086791992, | |
| "rewards/rejected": -10.422952651977539, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 1.9866071428571427e-07, | |
| "logits/chosen": -1.9903011322021484, | |
| "logits/rejected": -1.9317058324813843, | |
| "logps/chosen": -239.4633026123047, | |
| "logps/rejected": -167.33950805664062, | |
| "loss": 0.052, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": -1.950304627418518, | |
| "rewards/margins": 8.423638343811035, | |
| "rewards/rejected": -10.373943328857422, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.9547193877551018e-07, | |
| "logits/chosen": -2.1156675815582275, | |
| "logits/rejected": -1.9115331172943115, | |
| "logps/chosen": -283.9968566894531, | |
| "logps/rejected": -171.2687225341797, | |
| "loss": 0.0356, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.6293619275093079, | |
| "rewards/margins": 9.327872276306152, | |
| "rewards/rejected": -9.957234382629395, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.922831632653061e-07, | |
| "logits/chosen": -2.164762020111084, | |
| "logits/rejected": -1.9720051288604736, | |
| "logps/chosen": -332.24481201171875, | |
| "logps/rejected": -172.13970947265625, | |
| "loss": 0.0394, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.5399401783943176, | |
| "rewards/margins": 8.87366008758545, | |
| "rewards/rejected": -9.413599967956543, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 1.8909438775510204e-07, | |
| "logits/chosen": -2.0545859336853027, | |
| "logits/rejected": -1.916473627090454, | |
| "logps/chosen": -279.626953125, | |
| "logps/rejected": -176.50621032714844, | |
| "loss": 0.0147, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.018849492073059, | |
| "rewards/margins": 9.826019287109375, | |
| "rewards/rejected": -10.844868659973145, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.8590561224489794e-07, | |
| "logits/chosen": -2.088841438293457, | |
| "logits/rejected": -1.9257131814956665, | |
| "logps/chosen": -292.3592834472656, | |
| "logps/rejected": -173.84017944335938, | |
| "loss": 0.0291, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -1.085963487625122, | |
| "rewards/margins": 9.38125991821289, | |
| "rewards/rejected": -10.467223167419434, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.8271683673469387e-07, | |
| "logits/chosen": -2.1992461681365967, | |
| "logits/rejected": -2.0129992961883545, | |
| "logps/chosen": -298.34307861328125, | |
| "logps/rejected": -182.71925354003906, | |
| "loss": 0.039, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -1.1279481649398804, | |
| "rewards/margins": 9.488454818725586, | |
| "rewards/rejected": -10.61640453338623, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 1.795280612244898e-07, | |
| "logits/chosen": -2.144819974899292, | |
| "logits/rejected": -1.9659067392349243, | |
| "logps/chosen": -256.8332214355469, | |
| "logps/rejected": -152.0546112060547, | |
| "loss": 0.0389, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.4352855682373047, | |
| "rewards/margins": 7.898017883300781, | |
| "rewards/rejected": -8.33330249786377, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 1.763392857142857e-07, | |
| "logits/chosen": -2.188812494277954, | |
| "logits/rejected": -2.0105535984039307, | |
| "logps/chosen": -311.55633544921875, | |
| "logps/rejected": -167.11927795410156, | |
| "loss": 0.0237, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.43777579069137573, | |
| "rewards/margins": 8.915369987487793, | |
| "rewards/rejected": -9.35314655303955, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.731505102040816e-07, | |
| "logits/chosen": -2.087813377380371, | |
| "logits/rejected": -1.9207570552825928, | |
| "logps/chosen": -271.0974426269531, | |
| "logps/rejected": -211.95523071289062, | |
| "loss": 0.0268, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.8373743295669556, | |
| "rewards/margins": 13.245565414428711, | |
| "rewards/rejected": -14.082939147949219, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.6996173469387754e-07, | |
| "logits/chosen": -2.1115288734436035, | |
| "logits/rejected": -1.9904811382293701, | |
| "logps/chosen": -302.61138916015625, | |
| "logps/rejected": -177.2365264892578, | |
| "loss": 0.0235, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.5768619775772095, | |
| "rewards/margins": 8.965620040893555, | |
| "rewards/rejected": -9.54248332977295, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.6677295918367345e-07, | |
| "logits/chosen": -2.1301310062408447, | |
| "logits/rejected": -1.879533052444458, | |
| "logps/chosen": -332.2015686035156, | |
| "logps/rejected": -165.51576232910156, | |
| "loss": 0.0481, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.5949792861938477, | |
| "rewards/margins": 9.159207344055176, | |
| "rewards/rejected": -9.754185676574707, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 1.6358418367346938e-07, | |
| "logits/chosen": -2.1042232513427734, | |
| "logits/rejected": -1.8513002395629883, | |
| "logps/chosen": -344.7572937011719, | |
| "logps/rejected": -172.65744018554688, | |
| "loss": 0.0231, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.2580934464931488, | |
| "rewards/margins": 9.699871063232422, | |
| "rewards/rejected": -9.957964897155762, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 1.603954081632653e-07, | |
| "logits/chosen": -2.068237066268921, | |
| "logits/rejected": -1.8824779987335205, | |
| "logps/chosen": -284.03509521484375, | |
| "logps/rejected": -162.17001342773438, | |
| "loss": 0.0275, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.403529554605484, | |
| "rewards/margins": 9.167604446411133, | |
| "rewards/rejected": -9.571134567260742, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 1.572066326530612e-07, | |
| "logits/chosen": -2.1395645141601562, | |
| "logits/rejected": -1.9513561725616455, | |
| "logps/chosen": -342.67254638671875, | |
| "logps/rejected": -174.70993041992188, | |
| "loss": 0.0345, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.07841818779706955, | |
| "rewards/margins": 9.598522186279297, | |
| "rewards/rejected": -9.676939964294434, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 1.5401785714285714e-07, | |
| "logits/chosen": -2.126796245574951, | |
| "logits/rejected": -1.9961979389190674, | |
| "logps/chosen": -316.7900695800781, | |
| "logps/rejected": -176.53733825683594, | |
| "loss": 0.0412, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.37025541067123413, | |
| "rewards/margins": 9.903864860534668, | |
| "rewards/rejected": -10.27411937713623, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 1.5082908163265307e-07, | |
| "logits/chosen": -2.082777976989746, | |
| "logits/rejected": -1.9228655099868774, | |
| "logps/chosen": -297.30133056640625, | |
| "logps/rejected": -166.17623901367188, | |
| "loss": 0.0346, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -1.2105169296264648, | |
| "rewards/margins": 8.340936660766602, | |
| "rewards/rejected": -9.551454544067383, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 1.4764030612244895e-07, | |
| "logits/chosen": -2.1340670585632324, | |
| "logits/rejected": -2.0120952129364014, | |
| "logps/chosen": -306.83441162109375, | |
| "logps/rejected": -174.4154510498047, | |
| "loss": 0.0375, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.8796642422676086, | |
| "rewards/margins": 9.489983558654785, | |
| "rewards/rejected": -10.369647979736328, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 1.4445153061224488e-07, | |
| "logits/chosen": -2.180344581604004, | |
| "logits/rejected": -2.054753541946411, | |
| "logps/chosen": -319.8197326660156, | |
| "logps/rejected": -179.91517639160156, | |
| "loss": 0.0368, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.26657119393348694, | |
| "rewards/margins": 10.043919563293457, | |
| "rewards/rejected": -10.310490608215332, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 1.412627551020408e-07, | |
| "logits/chosen": -2.139643907546997, | |
| "logits/rejected": -1.9441308975219727, | |
| "logps/chosen": -280.61492919921875, | |
| "logps/rejected": -168.45120239257812, | |
| "loss": 0.0245, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.6699168086051941, | |
| "rewards/margins": 9.325021743774414, | |
| "rewards/rejected": -9.994939804077148, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 1.3807397959183672e-07, | |
| "logits/chosen": -2.2181859016418457, | |
| "logits/rejected": -2.0422186851501465, | |
| "logps/chosen": -348.22802734375, | |
| "logps/rejected": -184.8292999267578, | |
| "loss": 0.0389, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.8054157495498657, | |
| "rewards/margins": 9.742046356201172, | |
| "rewards/rejected": -10.547462463378906, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 1.3488520408163265e-07, | |
| "logits/chosen": -2.1537463665008545, | |
| "logits/rejected": -1.9965778589248657, | |
| "logps/chosen": -322.0302734375, | |
| "logps/rejected": -177.57180786132812, | |
| "loss": 0.039, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.8039339780807495, | |
| "rewards/margins": 9.590051651000977, | |
| "rewards/rejected": -10.3939847946167, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 1.3169642857142858e-07, | |
| "logits/chosen": -2.120084524154663, | |
| "logits/rejected": -1.944972276687622, | |
| "logps/chosen": -299.9647521972656, | |
| "logps/rejected": -174.93545532226562, | |
| "loss": 0.0365, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -1.0538573265075684, | |
| "rewards/margins": 9.612497329711914, | |
| "rewards/rejected": -10.666356086730957, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 1.2850765306122448e-07, | |
| "logits/chosen": -2.0994505882263184, | |
| "logits/rejected": -1.9836326837539673, | |
| "logps/chosen": -327.00103759765625, | |
| "logps/rejected": -176.6904754638672, | |
| "loss": 0.0426, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.0558558702468872, | |
| "rewards/margins": 9.380393981933594, | |
| "rewards/rejected": -10.436250686645508, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 1.253188775510204e-07, | |
| "logits/chosen": -2.063891887664795, | |
| "logits/rejected": -1.8763225078582764, | |
| "logps/chosen": -245.1425018310547, | |
| "logps/rejected": -155.8802032470703, | |
| "loss": 0.0527, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -1.3180797100067139, | |
| "rewards/margins": 8.703597068786621, | |
| "rewards/rejected": -10.021677017211914, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 1.2213010204081632e-07, | |
| "logits/chosen": -2.1077721118927, | |
| "logits/rejected": -1.9732635021209717, | |
| "logps/chosen": -272.14886474609375, | |
| "logps/rejected": -177.23715209960938, | |
| "loss": 0.038, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.5222896337509155, | |
| "rewards/margins": 9.948299407958984, | |
| "rewards/rejected": -10.470590591430664, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 1.1894132653061225e-07, | |
| "logits/chosen": -2.1442596912384033, | |
| "logits/rejected": -1.995617151260376, | |
| "logps/chosen": -297.60357666015625, | |
| "logps/rejected": -176.26931762695312, | |
| "loss": 0.0451, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -1.6867294311523438, | |
| "rewards/margins": 8.880840301513672, | |
| "rewards/rejected": -10.567569732666016, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 1.1575255102040815e-07, | |
| "logits/chosen": -2.1161892414093018, | |
| "logits/rejected": -1.9788262844085693, | |
| "logps/chosen": -291.28485107421875, | |
| "logps/rejected": -171.2949676513672, | |
| "loss": 0.0414, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -1.612140417098999, | |
| "rewards/margins": 8.652326583862305, | |
| "rewards/rejected": -10.264467239379883, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 1.1256377551020407e-07, | |
| "logits/chosen": -2.1823105812072754, | |
| "logits/rejected": -1.997688889503479, | |
| "logps/chosen": -314.90081787109375, | |
| "logps/rejected": -177.48538208007812, | |
| "loss": 0.0348, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.8158550262451172, | |
| "rewards/margins": 9.316900253295898, | |
| "rewards/rejected": -10.132753372192383, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 1.09375e-07, | |
| "logits/chosen": -2.1604926586151123, | |
| "logits/rejected": -1.92621648311615, | |
| "logps/chosen": -322.7879638671875, | |
| "logps/rejected": -174.97999572753906, | |
| "loss": 0.0337, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.6046790480613708, | |
| "rewards/margins": 9.716318130493164, | |
| "rewards/rejected": -10.32099723815918, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 1.0618622448979592e-07, | |
| "logits/chosen": -2.1750383377075195, | |
| "logits/rejected": -1.9606304168701172, | |
| "logps/chosen": -304.9856872558594, | |
| "logps/rejected": -166.8131103515625, | |
| "loss": 0.0206, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.726199746131897, | |
| "rewards/margins": 9.535404205322266, | |
| "rewards/rejected": -10.261604309082031, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 1.0299744897959183e-07, | |
| "logits/chosen": -2.1826305389404297, | |
| "logits/rejected": -1.9840447902679443, | |
| "logps/chosen": -287.8218078613281, | |
| "logps/rejected": -172.24215698242188, | |
| "loss": 0.0326, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -1.0618003606796265, | |
| "rewards/margins": 9.063058853149414, | |
| "rewards/rejected": -10.124858856201172, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 9.980867346938775e-08, | |
| "logits/chosen": -2.1716151237487793, | |
| "logits/rejected": -2.0073437690734863, | |
| "logps/chosen": -284.8751220703125, | |
| "logps/rejected": -177.5775146484375, | |
| "loss": 0.0339, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.5093313455581665, | |
| "rewards/margins": 10.088600158691406, | |
| "rewards/rejected": -10.597929954528809, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 9.661989795918367e-08, | |
| "logits/chosen": -2.136314868927002, | |
| "logits/rejected": -1.9978504180908203, | |
| "logps/chosen": -299.4578857421875, | |
| "logps/rejected": -177.2023468017578, | |
| "loss": 0.0253, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.22061769664287567, | |
| "rewards/margins": 10.269851684570312, | |
| "rewards/rejected": -10.490468978881836, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 9.343112244897958e-08, | |
| "logits/chosen": -2.1121840476989746, | |
| "logits/rejected": -1.8750556707382202, | |
| "logps/chosen": -294.89984130859375, | |
| "logps/rejected": -169.1930389404297, | |
| "loss": 0.0209, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.6526461839675903, | |
| "rewards/margins": 9.879607200622559, | |
| "rewards/rejected": -10.532255172729492, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 9.024234693877552e-08, | |
| "logits/chosen": -2.146242141723633, | |
| "logits/rejected": -1.8776633739471436, | |
| "logps/chosen": -313.9344177246094, | |
| "logps/rejected": -166.5650177001953, | |
| "loss": 0.0308, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": 0.13873246312141418, | |
| "rewards/margins": 9.890599250793457, | |
| "rewards/rejected": -9.751866340637207, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 8.705357142857142e-08, | |
| "logits/chosen": -2.079200506210327, | |
| "logits/rejected": -1.8900413513183594, | |
| "logps/chosen": -302.3304443359375, | |
| "logps/rejected": -163.3623809814453, | |
| "loss": 0.0126, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.21449720859527588, | |
| "rewards/margins": 9.112278938293457, | |
| "rewards/rejected": -9.326776504516602, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 8.386479591836734e-08, | |
| "logits/chosen": -2.063764810562134, | |
| "logits/rejected": -1.9468538761138916, | |
| "logps/chosen": -271.62640380859375, | |
| "logps/rejected": -183.78414916992188, | |
| "loss": 0.0364, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.5422567129135132, | |
| "rewards/margins": 10.014627456665039, | |
| "rewards/rejected": -10.556883811950684, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 8.067602040816327e-08, | |
| "logits/chosen": -2.0192930698394775, | |
| "logits/rejected": -1.8542463779449463, | |
| "logps/chosen": -282.27069091796875, | |
| "logps/rejected": -164.059326171875, | |
| "loss": 0.0239, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.4797702729701996, | |
| "rewards/margins": 9.287493705749512, | |
| "rewards/rejected": -9.767263412475586, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 7.748724489795918e-08, | |
| "logits/chosen": -2.097482204437256, | |
| "logits/rejected": -1.9764598608016968, | |
| "logps/chosen": -273.4100341796875, | |
| "logps/rejected": -175.4436492919922, | |
| "loss": 0.0318, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -0.9027210474014282, | |
| "rewards/margins": 9.17738151550293, | |
| "rewards/rejected": -10.080102920532227, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "eval_logits/chosen": -2.1461496353149414, | |
| "eval_logits/rejected": -1.9506442546844482, | |
| "eval_logps/chosen": -307.3617248535156, | |
| "eval_logps/rejected": -178.54031372070312, | |
| "eval_loss": 0.04423515126109123, | |
| "eval_rewards/accuracies": 0.9719827771186829, | |
| "eval_rewards/chosen": -0.8458480834960938, | |
| "eval_rewards/margins": 9.752861022949219, | |
| "eval_rewards/rejected": -10.598710060119629, | |
| "eval_runtime": 498.2368, | |
| "eval_samples_per_second": 3.699, | |
| "eval_steps_per_second": 0.116, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 7.429846938775509e-08, | |
| "logits/chosen": -2.1621012687683105, | |
| "logits/rejected": -1.9208993911743164, | |
| "logps/chosen": -343.2023010253906, | |
| "logps/rejected": -189.607666015625, | |
| "loss": 0.0304, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.39647844433784485, | |
| "rewards/margins": 10.542058944702148, | |
| "rewards/rejected": -10.938536643981934, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 7.110969387755102e-08, | |
| "logits/chosen": -2.077564239501953, | |
| "logits/rejected": -1.8879787921905518, | |
| "logps/chosen": -285.5229797363281, | |
| "logps/rejected": -171.03269958496094, | |
| "loss": 0.0418, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -1.0820773839950562, | |
| "rewards/margins": 9.324197769165039, | |
| "rewards/rejected": -10.406275749206543, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 6.792091836734694e-08, | |
| "logits/chosen": -2.1050283908843994, | |
| "logits/rejected": -1.9014123678207397, | |
| "logps/chosen": -291.99993896484375, | |
| "logps/rejected": -167.6908721923828, | |
| "loss": 0.0245, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": 0.04504892975091934, | |
| "rewards/margins": 10.022317886352539, | |
| "rewards/rejected": -9.97726821899414, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 6.473214285714285e-08, | |
| "logits/chosen": -2.155641794204712, | |
| "logits/rejected": -1.9214122295379639, | |
| "logps/chosen": -276.1907958984375, | |
| "logps/rejected": -168.76535034179688, | |
| "loss": 0.0403, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.5811057090759277, | |
| "rewards/margins": 9.711630821228027, | |
| "rewards/rejected": -10.292736053466797, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 6.154336734693877e-08, | |
| "logits/chosen": -2.0060572624206543, | |
| "logits/rejected": -1.8284757137298584, | |
| "logps/chosen": -251.60006713867188, | |
| "logps/rejected": -161.85813903808594, | |
| "loss": 0.0558, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -1.079087495803833, | |
| "rewards/margins": 8.86231803894043, | |
| "rewards/rejected": -9.94140625, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 5.835459183673469e-08, | |
| "logits/chosen": -2.1795308589935303, | |
| "logits/rejected": -1.9668638706207275, | |
| "logps/chosen": -307.10614013671875, | |
| "logps/rejected": -198.31088256835938, | |
| "loss": 0.0456, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.6056522130966187, | |
| "rewards/margins": 12.019081115722656, | |
| "rewards/rejected": -12.62473201751709, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 5.516581632653061e-08, | |
| "logits/chosen": -1.972507119178772, | |
| "logits/rejected": -1.8055419921875, | |
| "logps/chosen": -260.49920654296875, | |
| "logps/rejected": -156.6437225341797, | |
| "loss": 0.0285, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.8670014142990112, | |
| "rewards/margins": 8.884390830993652, | |
| "rewards/rejected": -9.751392364501953, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 5.197704081632653e-08, | |
| "logits/chosen": -2.0382626056671143, | |
| "logits/rejected": -1.8945573568344116, | |
| "logps/chosen": -251.30612182617188, | |
| "logps/rejected": -166.97303771972656, | |
| "loss": 0.0475, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.7538485527038574, | |
| "rewards/margins": 9.353151321411133, | |
| "rewards/rejected": -10.107000350952148, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 4.878826530612245e-08, | |
| "logits/chosen": -2.1863982677459717, | |
| "logits/rejected": -1.9323434829711914, | |
| "logps/chosen": -323.969970703125, | |
| "logps/rejected": -168.3311767578125, | |
| "loss": 0.0474, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -0.45526188611984253, | |
| "rewards/margins": 9.516814231872559, | |
| "rewards/rejected": -9.972076416015625, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 4.5599489795918364e-08, | |
| "logits/chosen": -2.0845093727111816, | |
| "logits/rejected": -1.8628953695297241, | |
| "logps/chosen": -283.7400817871094, | |
| "logps/rejected": -173.89602661132812, | |
| "loss": 0.0331, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.2650320529937744, | |
| "rewards/margins": 10.87845516204834, | |
| "rewards/rejected": -10.613424301147461, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 4.241071428571429e-08, | |
| "logits/chosen": -2.034493923187256, | |
| "logits/rejected": -1.8942550420761108, | |
| "logps/chosen": -270.16497802734375, | |
| "logps/rejected": -166.28965759277344, | |
| "loss": 0.0293, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.9170511960983276, | |
| "rewards/margins": 9.314929962158203, | |
| "rewards/rejected": -10.23198127746582, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 3.92219387755102e-08, | |
| "logits/chosen": -2.102139949798584, | |
| "logits/rejected": -1.9411756992340088, | |
| "logps/chosen": -301.23846435546875, | |
| "logps/rejected": -180.20574951171875, | |
| "loss": 0.0467, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -1.1675564050674438, | |
| "rewards/margins": 9.335586547851562, | |
| "rewards/rejected": -10.503143310546875, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 3.603316326530612e-08, | |
| "logits/chosen": -2.0670650005340576, | |
| "logits/rejected": -1.8098779916763306, | |
| "logps/chosen": -294.5066833496094, | |
| "logps/rejected": -173.11416625976562, | |
| "loss": 0.0246, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.8192850351333618, | |
| "rewards/margins": 9.408272743225098, | |
| "rewards/rejected": -10.227557182312012, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 3.284438775510204e-08, | |
| "logits/chosen": -2.138176918029785, | |
| "logits/rejected": -1.978822112083435, | |
| "logps/chosen": -329.6404724121094, | |
| "logps/rejected": -190.4500274658203, | |
| "loss": 0.0184, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.5538009405136108, | |
| "rewards/margins": 10.36497688293457, | |
| "rewards/rejected": -10.918777465820312, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 2.965561224489796e-08, | |
| "logits/chosen": -2.1873650550842285, | |
| "logits/rejected": -2.0006983280181885, | |
| "logps/chosen": -301.3078308105469, | |
| "logps/rejected": -175.1551055908203, | |
| "loss": 0.0297, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -1.220243215560913, | |
| "rewards/margins": 9.117000579833984, | |
| "rewards/rejected": -10.337244033813477, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 2.6466836734693878e-08, | |
| "logits/chosen": -2.089918851852417, | |
| "logits/rejected": -1.882301926612854, | |
| "logps/chosen": -287.04632568359375, | |
| "logps/rejected": -171.83804321289062, | |
| "loss": 0.0308, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.6570901274681091, | |
| "rewards/margins": 9.935832023620605, | |
| "rewards/rejected": -10.592921257019043, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2.3278061224489795e-08, | |
| "logits/chosen": -2.073734998703003, | |
| "logits/rejected": -1.8911638259887695, | |
| "logps/chosen": -297.2498474121094, | |
| "logps/rejected": -174.6929168701172, | |
| "loss": 0.0237, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": 0.06594938039779663, | |
| "rewards/margins": 10.632802963256836, | |
| "rewards/rejected": -10.566854476928711, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2.0089285714285716e-08, | |
| "logits/chosen": -2.1300342082977295, | |
| "logits/rejected": -2.0111982822418213, | |
| "logps/chosen": -284.3525695800781, | |
| "logps/rejected": -170.9279022216797, | |
| "loss": 0.0392, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.1906980723142624, | |
| "rewards/margins": 9.709856033325195, | |
| "rewards/rejected": -9.900552749633789, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 1.6900510204081633e-08, | |
| "logits/chosen": -2.114060878753662, | |
| "logits/rejected": -1.9764654636383057, | |
| "logps/chosen": -276.199462890625, | |
| "logps/rejected": -178.11782836914062, | |
| "loss": 0.0337, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.643802285194397, | |
| "rewards/margins": 9.700589179992676, | |
| "rewards/rejected": -10.344388961791992, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 1.371173469387755e-08, | |
| "logits/chosen": -1.897796392440796, | |
| "logits/rejected": -1.800377607345581, | |
| "logps/chosen": -251.58102416992188, | |
| "logps/rejected": -162.783203125, | |
| "loss": 0.0293, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.036834489554166794, | |
| "rewards/margins": 9.945042610168457, | |
| "rewards/rejected": -9.981878280639648, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 1.0522959183673468e-08, | |
| "logits/chosen": -2.0931904315948486, | |
| "logits/rejected": -1.8743259906768799, | |
| "logps/chosen": -295.89080810546875, | |
| "logps/rejected": -169.799560546875, | |
| "loss": 0.0309, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.7167753577232361, | |
| "rewards/margins": 9.488900184631348, | |
| "rewards/rejected": -10.205676078796387, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 7.334183673469387e-09, | |
| "logits/chosen": -2.086829662322998, | |
| "logits/rejected": -1.973637342453003, | |
| "logps/chosen": -272.9532165527344, | |
| "logps/rejected": -178.72866821289062, | |
| "loss": 0.0283, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -0.3609329164028168, | |
| "rewards/margins": 9.991233825683594, | |
| "rewards/rejected": -10.352167129516602, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 4.145408163265306e-09, | |
| "logits/chosen": -2.1046552658081055, | |
| "logits/rejected": -1.9208564758300781, | |
| "logps/chosen": -248.65567016601562, | |
| "logps/rejected": -149.6916046142578, | |
| "loss": 0.0201, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.8961359262466431, | |
| "rewards/margins": 8.393698692321777, | |
| "rewards/rejected": -9.289834976196289, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 9.566326530612245e-10, | |
| "logits/chosen": -2.1648898124694824, | |
| "logits/rejected": -1.9197168350219727, | |
| "logps/chosen": -293.1455078125, | |
| "logps/rejected": -166.76507568359375, | |
| "loss": 0.0269, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -0.503768265247345, | |
| "rewards/margins": 9.358155250549316, | |
| "rewards/rejected": -9.861923217773438, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 1743, | |
| "total_flos": 0.0, | |
| "train_loss": 0.06308119606848633, | |
| "train_runtime": 23531.0319, | |
| "train_samples_per_second": 2.37, | |
| "train_steps_per_second": 0.074 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1743, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "total_flos": 0.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |