| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 1250, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 4e-08, | |
| "logits/chosen": -0.00011328980326652527, | |
| "logits/rejected": 0.11862002313137054, | |
| "logps/chosen": -264.0264892578125, | |
| "logps/rejected": -239.23065185546875, | |
| "loss": 0.3534, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "logits/chosen": 0.12777267396450043, | |
| "logits/rejected": 0.19940675795078278, | |
| "logps/chosen": -355.0330810546875, | |
| "logps/rejected": -306.326904296875, | |
| "loss": 0.3805, | |
| "rewards/accuracies": 0.375, | |
| "rewards/chosen": 0.00010627516894601285, | |
| "rewards/margins": 7.693069346714765e-05, | |
| "rewards/rejected": 2.9344506401685067e-05, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 8.000000000000001e-07, | |
| "logits/chosen": 0.11426838487386703, | |
| "logits/rejected": 0.19905796647071838, | |
| "logps/chosen": -378.491943359375, | |
| "logps/rejected": -338.3055725097656, | |
| "loss": 0.3729, | |
| "rewards/accuracies": 0.4312500059604645, | |
| "rewards/chosen": -4.1492825403111055e-05, | |
| "rewards/margins": -9.77895688265562e-05, | |
| "rewards/rejected": 5.629673978546634e-05, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "logits/chosen": 0.18764689564704895, | |
| "logits/rejected": 0.17880608141422272, | |
| "logps/chosen": -301.9078674316406, | |
| "logps/rejected": -327.083740234375, | |
| "loss": 0.3733, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -5.114002124173567e-05, | |
| "rewards/margins": 5.109022094984539e-05, | |
| "rewards/rejected": -0.00010223023127764463, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.6000000000000001e-06, | |
| "logits/chosen": 0.1114225760102272, | |
| "logits/rejected": 0.15094885230064392, | |
| "logps/chosen": -296.6694641113281, | |
| "logps/rejected": -292.0389099121094, | |
| "loss": 0.3591, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": 0.0001755535922711715, | |
| "rewards/margins": 0.000263996102148667, | |
| "rewards/rejected": -8.844250987749547e-05, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "logits/chosen": 0.1339588612318039, | |
| "logits/rejected": 0.22528474032878876, | |
| "logps/chosen": -371.5467224121094, | |
| "logps/rejected": -314.2153625488281, | |
| "loss": 0.3859, | |
| "rewards/accuracies": 0.44999998807907104, | |
| "rewards/chosen": -4.210665792925283e-05, | |
| "rewards/margins": 9.823093569139019e-05, | |
| "rewards/rejected": -0.00014033759362064302, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "logits/chosen": 0.10837416350841522, | |
| "logits/rejected": 0.22042012214660645, | |
| "logps/chosen": -360.5331115722656, | |
| "logps/rejected": -354.9367980957031, | |
| "loss": 0.3783, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": 6.062311149435118e-05, | |
| "rewards/margins": 0.0006647260743193328, | |
| "rewards/rejected": -0.0006041030283086002, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "logits/chosen": 0.14214661717414856, | |
| "logits/rejected": 0.22639691829681396, | |
| "logps/chosen": -318.57574462890625, | |
| "logps/rejected": -312.8997802734375, | |
| "loss": 0.3718, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": 0.00033316004555672407, | |
| "rewards/margins": 0.000630541646387428, | |
| "rewards/rejected": -0.0002973815717268735, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 3.2000000000000003e-06, | |
| "logits/chosen": 0.1115114837884903, | |
| "logits/rejected": 0.18560612201690674, | |
| "logps/chosen": -322.22430419921875, | |
| "logps/rejected": -316.9699401855469, | |
| "loss": 0.3641, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": 0.0006448437925428152, | |
| "rewards/margins": 0.00127916329074651, | |
| "rewards/rejected": -0.0006343194982036948, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 3.6000000000000003e-06, | |
| "logits/chosen": 0.13246920704841614, | |
| "logits/rejected": 0.21337231993675232, | |
| "logps/chosen": -371.502685546875, | |
| "logps/rejected": -330.3704528808594, | |
| "loss": 0.3617, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.0013298379490152001, | |
| "rewards/margins": 0.0018641676288098097, | |
| "rewards/rejected": -0.0005343296797946095, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 4.000000000000001e-06, | |
| "logits/chosen": 0.11018741130828857, | |
| "logits/rejected": 0.17559225857257843, | |
| "logps/chosen": -362.69635009765625, | |
| "logps/rejected": -356.4203796386719, | |
| "loss": 0.3753, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": 0.001733560347929597, | |
| "rewards/margins": 0.003086336422711611, | |
| "rewards/rejected": -0.001352776074782014, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 4.4e-06, | |
| "logits/chosen": 0.09963714331388474, | |
| "logits/rejected": 0.2167224884033203, | |
| "logps/chosen": -389.74041748046875, | |
| "logps/rejected": -345.9913024902344, | |
| "loss": 0.3635, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.0017973188078030944, | |
| "rewards/margins": 0.005557372234761715, | |
| "rewards/rejected": -0.0037600535433739424, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.800000000000001e-06, | |
| "logits/chosen": 0.1074131578207016, | |
| "logits/rejected": 0.12421264499425888, | |
| "logps/chosen": -325.73028564453125, | |
| "logps/rejected": -332.47991943359375, | |
| "loss": 0.3537, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": 0.0013967979466542602, | |
| "rewards/margins": 0.008982640691101551, | |
| "rewards/rejected": -0.007585843093693256, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.999756310023261e-06, | |
| "logits/chosen": 0.09501053392887115, | |
| "logits/rejected": 0.15297293663024902, | |
| "logps/chosen": -330.9049377441406, | |
| "logps/rejected": -349.04486083984375, | |
| "loss": 0.3681, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.00485868938267231, | |
| "rewards/margins": 0.01828749105334282, | |
| "rewards/rejected": -0.02314617857336998, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 4.997807075247147e-06, | |
| "logits/chosen": 0.10013115406036377, | |
| "logits/rejected": 0.16476190090179443, | |
| "logps/chosen": -373.80352783203125, | |
| "logps/rejected": -409.95745849609375, | |
| "loss": 0.3555, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.030246371403336525, | |
| "rewards/margins": 0.03136765584349632, | |
| "rewards/rejected": -0.0616140253841877, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 4.993910125649561e-06, | |
| "logits/chosen": 0.10287214815616608, | |
| "logits/rejected": 0.15870796144008636, | |
| "logps/chosen": -432.9341735839844, | |
| "logps/rejected": -434.81744384765625, | |
| "loss": 0.3461, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.08363470435142517, | |
| "rewards/margins": 0.02743600867688656, | |
| "rewards/rejected": -0.11107071489095688, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.988068499954578e-06, | |
| "logits/chosen": 0.12284994125366211, | |
| "logits/rejected": 0.19283421337604523, | |
| "logps/chosen": -425.1427307128906, | |
| "logps/rejected": -479.8091735839844, | |
| "loss": 0.3394, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.09083442389965057, | |
| "rewards/margins": 0.05471948906779289, | |
| "rewards/rejected": -0.14555391669273376, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.980286753286196e-06, | |
| "logits/chosen": 0.17768949270248413, | |
| "logits/rejected": 0.2632974684238434, | |
| "logps/chosen": -457.4527282714844, | |
| "logps/rejected": -496.40234375, | |
| "loss": 0.3386, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.12117190659046173, | |
| "rewards/margins": 0.05782421678304672, | |
| "rewards/rejected": -0.17899611592292786, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.970570953616383e-06, | |
| "logits/chosen": 0.14909473061561584, | |
| "logits/rejected": 0.21552911400794983, | |
| "logps/chosen": -444.69708251953125, | |
| "logps/rejected": -533.7720947265625, | |
| "loss": 0.3142, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.12033964693546295, | |
| "rewards/margins": 0.08396399021148682, | |
| "rewards/rejected": -0.20430362224578857, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 4.958928677033465e-06, | |
| "logits/chosen": 0.18431484699249268, | |
| "logits/rejected": 0.21507856249809265, | |
| "logps/chosen": -475.87908935546875, | |
| "logps/rejected": -555.5066528320312, | |
| "loss": 0.2978, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.15021201968193054, | |
| "rewards/margins": 0.08976520597934723, | |
| "rewards/rejected": -0.23997724056243896, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 4.9453690018345144e-06, | |
| "logits/chosen": 0.17984744906425476, | |
| "logits/rejected": 0.26865434646606445, | |
| "logps/chosen": -501.103515625, | |
| "logps/rejected": -583.3973999023438, | |
| "loss": 0.274, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.17393726110458374, | |
| "rewards/margins": 0.10630029439926147, | |
| "rewards/rejected": -0.2802375257015228, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 4.9299025014463665e-06, | |
| "logits/chosen": 0.23147694766521454, | |
| "logits/rejected": 0.2544376850128174, | |
| "logps/chosen": -603.2955932617188, | |
| "logps/rejected": -713.1234130859375, | |
| "loss": 0.3164, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.24584278464317322, | |
| "rewards/margins": 0.12140010297298431, | |
| "rewards/rejected": -0.36724287271499634, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.912541236180779e-06, | |
| "logits/chosen": 0.2394331395626068, | |
| "logits/rejected": 0.34389573335647583, | |
| "logps/chosen": -596.3639526367188, | |
| "logps/rejected": -683.6994018554688, | |
| "loss": 0.2964, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.24022677540779114, | |
| "rewards/margins": 0.12776102125644684, | |
| "rewards/rejected": -0.3679877817630768, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.893298743830168e-06, | |
| "logits/chosen": 0.23847298324108124, | |
| "logits/rejected": 0.2884899973869324, | |
| "logps/chosen": -577.9203491210938, | |
| "logps/rejected": -740.6541748046875, | |
| "loss": 0.2913, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.25649553537368774, | |
| "rewards/margins": 0.15363454818725586, | |
| "rewards/rejected": -0.410130113363266, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.8721900291112415e-06, | |
| "logits/chosen": 0.15130439400672913, | |
| "logits/rejected": 0.2567364275455475, | |
| "logps/chosen": -567.4896240234375, | |
| "logps/rejected": -739.4119873046875, | |
| "loss": 0.2921, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.24494703114032745, | |
| "rewards/margins": 0.161117285490036, | |
| "rewards/rejected": -0.40606433153152466, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.849231551964771e-06, | |
| "logits/chosen": 0.12885180115699768, | |
| "logits/rejected": 0.2513241767883301, | |
| "logps/chosen": -499.4623107910156, | |
| "logps/rejected": -598.4744873046875, | |
| "loss": 0.2988, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.18550391495227814, | |
| "rewards/margins": 0.11325653642416, | |
| "rewards/rejected": -0.29876044392585754, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.824441214720629e-06, | |
| "logits/chosen": 0.11510957777500153, | |
| "logits/rejected": 0.16252097487449646, | |
| "logps/chosen": -528.8275146484375, | |
| "logps/rejected": -612.6734619140625, | |
| "loss": 0.301, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.20904381573200226, | |
| "rewards/margins": 0.11143267154693604, | |
| "rewards/rejected": -0.3204765021800995, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.7978383481380865e-06, | |
| "logits/chosen": 0.10701459646224976, | |
| "logits/rejected": 0.19334085285663605, | |
| "logps/chosen": -624.1452026367188, | |
| "logps/rejected": -728.2015380859375, | |
| "loss": 0.283, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.26597902178764343, | |
| "rewards/margins": 0.1281953901052475, | |
| "rewards/rejected": -0.3941744267940521, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.769443696332272e-06, | |
| "logits/chosen": 0.10034264624118805, | |
| "logits/rejected": 0.14515987038612366, | |
| "logps/chosen": -636.72509765625, | |
| "logps/rejected": -722.1334228515625, | |
| "loss": 0.3116, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.26666730642318726, | |
| "rewards/margins": 0.10862217098474503, | |
| "rewards/rejected": -0.3752894401550293, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.7392794005985324e-06, | |
| "logits/chosen": 0.05613882467150688, | |
| "logits/rejected": 0.1429525762796402, | |
| "logps/chosen": -570.8783569335938, | |
| "logps/rejected": -689.3982543945312, | |
| "loss": 0.2858, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.24179348349571228, | |
| "rewards/margins": 0.12780442833900452, | |
| "rewards/rejected": -0.3695979416370392, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 4.707368982147318e-06, | |
| "logits/chosen": 0.04738849401473999, | |
| "logits/rejected": 0.11430943012237549, | |
| "logps/chosen": -613.541259765625, | |
| "logps/rejected": -686.9055786132812, | |
| "loss": 0.3203, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.25295981764793396, | |
| "rewards/margins": 0.10196548700332642, | |
| "rewards/rejected": -0.35492533445358276, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 4.673737323763048e-06, | |
| "logits/chosen": 0.055924586951732635, | |
| "logits/rejected": 0.1546911895275116, | |
| "logps/chosen": -591.7320556640625, | |
| "logps/rejected": -695.1236572265625, | |
| "loss": 0.294, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.22645549476146698, | |
| "rewards/margins": 0.12102316319942474, | |
| "rewards/rejected": -0.3474786877632141, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.638410650401267e-06, | |
| "logits/chosen": 0.08233563601970673, | |
| "logits/rejected": 0.13290105760097504, | |
| "logps/chosen": -571.2188110351562, | |
| "logps/rejected": -663.9716796875, | |
| "loss": 0.2906, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.22465641796588898, | |
| "rewards/margins": 0.11367009580135345, | |
| "rewards/rejected": -0.3383265435695648, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.601416508739211e-06, | |
| "logits/chosen": 0.07142064720392227, | |
| "logits/rejected": 0.11630046367645264, | |
| "logps/chosen": -576.2288208007812, | |
| "logps/rejected": -685.9638061523438, | |
| "loss": 0.2786, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.2488282173871994, | |
| "rewards/margins": 0.13437636196613312, | |
| "rewards/rejected": -0.3832045793533325, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.562783745695738e-06, | |
| "logits/chosen": 0.06200198084115982, | |
| "logits/rejected": 0.1760326325893402, | |
| "logps/chosen": -616.6163330078125, | |
| "logps/rejected": -710.7817993164062, | |
| "loss": 0.3136, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.2617984414100647, | |
| "rewards/margins": 0.11515985429286957, | |
| "rewards/rejected": -0.37695831060409546, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 4.522542485937369e-06, | |
| "logits/chosen": 0.050816863775253296, | |
| "logits/rejected": 0.1229025274515152, | |
| "logps/chosen": -629.1052856445312, | |
| "logps/rejected": -734.9453125, | |
| "loss": 0.2544, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.2729812264442444, | |
| "rewards/margins": 0.1355685442686081, | |
| "rewards/rejected": -0.40854978561401367, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.4807241083879774e-06, | |
| "logits/chosen": 0.11080716550350189, | |
| "logits/rejected": 0.19923880696296692, | |
| "logps/chosen": -633.8490600585938, | |
| "logps/rejected": -751.9637451171875, | |
| "loss": 0.2693, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.28209587931632996, | |
| "rewards/margins": 0.14658969640731812, | |
| "rewards/rejected": -0.42868557572364807, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 4.437361221760449e-06, | |
| "logits/chosen": 0.09759480506181717, | |
| "logits/rejected": 0.15628591179847717, | |
| "logps/chosen": -621.000732421875, | |
| "logps/rejected": -775.7725219726562, | |
| "loss": 0.2512, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.28890570998191833, | |
| "rewards/margins": 0.16257989406585693, | |
| "rewards/rejected": -0.4514855444431305, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 4.3924876391293915e-06, | |
| "logits/chosen": 0.11163793504238129, | |
| "logits/rejected": 0.1869780719280243, | |
| "logps/chosen": -637.1199340820312, | |
| "logps/rejected": -713.552978515625, | |
| "loss": 0.2996, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.27824416756629944, | |
| "rewards/margins": 0.1207936555147171, | |
| "rewards/rejected": -0.39903780817985535, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 4.346138351564711e-06, | |
| "logits/chosen": 0.09762606769800186, | |
| "logits/rejected": 0.12994861602783203, | |
| "logps/chosen": -524.1798095703125, | |
| "logps/rejected": -645.6965942382812, | |
| "loss": 0.2902, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.19974325597286224, | |
| "rewards/margins": 0.13353568315505981, | |
| "rewards/rejected": -0.33327895402908325, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 4.2983495008466285e-06, | |
| "logits/chosen": 0.06368868052959442, | |
| "logits/rejected": 0.14533713459968567, | |
| "logps/chosen": -527.1744384765625, | |
| "logps/rejected": -576.6195068359375, | |
| "loss": 0.3164, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.1860881745815277, | |
| "rewards/margins": 0.07743660360574722, | |
| "rewards/rejected": -0.2635248005390167, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 4.249158351283414e-06, | |
| "logits/chosen": 0.09450914710760117, | |
| "logits/rejected": 0.12860049307346344, | |
| "logps/chosen": -548.9276123046875, | |
| "logps/rejected": -644.9456787109375, | |
| "loss": 0.2733, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.198289155960083, | |
| "rewards/margins": 0.12718045711517334, | |
| "rewards/rejected": -0.32546961307525635, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 4.198603260653792e-06, | |
| "logits/chosen": 0.017919300124049187, | |
| "logits/rejected": 0.1374201476573944, | |
| "logps/chosen": -593.1436157226562, | |
| "logps/rejected": -704.8626708984375, | |
| "loss": 0.2841, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.25730082392692566, | |
| "rewards/margins": 0.13911226391792297, | |
| "rewards/rejected": -0.39641305804252625, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 4.146723650296701e-06, | |
| "logits/chosen": 0.08983961492776871, | |
| "logits/rejected": 0.16111747920513153, | |
| "logps/chosen": -659.6158447265625, | |
| "logps/rejected": -744.7501220703125, | |
| "loss": 0.3244, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.3095107674598694, | |
| "rewards/margins": 0.11083928495645523, | |
| "rewards/rejected": -0.4203500747680664, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 4.093559974371725e-06, | |
| "logits/chosen": 0.09062441438436508, | |
| "logits/rejected": 0.11453583091497421, | |
| "logps/chosen": -580.1378784179688, | |
| "logps/rejected": -689.7623901367188, | |
| "loss": 0.3067, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.2507023215293884, | |
| "rewards/margins": 0.12082631886005402, | |
| "rewards/rejected": -0.37152862548828125, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 4.039153688314146e-06, | |
| "logits/chosen": 0.08577083796262741, | |
| "logits/rejected": 0.09571922570466995, | |
| "logps/chosen": -607.8924560546875, | |
| "logps/rejected": -690.2274169921875, | |
| "loss": 0.3173, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.28040188550949097, | |
| "rewards/margins": 0.09828029572963715, | |
| "rewards/rejected": -0.37868213653564453, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 3.983547216509254e-06, | |
| "logits/chosen": 0.09381420165300369, | |
| "logits/rejected": 0.1439538300037384, | |
| "logps/chosen": -564.9346923828125, | |
| "logps/rejected": -690.1428833007812, | |
| "loss": 0.2688, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.24971938133239746, | |
| "rewards/margins": 0.1450313776731491, | |
| "rewards/rejected": -0.3947507441043854, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.92678391921108e-06, | |
| "logits/chosen": 0.13807064294815063, | |
| "logits/rejected": 0.18901431560516357, | |
| "logps/chosen": -603.8775024414062, | |
| "logps/rejected": -746.001708984375, | |
| "loss": 0.2733, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.2938609719276428, | |
| "rewards/margins": 0.1443910002708435, | |
| "rewards/rejected": -0.43825197219848633, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.868908058731376e-06, | |
| "logits/chosen": 0.11844799667596817, | |
| "logits/rejected": 0.17255009710788727, | |
| "logps/chosen": -586.2673950195312, | |
| "logps/rejected": -730.5003662109375, | |
| "loss": 0.2908, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.281721293926239, | |
| "rewards/margins": 0.1272607296705246, | |
| "rewards/rejected": -0.4089820384979248, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 3.8099647649251984e-06, | |
| "logits/chosen": 0.09214349836111069, | |
| "logits/rejected": 0.149861678481102, | |
| "logps/chosen": -608.2205810546875, | |
| "logps/rejected": -698.6951293945312, | |
| "loss": 0.2973, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.24310803413391113, | |
| "rewards/margins": 0.12697242200374603, | |
| "rewards/rejected": -0.37008044123649597, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "logits/chosen": 0.09263923764228821, | |
| "logits/rejected": 0.17302937805652618, | |
| "logps/chosen": -586.6901245117188, | |
| "logps/rejected": -707.1153564453125, | |
| "loss": 0.2892, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.2508801817893982, | |
| "rewards/margins": 0.13355056941509247, | |
| "rewards/rejected": -0.38443079590797424, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 3.689060522675689e-06, | |
| "logits/chosen": 0.10364197194576263, | |
| "logits/rejected": 0.13604341447353363, | |
| "logps/chosen": -546.9564819335938, | |
| "logps/rejected": -676.0130615234375, | |
| "loss": 0.2864, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.24754706025123596, | |
| "rewards/margins": 0.13658317923545837, | |
| "rewards/rejected": -0.38413023948669434, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.627193851723577e-06, | |
| "logits/chosen": 0.10513228178024292, | |
| "logits/rejected": 0.16634628176689148, | |
| "logps/chosen": -704.7044067382812, | |
| "logps/rejected": -804.184326171875, | |
| "loss": 0.2799, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.3062405586242676, | |
| "rewards/margins": 0.14433066546916962, | |
| "rewards/rejected": -0.4505712389945984, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.564448228912682e-06, | |
| "logits/chosen": 0.12000527232885361, | |
| "logits/rejected": 0.16995035111904144, | |
| "logps/chosen": -625.014892578125, | |
| "logps/rejected": -774.2705688476562, | |
| "loss": 0.249, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.2779640853404999, | |
| "rewards/margins": 0.17402225732803345, | |
| "rewards/rejected": -0.4519864022731781, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 3.5008725813922383e-06, | |
| "logits/chosen": 0.13057418167591095, | |
| "logits/rejected": 0.21256506443023682, | |
| "logps/chosen": -691.3427734375, | |
| "logps/rejected": -856.4500732421875, | |
| "loss": 0.2642, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.3297198712825775, | |
| "rewards/margins": 0.1787293404340744, | |
| "rewards/rejected": -0.5084491968154907, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 3.436516483539781e-06, | |
| "logits/chosen": 0.16701681911945343, | |
| "logits/rejected": 0.16362786293029785, | |
| "logps/chosen": -669.8779296875, | |
| "logps/rejected": -775.9328002929688, | |
| "loss": 0.2979, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.31730252504348755, | |
| "rewards/margins": 0.12564435601234436, | |
| "rewards/rejected": -0.4429469108581543, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 3.3714301183045382e-06, | |
| "logits/chosen": 0.10453770309686661, | |
| "logits/rejected": 0.16834238171577454, | |
| "logps/chosen": -637.10986328125, | |
| "logps/rejected": -716.6213989257812, | |
| "loss": 0.286, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.2491428554058075, | |
| "rewards/margins": 0.12585783004760742, | |
| "rewards/rejected": -0.3750006854534149, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 3.3056642380762783e-06, | |
| "logits/chosen": 0.12909114360809326, | |
| "logits/rejected": 0.17991405725479126, | |
| "logps/chosen": -595.2734375, | |
| "logps/rejected": -669.8192749023438, | |
| "loss": 0.3194, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.2555830478668213, | |
| "rewards/margins": 0.08603434264659882, | |
| "rewards/rejected": -0.3416173756122589, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 3.2392701251101172e-06, | |
| "logits/chosen": 0.11188284307718277, | |
| "logits/rejected": 0.15505322813987732, | |
| "logps/chosen": -611.50146484375, | |
| "logps/rejected": -713.7044677734375, | |
| "loss": 0.286, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.23539476096630096, | |
| "rewards/margins": 0.13384294509887695, | |
| "rewards/rejected": -0.3692377209663391, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 3.1722995515381644e-06, | |
| "logits/chosen": 0.12050177156925201, | |
| "logits/rejected": 0.1782829463481903, | |
| "logps/chosen": -593.8160400390625, | |
| "logps/rejected": -679.1887817382812, | |
| "loss": 0.2833, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.251537024974823, | |
| "rewards/margins": 0.12410496175289154, | |
| "rewards/rejected": -0.37564200162887573, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 3.1048047389991693e-06, | |
| "logits/chosen": 0.18543259799480438, | |
| "logits/rejected": 0.25393250584602356, | |
| "logps/chosen": -636.8171997070312, | |
| "logps/rejected": -729.5845947265625, | |
| "loss": 0.2827, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.262044221162796, | |
| "rewards/margins": 0.13976578414440155, | |
| "rewards/rejected": -0.4018099904060364, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 3.0368383179176584e-06, | |
| "logits/chosen": 0.06445817649364471, | |
| "logits/rejected": 0.16580384969711304, | |
| "logps/chosen": -591.1585693359375, | |
| "logps/rejected": -707.2025146484375, | |
| "loss": 0.2758, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.25943881273269653, | |
| "rewards/margins": 0.13162514567375183, | |
| "rewards/rejected": -0.39106395840644836, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.9684532864643123e-06, | |
| "logits/chosen": 0.13952665030956268, | |
| "logits/rejected": 0.22097864747047424, | |
| "logps/chosen": -617.9655151367188, | |
| "logps/rejected": -701.77099609375, | |
| "loss": 0.3062, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.27600663900375366, | |
| "rewards/margins": 0.1049640029668808, | |
| "rewards/rejected": -0.38097062706947327, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.8997029692295875e-06, | |
| "logits/chosen": 0.1246538907289505, | |
| "logits/rejected": 0.26362496614456177, | |
| "logps/chosen": -671.7306518554688, | |
| "logps/rejected": -750.6450805664062, | |
| "loss": 0.247, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.2702576816082001, | |
| "rewards/margins": 0.15437254309654236, | |
| "rewards/rejected": -0.42463022470474243, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 2.8306409756428067e-06, | |
| "logits/chosen": 0.09484218060970306, | |
| "logits/rejected": 0.14448456466197968, | |
| "logps/chosen": -620.1800537109375, | |
| "logps/rejected": -763.6259765625, | |
| "loss": 0.2758, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.2843259274959564, | |
| "rewards/margins": 0.15226764976978302, | |
| "rewards/rejected": -0.43659359216690063, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 2.761321158169134e-06, | |
| "logits/chosen": 0.1688222438097, | |
| "logits/rejected": 0.20292790234088898, | |
| "logps/chosen": -661.6067504882812, | |
| "logps/rejected": -832.15966796875, | |
| "loss": 0.2399, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.293056845664978, | |
| "rewards/margins": 0.1886308789253235, | |
| "rewards/rejected": -0.4816876947879791, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 2.6917975703170466e-06, | |
| "logits/chosen": 0.1265597641468048, | |
| "logits/rejected": 0.19089344143867493, | |
| "logps/chosen": -688.8438720703125, | |
| "logps/rejected": -787.9508056640625, | |
| "loss": 0.3184, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.33772650361061096, | |
| "rewards/margins": 0.11141663789749146, | |
| "rewards/rejected": -0.44914311170578003, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.6221244244890336e-06, | |
| "logits/chosen": 0.09308207035064697, | |
| "logits/rejected": 0.16849645972251892, | |
| "logps/chosen": -662.6124877929688, | |
| "logps/rejected": -801.0103759765625, | |
| "loss": 0.2639, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.31664609909057617, | |
| "rewards/margins": 0.15342995524406433, | |
| "rewards/rejected": -0.4700759947299957, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.5523560497083927e-06, | |
| "logits/chosen": 0.07288189977407455, | |
| "logits/rejected": 0.15222066640853882, | |
| "logps/chosen": -677.3829345703125, | |
| "logps/rejected": -765.7808837890625, | |
| "loss": 0.305, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.321521520614624, | |
| "rewards/margins": 0.12811064720153809, | |
| "rewards/rejected": -0.4496321678161621, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 2.482546849255096e-06, | |
| "logits/chosen": 0.07991471141576767, | |
| "logits/rejected": 0.1225459948182106, | |
| "logps/chosen": -584.5565185546875, | |
| "logps/rejected": -719.6222534179688, | |
| "loss": 0.2702, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.2603689134120941, | |
| "rewards/margins": 0.1404421031475067, | |
| "rewards/rejected": -0.40081101655960083, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 2.4127512582437486e-06, | |
| "logits/chosen": 0.026161545887589455, | |
| "logits/rejected": 0.1416565477848053, | |
| "logps/chosen": -630.5640869140625, | |
| "logps/rejected": -756.4042358398438, | |
| "loss": 0.2569, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.26197516918182373, | |
| "rewards/margins": 0.154434934258461, | |
| "rewards/rejected": -0.4164101183414459, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 2.3430237011767166e-06, | |
| "logits/chosen": 0.11102084815502167, | |
| "logits/rejected": 0.16481521725654602, | |
| "logps/chosen": -616.1046142578125, | |
| "logps/rejected": -770.7633666992188, | |
| "loss": 0.2615, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.2678528130054474, | |
| "rewards/margins": 0.15996505320072174, | |
| "rewards/rejected": -0.4278178811073303, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2.2734185495055503e-06, | |
| "logits/chosen": 0.07429052889347076, | |
| "logits/rejected": 0.15507885813713074, | |
| "logps/chosen": -681.5098876953125, | |
| "logps/rejected": -819.2258911132812, | |
| "loss": 0.2728, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.2998114228248596, | |
| "rewards/margins": 0.16760887205600739, | |
| "rewards/rejected": -0.4674202501773834, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2.2039900792337477e-06, | |
| "logits/chosen": 0.0644397884607315, | |
| "logits/rejected": 0.17169585824012756, | |
| "logps/chosen": -622.50634765625, | |
| "logps/rejected": -755.0499267578125, | |
| "loss": 0.2805, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.29114142060279846, | |
| "rewards/margins": 0.132626011967659, | |
| "rewards/rejected": -0.42376741766929626, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 2.134792428593971e-06, | |
| "logits/chosen": 0.06624677032232285, | |
| "logits/rejected": 0.1864895075559616, | |
| "logps/chosen": -671.3873901367188, | |
| "logps/rejected": -820.2628784179688, | |
| "loss": 0.2597, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -0.31325167417526245, | |
| "rewards/margins": 0.18044736981391907, | |
| "rewards/rejected": -0.49369901418685913, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 2.0658795558326745e-06, | |
| "logits/chosen": 0.0903557687997818, | |
| "logits/rejected": 0.16360041499137878, | |
| "logps/chosen": -650.1401977539062, | |
| "logps/rejected": -785.5604248046875, | |
| "loss": 0.2643, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.31416067481040955, | |
| "rewards/margins": 0.15609334409236908, | |
| "rewards/rejected": -0.4702540338039398, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 1.997305197135089e-06, | |
| "logits/chosen": 0.05722861737012863, | |
| "logits/rejected": 0.1671527922153473, | |
| "logps/chosen": -684.0796508789062, | |
| "logps/rejected": -822.2579345703125, | |
| "loss": 0.263, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.3124196529388428, | |
| "rewards/margins": 0.17475970089435577, | |
| "rewards/rejected": -0.48717936873435974, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 1.9291228247233607e-06, | |
| "logits/chosen": 0.07877279818058014, | |
| "logits/rejected": 0.1802339255809784, | |
| "logps/chosen": -685.1399536132812, | |
| "logps/rejected": -797.7412719726562, | |
| "loss": 0.2755, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.30405086278915405, | |
| "rewards/margins": 0.14542993903160095, | |
| "rewards/rejected": -0.4494808614253998, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 1.8613856051605242e-06, | |
| "logits/chosen": 0.07512859255075455, | |
| "logits/rejected": 0.15511508285999298, | |
| "logps/chosen": -653.9706420898438, | |
| "logps/rejected": -778.0037841796875, | |
| "loss": 0.2942, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.3025520443916321, | |
| "rewards/margins": 0.13454928994178772, | |
| "rewards/rejected": -0.4371013641357422, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 1.7941463578928088e-06, | |
| "logits/chosen": 0.04881986230611801, | |
| "logits/rejected": 0.1380804181098938, | |
| "logps/chosen": -626.0306396484375, | |
| "logps/rejected": -762.0904541015625, | |
| "loss": 0.259, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.2857394218444824, | |
| "rewards/margins": 0.15368524193763733, | |
| "rewards/rejected": -0.43942466378211975, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 1.7274575140626318e-06, | |
| "logits/chosen": 0.1028677225112915, | |
| "logits/rejected": 0.1538250893354416, | |
| "logps/chosen": -597.4656372070312, | |
| "logps/rejected": -729.7425537109375, | |
| "loss": 0.2688, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.26536160707473755, | |
| "rewards/margins": 0.16551825404167175, | |
| "rewards/rejected": -0.4308798909187317, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.661371075624363e-06, | |
| "logits/chosen": 0.08805406838655472, | |
| "logits/rejected": 0.09362033009529114, | |
| "logps/chosen": -618.3331298828125, | |
| "logps/rejected": -738.7986450195312, | |
| "loss": 0.285, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.27929025888442993, | |
| "rewards/margins": 0.15323522686958313, | |
| "rewards/rejected": -0.4325254559516907, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 1.5959385747947697e-06, | |
| "logits/chosen": 0.1377531737089157, | |
| "logits/rejected": 0.20266500115394592, | |
| "logps/chosen": -666.70458984375, | |
| "logps/rejected": -725.9954223632812, | |
| "loss": 0.3019, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.29092544317245483, | |
| "rewards/margins": 0.1138964295387268, | |
| "rewards/rejected": -0.40482187271118164, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 1.5312110338697427e-06, | |
| "logits/chosen": 0.0972646176815033, | |
| "logits/rejected": 0.19206203520298004, | |
| "logps/chosen": -619.4850463867188, | |
| "logps/rejected": -751.2019653320312, | |
| "loss": 0.2679, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.26260116696357727, | |
| "rewards/margins": 0.15384645760059357, | |
| "rewards/rejected": -0.41644757986068726, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.467238925438646e-06, | |
| "logits/chosen": 0.054350584745407104, | |
| "logits/rejected": 0.14828361570835114, | |
| "logps/chosen": -667.3047485351562, | |
| "logps/rejected": -786.0614013671875, | |
| "loss": 0.2885, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.3170751929283142, | |
| "rewards/margins": 0.13434354960918427, | |
| "rewards/rejected": -0.4514187276363373, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 1.4040721330273063e-06, | |
| "logits/chosen": 0.03969626501202583, | |
| "logits/rejected": 0.1262727826833725, | |
| "logps/chosen": -616.1150512695312, | |
| "logps/rejected": -760.9036865234375, | |
| "loss": 0.2595, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.30283650755882263, | |
| "rewards/margins": 0.1451011449098587, | |
| "rewards/rejected": -0.44793766736984253, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.3417599122003464e-06, | |
| "logits/chosen": 0.03456869348883629, | |
| "logits/rejected": 0.14985020458698273, | |
| "logps/chosen": -642.2772216796875, | |
| "logps/rejected": -767.5933837890625, | |
| "loss": 0.283, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.29978328943252563, | |
| "rewards/margins": 0.13771583139896393, | |
| "rewards/rejected": -0.43749913573265076, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.280350852153168e-06, | |
| "logits/chosen": 0.05811434984207153, | |
| "logits/rejected": 0.1481775939464569, | |
| "logps/chosen": -623.8426513671875, | |
| "logps/rejected": -765.2539672851562, | |
| "loss": 0.2507, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.2784281373023987, | |
| "rewards/margins": 0.17238983511924744, | |
| "rewards/rejected": -0.4508179724216461, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.2198928378235717e-06, | |
| "logits/chosen": 0.1006983295083046, | |
| "logits/rejected": 0.1336323320865631, | |
| "logps/chosen": -656.2332153320312, | |
| "logps/rejected": -837.9879760742188, | |
| "loss": 0.264, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.331503301858902, | |
| "rewards/margins": 0.17370417714118958, | |
| "rewards/rejected": -0.5052074193954468, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 1.160433012552508e-06, | |
| "logits/chosen": 0.008404137566685677, | |
| "logits/rejected": 0.13712240755558014, | |
| "logps/chosen": -690.4094848632812, | |
| "logps/rejected": -833.4163818359375, | |
| "loss": 0.248, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.31104812026023865, | |
| "rewards/margins": 0.1767706573009491, | |
| "rewards/rejected": -0.48781880736351013, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 1.1020177413231334e-06, | |
| "logits/chosen": 0.06463897228240967, | |
| "logits/rejected": 0.16792722046375275, | |
| "logps/chosen": -704.4422607421875, | |
| "logps/rejected": -815.9893798828125, | |
| "loss": 0.26, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.3422713875770569, | |
| "rewards/margins": 0.14494988322257996, | |
| "rewards/rejected": -0.4872213304042816, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 1.0446925746067768e-06, | |
| "logits/chosen": 0.07121414691209793, | |
| "logits/rejected": 0.10987003892660141, | |
| "logps/chosen": -679.2484130859375, | |
| "logps/rejected": -791.3287963867188, | |
| "loss": 0.2832, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.315243661403656, | |
| "rewards/margins": 0.15550537407398224, | |
| "rewards/rejected": -0.47074899077415466, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 9.88502212844063e-07, | |
| "logits/chosen": 0.09697024524211884, | |
| "logits/rejected": 0.09367278963327408, | |
| "logps/chosen": -693.0015869140625, | |
| "logps/rejected": -857.8087768554688, | |
| "loss": 0.2695, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.3296516239643097, | |
| "rewards/margins": 0.1804228574037552, | |
| "rewards/rejected": -0.5100744962692261, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 9.334904715888496e-07, | |
| "logits/chosen": 0.06400808691978455, | |
| "logits/rejected": 0.13707710802555084, | |
| "logps/chosen": -645.275146484375, | |
| "logps/rejected": -822.3630981445312, | |
| "loss": 0.2545, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.29916906356811523, | |
| "rewards/margins": 0.17705786228179932, | |
| "rewards/rejected": -0.47622689604759216, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 8.797002473421729e-07, | |
| "logits/chosen": 0.06906579434871674, | |
| "logits/rejected": 0.15590672194957733, | |
| "logps/chosen": -628.9729614257812, | |
| "logps/rejected": -791.899169921875, | |
| "loss": 0.276, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.3074602484703064, | |
| "rewards/margins": 0.15690168738365173, | |
| "rewards/rejected": -0.46436190605163574, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 8.271734841028553e-07, | |
| "logits/chosen": 0.04669260233640671, | |
| "logits/rejected": 0.13272352516651154, | |
| "logps/chosen": -666.2098999023438, | |
| "logps/rejected": -846.0867309570312, | |
| "loss": 0.2306, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.31410351395606995, | |
| "rewards/margins": 0.18806637823581696, | |
| "rewards/rejected": -0.5021698474884033, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 7.759511406608255e-07, | |
| "logits/chosen": 0.12260327488183975, | |
| "logits/rejected": 0.158196821808815, | |
| "logps/chosen": -710.1339721679688, | |
| "logps/rejected": -857.9275512695312, | |
| "loss": 0.28, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.35299620032310486, | |
| "rewards/margins": 0.1603255420923233, | |
| "rewards/rejected": -0.5133217573165894, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 7.260731586586983e-07, | |
| "logits/chosen": 0.026339268311858177, | |
| "logits/rejected": 0.06290546804666519, | |
| "logps/chosen": -655.7492065429688, | |
| "logps/rejected": -841.7448120117188, | |
| "loss": 0.2744, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.3233845829963684, | |
| "rewards/margins": 0.16933201253414154, | |
| "rewards/rejected": -0.49271655082702637, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 6.775784314464717e-07, | |
| "logits/chosen": 0.047948818653821945, | |
| "logits/rejected": 0.07574843615293503, | |
| "logps/chosen": -642.6578979492188, | |
| "logps/rejected": -803.4212646484375, | |
| "loss": 0.2468, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.3140692412853241, | |
| "rewards/margins": 0.17406299710273743, | |
| "rewards/rejected": -0.4881322979927063, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 6.305047737536707e-07, | |
| "logits/chosen": 0.044209472835063934, | |
| "logits/rejected": 0.135572150349617, | |
| "logps/chosen": -657.6187133789062, | |
| "logps/rejected": -825.8516845703125, | |
| "loss": 0.2546, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.321727991104126, | |
| "rewards/margins": 0.1708468794822693, | |
| "rewards/rejected": -0.4925749897956848, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 5.848888922025553e-07, | |
| "logits/chosen": 0.1005764827132225, | |
| "logits/rejected": 0.16469302773475647, | |
| "logps/chosen": -592.0925903320312, | |
| "logps/rejected": -742.73046875, | |
| "loss": 0.253, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.28060537576675415, | |
| "rewards/margins": 0.1545715630054474, | |
| "rewards/rejected": -0.43517693877220154, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 5.407663566854008e-07, | |
| "logits/chosen": 0.06993698328733444, | |
| "logits/rejected": 0.1429472416639328, | |
| "logps/chosen": -666.2335205078125, | |
| "logps/rejected": -834.9615478515625, | |
| "loss": 0.2547, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.3260495364665985, | |
| "rewards/margins": 0.16703951358795166, | |
| "rewards/rejected": -0.49308905005455017, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 4.981715726281666e-07, | |
| "logits/chosen": 0.0734253004193306, | |
| "logits/rejected": 0.08005217462778091, | |
| "logps/chosen": -614.7472534179688, | |
| "logps/rejected": -813.3499755859375, | |
| "loss": 0.2446, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.3050769567489624, | |
| "rewards/margins": 0.18747222423553467, | |
| "rewards/rejected": -0.49254918098449707, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 4.5713775416217884e-07, | |
| "logits/chosen": 0.042130667716264725, | |
| "logits/rejected": 0.15730439126491547, | |
| "logps/chosen": -690.7484741210938, | |
| "logps/rejected": -863.3248901367188, | |
| "loss": 0.2551, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.32108041644096375, | |
| "rewards/margins": 0.18652570247650146, | |
| "rewards/rejected": -0.5076061487197876, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 4.1769689822475147e-07, | |
| "logits/chosen": 0.019676199182868004, | |
| "logits/rejected": 0.13735586404800415, | |
| "logps/chosen": -669.737060546875, | |
| "logps/rejected": -836.93798828125, | |
| "loss": 0.2325, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.3082973062992096, | |
| "rewards/margins": 0.18536880612373352, | |
| "rewards/rejected": -0.4936660826206207, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 3.798797596089351e-07, | |
| "logits/chosen": 0.07972914725542068, | |
| "logits/rejected": 0.11014799773693085, | |
| "logps/chosen": -691.45947265625, | |
| "logps/rejected": -789.2271728515625, | |
| "loss": 0.2931, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.340253084897995, | |
| "rewards/margins": 0.12567462027072906, | |
| "rewards/rejected": -0.46592769026756287, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 3.4371582698185636e-07, | |
| "logits/chosen": 0.013631382957100868, | |
| "logits/rejected": 0.12704741954803467, | |
| "logps/chosen": -667.0838012695312, | |
| "logps/rejected": -825.3572998046875, | |
| "loss": 0.2415, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.3246777653694153, | |
| "rewards/margins": 0.18739624321460724, | |
| "rewards/rejected": -0.5120740532875061, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 3.092332998903416e-07, | |
| "logits/chosen": 0.05157908797264099, | |
| "logits/rejected": 0.17861375212669373, | |
| "logps/chosen": -696.72265625, | |
| "logps/rejected": -837.0006713867188, | |
| "loss": 0.2547, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.33154386281967163, | |
| "rewards/margins": 0.17051096260547638, | |
| "rewards/rejected": -0.5020548105239868, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 2.764590667717562e-07, | |
| "logits/chosen": 0.04824944958090782, | |
| "logits/rejected": 0.11944194883108139, | |
| "logps/chosen": -641.5201416015625, | |
| "logps/rejected": -787.1324462890625, | |
| "loss": 0.2571, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.29287251830101013, | |
| "rewards/margins": 0.1719922125339508, | |
| "rewards/rejected": -0.46486473083496094, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 2.454186839872158e-07, | |
| "logits/chosen": 0.019634583964943886, | |
| "logits/rejected": 0.12763111293315887, | |
| "logps/chosen": -705.9700317382812, | |
| "logps/rejected": -854.2298583984375, | |
| "loss": 0.2456, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.32962867617607117, | |
| "rewards/margins": 0.18715420365333557, | |
| "rewards/rejected": -0.5167828798294067, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 2.1613635589349756e-07, | |
| "logits/chosen": 0.0320691280066967, | |
| "logits/rejected": 0.10501092672348022, | |
| "logps/chosen": -677.7659912109375, | |
| "logps/rejected": -816.9409790039062, | |
| "loss": 0.2671, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -0.3169502913951874, | |
| "rewards/margins": 0.18356899917125702, | |
| "rewards/rejected": -0.500519335269928, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 1.8863491596921745e-07, | |
| "logits/chosen": 0.024208758026361465, | |
| "logits/rejected": 0.16221018135547638, | |
| "logps/chosen": -642.9352416992188, | |
| "logps/rejected": -814.4283447265625, | |
| "loss": 0.2179, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.28280672430992126, | |
| "rewards/margins": 0.20693311095237732, | |
| "rewards/rejected": -0.4897398352622986, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.629358090099639e-07, | |
| "logits/chosen": 0.022695127874612808, | |
| "logits/rejected": 0.1417960822582245, | |
| "logps/chosen": -699.5377197265625, | |
| "logps/rejected": -843.7369384765625, | |
| "loss": 0.2683, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.3325157165527344, | |
| "rewards/margins": 0.1784597933292389, | |
| "rewards/rejected": -0.5109754800796509, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.3905907440629752e-07, | |
| "logits/chosen": 0.03463418409228325, | |
| "logits/rejected": 0.11470979452133179, | |
| "logps/chosen": -696.8150634765625, | |
| "logps/rejected": -839.8903198242188, | |
| "loss": 0.2622, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.33611923456192017, | |
| "rewards/margins": 0.16319701075553894, | |
| "rewards/rejected": -0.4993163049221039, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 1.1702333051763271e-07, | |
| "logits/chosen": 0.06853802502155304, | |
| "logits/rejected": 0.11035279929637909, | |
| "logps/chosen": -630.0426635742188, | |
| "logps/rejected": -830.0460815429688, | |
| "loss": 0.2218, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.3022988438606262, | |
| "rewards/margins": 0.1969214379787445, | |
| "rewards/rejected": -0.4992203116416931, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 9.684576015420277e-08, | |
| "logits/chosen": 0.07808580249547958, | |
| "logits/rejected": 0.14167234301567078, | |
| "logps/chosen": -714.4031982421875, | |
| "logps/rejected": -866.4649658203125, | |
| "loss": 0.2511, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.3143441081047058, | |
| "rewards/margins": 0.19203931093215942, | |
| "rewards/rejected": -0.50638347864151, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 7.854209717842231e-08, | |
| "logits/chosen": 0.03158559277653694, | |
| "logits/rejected": 0.13476143777370453, | |
| "logps/chosen": -654.6509399414062, | |
| "logps/rejected": -857.1920166015625, | |
| "loss": 0.27, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.3291003406047821, | |
| "rewards/margins": 0.1921570599079132, | |
| "rewards/rejected": -0.5212574005126953, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 6.212661423609184e-08, | |
| "logits/chosen": 0.01789831928908825, | |
| "logits/rejected": 0.11147431284189224, | |
| "logps/chosen": -667.9694213867188, | |
| "logps/rejected": -807.9912719726562, | |
| "loss": 0.2908, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.3370179831981659, | |
| "rewards/margins": 0.140846848487854, | |
| "rewards/rejected": -0.4778648316860199, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 4.761211162702117e-08, | |
| "logits/chosen": 0.028112689033150673, | |
| "logits/rejected": 0.09534429758787155, | |
| "logps/chosen": -656.6475830078125, | |
| "logps/rejected": -782.1783447265625, | |
| "loss": 0.2762, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.31203871965408325, | |
| "rewards/margins": 0.1346421241760254, | |
| "rewards/rejected": -0.44668087363243103, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 3.5009907323737826e-08, | |
| "logits/chosen": 0.12954480946063995, | |
| "logits/rejected": 0.14822609722614288, | |
| "logps/chosen": -681.8800659179688, | |
| "logps/rejected": -832.7755737304688, | |
| "loss": 0.2708, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.33531031012535095, | |
| "rewards/margins": 0.16368630528450012, | |
| "rewards/rejected": -0.4989965856075287, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2.4329828146074096e-08, | |
| "logits/chosen": 0.017124827951192856, | |
| "logits/rejected": 0.1074015349149704, | |
| "logps/chosen": -683.5274658203125, | |
| "logps/rejected": -843.3143310546875, | |
| "loss": 0.2659, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -0.31970110535621643, | |
| "rewards/margins": 0.19150231778621674, | |
| "rewards/rejected": -0.511203408241272, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 1.5580202098509078e-08, | |
| "logits/chosen": 0.041600506752729416, | |
| "logits/rejected": 0.14631414413452148, | |
| "logps/chosen": -659.5726318359375, | |
| "logps/rejected": -816.1785888671875, | |
| "loss": 0.2445, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.30856478214263916, | |
| "rewards/margins": 0.18453359603881836, | |
| "rewards/rejected": -0.49309834837913513, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 8.767851876239075e-09, | |
| "logits/chosen": 0.04737471044063568, | |
| "logits/rejected": 0.10654322057962418, | |
| "logps/chosen": -632.4381103515625, | |
| "logps/rejected": -729.2340087890625, | |
| "loss": 0.2933, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.3016565442085266, | |
| "rewards/margins": 0.13306598365306854, | |
| "rewards/rejected": -0.43472257256507874, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 3.8980895450474455e-09, | |
| "logits/chosen": 0.04944654554128647, | |
| "logits/rejected": 0.12228293716907501, | |
| "logps/chosen": -684.4515991210938, | |
| "logps/rejected": -818.5169067382812, | |
| "loss": 0.2597, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.32826223969459534, | |
| "rewards/margins": 0.16636613011360168, | |
| "rewards/rejected": -0.494628369808197, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 9.747123991141193e-10, | |
| "logits/chosen": 0.07023318111896515, | |
| "logits/rejected": 0.13329362869262695, | |
| "logps/chosen": -622.5777587890625, | |
| "logps/rejected": -774.9669799804688, | |
| "loss": 0.2367, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.2667763829231262, | |
| "rewards/margins": 0.18331165611743927, | |
| "rewards/rejected": -0.4500879645347595, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 0.0, | |
| "logits/chosen": 0.07397739589214325, | |
| "logits/rejected": 0.143302321434021, | |
| "logps/chosen": -675.6883544921875, | |
| "logps/rejected": -813.1781005859375, | |
| "loss": 0.2658, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.31572216749191284, | |
| "rewards/margins": 0.1733875572681427, | |
| "rewards/rejected": -0.48910975456237793, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 1250, | |
| "total_flos": 0.0, | |
| "train_loss": 0.2873324268341064, | |
| "train_runtime": 11591.8245, | |
| "train_samples_per_second": 1.725, | |
| "train_steps_per_second": 0.108 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1250, | |
| "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 | |
| } | |