| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.6047904191616766, | |
| "eval_steps": 500, | |
| "global_step": 303, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.001996007984031936, | |
| "grad_norm": 3.4413226850733536, | |
| "learning_rate": 5.88235294117647e-09, | |
| "logits/chosen": -0.2126132994890213, | |
| "logits/rejected": -0.06049485132098198, | |
| "logps/chosen": -312.6929931640625, | |
| "logps/rejected": -303.399658203125, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00998003992015968, | |
| "grad_norm": 3.5626484310109503, | |
| "learning_rate": 2.941176470588235e-08, | |
| "logits/chosen": -0.15043149888515472, | |
| "logits/rejected": -0.05884423851966858, | |
| "logps/chosen": -271.6809387207031, | |
| "logps/rejected": -308.14019775390625, | |
| "loss": 0.6932, | |
| "rewards/accuracies": 0.34375, | |
| "rewards/chosen": 0.00032658452983014286, | |
| "rewards/margins": -0.00044720349251292646, | |
| "rewards/rejected": 0.0007737880805507302, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.01996007984031936, | |
| "grad_norm": 3.444766226961807, | |
| "learning_rate": 5.88235294117647e-08, | |
| "logits/chosen": -0.04885115846991539, | |
| "logits/rejected": 0.02876092866063118, | |
| "logps/chosen": -314.44207763671875, | |
| "logps/rejected": -324.5089111328125, | |
| "loss": 0.6935, | |
| "rewards/accuracies": 0.4124999940395355, | |
| "rewards/chosen": -0.0015638660406693816, | |
| "rewards/margins": -0.0018882350996136665, | |
| "rewards/rejected": 0.0003243688843213022, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.029940119760479042, | |
| "grad_norm": 3.351776759283806, | |
| "learning_rate": 8.823529411764706e-08, | |
| "logits/chosen": -0.14645501971244812, | |
| "logits/rejected": -0.10253484547138214, | |
| "logps/chosen": -325.3197021484375, | |
| "logps/rejected": -335.1691589355469, | |
| "loss": 0.6937, | |
| "rewards/accuracies": 0.3375000059604645, | |
| "rewards/chosen": -0.0009325200808234513, | |
| "rewards/margins": -0.0027469159103929996, | |
| "rewards/rejected": 0.0018143958877772093, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.03992015968063872, | |
| "grad_norm": 3.2486760323756765, | |
| "learning_rate": 1.176470588235294e-07, | |
| "logits/chosen": -0.17513605952262878, | |
| "logits/rejected": -0.14125120639801025, | |
| "logps/chosen": -236.529052734375, | |
| "logps/rejected": -224.99609375, | |
| "loss": 0.6932, | |
| "rewards/accuracies": 0.44999998807907104, | |
| "rewards/chosen": -0.00048039742978289723, | |
| "rewards/margins": -0.0017443184042349458, | |
| "rewards/rejected": 0.0012639210326597095, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0499001996007984, | |
| "grad_norm": 3.682895506825886, | |
| "learning_rate": 1.4705882352941175e-07, | |
| "logits/chosen": 0.02582230605185032, | |
| "logits/rejected": -0.026322145015001297, | |
| "logps/chosen": -276.4084777832031, | |
| "logps/rejected": -271.5510559082031, | |
| "loss": 0.6928, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.00010870690312003717, | |
| "rewards/margins": -2.2810349037172273e-05, | |
| "rewards/rejected": -8.589646313339472e-05, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.059880239520958084, | |
| "grad_norm": 3.521568498969969, | |
| "learning_rate": 1.764705882352941e-07, | |
| "logits/chosen": -0.029633093625307083, | |
| "logits/rejected": 0.017462745308876038, | |
| "logps/chosen": -241.04345703125, | |
| "logps/rejected": -254.0708465576172, | |
| "loss": 0.6929, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.001156697398982942, | |
| "rewards/margins": 0.0012088593794032931, | |
| "rewards/rejected": -5.2162260544719175e-05, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.06986027944111776, | |
| "grad_norm": 3.4756100770970852, | |
| "learning_rate": 2.0588235294117647e-07, | |
| "logits/chosen": -0.09809896349906921, | |
| "logits/rejected": -0.12783308327198029, | |
| "logps/chosen": -316.04541015625, | |
| "logps/rejected": -303.2275085449219, | |
| "loss": 0.6926, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.0019062187056988478, | |
| "rewards/margins": 0.0034712664783000946, | |
| "rewards/rejected": -0.0015650471905246377, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.07984031936127745, | |
| "grad_norm": 3.5251485554920383, | |
| "learning_rate": 2.352941176470588e-07, | |
| "logits/chosen": -0.10866469144821167, | |
| "logits/rejected": -0.08842261880636215, | |
| "logps/chosen": -300.94866943359375, | |
| "logps/rejected": -301.57635498046875, | |
| "loss": 0.6927, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": 0.0005061564734205604, | |
| "rewards/margins": 0.0017319575417786837, | |
| "rewards/rejected": -0.0012258009519428015, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.08982035928143713, | |
| "grad_norm": 3.562421126179782, | |
| "learning_rate": 2.6470588235294114e-07, | |
| "logits/chosen": -0.04412439838051796, | |
| "logits/rejected": 0.046757772564888, | |
| "logps/chosen": -276.1053771972656, | |
| "logps/rejected": -299.3190612792969, | |
| "loss": 0.692, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.002857887651771307, | |
| "rewards/margins": 0.0004688422195613384, | |
| "rewards/rejected": -0.003326730104163289, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.0998003992015968, | |
| "grad_norm": 3.174201376940309, | |
| "learning_rate": 2.941176470588235e-07, | |
| "logits/chosen": -0.08824320137500763, | |
| "logits/rejected": -0.043245576322078705, | |
| "logps/chosen": -271.13604736328125, | |
| "logps/rejected": -272.74102783203125, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.0016361953457817435, | |
| "rewards/margins": 0.005965009797364473, | |
| "rewards/rejected": -0.007601206190884113, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.10978043912175649, | |
| "grad_norm": 3.3426951285606368, | |
| "learning_rate": 2.999415172560531e-07, | |
| "logits/chosen": 0.032454706728458405, | |
| "logits/rejected": 0.15314094722270966, | |
| "logps/chosen": -301.4625244140625, | |
| "logps/rejected": -304.8877258300781, | |
| "loss": 0.691, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.007659287191927433, | |
| "rewards/margins": 0.0023522169794887304, | |
| "rewards/rejected": -0.010011503472924232, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.11976047904191617, | |
| "grad_norm": 3.2266501701625283, | |
| "learning_rate": 2.997040092642407e-07, | |
| "logits/chosen": -0.13825412094593048, | |
| "logits/rejected": -0.06802596896886826, | |
| "logps/chosen": -273.85198974609375, | |
| "logps/rejected": -256.1399841308594, | |
| "loss": 0.6905, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.0074735768139362335, | |
| "rewards/margins": 0.005476811900734901, | |
| "rewards/rejected": -0.012950388714671135, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.12974051896207583, | |
| "grad_norm": 3.129826931598691, | |
| "learning_rate": 2.9928410999727467e-07, | |
| "logits/chosen": -0.03806266561150551, | |
| "logits/rejected": -0.06114925071597099, | |
| "logps/chosen": -273.815673828125, | |
| "logps/rejected": -257.2196350097656, | |
| "loss": 0.6897, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.010186719708144665, | |
| "rewards/margins": 0.00584446731954813, | |
| "rewards/rejected": -0.016031187027692795, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.13972055888223553, | |
| "grad_norm": 3.2696148017727764, | |
| "learning_rate": 2.9868233103773125e-07, | |
| "logits/chosen": -0.19126296043395996, | |
| "logits/rejected": -0.1384088099002838, | |
| "logps/chosen": -289.3074645996094, | |
| "logps/rejected": -304.941650390625, | |
| "loss": 0.6901, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.012994937598705292, | |
| "rewards/margins": 0.006138815078884363, | |
| "rewards/rejected": -0.019133752211928368, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.1497005988023952, | |
| "grad_norm": 3.158441752144902, | |
| "learning_rate": 2.978994055605757e-07, | |
| "logits/chosen": -0.3353852927684784, | |
| "logits/rejected": -0.20957405865192413, | |
| "logps/chosen": -282.4247131347656, | |
| "logps/rejected": -297.1376647949219, | |
| "loss": 0.6887, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.013628537766635418, | |
| "rewards/margins": 0.012240052223205566, | |
| "rewards/rejected": -0.02586858905851841, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.1596806387225549, | |
| "grad_norm": 3.2922551208958932, | |
| "learning_rate": 2.9693628743990157e-07, | |
| "logits/chosen": -0.2210439145565033, | |
| "logits/rejected": -0.13906380534172058, | |
| "logps/chosen": -286.2383728027344, | |
| "logps/rejected": -284.0565490722656, | |
| "loss": 0.6885, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -0.012788718566298485, | |
| "rewards/margins": 0.01518750749528408, | |
| "rewards/rejected": -0.027976226061582565, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.16966067864271456, | |
| "grad_norm": 3.143746580193799, | |
| "learning_rate": 2.9579415008678196e-07, | |
| "logits/chosen": 0.061027251183986664, | |
| "logits/rejected": 0.020937766879796982, | |
| "logps/chosen": -261.85797119140625, | |
| "logps/rejected": -245.4117889404297, | |
| "loss": 0.6879, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.020817553624510765, | |
| "rewards/margins": 0.01146668754518032, | |
| "rewards/rejected": -0.032284241169691086, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.17964071856287425, | |
| "grad_norm": 3.1363791165769634, | |
| "learning_rate": 2.944743850196487e-07, | |
| "logits/chosen": -0.18295350670814514, | |
| "logits/rejected": -0.138901025056839, | |
| "logps/chosen": -302.30377197265625, | |
| "logps/rejected": -297.4572448730469, | |
| "loss": 0.6867, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.022789286449551582, | |
| "rewards/margins": 0.014456982724368572, | |
| "rewards/rejected": -0.03724626824259758, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.18962075848303392, | |
| "grad_norm": 3.375624228841244, | |
| "learning_rate": 2.92978600168942e-07, | |
| "logits/chosen": -0.21953599154949188, | |
| "logits/rejected": -0.19249823689460754, | |
| "logps/chosen": -282.0645446777344, | |
| "logps/rejected": -293.2350158691406, | |
| "loss": 0.6865, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -0.028345201164484024, | |
| "rewards/margins": 0.015181821770966053, | |
| "rewards/rejected": -0.0435270220041275, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.1996007984031936, | |
| "grad_norm": 3.4004680540887673, | |
| "learning_rate": 2.913086179180945e-07, | |
| "logits/chosen": -0.26934295892715454, | |
| "logits/rejected": -0.24918439984321594, | |
| "logps/chosen": -267.2822570800781, | |
| "logps/rejected": -273.8163757324219, | |
| "loss": 0.6847, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.02366602048277855, | |
| "rewards/margins": 0.018390221521258354, | |
| "rewards/rejected": -0.042056240141391754, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.20958083832335328, | |
| "grad_norm": 3.348944145808141, | |
| "learning_rate": 2.894664728832377e-07, | |
| "logits/chosen": -0.12691783905029297, | |
| "logits/rejected": -0.10880570113658905, | |
| "logps/chosen": -306.7061462402344, | |
| "logps/rejected": -296.7961120605469, | |
| "loss": 0.684, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -0.02831300161778927, | |
| "rewards/margins": 0.024948222562670708, | |
| "rewards/rejected": -0.053261227905750275, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.21956087824351297, | |
| "grad_norm": 3.3068624575557677, | |
| "learning_rate": 2.8745440943433595e-07, | |
| "logits/chosen": -0.39724451303482056, | |
| "logits/rejected": -0.3233721852302551, | |
| "logps/chosen": -299.71356201171875, | |
| "logps/rejected": -307.81011962890625, | |
| "loss": 0.6826, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.037553686648607254, | |
| "rewards/margins": 0.01620485447347164, | |
| "rewards/rejected": -0.053758542984724045, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.22954091816367264, | |
| "grad_norm": 3.5399975333931057, | |
| "learning_rate": 2.8527487896076705e-07, | |
| "logits/chosen": -0.2971861958503723, | |
| "logits/rejected": -0.26923665404319763, | |
| "logps/chosen": -323.73760986328125, | |
| "logps/rejected": -317.88238525390625, | |
| "loss": 0.6812, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -0.03361167758703232, | |
| "rewards/margins": 0.025083836168050766, | |
| "rewards/rejected": -0.058695513755083084, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.23952095808383234, | |
| "grad_norm": 3.387251098081717, | |
| "learning_rate": 2.829305368846822e-07, | |
| "logits/chosen": -0.08930722624063492, | |
| "logits/rejected": -0.093043752014637, | |
| "logps/chosen": -298.3551330566406, | |
| "logps/rejected": -301.565185546875, | |
| "loss": 0.6807, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.04334830120205879, | |
| "rewards/margins": 0.02416048012673855, | |
| "rewards/rejected": -0.06750877946615219, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.249500998003992, | |
| "grad_norm": 3.458206539528906, | |
| "learning_rate": 2.8042423942578286e-07, | |
| "logits/chosen": -0.06799415498971939, | |
| "logits/rejected": 0.05015769600868225, | |
| "logps/chosen": -252.72793579101562, | |
| "logps/rejected": -279.40631103515625, | |
| "loss": 0.6788, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": -0.0386417880654335, | |
| "rewards/margins": 0.03730592131614685, | |
| "rewards/rejected": -0.07594770938158035, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.25948103792415167, | |
| "grad_norm": 3.4555138461844757, | |
| "learning_rate": 2.7775904012145714e-07, | |
| "logits/chosen": -0.24486279487609863, | |
| "logits/rejected": -0.1943987011909485, | |
| "logps/chosen": -326.3341369628906, | |
| "logps/rejected": -326.3017272949219, | |
| "loss": 0.6792, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.04846924915909767, | |
| "rewards/margins": 0.03331124037504196, | |
| "rewards/rejected": -0.08178049325942993, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2694610778443114, | |
| "grad_norm": 3.4103149118409104, | |
| "learning_rate": 2.749381861065149e-07, | |
| "logits/chosen": -0.11615544557571411, | |
| "logits/rejected": -0.02477203868329525, | |
| "logps/chosen": -281.5689697265625, | |
| "logps/rejected": -289.6168518066406, | |
| "loss": 0.6756, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.04677049443125725, | |
| "rewards/margins": 0.03843165561556816, | |
| "rewards/rejected": -0.085202157497406, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.27944111776447106, | |
| "grad_norm": 3.3294732309430946, | |
| "learning_rate": 2.719651141570541e-07, | |
| "logits/chosen": -0.07356293499469757, | |
| "logits/rejected": 0.01894739270210266, | |
| "logps/chosen": -295.6517333984375, | |
| "logps/rejected": -294.22064208984375, | |
| "loss": 0.6733, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -0.04726024717092514, | |
| "rewards/margins": 0.05050128698348999, | |
| "rewards/rejected": -0.09776153415441513, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.2894211576846307, | |
| "grad_norm": 3.335787112423438, | |
| "learning_rate": 2.6884344650327857e-07, | |
| "logits/chosen": -0.3144836127758026, | |
| "logits/rejected": -0.2471332997083664, | |
| "logps/chosen": -275.6163635253906, | |
| "logps/rejected": -283.7267150878906, | |
| "loss": 0.6741, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.06026310846209526, | |
| "rewards/margins": 0.03206663206219673, | |
| "rewards/rejected": -0.0923297330737114, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.2994011976047904, | |
| "grad_norm": 3.351780879830391, | |
| "learning_rate": 2.6557698641636836e-07, | |
| "logits/chosen": -0.09479612112045288, | |
| "logits/rejected": -0.05155174061655998, | |
| "logps/chosen": -273.46527099609375, | |
| "logps/rejected": -285.8160095214844, | |
| "loss": 0.6719, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.057442475110292435, | |
| "rewards/margins": 0.04023780673742294, | |
| "rewards/rejected": -0.09768027812242508, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3093812375249501, | |
| "grad_norm": 3.3953812787691744, | |
| "learning_rate": 2.6216971357477977e-07, | |
| "logits/chosen": -0.20542898774147034, | |
| "logits/rejected": -0.16760793328285217, | |
| "logps/chosen": -319.43438720703125, | |
| "logps/rejected": -337.89678955078125, | |
| "loss": 0.6677, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.0877736285328865, | |
| "rewards/margins": 0.05791449546813965, | |
| "rewards/rejected": -0.14568811655044556, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.3193612774451098, | |
| "grad_norm": 3.7295101426966957, | |
| "learning_rate": 2.5862577921562015e-07, | |
| "logits/chosen": -0.26064497232437134, | |
| "logits/rejected": -0.1645033359527588, | |
| "logps/chosen": -292.02484130859375, | |
| "logps/rejected": -300.4161376953125, | |
| "loss": 0.665, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.08468987792730331, | |
| "rewards/margins": 0.062365107238292694, | |
| "rewards/rejected": -0.1470550000667572, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.32934131736526945, | |
| "grad_norm": 3.5278102635310775, | |
| "learning_rate": 2.549495010770048e-07, | |
| "logits/chosen": -0.2758394777774811, | |
| "logits/rejected": -0.20927894115447998, | |
| "logps/chosen": -288.8243713378906, | |
| "logps/rejected": -291.8514404296875, | |
| "loss": 0.6655, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.10788210481405258, | |
| "rewards/margins": 0.06151789426803589, | |
| "rewards/rejected": -0.16940000653266907, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3393213572854291, | |
| "grad_norm": 3.512512373193113, | |
| "learning_rate": 2.511453581375585e-07, | |
| "logits/chosen": -0.39803841710090637, | |
| "logits/rejected": -0.3213597238063812, | |
| "logps/chosen": -258.41326904296875, | |
| "logps/rejected": -263.5000915527344, | |
| "loss": 0.6625, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.10974812507629395, | |
| "rewards/margins": 0.04202640801668167, | |
| "rewards/rejected": -0.15177452564239502, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.34930139720558884, | |
| "grad_norm": 3.7824536020137383, | |
| "learning_rate": 2.4721798515946963e-07, | |
| "logits/chosen": -0.2360338717699051, | |
| "logits/rejected": -0.17899933457374573, | |
| "logps/chosen": -274.8690185546875, | |
| "logps/rejected": -291.98419189453125, | |
| "loss": 0.6587, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -0.1216902956366539, | |
| "rewards/margins": 0.0732160434126854, | |
| "rewards/rejected": -0.1949063539505005, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3592814371257485, | |
| "grad_norm": 3.6570477679381033, | |
| "learning_rate": 2.4317216704174655e-07, | |
| "logits/chosen": -0.2678098976612091, | |
| "logits/rejected": -0.22654588520526886, | |
| "logps/chosen": -285.76885986328125, | |
| "logps/rejected": -300.3975524902344, | |
| "loss": 0.6575, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.13574646413326263, | |
| "rewards/margins": 0.06750482320785522, | |
| "rewards/rejected": -0.20325128734111786, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.36926147704590817, | |
| "grad_norm": 3.705774667801494, | |
| "learning_rate": 2.3901283299055524e-07, | |
| "logits/chosen": -0.4732998013496399, | |
| "logits/rejected": -0.39881661534309387, | |
| "logps/chosen": -321.14984130859375, | |
| "logps/rejected": -326.15032958984375, | |
| "loss": 0.6514, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -0.17002463340759277, | |
| "rewards/margins": 0.08681143075227737, | |
| "rewards/rejected": -0.25683605670928955, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.37924151696606784, | |
| "grad_norm": 3.4125215861302176, | |
| "learning_rate": 2.3474505051374067e-07, | |
| "logits/chosen": -0.34785977005958557, | |
| "logits/rejected": -0.24581901729106903, | |
| "logps/chosen": -323.81732177734375, | |
| "logps/rejected": -328.8360900878906, | |
| "loss": 0.6496, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.20683076977729797, | |
| "rewards/margins": 0.07259462773799896, | |
| "rewards/rejected": -0.2794254422187805, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.38922155688622756, | |
| "grad_norm": 3.6035947439545715, | |
| "learning_rate": 2.3037401924684947e-07, | |
| "logits/chosen": -0.2815292775630951, | |
| "logits/rejected": -0.2620851397514343, | |
| "logps/chosen": -319.5675048828125, | |
| "logps/rejected": -332.9228210449219, | |
| "loss": 0.6501, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.2412305772304535, | |
| "rewards/margins": 0.09057402610778809, | |
| "rewards/rejected": -0.3318046033382416, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.3992015968063872, | |
| "grad_norm": 3.6822468215438624, | |
| "learning_rate": 2.2590506461817452e-07, | |
| "logits/chosen": -0.259944349527359, | |
| "logits/rejected": -0.20725861191749573, | |
| "logps/chosen": -298.30303955078125, | |
| "logps/rejected": -305.5474548339844, | |
| "loss": 0.6419, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -0.24381069839000702, | |
| "rewards/margins": 0.11996382474899292, | |
| "rewards/rejected": -0.36377447843551636, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4091816367265469, | |
| "grad_norm": 3.7132773714988927, | |
| "learning_rate": 2.213436313605413e-07, | |
| "logits/chosen": -0.5560285449028015, | |
| "logits/rejected": -0.49163347482681274, | |
| "logps/chosen": -387.44171142578125, | |
| "logps/rejected": -402.449462890625, | |
| "loss": 0.6434, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.3169836699962616, | |
| "rewards/margins": 0.07618891447782516, | |
| "rewards/rejected": -0.39317259192466736, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.41916167664670656, | |
| "grad_norm": 3.775451248470663, | |
| "learning_rate": 2.166952768777391e-07, | |
| "logits/chosen": -0.2941240966320038, | |
| "logits/rejected": -0.3131045997142792, | |
| "logps/chosen": -303.090087890625, | |
| "logps/rejected": -298.85699462890625, | |
| "loss": 0.6404, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.2724608778953552, | |
| "rewards/margins": 0.1073354110121727, | |
| "rewards/rejected": -0.3797963261604309, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.4291417165668663, | |
| "grad_norm": 3.7925188623151422, | |
| "learning_rate": 2.1196566447368126e-07, | |
| "logits/chosen": -0.26538053154945374, | |
| "logits/rejected": -0.25145062804222107, | |
| "logps/chosen": -334.455078125, | |
| "logps/rejected": -356.2786865234375, | |
| "loss": 0.6355, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.34816819429397583, | |
| "rewards/margins": 0.12635044753551483, | |
| "rewards/rejected": -0.4745185971260071, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.43912175648702595, | |
| "grad_norm": 3.8981243362393565, | |
| "learning_rate": 2.0716055645254113e-07, | |
| "logits/chosen": -0.46777939796447754, | |
| "logits/rejected": -0.4300519824028015, | |
| "logps/chosen": -325.1937255859375, | |
| "logps/rejected": -333.54022216796875, | |
| "loss": 0.641, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.4105430245399475, | |
| "rewards/margins": 0.10119415819644928, | |
| "rewards/rejected": -0.5117372274398804, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4491017964071856, | |
| "grad_norm": 3.7392987472200354, | |
| "learning_rate": 2.0228580709827226e-07, | |
| "logits/chosen": -0.5368104577064514, | |
| "logits/rejected": -0.45491117238998413, | |
| "logps/chosen": -349.7928771972656, | |
| "logps/rejected": -371.95050048828125, | |
| "loss": 0.6326, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -0.3743346631526947, | |
| "rewards/margins": 0.1426324099302292, | |
| "rewards/rejected": -0.5169671177864075, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.4590818363273453, | |
| "grad_norm": 3.6974655152807174, | |
| "learning_rate": 1.9734735554206537e-07, | |
| "logits/chosen": -0.31408676505088806, | |
| "logits/rejected": -0.2642524242401123, | |
| "logps/chosen": -349.4893798828125, | |
| "logps/rejected": -369.08990478515625, | |
| "loss": 0.6238, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.47439342737197876, | |
| "rewards/margins": 0.15798419713974, | |
| "rewards/rejected": -0.632377564907074, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.469061876247505, | |
| "grad_norm": 3.9074441662298662, | |
| "learning_rate": 1.9235121852643148e-07, | |
| "logits/chosen": -0.4889460504055023, | |
| "logits/rejected": -0.48752936720848083, | |
| "logps/chosen": -331.08624267578125, | |
| "logps/rejected": -348.8202819824219, | |
| "loss": 0.6259, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -0.41649022698402405, | |
| "rewards/margins": 0.1930549144744873, | |
| "rewards/rejected": -0.609545111656189, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.47904191616766467, | |
| "grad_norm": 3.7710978575315757, | |
| "learning_rate": 1.8730348307472826e-07, | |
| "logits/chosen": -0.4119279384613037, | |
| "logits/rejected": -0.444758802652359, | |
| "logps/chosen": -348.2624206542969, | |
| "logps/rejected": -369.9598083496094, | |
| "loss": 0.6249, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.5225319862365723, | |
| "rewards/margins": 0.13163016736507416, | |
| "rewards/rejected": -0.6541622281074524, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.48902195608782434, | |
| "grad_norm": 4.08348789093204, | |
| "learning_rate": 1.822102990750595e-07, | |
| "logits/chosen": -0.5410420298576355, | |
| "logits/rejected": -0.4581461548805237, | |
| "logps/chosen": -300.3963317871094, | |
| "logps/rejected": -326.97357177734375, | |
| "loss": 0.6205, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -0.4268406927585602, | |
| "rewards/margins": 0.17870576679706573, | |
| "rewards/rejected": -0.6055464744567871, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.499001996007984, | |
| "grad_norm": 3.8249633059365977, | |
| "learning_rate": 1.7707787178758402e-07, | |
| "logits/chosen": -0.5376532673835754, | |
| "logits/rejected": -0.5166374444961548, | |
| "logps/chosen": -334.3974609375, | |
| "logps/rejected": -363.59832763671875, | |
| "loss": 0.6097, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -0.5467029809951782, | |
| "rewards/margins": 0.219395712018013, | |
| "rewards/rejected": -0.7660986185073853, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5089820359281437, | |
| "grad_norm": 4.084608415013939, | |
| "learning_rate": 1.7191245428436176e-07, | |
| "logits/chosen": -0.6120297908782959, | |
| "logits/rejected": -0.5665668249130249, | |
| "logps/chosen": -388.49237060546875, | |
| "logps/rejected": -404.17913818359375, | |
| "loss": 0.6066, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -0.5541552305221558, | |
| "rewards/margins": 0.18654166162014008, | |
| "rewards/rejected": -0.7406969666481018, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5189620758483033, | |
| "grad_norm": 4.127013078261958, | |
| "learning_rate": 1.667203398309488e-07, | |
| "logits/chosen": -0.602649450302124, | |
| "logits/rejected": -0.47091540694236755, | |
| "logps/chosen": -342.5652770996094, | |
| "logps/rejected": -363.1390380859375, | |
| "loss": 0.6099, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.5999665260314941, | |
| "rewards/margins": 0.22458717226982117, | |
| "rewards/rejected": -0.8245537877082825, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5289421157684631, | |
| "grad_norm": 4.181662944251744, | |
| "learning_rate": 1.6150785421902278e-07, | |
| "logits/chosen": -0.49136123061180115, | |
| "logits/rejected": -0.4355231821537018, | |
| "logps/chosen": -355.0814514160156, | |
| "logps/rejected": -400.48486328125, | |
| "loss": 0.6027, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.6992485523223877, | |
| "rewards/margins": 0.2647705674171448, | |
| "rewards/rejected": -0.9640190005302429, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5389221556886228, | |
| "grad_norm": 4.187328944312369, | |
| "learning_rate": 1.5628134805937993e-07, | |
| "logits/chosen": -0.6901322603225708, | |
| "logits/rejected": -0.7155795693397522, | |
| "logps/chosen": -338.4058532714844, | |
| "logps/rejected": -357.05145263671875, | |
| "loss": 0.6044, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -0.6499531865119934, | |
| "rewards/margins": 0.2078942060470581, | |
| "rewards/rejected": -0.8578473329544067, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5489021956087824, | |
| "grad_norm": 4.480820141822546, | |
| "learning_rate": 1.5104718904469424e-07, | |
| "logits/chosen": -0.49210062623023987, | |
| "logits/rejected": -0.4995419979095459, | |
| "logps/chosen": -362.78936767578125, | |
| "logps/rejected": -376.6024475097656, | |
| "loss": 0.6022, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.684400200843811, | |
| "rewards/margins": 0.22897310554981232, | |
| "rewards/rejected": -0.9133733510971069, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5588822355289421, | |
| "grad_norm": 4.335560690928003, | |
| "learning_rate": 1.458117541914647e-07, | |
| "logits/chosen": -0.6205674409866333, | |
| "logits/rejected": -0.5415416955947876, | |
| "logps/chosen": -363.15484619140625, | |
| "logps/rejected": -380.423828125, | |
| "loss": 0.601, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.7319628000259399, | |
| "rewards/margins": 0.22972945868968964, | |
| "rewards/rejected": -0.9616923332214355, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5688622754491018, | |
| "grad_norm": 4.236266469806635, | |
| "learning_rate": 1.4058142207060298e-07, | |
| "logits/chosen": -0.5227392315864563, | |
| "logits/rejected": -0.44109734892845154, | |
| "logps/chosen": -350.81268310546875, | |
| "logps/rejected": -387.339111328125, | |
| "loss": 0.599, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -0.7962437272071838, | |
| "rewards/margins": 0.2677839994430542, | |
| "rewards/rejected": -1.0640277862548828, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.5788423153692615, | |
| "grad_norm": 4.1859368563558546, | |
| "learning_rate": 1.353625650361276e-07, | |
| "logits/chosen": -0.4567599296569824, | |
| "logits/rejected": -0.37081751227378845, | |
| "logps/chosen": -292.90582275390625, | |
| "logps/rejected": -351.08428955078125, | |
| "loss": 0.5976, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -0.7117597460746765, | |
| "rewards/margins": 0.23347418010234833, | |
| "rewards/rejected": -0.9452340006828308, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.5888223552894212, | |
| "grad_norm": 4.963202872238042, | |
| "learning_rate": 1.301615414614316e-07, | |
| "logits/chosen": -0.5814245939254761, | |
| "logits/rejected": -0.477074533700943, | |
| "logps/chosen": -359.9764404296875, | |
| "logps/rejected": -408.649169921875, | |
| "loss": 0.5854, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -0.8252065777778625, | |
| "rewards/margins": 0.3438906669616699, | |
| "rewards/rejected": -1.1690973043441772, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.5988023952095808, | |
| "grad_norm": 4.522965998928778, | |
| "learning_rate": 1.2498468799258466e-07, | |
| "logits/chosen": -0.6931585669517517, | |
| "logits/rejected": -0.6726911067962646, | |
| "logps/chosen": -447.56134033203125, | |
| "logps/rejected": -467.0315856933594, | |
| "loss": 0.5951, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -1.0262882709503174, | |
| "rewards/margins": 0.3147868514060974, | |
| "rewards/rejected": -1.3410751819610596, | |
| "step": 300 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 501, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 101, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |