Invalid JSON:Unexpected token 'N', ..."ad_norm": NaN,
"... is not valid JSON
| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 1407, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.010666666666666666, | |
| "grad_norm": NaN, | |
| "learning_rate": 4.651162790697675e-07, | |
| "logits/chosen": -1.3544222116470337, | |
| "logits/rejected": -1.2236188650131226, | |
| "logps/chosen": -231.80038452148438, | |
| "logps/rejected": -251.7948760986328, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.3125, | |
| "rewards/chosen": -7.575037307105958e-05, | |
| "rewards/margins": 6.424905586754903e-05, | |
| "rewards/rejected": -0.00013999939255882055, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.021333333333333333, | |
| "grad_norm": 3.109471559524536, | |
| "learning_rate": 1.0465116279069768e-06, | |
| "logits/chosen": -1.308906078338623, | |
| "logits/rejected": -1.306926965713501, | |
| "logps/chosen": -224.56344604492188, | |
| "logps/rejected": -251.70864868164062, | |
| "loss": 0.6922, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.0001932097366079688, | |
| "rewards/margins": 0.0019199944799765944, | |
| "rewards/rejected": -0.0021132039837539196, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.032, | |
| "grad_norm": NaN, | |
| "learning_rate": 1.6279069767441862e-06, | |
| "logits/chosen": -1.3518203496932983, | |
| "logits/rejected": -1.2913354635238647, | |
| "logps/chosen": -241.5867156982422, | |
| "logps/rejected": -274.6536865234375, | |
| "loss": 0.688, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -0.0011881353566423059, | |
| "rewards/margins": 0.010762467980384827, | |
| "rewards/rejected": -0.011950601823627949, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.042666666666666665, | |
| "grad_norm": 2.744222402572632, | |
| "learning_rate": 2.2093023255813954e-06, | |
| "logits/chosen": -1.3305720090866089, | |
| "logits/rejected": -1.2945411205291748, | |
| "logps/chosen": -238.5529022216797, | |
| "logps/rejected": -261.58392333984375, | |
| "loss": 0.6826, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -0.005761843174695969, | |
| "rewards/margins": 0.02146998979151249, | |
| "rewards/rejected": -0.02723183110356331, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05333333333333334, | |
| "grad_norm": 4.8908371925354, | |
| "learning_rate": 2.790697674418605e-06, | |
| "logits/chosen": -1.3832628726959229, | |
| "logits/rejected": -1.3136870861053467, | |
| "logps/chosen": -220.1907958984375, | |
| "logps/rejected": -271.0519104003906, | |
| "loss": 0.6735, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.013803968206048012, | |
| "rewards/margins": 0.04084903746843338, | |
| "rewards/rejected": -0.05465300753712654, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 3.333070755004883, | |
| "learning_rate": 3.372093023255814e-06, | |
| "logits/chosen": -1.3710800409317017, | |
| "logits/rejected": -1.2284512519836426, | |
| "logps/chosen": -249.5676727294922, | |
| "logps/rejected": -253.8758544921875, | |
| "loss": 0.6666, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.03833365440368652, | |
| "rewards/margins": 0.05555647611618042, | |
| "rewards/rejected": -0.09389013051986694, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07466666666666667, | |
| "grad_norm": 3.271761417388916, | |
| "learning_rate": 3.953488372093024e-06, | |
| "logits/chosen": -1.4493813514709473, | |
| "logits/rejected": -1.3014929294586182, | |
| "logps/chosen": -238.13302612304688, | |
| "logps/rejected": -246.8301239013672, | |
| "loss": 0.6531, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.05066124349832535, | |
| "rewards/margins": 0.08616013824939728, | |
| "rewards/rejected": -0.13682137429714203, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08533333333333333, | |
| "grad_norm": 2.7482733726501465, | |
| "learning_rate": 4.5348837209302326e-06, | |
| "logits/chosen": -1.3479039669036865, | |
| "logits/rejected": -1.28487229347229, | |
| "logps/chosen": -239.9534149169922, | |
| "logps/rejected": -261.44146728515625, | |
| "loss": 0.637, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.07571917772293091, | |
| "rewards/margins": 0.12467632442712784, | |
| "rewards/rejected": -0.20039549469947815, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 3.2934110164642334, | |
| "learning_rate": 4.99999336897255e-06, | |
| "logits/chosen": -1.452958345413208, | |
| "logits/rejected": -1.312406301498413, | |
| "logps/chosen": -239.06747436523438, | |
| "logps/rejected": -248.6891632080078, | |
| "loss": 0.6173, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -0.09406520426273346, | |
| "rewards/margins": 0.16908395290374756, | |
| "rewards/rejected": -0.2631491422653198, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10666666666666667, | |
| "grad_norm": 2.822927951812744, | |
| "learning_rate": 4.999761286705285e-06, | |
| "logits/chosen": -1.401163101196289, | |
| "logits/rejected": -1.3909813165664673, | |
| "logps/chosen": -242.43093872070312, | |
| "logps/rejected": -258.1756286621094, | |
| "loss": 0.5937, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.12050650268793106, | |
| "rewards/margins": 0.23688121140003204, | |
| "rewards/rejected": -0.3573877215385437, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11733333333333333, | |
| "grad_norm": 3.2107081413269043, | |
| "learning_rate": 4.999197688241076e-06, | |
| "logits/chosen": -1.420990228652954, | |
| "logits/rejected": -1.3618907928466797, | |
| "logps/chosen": -232.1669921875, | |
| "logps/rejected": -267.45782470703125, | |
| "loss": 0.5926, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.15464790165424347, | |
| "rewards/margins": 0.25124019384384155, | |
| "rewards/rejected": -0.4058881402015686, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 3.5179314613342285, | |
| "learning_rate": 4.9983026483238685e-06, | |
| "logits/chosen": -1.4916331768035889, | |
| "logits/rejected": -1.3871724605560303, | |
| "logps/chosen": -236.6553497314453, | |
| "logps/rejected": -259.3423767089844, | |
| "loss": 0.5453, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -0.189641535282135, | |
| "rewards/margins": 0.38970229029655457, | |
| "rewards/rejected": -0.5793437957763672, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13866666666666666, | |
| "grad_norm": 2.8678946495056152, | |
| "learning_rate": 4.997076285653087e-06, | |
| "logits/chosen": -1.5075359344482422, | |
| "logits/rejected": -1.4784154891967773, | |
| "logps/chosen": -250.3011016845703, | |
| "logps/rejected": -271.4066467285156, | |
| "loss": 0.5349, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": -0.25288641452789307, | |
| "rewards/margins": 0.4173409938812256, | |
| "rewards/rejected": -0.6702274084091187, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14933333333333335, | |
| "grad_norm": 3.1192572116851807, | |
| "learning_rate": 4.995518762867897e-06, | |
| "logits/chosen": -1.4870703220367432, | |
| "logits/rejected": -1.4476995468139648, | |
| "logps/chosen": -215.5365753173828, | |
| "logps/rejected": -257.3900146484375, | |
| "loss": 0.5312, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -0.3431021571159363, | |
| "rewards/margins": 0.4714565873146057, | |
| "rewards/rejected": -0.8145586848258972, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 2.9746811389923096, | |
| "learning_rate": 4.993630286525634e-06, | |
| "logits/chosen": -1.5514018535614014, | |
| "logits/rejected": -1.3919061422348022, | |
| "logps/chosen": -232.2204132080078, | |
| "logps/rejected": -262.62286376953125, | |
| "loss": 0.482, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -0.3241270184516907, | |
| "rewards/margins": 0.6433579325675964, | |
| "rewards/rejected": -0.9674848318099976, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.17066666666666666, | |
| "grad_norm": 2.728930711746216, | |
| "learning_rate": 4.99141110707441e-06, | |
| "logits/chosen": -1.463017225265503, | |
| "logits/rejected": -1.3513896465301514, | |
| "logps/chosen": -257.9603576660156, | |
| "logps/rejected": -288.959716796875, | |
| "loss": 0.4826, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -0.379680335521698, | |
| "rewards/margins": 0.6570689082145691, | |
| "rewards/rejected": -1.036749243736267, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.18133333333333335, | |
| "grad_norm": 2.79579758644104, | |
| "learning_rate": 4.9888615188199e-06, | |
| "logits/chosen": -1.5496175289154053, | |
| "logits/rejected": -1.4801855087280273, | |
| "logps/chosen": -238.9860382080078, | |
| "logps/rejected": -276.8408203125, | |
| "loss": 0.4282, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -0.3846990466117859, | |
| "rewards/margins": 0.8239493370056152, | |
| "rewards/rejected": -1.208648443222046, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 2.9647085666656494, | |
| "learning_rate": 4.9859818598863104e-06, | |
| "logits/chosen": -1.4761499166488647, | |
| "logits/rejected": -1.3574707508087158, | |
| "logps/chosen": -235.34024047851562, | |
| "logps/rejected": -261.658203125, | |
| "loss": 0.462, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -0.4849798083305359, | |
| "rewards/margins": 0.7756170034408569, | |
| "rewards/rejected": -1.2605968713760376, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.20266666666666666, | |
| "grad_norm": 3.024820327758789, | |
| "learning_rate": 4.982772512171538e-06, | |
| "logits/chosen": -1.5438058376312256, | |
| "logits/rejected": -1.4416005611419678, | |
| "logps/chosen": -239.1410369873047, | |
| "logps/rejected": -278.98419189453125, | |
| "loss": 0.4506, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -0.6898207664489746, | |
| "rewards/margins": 0.8674094080924988, | |
| "rewards/rejected": -1.5572302341461182, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.21333333333333335, | |
| "grad_norm": 2.9155285358428955, | |
| "learning_rate": 4.979233901296523e-06, | |
| "logits/chosen": -1.5428584814071655, | |
| "logits/rejected": -1.4572758674621582, | |
| "logps/chosen": -210.7685546875, | |
| "logps/rejected": -254.948974609375, | |
| "loss": 0.3884, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -0.6256662011146545, | |
| "rewards/margins": 1.2459337711334229, | |
| "rewards/rejected": -1.8715999126434326, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.224, | |
| "grad_norm": 2.731834650039673, | |
| "learning_rate": 4.975366496548804e-06, | |
| "logits/chosen": -1.5170609951019287, | |
| "logits/rejected": -1.4053804874420166, | |
| "logps/chosen": -241.0113983154297, | |
| "logps/rejected": -256.126220703125, | |
| "loss": 0.4358, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -0.6871055364608765, | |
| "rewards/margins": 0.9470389485359192, | |
| "rewards/rejected": -1.6341444253921509, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23466666666666666, | |
| "grad_norm": 2.4331085681915283, | |
| "learning_rate": 4.971170810820279e-06, | |
| "logits/chosen": -1.425811529159546, | |
| "logits/rejected": -1.4387760162353516, | |
| "logps/chosen": -232.7527313232422, | |
| "logps/rejected": -277.49896240234375, | |
| "loss": 0.3911, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -0.6859896779060364, | |
| "rewards/margins": 1.2273075580596924, | |
| "rewards/rejected": -1.9132972955703735, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.24533333333333332, | |
| "grad_norm": 2.802173137664795, | |
| "learning_rate": 4.96664740053919e-06, | |
| "logits/chosen": -1.4978256225585938, | |
| "logits/rejected": -1.4406821727752686, | |
| "logps/chosen": -230.27685546875, | |
| "logps/rejected": -288.71697998046875, | |
| "loss": 0.3774, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -0.8096278309822083, | |
| "rewards/margins": 1.2959668636322021, | |
| "rewards/rejected": -2.1055946350097656, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 2.4168198108673096, | |
| "learning_rate": 4.961796865596328e-06, | |
| "logits/chosen": -1.4427374601364136, | |
| "logits/rejected": -1.352508306503296, | |
| "logps/chosen": -242.40817260742188, | |
| "logps/rejected": -288.6490173339844, | |
| "loss": 0.3858, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": -0.9087854623794556, | |
| "rewards/margins": 1.4612674713134766, | |
| "rewards/rejected": -2.3700530529022217, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.26666666666666666, | |
| "grad_norm": 3.649536371231079, | |
| "learning_rate": 4.956619849265473e-06, | |
| "logits/chosen": -1.5077062845230103, | |
| "logits/rejected": -1.373518466949463, | |
| "logps/chosen": -241.6128387451172, | |
| "logps/rejected": -282.96295166015625, | |
| "loss": 0.4056, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -1.0533024072647095, | |
| "rewards/margins": 1.3088269233703613, | |
| "rewards/rejected": -2.3621296882629395, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2773333333333333, | |
| "grad_norm": 2.843197822570801, | |
| "learning_rate": 4.951117038118091e-06, | |
| "logits/chosen": -1.6640141010284424, | |
| "logits/rejected": -1.5276637077331543, | |
| "logps/chosen": -234.8096466064453, | |
| "logps/rejected": -284.8505554199219, | |
| "loss": 0.4381, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -0.9368289709091187, | |
| "rewards/margins": 1.2585748434066772, | |
| "rewards/rejected": -2.195403575897217, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 2.889314889907837, | |
| "learning_rate": 4.945289161932273e-06, | |
| "logits/chosen": -1.5789138078689575, | |
| "logits/rejected": -1.524099588394165, | |
| "logps/chosen": -248.79867553710938, | |
| "logps/rejected": -292.62481689453125, | |
| "loss": 0.3831, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -0.8927059173583984, | |
| "rewards/margins": 1.4558749198913574, | |
| "rewards/rejected": -2.348580837249756, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.2986666666666667, | |
| "grad_norm": 3.7556686401367188, | |
| "learning_rate": 4.93913699359596e-06, | |
| "logits/chosen": -1.6477344036102295, | |
| "logits/rejected": -1.4557158946990967, | |
| "logps/chosen": -230.6923828125, | |
| "logps/rejected": -281.36468505859375, | |
| "loss": 0.3376, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -0.8048444986343384, | |
| "rewards/margins": 1.6911342144012451, | |
| "rewards/rejected": -2.495978593826294, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.30933333333333335, | |
| "grad_norm": 2.8196702003479004, | |
| "learning_rate": 4.9326613490044364e-06, | |
| "logits/chosen": -1.5528147220611572, | |
| "logits/rejected": -1.4781819581985474, | |
| "logps/chosen": -224.55899047851562, | |
| "logps/rejected": -263.1064147949219, | |
| "loss": 0.3614, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -0.56626957654953, | |
| "rewards/margins": 1.4652245044708252, | |
| "rewards/rejected": -2.031494140625, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 3.5620477199554443, | |
| "learning_rate": 4.925863086952132e-06, | |
| "logits/chosen": -1.6504319906234741, | |
| "logits/rejected": -1.5435733795166016, | |
| "logps/chosen": -255.0889129638672, | |
| "logps/rejected": -297.24017333984375, | |
| "loss": 0.3755, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -0.7335026264190674, | |
| "rewards/margins": 1.4152415990829468, | |
| "rewards/rejected": -2.1487443447113037, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.33066666666666666, | |
| "grad_norm": 2.918090343475342, | |
| "learning_rate": 4.918743109018727e-06, | |
| "logits/chosen": -1.659012794494629, | |
| "logits/rejected": -1.5104398727416992, | |
| "logps/chosen": -234.96072387695312, | |
| "logps/rejected": -279.97979736328125, | |
| "loss": 0.3624, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -0.7696214914321899, | |
| "rewards/margins": 1.4934290647506714, | |
| "rewards/rejected": -2.2630505561828613, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.3413333333333333, | |
| "grad_norm": 2.317545175552368, | |
| "learning_rate": 4.911302359449585e-06, | |
| "logits/chosen": -1.6991151571273804, | |
| "logits/rejected": -1.5082671642303467, | |
| "logps/chosen": -232.7823486328125, | |
| "logps/rejected": -265.2646789550781, | |
| "loss": 0.3063, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": -0.5311766862869263, | |
| "rewards/margins": 1.7594959735870361, | |
| "rewards/rejected": -2.290672779083252, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.352, | |
| "grad_norm": 2.7809877395629883, | |
| "learning_rate": 4.9035418250305314e-06, | |
| "logits/chosen": -1.5785572528839111, | |
| "logits/rejected": -1.440476655960083, | |
| "logps/chosen": -239.0006103515625, | |
| "logps/rejected": -274.723876953125, | |
| "loss": 0.3032, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -0.5675729513168335, | |
| "rewards/margins": 1.8231475353240967, | |
| "rewards/rejected": -2.3907203674316406, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3626666666666667, | |
| "grad_norm": 5.299256801605225, | |
| "learning_rate": 4.895462534956981e-06, | |
| "logits/chosen": -1.5343797206878662, | |
| "logits/rejected": -1.501845121383667, | |
| "logps/chosen": -249.54916381835938, | |
| "logps/rejected": -271.220703125, | |
| "loss": 0.4059, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -0.8906148672103882, | |
| "rewards/margins": 1.5632604360580444, | |
| "rewards/rejected": -2.4538753032684326, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.37333333333333335, | |
| "grad_norm": 3.4545021057128906, | |
| "learning_rate": 4.887065560697449e-06, | |
| "logits/chosen": -1.5327939987182617, | |
| "logits/rejected": -1.5170584917068481, | |
| "logps/chosen": -257.06146240234375, | |
| "logps/rejected": -294.1949462890625, | |
| "loss": 0.3698, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -0.8633643984794617, | |
| "rewards/margins": 1.7004616260528564, | |
| "rewards/rejected": -2.563826084136963, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 3.70247745513916, | |
| "learning_rate": 4.878352015851459e-06, | |
| "logits/chosen": -1.652748703956604, | |
| "logits/rejected": -1.486902117729187, | |
| "logps/chosen": -253.253662109375, | |
| "logps/rejected": -273.6691589355469, | |
| "loss": 0.3075, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": -0.6801033020019531, | |
| "rewards/margins": 1.8728739023208618, | |
| "rewards/rejected": -2.5529773235321045, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.39466666666666667, | |
| "grad_norm": 2.5748231410980225, | |
| "learning_rate": 4.869323056001852e-06, | |
| "logits/chosen": -1.6040265560150146, | |
| "logits/rejected": -1.5244295597076416, | |
| "logps/chosen": -226.2984619140625, | |
| "logps/rejected": -279.1867980957031, | |
| "loss": 0.3129, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": -0.4969603419303894, | |
| "rewards/margins": 1.9832227230072021, | |
| "rewards/rejected": -2.4801833629608154, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.4053333333333333, | |
| "grad_norm": 3.6774702072143555, | |
| "learning_rate": 4.859979878561536e-06, | |
| "logits/chosen": -1.6129461526870728, | |
| "logits/rejected": -1.516275405883789, | |
| "logps/chosen": -237.805908203125, | |
| "logps/rejected": -278.765380859375, | |
| "loss": 0.3054, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": -0.6979565620422363, | |
| "rewards/margins": 2.057680130004883, | |
| "rewards/rejected": -2.755636692047119, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.416, | |
| "grad_norm": 3.621861219406128, | |
| "learning_rate": 4.850323722614685e-06, | |
| "logits/chosen": -1.588699460029602, | |
| "logits/rejected": -1.4784739017486572, | |
| "logps/chosen": -227.1220703125, | |
| "logps/rejected": -269.6943359375, | |
| "loss": 0.3308, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -0.7699406743049622, | |
| "rewards/margins": 1.8302541971206665, | |
| "rewards/rejected": -2.6001949310302734, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.4266666666666667, | |
| "grad_norm": 3.025399923324585, | |
| "learning_rate": 4.840355868752416e-06, | |
| "logits/chosen": -1.523004174232483, | |
| "logits/rejected": -1.4143463373184204, | |
| "logps/chosen": -284.00787353515625, | |
| "logps/rejected": -284.97100830078125, | |
| "loss": 0.3458, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": -0.6804647445678711, | |
| "rewards/margins": 1.930129051208496, | |
| "rewards/rejected": -2.610593795776367, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.43733333333333335, | |
| "grad_norm": 3.7301924228668213, | |
| "learning_rate": 4.830077638902955e-06, | |
| "logits/chosen": -1.6758838891983032, | |
| "logits/rejected": -1.541736364364624, | |
| "logps/chosen": -213.0896759033203, | |
| "logps/rejected": -273.7264099121094, | |
| "loss": 0.3491, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -0.8146641850471497, | |
| "rewards/margins": 2.013720989227295, | |
| "rewards/rejected": -2.828385353088379, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 3.1502671241760254, | |
| "learning_rate": 4.8194903961563255e-06, | |
| "logits/chosen": -1.65460205078125, | |
| "logits/rejected": -1.5929137468338013, | |
| "logps/chosen": -220.838134765625, | |
| "logps/rejected": -291.26788330078125, | |
| "loss": 0.2393, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -0.652218222618103, | |
| "rewards/margins": 2.579854726791382, | |
| "rewards/rejected": -3.2320728302001953, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.45866666666666667, | |
| "grad_norm": 3.069420337677002, | |
| "learning_rate": 4.808595544583578e-06, | |
| "logits/chosen": -1.6939719915390015, | |
| "logits/rejected": -1.6586183309555054, | |
| "logps/chosen": -225.1257781982422, | |
| "logps/rejected": -302.8365478515625, | |
| "loss": 0.274, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": -0.784705638885498, | |
| "rewards/margins": 2.293743371963501, | |
| "rewards/rejected": -3.078449010848999, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.4693333333333333, | |
| "grad_norm": 3.317854404449463, | |
| "learning_rate": 4.797394529050577e-06, | |
| "logits/chosen": -1.5986617803573608, | |
| "logits/rejected": -1.5012736320495605, | |
| "logps/chosen": -250.2396697998047, | |
| "logps/rejected": -291.0882263183594, | |
| "loss": 0.276, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": -0.6932470798492432, | |
| "rewards/margins": 2.4106459617614746, | |
| "rewards/rejected": -3.1038930416107178, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 14.15256118774414, | |
| "learning_rate": 4.785888835026393e-06, | |
| "logits/chosen": -1.5902974605560303, | |
| "logits/rejected": -1.534092664718628, | |
| "logps/chosen": -229.6944580078125, | |
| "logps/rejected": -281.76263427734375, | |
| "loss": 0.2861, | |
| "rewards/accuracies": 0.8812500238418579, | |
| "rewards/chosen": -0.8665626645088196, | |
| "rewards/margins": 2.3249261379241943, | |
| "rewards/rejected": -3.1914889812469482, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.49066666666666664, | |
| "grad_norm": 5.102518081665039, | |
| "learning_rate": 4.7740799883862966e-06, | |
| "logits/chosen": -1.6663917303085327, | |
| "logits/rejected": -1.5469465255737305, | |
| "logps/chosen": -239.5758819580078, | |
| "logps/rejected": -291.83526611328125, | |
| "loss": 0.3142, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -1.139182686805725, | |
| "rewards/margins": 2.0994646549224854, | |
| "rewards/rejected": -3.2386474609375, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5013333333333333, | |
| "grad_norm": 2.747270345687866, | |
| "learning_rate": 4.761969555209394e-06, | |
| "logits/chosen": -1.6959060430526733, | |
| "logits/rejected": -1.7034603357315063, | |
| "logps/chosen": -218.6403350830078, | |
| "logps/rejected": -296.04327392578125, | |
| "loss": 0.3023, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -1.0236613750457764, | |
| "rewards/margins": 2.3380959033966064, | |
| "rewards/rejected": -3.361757278442383, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 2.4223251342773438, | |
| "learning_rate": 4.7495591415709435e-06, | |
| "logits/chosen": -1.7274869680404663, | |
| "logits/rejected": -1.5388081073760986, | |
| "logps/chosen": -227.67236328125, | |
| "logps/rejected": -291.2453308105469, | |
| "loss": 0.2354, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": -0.7354499101638794, | |
| "rewards/margins": 2.73541522026062, | |
| "rewards/rejected": -3.470865249633789, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5226666666666666, | |
| "grad_norm": 4.390448570251465, | |
| "learning_rate": 4.736850393329353e-06, | |
| "logits/chosen": -1.6266454458236694, | |
| "logits/rejected": -1.512935996055603, | |
| "logps/chosen": -244.3636474609375, | |
| "logps/rejected": -295.836181640625, | |
| "loss": 0.2993, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -0.9331833124160767, | |
| "rewards/margins": 2.6201255321502686, | |
| "rewards/rejected": -3.5533084869384766, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5333333333333333, | |
| "grad_norm": 3.507237672805786, | |
| "learning_rate": 4.7238449959079095e-06, | |
| "logits/chosen": -1.753446340560913, | |
| "logits/rejected": -1.6419095993041992, | |
| "logps/chosen": -233.2483367919922, | |
| "logps/rejected": -294.78887939453125, | |
| "loss": 0.2506, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": -0.8268024325370789, | |
| "rewards/margins": 2.761507272720337, | |
| "rewards/rejected": -3.5883095264434814, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.544, | |
| "grad_norm": 2.5926451683044434, | |
| "learning_rate": 4.710544674071259e-06, | |
| "logits/chosen": -1.6371052265167236, | |
| "logits/rejected": -1.5320228338241577, | |
| "logps/chosen": -230.4346923828125, | |
| "logps/rejected": -286.22198486328125, | |
| "loss": 0.2302, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -0.7810002565383911, | |
| "rewards/margins": 2.938009262084961, | |
| "rewards/rejected": -3.7190093994140625, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5546666666666666, | |
| "grad_norm": 2.7341744899749756, | |
| "learning_rate": 4.696951191696673e-06, | |
| "logits/chosen": -1.6878139972686768, | |
| "logits/rejected": -1.575380563735962, | |
| "logps/chosen": -234.5509490966797, | |
| "logps/rejected": -295.4001159667969, | |
| "loss": 0.2846, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -1.2288405895233154, | |
| "rewards/margins": 2.4302990436553955, | |
| "rewards/rejected": -3.659140110015869, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5653333333333334, | |
| "grad_norm": 3.923440933227539, | |
| "learning_rate": 4.6830663515401196e-06, | |
| "logits/chosen": -1.7157907485961914, | |
| "logits/rejected": -1.5281307697296143, | |
| "logps/chosen": -255.41696166992188, | |
| "logps/rejected": -287.6669006347656, | |
| "loss": 0.2849, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": -0.8545368909835815, | |
| "rewards/margins": 2.664400815963745, | |
| "rewards/rejected": -3.518937587738037, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 2.7870829105377197, | |
| "learning_rate": 4.6688919949971855e-06, | |
| "logits/chosen": -1.6280025243759155, | |
| "logits/rejected": -1.5273808240890503, | |
| "logps/chosen": -256.6709899902344, | |
| "logps/rejected": -299.0733947753906, | |
| "loss": 0.2483, | |
| "rewards/accuracies": 0.8812500238418579, | |
| "rewards/chosen": -0.799675703048706, | |
| "rewards/margins": 2.8166801929473877, | |
| "rewards/rejected": -3.6163558959960938, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5866666666666667, | |
| "grad_norm": 3.013019561767578, | |
| "learning_rate": 4.654430001858874e-06, | |
| "logits/chosen": -1.5734882354736328, | |
| "logits/rejected": -1.5436983108520508, | |
| "logps/chosen": -248.3529815673828, | |
| "logps/rejected": -302.51617431640625, | |
| "loss": 0.2643, | |
| "rewards/accuracies": 0.8812500238418579, | |
| "rewards/chosen": -0.6249435544013977, | |
| "rewards/margins": 2.736476421356201, | |
| "rewards/rejected": -3.361420154571533, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5973333333333334, | |
| "grad_norm": 6.8396077156066895, | |
| "learning_rate": 4.6396822900623075e-06, | |
| "logits/chosen": -1.6659371852874756, | |
| "logits/rejected": -1.506557822227478, | |
| "logps/chosen": -238.4332733154297, | |
| "logps/rejected": -282.52227783203125, | |
| "loss": 0.3028, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -0.7782389521598816, | |
| "rewards/margins": 2.5956223011016846, | |
| "rewards/rejected": -3.373861312866211, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.608, | |
| "grad_norm": 3.025742292404175, | |
| "learning_rate": 4.624650815436368e-06, | |
| "logits/chosen": -1.6658992767333984, | |
| "logits/rejected": -1.6078733205795288, | |
| "logps/chosen": -248.2996368408203, | |
| "logps/rejected": -303.9361572265625, | |
| "loss": 0.2461, | |
| "rewards/accuracies": 0.9125000238418579, | |
| "rewards/chosen": -0.6723818778991699, | |
| "rewards/margins": 2.8119044303894043, | |
| "rewards/rejected": -3.4842865467071533, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6186666666666667, | |
| "grad_norm": 2.623469591140747, | |
| "learning_rate": 4.609337571442324e-06, | |
| "logits/chosen": -1.6829065084457397, | |
| "logits/rejected": -1.5887826681137085, | |
| "logps/chosen": -226.87155151367188, | |
| "logps/rejected": -293.0638732910156, | |
| "loss": 0.2396, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -0.7441937923431396, | |
| "rewards/margins": 2.8528599739074707, | |
| "rewards/rejected": -3.5970540046691895, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6293333333333333, | |
| "grad_norm": 3.053109645843506, | |
| "learning_rate": 4.5937445889094535e-06, | |
| "logits/chosen": -1.663840651512146, | |
| "logits/rejected": -1.557969331741333, | |
| "logps/chosen": -237.5575714111328, | |
| "logps/rejected": -270.58392333984375, | |
| "loss": 0.3104, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -0.4625615179538727, | |
| "rewards/margins": 2.317203998565674, | |
| "rewards/rejected": -2.7797656059265137, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 2.9096717834472656, | |
| "learning_rate": 4.577873935765722e-06, | |
| "logits/chosen": -1.7499110698699951, | |
| "logits/rejected": -1.525377631187439, | |
| "logps/chosen": -231.7384490966797, | |
| "logps/rejected": -274.1564636230469, | |
| "loss": 0.2208, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/chosen": -0.593193531036377, | |
| "rewards/margins": 2.9766085147857666, | |
| "rewards/rejected": -3.5698020458221436, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6506666666666666, | |
| "grad_norm": 3.812568426132202, | |
| "learning_rate": 4.561727716763532e-06, | |
| "logits/chosen": -1.6329309940338135, | |
| "logits/rejected": -1.5051950216293335, | |
| "logps/chosen": -255.3319091796875, | |
| "logps/rejected": -280.5289001464844, | |
| "loss": 0.2537, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": -0.8085314035415649, | |
| "rewards/margins": 2.630579948425293, | |
| "rewards/rejected": -3.4391112327575684, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6613333333333333, | |
| "grad_norm": 4.5825605392456055, | |
| "learning_rate": 4.545308073200591e-06, | |
| "logits/chosen": -1.618645429611206, | |
| "logits/rejected": -1.5387064218521118, | |
| "logps/chosen": -246.5149383544922, | |
| "logps/rejected": -281.2253112792969, | |
| "loss": 0.2933, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": -0.7762895822525024, | |
| "rewards/margins": 2.5413150787353516, | |
| "rewards/rejected": -3.3176045417785645, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.672, | |
| "grad_norm": 3.3296334743499756, | |
| "learning_rate": 4.528617182635943e-06, | |
| "logits/chosen": -1.7562048435211182, | |
| "logits/rejected": -1.7319389581680298, | |
| "logps/chosen": -232.3672637939453, | |
| "logps/rejected": -317.72369384765625, | |
| "loss": 0.2532, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -0.7930496335029602, | |
| "rewards/margins": 3.444871425628662, | |
| "rewards/rejected": -4.237921714782715, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6826666666666666, | |
| "grad_norm": 6.051812648773193, | |
| "learning_rate": 4.511657258601171e-06, | |
| "logits/chosen": -1.6694326400756836, | |
| "logits/rejected": -1.5220441818237305, | |
| "logps/chosen": -235.4072265625, | |
| "logps/rejected": -296.1033020019531, | |
| "loss": 0.2432, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": -0.9599354863166809, | |
| "rewards/margins": 3.038886308670044, | |
| "rewards/rejected": -3.998821973800659, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6933333333333334, | |
| "grad_norm": 5.517160892486572, | |
| "learning_rate": 4.494430550306848e-06, | |
| "logits/chosen": -1.5575799942016602, | |
| "logits/rejected": -1.5139790773391724, | |
| "logps/chosen": -254.32772827148438, | |
| "logps/rejected": -290.1891174316406, | |
| "loss": 0.3125, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -0.8352974057197571, | |
| "rewards/margins": 2.8006937503814697, | |
| "rewards/rejected": -3.635990858078003, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 2.6814022064208984, | |
| "learning_rate": 4.476939342344246e-06, | |
| "logits/chosen": -1.7075474262237549, | |
| "logits/rejected": -1.6463184356689453, | |
| "logps/chosen": -248.3602294921875, | |
| "logps/rejected": -305.9772033691406, | |
| "loss": 0.2713, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": -1.1044809818267822, | |
| "rewards/margins": 2.6876957416534424, | |
| "rewards/rejected": -3.7921767234802246, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7146666666666667, | |
| "grad_norm": 6.655895709991455, | |
| "learning_rate": 4.45918595438236e-06, | |
| "logits/chosen": -1.6074548959732056, | |
| "logits/rejected": -1.5420688390731812, | |
| "logps/chosen": -235.84518432617188, | |
| "logps/rejected": -275.97943115234375, | |
| "loss": 0.3032, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": -0.7538453340530396, | |
| "rewards/margins": 2.4696977138519287, | |
| "rewards/rejected": -3.223543643951416, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7253333333333334, | |
| "grad_norm": 2.56044602394104, | |
| "learning_rate": 4.441172740860267e-06, | |
| "logits/chosen": -1.6362565755844116, | |
| "logits/rejected": -1.5230262279510498, | |
| "logps/chosen": -252.67575073242188, | |
| "logps/rejected": -284.8129577636719, | |
| "loss": 0.2388, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": -0.6972413063049316, | |
| "rewards/margins": 2.9790282249450684, | |
| "rewards/rejected": -3.676270008087158, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.736, | |
| "grad_norm": 4.380439281463623, | |
| "learning_rate": 4.422902090674886e-06, | |
| "logits/chosen": -1.5679028034210205, | |
| "logits/rejected": -1.4958457946777344, | |
| "logps/chosen": -256.5689697265625, | |
| "logps/rejected": -279.14691162109375, | |
| "loss": 0.2577, | |
| "rewards/accuracies": 0.8812500238418579, | |
| "rewards/chosen": -0.9045771360397339, | |
| "rewards/margins": 2.8589415550231934, | |
| "rewards/rejected": -3.763519287109375, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7466666666666667, | |
| "grad_norm": 3.181389808654785, | |
| "learning_rate": 4.4043764268641705e-06, | |
| "logits/chosen": -1.6590477228164673, | |
| "logits/rejected": -1.5410702228546143, | |
| "logps/chosen": -251.5687255859375, | |
| "logps/rejected": -301.7498779296875, | |
| "loss": 0.2393, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": -0.5267749428749084, | |
| "rewards/margins": 3.0952720642089844, | |
| "rewards/rejected": -3.622046947479248, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7573333333333333, | |
| "grad_norm": 3.395481824874878, | |
| "learning_rate": 4.385598206285758e-06, | |
| "logits/chosen": -1.7113549709320068, | |
| "logits/rejected": -1.656198501586914, | |
| "logps/chosen": -247.5892791748047, | |
| "logps/rejected": -303.3941345214844, | |
| "loss": 0.275, | |
| "rewards/accuracies": 0.8812500238418579, | |
| "rewards/chosen": -0.4766809344291687, | |
| "rewards/margins": 2.8971753120422363, | |
| "rewards/rejected": -3.3738560676574707, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 71.72124481201172, | |
| "learning_rate": 4.3665699192911495e-06, | |
| "logits/chosen": -1.5628933906555176, | |
| "logits/rejected": -1.5330002307891846, | |
| "logps/chosen": -289.46942138671875, | |
| "logps/rejected": -281.2159423828125, | |
| "loss": 0.2754, | |
| "rewards/accuracies": 0.8812500238418579, | |
| "rewards/chosen": -0.38477277755737305, | |
| "rewards/margins": 2.6930577754974365, | |
| "rewards/rejected": -3.0778305530548096, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7786666666666666, | |
| "grad_norm": 2.965637445449829, | |
| "learning_rate": 4.347294089395438e-06, | |
| "logits/chosen": -1.672273874282837, | |
| "logits/rejected": -1.5576097965240479, | |
| "logps/chosen": -253.80630493164062, | |
| "logps/rejected": -281.87152099609375, | |
| "loss": 0.2398, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": -0.35890644788742065, | |
| "rewards/margins": 2.8527824878692627, | |
| "rewards/rejected": -3.211688995361328, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7893333333333333, | |
| "grad_norm": 4.277189254760742, | |
| "learning_rate": 4.327773272942647e-06, | |
| "logits/chosen": -1.6600908041000366, | |
| "logits/rejected": -1.590300440788269, | |
| "logps/chosen": -231.3988800048828, | |
| "logps/rejected": -298.57745361328125, | |
| "loss": 0.2088, | |
| "rewards/accuracies": 0.9125000238418579, | |
| "rewards/chosen": -0.40050333738327026, | |
| "rewards/margins": 3.0887906551361084, | |
| "rewards/rejected": -3.4892940521240234, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 2.961402654647827, | |
| "learning_rate": 4.308010058766703e-06, | |
| "logits/chosen": -1.660800576210022, | |
| "logits/rejected": -1.5310600996017456, | |
| "logps/chosen": -255.83114624023438, | |
| "logps/rejected": -295.5674743652344, | |
| "loss": 0.2092, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/chosen": -0.6556711792945862, | |
| "rewards/margins": 3.229142427444458, | |
| "rewards/rejected": -3.8848137855529785, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.8106666666666666, | |
| "grad_norm": 3.6059043407440186, | |
| "learning_rate": 4.2880070678481135e-06, | |
| "logits/chosen": -1.7003570795059204, | |
| "logits/rejected": -1.5770372152328491, | |
| "logps/chosen": -223.58554077148438, | |
| "logps/rejected": -292.36114501953125, | |
| "loss": 0.2584, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -0.6750256419181824, | |
| "rewards/margins": 3.2846837043762207, | |
| "rewards/rejected": -3.9597091674804688, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8213333333333334, | |
| "grad_norm": 3.216796875, | |
| "learning_rate": 4.267766952966369e-06, | |
| "logits/chosen": -1.6272624731063843, | |
| "logits/rejected": -1.571340799331665, | |
| "logps/chosen": -241.29873657226562, | |
| "logps/rejected": -301.0836486816406, | |
| "loss": 0.2603, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -0.8578468561172485, | |
| "rewards/margins": 3.2342536449432373, | |
| "rewards/rejected": -4.092100620269775, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 4.120914936065674, | |
| "learning_rate": 4.24729239834814e-06, | |
| "logits/chosen": -1.626132607460022, | |
| "logits/rejected": -1.52275550365448, | |
| "logps/chosen": -259.8028869628906, | |
| "logps/rejected": -295.94512939453125, | |
| "loss": 0.281, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -0.6032441854476929, | |
| "rewards/margins": 3.1042075157165527, | |
| "rewards/rejected": -3.707451581954956, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8426666666666667, | |
| "grad_norm": 3.3638036251068115, | |
| "learning_rate": 4.226586119311291e-06, | |
| "logits/chosen": -1.6753953695297241, | |
| "logits/rejected": -1.6117645502090454, | |
| "logps/chosen": -241.83935546875, | |
| "logps/rejected": -291.6381530761719, | |
| "loss": 0.2388, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/chosen": -0.4112634062767029, | |
| "rewards/margins": 3.3937580585479736, | |
| "rewards/rejected": -3.8050217628479004, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8533333333333334, | |
| "grad_norm": 2.242154359817505, | |
| "learning_rate": 4.205650861904782e-06, | |
| "logits/chosen": -1.6429239511489868, | |
| "logits/rejected": -1.5123769044876099, | |
| "logps/chosen": -222.09957885742188, | |
| "logps/rejected": -289.91717529296875, | |
| "loss": 0.2233, | |
| "rewards/accuracies": 0.9125000238418579, | |
| "rewards/chosen": -0.8481672406196594, | |
| "rewards/margins": 3.0394561290740967, | |
| "rewards/rejected": -3.8876235485076904, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.864, | |
| "grad_norm": 4.096096992492676, | |
| "learning_rate": 4.184489402544486e-06, | |
| "logits/chosen": -1.7357295751571655, | |
| "logits/rejected": -1.6345306634902954, | |
| "logps/chosen": -232.3779754638672, | |
| "logps/rejected": -294.068115234375, | |
| "loss": 0.2335, | |
| "rewards/accuracies": 0.918749988079071, | |
| "rewards/chosen": -0.7445335388183594, | |
| "rewards/margins": 3.0647544860839844, | |
| "rewards/rejected": -3.809288501739502, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8746666666666667, | |
| "grad_norm": 3.3298556804656982, | |
| "learning_rate": 4.1631045476449865e-06, | |
| "logits/chosen": -1.7491382360458374, | |
| "logits/rejected": -1.6363424062728882, | |
| "logps/chosen": -240.1590576171875, | |
| "logps/rejected": -309.88482666015625, | |
| "loss": 0.2004, | |
| "rewards/accuracies": 0.918749988079071, | |
| "rewards/chosen": -0.7465359568595886, | |
| "rewards/margins": 3.701233386993408, | |
| "rewards/rejected": -4.4477691650390625, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8853333333333333, | |
| "grad_norm": 2.753004789352417, | |
| "learning_rate": 4.141499133247391e-06, | |
| "logits/chosen": -1.5934135913848877, | |
| "logits/rejected": -1.5748265981674194, | |
| "logps/chosen": -242.0540771484375, | |
| "logps/rejected": -309.4254455566406, | |
| "loss": 0.2532, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": -0.6847050786018372, | |
| "rewards/margins": 3.4728336334228516, | |
| "rewards/rejected": -4.157538414001465, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 3.4208219051361084, | |
| "learning_rate": 4.119676024643217e-06, | |
| "logits/chosen": -1.5917304754257202, | |
| "logits/rejected": -1.5950918197631836, | |
| "logps/chosen": -243.2762908935547, | |
| "logps/rejected": -299.3033447265625, | |
| "loss": 0.1765, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -0.5414136648178101, | |
| "rewards/margins": 3.837564468383789, | |
| "rewards/rejected": -4.378978252410889, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9066666666666666, | |
| "grad_norm": 5.384462833404541, | |
| "learning_rate": 4.0976381159944055e-06, | |
| "logits/chosen": -1.703701376914978, | |
| "logits/rejected": -1.5836318731307983, | |
| "logps/chosen": -228.7626190185547, | |
| "logps/rejected": -318.1847229003906, | |
| "loss": 0.2142, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -0.7460819482803345, | |
| "rewards/margins": 3.7437853813171387, | |
| "rewards/rejected": -4.489867210388184, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9173333333333333, | |
| "grad_norm": 2.9159977436065674, | |
| "learning_rate": 4.07538832994949e-06, | |
| "logits/chosen": -1.6830871105194092, | |
| "logits/rejected": -1.5406839847564697, | |
| "logps/chosen": -243.9796142578125, | |
| "logps/rejected": -309.07666015625, | |
| "loss": 0.2847, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/chosen": -1.0015039443969727, | |
| "rewards/margins": 3.680140733718872, | |
| "rewards/rejected": -4.681644439697266, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.928, | |
| "grad_norm": 3.914306402206421, | |
| "learning_rate": 4.0529296172560055e-06, | |
| "logits/chosen": -1.7112497091293335, | |
| "logits/rejected": -1.5582363605499268, | |
| "logps/chosen": -237.08541870117188, | |
| "logps/rejected": -321.8060302734375, | |
| "loss": 0.257, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -1.2424498796463013, | |
| "rewards/margins": 3.5792899131774902, | |
| "rewards/rejected": -4.82174015045166, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9386666666666666, | |
| "grad_norm": 4.649114608764648, | |
| "learning_rate": 4.030264956369158e-06, | |
| "logits/chosen": -1.5580627918243408, | |
| "logits/rejected": -1.4607374668121338, | |
| "logps/chosen": -234.0313262939453, | |
| "logps/rejected": -297.1539306640625, | |
| "loss": 0.2303, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": -1.0807582139968872, | |
| "rewards/margins": 3.3951945304870605, | |
| "rewards/rejected": -4.475952625274658, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9493333333333334, | |
| "grad_norm": 4.41303014755249, | |
| "learning_rate": 4.0073973530568274e-06, | |
| "logits/chosen": -1.495747447013855, | |
| "logits/rejected": -1.4580141305923462, | |
| "logps/chosen": -226.8778839111328, | |
| "logps/rejected": -291.4224548339844, | |
| "loss": 0.2653, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": -0.8967695236206055, | |
| "rewards/margins": 3.4394097328186035, | |
| "rewards/rejected": -4.336178779602051, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 2.877248525619507, | |
| "learning_rate": 3.9843298400009435e-06, | |
| "logits/chosen": -1.7053565979003906, | |
| "logits/rejected": -1.6477537155151367, | |
| "logps/chosen": -239.79800415039062, | |
| "logps/rejected": -309.544189453125, | |
| "loss": 0.2526, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -1.0235546827316284, | |
| "rewards/margins": 3.4787325859069824, | |
| "rewards/rejected": -4.502286911010742, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9706666666666667, | |
| "grad_norm": 6.4316725730896, | |
| "learning_rate": 3.961065476395296e-06, | |
| "logits/chosen": -1.768538236618042, | |
| "logits/rejected": -1.632031798362732, | |
| "logps/chosen": -250.4674530029297, | |
| "logps/rejected": -336.96759033203125, | |
| "loss": 0.1837, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/chosen": -0.5413009524345398, | |
| "rewards/margins": 4.095879554748535, | |
| "rewards/rejected": -4.637180805206299, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9813333333333333, | |
| "grad_norm": 3.302759885787964, | |
| "learning_rate": 3.937607347539823e-06, | |
| "logits/chosen": -1.685744047164917, | |
| "logits/rejected": -1.5522499084472656, | |
| "logps/chosen": -227.767578125, | |
| "logps/rejected": -318.6248474121094, | |
| "loss": 0.2087, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": -0.8970295190811157, | |
| "rewards/margins": 3.9703762531280518, | |
| "rewards/rejected": -4.867405891418457, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.992, | |
| "grad_norm": 4.109023571014404, | |
| "learning_rate": 3.913958564431447e-06, | |
| "logits/chosen": -1.8247349262237549, | |
| "logits/rejected": -1.6116182804107666, | |
| "logps/chosen": -255.1387176513672, | |
| "logps/rejected": -299.3416748046875, | |
| "loss": 0.2102, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": -0.7889047861099243, | |
| "rewards/margins": 4.012124538421631, | |
| "rewards/rejected": -4.801029682159424, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.0021333333333333, | |
| "grad_norm": 4.297785758972168, | |
| "learning_rate": 3.89012226335149e-06, | |
| "logits/chosen": -1.5690182447433472, | |
| "logits/rejected": -1.5636067390441895, | |
| "logps/chosen": -235.01690673828125, | |
| "logps/rejected": -301.91552734375, | |
| "loss": 0.2782, | |
| "rewards/accuracies": 0.8618420958518982, | |
| "rewards/chosen": -1.062803030014038, | |
| "rewards/margins": 3.59759521484375, | |
| "rewards/rejected": -4.660398006439209, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.0128, | |
| "grad_norm": 2.434410572052002, | |
| "learning_rate": 3.86610160544975e-06, | |
| "logits/chosen": -1.6619993448257446, | |
| "logits/rejected": -1.6416118144989014, | |
| "logps/chosen": -253.34414672851562, | |
| "logps/rejected": -312.0550842285156, | |
| "loss": 0.2136, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": -0.7877097725868225, | |
| "rewards/margins": 3.9225335121154785, | |
| "rewards/rejected": -4.71024227142334, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.0234666666666667, | |
| "grad_norm": 3.1037039756774902, | |
| "learning_rate": 3.841899776325267e-06, | |
| "logits/chosen": -1.5927152633666992, | |
| "logits/rejected": -1.4927324056625366, | |
| "logps/chosen": -250.84481811523438, | |
| "logps/rejected": -291.4174499511719, | |
| "loss": 0.1873, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": -0.5820282101631165, | |
| "rewards/margins": 3.742619276046753, | |
| "rewards/rejected": -4.324646949768066, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.0341333333333333, | |
| "grad_norm": 2.221498489379883, | |
| "learning_rate": 3.817519985603855e-06, | |
| "logits/chosen": -1.7725327014923096, | |
| "logits/rejected": -1.7125329971313477, | |
| "logps/chosen": -221.89370727539062, | |
| "logps/rejected": -311.96270751953125, | |
| "loss": 0.1119, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -0.5550159215927124, | |
| "rewards/margins": 4.137165546417236, | |
| "rewards/rejected": -4.692181587219238, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.0448, | |
| "grad_norm": 3.163022756576538, | |
| "learning_rate": 3.792965466512437e-06, | |
| "logits/chosen": -1.7214984893798828, | |
| "logits/rejected": -1.6236610412597656, | |
| "logps/chosen": -243.5190887451172, | |
| "logps/rejected": -310.9081115722656, | |
| "loss": 0.2174, | |
| "rewards/accuracies": 0.918749988079071, | |
| "rewards/chosen": -0.657336950302124, | |
| "rewards/margins": 3.4691410064697266, | |
| "rewards/rejected": -4.1264777183532715, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.0554666666666668, | |
| "grad_norm": 1.9490315914154053, | |
| "learning_rate": 3.7682394754502687e-06, | |
| "logits/chosen": -1.7172479629516602, | |
| "logits/rejected": -1.6332595348358154, | |
| "logps/chosen": -227.83163452148438, | |
| "logps/rejected": -305.6872863769531, | |
| "loss": 0.189, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/chosen": -0.819398045539856, | |
| "rewards/margins": 4.044649124145508, | |
| "rewards/rejected": -4.864047050476074, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.0661333333333334, | |
| "grad_norm": 2.395745277404785, | |
| "learning_rate": 3.7433452915570685e-06, | |
| "logits/chosen": -1.7560336589813232, | |
| "logits/rejected": -1.5769942998886108, | |
| "logps/chosen": -240.0421905517578, | |
| "logps/rejected": -305.04827880859375, | |
| "loss": 0.1668, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -0.8382463455200195, | |
| "rewards/margins": 4.010151386260986, | |
| "rewards/rejected": -4.848397254943848, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0768, | |
| "grad_norm": 3.2161705493927, | |
| "learning_rate": 3.718286216278143e-06, | |
| "logits/chosen": -1.7935104370117188, | |
| "logits/rejected": -1.6894598007202148, | |
| "logps/chosen": -250.8125762939453, | |
| "logps/rejected": -325.422607421875, | |
| "loss": 0.1802, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -1.3864238262176514, | |
| "rewards/margins": 4.100071907043457, | |
| "rewards/rejected": -5.486495018005371, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.0874666666666666, | |
| "grad_norm": 3.6139273643493652, | |
| "learning_rate": 3.6930655729265557e-06, | |
| "logits/chosen": -1.659446120262146, | |
| "logits/rejected": -1.6318769454956055, | |
| "logps/chosen": -251.3518829345703, | |
| "logps/rejected": -334.08917236328125, | |
| "loss": 0.1798, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": -1.2920472621917725, | |
| "rewards/margins": 4.446113586425781, | |
| "rewards/rejected": -5.738161087036133, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.0981333333333334, | |
| "grad_norm": 4.280843257904053, | |
| "learning_rate": 3.667686706242386e-06, | |
| "logits/chosen": -1.5787420272827148, | |
| "logits/rejected": -1.4920685291290283, | |
| "logps/chosen": -254.0977325439453, | |
| "logps/rejected": -299.56256103515625, | |
| "loss": 0.2441, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": -1.4882808923721313, | |
| "rewards/margins": 4.044205188751221, | |
| "rewards/rejected": -5.5324859619140625, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.1088, | |
| "grad_norm": 2.1550538539886475, | |
| "learning_rate": 3.6421529819491576e-06, | |
| "logits/chosen": -1.6968841552734375, | |
| "logits/rejected": -1.6490501165390015, | |
| "logps/chosen": -250.90933227539062, | |
| "logps/rejected": -326.4671936035156, | |
| "loss": 0.188, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": -1.3018690347671509, | |
| "rewards/margins": 4.414233207702637, | |
| "rewards/rejected": -5.716102600097656, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.1194666666666666, | |
| "grad_norm": 1.9021104574203491, | |
| "learning_rate": 3.6164677863074783e-06, | |
| "logits/chosen": -1.6301161050796509, | |
| "logits/rejected": -1.6059134006500244, | |
| "logps/chosen": -228.54275512695312, | |
| "logps/rejected": -297.1298522949219, | |
| "loss": 0.2367, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -1.2790645360946655, | |
| "rewards/margins": 3.9954991340637207, | |
| "rewards/rejected": -5.274563789367676, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.1301333333333332, | |
| "grad_norm": 3.255465030670166, | |
| "learning_rate": 3.590634525665955e-06, | |
| "logits/chosen": -1.7874761819839478, | |
| "logits/rejected": -1.6896089315414429, | |
| "logps/chosen": -233.70816040039062, | |
| "logps/rejected": -306.7872314453125, | |
| "loss": 0.1776, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -1.035360336303711, | |
| "rewards/margins": 4.155150413513184, | |
| "rewards/rejected": -5.190511703491211, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.1408, | |
| "grad_norm": 2.013275623321533, | |
| "learning_rate": 3.564656626009446e-06, | |
| "logits/chosen": -1.8429739475250244, | |
| "logits/rejected": -1.6779648065567017, | |
| "logps/chosen": -247.69906616210938, | |
| "logps/rejected": -327.61865234375, | |
| "loss": 0.1586, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": -1.3286124467849731, | |
| "rewards/margins": 4.181220054626465, | |
| "rewards/rejected": -5.509832382202148, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.1514666666666666, | |
| "grad_norm": 4.79525089263916, | |
| "learning_rate": 3.5385375325047167e-06, | |
| "logits/chosen": -1.753906011581421, | |
| "logits/rejected": -1.6913907527923584, | |
| "logps/chosen": -251.79751586914062, | |
| "logps/rejected": -307.3892822265625, | |
| "loss": 0.1615, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -1.2266600131988525, | |
| "rewards/margins": 4.172353267669678, | |
| "rewards/rejected": -5.399013519287109, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.1621333333333332, | |
| "grad_norm": 3.9382004737854004, | |
| "learning_rate": 3.512280709043534e-06, | |
| "logits/chosen": -1.6286598443984985, | |
| "logits/rejected": -1.4853745698928833, | |
| "logps/chosen": -248.4619140625, | |
| "logps/rejected": -293.0491943359375, | |
| "loss": 0.2146, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/chosen": -1.1233129501342773, | |
| "rewards/margins": 3.978351593017578, | |
| "rewards/rejected": -5.101665019989014, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.1728, | |
| "grad_norm": 4.661540985107422, | |
| "learning_rate": 3.4858896377832966e-06, | |
| "logits/chosen": -1.7429920434951782, | |
| "logits/rejected": -1.6357555389404297, | |
| "logps/chosen": -245.42489624023438, | |
| "logps/rejected": -297.8553771972656, | |
| "loss": 0.1958, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -1.1029307842254639, | |
| "rewards/margins": 3.8211007118225098, | |
| "rewards/rejected": -4.9240312576293945, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.1834666666666667, | |
| "grad_norm": 2.63988995552063, | |
| "learning_rate": 3.459367818685231e-06, | |
| "logits/chosen": -1.7078949213027954, | |
| "logits/rejected": -1.5824353694915771, | |
| "logps/chosen": -244.9107666015625, | |
| "logps/rejected": -318.76898193359375, | |
| "loss": 0.1662, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/chosen": -0.956513524055481, | |
| "rewards/margins": 4.618618011474609, | |
| "rewards/rejected": -5.575131416320801, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.1941333333333333, | |
| "grad_norm": 4.890956878662109, | |
| "learning_rate": 3.432718769050228e-06, | |
| "logits/chosen": -1.6049926280975342, | |
| "logits/rejected": -1.4823522567749023, | |
| "logps/chosen": -243.07571411132812, | |
| "logps/rejected": -317.7985534667969, | |
| "loss": 0.1449, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -0.7478665113449097, | |
| "rewards/margins": 4.851959228515625, | |
| "rewards/rejected": -5.599825382232666, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.2048, | |
| "grad_norm": 3.967740774154663, | |
| "learning_rate": 3.405946023052386e-06, | |
| "logits/chosen": -1.7225697040557861, | |
| "logits/rejected": -1.6431505680084229, | |
| "logps/chosen": -238.2488555908203, | |
| "logps/rejected": -307.4492492675781, | |
| "loss": 0.2182, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": -1.2250194549560547, | |
| "rewards/margins": 4.405190467834473, | |
| "rewards/rejected": -5.630209922790527, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.2154666666666667, | |
| "grad_norm": 3.474419355392456, | |
| "learning_rate": 3.3790531312703055e-06, | |
| "logits/chosen": -1.659584403038025, | |
| "logits/rejected": -1.6280295848846436, | |
| "logps/chosen": -239.4110107421875, | |
| "logps/rejected": -326.34466552734375, | |
| "loss": 0.1538, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -0.9404169917106628, | |
| "rewards/margins": 4.299707412719727, | |
| "rewards/rejected": -5.240124225616455, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.2261333333333333, | |
| "grad_norm": 4.2344183921813965, | |
| "learning_rate": 3.35204366021622e-06, | |
| "logits/chosen": -1.6510120630264282, | |
| "logits/rejected": -1.5048983097076416, | |
| "logps/chosen": -240.78897094726562, | |
| "logps/rejected": -310.3196105957031, | |
| "loss": 0.2342, | |
| "rewards/accuracies": 0.8812500238418579, | |
| "rewards/chosen": -0.8230067491531372, | |
| "rewards/margins": 3.923809051513672, | |
| "rewards/rejected": -4.7468156814575195, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.2368000000000001, | |
| "grad_norm": 3.379074811935425, | |
| "learning_rate": 3.3249211918630055e-06, | |
| "logits/chosen": -1.7805049419403076, | |
| "logits/rejected": -1.6977018117904663, | |
| "logps/chosen": -238.62423706054688, | |
| "logps/rejected": -301.8321228027344, | |
| "loss": 0.1984, | |
| "rewards/accuracies": 0.918749988079071, | |
| "rewards/chosen": -1.016635537147522, | |
| "rewards/margins": 4.349543571472168, | |
| "rewards/rejected": -5.3661789894104, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.2474666666666667, | |
| "grad_norm": 2.739449977874756, | |
| "learning_rate": 3.2976893231691424e-06, | |
| "logits/chosen": -1.6601759195327759, | |
| "logits/rejected": -1.619744062423706, | |
| "logps/chosen": -230.4942626953125, | |
| "logps/rejected": -320.5411071777344, | |
| "loss": 0.1424, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -0.6311012506484985, | |
| "rewards/margins": 4.800417900085449, | |
| "rewards/rejected": -5.431519508361816, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.2581333333333333, | |
| "grad_norm": 3.436858654022217, | |
| "learning_rate": 3.270351665601691e-06, | |
| "logits/chosen": -1.682958960533142, | |
| "logits/rejected": -1.6138877868652344, | |
| "logps/chosen": -236.56436157226562, | |
| "logps/rejected": -306.29205322265625, | |
| "loss": 0.1444, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -1.0519824028015137, | |
| "rewards/margins": 4.614463806152344, | |
| "rewards/rejected": -5.666445732116699, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.2688, | |
| "grad_norm": 1.9738092422485352, | |
| "learning_rate": 3.2429118446573417e-06, | |
| "logits/chosen": -1.6976039409637451, | |
| "logits/rejected": -1.568501353263855, | |
| "logps/chosen": -239.57424926757812, | |
| "logps/rejected": -324.35601806640625, | |
| "loss": 0.1958, | |
| "rewards/accuracies": 0.9125000238418579, | |
| "rewards/chosen": -1.3311951160430908, | |
| "rewards/margins": 4.367763042449951, | |
| "rewards/rejected": -5.698958396911621, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.2794666666666665, | |
| "grad_norm": 3.606091260910034, | |
| "learning_rate": 3.215373499381602e-06, | |
| "logits/chosen": -1.6570765972137451, | |
| "logits/rejected": -1.557324767112732, | |
| "logps/chosen": -250.1347198486328, | |
| "logps/rejected": -305.9236755371094, | |
| "loss": 0.2033, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/chosen": -1.1782742738723755, | |
| "rewards/margins": 4.263705730438232, | |
| "rewards/rejected": -5.441979885101318, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.2901333333333334, | |
| "grad_norm": 4.431084156036377, | |
| "learning_rate": 3.1877402818861954e-06, | |
| "logits/chosen": -1.783603310585022, | |
| "logits/rejected": -1.6870628595352173, | |
| "logps/chosen": -252.9153289794922, | |
| "logps/rejected": -331.8052673339844, | |
| "loss": 0.1427, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -1.4646859169006348, | |
| "rewards/margins": 4.739185333251953, | |
| "rewards/rejected": -6.203871250152588, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.3008, | |
| "grad_norm": 2.1071999073028564, | |
| "learning_rate": 3.1600158568647145e-06, | |
| "logits/chosen": -1.6523160934448242, | |
| "logits/rejected": -1.64646315574646, | |
| "logps/chosen": -246.16506958007812, | |
| "logps/rejected": -309.9844665527344, | |
| "loss": 0.2131, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -1.479095220565796, | |
| "rewards/margins": 4.29295015335083, | |
| "rewards/rejected": -5.772045135498047, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.3114666666666666, | |
| "grad_norm": 2.0734453201293945, | |
| "learning_rate": 3.1322039011066172e-06, | |
| "logits/chosen": -1.5979714393615723, | |
| "logits/rejected": -1.5535624027252197, | |
| "logps/chosen": -251.6847686767578, | |
| "logps/rejected": -300.205810546875, | |
| "loss": 0.1771, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/chosen": -1.2251965999603271, | |
| "rewards/margins": 4.506723880767822, | |
| "rewards/rejected": -5.7319207191467285, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.3221333333333334, | |
| "grad_norm": 3.522108554840088, | |
| "learning_rate": 3.1043081030096125e-06, | |
| "logits/chosen": -1.583038091659546, | |
| "logits/rejected": -1.4954283237457275, | |
| "logps/chosen": -244.6840362548828, | |
| "logps/rejected": -288.87139892578125, | |
| "loss": 0.1722, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -1.3493521213531494, | |
| "rewards/margins": 4.36538028717041, | |
| "rewards/rejected": -5.7147321701049805, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.3328, | |
| "grad_norm": 3.5037546157836914, | |
| "learning_rate": 3.0763321620905072e-06, | |
| "logits/chosen": -1.68924880027771, | |
| "logits/rejected": -1.6171550750732422, | |
| "logps/chosen": -247.78756713867188, | |
| "logps/rejected": -317.5525207519531, | |
| "loss": 0.1445, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -1.0897579193115234, | |
| "rewards/margins": 4.7230424880981445, | |
| "rewards/rejected": -5.812800407409668, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.3434666666666666, | |
| "grad_norm": 3.3278591632843018, | |
| "learning_rate": 3.048279788494583e-06, | |
| "logits/chosen": -1.7668014764785767, | |
| "logits/rejected": -1.611505150794983, | |
| "logps/chosen": -235.28744506835938, | |
| "logps/rejected": -308.88848876953125, | |
| "loss": 0.1671, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -1.3909533023834229, | |
| "rewards/margins": 4.262879848480225, | |
| "rewards/rejected": -5.653833389282227, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.3541333333333334, | |
| "grad_norm": 2.192690372467041, | |
| "learning_rate": 3.0201547025035553e-06, | |
| "logits/chosen": -1.7089593410491943, | |
| "logits/rejected": -1.5918943881988525, | |
| "logps/chosen": -247.35800170898438, | |
| "logps/rejected": -318.1089172363281, | |
| "loss": 0.1385, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -1.4722728729248047, | |
| "rewards/margins": 4.720250129699707, | |
| "rewards/rejected": -6.192523002624512, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 1.3648, | |
| "grad_norm": 7.483966827392578, | |
| "learning_rate": 2.9919606340421963e-06, | |
| "logits/chosen": -1.6352918148040771, | |
| "logits/rejected": -1.5761418342590332, | |
| "logps/chosen": -245.5006103515625, | |
| "logps/rejected": -314.25115966796875, | |
| "loss": 0.1857, | |
| "rewards/accuracies": 0.918749988079071, | |
| "rewards/chosen": -1.7750804424285889, | |
| "rewards/margins": 4.828191757202148, | |
| "rewards/rejected": -6.603272914886475, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.3754666666666666, | |
| "grad_norm": 3.6300911903381348, | |
| "learning_rate": 2.9637013221836723e-06, | |
| "logits/chosen": -1.6962321996688843, | |
| "logits/rejected": -1.5770907402038574, | |
| "logps/chosen": -268.17938232421875, | |
| "logps/rejected": -325.52020263671875, | |
| "loss": 0.2036, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/chosen": -1.6570656299591064, | |
| "rewards/margins": 4.438990592956543, | |
| "rewards/rejected": -6.0960564613342285, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.3861333333333334, | |
| "grad_norm": 5.613654613494873, | |
| "learning_rate": 2.9353805146536762e-06, | |
| "logits/chosen": -1.6913630962371826, | |
| "logits/rejected": -1.583839774131775, | |
| "logps/chosen": -252.5726776123047, | |
| "logps/rejected": -320.84478759765625, | |
| "loss": 0.1636, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -1.6463943719863892, | |
| "rewards/margins": 4.656146049499512, | |
| "rewards/rejected": -6.302539825439453, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.3968, | |
| "grad_norm": 1.9233977794647217, | |
| "learning_rate": 2.907001967333401e-06, | |
| "logits/chosen": -1.7078460454940796, | |
| "logits/rejected": -1.6775310039520264, | |
| "logps/chosen": -257.7984924316406, | |
| "logps/rejected": -340.76470947265625, | |
| "loss": 0.1728, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -1.483618140220642, | |
| "rewards/margins": 4.942142009735107, | |
| "rewards/rejected": -6.425759792327881, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.4074666666666666, | |
| "grad_norm": 3.366171360015869, | |
| "learning_rate": 2.878569443761442e-06, | |
| "logits/chosen": -1.6747850179672241, | |
| "logits/rejected": -1.621994972229004, | |
| "logps/chosen": -237.207275390625, | |
| "logps/rejected": -312.04949951171875, | |
| "loss": 0.2247, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -1.444260835647583, | |
| "rewards/margins": 4.030089378356934, | |
| "rewards/rejected": -5.474350452423096, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.4181333333333335, | |
| "grad_norm": 3.9642248153686523, | |
| "learning_rate": 2.85008671463468e-06, | |
| "logits/chosen": -1.74550461769104, | |
| "logits/rejected": -1.6358397006988525, | |
| "logps/chosen": -243.82864379882812, | |
| "logps/rejected": -310.81390380859375, | |
| "loss": 0.1717, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -1.506588101387024, | |
| "rewards/margins": 4.499510765075684, | |
| "rewards/rejected": -6.006098747253418, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.4288, | |
| "grad_norm": 2.471996307373047, | |
| "learning_rate": 2.8215575573082127e-06, | |
| "logits/chosen": -1.6209523677825928, | |
| "logits/rejected": -1.6456940174102783, | |
| "logps/chosen": -247.5177764892578, | |
| "logps/rejected": -340.74420166015625, | |
| "loss": 0.1342, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -1.7314220666885376, | |
| "rewards/margins": 4.739981651306152, | |
| "rewards/rejected": -6.471404075622559, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.4394666666666667, | |
| "grad_norm": 3.4441559314727783, | |
| "learning_rate": 2.792985755294404e-06, | |
| "logits/chosen": -1.8194706439971924, | |
| "logits/rejected": -1.696157455444336, | |
| "logps/chosen": -232.4257049560547, | |
| "logps/rejected": -337.1112365722656, | |
| "loss": 0.1154, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -1.5092469453811646, | |
| "rewards/margins": 5.188037872314453, | |
| "rewards/rejected": -6.697284698486328, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.4501333333333333, | |
| "grad_norm": 1.8599908351898193, | |
| "learning_rate": 2.764375097761122e-06, | |
| "logits/chosen": -1.6593608856201172, | |
| "logits/rejected": -1.546225905418396, | |
| "logps/chosen": -250.6788787841797, | |
| "logps/rejected": -318.9678649902344, | |
| "loss": 0.1712, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -1.2948193550109863, | |
| "rewards/margins": 4.342173099517822, | |
| "rewards/rejected": -5.636992454528809, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.4607999999999999, | |
| "grad_norm": 2.777001142501831, | |
| "learning_rate": 2.73572937902922e-06, | |
| "logits/chosen": -1.7056080102920532, | |
| "logits/rejected": -1.5910823345184326, | |
| "logps/chosen": -234.8707275390625, | |
| "logps/rejected": -323.86993408203125, | |
| "loss": 0.1672, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -1.3926390409469604, | |
| "rewards/margins": 4.474860191345215, | |
| "rewards/rejected": -5.867499351501465, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.4714666666666667, | |
| "grad_norm": 2.0958902835845947, | |
| "learning_rate": 2.7070523980693362e-06, | |
| "logits/chosen": -1.5597681999206543, | |
| "logits/rejected": -1.4919359683990479, | |
| "logps/chosen": -258.60162353515625, | |
| "logps/rejected": -311.6105651855469, | |
| "loss": 0.1274, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -1.177411675453186, | |
| "rewards/margins": 4.901284217834473, | |
| "rewards/rejected": -6.078696250915527, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.4821333333333333, | |
| "grad_norm": 4.915290832519531, | |
| "learning_rate": 2.678347957998081e-06, | |
| "logits/chosen": -1.740426778793335, | |
| "logits/rejected": -1.652065634727478, | |
| "logps/chosen": -250.73391723632812, | |
| "logps/rejected": -322.7600402832031, | |
| "loss": 0.1315, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -1.6637054681777954, | |
| "rewards/margins": 5.201625823974609, | |
| "rewards/rejected": -6.865331172943115, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.4928, | |
| "grad_norm": 1.5418274402618408, | |
| "learning_rate": 2.649619865573666e-06, | |
| "logits/chosen": -1.5983693599700928, | |
| "logits/rejected": -1.5417187213897705, | |
| "logps/chosen": -245.4449462890625, | |
| "logps/rejected": -339.23028564453125, | |
| "loss": 0.1787, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/chosen": -1.7123286724090576, | |
| "rewards/margins": 5.181328296661377, | |
| "rewards/rejected": -6.8936567306518555, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.5034666666666667, | |
| "grad_norm": 2.4393539428710938, | |
| "learning_rate": 2.620871930691058e-06, | |
| "logits/chosen": -1.6640474796295166, | |
| "logits/rejected": -1.6505119800567627, | |
| "logps/chosen": -249.7317352294922, | |
| "logps/rejected": -331.77996826171875, | |
| "loss": 0.1866, | |
| "rewards/accuracies": 0.9125000238418579, | |
| "rewards/chosen": -1.970604658126831, | |
| "rewards/margins": 4.820348262786865, | |
| "rewards/rejected": -6.790953159332275, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.5141333333333333, | |
| "grad_norm": 1.9373661279678345, | |
| "learning_rate": 2.5921079658767152e-06, | |
| "logits/chosen": -1.731371283531189, | |
| "logits/rejected": -1.6103847026824951, | |
| "logps/chosen": -265.63885498046875, | |
| "logps/rejected": -307.3009338378906, | |
| "loss": 0.15, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -1.724918007850647, | |
| "rewards/margins": 4.603318214416504, | |
| "rewards/rejected": -6.3282365798950195, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.5248, | |
| "grad_norm": 3.6780591011047363, | |
| "learning_rate": 2.56333178578297e-06, | |
| "logits/chosen": -1.6388132572174072, | |
| "logits/rejected": -1.5583611726760864, | |
| "logps/chosen": -245.09841918945312, | |
| "logps/rejected": -328.18377685546875, | |
| "loss": 0.141, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -1.9475570917129517, | |
| "rewards/margins": 5.37711238861084, | |
| "rewards/rejected": -7.32466983795166, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.5354666666666668, | |
| "grad_norm": 5.639969348907471, | |
| "learning_rate": 2.534547206682136e-06, | |
| "logits/chosen": -1.676372766494751, | |
| "logits/rejected": -1.610573172569275, | |
| "logps/chosen": -261.094482421875, | |
| "logps/rejected": -323.09112548828125, | |
| "loss": 0.1778, | |
| "rewards/accuracies": 0.918749988079071, | |
| "rewards/chosen": -1.8512113094329834, | |
| "rewards/margins": 4.54202127456665, | |
| "rewards/rejected": -6.3932318687438965, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.5461333333333334, | |
| "grad_norm": 4.020143508911133, | |
| "learning_rate": 2.5057580459603947e-06, | |
| "logits/chosen": -1.7616875171661377, | |
| "logits/rejected": -1.6894457340240479, | |
| "logps/chosen": -235.62210083007812, | |
| "logps/rejected": -308.05975341796875, | |
| "loss": 0.215, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": -1.8900903463363647, | |
| "rewards/margins": 4.641322135925293, | |
| "rewards/rejected": -6.5314130783081055, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.5568, | |
| "grad_norm": 4.262376308441162, | |
| "learning_rate": 2.4769681216115384e-06, | |
| "logits/chosen": -1.7237647771835327, | |
| "logits/rejected": -1.6486454010009766, | |
| "logps/chosen": -244.239013671875, | |
| "logps/rejected": -324.2884216308594, | |
| "loss": 0.1723, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -1.6464115381240845, | |
| "rewards/margins": 4.711733818054199, | |
| "rewards/rejected": -6.358145713806152, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.5674666666666668, | |
| "grad_norm": 1.481379508972168, | |
| "learning_rate": 2.448181251730632e-06, | |
| "logits/chosen": -1.6676714420318604, | |
| "logits/rejected": -1.584498643875122, | |
| "logps/chosen": -249.21066284179688, | |
| "logps/rejected": -323.2672119140625, | |
| "loss": 0.2, | |
| "rewards/accuracies": 0.9125000238418579, | |
| "rewards/chosen": -1.6483440399169922, | |
| "rewards/margins": 4.992440700531006, | |
| "rewards/rejected": -6.64078426361084, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.5781333333333334, | |
| "grad_norm": 3.6630077362060547, | |
| "learning_rate": 2.419401254007658e-06, | |
| "logits/chosen": -1.5993918180465698, | |
| "logits/rejected": -1.5106697082519531, | |
| "logps/chosen": -246.16738891601562, | |
| "logps/rejected": -303.5965881347656, | |
| "loss": 0.1411, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -1.581565260887146, | |
| "rewards/margins": 4.666982650756836, | |
| "rewards/rejected": -6.248547554016113, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.5888, | |
| "grad_norm": 4.977443218231201, | |
| "learning_rate": 2.390631945221221e-06, | |
| "logits/chosen": -1.6648800373077393, | |
| "logits/rejected": -1.535312294960022, | |
| "logps/chosen": -259.2499084472656, | |
| "logps/rejected": -321.27313232421875, | |
| "loss": 0.1764, | |
| "rewards/accuracies": 0.9125000238418579, | |
| "rewards/chosen": -1.6946322917938232, | |
| "rewards/margins": 4.95900297164917, | |
| "rewards/rejected": -6.653635501861572, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.5994666666666668, | |
| "grad_norm": 3.6690218448638916, | |
| "learning_rate": 2.361877140732366e-06, | |
| "logits/chosen": -1.7277066707611084, | |
| "logits/rejected": -1.6841871738433838, | |
| "logps/chosen": -224.5117950439453, | |
| "logps/rejected": -330.66082763671875, | |
| "loss": 0.1727, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -1.3350045680999756, | |
| "rewards/margins": 5.390060901641846, | |
| "rewards/rejected": -6.7250657081604, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.6101333333333332, | |
| "grad_norm": 3.617159843444824, | |
| "learning_rate": 2.333140653978592e-06, | |
| "logits/chosen": -1.7293851375579834, | |
| "logits/rejected": -1.5459930896759033, | |
| "logps/chosen": -260.37548828125, | |
| "logps/rejected": -319.6600646972656, | |
| "loss": 0.1591, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -1.4235215187072754, | |
| "rewards/margins": 4.943792343139648, | |
| "rewards/rejected": -6.367314338684082, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 1.6208, | |
| "grad_norm": 3.683688163757324, | |
| "learning_rate": 2.3044262959681156e-06, | |
| "logits/chosen": -1.6138556003570557, | |
| "logits/rejected": -1.472929835319519, | |
| "logps/chosen": -256.35125732421875, | |
| "logps/rejected": -320.8252868652344, | |
| "loss": 0.1434, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -1.7510179281234741, | |
| "rewards/margins": 5.087538242340088, | |
| "rewards/rejected": -6.838555812835693, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.6314666666666666, | |
| "grad_norm": 2.7799344062805176, | |
| "learning_rate": 2.2757378747744575e-06, | |
| "logits/chosen": -1.6599740982055664, | |
| "logits/rejected": -1.521399736404419, | |
| "logps/chosen": -232.0360870361328, | |
| "logps/rejected": -297.9733581542969, | |
| "loss": 0.1437, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -1.280116319656372, | |
| "rewards/margins": 4.649960517883301, | |
| "rewards/rejected": -5.930076599121094, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 1.6421333333333332, | |
| "grad_norm": 3.1140360832214355, | |
| "learning_rate": 2.24707919503142e-06, | |
| "logits/chosen": -1.7199656963348389, | |
| "logits/rejected": -1.620208501815796, | |
| "logps/chosen": -242.5529327392578, | |
| "logps/rejected": -318.6707763671875, | |
| "loss": 0.1247, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -1.4802429676055908, | |
| "rewards/margins": 4.666882514953613, | |
| "rewards/rejected": -6.147125244140625, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.6528, | |
| "grad_norm": 6.060303688049316, | |
| "learning_rate": 2.2184540574285213e-06, | |
| "logits/chosen": -1.6107571125030518, | |
| "logits/rejected": -1.4901559352874756, | |
| "logps/chosen": -266.9434509277344, | |
| "logps/rejected": -303.53179931640625, | |
| "loss": 0.2051, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -1.3911632299423218, | |
| "rewards/margins": 4.4641523361206055, | |
| "rewards/rejected": -5.855315208435059, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.6634666666666666, | |
| "grad_norm": 2.654046058654785, | |
| "learning_rate": 2.1898662582069486e-06, | |
| "logits/chosen": -1.6774696111679077, | |
| "logits/rejected": -1.5776320695877075, | |
| "logps/chosen": -250.61331176757812, | |
| "logps/rejected": -306.73663330078125, | |
| "loss": 0.1373, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -1.7681554555892944, | |
| "rewards/margins": 5.010857582092285, | |
| "rewards/rejected": -6.779012203216553, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.6741333333333333, | |
| "grad_norm": 3.1174325942993164, | |
| "learning_rate": 2.1613195886561044e-06, | |
| "logits/chosen": -1.7513983249664307, | |
| "logits/rejected": -1.6123911142349243, | |
| "logps/chosen": -230.7318878173828, | |
| "logps/rejected": -302.77874755859375, | |
| "loss": 0.113, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -1.284829020500183, | |
| "rewards/margins": 5.273435115814209, | |
| "rewards/rejected": -6.55826473236084, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 1.6848, | |
| "grad_norm": 5.556603908538818, | |
| "learning_rate": 2.132817834610811e-06, | |
| "logits/chosen": -1.6932178735733032, | |
| "logits/rejected": -1.552677869796753, | |
| "logps/chosen": -241.80630493164062, | |
| "logps/rejected": -315.0558776855469, | |
| "loss": 0.1641, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -1.5892698764801025, | |
| "rewards/margins": 4.98702335357666, | |
| "rewards/rejected": -6.576292991638184, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.6954666666666667, | |
| "grad_norm": 2.940763235092163, | |
| "learning_rate": 2.1043647759492338e-06, | |
| "logits/chosen": -1.5830045938491821, | |
| "logits/rejected": -1.4869303703308105, | |
| "logps/chosen": -264.3173828125, | |
| "logps/rejected": -319.78924560546875, | |
| "loss": 0.1699, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -1.4739254713058472, | |
| "rewards/margins": 4.805297374725342, | |
| "rewards/rejected": -6.279223442077637, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 1.7061333333333333, | |
| "grad_norm": 4.358272075653076, | |
| "learning_rate": 2.075964186091599e-06, | |
| "logits/chosen": -1.7524511814117432, | |
| "logits/rejected": -1.6442880630493164, | |
| "logps/chosen": -233.0540008544922, | |
| "logps/rejected": -352.62567138671875, | |
| "loss": 0.1132, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -1.8809230327606201, | |
| "rewards/margins": 5.72424840927124, | |
| "rewards/rejected": -7.605171203613281, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.7168, | |
| "grad_norm": 3.5285327434539795, | |
| "learning_rate": 2.047619831499765e-06, | |
| "logits/chosen": -1.5862383842468262, | |
| "logits/rejected": -1.5314266681671143, | |
| "logps/chosen": -239.5761260986328, | |
| "logps/rejected": -314.5472106933594, | |
| "loss": 0.1659, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -1.4697277545928955, | |
| "rewards/margins": 4.891806125640869, | |
| "rewards/rejected": -6.3615336418151855, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 1.7274666666666667, | |
| "grad_norm": 3.361513614654541, | |
| "learning_rate": 2.0193354711777155e-06, | |
| "logits/chosen": -1.7188619375228882, | |
| "logits/rejected": -1.5833302736282349, | |
| "logps/chosen": -264.5706481933594, | |
| "logps/rejected": -334.4622497558594, | |
| "loss": 0.1704, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -1.8755519390106201, | |
| "rewards/margins": 5.095242500305176, | |
| "rewards/rejected": -6.970794677734375, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.7381333333333333, | |
| "grad_norm": 2.792012929916382, | |
| "learning_rate": 1.991114856173048e-06, | |
| "logits/chosen": -1.6884195804595947, | |
| "logits/rejected": -1.6442887783050537, | |
| "logps/chosen": -230.58377075195312, | |
| "logps/rejected": -339.1591796875, | |
| "loss": 0.171, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": -1.912938117980957, | |
| "rewards/margins": 5.23507022857666, | |
| "rewards/rejected": -7.148007869720459, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 1.7488000000000001, | |
| "grad_norm": 4.240962505340576, | |
| "learning_rate": 1.9629617290795115e-06, | |
| "logits/chosen": -1.5643692016601562, | |
| "logits/rejected": -1.526997685432434, | |
| "logps/chosen": -248.695556640625, | |
| "logps/rejected": -323.1645202636719, | |
| "loss": 0.1728, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": -1.9166147708892822, | |
| "rewards/margins": 4.900598049163818, | |
| "rewards/rejected": -6.817213535308838, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.7594666666666665, | |
| "grad_norm": 1.9399546384811401, | |
| "learning_rate": 1.934879823540663e-06, | |
| "logits/chosen": -1.6955827474594116, | |
| "logits/rejected": -1.568052053451538, | |
| "logps/chosen": -247.80142211914062, | |
| "logps/rejected": -331.73883056640625, | |
| "loss": 0.2002, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -2.013258934020996, | |
| "rewards/margins": 5.11118745803833, | |
| "rewards/rejected": -7.124446868896484, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.7701333333333333, | |
| "grad_norm": 2.1034741401672363, | |
| "learning_rate": 1.9068728637547195e-06, | |
| "logits/chosen": -1.6912727355957031, | |
| "logits/rejected": -1.579400897026062, | |
| "logps/chosen": -254.18777465820312, | |
| "logps/rejected": -339.73016357421875, | |
| "loss": 0.1794, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -1.799628496170044, | |
| "rewards/margins": 5.339221000671387, | |
| "rewards/rejected": -7.138850212097168, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.7808000000000002, | |
| "grad_norm": 2.8324687480926514, | |
| "learning_rate": 1.878944563980658e-06, | |
| "logits/chosen": -1.681926965713501, | |
| "logits/rejected": -1.5870113372802734, | |
| "logps/chosen": -248.7440948486328, | |
| "logps/rejected": -322.190673828125, | |
| "loss": 0.1347, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -1.6237030029296875, | |
| "rewards/margins": 5.433587551116943, | |
| "rewards/rejected": -7.057290554046631, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 1.7914666666666665, | |
| "grad_norm": 2.730384111404419, | |
| "learning_rate": 1.85109862804563e-06, | |
| "logits/chosen": -1.6680930852890015, | |
| "logits/rejected": -1.6028831005096436, | |
| "logps/chosen": -299.27667236328125, | |
| "logps/rejected": -340.4490966796875, | |
| "loss": 0.1411, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -1.330984354019165, | |
| "rewards/margins": 5.942017555236816, | |
| "rewards/rejected": -7.273001194000244, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.8021333333333334, | |
| "grad_norm": 3.7725889682769775, | |
| "learning_rate": 1.823338748853766e-06, | |
| "logits/chosen": -1.6155637502670288, | |
| "logits/rejected": -1.6243375539779663, | |
| "logps/chosen": -263.1604919433594, | |
| "logps/rejected": -347.5821228027344, | |
| "loss": 0.1621, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -2.142850399017334, | |
| "rewards/margins": 5.3936767578125, | |
| "rewards/rejected": -7.536526679992676, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 1.8128, | |
| "grad_norm": 5.143514633178711, | |
| "learning_rate": 1.7956686078964257e-06, | |
| "logits/chosen": -1.7049986124038696, | |
| "logits/rejected": -1.6571338176727295, | |
| "logps/chosen": -238.5912322998047, | |
| "logps/rejected": -330.8258972167969, | |
| "loss": 0.2134, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -1.6940962076187134, | |
| "rewards/margins": 4.851913928985596, | |
| "rewards/rejected": -6.5460100173950195, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.8234666666666666, | |
| "grad_norm": 4.174952030181885, | |
| "learning_rate": 1.7680918747639608e-06, | |
| "logits/chosen": -1.7732454538345337, | |
| "logits/rejected": -1.6001640558242798, | |
| "logps/chosen": -235.79336547851562, | |
| "logps/rejected": -340.64886474609375, | |
| "loss": 0.1289, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -1.5047314167022705, | |
| "rewards/margins": 5.615990161895752, | |
| "rewards/rejected": -7.120721340179443, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 1.8341333333333334, | |
| "grad_norm": 2.559386968612671, | |
| "learning_rate": 1.740612206659057e-06, | |
| "logits/chosen": -1.6925417184829712, | |
| "logits/rejected": -1.5877892971038818, | |
| "logps/chosen": -252.12435913085938, | |
| "logps/rejected": -337.1178894042969, | |
| "loss": 0.1112, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -1.6748367547988892, | |
| "rewards/margins": 5.5472025871276855, | |
| "rewards/rejected": -7.222039222717285, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.8448, | |
| "grad_norm": 5.024665832519531, | |
| "learning_rate": 1.7132332479117204e-06, | |
| "logits/chosen": -1.757213830947876, | |
| "logits/rejected": -1.6775776147842407, | |
| "logps/chosen": -249.8083953857422, | |
| "logps/rejected": -341.9578857421875, | |
| "loss": 0.1382, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -1.523437738418579, | |
| "rewards/margins": 5.880698204040527, | |
| "rewards/rejected": -7.404136657714844, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 1.8554666666666666, | |
| "grad_norm": 1.7119357585906982, | |
| "learning_rate": 1.6859586294959666e-06, | |
| "logits/chosen": -1.7067601680755615, | |
| "logits/rejected": -1.6395823955535889, | |
| "logps/chosen": -261.8993835449219, | |
| "logps/rejected": -340.2198791503906, | |
| "loss": 0.1338, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -1.6026289463043213, | |
| "rewards/margins": 5.703084468841553, | |
| "rewards/rejected": -7.305713653564453, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.8661333333333334, | |
| "grad_norm": 2.4370765686035156, | |
| "learning_rate": 1.6587919685482884e-06, | |
| "logits/chosen": -1.715688705444336, | |
| "logits/rejected": -1.5712188482284546, | |
| "logps/chosen": -242.3165740966797, | |
| "logps/rejected": -318.1778869628906, | |
| "loss": 0.128, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -1.9740794897079468, | |
| "rewards/margins": 5.140053749084473, | |
| "rewards/rejected": -7.114132881164551, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.8768, | |
| "grad_norm": 5.989840984344482, | |
| "learning_rate": 1.6317368678879497e-06, | |
| "logits/chosen": -1.6727215051651, | |
| "logits/rejected": -1.5320321321487427, | |
| "logps/chosen": -242.89840698242188, | |
| "logps/rejected": -302.96954345703125, | |
| "loss": 0.1445, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -1.4183199405670166, | |
| "rewards/margins": 4.921168327331543, | |
| "rewards/rejected": -6.3394880294799805, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.8874666666666666, | |
| "grad_norm": 3.716749429702759, | |
| "learning_rate": 1.6047969155391856e-06, | |
| "logits/chosen": -1.554337501525879, | |
| "logits/rejected": -1.4884510040283203, | |
| "logps/chosen": -256.3968200683594, | |
| "logps/rejected": -308.56573486328125, | |
| "loss": 0.1446, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -1.6531686782836914, | |
| "rewards/margins": 5.160852432250977, | |
| "rewards/rejected": -6.814022064208984, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 1.8981333333333335, | |
| "grad_norm": 3.8302741050720215, | |
| "learning_rate": 1.5779756842553617e-06, | |
| "logits/chosen": -1.6652368307113647, | |
| "logits/rejected": -1.6241079568862915, | |
| "logps/chosen": -263.47515869140625, | |
| "logps/rejected": -331.0129089355469, | |
| "loss": 0.1351, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -2.0062878131866455, | |
| "rewards/margins": 5.339743614196777, | |
| "rewards/rejected": -7.34603214263916, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.9088, | |
| "grad_norm": 7.920835971832275, | |
| "learning_rate": 1.551276731045157e-06, | |
| "logits/chosen": -1.5562691688537598, | |
| "logits/rejected": -1.5247366428375244, | |
| "logps/chosen": -275.4099426269531, | |
| "logps/rejected": -347.5796203613281, | |
| "loss": 0.1669, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -1.7762320041656494, | |
| "rewards/margins": 5.540648460388184, | |
| "rewards/rejected": -7.316880702972412, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 1.9194666666666667, | |
| "grad_norm": 2.614849805831909, | |
| "learning_rate": 1.5247035967008369e-06, | |
| "logits/chosen": -1.5560858249664307, | |
| "logits/rejected": -1.503138542175293, | |
| "logps/chosen": -237.936767578125, | |
| "logps/rejected": -330.28277587890625, | |
| "loss": 0.152, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -1.6794102191925049, | |
| "rewards/margins": 5.538970947265625, | |
| "rewards/rejected": -7.218381404876709, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.9301333333333335, | |
| "grad_norm": 3.9147305488586426, | |
| "learning_rate": 1.498259805328678e-06, | |
| "logits/chosen": -1.6482226848602295, | |
| "logits/rejected": -1.59519362449646, | |
| "logps/chosen": -239.3106689453125, | |
| "logps/rejected": -328.4599609375, | |
| "loss": 0.1567, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -2.219871997833252, | |
| "rewards/margins": 5.234225273132324, | |
| "rewards/rejected": -7.454097747802734, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 1.9407999999999999, | |
| "grad_norm": 4.330874443054199, | |
| "learning_rate": 1.471948863881601e-06, | |
| "logits/chosen": -1.7839887142181396, | |
| "logits/rejected": -1.594058871269226, | |
| "logps/chosen": -255.79690551757812, | |
| "logps/rejected": -316.6875, | |
| "loss": 0.1776, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": -1.6719051599502563, | |
| "rewards/margins": 5.126084327697754, | |
| "rewards/rejected": -6.797989845275879, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.9514666666666667, | |
| "grad_norm": 3.5629780292510986, | |
| "learning_rate": 1.4457742616940864e-06, | |
| "logits/chosen": -1.6401689052581787, | |
| "logits/rejected": -1.5712366104125977, | |
| "logps/chosen": -250.49868774414062, | |
| "logps/rejected": -326.28741455078125, | |
| "loss": 0.1683, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -1.7338117361068726, | |
| "rewards/margins": 5.273244380950928, | |
| "rewards/rejected": -7.00705623626709, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 1.9621333333333333, | |
| "grad_norm": 3.6759705543518066, | |
| "learning_rate": 1.4197394700194165e-06, | |
| "logits/chosen": -1.6852229833602905, | |
| "logits/rejected": -1.6576778888702393, | |
| "logps/chosen": -249.3162841796875, | |
| "logps/rejected": -325.59326171875, | |
| "loss": 0.1717, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": -1.5520923137664795, | |
| "rewards/margins": 5.2563347816467285, | |
| "rewards/rejected": -6.808426856994629, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.9727999999999999, | |
| "grad_norm": 3.09043288230896, | |
| "learning_rate": 1.3938479415693244e-06, | |
| "logits/chosen": -1.6909940242767334, | |
| "logits/rejected": -1.6325387954711914, | |
| "logps/chosen": -250.38870239257812, | |
| "logps/rejected": -348.135986328125, | |
| "loss": 0.109, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -1.6463645696640015, | |
| "rewards/margins": 5.957003593444824, | |
| "rewards/rejected": -7.603367805480957, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.9834666666666667, | |
| "grad_norm": 3.553077220916748, | |
| "learning_rate": 1.368103110056096e-06, | |
| "logits/chosen": -1.7662723064422607, | |
| "logits/rejected": -1.6744906902313232, | |
| "logps/chosen": -252.43728637695312, | |
| "logps/rejected": -321.92413330078125, | |
| "loss": 0.1539, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -1.9222122430801392, | |
| "rewards/margins": 5.75597620010376, | |
| "rewards/rejected": -7.678188323974609, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.9941333333333333, | |
| "grad_norm": 4.891976356506348, | |
| "learning_rate": 1.3425083897371983e-06, | |
| "logits/chosen": -1.6703214645385742, | |
| "logits/rejected": -1.5516163110733032, | |
| "logps/chosen": -262.6437683105469, | |
| "logps/rejected": -329.6023254394531, | |
| "loss": 0.1595, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -1.671091079711914, | |
| "rewards/margins": 5.250547409057617, | |
| "rewards/rejected": -6.921639442443848, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 2.0042666666666666, | |
| "grad_norm": 3.132736921310425, | |
| "learning_rate": 1.317067174962478e-06, | |
| "logits/chosen": -1.6124461889266968, | |
| "logits/rejected": -1.465229868888855, | |
| "logps/chosen": -258.7012939453125, | |
| "logps/rejected": -305.882568359375, | |
| "loss": 0.176, | |
| "rewards/accuracies": 0.9539473652839661, | |
| "rewards/chosen": -2.025094985961914, | |
| "rewards/margins": 5.082742214202881, | |
| "rewards/rejected": -7.107837677001953, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 2.0149333333333335, | |
| "grad_norm": 3.199899435043335, | |
| "learning_rate": 1.291782839724009e-06, | |
| "logits/chosen": -1.635519027709961, | |
| "logits/rejected": -1.5712947845458984, | |
| "logps/chosen": -258.9504699707031, | |
| "logps/rejected": -322.48211669921875, | |
| "loss": 0.1754, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -1.844058632850647, | |
| "rewards/margins": 4.864871978759766, | |
| "rewards/rejected": -6.708930969238281, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 2.0256, | |
| "grad_norm": 3.8576583862304688, | |
| "learning_rate": 1.2666587372086352e-06, | |
| "logits/chosen": -1.575919508934021, | |
| "logits/rejected": -1.6220824718475342, | |
| "logps/chosen": -236.1714324951172, | |
| "logps/rejected": -340.546875, | |
| "loss": 0.1122, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -1.6586267948150635, | |
| "rewards/margins": 5.464929580688477, | |
| "rewards/rejected": -7.123556613922119, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.0362666666666667, | |
| "grad_norm": 3.3838019371032715, | |
| "learning_rate": 1.241698199353275e-06, | |
| "logits/chosen": -1.6258891820907593, | |
| "logits/rejected": -1.5644147396087646, | |
| "logps/chosen": -243.3029022216797, | |
| "logps/rejected": -324.55096435546875, | |
| "loss": 0.145, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -1.659441590309143, | |
| "rewards/margins": 4.9935383796691895, | |
| "rewards/rejected": -6.652979850769043, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 2.0469333333333335, | |
| "grad_norm": 4.255578994750977, | |
| "learning_rate": 1.2169045364030404e-06, | |
| "logits/chosen": -1.7650766372680664, | |
| "logits/rejected": -1.6293489933013916, | |
| "logps/chosen": -236.46322631835938, | |
| "logps/rejected": -319.8513488769531, | |
| "loss": 0.1216, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -1.9534380435943604, | |
| "rewards/margins": 5.467280864715576, | |
| "rewards/rejected": -7.420719146728516, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 2.0576, | |
| "grad_norm": 6.157327651977539, | |
| "learning_rate": 1.1922810364722402e-06, | |
| "logits/chosen": -1.645727515220642, | |
| "logits/rejected": -1.5266873836517334, | |
| "logps/chosen": -258.87347412109375, | |
| "logps/rejected": -314.80267333984375, | |
| "loss": 0.1738, | |
| "rewards/accuracies": 0.9125000238418579, | |
| "rewards/chosen": -1.7486776113510132, | |
| "rewards/margins": 4.593234539031982, | |
| "rewards/rejected": -6.341912269592285, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 2.0682666666666667, | |
| "grad_norm": 3.139277696609497, | |
| "learning_rate": 1.1678309651083055e-06, | |
| "logits/chosen": -1.7026522159576416, | |
| "logits/rejected": -1.6696736812591553, | |
| "logps/chosen": -236.3257293701172, | |
| "logps/rejected": -341.9847412109375, | |
| "loss": 0.1023, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -1.863978624343872, | |
| "rewards/margins": 5.8271098136901855, | |
| "rewards/rejected": -7.691088676452637, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 2.0789333333333335, | |
| "grad_norm": 1.7706592082977295, | |
| "learning_rate": 1.1435575648587224e-06, | |
| "logits/chosen": -1.6053731441497803, | |
| "logits/rejected": -1.5967332124710083, | |
| "logps/chosen": -249.1739501953125, | |
| "logps/rejected": -336.7267761230469, | |
| "loss": 0.1229, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -1.9126554727554321, | |
| "rewards/margins": 5.3537068367004395, | |
| "rewards/rejected": -7.266362190246582, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 2.0896, | |
| "grad_norm": 3.7354724407196045, | |
| "learning_rate": 1.1194640548410039e-06, | |
| "logits/chosen": -1.5527129173278809, | |
| "logits/rejected": -1.512411117553711, | |
| "logps/chosen": -259.1437683105469, | |
| "logps/rejected": -303.2730712890625, | |
| "loss": 0.1752, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -1.8620319366455078, | |
| "rewards/margins": 5.097782135009766, | |
| "rewards/rejected": -6.959813594818115, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 2.1002666666666667, | |
| "grad_norm": 1.9574644565582275, | |
| "learning_rate": 1.095553630315774e-06, | |
| "logits/chosen": -1.608660101890564, | |
| "logits/rejected": -1.5342473983764648, | |
| "logps/chosen": -256.6988220214844, | |
| "logps/rejected": -332.3222351074219, | |
| "loss": 0.1137, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -1.5717804431915283, | |
| "rewards/margins": 5.8155951499938965, | |
| "rewards/rejected": -7.387376308441162, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 2.1109333333333336, | |
| "grad_norm": 2.388463258743286, | |
| "learning_rate": 1.0718294622630188e-06, | |
| "logits/chosen": -1.7099485397338867, | |
| "logits/rejected": -1.5633583068847656, | |
| "logps/chosen": -273.2356262207031, | |
| "logps/rejected": -345.201171875, | |
| "loss": 0.1278, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -1.8504549264907837, | |
| "rewards/margins": 5.3531904220581055, | |
| "rewards/rejected": -7.2036452293396, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 2.1216, | |
| "grad_norm": 3.036656618118286, | |
| "learning_rate": 1.0482946969615495e-06, | |
| "logits/chosen": -1.7514898777008057, | |
| "logits/rejected": -1.6838487386703491, | |
| "logps/chosen": -233.05953979492188, | |
| "logps/rejected": -331.9215393066406, | |
| "loss": 0.0959, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -2.013244867324829, | |
| "rewards/margins": 5.820936679840088, | |
| "rewards/rejected": -7.834181785583496, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 2.1322666666666668, | |
| "grad_norm": 8.45803451538086, | |
| "learning_rate": 1.0249524555717461e-06, | |
| "logits/chosen": -1.6929506063461304, | |
| "logits/rejected": -1.5387563705444336, | |
| "logps/chosen": -242.8338623046875, | |
| "logps/rejected": -329.45965576171875, | |
| "loss": 0.1429, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -1.9728199243545532, | |
| "rewards/margins": 5.124451637268066, | |
| "rewards/rejected": -7.097271919250488, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.142933333333333, | |
| "grad_norm": 1.7584189176559448, | |
| "learning_rate": 1.0018058337216327e-06, | |
| "logits/chosen": -1.6556212902069092, | |
| "logits/rejected": -1.6083987951278687, | |
| "logps/chosen": -238.981689453125, | |
| "logps/rejected": -326.83062744140625, | |
| "loss": 0.1182, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -1.6057265996932983, | |
| "rewards/margins": 5.429762363433838, | |
| "rewards/rejected": -7.035489559173584, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 2.1536, | |
| "grad_norm": 4.888096809387207, | |
| "learning_rate": 9.788579010963387e-07, | |
| "logits/chosen": -1.6201951503753662, | |
| "logits/rejected": -1.5311352014541626, | |
| "logps/chosen": -267.74908447265625, | |
| "logps/rejected": -334.7906494140625, | |
| "loss": 0.1455, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -2.1402487754821777, | |
| "rewards/margins": 5.422475337982178, | |
| "rewards/rejected": -7.5627241134643555, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 2.164266666666667, | |
| "grad_norm": 2.1864840984344482, | |
| "learning_rate": 9.56111701030998e-07, | |
| "logits/chosen": -1.7070671319961548, | |
| "logits/rejected": -1.558129906654358, | |
| "logps/chosen": -247.9465789794922, | |
| "logps/rejected": -343.86029052734375, | |
| "loss": 0.1001, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -1.7870800495147705, | |
| "rewards/margins": 6.4040422439575195, | |
| "rewards/rejected": -8.191122055053711, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 2.174933333333333, | |
| "grad_norm": 3.70732045173645, | |
| "learning_rate": 9.335702501071459e-07, | |
| "logits/chosen": -1.6212594509124756, | |
| "logits/rejected": -1.4471334218978882, | |
| "logps/chosen": -265.968994140625, | |
| "logps/rejected": -326.27630615234375, | |
| "loss": 0.1046, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -1.8242906332015991, | |
| "rewards/margins": 5.466552257537842, | |
| "rewards/rejected": -7.2908430099487305, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 2.1856, | |
| "grad_norm": 4.0817084312438965, | |
| "learning_rate": 9.112365377526636e-07, | |
| "logits/chosen": -1.6601269245147705, | |
| "logits/rejected": -1.6235460042953491, | |
| "logps/chosen": -234.23635864257812, | |
| "logps/rejected": -316.6658630371094, | |
| "loss": 0.1544, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -2.0195419788360596, | |
| "rewards/margins": 5.3594865798950195, | |
| "rewards/rejected": -7.379029273986816, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 2.196266666666667, | |
| "grad_norm": 1.6365838050842285, | |
| "learning_rate": 8.891135258453195e-07, | |
| "logits/chosen": -1.603167176246643, | |
| "logits/rejected": -1.5676203966140747, | |
| "logps/chosen": -243.29550170898438, | |
| "logps/rejected": -324.26318359375, | |
| "loss": 0.1142, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -2.0350539684295654, | |
| "rewards/margins": 5.261748313903809, | |
| "rewards/rejected": -7.2968034744262695, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 2.206933333333333, | |
| "grad_norm": 3.2687931060791016, | |
| "learning_rate": 8.672041483199709e-07, | |
| "logits/chosen": -1.7075221538543701, | |
| "logits/rejected": -1.62152099609375, | |
| "logps/chosen": -260.42828369140625, | |
| "logps/rejected": -325.1893310546875, | |
| "loss": 0.1046, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -1.961230993270874, | |
| "rewards/margins": 5.438101291656494, | |
| "rewards/rejected": -7.399331569671631, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 2.2176, | |
| "grad_norm": 4.33626651763916, | |
| "learning_rate": 8.455113107794652e-07, | |
| "logits/chosen": -1.5977696180343628, | |
| "logits/rejected": -1.5143213272094727, | |
| "logps/chosen": -243.13815307617188, | |
| "logps/rejected": -315.7250061035156, | |
| "loss": 0.1049, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -2.2392735481262207, | |
| "rewards/margins": 5.705511093139648, | |
| "rewards/rejected": -7.944784641265869, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 2.228266666666667, | |
| "grad_norm": 3.568450927734375, | |
| "learning_rate": 8.240378901093035e-07, | |
| "logits/chosen": -1.7558826208114624, | |
| "logits/rejected": -1.6090519428253174, | |
| "logps/chosen": -253.54891967773438, | |
| "logps/rejected": -336.08343505859375, | |
| "loss": 0.1332, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -1.987137794494629, | |
| "rewards/margins": 5.502018928527832, | |
| "rewards/rejected": -7.489157199859619, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 2.238933333333333, | |
| "grad_norm": 2.2904350757598877, | |
| "learning_rate": 8.027867340961113e-07, | |
| "logits/chosen": -1.613086462020874, | |
| "logits/rejected": -1.5385499000549316, | |
| "logps/chosen": -251.39895629882812, | |
| "logps/rejected": -336.3780212402344, | |
| "loss": 0.1591, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -2.0407490730285645, | |
| "rewards/margins": 5.212506294250488, | |
| "rewards/rejected": -7.253255367279053, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.2496, | |
| "grad_norm": 3.7621541023254395, | |
| "learning_rate": 7.817606610499639e-07, | |
| "logits/chosen": -1.7154810428619385, | |
| "logits/rejected": -1.6074755191802979, | |
| "logps/chosen": -250.9788360595703, | |
| "logps/rejected": -348.1394958496094, | |
| "loss": 0.1059, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -1.8897058963775635, | |
| "rewards/margins": 6.101401329040527, | |
| "rewards/rejected": -7.991107940673828, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 2.2602666666666664, | |
| "grad_norm": 2.9305503368377686, | |
| "learning_rate": 7.609624594306278e-07, | |
| "logits/chosen": -1.6567564010620117, | |
| "logits/rejected": -1.6048027276992798, | |
| "logps/chosen": -247.90121459960938, | |
| "logps/rejected": -337.5819396972656, | |
| "loss": 0.1433, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -1.9860131740570068, | |
| "rewards/margins": 6.067047119140625, | |
| "rewards/rejected": -8.053060531616211, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 2.2709333333333332, | |
| "grad_norm": 4.816735744476318, | |
| "learning_rate": 7.403948874777556e-07, | |
| "logits/chosen": -1.6866226196289062, | |
| "logits/rejected": -1.6687873601913452, | |
| "logps/chosen": -259.2507019042969, | |
| "logps/rejected": -353.58197021484375, | |
| "loss": 0.1308, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -1.9822838306427002, | |
| "rewards/margins": 5.897215843200684, | |
| "rewards/rejected": -7.879499912261963, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 2.2816, | |
| "grad_norm": 5.037070274353027, | |
| "learning_rate": 7.20060672845091e-07, | |
| "logits/chosen": -1.5634400844573975, | |
| "logits/rejected": -1.5417492389678955, | |
| "logps/chosen": -253.765869140625, | |
| "logps/rejected": -333.6149597167969, | |
| "loss": 0.1106, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -2.216799259185791, | |
| "rewards/margins": 5.854150295257568, | |
| "rewards/rejected": -8.070950508117676, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 2.2922666666666665, | |
| "grad_norm": 3.694530487060547, | |
| "learning_rate": 6.99962512238733e-07, | |
| "logits/chosen": -1.507551908493042, | |
| "logits/rejected": -1.4376182556152344, | |
| "logps/chosen": -251.04052734375, | |
| "logps/rejected": -307.02593994140625, | |
| "loss": 0.1729, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/chosen": -1.9417812824249268, | |
| "rewards/margins": 5.363430023193359, | |
| "rewards/rejected": -7.305211067199707, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 2.3029333333333333, | |
| "grad_norm": 1.9394572973251343, | |
| "learning_rate": 6.801030710594949e-07, | |
| "logits/chosen": -1.5492713451385498, | |
| "logits/rejected": -1.4976288080215454, | |
| "logps/chosen": -252.68612670898438, | |
| "logps/rejected": -325.3870544433594, | |
| "loss": 0.171, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -2.119556188583374, | |
| "rewards/margins": 5.055575847625732, | |
| "rewards/rejected": -7.175132751464844, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 2.3136, | |
| "grad_norm": 2.8549582958221436, | |
| "learning_rate": 6.604849830494264e-07, | |
| "logits/chosen": -1.7532446384429932, | |
| "logits/rejected": -1.6480261087417603, | |
| "logps/chosen": -248.76596069335938, | |
| "logps/rejected": -333.56781005859375, | |
| "loss": 0.1086, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -1.9362081289291382, | |
| "rewards/margins": 6.194217205047607, | |
| "rewards/rejected": -8.130426406860352, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 2.3242666666666665, | |
| "grad_norm": 2.9997100830078125, | |
| "learning_rate": 6.411108499425253e-07, | |
| "logits/chosen": -1.733490228652954, | |
| "logits/rejected": -1.5960662364959717, | |
| "logps/chosen": -264.21929931640625, | |
| "logps/rejected": -333.2695007324219, | |
| "loss": 0.1262, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -1.9376022815704346, | |
| "rewards/margins": 5.7414021492004395, | |
| "rewards/rejected": -7.6790056228637695, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 2.3349333333333333, | |
| "grad_norm": 3.0645434856414795, | |
| "learning_rate": 6.219832411197013e-07, | |
| "logits/chosen": -1.642669677734375, | |
| "logits/rejected": -1.4960535764694214, | |
| "logps/chosen": -248.64218139648438, | |
| "logps/rejected": -323.8027038574219, | |
| "loss": 0.1434, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -1.804839849472046, | |
| "rewards/margins": 5.383707523345947, | |
| "rewards/rejected": -7.188547611236572, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 2.3456, | |
| "grad_norm": 4.203052043914795, | |
| "learning_rate": 6.031046932680229e-07, | |
| "logits/chosen": -1.6723241806030273, | |
| "logits/rejected": -1.5729491710662842, | |
| "logps/chosen": -256.49395751953125, | |
| "logps/rejected": -353.57769775390625, | |
| "loss": 0.1144, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -1.992719292640686, | |
| "rewards/margins": 5.605996608734131, | |
| "rewards/rejected": -7.598715782165527, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.3562666666666665, | |
| "grad_norm": 3.4321813583374023, | |
| "learning_rate": 5.844777100443091e-07, | |
| "logits/chosen": -1.6482032537460327, | |
| "logits/rejected": -1.5842411518096924, | |
| "logps/chosen": -240.6206817626953, | |
| "logps/rejected": -331.48101806640625, | |
| "loss": 0.1226, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -1.8029711246490479, | |
| "rewards/margins": 5.810032367706299, | |
| "rewards/rejected": -7.613003730773926, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 2.3669333333333333, | |
| "grad_norm": 4.681034564971924, | |
| "learning_rate": 5.66104761743092e-07, | |
| "logits/chosen": -1.7696841955184937, | |
| "logits/rejected": -1.630789041519165, | |
| "logps/chosen": -255.98580932617188, | |
| "logps/rejected": -314.320556640625, | |
| "loss": 0.1541, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -2.265458583831787, | |
| "rewards/margins": 5.304887294769287, | |
| "rewards/rejected": -7.570345401763916, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 2.3776, | |
| "grad_norm": 2.437649726867676, | |
| "learning_rate": 5.479882849690102e-07, | |
| "logits/chosen": -1.6937763690948486, | |
| "logits/rejected": -1.6039565801620483, | |
| "logps/chosen": -248.5046844482422, | |
| "logps/rejected": -330.6466064453125, | |
| "loss": 0.1357, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -1.949684500694275, | |
| "rewards/margins": 5.519129276275635, | |
| "rewards/rejected": -7.468813896179199, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 2.3882666666666665, | |
| "grad_norm": 1.7076597213745117, | |
| "learning_rate": 5.301306823136687e-07, | |
| "logits/chosen": -1.5468543767929077, | |
| "logits/rejected": -1.4817711114883423, | |
| "logps/chosen": -242.590576171875, | |
| "logps/rejected": -331.3986511230469, | |
| "loss": 0.1559, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -1.899458885192871, | |
| "rewards/margins": 5.719172477722168, | |
| "rewards/rejected": -7.6186323165893555, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 2.3989333333333334, | |
| "grad_norm": 3.783508539199829, | |
| "learning_rate": 5.125343220370088e-07, | |
| "logits/chosen": -1.5996062755584717, | |
| "logits/rejected": -1.5361487865447998, | |
| "logps/chosen": -247.81689453125, | |
| "logps/rejected": -312.226806640625, | |
| "loss": 0.1241, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -1.627031683921814, | |
| "rewards/margins": 5.376161098480225, | |
| "rewards/rejected": -7.003192901611328, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 2.4096, | |
| "grad_norm": 3.6416852474212646, | |
| "learning_rate": 4.952015377532329e-07, | |
| "logits/chosen": -1.8114979267120361, | |
| "logits/rejected": -1.6875921487808228, | |
| "logps/chosen": -250.836669921875, | |
| "logps/rejected": -349.7911071777344, | |
| "loss": 0.1159, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -2.255504608154297, | |
| "rewards/margins": 5.732514381408691, | |
| "rewards/rejected": -7.9880194664001465, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 2.4202666666666666, | |
| "grad_norm": 3.7079567909240723, | |
| "learning_rate": 4.781346281213204e-07, | |
| "logits/chosen": -1.683234453201294, | |
| "logits/rejected": -1.6827503442764282, | |
| "logps/chosen": -234.7036895751953, | |
| "logps/rejected": -349.72784423828125, | |
| "loss": 0.1055, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -2.1054515838623047, | |
| "rewards/margins": 6.153875827789307, | |
| "rewards/rejected": -8.259326934814453, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 2.4309333333333334, | |
| "grad_norm": 3.7606658935546875, | |
| "learning_rate": 4.613358565401818e-07, | |
| "logits/chosen": -1.6943823099136353, | |
| "logits/rejected": -1.5724833011627197, | |
| "logps/chosen": -265.177734375, | |
| "logps/rejected": -340.174072265625, | |
| "loss": 0.1427, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -2.0506272315979004, | |
| "rewards/margins": 5.461020469665527, | |
| "rewards/rejected": -7.511648654937744, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 2.4416, | |
| "grad_norm": 3.089946985244751, | |
| "learning_rate": 4.448074508484906e-07, | |
| "logits/chosen": -1.686182975769043, | |
| "logits/rejected": -1.5479154586791992, | |
| "logps/chosen": -263.35321044921875, | |
| "logps/rejected": -339.89508056640625, | |
| "loss": 0.1014, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -1.8184881210327148, | |
| "rewards/margins": 5.920653820037842, | |
| "rewards/rejected": -7.739141941070557, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 2.4522666666666666, | |
| "grad_norm": 2.193614959716797, | |
| "learning_rate": 4.2855160302922777e-07, | |
| "logits/chosen": -1.6815516948699951, | |
| "logits/rejected": -1.5796687602996826, | |
| "logps/chosen": -254.398193359375, | |
| "logps/rejected": -342.8807678222656, | |
| "loss": 0.0996, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -2.013457775115967, | |
| "rewards/margins": 6.017300605773926, | |
| "rewards/rejected": -8.03075885772705, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.4629333333333334, | |
| "grad_norm": 6.938134670257568, | |
| "learning_rate": 4.125704689189819e-07, | |
| "logits/chosen": -1.5728188753128052, | |
| "logits/rejected": -1.5077226161956787, | |
| "logps/chosen": -253.8098602294922, | |
| "logps/rejected": -345.307373046875, | |
| "loss": 0.2014, | |
| "rewards/accuracies": 0.918749988079071, | |
| "rewards/chosen": -1.9437261819839478, | |
| "rewards/margins": 5.033670902252197, | |
| "rewards/rejected": -6.977396488189697, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 2.4736000000000002, | |
| "grad_norm": 3.552980899810791, | |
| "learning_rate": 3.9686616792204677e-07, | |
| "logits/chosen": -1.617547631263733, | |
| "logits/rejected": -1.5896202325820923, | |
| "logps/chosen": -245.8852081298828, | |
| "logps/rejected": -349.4147644042969, | |
| "loss": 0.0944, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -1.8278429508209229, | |
| "rewards/margins": 5.722242832183838, | |
| "rewards/rejected": -7.550085544586182, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 2.4842666666666666, | |
| "grad_norm": 2.6436843872070312, | |
| "learning_rate": 3.8144078272934676e-07, | |
| "logits/chosen": -1.727230429649353, | |
| "logits/rejected": -1.6059099435806274, | |
| "logps/chosen": -269.52691650390625, | |
| "logps/rejected": -352.0751647949219, | |
| "loss": 0.0784, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -1.8891433477401733, | |
| "rewards/margins": 6.365373611450195, | |
| "rewards/rejected": -8.2545166015625, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 2.4949333333333334, | |
| "grad_norm": 3.0084333419799805, | |
| "learning_rate": 3.6629635904222943e-07, | |
| "logits/chosen": -1.6209266185760498, | |
| "logits/rejected": -1.6020692586898804, | |
| "logps/chosen": -255.9269561767578, | |
| "logps/rejected": -354.1142578125, | |
| "loss": 0.12, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -2.0139718055725098, | |
| "rewards/margins": 5.429463863372803, | |
| "rewards/rejected": -7.4434356689453125, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 2.5056000000000003, | |
| "grad_norm": 3.403123378753662, | |
| "learning_rate": 3.514349053011698e-07, | |
| "logits/chosen": -1.653505563735962, | |
| "logits/rejected": -1.4818648099899292, | |
| "logps/chosen": -240.80331420898438, | |
| "logps/rejected": -323.3808288574219, | |
| "loss": 0.12, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -1.5888588428497314, | |
| "rewards/margins": 6.088988304138184, | |
| "rewards/rejected": -7.677847385406494, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 2.5162666666666667, | |
| "grad_norm": 4.319390296936035, | |
| "learning_rate": 3.3685839241941133e-07, | |
| "logits/chosen": -1.6171585321426392, | |
| "logits/rejected": -1.6007347106933594, | |
| "logps/chosen": -258.7629089355469, | |
| "logps/rejected": -342.47998046875, | |
| "loss": 0.1118, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -2.1584792137145996, | |
| "rewards/margins": 5.684945583343506, | |
| "rewards/rejected": -7.843425750732422, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 2.5269333333333335, | |
| "grad_norm": 4.567004680633545, | |
| "learning_rate": 3.2256875352158435e-07, | |
| "logits/chosen": -1.6689363718032837, | |
| "logits/rejected": -1.5358021259307861, | |
| "logps/chosen": -236.81808471679688, | |
| "logps/rejected": -322.1048583984375, | |
| "loss": 0.1405, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -1.7008771896362305, | |
| "rewards/margins": 5.543111801147461, | |
| "rewards/rejected": -7.243988990783691, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 2.5376, | |
| "grad_norm": 6.191208362579346, | |
| "learning_rate": 3.085678836873399e-07, | |
| "logits/chosen": -1.702701210975647, | |
| "logits/rejected": -1.6787986755371094, | |
| "logps/chosen": -249.1476593017578, | |
| "logps/rejected": -344.263671875, | |
| "loss": 0.1209, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -1.9360427856445312, | |
| "rewards/margins": 5.811527729034424, | |
| "rewards/rejected": -7.7475690841674805, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 2.5482666666666667, | |
| "grad_norm": 2.056534767150879, | |
| "learning_rate": 2.9485763970002106e-07, | |
| "logits/chosen": -1.675588607788086, | |
| "logits/rejected": -1.6271533966064453, | |
| "logps/chosen": -233.121826171875, | |
| "logps/rejected": -334.0964050292969, | |
| "loss": 0.1223, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -1.9534991979599, | |
| "rewards/margins": 5.506706237792969, | |
| "rewards/rejected": -7.460206508636475, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 2.558933333333333, | |
| "grad_norm": 6.617891788482666, | |
| "learning_rate": 2.8143983980042174e-07, | |
| "logits/chosen": -1.6359453201293945, | |
| "logits/rejected": -1.5568664073944092, | |
| "logps/chosen": -262.3926086425781, | |
| "logps/rejected": -335.15667724609375, | |
| "loss": 0.1334, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -2.1467742919921875, | |
| "rewards/margins": 5.749007701873779, | |
| "rewards/rejected": -7.895781517028809, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.5696, | |
| "grad_norm": 3.7044594287872314, | |
| "learning_rate": 2.683162634456515e-07, | |
| "logits/chosen": -1.583653211593628, | |
| "logits/rejected": -1.6004180908203125, | |
| "logps/chosen": -250.63845825195312, | |
| "logps/rejected": -325.78436279296875, | |
| "loss": 0.1644, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -1.9192173480987549, | |
| "rewards/margins": 5.139154434204102, | |
| "rewards/rejected": -7.058372497558594, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 2.5802666666666667, | |
| "grad_norm": 2.7366485595703125, | |
| "learning_rate": 2.5548865107314606e-07, | |
| "logits/chosen": -1.557058572769165, | |
| "logits/rejected": -1.5571913719177246, | |
| "logps/chosen": -270.0423583984375, | |
| "logps/rejected": -337.73614501953125, | |
| "loss": 0.1198, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -1.8408619165420532, | |
| "rewards/margins": 5.491454601287842, | |
| "rewards/rejected": -7.3323163986206055, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 2.590933333333333, | |
| "grad_norm": 2.9448840618133545, | |
| "learning_rate": 2.429587038698511e-07, | |
| "logits/chosen": -1.641669511795044, | |
| "logits/rejected": -1.5695855617523193, | |
| "logps/chosen": -234.04360961914062, | |
| "logps/rejected": -311.2100524902344, | |
| "loss": 0.122, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -2.0470471382141113, | |
| "rewards/margins": 5.0656328201293945, | |
| "rewards/rejected": -7.112679481506348, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 2.6016, | |
| "grad_norm": 2.9614179134368896, | |
| "learning_rate": 2.307280835466144e-07, | |
| "logits/chosen": -1.749356985092163, | |
| "logits/rejected": -1.6079270839691162, | |
| "logps/chosen": -251.8319854736328, | |
| "logps/rejected": -315.96710205078125, | |
| "loss": 0.1393, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -2.129053831100464, | |
| "rewards/margins": 5.405560493469238, | |
| "rewards/rejected": -7.534613609313965, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 2.6122666666666667, | |
| "grad_norm": 2.680088996887207, | |
| "learning_rate": 2.1879841211780856e-07, | |
| "logits/chosen": -1.5628279447555542, | |
| "logits/rejected": -1.4973697662353516, | |
| "logps/chosen": -252.55484008789062, | |
| "logps/rejected": -329.0304260253906, | |
| "loss": 0.1142, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -2.107846736907959, | |
| "rewards/margins": 5.468822002410889, | |
| "rewards/rejected": -7.576669216156006, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 2.622933333333333, | |
| "grad_norm": 3.77762508392334, | |
| "learning_rate": 2.0717127168622202e-07, | |
| "logits/chosen": -1.564953088760376, | |
| "logits/rejected": -1.4988367557525635, | |
| "logps/chosen": -257.25506591796875, | |
| "logps/rejected": -330.61724853515625, | |
| "loss": 0.1235, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -2.027205467224121, | |
| "rewards/margins": 5.905071258544922, | |
| "rewards/rejected": -7.932276248931885, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 2.6336, | |
| "grad_norm": 3.186311721801758, | |
| "learning_rate": 1.958482042332424e-07, | |
| "logits/chosen": -1.663132667541504, | |
| "logits/rejected": -1.5677236318588257, | |
| "logps/chosen": -243.77468872070312, | |
| "logps/rejected": -324.66485595703125, | |
| "loss": 0.1043, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -1.8823124170303345, | |
| "rewards/margins": 5.564373970031738, | |
| "rewards/rejected": -7.446686744689941, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 2.6442666666666668, | |
| "grad_norm": 1.523322343826294, | |
| "learning_rate": 1.8483071141435937e-07, | |
| "logits/chosen": -1.591597318649292, | |
| "logits/rejected": -1.5641478300094604, | |
| "logps/chosen": -278.78485107421875, | |
| "logps/rejected": -347.0000305175781, | |
| "loss": 0.116, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -2.0552477836608887, | |
| "rewards/margins": 5.445364475250244, | |
| "rewards/rejected": -7.500611782073975, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 2.654933333333333, | |
| "grad_norm": 2.483313798904419, | |
| "learning_rate": 1.741202543600176e-07, | |
| "logits/chosen": -1.658911108970642, | |
| "logits/rejected": -1.600517988204956, | |
| "logps/chosen": -239.78817749023438, | |
| "logps/rejected": -339.4964294433594, | |
| "loss": 0.1052, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -2.173642635345459, | |
| "rewards/margins": 6.087950706481934, | |
| "rewards/rejected": -8.261592864990234, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 2.6656, | |
| "grad_norm": 2.3703670501708984, | |
| "learning_rate": 1.6371825348184155e-07, | |
| "logits/chosen": -1.5409141778945923, | |
| "logits/rejected": -1.4761722087860107, | |
| "logps/chosen": -240.65487670898438, | |
| "logps/rejected": -336.3206787109375, | |
| "loss": 0.0976, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -1.9273293018341064, | |
| "rewards/margins": 6.195014953613281, | |
| "rewards/rejected": -8.122344017028809, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.676266666666667, | |
| "grad_norm": 3.7135093212127686, | |
| "learning_rate": 1.5362608828426246e-07, | |
| "logits/chosen": -1.6882669925689697, | |
| "logits/rejected": -1.6413158178329468, | |
| "logps/chosen": -259.45892333984375, | |
| "logps/rejected": -355.84356689453125, | |
| "loss": 0.0932, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -2.087570905685425, | |
| "rewards/margins": 5.9243245124816895, | |
| "rewards/rejected": -8.011896133422852, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 2.686933333333333, | |
| "grad_norm": 4.499937057495117, | |
| "learning_rate": 1.438450971815686e-07, | |
| "logits/chosen": -1.6984704732894897, | |
| "logits/rejected": -1.5697863101959229, | |
| "logps/chosen": -293.70849609375, | |
| "logps/rejected": -330.1623229980469, | |
| "loss": 0.0973, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -1.8370702266693115, | |
| "rewards/margins": 6.17175817489624, | |
| "rewards/rejected": -8.008829116821289, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 2.6976, | |
| "grad_norm": 2.924912691116333, | |
| "learning_rate": 1.3437657732040783e-07, | |
| "logits/chosen": -1.6595064401626587, | |
| "logits/rejected": -1.5919541120529175, | |
| "logps/chosen": -272.3826599121094, | |
| "logps/rejected": -333.85504150390625, | |
| "loss": 0.1171, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -2.263150691986084, | |
| "rewards/margins": 6.062061309814453, | |
| "rewards/rejected": -8.325211524963379, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 2.708266666666667, | |
| "grad_norm": 3.1616270542144775, | |
| "learning_rate": 1.2522178440776022e-07, | |
| "logits/chosen": -1.5888944864273071, | |
| "logits/rejected": -1.5637750625610352, | |
| "logps/chosen": -259.8056640625, | |
| "logps/rejected": -316.7781982421875, | |
| "loss": 0.1181, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -2.2415337562561035, | |
| "rewards/margins": 5.334750175476074, | |
| "rewards/rejected": -7.576283931732178, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 2.718933333333333, | |
| "grad_norm": 2.0143356323242188, | |
| "learning_rate": 1.1638193254440655e-07, | |
| "logits/chosen": -1.6921374797821045, | |
| "logits/rejected": -1.6408201456069946, | |
| "logps/chosen": -256.64093017578125, | |
| "logps/rejected": -332.04278564453125, | |
| "loss": 0.1319, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": -2.222224473953247, | |
| "rewards/margins": 5.89315938949585, | |
| "rewards/rejected": -8.115384101867676, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 2.7296, | |
| "grad_norm": 2.776367425918579, | |
| "learning_rate": 1.0785819406391562e-07, | |
| "logits/chosen": -1.6165263652801514, | |
| "logits/rejected": -1.5132572650909424, | |
| "logps/chosen": -248.881103515625, | |
| "logps/rejected": -339.60540771484375, | |
| "loss": 0.112, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -2.215090274810791, | |
| "rewards/margins": 5.569876670837402, | |
| "rewards/rejected": -7.784966468811035, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 2.740266666666667, | |
| "grad_norm": 2.7171905040740967, | |
| "learning_rate": 9.965169937717095e-08, | |
| "logits/chosen": -1.6130340099334717, | |
| "logits/rejected": -1.5577963590621948, | |
| "logps/chosen": -260.16143798828125, | |
| "logps/rejected": -337.82611083984375, | |
| "loss": 0.098, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -2.314629316329956, | |
| "rewards/margins": 5.716673851013184, | |
| "rewards/rejected": -8.031302452087402, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 2.7509333333333332, | |
| "grad_norm": 2.7834761142730713, | |
| "learning_rate": 9.176353682245676e-08, | |
| "logits/chosen": -1.6675822734832764, | |
| "logits/rejected": -1.5338008403778076, | |
| "logps/chosen": -257.9503479003906, | |
| "logps/rejected": -337.56756591796875, | |
| "loss": 0.1559, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/chosen": -2.267277717590332, | |
| "rewards/margins": 5.890148162841797, | |
| "rewards/rejected": -8.157425880432129, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.7616, | |
| "grad_norm": 1.5055266618728638, | |
| "learning_rate": 8.41947525211223e-08, | |
| "logits/chosen": -1.5467194318771362, | |
| "logits/rejected": -1.4370124340057373, | |
| "logps/chosen": -245.481689453125, | |
| "logps/rejected": -307.88970947265625, | |
| "loss": 0.1115, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -2.0703835487365723, | |
| "rewards/margins": 5.803430557250977, | |
| "rewards/rejected": -7.873814582824707, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 2.772266666666667, | |
| "grad_norm": 3.726200580596924, | |
| "learning_rate": 7.694635023884789e-08, | |
| "logits/chosen": -1.615465521812439, | |
| "logits/rejected": -1.5055599212646484, | |
| "logps/chosen": -261.68084716796875, | |
| "logps/rejected": -333.84735107421875, | |
| "loss": 0.1108, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": -1.922776460647583, | |
| "rewards/margins": 5.716401100158691, | |
| "rewards/rejected": -7.6391777992248535, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.7829333333333333, | |
| "grad_norm": 3.27950382232666, | |
| "learning_rate": 7.00192912525241e-08, | |
| "logits/chosen": -1.7413036823272705, | |
| "logits/rejected": -1.6767768859863281, | |
| "logps/chosen": -249.2274627685547, | |
| "logps/rejected": -324.9376525878906, | |
| "loss": 0.1504, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": -2.0574612617492676, | |
| "rewards/margins": 5.184039115905762, | |
| "rewards/rejected": -7.2415008544921875, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 2.7936, | |
| "grad_norm": 2.958669900894165, | |
| "learning_rate": 6.341449422277047e-08, | |
| "logits/chosen": -1.6207433938980103, | |
| "logits/rejected": -1.5499709844589233, | |
| "logps/chosen": -271.83709716796875, | |
| "logps/rejected": -333.00177001953125, | |
| "loss": 0.0846, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -2.2863690853118896, | |
| "rewards/margins": 5.1181640625, | |
| "rewards/rejected": -7.404533386230469, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.804266666666667, | |
| "grad_norm": 3.068798303604126, | |
| "learning_rate": 5.7132835072101486e-08, | |
| "logits/chosen": -1.7072492837905884, | |
| "logits/rejected": -1.6401913166046143, | |
| "logps/chosen": -248.870849609375, | |
| "logps/rejected": -345.37939453125, | |
| "loss": 0.1536, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -2.2537102699279785, | |
| "rewards/margins": 5.832470893859863, | |
| "rewards/rejected": -8.086182594299316, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 2.8149333333333333, | |
| "grad_norm": 2.465273141860962, | |
| "learning_rate": 5.117514686876379e-08, | |
| "logits/chosen": -1.6905529499053955, | |
| "logits/rejected": -1.5897334814071655, | |
| "logps/chosen": -234.72500610351562, | |
| "logps/rejected": -331.4391174316406, | |
| "loss": 0.1376, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -1.9037342071533203, | |
| "rewards/margins": 5.567811965942383, | |
| "rewards/rejected": -7.4715471267700195, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.8256, | |
| "grad_norm": 5.49542760848999, | |
| "learning_rate": 4.5542219716255274e-08, | |
| "logits/chosen": -1.62538743019104, | |
| "logits/rejected": -1.5213114023208618, | |
| "logps/chosen": -267.12176513671875, | |
| "logps/rejected": -337.9035949707031, | |
| "loss": 0.1046, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -2.3189761638641357, | |
| "rewards/margins": 5.719181060791016, | |
| "rewards/rejected": -8.03815746307373, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 2.836266666666667, | |
| "grad_norm": 2.2211387157440186, | |
| "learning_rate": 4.023480064854174e-08, | |
| "logits/chosen": -1.6421716213226318, | |
| "logits/rejected": -1.6016359329223633, | |
| "logps/chosen": -257.92681884765625, | |
| "logps/rejected": -335.91326904296875, | |
| "loss": 0.1023, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -2.207742214202881, | |
| "rewards/margins": 5.49573278427124, | |
| "rewards/rejected": -7.703474998474121, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.8469333333333333, | |
| "grad_norm": 1.8862336874008179, | |
| "learning_rate": 3.525359353098584e-08, | |
| "logits/chosen": -1.7027902603149414, | |
| "logits/rejected": -1.6094478368759155, | |
| "logps/chosen": -240.2790069580078, | |
| "logps/rejected": -345.55712890625, | |
| "loss": 0.0974, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -1.9166862964630127, | |
| "rewards/margins": 6.030681610107422, | |
| "rewards/rejected": -7.947368621826172, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 2.8576, | |
| "grad_norm": 1.444413661956787, | |
| "learning_rate": 3.059925896700094e-08, | |
| "logits/chosen": -1.6089794635772705, | |
| "logits/rejected": -1.5267149209976196, | |
| "logps/chosen": -269.1159362792969, | |
| "logps/rejected": -328.16534423828125, | |
| "loss": 0.1216, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -1.9404586553573608, | |
| "rewards/margins": 5.162356376647949, | |
| "rewards/rejected": -7.1028151512146, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.8682666666666665, | |
| "grad_norm": 5.427722454071045, | |
| "learning_rate": 2.6272414210443964e-08, | |
| "logits/chosen": -1.6677299737930298, | |
| "logits/rejected": -1.625317931175232, | |
| "logps/chosen": -269.1617736816406, | |
| "logps/rejected": -345.4956970214844, | |
| "loss": 0.1349, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -2.075646162033081, | |
| "rewards/margins": 5.521030426025391, | |
| "rewards/rejected": -7.596676826477051, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 2.8789333333333333, | |
| "grad_norm": 1.782738208770752, | |
| "learning_rate": 2.2273633083753632e-08, | |
| "logits/chosen": -1.7253984212875366, | |
| "logits/rejected": -1.5926178693771362, | |
| "logps/chosen": -249.1942596435547, | |
| "logps/rejected": -338.21722412109375, | |
| "loss": 0.1067, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -2.3312084674835205, | |
| "rewards/margins": 6.381017208099365, | |
| "rewards/rejected": -8.712225914001465, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.8895999999999997, | |
| "grad_norm": 5.87308931350708, | |
| "learning_rate": 1.8603445901851647e-08, | |
| "logits/chosen": -1.621783971786499, | |
| "logits/rejected": -1.5659376382827759, | |
| "logps/chosen": -252.837646484375, | |
| "logps/rejected": -325.2681884765625, | |
| "loss": 0.1463, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -2.2585935592651367, | |
| "rewards/margins": 5.560887336730957, | |
| "rewards/rejected": -7.819480895996094, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 2.9002666666666665, | |
| "grad_norm": 6.012791633605957, | |
| "learning_rate": 1.5262339401813375e-08, | |
| "logits/chosen": -1.684981346130371, | |
| "logits/rejected": -1.6372852325439453, | |
| "logps/chosen": -250.5609130859375, | |
| "logps/rejected": -345.0445556640625, | |
| "loss": 0.145, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": -2.3670361042022705, | |
| "rewards/margins": 5.3330302238464355, | |
| "rewards/rejected": -7.700067043304443, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.9109333333333334, | |
| "grad_norm": 3.0086324214935303, | |
| "learning_rate": 1.2250756678315601e-08, | |
| "logits/chosen": -1.6575305461883545, | |
| "logits/rejected": -1.5837920904159546, | |
| "logps/chosen": -269.42578125, | |
| "logps/rejected": -344.31793212890625, | |
| "loss": 0.092, | |
| "rewards/accuracies": 0.987500011920929, | |
| "rewards/chosen": -2.132127046585083, | |
| "rewards/margins": 5.720582962036133, | |
| "rewards/rejected": -7.852709770202637, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 2.9215999999999998, | |
| "grad_norm": 2.5158169269561768, | |
| "learning_rate": 9.569097124875193e-09, | |
| "logits/chosen": -1.636508584022522, | |
| "logits/rejected": -1.567469596862793, | |
| "logps/chosen": -243.44912719726562, | |
| "logps/rejected": -332.2774353027344, | |
| "loss": 0.1252, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": -2.221357822418213, | |
| "rewards/margins": 5.588191032409668, | |
| "rewards/rejected": -7.809549808502197, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.9322666666666666, | |
| "grad_norm": 3.4159395694732666, | |
| "learning_rate": 7.217716380881479e-09, | |
| "logits/chosen": -1.6869888305664062, | |
| "logits/rejected": -1.615618348121643, | |
| "logps/chosen": -250.40011596679688, | |
| "logps/rejected": -348.78521728515625, | |
| "loss": 0.0966, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -2.015139102935791, | |
| "rewards/margins": 5.993285179138184, | |
| "rewards/rejected": -8.008424758911133, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 2.9429333333333334, | |
| "grad_norm": 2.5707411766052246, | |
| "learning_rate": 5.196926284430359e-09, | |
| "logits/chosen": -1.6021541357040405, | |
| "logits/rejected": -1.5102264881134033, | |
| "logps/chosen": -261.15020751953125, | |
| "logps/rejected": -347.87188720703125, | |
| "loss": 0.098, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -2.2800986766815186, | |
| "rewards/margins": 6.126519680023193, | |
| "rewards/rejected": -8.40661907196045, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.9536, | |
| "grad_norm": 2.2150466442108154, | |
| "learning_rate": 3.5069948309712775e-09, | |
| "logits/chosen": -1.6135914325714111, | |
| "logits/rejected": -1.5576746463775635, | |
| "logps/chosen": -262.16571044921875, | |
| "logps/rejected": -354.02056884765625, | |
| "loss": 0.1059, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -1.9849374294281006, | |
| "rewards/margins": 5.869291305541992, | |
| "rewards/rejected": -7.8542280197143555, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 2.9642666666666666, | |
| "grad_norm": 3.714444637298584, | |
| "learning_rate": 2.1481461377634294e-09, | |
| "logits/chosen": -1.6816928386688232, | |
| "logits/rejected": -1.6221061944961548, | |
| "logps/chosen": -265.137451171875, | |
| "logps/rejected": -331.76727294921875, | |
| "loss": 0.1451, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -2.0669548511505127, | |
| "rewards/margins": 5.562548637390137, | |
| "rewards/rejected": -7.629502773284912, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.9749333333333334, | |
| "grad_norm": 2.0670387744903564, | |
| "learning_rate": 1.120560414154259e-09, | |
| "logits/chosen": -1.5556447505950928, | |
| "logits/rejected": -1.5012569427490234, | |
| "logps/chosen": -235.62307739257812, | |
| "logps/rejected": -331.75531005859375, | |
| "loss": 0.1097, | |
| "rewards/accuracies": 0.96875, | |
| "rewards/chosen": -1.9400962591171265, | |
| "rewards/margins": 5.777553081512451, | |
| "rewards/rejected": -7.717649936676025, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 2.9856, | |
| "grad_norm": 3.319321632385254, | |
| "learning_rate": 4.2437393768079983e-10, | |
| "logits/chosen": -1.6425552368164062, | |
| "logits/rejected": -1.5908761024475098, | |
| "logps/chosen": -259.075439453125, | |
| "logps/rejected": -341.9222412109375, | |
| "loss": 0.1158, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": -1.9800103902816772, | |
| "rewards/margins": 6.281598091125488, | |
| "rewards/rejected": -8.261608123779297, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.9962666666666666, | |
| "grad_norm": 3.756913423538208, | |
| "learning_rate": 5.967903599635171e-11, | |
| "logits/chosen": -1.6659742593765259, | |
| "logits/rejected": -1.5326024293899536, | |
| "logps/chosen": -264.480712890625, | |
| "logps/rejected": -333.2155456542969, | |
| "loss": 0.1354, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": -2.168793201446533, | |
| "rewards/margins": 5.448444843292236, | |
| "rewards/rejected": -7.617238521575928, | |
| "step": 1405 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 1407, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |