| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9982631930527722, | |
| "eval_steps": 400, | |
| "global_step": 467, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01068804275217101, | |
| "grad_norm": 53.52218298444476, | |
| "learning_rate": 1.0638297872340425e-07, | |
| "logits/chosen": -1.0146243572235107, | |
| "logits/rejected": -0.9850981831550598, | |
| "logps/chosen": -0.27403339743614197, | |
| "logps/rejected": -0.2716384530067444, | |
| "loss": 3.0444, | |
| "rewards/accuracies": 0.4437499940395355, | |
| "rewards/chosen": -2.7403340339660645, | |
| "rewards/margins": -0.02394939959049225, | |
| "rewards/rejected": -2.7163848876953125, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02137608550434202, | |
| "grad_norm": 39.10999969888965, | |
| "learning_rate": 2.127659574468085e-07, | |
| "logits/chosen": -1.0449364185333252, | |
| "logits/rejected": -0.9776930809020996, | |
| "logps/chosen": -0.29451489448547363, | |
| "logps/rejected": -0.2995792329311371, | |
| "loss": 3.0211, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -2.9451489448547363, | |
| "rewards/margins": 0.05064352601766586, | |
| "rewards/rejected": -2.9957923889160156, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03206412825651302, | |
| "grad_norm": 53.821066581509214, | |
| "learning_rate": 3.1914893617021275e-07, | |
| "logits/chosen": -0.9672248959541321, | |
| "logits/rejected": -0.9867329597473145, | |
| "logps/chosen": -0.26386433839797974, | |
| "logps/rejected": -0.30063143372535706, | |
| "loss": 3.0404, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -2.638643503189087, | |
| "rewards/margins": 0.3676711320877075, | |
| "rewards/rejected": -3.006314516067505, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04275217100868404, | |
| "grad_norm": 86.6542555553414, | |
| "learning_rate": 4.25531914893617e-07, | |
| "logits/chosen": -0.9602643847465515, | |
| "logits/rejected": -0.9344671964645386, | |
| "logps/chosen": -0.2776374816894531, | |
| "logps/rejected": -0.29131022095680237, | |
| "loss": 2.9793, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -2.7763748168945312, | |
| "rewards/margins": 0.13672712445259094, | |
| "rewards/rejected": -2.913102149963379, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.053440213760855046, | |
| "grad_norm": 56.919799993589805, | |
| "learning_rate": 5.319148936170212e-07, | |
| "logits/chosen": -1.0135596990585327, | |
| "logits/rejected": -0.9844949841499329, | |
| "logps/chosen": -0.2717221677303314, | |
| "logps/rejected": -0.2782990336418152, | |
| "loss": 3.124, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -2.717221736907959, | |
| "rewards/margins": 0.06576814502477646, | |
| "rewards/rejected": -2.7829902172088623, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06412825651302605, | |
| "grad_norm": 45.796379698409524, | |
| "learning_rate": 6.382978723404255e-07, | |
| "logits/chosen": -0.9898878931999207, | |
| "logits/rejected": -0.9455238580703735, | |
| "logps/chosen": -0.2733747959136963, | |
| "logps/rejected": -0.279060035943985, | |
| "loss": 2.8977, | |
| "rewards/accuracies": 0.4437499940395355, | |
| "rewards/chosen": -2.733747959136963, | |
| "rewards/margins": 0.05685253068804741, | |
| "rewards/rejected": -2.790600299835205, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07481629926519706, | |
| "grad_norm": 64.64288788170485, | |
| "learning_rate": 7.446808510638297e-07, | |
| "logits/chosen": -1.0491113662719727, | |
| "logits/rejected": -0.9738750457763672, | |
| "logps/chosen": -0.2941775918006897, | |
| "logps/rejected": -0.32069069147109985, | |
| "loss": 2.9119, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -2.9417757987976074, | |
| "rewards/margins": 0.2651310861110687, | |
| "rewards/rejected": -3.206906795501709, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08550434201736808, | |
| "grad_norm": 60.56769615337976, | |
| "learning_rate": 8.51063829787234e-07, | |
| "logits/chosen": -1.0074384212493896, | |
| "logits/rejected": -0.963466465473175, | |
| "logps/chosen": -0.2797192931175232, | |
| "logps/rejected": -0.3225395083427429, | |
| "loss": 2.9345, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -2.7971930503845215, | |
| "rewards/margins": 0.4282020032405853, | |
| "rewards/rejected": -3.2253952026367188, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09619238476953908, | |
| "grad_norm": 48.675093440338955, | |
| "learning_rate": 9.574468085106384e-07, | |
| "logits/chosen": -1.0469945669174194, | |
| "logits/rejected": -1.0040814876556396, | |
| "logps/chosen": -0.33255186676979065, | |
| "logps/rejected": -0.38402628898620605, | |
| "loss": 2.9815, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -3.3255183696746826, | |
| "rewards/margins": 0.5147446393966675, | |
| "rewards/rejected": -3.8402628898620605, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10688042752171009, | |
| "grad_norm": 92.08652708998007, | |
| "learning_rate": 9.998741174712533e-07, | |
| "logits/chosen": -1.038892388343811, | |
| "logits/rejected": -0.988103985786438, | |
| "logps/chosen": -0.34245526790618896, | |
| "logps/rejected": -0.38594862818717957, | |
| "loss": 3.0508, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -3.4245529174804688, | |
| "rewards/margins": 0.4349338412284851, | |
| "rewards/rejected": -3.8594863414764404, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11756847027388109, | |
| "grad_norm": 72.54827446103837, | |
| "learning_rate": 9.991050648838675e-07, | |
| "logits/chosen": -1.0567952394485474, | |
| "logits/rejected": -1.0215675830841064, | |
| "logps/chosen": -0.28753459453582764, | |
| "logps/rejected": -0.3490275740623474, | |
| "loss": 2.7982, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -2.8753461837768555, | |
| "rewards/margins": 0.6149295568466187, | |
| "rewards/rejected": -3.4902758598327637, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.1282565130260521, | |
| "grad_norm": 55.56312267177659, | |
| "learning_rate": 9.97637968732563e-07, | |
| "logits/chosen": -1.0922194719314575, | |
| "logits/rejected": -1.059291958808899, | |
| "logps/chosen": -0.3225264847278595, | |
| "logps/rejected": -0.3470703959465027, | |
| "loss": 2.8716, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -3.22526478767395, | |
| "rewards/margins": 0.2454390972852707, | |
| "rewards/rejected": -3.4707038402557373, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13894455577822312, | |
| "grad_norm": 54.05440384507174, | |
| "learning_rate": 9.954748808839674e-07, | |
| "logits/chosen": -0.9975064992904663, | |
| "logits/rejected": -0.9689160585403442, | |
| "logps/chosen": -0.37468865513801575, | |
| "logps/rejected": -0.43205341696739197, | |
| "loss": 2.7901, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -3.746886730194092, | |
| "rewards/margins": 0.5736472010612488, | |
| "rewards/rejected": -4.3205342292785645, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14963259853039412, | |
| "grad_norm": 37.66775098927071, | |
| "learning_rate": 9.926188266120295e-07, | |
| "logits/chosen": -1.0229814052581787, | |
| "logits/rejected": -0.9982998967170715, | |
| "logps/chosen": -0.3514581620693207, | |
| "logps/rejected": -0.4274352192878723, | |
| "loss": 2.8718, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -3.5145821571350098, | |
| "rewards/margins": 0.7597699761390686, | |
| "rewards/rejected": -4.274352073669434, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.16032064128256512, | |
| "grad_norm": 51.934633835606974, | |
| "learning_rate": 9.890738003669027e-07, | |
| "logits/chosen": -0.9838461875915527, | |
| "logits/rejected": -0.9134309887886047, | |
| "logps/chosen": -0.35928577184677124, | |
| "logps/rejected": -0.4099213182926178, | |
| "loss": 2.8345, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -3.592857837677002, | |
| "rewards/margins": 0.5063551664352417, | |
| "rewards/rejected": -4.099213123321533, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.17100868403473615, | |
| "grad_norm": 46.83578017177419, | |
| "learning_rate": 9.848447601883433e-07, | |
| "logits/chosen": -0.9681940078735352, | |
| "logits/rejected": -0.9539217948913574, | |
| "logps/chosen": -0.353752076625824, | |
| "logps/rejected": -0.4523216187953949, | |
| "loss": 2.7878, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -3.53752064704895, | |
| "rewards/margins": 0.985695481300354, | |
| "rewards/rejected": -4.523216247558594, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.18169672678690715, | |
| "grad_norm": 54.174949491419966, | |
| "learning_rate": 9.799376207714444e-07, | |
| "logits/chosen": -0.9862138628959656, | |
| "logits/rejected": -0.9641338586807251, | |
| "logps/chosen": -0.3405635952949524, | |
| "logps/rejected": -0.39860305190086365, | |
| "loss": 2.6715, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -3.4056358337402344, | |
| "rewards/margins": 0.5803946852684021, | |
| "rewards/rejected": -3.9860305786132812, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.19238476953907815, | |
| "grad_norm": 62.18682762469074, | |
| "learning_rate": 9.743592451943998e-07, | |
| "logits/chosen": -1.028374195098877, | |
| "logits/rejected": -0.9928615689277649, | |
| "logps/chosen": -0.4192899763584137, | |
| "logps/rejected": -0.5028694868087769, | |
| "loss": 2.8803, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -4.192899703979492, | |
| "rewards/margins": 0.8357950448989868, | |
| "rewards/rejected": -5.028695106506348, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.20307281229124916, | |
| "grad_norm": 56.712862810919404, | |
| "learning_rate": 9.681174353198686e-07, | |
| "logits/chosen": -1.102429747581482, | |
| "logits/rejected": -1.017956256866455, | |
| "logps/chosen": -0.4515204429626465, | |
| "logps/rejected": -0.49105948209762573, | |
| "loss": 2.7854, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -4.515204429626465, | |
| "rewards/margins": 0.39539000391960144, | |
| "rewards/rejected": -4.910594463348389, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.21376085504342018, | |
| "grad_norm": 77.56651991727357, | |
| "learning_rate": 9.612209208833646e-07, | |
| "logits/chosen": -1.0002816915512085, | |
| "logits/rejected": -0.9756115078926086, | |
| "logps/chosen": -0.4405655860900879, | |
| "logps/rejected": -0.5030835866928101, | |
| "loss": 2.8381, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -4.405655860900879, | |
| "rewards/margins": 0.6251801252365112, | |
| "rewards/rejected": -5.0308356285095215, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.22444889779559118, | |
| "grad_norm": 63.78609875386195, | |
| "learning_rate": 9.536793472839324e-07, | |
| "logits/chosen": -1.0079588890075684, | |
| "logits/rejected": -0.9540907144546509, | |
| "logps/chosen": -0.41310757398605347, | |
| "logps/rejected": -0.5235550999641418, | |
| "loss": 2.7704, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -4.131075859069824, | |
| "rewards/margins": 1.1044747829437256, | |
| "rewards/rejected": -5.235550880432129, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23513694054776219, | |
| "grad_norm": 59.92913033519696, | |
| "learning_rate": 9.455032620941839e-07, | |
| "logits/chosen": -0.9624613523483276, | |
| "logits/rejected": -0.9022065997123718, | |
| "logps/chosen": -0.4771413207054138, | |
| "logps/rejected": -0.6054214239120483, | |
| "loss": 2.6684, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -4.771413326263428, | |
| "rewards/margins": 1.2828001976013184, | |
| "rewards/rejected": -6.0542144775390625, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2458249832999332, | |
| "grad_norm": 57.71552130623015, | |
| "learning_rate": 9.367041003085648e-07, | |
| "logits/chosen": -1.0269968509674072, | |
| "logits/rejected": -0.9661616086959839, | |
| "logps/chosen": -0.5121074914932251, | |
| "logps/rejected": -0.578630268573761, | |
| "loss": 2.5559, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -5.121075630187988, | |
| "rewards/margins": 0.6652273535728455, | |
| "rewards/rejected": -5.7863030433654785, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2565130260521042, | |
| "grad_norm": 64.09249680400335, | |
| "learning_rate": 9.272941683504808e-07, | |
| "logits/chosen": -0.998211681842804, | |
| "logits/rejected": -0.9050429463386536, | |
| "logps/chosen": -0.5254617929458618, | |
| "logps/rejected": -0.7217136025428772, | |
| "loss": 2.4049, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -5.2546186447143555, | |
| "rewards/margins": 1.9625177383422852, | |
| "rewards/rejected": -7.217136383056641, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.26720106880427524, | |
| "grad_norm": 54.10213565718134, | |
| "learning_rate": 9.172866268606513e-07, | |
| "logits/chosen": -1.06635320186615, | |
| "logits/rejected": -1.0216171741485596, | |
| "logps/chosen": -0.5953704714775085, | |
| "logps/rejected": -0.6902128458023071, | |
| "loss": 2.3251, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -5.953704833984375, | |
| "rewards/margins": 0.9484230875968933, | |
| "rewards/rejected": -6.902127742767334, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.27788911155644624, | |
| "grad_norm": 87.63946362541415, | |
| "learning_rate": 9.066954722907638e-07, | |
| "logits/chosen": -1.0916361808776855, | |
| "logits/rejected": -1.085458517074585, | |
| "logps/chosen": -0.588487446308136, | |
| "logps/rejected": -0.8501450419425964, | |
| "loss": 2.1826, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -5.884873867034912, | |
| "rewards/margins": 2.6165759563446045, | |
| "rewards/rejected": -8.501450538635254, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.28857715430861725, | |
| "grad_norm": 66.78226800807278, | |
| "learning_rate": 8.955355173281707e-07, | |
| "logits/chosen": -1.067176342010498, | |
| "logits/rejected": -1.0195186138153076, | |
| "logps/chosen": -0.6727192401885986, | |
| "logps/rejected": -0.8245170712471008, | |
| "loss": 2.1861, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -6.7271928787231445, | |
| "rewards/margins": 1.5179781913757324, | |
| "rewards/rejected": -8.245170593261719, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.29926519706078825, | |
| "grad_norm": 77.9071558548112, | |
| "learning_rate": 8.838223701790055e-07, | |
| "logits/chosen": -1.1569595336914062, | |
| "logits/rejected": -1.1336597204208374, | |
| "logps/chosen": -0.785293698310852, | |
| "logps/rejected": -0.9337224960327148, | |
| "loss": 2.1564, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -7.852936744689941, | |
| "rewards/margins": 1.4842884540557861, | |
| "rewards/rejected": -9.337224960327148, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.30995323981295925, | |
| "grad_norm": 109.08809267522398, | |
| "learning_rate": 8.71572412738697e-07, | |
| "logits/chosen": -1.0582095384597778, | |
| "logits/rejected": -1.03193998336792, | |
| "logps/chosen": -0.827114462852478, | |
| "logps/rejected": -1.0477594137191772, | |
| "loss": 2.0118, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -8.27114486694336, | |
| "rewards/margins": 2.206449508666992, | |
| "rewards/rejected": -10.477594375610352, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.32064128256513025, | |
| "grad_norm": 72.02715367718524, | |
| "learning_rate": 8.588027776804058e-07, | |
| "logits/chosen": -1.0821495056152344, | |
| "logits/rejected": -1.0622715950012207, | |
| "logps/chosen": -0.9100320935249329, | |
| "logps/rejected": -1.1453698873519897, | |
| "loss": 2.0273, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": -9.100319862365723, | |
| "rewards/margins": 2.353379011154175, | |
| "rewards/rejected": -11.453699111938477, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.33132932531730125, | |
| "grad_norm": 70.83089987980944, | |
| "learning_rate": 8.455313244934324e-07, | |
| "logits/chosen": -1.092185139656067, | |
| "logits/rejected": -1.070657730102539, | |
| "logps/chosen": -0.9754332304000854, | |
| "logps/rejected": -1.2774028778076172, | |
| "loss": 2.0633, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -9.754331588745117, | |
| "rewards/margins": 3.0196967124938965, | |
| "rewards/rejected": -12.774029731750488, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.3420173680694723, | |
| "grad_norm": 89.85353120616982, | |
| "learning_rate": 8.317766145051057e-07, | |
| "logits/chosen": -1.107634425163269, | |
| "logits/rejected": -1.0893046855926514, | |
| "logps/chosen": -1.0988253355026245, | |
| "logps/rejected": -1.4862325191497803, | |
| "loss": 2.0523, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -10.988253593444824, | |
| "rewards/margins": 3.8740711212158203, | |
| "rewards/rejected": -14.862322807312012, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3527054108216433, | |
| "grad_norm": 60.873789300571126, | |
| "learning_rate": 8.175578849210894e-07, | |
| "logits/chosen": -1.1269104480743408, | |
| "logits/rejected": -1.1010853052139282, | |
| "logps/chosen": -1.0796130895614624, | |
| "logps/rejected": -1.4461021423339844, | |
| "loss": 1.8838, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -10.796131134033203, | |
| "rewards/margins": 3.6648898124694824, | |
| "rewards/rejected": -14.461019515991211, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3633934535738143, | |
| "grad_norm": 77.23211870911884, | |
| "learning_rate": 8.028950219204099e-07, | |
| "logits/chosen": -1.1438876390457153, | |
| "logits/rejected": -1.1206210851669312, | |
| "logps/chosen": -1.0647801160812378, | |
| "logps/rejected": -1.4476187229156494, | |
| "loss": 1.8488, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -10.647802352905273, | |
| "rewards/margins": 3.828385591506958, | |
| "rewards/rejected": -14.476186752319336, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.3740814963259853, | |
| "grad_norm": 97.35090322598491, | |
| "learning_rate": 7.878085328428368e-07, | |
| "logits/chosen": -1.157462239265442, | |
| "logits/rejected": -1.1056431531906128, | |
| "logps/chosen": -1.1336826086044312, | |
| "logps/rejected": -1.3956897258758545, | |
| "loss": 1.7083, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -11.336827278137207, | |
| "rewards/margins": 2.620070695877075, | |
| "rewards/rejected": -13.956896781921387, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3847695390781563, | |
| "grad_norm": 70.60533034676232, | |
| "learning_rate": 7.723195175075135e-07, | |
| "logits/chosen": -1.1124871969223022, | |
| "logits/rejected": -1.0904567241668701, | |
| "logps/chosen": -1.0966602563858032, | |
| "logps/rejected": -1.4549492597579956, | |
| "loss": 1.6569, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -10.966601371765137, | |
| "rewards/margins": 3.5828919410705566, | |
| "rewards/rejected": -14.549494743347168, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3954575818303273, | |
| "grad_norm": 84.92007593834019, | |
| "learning_rate": 7.564496387029531e-07, | |
| "logits/chosen": -1.1567202806472778, | |
| "logits/rejected": -1.0984870195388794, | |
| "logps/chosen": -1.1207507848739624, | |
| "logps/rejected": -1.5250511169433594, | |
| "loss": 1.6817, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -11.20750904083252, | |
| "rewards/margins": 4.043001651763916, | |
| "rewards/rejected": -15.250509262084961, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.4061456245824983, | |
| "grad_norm": 82.82215861540205, | |
| "learning_rate": 7.402210918896689e-07, | |
| "logits/chosen": -1.1628299951553345, | |
| "logits/rejected": -1.170377492904663, | |
| "logps/chosen": -1.235033392906189, | |
| "logps/rejected": -1.7156970500946045, | |
| "loss": 1.5387, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -12.350334167480469, | |
| "rewards/margins": 4.806637763977051, | |
| "rewards/rejected": -17.156970977783203, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4168336673346693, | |
| "grad_norm": 70.50682719627838, | |
| "learning_rate": 7.236565741578162e-07, | |
| "logits/chosen": -1.1164333820343018, | |
| "logits/rejected": -1.0961400270462036, | |
| "logps/chosen": -1.2593460083007812, | |
| "logps/rejected": -1.6189504861831665, | |
| "loss": 1.6047, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -12.593461036682129, | |
| "rewards/margins": 3.5960440635681152, | |
| "rewards/rejected": -16.189504623413086, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.42752171008684037, | |
| "grad_norm": 97.28442308133118, | |
| "learning_rate": 7.067792524832603e-07, | |
| "logits/chosen": -1.1036303043365479, | |
| "logits/rejected": -1.0911258459091187, | |
| "logps/chosen": -1.324706792831421, | |
| "logps/rejected": -1.7423721551895142, | |
| "loss": 1.5626, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -13.247068405151367, | |
| "rewards/margins": 4.176652908325195, | |
| "rewards/rejected": -17.42371940612793, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.43820975283901137, | |
| "grad_norm": 85.54406338680343, | |
| "learning_rate": 6.896127313264642e-07, | |
| "logits/chosen": -1.154517650604248, | |
| "logits/rejected": -1.101162314414978, | |
| "logps/chosen": -1.40175461769104, | |
| "logps/rejected": -1.8435806035995483, | |
| "loss": 1.7321, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -14.017547607421875, | |
| "rewards/margins": 4.418261528015137, | |
| "rewards/rejected": -18.435808181762695, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.44889779559118237, | |
| "grad_norm": 103.7420052940262, | |
| "learning_rate": 6.721810196195174e-07, | |
| "logits/chosen": -1.1762125492095947, | |
| "logits/rejected": -1.1645376682281494, | |
| "logps/chosen": -1.4059008359909058, | |
| "logps/rejected": -1.826703429222107, | |
| "loss": 1.6602, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -14.05901050567627, | |
| "rewards/margins": 4.208024024963379, | |
| "rewards/rejected": -18.267032623291016, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.45958583834335337, | |
| "grad_norm": 115.44925865991426, | |
| "learning_rate": 6.545084971874736e-07, | |
| "logits/chosen": -1.1282669305801392, | |
| "logits/rejected": -1.1098558902740479, | |
| "logps/chosen": -1.4084670543670654, | |
| "logps/rejected": -1.8751609325408936, | |
| "loss": 1.5529, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -14.084672927856445, | |
| "rewards/margins": 4.666939735412598, | |
| "rewards/rejected": -18.751609802246094, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.47027388109552437, | |
| "grad_norm": 94.83729222797992, | |
| "learning_rate": 6.3661988065096e-07, | |
| "logits/chosen": -1.191811442375183, | |
| "logits/rejected": -1.1707171201705933, | |
| "logps/chosen": -1.4761518239974976, | |
| "logps/rejected": -1.9565551280975342, | |
| "loss": 1.5052, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -14.761518478393555, | |
| "rewards/margins": 4.804032325744629, | |
| "rewards/rejected": -19.5655517578125, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.48096192384769537, | |
| "grad_norm": 69.08600083744463, | |
| "learning_rate": 6.185401888577487e-07, | |
| "logits/chosen": -1.171081304550171, | |
| "logits/rejected": -1.1351138353347778, | |
| "logps/chosen": -1.4978052377700806, | |
| "logps/rejected": -1.93888258934021, | |
| "loss": 1.4742, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -14.978052139282227, | |
| "rewards/margins": 4.410772800445557, | |
| "rewards/rejected": -19.388826370239258, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.4916499665998664, | |
| "grad_norm": 72.13177261697588, | |
| "learning_rate": 6.002947078916364e-07, | |
| "logits/chosen": -1.245228886604309, | |
| "logits/rejected": -1.1923692226409912, | |
| "logps/chosen": -1.4302809238433838, | |
| "logps/rejected": -1.8505923748016357, | |
| "loss": 1.4317, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -14.302810668945312, | |
| "rewards/margins": 4.203113555908203, | |
| "rewards/rejected": -18.505924224853516, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5023380093520374, | |
| "grad_norm": 78.71892029667256, | |
| "learning_rate": 5.819089557075688e-07, | |
| "logits/chosen": -1.2821385860443115, | |
| "logits/rejected": -1.2519080638885498, | |
| "logps/chosen": -1.4527919292449951, | |
| "logps/rejected": -1.9279251098632812, | |
| "loss": 1.4134, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": -14.527920722961426, | |
| "rewards/margins": 4.751331329345703, | |
| "rewards/rejected": -19.279251098632812, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5130260521042084, | |
| "grad_norm": 87.04649782214463, | |
| "learning_rate": 5.634086464424742e-07, | |
| "logits/chosen": -1.245603322982788, | |
| "logits/rejected": -1.247234582901001, | |
| "logps/chosen": -1.3713314533233643, | |
| "logps/rejected": -1.8449758291244507, | |
| "loss": 1.4346, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -13.7133150100708, | |
| "rewards/margins": 4.736443042755127, | |
| "rewards/rejected": -18.449758529663086, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5237140948563794, | |
| "grad_norm": 120.6181547874012, | |
| "learning_rate": 5.448196544517167e-07, | |
| "logits/chosen": -1.3496326208114624, | |
| "logits/rejected": -1.2907614707946777, | |
| "logps/chosen": -1.440033197402954, | |
| "logps/rejected": -2.0060055255889893, | |
| "loss": 1.4071, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": -14.4003324508667, | |
| "rewards/margins": 5.659722328186035, | |
| "rewards/rejected": -20.060054779052734, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5344021376085505, | |
| "grad_norm": 127.16635817286267, | |
| "learning_rate": 5.26167978121472e-07, | |
| "logits/chosen": -1.2846823930740356, | |
| "logits/rejected": -1.2672080993652344, | |
| "logps/chosen": -1.5308005809783936, | |
| "logps/rejected": -2.1003577709198, | |
| "loss": 1.3208, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -15.308004379272461, | |
| "rewards/margins": 5.695572853088379, | |
| "rewards/rejected": -21.003578186035156, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5450901803607214, | |
| "grad_norm": 306.6500775815346, | |
| "learning_rate": 5.074797035076318e-07, | |
| "logits/chosen": -1.3492941856384277, | |
| "logits/rejected": -1.3214812278747559, | |
| "logps/chosen": -1.6527442932128906, | |
| "logps/rejected": -2.1238582134246826, | |
| "loss": 1.4957, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -16.527442932128906, | |
| "rewards/margins": 4.711141586303711, | |
| "rewards/rejected": -21.238582611083984, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5557782231128925, | |
| "grad_norm": 89.88872208917493, | |
| "learning_rate": 4.887809678520975e-07, | |
| "logits/chosen": -1.3080346584320068, | |
| "logits/rejected": -1.2766286134719849, | |
| "logps/chosen": -1.5681380033493042, | |
| "logps/rejected": -2.0582587718963623, | |
| "loss": 1.4255, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -15.681379318237305, | |
| "rewards/margins": 4.901208877563477, | |
| "rewards/rejected": -20.58258819580078, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5664662658650634, | |
| "grad_norm": 86.94866969630735, | |
| "learning_rate": 4.700979230274829e-07, | |
| "logits/chosen": -1.2753899097442627, | |
| "logits/rejected": -1.2558867931365967, | |
| "logps/chosen": -1.668534278869629, | |
| "logps/rejected": -2.181380271911621, | |
| "loss": 1.4204, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -16.68534278869629, | |
| "rewards/margins": 5.1284589767456055, | |
| "rewards/rejected": -21.813800811767578, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5771543086172345, | |
| "grad_norm": 144.84472573271995, | |
| "learning_rate": 4.514566989613559e-07, | |
| "logits/chosen": -1.2718496322631836, | |
| "logits/rejected": -1.2400305271148682, | |
| "logps/chosen": -1.4821763038635254, | |
| "logps/rejected": -2.010958194732666, | |
| "loss": 1.3379, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -14.82176399230957, | |
| "rewards/margins": 5.287820816040039, | |
| "rewards/rejected": -20.109582901000977, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5878423513694054, | |
| "grad_norm": 73.43309027045284, | |
| "learning_rate": 4.328833670911724e-07, | |
| "logits/chosen": -1.2501633167266846, | |
| "logits/rejected": -1.2101550102233887, | |
| "logps/chosen": -1.4654467105865479, | |
| "logps/rejected": -1.9191405773162842, | |
| "loss": 1.4963, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -14.654467582702637, | |
| "rewards/margins": 4.536937713623047, | |
| "rewards/rejected": -19.191404342651367, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5985303941215765, | |
| "grad_norm": 118.57888241178858, | |
| "learning_rate": 4.144039039010124e-07, | |
| "logits/chosen": -1.3355966806411743, | |
| "logits/rejected": -1.3093878030776978, | |
| "logps/chosen": -1.5047754049301147, | |
| "logps/rejected": -2.050473213195801, | |
| "loss": 1.3782, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -15.047755241394043, | |
| "rewards/margins": 5.456977844238281, | |
| "rewards/rejected": -20.50473403930664, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6092184368737475, | |
| "grad_norm": 100.79377019073691, | |
| "learning_rate": 3.960441545911204e-07, | |
| "logits/chosen": -1.307716965675354, | |
| "logits/rejected": -1.2712657451629639, | |
| "logps/chosen": -1.5421284437179565, | |
| "logps/rejected": -2.111297845840454, | |
| "loss": 1.1768, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": -15.421285629272461, | |
| "rewards/margins": 5.6916913986206055, | |
| "rewards/rejected": -21.112977981567383, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6199064796259185, | |
| "grad_norm": 102.72150408454053, | |
| "learning_rate": 3.778297969310529e-07, | |
| "logits/chosen": -1.333150863647461, | |
| "logits/rejected": -1.2860305309295654, | |
| "logps/chosen": -1.5572869777679443, | |
| "logps/rejected": -2.028750419616699, | |
| "loss": 1.3993, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -15.572871208190918, | |
| "rewards/margins": 4.714633464813232, | |
| "rewards/rejected": -20.287504196166992, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6305945223780896, | |
| "grad_norm": 96.32710532692002, | |
| "learning_rate": 3.5978630534699865e-07, | |
| "logits/chosen": -1.2499698400497437, | |
| "logits/rejected": -1.2331962585449219, | |
| "logps/chosen": -1.5715104341506958, | |
| "logps/rejected": -2.08168625831604, | |
| "loss": 1.2236, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -15.715105056762695, | |
| "rewards/margins": 5.101758003234863, | |
| "rewards/rejected": -20.81686019897461, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6412825651302605, | |
| "grad_norm": 86.57462147935358, | |
| "learning_rate": 3.4193891529348795e-07, | |
| "logits/chosen": -1.1837140321731567, | |
| "logits/rejected": -1.156435251235962, | |
| "logps/chosen": -1.662043809890747, | |
| "logps/rejected": -2.1141371726989746, | |
| "loss": 1.6291, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -16.620437622070312, | |
| "rewards/margins": 4.52093505859375, | |
| "rewards/rejected": -21.141372680664062, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6519706078824316, | |
| "grad_norm": 80.28274687652879, | |
| "learning_rate": 3.243125879593286e-07, | |
| "logits/chosen": -1.2831798791885376, | |
| "logits/rejected": -1.2358052730560303, | |
| "logps/chosen": -1.636275053024292, | |
| "logps/rejected": -2.093479871749878, | |
| "loss": 1.3641, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -16.362751007080078, | |
| "rewards/margins": 4.572048664093018, | |
| "rewards/rejected": -20.93480110168457, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6626586506346025, | |
| "grad_norm": 96.95154393343023, | |
| "learning_rate": 3.069319753571269e-07, | |
| "logits/chosen": -1.3118140697479248, | |
| "logits/rejected": -1.2903715372085571, | |
| "logps/chosen": -1.6529546976089478, | |
| "logps/rejected": -2.148355007171631, | |
| "loss": 1.4766, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -16.529544830322266, | |
| "rewards/margins": 4.954004764556885, | |
| "rewards/rejected": -21.483551025390625, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6733466933867736, | |
| "grad_norm": 91.51736686071692, | |
| "learning_rate": 2.898213858452173e-07, | |
| "logits/chosen": -1.3063311576843262, | |
| "logits/rejected": -1.2485519647598267, | |
| "logps/chosen": -1.6333932876586914, | |
| "logps/rejected": -2.1507859230041504, | |
| "loss": 1.3963, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -16.333934783935547, | |
| "rewards/margins": 5.173925876617432, | |
| "rewards/rejected": -21.507858276367188, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6840347361389446, | |
| "grad_norm": 97.07913178610919, | |
| "learning_rate": 2.730047501302266e-07, | |
| "logits/chosen": -1.2934232950210571, | |
| "logits/rejected": -1.2893450260162354, | |
| "logps/chosen": -1.6584867238998413, | |
| "logps/rejected": -2.2650082111358643, | |
| "loss": 1.3115, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": -16.584867477416992, | |
| "rewards/margins": 6.065215110778809, | |
| "rewards/rejected": -22.650081634521484, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6947227788911156, | |
| "grad_norm": 65.85264295945626, | |
| "learning_rate": 2.5650558779781635e-07, | |
| "logits/chosen": -1.326992392539978, | |
| "logits/rejected": -1.2749508619308472, | |
| "logps/chosen": -1.7087081670761108, | |
| "logps/rejected": -2.377331256866455, | |
| "loss": 1.307, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -17.087081909179688, | |
| "rewards/margins": 6.6862287521362305, | |
| "rewards/rejected": -23.773311614990234, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.7054108216432866, | |
| "grad_norm": 71.08108071468983, | |
| "learning_rate": 2.403469744184154e-07, | |
| "logits/chosen": -1.2321017980575562, | |
| "logits/rejected": -1.1879392862319946, | |
| "logps/chosen": -1.6843183040618896, | |
| "logps/rejected": -2.170222759246826, | |
| "loss": 1.3597, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -16.843183517456055, | |
| "rewards/margins": 4.859041690826416, | |
| "rewards/rejected": -21.702226638793945, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7160988643954576, | |
| "grad_norm": 85.80290375242986, | |
| "learning_rate": 2.2455150927394878e-07, | |
| "logits/chosen": -1.2848079204559326, | |
| "logits/rejected": -1.2643808126449585, | |
| "logps/chosen": -1.6716737747192383, | |
| "logps/rejected": -2.2179079055786133, | |
| "loss": 1.2118, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -16.716739654541016, | |
| "rewards/margins": 5.46234130859375, | |
| "rewards/rejected": -22.179079055786133, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7267869071476286, | |
| "grad_norm": 106.87884023285183, | |
| "learning_rate": 2.0914128375069722e-07, | |
| "logits/chosen": -1.3009603023529053, | |
| "logits/rejected": -1.2630964517593384, | |
| "logps/chosen": -1.5984188318252563, | |
| "logps/rejected": -2.146073579788208, | |
| "loss": 1.3799, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -15.984187126159668, | |
| "rewards/margins": 5.476546287536621, | |
| "rewards/rejected": -21.460735321044922, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7374749498997996, | |
| "grad_norm": 81.57738599240237, | |
| "learning_rate": 1.9413785044249676e-07, | |
| "logits/chosen": -1.3159044981002808, | |
| "logits/rejected": -1.2908227443695068, | |
| "logps/chosen": -1.6741054058074951, | |
| "logps/rejected": -2.3151228427886963, | |
| "loss": 1.4007, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -16.741056442260742, | |
| "rewards/margins": 6.4101715087890625, | |
| "rewards/rejected": -23.151227951049805, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7481629926519706, | |
| "grad_norm": 137.46788470613842, | |
| "learning_rate": 1.7956219300748792e-07, | |
| "logits/chosen": -1.3111270666122437, | |
| "logits/rejected": -1.3133299350738525, | |
| "logps/chosen": -1.55172860622406, | |
| "logps/rejected": -2.0665595531463623, | |
| "loss": 1.3291, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -15.51728630065918, | |
| "rewards/margins": 5.148309707641602, | |
| "rewards/rejected": -20.66559410095215, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7588510354041417, | |
| "grad_norm": 73.0747912837978, | |
| "learning_rate": 1.6543469682057104e-07, | |
| "logits/chosen": -1.2305195331573486, | |
| "logits/rejected": -1.2432688474655151, | |
| "logps/chosen": -1.5626884698867798, | |
| "logps/rejected": -2.1072001457214355, | |
| "loss": 1.159, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -15.626884460449219, | |
| "rewards/margins": 5.445114612579346, | |
| "rewards/rejected": -21.071998596191406, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7695390781563126, | |
| "grad_norm": 82.19549372560476, | |
| "learning_rate": 1.5177512046261666e-07, | |
| "logits/chosen": -1.2950479984283447, | |
| "logits/rejected": -1.2918254137039185, | |
| "logps/chosen": -1.5626431703567505, | |
| "logps/rejected": -2.192157030105591, | |
| "loss": 1.3653, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -15.626432418823242, | |
| "rewards/margins": 6.295140266418457, | |
| "rewards/rejected": -21.921573638916016, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7802271209084837, | |
| "grad_norm": 82.26556152038766, | |
| "learning_rate": 1.3860256808630427e-07, | |
| "logits/chosen": -1.3408092260360718, | |
| "logits/rejected": -1.2676836252212524, | |
| "logps/chosen": -1.621119737625122, | |
| "logps/rejected": -2.2568397521972656, | |
| "loss": 1.2936, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -16.211196899414062, | |
| "rewards/margins": 6.35720157623291, | |
| "rewards/rejected": -22.56839942932129, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7909151636606546, | |
| "grad_norm": 97.91298047906564, | |
| "learning_rate": 1.2593546269723647e-07, | |
| "logits/chosen": -1.2643686532974243, | |
| "logits/rejected": -1.2498524188995361, | |
| "logps/chosen": -1.578148603439331, | |
| "logps/rejected": -2.067432403564453, | |
| "loss": 1.3095, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -15.781486511230469, | |
| "rewards/margins": 4.892836093902588, | |
| "rewards/rejected": -20.6743221282959, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.8016032064128257, | |
| "grad_norm": 76.40375667456833, | |
| "learning_rate": 1.1379152038770029e-07, | |
| "logits/chosen": -1.2894870042800903, | |
| "logits/rejected": -1.2930238246917725, | |
| "logps/chosen": -1.7195911407470703, | |
| "logps/rejected": -2.293926954269409, | |
| "loss": 1.2661, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -17.195911407470703, | |
| "rewards/margins": 5.7433576583862305, | |
| "rewards/rejected": -22.939268112182617, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.8122912491649966, | |
| "grad_norm": 128.55014662844385, | |
| "learning_rate": 1.0218772555910954e-07, | |
| "logits/chosen": -1.3014891147613525, | |
| "logits/rejected": -1.2802826166152954, | |
| "logps/chosen": -1.5879671573638916, | |
| "logps/rejected": -2.113447666168213, | |
| "loss": 1.4202, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": -15.879669189453125, | |
| "rewards/margins": 5.25480842590332, | |
| "rewards/rejected": -21.134477615356445, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8229792919171677, | |
| "grad_norm": 77.57546829061782, | |
| "learning_rate": 9.114030716778432e-08, | |
| "logits/chosen": -1.310450792312622, | |
| "logits/rejected": -1.2848607301712036, | |
| "logps/chosen": -1.6349436044692993, | |
| "logps/rejected": -2.3224172592163086, | |
| "loss": 1.1354, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -16.349435806274414, | |
| "rewards/margins": 6.8747382164001465, | |
| "rewards/rejected": -23.224172592163086, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.8336673346693386, | |
| "grad_norm": 75.76498018298135, | |
| "learning_rate": 8.066471602728803e-08, | |
| "logits/chosen": -1.3069926500320435, | |
| "logits/rejected": -1.289568305015564, | |
| "logps/chosen": -1.699163794517517, | |
| "logps/rejected": -2.2884535789489746, | |
| "loss": 1.3105, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -16.99163818359375, | |
| "rewards/margins": 5.8928985595703125, | |
| "rewards/rejected": -22.884536743164062, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8443553774215097, | |
| "grad_norm": 74.3951066976334, | |
| "learning_rate": 7.077560319906694e-08, | |
| "logits/chosen": -1.3087493181228638, | |
| "logits/rejected": -1.2855933904647827, | |
| "logps/chosen": -1.6192277669906616, | |
| "logps/rejected": -2.1721370220184326, | |
| "loss": 1.2688, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -16.192277908325195, | |
| "rewards/margins": 5.5290937423706055, | |
| "rewards/rejected": -21.721370697021484, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8550434201736807, | |
| "grad_norm": 63.10639530225684, | |
| "learning_rate": 6.148679950161672e-08, | |
| "logits/chosen": -1.3169952630996704, | |
| "logits/rejected": -1.2985506057739258, | |
| "logps/chosen": -1.6467092037200928, | |
| "logps/rejected": -2.164301633834839, | |
| "loss": 1.2114, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -16.467090606689453, | |
| "rewards/margins": 5.17592716217041, | |
| "rewards/rejected": -21.643016815185547, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8550434201736807, | |
| "eval_logits/chosen": -1.4850261211395264, | |
| "eval_logits/rejected": -1.493988037109375, | |
| "eval_logps/chosen": -1.664995551109314, | |
| "eval_logps/rejected": -2.2206830978393555, | |
| "eval_loss": 1.2490928173065186, | |
| "eval_rewards/accuracies": 0.8414633870124817, | |
| "eval_rewards/chosen": -16.64995574951172, | |
| "eval_rewards/margins": 5.556875228881836, | |
| "eval_rewards/rejected": -22.206830978393555, | |
| "eval_runtime": 95.4555, | |
| "eval_samples_per_second": 20.544, | |
| "eval_steps_per_second": 1.289, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8657314629258517, | |
| "grad_norm": 102.60713365281785, | |
| "learning_rate": 5.2811296166831666e-08, | |
| "logits/chosen": -1.267327904701233, | |
| "logits/rejected": -1.2850300073623657, | |
| "logps/chosen": -1.7324796915054321, | |
| "logps/rejected": -2.2837843894958496, | |
| "loss": 1.2554, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -17.32479476928711, | |
| "rewards/margins": 5.513047218322754, | |
| "rewards/rejected": -22.83784294128418, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8764195056780227, | |
| "grad_norm": 159.07261192162792, | |
| "learning_rate": 4.4761226670592066e-08, | |
| "logits/chosen": -1.2908105850219727, | |
| "logits/rejected": -1.2769014835357666, | |
| "logps/chosen": -1.6668212413787842, | |
| "logps/rejected": -2.2075092792510986, | |
| "loss": 1.3804, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -16.668210983276367, | |
| "rewards/margins": 5.406882286071777, | |
| "rewards/rejected": -22.075092315673828, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8871075484301937, | |
| "grad_norm": 76.85488373819665, | |
| "learning_rate": 3.734784976300165e-08, | |
| "logits/chosen": -1.2890928983688354, | |
| "logits/rejected": -1.2320820093154907, | |
| "logps/chosen": -1.5973718166351318, | |
| "logps/rejected": -2.237947940826416, | |
| "loss": 1.4163, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -15.973716735839844, | |
| "rewards/margins": 6.40576171875, | |
| "rewards/rejected": -22.37947654724121, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.8977955911823647, | |
| "grad_norm": 92.42320617715352, | |
| "learning_rate": 3.058153372200695e-08, | |
| "logits/chosen": -1.3191107511520386, | |
| "logits/rejected": -1.2656759023666382, | |
| "logps/chosen": -1.5610657930374146, | |
| "logps/rejected": -2.152204990386963, | |
| "loss": 1.2658, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -15.610658645629883, | |
| "rewards/margins": 5.911390781402588, | |
| "rewards/rejected": -21.522048950195312, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9084836339345357, | |
| "grad_norm": 102.84147971960329, | |
| "learning_rate": 2.4471741852423233e-08, | |
| "logits/chosen": -1.3186463117599487, | |
| "logits/rejected": -1.3073859214782715, | |
| "logps/chosen": -1.736202597618103, | |
| "logps/rejected": -2.2703185081481934, | |
| "loss": 1.4248, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -17.362024307250977, | |
| "rewards/margins": 5.34116268157959, | |
| "rewards/rejected": -22.70318603515625, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9191716766867067, | |
| "grad_norm": 108.04777919102577, | |
| "learning_rate": 1.9027019250647036e-08, | |
| "logits/chosen": -1.2982522249221802, | |
| "logits/rejected": -1.2813619375228882, | |
| "logps/chosen": -1.7414271831512451, | |
| "logps/rejected": -2.3307671546936035, | |
| "loss": 1.2802, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -17.41427230834961, | |
| "rewards/margins": 5.893403053283691, | |
| "rewards/rejected": -23.30767250061035, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9298597194388778, | |
| "grad_norm": 85.6236171514638, | |
| "learning_rate": 1.4254980853566246e-08, | |
| "logits/chosen": -1.262458324432373, | |
| "logits/rejected": -1.2183687686920166, | |
| "logps/chosen": -1.5962882041931152, | |
| "logps/rejected": -2.182863712310791, | |
| "loss": 1.222, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": -15.962882995605469, | |
| "rewards/margins": 5.865753650665283, | |
| "rewards/rejected": -21.828638076782227, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9405477621910487, | |
| "grad_norm": 88.93173263482028, | |
| "learning_rate": 1.016230078838226e-08, | |
| "logits/chosen": -1.2786242961883545, | |
| "logits/rejected": -1.2167497873306274, | |
| "logps/chosen": -1.7170331478118896, | |
| "logps/rejected": -2.2510578632354736, | |
| "loss": 1.2694, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -17.170331954956055, | |
| "rewards/margins": 5.340245723724365, | |
| "rewards/rejected": -22.510578155517578, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9512358049432198, | |
| "grad_norm": 80.06878550984797, | |
| "learning_rate": 6.754703038239329e-09, | |
| "logits/chosen": -1.229853868484497, | |
| "logits/rejected": -1.2106773853302002, | |
| "logps/chosen": -1.688746690750122, | |
| "logps/rejected": -2.332123279571533, | |
| "loss": 1.1496, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -16.88746452331543, | |
| "rewards/margins": 6.433764457702637, | |
| "rewards/rejected": -23.321231842041016, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.9619238476953907, | |
| "grad_norm": 87.87225651237878, | |
| "learning_rate": 4.036953436716895e-09, | |
| "logits/chosen": -1.3426064252853394, | |
| "logits/rejected": -1.3202402591705322, | |
| "logps/chosen": -1.6350570917129517, | |
| "logps/rejected": -2.1853957176208496, | |
| "loss": 1.3199, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -16.350570678710938, | |
| "rewards/margins": 5.503388404846191, | |
| "rewards/rejected": -21.853958129882812, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9726118904475618, | |
| "grad_norm": 97.584405727653, | |
| "learning_rate": 2.0128530023804656e-09, | |
| "logits/chosen": -1.3094408512115479, | |
| "logits/rejected": -1.2717828750610352, | |
| "logps/chosen": -1.6554279327392578, | |
| "logps/rejected": -2.304875373840332, | |
| "loss": 1.0871, | |
| "rewards/accuracies": 0.9125000238418579, | |
| "rewards/chosen": -16.554279327392578, | |
| "rewards/margins": 6.494471549987793, | |
| "rewards/rejected": -23.048751831054688, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9832999331997327, | |
| "grad_norm": 94.99231466494224, | |
| "learning_rate": 6.852326227130833e-10, | |
| "logits/chosen": -1.3035484552383423, | |
| "logits/rejected": -1.2918545007705688, | |
| "logps/chosen": -1.7271077632904053, | |
| "logps/rejected": -2.327470541000366, | |
| "loss": 1.2419, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -17.27107810974121, | |
| "rewards/margins": 6.003628730773926, | |
| "rewards/rejected": -23.27470588684082, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9939879759519038, | |
| "grad_norm": 86.3036149732278, | |
| "learning_rate": 5.594909486328348e-11, | |
| "logits/chosen": -1.286787986755371, | |
| "logits/rejected": -1.2908694744110107, | |
| "logps/chosen": -1.7435226440429688, | |
| "logps/rejected": -2.3610475063323975, | |
| "loss": 1.3748, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -17.435226440429688, | |
| "rewards/margins": 6.175250053405762, | |
| "rewards/rejected": -23.610477447509766, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.9982631930527722, | |
| "step": 467, | |
| "total_flos": 0.0, | |
| "train_loss": 1.8360214427400707, | |
| "train_runtime": 11486.9698, | |
| "train_samples_per_second": 5.213, | |
| "train_steps_per_second": 0.041 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 467, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 1000000, | |
| "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": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |