| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9995419147961521, | |
| "eval_steps": 100, | |
| "global_step": 1091, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0009161704076958314, | |
| "grad_norm": 0.8362945914268494, | |
| "learning_rate": 9.09090909090909e-09, | |
| "logits/chosen": -0.33099836111068726, | |
| "logits/rejected": -0.6684360504150391, | |
| "logps/chosen": -201.79928588867188, | |
| "logps/rejected": -183.10525512695312, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.009161704076958314, | |
| "grad_norm": 0.7959509491920471, | |
| "learning_rate": 9.09090909090909e-08, | |
| "logits/chosen": -0.8238513469696045, | |
| "logits/rejected": -0.6410912275314331, | |
| "logps/chosen": -101.17930603027344, | |
| "logps/rejected": -201.2146759033203, | |
| "loss": 0.693, | |
| "rewards/accuracies": 0.5833333134651184, | |
| "rewards/chosen": 0.0013685817830264568, | |
| "rewards/margins": 0.003857892006635666, | |
| "rewards/rejected": -0.002489310223609209, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01832340815391663, | |
| "grad_norm": 0.791206419467926, | |
| "learning_rate": 1.818181818181818e-07, | |
| "logits/chosen": -0.696135401725769, | |
| "logits/rejected": -0.6421085000038147, | |
| "logps/chosen": -103.54312896728516, | |
| "logps/rejected": -195.77432250976562, | |
| "loss": 0.6932, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.0055698854848742485, | |
| "rewards/margins": -0.009762462228536606, | |
| "rewards/rejected": 0.004192576743662357, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.027485112230874943, | |
| "grad_norm": 0.8699631690979004, | |
| "learning_rate": 2.727272727272727e-07, | |
| "logits/chosen": -0.7419141530990601, | |
| "logits/rejected": -0.6500910520553589, | |
| "logps/chosen": -137.67446899414062, | |
| "logps/rejected": -262.87994384765625, | |
| "loss": 0.6912, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": 0.00041480286745354533, | |
| "rewards/margins": 0.004314146004617214, | |
| "rewards/rejected": -0.0038993440102785826, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03664681630783326, | |
| "grad_norm": 0.8166934847831726, | |
| "learning_rate": 3.636363636363636e-07, | |
| "logits/chosen": -0.7547220587730408, | |
| "logits/rejected": -0.5034279823303223, | |
| "logps/chosen": -94.3946762084961, | |
| "logps/rejected": -208.92074584960938, | |
| "loss": 0.6937, | |
| "rewards/accuracies": 0.44999998807907104, | |
| "rewards/chosen": -0.002515423344448209, | |
| "rewards/margins": -0.0004801487666554749, | |
| "rewards/rejected": -0.002035274403169751, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.04580852038479157, | |
| "grad_norm": 0.8770909309387207, | |
| "learning_rate": 4.545454545454545e-07, | |
| "logits/chosen": -0.7312489151954651, | |
| "logits/rejected": -0.521498441696167, | |
| "logps/chosen": -132.73512268066406, | |
| "logps/rejected": -239.8402862548828, | |
| "loss": 0.6924, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -7.614419155288488e-05, | |
| "rewards/margins": 0.0043979184702038765, | |
| "rewards/rejected": -0.004474063403904438, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.054970224461749886, | |
| "grad_norm": 0.8896933794021606, | |
| "learning_rate": 5.454545454545454e-07, | |
| "logits/chosen": -0.7862143516540527, | |
| "logits/rejected": -0.5695743560791016, | |
| "logps/chosen": -166.35006713867188, | |
| "logps/rejected": -254.9455108642578, | |
| "loss": 0.6945, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.003317029681056738, | |
| "rewards/margins": 0.011281955055892467, | |
| "rewards/rejected": -0.007964925840497017, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.0641319285387082, | |
| "grad_norm": 1.035460114479065, | |
| "learning_rate": 6.363636363636363e-07, | |
| "logits/chosen": -0.8119982481002808, | |
| "logits/rejected": -0.6650353670120239, | |
| "logps/chosen": -91.4764404296875, | |
| "logps/rejected": -195.1119842529297, | |
| "loss": 0.6915, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": 0.005646915640681982, | |
| "rewards/margins": 0.004837614018470049, | |
| "rewards/rejected": 0.000809302378911525, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.07329363261566652, | |
| "grad_norm": 1.1378659009933472, | |
| "learning_rate": 7.272727272727272e-07, | |
| "logits/chosen": -0.8396841287612915, | |
| "logits/rejected": -0.739294707775116, | |
| "logps/chosen": -128.26870727539062, | |
| "logps/rejected": -237.46743774414062, | |
| "loss": 0.688, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.007331106811761856, | |
| "rewards/margins": 0.0071782395243644714, | |
| "rewards/rejected": 0.0001528669090475887, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.08245533669262482, | |
| "grad_norm": 0.8795006275177002, | |
| "learning_rate": 8.181818181818182e-07, | |
| "logits/chosen": -0.7932317852973938, | |
| "logits/rejected": -0.634912371635437, | |
| "logps/chosen": -111.9941177368164, | |
| "logps/rejected": -242.5846710205078, | |
| "loss": 0.6868, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": 0.021863078698515892, | |
| "rewards/margins": 0.009898947551846504, | |
| "rewards/rejected": 0.011964133009314537, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.09161704076958314, | |
| "grad_norm": 0.9012777209281921, | |
| "learning_rate": 9.09090909090909e-07, | |
| "logits/chosen": -0.7805367708206177, | |
| "logits/rejected": -0.49868839979171753, | |
| "logps/chosen": -140.55088806152344, | |
| "logps/rejected": -290.98150634765625, | |
| "loss": 0.6806, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": 0.02105652168393135, | |
| "rewards/margins": 0.013599460944533348, | |
| "rewards/rejected": 0.00745706120505929, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.09161704076958314, | |
| "eval_logits/chosen": -0.7638165950775146, | |
| "eval_logits/rejected": -0.5878660082817078, | |
| "eval_logps/chosen": -139.51101684570312, | |
| "eval_logps/rejected": -264.5260009765625, | |
| "eval_loss": 0.6816309690475464, | |
| "eval_rewards/accuracies": 0.6445086598396301, | |
| "eval_rewards/chosen": 0.030344096943736076, | |
| "eval_rewards/margins": 0.02048710361123085, | |
| "eval_rewards/rejected": 0.009856995195150375, | |
| "eval_runtime": 253.9092, | |
| "eval_samples_per_second": 10.882, | |
| "eval_steps_per_second": 1.363, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.10077874484654145, | |
| "grad_norm": 1.0326433181762695, | |
| "learning_rate": 1e-06, | |
| "logits/chosen": -0.8220171928405762, | |
| "logits/rejected": -0.720903754234314, | |
| "logps/chosen": -186.0214385986328, | |
| "logps/rejected": -243.47702026367188, | |
| "loss": 0.6773, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.05213131755590439, | |
| "rewards/margins": 0.03893345221877098, | |
| "rewards/rejected": 0.013197864405810833, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.10994044892349977, | |
| "grad_norm": 1.0610991716384888, | |
| "learning_rate": 9.997436315234263e-07, | |
| "logits/chosen": -0.687503457069397, | |
| "logits/rejected": -0.6214867830276489, | |
| "logps/chosen": -119.43013000488281, | |
| "logps/rejected": -170.53787231445312, | |
| "loss": 0.6699, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": 0.09322745352983475, | |
| "rewards/margins": 0.058328378945589066, | |
| "rewards/rejected": 0.034899067133665085, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.11910215300045808, | |
| "grad_norm": 0.9824616312980652, | |
| "learning_rate": 9.989747889928883e-07, | |
| "logits/chosen": -0.7877914905548096, | |
| "logits/rejected": -0.6358338594436646, | |
| "logps/chosen": -128.0840606689453, | |
| "logps/rejected": -254.0060272216797, | |
| "loss": 0.656, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.102009616792202, | |
| "rewards/margins": 0.07630841434001923, | |
| "rewards/rejected": 0.02570119872689247, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.1282638570774164, | |
| "grad_norm": 1.0044574737548828, | |
| "learning_rate": 9.976942608363393e-07, | |
| "logits/chosen": -0.8087643384933472, | |
| "logits/rejected": -0.5704258680343628, | |
| "logps/chosen": -112.33183288574219, | |
| "logps/rejected": -218.6199951171875, | |
| "loss": 0.6455, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": 0.18401500582695007, | |
| "rewards/margins": 0.12449517101049423, | |
| "rewards/rejected": 0.05951983854174614, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.1374255611543747, | |
| "grad_norm": 1.028503179550171, | |
| "learning_rate": 9.9590336020199e-07, | |
| "logits/chosen": -0.7007342576980591, | |
| "logits/rejected": -0.5621221661567688, | |
| "logps/chosen": -114.86798095703125, | |
| "logps/rejected": -248.943603515625, | |
| "loss": 0.6348, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": 0.24283933639526367, | |
| "rewards/margins": 0.14451688528060913, | |
| "rewards/rejected": 0.09832244366407394, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.14658726523133303, | |
| "grad_norm": 0.8501743674278259, | |
| "learning_rate": 9.936039236117095e-07, | |
| "logits/chosen": -0.7554874420166016, | |
| "logits/rejected": -0.7505234479904175, | |
| "logps/chosen": -147.8970184326172, | |
| "logps/rejected": -204.843505859375, | |
| "loss": 0.6207, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 0.26698964834213257, | |
| "rewards/margins": 0.1363282948732376, | |
| "rewards/rejected": 0.13066136837005615, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.15574896930829135, | |
| "grad_norm": 0.8963977098464966, | |
| "learning_rate": 9.907983090777206e-07, | |
| "logits/chosen": -0.7809524536132812, | |
| "logits/rejected": -0.6654596924781799, | |
| "logps/chosen": -131.55320739746094, | |
| "logps/rejected": -207.1016082763672, | |
| "loss": 0.6215, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": 0.30690616369247437, | |
| "rewards/margins": 0.15369756519794464, | |
| "rewards/rejected": 0.15320861339569092, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.16491067338524965, | |
| "grad_norm": 0.9032484292984009, | |
| "learning_rate": 9.874893936845187e-07, | |
| "logits/chosen": -0.6498181819915771, | |
| "logits/rejected": -0.5487117767333984, | |
| "logps/chosen": -152.4180908203125, | |
| "logps/rejected": -292.68438720703125, | |
| "loss": 0.6068, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": 0.3112695813179016, | |
| "rewards/margins": 0.18446235358715057, | |
| "rewards/rejected": 0.12680718302726746, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.17407237746220797, | |
| "grad_norm": 0.7833600044250488, | |
| "learning_rate": 9.836805706384983e-07, | |
| "logits/chosen": -0.8308868408203125, | |
| "logits/rejected": -0.66489177942276, | |
| "logps/chosen": -97.64247131347656, | |
| "logps/rejected": -185.3607635498047, | |
| "loss": 0.5958, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": 0.47603702545166016, | |
| "rewards/margins": 0.26404884457588196, | |
| "rewards/rejected": 0.21198812127113342, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.1832340815391663, | |
| "grad_norm": 0.9265146255493164, | |
| "learning_rate": 9.793757457883061e-07, | |
| "logits/chosen": -0.8464080691337585, | |
| "logits/rejected": -0.6489285826683044, | |
| "logps/chosen": -81.71176147460938, | |
| "logps/rejected": -186.20758056640625, | |
| "loss": 0.5704, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.5865594148635864, | |
| "rewards/margins": 0.3980158567428589, | |
| "rewards/rejected": 0.18854358792304993, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.1832340815391663, | |
| "eval_logits/chosen": -0.7638057470321655, | |
| "eval_logits/rejected": -0.5881015062332153, | |
| "eval_logps/chosen": -133.12771606445312, | |
| "eval_logps/rejected": -261.77801513671875, | |
| "eval_loss": 0.5992826819419861, | |
| "eval_rewards/accuracies": 0.823699414730072, | |
| "eval_rewards/chosen": 0.34951016306877136, | |
| "eval_rewards/margins": 0.20225529372692108, | |
| "eval_rewards/rejected": 0.14725488424301147, | |
| "eval_runtime": 252.1498, | |
| "eval_samples_per_second": 10.958, | |
| "eval_steps_per_second": 1.372, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.1923957856161246, | |
| "grad_norm": 0.8853036165237427, | |
| "learning_rate": 9.745793336194975e-07, | |
| "logits/chosen": -0.674169659614563, | |
| "logits/rejected": -0.5945926904678345, | |
| "logps/chosen": -106.3974380493164, | |
| "logps/rejected": -215.2452850341797, | |
| "loss": 0.574, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 0.6075299978256226, | |
| "rewards/margins": 0.40666985511779785, | |
| "rewards/rejected": 0.20086011290550232, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.2015574896930829, | |
| "grad_norm": 0.8835374116897583, | |
| "learning_rate": 9.69296252727595e-07, | |
| "logits/chosen": -0.6831656098365784, | |
| "logits/rejected": -0.6844693422317505, | |
| "logps/chosen": -99.75518035888672, | |
| "logps/rejected": -196.4925994873047, | |
| "loss": 0.5604, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": 0.5753337740898132, | |
| "rewards/margins": 0.31991419196128845, | |
| "rewards/rejected": 0.25541964173316956, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.21071919377004122, | |
| "grad_norm": 0.8146816492080688, | |
| "learning_rate": 9.63531920774199e-07, | |
| "logits/chosen": -0.8509872555732727, | |
| "logits/rejected": -0.7722374796867371, | |
| "logps/chosen": -84.59940338134766, | |
| "logps/rejected": -179.6902618408203, | |
| "loss": 0.5335, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.8087629079818726, | |
| "rewards/margins": 0.43924540281295776, | |
| "rewards/rejected": 0.36951744556427, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.21988089784699955, | |
| "grad_norm": 0.7882716059684753, | |
| "learning_rate": 9.572922489313142e-07, | |
| "logits/chosen": -0.7912776470184326, | |
| "logits/rejected": -0.7742729187011719, | |
| "logps/chosen": -122.55216217041016, | |
| "logps/rejected": -203.78738403320312, | |
| "loss": 0.5186, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": 0.6593917608261108, | |
| "rewards/margins": 0.20427684485912323, | |
| "rewards/rejected": 0.4551149010658264, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.22904260192395787, | |
| "grad_norm": 0.778360903263092, | |
| "learning_rate": 9.505836358195993e-07, | |
| "logits/chosen": -0.8406602144241333, | |
| "logits/rejected": -0.6831453442573547, | |
| "logps/chosen": -95.79948425292969, | |
| "logps/rejected": -225.69931030273438, | |
| "loss": 0.5362, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": 0.8291288614273071, | |
| "rewards/margins": 0.5095328092575073, | |
| "rewards/rejected": 0.3195960223674774, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.23820430600091616, | |
| "grad_norm": 0.8873792290687561, | |
| "learning_rate": 9.434129609467483e-07, | |
| "logits/chosen": -0.6178407669067383, | |
| "logits/rejected": -0.6204096674919128, | |
| "logps/chosen": -194.87338256835938, | |
| "logps/rejected": -260.9487609863281, | |
| "loss": 0.5372, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 0.5945974588394165, | |
| "rewards/margins": 0.42120856046676636, | |
| "rewards/rejected": 0.17338886857032776, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.24736601007787448, | |
| "grad_norm": 0.833759069442749, | |
| "learning_rate": 9.357875776527333e-07, | |
| "logits/chosen": -0.7258036732673645, | |
| "logits/rejected": -0.6290136575698853, | |
| "logps/chosen": -110.7699966430664, | |
| "logps/rejected": -214.27734375, | |
| "loss": 0.5062, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 0.6229942440986633, | |
| "rewards/margins": 0.48382797837257385, | |
| "rewards/rejected": 0.13916628062725067, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.2565277141548328, | |
| "grad_norm": 0.7017952799797058, | |
| "learning_rate": 9.27715305569148e-07, | |
| "logits/chosen": -0.7022525072097778, | |
| "logits/rejected": -0.5847650766372681, | |
| "logps/chosen": -103.59033203125, | |
| "logps/rejected": -211.4336700439453, | |
| "loss": 0.5042, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 0.7977592349052429, | |
| "rewards/margins": 0.5717019438743591, | |
| "rewards/rejected": 0.2260572463274002, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.2656894182317911, | |
| "grad_norm": 0.8911390900611877, | |
| "learning_rate": 9.192044226003788e-07, | |
| "logits/chosen": -0.7471307516098022, | |
| "logits/rejected": -0.6816040277481079, | |
| "logps/chosen": -121.2009048461914, | |
| "logps/rejected": -197.65383911132812, | |
| "loss": 0.5119, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.9325343370437622, | |
| "rewards/margins": 0.6139361262321472, | |
| "rewards/rejected": 0.3185981810092926, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.2748511223087494, | |
| "grad_norm": 0.7208541631698608, | |
| "learning_rate": 9.102636564348294e-07, | |
| "logits/chosen": -0.7008062601089478, | |
| "logits/rejected": -0.6801473498344421, | |
| "logps/chosen": -114.77880859375, | |
| "logps/rejected": -210.2012176513672, | |
| "loss": 0.5032, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": 0.8345085978507996, | |
| "rewards/margins": 0.5057986378669739, | |
| "rewards/rejected": 0.3287099599838257, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.2748511223087494, | |
| "eval_logits/chosen": -0.7651150226593018, | |
| "eval_logits/rejected": -0.5892514586448669, | |
| "eval_logps/chosen": -128.52711486816406, | |
| "eval_logps/rejected": -261.1383056640625, | |
| "eval_loss": 0.5312727093696594, | |
| "eval_rewards/accuracies": 0.852601170539856, | |
| "eval_rewards/chosen": 0.5795396566390991, | |
| "eval_rewards/margins": 0.40029817819595337, | |
| "eval_rewards/rejected": 0.17924147844314575, | |
| "eval_runtime": 252.4578, | |
| "eval_samples_per_second": 10.944, | |
| "eval_steps_per_second": 1.371, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.28401282638570774, | |
| "grad_norm": 0.8867717385292053, | |
| "learning_rate": 9.009021755949051e-07, | |
| "logits/chosen": -0.7086865901947021, | |
| "logits/rejected": -0.7210379242897034, | |
| "logps/chosen": -109.81558990478516, | |
| "logps/rejected": -152.71084594726562, | |
| "loss": 0.5059, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.8528501391410828, | |
| "rewards/margins": 0.4039060175418854, | |
| "rewards/rejected": 0.4489441514015198, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.29317453046266606, | |
| "grad_norm": 0.9950509071350098, | |
| "learning_rate": 8.911295800349314e-07, | |
| "logits/chosen": -0.6988198757171631, | |
| "logits/rejected": -0.6140611171722412, | |
| "logps/chosen": -160.05844116210938, | |
| "logps/rejected": -236.6884307861328, | |
| "loss": 0.4987, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": 0.7523847818374634, | |
| "rewards/margins": 0.4543587267398834, | |
| "rewards/rejected": 0.2980259656906128, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.3023362345396244, | |
| "grad_norm": 0.9498128294944763, | |
| "learning_rate": 8.809558912966519e-07, | |
| "logits/chosen": -0.6651901006698608, | |
| "logits/rejected": -0.6055413484573364, | |
| "logps/chosen": -84.35199737548828, | |
| "logps/rejected": -208.5164794921875, | |
| "loss": 0.4501, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 0.9873196482658386, | |
| "rewards/margins": 0.7461606860160828, | |
| "rewards/rejected": 0.24115900695323944, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.3114979386165827, | |
| "grad_norm": 0.7728273272514343, | |
| "learning_rate": 8.703915422323984e-07, | |
| "logits/chosen": -0.5462226867675781, | |
| "logits/rejected": -0.46084174513816833, | |
| "logps/chosen": -129.86944580078125, | |
| "logps/rejected": -210.6329345703125, | |
| "loss": 0.4827, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": 0.8226272463798523, | |
| "rewards/margins": 0.4101462960243225, | |
| "rewards/rejected": 0.4124809801578522, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.320659642693541, | |
| "grad_norm": 0.6749230623245239, | |
| "learning_rate": 8.594473663064734e-07, | |
| "logits/chosen": -0.8189064264297485, | |
| "logits/rejected": -0.6820628046989441, | |
| "logps/chosen": -83.3464126586914, | |
| "logps/rejected": -180.37059020996094, | |
| "loss": 0.4865, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": 1.0249218940734863, | |
| "rewards/margins": 0.6868400573730469, | |
| "rewards/rejected": 0.3380817770957947, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.3298213467704993, | |
| "grad_norm": 0.917298436164856, | |
| "learning_rate": 8.481345864857146e-07, | |
| "logits/chosen": -0.7375888824462891, | |
| "logits/rejected": -0.5009106993675232, | |
| "logps/chosen": -119.41560363769531, | |
| "logps/rejected": -241.78866577148438, | |
| "loss": 0.4839, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": 0.7711573839187622, | |
| "rewards/margins": 0.5962369441986084, | |
| "rewards/rejected": 0.1749204695224762, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.3389830508474576, | |
| "grad_norm": 0.6750978827476501, | |
| "learning_rate": 8.36464803730636e-07, | |
| "logits/chosen": -0.8278859853744507, | |
| "logits/rejected": -0.761436402797699, | |
| "logps/chosen": -101.12157440185547, | |
| "logps/rejected": -170.44569396972656, | |
| "loss": 0.4589, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 1.0102306604385376, | |
| "rewards/margins": 0.6717247366905212, | |
| "rewards/rejected": 0.33850589394569397, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.34814475492441593, | |
| "grad_norm": 0.8837040066719055, | |
| "learning_rate": 8.244499850989451e-07, | |
| "logits/chosen": -0.8623167276382446, | |
| "logits/rejected": -0.699422299861908, | |
| "logps/chosen": -77.54528045654297, | |
| "logps/rejected": -172.49044799804688, | |
| "loss": 0.456, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": 1.0262943506240845, | |
| "rewards/margins": 0.7692972421646118, | |
| "rewards/rejected": 0.25699710845947266, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.35730645900137425, | |
| "grad_norm": 0.6525785326957703, | |
| "learning_rate": 8.121024514736377e-07, | |
| "logits/chosen": -0.7333749532699585, | |
| "logits/rejected": -0.5934922099113464, | |
| "logps/chosen": -71.39646911621094, | |
| "logps/rejected": -198.8276824951172, | |
| "loss": 0.425, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 0.9156519770622253, | |
| "rewards/margins": 0.7522441148757935, | |
| "rewards/rejected": 0.16340789198875427, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.3664681630783326, | |
| "grad_norm": 0.795269250869751, | |
| "learning_rate": 7.994348649282532e-07, | |
| "logits/chosen": -0.7343798279762268, | |
| "logits/rejected": -0.5222686529159546, | |
| "logps/chosen": -116.50679779052734, | |
| "logps/rejected": -223.85409545898438, | |
| "loss": 0.4548, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.9559024572372437, | |
| "rewards/margins": 0.7852972745895386, | |
| "rewards/rejected": 0.17060522735118866, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.3664681630783326, | |
| "eval_logits/chosen": -0.7659581303596497, | |
| "eval_logits/rejected": -0.5900666117668152, | |
| "eval_logps/chosen": -127.30506896972656, | |
| "eval_logps/rejected": -263.6780090332031, | |
| "eval_loss": 0.47267913818359375, | |
| "eval_rewards/accuracies": 0.884393036365509, | |
| "eval_rewards/chosen": 0.6406417489051819, | |
| "eval_rewards/margins": 0.5883842706680298, | |
| "eval_rewards/rejected": 0.05225752666592598, | |
| "eval_runtime": 252.1281, | |
| "eval_samples_per_second": 10.959, | |
| "eval_steps_per_second": 1.372, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.3756298671552909, | |
| "grad_norm": 0.6601365208625793, | |
| "learning_rate": 7.8646021574225e-07, | |
| "logits/chosen": -0.6952235102653503, | |
| "logits/rejected": -0.46782374382019043, | |
| "logps/chosen": -104.0670394897461, | |
| "logps/rejected": -239.7484130859375, | |
| "loss": 0.4482, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 0.9586408734321594, | |
| "rewards/margins": 0.7240998148918152, | |
| "rewards/rejected": 0.23454102873802185, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.3847915712322492, | |
| "grad_norm": 0.7361147999763489, | |
| "learning_rate": 7.731918090798113e-07, | |
| "logits/chosen": -0.8017680048942566, | |
| "logits/rejected": -0.6514943838119507, | |
| "logps/chosen": -97.84058380126953, | |
| "logps/rejected": -181.79403686523438, | |
| "loss": 0.4194, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 1.1092697381973267, | |
| "rewards/margins": 0.7227686047554016, | |
| "rewards/rejected": 0.3865012228488922, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.39395327530920754, | |
| "grad_norm": 0.8337231278419495, | |
| "learning_rate": 7.596432513457482e-07, | |
| "logits/chosen": -0.7028661966323853, | |
| "logits/rejected": -0.6256899833679199, | |
| "logps/chosen": -96.13346862792969, | |
| "logps/rejected": -165.75747680664062, | |
| "loss": 0.426, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": 1.0767663717269897, | |
| "rewards/margins": 0.7708722949028015, | |
| "rewards/rejected": 0.30589404702186584, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.4031149793861658, | |
| "grad_norm": 0.826393723487854, | |
| "learning_rate": 7.458284362324842e-07, | |
| "logits/chosen": -0.7026184797286987, | |
| "logits/rejected": -0.5175060033798218, | |
| "logps/chosen": -73.34812927246094, | |
| "logps/rejected": -228.7108917236328, | |
| "loss": 0.4312, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": 1.0083404779434204, | |
| "rewards/margins": 0.9987894296646118, | |
| "rewards/rejected": 0.009551113471388817, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.4122766834631241, | |
| "grad_norm": 0.7044612765312195, | |
| "learning_rate": 7.317615304724387e-07, | |
| "logits/chosen": -0.6982114911079407, | |
| "logits/rejected": -0.6542345881462097, | |
| "logps/chosen": -97.9085922241211, | |
| "logps/rejected": -188.80284118652344, | |
| "loss": 0.4191, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 1.0013755559921265, | |
| "rewards/margins": 0.9130918383598328, | |
| "rewards/rejected": 0.08828376233577728, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.42143838754008245, | |
| "grad_norm": 0.8462466597557068, | |
| "learning_rate": 7.174569593104108e-07, | |
| "logits/chosen": -0.8011992573738098, | |
| "logits/rejected": -0.7858909368515015, | |
| "logps/chosen": -111.00840759277344, | |
| "logps/rejected": -198.2548370361328, | |
| "loss": 0.4307, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": 1.0858404636383057, | |
| "rewards/margins": 0.7915989756584167, | |
| "rewards/rejected": 0.29424166679382324, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.43060009161704077, | |
| "grad_norm": 1.0214639902114868, | |
| "learning_rate": 7.029293917108677e-07, | |
| "logits/chosen": -0.6210706233978271, | |
| "logits/rejected": -0.5300668478012085, | |
| "logps/chosen": -198.18032836914062, | |
| "logps/rejected": -255.27877807617188, | |
| "loss": 0.4096, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.7058089971542358, | |
| "rewards/margins": 0.7782974243164062, | |
| "rewards/rejected": -0.0724884420633316, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.4397617956939991, | |
| "grad_norm": 0.8063747882843018, | |
| "learning_rate": 6.881937253153051e-07, | |
| "logits/chosen": -0.8666943311691284, | |
| "logits/rejected": -0.8363106846809387, | |
| "logps/chosen": -119.9945068359375, | |
| "logps/rejected": -200.77627563476562, | |
| "loss": 0.4004, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": 1.2585500478744507, | |
| "rewards/margins": 1.000775694847107, | |
| "rewards/rejected": 0.25777435302734375, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.4489234997709574, | |
| "grad_norm": 0.7443576455116272, | |
| "learning_rate": 6.732650711651031e-07, | |
| "logits/chosen": -0.7361257076263428, | |
| "logits/rejected": -0.5734608769416809, | |
| "logps/chosen": -127.10151672363281, | |
| "logps/rejected": -238.904296875, | |
| "loss": 0.4098, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 0.6828312873840332, | |
| "rewards/margins": 0.7480920553207397, | |
| "rewards/rejected": -0.06526081264019012, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.45808520384791573, | |
| "grad_norm": 0.6595803499221802, | |
| "learning_rate": 6.581587382055491e-07, | |
| "logits/chosen": -0.8137266039848328, | |
| "logits/rejected": -0.7571356296539307, | |
| "logps/chosen": -104.3070297241211, | |
| "logps/rejected": -201.35647583007812, | |
| "loss": 0.3823, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": 1.131914496421814, | |
| "rewards/margins": 1.0377442836761475, | |
| "rewards/rejected": 0.09417015314102173, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.45808520384791573, | |
| "eval_logits/chosen": -0.7672035098075867, | |
| "eval_logits/rejected": -0.591365396976471, | |
| "eval_logps/chosen": -127.29341125488281, | |
| "eval_logps/rejected": -267.3507080078125, | |
| "eval_loss": 0.4234526753425598, | |
| "eval_rewards/accuracies": 0.8930636048316956, | |
| "eval_rewards/chosen": 0.6412254571914673, | |
| "eval_rewards/margins": 0.7726045250892639, | |
| "eval_rewards/rejected": -0.13137909770011902, | |
| "eval_runtime": 252.2594, | |
| "eval_samples_per_second": 10.953, | |
| "eval_steps_per_second": 1.372, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.467246907924874, | |
| "grad_norm": 0.7072412371635437, | |
| "learning_rate": 6.428902175869126e-07, | |
| "logits/chosen": -0.7340083122253418, | |
| "logits/rejected": -0.6606656908988953, | |
| "logps/chosen": -126.58953857421875, | |
| "logps/rejected": -193.69363403320312, | |
| "loss": 0.3779, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 1.1136077642440796, | |
| "rewards/margins": 1.0119339227676392, | |
| "rewards/rejected": 0.10167402029037476, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.4764086120018323, | |
| "grad_norm": 0.7274028658866882, | |
| "learning_rate": 6.274751667786761e-07, | |
| "logits/chosen": -0.6646409630775452, | |
| "logits/rejected": -0.5588616728782654, | |
| "logps/chosen": -177.66925048828125, | |
| "logps/rejected": -301.41876220703125, | |
| "loss": 0.4, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 0.8406426310539246, | |
| "rewards/margins": 0.9421254992485046, | |
| "rewards/rejected": -0.10148291289806366, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.48557031607879064, | |
| "grad_norm": 0.8565487861633301, | |
| "learning_rate": 6.119293935132075e-07, | |
| "logits/chosen": -0.7664127349853516, | |
| "logits/rejected": -0.6106132864952087, | |
| "logps/chosen": -105.25531005859375, | |
| "logps/rejected": -184.4231414794922, | |
| "loss": 0.3778, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": 0.9561585187911987, | |
| "rewards/margins": 0.908736526966095, | |
| "rewards/rejected": 0.04742204025387764, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.49473202015574896, | |
| "grad_norm": 0.8583614826202393, | |
| "learning_rate": 5.962688395753437e-07, | |
| "logits/chosen": -0.9779294729232788, | |
| "logits/rejected": -0.892353355884552, | |
| "logps/chosen": -84.86984252929688, | |
| "logps/rejected": -183.84295654296875, | |
| "loss": 0.3823, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 0.9522911310195923, | |
| "rewards/margins": 0.9800866842269897, | |
| "rewards/rejected": -0.027795499190688133, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.5038937242327073, | |
| "grad_norm": 0.9459156394004822, | |
| "learning_rate": 5.80509564454506e-07, | |
| "logits/chosen": -0.7560229897499084, | |
| "logits/rejected": -0.6336179971694946, | |
| "logps/chosen": -60.79301834106445, | |
| "logps/rejected": -198.7034912109375, | |
| "loss": 0.3682, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": 1.1121090650558472, | |
| "rewards/margins": 0.9977458715438843, | |
| "rewards/rejected": 0.11436320841312408, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.5130554283096656, | |
| "grad_norm": 0.7807260155677795, | |
| "learning_rate": 5.646677288761132e-07, | |
| "logits/chosen": -0.7362383604049683, | |
| "logits/rejected": -0.6392463445663452, | |
| "logps/chosen": -94.71185302734375, | |
| "logps/rejected": -229.9622802734375, | |
| "loss": 0.3765, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 0.8846451640129089, | |
| "rewards/margins": 1.0037345886230469, | |
| "rewards/rejected": -0.11908946186304092, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.5222171323866239, | |
| "grad_norm": 0.9470974802970886, | |
| "learning_rate": 5.487595782291784e-07, | |
| "logits/chosen": -0.807681679725647, | |
| "logits/rejected": -0.6749477982521057, | |
| "logps/chosen": -118.90516662597656, | |
| "logps/rejected": -185.31967163085938, | |
| "loss": 0.3591, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": 1.0300986766815186, | |
| "rewards/margins": 1.0025510787963867, | |
| "rewards/rejected": 0.02754763886332512, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.5313788364635822, | |
| "grad_norm": 0.776768684387207, | |
| "learning_rate": 5.328014259070878e-07, | |
| "logits/chosen": -0.7403959035873413, | |
| "logits/rejected": -0.5926883816719055, | |
| "logps/chosen": -135.42613220214844, | |
| "logps/rejected": -229.9550323486328, | |
| "loss": 0.3776, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": 0.7646117806434631, | |
| "rewards/margins": 0.9938881993293762, | |
| "rewards/rejected": -0.22927632927894592, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.5405405405405406, | |
| "grad_norm": 0.890230119228363, | |
| "learning_rate": 5.168096365786402e-07, | |
| "logits/chosen": -0.8721317052841187, | |
| "logits/rejected": -0.6174881458282471, | |
| "logps/chosen": -113.02799987792969, | |
| "logps/rejected": -223.82821655273438, | |
| "loss": 0.3664, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 0.9226465225219727, | |
| "rewards/margins": 1.041733980178833, | |
| "rewards/rejected": -0.11908741295337677, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.5497022446174988, | |
| "grad_norm": 0.7550140023231506, | |
| "learning_rate": 5.008006094065069e-07, | |
| "logits/chosen": -0.6947864294052124, | |
| "logits/rejected": -0.6369723081588745, | |
| "logps/chosen": -108.70173645019531, | |
| "logps/rejected": -202.474365234375, | |
| "loss": 0.3513, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": 0.9735838174819946, | |
| "rewards/margins": 0.9605382680892944, | |
| "rewards/rejected": 0.013045459985733032, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.5497022446174988, | |
| "eval_logits/chosen": -0.7693285346031189, | |
| "eval_logits/rejected": -0.5936357975006104, | |
| "eval_logps/chosen": -127.94479370117188, | |
| "eval_logps/rejected": -271.55316162109375, | |
| "eval_loss": 0.3842833936214447, | |
| "eval_rewards/accuracies": 0.913294792175293, | |
| "eval_rewards/chosen": 0.608656644821167, | |
| "eval_rewards/margins": 0.9501577615737915, | |
| "eval_rewards/rejected": -0.3415011465549469, | |
| "eval_runtime": 252.1964, | |
| "eval_samples_per_second": 10.956, | |
| "eval_steps_per_second": 1.372, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.5588639486944572, | |
| "grad_norm": 0.7903901934623718, | |
| "learning_rate": 4.847907612303182e-07, | |
| "logits/chosen": -0.7707523107528687, | |
| "logits/rejected": -0.6204859614372253, | |
| "logps/chosen": -131.9110870361328, | |
| "logps/rejected": -249.21353149414062, | |
| "loss": 0.3699, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": 0.5509034991264343, | |
| "rewards/margins": 0.738714337348938, | |
| "rewards/rejected": -0.18781085312366486, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.5680256527714155, | |
| "grad_norm": 0.7700456380844116, | |
| "learning_rate": 4.687965097316223e-07, | |
| "logits/chosen": -0.8288325071334839, | |
| "logits/rejected": -0.6177199482917786, | |
| "logps/chosen": -75.93597412109375, | |
| "logps/rejected": -225.97067260742188, | |
| "loss": 0.3304, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 1.001242756843567, | |
| "rewards/margins": 1.3316209316253662, | |
| "rewards/rejected": -0.33037835359573364, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.5771873568483737, | |
| "grad_norm": 0.9662141799926758, | |
| "learning_rate": 4.5283425659798175e-07, | |
| "logits/chosen": -0.8291882276535034, | |
| "logits/rejected": -0.8248087167739868, | |
| "logps/chosen": -143.07180786132812, | |
| "logps/rejected": -264.81463623046875, | |
| "loss": 0.3757, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": 0.9661464691162109, | |
| "rewards/margins": 1.1662724018096924, | |
| "rewards/rejected": -0.20012597739696503, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.5863490609253321, | |
| "grad_norm": 0.7089081406593323, | |
| "learning_rate": 4.3692037070347123e-07, | |
| "logits/chosen": -0.8086493611335754, | |
| "logits/rejected": -0.6815121173858643, | |
| "logps/chosen": -89.39459228515625, | |
| "logps/rejected": -196.5904541015625, | |
| "loss": 0.3598, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 0.9766321182250977, | |
| "rewards/margins": 1.233741044998169, | |
| "rewards/rejected": -0.2571088671684265, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.5955107650022904, | |
| "grad_norm": 0.6838498115539551, | |
| "learning_rate": 4.21071171322823e-07, | |
| "logits/chosen": -0.6819313764572144, | |
| "logits/rejected": -0.649137556552887, | |
| "logps/chosen": -202.49642944335938, | |
| "logps/rejected": -305.983154296875, | |
| "loss": 0.358, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.5383878946304321, | |
| "rewards/margins": 0.8751532435417175, | |
| "rewards/rejected": -0.3367653489112854, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.6046724690792488, | |
| "grad_norm": 0.8088337779045105, | |
| "learning_rate": 4.0530291139643755e-07, | |
| "logits/chosen": -0.8444887399673462, | |
| "logits/rejected": -0.7773497700691223, | |
| "logps/chosen": -92.40638732910156, | |
| "logps/rejected": -210.07962036132812, | |
| "loss": 0.3479, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 1.1181201934814453, | |
| "rewards/margins": 1.1484359502792358, | |
| "rewards/rejected": -0.030315812677145004, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.613834173156207, | |
| "grad_norm": 0.743868350982666, | |
| "learning_rate": 3.8963176086341727e-07, | |
| "logits/chosen": -0.7523199319839478, | |
| "logits/rejected": -0.545455813407898, | |
| "logps/chosen": -94.30660247802734, | |
| "logps/rejected": -229.30667114257812, | |
| "loss": 0.3394, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": 0.5909605026245117, | |
| "rewards/margins": 1.0719151496887207, | |
| "rewards/rejected": -0.4809547960758209, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.6229958772331654, | |
| "grad_norm": 0.7818168997764587, | |
| "learning_rate": 3.7407379007971506e-07, | |
| "logits/chosen": -0.6859625577926636, | |
| "logits/rejected": -0.5579516887664795, | |
| "logps/chosen": -143.53274536132812, | |
| "logps/rejected": -291.33453369140625, | |
| "loss": 0.3549, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 0.8891765475273132, | |
| "rewards/margins": 1.2517534494400024, | |
| "rewards/rejected": -0.36257684230804443, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.6321575813101237, | |
| "grad_norm": 0.8612805604934692, | |
| "learning_rate": 3.586449533384048e-07, | |
| "logits/chosen": -0.6445218324661255, | |
| "logits/rejected": -0.5638427138328552, | |
| "logps/chosen": -84.69303894042969, | |
| "logps/rejected": -175.0728302001953, | |
| "loss": 0.3413, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 0.8365532159805298, | |
| "rewards/margins": 1.054081916809082, | |
| "rewards/rejected": -0.21752884984016418, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.641319285387082, | |
| "grad_norm": 0.6171064972877502, | |
| "learning_rate": 3.433610725089692e-07, | |
| "logits/chosen": -0.7035859823226929, | |
| "logits/rejected": -0.5256271362304688, | |
| "logps/chosen": -138.380615234375, | |
| "logps/rejected": -248.87451171875, | |
| "loss": 0.3444, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.8877826929092407, | |
| "rewards/margins": 1.0520169734954834, | |
| "rewards/rejected": -0.16423439979553223, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.641319285387082, | |
| "eval_logits/chosen": -0.7720704078674316, | |
| "eval_logits/rejected": -0.5964747667312622, | |
| "eval_logps/chosen": -128.3763427734375, | |
| "eval_logps/rejected": -274.77838134765625, | |
| "eval_loss": 0.35713571310043335, | |
| "eval_rewards/accuracies": 0.910404622554779, | |
| "eval_rewards/chosen": 0.5870785117149353, | |
| "eval_rewards/margins": 1.0898393392562866, | |
| "eval_rewards/rejected": -0.5027608871459961, | |
| "eval_runtime": 252.0905, | |
| "eval_samples_per_second": 10.96, | |
| "eval_steps_per_second": 1.373, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.6504809894640403, | |
| "grad_norm": 0.8962298631668091, | |
| "learning_rate": 3.2823782081238555e-07, | |
| "logits/chosen": -0.7447826266288757, | |
| "logits/rejected": -0.7338870167732239, | |
| "logps/chosen": -95.43080139160156, | |
| "logps/rejected": -192.2973175048828, | |
| "loss": 0.3405, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 1.0037095546722412, | |
| "rewards/margins": 1.1646350622177124, | |
| "rewards/rejected": -0.16092538833618164, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.6596426935409986, | |
| "grad_norm": 0.7918722033500671, | |
| "learning_rate": 3.132907067486471e-07, | |
| "logits/chosen": -0.7636629343032837, | |
| "logits/rejected": -0.6649468541145325, | |
| "logps/chosen": -114.23177337646484, | |
| "logps/rejected": -215.4693145751953, | |
| "loss": 0.3439, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": 0.6278459429740906, | |
| "rewards/margins": 0.8962922096252441, | |
| "rewards/rejected": -0.26844626665115356, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.668804397617957, | |
| "grad_norm": 0.8416666388511658, | |
| "learning_rate": 2.985350581932005e-07, | |
| "logits/chosen": -0.776689350605011, | |
| "logits/rejected": -0.6382275819778442, | |
| "logps/chosen": -142.0974884033203, | |
| "logps/rejected": -211.6706085205078, | |
| "loss": 0.3482, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": 1.0184391736984253, | |
| "rewards/margins": 1.137699842453003, | |
| "rewards/rejected": -0.11926088482141495, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.6779661016949152, | |
| "grad_norm": 0.8943964242935181, | |
| "learning_rate": 2.839860066786103e-07, | |
| "logits/chosen": -0.7382106781005859, | |
| "logits/rejected": -0.6326214075088501, | |
| "logps/chosen": -84.48265075683594, | |
| "logps/rejected": -175.4539031982422, | |
| "loss": 0.3261, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 1.1386158466339111, | |
| "rewards/margins": 1.2899872064590454, | |
| "rewards/rejected": -0.1513712853193283, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.6871278057718736, | |
| "grad_norm": 0.8377578258514404, | |
| "learning_rate": 2.6965847187756553e-07, | |
| "logits/chosen": -0.7496690154075623, | |
| "logits/rejected": -0.6893197298049927, | |
| "logps/chosen": -96.63633728027344, | |
| "logps/rejected": -162.9729461669922, | |
| "loss": 0.3157, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 1.0216400623321533, | |
| "rewards/margins": 1.3754501342773438, | |
| "rewards/rejected": -0.3538101315498352, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.6962895098488319, | |
| "grad_norm": 0.8681157827377319, | |
| "learning_rate": 2.5556714630314613e-07, | |
| "logits/chosen": -0.8700932264328003, | |
| "logits/rejected": -0.7129749059677124, | |
| "logps/chosen": -68.58832550048828, | |
| "logps/rejected": -177.4335479736328, | |
| "loss": 0.3173, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 1.1690860986709595, | |
| "rewards/margins": 1.593080759048462, | |
| "rewards/rejected": -0.42399463057518005, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.7054512139257902, | |
| "grad_norm": 0.7554333209991455, | |
| "learning_rate": 2.417264802420343e-07, | |
| "logits/chosen": -0.8102506399154663, | |
| "logits/rejected": -0.6963295936584473, | |
| "logps/chosen": -94.93528747558594, | |
| "logps/rejected": -227.2413787841797, | |
| "loss": 0.3225, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": 0.9498472213745117, | |
| "rewards/margins": 1.3431434631347656, | |
| "rewards/rejected": -0.39329633116722107, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.7146129180027485, | |
| "grad_norm": 0.9492266178131104, | |
| "learning_rate": 2.2815066693612117e-07, | |
| "logits/chosen": -0.7795412540435791, | |
| "logits/rejected": -0.6747149229049683, | |
| "logps/chosen": -101.27651977539062, | |
| "logps/rejected": -204.02151489257812, | |
| "loss": 0.321, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 1.01050865650177, | |
| "rewards/margins": 1.2811126708984375, | |
| "rewards/rejected": -0.2706039547920227, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.7237746220797068, | |
| "grad_norm": 0.8731204867362976, | |
| "learning_rate": 2.1485362802770862e-07, | |
| "logits/chosen": -0.7779867053031921, | |
| "logits/rejected": -0.6321139335632324, | |
| "logps/chosen": -139.72247314453125, | |
| "logps/rejected": -309.1371154785156, | |
| "loss": 0.3098, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 0.7068808078765869, | |
| "rewards/margins": 1.2700955867767334, | |
| "rewards/rejected": -0.563214898109436, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.7329363261566652, | |
| "grad_norm": 0.9384542107582092, | |
| "learning_rate": 2.018489992832283e-07, | |
| "logits/chosen": -0.7126377820968628, | |
| "logits/rejected": -0.5895640850067139, | |
| "logps/chosen": -134.41358947753906, | |
| "logps/rejected": -231.8523712158203, | |
| "loss": 0.3486, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": 0.8817362785339355, | |
| "rewards/margins": 1.268080234527588, | |
| "rewards/rejected": -0.3863441050052643, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.7329363261566652, | |
| "eval_logits/chosen": -0.7724946141242981, | |
| "eval_logits/rejected": -0.5970612168312073, | |
| "eval_logps/chosen": -128.75588989257812, | |
| "eval_logps/rejected": -277.0340576171875, | |
| "eval_loss": 0.3426840603351593, | |
| "eval_rewards/accuracies": 0.910404622554779, | |
| "eval_rewards/chosen": 0.5681010484695435, | |
| "eval_rewards/margins": 1.1836459636688232, | |
| "eval_rewards/rejected": -0.6155449151992798, | |
| "eval_runtime": 251.9975, | |
| "eval_samples_per_second": 10.964, | |
| "eval_steps_per_second": 1.373, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.7420980302336234, | |
| "grad_norm": 0.9169598817825317, | |
| "learning_rate": 1.891501166101187e-07, | |
| "logits/chosen": -0.7657705545425415, | |
| "logits/rejected": -0.6736264228820801, | |
| "logps/chosen": -82.07408905029297, | |
| "logps/rejected": -185.07350158691406, | |
| "loss": 0.3199, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 0.8503938913345337, | |
| "rewards/margins": 1.3659098148345947, | |
| "rewards/rejected": -0.515515923500061, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.7512597343105818, | |
| "grad_norm": 0.8475902080535889, | |
| "learning_rate": 1.767700023812e-07, | |
| "logits/chosen": -0.8032963871955872, | |
| "logits/rejected": -0.5944005846977234, | |
| "logps/chosen": -111.13800048828125, | |
| "logps/rejected": -270.47247314453125, | |
| "loss": 0.3154, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 0.8725293278694153, | |
| "rewards/margins": 1.7224490642547607, | |
| "rewards/rejected": -0.8499197959899902, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.7604214383875401, | |
| "grad_norm": 0.9381293058395386, | |
| "learning_rate": 1.6472135208057125e-07, | |
| "logits/chosen": -0.78323894739151, | |
| "logits/rejected": -0.6326896548271179, | |
| "logps/chosen": -106.88468170166016, | |
| "logps/rejected": -183.2010498046875, | |
| "loss": 0.3454, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.8930851221084595, | |
| "rewards/margins": 1.1470720767974854, | |
| "rewards/rejected": -0.25398701429367065, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.7695831424644984, | |
| "grad_norm": 0.8112012147903442, | |
| "learning_rate": 1.530165212847217e-07, | |
| "logits/chosen": -0.7700347900390625, | |
| "logits/rejected": -0.7001415491104126, | |
| "logps/chosen": -93.79769134521484, | |
| "logps/rejected": -191.95217895507812, | |
| "loss": 0.3337, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 1.0522786378860474, | |
| "rewards/margins": 1.627384901046753, | |
| "rewards/rejected": -0.5751065015792847, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.7787448465414567, | |
| "grad_norm": 0.674938976764679, | |
| "learning_rate": 1.4166751299221003e-07, | |
| "logits/chosen": -0.70830899477005, | |
| "logits/rejected": -0.6749675273895264, | |
| "logps/chosen": -106.66004943847656, | |
| "logps/rejected": -215.84115600585938, | |
| "loss": 0.3163, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 1.1951297521591187, | |
| "rewards/margins": 1.7206119298934937, | |
| "rewards/rejected": -0.5254822969436646, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.7879065506184151, | |
| "grad_norm": 0.7585360407829285, | |
| "learning_rate": 1.306859653149025e-07, | |
| "logits/chosen": -0.7158334851264954, | |
| "logits/rejected": -0.6363730430603027, | |
| "logps/chosen": -111.70841979980469, | |
| "logps/rejected": -218.56442260742188, | |
| "loss": 0.3239, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 0.8049896359443665, | |
| "rewards/margins": 1.498777985572815, | |
| "rewards/rejected": -0.6937884092330933, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.7970682546953733, | |
| "grad_norm": 0.8166142106056213, | |
| "learning_rate": 1.2008313954339305e-07, | |
| "logits/chosen": -0.5638805627822876, | |
| "logits/rejected": -0.42692404985427856, | |
| "logps/chosen": -148.00503540039062, | |
| "logps/rejected": -250.53756713867188, | |
| "loss": 0.3133, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 0.6730099320411682, | |
| "rewards/margins": 1.2878690958023071, | |
| "rewards/rejected": -0.6148591637611389, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.8062299587723316, | |
| "grad_norm": 0.6894632577896118, | |
| "learning_rate": 1.098699085988432e-07, | |
| "logits/chosen": -0.6931108236312866, | |
| "logits/rejected": -0.6952135562896729, | |
| "logps/chosen": -102.37767791748047, | |
| "logps/rejected": -263.8074035644531, | |
| "loss": 0.3437, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 0.7925483584403992, | |
| "rewards/margins": 1.331838607788086, | |
| "rewards/rejected": -0.5392903089523315, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.81539166284929, | |
| "grad_norm": 0.8443871736526489, | |
| "learning_rate": 1.0005674588308566e-07, | |
| "logits/chosen": -0.8288379907608032, | |
| "logits/rejected": -0.7619950175285339, | |
| "logps/chosen": -87.10929870605469, | |
| "logps/rejected": -192.11410522460938, | |
| "loss": 0.2805, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 1.0762431621551514, | |
| "rewards/margins": 1.3870527744293213, | |
| "rewards/rejected": -0.31080955266952515, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.8245533669262483, | |
| "grad_norm": 1.089480996131897, | |
| "learning_rate": 9.065371453842358e-08, | |
| "logits/chosen": -0.7295564413070679, | |
| "logits/rejected": -0.66633141040802, | |
| "logps/chosen": -80.64302062988281, | |
| "logps/rejected": -198.11285400390625, | |
| "loss": 0.3317, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 0.9881649017333984, | |
| "rewards/margins": 1.6359241008758545, | |
| "rewards/rejected": -0.6477591395378113, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.8245533669262483, | |
| "eval_logits/chosen": -0.7748228907585144, | |
| "eval_logits/rejected": -0.5993096232414246, | |
| "eval_logps/chosen": -128.94505310058594, | |
| "eval_logps/rejected": -278.2012939453125, | |
| "eval_loss": 0.33487406373023987, | |
| "eval_rewards/accuracies": 0.913294792175293, | |
| "eval_rewards/chosen": 0.5586429238319397, | |
| "eval_rewards/margins": 1.2325509786605835, | |
| "eval_rewards/rejected": -0.6739080548286438, | |
| "eval_runtime": 251.978, | |
| "eval_samples_per_second": 10.965, | |
| "eval_steps_per_second": 1.373, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.8337150710032066, | |
| "grad_norm": 0.7584892511367798, | |
| "learning_rate": 8.167045712814108e-08, | |
| "logits/chosen": -0.6891669034957886, | |
| "logits/rejected": -0.509884774684906, | |
| "logps/chosen": -120.35499572753906, | |
| "logps/rejected": -280.7765197753906, | |
| "loss": 0.3101, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 0.7489343881607056, | |
| "rewards/margins": 1.3420490026474, | |
| "rewards/rejected": -0.5931144952774048, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.8428767750801649, | |
| "grad_norm": 0.6903491020202637, | |
| "learning_rate": 7.311618574830569e-08, | |
| "logits/chosen": -0.6809369921684265, | |
| "logits/rejected": -0.5410372614860535, | |
| "logps/chosen": -115.5818862915039, | |
| "logps/rejected": -269.8819885253906, | |
| "loss": 0.2988, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": 1.0188477039337158, | |
| "rewards/margins": 1.640873908996582, | |
| "rewards/rejected": -0.6220260262489319, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.8520384791571233, | |
| "grad_norm": 0.837735116481781, | |
| "learning_rate": 6.499967258100514e-08, | |
| "logits/chosen": -0.7608433365821838, | |
| "logits/rejected": -0.6938787698745728, | |
| "logps/chosen": -108.01826477050781, | |
| "logps/rejected": -226.8501434326172, | |
| "loss": 0.3083, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": 0.8835248947143555, | |
| "rewards/margins": 1.5439478158950806, | |
| "rewards/rejected": -0.6604230403900146, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.8612001832340815, | |
| "grad_norm": 0.5987643599510193, | |
| "learning_rate": 5.732924089870245e-08, | |
| "logits/chosen": -0.6554332971572876, | |
| "logits/rejected": -0.46006011962890625, | |
| "logps/chosen": -141.659423828125, | |
| "logps/rejected": -257.38958740234375, | |
| "loss": 0.2997, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 0.7329229712486267, | |
| "rewards/margins": 1.5281693935394287, | |
| "rewards/rejected": -0.795246422290802, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.8703618873110398, | |
| "grad_norm": 0.7886483073234558, | |
| "learning_rate": 5.011275652893782e-08, | |
| "logits/chosen": -0.696278989315033, | |
| "logits/rejected": -0.6031322479248047, | |
| "logps/chosen": -92.44393157958984, | |
| "logps/rejected": -203.4006805419922, | |
| "loss": 0.2837, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": 0.9351547360420227, | |
| "rewards/margins": 1.3968000411987305, | |
| "rewards/rejected": -0.46164530515670776, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.8795235913879982, | |
| "grad_norm": 0.8714445233345032, | |
| "learning_rate": 4.3357619788127634e-08, | |
| "logits/chosen": -0.7752223014831543, | |
| "logits/rejected": -0.7101688981056213, | |
| "logps/chosen": -135.9874725341797, | |
| "logps/rejected": -251.438232421875, | |
| "loss": 0.3198, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.7239378094673157, | |
| "rewards/margins": 1.1686517000198364, | |
| "rewards/rejected": -0.444713830947876, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.8886852954649564, | |
| "grad_norm": 0.7179840207099915, | |
| "learning_rate": 3.707075789273306e-08, | |
| "logits/chosen": -0.7246268391609192, | |
| "logits/rejected": -0.726759135723114, | |
| "logps/chosen": -98.63682556152344, | |
| "logps/rejected": -221.41909790039062, | |
| "loss": 0.3284, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 1.0240366458892822, | |
| "rewards/margins": 1.735687255859375, | |
| "rewards/rejected": -0.711650550365448, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.8978469995419148, | |
| "grad_norm": 0.8668420314788818, | |
| "learning_rate": 3.125861785558015e-08, | |
| "logits/chosen": -0.8971655964851379, | |
| "logits/rejected": -0.7343107461929321, | |
| "logps/chosen": -121.60533142089844, | |
| "logps/rejected": -274.42510986328125, | |
| "loss": 0.3273, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.9639977216720581, | |
| "rewards/margins": 1.4876940250396729, | |
| "rewards/rejected": -0.52369624376297, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.9070087036188731, | |
| "grad_norm": 0.8045462965965271, | |
| "learning_rate": 2.592715987461702e-08, | |
| "logits/chosen": -0.707330584526062, | |
| "logits/rejected": -0.6783408522605896, | |
| "logps/chosen": -161.0232391357422, | |
| "logps/rejected": -251.8162384033203, | |
| "loss": 0.3285, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 0.64153653383255, | |
| "rewards/margins": 1.0937219858169556, | |
| "rewards/rejected": -0.45218539237976074, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.9161704076958315, | |
| "grad_norm": 0.7876152992248535, | |
| "learning_rate": 2.108185122088546e-08, | |
| "logits/chosen": -0.8192548751831055, | |
| "logits/rejected": -0.7370203733444214, | |
| "logps/chosen": -109.0053939819336, | |
| "logps/rejected": -241.18704223632812, | |
| "loss": 0.3077, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": 0.7523542642593384, | |
| "rewards/margins": 1.3361942768096924, | |
| "rewards/rejected": -0.5838399529457092, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.9161704076958315, | |
| "eval_logits/chosen": -0.7753556370735168, | |
| "eval_logits/rejected": -0.5998021960258484, | |
| "eval_logps/chosen": -129.05845642089844, | |
| "eval_logps/rejected": -278.6715087890625, | |
| "eval_loss": 0.33283570408821106, | |
| "eval_rewards/accuracies": 0.9075144529342651, | |
| "eval_rewards/chosen": 0.5529729723930359, | |
| "eval_rewards/margins": 1.2503920793533325, | |
| "eval_rewards/rejected": -0.697419285774231, | |
| "eval_runtime": 411.7768, | |
| "eval_samples_per_second": 6.71, | |
| "eval_steps_per_second": 0.84, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.9253321117727897, | |
| "grad_norm": 0.9411417245864868, | |
| "learning_rate": 1.672766063197789e-08, | |
| "logits/chosen": -0.6959950923919678, | |
| "logits/rejected": -0.671855092048645, | |
| "logps/chosen": -124.23388671875, | |
| "logps/rejected": -235.0277862548828, | |
| "loss": 0.3114, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.9731966257095337, | |
| "rewards/margins": 1.4448806047439575, | |
| "rewards/rejected": -0.4716837406158447, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.934493815849748, | |
| "grad_norm": 0.7529547214508057, | |
| "learning_rate": 1.286905321672621e-08, | |
| "logits/chosen": -0.7256031632423401, | |
| "logits/rejected": -0.6639115214347839, | |
| "logps/chosen": -74.08079528808594, | |
| "logps/rejected": -190.6580047607422, | |
| "loss": 0.2967, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 1.259799838066101, | |
| "rewards/margins": 1.864816665649414, | |
| "rewards/rejected": -0.605016827583313, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.9436555199267064, | |
| "grad_norm": 0.8753419518470764, | |
| "learning_rate": 9.509985876349491e-09, | |
| "logits/chosen": -0.6738966703414917, | |
| "logits/rejected": -0.5317025184631348, | |
| "logps/chosen": -84.36373138427734, | |
| "logps/rejected": -239.25881958007812, | |
| "loss": 0.2962, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.9429308772087097, | |
| "rewards/margins": 1.350412368774414, | |
| "rewards/rejected": -0.40748143196105957, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.9528172240036646, | |
| "grad_norm": 0.8753317594528198, | |
| "learning_rate": 6.6539032467546885e-09, | |
| "logits/chosen": -0.7684177756309509, | |
| "logits/rejected": -0.7221516370773315, | |
| "logps/chosen": -195.75209045410156, | |
| "logps/rejected": -291.89117431640625, | |
| "loss": 0.2878, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 0.786228597164154, | |
| "rewards/margins": 1.300559163093567, | |
| "rewards/rejected": -0.5143305063247681, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.961978928080623, | |
| "grad_norm": 0.7298064827919006, | |
| "learning_rate": 4.303734166152706e-09, | |
| "logits/chosen": -0.6869091987609863, | |
| "logits/rejected": -0.7067757844924927, | |
| "logps/chosen": -120.041748046875, | |
| "logps/rejected": -194.99925231933594, | |
| "loss": 0.3014, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 1.13614022731781, | |
| "rewards/margins": 1.6430591344833374, | |
| "rewards/rejected": -0.5069188475608826, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.9711406321575813, | |
| "grad_norm": 0.6257432103157043, | |
| "learning_rate": 2.4618886716110676e-09, | |
| "logits/chosen": -0.7349929213523865, | |
| "logits/rejected": -0.6031204462051392, | |
| "logps/chosen": -102.07683563232422, | |
| "logps/rejected": -239.6611785888672, | |
| "loss": 0.3347, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.5327378511428833, | |
| "rewards/margins": 1.1345620155334473, | |
| "rewards/rejected": -0.6018242835998535, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.9803023362345397, | |
| "grad_norm": 0.6151149868965149, | |
| "learning_rate": 1.1302555276238579e-09, | |
| "logits/chosen": -0.8125348091125488, | |
| "logits/rejected": -0.6750482320785522, | |
| "logps/chosen": -64.71864318847656, | |
| "logps/rejected": -191.86572265625, | |
| "loss": 0.2674, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 1.005023717880249, | |
| "rewards/margins": 1.796847939491272, | |
| "rewards/rejected": -0.7918239831924438, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.9894640403114979, | |
| "grad_norm": 0.7649210691452026, | |
| "learning_rate": 3.102002892329536e-10, | |
| "logits/chosen": -0.6314386129379272, | |
| "logits/rejected": -0.5421333909034729, | |
| "logps/chosen": -96.51834869384766, | |
| "logps/rejected": -234.3397979736328, | |
| "loss": 0.2975, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": 1.1401491165161133, | |
| "rewards/margins": 1.9231317043304443, | |
| "rewards/rejected": -0.7829826474189758, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.9986257443884563, | |
| "grad_norm": 0.7103332877159119, | |
| "learning_rate": 2.5639016871248366e-12, | |
| "logits/chosen": -0.7133598923683167, | |
| "logits/rejected": -0.5197003483772278, | |
| "logps/chosen": -163.91204833984375, | |
| "logps/rejected": -249.02682495117188, | |
| "loss": 0.3111, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": 0.7740644216537476, | |
| "rewards/margins": 1.3091305494308472, | |
| "rewards/rejected": -0.5350660681724548, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.9995419147961521, | |
| "step": 1091, | |
| "total_flos": 0.0, | |
| "train_loss": 0.4306906612229719, | |
| "train_runtime": 8689.0202, | |
| "train_samples_per_second": 4.019, | |
| "train_steps_per_second": 0.126 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1091, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "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 | |
| } | |