| { | |
| "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": 47.431165475675876, | |
| "learning_rate": 1.0638297872340425e-07, | |
| "logits/chosen": -1.0093685388565063, | |
| "logits/rejected": -0.9801958203315735, | |
| "logps/chosen": -0.27410927414894104, | |
| "logps/rejected": -0.2716289460659027, | |
| "loss": 3.0415, | |
| "rewards/accuracies": 0.4312500059604645, | |
| "rewards/chosen": -2.7410929203033447, | |
| "rewards/margins": -0.024803416803479195, | |
| "rewards/rejected": -2.7162892818450928, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02137608550434202, | |
| "grad_norm": 41.239549738008755, | |
| "learning_rate": 2.127659574468085e-07, | |
| "logits/chosen": -1.0442249774932861, | |
| "logits/rejected": -0.9774014353752136, | |
| "logps/chosen": -0.29465872049331665, | |
| "logps/rejected": -0.29961076378822327, | |
| "loss": 3.0332, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -2.946587085723877, | |
| "rewards/margins": 0.04952071234583855, | |
| "rewards/rejected": -2.996107578277588, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03206412825651302, | |
| "grad_norm": 43.03376486032676, | |
| "learning_rate": 3.1914893617021275e-07, | |
| "logits/chosen": -0.9605627059936523, | |
| "logits/rejected": -0.9794631004333496, | |
| "logps/chosen": -0.26438188552856445, | |
| "logps/rejected": -0.30044788122177124, | |
| "loss": 3.047, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -2.6438186168670654, | |
| "rewards/margins": 0.36066022515296936, | |
| "rewards/rejected": -3.004478931427002, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04275217100868404, | |
| "grad_norm": 55.729867353234994, | |
| "learning_rate": 4.25531914893617e-07, | |
| "logits/chosen": -0.9706584215164185, | |
| "logits/rejected": -0.9438554048538208, | |
| "logps/chosen": -0.2775923013687134, | |
| "logps/rejected": -0.29160231351852417, | |
| "loss": 2.9682, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -2.775923252105713, | |
| "rewards/margins": 0.14010018110275269, | |
| "rewards/rejected": -2.916023015975952, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.053440213760855046, | |
| "grad_norm": 53.66865278898649, | |
| "learning_rate": 5.319148936170212e-07, | |
| "logits/chosen": -1.0153733491897583, | |
| "logits/rejected": -0.9855473637580872, | |
| "logps/chosen": -0.2714543342590332, | |
| "logps/rejected": -0.2778756022453308, | |
| "loss": 3.1483, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -2.714543104171753, | |
| "rewards/margins": 0.0642128437757492, | |
| "rewards/rejected": -2.7787561416625977, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06412825651302605, | |
| "grad_norm": 43.4911593109138, | |
| "learning_rate": 6.382978723404255e-07, | |
| "logits/chosen": -1.0054885149002075, | |
| "logits/rejected": -0.9609144330024719, | |
| "logps/chosen": -0.27275633811950684, | |
| "logps/rejected": -0.2785263657569885, | |
| "loss": 2.9455, | |
| "rewards/accuracies": 0.44999998807907104, | |
| "rewards/chosen": -2.7275633811950684, | |
| "rewards/margins": 0.05770007520914078, | |
| "rewards/rejected": -2.7852635383605957, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07481629926519706, | |
| "grad_norm": 62.11467926433348, | |
| "learning_rate": 7.446808510638297e-07, | |
| "logits/chosen": -1.048362374305725, | |
| "logits/rejected": -0.9739812612533569, | |
| "logps/chosen": -0.2923530042171478, | |
| "logps/rejected": -0.31766629219055176, | |
| "loss": 2.9148, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -2.923529863357544, | |
| "rewards/margins": 0.253133088350296, | |
| "rewards/rejected": -3.1766631603240967, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08550434201736808, | |
| "grad_norm": 66.19354446783973, | |
| "learning_rate": 8.51063829787234e-07, | |
| "logits/chosen": -1.027752161026001, | |
| "logits/rejected": -0.982993483543396, | |
| "logps/chosen": -0.28172561526298523, | |
| "logps/rejected": -0.3218705952167511, | |
| "loss": 2.9039, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -2.817256212234497, | |
| "rewards/margins": 0.40144944190979004, | |
| "rewards/rejected": -3.218705654144287, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09619238476953908, | |
| "grad_norm": 40.808829184419544, | |
| "learning_rate": 9.574468085106384e-07, | |
| "logits/chosen": -1.064257025718689, | |
| "logits/rejected": -1.022287130355835, | |
| "logps/chosen": -0.33324795961380005, | |
| "logps/rejected": -0.3929290771484375, | |
| "loss": 2.9539, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -3.332479476928711, | |
| "rewards/margins": 0.5968114137649536, | |
| "rewards/rejected": -3.929290771484375, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10688042752171009, | |
| "grad_norm": 51.85470904227569, | |
| "learning_rate": 9.998741174712533e-07, | |
| "logits/chosen": -1.0401779413223267, | |
| "logits/rejected": -0.990446925163269, | |
| "logps/chosen": -0.321146696805954, | |
| "logps/rejected": -0.37150952219963074, | |
| "loss": 2.9887, | |
| "rewards/accuracies": 0.48124998807907104, | |
| "rewards/chosen": -3.2114670276641846, | |
| "rewards/margins": 0.5036287307739258, | |
| "rewards/rejected": -3.7150955200195312, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11756847027388109, | |
| "grad_norm": 75.50941924220831, | |
| "learning_rate": 9.991050648838675e-07, | |
| "logits/chosen": -1.0736358165740967, | |
| "logits/rejected": -1.0379724502563477, | |
| "logps/chosen": -0.28898900747299194, | |
| "logps/rejected": -0.3547995686531067, | |
| "loss": 2.7366, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -2.88988995552063, | |
| "rewards/margins": 0.658105731010437, | |
| "rewards/rejected": -3.5479958057403564, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.1282565130260521, | |
| "grad_norm": 53.536746683672696, | |
| "learning_rate": 9.97637968732563e-07, | |
| "logits/chosen": -1.1059926748275757, | |
| "logits/rejected": -1.0733429193496704, | |
| "logps/chosen": -0.3318749964237213, | |
| "logps/rejected": -0.3473610281944275, | |
| "loss": 2.8777, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -3.3187499046325684, | |
| "rewards/margins": 0.15486064553260803, | |
| "rewards/rejected": -3.4736104011535645, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13894455577822312, | |
| "grad_norm": 57.894224289101096, | |
| "learning_rate": 9.954748808839674e-07, | |
| "logits/chosen": -1.0196824073791504, | |
| "logits/rejected": -0.9911971092224121, | |
| "logps/chosen": -0.3787004351615906, | |
| "logps/rejected": -0.45460987091064453, | |
| "loss": 2.7406, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -3.787003993988037, | |
| "rewards/margins": 0.7590948343276978, | |
| "rewards/rejected": -4.5460991859436035, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14963259853039412, | |
| "grad_norm": 40.680575216186135, | |
| "learning_rate": 9.926188266120295e-07, | |
| "logits/chosen": -1.0517082214355469, | |
| "logits/rejected": -1.0281683206558228, | |
| "logps/chosen": -0.3594748377799988, | |
| "logps/rejected": -0.4522577226161957, | |
| "loss": 2.8528, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -3.594748020172119, | |
| "rewards/margins": 0.9278289675712585, | |
| "rewards/rejected": -4.522576808929443, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.16032064128256512, | |
| "grad_norm": 58.56175042188981, | |
| "learning_rate": 9.890738003669027e-07, | |
| "logits/chosen": -1.0064220428466797, | |
| "logits/rejected": -0.9376434087753296, | |
| "logps/chosen": -0.36788100004196167, | |
| "logps/rejected": -0.42331838607788086, | |
| "loss": 2.7642, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -3.6788101196289062, | |
| "rewards/margins": 0.5543740391731262, | |
| "rewards/rejected": -4.233183860778809, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.17100868403473615, | |
| "grad_norm": 47.18670280685341, | |
| "learning_rate": 9.848447601883433e-07, | |
| "logits/chosen": -0.9864493608474731, | |
| "logits/rejected": -0.9741519689559937, | |
| "logps/chosen": -0.3633944094181061, | |
| "logps/rejected": -0.47577548027038574, | |
| "loss": 2.6618, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -3.633944272994995, | |
| "rewards/margins": 1.1238101720809937, | |
| "rewards/rejected": -4.757754325866699, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.18169672678690715, | |
| "grad_norm": 56.89275049644848, | |
| "learning_rate": 9.799376207714444e-07, | |
| "logits/chosen": -0.9944761395454407, | |
| "logits/rejected": -0.9737129211425781, | |
| "logps/chosen": -0.348540723323822, | |
| "logps/rejected": -0.4202929437160492, | |
| "loss": 2.5907, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -3.4854073524475098, | |
| "rewards/margins": 0.7175217866897583, | |
| "rewards/rejected": -4.2029290199279785, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.19238476953907815, | |
| "grad_norm": 64.95639784576333, | |
| "learning_rate": 9.743592451943998e-07, | |
| "logits/chosen": -1.039825677871704, | |
| "logits/rejected": -1.0079519748687744, | |
| "logps/chosen": -0.4592970013618469, | |
| "logps/rejected": -0.5395578145980835, | |
| "loss": 2.8586, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -4.5929694175720215, | |
| "rewards/margins": 0.8026081919670105, | |
| "rewards/rejected": -5.395577430725098, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.20307281229124916, | |
| "grad_norm": 58.58608272426839, | |
| "learning_rate": 9.681174353198686e-07, | |
| "logits/chosen": -1.1195189952850342, | |
| "logits/rejected": -1.0406802892684937, | |
| "logps/chosen": -0.47035104036331177, | |
| "logps/rejected": -0.5184075236320496, | |
| "loss": 2.6804, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -4.703510284423828, | |
| "rewards/margins": 0.4805658459663391, | |
| "rewards/rejected": -5.184075832366943, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.21376085504342018, | |
| "grad_norm": 65.6357725034444, | |
| "learning_rate": 9.612209208833646e-07, | |
| "logits/chosen": -1.018027901649475, | |
| "logits/rejected": -0.9948932528495789, | |
| "logps/chosen": -0.4622405171394348, | |
| "logps/rejected": -0.528734564781189, | |
| "loss": 2.7539, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -4.622405052185059, | |
| "rewards/margins": 0.6649408340454102, | |
| "rewards/rejected": -5.287345886230469, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.22444889779559118, | |
| "grad_norm": 60.841137599337706, | |
| "learning_rate": 9.536793472839324e-07, | |
| "logits/chosen": -1.0319901704788208, | |
| "logits/rejected": -0.9800949096679688, | |
| "logps/chosen": -0.43034663796424866, | |
| "logps/rejected": -0.5452305674552917, | |
| "loss": 2.6903, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -4.303465843200684, | |
| "rewards/margins": 1.1488397121429443, | |
| "rewards/rejected": -5.452306270599365, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23513694054776219, | |
| "grad_norm": 66.66409649647449, | |
| "learning_rate": 9.455032620941839e-07, | |
| "logits/chosen": -0.985127329826355, | |
| "logits/rejected": -0.9267324209213257, | |
| "logps/chosen": -0.48107171058654785, | |
| "logps/rejected": -0.6188473701477051, | |
| "loss": 2.6025, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -4.8107171058654785, | |
| "rewards/margins": 1.3777568340301514, | |
| "rewards/rejected": -6.188473701477051, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2458249832999332, | |
| "grad_norm": 79.76554466353629, | |
| "learning_rate": 9.367041003085648e-07, | |
| "logits/chosen": -1.0459407567977905, | |
| "logits/rejected": -0.987104058265686, | |
| "logps/chosen": -0.5227265357971191, | |
| "logps/rejected": -0.5985993146896362, | |
| "loss": 2.4679, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -5.227265357971191, | |
| "rewards/margins": 0.7587274312973022, | |
| "rewards/rejected": -5.985992908477783, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2565130260521042, | |
| "grad_norm": 71.0699167092581, | |
| "learning_rate": 9.272941683504808e-07, | |
| "logits/chosen": -1.0006247758865356, | |
| "logits/rejected": -0.9123749732971191, | |
| "logps/chosen": -0.5375245213508606, | |
| "logps/rejected": -0.7409927845001221, | |
| "loss": 2.3789, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -5.375245094299316, | |
| "rewards/margins": 2.0346832275390625, | |
| "rewards/rejected": -7.409928321838379, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.26720106880427524, | |
| "grad_norm": 71.80653378932202, | |
| "learning_rate": 9.172866268606513e-07, | |
| "logits/chosen": -1.0720800161361694, | |
| "logits/rejected": -1.0307856798171997, | |
| "logps/chosen": -0.6043092608451843, | |
| "logps/rejected": -0.7186105847358704, | |
| "loss": 2.2873, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -6.043093681335449, | |
| "rewards/margins": 1.1430130004882812, | |
| "rewards/rejected": -7.186106204986572, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.27788911155644624, | |
| "grad_norm": 76.52309149449971, | |
| "learning_rate": 9.066954722907638e-07, | |
| "logits/chosen": -1.0935930013656616, | |
| "logits/rejected": -1.087660551071167, | |
| "logps/chosen": -0.6125985383987427, | |
| "logps/rejected": -0.9003360867500305, | |
| "loss": 2.0828, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -6.125986099243164, | |
| "rewards/margins": 2.8773741722106934, | |
| "rewards/rejected": -9.003360748291016, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.28857715430861725, | |
| "grad_norm": 74.74165844394271, | |
| "learning_rate": 8.955355173281707e-07, | |
| "logits/chosen": -1.0836272239685059, | |
| "logits/rejected": -1.0363242626190186, | |
| "logps/chosen": -0.705600917339325, | |
| "logps/rejected": -0.8650037050247192, | |
| "loss": 2.1181, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -7.056008815765381, | |
| "rewards/margins": 1.5940272808074951, | |
| "rewards/rejected": -8.650036811828613, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.29926519706078825, | |
| "grad_norm": 82.01125840970528, | |
| "learning_rate": 8.838223701790055e-07, | |
| "logits/chosen": -1.1484724283218384, | |
| "logits/rejected": -1.1269563436508179, | |
| "logps/chosen": -0.8566378355026245, | |
| "logps/rejected": -1.011114239692688, | |
| "loss": 2.0912, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -8.566378593444824, | |
| "rewards/margins": 1.5447633266448975, | |
| "rewards/rejected": -10.1111421585083, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.30995323981295925, | |
| "grad_norm": 85.34384972694933, | |
| "learning_rate": 8.71572412738697e-07, | |
| "logits/chosen": -1.0586411952972412, | |
| "logits/rejected": -1.0339049100875854, | |
| "logps/chosen": -0.8592250943183899, | |
| "logps/rejected": -1.117244005203247, | |
| "loss": 1.9373, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -8.592249870300293, | |
| "rewards/margins": 2.5801892280578613, | |
| "rewards/rejected": -11.172440528869629, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.32064128256513025, | |
| "grad_norm": 83.57155238411846, | |
| "learning_rate": 8.588027776804058e-07, | |
| "logits/chosen": -1.0752789974212646, | |
| "logits/rejected": -1.057715654373169, | |
| "logps/chosen": -0.9481467008590698, | |
| "logps/rejected": -1.1904191970825195, | |
| "loss": 1.9564, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -9.481468200683594, | |
| "rewards/margins": 2.422724723815918, | |
| "rewards/rejected": -11.904191970825195, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.33132932531730125, | |
| "grad_norm": 87.97325530333627, | |
| "learning_rate": 8.455313244934324e-07, | |
| "logits/chosen": -1.084198236465454, | |
| "logits/rejected": -1.0641534328460693, | |
| "logps/chosen": -1.0298274755477905, | |
| "logps/rejected": -1.338127851486206, | |
| "loss": 2.0325, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -10.298274993896484, | |
| "rewards/margins": 3.0830020904541016, | |
| "rewards/rejected": -13.381277084350586, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.3420173680694723, | |
| "grad_norm": 99.47098531695673, | |
| "learning_rate": 8.317766145051057e-07, | |
| "logits/chosen": -1.0900102853775024, | |
| "logits/rejected": -1.0740206241607666, | |
| "logps/chosen": -1.147585153579712, | |
| "logps/rejected": -1.567286729812622, | |
| "loss": 1.97, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -11.475851058959961, | |
| "rewards/margins": 4.197016716003418, | |
| "rewards/rejected": -15.672868728637695, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3527054108216433, | |
| "grad_norm": 63.97740340024748, | |
| "learning_rate": 8.175578849210894e-07, | |
| "logits/chosen": -1.114548921585083, | |
| "logits/rejected": -1.0925266742706299, | |
| "logps/chosen": -1.1195895671844482, | |
| "logps/rejected": -1.5141029357910156, | |
| "loss": 1.8483, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -11.195895195007324, | |
| "rewards/margins": 3.9451332092285156, | |
| "rewards/rejected": -15.141029357910156, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3633934535738143, | |
| "grad_norm": 96.40902503969279, | |
| "learning_rate": 8.028950219204099e-07, | |
| "logits/chosen": -1.1151658296585083, | |
| "logits/rejected": -1.0936176776885986, | |
| "logps/chosen": -1.0980020761489868, | |
| "logps/rejected": -1.5130436420440674, | |
| "loss": 1.8009, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -10.980019569396973, | |
| "rewards/margins": 4.150416374206543, | |
| "rewards/rejected": -15.1304349899292, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.3740814963259853, | |
| "grad_norm": 88.267242124785, | |
| "learning_rate": 7.878085328428368e-07, | |
| "logits/chosen": -1.1430907249450684, | |
| "logits/rejected": -1.0969905853271484, | |
| "logps/chosen": -1.1937079429626465, | |
| "logps/rejected": -1.4644215106964111, | |
| "loss": 1.6648, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -11.937080383300781, | |
| "rewards/margins": 2.7071356773376465, | |
| "rewards/rejected": -14.644216537475586, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3847695390781563, | |
| "grad_norm": 97.14634805094629, | |
| "learning_rate": 7.723195175075135e-07, | |
| "logits/chosen": -1.100259780883789, | |
| "logits/rejected": -1.0826084613800049, | |
| "logps/chosen": -1.1161882877349854, | |
| "logps/rejected": -1.483745813369751, | |
| "loss": 1.6508, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -11.161882400512695, | |
| "rewards/margins": 3.6755752563476562, | |
| "rewards/rejected": -14.837457656860352, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3954575818303273, | |
| "grad_norm": 99.49435944644794, | |
| "learning_rate": 7.564496387029531e-07, | |
| "logits/chosen": -1.1444669961929321, | |
| "logits/rejected": -1.0918629169464111, | |
| "logps/chosen": -1.1615946292877197, | |
| "logps/rejected": -1.5793178081512451, | |
| "loss": 1.7334, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -11.615945816040039, | |
| "rewards/margins": 4.177231788635254, | |
| "rewards/rejected": -15.793177604675293, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.4061456245824983, | |
| "grad_norm": 90.47582983297573, | |
| "learning_rate": 7.402210918896689e-07, | |
| "logits/chosen": -1.1339255571365356, | |
| "logits/rejected": -1.143362283706665, | |
| "logps/chosen": -1.2574008703231812, | |
| "logps/rejected": -1.757472276687622, | |
| "loss": 1.565, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -12.574009895324707, | |
| "rewards/margins": 5.000711917877197, | |
| "rewards/rejected": -17.57472038269043, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4168336673346693, | |
| "grad_norm": 69.29879834232428, | |
| "learning_rate": 7.236565741578162e-07, | |
| "logits/chosen": -1.0910736322402954, | |
| "logits/rejected": -1.073672890663147, | |
| "logps/chosen": -1.2707961797714233, | |
| "logps/rejected": -1.6355140209197998, | |
| "loss": 1.6663, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -12.707962036132812, | |
| "rewards/margins": 3.6471798419952393, | |
| "rewards/rejected": -16.355140686035156, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.42752171008684037, | |
| "grad_norm": 114.06338305078152, | |
| "learning_rate": 7.067792524832603e-07, | |
| "logits/chosen": -1.0798423290252686, | |
| "logits/rejected": -1.0704095363616943, | |
| "logps/chosen": -1.2981719970703125, | |
| "logps/rejected": -1.7119674682617188, | |
| "loss": 1.5558, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -12.981719970703125, | |
| "rewards/margins": 4.137955665588379, | |
| "rewards/rejected": -17.11967658996582, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.43820975283901137, | |
| "grad_norm": 90.04523554097818, | |
| "learning_rate": 6.896127313264642e-07, | |
| "logits/chosen": -1.1306648254394531, | |
| "logits/rejected": -1.082484483718872, | |
| "logps/chosen": -1.3895291090011597, | |
| "logps/rejected": -1.8172976970672607, | |
| "loss": 1.7407, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -13.895291328430176, | |
| "rewards/margins": 4.277683734893799, | |
| "rewards/rejected": -18.172977447509766, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.44889779559118237, | |
| "grad_norm": 90.2095455226181, | |
| "learning_rate": 6.721810196195174e-07, | |
| "logits/chosen": -1.156951904296875, | |
| "logits/rejected": -1.1476523876190186, | |
| "logps/chosen": -1.405803918838501, | |
| "logps/rejected": -1.8274023532867432, | |
| "loss": 1.6338, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -14.058039665222168, | |
| "rewards/margins": 4.215982913970947, | |
| "rewards/rejected": -18.274023056030273, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.45958583834335337, | |
| "grad_norm": 121.41353956332256, | |
| "learning_rate": 6.545084971874736e-07, | |
| "logits/chosen": -1.098239541053772, | |
| "logits/rejected": -1.082698941230774, | |
| "logps/chosen": -1.419948697090149, | |
| "logps/rejected": -1.8909088373184204, | |
| "loss": 1.4946, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -14.199487686157227, | |
| "rewards/margins": 4.709600925445557, | |
| "rewards/rejected": -18.909086227416992, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.47027388109552437, | |
| "grad_norm": 101.41780500317314, | |
| "learning_rate": 6.3661988065096e-07, | |
| "logits/chosen": -1.1594250202178955, | |
| "logits/rejected": -1.142090082168579, | |
| "logps/chosen": -1.455776333808899, | |
| "logps/rejected": -1.9567766189575195, | |
| "loss": 1.4418, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -14.557762145996094, | |
| "rewards/margins": 5.01000452041626, | |
| "rewards/rejected": -19.567766189575195, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.48096192384769537, | |
| "grad_norm": 75.55567519433296, | |
| "learning_rate": 6.185401888577487e-07, | |
| "logits/chosen": -1.139162302017212, | |
| "logits/rejected": -1.1079540252685547, | |
| "logps/chosen": -1.5027626752853394, | |
| "logps/rejected": -1.9599437713623047, | |
| "loss": 1.4648, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -15.027628898620605, | |
| "rewards/margins": 4.571808815002441, | |
| "rewards/rejected": -19.599437713623047, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.4916499665998664, | |
| "grad_norm": 74.26739925465552, | |
| "learning_rate": 6.002947078916364e-07, | |
| "logits/chosen": -1.2207891941070557, | |
| "logits/rejected": -1.1718138456344604, | |
| "logps/chosen": -1.4501031637191772, | |
| "logps/rejected": -1.89298415184021, | |
| "loss": 1.4109, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -14.501032829284668, | |
| "rewards/margins": 4.428809642791748, | |
| "rewards/rejected": -18.929840087890625, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5023380093520374, | |
| "grad_norm": 94.05649604378364, | |
| "learning_rate": 5.819089557075688e-07, | |
| "logits/chosen": -1.247056007385254, | |
| "logits/rejected": -1.2196362018585205, | |
| "logps/chosen": -1.4864906072616577, | |
| "logps/rejected": -1.998369812965393, | |
| "loss": 1.3624, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -14.864908218383789, | |
| "rewards/margins": 5.1187896728515625, | |
| "rewards/rejected": -19.98369789123535, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5130260521042084, | |
| "grad_norm": 83.1982843702179, | |
| "learning_rate": 5.634086464424742e-07, | |
| "logits/chosen": -1.219469428062439, | |
| "logits/rejected": -1.2216647863388062, | |
| "logps/chosen": -1.4312059879302979, | |
| "logps/rejected": -1.9184064865112305, | |
| "loss": 1.4426, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -14.31205940246582, | |
| "rewards/margins": 4.872005462646484, | |
| "rewards/rejected": -19.184062957763672, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5237140948563794, | |
| "grad_norm": 91.50503743289462, | |
| "learning_rate": 5.448196544517167e-07, | |
| "logits/chosen": -1.3046424388885498, | |
| "logits/rejected": -1.2480084896087646, | |
| "logps/chosen": -1.4915251731872559, | |
| "logps/rejected": -2.1005947589874268, | |
| "loss": 1.3463, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": -14.915250778198242, | |
| "rewards/margins": 6.090696334838867, | |
| "rewards/rejected": -21.00594711303711, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5344021376085505, | |
| "grad_norm": 93.25169357476504, | |
| "learning_rate": 5.26167978121472e-07, | |
| "logits/chosen": -1.2517807483673096, | |
| "logits/rejected": -1.2366987466812134, | |
| "logps/chosen": -1.5615525245666504, | |
| "logps/rejected": -2.1383166313171387, | |
| "loss": 1.3433, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -15.61552619934082, | |
| "rewards/margins": 5.767637729644775, | |
| "rewards/rejected": -21.383163452148438, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5450901803607214, | |
| "grad_norm": 84.64245243867487, | |
| "learning_rate": 5.074797035076318e-07, | |
| "logits/chosen": -1.3026002645492554, | |
| "logits/rejected": -1.279266595840454, | |
| "logps/chosen": -1.6704552173614502, | |
| "logps/rejected": -2.1482787132263184, | |
| "loss": 1.4096, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -16.70454978942871, | |
| "rewards/margins": 4.778237342834473, | |
| "rewards/rejected": -21.4827880859375, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5557782231128925, | |
| "grad_norm": 95.87257492049523, | |
| "learning_rate": 4.887809678520975e-07, | |
| "logits/chosen": -1.2724745273590088, | |
| "logits/rejected": -1.2435853481292725, | |
| "logps/chosen": -1.5778831243515015, | |
| "logps/rejected": -2.0692591667175293, | |
| "loss": 1.4179, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -15.778831481933594, | |
| "rewards/margins": 4.913762092590332, | |
| "rewards/rejected": -20.69259262084961, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5664662658650634, | |
| "grad_norm": 75.03721390019906, | |
| "learning_rate": 4.700979230274829e-07, | |
| "logits/chosen": -1.239183783531189, | |
| "logits/rejected": -1.222015619277954, | |
| "logps/chosen": -1.6615720987319946, | |
| "logps/rejected": -2.201857566833496, | |
| "loss": 1.3249, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -16.615720748901367, | |
| "rewards/margins": 5.402855396270752, | |
| "rewards/rejected": -22.018573760986328, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5771543086172345, | |
| "grad_norm": 164.9017663104039, | |
| "learning_rate": 4.514566989613559e-07, | |
| "logits/chosen": -1.235475778579712, | |
| "logits/rejected": -1.2083224058151245, | |
| "logps/chosen": -1.4797016382217407, | |
| "logps/rejected": -2.020514965057373, | |
| "loss": 1.2676, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -14.797017097473145, | |
| "rewards/margins": 5.408134460449219, | |
| "rewards/rejected": -20.205150604248047, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5878423513694054, | |
| "grad_norm": 81.15228217715286, | |
| "learning_rate": 4.328833670911724e-07, | |
| "logits/chosen": -1.2022705078125, | |
| "logits/rejected": -1.1674965620040894, | |
| "logps/chosen": -1.4664252996444702, | |
| "logps/rejected": -1.931023359298706, | |
| "loss": 1.4248, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -14.664253234863281, | |
| "rewards/margins": 4.645982265472412, | |
| "rewards/rejected": -19.31023597717285, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5985303941215765, | |
| "grad_norm": 104.47138977370568, | |
| "learning_rate": 4.144039039010124e-07, | |
| "logits/chosen": -1.28118097782135, | |
| "logits/rejected": -1.2569917440414429, | |
| "logps/chosen": -1.5222523212432861, | |
| "logps/rejected": -2.0608773231506348, | |
| "loss": 1.3366, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -15.22252368927002, | |
| "rewards/margins": 5.386250972747803, | |
| "rewards/rejected": -20.608774185180664, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6092184368737475, | |
| "grad_norm": 93.20515844216109, | |
| "learning_rate": 3.960441545911204e-07, | |
| "logits/chosen": -1.251189947128296, | |
| "logits/rejected": -1.2210452556610107, | |
| "logps/chosen": -1.5165798664093018, | |
| "logps/rejected": -2.1089890003204346, | |
| "loss": 1.0744, | |
| "rewards/accuracies": 0.8812500238418579, | |
| "rewards/chosen": -15.165797233581543, | |
| "rewards/margins": 5.924090385437012, | |
| "rewards/rejected": -21.089889526367188, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6199064796259185, | |
| "grad_norm": 81.40609622602612, | |
| "learning_rate": 3.778297969310529e-07, | |
| "logits/chosen": -1.2748587131500244, | |
| "logits/rejected": -1.2314293384552002, | |
| "logps/chosen": -1.5472508668899536, | |
| "logps/rejected": -2.0469508171081543, | |
| "loss": 1.3477, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -15.472508430480957, | |
| "rewards/margins": 4.996997356414795, | |
| "rewards/rejected": -20.469507217407227, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6305945223780896, | |
| "grad_norm": 99.64164292903165, | |
| "learning_rate": 3.5978630534699865e-07, | |
| "logits/chosen": -1.2144112586975098, | |
| "logits/rejected": -1.2017982006072998, | |
| "logps/chosen": -1.56182062625885, | |
| "logps/rejected": -2.0901787281036377, | |
| "loss": 1.1901, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -15.618207931518555, | |
| "rewards/margins": 5.283578872680664, | |
| "rewards/rejected": -20.901784896850586, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6412825651302605, | |
| "grad_norm": 88.30403958670574, | |
| "learning_rate": 3.4193891529348795e-07, | |
| "logits/chosen": -1.1442140340805054, | |
| "logits/rejected": -1.1183984279632568, | |
| "logps/chosen": -1.6569198369979858, | |
| "logps/rejected": -2.119126558303833, | |
| "loss": 1.5134, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -16.569198608398438, | |
| "rewards/margins": 4.622069358825684, | |
| "rewards/rejected": -21.191268920898438, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6519706078824316, | |
| "grad_norm": 76.57630583970531, | |
| "learning_rate": 3.243125879593286e-07, | |
| "logits/chosen": -1.2529053688049316, | |
| "logits/rejected": -1.2107479572296143, | |
| "logps/chosen": -1.6253858804702759, | |
| "logps/rejected": -2.088477849960327, | |
| "loss": 1.3079, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -16.253854751586914, | |
| "rewards/margins": 4.630920886993408, | |
| "rewards/rejected": -20.884777069091797, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6626586506346025, | |
| "grad_norm": 115.60549772909806, | |
| "learning_rate": 3.069319753571269e-07, | |
| "logits/chosen": -1.2966339588165283, | |
| "logits/rejected": -1.2777100801467896, | |
| "logps/chosen": -1.6359281539916992, | |
| "logps/rejected": -2.1718597412109375, | |
| "loss": 1.325, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -16.35927963256836, | |
| "rewards/margins": 5.359316825866699, | |
| "rewards/rejected": -21.71860122680664, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6733466933867736, | |
| "grad_norm": 88.88632603422757, | |
| "learning_rate": 2.898213858452173e-07, | |
| "logits/chosen": -1.2804549932479858, | |
| "logits/rejected": -1.2265164852142334, | |
| "logps/chosen": -1.6381988525390625, | |
| "logps/rejected": -2.162752628326416, | |
| "loss": 1.3169, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -16.381986618041992, | |
| "rewards/margins": 5.245541572570801, | |
| "rewards/rejected": -21.62752914428711, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6840347361389446, | |
| "grad_norm": 87.7153445266354, | |
| "learning_rate": 2.730047501302266e-07, | |
| "logits/chosen": -1.2721607685089111, | |
| "logits/rejected": -1.268053650856018, | |
| "logps/chosen": -1.6763546466827393, | |
| "logps/rejected": -2.327500820159912, | |
| "loss": 1.233, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": -16.763545989990234, | |
| "rewards/margins": 6.511463165283203, | |
| "rewards/rejected": -23.275009155273438, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6947227788911156, | |
| "grad_norm": 70.42378282195334, | |
| "learning_rate": 2.5650558779781635e-07, | |
| "logits/chosen": -1.3053241968154907, | |
| "logits/rejected": -1.2554945945739746, | |
| "logps/chosen": -1.7189693450927734, | |
| "logps/rejected": -2.395566463470459, | |
| "loss": 1.3248, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -17.189693450927734, | |
| "rewards/margins": 6.765972137451172, | |
| "rewards/rejected": -23.95566749572754, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.7054108216432866, | |
| "grad_norm": 73.22054751663588, | |
| "learning_rate": 2.403469744184154e-07, | |
| "logits/chosen": -1.2106170654296875, | |
| "logits/rejected": -1.1687678098678589, | |
| "logps/chosen": -1.6829122304916382, | |
| "logps/rejected": -2.164522647857666, | |
| "loss": 1.3184, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -16.829120635986328, | |
| "rewards/margins": 4.816105842590332, | |
| "rewards/rejected": -21.64522933959961, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7160988643954576, | |
| "grad_norm": 97.2368838538889, | |
| "learning_rate": 2.2455150927394878e-07, | |
| "logits/chosen": -1.2514355182647705, | |
| "logits/rejected": -1.2357397079467773, | |
| "logps/chosen": -1.652038335800171, | |
| "logps/rejected": -2.202007532119751, | |
| "loss": 1.1712, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -16.520383834838867, | |
| "rewards/margins": 5.499691009521484, | |
| "rewards/rejected": -22.02007484436035, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7267869071476286, | |
| "grad_norm": 101.12187857047496, | |
| "learning_rate": 2.0914128375069722e-07, | |
| "logits/chosen": -1.2703526020050049, | |
| "logits/rejected": -1.237259864807129, | |
| "logps/chosen": -1.5824345350265503, | |
| "logps/rejected": -2.1380550861358643, | |
| "loss": 1.275, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -15.824345588684082, | |
| "rewards/margins": 5.556204795837402, | |
| "rewards/rejected": -21.380550384521484, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7374749498997996, | |
| "grad_norm": 83.8607340097021, | |
| "learning_rate": 1.9413785044249676e-07, | |
| "logits/chosen": -1.3083045482635498, | |
| "logits/rejected": -1.2853591442108154, | |
| "logps/chosen": -1.6669658422470093, | |
| "logps/rejected": -2.3121345043182373, | |
| "loss": 1.3494, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -16.669658660888672, | |
| "rewards/margins": 6.45168924331665, | |
| "rewards/rejected": -23.121349334716797, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7481629926519706, | |
| "grad_norm": 108.91802839900684, | |
| "learning_rate": 1.7956219300748792e-07, | |
| "logits/chosen": -1.2830331325531006, | |
| "logits/rejected": -1.2870824337005615, | |
| "logps/chosen": -1.5124456882476807, | |
| "logps/rejected": -2.100379467010498, | |
| "loss": 1.2289, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -15.124456405639648, | |
| "rewards/margins": 5.879337310791016, | |
| "rewards/rejected": -21.003795623779297, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7588510354041417, | |
| "grad_norm": 92.29342846486482, | |
| "learning_rate": 1.6543469682057104e-07, | |
| "logits/chosen": -1.1956264972686768, | |
| "logits/rejected": -1.2103073596954346, | |
| "logps/chosen": -1.5397236347198486, | |
| "logps/rejected": -2.1057169437408447, | |
| "loss": 1.1804, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": -15.397236824035645, | |
| "rewards/margins": 5.659933090209961, | |
| "rewards/rejected": -21.05716896057129, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7695390781563126, | |
| "grad_norm": 80.96407512590481, | |
| "learning_rate": 1.5177512046261666e-07, | |
| "logits/chosen": -1.2520828247070312, | |
| "logits/rejected": -1.252455472946167, | |
| "logps/chosen": -1.548658847808838, | |
| "logps/rejected": -2.2058348655700684, | |
| "loss": 1.2696, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -15.486590385437012, | |
| "rewards/margins": 6.571758270263672, | |
| "rewards/rejected": -22.058345794677734, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7802271209084837, | |
| "grad_norm": 80.42780807489935, | |
| "learning_rate": 1.3860256808630427e-07, | |
| "logits/chosen": -1.2946264743804932, | |
| "logits/rejected": -1.2320127487182617, | |
| "logps/chosen": -1.6179031133651733, | |
| "logps/rejected": -2.2405967712402344, | |
| "loss": 1.2432, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -16.179031372070312, | |
| "rewards/margins": 6.226935863494873, | |
| "rewards/rejected": -22.405969619750977, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7909151636606546, | |
| "grad_norm": 101.37315779363591, | |
| "learning_rate": 1.2593546269723647e-07, | |
| "logits/chosen": -1.2141435146331787, | |
| "logits/rejected": -1.20265793800354, | |
| "logps/chosen": -1.5797102451324463, | |
| "logps/rejected": -2.090893507003784, | |
| "loss": 1.2071, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -15.797101974487305, | |
| "rewards/margins": 5.111833095550537, | |
| "rewards/rejected": -20.908937454223633, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.8016032064128257, | |
| "grad_norm": 77.11466865502938, | |
| "learning_rate": 1.1379152038770029e-07, | |
| "logits/chosen": -1.2475601434707642, | |
| "logits/rejected": -1.2534716129302979, | |
| "logps/chosen": -1.711700439453125, | |
| "logps/rejected": -2.310718536376953, | |
| "loss": 1.2262, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -17.11700439453125, | |
| "rewards/margins": 5.990179538726807, | |
| "rewards/rejected": -23.1071834564209, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.8122912491649966, | |
| "grad_norm": 117.37788053060973, | |
| "learning_rate": 1.0218772555910954e-07, | |
| "logits/chosen": -1.2747116088867188, | |
| "logits/rejected": -1.255689024925232, | |
| "logps/chosen": -1.5994197130203247, | |
| "logps/rejected": -2.137598752975464, | |
| "loss": 1.4027, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -15.994195938110352, | |
| "rewards/margins": 5.381791114807129, | |
| "rewards/rejected": -21.375986099243164, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8229792919171677, | |
| "grad_norm": 78.98597874286344, | |
| "learning_rate": 9.114030716778432e-08, | |
| "logits/chosen": -1.2616714239120483, | |
| "logits/rejected": -1.2391117811203003, | |
| "logps/chosen": -1.6102495193481445, | |
| "logps/rejected": -2.3222262859344482, | |
| "loss": 1.0892, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -16.102497100830078, | |
| "rewards/margins": 7.1197686195373535, | |
| "rewards/rejected": -23.222265243530273, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.8336673346693386, | |
| "grad_norm": 84.54508830914433, | |
| "learning_rate": 8.066471602728803e-08, | |
| "logits/chosen": -1.2698580026626587, | |
| "logits/rejected": -1.2559095621109009, | |
| "logps/chosen": -1.6754401922225952, | |
| "logps/rejected": -2.296884298324585, | |
| "loss": 1.2113, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -16.75440216064453, | |
| "rewards/margins": 6.214441299438477, | |
| "rewards/rejected": -22.96884536743164, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8443553774215097, | |
| "grad_norm": 80.87157072797021, | |
| "learning_rate": 7.077560319906694e-08, | |
| "logits/chosen": -1.271916389465332, | |
| "logits/rejected": -1.252569317817688, | |
| "logps/chosen": -1.5791504383087158, | |
| "logps/rejected": -2.1556344032287598, | |
| "loss": 1.26, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -15.79150390625, | |
| "rewards/margins": 5.764839172363281, | |
| "rewards/rejected": -21.55634117126465, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8550434201736807, | |
| "grad_norm": 65.17427856770865, | |
| "learning_rate": 6.148679950161672e-08, | |
| "logits/chosen": -1.2736351490020752, | |
| "logits/rejected": -1.2584983110427856, | |
| "logps/chosen": -1.6285731792449951, | |
| "logps/rejected": -2.164571762084961, | |
| "loss": 1.1724, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -16.28573226928711, | |
| "rewards/margins": 5.359988212585449, | |
| "rewards/rejected": -21.645719528198242, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8550434201736807, | |
| "eval_logits/chosen": -1.460854172706604, | |
| "eval_logits/rejected": -1.470569372177124, | |
| "eval_logps/chosen": -1.6253507137298584, | |
| "eval_logps/rejected": -2.183858871459961, | |
| "eval_loss": 1.351211428642273, | |
| "eval_rewards/accuracies": 0.8394308686256409, | |
| "eval_rewards/chosen": -16.253507614135742, | |
| "eval_rewards/margins": 5.5850830078125, | |
| "eval_rewards/rejected": -21.838590621948242, | |
| "eval_runtime": 98.0395, | |
| "eval_samples_per_second": 20.002, | |
| "eval_steps_per_second": 1.255, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8657314629258517, | |
| "grad_norm": 113.62350381683838, | |
| "learning_rate": 5.2811296166831666e-08, | |
| "logits/chosen": -1.2400482892990112, | |
| "logits/rejected": -1.2593395709991455, | |
| "logps/chosen": -1.7036975622177124, | |
| "logps/rejected": -2.276524066925049, | |
| "loss": 1.1972, | |
| "rewards/accuracies": 0.8812500238418579, | |
| "rewards/chosen": -17.036977767944336, | |
| "rewards/margins": 5.728262901306152, | |
| "rewards/rejected": -22.765239715576172, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8764195056780227, | |
| "grad_norm": 144.61743799041326, | |
| "learning_rate": 4.4761226670592066e-08, | |
| "logits/chosen": -1.2617685794830322, | |
| "logits/rejected": -1.249569296836853, | |
| "logps/chosen": -1.6581430435180664, | |
| "logps/rejected": -2.1955461502075195, | |
| "loss": 1.3257, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -16.581430435180664, | |
| "rewards/margins": 5.374030113220215, | |
| "rewards/rejected": -21.955459594726562, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8871075484301937, | |
| "grad_norm": 88.89080057583915, | |
| "learning_rate": 3.734784976300165e-08, | |
| "logits/chosen": -1.2503310441970825, | |
| "logits/rejected": -1.1999974250793457, | |
| "logps/chosen": -1.5886937379837036, | |
| "logps/rejected": -2.2466578483581543, | |
| "loss": 1.3331, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -15.886937141418457, | |
| "rewards/margins": 6.579639434814453, | |
| "rewards/rejected": -22.466577529907227, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.8977955911823647, | |
| "grad_norm": 85.80208344095527, | |
| "learning_rate": 3.058153372200695e-08, | |
| "logits/chosen": -1.2792952060699463, | |
| "logits/rejected": -1.2328004837036133, | |
| "logps/chosen": -1.524823546409607, | |
| "logps/rejected": -2.1325583457946777, | |
| "loss": 1.1978, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -15.248235702514648, | |
| "rewards/margins": 6.077347755432129, | |
| "rewards/rejected": -21.325584411621094, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9084836339345357, | |
| "grad_norm": 107.0115137921273, | |
| "learning_rate": 2.4471741852423233e-08, | |
| "logits/chosen": -1.2821718454360962, | |
| "logits/rejected": -1.273240089416504, | |
| "logps/chosen": -1.7154676914215088, | |
| "logps/rejected": -2.2665412425994873, | |
| "loss": 1.4189, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -17.154674530029297, | |
| "rewards/margins": 5.510737419128418, | |
| "rewards/rejected": -22.6654109954834, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9191716766867067, | |
| "grad_norm": 106.31448672736177, | |
| "learning_rate": 1.9027019250647036e-08, | |
| "logits/chosen": -1.267595887184143, | |
| "logits/rejected": -1.2530834674835205, | |
| "logps/chosen": -1.7077808380126953, | |
| "logps/rejected": -2.310647487640381, | |
| "loss": 1.253, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": -17.077808380126953, | |
| "rewards/margins": 6.028668403625488, | |
| "rewards/rejected": -23.106477737426758, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9298597194388778, | |
| "grad_norm": 82.83256823235773, | |
| "learning_rate": 1.4254980853566246e-08, | |
| "logits/chosen": -1.2165040969848633, | |
| "logits/rejected": -1.180338740348816, | |
| "logps/chosen": -1.5563018321990967, | |
| "logps/rejected": -2.143784284591675, | |
| "loss": 1.1374, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -15.563015937805176, | |
| "rewards/margins": 5.874825477600098, | |
| "rewards/rejected": -21.437841415405273, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9405477621910487, | |
| "grad_norm": 83.84639598697997, | |
| "learning_rate": 1.016230078838226e-08, | |
| "logits/chosen": -1.2405247688293457, | |
| "logits/rejected": -1.1848217248916626, | |
| "logps/chosen": -1.6584663391113281, | |
| "logps/rejected": -2.197697401046753, | |
| "loss": 1.2363, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -16.58466148376465, | |
| "rewards/margins": 5.392308235168457, | |
| "rewards/rejected": -21.976970672607422, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9512358049432198, | |
| "grad_norm": 80.97147542055019, | |
| "learning_rate": 6.754703038239329e-09, | |
| "logits/chosen": -1.1893346309661865, | |
| "logits/rejected": -1.1745051145553589, | |
| "logps/chosen": -1.6517200469970703, | |
| "logps/rejected": -2.3159308433532715, | |
| "loss": 1.0816, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -16.517200469970703, | |
| "rewards/margins": 6.642111301422119, | |
| "rewards/rejected": -23.159311294555664, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.9619238476953907, | |
| "grad_norm": 101.25047268904304, | |
| "learning_rate": 4.036953436716895e-09, | |
| "logits/chosen": -1.2991673946380615, | |
| "logits/rejected": -1.2805778980255127, | |
| "logps/chosen": -1.6150133609771729, | |
| "logps/rejected": -2.1743900775909424, | |
| "loss": 1.3429, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -16.150135040283203, | |
| "rewards/margins": 5.593768119812012, | |
| "rewards/rejected": -21.7439022064209, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9726118904475618, | |
| "grad_norm": 104.9392779126059, | |
| "learning_rate": 2.0128530023804656e-09, | |
| "logits/chosen": -1.2574083805084229, | |
| "logits/rejected": -1.22499680519104, | |
| "logps/chosen": -1.6232726573944092, | |
| "logps/rejected": -2.2851340770721436, | |
| "loss": 1.0189, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": -16.232728958129883, | |
| "rewards/margins": 6.618613243103027, | |
| "rewards/rejected": -22.851343154907227, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9832999331997327, | |
| "grad_norm": 89.08039811332223, | |
| "learning_rate": 6.852326227130833e-10, | |
| "logits/chosen": -1.2663607597351074, | |
| "logits/rejected": -1.2550715208053589, | |
| "logps/chosen": -1.699883222579956, | |
| "logps/rejected": -2.3202741146087646, | |
| "loss": 1.1666, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -16.99883460998535, | |
| "rewards/margins": 6.203906536102295, | |
| "rewards/rejected": -23.202741622924805, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9939879759519038, | |
| "grad_norm": 104.72078285232733, | |
| "learning_rate": 5.594909486328348e-11, | |
| "logits/chosen": -1.2424919605255127, | |
| "logits/rejected": -1.251131296157837, | |
| "logps/chosen": -1.704716682434082, | |
| "logps/rejected": -2.33524489402771, | |
| "loss": 1.3005, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -17.047168731689453, | |
| "rewards/margins": 6.305283546447754, | |
| "rewards/rejected": -23.352449417114258, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.9982631930527722, | |
| "step": 467, | |
| "total_flos": 0.0, | |
| "train_loss": 1.7901343150618756, | |
| "train_runtime": 11462.4072, | |
| "train_samples_per_second": 5.224, | |
| "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 | |
| } | |