| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.3420996365191362, | |
| "eval_steps": 100, | |
| "global_step": 300, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.002280664243460908, | |
| "grad_norm": 104.30809234123367, | |
| "learning_rate": 4e-09, | |
| "logits/chosen": -1.353676199913025, | |
| "logits/rejected": -1.4426417350769043, | |
| "logps/chosen": -169.59426879882812, | |
| "logps/rejected": -212.90069580078125, | |
| "loss": 0.7312, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": 0.23574542999267578, | |
| "rewards/margins": -0.06179435923695564, | |
| "rewards/rejected": 0.29753977060317993, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.004561328486921816, | |
| "grad_norm": 92.96399840391902, | |
| "learning_rate": 8e-09, | |
| "logits/chosen": -1.2849147319793701, | |
| "logits/rejected": -1.3181504011154175, | |
| "logps/chosen": -156.07391357421875, | |
| "logps/rejected": -180.74073791503906, | |
| "loss": 0.6935, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.32024019956588745, | |
| "rewards/margins": 0.08485272526741028, | |
| "rewards/rejected": 0.23538745939731598, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.006841992730382724, | |
| "grad_norm": 98.95959882761899, | |
| "learning_rate": 1.1999999999999998e-08, | |
| "logits/chosen": -1.4182077646255493, | |
| "logits/rejected": -1.4656966924667358, | |
| "logps/chosen": -162.13934326171875, | |
| "logps/rejected": -188.761474609375, | |
| "loss": 0.6766, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.4807170629501343, | |
| "rewards/margins": 0.1233808621764183, | |
| "rewards/rejected": 0.35733622312545776, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.009122656973843632, | |
| "grad_norm": 100.15592083854503, | |
| "learning_rate": 1.6e-08, | |
| "logits/chosen": -1.3926836252212524, | |
| "logits/rejected": -1.401774287223816, | |
| "logps/chosen": -170.7425079345703, | |
| "logps/rejected": -190.57833862304688, | |
| "loss": 0.7039, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 0.4466077983379364, | |
| "rewards/margins": 0.06900040060281754, | |
| "rewards/rejected": 0.37760740518569946, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.01140332121730454, | |
| "grad_norm": 97.0889776355223, | |
| "learning_rate": 2e-08, | |
| "logits/chosen": -1.376230239868164, | |
| "logits/rejected": -1.409171223640442, | |
| "logps/chosen": -169.57582092285156, | |
| "logps/rejected": -185.82571411132812, | |
| "loss": 0.7113, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.39193734526634216, | |
| "rewards/margins": 0.3205062747001648, | |
| "rewards/rejected": 0.07143110036849976, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.013683985460765448, | |
| "grad_norm": 92.66648488760384, | |
| "learning_rate": 2.3999999999999997e-08, | |
| "logits/chosen": -1.3613837957382202, | |
| "logits/rejected": -1.4228042364120483, | |
| "logps/chosen": -142.47850036621094, | |
| "logps/rejected": -195.29649353027344, | |
| "loss": 0.6867, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.19295203685760498, | |
| "rewards/margins": -0.01336541399359703, | |
| "rewards/rejected": 0.20631742477416992, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.015964649704226355, | |
| "grad_norm": 88.1654009314199, | |
| "learning_rate": 2.8000000000000003e-08, | |
| "logits/chosen": -1.5071678161621094, | |
| "logits/rejected": -1.5190939903259277, | |
| "logps/chosen": -128.34288024902344, | |
| "logps/rejected": -131.07789611816406, | |
| "loss": 0.6954, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.23906439542770386, | |
| "rewards/margins": 0.07392804324626923, | |
| "rewards/rejected": 0.16513636708259583, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.018245313947687265, | |
| "grad_norm": 83.0913005177034, | |
| "learning_rate": 3.2e-08, | |
| "logits/chosen": -1.2882771492004395, | |
| "logits/rejected": -1.3209519386291504, | |
| "logps/chosen": -125.15751647949219, | |
| "logps/rejected": -132.21255493164062, | |
| "loss": 0.6652, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": 0.18403516709804535, | |
| "rewards/margins": -0.032971642911434174, | |
| "rewards/rejected": 0.21700681746006012, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.02052597819114817, | |
| "grad_norm": 92.77604284067941, | |
| "learning_rate": 3.6e-08, | |
| "logits/chosen": -1.3992947340011597, | |
| "logits/rejected": -1.473150610923767, | |
| "logps/chosen": -173.55206298828125, | |
| "logps/rejected": -202.7044677734375, | |
| "loss": 0.7212, | |
| "rewards/accuracies": 0.34375, | |
| "rewards/chosen": 0.4242628216743469, | |
| "rewards/margins": -0.0976889505982399, | |
| "rewards/rejected": 0.5219517350196838, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.02280664243460908, | |
| "grad_norm": 88.66990299818134, | |
| "learning_rate": 4e-08, | |
| "logits/chosen": -1.432981014251709, | |
| "logits/rejected": -1.5539088249206543, | |
| "logps/chosen": -182.9879913330078, | |
| "logps/rejected": -203.99020385742188, | |
| "loss": 0.7184, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.4513806104660034, | |
| "rewards/margins": 0.20164981484413147, | |
| "rewards/rejected": 0.24973079562187195, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02508730667806999, | |
| "grad_norm": 77.91111396247051, | |
| "learning_rate": 4.4e-08, | |
| "logits/chosen": -1.3197808265686035, | |
| "logits/rejected": -1.3996787071228027, | |
| "logps/chosen": -154.55003356933594, | |
| "logps/rejected": -232.81744384765625, | |
| "loss": 0.6479, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.3307289481163025, | |
| "rewards/margins": 0.18726389110088348, | |
| "rewards/rejected": 0.14346502721309662, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.027367970921530895, | |
| "grad_norm": 99.08871355118758, | |
| "learning_rate": 4.799999999999999e-08, | |
| "logits/chosen": -1.3024228811264038, | |
| "logits/rejected": -1.3258510828018188, | |
| "logps/chosen": -131.18356323242188, | |
| "logps/rejected": -148.7147216796875, | |
| "loss": 0.7046, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.421053409576416, | |
| "rewards/margins": 0.3122374713420868, | |
| "rewards/rejected": 0.10881592333316803, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.029648635164991805, | |
| "grad_norm": 89.36423052851278, | |
| "learning_rate": 5.2e-08, | |
| "logits/chosen": -1.329531192779541, | |
| "logits/rejected": -1.4354490041732788, | |
| "logps/chosen": -168.1116180419922, | |
| "logps/rejected": -207.8105010986328, | |
| "loss": 0.6619, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.40395674109458923, | |
| "rewards/margins": 0.26189538836479187, | |
| "rewards/rejected": 0.14206132292747498, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.03192929940845271, | |
| "grad_norm": 99.33427508045129, | |
| "learning_rate": 5.6000000000000005e-08, | |
| "logits/chosen": -1.3796099424362183, | |
| "logits/rejected": -1.3620996475219727, | |
| "logps/chosen": -131.89183044433594, | |
| "logps/rejected": -133.18844604492188, | |
| "loss": 0.6782, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.37585678696632385, | |
| "rewards/margins": 0.27094465494155884, | |
| "rewards/rejected": 0.10491211712360382, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.03420996365191362, | |
| "grad_norm": 98.1725487266389, | |
| "learning_rate": 6e-08, | |
| "logits/chosen": -1.4092419147491455, | |
| "logits/rejected": -1.4259543418884277, | |
| "logps/chosen": -228.17083740234375, | |
| "logps/rejected": -249.2948455810547, | |
| "loss": 0.6923, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.4407646358013153, | |
| "rewards/margins": 0.335997998714447, | |
| "rewards/rejected": 0.10476663708686829, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03649062789537453, | |
| "grad_norm": 75.77879619863296, | |
| "learning_rate": 6.4e-08, | |
| "logits/chosen": -1.364654541015625, | |
| "logits/rejected": -1.4574649333953857, | |
| "logps/chosen": -134.63072204589844, | |
| "logps/rejected": -159.65719604492188, | |
| "loss": 0.6454, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.28305765986442566, | |
| "rewards/margins": 0.14592355489730835, | |
| "rewards/rejected": 0.1371341049671173, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.038771292138835435, | |
| "grad_norm": 100.55666520099189, | |
| "learning_rate": 6.8e-08, | |
| "logits/chosen": -1.3500841856002808, | |
| "logits/rejected": -1.4460691213607788, | |
| "logps/chosen": -134.4729461669922, | |
| "logps/rejected": -159.3612823486328, | |
| "loss": 0.716, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.44891929626464844, | |
| "rewards/margins": 0.12445079535245895, | |
| "rewards/rejected": 0.3244684934616089, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.04105195638229634, | |
| "grad_norm": 104.71963688579538, | |
| "learning_rate": 7.2e-08, | |
| "logits/chosen": -1.3853086233139038, | |
| "logits/rejected": -1.434361457824707, | |
| "logps/chosen": -139.8603057861328, | |
| "logps/rejected": -171.86366271972656, | |
| "loss": 0.6988, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.3810410499572754, | |
| "rewards/margins": 0.2534639537334442, | |
| "rewards/rejected": 0.12757712602615356, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.043332620625757254, | |
| "grad_norm": 77.00258209889141, | |
| "learning_rate": 7.599999999999999e-08, | |
| "logits/chosen": -1.3689879179000854, | |
| "logits/rejected": -1.3514069318771362, | |
| "logps/chosen": -208.98460388183594, | |
| "logps/rejected": -216.9095458984375, | |
| "loss": 0.6691, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.5943199396133423, | |
| "rewards/margins": 0.26101234555244446, | |
| "rewards/rejected": 0.33330756425857544, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.04561328486921816, | |
| "grad_norm": 83.25872912559916, | |
| "learning_rate": 8e-08, | |
| "logits/chosen": -1.3037664890289307, | |
| "logits/rejected": -1.3465715646743774, | |
| "logps/chosen": -160.97340393066406, | |
| "logps/rejected": -178.60267639160156, | |
| "loss": 0.6319, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.4052881896495819, | |
| "rewards/margins": 0.16212955117225647, | |
| "rewards/rejected": 0.24315865337848663, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.047893949112679066, | |
| "grad_norm": 113.55879280714245, | |
| "learning_rate": 8.4e-08, | |
| "logits/chosen": -1.3739519119262695, | |
| "logits/rejected": -1.442929744720459, | |
| "logps/chosen": -163.52279663085938, | |
| "logps/rejected": -217.51824951171875, | |
| "loss": 0.7171, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.13579119741916656, | |
| "rewards/margins": 0.14747555553913116, | |
| "rewards/rejected": -0.011684387922286987, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.05017461335613998, | |
| "grad_norm": 81.14122988211874, | |
| "learning_rate": 8.8e-08, | |
| "logits/chosen": -1.373199701309204, | |
| "logits/rejected": -1.4750025272369385, | |
| "logps/chosen": -132.52114868164062, | |
| "logps/rejected": -154.50648498535156, | |
| "loss": 0.6947, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.4790118932723999, | |
| "rewards/margins": 0.16499680280685425, | |
| "rewards/rejected": 0.31401512026786804, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.052455277599600884, | |
| "grad_norm": 89.9603968710281, | |
| "learning_rate": 9.2e-08, | |
| "logits/chosen": -1.3388586044311523, | |
| "logits/rejected": -1.4220290184020996, | |
| "logps/chosen": -181.95082092285156, | |
| "logps/rejected": -226.30923461914062, | |
| "loss": 0.6941, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": 0.4439522325992584, | |
| "rewards/margins": 0.17398414015769958, | |
| "rewards/rejected": 0.26996806263923645, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.05473594184306179, | |
| "grad_norm": 104.88400068525443, | |
| "learning_rate": 9.599999999999999e-08, | |
| "logits/chosen": -1.456312894821167, | |
| "logits/rejected": -1.451519250869751, | |
| "logps/chosen": -199.40493774414062, | |
| "logps/rejected": -216.68386840820312, | |
| "loss": 0.7537, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.4792614281177521, | |
| "rewards/margins": 0.112982377409935, | |
| "rewards/rejected": 0.36627906560897827, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.0570166060865227, | |
| "grad_norm": 73.96994315968016, | |
| "learning_rate": 1e-07, | |
| "logits/chosen": -1.3086615800857544, | |
| "logits/rejected": -1.3628607988357544, | |
| "logps/chosen": -151.69989013671875, | |
| "logps/rejected": -171.46029663085938, | |
| "loss": 0.6314, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.1983395218849182, | |
| "rewards/margins": 0.22184070944786072, | |
| "rewards/rejected": -0.023501206189393997, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05929727032998361, | |
| "grad_norm": 100.64321021539473, | |
| "learning_rate": 1.04e-07, | |
| "logits/chosen": -1.5143202543258667, | |
| "logits/rejected": -1.5794754028320312, | |
| "logps/chosen": -118.7998046875, | |
| "logps/rejected": -149.1954803466797, | |
| "loss": 0.6704, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.34698718786239624, | |
| "rewards/margins": 0.06618239730596542, | |
| "rewards/rejected": 0.2808048129081726, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.061577934573444515, | |
| "grad_norm": 96.37635530030595, | |
| "learning_rate": 1.08e-07, | |
| "logits/chosen": -1.254115104675293, | |
| "logits/rejected": -1.3123385906219482, | |
| "logps/chosen": -152.57516479492188, | |
| "logps/rejected": -172.66200256347656, | |
| "loss": 0.6966, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": 0.28736555576324463, | |
| "rewards/margins": 0.0115619245916605, | |
| "rewards/rejected": 0.2758035957813263, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.06385859881690542, | |
| "grad_norm": 99.11153586277212, | |
| "learning_rate": 1.1200000000000001e-07, | |
| "logits/chosen": -1.3405977487564087, | |
| "logits/rejected": -1.4054450988769531, | |
| "logps/chosen": -194.3892822265625, | |
| "logps/rejected": -219.24171447753906, | |
| "loss": 0.6625, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.382589727640152, | |
| "rewards/margins": 0.24182234704494476, | |
| "rewards/rejected": 0.14076738059520721, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.06613926306036633, | |
| "grad_norm": 82.95051663248452, | |
| "learning_rate": 1.1599999999999999e-07, | |
| "logits/chosen": -1.4561753273010254, | |
| "logits/rejected": -1.4183000326156616, | |
| "logps/chosen": -189.18923950195312, | |
| "logps/rejected": -186.99264526367188, | |
| "loss": 0.6206, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.4602745771408081, | |
| "rewards/margins": 0.18983443081378937, | |
| "rewards/rejected": 0.27044013142585754, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.06841992730382725, | |
| "grad_norm": 85.63947077517764, | |
| "learning_rate": 1.2e-07, | |
| "logits/chosen": -1.4239516258239746, | |
| "logits/rejected": -1.5685184001922607, | |
| "logps/chosen": -140.7595977783203, | |
| "logps/rejected": -184.33067321777344, | |
| "loss": 0.7255, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": 0.2823754847049713, | |
| "rewards/margins": 0.02492373436689377, | |
| "rewards/rejected": 0.25745177268981934, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.07070059154728815, | |
| "grad_norm": 97.67365684595, | |
| "learning_rate": 1.24e-07, | |
| "logits/chosen": -1.350022554397583, | |
| "logits/rejected": -1.3478128910064697, | |
| "logps/chosen": -171.09677124023438, | |
| "logps/rejected": -165.71694946289062, | |
| "loss": 0.6819, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.3082874119281769, | |
| "rewards/margins": 0.11261623352766037, | |
| "rewards/rejected": 0.1956711858510971, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.07298125579074906, | |
| "grad_norm": 99.3804737289846, | |
| "learning_rate": 1.28e-07, | |
| "logits/chosen": -1.4815456867218018, | |
| "logits/rejected": -1.4830272197723389, | |
| "logps/chosen": -107.2784194946289, | |
| "logps/rejected": -121.1039047241211, | |
| "loss": 0.7165, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.31323567032814026, | |
| "rewards/margins": 0.11968746036291122, | |
| "rewards/rejected": 0.19354820251464844, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.07526192003420996, | |
| "grad_norm": 94.44141503559203, | |
| "learning_rate": 1.32e-07, | |
| "logits/chosen": -1.493153691291809, | |
| "logits/rejected": -1.4850858449935913, | |
| "logps/chosen": -164.4472198486328, | |
| "logps/rejected": -169.9815673828125, | |
| "loss": 0.6558, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.5161038637161255, | |
| "rewards/margins": 0.30025073885917664, | |
| "rewards/rejected": 0.21585312485694885, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.07754258427767087, | |
| "grad_norm": 86.21425469616315, | |
| "learning_rate": 1.36e-07, | |
| "logits/chosen": -1.333409070968628, | |
| "logits/rejected": -1.3404844999313354, | |
| "logps/chosen": -153.38101196289062, | |
| "logps/rejected": -159.85690307617188, | |
| "loss": 0.6781, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.3244006931781769, | |
| "rewards/margins": 0.22184959053993225, | |
| "rewards/rejected": 0.10255111753940582, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.07982324852113178, | |
| "grad_norm": 90.03565465587579, | |
| "learning_rate": 1.3999999999999998e-07, | |
| "logits/chosen": -1.5341051816940308, | |
| "logits/rejected": -1.5601752996444702, | |
| "logps/chosen": -149.2415008544922, | |
| "logps/rejected": -156.4012908935547, | |
| "loss": 0.6936, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": 0.27590540051460266, | |
| "rewards/margins": 0.12058861553668976, | |
| "rewards/rejected": 0.1553167849779129, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.08210391276459268, | |
| "grad_norm": 80.65697099385187, | |
| "learning_rate": 1.44e-07, | |
| "logits/chosen": -1.3111834526062012, | |
| "logits/rejected": -1.3080787658691406, | |
| "logps/chosen": -152.25611877441406, | |
| "logps/rejected": -169.93919372558594, | |
| "loss": 0.6471, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 0.34636569023132324, | |
| "rewards/margins": 0.3154309093952179, | |
| "rewards/rejected": 0.03093479573726654, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.0843845770080536, | |
| "grad_norm": 116.72716177672243, | |
| "learning_rate": 1.48e-07, | |
| "logits/chosen": -1.3597153425216675, | |
| "logits/rejected": -1.4166897535324097, | |
| "logps/chosen": -174.33534240722656, | |
| "logps/rejected": -200.9926300048828, | |
| "loss": 0.6831, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.2202744334936142, | |
| "rewards/margins": 0.13681186735630035, | |
| "rewards/rejected": 0.08346255123615265, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.08666524125151451, | |
| "grad_norm": 99.62367819711024, | |
| "learning_rate": 1.5199999999999998e-07, | |
| "logits/chosen": -1.3973196744918823, | |
| "logits/rejected": -1.4653477668762207, | |
| "logps/chosen": -189.6744384765625, | |
| "logps/rejected": -224.8593292236328, | |
| "loss": 0.7086, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.33062875270843506, | |
| "rewards/margins": 0.11676961928606033, | |
| "rewards/rejected": 0.21385914087295532, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.08894590549497541, | |
| "grad_norm": 86.47188266565705, | |
| "learning_rate": 1.56e-07, | |
| "logits/chosen": -1.3268086910247803, | |
| "logits/rejected": -1.302788257598877, | |
| "logps/chosen": -135.62147521972656, | |
| "logps/rejected": -153.5860137939453, | |
| "loss": 0.6835, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.2682960629463196, | |
| "rewards/margins": 0.05024400353431702, | |
| "rewards/rejected": 0.21805202960968018, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.09122656973843632, | |
| "grad_norm": 125.16786139574279, | |
| "learning_rate": 1.6e-07, | |
| "logits/chosen": -1.2104482650756836, | |
| "logits/rejected": -1.3274182081222534, | |
| "logps/chosen": -161.164794921875, | |
| "logps/rejected": -210.53590393066406, | |
| "loss": 0.7174, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": 0.2355586439371109, | |
| "rewards/margins": 0.07326290011405945, | |
| "rewards/rejected": 0.16229577362537384, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.09350723398189723, | |
| "grad_norm": 98.85832957481998, | |
| "learning_rate": 1.6399999999999999e-07, | |
| "logits/chosen": -1.3579671382904053, | |
| "logits/rejected": -1.4671956300735474, | |
| "logps/chosen": -153.21591186523438, | |
| "logps/rejected": -173.60589599609375, | |
| "loss": 0.6862, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": 0.2675905227661133, | |
| "rewards/margins": 0.046474047005176544, | |
| "rewards/rejected": 0.22111651301383972, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.09578789822535813, | |
| "grad_norm": 83.11445245790541, | |
| "learning_rate": 1.68e-07, | |
| "logits/chosen": -1.2888550758361816, | |
| "logits/rejected": -1.3688700199127197, | |
| "logps/chosen": -145.24754333496094, | |
| "logps/rejected": -174.91700744628906, | |
| "loss": 0.637, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.36274462938308716, | |
| "rewards/margins": 0.15029877424240112, | |
| "rewards/rejected": 0.21244585514068604, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.09806856246881904, | |
| "grad_norm": 82.49239750970325, | |
| "learning_rate": 1.7199999999999998e-07, | |
| "logits/chosen": -1.3739551305770874, | |
| "logits/rejected": -1.3799883127212524, | |
| "logps/chosen": -178.80386352539062, | |
| "logps/rejected": -215.7256317138672, | |
| "loss": 0.6907, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.32459211349487305, | |
| "rewards/margins": 0.16312535107135773, | |
| "rewards/rejected": 0.1614667773246765, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.10034922671227996, | |
| "grad_norm": 105.64550946139919, | |
| "learning_rate": 1.76e-07, | |
| "logits/chosen": -1.4653171300888062, | |
| "logits/rejected": -1.508003830909729, | |
| "logps/chosen": -192.55154418945312, | |
| "logps/rejected": -214.61546325683594, | |
| "loss": 0.7401, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.43679386377334595, | |
| "rewards/margins": 0.1878117173910141, | |
| "rewards/rejected": 0.24898216128349304, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.10262989095574086, | |
| "grad_norm": 106.99677192741333, | |
| "learning_rate": 1.8e-07, | |
| "logits/chosen": -1.3307772874832153, | |
| "logits/rejected": -1.4872570037841797, | |
| "logps/chosen": -226.3917236328125, | |
| "logps/rejected": -265.8712158203125, | |
| "loss": 0.6758, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": 0.42067086696624756, | |
| "rewards/margins": 0.07930833846330643, | |
| "rewards/rejected": 0.34136250615119934, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.10491055519920177, | |
| "grad_norm": 104.67741510645097, | |
| "learning_rate": 1.84e-07, | |
| "logits/chosen": -1.3470810651779175, | |
| "logits/rejected": -1.3835241794586182, | |
| "logps/chosen": -162.22958374023438, | |
| "logps/rejected": -184.17697143554688, | |
| "loss": 0.6993, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.25471583008766174, | |
| "rewards/margins": 0.08957283943891525, | |
| "rewards/rejected": 0.1651429831981659, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.10719121944266267, | |
| "grad_norm": 82.2596269922547, | |
| "learning_rate": 1.88e-07, | |
| "logits/chosen": -1.3317389488220215, | |
| "logits/rejected": -1.367332935333252, | |
| "logps/chosen": -138.1045684814453, | |
| "logps/rejected": -213.2273712158203, | |
| "loss": 0.5933, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.27782779932022095, | |
| "rewards/margins": 0.2610364258289337, | |
| "rewards/rejected": 0.01679137721657753, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.10947188368612358, | |
| "grad_norm": 106.46518450026352, | |
| "learning_rate": 1.9199999999999997e-07, | |
| "logits/chosen": -1.3722081184387207, | |
| "logits/rejected": -1.4072837829589844, | |
| "logps/chosen": -185.40281677246094, | |
| "logps/rejected": -205.63037109375, | |
| "loss": 0.7079, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 0.4706501066684723, | |
| "rewards/margins": 0.29465728998184204, | |
| "rewards/rejected": 0.17599281668663025, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.11175254792958449, | |
| "grad_norm": 79.05608727628055, | |
| "learning_rate": 1.9599999999999998e-07, | |
| "logits/chosen": -1.2723050117492676, | |
| "logits/rejected": -1.2712621688842773, | |
| "logps/chosen": -170.6747589111328, | |
| "logps/rejected": -172.70376586914062, | |
| "loss": 0.6713, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.4158933460712433, | |
| "rewards/margins": 0.22599810361862183, | |
| "rewards/rejected": 0.18989527225494385, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.1140332121730454, | |
| "grad_norm": 102.50537690642588, | |
| "learning_rate": 2e-07, | |
| "logits/chosen": -1.2609540224075317, | |
| "logits/rejected": -1.3507779836654663, | |
| "logps/chosen": -126.81893157958984, | |
| "logps/rejected": -172.69830322265625, | |
| "loss": 0.7344, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.2087049037218094, | |
| "rewards/margins": 0.05775710940361023, | |
| "rewards/rejected": 0.15094780921936035, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.1140332121730454, | |
| "eval_logits/chosen": -1.465891718864441, | |
| "eval_logits/rejected": -1.452298879623413, | |
| "eval_logps/chosen": -126.79114532470703, | |
| "eval_logps/rejected": -128.49049377441406, | |
| "eval_loss": 0.7120834589004517, | |
| "eval_rewards/accuracies": 0.4000000059604645, | |
| "eval_rewards/chosen": 0.3106829524040222, | |
| "eval_rewards/margins": -0.04654194042086601, | |
| "eval_rewards/rejected": 0.3572249114513397, | |
| "eval_runtime": 22.2628, | |
| "eval_samples_per_second": 4.492, | |
| "eval_steps_per_second": 1.123, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.11631387641650631, | |
| "grad_norm": 115.65869747987765, | |
| "learning_rate": 1.9999927671554446e-07, | |
| "logits/chosen": -1.4759191274642944, | |
| "logits/rejected": -1.458261251449585, | |
| "logps/chosen": -200.50059509277344, | |
| "logps/rejected": -213.1103973388672, | |
| "loss": 0.6618, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": 0.3923037648200989, | |
| "rewards/margins": 0.1664755940437317, | |
| "rewards/rejected": 0.2258281409740448, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.11859454065996722, | |
| "grad_norm": 93.12582037555201, | |
| "learning_rate": 1.9999710687264071e-07, | |
| "logits/chosen": -1.311631441116333, | |
| "logits/rejected": -1.3278183937072754, | |
| "logps/chosen": -132.20401000976562, | |
| "logps/rejected": -157.69627380371094, | |
| "loss": 0.7135, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.19503216445446014, | |
| "rewards/margins": -0.017576567828655243, | |
| "rewards/rejected": 0.21260873973369598, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.12087520490342812, | |
| "grad_norm": 76.40018414480076, | |
| "learning_rate": 1.9999349050267698e-07, | |
| "logits/chosen": -1.291077733039856, | |
| "logits/rejected": -1.3958327770233154, | |
| "logps/chosen": -158.81573486328125, | |
| "logps/rejected": -199.284423828125, | |
| "loss": 0.6252, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.25081124901771545, | |
| "rewards/margins": 0.25090575218200684, | |
| "rewards/rejected": -9.45068895816803e-05, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.12315586914688903, | |
| "grad_norm": 90.49004280394047, | |
| "learning_rate": 1.9998842765796658e-07, | |
| "logits/chosen": -1.4403332471847534, | |
| "logits/rejected": -1.4392634630203247, | |
| "logps/chosen": -143.04052734375, | |
| "logps/rejected": -156.4816436767578, | |
| "loss": 0.6816, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.10938511788845062, | |
| "rewards/margins": 0.17133575677871704, | |
| "rewards/rejected": -0.06195063889026642, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.12543653339034994, | |
| "grad_norm": 81.5018944914688, | |
| "learning_rate": 1.9998191841174702e-07, | |
| "logits/chosen": -1.4619545936584473, | |
| "logits/rejected": -1.483961582183838, | |
| "logps/chosen": -120.11573028564453, | |
| "logps/rejected": -121.2934341430664, | |
| "loss": 0.6783, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.4725874662399292, | |
| "rewards/margins": 0.37433862686157227, | |
| "rewards/rejected": 0.09824882447719574, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.12771719763381084, | |
| "grad_norm": 91.38630396004429, | |
| "learning_rate": 1.9997396285817904e-07, | |
| "logits/chosen": -1.152282953262329, | |
| "logits/rejected": -1.252787470817566, | |
| "logps/chosen": -145.3370361328125, | |
| "logps/rejected": -174.6603240966797, | |
| "loss": 0.6269, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.40825897455215454, | |
| "rewards/margins": 0.25122010707855225, | |
| "rewards/rejected": 0.1570388376712799, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.12999786187727175, | |
| "grad_norm": 97.62629025517847, | |
| "learning_rate": 1.9996456111234526e-07, | |
| "logits/chosen": -1.393836259841919, | |
| "logits/rejected": -1.5118399858474731, | |
| "logps/chosen": -173.96847534179688, | |
| "logps/rejected": -229.94232177734375, | |
| "loss": 0.6697, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.21259565651416779, | |
| "rewards/margins": 0.18360112607479095, | |
| "rewards/rejected": 0.028994524851441383, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.13227852612073265, | |
| "grad_norm": 96.19706483741392, | |
| "learning_rate": 1.9995371331024833e-07, | |
| "logits/chosen": -1.299769401550293, | |
| "logits/rejected": -1.3910255432128906, | |
| "logps/chosen": -112.2242660522461, | |
| "logps/rejected": -130.05230712890625, | |
| "loss": 0.7283, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": 0.14800208806991577, | |
| "rewards/margins": -0.08275800198316574, | |
| "rewards/rejected": 0.23076008260250092, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.13455919036419356, | |
| "grad_norm": 114.37970019595372, | |
| "learning_rate": 1.999414196088092e-07, | |
| "logits/chosen": -1.3472720384597778, | |
| "logits/rejected": -1.4912211894989014, | |
| "logps/chosen": -202.3058319091797, | |
| "logps/rejected": -244.92904663085938, | |
| "loss": 0.7194, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": 0.3503240644931793, | |
| "rewards/margins": 0.024903126060962677, | |
| "rewards/rejected": 0.32542091608047485, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.1368398546076545, | |
| "grad_norm": 87.71566928589093, | |
| "learning_rate": 1.9992768018586478e-07, | |
| "logits/chosen": -1.3490660190582275, | |
| "logits/rejected": -1.3772616386413574, | |
| "logps/chosen": -146.59048461914062, | |
| "logps/rejected": -159.0959930419922, | |
| "loss": 0.6736, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.3947565257549286, | |
| "rewards/margins": 0.15041925013065338, | |
| "rewards/rejected": 0.244337260723114, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.1391205188511154, | |
| "grad_norm": 109.98286067474862, | |
| "learning_rate": 1.9991249524016528e-07, | |
| "logits/chosen": -1.2736543416976929, | |
| "logits/rejected": -1.4115746021270752, | |
| "logps/chosen": -150.3526611328125, | |
| "logps/rejected": -189.20545959472656, | |
| "loss": 0.761, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": 0.28761598467826843, | |
| "rewards/margins": -0.15278732776641846, | |
| "rewards/rejected": 0.44040337204933167, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.1414011830945763, | |
| "grad_norm": 108.93431507211488, | |
| "learning_rate": 1.9989586499137135e-07, | |
| "logits/chosen": -1.2978495359420776, | |
| "logits/rejected": -1.3472024202346802, | |
| "logps/chosen": -179.5664825439453, | |
| "logps/rejected": -203.53968811035156, | |
| "loss": 0.7215, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.39709562063217163, | |
| "rewards/margins": 0.20268934965133667, | |
| "rewards/rejected": 0.19440627098083496, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.1436818473380372, | |
| "grad_norm": 99.31026397357329, | |
| "learning_rate": 1.9987778968005106e-07, | |
| "logits/chosen": -1.2714462280273438, | |
| "logits/rejected": -1.4020087718963623, | |
| "logps/chosen": -142.49758911132812, | |
| "logps/rejected": -262.75775146484375, | |
| "loss": 0.7111, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.1395024061203003, | |
| "rewards/margins": -0.11353214085102081, | |
| "rewards/rejected": 0.2530345320701599, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.14596251158149812, | |
| "grad_norm": 92.0460328359884, | |
| "learning_rate": 1.9985826956767617e-07, | |
| "logits/chosen": -1.474685788154602, | |
| "logits/rejected": -1.5739442110061646, | |
| "logps/chosen": -135.28616333007812, | |
| "logps/rejected": -165.1414031982422, | |
| "loss": 0.6874, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.14916732907295227, | |
| "rewards/margins": 0.06242816895246506, | |
| "rewards/rejected": 0.0867391973733902, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.14824317582495902, | |
| "grad_norm": 96.22054918626448, | |
| "learning_rate": 1.9983730493661865e-07, | |
| "logits/chosen": -1.276041030883789, | |
| "logits/rejected": -1.3194851875305176, | |
| "logps/chosen": -221.15992736816406, | |
| "logps/rejected": -242.77647399902344, | |
| "loss": 0.6457, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.260553777217865, | |
| "rewards/margins": 0.26679596304893494, | |
| "rewards/rejected": -0.006242218893021345, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.15052384006841993, | |
| "grad_norm": 78.01858280488834, | |
| "learning_rate": 1.998148960901463e-07, | |
| "logits/chosen": -1.393154263496399, | |
| "logits/rejected": -1.423519492149353, | |
| "logps/chosen": -201.9635467529297, | |
| "logps/rejected": -235.25816345214844, | |
| "loss": 0.6893, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.382313072681427, | |
| "rewards/margins": 0.3356338441371918, | |
| "rewards/rejected": 0.04667920991778374, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.15280450431188083, | |
| "grad_norm": 103.00857915327929, | |
| "learning_rate": 1.997910433524185e-07, | |
| "logits/chosen": -1.3205957412719727, | |
| "logits/rejected": -1.3766120672225952, | |
| "logps/chosen": -189.32142639160156, | |
| "logps/rejected": -217.68544006347656, | |
| "loss": 0.724, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.25681746006011963, | |
| "rewards/margins": 0.15590888261795044, | |
| "rewards/rejected": 0.10090853273868561, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.15508516855534174, | |
| "grad_norm": 90.52690719042177, | |
| "learning_rate": 1.9976574706848153e-07, | |
| "logits/chosen": -1.4535361528396606, | |
| "logits/rejected": -1.4828790426254272, | |
| "logps/chosen": -171.7893829345703, | |
| "logps/rejected": -213.03846740722656, | |
| "loss": 0.6788, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": 0.33188748359680176, | |
| "rewards/margins": 0.07647443562746048, | |
| "rewards/rejected": 0.2554129958152771, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.15736583279880265, | |
| "grad_norm": 101.69492728925366, | |
| "learning_rate": 1.9973900760426363e-07, | |
| "logits/chosen": -1.4226858615875244, | |
| "logits/rejected": -1.4269222021102905, | |
| "logps/chosen": -166.29103088378906, | |
| "logps/rejected": -268.772705078125, | |
| "loss": 0.678, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": 0.3539034128189087, | |
| "rewards/margins": 0.052625708281993866, | |
| "rewards/rejected": 0.3012777268886566, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.15964649704226355, | |
| "grad_norm": 83.07835169069685, | |
| "learning_rate": 1.9971082534656955e-07, | |
| "logits/chosen": -1.37977933883667, | |
| "logits/rejected": -1.466927409172058, | |
| "logps/chosen": -113.37686157226562, | |
| "logps/rejected": -142.82138061523438, | |
| "loss": 0.6442, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.15869390964508057, | |
| "rewards/margins": 0.20419231057167053, | |
| "rewards/rejected": -0.045498400926589966, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.16192716128572446, | |
| "grad_norm": 103.54665420816978, | |
| "learning_rate": 1.99681200703075e-07, | |
| "logits/chosen": -1.4178937673568726, | |
| "logits/rejected": -1.4681955575942993, | |
| "logps/chosen": -158.346435546875, | |
| "logps/rejected": -180.1523895263672, | |
| "loss": 0.7127, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.2817228436470032, | |
| "rewards/margins": 0.11103209853172302, | |
| "rewards/rejected": 0.17069074511528015, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.16420782552918536, | |
| "grad_norm": 77.71797198863521, | |
| "learning_rate": 1.9965013410232097e-07, | |
| "logits/chosen": -1.40109384059906, | |
| "logits/rejected": -1.4553627967834473, | |
| "logps/chosen": -172.07899475097656, | |
| "logps/rejected": -222.21438598632812, | |
| "loss": 0.6728, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.27610713243484497, | |
| "rewards/margins": 0.35402336716651917, | |
| "rewards/rejected": -0.077916219830513, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.16648848977264627, | |
| "grad_norm": 91.27829192980697, | |
| "learning_rate": 1.996176259937072e-07, | |
| "logits/chosen": -1.3127285242080688, | |
| "logits/rejected": -1.3604404926300049, | |
| "logps/chosen": -168.33004760742188, | |
| "logps/rejected": -193.3489227294922, | |
| "loss": 0.679, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.18943265080451965, | |
| "rewards/margins": 0.11781211942434311, | |
| "rewards/rejected": 0.07162053138017654, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.1687691540161072, | |
| "grad_norm": 105.10949942036537, | |
| "learning_rate": 1.9958367684748586e-07, | |
| "logits/chosen": -1.301593542098999, | |
| "logits/rejected": -1.3123453855514526, | |
| "logps/chosen": -153.8624267578125, | |
| "logps/rejected": -172.3568115234375, | |
| "loss": 0.6824, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": 0.022907603532075882, | |
| "rewards/margins": 0.08112253248691559, | |
| "rewards/rejected": -0.058214932680130005, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.1710498182595681, | |
| "grad_norm": 99.17231684606652, | |
| "learning_rate": 1.995482871547548e-07, | |
| "logits/chosen": -1.2876176834106445, | |
| "logits/rejected": -1.3324649333953857, | |
| "logps/chosen": -127.41693115234375, | |
| "logps/rejected": -158.8002166748047, | |
| "loss": 0.6995, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 0.21076691150665283, | |
| "rewards/margins": 0.24259139597415924, | |
| "rewards/rejected": -0.031824491918087006, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.17333048250302902, | |
| "grad_norm": 91.41364046696481, | |
| "learning_rate": 1.9951145742745024e-07, | |
| "logits/chosen": -1.2649379968643188, | |
| "logits/rejected": -1.4087891578674316, | |
| "logps/chosen": -130.08468627929688, | |
| "logps/rejected": -158.69525146484375, | |
| "loss": 0.7016, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.23989444971084595, | |
| "rewards/margins": 0.030172180384397507, | |
| "rewards/rejected": 0.20972229540348053, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.17561114674648992, | |
| "grad_norm": 75.7202165191751, | |
| "learning_rate": 1.9947318819833962e-07, | |
| "logits/chosen": -1.3358944654464722, | |
| "logits/rejected": -1.3563066720962524, | |
| "logps/chosen": -106.91460418701172, | |
| "logps/rejected": -113.2139892578125, | |
| "loss": 0.6607, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": 0.14014360308647156, | |
| "rewards/margins": -0.031036917120218277, | |
| "rewards/rejected": 0.17118054628372192, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.17789181098995083, | |
| "grad_norm": 109.3316094106793, | |
| "learning_rate": 1.994334800210137e-07, | |
| "logits/chosen": -1.3152189254760742, | |
| "logits/rejected": -1.3776334524154663, | |
| "logps/chosen": -160.3906707763672, | |
| "logps/rejected": -232.65725708007812, | |
| "loss": 0.7105, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": 0.10319292545318604, | |
| "rewards/margins": -0.037254467606544495, | |
| "rewards/rejected": 0.14044740796089172, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.18017247523341173, | |
| "grad_norm": 84.0665837850057, | |
| "learning_rate": 1.9939233346987862e-07, | |
| "logits/chosen": -1.4214388132095337, | |
| "logits/rejected": -1.437930703163147, | |
| "logps/chosen": -172.41867065429688, | |
| "logps/rejected": -182.8797607421875, | |
| "loss": 0.6807, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.16843558847904205, | |
| "rewards/margins": 0.16182081401348114, | |
| "rewards/rejected": 0.0066147902980446815, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.18245313947687264, | |
| "grad_norm": 93.91811474012694, | |
| "learning_rate": 1.9934974914014762e-07, | |
| "logits/chosen": -1.3304110765457153, | |
| "logits/rejected": -1.386664628982544, | |
| "logps/chosen": -112.3897933959961, | |
| "logps/rejected": -133.53810119628906, | |
| "loss": 0.7031, | |
| "rewards/accuracies": 0.40625, | |
| "rewards/chosen": 0.19350233674049377, | |
| "rewards/margins": -0.08095596730709076, | |
| "rewards/rejected": 0.27445828914642334, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.18473380372033354, | |
| "grad_norm": 90.17492690496479, | |
| "learning_rate": 1.9930572764783236e-07, | |
| "logits/chosen": -1.2432975769042969, | |
| "logits/rejected": -1.2546998262405396, | |
| "logps/chosen": -167.08847045898438, | |
| "logps/rejected": -183.3931884765625, | |
| "loss": 0.6979, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.05589159578084946, | |
| "rewards/margins": 0.17001007497310638, | |
| "rewards/rejected": -0.11411847919225693, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.18701446796379445, | |
| "grad_norm": 95.44656254255437, | |
| "learning_rate": 1.9926026962973403e-07, | |
| "logits/chosen": -1.326186180114746, | |
| "logits/rejected": -1.4297109842300415, | |
| "logps/chosen": -183.224853515625, | |
| "logps/rejected": -239.97793579101562, | |
| "loss": 0.6436, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.16740243136882782, | |
| "rewards/margins": 0.23012831807136536, | |
| "rewards/rejected": -0.06272587180137634, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.18929513220725536, | |
| "grad_norm": 92.54409107962749, | |
| "learning_rate": 1.992133757434342e-07, | |
| "logits/chosen": -1.3230640888214111, | |
| "logits/rejected": -1.31871497631073, | |
| "logps/chosen": -226.2008819580078, | |
| "logps/rejected": -243.1304931640625, | |
| "loss": 0.6795, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.1608559638261795, | |
| "rewards/margins": 0.35546940565109253, | |
| "rewards/rejected": -0.1946134865283966, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.19157579645071626, | |
| "grad_norm": 93.78690688865437, | |
| "learning_rate": 1.991650466672853e-07, | |
| "logits/chosen": -1.2681585550308228, | |
| "logits/rejected": -1.27677321434021, | |
| "logps/chosen": -151.71714782714844, | |
| "logps/rejected": -171.88832092285156, | |
| "loss": 0.6772, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.21696476638317108, | |
| "rewards/margins": 0.30003517866134644, | |
| "rewards/rejected": -0.08307043462991714, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.19385646069417717, | |
| "grad_norm": 79.03039382866096, | |
| "learning_rate": 1.991152831004007e-07, | |
| "logits/chosen": -1.4594199657440186, | |
| "logits/rejected": -1.456141710281372, | |
| "logps/chosen": -169.4336700439453, | |
| "logps/rejected": -180.23756408691406, | |
| "loss": 0.695, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.091593436896801, | |
| "rewards/margins": 0.008565463125705719, | |
| "rewards/rejected": 0.08302795141935349, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.19613712493763807, | |
| "grad_norm": 84.01165418355862, | |
| "learning_rate": 1.9906408576264467e-07, | |
| "logits/chosen": -1.272233247756958, | |
| "logits/rejected": -1.309181809425354, | |
| "logps/chosen": -182.87887573242188, | |
| "logps/rejected": -218.7847442626953, | |
| "loss": 0.6868, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.13641594350337982, | |
| "rewards/margins": 0.20986737310886383, | |
| "rewards/rejected": -0.07345142215490341, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.198417789181099, | |
| "grad_norm": 95.73538074987611, | |
| "learning_rate": 1.9901145539462197e-07, | |
| "logits/chosen": -1.2831114530563354, | |
| "logits/rejected": -1.302902340888977, | |
| "logps/chosen": -147.5471954345703, | |
| "logps/rejected": -178.70321655273438, | |
| "loss": 0.6013, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.2684887647628784, | |
| "rewards/margins": 0.47333818674087524, | |
| "rewards/rejected": -0.20484942197799683, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.20069845342455991, | |
| "grad_norm": 102.90991917593935, | |
| "learning_rate": 1.9895739275766715e-07, | |
| "logits/chosen": -1.3818506002426147, | |
| "logits/rejected": -1.4092711210250854, | |
| "logps/chosen": -227.61679077148438, | |
| "logps/rejected": -226.88558959960938, | |
| "loss": 0.737, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.13025416433811188, | |
| "rewards/margins": 0.26329731941223145, | |
| "rewards/rejected": -0.13304319977760315, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.20297911766802082, | |
| "grad_norm": 105.89850456514735, | |
| "learning_rate": 1.9890189863383353e-07, | |
| "logits/chosen": -1.4667214155197144, | |
| "logits/rejected": -1.5101617574691772, | |
| "logps/chosen": -185.08575439453125, | |
| "logps/rejected": -237.65597534179688, | |
| "loss": 0.7025, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.2359275072813034, | |
| "rewards/margins": 0.08144809305667877, | |
| "rewards/rejected": 0.15447941422462463, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.20525978191148173, | |
| "grad_norm": 82.46406340773291, | |
| "learning_rate": 1.9884497382588183e-07, | |
| "logits/chosen": -1.2853760719299316, | |
| "logits/rejected": -1.3631393909454346, | |
| "logps/chosen": -140.35952758789062, | |
| "logps/rejected": -153.50978088378906, | |
| "loss": 0.6683, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.1511276513338089, | |
| "rewards/margins": 0.06823855638504028, | |
| "rewards/rejected": 0.08288908749818802, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.20754044615494263, | |
| "grad_norm": 90.32390874613019, | |
| "learning_rate": 1.9878661915726865e-07, | |
| "logits/chosen": -1.4392482042312622, | |
| "logits/rejected": -1.4409726858139038, | |
| "logps/chosen": -171.09303283691406, | |
| "logps/rejected": -182.13232421875, | |
| "loss": 0.6691, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": 0.20515748858451843, | |
| "rewards/margins": 0.10885492712259293, | |
| "rewards/rejected": 0.0963025838136673, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.20982111039840354, | |
| "grad_norm": 75.10429123148643, | |
| "learning_rate": 1.9872683547213446e-07, | |
| "logits/chosen": -1.3081107139587402, | |
| "logits/rejected": -1.317389965057373, | |
| "logps/chosen": -174.84463500976562, | |
| "logps/rejected": -210.18130493164062, | |
| "loss": 0.6158, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.18892019987106323, | |
| "rewards/margins": 0.4061851501464844, | |
| "rewards/rejected": -0.21726495027542114, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.21210177464186444, | |
| "grad_norm": 107.2216966034497, | |
| "learning_rate": 1.9866562363529144e-07, | |
| "logits/chosen": -1.2954977750778198, | |
| "logits/rejected": -1.3127189874649048, | |
| "logps/chosen": -197.97561645507812, | |
| "logps/rejected": -204.29013061523438, | |
| "loss": 0.7154, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": 0.17145520448684692, | |
| "rewards/margins": -0.03724336624145508, | |
| "rewards/rejected": 0.208698570728302, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.21438243888532535, | |
| "grad_norm": 96.60657963150389, | |
| "learning_rate": 1.9860298453221107e-07, | |
| "logits/chosen": -1.3616856336593628, | |
| "logits/rejected": -1.3934649229049683, | |
| "logps/chosen": -179.2027130126953, | |
| "logps/rejected": -187.0948028564453, | |
| "loss": 0.6701, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.0954272449016571, | |
| "rewards/margins": -0.021092543378472328, | |
| "rewards/rejected": 0.11651977896690369, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.21666310312878626, | |
| "grad_norm": 87.8142315385882, | |
| "learning_rate": 1.9853891906901108e-07, | |
| "logits/chosen": -1.3980423212051392, | |
| "logits/rejected": -1.4947185516357422, | |
| "logps/chosen": -142.8896942138672, | |
| "logps/rejected": -180.8528594970703, | |
| "loss": 0.6415, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.13048064708709717, | |
| "rewards/margins": 0.3490259051322937, | |
| "rewards/rejected": -0.21854525804519653, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.21894376737224716, | |
| "grad_norm": 91.22189567457308, | |
| "learning_rate": 1.9847342817244254e-07, | |
| "logits/chosen": -1.3708425760269165, | |
| "logits/rejected": -1.3748908042907715, | |
| "logps/chosen": -155.70945739746094, | |
| "logps/rejected": -175.20013427734375, | |
| "loss": 0.6507, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.3499588370323181, | |
| "rewards/margins": 0.12689019739627838, | |
| "rewards/rejected": 0.22306862473487854, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.22122443161570807, | |
| "grad_norm": 103.90331869438803, | |
| "learning_rate": 1.9840651278987642e-07, | |
| "logits/chosen": -1.3758065700531006, | |
| "logits/rejected": -1.3824987411499023, | |
| "logps/chosen": -181.644775390625, | |
| "logps/rejected": -178.72109985351562, | |
| "loss": 0.6563, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 0.29924964904785156, | |
| "rewards/margins": 0.3290708065032959, | |
| "rewards/rejected": -0.02982119843363762, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.22350509585916897, | |
| "grad_norm": 103.8705512922574, | |
| "learning_rate": 1.9833817388928984e-07, | |
| "logits/chosen": -1.4562551975250244, | |
| "logits/rejected": -1.511305809020996, | |
| "logps/chosen": -168.09715270996094, | |
| "logps/rejected": -189.35787963867188, | |
| "loss": 0.6926, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.31778958439826965, | |
| "rewards/margins": 0.04199263080954552, | |
| "rewards/rejected": 0.27579694986343384, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.22578576010262988, | |
| "grad_norm": 103.40331674623164, | |
| "learning_rate": 1.982684124592521e-07, | |
| "logits/chosen": -1.3754979372024536, | |
| "logits/rejected": -1.347690463066101, | |
| "logps/chosen": -168.61148071289062, | |
| "logps/rejected": -186.8193359375, | |
| "loss": 0.7315, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.14702042937278748, | |
| "rewards/margins": 0.059913113713264465, | |
| "rewards/rejected": 0.08710730820894241, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.2280664243460908, | |
| "grad_norm": 102.65405433185128, | |
| "learning_rate": 1.981972295089103e-07, | |
| "logits/chosen": -1.4264771938323975, | |
| "logits/rejected": -1.5102254152297974, | |
| "logps/chosen": -160.4220428466797, | |
| "logps/rejected": -173.62269592285156, | |
| "loss": 0.6553, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.3021972179412842, | |
| "rewards/margins": 0.10177014023065567, | |
| "rewards/rejected": 0.2004270851612091, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.2280664243460908, | |
| "eval_logits/chosen": -1.4570631980895996, | |
| "eval_logits/rejected": -1.442795753479004, | |
| "eval_logps/chosen": -127.66089630126953, | |
| "eval_logps/rejected": -129.5493621826172, | |
| "eval_loss": 0.6965904235839844, | |
| "eval_rewards/accuracies": 0.5199999809265137, | |
| "eval_rewards/chosen": 0.22370710968971252, | |
| "eval_rewards/margins": -0.02763114497065544, | |
| "eval_rewards/rejected": 0.2513382136821747, | |
| "eval_runtime": 22.1102, | |
| "eval_samples_per_second": 4.523, | |
| "eval_steps_per_second": 1.131, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.23034708858955172, | |
| "grad_norm": 70.85461032661505, | |
| "learning_rate": 1.9812462606797498e-07, | |
| "logits/chosen": -1.1571446657180786, | |
| "logits/rejected": -1.2567036151885986, | |
| "logps/chosen": -170.15008544921875, | |
| "logps/rejected": -181.82806396484375, | |
| "loss": 0.6236, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.13917526602745056, | |
| "rewards/margins": 0.30568423867225647, | |
| "rewards/rejected": -0.1665089875459671, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.23262775283301262, | |
| "grad_norm": 101.20833615282186, | |
| "learning_rate": 1.9805060318670485e-07, | |
| "logits/chosen": -1.3708992004394531, | |
| "logits/rejected": -1.3186278343200684, | |
| "logps/chosen": -154.29226684570312, | |
| "logps/rejected": -148.5332794189453, | |
| "loss": 0.6319, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.09488646686077118, | |
| "rewards/margins": 0.11530336737632751, | |
| "rewards/rejected": -0.020416898652911186, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.23490841707647353, | |
| "grad_norm": 96.32887019136864, | |
| "learning_rate": 1.9797516193589192e-07, | |
| "logits/chosen": -1.3597060441970825, | |
| "logits/rejected": -1.3876497745513916, | |
| "logps/chosen": -101.66006469726562, | |
| "logps/rejected": -131.19041442871094, | |
| "loss": 0.7312, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": 0.20645485818386078, | |
| "rewards/margins": 0.06082191318273544, | |
| "rewards/rejected": 0.14563296735286713, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.23718908131993444, | |
| "grad_norm": 91.31654055574913, | |
| "learning_rate": 1.978983034068459e-07, | |
| "logits/chosen": -1.3642208576202393, | |
| "logits/rejected": -1.405060052871704, | |
| "logps/chosen": -188.59420776367188, | |
| "logps/rejected": -199.42637634277344, | |
| "loss": 0.6922, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.13877716660499573, | |
| "rewards/margins": 0.17159023880958557, | |
| "rewards/rejected": -0.03281306475400925, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.23946974556339534, | |
| "grad_norm": 99.31042679769261, | |
| "learning_rate": 1.9782002871137832e-07, | |
| "logits/chosen": -1.3214318752288818, | |
| "logits/rejected": -1.3617138862609863, | |
| "logps/chosen": -195.82009887695312, | |
| "logps/rejected": -214.06614685058594, | |
| "loss": 0.6673, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.16799718141555786, | |
| "rewards/margins": 0.25821763277053833, | |
| "rewards/rejected": -0.09022042155265808, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.24175040980685625, | |
| "grad_norm": 103.3617201342953, | |
| "learning_rate": 1.9774033898178666e-07, | |
| "logits/chosen": -1.4174686670303345, | |
| "logits/rejected": -1.420326590538025, | |
| "logps/chosen": -129.22848510742188, | |
| "logps/rejected": -145.9871826171875, | |
| "loss": 0.6716, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.3314517140388489, | |
| "rewards/margins": 0.1635245978832245, | |
| "rewards/rejected": 0.167927086353302, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.24403107405031715, | |
| "grad_norm": 85.24728778878888, | |
| "learning_rate": 1.9765923537083774e-07, | |
| "logits/chosen": -1.243302583694458, | |
| "logits/rejected": -1.2585264444351196, | |
| "logps/chosen": -181.84434509277344, | |
| "logps/rejected": -238.97323608398438, | |
| "loss": 0.6159, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.0770760029554367, | |
| "rewards/margins": 0.2519676089286804, | |
| "rewards/rejected": -0.17489157617092133, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.24631173829377806, | |
| "grad_norm": 96.37551956943565, | |
| "learning_rate": 1.9757671905175115e-07, | |
| "logits/chosen": -1.3951995372772217, | |
| "logits/rejected": -1.4096182584762573, | |
| "logps/chosen": -119.3514404296875, | |
| "logps/rejected": -131.9920196533203, | |
| "loss": 0.6021, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.2738375961780548, | |
| "rewards/margins": 0.2932060658931732, | |
| "rewards/rejected": -0.01936846226453781, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.24859240253723897, | |
| "grad_norm": 90.01951430126202, | |
| "learning_rate": 1.9749279121818234e-07, | |
| "logits/chosen": -1.3084944486618042, | |
| "logits/rejected": -1.3836894035339355, | |
| "logps/chosen": -155.07765197753906, | |
| "logps/rejected": -191.67471313476562, | |
| "loss": 0.6953, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.12750835716724396, | |
| "rewards/margins": 0.31851378083229065, | |
| "rewards/rejected": -0.1910054087638855, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.25087306678069987, | |
| "grad_norm": 107.56947817547281, | |
| "learning_rate": 1.9740745308420528e-07, | |
| "logits/chosen": -1.4428541660308838, | |
| "logits/rejected": -1.4522430896759033, | |
| "logps/chosen": -159.34136962890625, | |
| "logps/rejected": -201.03256225585938, | |
| "loss": 0.7085, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": 0.08158425986766815, | |
| "rewards/margins": 0.062202394008636475, | |
| "rewards/rejected": 0.019381869584321976, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.2531537310241608, | |
| "grad_norm": 83.84650346626758, | |
| "learning_rate": 1.9732070588429487e-07, | |
| "logits/chosen": -1.4249211549758911, | |
| "logits/rejected": -1.4040560722351074, | |
| "logps/chosen": -139.1720733642578, | |
| "logps/rejected": -150.09683227539062, | |
| "loss": 0.6528, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.25594449043273926, | |
| "rewards/margins": 0.27077823877334595, | |
| "rewards/rejected": -0.014833783730864525, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.2554343952676217, | |
| "grad_norm": 95.84521399355921, | |
| "learning_rate": 1.972325508733091e-07, | |
| "logits/chosen": -1.4068944454193115, | |
| "logits/rejected": -1.4884459972381592, | |
| "logps/chosen": -178.6918182373047, | |
| "logps/rejected": -178.89633178710938, | |
| "loss": 0.6725, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.17113548517227173, | |
| "rewards/margins": 0.35012179613113403, | |
| "rewards/rejected": -0.1789863258600235, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.2577150595110826, | |
| "grad_norm": 90.48418130257966, | |
| "learning_rate": 1.9714298932647098e-07, | |
| "logits/chosen": -1.3818295001983643, | |
| "logits/rejected": -1.4686387777328491, | |
| "logps/chosen": -203.4779510498047, | |
| "logps/rejected": -231.25755310058594, | |
| "loss": 0.6145, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.2646425664424896, | |
| "rewards/margins": 0.30655670166015625, | |
| "rewards/rejected": -0.04191412031650543, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.2599957237545435, | |
| "grad_norm": 96.02415026722944, | |
| "learning_rate": 1.9705202253935004e-07, | |
| "logits/chosen": -1.459298849105835, | |
| "logits/rejected": -1.4874234199523926, | |
| "logps/chosen": -190.43370056152344, | |
| "logps/rejected": -208.5654296875, | |
| "loss": 0.6259, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.2924637198448181, | |
| "rewards/margins": 0.3736059367656708, | |
| "rewards/rejected": -0.08114223182201385, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.2622763879980044, | |
| "grad_norm": 94.75504400212658, | |
| "learning_rate": 1.9695965182784343e-07, | |
| "logits/chosen": -1.37540864944458, | |
| "logits/rejected": -1.4327397346496582, | |
| "logps/chosen": -158.9178924560547, | |
| "logps/rejected": -187.00572204589844, | |
| "loss": 0.6413, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.031393878161907196, | |
| "rewards/margins": 0.2354445904493332, | |
| "rewards/rejected": -0.2040507048368454, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.2645570522414653, | |
| "grad_norm": 88.68651480780905, | |
| "learning_rate": 1.968658785281573e-07, | |
| "logits/chosen": -1.4091265201568604, | |
| "logits/rejected": -1.3750542402267456, | |
| "logps/chosen": -159.5924530029297, | |
| "logps/rejected": -157.25701904296875, | |
| "loss": 0.6956, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.05787750333547592, | |
| "rewards/margins": 0.07036175578832626, | |
| "rewards/rejected": -0.012484237551689148, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.2668377164849262, | |
| "grad_norm": 87.80878700459785, | |
| "learning_rate": 1.9677070399678693e-07, | |
| "logits/chosen": -1.3906813859939575, | |
| "logits/rejected": -1.4185580015182495, | |
| "logps/chosen": -170.7882843017578, | |
| "logps/rejected": -192.6103515625, | |
| "loss": 0.6498, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.026271583512425423, | |
| "rewards/margins": 0.08472327888011932, | |
| "rewards/rejected": -0.058451712131500244, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.2691183807283871, | |
| "grad_norm": 95.5482100990242, | |
| "learning_rate": 1.9667412961049754e-07, | |
| "logits/chosen": -1.4015161991119385, | |
| "logits/rejected": -1.4669052362442017, | |
| "logps/chosen": -180.34584045410156, | |
| "logps/rejected": -205.79164123535156, | |
| "loss": 0.6991, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 0.12791365385055542, | |
| "rewards/margins": 0.28303825855255127, | |
| "rewards/rejected": -0.15512457489967346, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.271399044971848, | |
| "grad_norm": 88.45630335400043, | |
| "learning_rate": 1.9657615676630418e-07, | |
| "logits/chosen": -1.2741804122924805, | |
| "logits/rejected": -1.2581013441085815, | |
| "logps/chosen": -184.23312377929688, | |
| "logps/rejected": -201.00103759765625, | |
| "loss": 0.6514, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.036131352186203, | |
| "rewards/margins": 0.508301317691803, | |
| "rewards/rejected": -0.4721699059009552, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.273679709215309, | |
| "grad_norm": 99.3871661426225, | |
| "learning_rate": 1.9647678688145159e-07, | |
| "logits/chosen": -1.327433466911316, | |
| "logits/rejected": -1.4025075435638428, | |
| "logps/chosen": -167.82005310058594, | |
| "logps/rejected": -194.77706909179688, | |
| "loss": 0.6544, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.01046083215624094, | |
| "rewards/margins": 0.2960438132286072, | |
| "rewards/rejected": -0.30650466680526733, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.2759603734587699, | |
| "grad_norm": 129.18874275194366, | |
| "learning_rate": 1.9637602139339355e-07, | |
| "logits/chosen": -1.1904761791229248, | |
| "logits/rejected": -1.2964377403259277, | |
| "logps/chosen": -198.83749389648438, | |
| "logps/rejected": -213.1676025390625, | |
| "loss": 0.7065, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.018888656049966812, | |
| "rewards/margins": 0.0817704051733017, | |
| "rewards/rejected": -0.06288175284862518, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.2782410377022308, | |
| "grad_norm": 78.46019369342993, | |
| "learning_rate": 1.9627386175977238e-07, | |
| "logits/chosen": -1.363703966140747, | |
| "logits/rejected": -1.385565161705017, | |
| "logps/chosen": -158.67543029785156, | |
| "logps/rejected": -171.18373107910156, | |
| "loss": 0.6328, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.0427667461335659, | |
| "rewards/margins": 0.24535968899726868, | |
| "rewards/rejected": -0.20259293913841248, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.2805217019456917, | |
| "grad_norm": 93.42637414011723, | |
| "learning_rate": 1.9617030945839747e-07, | |
| "logits/chosen": -1.4135775566101074, | |
| "logits/rejected": -1.482433557510376, | |
| "logps/chosen": -201.14869689941406, | |
| "logps/rejected": -240.91909790039062, | |
| "loss": 0.6572, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.14201030135154724, | |
| "rewards/margins": 0.3642757833003998, | |
| "rewards/rejected": -0.22226551175117493, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.2828023661891526, | |
| "grad_norm": 94.8034192613356, | |
| "learning_rate": 1.9606536598722431e-07, | |
| "logits/chosen": -1.3897498846054077, | |
| "logits/rejected": -1.4480516910552979, | |
| "logps/chosen": -211.97976684570312, | |
| "logps/rejected": -213.59344482421875, | |
| "loss": 0.6947, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.1269560605287552, | |
| "rewards/margins": 0.18549999594688416, | |
| "rewards/rejected": -0.05854398384690285, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.2850830304326135, | |
| "grad_norm": 107.48956854367718, | |
| "learning_rate": 1.9595903286433253e-07, | |
| "logits/chosen": -1.3192163705825806, | |
| "logits/rejected": -1.3453290462493896, | |
| "logps/chosen": -166.43765258789062, | |
| "logps/rejected": -185.8678436279297, | |
| "loss": 0.6446, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.22617535293102264, | |
| "rewards/margins": 0.19618834555149078, | |
| "rewards/rejected": -0.4223636984825134, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.2873636946760744, | |
| "grad_norm": 97.18344743194635, | |
| "learning_rate": 1.9585131162790395e-07, | |
| "logits/chosen": -1.2871501445770264, | |
| "logits/rejected": -1.3303256034851074, | |
| "logps/chosen": -169.388671875, | |
| "logps/rejected": -222.89968872070312, | |
| "loss": 0.6597, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.016080807894468307, | |
| "rewards/margins": 0.22609557211399078, | |
| "rewards/rejected": -0.21001477539539337, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.2896443589195353, | |
| "grad_norm": 102.3447949736702, | |
| "learning_rate": 1.9574220383620054e-07, | |
| "logits/chosen": -1.332809329032898, | |
| "logits/rejected": -1.3781144618988037, | |
| "logps/chosen": -192.14796447753906, | |
| "logps/rejected": -229.9279327392578, | |
| "loss": 0.6849, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": -0.12542986869812012, | |
| "rewards/margins": 0.2946632206439972, | |
| "rewards/rejected": -0.4200930893421173, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.29192502316299623, | |
| "grad_norm": 95.89945385250834, | |
| "learning_rate": 1.956317110675417e-07, | |
| "logits/chosen": -1.3515238761901855, | |
| "logits/rejected": -1.3907783031463623, | |
| "logps/chosen": -135.85316467285156, | |
| "logps/rejected": -168.83544921875, | |
| "loss": 0.6599, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.2081318199634552, | |
| "rewards/margins": 0.3642672896385193, | |
| "rewards/rejected": -0.15613549947738647, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.29420568740645714, | |
| "grad_norm": 104.38674521939737, | |
| "learning_rate": 1.9551983492028144e-07, | |
| "logits/chosen": -1.3251550197601318, | |
| "logits/rejected": -1.418371319770813, | |
| "logps/chosen": -172.90769958496094, | |
| "logps/rejected": -200.890380859375, | |
| "loss": 0.6987, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.051881127059459686, | |
| "rewards/margins": 0.18907570838928223, | |
| "rewards/rejected": -0.24095678329467773, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.29648635164991805, | |
| "grad_norm": 110.83873967012946, | |
| "learning_rate": 1.9540657701278533e-07, | |
| "logits/chosen": -1.4099152088165283, | |
| "logits/rejected": -1.4348920583724976, | |
| "logps/chosen": -196.08592224121094, | |
| "logps/rejected": -237.6608428955078, | |
| "loss": 0.6704, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.22441606223583221, | |
| "rewards/margins": 0.2391555905342102, | |
| "rewards/rejected": -0.4635716676712036, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.29876701589337895, | |
| "grad_norm": 103.49844878423605, | |
| "learning_rate": 1.952919389834071e-07, | |
| "logits/chosen": -1.2992289066314697, | |
| "logits/rejected": -1.3971199989318848, | |
| "logps/chosen": -173.28128051757812, | |
| "logps/rejected": -213.29893493652344, | |
| "loss": 0.6847, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.19963645935058594, | |
| "rewards/margins": 0.11320970952510834, | |
| "rewards/rejected": 0.0864267572760582, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.30104768013683986, | |
| "grad_norm": 99.13994519459824, | |
| "learning_rate": 1.9517592249046475e-07, | |
| "logits/chosen": -1.2953108549118042, | |
| "logits/rejected": -1.3212617635726929, | |
| "logps/chosen": -133.38534545898438, | |
| "logps/rejected": -153.64303588867188, | |
| "loss": 0.6807, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 0.2558751106262207, | |
| "rewards/margins": 0.31515762209892273, | |
| "rewards/rejected": -0.059282511472702026, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.30332834438030076, | |
| "grad_norm": 105.16807067574956, | |
| "learning_rate": 1.950585292122169e-07, | |
| "logits/chosen": -1.380258560180664, | |
| "logits/rejected": -1.381796956062317, | |
| "logps/chosen": -265.73822021484375, | |
| "logps/rejected": -279.2653503417969, | |
| "loss": 0.6065, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.020402366295456886, | |
| "rewards/margins": 0.3811405599117279, | |
| "rewards/rejected": -0.36073821783065796, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.30560900862376167, | |
| "grad_norm": 83.84465669976083, | |
| "learning_rate": 1.9493976084683813e-07, | |
| "logits/chosen": -1.3013784885406494, | |
| "logits/rejected": -1.3473682403564453, | |
| "logps/chosen": -143.68994140625, | |
| "logps/rejected": -161.75987243652344, | |
| "loss": 0.6565, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.09173958003520966, | |
| "rewards/margins": 0.3339024484157562, | |
| "rewards/rejected": -0.4256420433521271, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.3078896728672226, | |
| "grad_norm": 97.4032150220662, | |
| "learning_rate": 1.9481961911239475e-07, | |
| "logits/chosen": -1.3531267642974854, | |
| "logits/rejected": -1.3918657302856445, | |
| "logps/chosen": -158.61968994140625, | |
| "logps/rejected": -196.80203247070312, | |
| "loss": 0.6872, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.018977202475070953, | |
| "rewards/margins": 0.31147533655166626, | |
| "rewards/rejected": -0.330452561378479, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.3101703371106835, | |
| "grad_norm": 102.05098535340969, | |
| "learning_rate": 1.9469810574681968e-07, | |
| "logits/chosen": -1.2518314123153687, | |
| "logits/rejected": -1.373193383216858, | |
| "logps/chosen": -208.13992309570312, | |
| "logps/rejected": -265.2911376953125, | |
| "loss": 0.6599, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -0.10957180708646774, | |
| "rewards/margins": 0.4111281931400299, | |
| "rewards/rejected": -0.5207000374794006, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.3124510013541444, | |
| "grad_norm": 101.2119365968094, | |
| "learning_rate": 1.9457522250788756e-07, | |
| "logits/chosen": -1.4733508825302124, | |
| "logits/rejected": -1.4335081577301025, | |
| "logps/chosen": -152.27394104003906, | |
| "logps/rejected": -149.08934020996094, | |
| "loss": 0.6512, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": 0.0012259185314178467, | |
| "rewards/margins": 0.06475882232189178, | |
| "rewards/rejected": -0.06353290379047394, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.3147316655976053, | |
| "grad_norm": 102.70187300991479, | |
| "learning_rate": 1.9445097117318912e-07, | |
| "logits/chosen": -1.2339739799499512, | |
| "logits/rejected": -1.2878998517990112, | |
| "logps/chosen": -154.39866638183594, | |
| "logps/rejected": -169.81304931640625, | |
| "loss": 0.6069, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.030073221772909164, | |
| "rewards/margins": 0.06755417585372925, | |
| "rewards/rejected": -0.09762738645076752, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.3170123298410662, | |
| "grad_norm": 92.89132088481786, | |
| "learning_rate": 1.9432535354010545e-07, | |
| "logits/chosen": -1.381090521812439, | |
| "logits/rejected": -1.408646583557129, | |
| "logps/chosen": -193.48159790039062, | |
| "logps/rejected": -218.6549530029297, | |
| "loss": 0.721, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.018713481724262238, | |
| "rewards/margins": 0.14033427834510803, | |
| "rewards/rejected": -0.15904778242111206, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.3192929940845271, | |
| "grad_norm": 92.57611013262112, | |
| "learning_rate": 1.9419837142578226e-07, | |
| "logits/chosen": -1.3027143478393555, | |
| "logits/rejected": -1.2925523519515991, | |
| "logps/chosen": -159.863525390625, | |
| "logps/rejected": -182.41249084472656, | |
| "loss": 0.6453, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 0.16124297678470612, | |
| "rewards/margins": 0.37421250343322754, | |
| "rewards/rejected": -0.21296954154968262, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.321573658327988, | |
| "grad_norm": 88.34893118769534, | |
| "learning_rate": 1.9407002666710333e-07, | |
| "logits/chosen": -1.2514938116073608, | |
| "logits/rejected": -1.2840971946716309, | |
| "logps/chosen": -167.23568725585938, | |
| "logps/rejected": -185.19515991210938, | |
| "loss": 0.6776, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.0749707818031311, | |
| "rewards/margins": 0.2606170177459717, | |
| "rewards/rejected": -0.18564626574516296, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.3238543225714489, | |
| "grad_norm": 87.77224347782207, | |
| "learning_rate": 1.93940321120664e-07, | |
| "logits/chosen": -1.234694242477417, | |
| "logits/rejected": -1.245900273323059, | |
| "logps/chosen": -181.2189178466797, | |
| "logps/rejected": -201.54295349121094, | |
| "loss": 0.656, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.03384766727685928, | |
| "rewards/margins": 0.15053534507751465, | |
| "rewards/rejected": -0.18438303470611572, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.3261349868149098, | |
| "grad_norm": 102.48592883969225, | |
| "learning_rate": 1.9380925666274443e-07, | |
| "logits/chosen": -1.29789400100708, | |
| "logits/rejected": -1.3402178287506104, | |
| "logps/chosen": -200.9523468017578, | |
| "logps/rejected": -274.8287658691406, | |
| "loss": 0.6817, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.21383914351463318, | |
| "rewards/margins": 0.0331435352563858, | |
| "rewards/rejected": -0.24698270857334137, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.32841565105837073, | |
| "grad_norm": 98.03735612419584, | |
| "learning_rate": 1.9367683518928225e-07, | |
| "logits/chosen": -1.3940753936767578, | |
| "logits/rejected": -1.464486837387085, | |
| "logps/chosen": -124.9957504272461, | |
| "logps/rejected": -153.5186767578125, | |
| "loss": 0.6812, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.005818367004394531, | |
| "rewards/margins": 0.20061178505420685, | |
| "rewards/rejected": -0.20643013715744019, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.33069631530183163, | |
| "grad_norm": 103.78616077769101, | |
| "learning_rate": 1.935430586158454e-07, | |
| "logits/chosen": -1.1263610124588013, | |
| "logits/rejected": -1.1101529598236084, | |
| "logps/chosen": -117.66938018798828, | |
| "logps/rejected": -164.22933959960938, | |
| "loss": 0.6884, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.0009657144546508789, | |
| "rewards/margins": 0.10708783566951752, | |
| "rewards/rejected": -0.1080535426735878, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.33297697954529254, | |
| "grad_norm": 110.5063902175955, | |
| "learning_rate": 1.9340792887760412e-07, | |
| "logits/chosen": -1.324884295463562, | |
| "logits/rejected": -1.3702059984207153, | |
| "logps/chosen": -194.0547332763672, | |
| "logps/rejected": -223.9933624267578, | |
| "loss": 0.6334, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.0599634051322937, | |
| "rewards/margins": 0.17343786358833313, | |
| "rewards/rejected": -0.23340126872062683, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.3352576437887535, | |
| "grad_norm": 85.87924361401205, | |
| "learning_rate": 1.932714479293033e-07, | |
| "logits/chosen": -1.2896180152893066, | |
| "logits/rejected": -1.3805060386657715, | |
| "logps/chosen": -182.65460205078125, | |
| "logps/rejected": -212.5992889404297, | |
| "loss": 0.6151, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.10547913610935211, | |
| "rewards/margins": 0.20131757855415344, | |
| "rewards/rejected": -0.30679669976234436, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.3375383080322144, | |
| "grad_norm": 108.47177643694172, | |
| "learning_rate": 1.9313361774523386e-07, | |
| "logits/chosen": -1.4138604402542114, | |
| "logits/rejected": -1.402896523475647, | |
| "logps/chosen": -147.80795288085938, | |
| "logps/rejected": -165.24566650390625, | |
| "loss": 0.7377, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.21476852893829346, | |
| "rewards/margins": 0.08000420778989792, | |
| "rewards/rejected": -0.2947727143764496, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.3398189722756753, | |
| "grad_norm": 111.8896748854085, | |
| "learning_rate": 1.9299444031920436e-07, | |
| "logits/chosen": -1.2892051935195923, | |
| "logits/rejected": -1.380386233329773, | |
| "logps/chosen": -190.3002471923828, | |
| "logps/rejected": -246.29263305664062, | |
| "loss": 0.6347, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.08067283034324646, | |
| "rewards/margins": 0.4462183713912964, | |
| "rewards/rejected": -0.5268911719322205, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.3420996365191362, | |
| "grad_norm": 96.80869969762502, | |
| "learning_rate": 1.9285391766451217e-07, | |
| "logits/chosen": -1.3364348411560059, | |
| "logits/rejected": -1.3174083232879639, | |
| "logps/chosen": -161.74612426757812, | |
| "logps/rejected": -179.8140106201172, | |
| "loss": 0.669, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.1118321567773819, | |
| "rewards/margins": 0.23199215531349182, | |
| "rewards/rejected": -0.34382426738739014, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.3420996365191362, | |
| "eval_logits/chosen": -1.4281266927719116, | |
| "eval_logits/rejected": -1.4122414588928223, | |
| "eval_logps/chosen": -128.86663818359375, | |
| "eval_logps/rejected": -131.4814910888672, | |
| "eval_loss": 0.6797733306884766, | |
| "eval_rewards/accuracies": 0.5199999809265137, | |
| "eval_rewards/chosen": 0.10313291847705841, | |
| "eval_rewards/margins": 0.045006848871707916, | |
| "eval_rewards/rejected": 0.05812607705593109, | |
| "eval_runtime": 21.8566, | |
| "eval_samples_per_second": 4.575, | |
| "eval_steps_per_second": 1.144, | |
| "step": 300 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 1752, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 300, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |