| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.998691442030882, | |
| "eval_steps": 500, | |
| "global_step": 477, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.002093692750588851, | |
| "grad_norm": 3522.287948369491, | |
| "learning_rate": 1.0416666666666666e-08, | |
| "logits/chosen": 5002.53564453125, | |
| "logits/rejected": 4591.72021484375, | |
| "logps/chosen": -265.6396789550781, | |
| "logps/rejected": -206.22401428222656, | |
| "loss": 222.1204, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02093692750588851, | |
| "grad_norm": 3570.1637579903136, | |
| "learning_rate": 1.0416666666666667e-07, | |
| "logits/chosen": 5976.8369140625, | |
| "logits/rejected": 5015.12548828125, | |
| "logps/chosen": -292.30426025390625, | |
| "logps/rejected": -259.09771728515625, | |
| "loss": 261.6898, | |
| "rewards/accuracies": 0.4583333432674408, | |
| "rewards/chosen": 0.002565481700003147, | |
| "rewards/margins": 0.00030597145087085664, | |
| "rewards/rejected": 0.002259510336443782, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.04187385501177702, | |
| "grad_norm": 1740.0232034178, | |
| "learning_rate": 2.0833333333333333e-07, | |
| "logits/chosen": 5865.3740234375, | |
| "logits/rejected": 4845.28564453125, | |
| "logps/chosen": -283.95458984375, | |
| "logps/rejected": -243.2721710205078, | |
| "loss": 247.7889, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": 0.06638871878385544, | |
| "rewards/margins": 0.0017764812801033258, | |
| "rewards/rejected": 0.0646122470498085, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06281078251766553, | |
| "grad_norm": 1350.7689845691493, | |
| "learning_rate": 3.1249999999999997e-07, | |
| "logits/chosen": 5561.669921875, | |
| "logits/rejected": 4845.861328125, | |
| "logps/chosen": -256.84765625, | |
| "logps/rejected": -223.6597137451172, | |
| "loss": 235.7881, | |
| "rewards/accuracies": 0.528124988079071, | |
| "rewards/chosen": 0.15389783680438995, | |
| "rewards/margins": 0.002473007421940565, | |
| "rewards/rejected": 0.1514248251914978, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.08374771002355404, | |
| "grad_norm": 1024.0536736135682, | |
| "learning_rate": 4.1666666666666667e-07, | |
| "logits/chosen": 5555.7802734375, | |
| "logits/rejected": 4829.6064453125, | |
| "logps/chosen": -253.9397735595703, | |
| "logps/rejected": -237.85189819335938, | |
| "loss": 235.6024, | |
| "rewards/accuracies": 0.4781250059604645, | |
| "rewards/chosen": 0.20017290115356445, | |
| "rewards/margins": -0.005847611464560032, | |
| "rewards/rejected": 0.2060205191373825, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.10468463752944256, | |
| "grad_norm": 1139.3651978834764, | |
| "learning_rate": 4.999731868769026e-07, | |
| "logits/chosen": 5772.9248046875, | |
| "logits/rejected": 5017.52197265625, | |
| "logps/chosen": -256.2254943847656, | |
| "logps/rejected": -243.820556640625, | |
| "loss": 238.5566, | |
| "rewards/accuracies": 0.4625000059604645, | |
| "rewards/chosen": 0.2239224910736084, | |
| "rewards/margins": -0.01306430995464325, | |
| "rewards/rejected": 0.23698678612709045, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.12562156503533106, | |
| "grad_norm": 949.2424330626906, | |
| "learning_rate": 4.990353313429303e-07, | |
| "logits/chosen": 5492.53173828125, | |
| "logits/rejected": 4689.26611328125, | |
| "logps/chosen": -251.794921875, | |
| "logps/rejected": -223.1071014404297, | |
| "loss": 226.0058, | |
| "rewards/accuracies": 0.4593749940395355, | |
| "rewards/chosen": 0.266275554895401, | |
| "rewards/margins": -0.017442751675844193, | |
| "rewards/rejected": 0.2837182879447937, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.14655849254121958, | |
| "grad_norm": 954.6500714792113, | |
| "learning_rate": 4.967625656594781e-07, | |
| "logits/chosen": 5274.14453125, | |
| "logits/rejected": 4874.0908203125, | |
| "logps/chosen": -250.61135864257812, | |
| "logps/rejected": -238.37997436523438, | |
| "loss": 225.2396, | |
| "rewards/accuracies": 0.4625000059604645, | |
| "rewards/chosen": 0.27696576714515686, | |
| "rewards/margins": -0.007267093751579523, | |
| "rewards/rejected": 0.28423285484313965, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.16749542004710807, | |
| "grad_norm": 932.4135888762232, | |
| "learning_rate": 4.93167072587771e-07, | |
| "logits/chosen": 5357.58056640625, | |
| "logits/rejected": 4600.1162109375, | |
| "logps/chosen": -245.84445190429688, | |
| "logps/rejected": -223.32791137695312, | |
| "loss": 224.8782, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": 0.3024243712425232, | |
| "rewards/margins": 0.004978764336556196, | |
| "rewards/rejected": 0.2974456250667572, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.1884323475529966, | |
| "grad_norm": 1000.848976733069, | |
| "learning_rate": 4.882681251368548e-07, | |
| "logits/chosen": 5479.4873046875, | |
| "logits/rejected": 4642.76416015625, | |
| "logps/chosen": -245.2469482421875, | |
| "logps/rejected": -221.87753295898438, | |
| "loss": 225.3665, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": 0.32460635900497437, | |
| "rewards/margins": -0.0063544088043272495, | |
| "rewards/rejected": 0.33096081018447876, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.2093692750588851, | |
| "grad_norm": 988.1458235134902, | |
| "learning_rate": 4.820919832540181e-07, | |
| "logits/chosen": 5540.859375, | |
| "logits/rejected": 5128.6162109375, | |
| "logps/chosen": -247.3248748779297, | |
| "logps/rejected": -237.8451385498047, | |
| "loss": 221.1634, | |
| "rewards/accuracies": 0.43437498807907104, | |
| "rewards/chosen": 0.32543858885765076, | |
| "rewards/margins": -0.04126424342393875, | |
| "rewards/rejected": 0.3667028546333313, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.23030620256477363, | |
| "grad_norm": 968.9316114751973, | |
| "learning_rate": 4.7467175306295647e-07, | |
| "logits/chosen": 5704.5849609375, | |
| "logits/rejected": 4630.22412109375, | |
| "logps/chosen": -239.6920928955078, | |
| "logps/rejected": -225.50222778320312, | |
| "loss": 225.2349, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": 0.3493049740791321, | |
| "rewards/margins": 0.01561106275767088, | |
| "rewards/rejected": 0.33369389176368713, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2512431300706621, | |
| "grad_norm": 944.9549945207706, | |
| "learning_rate": 4.6604720940421207e-07, | |
| "logits/chosen": 5287.7451171875, | |
| "logits/rejected": 4921.56884765625, | |
| "logps/chosen": -233.7930908203125, | |
| "logps/rejected": -231.26657104492188, | |
| "loss": 229.5115, | |
| "rewards/accuracies": 0.5093749761581421, | |
| "rewards/chosen": 0.33385908603668213, | |
| "rewards/margins": -0.022233102470636368, | |
| "rewards/rejected": 0.356092244386673, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2721800575765506, | |
| "grad_norm": 1276.812459526022, | |
| "learning_rate": 4.5626458262912735e-07, | |
| "logits/chosen": 5355.15576171875, | |
| "logits/rejected": 4776.7998046875, | |
| "logps/chosen": -236.29244995117188, | |
| "logps/rejected": -233.30307006835938, | |
| "loss": 221.8784, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": 0.322248637676239, | |
| "rewards/margins": -0.03480709344148636, | |
| "rewards/rejected": 0.3570557236671448, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.29311698508243916, | |
| "grad_norm": 899.787701651727, | |
| "learning_rate": 4.453763107901675e-07, | |
| "logits/chosen": 5164.56640625, | |
| "logits/rejected": 4602.421875, | |
| "logps/chosen": -238.9141845703125, | |
| "logps/rejected": -216.98031616210938, | |
| "loss": 229.5713, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": 0.3299064338207245, | |
| "rewards/margins": -0.011618844233453274, | |
| "rewards/rejected": 0.34152525663375854, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.31405391258832765, | |
| "grad_norm": 1078.422410282313, | |
| "learning_rate": 4.3344075855595097e-07, | |
| "logits/chosen": 5692.89453125, | |
| "logits/rejected": 4704.0322265625, | |
| "logps/chosen": -256.35137939453125, | |
| "logps/rejected": -216.76150512695312, | |
| "loss": 214.8497, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": 0.32763591408729553, | |
| "rewards/margins": -0.019656497985124588, | |
| "rewards/rejected": 0.347292423248291, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.33499084009421615, | |
| "grad_norm": 912.0955307736757, | |
| "learning_rate": 4.2052190435769554e-07, | |
| "logits/chosen": 5401.46875, | |
| "logits/rejected": 4505.7705078125, | |
| "logps/chosen": -245.025146484375, | |
| "logps/rejected": -219.9535369873047, | |
| "loss": 229.4244, | |
| "rewards/accuracies": 0.5218750238418579, | |
| "rewards/chosen": 0.3956901431083679, | |
| "rewards/margins": 0.03144276887178421, | |
| "rewards/rejected": 0.3642473816871643, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3559277676001047, | |
| "grad_norm": 954.243569931957, | |
| "learning_rate": 4.0668899744407567e-07, | |
| "logits/chosen": 5647.60400390625, | |
| "logits/rejected": 4826.2236328125, | |
| "logps/chosen": -251.16824340820312, | |
| "logps/rejected": -207.45095825195312, | |
| "loss": 210.6987, | |
| "rewards/accuracies": 0.515625, | |
| "rewards/chosen": 0.3114745318889618, | |
| "rewards/margins": -0.020466681569814682, | |
| "rewards/rejected": 0.3319412171840668, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.3768646951059932, | |
| "grad_norm": 874.3975161439175, | |
| "learning_rate": 3.920161866827889e-07, | |
| "logits/chosen": 5324.474609375, | |
| "logits/rejected": 4590.5234375, | |
| "logps/chosen": -253.6103057861328, | |
| "logps/rejected": -212.4313201904297, | |
| "loss": 219.7222, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": 0.33684980869293213, | |
| "rewards/margins": -0.03864480182528496, | |
| "rewards/rejected": 0.3754945993423462, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.39780162261188173, | |
| "grad_norm": 890.3703764703043, | |
| "learning_rate": 3.765821230985757e-07, | |
| "logits/chosen": 5454.69677734375, | |
| "logits/rejected": 4614.9931640625, | |
| "logps/chosen": -234.9591827392578, | |
| "logps/rejected": -210.7035369873047, | |
| "loss": 217.6231, | |
| "rewards/accuracies": 0.46562498807907104, | |
| "rewards/chosen": 0.3479231297969818, | |
| "rewards/margins": -0.023874130100011826, | |
| "rewards/rejected": 0.3717973232269287, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4187385501177702, | |
| "grad_norm": 959.127096720511, | |
| "learning_rate": 3.604695382782159e-07, | |
| "logits/chosen": 5091.693359375, | |
| "logits/rejected": 4443.45703125, | |
| "logps/chosen": -221.46017456054688, | |
| "logps/rejected": -207.1079864501953, | |
| "loss": 222.0456, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": 0.32890671491622925, | |
| "rewards/margins": -0.028620053082704544, | |
| "rewards/rejected": 0.3575268089771271, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4396754776236587, | |
| "grad_norm": 977.743288922772, | |
| "learning_rate": 3.4376480090239047e-07, | |
| "logits/chosen": 5715.4208984375, | |
| "logits/rejected": 4880.05615234375, | |
| "logps/chosen": -251.9993896484375, | |
| "logps/rejected": -224.98220825195312, | |
| "loss": 225.9833, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": 0.3476799726486206, | |
| "rewards/margins": -0.019533518701791763, | |
| "rewards/rejected": 0.36721348762512207, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.46061240512954726, | |
| "grad_norm": 946.0452194833395, | |
| "learning_rate": 3.265574537815398e-07, | |
| "logits/chosen": 5137.0146484375, | |
| "logits/rejected": 4787.4306640625, | |
| "logps/chosen": -237.6749725341797, | |
| "logps/rejected": -222.62771606445312, | |
| "loss": 228.3879, | |
| "rewards/accuracies": 0.484375, | |
| "rewards/chosen": 0.3344074785709381, | |
| "rewards/margins": -0.04275032505393028, | |
| "rewards/rejected": 0.3771578073501587, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.48154933263543576, | |
| "grad_norm": 1021.1755452410538, | |
| "learning_rate": 3.0893973387735683e-07, | |
| "logits/chosen": 5326.5107421875, | |
| "logits/rejected": 4527.2001953125, | |
| "logps/chosen": -231.9087677001953, | |
| "logps/rejected": -209.6526641845703, | |
| "loss": 212.3704, | |
| "rewards/accuracies": 0.4906249940395355, | |
| "rewards/chosen": 0.36360007524490356, | |
| "rewards/margins": 0.00580978998914361, | |
| "rewards/rejected": 0.3577902913093567, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5024862601413242, | |
| "grad_norm": 931.7653919091188, | |
| "learning_rate": 2.910060778827554e-07, | |
| "logits/chosen": 5480.38134765625, | |
| "logits/rejected": 4673.6044921875, | |
| "logps/chosen": -230.7592010498047, | |
| "logps/rejected": -211.5625457763672, | |
| "loss": 222.4007, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": 0.3481771647930145, | |
| "rewards/margins": -0.034006811678409576, | |
| "rewards/rejected": 0.3821839988231659, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5234231876472127, | |
| "grad_norm": 893.8551271831342, | |
| "learning_rate": 2.7285261601056697e-07, | |
| "logits/chosen": 5610.1064453125, | |
| "logits/rejected": 4743.2900390625, | |
| "logps/chosen": -247.9258270263672, | |
| "logps/rejected": -224.725341796875, | |
| "loss": 218.3257, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.38022834062576294, | |
| "rewards/margins": 0.006729437503963709, | |
| "rewards/rejected": 0.37349894642829895, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5443601151531012, | |
| "grad_norm": 954.9590918842988, | |
| "learning_rate": 2.5457665670441937e-07, | |
| "logits/chosen": 5426.6806640625, | |
| "logits/rejected": 4926.78076171875, | |
| "logps/chosen": -248.59634399414062, | |
| "logps/rejected": -222.47549438476562, | |
| "loss": 216.7827, | |
| "rewards/accuracies": 0.4593749940395355, | |
| "rewards/chosen": 0.37472763657569885, | |
| "rewards/margins": -0.011340203694999218, | |
| "rewards/rejected": 0.38606783747673035, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5652970426589898, | |
| "grad_norm": 980.2338179672311, | |
| "learning_rate": 2.3627616503391812e-07, | |
| "logits/chosen": 5509.77490234375, | |
| "logits/rejected": 5063.6748046875, | |
| "logps/chosen": -237.590576171875, | |
| "logps/rejected": -221.60397338867188, | |
| "loss": 214.9091, | |
| "rewards/accuracies": 0.515625, | |
| "rewards/chosen": 0.37023288011550903, | |
| "rewards/margins": -0.017692191526293755, | |
| "rewards/rejected": 0.38792508840560913, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5862339701648783, | |
| "grad_norm": 882.5545431851971, | |
| "learning_rate": 2.1804923757009882e-07, | |
| "logits/chosen": 5514.5966796875, | |
| "logits/rejected": 4599.09765625, | |
| "logps/chosen": -245.1168212890625, | |
| "logps/rejected": -208.6852569580078, | |
| "loss": 216.9057, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": 0.35608822107315063, | |
| "rewards/margins": -0.02490018866956234, | |
| "rewards/rejected": 0.3809884190559387, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6071708976707668, | |
| "grad_norm": 1027.801450351746, | |
| "learning_rate": 1.9999357655598891e-07, | |
| "logits/chosen": 5770.27978515625, | |
| "logits/rejected": 5421.36083984375, | |
| "logps/chosen": -252.1714324951172, | |
| "logps/rejected": -246.72555541992188, | |
| "loss": 225.9217, | |
| "rewards/accuracies": 0.4593749940395355, | |
| "rewards/chosen": 0.3926265239715576, | |
| "rewards/margins": -0.11933328956365585, | |
| "rewards/rejected": 0.5119598507881165, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6281078251766553, | |
| "grad_norm": 983.8896941735266, | |
| "learning_rate": 1.8220596619089573e-07, | |
| "logits/chosen": 5291.69921875, | |
| "logits/rejected": 4989.41064453125, | |
| "logps/chosen": -236.5386962890625, | |
| "logps/rejected": -229.07266235351562, | |
| "loss": 226.711, | |
| "rewards/accuracies": 0.48124998807907104, | |
| "rewards/chosen": 0.3481127917766571, | |
| "rewards/margins": -0.07034964859485626, | |
| "rewards/rejected": 0.4184624254703522, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6490447526825438, | |
| "grad_norm": 972.552350701513, | |
| "learning_rate": 1.647817538357072e-07, | |
| "logits/chosen": 5905.65380859375, | |
| "logits/rejected": 5196.03125, | |
| "logps/chosen": -250.03976440429688, | |
| "logps/rejected": -230.9469757080078, | |
| "loss": 216.3124, | |
| "rewards/accuracies": 0.528124988079071, | |
| "rewards/chosen": 0.40789794921875, | |
| "rewards/margins": 0.021800601854920387, | |
| "rewards/rejected": 0.38609737157821655, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6699816801884323, | |
| "grad_norm": 923.1332436203306, | |
| "learning_rate": 1.478143389201113e-07, | |
| "logits/chosen": 5463.5693359375, | |
| "logits/rejected": 4749.58837890625, | |
| "logps/chosen": -235.2841796875, | |
| "logps/rejected": -219.2712860107422, | |
| "loss": 217.7832, | |
| "rewards/accuracies": 0.4593749940395355, | |
| "rewards/chosen": 0.3657926023006439, | |
| "rewards/margins": -0.039618875831365585, | |
| "rewards/rejected": 0.4054114818572998, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6909186076943209, | |
| "grad_norm": 965.0967343991006, | |
| "learning_rate": 1.3139467229135998e-07, | |
| "logits/chosen": 5287.45947265625, | |
| "logits/rejected": 4928.7021484375, | |
| "logps/chosen": -241.1490478515625, | |
| "logps/rejected": -224.2013397216797, | |
| "loss": 219.6799, | |
| "rewards/accuracies": 0.453125, | |
| "rewards/chosen": 0.35635194182395935, | |
| "rewards/margins": -0.044928766787052155, | |
| "rewards/rejected": 0.4012807309627533, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7118555352002094, | |
| "grad_norm": 915.8564126507746, | |
| "learning_rate": 1.1561076868822755e-07, | |
| "logits/chosen": 5126.6953125, | |
| "logits/rejected": 4468.2080078125, | |
| "logps/chosen": -218.98098754882812, | |
| "logps/rejected": -202.6676483154297, | |
| "loss": 211.1805, | |
| "rewards/accuracies": 0.515625, | |
| "rewards/chosen": 0.3858596980571747, | |
| "rewards/margins": 0.0005099147674627602, | |
| "rewards/rejected": 0.3853498101234436, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7327924627060979, | |
| "grad_norm": 914.8488378330155, | |
| "learning_rate": 1.0054723495346482e-07, | |
| "logits/chosen": 5789.267578125, | |
| "logits/rejected": 4720.3525390625, | |
| "logps/chosen": -260.73028564453125, | |
| "logps/rejected": -228.9587860107422, | |
| "loss": 222.0229, | |
| "rewards/accuracies": 0.5218750238418579, | |
| "rewards/chosen": 0.40184181928634644, | |
| "rewards/margins": 0.008045284077525139, | |
| "rewards/rejected": 0.39379650354385376, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7537293902119864, | |
| "grad_norm": 876.5437019392028, | |
| "learning_rate": 8.628481651367875e-08, | |
| "logits/chosen": 5485.41943359375, | |
| "logits/rejected": 4798.8369140625, | |
| "logps/chosen": -239.21206665039062, | |
| "logps/rejected": -218.1109161376953, | |
| "loss": 212.415, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.3851151466369629, | |
| "rewards/margins": -0.016298171132802963, | |
| "rewards/rejected": 0.40141329169273376, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7746663177178749, | |
| "grad_norm": 862.6497945839665, | |
| "learning_rate": 7.289996455765748e-08, | |
| "logits/chosen": 5362.8544921875, | |
| "logits/rejected": 4595.103515625, | |
| "logps/chosen": -254.4463653564453, | |
| "logps/rejected": -221.9584197998047, | |
| "loss": 216.0785, | |
| "rewards/accuracies": 0.528124988079071, | |
| "rewards/chosen": 0.34346675872802734, | |
| "rewards/margins": -0.006593307945877314, | |
| "rewards/rejected": 0.3500600755214691, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7956032452237635, | |
| "grad_norm": 932.2858564856045, | |
| "learning_rate": 6.046442623320145e-08, | |
| "logits/chosen": 5608.267578125, | |
| "logits/rejected": 4909.3837890625, | |
| "logps/chosen": -249.43460083007812, | |
| "logps/rejected": -211.33786010742188, | |
| "loss": 222.1436, | |
| "rewards/accuracies": 0.47187501192092896, | |
| "rewards/chosen": 0.3526684641838074, | |
| "rewards/margins": -0.04252925515174866, | |
| "rewards/rejected": 0.39519768953323364, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.816540172729652, | |
| "grad_norm": 957.8822198142358, | |
| "learning_rate": 4.904486005914027e-08, | |
| "logits/chosen": 5878.46630859375, | |
| "logits/rejected": 5202.1943359375, | |
| "logps/chosen": -251.19577026367188, | |
| "logps/rejected": -233.38088989257812, | |
| "loss": 223.1316, | |
| "rewards/accuracies": 0.49687498807907104, | |
| "rewards/chosen": 0.40553656220436096, | |
| "rewards/margins": -0.008421207778155804, | |
| "rewards/rejected": 0.41395774483680725, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8374771002355405, | |
| "grad_norm": 893.9334894466286, | |
| "learning_rate": 3.8702478614051345e-08, | |
| "logits/chosen": 5456.71435546875, | |
| "logits/rejected": 4864.9521484375, | |
| "logps/chosen": -250.51095581054688, | |
| "logps/rejected": -213.61160278320312, | |
| "loss": 220.8271, | |
| "rewards/accuracies": 0.4906249940395355, | |
| "rewards/chosen": 0.4039316773414612, | |
| "rewards/margins": -0.01853012666106224, | |
| "rewards/rejected": 0.4224618077278137, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8584140277414289, | |
| "grad_norm": 917.1537249021089, | |
| "learning_rate": 2.9492720416985e-08, | |
| "logits/chosen": 5479.6455078125, | |
| "logits/rejected": 4854.544921875, | |
| "logps/chosen": -236.3294677734375, | |
| "logps/rejected": -217.6251220703125, | |
| "loss": 221.5988, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": 0.4553787112236023, | |
| "rewards/margins": -0.1794721484184265, | |
| "rewards/rejected": 0.6348508596420288, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8793509552473174, | |
| "grad_norm": 963.815826471618, | |
| "learning_rate": 2.1464952759020856e-08, | |
| "logits/chosen": 5807.9521484375, | |
| "logits/rejected": 5086.6845703125, | |
| "logps/chosen": -253.22854614257812, | |
| "logps/rejected": -232.03866577148438, | |
| "loss": 218.956, | |
| "rewards/accuracies": 0.4312500059604645, | |
| "rewards/chosen": 0.38750648498535156, | |
| "rewards/margins": -0.016989920288324356, | |
| "rewards/rejected": 0.40449637174606323, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9002878827532059, | |
| "grad_norm": 1577.2268425412074, | |
| "learning_rate": 1.4662207078575684e-08, | |
| "logits/chosen": 5714.33251953125, | |
| "logits/rejected": 4623.4052734375, | |
| "logps/chosen": -248.07272338867188, | |
| "logps/rejected": -219.09396362304688, | |
| "loss": 211.869, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": 0.39529919624328613, | |
| "rewards/margins": 0.011650885455310345, | |
| "rewards/rejected": 0.3836483359336853, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9212248102590945, | |
| "grad_norm": 890.353373232557, | |
| "learning_rate": 9.12094829893642e-09, | |
| "logits/chosen": 5540.88134765625, | |
| "logits/rejected": 5293.84912109375, | |
| "logps/chosen": -243.41299438476562, | |
| "logps/rejected": -236.95791625976562, | |
| "loss": 216.8079, | |
| "rewards/accuracies": 0.4468750059604645, | |
| "rewards/chosen": 0.35130536556243896, | |
| "rewards/margins": -0.039575204253196716, | |
| "rewards/rejected": 0.3908805251121521, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.942161737764983, | |
| "grad_norm": 910.2914323549969, | |
| "learning_rate": 4.8708793644441086e-09, | |
| "logits/chosen": 5690.5556640625, | |
| "logits/rejected": 4916.1923828125, | |
| "logps/chosen": -252.1903839111328, | |
| "logps/rejected": -218.7462158203125, | |
| "loss": 213.594, | |
| "rewards/accuracies": 0.45625001192092896, | |
| "rewards/chosen": 0.3814484477043152, | |
| "rewards/margins": -0.04559388756752014, | |
| "rewards/rejected": 0.4270423352718353, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9630986652708715, | |
| "grad_norm": 930.7189662835303, | |
| "learning_rate": 1.9347820230782295e-09, | |
| "logits/chosen": 5878.43408203125, | |
| "logits/rejected": 4987.50341796875, | |
| "logps/chosen": -272.1065979003906, | |
| "logps/rejected": -240.8032989501953, | |
| "loss": 232.251, | |
| "rewards/accuracies": 0.4781250059604645, | |
| "rewards/chosen": 0.4089243412017822, | |
| "rewards/margins": -0.040169842541217804, | |
| "rewards/rejected": 0.44909414649009705, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.98403559277676, | |
| "grad_norm": 942.5169399812006, | |
| "learning_rate": 3.2839470889836627e-10, | |
| "logits/chosen": 5481.58837890625, | |
| "logits/rejected": 4879.6376953125, | |
| "logps/chosen": -240.8720245361328, | |
| "logps/rejected": -215.34402465820312, | |
| "loss": 214.6576, | |
| "rewards/accuracies": 0.4781250059604645, | |
| "rewards/chosen": 0.3996717035770416, | |
| "rewards/margins": -0.024916449561715126, | |
| "rewards/rejected": 0.424588143825531, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.998691442030882, | |
| "step": 477, | |
| "total_flos": 0.0, | |
| "train_loss": 222.93432777132878, | |
| "train_runtime": 12805.5892, | |
| "train_samples_per_second": 4.774, | |
| "train_steps_per_second": 0.037 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 477, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |