| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9984, | |
| "eval_steps": 500, | |
| "global_step": 468, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 1.0638297872340426e-07, | |
| "logits/chosen": 0.1422731578350067, | |
| "logits/rejected": 0.3313992917537689, | |
| "logps/chosen": -382.77789306640625, | |
| "logps/rejected": -331.4576416015625, | |
| "loss": 0.4045, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 1.0638297872340427e-06, | |
| "logits/chosen": 0.11944054067134857, | |
| "logits/rejected": 0.18962062895298004, | |
| "logps/chosen": -326.4944763183594, | |
| "logps/rejected": -291.0369873046875, | |
| "loss": 0.3779, | |
| "rewards/accuracies": 0.4236111044883728, | |
| "rewards/chosen": 5.325177335180342e-05, | |
| "rewards/margins": 5.4291256674332544e-05, | |
| "rewards/rejected": -1.039480139297666e-06, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 2.1276595744680853e-06, | |
| "logits/chosen": 0.12437532097101212, | |
| "logits/rejected": 0.17575038969516754, | |
| "logps/chosen": -355.1729736328125, | |
| "logps/rejected": -326.6689453125, | |
| "loss": 0.379, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": 0.0001221647544298321, | |
| "rewards/margins": 0.00025352081866003573, | |
| "rewards/rejected": -0.0001313560496782884, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 3.191489361702128e-06, | |
| "logits/chosen": 0.14455276727676392, | |
| "logits/rejected": 0.17237640917301178, | |
| "logps/chosen": -338.4366455078125, | |
| "logps/rejected": -344.5356140136719, | |
| "loss": 0.3714, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -8.463774429401383e-05, | |
| "rewards/margins": 0.00026104756398126483, | |
| "rewards/rejected": -0.00034568531555123627, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 4.255319148936171e-06, | |
| "logits/chosen": 0.09771748632192612, | |
| "logits/rejected": 0.20755767822265625, | |
| "logps/chosen": -372.2322692871094, | |
| "logps/rejected": -351.06793212890625, | |
| "loss": 0.3867, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -4.432376954355277e-05, | |
| "rewards/margins": 0.0008093343349173665, | |
| "rewards/rejected": -0.00085365807171911, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 4.999373573764188e-06, | |
| "logits/chosen": 0.07760350406169891, | |
| "logits/rejected": 0.154631569981575, | |
| "logps/chosen": -344.04986572265625, | |
| "logps/rejected": -310.27630615234375, | |
| "loss": 0.3668, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": 0.0006088110385462642, | |
| "rewards/margins": 0.0018962211906909943, | |
| "rewards/rejected": -0.0012874103849753737, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.988245838331339e-06, | |
| "logits/chosen": 0.14376303553581238, | |
| "logits/rejected": 0.1840187907218933, | |
| "logps/chosen": -348.7585754394531, | |
| "logps/rejected": -347.42633056640625, | |
| "loss": 0.3701, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": 0.0007757244748063385, | |
| "rewards/margins": 0.0029361725319176912, | |
| "rewards/rejected": -0.00216044788248837, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 4.963268819535228e-06, | |
| "logits/chosen": 0.10374955832958221, | |
| "logits/rejected": 0.20326051115989685, | |
| "logps/chosen": -330.6069641113281, | |
| "logps/rejected": -329.6276550292969, | |
| "loss": 0.3494, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": 0.0009276155615225434, | |
| "rewards/margins": 0.005808635149151087, | |
| "rewards/rejected": -0.004881019704043865, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 4.9245815365216115e-06, | |
| "logits/chosen": 0.11984933912754059, | |
| "logits/rejected": 0.19475053250789642, | |
| "logps/chosen": -341.2612609863281, | |
| "logps/rejected": -341.21063232421875, | |
| "loss": 0.3772, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 0.0007902610814198852, | |
| "rewards/margins": 0.014523616060614586, | |
| "rewards/rejected": -0.013733354397118092, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.872399318152594e-06, | |
| "logits/chosen": 0.09717285633087158, | |
| "logits/rejected": 0.19757065176963806, | |
| "logps/chosen": -383.49658203125, | |
| "logps/rejected": -375.48980712890625, | |
| "loss": 0.3443, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.013060875236988068, | |
| "rewards/margins": 0.018941307440400124, | |
| "rewards/rejected": -0.03200218454003334, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.807012604511542e-06, | |
| "logits/chosen": 0.09459100663661957, | |
| "logits/rejected": 0.172474667429924, | |
| "logps/chosen": -414.6573181152344, | |
| "logps/rejected": -423.6068420410156, | |
| "loss": 0.3207, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.0480206236243248, | |
| "rewards/margins": 0.05255536362528801, | |
| "rewards/rejected": -0.10057598352432251, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.728785330347771e-06, | |
| "logits/chosen": 0.17286503314971924, | |
| "logits/rejected": 0.2384227067232132, | |
| "logps/chosen": -535.0108032226562, | |
| "logps/rejected": -592.4139404296875, | |
| "loss": 0.3274, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.14901921153068542, | |
| "rewards/margins": 0.07979042828083038, | |
| "rewards/rejected": -0.22880962491035461, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.63815289945858e-06, | |
| "logits/chosen": 0.17661482095718384, | |
| "logits/rejected": 0.20918302237987518, | |
| "logps/chosen": -542.838134765625, | |
| "logps/rejected": -599.6415405273438, | |
| "loss": 0.3159, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.1861373484134674, | |
| "rewards/margins": 0.09151609241962433, | |
| "rewards/rejected": -0.27765345573425293, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 4.535619761282989e-06, | |
| "logits/chosen": 0.13282707333564758, | |
| "logits/rejected": 0.18375340104103088, | |
| "logps/chosen": -484.95953369140625, | |
| "logps/rejected": -634.0150146484375, | |
| "loss": 0.2988, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.15236204862594604, | |
| "rewards/margins": 0.14143863320350647, | |
| "rewards/rejected": -0.2938006520271301, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 4.42175660319555e-06, | |
| "logits/chosen": 0.18953588604927063, | |
| "logits/rejected": 0.19973225891590118, | |
| "logps/chosen": -518.3944091796875, | |
| "logps/rejected": -572.0684814453125, | |
| "loss": 0.3261, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.17030271887779236, | |
| "rewards/margins": 0.08826728165149689, | |
| "rewards/rejected": -0.25857001543045044, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 4.297197174127619e-06, | |
| "logits/chosen": 0.14177818596363068, | |
| "logits/rejected": 0.19645562767982483, | |
| "logps/chosen": -483.99383544921875, | |
| "logps/rejected": -587.1974487304688, | |
| "loss": 0.3197, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.1454397737979889, | |
| "rewards/margins": 0.09491118788719177, | |
| "rewards/rejected": -0.24035099148750305, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 4.162634757195418e-06, | |
| "logits/chosen": 0.1273745745420456, | |
| "logits/rejected": 0.21790139377117157, | |
| "logps/chosen": -491.3045959472656, | |
| "logps/rejected": -556.7537841796875, | |
| "loss": 0.3097, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.15721774101257324, | |
| "rewards/margins": 0.08056008815765381, | |
| "rewards/rejected": -0.23777782917022705, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 4.018818310967843e-06, | |
| "logits/chosen": 0.16942057013511658, | |
| "logits/rejected": 0.25549784302711487, | |
| "logps/chosen": -509.8619079589844, | |
| "logps/rejected": -641.41748046875, | |
| "loss": 0.3034, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.1676618903875351, | |
| "rewards/margins": 0.12202789634466171, | |
| "rewards/rejected": -0.2896897792816162, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.866548300851254e-06, | |
| "logits/chosen": 0.15910522639751434, | |
| "logits/rejected": 0.1666603833436966, | |
| "logps/chosen": -499.43865966796875, | |
| "logps/rejected": -589.9558715820312, | |
| "loss": 0.3048, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.17921218276023865, | |
| "rewards/margins": 0.09624715149402618, | |
| "rewards/rejected": -0.275459349155426, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 3.706672243793271e-06, | |
| "logits/chosen": 0.11776936054229736, | |
| "logits/rejected": 0.15903571248054504, | |
| "logps/chosen": -492.0912170410156, | |
| "logps/rejected": -582.832275390625, | |
| "loss": 0.3032, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.18043668568134308, | |
| "rewards/margins": 0.09369800984859467, | |
| "rewards/rejected": -0.27413469552993774, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 3.5400799911032357e-06, | |
| "logits/chosen": 0.1436721682548523, | |
| "logits/rejected": 0.1706967055797577, | |
| "logps/chosen": -506.0414123535156, | |
| "logps/rejected": -571.4830322265625, | |
| "loss": 0.3173, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.15588083863258362, | |
| "rewards/margins": 0.07620591670274734, | |
| "rewards/rejected": -0.23208674788475037, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 3.3676987756445894e-06, | |
| "logits/chosen": 0.1567540317773819, | |
| "logits/rejected": 0.21821483969688416, | |
| "logps/chosen": -516.2056884765625, | |
| "logps/rejected": -605.7388916015625, | |
| "loss": 0.2942, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.16281944513320923, | |
| "rewards/margins": 0.11313279718160629, | |
| "rewards/rejected": -0.2759522497653961, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 3.1904880509659397e-06, | |
| "logits/chosen": 0.09412068128585815, | |
| "logits/rejected": 0.26981237530708313, | |
| "logps/chosen": -573.1748046875, | |
| "logps/rejected": -600.4771728515625, | |
| "loss": 0.3023, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.1811535358428955, | |
| "rewards/margins": 0.09422539919614792, | |
| "rewards/rejected": -0.275378942489624, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 3.0094341510955697e-06, | |
| "logits/chosen": 0.11837532371282578, | |
| "logits/rejected": 0.19376713037490845, | |
| "logps/chosen": -569.38037109375, | |
| "logps/rejected": -616.151123046875, | |
| "loss": 0.3027, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.21316485106945038, | |
| "rewards/margins": 0.07764045149087906, | |
| "rewards/rejected": -0.29080528020858765, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 2.825544800722376e-06, | |
| "logits/chosen": 0.1659005582332611, | |
| "logits/rejected": 0.2277708798646927, | |
| "logps/chosen": -571.8699951171875, | |
| "logps/rejected": -628.5322875976562, | |
| "loss": 0.2852, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.20772342383861542, | |
| "rewards/margins": 0.10460670292377472, | |
| "rewards/rejected": -0.31233012676239014, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 2.639843506318899e-06, | |
| "logits/chosen": 0.1882917284965515, | |
| "logits/rejected": 0.2411785125732422, | |
| "logps/chosen": -557.4056396484375, | |
| "logps/rejected": -673.3493041992188, | |
| "loss": 0.2959, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.20086392760276794, | |
| "rewards/margins": 0.11475809663534164, | |
| "rewards/rejected": -0.3156220316886902, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 2.4533638594248094e-06, | |
| "logits/chosen": 0.13379403948783875, | |
| "logits/rejected": 0.19422146677970886, | |
| "logps/chosen": -565.0435180664062, | |
| "logps/rejected": -694.2730712890625, | |
| "loss": 0.2895, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.23057802021503448, | |
| "rewards/margins": 0.11758030951023102, | |
| "rewards/rejected": -0.3481583297252655, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2.2671437837980943e-06, | |
| "logits/chosen": 0.1472124606370926, | |
| "logits/rejected": 0.21264033019542694, | |
| "logps/chosen": -553.3011474609375, | |
| "logps/rejected": -670.6328125, | |
| "loss": 0.285, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.21868601441383362, | |
| "rewards/margins": 0.12228033691644669, | |
| "rewards/rejected": -0.3409663736820221, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 2.082219758453629e-06, | |
| "logits/chosen": 0.15410706400871277, | |
| "logits/rejected": 0.17184455692768097, | |
| "logps/chosen": -515.2255859375, | |
| "logps/rejected": -623.633544921875, | |
| "loss": 0.296, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.21407786011695862, | |
| "rewards/margins": 0.105091392993927, | |
| "rewards/rejected": -0.3191692531108856, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 1.899621048743019e-06, | |
| "logits/chosen": 0.08964169770479202, | |
| "logits/rejected": 0.16084159910678864, | |
| "logps/chosen": -526.2240600585938, | |
| "logps/rejected": -663.767822265625, | |
| "loss": 0.3056, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.21246090531349182, | |
| "rewards/margins": 0.12431824207305908, | |
| "rewards/rejected": -0.3367791175842285, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 1.7203639775848423e-06, | |
| "logits/chosen": 0.06412859261035919, | |
| "logits/rejected": 0.1526188850402832, | |
| "logps/chosen": -508.3192443847656, | |
| "logps/rejected": -625.1074829101562, | |
| "loss": 0.2862, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.18599674105644226, | |
| "rewards/margins": 0.11884822696447372, | |
| "rewards/rejected": -0.3048449456691742, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 1.5454462687309445e-06, | |
| "logits/chosen": 0.12014339864253998, | |
| "logits/rejected": 0.17034398019313812, | |
| "logps/chosen": -528.6392822265625, | |
| "logps/rejected": -658.5401611328125, | |
| "loss": 0.3008, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.2073144167661667, | |
| "rewards/margins": 0.1150335818529129, | |
| "rewards/rejected": -0.3223479390144348, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 1.3758414935535147e-06, | |
| "logits/chosen": 0.1499626636505127, | |
| "logits/rejected": 0.1570507287979126, | |
| "logps/chosen": -577.6300659179688, | |
| "logps/rejected": -674.1231079101562, | |
| "loss": 0.2964, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.2138519585132599, | |
| "rewards/margins": 0.10956285148859024, | |
| "rewards/rejected": -0.32341480255126953, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.2124936522614622e-06, | |
| "logits/chosen": 0.08589409291744232, | |
| "logits/rejected": 0.18073198199272156, | |
| "logps/chosen": -520.1068115234375, | |
| "logps/rejected": -660.3935546875, | |
| "loss": 0.284, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.17694085836410522, | |
| "rewards/margins": 0.15071728825569153, | |
| "rewards/rejected": -0.32765814661979675, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 1.0563119197063934e-06, | |
| "logits/chosen": 0.09931635111570358, | |
| "logits/rejected": 0.19644983112812042, | |
| "logps/chosen": -560.4113159179688, | |
| "logps/rejected": -655.904296875, | |
| "loss": 0.2886, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.21122217178344727, | |
| "rewards/margins": 0.09574859589338303, | |
| "rewards/rejected": -0.3069707751274109, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 9.081655850224449e-07, | |
| "logits/chosen": 0.10707017034292221, | |
| "logits/rejected": 0.1817469596862793, | |
| "logps/chosen": -537.7971801757812, | |
| "logps/rejected": -607.695068359375, | |
| "loss": 0.3062, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.21373042464256287, | |
| "rewards/margins": 0.07750532776117325, | |
| "rewards/rejected": -0.2912357449531555, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 7.688792132653111e-07, | |
| "logits/chosen": 0.09279724210500717, | |
| "logits/rejected": 0.14600899815559387, | |
| "logps/chosen": -518.6719970703125, | |
| "logps/rejected": -613.5050048828125, | |
| "loss": 0.3102, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.20901672542095184, | |
| "rewards/margins": 0.10807285457849503, | |
| "rewards/rejected": -0.31708958745002747, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 6.392280559802341e-07, | |
| "logits/chosen": 0.09189110994338989, | |
| "logits/rejected": 0.14240947365760803, | |
| "logps/chosen": -499.902587890625, | |
| "logps/rejected": -645.5587158203125, | |
| "loss": 0.2901, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.19295868277549744, | |
| "rewards/margins": 0.1267748475074768, | |
| "rewards/rejected": -0.31973353028297424, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 5.199337362431792e-07, | |
| "logits/chosen": 0.10996093600988388, | |
| "logits/rejected": 0.17150555551052094, | |
| "logps/chosen": -541.5203857421875, | |
| "logps/rejected": -660.8944091796875, | |
| "loss": 0.2869, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.19876644015312195, | |
| "rewards/margins": 0.13654473423957825, | |
| "rewards/rejected": -0.3353111743927002, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 4.1166023219176176e-07, | |
| "logits/chosen": 0.08110027015209198, | |
| "logits/rejected": 0.1370091438293457, | |
| "logps/chosen": -551.21923828125, | |
| "logps/rejected": -639.4774780273438, | |
| "loss": 0.2989, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.2069006711244583, | |
| "rewards/margins": 0.1049899011850357, | |
| "rewards/rejected": -0.31189054250717163, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 3.150101814011136e-07, | |
| "logits/chosen": 0.09940965473651886, | |
| "logits/rejected": 0.14982450008392334, | |
| "logps/chosen": -575.6122436523438, | |
| "logps/rejected": -670.1174926757812, | |
| "loss": 0.3034, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.2270297110080719, | |
| "rewards/margins": 0.09964431822299957, | |
| "rewards/rejected": -0.3266740143299103, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 2.3052152667409289e-07, | |
| "logits/chosen": 0.12830981612205505, | |
| "logits/rejected": 0.12991704046726227, | |
| "logps/chosen": -512.4361572265625, | |
| "logps/rejected": -703.7268676757812, | |
| "loss": 0.2663, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.19513638317584991, | |
| "rewards/margins": 0.15631332993507385, | |
| "rewards/rejected": -0.3514496684074402, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.5866452191498488e-07, | |
| "logits/chosen": 0.12449488788843155, | |
| "logits/rejected": 0.12310103327035904, | |
| "logps/chosen": -507.05352783203125, | |
| "logps/rejected": -690.96337890625, | |
| "loss": 0.3093, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.19581075012683868, | |
| "rewards/margins": 0.1489998996257782, | |
| "rewards/rejected": -0.3448106646537781, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 9.983911475163727e-08, | |
| "logits/chosen": 0.12271595001220703, | |
| "logits/rejected": 0.11245056241750717, | |
| "logps/chosen": -534.17724609375, | |
| "logps/rejected": -601.7576904296875, | |
| "loss": 0.2942, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.20346157252788544, | |
| "rewards/margins": 0.10961834341287613, | |
| "rewards/rejected": -0.3130798935890198, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 5.437272047405712e-08, | |
| "logits/chosen": 0.10363437235355377, | |
| "logits/rejected": 0.19413235783576965, | |
| "logps/chosen": -529.4188232421875, | |
| "logps/rejected": -675.1129150390625, | |
| "loss": 0.2833, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.20683610439300537, | |
| "rewards/margins": 0.13403114676475525, | |
| "rewards/rejected": -0.3408672511577606, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2.251839967945535e-08, | |
| "logits/chosen": 0.08495406061410904, | |
| "logits/rejected": 0.15138773620128632, | |
| "logps/chosen": -553.8279418945312, | |
| "logps/rejected": -693.7943725585938, | |
| "loss": 0.3031, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.22638258337974548, | |
| "rewards/margins": 0.11836589872837067, | |
| "rewards/rejected": -0.34474846720695496, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 4.453449766758933e-09, | |
| "logits/chosen": 0.11232316493988037, | |
| "logits/rejected": 0.1733480840921402, | |
| "logps/chosen": -540.5206298828125, | |
| "logps/rejected": -626.1546630859375, | |
| "loss": 0.2912, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.18953736126422882, | |
| "rewards/margins": 0.12424160540103912, | |
| "rewards/rejected": -0.31377896666526794, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 468, | |
| "total_flos": 0.0, | |
| "train_loss": 0.105732237171923, | |
| "train_runtime": 2052.0868, | |
| "train_samples_per_second": 7.31, | |
| "train_steps_per_second": 0.228 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 468, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "total_flos": 0.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |