| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9881235154394297, | |
| "eval_steps": 500, | |
| "global_step": 234, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.06334125098970704, | |
| "grad_norm": 302.0, | |
| "learning_rate": 1.0416666666666668e-05, | |
| "log_odds_chosen": 3.157041072845459, | |
| "log_odds_ratio": -9.87833023071289, | |
| "logps/chosen": -23.06415557861328, | |
| "logps/rejected": -26.220584869384766, | |
| "loss": 170.2786, | |
| "nll_loss": 10.642413139343262, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -1.1532078981399536, | |
| "rewards/margins": 0.15782156586647034, | |
| "rewards/rejected": -1.3110291957855225, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.12668250197941408, | |
| "grad_norm": 260.0, | |
| "learning_rate": 2.0833333333333336e-05, | |
| "log_odds_chosen": 6.424581050872803, | |
| "log_odds_ratio": -9.701116561889648, | |
| "logps/chosen": -22.176563262939453, | |
| "logps/rejected": -28.598962783813477, | |
| "loss": 168.5609, | |
| "nll_loss": 10.535055160522461, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -1.1088283061981201, | |
| "rewards/margins": 0.32112011313438416, | |
| "rewards/rejected": -1.4299482107162476, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.19002375296912113, | |
| "grad_norm": 370.0, | |
| "learning_rate": 3.125e-05, | |
| "log_odds_chosen": 5.751986503601074, | |
| "log_odds_ratio": -6.494493007659912, | |
| "logps/chosen": -18.182918548583984, | |
| "logps/rejected": -23.933364868164062, | |
| "loss": 135.9466, | |
| "nll_loss": 8.496664047241211, | |
| "rewards/accuracies": 0.546875, | |
| "rewards/chosen": -0.9091460108757019, | |
| "rewards/margins": 0.28752225637435913, | |
| "rewards/rejected": -1.196668267250061, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.25336500395882816, | |
| "grad_norm": 174.0, | |
| "learning_rate": 4.166666666666667e-05, | |
| "log_odds_chosen": 1.3518788814544678, | |
| "log_odds_ratio": -3.2631962299346924, | |
| "logps/chosen": -8.165756225585938, | |
| "logps/rejected": -9.506813049316406, | |
| "loss": 67.1286, | |
| "nll_loss": 4.195539951324463, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.40828782320022583, | |
| "rewards/margins": 0.06705284118652344, | |
| "rewards/rejected": -0.47534069418907166, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.3167062549485352, | |
| "grad_norm": 226.0, | |
| "learning_rate": 4.999720254525684e-05, | |
| "log_odds_chosen": 0.28868618607521057, | |
| "log_odds_ratio": -0.8123720288276672, | |
| "logps/chosen": -1.6918294429779053, | |
| "logps/rejected": -1.9460710287094116, | |
| "loss": 32.5673, | |
| "nll_loss": 2.0354533195495605, | |
| "rewards/accuracies": 0.565625011920929, | |
| "rewards/chosen": -0.08459147810935974, | |
| "rewards/margins": 0.012712081894278526, | |
| "rewards/rejected": -0.09730355441570282, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.38004750593824227, | |
| "grad_norm": 85.5, | |
| "learning_rate": 4.989935734988098e-05, | |
| "log_odds_chosen": 0.13547295331954956, | |
| "log_odds_ratio": -0.8609800338745117, | |
| "logps/chosen": -1.710571050643921, | |
| "logps/rejected": -1.8393272161483765, | |
| "loss": 31.734, | |
| "nll_loss": 1.9833742380142212, | |
| "rewards/accuracies": 0.546875, | |
| "rewards/chosen": -0.08552855998277664, | |
| "rewards/margins": 0.0064378054812550545, | |
| "rewards/rejected": -0.09196637570858002, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.4433887569279493, | |
| "grad_norm": 604.0, | |
| "learning_rate": 4.9662264848088034e-05, | |
| "log_odds_chosen": 0.038943640887737274, | |
| "log_odds_ratio": -0.864970862865448, | |
| "logps/chosen": -1.499442219734192, | |
| "logps/rejected": -1.5386669635772705, | |
| "loss": 29.0551, | |
| "nll_loss": 1.815943956375122, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": -0.07497210055589676, | |
| "rewards/margins": 0.0019612452015280724, | |
| "rewards/rejected": -0.076933354139328, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.5067300079176563, | |
| "grad_norm": 69.5, | |
| "learning_rate": 4.928725095732169e-05, | |
| "log_odds_chosen": 0.14633053541183472, | |
| "log_odds_ratio": -0.7850778102874756, | |
| "logps/chosen": -1.3732013702392578, | |
| "logps/rejected": -1.5010284185409546, | |
| "loss": 27.5485, | |
| "nll_loss": 1.7217798233032227, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.06866006553173065, | |
| "rewards/margins": 0.006391354836523533, | |
| "rewards/rejected": -0.07505142688751221, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5700712589073634, | |
| "grad_norm": 42.25, | |
| "learning_rate": 4.877641290737884e-05, | |
| "log_odds_chosen": 0.1304132044315338, | |
| "log_odds_ratio": -0.7902301549911499, | |
| "logps/chosen": -1.3314447402954102, | |
| "logps/rejected": -1.4372903108596802, | |
| "loss": 26.0469, | |
| "nll_loss": 1.6279296875, | |
| "rewards/accuracies": 0.534375011920929, | |
| "rewards/chosen": -0.06657223403453827, | |
| "rewards/margins": 0.0052922810427844524, | |
| "rewards/rejected": -0.07186451554298401, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.6334125098970704, | |
| "grad_norm": 31.0, | |
| "learning_rate": 4.813260751184992e-05, | |
| "log_odds_chosen": 0.274090439081192, | |
| "log_odds_ratio": -0.7146386504173279, | |
| "logps/chosen": -1.2023499011993408, | |
| "logps/rejected": -1.4235477447509766, | |
| "loss": 24.6248, | |
| "nll_loss": 1.539048433303833, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.06011749058961868, | |
| "rewards/margins": 0.011059902608394623, | |
| "rewards/rejected": -0.0711773931980133, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6967537608867775, | |
| "grad_norm": 50.75, | |
| "learning_rate": 4.7359435191658425e-05, | |
| "log_odds_chosen": 0.10976365953683853, | |
| "log_odds_ratio": -0.7592318058013916, | |
| "logps/chosen": -1.2341073751449585, | |
| "logps/rejected": -1.3237587213516235, | |
| "loss": 24.4883, | |
| "nll_loss": 1.5305156707763672, | |
| "rewards/accuracies": 0.5093749761581421, | |
| "rewards/chosen": -0.06170536205172539, | |
| "rewards/margins": 0.004482570104300976, | |
| "rewards/rejected": -0.06618793308734894, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7600950118764845, | |
| "grad_norm": 73.5, | |
| "learning_rate": 4.6461219840046654e-05, | |
| "log_odds_chosen": 0.19487467408180237, | |
| "log_odds_ratio": -0.701287567615509, | |
| "logps/chosen": -1.1585313081741333, | |
| "logps/rejected": -1.3133569955825806, | |
| "loss": 24.4529, | |
| "nll_loss": 1.5283066034317017, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.05792655795812607, | |
| "rewards/margins": 0.007741288281977177, | |
| "rewards/rejected": -0.06566785275936127, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.8234362628661916, | |
| "grad_norm": 52.25, | |
| "learning_rate": 4.544298464161079e-05, | |
| "log_odds_chosen": 0.24091807007789612, | |
| "log_odds_ratio": -0.6930054426193237, | |
| "logps/chosen": -1.125671148300171, | |
| "logps/rejected": -1.3065881729125977, | |
| "loss": 24.4496, | |
| "nll_loss": 1.5281010866165161, | |
| "rewards/accuracies": 0.578125, | |
| "rewards/chosen": -0.05628355219960213, | |
| "rewards/margins": 0.009045850485563278, | |
| "rewards/rejected": -0.06532939523458481, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8867775138558986, | |
| "grad_norm": 43.75, | |
| "learning_rate": 4.431042398061499e-05, | |
| "log_odds_chosen": 0.09587414562702179, | |
| "log_odds_ratio": -0.7694802284240723, | |
| "logps/chosen": -1.169614553451538, | |
| "logps/rejected": -1.2482807636260986, | |
| "loss": 24.5127, | |
| "nll_loss": 1.5320452451705933, | |
| "rewards/accuracies": 0.5218750238418579, | |
| "rewards/chosen": -0.058480728417634964, | |
| "rewards/margins": 0.003933308180421591, | |
| "rewards/rejected": -0.06241403892636299, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9501187648456056, | |
| "grad_norm": 38.5, | |
| "learning_rate": 4.306987159568479e-05, | |
| "log_odds_chosen": 0.16929097473621368, | |
| "log_odds_ratio": -0.7172764539718628, | |
| "logps/chosen": -1.0788049697875977, | |
| "logps/rejected": -1.1940680742263794, | |
| "loss": 22.7263, | |
| "nll_loss": 1.4203910827636719, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.05394024774432182, | |
| "rewards/margins": 0.005763149820268154, | |
| "rewards/rejected": -0.05970339849591255, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.0253365003958828, | |
| "grad_norm": 181.0, | |
| "learning_rate": 4.172826515897146e-05, | |
| "log_odds_chosen": 0.32294443249702454, | |
| "log_odds_ratio": -0.6628458499908447, | |
| "logps/chosen": -1.0215741395950317, | |
| "logps/rejected": -1.248105525970459, | |
| "loss": 25.0471, | |
| "nll_loss": 1.3915075063705444, | |
| "rewards/accuracies": 0.605555534362793, | |
| "rewards/chosen": -0.05107870697975159, | |
| "rewards/margins": 0.011326568201184273, | |
| "rewards/rejected": -0.06240528076887131, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0886777513855899, | |
| "grad_norm": 24.625, | |
| "learning_rate": 4.029310747787516e-05, | |
| "log_odds_chosen": 0.540235698223114, | |
| "log_odds_ratio": -0.5549238324165344, | |
| "logps/chosen": -0.8654016256332397, | |
| "logps/rejected": -1.20816969871521, | |
| "loss": 19.002, | |
| "nll_loss": 1.1876252889633179, | |
| "rewards/accuracies": 0.7093750238418579, | |
| "rewards/chosen": -0.043270088732242584, | |
| "rewards/margins": 0.01713840290904045, | |
| "rewards/rejected": -0.06040848419070244, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.152019002375297, | |
| "grad_norm": 31.5, | |
| "learning_rate": 3.8772424536302564e-05, | |
| "log_odds_chosen": 0.6173906326293945, | |
| "log_odds_ratio": -0.5311087965965271, | |
| "logps/chosen": -0.7949373722076416, | |
| "logps/rejected": -1.1982877254486084, | |
| "loss": 18.4072, | |
| "nll_loss": 1.1504485607147217, | |
| "rewards/accuracies": 0.7281249761581421, | |
| "rewards/chosen": -0.03974687308073044, | |
| "rewards/margins": 0.02016752026975155, | |
| "rewards/rejected": -0.05991438776254654, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.2153602533650039, | |
| "grad_norm": 58.5, | |
| "learning_rate": 3.717472061010918e-05, | |
| "log_odds_chosen": 0.5371578931808472, | |
| "log_odds_ratio": -0.5785112380981445, | |
| "logps/chosen": -0.8111588358879089, | |
| "logps/rejected": -1.1321982145309448, | |
| "loss": 18.4027, | |
| "nll_loss": 1.1501662731170654, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.040557943284511566, | |
| "rewards/margins": 0.016051968559622765, | |
| "rewards/rejected": -0.05660990998148918, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.278701504354711, | |
| "grad_norm": 43.25, | |
| "learning_rate": 3.550893070773914e-05, | |
| "log_odds_chosen": 0.5756627321243286, | |
| "log_odds_ratio": -0.5552070736885071, | |
| "logps/chosen": -0.8670624494552612, | |
| "logps/rejected": -1.227243423461914, | |
| "loss": 18.8739, | |
| "nll_loss": 1.1796166896820068, | |
| "rewards/accuracies": 0.690625011920929, | |
| "rewards/chosen": -0.0433531180024147, | |
| "rewards/margins": 0.01800905168056488, | |
| "rewards/rejected": -0.061362169682979584, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.342042755344418, | |
| "grad_norm": 24.25, | |
| "learning_rate": 3.378437060203357e-05, | |
| "log_odds_chosen": 0.4615544378757477, | |
| "log_odds_ratio": -0.5932336449623108, | |
| "logps/chosen": -0.8293741941452026, | |
| "logps/rejected": -1.1050505638122559, | |
| "loss": 18.096, | |
| "nll_loss": 1.1310014724731445, | |
| "rewards/accuracies": 0.671875, | |
| "rewards/chosen": -0.04146871715784073, | |
| "rewards/margins": 0.01378381997346878, | |
| "rewards/rejected": -0.05525252968072891, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.405384006334125, | |
| "grad_norm": 54.5, | |
| "learning_rate": 3.201068473265007e-05, | |
| "log_odds_chosen": 0.49423256516456604, | |
| "log_odds_ratio": -0.5842307209968567, | |
| "logps/chosen": -0.8439785242080688, | |
| "logps/rejected": -1.1538585424423218, | |
| "loss": 19.262, | |
| "nll_loss": 1.2038729190826416, | |
| "rewards/accuracies": 0.671875, | |
| "rewards/chosen": -0.04219892621040344, | |
| "rewards/margins": 0.015493998304009438, | |
| "rewards/rejected": -0.05769292637705803, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.468725257323832, | |
| "grad_norm": 28.5, | |
| "learning_rate": 3.0197792270443982e-05, | |
| "log_odds_chosen": 0.5744311809539795, | |
| "log_odds_ratio": -0.5467215776443481, | |
| "logps/chosen": -0.7895219326019287, | |
| "logps/rejected": -1.148335337638855, | |
| "loss": 18.1877, | |
| "nll_loss": 1.1367297172546387, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.039476100355386734, | |
| "rewards/margins": 0.017940666526556015, | |
| "rewards/rejected": -0.05741675943136215, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.5320665083135392, | |
| "grad_norm": 56.75, | |
| "learning_rate": 2.8355831645441388e-05, | |
| "log_odds_chosen": 0.6337670683860779, | |
| "log_odds_ratio": -0.5344417691230774, | |
| "logps/chosen": -0.7838994860649109, | |
| "logps/rejected": -1.1551088094711304, | |
| "loss": 18.0842, | |
| "nll_loss": 1.1302621364593506, | |
| "rewards/accuracies": 0.6968749761581421, | |
| "rewards/chosen": -0.0391949787735939, | |
| "rewards/margins": 0.018560461699962616, | |
| "rewards/rejected": -0.05775544047355652, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5954077593032463, | |
| "grad_norm": 41.75, | |
| "learning_rate": 2.649510384862586e-05, | |
| "log_odds_chosen": 0.5285617709159851, | |
| "log_odds_ratio": -0.5650445818901062, | |
| "logps/chosen": -0.7971245646476746, | |
| "logps/rejected": -1.1233524084091187, | |
| "loss": 18.0659, | |
| "nll_loss": 1.12911856174469, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.03985622525215149, | |
| "rewards/margins": 0.016311392188072205, | |
| "rewards/rejected": -0.05616762116551399, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6587490102929534, | |
| "grad_norm": 52.25, | |
| "learning_rate": 2.4626014824618415e-05, | |
| "log_odds_chosen": 0.4448211193084717, | |
| "log_odds_ratio": -0.5954711437225342, | |
| "logps/chosen": -0.7822165489196777, | |
| "logps/rejected": -1.0468573570251465, | |
| "loss": 18.0461, | |
| "nll_loss": 1.1278802156448364, | |
| "rewards/accuracies": 0.653124988079071, | |
| "rewards/chosen": -0.039110831916332245, | |
| "rewards/margins": 0.013232040219008923, | |
| "rewards/rejected": -0.052342869341373444, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.7220902612826603, | |
| "grad_norm": 38.75, | |
| "learning_rate": 2.2759017277414166e-05, | |
| "log_odds_chosen": 0.532827615737915, | |
| "log_odds_ratio": -0.5476608276367188, | |
| "logps/chosen": -0.8188691139221191, | |
| "logps/rejected": -1.1497671604156494, | |
| "loss": 18.2174, | |
| "nll_loss": 1.1385877132415771, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.040943462401628494, | |
| "rewards/margins": 0.016544902697205544, | |
| "rewards/rejected": -0.05748836323618889, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.7854315122723674, | |
| "grad_norm": 38.25, | |
| "learning_rate": 2.090455221462156e-05, | |
| "log_odds_chosen": 0.551643967628479, | |
| "log_odds_ratio": -0.5697082281112671, | |
| "logps/chosen": -0.7936131358146667, | |
| "logps/rejected": -1.1227937936782837, | |
| "loss": 17.6124, | |
| "nll_loss": 1.1007720232009888, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.03968065604567528, | |
| "rewards/margins": 0.016459036618471146, | |
| "rewards/rejected": -0.05613969638943672, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.8487727632620743, | |
| "grad_norm": 21.25, | |
| "learning_rate": 1.9072990557112564e-05, | |
| "log_odds_chosen": 0.7169318795204163, | |
| "log_odds_ratio": -0.5285918712615967, | |
| "logps/chosen": -0.768589198589325, | |
| "logps/rejected": -1.2199491262435913, | |
| "loss": 17.5301, | |
| "nll_loss": 1.0956323146820068, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.03842946141958237, | |
| "rewards/margins": 0.022568000480532646, | |
| "rewards/rejected": -0.06099746376276016, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.9121140142517814, | |
| "grad_norm": 20.625, | |
| "learning_rate": 1.7274575140626318e-05, | |
| "log_odds_chosen": 0.5951381921768188, | |
| "log_odds_ratio": -0.5281034708023071, | |
| "logps/chosen": -0.7978143692016602, | |
| "logps/rejected": -1.146867036819458, | |
| "loss": 17.1084, | |
| "nll_loss": 1.0692765712738037, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.03989071771502495, | |
| "rewards/margins": 0.01745263859629631, | |
| "rewards/rejected": -0.05734335258603096, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9754552652414885, | |
| "grad_norm": 25.875, | |
| "learning_rate": 1.551936343367679e-05, | |
| "log_odds_chosen": 0.5835083723068237, | |
| "log_odds_ratio": -0.5596362948417664, | |
| "logps/chosen": -0.793925404548645, | |
| "logps/rejected": -1.142182469367981, | |
| "loss": 18.0845, | |
| "nll_loss": 1.1302804946899414, | |
| "rewards/accuracies": 0.690625011920929, | |
| "rewards/chosen": -0.03969626873731613, | |
| "rewards/margins": 0.017412852495908737, | |
| "rewards/rejected": -0.05710912495851517, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.0506730007917655, | |
| "grad_norm": 20.5, | |
| "learning_rate": 1.3817171292109183e-05, | |
| "log_odds_chosen": 0.8345562219619751, | |
| "log_odds_ratio": -0.46272528171539307, | |
| "logps/chosen": -0.6592908501625061, | |
| "logps/rejected": -1.1182148456573486, | |
| "loss": 17.5074, | |
| "nll_loss": 0.9726357460021973, | |
| "rewards/accuracies": 0.7972221970558167, | |
| "rewards/chosen": -0.032964546233415604, | |
| "rewards/margins": 0.022946199402213097, | |
| "rewards/rejected": -0.055910736322402954, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.1140142517814726, | |
| "grad_norm": 39.5, | |
| "learning_rate": 1.217751806485235e-05, | |
| "log_odds_chosen": 1.2292115688323975, | |
| "log_odds_ratio": -0.3634593188762665, | |
| "logps/chosen": -0.5913038849830627, | |
| "logps/rejected": -1.2762850522994995, | |
| "loss": 14.5297, | |
| "nll_loss": 0.9081088304519653, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -0.029565196484327316, | |
| "rewards/margins": 0.034249067306518555, | |
| "rewards/rejected": -0.06381426751613617, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.1773555027711797, | |
| "grad_norm": 24.375, | |
| "learning_rate": 1.0609573357858166e-05, | |
| "log_odds_chosen": 1.1033363342285156, | |
| "log_odds_ratio": -0.3850035071372986, | |
| "logps/chosen": -0.5864080786705017, | |
| "logps/rejected": -1.1940311193466187, | |
| "loss": 13.3906, | |
| "nll_loss": 0.8369148969650269, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -0.029320403933525085, | |
| "rewards/margins": 0.030381154268980026, | |
| "rewards/rejected": -0.05970155447721481, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.240696753760887, | |
| "grad_norm": 26.125, | |
| "learning_rate": 9.122105753945531e-06, | |
| "log_odds_chosen": 1.1297001838684082, | |
| "log_odds_ratio": -0.3876176178455353, | |
| "logps/chosen": -0.5929871797561646, | |
| "logps/rejected": -1.2141138315200806, | |
| "loss": 13.93, | |
| "nll_loss": 0.8706231117248535, | |
| "rewards/accuracies": 0.8218749761581421, | |
| "rewards/chosen": -0.02964935638010502, | |
| "rewards/margins": 0.03105633333325386, | |
| "rewards/rejected": -0.06070569157600403, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.304038004750594, | |
| "grad_norm": 23.0, | |
| "learning_rate": 7.723433775328384e-06, | |
| "log_odds_chosen": 1.1585485935211182, | |
| "log_odds_ratio": -0.38127031922340393, | |
| "logps/chosen": -0.5841962695121765, | |
| "logps/rejected": -1.200871467590332, | |
| "loss": 13.7727, | |
| "nll_loss": 0.8607965707778931, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -0.029209813103079796, | |
| "rewards/margins": 0.030833765864372253, | |
| "rewards/rejected": -0.0600435733795166, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.367379255740301, | |
| "grad_norm": 42.0, | |
| "learning_rate": 6.421379363065142e-06, | |
| "log_odds_chosen": 1.0362011194229126, | |
| "log_odds_ratio": -0.39413753151893616, | |
| "logps/chosen": -0.6054643988609314, | |
| "logps/rejected": -1.1436563730239868, | |
| "loss": 13.6624, | |
| "nll_loss": 0.8538990020751953, | |
| "rewards/accuracies": 0.8343750238418579, | |
| "rewards/chosen": -0.03027322329580784, | |
| "rewards/margins": 0.0269095990806818, | |
| "rewards/rejected": -0.05718282610177994, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.4307205067300077, | |
| "grad_norm": 25.5, | |
| "learning_rate": 5.223224133591476e-06, | |
| "log_odds_chosen": 1.0900369882583618, | |
| "log_odds_ratio": -0.4055331349372864, | |
| "logps/chosen": -0.5883392095565796, | |
| "logps/rejected": -1.1630512475967407, | |
| "loss": 13.5892, | |
| "nll_loss": 0.849326491355896, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -0.02941695973277092, | |
| "rewards/margins": 0.028735607862472534, | |
| "rewards/rejected": -0.058152567595243454, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.494061757719715, | |
| "grad_norm": 22.625, | |
| "learning_rate": 4.135668656967434e-06, | |
| "log_odds_chosen": 1.1677565574645996, | |
| "log_odds_ratio": -0.381297767162323, | |
| "logps/chosen": -0.594551682472229, | |
| "logps/rejected": -1.240464687347412, | |
| "loss": 13.8641, | |
| "nll_loss": 0.8665077090263367, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -0.02972758375108242, | |
| "rewards/margins": 0.032295651733875275, | |
| "rewards/rejected": -0.062023233622312546, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.557403008709422, | |
| "grad_norm": 24.875, | |
| "learning_rate": 3.164794984571759e-06, | |
| "log_odds_chosen": 1.1639173030853271, | |
| "log_odds_ratio": -0.39818206429481506, | |
| "logps/chosen": -0.5963728427886963, | |
| "logps/rejected": -1.2124865055084229, | |
| "loss": 14.1579, | |
| "nll_loss": 0.8848661184310913, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -0.029818642884492874, | |
| "rewards/margins": 0.030805686488747597, | |
| "rewards/rejected": -0.06062433123588562, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.620744259699129, | |
| "grad_norm": 21.0, | |
| "learning_rate": 2.316032635803378e-06, | |
| "log_odds_chosen": 1.259031057357788, | |
| "log_odds_ratio": -0.3424246907234192, | |
| "logps/chosen": -0.546891450881958, | |
| "logps/rejected": -1.2267786264419556, | |
| "loss": 13.1764, | |
| "nll_loss": 0.823525607585907, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": -0.02734457328915596, | |
| "rewards/margins": 0.03399435430765152, | |
| "rewards/rejected": -0.06133893132209778, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.684085510688836, | |
| "grad_norm": 25.625, | |
| "learning_rate": 1.59412823400657e-06, | |
| "log_odds_chosen": 1.2532789707183838, | |
| "log_odds_ratio": -0.3643453121185303, | |
| "logps/chosen": -0.5780880451202393, | |
| "logps/rejected": -1.268188714981079, | |
| "loss": 13.6232, | |
| "nll_loss": 0.8514481782913208, | |
| "rewards/accuracies": 0.8656250238418579, | |
| "rewards/chosen": -0.028904402628540993, | |
| "rewards/margins": 0.03450504317879677, | |
| "rewards/rejected": -0.06340944021940231, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.7474267616785433, | |
| "grad_norm": 27.625, | |
| "learning_rate": 1.0031189614277765e-06, | |
| "log_odds_chosen": 1.122249722480774, | |
| "log_odds_ratio": -0.4062274396419525, | |
| "logps/chosen": -0.5947844386100769, | |
| "logps/rejected": -1.1917502880096436, | |
| "loss": 13.6413, | |
| "nll_loss": 0.8525827527046204, | |
| "rewards/accuracies": 0.8656250238418579, | |
| "rewards/chosen": -0.029739221557974815, | |
| "rewards/margins": 0.029848288744688034, | |
| "rewards/rejected": -0.0595875158905983, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.81076801266825, | |
| "grad_norm": 27.125, | |
| "learning_rate": 5.463099816548579e-07, | |
| "log_odds_chosen": 1.1767040491104126, | |
| "log_odds_ratio": -0.3847493529319763, | |
| "logps/chosen": -0.5844542384147644, | |
| "logps/rejected": -1.234933853149414, | |
| "loss": 13.8583, | |
| "nll_loss": 0.8661431074142456, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -0.02922271564602852, | |
| "rewards/margins": 0.032523978501558304, | |
| "rewards/rejected": -0.06174669414758682, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.8741092636579575, | |
| "grad_norm": 26.0, | |
| "learning_rate": 2.262559558016325e-07, | |
| "log_odds_chosen": 1.1544322967529297, | |
| "log_odds_ratio": -0.37349843978881836, | |
| "logps/chosen": -0.5899974703788757, | |
| "logps/rejected": -1.2202789783477783, | |
| "loss": 13.6775, | |
| "nll_loss": 0.8548417091369629, | |
| "rewards/accuracies": 0.859375, | |
| "rewards/chosen": -0.029499869793653488, | |
| "rewards/margins": 0.03151407092809677, | |
| "rewards/rejected": -0.061013948172330856, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.937450514647664, | |
| "grad_norm": 20.375, | |
| "learning_rate": 4.474675580662113e-08, | |
| "log_odds_chosen": 1.106380820274353, | |
| "log_odds_ratio": -0.38155874609947205, | |
| "logps/chosen": -0.5873739719390869, | |
| "logps/rejected": -1.173477053642273, | |
| "loss": 13.7756, | |
| "nll_loss": 0.8609777688980103, | |
| "rewards/accuracies": 0.859375, | |
| "rewards/chosen": -0.029368694871664047, | |
| "rewards/margins": 0.02930515632033348, | |
| "rewards/rejected": -0.05867385119199753, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.9881235154394297, | |
| "step": 234, | |
| "total_flos": 0.0, | |
| "train_loss": 28.92138784359663, | |
| "train_runtime": 2488.1916, | |
| "train_samples_per_second": 6.089, | |
| "train_steps_per_second": 0.094 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 234, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |