{"canonical_length": 14, "canonical_nll": 1.7957624707903181, "generation_rouge_l_f1": 0.3157894736842105, "generation_token_f1": 0.3157894736842105, "prompt_id": "54ef1334aadf1dfd", "reading_canonical_log_prob": -25.140674591064453, "reading_distances": [37, 36, 33, 38], "reading_log_probs": [-27.395549774169922, -27.970338821411133, -26.900575637817383, -27.00428581237793], "reading_logprob_std": 0.48449946833697066, "tokenization_path_logprob_std": 22.434361871529376, "trajectory_balance_rmse": 241.52773078354207} {"canonical_length": 23, "canonical_nll": 0.7028683372165846, "generation_rouge_l_f1": 0.08396946564885495, "generation_token_f1": 0.09923664122137404, "prompt_id": "d109cb9b45bb2b84", "reading_canonical_log_prob": -16.165971755981445, "reading_distances": [12, 12, 9, 10], "reading_log_probs": [-17.867921829223633, -19.445167541503906, -17.063323974609375, -20.245758056640625], "reading_logprob_std": 1.450034410670165, "tokenization_path_logprob_std": 45.88133299469631, "trajectory_balance_rmse": 228.37852521725728} {"canonical_length": 79, "canonical_nll": 1.765072979504549, "generation_rouge_l_f1": 0.3431952662721893, "generation_token_f1": 0.485207100591716, "prompt_id": "f461f81672c6f974", "reading_canonical_log_prob": -139.44076538085938, "reading_distances": [24, 27, 34, 27], "reading_log_probs": [-133.92947387695312, -134.45257568359375, -132.2335968017578, -131.48997497558594], "reading_logprob_std": 1.395069684565878, "tokenization_path_logprob_std": 133.1525647888817, "trajectory_balance_rmse": 1903.0145508293454} {"canonical_length": 22, "canonical_nll": 0.7461254813454368, "generation_rouge_l_f1": 0.05185185185185185, "generation_token_f1": 0.05925925925925926, "prompt_id": "db2b92b024f2b9fd", "reading_canonical_log_prob": -16.41476058959961, "reading_distances": [20, 13, 17, 14], "reading_log_probs": [-17.054819107055664, -16.764305114746094, -15.990623474121094, -16.163978576660156], "reading_logprob_std": 0.4999612318659295, "tokenization_path_logprob_std": 52.658587915191546, "trajectory_balance_rmse": 191.08608450719188} {"canonical_length": 22, "canonical_nll": 1.7958963567560369, "generation_rouge_l_f1": 0.4666666666666667, "generation_token_f1": 0.4666666666666667, "prompt_id": "a3ca3e528a4c44fb", "reading_canonical_log_prob": -39.50971984863281, "reading_distances": [27, 30, 28, 27], "reading_log_probs": [-45.53224182128906, -46.677490234375, -45.36222839355469, -45.15628433227539], "reading_logprob_std": 0.6811899263854544, "tokenization_path_logprob_std": 107.00036406279854, "trajectory_balance_rmse": 664.208740243141} {"canonical_length": 46, "canonical_nll": 1.1380810944930366, "generation_rouge_l_f1": 0.30303030303030304, "generation_token_f1": 0.36363636363636365, "prompt_id": "bea2699cff758312", "reading_canonical_log_prob": -52.35173034667969, "reading_distances": [21, 15, 20, 25], "reading_log_probs": [-50.47179412841797, -50.865089416503906, -50.701927185058594, -50.05299377441406], "reading_logprob_std": 0.3524050653515814, "tokenization_path_logprob_std": 77.14120599508306, "trajectory_balance_rmse": 1264.846983617254} {"canonical_length": 41, "canonical_nll": 1.34303972197742, "generation_rouge_l_f1": 0.19811320754716982, "generation_token_f1": 0.19811320754716982, "prompt_id": "cf9306b8e22601b5", "reading_canonical_log_prob": -55.06462860107422, "reading_distances": [16, 21, 21, 18], "reading_log_probs": [-54.334259033203125, -53.85585021972656, -53.587562561035156, -53.98503494262695], "reading_logprob_std": 0.31024712386044223, "tokenization_path_logprob_std": 54.21394837015288, "trajectory_balance_rmse": 377.50445357182923} {"canonical_length": 231, "canonical_nll": 1.1898460223045184, "generation_rouge_l_f1": 0.22587268993839837, "generation_token_f1": 0.4928131416837782, "prompt_id": "ff6c469935e35f5e", "reading_canonical_log_prob": -274.85443115234375, "reading_distances": [17, 21, 23, 17], "reading_log_probs": [-267.4231262207031, -268.06463623046875, -266.8033142089844, -264.3762512207031], "reading_logprob_std": 1.6115443229853377, "tokenization_path_logprob_std": 222.85973207816508, "trajectory_balance_rmse": 4997.198875133044} {"canonical_length": 203, "canonical_nll": 1.4607047846751848, "generation_rouge_l_f1": 0.31382978723404253, "generation_token_f1": 0.5904255319148936, "prompt_id": "41c1bf9aacb45501", "reading_canonical_log_prob": -296.5230712890625, "reading_distances": [23, 22, 21, 22], "reading_log_probs": [-298.8663330078125, -299.314697265625, -298.79364013671875, -300.1805419921875], "reading_logprob_std": 0.6375829615696005, "tokenization_path_logprob_std": 140.25270065669875, "trajectory_balance_rmse": 4860.009041394161} {"canonical_length": 33, "canonical_nll": 1.3100093494762073, "generation_rouge_l_f1": 0.21212121212121213, "generation_token_f1": 0.24242424242424243, "prompt_id": "aff22e4e1c81f916", "reading_canonical_log_prob": -43.230308532714844, "reading_distances": [25, 29, 34, 25], "reading_log_probs": [-44.09911346435547, -42.4903564453125, -43.637046813964844, -43.565040588378906], "reading_logprob_std": 0.6808002814102107, "tokenization_path_logprob_std": 155.71491477107244, "trajectory_balance_rmse": 850.4358725863245} {"canonical_length": 10, "canonical_nll": 1.0396286010742188, "generation_rouge_l_f1": 0.14285714285714288, "generation_token_f1": 0.14285714285714288, "prompt_id": "f599bb889055c49d", "reading_canonical_log_prob": -10.396286010742188, "reading_distances": [63, 53, 60, 54], "reading_log_probs": [-12.209800720214844, -11.048086166381836, -13.041303634643555, -11.127355575561523], "reading_logprob_std": 0.9510985744720698, "tokenization_path_logprob_std": 45.28929129433415, "trajectory_balance_rmse": 191.75396872926828} {"canonical_length": 90, "canonical_nll": 1.2788369920518663, "generation_rouge_l_f1": 0.21369863013698628, "generation_token_f1": 0.31780821917808216, "prompt_id": "7fbc599a6c4c6b54", "reading_canonical_log_prob": -115.09532928466797, "reading_distances": [21, 34, 28, 33], "reading_log_probs": [-115.59009552001953, -122.91690063476562, -119.85042572021484, -123.67354583740234], "reading_logprob_std": 3.671587797221253, "tokenization_path_logprob_std": 125.63156452076642, "trajectory_balance_rmse": 1940.7353349606622} {"canonical_length": 294, "canonical_nll": 1.6796504429408483, "generation_rouge_l_f1": 0.19285714285714287, "generation_token_f1": 0.36428571428571427, "prompt_id": "9e624d36638d2bbd", "reading_canonical_log_prob": -493.8172302246094, "reading_distances": [20, 13, 22, 16], "reading_log_probs": [-482.8519287109375, -480.8216247558594, -481.590087890625, -485.073974609375], "reading_logprob_std": 1.858815615650786, "tokenization_path_logprob_std": 279.96852544046084, "trajectory_balance_rmse": 6030.269514927361} {"canonical_length": 27, "canonical_nll": 0.937792389481156, "generation_rouge_l_f1": 0.0859375, "generation_token_f1": 0.09375, "prompt_id": "3982276a331d29f5", "reading_canonical_log_prob": -25.32039451599121, "reading_distances": [26, 23, 24, 25], "reading_log_probs": [-23.38796043395996, -24.75791358947754, -26.0234375, -23.50422477722168], "reading_logprob_std": 1.2367884338121249, "tokenization_path_logprob_std": 37.655425799035385, "trajectory_balance_rmse": 340.61991000986103} {"canonical_length": 10, "canonical_nll": 1.4250356674194335, "generation_rouge_l_f1": 0.4615384615384615, "generation_token_f1": 0.6153846153846154, "prompt_id": "3b79dbc85baf9416", "reading_canonical_log_prob": -14.250356674194336, "reading_distances": [21, 25, 21, 19], "reading_log_probs": [-14.593990325927734, -13.44955825805664, -12.866703987121582, -12.792791366577148], "reading_logprob_std": 0.8323703838902048, "tokenization_path_logprob_std": 21.823851842700897, "trajectory_balance_rmse": 200.77342373848006} {"canonical_length": 174, "canonical_nll": 1.40594447344199, "generation_rouge_l_f1": 0.21967963386727687, "generation_token_f1": 0.4302059496567505, "prompt_id": "bc017f1a688cb1fb", "reading_canonical_log_prob": -244.63433837890625, "reading_distances": [23, 20, 25, 28], "reading_log_probs": [-242.05062866210938, -242.57948303222656, -241.63385009765625, -241.38455200195312], "reading_logprob_std": 0.5229149698533025, "tokenization_path_logprob_std": 229.53904714872425, "trajectory_balance_rmse": 4209.353421401069} {"canonical_length": 64, "canonical_nll": 0.9261763095855713, "generation_rouge_l_f1": 0.2138364779874214, "generation_token_f1": 0.31446540880503143, "prompt_id": "2060bf62bca6703a", "reading_canonical_log_prob": -59.27528381347656, "reading_distances": [17, 14, 23, 18], "reading_log_probs": [-58.740997314453125, -57.46513748168945, -57.99104690551758, -57.478965759277344], "reading_logprob_std": 0.6001365552000596, "tokenization_path_logprob_std": 76.17472057426647, "trajectory_balance_rmse": 1256.4096093356463} {"canonical_length": 12, "canonical_nll": 0.8607707023620605, "generation_rouge_l_f1": 0.06666666666666667, "generation_token_f1": 0.06666666666666667, "prompt_id": "8300e5ea4c06c29e", "reading_canonical_log_prob": -10.329248428344727, "reading_distances": [27, 29, 30, 27], "reading_log_probs": [-10.631999969482422, -10.08415412902832, -11.14344310760498, -9.93612289428711], "reading_logprob_std": 0.5513304012641297, "tokenization_path_logprob_std": 14.978004150048816, "trajectory_balance_rmse": 136.48631704238232} {"canonical_length": 142, "canonical_nll": 1.3117555430237675, "generation_rouge_l_f1": 0.22596153846153846, "generation_token_f1": 0.48557692307692313, "prompt_id": "5c25240178b5a183", "reading_canonical_log_prob": -186.269287109375, "reading_distances": [20, 19, 20, 21], "reading_log_probs": [-187.70506286621094, -190.72068786621094, -190.61532592773438, -191.8138427734375], "reading_logprob_std": 1.7580346264277786, "tokenization_path_logprob_std": 83.25830698604739, "trajectory_balance_rmse": 2863.0370163590983} {"canonical_length": 14, "canonical_nll": 1.5132596152169364, "generation_rouge_l_f1": 0.4444444444444444, "generation_token_f1": 0.4444444444444444, "prompt_id": "571e075c79645e81", "reading_canonical_log_prob": -21.18563461303711, "reading_distances": [28, 15, 24, 19], "reading_log_probs": [-23.560871124267578, -21.255563735961914, -23.31977081298828, -23.383926391601562], "reading_logprob_std": 1.087768179275357, "tokenization_path_logprob_std": 13.503793167026622, "trajectory_balance_rmse": 326.0058977414263} {"canonical_length": 153, "canonical_nll": 1.4784915899139603, "generation_rouge_l_f1": 0.18784530386740333, "generation_token_f1": 0.4806629834254144, "prompt_id": "fc69046f95700385", "reading_canonical_log_prob": -226.20921325683594, "reading_distances": [39, 41, 39, 38], "reading_log_probs": [-231.6458740234375, -226.41082763671875, -233.51475524902344, -229.40786743164062], "reading_logprob_std": 3.058072157001703, "tokenization_path_logprob_std": 201.55148164962475, "trajectory_balance_rmse": 3482.3552015564305} {"canonical_length": 56, "canonical_nll": 1.7841008050101144, "generation_rouge_l_f1": 0.16393442622950824, "generation_token_f1": 0.22950819672131142, "prompt_id": "fcbb32fd544b170a", "reading_canonical_log_prob": -99.9096450805664, "reading_distances": [24, 22, 24, 25], "reading_log_probs": [-101.06997680664062, -99.90385437011719, -100.38185119628906, -98.92610168457031], "reading_logprob_std": 0.9006130713950247, "tokenization_path_logprob_std": 65.0927418311235, "trajectory_balance_rmse": 1165.320544344452} {"canonical_length": 83, "canonical_nll": 0.9910941985716303, "generation_rouge_l_f1": 0.2681564245810056, "generation_token_f1": 0.5027932960893854, "prompt_id": "f6d789446e38536d", "reading_canonical_log_prob": -82.26081848144531, "reading_distances": [8, 7, 8, 9], "reading_log_probs": [-80.7197036743164, -78.26274108886719, -80.65573120117188, -78.40763092041016], "reading_logprob_std": 1.3597728871704593, "tokenization_path_logprob_std": 145.9240446600778, "trajectory_balance_rmse": 1695.2870585090566} {"canonical_length": 270, "canonical_nll": 1.3740394874855324, "generation_rouge_l_f1": 0.15544041450777205, "generation_token_f1": 0.2849740932642487, "prompt_id": "86891b80d6ccb140", "reading_canonical_log_prob": -370.99066162109375, "reading_distances": [20, 24, 27, 23], "reading_log_probs": [-368.32818603515625, -369.86224365234375, -368.0935363769531, -373.20806884765625], "reading_logprob_std": 2.357664161060035, "tokenization_path_logprob_std": 140.25070150191996, "trajectory_balance_rmse": 3424.9782029432586} {"canonical_length": 27, "canonical_nll": 1.4737110844364874, "generation_rouge_l_f1": 0.3673469387755102, "generation_token_f1": 0.5306122448979592, "prompt_id": "01a2b8a844748c41", "reading_canonical_log_prob": -39.790199279785156, "reading_distances": [33, 37, 33, 36], "reading_log_probs": [-39.739898681640625, -37.78455352783203, -37.914310455322266, -39.59626007080078], "reading_logprob_std": 1.052966090788096, "tokenization_path_logprob_std": 21.261394369156534, "trajectory_balance_rmse": 683.592388199804} {"canonical_length": 66, "canonical_nll": 1.1143560698538115, "generation_rouge_l_f1": 0.14330218068535827, "generation_token_f1": 0.21806853582554517, "prompt_id": "fc8d39645d6796a9", "reading_canonical_log_prob": -73.54750061035156, "reading_distances": [28, 32, 30, 26], "reading_log_probs": [-73.9703140258789, -73.91653442382812, -73.97151184082031, -71.90560913085938], "reading_logprob_std": 1.0239098691282809, "tokenization_path_logprob_std": 90.46727648201023, "trajectory_balance_rmse": 1477.0301054342694} {"canonical_length": 340, "canonical_nll": 1.29162696389591, "generation_rouge_l_f1": 0.19589257503949448, "generation_token_f1": 0.3696682464454976, "prompt_id": "7a3c4e3465a24fae", "reading_canonical_log_prob": -439.1531677246094, "reading_distances": [28, 28, 21, 20], "reading_log_probs": [-436.05438232421875, -434.68084716796875, -436.67071533203125, -437.52447509765625], "reading_logprob_std": 1.1973003944546599, "tokenization_path_logprob_std": 203.38608523315526, "trajectory_balance_rmse": 7229.79344816856} {"canonical_length": 22, "canonical_nll": 0.9568131186745383, "generation_rouge_l_f1": 0.8333333333333334, "generation_token_f1": 0.8333333333333334, "prompt_id": "9652191d0c5b1125", "reading_canonical_log_prob": -21.049888610839844, "reading_distances": [20, 17, 15, 23], "reading_log_probs": [-22.16461944580078, -23.317127227783203, -22.235198974609375, -23.207141876220703], "reading_logprob_std": 0.615592343594924, "tokenization_path_logprob_std": 36.988854739019146, "trajectory_balance_rmse": 427.90587346825157} {"canonical_length": 229, "canonical_nll": 0.9750522130441457, "generation_rouge_l_f1": 0.33191489361702126, "generation_token_f1": 0.5446808510638298, "prompt_id": "1fb3dd895fb36ca1", "reading_canonical_log_prob": -223.28695678710938, "reading_distances": [46, 47, 47, 53], "reading_log_probs": [-223.36138916015625, -219.49484252929688, -219.855224609375, -220.23672485351562], "reading_logprob_std": 1.7755916102778255, "tokenization_path_logprob_std": 218.83004221513846, "trajectory_balance_rmse": 4686.672205246182} {"canonical_length": 84, "canonical_nll": 0.19835578827630906, "generation_rouge_l_f1": 0.42500000000000004, "generation_token_f1": 0.6375, "prompt_id": "fba6cc8156910498", "reading_canonical_log_prob": -16.66188621520996, "reading_distances": [22, 14, 24, 27], "reading_log_probs": [-18.5472412109375, -17.945585250854492, -18.247055053710938, -18.60556983947754], "reading_logprob_std": 0.30420667358841297, "tokenization_path_logprob_std": 179.909744886188, "trajectory_balance_rmse": 1433.6562480743316} {"canonical_length": 24, "canonical_nll": 0.525162140528361, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "02b01e693f3f83a0", "reading_canonical_log_prob": -12.603891372680664, "reading_distances": [34, 32, 25, 24], "reading_log_probs": [-10.562461853027344, -10.704448699951172, -11.127869606018066, -10.37360668182373], "reading_logprob_std": 0.3205680805225667, "tokenization_path_logprob_std": 48.007183120620624, "trajectory_balance_rmse": 343.7760389720533} {"canonical_length": 331, "canonical_nll": 0.8871736555301171, "generation_rouge_l_f1": 0.2433460076045627, "generation_token_f1": 0.44106463878326996, "prompt_id": "40bc9c44485ded59", "reading_canonical_log_prob": -293.65447998046875, "reading_distances": [27, 34, 28, 33], "reading_log_probs": [-298.4508056640625, -298.9949951171875, -298.4354248046875, -299.11468505859375], "reading_logprob_std": 0.3565989450644264, "tokenization_path_logprob_std": 212.53851652850562, "trajectory_balance_rmse": 4637.112572895996} {"canonical_length": 11, "canonical_nll": 1.548527110706676, "generation_rouge_l_f1": 0.3, "generation_token_f1": 0.3, "prompt_id": "f0c9ccf7c1a3a5c6", "reading_canonical_log_prob": -17.033798217773438, "reading_distances": [22, 26, 27, 28], "reading_log_probs": [-15.821548461914062, -17.224573135375977, -17.57143211364746, -17.087575912475586], "reading_logprob_std": 0.7641221950613478, "tokenization_path_logprob_std": 36.523272166722585, "trajectory_balance_rmse": 248.83141314539736} {"canonical_length": 149, "canonical_nll": 1.8418419345113255, "generation_rouge_l_f1": 0.2961165048543689, "generation_token_f1": 0.4951456310679611, "prompt_id": "7b4b39937f27934d", "reading_canonical_log_prob": -274.4344482421875, "reading_distances": [27, 27, 23, 30], "reading_log_probs": [-269.57598876953125, -267.4193115234375, -268.0469665527344, -269.330322265625], "reading_logprob_std": 1.03047046265226, "tokenization_path_logprob_std": 115.44605115151363, "trajectory_balance_rmse": 3268.1216851410463} {"canonical_length": 10, "canonical_nll": 0.7008621215820312, "generation_rouge_l_f1": 0.2318840579710145, "generation_token_f1": 0.2318840579710145, "prompt_id": "177155c261ac8893", "reading_canonical_log_prob": -7.0086212158203125, "reading_distances": [18, 15, 11, 15], "reading_log_probs": [-6.522451877593994, -6.365312576293945, -6.516655445098877, -6.641888618469238], "reading_logprob_std": 0.11331144854982121, "tokenization_path_logprob_std": 34.199370820303805, "trajectory_balance_rmse": 206.41821494907234} {"canonical_length": 9, "canonical_nll": 0.257839388317532, "generation_rouge_l_f1": 0.7058823529411765, "generation_token_f1": 0.823529411764706, "prompt_id": "316f7da66f20e71c", "reading_canonical_log_prob": -2.320554494857788, "reading_distances": [31, 31, 30, 26], "reading_log_probs": [-2.1214687824249268, -2.099520206451416, -2.0982930660247803, -2.048323154449463], "reading_logprob_std": 0.03094184434819129, "tokenization_path_logprob_std": 35.910490818813344, "trajectory_balance_rmse": 208.7133983457232} {"canonical_length": 10, "canonical_nll": 1.5077809333801269, "generation_rouge_l_f1": 0.04669260700389105, "generation_token_f1": 0.054474708171206226, "prompt_id": "4c771fd22dee11bd", "reading_canonical_log_prob": -15.07780933380127, "reading_distances": [25, 31, 29, 30], "reading_log_probs": [-15.181446075439453, -14.875438690185547, -15.007551193237305, -16.479494094848633], "reading_logprob_std": 0.7396999375238561, "tokenization_path_logprob_std": 35.27983613697859, "trajectory_balance_rmse": 288.39006134203595} {"canonical_length": 150, "canonical_nll": 1.093818155924479, "generation_rouge_l_f1": 0.3138173302107728, "generation_token_f1": 0.47306791569086654, "prompt_id": "fb7e151d45f5c282", "reading_canonical_log_prob": -164.07272338867188, "reading_distances": [37, 33, 38, 39], "reading_log_probs": [-163.52581787109375, -165.6668243408203, -163.05316162109375, -165.00608825683594], "reading_logprob_std": 1.2274716724069827, "tokenization_path_logprob_std": 133.59516463519734, "trajectory_balance_rmse": 2837.089129049847} {"canonical_length": 186, "canonical_nll": 1.3087493732411375, "generation_rouge_l_f1": 0.2388758782201405, "generation_token_f1": 0.4402810304449649, "prompt_id": "c6ce7ec89de605a5", "reading_canonical_log_prob": -243.42738342285156, "reading_distances": [17, 19, 20, 19], "reading_log_probs": [-240.34571838378906, -240.00729370117188, -239.5186004638672, -239.3808135986328], "reading_logprob_std": 0.4453444531084589, "tokenization_path_logprob_std": 191.4450928191304, "trajectory_balance_rmse": 3996.498829565132} {"canonical_length": 220, "canonical_nll": 0.9360420920632102, "generation_rouge_l_f1": 0.3111111111111111, "generation_token_f1": 0.5688888888888889, "prompt_id": "320fa6e957218c02", "reading_canonical_log_prob": -205.92926025390625, "reading_distances": [14, 13, 14, 13], "reading_log_probs": [-207.85244750976562, -203.8854217529297, -205.02647399902344, -208.76669311523438], "reading_logprob_std": 2.303571756424798, "tokenization_path_logprob_std": 202.51435936419807, "trajectory_balance_rmse": 3966.5757698119014} {"canonical_length": 58, "canonical_nll": 1.231512398555361, "generation_rouge_l_f1": 0.0909090909090909, "generation_token_f1": 0.0909090909090909, "prompt_id": "d3de64d71666f0d5", "reading_canonical_log_prob": -71.42771911621094, "reading_distances": [22, 20, 23, 23], "reading_log_probs": [-67.35636138916016, -69.43415832519531, -68.6319580078125, -66.8252944946289], "reading_logprob_std": 1.1881409769361349, "tokenization_path_logprob_std": 40.43546368376671, "trajectory_balance_rmse": 912.762218709838} {"canonical_length": 11, "canonical_nll": 0.09240181879563765, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "bd4efb1cb33c7494", "reading_canonical_log_prob": -1.0164200067520142, "reading_distances": [44, 47, 47, 43], "reading_log_probs": [-1.659365177154541, -1.3994286060333252, -1.44678795337677, -1.2208507061004639], "reading_logprob_std": 0.1803310964659126, "tokenization_path_logprob_std": 43.464682824320285, "trajectory_balance_rmse": 188.91560684576865} {"canonical_length": 38, "canonical_nll": 0.7006138751381322, "generation_rouge_l_f1": 0.5333333333333333, "generation_token_f1": 0.5666666666666667, "prompt_id": "d19ca6fde647262d", "reading_canonical_log_prob": -26.623327255249023, "reading_distances": [17, 15, 15, 17], "reading_log_probs": [-26.394237518310547, -24.47658348083496, -26.018278121948242, -24.86062240600586], "reading_logprob_std": 0.9144758429006724, "tokenization_path_logprob_std": 53.09359299333357, "trajectory_balance_rmse": 920.2193010505531} {"canonical_length": 88, "canonical_nll": 1.244468168778853, "generation_rouge_l_f1": 0.28571428571428575, "generation_token_f1": 0.5028571428571429, "prompt_id": "b0e33c086e3feda9", "reading_canonical_log_prob": -109.51319885253906, "reading_distances": [14, 17, 19, 20], "reading_log_probs": [-115.42072296142578, -120.48368072509766, -115.9695816040039, -115.0876235961914], "reading_logprob_std": 2.5218729094068406, "tokenization_path_logprob_std": 95.93661946424321, "trajectory_balance_rmse": 1816.3290453731179} {"canonical_length": 53, "canonical_nll": 0.9120993704166053, "generation_rouge_l_f1": 0.49438202247191015, "generation_token_f1": 0.8089887640449438, "prompt_id": "85224f0bc07faa4c", "reading_canonical_log_prob": -48.34126663208008, "reading_distances": [18, 20, 17, 19], "reading_log_probs": [-53.51824951171875, -53.18374252319336, -52.46519088745117, -53.75180435180664], "reading_logprob_std": 0.5604878942984369, "tokenization_path_logprob_std": 99.42926680374171, "trajectory_balance_rmse": 1030.5828993266018} {"canonical_length": 26, "canonical_nll": 1.3719717172475963, "generation_rouge_l_f1": 0.27272727272727276, "generation_token_f1": 0.3636363636363636, "prompt_id": "7327a63351d4e2c9", "reading_canonical_log_prob": -35.6712646484375, "reading_distances": [40, 29, 31, 36], "reading_log_probs": [-43.0477294921875, -43.19906234741211, -41.44890594482422, -46.99321746826172], "reading_logprob_std": 2.3513128340271687, "tokenization_path_logprob_std": 107.41511989389879, "trajectory_balance_rmse": 649.8105000847681} {"canonical_length": 84, "canonical_nll": 1.3316694895426433, "generation_rouge_l_f1": 0.30088495575221236, "generation_token_f1": 0.4070796460176992, "prompt_id": "7fabfcdc4a3bff58", "reading_canonical_log_prob": -111.86023712158203, "reading_distances": [29, 25, 32, 33], "reading_log_probs": [-112.69515991210938, -111.9666748046875, -112.42082214355469, -114.1771469116211], "reading_logprob_std": 0.9565280385142872, "tokenization_path_logprob_std": 114.13046298052996, "trajectory_balance_rmse": 1765.6089951250121} {"canonical_length": 13, "canonical_nll": 0.4371213912963867, "generation_rouge_l_f1": 0.6250000000000001, "generation_token_f1": 0.6250000000000001, "prompt_id": "173c8d85c343f33f", "reading_canonical_log_prob": -5.682578086853027, "reading_distances": [24, 32, 28, 32], "reading_log_probs": [-6.756701946258545, -6.314267158508301, -6.383636474609375, -6.056057929992676], "reading_logprob_std": 0.28934123966891706, "tokenization_path_logprob_std": 56.70851235584728, "trajectory_balance_rmse": 202.93181351812143} {"canonical_length": 118, "canonical_nll": 1.2424263388423595, "generation_rouge_l_f1": 0.3939393939393939, "generation_token_f1": 0.5353535353535354, "prompt_id": "e1f635b187b10d1b", "reading_canonical_log_prob": -146.60630798339844, "reading_distances": [29, 26, 23, 20], "reading_log_probs": [-143.16915893554688, -142.1563262939453, -142.65957641601562, -144.67213439941406], "reading_logprob_std": 1.0869445054271414, "tokenization_path_logprob_std": 108.32449843273007, "trajectory_balance_rmse": 2270.6830918242504} {"canonical_length": 242, "canonical_nll": 0.7819334928654442, "generation_rouge_l_f1": 0.38396624472573837, "generation_token_f1": 0.6033755274261603, "prompt_id": "5ff715a6459080e1", "reading_canonical_log_prob": -189.2279052734375, "reading_distances": [20, 30, 28, 26], "reading_log_probs": [-179.35308837890625, -183.17526245117188, -182.75241088867188, -183.3269500732422], "reading_logprob_std": 1.8816635543895306, "tokenization_path_logprob_std": 136.9391277351569, "trajectory_balance_rmse": 2933.927971645649} {"canonical_length": 147, "canonical_nll": 0.7359950266727785, "generation_rouge_l_f1": 0.39024390243902435, "generation_token_f1": 0.4738675958188153, "prompt_id": "3e82cf9b74d0c14b", "reading_canonical_log_prob": -108.19126892089844, "reading_distances": [24, 26, 25, 23], "reading_log_probs": [-113.65472412109375, -112.2670669555664, -112.71884155273438, -113.58111572265625], "reading_logprob_std": 0.675846874926188, "tokenization_path_logprob_std": 135.57836494876656, "trajectory_balance_rmse": 1595.0634056519968} {"canonical_length": 42, "canonical_nll": 1.3925120944068545, "generation_rouge_l_f1": 0.26436781609195403, "generation_token_f1": 0.3563218390804598, "prompt_id": "b1354ebe20eb5cb8", "reading_canonical_log_prob": -58.48550796508789, "reading_distances": [46, 54, 47, 50], "reading_log_probs": [-58.863800048828125, -59.52105712890625, -59.75703811645508, -59.28961944580078], "reading_logprob_std": 0.3806697951123881, "tokenization_path_logprob_std": 43.83547934397724, "trajectory_balance_rmse": 869.3262580660416} {"canonical_length": 124, "canonical_nll": 1.1451034545898438, "generation_rouge_l_f1": 0.1850899742930591, "generation_token_f1": 0.4061696658097686, "prompt_id": "751bafc67843f928", "reading_canonical_log_prob": -141.99282836914062, "reading_distances": [31, 31, 34, 36], "reading_log_probs": [-140.54273986816406, -140.62771606445312, -140.6419219970703, -140.69277954101562], "reading_logprob_std": 0.06231071760380233, "tokenization_path_logprob_std": 66.6055562325628, "trajectory_balance_rmse": 2520.3098561438173} {"canonical_length": 283, "canonical_nll": 1.1705982221731448, "generation_rouge_l_f1": 0.2813688212927757, "generation_token_f1": 0.5247148288973384, "prompt_id": "2c0f185491268c7a", "reading_canonical_log_prob": -331.279296875, "reading_distances": [23, 30, 38, 21], "reading_log_probs": [-328.66778564453125, -327.52593994140625, -326.7391357421875, -324.75860595703125], "reading_logprob_std": 1.6458251310410206, "tokenization_path_logprob_std": 140.51067575798984, "trajectory_balance_rmse": 4375.108719229346} {"canonical_length": 384, "canonical_nll": 1.4291140238444011, "generation_rouge_l_f1": 0.23011844331641287, "generation_token_f1": 0.4060913705583757, "prompt_id": "d93394062daf791b", "reading_canonical_log_prob": -548.77978515625, "reading_distances": [25, 28, 29, 25], "reading_log_probs": [-547.7852172851562, -536.4844970703125, -535.9105834960938, -539.1705322265625], "reading_logprob_std": 5.485567663073179, "tokenization_path_logprob_std": 158.7109256209528, "trajectory_balance_rmse": 6820.064319935724} {"canonical_length": 13, "canonical_nll": 2.224029247577374, "generation_rouge_l_f1": 0.16, "generation_token_f1": 0.23999999999999996, "prompt_id": "1bc766d97da83b42", "reading_canonical_log_prob": -28.91238021850586, "reading_distances": [20, 20, 19, 19], "reading_log_probs": [-27.074966430664062, -27.680404663085938, -27.955238342285156, -26.760644912719727], "reading_logprob_std": 0.5468681376626057, "tokenization_path_logprob_std": 37.530750070672646, "trajectory_balance_rmse": 309.6135981336395} {"canonical_length": 5, "canonical_nll": 0.6718472480773926, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.6666666666666666, "prompt_id": "03cbf10fd71003f1", "reading_canonical_log_prob": -3.359236240386963, "reading_distances": [27, 20, 23, 23], "reading_log_probs": [-4.6777544021606445, -5.253540992736816, -5.606389045715332, -5.133317947387695], "reading_logprob_std": 0.3834255942632781, "tokenization_path_logprob_std": 20.459999230021214, "trajectory_balance_rmse": 98.04568560010235} {"canonical_length": 218, "canonical_nll": 1.2465891706834145, "generation_rouge_l_f1": 0.23744292237442924, "generation_token_f1": 0.43378995433789946, "prompt_id": "3b33c4e6381082ec", "reading_canonical_log_prob": -271.7564392089844, "reading_distances": [23, 20, 19, 25], "reading_log_probs": [-274.34844970703125, -276.17510986328125, -272.0177307128906, -272.4735107421875], "reading_logprob_std": 1.9034553529143403, "tokenization_path_logprob_std": 128.0885416722249, "trajectory_balance_rmse": 3461.798870813108} {"canonical_length": 14, "canonical_nll": 1.07881532396589, "generation_rouge_l_f1": 0.64, "generation_token_f1": 0.64, "prompt_id": "275fffbea930b52f", "reading_canonical_log_prob": -15.103414535522461, "reading_distances": [12, 17, 19, 17], "reading_log_probs": [-18.732810974121094, -21.31572723388672, -25.524442672729492, -17.799327850341797], "reading_logprob_std": 3.45717680540191, "tokenization_path_logprob_std": 21.478554371131416, "trajectory_balance_rmse": 278.49247560600213} {"canonical_length": 181, "canonical_nll": 0.9366618625366885, "generation_rouge_l_f1": 0.2296650717703349, "generation_token_f1": 0.4880382775119617, "prompt_id": "a7f0f685867c31e2", "reading_canonical_log_prob": -169.53579711914062, "reading_distances": [81, 74, 72, 75], "reading_log_probs": [-170.00460815429688, -172.20333862304688, -173.40371704101562, -169.19729614257812], "reading_logprob_std": 1.9410325022558863, "tokenization_path_logprob_std": 89.48596057768768, "trajectory_balance_rmse": 2840.3249650754524} {"canonical_length": 128, "canonical_nll": 1.204847812652588, "generation_rouge_l_f1": 0.2631578947368421, "generation_token_f1": 0.4368421052631579, "prompt_id": "858a5416234d0d2a", "reading_canonical_log_prob": -154.22052001953125, "reading_distances": [19, 21, 21, 19], "reading_log_probs": [-157.47935485839844, -155.57717895507812, -153.30841064453125, -156.02163696289062], "reading_logprob_std": 1.7283527776358016, "tokenization_path_logprob_std": 100.95192852341033, "trajectory_balance_rmse": 2645.9998003711025} {"canonical_length": 205, "canonical_nll": 0.8644543159298781, "generation_rouge_l_f1": 0.35989717223650386, "generation_token_f1": 0.5347043701799485, "prompt_id": "17197512e608a44f", "reading_canonical_log_prob": -177.213134765625, "reading_distances": [16, 14, 12, 13], "reading_log_probs": [-207.347412109375, -195.3504180908203, -197.281982421875, -203.52674865722656], "reading_logprob_std": 5.548412964684624, "tokenization_path_logprob_std": 144.93185771211915, "trajectory_balance_rmse": 3646.6694256839296} {"canonical_length": 160, "canonical_nll": 1.6356739044189452, "generation_rouge_l_f1": 0.15517241379310345, "generation_token_f1": 0.2931034482758621, "prompt_id": "7c90d4dc6f1c3493", "reading_canonical_log_prob": -261.70782470703125, "reading_distances": [22, 24, 15, 17], "reading_log_probs": [-263.2652282714844, -265.34783935546875, -261.9654235839844, -262.3727722167969], "reading_logprob_std": 1.507783494675093, "tokenization_path_logprob_std": 75.28200507069047, "trajectory_balance_rmse": 3120.6517734087497} {"canonical_length": 23, "canonical_nll": 3.2361672442892324, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "d1e684b980513357", "reading_canonical_log_prob": -74.43184661865234, "reading_distances": [9, 19, 23, 13], "reading_log_probs": [-75.12666320800781, -75.0389404296875, -75.38800048828125, -74.22470092773438], "reading_logprob_std": 0.502296656497923, "tokenization_path_logprob_std": 54.86957343573098, "trajectory_balance_rmse": 381.88741824473135} {"canonical_length": 392, "canonical_nll": 1.0540573742924904, "generation_rouge_l_f1": 0.28396946564885495, "generation_token_f1": 0.5374045801526719, "prompt_id": "5b2aa008c5db43fd", "reading_canonical_log_prob": -413.19049072265625, "reading_distances": [7, 12, 10, 9], "reading_log_probs": [-407.9281005859375, -402.5000305175781, -401.62506103515625, -400.97869873046875], "reading_logprob_std": 3.175219738132099, "tokenization_path_logprob_std": 99.01127477735888, "trajectory_balance_rmse": 7014.749153467996} {"canonical_length": 82, "canonical_nll": 1.0695235089557926, "generation_rouge_l_f1": 0.3076923076923077, "generation_token_f1": 0.3812709030100334, "prompt_id": "180245122ca6704c", "reading_canonical_log_prob": -87.700927734375, "reading_distances": [29, 24, 35, 31], "reading_log_probs": [-92.43093872070312, -90.84366607666016, -88.7103271484375, -87.0949935913086], "reading_logprob_std": 2.3460548130141454, "tokenization_path_logprob_std": 55.132121520980334, "trajectory_balance_rmse": 1165.079247193753} {"canonical_length": 99, "canonical_nll": 2.020624796549479, "generation_rouge_l_f1": 0.24369747899159663, "generation_token_f1": 0.4117647058823529, "prompt_id": "0280ead696ae0c9e", "reading_canonical_log_prob": -200.04185485839844, "reading_distances": [34, 24, 24, 32], "reading_log_probs": [-199.5795135498047, -199.24862670898438, -199.53778076171875, -199.59466552734375], "reading_logprob_std": 0.16280000922330395, "tokenization_path_logprob_std": 127.49380554309151, "trajectory_balance_rmse": 2539.943567646873} {"canonical_length": 148, "canonical_nll": 1.5487266746727195, "generation_rouge_l_f1": 0.1865284974093264, "generation_token_f1": 0.27979274611398963, "prompt_id": "1d5af645718b1db0", "reading_canonical_log_prob": -229.2115478515625, "reading_distances": [58, 46, 51, 52], "reading_log_probs": [-227.35678100585938, -218.41897583007812, -231.01602172851562, -227.287353515625], "reading_logprob_std": 5.3581392835700274, "tokenization_path_logprob_std": 101.29908899212185, "trajectory_balance_rmse": 2469.3919247394406} {"canonical_length": 120, "canonical_nll": 1.4236101786295572, "generation_rouge_l_f1": 0.27149321266968324, "generation_token_f1": 0.4343891402714932, "prompt_id": "0211ac7bc9c6651a", "reading_canonical_log_prob": -170.83322143554688, "reading_distances": [17, 21, 20, 24], "reading_log_probs": [-173.76637268066406, -173.92689514160156, -175.009033203125, -176.71832275390625], "reading_logprob_std": 1.3591892049063539, "tokenization_path_logprob_std": 133.1580424567312, "trajectory_balance_rmse": 2404.293493755267} {"canonical_length": 115, "canonical_nll": 0.6802046402640964, "generation_rouge_l_f1": 0.2709677419354839, "generation_token_f1": 0.40645161290322573, "prompt_id": "0da3d3229c8dfac7", "reading_canonical_log_prob": -78.2235336303711, "reading_distances": [14, 15, 19, 15], "reading_log_probs": [-76.73690795898438, -78.16303253173828, -78.54350280761719, -78.12112426757812], "reading_logprob_std": 0.7926006033733642, "tokenization_path_logprob_std": 123.01309242687653, "trajectory_balance_rmse": 1483.1969065132073} {"canonical_length": 196, "canonical_nll": 1.733198204819037, "generation_rouge_l_f1": 0.20750551876379691, "generation_token_f1": 0.48565121412803536, "prompt_id": "dd26708537c260a2", "reading_canonical_log_prob": -339.70684814453125, "reading_distances": [35, 38, 37, 36], "reading_log_probs": [-335.4466247558594, -337.1091003417969, -332.3307189941406, -334.34503173828125], "reading_logprob_std": 2.0045086751684007, "tokenization_path_logprob_std": 217.3225483482035, "trajectory_balance_rmse": 4675.077814746923} {"canonical_length": 95, "canonical_nll": 1.6129911723889803, "generation_rouge_l_f1": 0.3163841807909605, "generation_token_f1": 0.5084745762711864, "prompt_id": "9fc2e86bc1fbb199", "reading_canonical_log_prob": -153.23416137695312, "reading_distances": [13, 14, 14, 17], "reading_log_probs": [-149.705322265625, -150.27606201171875, -149.64044189453125, -150.45114135742188], "reading_logprob_std": 0.40600685213938303, "tokenization_path_logprob_std": 67.43031207982725, "trajectory_balance_rmse": 2260.505241934343} {"canonical_length": 178, "canonical_nll": 1.2404795443074088, "generation_rouge_l_f1": 0.2791762013729977, "generation_token_f1": 0.4897025171624714, "prompt_id": "0d0deb69605ca531", "reading_canonical_log_prob": -220.80535888671875, "reading_distances": [17, 19, 18, 14], "reading_log_probs": [-221.1854248046875, -220.2928466796875, -222.034912109375, -223.65065002441406], "reading_logprob_std": 1.4293327128213125, "tokenization_path_logprob_std": 77.99659123420233, "trajectory_balance_rmse": 4095.989547547797} {"canonical_length": 20, "canonical_nll": 0.599433708190918, "generation_rouge_l_f1": 0.5853658536585366, "generation_token_f1": 0.5853658536585366, "prompt_id": "5b98bbd785876818", "reading_canonical_log_prob": -11.98867416381836, "reading_distances": [25, 35, 29, 29], "reading_log_probs": [-12.592450141906738, -13.159380912780762, -11.832374572753906, -12.528953552246094], "reading_logprob_std": 0.5436576699187569, "tokenization_path_logprob_std": 16.482031994062417, "trajectory_balance_rmse": 259.6944088756992} {"canonical_length": 322, "canonical_nll": 1.0440785662728067, "generation_rouge_l_f1": 0.26788990825688075, "generation_token_f1": 0.5064220183486238, "prompt_id": "9d8ebc1fd91528b6", "reading_canonical_log_prob": -336.19329833984375, "reading_distances": [21, 16, 23, 24], "reading_log_probs": [-332.77203369140625, -330.229736328125, -333.62005615234375, -332.83099365234375], "reading_logprob_std": 1.4739211325800767, "tokenization_path_logprob_std": 198.66954748295626, "trajectory_balance_rmse": 5606.648494142218} {"canonical_length": 202, "canonical_nll": 1.050861925181776, "generation_rouge_l_f1": 0.24944320712694878, "generation_token_f1": 0.45879732739420936, "prompt_id": "5f17b0be639454cf", "reading_canonical_log_prob": -212.27410888671875, "reading_distances": [20, 15, 22, 24], "reading_log_probs": [-208.61997985839844, -208.41717529296875, -210.2152099609375, -208.72845458984375], "reading_logprob_std": 0.8235047328605468, "tokenization_path_logprob_std": 266.95942879720184, "trajectory_balance_rmse": 4199.898210608302} {"canonical_length": 11, "canonical_nll": 0.7897185412320223, "generation_rouge_l_f1": 0.9, "generation_token_f1": 0.9, "prompt_id": "1932b722766373ae", "reading_canonical_log_prob": -8.686903953552246, "reading_distances": [39, 43, 40, 40], "reading_log_probs": [-9.019661903381348, -9.623262405395508, -9.847228050231934, -8.85777759552002], "reading_logprob_std": 0.47351043891373495, "tokenization_path_logprob_std": 17.051907015777825, "trajectory_balance_rmse": 177.57623665984337} {"canonical_length": 31, "canonical_nll": 1.212675033077117, "generation_rouge_l_f1": 0.23703703703703705, "generation_token_f1": 0.37037037037037046, "prompt_id": "3f8935161866e686", "reading_canonical_log_prob": -37.592926025390625, "reading_distances": [29, 21, 32, 27], "reading_log_probs": [-35.35559844970703, -41.27239227294922, -35.18592071533203, -34.67133331298828], "reading_logprob_std": 3.114340139671681, "tokenization_path_logprob_std": 76.56293926784618, "trajectory_balance_rmse": 522.5415762963511} {"canonical_length": 45, "canonical_nll": 0.9914425320095486, "generation_rouge_l_f1": 0.39999999999999997, "generation_token_f1": 0.4999999999999999, "prompt_id": "88f54467af0c5001", "reading_canonical_log_prob": -44.61491394042969, "reading_distances": [23, 28, 28, 26], "reading_log_probs": [-47.346435546875, -52.735408782958984, -47.28974151611328, -47.541587829589844], "reading_logprob_std": 2.673587256357164, "tokenization_path_logprob_std": 79.56932801069269, "trajectory_balance_rmse": 854.5182971866304} {"canonical_length": 272, "canonical_nll": 1.3404483795166016, "generation_rouge_l_f1": 0.24474187380497134, "generation_token_f1": 0.5889101338432121, "prompt_id": "05940c01ebeedfd9", "reading_canonical_log_prob": -364.6019592285156, "reading_distances": [21, 29, 30, 24], "reading_log_probs": [-361.86248779296875, -363.33050537109375, -364.7895202636719, -365.4743347167969], "reading_logprob_std": 1.606282780492294, "tokenization_path_logprob_std": 195.18264957938908, "trajectory_balance_rmse": 6510.813543744098} {"canonical_length": 26, "canonical_nll": 1.091227311354417, "generation_rouge_l_f1": 0.23076923076923078, "generation_token_f1": 0.23076923076923078, "prompt_id": "cc3b11b7124977c4", "reading_canonical_log_prob": -28.371910095214844, "reading_distances": [22, 25, 18, 24], "reading_log_probs": [-27.618709564208984, -27.238719940185547, -29.844331741333008, -27.185688018798828], "reading_logprob_std": 1.2631213169102309, "tokenization_path_logprob_std": 64.35503087879106, "trajectory_balance_rmse": 547.8061168854119} {"canonical_length": 25, "canonical_nll": 0.7372576904296875, "generation_rouge_l_f1": 0.7741935483870969, "generation_token_f1": 0.9032258064516129, "prompt_id": "9335f10e0202a2ca", "reading_canonical_log_prob": -18.431442260742188, "reading_distances": [21, 24, 28, 25], "reading_log_probs": [-19.894107818603516, -19.443138122558594, -19.95993995666504, -20.182601928710938], "reading_logprob_std": 0.31016200052818355, "tokenization_path_logprob_std": 20.39767539872382, "trajectory_balance_rmse": 246.51662373752313} {"canonical_length": 160, "canonical_nll": 1.2957616806030274, "generation_rouge_l_f1": 0.26252983293556087, "generation_token_f1": 0.43914081145584727, "prompt_id": "ee0b55cb338c65fa", "reading_canonical_log_prob": -207.32186889648438, "reading_distances": [31, 27, 24, 29], "reading_log_probs": [-214.81182861328125, -221.18484497070312, -221.09832763671875, -216.34906005859375], "reading_logprob_std": 3.27167555790096, "tokenization_path_logprob_std": 194.28739937808498, "trajectory_balance_rmse": 3106.5942508955195} {"canonical_length": 205, "canonical_nll": 0.4905602431878811, "generation_rouge_l_f1": 0.4208333333333334, "generation_token_f1": 0.6416666666666667, "prompt_id": "9ed2cebe3982e4a6", "reading_canonical_log_prob": -100.56484985351562, "reading_distances": [11, 9, 12, 13], "reading_log_probs": [-101.34492492675781, -100.3199462890625, -104.6386947631836, -103.009033203125], "reading_logprob_std": 1.897527578038325, "tokenization_path_logprob_std": 224.87558749746643, "trajectory_balance_rmse": 2018.6164480797606} {"canonical_length": 67, "canonical_nll": 1.3880849809788947, "generation_rouge_l_f1": 0.28220858895705525, "generation_token_f1": 0.4662576687116564, "prompt_id": "57c5e7bb82e63751", "reading_canonical_log_prob": -93.00169372558594, "reading_distances": [25, 31, 31, 31], "reading_log_probs": [-90.83477783203125, -87.7013168334961, -90.21446228027344, -87.32151794433594], "reading_logprob_std": 1.7648332880492392, "tokenization_path_logprob_std": 147.20151924228873, "trajectory_balance_rmse": 1344.1025943686795} {"canonical_length": 233, "canonical_nll": 1.3199225822743428, "generation_rouge_l_f1": 0.22356495468277948, "generation_token_f1": 0.4229607250755287, "prompt_id": "2049d32643760e75", "reading_canonical_log_prob": -307.5419616699219, "reading_distances": [33, 23, 35, 30], "reading_log_probs": [-304.9925842285156, -307.6387634277344, -304.6796875, -304.8675842285156], "reading_logprob_std": 1.4019826258421906, "tokenization_path_logprob_std": 244.75910153523853, "trajectory_balance_rmse": 4319.398832588207} {"canonical_length": 65, "canonical_nll": 1.1510049673227163, "generation_rouge_l_f1": 0.5362318840579711, "generation_token_f1": 0.5942028985507245, "prompt_id": "9327dc9ed2ef5314", "reading_canonical_log_prob": -74.81532287597656, "reading_distances": [14, 17, 18, 19], "reading_log_probs": [-75.44464111328125, -75.15592956542969, -75.15167236328125, -75.29086303710938], "reading_logprob_std": 0.13857367242761873, "tokenization_path_logprob_std": 77.8157162374662, "trajectory_balance_rmse": 1312.8252508442022} {"canonical_length": 27, "canonical_nll": 1.609988318549262, "generation_rouge_l_f1": 0.3829787234042554, "generation_token_f1": 0.4680851063829788, "prompt_id": "fb2aa3ee59bdbf8a", "reading_canonical_log_prob": -43.46968460083008, "reading_distances": [103, 96, 95, 98], "reading_log_probs": [-42.785030364990234, -45.30641174316406, -43.46092224121094, -43.81301498413086], "reading_logprob_std": 1.065793168687822, "tokenization_path_logprob_std": 127.95527698629844, "trajectory_balance_rmse": 671.3341714928815} {"canonical_length": 176, "canonical_nll": 0.7241646159778942, "generation_rouge_l_f1": 0.4488778054862843, "generation_token_f1": 0.5885286783042395, "prompt_id": "24055c0763c8403a", "reading_canonical_log_prob": -127.45297241210938, "reading_distances": [140, 136, 147, 136], "reading_log_probs": [-127.16070556640625, -127.42794799804688, -127.98704528808594, -124.55833435058594], "reading_logprob_std": 1.5228789281100716, "tokenization_path_logprob_std": 70.70130133059571, "trajectory_balance_rmse": 3244.8065847227176} {"canonical_length": 73, "canonical_nll": 0.46111438698964574, "generation_rouge_l_f1": 0.6837606837606838, "generation_token_f1": 0.7521367521367521, "prompt_id": "c59968b8c3d9c69d", "reading_canonical_log_prob": -33.66135025024414, "reading_distances": [15, 15, 12, 15], "reading_log_probs": [-33.803104400634766, -34.3182373046875, -33.28054428100586, -34.02881622314453], "reading_logprob_std": 0.4387348371089391, "tokenization_path_logprob_std": 151.8956875100489, "trajectory_balance_rmse": 1093.9985198740908} {"canonical_length": 283, "canonical_nll": 1.4628198846068905, "generation_rouge_l_f1": 0.3238095238095238, "generation_token_f1": 0.518095238095238, "prompt_id": "c3213521900f0885", "reading_canonical_log_prob": -413.97802734375, "reading_distances": [39, 42, 35, 42], "reading_log_probs": [-409.0291442871094, -406.9076843261719, -404.288330078125, -407.26422119140625], "reading_logprob_std": 1.9565047149371178, "tokenization_path_logprob_std": 127.8861303984847, "trajectory_balance_rmse": 4868.412437095462} {"canonical_length": 70, "canonical_nll": 1.530073983328683, "generation_rouge_l_f1": 0.20567375886524825, "generation_token_f1": 0.2907801418439716, "prompt_id": "48bb22e3d5050293", "reading_canonical_log_prob": -107.10517883300781, "reading_distances": [16, 17, 20, 23], "reading_log_probs": [-105.99189758300781, -105.58683776855469, -106.35240936279297, -110.5612564086914], "reading_logprob_std": 2.313338431854013, "tokenization_path_logprob_std": 38.07534909834708, "trajectory_balance_rmse": 1156.997892336035} {"canonical_length": 183, "canonical_nll": 1.3010489041688011, "generation_rouge_l_f1": 0.21923937360178972, "generation_token_f1": 0.447427293064877, "prompt_id": "5a342c18ce379992", "reading_canonical_log_prob": -238.09194946289062, "reading_distances": [30, 32, 26, 26], "reading_log_probs": [-232.9976806640625, -230.23797607421875, -231.8923797607422, -229.7427215576172], "reading_logprob_std": 1.50099348430675, "tokenization_path_logprob_std": 187.27402830100334, "trajectory_balance_rmse": 3642.7202254613744} {"canonical_length": 280, "canonical_nll": 1.5802592686244419, "generation_rouge_l_f1": 0.2788844621513944, "generation_token_f1": 0.5219123505976095, "prompt_id": "66f2043754e3c9e5", "reading_canonical_log_prob": -442.47259521484375, "reading_distances": [21, 15, 28, 21], "reading_log_probs": [-450.23614501953125, -445.07891845703125, -446.7733459472656, -446.2301940917969], "reading_logprob_std": 2.2197436059190343, "tokenization_path_logprob_std": 230.30673843878145, "trajectory_balance_rmse": 5490.9863557381195} {"canonical_length": 61, "canonical_nll": 1.7787733234342982, "generation_rouge_l_f1": 0.20270270270270271, "generation_token_f1": 0.3783783783783784, "prompt_id": "b6ca714dbce654ce", "reading_canonical_log_prob": -108.50517272949219, "reading_distances": [44, 40, 44, 45], "reading_log_probs": [-107.97023010253906, -109.13961791992188, -108.34403991699219, -109.75206756591797], "reading_logprob_std": 0.7996211174931114, "tokenization_path_logprob_std": 144.27621889539145, "trajectory_balance_rmse": 1621.0341918145527} {"canonical_length": 38, "canonical_nll": 1.0382431431820518, "generation_rouge_l_f1": 0.19387755102040816, "generation_token_f1": 0.23469387755102042, "prompt_id": "4112128a24dc114d", "reading_canonical_log_prob": -39.45323944091797, "reading_distances": [28, 33, 29, 29], "reading_log_probs": [-39.23197937011719, -39.64467239379883, -39.368873596191406, -39.76026916503906], "reading_logprob_std": 0.243372785034091, "tokenization_path_logprob_std": 151.45692945604875, "trajectory_balance_rmse": 796.0351260079174} {"canonical_length": 24, "canonical_nll": 0.6601370970408121, "generation_rouge_l_f1": 0.7804878048780488, "generation_token_f1": 0.7804878048780488, "prompt_id": "07948a71ef97c7bb", "reading_canonical_log_prob": -15.843290328979492, "reading_distances": [22, 17, 20, 21], "reading_log_probs": [-14.97425651550293, -16.19293975830078, -14.595972061157227, -16.553430557250977], "reading_logprob_std": 0.9413634541886624, "tokenization_path_logprob_std": 85.68001633187342, "trajectory_balance_rmse": 625.8870465221017} {"canonical_length": 130, "canonical_nll": 1.4711888239933895, "generation_rouge_l_f1": 0.1836228287841191, "generation_token_f1": 0.35732009925558317, "prompt_id": "d894cb33b4a83ecb", "reading_canonical_log_prob": -191.25454711914062, "reading_distances": [22, 32, 25, 33], "reading_log_probs": [-192.437255859375, -196.80535888671875, -200.76309204101562, -193.99169921875], "reading_logprob_std": 3.654315480128893, "tokenization_path_logprob_std": 74.19565703286696, "trajectory_balance_rmse": 2922.8426433074505} {"canonical_length": 49, "canonical_nll": 0.39049876466089367, "generation_rouge_l_f1": 0.29508196721311475, "generation_token_f1": 0.32786885245901637, "prompt_id": "98f5a9670dc63974", "reading_canonical_log_prob": -19.13443946838379, "reading_distances": [14, 19, 21, 21], "reading_log_probs": [-20.23666763305664, -19.336862564086914, -19.838329315185547, -21.21979522705078], "reading_logprob_std": 0.7979251648838983, "tokenization_path_logprob_std": 57.075372468745385, "trajectory_balance_rmse": 693.4251716098094} {"canonical_length": 155, "canonical_nll": 1.3608538227696572, "generation_rouge_l_f1": 0.2115869017632242, "generation_token_f1": 0.36272040302267006, "prompt_id": "9866aa5bc569efc5", "reading_canonical_log_prob": -210.93234252929688, "reading_distances": [11, 8, 10, 13], "reading_log_probs": [-210.68362426757812, -209.71153259277344, -208.9783477783203, -209.95863342285156], "reading_logprob_std": 0.7034510520297638, "tokenization_path_logprob_std": 112.52653763566926, "trajectory_balance_rmse": 3140.336202966883} {"canonical_length": 173, "canonical_nll": 2.0269874175848988, "generation_rouge_l_f1": 0.1592039800995025, "generation_token_f1": 0.2835820895522388, "prompt_id": "63dbcd016e93c526", "reading_canonical_log_prob": -350.6688232421875, "reading_distances": [23, 18, 27, 28], "reading_log_probs": [-339.8346862792969, -345.1296691894531, -338.4852294921875, -338.6806640625], "reading_logprob_std": 3.1220464122821125, "tokenization_path_logprob_std": 175.0181983987535, "trajectory_balance_rmse": 3683.164421812736} {"canonical_length": 45, "canonical_nll": 0.8851430257161458, "generation_rouge_l_f1": 0.3278688524590164, "generation_token_f1": 0.4262295081967213, "prompt_id": "bb113220673ce1f4", "reading_canonical_log_prob": -39.83143615722656, "reading_distances": [24, 30, 27, 22], "reading_log_probs": [-40.84638977050781, -41.59039306640625, -41.45060729980469, -41.2213249206543], "reading_logprob_std": 0.32500228973228784, "tokenization_path_logprob_std": 17.92314602530419, "trajectory_balance_rmse": 574.5352947828574} {"canonical_length": 16, "canonical_nll": 1.895113229751587, "generation_rouge_l_f1": 0.18181818181818185, "generation_token_f1": 0.3636363636363637, "prompt_id": "5d38e8c442eebce3", "reading_canonical_log_prob": -30.32181167602539, "reading_distances": [37, 40, 42, 45], "reading_log_probs": [-31.966629028320312, -32.36606979370117, -32.16047286987305, -32.13336944580078], "reading_logprob_std": 0.16383041775326732, "tokenization_path_logprob_std": 45.255950800267144, "trajectory_balance_rmse": 356.1220371856721} {"canonical_length": 7, "canonical_nll": 0.9223908015659877, "generation_rouge_l_f1": 0.04201680672268908, "generation_token_f1": 0.04201680672268908, "prompt_id": "8ee1f7d9e89653dc", "reading_canonical_log_prob": -6.456735610961914, "reading_distances": [21, 23, 24, 19], "reading_log_probs": [-6.67614221572876, -6.735440254211426, -5.693299293518066, -6.136295318603516], "reading_logprob_std": 0.4917832393023283, "tokenization_path_logprob_std": 39.799398450598694, "trajectory_balance_rmse": 123.86257100473898} {"canonical_length": 220, "canonical_nll": 1.7190202192826705, "generation_rouge_l_f1": 0.2028397565922921, "generation_token_f1": 0.4016227180527383, "prompt_id": "8768fc8d898485e9", "reading_canonical_log_prob": -378.1844482421875, "reading_distances": [16, 15, 12, 16], "reading_log_probs": [-376.952392578125, -379.24627685546875, -376.075927734375, -377.2974853515625], "reading_logprob_std": 1.3382292283050532, "tokenization_path_logprob_std": 92.91831819327442, "trajectory_balance_rmse": 4570.000584287723} {"canonical_length": 238, "canonical_nll": 1.0453161672383797, "generation_rouge_l_f1": 0.2988505747126437, "generation_token_f1": 0.5379310344827587, "prompt_id": "f7a53c2bee8bfe30", "reading_canonical_log_prob": -248.78524780273438, "reading_distances": [14, 15, 19, 24], "reading_log_probs": [-245.5338134765625, -246.18563842773438, -249.3240203857422, -248.135009765625], "reading_logprob_std": 1.7469020793413617, "tokenization_path_logprob_std": 53.29034892704369, "trajectory_balance_rmse": 4228.918957271204} {"canonical_length": 218, "canonical_nll": 1.2449903575652237, "generation_rouge_l_f1": 0.2630480167014614, "generation_token_f1": 0.5469728601252609, "prompt_id": "90b0a660ae2019c3", "reading_canonical_log_prob": -271.40789794921875, "reading_distances": [31, 29, 40, 26], "reading_log_probs": [-269.5914611816406, -270.48626708984375, -267.58135986328125, -271.271728515625], "reading_logprob_std": 1.590040347328935, "tokenization_path_logprob_std": 174.838685563199, "trajectory_balance_rmse": 3686.7719516050965} {"canonical_length": 153, "canonical_nll": 1.1447825712316178, "generation_rouge_l_f1": 0.2845528455284553, "generation_token_f1": 0.35772357723577236, "prompt_id": "e11b424a8c471248", "reading_canonical_log_prob": -175.1517333984375, "reading_distances": [38, 51, 52, 57], "reading_log_probs": [-165.919677734375, -163.4603271484375, -163.75344848632812, -166.01080322265625], "reading_logprob_std": 1.36735014546907, "tokenization_path_logprob_std": 224.4969083583907, "trajectory_balance_rmse": 2706.6307433525794} {"canonical_length": 69, "canonical_nll": 1.767534007196841, "generation_rouge_l_f1": 0.23841059602649003, "generation_token_f1": 0.35761589403973504, "prompt_id": "cc633ec6fcbbe3de", "reading_canonical_log_prob": -121.95984649658203, "reading_distances": [18, 20, 20, 22], "reading_log_probs": [-120.97756958007812, -122.15182495117188, -122.74288177490234, -122.44500732421875], "reading_logprob_std": 0.7731218972333291, "tokenization_path_logprob_std": 129.79775613116942, "trajectory_balance_rmse": 1240.28014080103} {"canonical_length": 51, "canonical_nll": 1.737001755658318, "generation_rouge_l_f1": 0.21153846153846154, "generation_token_f1": 0.33653846153846156, "prompt_id": "66e12e5aa4182ef3", "reading_canonical_log_prob": -88.58708953857422, "reading_distances": [33, 39, 39, 49], "reading_log_probs": [-86.601806640625, -85.30549621582031, -86.89091491699219, -86.32105255126953], "reading_logprob_std": 0.6899563643191936, "tokenization_path_logprob_std": 107.32008841960197, "trajectory_balance_rmse": 884.418713351886} {"canonical_length": 147, "canonical_nll": 1.5488357024938881, "generation_rouge_l_f1": 0.20098039215686278, "generation_token_f1": 0.42156862745098034, "prompt_id": "831cd880037811e2", "reading_canonical_log_prob": -227.67884826660156, "reading_distances": [19, 19, 28, 19], "reading_log_probs": [-224.1432342529297, -223.71685791015625, -223.01190185546875, -223.23654174804688], "reading_logprob_std": 0.5051345376558762, "tokenization_path_logprob_std": 138.90284794872255, "trajectory_balance_rmse": 3376.1986638296603} {"canonical_length": 239, "canonical_nll": 0.9889829627639579, "generation_rouge_l_f1": 0.2653465346534653, "generation_token_f1": 0.5148514851485149, "prompt_id": "61ee3d334ec189e4", "reading_canonical_log_prob": -236.36692810058594, "reading_distances": [22, 26, 31, 29], "reading_log_probs": [-232.6068572998047, -231.21282958984375, -232.8525390625, -231.19586181640625], "reading_logprob_std": 0.8863828930269374, "tokenization_path_logprob_std": 262.9826765408821, "trajectory_balance_rmse": 5373.930758014319} {"canonical_length": 239, "canonical_nll": 1.428285606735421, "generation_rouge_l_f1": 0.326963906581741, "generation_token_f1": 0.5435244161358812, "prompt_id": "978f2098f6623381", "reading_canonical_log_prob": -341.3602600097656, "reading_distances": [16, 19, 23, 20], "reading_log_probs": [-340.5203552246094, -343.4427185058594, -344.2750244140625, -339.9326477050781], "reading_logprob_std": 2.1380036177159396, "tokenization_path_logprob_std": 338.25373165259487, "trajectory_balance_rmse": 4690.757375481203} {"canonical_length": 366, "canonical_nll": 1.863008926474983, "generation_rouge_l_f1": 0.2240527182866557, "generation_token_f1": 0.42504118616144976, "prompt_id": "f4a2dbf6cf78f4f4", "reading_canonical_log_prob": -681.8612670898438, "reading_distances": [41, 37, 37, 36], "reading_log_probs": [-675.8961791992188, -677.1406860351562, -678.47998046875, -676.380859375], "reading_logprob_std": 1.1268354152776487, "tokenization_path_logprob_std": 107.09655936431551, "trajectory_balance_rmse": 7336.50158647485} {"canonical_length": 323, "canonical_nll": 1.5700694931549923, "generation_rouge_l_f1": 0.2064056939501779, "generation_token_f1": 0.4092526690391459, "prompt_id": "607d603874554f5a", "reading_canonical_log_prob": -507.1324462890625, "reading_distances": [21, 20, 22, 23], "reading_log_probs": [-504.7780456542969, -505.03289794921875, -504.1675720214844, -503.8044738769531], "reading_logprob_std": 0.5608865511639854, "tokenization_path_logprob_std": 205.89083523365758, "trajectory_balance_rmse": 5710.217493477298} {"canonical_length": 16, "canonical_nll": 1.3983162641525269, "generation_rouge_l_f1": 0.16, "generation_token_f1": 0.16, "prompt_id": "65ee5494d4aae49e", "reading_canonical_log_prob": -22.37306022644043, "reading_distances": [15, 17, 22, 18], "reading_log_probs": [-23.841833114624023, -25.46994972229004, -25.718929290771484, -25.12613296508789], "reading_logprob_std": 0.8344303506167647, "tokenization_path_logprob_std": 45.30734602954628, "trajectory_balance_rmse": 334.4501746201457} {"canonical_length": 120, "canonical_nll": 1.887964630126953, "generation_rouge_l_f1": 0.31460674157303375, "generation_token_f1": 0.5243445692883896, "prompt_id": "bc654a75738c95db", "reading_canonical_log_prob": -226.55575561523438, "reading_distances": [27, 29, 26, 27], "reading_log_probs": [-217.74169921875, -218.09432983398438, -217.6043701171875, -214.77540588378906], "reading_logprob_std": 1.5329834508533124, "tokenization_path_logprob_std": 160.56590694685033, "trajectory_balance_rmse": 2294.234298660827} {"canonical_length": 166, "canonical_nll": 1.04591746502612, "generation_rouge_l_f1": 0.3020833333333333, "generation_token_f1": 0.5364583333333333, "prompt_id": "1277e399b9abf4ff", "reading_canonical_log_prob": -173.62229919433594, "reading_distances": [28, 32, 24, 24], "reading_log_probs": [-170.69110107421875, -167.93600463867188, -169.103759765625, -170.81201171875], "reading_logprob_std": 1.3747127481657264, "tokenization_path_logprob_std": 122.69291453786737, "trajectory_balance_rmse": 3886.8186783754554} {"canonical_length": 13, "canonical_nll": 2.4300277416522684, "generation_rouge_l_f1": 0.1904761904761905, "generation_token_f1": 0.1904761904761905, "prompt_id": "888761fd7cf1d9af", "reading_canonical_log_prob": -31.590360641479492, "reading_distances": [29, 23, 26, 29], "reading_log_probs": [-30.236083984375, -30.867198944091797, -31.168678283691406, -29.65888214111328], "reading_logprob_std": 0.6727805404977876, "tokenization_path_logprob_std": 73.91193697037394, "trajectory_balance_rmse": 360.8593988246046} {"canonical_length": 196, "canonical_nll": 1.3216107426857462, "generation_rouge_l_f1": 0.17118997912317327, "generation_token_f1": 0.3507306889352818, "prompt_id": "4049144029e8d577", "reading_canonical_log_prob": -259.03570556640625, "reading_distances": [20, 24, 26, 23], "reading_log_probs": [-253.24209594726562, -253.94129943847656, -256.2216491699219, -254.45816040039062], "reading_logprob_std": 1.2722181572236113, "tokenization_path_logprob_std": 135.77770427197729, "trajectory_balance_rmse": 4863.402282062986} {"canonical_length": 105, "canonical_nll": 1.2319046747116815, "generation_rouge_l_f1": 0.3063583815028902, "generation_token_f1": 0.42196531791907516, "prompt_id": "35fc3a98ce123749", "reading_canonical_log_prob": -129.34999084472656, "reading_distances": [37, 34, 41, 33], "reading_log_probs": [-131.41790771484375, -129.67413330078125, -130.48886108398438, -128.66632080078125], "reading_logprob_std": 1.1717589298026714, "tokenization_path_logprob_std": 184.32703988108943, "trajectory_balance_rmse": 2021.5475965846122} {"canonical_length": 17, "canonical_nll": 3.2428086224724266, "generation_rouge_l_f1": 0.04405286343612335, "generation_token_f1": 0.05286343612334802, "prompt_id": "30594b15ee7b3362", "reading_canonical_log_prob": -55.12774658203125, "reading_distances": [20, 22, 21, 20], "reading_log_probs": [-51.178768157958984, -51.0335578918457, -51.779300689697266, -51.446170806884766], "reading_logprob_std": 0.3279464865823365, "tokenization_path_logprob_std": 79.2961479695723, "trajectory_balance_rmse": 379.6466946466306} {"canonical_length": 89, "canonical_nll": 1.1495485627249387, "generation_rouge_l_f1": 0.23140495867768596, "generation_token_f1": 0.3415977961432507, "prompt_id": "55622d0a348ebb7e", "reading_canonical_log_prob": -102.30982208251953, "reading_distances": [18, 15, 16, 18], "reading_log_probs": [-102.92790222167969, -102.08795166015625, -102.77961730957031, -101.79483032226562], "reading_logprob_std": 0.5435595789082849, "tokenization_path_logprob_std": 96.86014887492449, "trajectory_balance_rmse": 1591.6909164100718} {"canonical_length": 181, "canonical_nll": 1.213010840653056, "generation_rouge_l_f1": 0.27378190255220414, "generation_token_f1": 0.46867749419953597, "prompt_id": "96bafc4191db5d29", "reading_canonical_log_prob": -219.55496215820312, "reading_distances": [35, 26, 23, 33], "reading_log_probs": [-230.03173828125, -224.3702850341797, -224.292724609375, -231.8202667236328], "reading_logprob_std": 3.876846848691936, "tokenization_path_logprob_std": 103.04807742807677, "trajectory_balance_rmse": 3493.968841518867} {"canonical_length": 186, "canonical_nll": 1.2902062528876848, "generation_rouge_l_f1": 0.22014051522248246, "generation_token_f1": 0.4590163934426229, "prompt_id": "a69ab9afafd56708", "reading_canonical_log_prob": -239.97836303710938, "reading_distances": [26, 30, 32, 31], "reading_log_probs": [-241.3502197265625, -240.30911254882812, -241.23080444335938, -244.32135009765625], "reading_logprob_std": 1.7422403772487984, "tokenization_path_logprob_std": 123.27143631588942, "trajectory_balance_rmse": 3384.277553958019} {"canonical_length": 7, "canonical_nll": 1.098605019705636, "generation_rouge_l_f1": 0.163265306122449, "generation_token_f1": 0.163265306122449, "prompt_id": "7d277a000cf626ea", "reading_canonical_log_prob": -7.690235137939453, "reading_distances": [35, 27, 28, 28], "reading_log_probs": [-6.918483257293701, -7.72528600692749, -7.967165470123291, -7.286585807800293], "reading_logprob_std": 0.46550302220725726, "tokenization_path_logprob_std": 55.00068591932921, "trajectory_balance_rmse": 185.56726615179207} {"canonical_length": 8, "canonical_nll": 0.42872318625450134, "generation_rouge_l_f1": 0.2105263157894737, "generation_token_f1": 0.2105263157894737, "prompt_id": "4b1173c764ce0a81", "reading_canonical_log_prob": -3.4297854900360107, "reading_distances": [45, 57, 50, 50], "reading_log_probs": [-7.133310317993164, -5.712724208831787, -5.765099048614502, -7.011235237121582], "reading_logprob_std": 0.7717241903310984, "tokenization_path_logprob_std": 37.007258609183324, "trajectory_balance_rmse": 142.92855870083363} {"canonical_length": 48, "canonical_nll": 0.8580527305603027, "generation_rouge_l_f1": 0.3018867924528302, "generation_token_f1": 0.3490566037735849, "prompt_id": "c86574795b337627", "reading_canonical_log_prob": -41.18653106689453, "reading_distances": [8, 10, 9, 8], "reading_log_probs": [-43.20586395263672, -43.721229553222656, -42.40636444091797, -43.061370849609375], "reading_logprob_std": 0.5415253607515631, "tokenization_path_logprob_std": 11.684035595905433, "trajectory_balance_rmse": 497.16414771794626} {"canonical_length": 251, "canonical_nll": 1.2324724539342629, "generation_rouge_l_f1": 0.19764705882352943, "generation_token_f1": 0.40941176470588236, "prompt_id": "f0455cfad098aced", "reading_canonical_log_prob": -309.3505859375, "reading_distances": [40, 40, 42, 42], "reading_log_probs": [-310.1026306152344, -309.88836669921875, -306.6331787109375, -309.51788330078125], "reading_logprob_std": 1.6196697791507182, "tokenization_path_logprob_std": 176.76113924934663, "trajectory_balance_rmse": 5018.989711853039} {"canonical_length": 92, "canonical_nll": 1.4027915622877039, "generation_rouge_l_f1": 0.21864951768488747, "generation_token_f1": 0.3086816720257235, "prompt_id": "765c339341bf24b5", "reading_canonical_log_prob": -129.05682373046875, "reading_distances": [73, 64, 58, 62], "reading_log_probs": [-125.53901672363281, -125.57717895507812, -127.7451400756836, -129.4590301513672], "reading_logprob_std": 1.8916744169321027, "tokenization_path_logprob_std": 120.44676606835954, "trajectory_balance_rmse": 1282.3575675302122} {"canonical_length": 25, "canonical_nll": 2.0080667114257813, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "ef02d7967a69b355", "reading_canonical_log_prob": -50.20166778564453, "reading_distances": [34, 29, 29, 42], "reading_log_probs": [-51.76067352294922, -51.09850311279297, -51.11074447631836, -49.74715042114258], "reading_logprob_std": 0.846603281490829, "tokenization_path_logprob_std": 22.13221172670101, "trajectory_balance_rmse": 476.7302233493058} {"canonical_length": 26, "canonical_nll": 0.9295328580416166, "generation_rouge_l_f1": 0.15789473684210525, "generation_token_f1": 0.15789473684210525, "prompt_id": "27392a7bee3d725d", "reading_canonical_log_prob": -24.16785430908203, "reading_distances": [30, 32, 35, 30], "reading_log_probs": [-23.872161865234375, -23.87560272216797, -24.259017944335938, -22.591623306274414], "reading_logprob_std": 0.7283121880120222, "tokenization_path_logprob_std": 26.294970864680277, "trajectory_balance_rmse": 549.680845395493} {"canonical_length": 53, "canonical_nll": 1.098489005610628, "generation_rouge_l_f1": 0.30344827586206896, "generation_token_f1": 0.4413793103448275, "prompt_id": "deecb5d2b0490e50", "reading_canonical_log_prob": -58.21991729736328, "reading_distances": [15, 23, 21, 17], "reading_log_probs": [-60.32011413574219, -60.345741271972656, -60.555824279785156, -60.01353073120117], "reading_logprob_std": 0.22338129070849752, "tokenization_path_logprob_std": 94.57375087432969, "trajectory_balance_rmse": 1116.5084122841745} {"canonical_length": 124, "canonical_nll": 1.2128379575667843, "generation_rouge_l_f1": 0.22872340425531912, "generation_token_f1": 0.3457446808510638, "prompt_id": "8c1c19dea7da1533", "reading_canonical_log_prob": -150.39190673828125, "reading_distances": [26, 29, 26, 20], "reading_log_probs": [-142.71865844726562, -144.1075439453125, -143.74014282226562, -144.95323181152344], "reading_logprob_std": 0.9259008146770541, "tokenization_path_logprob_std": 205.30410662732595, "trajectory_balance_rmse": 2075.258183213802} {"canonical_length": 81, "canonical_nll": 1.3639872986593364, "generation_rouge_l_f1": 0.17391304347826086, "generation_token_f1": 0.3354037267080745, "prompt_id": "8ddcb00e61289cea", "reading_canonical_log_prob": -110.48297119140625, "reading_distances": [17, 20, 20, 22], "reading_log_probs": [-114.07015991210938, -113.78160858154297, -114.93978881835938, -115.54864501953125], "reading_logprob_std": 0.8093175945233817, "tokenization_path_logprob_std": 49.139123483611286, "trajectory_balance_rmse": 1663.2236533508544} {"canonical_length": 78, "canonical_nll": 0.7581288264347956, "generation_rouge_l_f1": 0.3670886075949368, "generation_token_f1": 0.41139240506329117, "prompt_id": "e509fb17129966c3", "reading_canonical_log_prob": -59.13404846191406, "reading_distances": [48, 40, 43, 52], "reading_log_probs": [-64.36631774902344, -64.30810546875, -63.34947204589844, -68.61187744140625], "reading_logprob_std": 2.3486961406601634, "tokenization_path_logprob_std": 60.893253976036924, "trajectory_balance_rmse": 1146.2517425649787} {"canonical_length": 170, "canonical_nll": 1.040674456428079, "generation_rouge_l_f1": 0.3219814241486068, "generation_token_f1": 0.5201238390092879, "prompt_id": "07a3f091145dd3c4", "reading_canonical_log_prob": -176.91465759277344, "reading_distances": [27, 25, 30, 29], "reading_log_probs": [-173.26976013183594, -175.09002685546875, -174.11740112304688, -173.47145080566406], "reading_logprob_std": 0.8193283011077188, "tokenization_path_logprob_std": 103.72005635251347, "trajectory_balance_rmse": 3881.185014728607} {"canonical_length": 159, "canonical_nll": 1.4140077027134925, "generation_rouge_l_f1": 0.3312883435582822, "generation_token_f1": 0.6319018404907975, "prompt_id": "1cfb5bc2275fa797", "reading_canonical_log_prob": -224.8272247314453, "reading_distances": [24, 27, 18, 22], "reading_log_probs": [-218.67434692382812, -219.86083984375, -222.27761840820312, -218.15023803710938], "reading_logprob_std": 1.8363974438749313, "tokenization_path_logprob_std": 84.58638193226251, "trajectory_balance_rmse": 3006.795272590224} {"canonical_length": 11, "canonical_nll": 0.04471214522014965, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "c9f832251ee1421d", "reading_canonical_log_prob": -0.4918335974216461, "reading_distances": [20, 22, 20, 18], "reading_log_probs": [-0.7424169182777405, -0.7245600819587708, -0.6896920204162598, -0.5888705849647522], "reading_logprob_std": 0.06859764396368019, "tokenization_path_logprob_std": 31.284933890305144, "trajectory_balance_rmse": 122.55141570974627} {"canonical_length": 174, "canonical_nll": 1.0650380452473958, "generation_rouge_l_f1": 0.3095823095823096, "generation_token_f1": 0.520884520884521, "prompt_id": "13d2c1dd5e0685fa", "reading_canonical_log_prob": -185.31661987304688, "reading_distances": [54, 61, 58, 50], "reading_log_probs": [-177.9232177734375, -178.81715393066406, -176.12847900390625, -175.6510772705078], "reading_logprob_std": 1.4906306904551965, "tokenization_path_logprob_std": 84.71512061962984, "trajectory_balance_rmse": 1868.204393487624} {"canonical_length": 266, "canonical_nll": 1.2053997068476856, "generation_rouge_l_f1": 0.24812030075187969, "generation_token_f1": 0.518796992481203, "prompt_id": "e2b16f584507a85d", "reading_canonical_log_prob": -320.6363220214844, "reading_distances": [32, 28, 27, 23], "reading_log_probs": [-320.15692138671875, -321.75982666015625, -321.6676025390625, -323.0338439941406], "reading_logprob_std": 1.1770863646497944, "tokenization_path_logprob_std": 447.408155452525, "trajectory_balance_rmse": 6038.136449691668} {"canonical_length": 279, "canonical_nll": 1.2512992394013216, "generation_rouge_l_f1": 0.3194192377495463, "generation_token_f1": 0.5952813067150636, "prompt_id": "76e96744ab3b125b", "reading_canonical_log_prob": -349.11248779296875, "reading_distances": [18, 25, 16, 24], "reading_log_probs": [-346.55694580078125, -348.1793212890625, -349.42047119140625, -349.49609375], "reading_logprob_std": 1.3769087394252095, "tokenization_path_logprob_std": 270.13251510514607, "trajectory_balance_rmse": 6638.454113294614} {"canonical_length": 157, "canonical_nll": 1.4866085173977408, "generation_rouge_l_f1": 0.3054892601431981, "generation_token_f1": 0.47255369928400953, "prompt_id": "2fad87f3c29a8aa9", "reading_canonical_log_prob": -233.3975372314453, "reading_distances": [27, 26, 28, 22], "reading_log_probs": [-227.47146606445312, -228.84214782714844, -228.9282989501953, -227.79588317871094], "reading_logprob_std": 0.7354609793449072, "tokenization_path_logprob_std": 74.34900827587653, "trajectory_balance_rmse": 3436.7744903352823} {"canonical_length": 135, "canonical_nll": 2.161320439091435, "generation_rouge_l_f1": 0.13612565445026178, "generation_token_f1": 0.2408376963350785, "prompt_id": "057edeee1c342070", "reading_canonical_log_prob": -291.77825927734375, "reading_distances": [19, 19, 23, 21], "reading_log_probs": [-287.4183654785156, -289.48565673828125, -288.4176025390625, -289.1221923828125], "reading_logprob_std": 0.9103335896692095, "tokenization_path_logprob_std": 62.186423289500325, "trajectory_balance_rmse": 3166.6905859841163} {"canonical_length": 273, "canonical_nll": 1.0226592458648123, "generation_rouge_l_f1": 0.24528301886792453, "generation_token_f1": 0.4830188679245283, "prompt_id": "b42b291887a21049", "reading_canonical_log_prob": -279.18597412109375, "reading_distances": [31, 28, 27, 29], "reading_log_probs": [-277.2994384765625, -277.63690185546875, -279.19439697265625, -280.85211181640625], "reading_logprob_std": 1.6288447268391628, "tokenization_path_logprob_std": 102.91900377750567, "trajectory_balance_rmse": 5587.791152547231} {"canonical_length": 8, "canonical_nll": 0.40243610739707947, "generation_rouge_l_f1": 0.923076923076923, "generation_token_f1": 0.923076923076923, "prompt_id": "751d7c04030e09ea", "reading_canonical_log_prob": -3.2194888591766357, "reading_distances": [22, 26, 26, 24], "reading_log_probs": [-3.092595100402832, -3.5714826583862305, -2.5349271297454834, -3.383406162261963], "reading_logprob_std": 0.4522752419119546, "tokenization_path_logprob_std": 73.13500113661732, "trajectory_balance_rmse": 206.6113586956776} {"canonical_length": 45, "canonical_nll": 1.292697991265191, "generation_rouge_l_f1": 0.6373626373626373, "generation_token_f1": 0.7912087912087912, "prompt_id": "bdda8d0e92726300", "reading_canonical_log_prob": -58.171409606933594, "reading_distances": [18, 23, 23, 24], "reading_log_probs": [-59.02554702758789, -58.61844253540039, -57.15440368652344, -58.019256591796875], "reading_logprob_std": 0.8129132797807348, "tokenization_path_logprob_std": 39.39665930456062, "trajectory_balance_rmse": 1010.0040674063795} {"canonical_length": 225, "canonical_nll": 0.7878788248697917, "generation_rouge_l_f1": 0.3612903225806452, "generation_token_f1": 0.6236559139784946, "prompt_id": "ce2f11428dfe632d", "reading_canonical_log_prob": -177.27273559570312, "reading_distances": [35, 33, 26, 26], "reading_log_probs": [-174.55882263183594, -176.89633178710938, -175.4510040283203, -176.02734375], "reading_logprob_std": 0.9828856742378966, "tokenization_path_logprob_std": 60.169374469416596, "trajectory_balance_rmse": 3451.6436936837486} {"canonical_length": 371, "canonical_nll": 1.1146552813342319, "generation_rouge_l_f1": 0.21441124780316342, "generation_token_f1": 0.3971880492091388, "prompt_id": "3ce75f08c1522ce9", "reading_canonical_log_prob": -413.537109375, "reading_distances": [20, 17, 24, 20], "reading_log_probs": [-419.60247802734375, -427.0223083496094, -415.178466796875, -423.1822509765625], "reading_logprob_std": 5.054071925158213, "tokenization_path_logprob_std": 316.46913048823126, "trajectory_balance_rmse": 5976.065705299198} {"canonical_length": 3, "canonical_nll": 1.5504660606384277, "generation_rouge_l_f1": 0.03389830508474576, "generation_token_f1": 0.03389830508474576, "prompt_id": "cc1031d1c53d6dff", "reading_canonical_log_prob": -4.651398181915283, "reading_distances": [40, 29, 40, 40], "reading_log_probs": [-5.9812846183776855, -5.147566318511963, -6.004363059997559, -5.709803581237793], "reading_logprob_std": 0.39857009566178553, "tokenization_path_logprob_std": 13.616322699954043, "trajectory_balance_rmse": 32.38074727460941} {"canonical_length": 14, "canonical_nll": 3.5997012002127513, "generation_rouge_l_f1": 0.07407407407407408, "generation_token_f1": 0.07407407407407408, "prompt_id": "6b9afce682e6088d", "reading_canonical_log_prob": -50.395816802978516, "reading_distances": [55, 50, 45, 53], "reading_log_probs": [-45.265220642089844, -45.018760681152344, -45.30120849609375, -44.638153076171875], "reading_logprob_std": 0.3054413169537541, "tokenization_path_logprob_std": 88.92487022460696, "trajectory_balance_rmse": 375.5143601799303} {"canonical_length": 15, "canonical_nll": 0.6664319356282552, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "16fc84421f512267", "reading_canonical_log_prob": -9.996479034423828, "reading_distances": [24, 30, 36, 24], "reading_log_probs": [-9.828088760375977, -8.184652328491211, -7.9740824699401855, -9.459173202514648], "reading_logprob_std": 0.9196266872615562, "tokenization_path_logprob_std": 13.830876006720803, "trajectory_balance_rmse": 282.15816306362046} {"canonical_length": 12, "canonical_nll": 0.31275741259257, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "7550a792cc616b59", "reading_canonical_log_prob": -3.75308895111084, "reading_distances": [24, 37, 26, 31], "reading_log_probs": [-6.934922218322754, -9.058382987976074, -7.6864013671875, -7.920326232910156], "reading_logprob_std": 0.8792541920920014, "tokenization_path_logprob_std": 27.78542344210494, "trajectory_balance_rmse": 181.5149707559672} {"canonical_length": 242, "canonical_nll": 1.5391107984810821, "generation_rouge_l_f1": 0.18897637795275593, "generation_token_f1": 0.3228346456692914, "prompt_id": "d8ba10df8798bb3c", "reading_canonical_log_prob": -372.4648132324219, "reading_distances": [33, 38, 35, 38], "reading_log_probs": [-373.3001708984375, -372.8436279296875, -372.00872802734375, -370.6244201660156], "reading_logprob_std": 1.175235698128965, "tokenization_path_logprob_std": 493.20818520992606, "trajectory_balance_rmse": 5422.730171729404} {"canonical_length": 15, "canonical_nll": 1.2801508585611978, "generation_rouge_l_f1": 0.3157894736842105, "generation_token_f1": 0.3157894736842105, "prompt_id": "c62abdde12be0560", "reading_canonical_log_prob": -19.20226287841797, "reading_distances": [25, 28, 29, 29], "reading_log_probs": [-20.471620559692383, -20.251522064208984, -20.787376403808594, -20.16604995727539], "reading_logprob_std": 0.2771898268578861, "tokenization_path_logprob_std": 57.934856916153535, "trajectory_balance_rmse": 273.0323117508966} {"canonical_length": 318, "canonical_nll": 1.547247736708923, "generation_rouge_l_f1": 0.21754385964912284, "generation_token_f1": 0.4771929824561404, "prompt_id": "e65b179e41f3017d", "reading_canonical_log_prob": -492.0247802734375, "reading_distances": [21, 27, 27, 30], "reading_log_probs": [-492.94091796875, -492.47607421875, -491.43280029296875, -492.376220703125], "reading_logprob_std": 0.6323115606526142, "tokenization_path_logprob_std": 301.77950076481454, "trajectory_balance_rmse": 6594.051731678063} {"canonical_length": 83, "canonical_nll": 1.2288500314735504, "generation_rouge_l_f1": 0.18028169014084505, "generation_token_f1": 0.3436619718309859, "prompt_id": "688f5dc10c3a0a47", "reading_canonical_log_prob": -101.99455261230469, "reading_distances": [33, 34, 24, 33], "reading_log_probs": [-114.20707702636719, -111.9814453125, -108.31524658203125, -108.39447021484375], "reading_logprob_std": 2.8833889251255, "tokenization_path_logprob_std": 29.52654207088559, "trajectory_balance_rmse": 1704.3756248445607} {"canonical_length": 324, "canonical_nll": 1.6067887294439622, "generation_rouge_l_f1": 0.3525305410122164, "generation_token_f1": 0.5549738219895288, "prompt_id": "d7836a3fff052985", "reading_canonical_log_prob": -520.5995483398438, "reading_distances": [22, 21, 22, 22], "reading_log_probs": [-520.4908447265625, -519.6460571289062, -520.212158203125, -520.6383056640625], "reading_logprob_std": 0.4377695580822843, "tokenization_path_logprob_std": 348.3616235947198, "trajectory_balance_rmse": 5845.158739321471} {"canonical_length": 18, "canonical_nll": 1.3522861268785265, "generation_rouge_l_f1": 0.5, "generation_token_f1": 0.5, "prompt_id": "97d1bc52cae659b0", "reading_canonical_log_prob": -24.341150283813477, "reading_distances": [9, 12, 13, 10], "reading_log_probs": [-24.828414916992188, -24.18732452392578, -25.400901794433594, -24.660140991210938], "reading_logprob_std": 0.501007817790199, "tokenization_path_logprob_std": 24.5182319915269, "trajectory_balance_rmse": 228.28081081077406} {"canonical_length": 311, "canonical_nll": 1.0164757633515875, "generation_rouge_l_f1": 0.32111692844677137, "generation_token_f1": 0.537521815008726, "prompt_id": "d655055a06f3b252", "reading_canonical_log_prob": -316.12396240234375, "reading_distances": [30, 34, 35, 31], "reading_log_probs": [-315.08984375, -312.06304931640625, -314.1181945800781, -314.83587646484375], "reading_logprob_std": 1.3722653342860767, "tokenization_path_logprob_std": 334.2584819215336, "trajectory_balance_rmse": 5849.494115725062} {"canonical_length": 270, "canonical_nll": 1.4002417670355902, "generation_rouge_l_f1": 0.27007299270072993, "generation_token_f1": 0.4671532846715329, "prompt_id": "62d551bf467199ab", "reading_canonical_log_prob": -378.0652770996094, "reading_distances": [18, 19, 24, 20], "reading_log_probs": [-373.412353515625, -370.8473205566406, -370.8311767578125, -367.4348449707031], "reading_logprob_std": 2.452087474570583, "tokenization_path_logprob_std": 301.8447035575873, "trajectory_balance_rmse": 5024.099491133763} {"canonical_length": 11, "canonical_nll": 0.43064455552534625, "generation_rouge_l_f1": 0.8, "generation_token_f1": 0.8, "prompt_id": "1237b2c3cf7851d2", "reading_canonical_log_prob": -4.737090110778809, "reading_distances": [20, 20, 16, 18], "reading_log_probs": [-4.257770538330078, -4.3414483070373535, -4.7629618644714355, -4.558579444885254], "reading_logprob_std": 0.22717431391022422, "tokenization_path_logprob_std": 64.35897335424613, "trajectory_balance_rmse": 296.6978647446075} {"canonical_length": 3, "canonical_nll": 2.06809933980306, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.6666666666666666, "prompt_id": "9af597ad9c807de8", "reading_canonical_log_prob": -6.20429801940918, "reading_distances": [17, 19, 21, 18], "reading_log_probs": [-5.924584865570068, -5.979708671569824, -6.143928050994873, -7.700678825378418], "reading_logprob_std": 0.8474392342753546, "tokenization_path_logprob_std": 13.806243674125422, "trajectory_balance_rmse": 37.14764642758037} {"canonical_length": 158, "canonical_nll": 1.3316026518616495, "generation_rouge_l_f1": 0.34626865671641793, "generation_token_f1": 0.6567164179104478, "prompt_id": "0366eeefae975865", "reading_canonical_log_prob": -210.39321899414062, "reading_distances": [31, 34, 31, 32], "reading_log_probs": [-204.9267578125, -203.13467407226562, -207.7503662109375, -203.97830200195312], "reading_logprob_std": 2.0068359703719856, "tokenization_path_logprob_std": 167.4262633220867, "trajectory_balance_rmse": 2777.1528047420898} {"canonical_length": 94, "canonical_nll": 1.7282159683552194, "generation_rouge_l_f1": 0.2028169014084507, "generation_token_f1": 0.3042253521126761, "prompt_id": "e99bd769bc8f61be", "reading_canonical_log_prob": -162.45230102539062, "reading_distances": [18, 16, 21, 22], "reading_log_probs": [-160.5147705078125, -163.4421844482422, -160.581298828125, -162.17584228515625], "reading_logprob_std": 1.4042849807902609, "tokenization_path_logprob_std": 265.0056241266746, "trajectory_balance_rmse": 2244.255217037264} {"canonical_length": 287, "canonical_nll": 0.3058682750741779, "generation_rouge_l_f1": 0.5729166666666666, "generation_token_f1": 0.6510416666666666, "prompt_id": "e1ea57924db21e38", "reading_canonical_log_prob": -87.78419494628906, "reading_distances": [20, 19, 16, 21], "reading_log_probs": [-88.73516845703125, -88.04244232177734, -88.55906677246094, -89.09984588623047], "reading_logprob_std": 0.43982156421585, "tokenization_path_logprob_std": 114.17597664157891, "trajectory_balance_rmse": 2421.6010213091376} {"canonical_length": 248, "canonical_nll": 0.7401150118920111, "generation_rouge_l_f1": 0.4601449275362319, "generation_token_f1": 0.6630434782608695, "prompt_id": "d13dad8f3ef63531", "reading_canonical_log_prob": -183.54852294921875, "reading_distances": [29, 34, 32, 32], "reading_log_probs": [-184.95559692382812, -184.3310546875, -188.99871826171875, -187.75465393066406], "reading_logprob_std": 2.229114442348991, "tokenization_path_logprob_std": 173.8348707446069, "trajectory_balance_rmse": 3362.605959533523} {"canonical_length": 21, "canonical_nll": 1.3726537795293898, "generation_rouge_l_f1": 0.5333333333333333, "generation_token_f1": 0.5333333333333333, "prompt_id": "146de30da91fe96b", "reading_canonical_log_prob": -28.825729370117188, "reading_distances": [19, 19, 18, 14], "reading_log_probs": [-28.28616714477539, -29.271753311157227, -27.919891357421875, -28.068544387817383], "reading_logprob_std": 0.6089771401510241, "tokenization_path_logprob_std": 51.36544385758203, "trajectory_balance_rmse": 445.1087767263322} {"canonical_length": 141, "canonical_nll": 1.0267124040752438, "generation_rouge_l_f1": 0.19473684210526315, "generation_token_f1": 0.4, "prompt_id": "e8d40d5d31440017", "reading_canonical_log_prob": -144.76644897460938, "reading_distances": [76, 71, 76, 68], "reading_log_probs": [-140.76434326171875, -142.80349731445312, -140.50767517089844, -142.34507751464844], "reading_logprob_std": 1.1394349454342325, "tokenization_path_logprob_std": 109.63598619211952, "trajectory_balance_rmse": 2635.2032075347515} {"canonical_length": 7, "canonical_nll": 0.047342538833618164, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "e77edca13c4c71bd", "reading_canonical_log_prob": -0.33139777183532715, "reading_distances": [13, 20, 17, 13], "reading_log_probs": [-0.2509951889514923, -0.312846302986145, -0.4967299699783325, -0.3969269394874573], "reading_logprob_std": 0.10659527039392583, "tokenization_path_logprob_std": 28.161090541950166, "trajectory_balance_rmse": 212.98499819884438} {"canonical_length": 255, "canonical_nll": 1.4205020680147058, "generation_rouge_l_f1": 0.174573055028463, "generation_token_f1": 0.34535104364326374, "prompt_id": "5dc83250d862853a", "reading_canonical_log_prob": -362.22802734375, "reading_distances": [31, 29, 31, 27], "reading_log_probs": [-363.780029296875, -361.35369873046875, -362.44610595703125, -357.83734130859375], "reading_logprob_std": 2.545924957519173, "tokenization_path_logprob_std": 360.1494387037322, "trajectory_balance_rmse": 5364.147020159841} {"canonical_length": 25, "canonical_nll": 1.3318600463867187, "generation_rouge_l_f1": 0.10362694300518134, "generation_token_f1": 0.1761658031088083, "prompt_id": "a9f2944f6ac15039", "reading_canonical_log_prob": -33.29650115966797, "reading_distances": [26, 27, 34, 28], "reading_log_probs": [-27.404953002929688, -27.286895751953125, -27.403133392333984, -28.433488845825195], "reading_logprob_std": 0.5370945042624522, "tokenization_path_logprob_std": 75.17315682685611, "trajectory_balance_rmse": 469.4756912425875} {"canonical_length": 51, "canonical_nll": 1.4384792552274817, "generation_rouge_l_f1": 0.28571428571428575, "generation_token_f1": 0.46616541353383456, "prompt_id": "0e9e30fa66e7e5e1", "reading_canonical_log_prob": -73.36244201660156, "reading_distances": [14, 18, 21, 19], "reading_log_probs": [-72.64622497558594, -74.8773193359375, -74.3154067993164, -72.18533325195312], "reading_logprob_std": 1.2934488547655651, "tokenization_path_logprob_std": 72.12133208380968, "trajectory_balance_rmse": 846.1343611611966} {"canonical_length": 102, "canonical_nll": 2.349698683794807, "generation_rouge_l_f1": 0.14769230769230768, "generation_token_f1": 0.2769230769230769, "prompt_id": "f6e26ca6d4af6e82", "reading_canonical_log_prob": -239.6692657470703, "reading_distances": [27, 34, 25, 22], "reading_log_probs": [-232.089599609375, -241.19476318359375, -246.56053161621094, -233.40072631835938], "reading_logprob_std": 6.811448005094234, "tokenization_path_logprob_std": 26.25352551612198, "trajectory_balance_rmse": 1993.1030031167563} {"canonical_length": 92, "canonical_nll": 1.7368008157481318, "generation_rouge_l_f1": 0.21848739495798317, "generation_token_f1": 0.3585434173669468, "prompt_id": "e5d1110bf220108e", "reading_canonical_log_prob": -159.78567504882812, "reading_distances": [30, 29, 24, 27], "reading_log_probs": [-159.1311798095703, -158.96188354492188, -159.30894470214844, -160.06475830078125], "reading_logprob_std": 0.48647274258640943, "tokenization_path_logprob_std": 139.9107394626763, "trajectory_balance_rmse": 1824.9458558950398} {"canonical_length": 10, "canonical_nll": 0.9542611122131348, "generation_rouge_l_f1": 0.0915032679738562, "generation_token_f1": 0.0915032679738562, "prompt_id": "dd04d981a73d4900", "reading_canonical_log_prob": -9.542611122131348, "reading_distances": [20, 14, 16, 15], "reading_log_probs": [-17.989009857177734, -14.909476280212402, -17.207763671875, -15.609817504882812], "reading_logprob_std": 1.416582260853529, "tokenization_path_logprob_std": 26.882790924225752, "trajectory_balance_rmse": 130.65158646009624} {"canonical_length": 299, "canonical_nll": 1.6040743314302885, "generation_rouge_l_f1": 0.2735346358792185, "generation_token_f1": 0.5186500888099467, "prompt_id": "d9940a473d55b366", "reading_canonical_log_prob": -479.61822509765625, "reading_distances": [26, 27, 16, 26], "reading_log_probs": [-476.2945556640625, -479.19439697265625, -478.34161376953125, -477.91357421875], "reading_logprob_std": 1.216953558347769, "tokenization_path_logprob_std": 278.8056362814835, "trajectory_balance_rmse": 5699.355466082518} {"canonical_length": 416, "canonical_nll": 1.1865677466759315, "generation_rouge_l_f1": 0.2835130970724191, "generation_token_f1": 0.4530046224961479, "prompt_id": "7c0396dbe5594b7c", "reading_canonical_log_prob": -493.6121826171875, "reading_distances": [25, 21, 23, 25], "reading_log_probs": [-492.1358642578125, -493.371826171875, -491.93524169921875, -491.9774169921875], "reading_logprob_std": 0.6833061455663305, "tokenization_path_logprob_std": 80.70699718241924, "trajectory_balance_rmse": 6018.458289037356} {"canonical_length": 138, "canonical_nll": 1.4021694902060688, "generation_rouge_l_f1": 0.22330097087378642, "generation_token_f1": 0.43203883495145634, "prompt_id": "5ee117416302abcd", "reading_canonical_log_prob": -193.4993896484375, "reading_distances": [20, 17, 19, 22], "reading_log_probs": [-203.03363037109375, -202.8465576171875, -206.1075897216797, -210.0800323486328], "reading_logprob_std": 3.389613923870928, "tokenization_path_logprob_std": 68.41874714157309, "trajectory_balance_rmse": 2656.7576092393656} {"canonical_length": 326, "canonical_nll": 1.5296420231918615, "generation_rouge_l_f1": 0.24738675958188155, "generation_token_f1": 0.5226480836236934, "prompt_id": "7d6d0adc840aad0a", "reading_canonical_log_prob": -498.6632995605469, "reading_distances": [26, 24, 27, 19], "reading_log_probs": [-497.711181640625, -500.45599365234375, -497.24853515625, -497.99981689453125], "reading_logprob_std": 1.435164438285581, "tokenization_path_logprob_std": 368.178633499075, "trajectory_balance_rmse": 6781.564774659471} {"canonical_length": 17, "canonical_nll": 1.7620152866139132, "generation_rouge_l_f1": 0.22222222222222224, "generation_token_f1": 0.3703703703703704, "prompt_id": "305ea9936db72d06", "reading_canonical_log_prob": -29.954259872436523, "reading_distances": [62, 54, 57, 53], "reading_log_probs": [-28.553115844726562, -25.476966857910156, -26.400144577026367, -27.090768814086914], "reading_logprob_std": 1.2964699636031805, "tokenization_path_logprob_std": 16.546687065825477, "trajectory_balance_rmse": 354.1122777571872} {"canonical_length": 8, "canonical_nll": 1.2508270740509033, "generation_rouge_l_f1": 0.5714285714285714, "generation_token_f1": 0.5714285714285714, "prompt_id": "f9a917228bf6b5fe", "reading_canonical_log_prob": -10.006616592407227, "reading_distances": [32, 38, 38, 37], "reading_log_probs": [-10.788261413574219, -10.185190200805664, -11.657782554626465, -10.55463695526123], "reading_logprob_std": 0.6255858353200551, "tokenization_path_logprob_std": 25.977653343592266, "trajectory_balance_rmse": 116.61116186301565} {"canonical_length": 206, "canonical_nll": 1.2554476839824789, "generation_rouge_l_f1": 0.4615384615384615, "generation_token_f1": 0.6307692307692306, "prompt_id": "6fb0440b40107b60", "reading_canonical_log_prob": -258.6222229003906, "reading_distances": [13, 11, 13, 12], "reading_log_probs": [-253.91705322265625, -255.56329345703125, -254.26666259765625, -256.01275634765625], "reading_logprob_std": 1.0064962771838464, "tokenization_path_logprob_std": 199.72422258053516, "trajectory_balance_rmse": 2845.4750219491416} {"canonical_length": 94, "canonical_nll": 1.9723059471617355, "generation_rouge_l_f1": 0.22608695652173913, "generation_token_f1": 0.3884057971014493, "prompt_id": "16cf88f25ee512b8", "reading_canonical_log_prob": -185.39675903320312, "reading_distances": [17, 17, 16, 22], "reading_log_probs": [-181.9580078125, -183.73826599121094, -183.50119018554688, -182.17491149902344], "reading_logprob_std": 0.9063238746339367, "tokenization_path_logprob_std": 41.88124038466525, "trajectory_balance_rmse": 2267.9299524909366} {"canonical_length": 12, "canonical_nll": 1.3389403025309246, "generation_rouge_l_f1": 0.1258741258741259, "generation_token_f1": 0.15384615384615385, "prompt_id": "024090bc27e03da4", "reading_canonical_log_prob": -16.067283630371094, "reading_distances": [26, 19, 17, 29], "reading_log_probs": [-13.667413711547852, -13.76064395904541, -12.693303108215332, -13.336714744567871], "reading_logprob_std": 0.483033483817547, "tokenization_path_logprob_std": 55.204724706141235, "trajectory_balance_rmse": 319.0403216104305} {"canonical_length": 58, "canonical_nll": 1.98306182335163, "generation_rouge_l_f1": 0.24242424242424243, "generation_token_f1": 0.303030303030303, "prompt_id": "d97dcaaab25ed5d3", "reading_canonical_log_prob": -115.01758575439453, "reading_distances": [11, 17, 10, 15], "reading_log_probs": [-124.83145141601562, -128.1802978515625, -132.96112060546875, -126.23906707763672], "reading_logprob_std": 3.5484541491401895, "tokenization_path_logprob_std": 115.13764669500067, "trajectory_balance_rmse": 1358.3177155958422} {"canonical_length": 386, "canonical_nll": 0.8431855038657707, "generation_rouge_l_f1": 0.3694915254237289, "generation_token_f1": 0.511864406779661, "prompt_id": "5a59813799d04978", "reading_canonical_log_prob": -325.4696044921875, "reading_distances": [9, 11, 12, 14], "reading_log_probs": [-323.8492736816406, -325.27923583984375, -324.09942626953125, -323.63238525390625], "reading_logprob_std": 0.7346543102079877, "tokenization_path_logprob_std": 245.33131951629045, "trajectory_balance_rmse": 5679.322876885531} {"canonical_length": 120, "canonical_nll": 1.0961560567220052, "generation_rouge_l_f1": 0.3636363636363636, "generation_token_f1": 0.3636363636363636, "prompt_id": "84eecbd6c7c38a00", "reading_canonical_log_prob": -131.53872680664062, "reading_distances": [20, 24, 27, 23], "reading_log_probs": [-135.21095275878906, -136.48788452148438, -133.67848205566406, -133.68896484375], "reading_logprob_std": 1.354691596583688, "tokenization_path_logprob_std": 130.12995494837963, "trajectory_balance_rmse": 2061.9704503414287} {"canonical_length": 12, "canonical_nll": 0.12283927202224731, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "1914d368729c6a5a", "reading_canonical_log_prob": -1.4740712642669678, "reading_distances": [23, 22, 23, 27], "reading_log_probs": [-1.781080961227417, -1.4454967975616455, -2.029163360595703, -1.9940476417541504], "reading_logprob_std": 0.2680678107575605, "tokenization_path_logprob_std": 21.981050572066817, "trajectory_balance_rmse": 261.69900387840283} {"canonical_length": 313, "canonical_nll": 1.1124924729807308, "generation_rouge_l_f1": 0.15280135823429541, "generation_token_f1": 0.37011884550084884, "prompt_id": "dc582c1cebaf17e1", "reading_canonical_log_prob": -348.21014404296875, "reading_distances": [43, 44, 34, 50], "reading_log_probs": [-350.2152099609375, -353.7485656738281, -349.2607116699219, -349.87786865234375], "reading_logprob_std": 2.0210038857234163, "tokenization_path_logprob_std": 225.79275399578816, "trajectory_balance_rmse": 6863.580219998492} {"canonical_length": 34, "canonical_nll": 1.1580199073342716, "generation_rouge_l_f1": 0.10526315789473684, "generation_token_f1": 0.12631578947368424, "prompt_id": "c32c06c0fe76b404", "reading_canonical_log_prob": -39.372676849365234, "reading_distances": [19, 19, 20, 22], "reading_log_probs": [-40.834922790527344, -42.836181640625, -40.379920959472656, -40.72123336791992], "reading_logprob_std": 1.112343840213214, "tokenization_path_logprob_std": 42.473391500703244, "trajectory_balance_rmse": 516.904610062173} {"canonical_length": 124, "canonical_nll": 1.7269725184286795, "generation_rouge_l_f1": 0.22222222222222218, "generation_token_f1": 0.3669250645994832, "prompt_id": "3415a33aecea8511", "reading_canonical_log_prob": -214.14459228515625, "reading_distances": [22, 20, 19, 23], "reading_log_probs": [-212.31015014648438, -212.57833862304688, -216.1121826171875, -213.84585571289062], "reading_logprob_std": 1.7348548964068666, "tokenization_path_logprob_std": 91.533747934592, "trajectory_balance_rmse": 2962.3207366416746} {"canonical_length": 333, "canonical_nll": 1.4876688822611674, "generation_rouge_l_f1": 0.22628951747088186, "generation_token_f1": 0.47920133111480856, "prompt_id": "f7e2c3707f2870f0", "reading_canonical_log_prob": -495.39373779296875, "reading_distances": [58, 53, 59, 53], "reading_log_probs": [-491.0830993652344, -489.60906982421875, -489.4990234375, -491.25531005859375], "reading_logprob_std": 0.9362369699787274, "tokenization_path_logprob_std": 194.8946496480197, "trajectory_balance_rmse": 6550.215362108625} {"canonical_length": 39, "canonical_nll": 1.5532901470477765, "generation_rouge_l_f1": 0.1015873015873016, "generation_token_f1": 0.15873015873015875, "prompt_id": "ec9f4061e1384314", "reading_canonical_log_prob": -60.57831573486328, "reading_distances": [19, 18, 19, 16], "reading_log_probs": [-58.93391418457031, -60.36697006225586, -59.99273681640625, -58.78003692626953], "reading_logprob_std": 0.7814239698997906, "tokenization_path_logprob_std": 71.80333187467288, "trajectory_balance_rmse": 814.930928326995} {"canonical_length": 14, "canonical_nll": 1.5227628435407365, "generation_rouge_l_f1": 0.45161290322580644, "generation_token_f1": 0.45161290322580644, "prompt_id": "a05868d3246d849c", "reading_canonical_log_prob": -21.318679809570312, "reading_distances": [10, 9, 8, 9], "reading_log_probs": [-20.26900863647461, -20.625085830688477, -20.992332458496094, -20.601417541503906], "reading_logprob_std": 0.29562486726730836, "tokenization_path_logprob_std": 103.97134674553904, "trajectory_balance_rmse": 405.7518231504156} {"canonical_length": 127, "canonical_nll": 1.4651920588936393, "generation_rouge_l_f1": 0.23762376237623767, "generation_token_f1": 0.41089108910891087, "prompt_id": "420ee5406fba3002", "reading_canonical_log_prob": -186.0793914794922, "reading_distances": [21, 24, 21, 25], "reading_log_probs": [-185.68710327148438, -184.96173095703125, -185.08761596679688, -184.25099182128906], "reading_logprob_std": 0.5894137791091367, "tokenization_path_logprob_std": 138.55666473334045, "trajectory_balance_rmse": 2800.1703547553266} {"canonical_length": 9, "canonical_nll": 1.486888779534234, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "0a4e72c4105acd5f", "reading_canonical_log_prob": -13.381999015808105, "reading_distances": [16, 8, 15, 16], "reading_log_probs": [-11.673103332519531, -12.500017166137695, -12.269834518432617, -12.337371826171875], "reading_logprob_std": 0.36114702384477537, "tokenization_path_logprob_std": 24.236107432894517, "trajectory_balance_rmse": 138.36373120158692} {"canonical_length": 16, "canonical_nll": 1.3252403736114502, "generation_rouge_l_f1": 0.4210526315789473, "generation_token_f1": 0.4210526315789473, "prompt_id": "0ce850d0751da7a2", "reading_canonical_log_prob": -21.203845977783203, "reading_distances": [61, 60, 73, 64], "reading_log_probs": [-24.344633102416992, -24.468395233154297, -23.680625915527344, -26.003334045410156], "reading_logprob_std": 0.9823060512267264, "tokenization_path_logprob_std": 28.483178212408987, "trajectory_balance_rmse": 373.0149459734815} {"canonical_length": 9, "canonical_nll": 3.758661058213976, "generation_rouge_l_f1": 0.22222222222222224, "generation_token_f1": 0.22222222222222224, "prompt_id": "eeee24292d24cae3", "reading_canonical_log_prob": -33.82794952392578, "reading_distances": [19, 19, 26, 23], "reading_log_probs": [-36.466190338134766, -36.3476448059082, -36.029850006103516, -35.510311126708984], "reading_logprob_std": 0.4272189980668307, "tokenization_path_logprob_std": 31.08305436990897, "trajectory_balance_rmse": 182.5067069725086} {"canonical_length": 20, "canonical_nll": 2.3494289398193358, "generation_rouge_l_f1": 0.09395973154362416, "generation_token_f1": 0.10067114093959731, "prompt_id": "aa417b16bf366ab6", "reading_canonical_log_prob": -46.98857879638672, "reading_distances": [31, 33, 27, 38], "reading_log_probs": [-37.783851623535156, -40.022911071777344, -39.78253173828125, -40.57493591308594], "reading_logprob_std": 1.2175343578609654, "tokenization_path_logprob_std": 52.9146821453107, "trajectory_balance_rmse": 402.26379922555185} {"canonical_length": 277, "canonical_nll": 1.3958286326714802, "generation_rouge_l_f1": 0.24277456647398843, "generation_token_f1": 0.47398843930635837, "prompt_id": "31ffef5f718912de", "reading_canonical_log_prob": -386.64453125, "reading_distances": [15, 18, 20, 15], "reading_log_probs": [-383.67376708984375, -389.1209716796875, -390.05633544921875, -385.61187744140625], "reading_logprob_std": 2.987576263716159, "tokenization_path_logprob_std": 252.41729516160697, "trajectory_balance_rmse": 5904.0354344199295} {"canonical_length": 170, "canonical_nll": 1.0835190716911764, "generation_rouge_l_f1": 0.15421686746987953, "generation_token_f1": 0.26987951807228916, "prompt_id": "fe5c1f86ece4d243", "reading_canonical_log_prob": -184.1982421875, "reading_distances": [18, 14, 18, 19], "reading_log_probs": [-184.27178955078125, -179.9618682861328, -183.45245361328125, -177.73988342285156], "reading_logprob_std": 3.0505107658305084, "tokenization_path_logprob_std": 55.161444369655214, "trajectory_balance_rmse": 2966.859350884013} {"canonical_length": 166, "canonical_nll": 1.3918933179005082, "generation_rouge_l_f1": 0.42297650130548303, "generation_token_f1": 0.5848563968668408, "prompt_id": "8db7426fa0df6c62", "reading_canonical_log_prob": -231.05429077148438, "reading_distances": [18, 17, 24, 22], "reading_log_probs": [-230.63165283203125, -230.40655517578125, -226.04949951171875, -230.34703063964844], "reading_logprob_std": 2.2095259450796694, "tokenization_path_logprob_std": 116.1203364323401, "trajectory_balance_rmse": 2148.4941448124714} {"canonical_length": 39, "canonical_nll": 0.610107666406876, "generation_rouge_l_f1": 0.3125, "generation_token_f1": 0.453125, "prompt_id": "6c3aa6c41d3147aa", "reading_canonical_log_prob": -23.794198989868164, "reading_distances": [49, 49, 41, 43], "reading_log_probs": [-23.481849670410156, -24.340970993041992, -24.944347381591797, -23.487770080566406], "reading_logprob_std": 0.7124287303910721, "tokenization_path_logprob_std": 88.01661677803703, "trajectory_balance_rmse": 708.9702456360968} {"canonical_length": 8, "canonical_nll": 1.6981130838394165, "generation_rouge_l_f1": 0.5454545454545454, "generation_token_f1": 0.5454545454545454, "prompt_id": "e07ed69174530f07", "reading_canonical_log_prob": -13.584904670715332, "reading_distances": [35, 38, 33, 38], "reading_log_probs": [-12.608579635620117, -13.058916091918945, -14.574502944946289, -13.203311920166016], "reading_logprob_std": 0.8475151111252333, "tokenization_path_logprob_std": 44.117091569869146, "trajectory_balance_rmse": 118.59627152503923} {"canonical_length": 57, "canonical_nll": 2.007944475140488, "generation_rouge_l_f1": 0.14814814814814814, "generation_token_f1": 0.2777777777777778, "prompt_id": "93c96f6a3b9968fc", "reading_canonical_log_prob": -114.45283508300781, "reading_distances": [25, 18, 29, 22], "reading_log_probs": [-115.96879577636719, -115.41654968261719, -114.74729919433594, -114.69346618652344], "reading_logprob_std": 0.6053331838627647, "tokenization_path_logprob_std": 126.99414205635972, "trajectory_balance_rmse": 1222.9126207265556} {"canonical_length": 264, "canonical_nll": 1.359983964399858, "generation_rouge_l_f1": 0.3092783505154639, "generation_token_f1": 0.4948453608247423, "prompt_id": "0a9cdb872ed6cdad", "reading_canonical_log_prob": -359.0357666015625, "reading_distances": [23, 20, 21, 19], "reading_log_probs": [-356.34405517578125, -354.6386413574219, -355.01123046875, -355.287353515625], "reading_logprob_std": 0.732425789027882, "tokenization_path_logprob_std": 246.22518020533818, "trajectory_balance_rmse": 4921.764934282386} {"canonical_length": 141, "canonical_nll": 1.523913877230164, "generation_rouge_l_f1": 0.3333333333333333, "generation_token_f1": 0.5246913580246915, "prompt_id": "c6291fd5f11c978c", "reading_canonical_log_prob": -214.87185668945312, "reading_distances": [23, 17, 21, 19], "reading_log_probs": [-214.98077392578125, -214.24237060546875, -217.7718505859375, -216.28038024902344], "reading_logprob_std": 1.550793371884606, "tokenization_path_logprob_std": 167.423903315993, "trajectory_balance_rmse": 3016.542192533821} {"canonical_length": 9, "canonical_nll": 0.11619548002878825, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "c5b68b2a56c5748c", "reading_canonical_log_prob": -1.0457593202590942, "reading_distances": [48, 43, 38, 41], "reading_log_probs": [-1.6907453536987305, -1.5679965019226074, -1.5111761093139648, -2.4084486961364746], "reading_logprob_std": 0.4160425684075534, "tokenization_path_logprob_std": 47.96713067976899, "trajectory_balance_rmse": 118.8628860094365} {"canonical_length": 101, "canonical_nll": 1.6801337818107982, "generation_rouge_l_f1": 0.2216358839050132, "generation_token_f1": 0.3799472295514512, "prompt_id": "4478e3a9a8e3ebd2", "reading_canonical_log_prob": -169.69351196289062, "reading_distances": [28, 21, 23, 19], "reading_log_probs": [-165.21365356445312, -167.90533447265625, -168.66143798828125, -166.27166748046875], "reading_logprob_std": 1.5600032168496107, "tokenization_path_logprob_std": 128.85833151596594, "trajectory_balance_rmse": 1239.339323736549} {"canonical_length": 153, "canonical_nll": 1.7376848607281454, "generation_rouge_l_f1": 0.24691358024691354, "generation_token_f1": 0.5185185185185185, "prompt_id": "1679ec014dbdb2b7", "reading_canonical_log_prob": -265.86578369140625, "reading_distances": [25, 25, 20, 28], "reading_log_probs": [-257.9296875, -256.4900207519531, -257.07525634765625, -258.8799743652344], "reading_logprob_std": 1.0415182893597181, "tokenization_path_logprob_std": 154.78165775976777, "trajectory_balance_rmse": 2974.101604001496} {"canonical_length": 6, "canonical_nll": 0.38346298535664874, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "d608b8ecb5eb0aa8", "reading_canonical_log_prob": -2.3007779121398926, "reading_distances": [6, 12, 17, 13], "reading_log_probs": [-2.084320545196533, -1.9651305675506592, -2.1854002475738525, -2.1230690479278564], "reading_logprob_std": 0.09276910290182741, "tokenization_path_logprob_std": 32.66830381512915, "trajectory_balance_rmse": 133.27555298188983} {"canonical_length": 8, "canonical_nll": 0.5341249108314514, "generation_rouge_l_f1": 0.923076923076923, "generation_token_f1": 0.923076923076923, "prompt_id": "c7c5da8841670598", "reading_canonical_log_prob": -4.272999286651611, "reading_distances": [23, 23, 28, 23], "reading_log_probs": [-5.457584857940674, -6.218646049499512, -4.7358293533325195, -3.221681833267212], "reading_logprob_std": 1.2771253547779828, "tokenization_path_logprob_std": 28.03008880843381, "trajectory_balance_rmse": 216.40516718910982} {"canonical_length": 72, "canonical_nll": 1.7188393274943035, "generation_rouge_l_f1": 0.18055555555555558, "generation_token_f1": 0.3055555555555555, "prompt_id": "8b986ced31f57ea9", "reading_canonical_log_prob": -123.75643157958984, "reading_distances": [31, 31, 35, 33], "reading_log_probs": [-122.9246826171875, -120.06498718261719, -121.96443176269531, -120.15867614746094], "reading_logprob_std": 1.4032151221835933, "tokenization_path_logprob_std": 140.24761062420188, "trajectory_balance_rmse": 1445.490670774335} {"canonical_length": 288, "canonical_nll": 1.2756819195217557, "generation_rouge_l_f1": 0.25193798449612403, "generation_token_f1": 0.5193798449612403, "prompt_id": "3d6847202e1bf0e3", "reading_canonical_log_prob": -367.3963928222656, "reading_distances": [24, 20, 25, 23], "reading_log_probs": [-367.2142639160156, -367.9244079589844, -366.7200927734375, -368.7885437011719], "reading_logprob_std": 0.8991876616045601, "tokenization_path_logprob_std": 271.33032974488265, "trajectory_balance_rmse": 5576.007330363882} {"canonical_length": 6, "canonical_nll": 0.8535130023956299, "generation_rouge_l_f1": 0.8000000000000002, "generation_token_f1": 0.8000000000000002, "prompt_id": "81930576a5952ad4", "reading_canonical_log_prob": -5.121078014373779, "reading_distances": [48, 43, 45, 39], "reading_log_probs": [-2.8109002113342285, -2.103595018386841, -3.2444307804107666, -3.4066317081451416], "reading_logprob_std": 0.5822975612686934, "tokenization_path_logprob_std": 20.339996428185373, "trajectory_balance_rmse": 85.17251723040384} {"canonical_length": 271, "canonical_nll": 1.1544065581036669, "generation_rouge_l_f1": 0.2756052141527002, "generation_token_f1": 0.5251396648044693, "prompt_id": "52dc7c85e884efbf", "reading_canonical_log_prob": -312.84417724609375, "reading_distances": [16, 11, 11, 19], "reading_log_probs": [-308.7207946777344, -308.5360412597656, -307.52581787109375, -309.7209777832031], "reading_logprob_std": 0.8993433964868086, "tokenization_path_logprob_std": 117.68416228198036, "trajectory_balance_rmse": 5665.219117168877} {"canonical_length": 220, "canonical_nll": 1.1032584450461649, "generation_rouge_l_f1": 0.255125284738041, "generation_token_f1": 0.45102505694760814, "prompt_id": "02a68089e544097d", "reading_canonical_log_prob": -242.71685791015625, "reading_distances": [27, 21, 23, 25], "reading_log_probs": [-238.73646545410156, -239.19775390625, -236.57589721679688, -238.23226928710938], "reading_logprob_std": 1.1432744422667265, "tokenization_path_logprob_std": 360.1635965888045, "trajectory_balance_rmse": 3856.7831716916908} {"canonical_length": 116, "canonical_nll": 1.674878350619612, "generation_rouge_l_f1": 0.2589285714285714, "generation_token_f1": 0.4375, "prompt_id": "4fcd7cfb63238642", "reading_canonical_log_prob": -194.285888671875, "reading_distances": [28, 26, 36, 35], "reading_log_probs": [-192.61578369140625, -192.4986114501953, -191.6081085205078, -192.8271484375], "reading_logprob_std": 0.5370291720099605, "tokenization_path_logprob_std": 78.86074383822927, "trajectory_balance_rmse": 2861.5076211182877} {"canonical_length": 260, "canonical_nll": 1.5189759474534255, "generation_rouge_l_f1": 0.1141804788213628, "generation_token_f1": 0.19521178637200737, "prompt_id": "0b49b7e13d1bd9b0", "reading_canonical_log_prob": -394.9337463378906, "reading_distances": [35, 29, 27, 24], "reading_log_probs": [-385.8768310546875, -384.32647705078125, -388.3924560546875, -387.608154296875], "reading_logprob_std": 1.8176488224726715, "tokenization_path_logprob_std": 252.79023116278285, "trajectory_balance_rmse": 4816.906526759201} {"canonical_length": 169, "canonical_nll": 1.2702531983866494, "generation_rouge_l_f1": 0.2631578947368421, "generation_token_f1": 0.48026315789473684, "prompt_id": "112cdc2afc089b69", "reading_canonical_log_prob": -214.67279052734375, "reading_distances": [23, 27, 25, 31], "reading_log_probs": [-212.2710723876953, -211.1374969482422, -211.86178588867188, -211.50787353515625], "reading_logprob_std": 0.48494058425276765, "tokenization_path_logprob_std": 87.05465185721036, "trajectory_balance_rmse": 2957.1420136408087} {"canonical_length": 250, "canonical_nll": 1.1495283203125, "generation_rouge_l_f1": 0.23333333333333334, "generation_token_f1": 0.4708333333333333, "prompt_id": "845212ac338f2658", "reading_canonical_log_prob": -287.382080078125, "reading_distances": [41, 42, 28, 40], "reading_log_probs": [-293.8902893066406, -296.3350830078125, -295.295654296875, -294.0539245605469], "reading_logprob_std": 1.147636416659661, "tokenization_path_logprob_std": 203.69886542242114, "trajectory_balance_rmse": 5147.629450586282} {"canonical_length": 30, "canonical_nll": 0.5065646489461263, "generation_rouge_l_f1": 0.32857142857142857, "generation_token_f1": 0.32857142857142857, "prompt_id": "9ca4c748ffa8cd48", "reading_canonical_log_prob": -15.196939468383789, "reading_distances": [46, 53, 45, 52], "reading_log_probs": [-15.218433380126953, -15.642875671386719, -15.951376914978027, -16.106538772583008], "reading_logprob_std": 0.39161209152713894, "tokenization_path_logprob_std": 140.33191015178133, "trajectory_balance_rmse": 677.6532306176513} {"canonical_length": 501, "canonical_nll": 0.5860478749532185, "generation_rouge_l_f1": 0.24850657108721627, "generation_token_f1": 0.40621266427718045, "prompt_id": "585bee96b6b0ece7", "reading_canonical_log_prob": -293.6099853515625, "reading_distances": [19, 22, 22, 17], "reading_log_probs": [-294.51751708984375, -295.0684814453125, -295.12335205078125, -296.2208557128906], "reading_logprob_std": 0.7134103371551955, "tokenization_path_logprob_std": 234.42068257374265, "trajectory_balance_rmse": 8128.765049169473} {"canonical_length": 168, "canonical_nll": 1.1162656148274739, "generation_rouge_l_f1": 0.228310502283105, "generation_token_f1": 0.406392694063927, "prompt_id": "b3cb4c0e99b9a63e", "reading_canonical_log_prob": -187.53262329101562, "reading_distances": [21, 25, 25, 26], "reading_log_probs": [-184.69732666015625, -184.2550506591797, -184.66009521484375, -185.15081787109375], "reading_logprob_std": 0.366278268002153, "tokenization_path_logprob_std": 150.55191612210172, "trajectory_balance_rmse": 3372.1659438863403} {"canonical_length": 430, "canonical_nll": 1.0091176587481832, "generation_rouge_l_f1": 0.22814814814814816, "generation_token_f1": 0.40888888888888886, "prompt_id": "2fa618fc412d053e", "reading_canonical_log_prob": -433.92059326171875, "reading_distances": [27, 29, 26, 25], "reading_log_probs": [-421.4955139160156, -424.5671691894531, -420.6670227050781, -421.08941650390625], "reading_logprob_std": 1.7741359816414544, "tokenization_path_logprob_std": 214.97377305392334, "trajectory_balance_rmse": 6668.8181471561875} {"canonical_length": 160, "canonical_nll": 1.0715887069702148, "generation_rouge_l_f1": 0.2703962703962704, "generation_token_f1": 0.5547785547785548, "prompt_id": "72f894b5c3eed051", "reading_canonical_log_prob": -171.45419311523438, "reading_distances": [27, 26, 28, 25], "reading_log_probs": [-173.0830841064453, -170.76260375976562, -171.22377014160156, -172.0589141845703], "reading_logprob_std": 1.0198513866311707, "tokenization_path_logprob_std": 237.43698651501634, "trajectory_balance_rmse": 3155.6094509570494} {"canonical_length": 98, "canonical_nll": 1.4699926570970185, "generation_rouge_l_f1": 0.2185792349726776, "generation_token_f1": 0.32786885245901637, "prompt_id": "9c00b56f3aa02b5b", "reading_canonical_log_prob": -144.0592803955078, "reading_distances": [37, 31, 42, 34], "reading_log_probs": [-138.3775634765625, -136.54212951660156, -137.32781982421875, -139.46533203125], "reading_logprob_std": 1.2710035609023997, "tokenization_path_logprob_std": 202.2434863908292, "trajectory_balance_rmse": 2072.1213341678945} {"canonical_length": 10, "canonical_nll": 0.2769942283630371, "generation_rouge_l_f1": 0.5333333333333333, "generation_token_f1": 0.5333333333333333, "prompt_id": "363eae7c1280c1a6", "reading_canonical_log_prob": -2.769942283630371, "reading_distances": [65, 53, 54, 43], "reading_log_probs": [-2.918337106704712, -4.183994293212891, -3.819265604019165, -3.2947049140930176], "reading_logprob_std": 0.5593329011289204, "tokenization_path_logprob_std": 60.91479514232366, "trajectory_balance_rmse": 210.52080573346322} {"canonical_length": 7, "canonical_nll": 1.1164043971470423, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.6666666666666666, "prompt_id": "86c615edc3ae2a5f", "reading_canonical_log_prob": -7.814830780029297, "reading_distances": [42, 39, 44, 43], "reading_log_probs": [-9.777790069580078, -9.187902450561523, -8.248485565185547, -9.032373428344727], "reading_logprob_std": 0.6300502084855919, "tokenization_path_logprob_std": 26.82406763814742, "trajectory_balance_rmse": 124.2593900469909} {"canonical_length": 28, "canonical_nll": 0.7688486916678292, "generation_rouge_l_f1": 0.5, "generation_token_f1": 0.6818181818181819, "prompt_id": "b424089c85e11874", "reading_canonical_log_prob": -21.52776336669922, "reading_distances": [23, 24, 20, 16], "reading_log_probs": [-20.57935333251953, -19.98345375061035, -19.93869972229004, -21.545827865600586], "reading_logprob_std": 0.74863670820015, "tokenization_path_logprob_std": 67.60424429475498, "trajectory_balance_rmse": 527.5563978779786} {"canonical_length": 7, "canonical_nll": 0.11484368358339582, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "69e0fa630dc3c290", "reading_canonical_log_prob": -0.8039057850837708, "reading_distances": [41, 32, 25, 23], "reading_log_probs": [-7.985820293426514, -7.377434730529785, -2.397149085998535, -2.1975178718566895], "reading_logprob_std": 3.1195947942932203, "tokenization_path_logprob_std": 14.056738623458696, "trajectory_balance_rmse": 95.06746310018484} {"canonical_length": 158, "canonical_nll": 1.3775028277047072, "generation_rouge_l_f1": 0.2615384615384615, "generation_token_f1": 0.44102564102564107, "prompt_id": "edfb4a427bd0504a", "reading_canonical_log_prob": -217.64544677734375, "reading_distances": [31, 25, 27, 23], "reading_log_probs": [-216.0948028564453, -214.662841796875, -209.65843200683594, -207.6669464111328], "reading_logprob_std": 4.004776636773924, "tokenization_path_logprob_std": 102.2253450360051, "trajectory_balance_rmse": 3072.3052986664375} {"canonical_length": 9, "canonical_nll": 0.39158860842386883, "generation_rouge_l_f1": 0.608695652173913, "generation_token_f1": 0.608695652173913, "prompt_id": "9a5bfd989c2d5e52", "reading_canonical_log_prob": -3.5242974758148193, "reading_distances": [37, 40, 36, 44], "reading_log_probs": [-3.127553939819336, -4.645318031311035, -3.981069564819336, -3.73576283454895], "reading_logprob_std": 0.6278738522637608, "tokenization_path_logprob_std": 38.29635775382983, "trajectory_balance_rmse": 125.75110205561957} {"canonical_length": 272, "canonical_nll": 0.8272898056927849, "generation_rouge_l_f1": 0.4322469982847342, "generation_token_f1": 0.614065180102916, "prompt_id": "a2cdcfce41950902", "reading_canonical_log_prob": -225.0228271484375, "reading_distances": [29, 18, 27, 24], "reading_log_probs": [-228.6250762939453, -226.56007385253906, -229.42388916015625, -227.48287963867188], "reading_logprob_std": 1.2592152607580465, "tokenization_path_logprob_std": 208.19829125051848, "trajectory_balance_rmse": 3527.3594970809613} {"canonical_length": 72, "canonical_nll": 1.360040134853787, "generation_rouge_l_f1": 0.174496644295302, "generation_token_f1": 0.2818791946308725, "prompt_id": "41dfa00e84da9369", "reading_canonical_log_prob": -97.92288970947266, "reading_distances": [44, 37, 39, 41], "reading_log_probs": [-104.15556335449219, -103.66942596435547, -111.90448760986328, -111.86766815185547], "reading_logprob_std": 4.607850981619919, "tokenization_path_logprob_std": 68.9914013180337, "trajectory_balance_rmse": 840.9923160370664} {"canonical_length": 13, "canonical_nll": 0.7804132608266977, "generation_rouge_l_f1": 0.15714285714285714, "generation_token_f1": 0.15714285714285714, "prompt_id": "398d73f7e39625a8", "reading_canonical_log_prob": -10.14537239074707, "reading_distances": [27, 27, 24, 31], "reading_log_probs": [-13.11671257019043, -11.510065078735352, -12.73363971710205, -11.980255126953125], "reading_logprob_std": 0.7248792089394811, "tokenization_path_logprob_std": 32.98056457041947, "trajectory_balance_rmse": 143.0976918571438} {"canonical_length": 89, "canonical_nll": 1.4379935961091117, "generation_rouge_l_f1": 0.3161512027491409, "generation_token_f1": 0.3436426116838488, "prompt_id": "d769bfcaa2a347d7", "reading_canonical_log_prob": -127.98143005371094, "reading_distances": [67, 74, 63, 79], "reading_log_probs": [-130.35781860351562, -132.77713012695312, -129.60101318359375, -128.68568420410156], "reading_logprob_std": 1.753266757108375, "tokenization_path_logprob_std": 106.82493111487075, "trajectory_balance_rmse": 1994.853667386639} {"canonical_length": 351, "canonical_nll": 1.3731452963630697, "generation_rouge_l_f1": 0.197869101978691, "generation_token_f1": 0.3074581430745814, "prompt_id": "026d0e5d489e80b2", "reading_canonical_log_prob": -481.9739990234375, "reading_distances": [7, 12, 10, 9], "reading_log_probs": [-473.92596435546875, -472.8057861328125, -470.7224426269531, -473.05126953125], "reading_logprob_std": 1.3572784890972065, "tokenization_path_logprob_std": 77.36126247880496, "trajectory_balance_rmse": 7260.873537061126} {"canonical_length": 55, "canonical_nll": 1.0118749445134942, "generation_rouge_l_f1": 0.2631578947368421, "generation_token_f1": 0.2982456140350877, "prompt_id": "4c12c24de5859788", "reading_canonical_log_prob": -55.65312194824219, "reading_distances": [20, 25, 24, 27], "reading_log_probs": [-51.163414001464844, -50.623043060302734, -51.403343200683594, -51.047752380371094], "reading_logprob_std": 0.3264245104673313, "tokenization_path_logprob_std": 104.10685246778027, "trajectory_balance_rmse": 1233.8423866120447} {"canonical_length": 298, "canonical_nll": 1.698729035038276, "generation_rouge_l_f1": 0.15384615384615385, "generation_token_f1": 0.32234432234432236, "prompt_id": "1259351bdd54ab2c", "reading_canonical_log_prob": -506.22125244140625, "reading_distances": [18, 17, 17, 15], "reading_log_probs": [-506.8831787109375, -519.0955810546875, -517.7701416015625, -516.15234375], "reading_logprob_std": 5.527377753608091, "tokenization_path_logprob_std": 63.330033004301534, "trajectory_balance_rmse": 5109.358388186007} {"canonical_length": 17, "canonical_nll": 0.9302282894358915, "generation_rouge_l_f1": 0.6, "generation_token_f1": 0.6, "prompt_id": "1e2afb46510ad456", "reading_canonical_log_prob": -15.813880920410156, "reading_distances": [26, 30, 36, 28], "reading_log_probs": [-23.469465255737305, -22.64242172241211, -25.97762680053711, -30.484561920166016], "reading_logprob_std": 3.5251675671641163, "tokenization_path_logprob_std": 64.12175248879636, "trajectory_balance_rmse": 306.0352913299452} {"canonical_length": 23, "canonical_nll": 1.0617152504299, "generation_rouge_l_f1": 0.09859154929577466, "generation_token_f1": 0.11267605633802817, "prompt_id": "dfa52b61eae824fc", "reading_canonical_log_prob": -24.419450759887695, "reading_distances": [14, 16, 20, 16], "reading_log_probs": [-18.98644256591797, -19.383609771728516, -20.673601150512695, -19.2208251953125], "reading_logprob_std": 0.7561038094274943, "tokenization_path_logprob_std": 45.55932030752192, "trajectory_balance_rmse": 273.03086045817986} {"canonical_length": 5, "canonical_nll": 0.38602304458618164, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "30fd5bfb1635c7a5", "reading_canonical_log_prob": -1.9301152229309082, "reading_distances": [19, 19, 28, 23], "reading_log_probs": [-4.880743026733398, -6.160131454467773, -5.053819179534912, -5.495510101318359], "reading_logprob_std": 0.5704872130321409, "tokenization_path_logprob_std": 8.647482086949372, "trajectory_balance_rmse": 10.46219999856223} {"canonical_length": 11, "canonical_nll": 0.1741601120341908, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "bd172f5e24a2f430", "reading_canonical_log_prob": -1.9157612323760986, "reading_distances": [36, 40, 44, 39], "reading_log_probs": [-2.3764593601226807, -2.51995587348938, -1.7138960361480713, -2.4563026428222656], "reading_logprob_std": 0.3731519040774456, "tokenization_path_logprob_std": 24.227262873743403, "trajectory_balance_rmse": 181.57495530661652} {"canonical_length": 8, "canonical_nll": 0.7283336520195007, "generation_rouge_l_f1": 0.06363636363636364, "generation_token_f1": 0.06363636363636364, "prompt_id": "74eb19d7e1933dab", "reading_canonical_log_prob": -5.826669216156006, "reading_distances": [27, 23, 22, 23], "reading_log_probs": [-4.623970031738281, -5.333925724029541, -5.018643379211426, -5.16981315612793], "reading_logprob_std": 0.3037182145256971, "tokenization_path_logprob_std": 36.40011984256034, "trajectory_balance_rmse": 166.5639405819777} {"canonical_length": 2, "canonical_nll": 0.5120294690132141, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "69770982bd8468eb", "reading_canonical_log_prob": -1.0240589380264282, "reading_distances": [27, 24, 27, 30], "reading_log_probs": [-1.1366413831710815, -1.2467031478881836, -1.2861586809158325, -1.05703604221344], "reading_logprob_std": 0.10441645716153297, "tokenization_path_logprob_std": 17.55562615751269, "trajectory_balance_rmse": 24.15937676513566} {"canonical_length": 33, "canonical_nll": 0.6928992415919448, "generation_rouge_l_f1": 0.43636363636363634, "generation_token_f1": 0.5818181818181818, "prompt_id": "decae069198bb200", "reading_canonical_log_prob": -22.86567497253418, "reading_distances": [16, 19, 18, 15], "reading_log_probs": [-24.002946853637695, -24.39689826965332, -24.797945022583008, -25.215621948242188], "reading_logprob_std": 0.5214883880160361, "tokenization_path_logprob_std": 74.24769680974747, "trajectory_balance_rmse": 476.5953398342264} {"canonical_length": 296, "canonical_nll": 1.5007011825973924, "generation_rouge_l_f1": 0.27075812274368233, "generation_token_f1": 0.5667870036101083, "prompt_id": "78378b5209c8e413", "reading_canonical_log_prob": -444.2075500488281, "reading_distances": [15, 20, 20, 20], "reading_log_probs": [-434.64215087890625, -431.9530029296875, -431.4856872558594, -434.50213623046875], "reading_logprob_std": 1.6590612892727918, "tokenization_path_logprob_std": 184.4125733384712, "trajectory_balance_rmse": 5710.2848363906405} {"canonical_length": 46, "canonical_nll": 0.9446693088697351, "generation_rouge_l_f1": 0.15162454873646208, "generation_token_f1": 0.23826714801444043, "prompt_id": "6c87b1ea5d43b53e", "reading_canonical_log_prob": -43.45478820800781, "reading_distances": [22, 18, 15, 22], "reading_log_probs": [-44.038665771484375, -40.88172912597656, -44.253753662109375, -43.561187744140625], "reading_logprob_std": 1.5617883061976043, "tokenization_path_logprob_std": 81.58761614313579, "trajectory_balance_rmse": 634.6384434558079} {"canonical_length": 120, "canonical_nll": 2.1460528055826824, "generation_rouge_l_f1": 0.1932114882506527, "generation_token_f1": 0.35509138381201044, "prompt_id": "c823f34b5de71522", "reading_canonical_log_prob": -257.5263366699219, "reading_distances": [21, 26, 20, 22], "reading_log_probs": [-253.642822265625, -254.45848083496094, -257.0649719238281, -255.0542755126953], "reading_logprob_std": 1.4594597917734058, "tokenization_path_logprob_std": 110.17191937585056, "trajectory_balance_rmse": 2715.9165683016377} {"canonical_length": 23, "canonical_nll": 0.940280582593835, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.8666666666666667, "prompt_id": "95ee759ef85d853e", "reading_canonical_log_prob": -21.626453399658203, "reading_distances": [29, 27, 29, 26], "reading_log_probs": [-23.80484390258789, -22.54125213623047, -24.236053466796875, -23.01446533203125], "reading_logprob_std": 0.7635369438145816, "tokenization_path_logprob_std": 45.11605353951324, "trajectory_balance_rmse": 217.99106980605012} {"canonical_length": 11, "canonical_nll": 0.8789038224653765, "generation_rouge_l_f1": 0.823529411764706, "generation_token_f1": 0.823529411764706, "prompt_id": "036acca12e6542a8", "reading_canonical_log_prob": -9.66794204711914, "reading_distances": [18, 20, 25, 21], "reading_log_probs": [-9.204654693603516, -9.767316818237305, -8.940288543701172, -9.066254615783691], "reading_logprob_std": 0.36480178855091455, "tokenization_path_logprob_std": 28.157458144645993, "trajectory_balance_rmse": 192.43465139789606} {"canonical_length": 13, "canonical_nll": 1.2732011354886568, "generation_rouge_l_f1": 0.5555555555555556, "generation_token_f1": 0.5555555555555556, "prompt_id": "5e5a1ac1eccda1ac", "reading_canonical_log_prob": -16.55161476135254, "reading_distances": [20, 36, 27, 27], "reading_log_probs": [-15.632772445678711, -15.187163352966309, -14.358865737915039, -15.553766250610352], "reading_logprob_std": 0.5828028255302544, "tokenization_path_logprob_std": 103.66574074184302, "trajectory_balance_rmse": 362.07527861443066} {"canonical_length": 8, "canonical_nll": 0.05620970577001572, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "05b4960bf40b9d90", "reading_canonical_log_prob": -0.44967764616012573, "reading_distances": [24, 30, 24, 25], "reading_log_probs": [-0.9784222841262817, -0.9011425375938416, -0.946128249168396, -0.7024587392807007], "reading_logprob_std": 0.12384289039230104, "tokenization_path_logprob_std": 38.096559439545416, "trajectory_balance_rmse": 143.74751580801495} {"canonical_length": 199, "canonical_nll": 1.2348480224609375, "generation_rouge_l_f1": 0.2652173913043478, "generation_token_f1": 0.5565217391304347, "prompt_id": "55f01821168c15c2", "reading_canonical_log_prob": -245.73475646972656, "reading_distances": [29, 33, 37, 34], "reading_log_probs": [-243.60589599609375, -242.13046264648438, -242.45291137695312, -243.30178833007812], "reading_logprob_std": 0.6949417059496142, "tokenization_path_logprob_std": 87.39499499373949, "trajectory_balance_rmse": 3767.767526851202} {"canonical_length": 181, "canonical_nll": 0.9319345885218836, "generation_rouge_l_f1": 0.2942528735632184, "generation_token_f1": 0.4689655172413793, "prompt_id": "28aed3c6f636dfb4", "reading_canonical_log_prob": -168.68016052246094, "reading_distances": [13, 12, 11, 15], "reading_log_probs": [-167.04977416992188, -167.72802734375, -167.60008239746094, -166.8643798828125], "reading_logprob_std": 0.4184049664024316, "tokenization_path_logprob_std": 109.35653219430976, "trajectory_balance_rmse": 3056.98132227823} {"canonical_length": 102, "canonical_nll": 1.966592227711397, "generation_rouge_l_f1": 0.23404255319148937, "generation_token_f1": 0.4397163120567376, "prompt_id": "33e9d73a4126f71a", "reading_canonical_log_prob": -200.5924072265625, "reading_distances": [18, 20, 12, 20], "reading_log_probs": [-195.61959838867188, -196.0882110595703, -198.56930541992188, -197.23085021972656], "reading_logprob_std": 1.3155929482819437, "tokenization_path_logprob_std": 218.61898476190265, "trajectory_balance_rmse": 2423.093369148674} {"canonical_length": 169, "canonical_nll": 1.9361976758968196, "generation_rouge_l_f1": 0.2382022471910112, "generation_token_f1": 0.42696629213483145, "prompt_id": "ba9c16c03160cc50", "reading_canonical_log_prob": -327.2174072265625, "reading_distances": [23, 21, 23, 16], "reading_log_probs": [-321.6386413574219, -322.685791015625, -320.1856689453125, -320.63555908203125], "reading_logprob_std": 1.1131894032963998, "tokenization_path_logprob_std": 256.4766192195963, "trajectory_balance_rmse": 3116.823172261451} {"canonical_length": 278, "canonical_nll": 0.6843491149463242, "generation_rouge_l_f1": 0.22823984526112184, "generation_token_f1": 0.36750483558994196, "prompt_id": "d04a82b0b7e2a434", "reading_canonical_log_prob": -190.24905395507812, "reading_distances": [23, 27, 24, 28], "reading_log_probs": [-191.13394165039062, -193.78347778320312, -196.14361572265625, -194.51611328125], "reading_logprob_std": 2.087896290388709, "tokenization_path_logprob_std": 119.73895916449084, "trajectory_balance_rmse": 3242.3903146244797} {"canonical_length": 38, "canonical_nll": 0.7527858332583779, "generation_rouge_l_f1": 0.42857142857142855, "generation_token_f1": 0.5357142857142857, "prompt_id": "e1a492712f46f15d", "reading_canonical_log_prob": -28.60586166381836, "reading_distances": [19, 21, 19, 17], "reading_log_probs": [-28.388099670410156, -29.362445831298828, -28.619041442871094, -28.75962257385254], "reading_logprob_std": 0.41598505883905074, "tokenization_path_logprob_std": 79.20250149241758, "trajectory_balance_rmse": 562.6952105087378} {"canonical_length": 25, "canonical_nll": 0.9377142333984375, "generation_rouge_l_f1": 0.13745704467353953, "generation_token_f1": 0.15120274914089346, "prompt_id": "fcc6cd8702ae37ce", "reading_canonical_log_prob": -23.442855834960938, "reading_distances": [23, 26, 26, 22], "reading_log_probs": [-23.762405395507812, -24.402647018432617, -24.755638122558594, -24.750755310058594], "reading_logprob_std": 0.4671778822132777, "tokenization_path_logprob_std": 23.08415030966486, "trajectory_balance_rmse": 303.07140918714344} {"canonical_length": 342, "canonical_nll": 1.2124011837251005, "generation_rouge_l_f1": 0.18985270049099834, "generation_token_f1": 0.3502454991816694, "prompt_id": "38014139c1d9865d", "reading_canonical_log_prob": -414.6412048339844, "reading_distances": [32, 30, 25, 38], "reading_log_probs": [-419.85455322265625, -422.7305908203125, -421.12506103515625, -421.658447265625], "reading_logprob_std": 1.1955310760895308, "tokenization_path_logprob_std": 202.95840228306756, "trajectory_balance_rmse": 5736.496328375725} {"canonical_length": 6, "canonical_nll": 0.5416318575541178, "generation_rouge_l_f1": 0.6666666666666665, "generation_token_f1": 0.6666666666666665, "prompt_id": "bd70ea68bde60a24", "reading_canonical_log_prob": -3.249791145324707, "reading_distances": [25, 25, 19, 25], "reading_log_probs": [-3.7779784202575684, -4.09656286239624, -3.549164295196533, -3.625831365585327], "reading_logprob_std": 0.24223126862391736, "tokenization_path_logprob_std": 15.843662037845391, "trajectory_balance_rmse": 103.45666502243348} {"canonical_length": 11, "canonical_nll": 2.021595001220703, "generation_rouge_l_f1": 0.45454545454545453, "generation_token_f1": 0.45454545454545453, "prompt_id": "57765c15fcf7839d", "reading_canonical_log_prob": -22.237545013427734, "reading_distances": [10, 12, 12, 14], "reading_log_probs": [-20.829797744750977, -20.983692169189453, -20.7364501953125, -21.967945098876953], "reading_logprob_std": 0.568201752557546, "tokenization_path_logprob_std": 45.788461586098954, "trajectory_balance_rmse": 222.03966903369928} {"canonical_length": 50, "canonical_nll": 1.271132354736328, "generation_rouge_l_f1": 0.28571428571428575, "generation_token_f1": 0.3214285714285714, "prompt_id": "253987951dace9e0", "reading_canonical_log_prob": -63.556617736816406, "reading_distances": [32, 23, 29, 26], "reading_log_probs": [-63.68841552734375, -62.33833312988281, -62.89625930786133, -63.19326400756836], "reading_logprob_std": 0.5646392064020759, "tokenization_path_logprob_std": 130.3875988516492, "trajectory_balance_rmse": 1098.8031514872043} {"canonical_length": 38, "canonical_nll": 1.1979293823242188, "generation_rouge_l_f1": 0.21969696969696967, "generation_token_f1": 0.25, "prompt_id": "e17273b705c2c377", "reading_canonical_log_prob": -45.52131652832031, "reading_distances": [29, 22, 28, 29], "reading_log_probs": [-44.799278259277344, -43.79157257080078, -43.366817474365234, -44.28924560546875], "reading_logprob_std": 0.6195776088920578, "tokenization_path_logprob_std": 91.79328602490396, "trajectory_balance_rmse": 766.8796514033381} {"canonical_length": 344, "canonical_nll": 1.596207019894622, "generation_rouge_l_f1": 0.23688663282571915, "generation_token_f1": 0.46362098138747887, "prompt_id": "3e31d044adc4af42", "reading_canonical_log_prob": -549.09521484375, "reading_distances": [23, 27, 31, 26], "reading_log_probs": [-541.18896484375, -546.50732421875, -545.539306640625, -543.0997314453125], "reading_logprob_std": 2.404192915021262, "tokenization_path_logprob_std": 177.00379098430727, "trajectory_balance_rmse": 7764.5382769853695} {"canonical_length": 23, "canonical_nll": 0.8065088520879331, "generation_rouge_l_f1": 0.43243243243243246, "generation_token_f1": 0.43243243243243246, "prompt_id": "31f5a769dd1b152b", "reading_canonical_log_prob": -18.54970359802246, "reading_distances": [24, 19, 24, 19], "reading_log_probs": [-29.564090728759766, -22.484783172607422, -29.93149757385254, -33.112335205078125], "reading_logprob_std": 4.484768367680253, "tokenization_path_logprob_std": 37.40608359923424, "trajectory_balance_rmse": 470.09405841901497} {"canonical_length": 22, "canonical_nll": 0.986086065118963, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.7333333333333333, "prompt_id": "74fb912570f58841", "reading_canonical_log_prob": -21.693893432617188, "reading_distances": [26, 24, 17, 29], "reading_log_probs": [-21.48095703125, -21.712467193603516, -21.05498504638672, -21.514972686767578], "reading_logprob_std": 0.27674922996338763, "tokenization_path_logprob_std": 38.120724464611335, "trajectory_balance_rmse": 227.4761744475657} {"canonical_length": 10, "canonical_nll": 0.4615880012512207, "generation_rouge_l_f1": 0.11612903225806451, "generation_token_f1": 0.11612903225806451, "prompt_id": "434a6990c55f121a", "reading_canonical_log_prob": -4.615880012512207, "reading_distances": [14, 20, 19, 15], "reading_log_probs": [-5.074979782104492, -4.612740516662598, -4.812633991241455, -4.7469401359558105], "reading_logprob_std": 0.19416124818586347, "tokenization_path_logprob_std": 47.15030091781882, "trajectory_balance_rmse": 255.7822348490899} {"canonical_length": 222, "canonical_nll": 1.1214162465688344, "generation_rouge_l_f1": 0.2289156626506024, "generation_token_f1": 0.4779116465863454, "prompt_id": "2e39c9a4bb908fa5", "reading_canonical_log_prob": -248.95440673828125, "reading_distances": [16, 23, 18, 23], "reading_log_probs": [-242.63558959960938, -241.81887817382812, -243.37474060058594, -242.66165161132812], "reading_logprob_std": 0.635971163409037, "tokenization_path_logprob_std": 105.3505730502942, "trajectory_balance_rmse": 5315.698727273289} {"canonical_length": 226, "canonical_nll": 1.4074631412472345, "generation_rouge_l_f1": 0.19172113289760348, "generation_token_f1": 0.39215686274509803, "prompt_id": "1a576b185bace9d2", "reading_canonical_log_prob": -318.086669921875, "reading_distances": [10, 16, 12, 16], "reading_log_probs": [-311.46868896484375, -307.5743408203125, -306.19873046875, -312.57012939453125], "reading_logprob_std": 3.0495424879614026, "tokenization_path_logprob_std": 224.99244237322148, "trajectory_balance_rmse": 4061.945813182981} {"canonical_length": 270, "canonical_nll": 1.2257021303530093, "generation_rouge_l_f1": 0.24344569288389514, "generation_token_f1": 0.49812734082397003, "prompt_id": "c2df8f1e4778ecd3", "reading_canonical_log_prob": -330.9395751953125, "reading_distances": [35, 28, 35, 35], "reading_log_probs": [-328.90972900390625, -329.15789794921875, -328.0093994140625, -331.67041015625], "reading_logprob_std": 1.5686593997960252, "tokenization_path_logprob_std": 191.13066878788578, "trajectory_balance_rmse": 5483.5144789713095} {"canonical_length": 2, "canonical_nll": 0.08440311998128891, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "42c831675952234b", "reading_canonical_log_prob": -0.16880623996257782, "reading_distances": [68, 67, 67, 49], "reading_log_probs": [-1.2818557024002075, -1.200981616973877, -1.3155245780944824, -1.195085048675537], "reading_logprob_std": 0.05976613933944198, "tokenization_path_logprob_std": 6.773033946398836, "trajectory_balance_rmse": 22.28982074238248} {"canonical_length": 16, "canonical_nll": 2.721146821975708, "generation_rouge_l_f1": 0.07692307692307691, "generation_token_f1": 0.15384615384615383, "prompt_id": "a9f4c15e6cb406fd", "reading_canonical_log_prob": -43.53834915161133, "reading_distances": [21, 18, 20, 19], "reading_log_probs": [-43.54096984863281, -43.568359375, -43.59728240966797, -43.07426834106445], "reading_logprob_std": 0.24836763670659698, "tokenization_path_logprob_std": 60.89841750687011, "trajectory_balance_rmse": 440.9838857495545} {"canonical_length": 161, "canonical_nll": 1.732330890916149, "generation_rouge_l_f1": 0.17647058823529413, "generation_token_f1": 0.3186274509803922, "prompt_id": "25e2da0cacffb5d9", "reading_canonical_log_prob": -278.9052734375, "reading_distances": [23, 21, 19, 23], "reading_log_probs": [-275.09161376953125, -274.5348205566406, -275.04974365234375, -273.9009704589844], "reading_logprob_std": 0.5564766296495003, "tokenization_path_logprob_std": 210.64495035193778, "trajectory_balance_rmse": 3350.5601356224056} {"canonical_length": 116, "canonical_nll": 1.1864785819218076, "generation_rouge_l_f1": 0.26809651474530827, "generation_token_f1": 0.38605898123324395, "prompt_id": "d71e3178e4fc08e6", "reading_canonical_log_prob": -137.6315155029297, "reading_distances": [36, 35, 42, 35], "reading_log_probs": [-133.86685180664062, -133.84585571289062, -133.28396606445312, -134.07586669921875], "reading_logprob_std": 0.33906925250268205, "tokenization_path_logprob_std": 90.02355048378257, "trajectory_balance_rmse": 2841.216159999138} {"canonical_length": 9, "canonical_nll": 0.9277390374077691, "generation_rouge_l_f1": 0.16438356164383564, "generation_token_f1": 0.16438356164383564, "prompt_id": "c69244953386b7ea", "reading_canonical_log_prob": -8.349651336669922, "reading_distances": [16, 19, 15, 5], "reading_log_probs": [-9.154403686523438, -8.7122802734375, -8.874781608581543, -8.028200149536133], "reading_logprob_std": 0.4789805573849857, "tokenization_path_logprob_std": 20.305712216406587, "trajectory_balance_rmse": 221.52974991712293} {"canonical_length": 16, "canonical_nll": 2.284400701522827, "generation_rouge_l_f1": 0.4444444444444445, "generation_token_f1": 0.4444444444444445, "prompt_id": "44b99fb81cd83d68", "reading_canonical_log_prob": -36.550411224365234, "reading_distances": [25, 21, 24, 30], "reading_log_probs": [-43.21766662597656, -39.262451171875, -38.298057556152344, -39.24815368652344], "reading_logprob_std": 2.1877739406724674, "tokenization_path_logprob_std": 74.74932922998379, "trajectory_balance_rmse": 398.0007397140347} {"canonical_length": 23, "canonical_nll": 0.2915528131567914, "generation_rouge_l_f1": 0.07272727272727274, "generation_token_f1": 0.1090909090909091, "prompt_id": "f3f998d8d70ee428", "reading_canonical_log_prob": -6.705714702606201, "reading_distances": [51, 41, 40, 42], "reading_log_probs": [-8.768011093139648, -9.591313362121582, -8.373916625976562, -8.925651550292969], "reading_logprob_std": 0.507244145399497, "tokenization_path_logprob_std": 20.74114865130012, "trajectory_balance_rmse": 290.1343297503314} {"canonical_length": 45, "canonical_nll": 0.7577724880642361, "generation_rouge_l_f1": 0.3310344827586207, "generation_token_f1": 0.4413793103448276, "prompt_id": "8565123aabdca8e0", "reading_canonical_log_prob": -34.099761962890625, "reading_distances": [21, 22, 24, 22], "reading_log_probs": [-36.227745056152344, -36.19975280761719, -35.93238830566406, -36.67306137084961], "reading_logprob_std": 0.3069238720211876, "tokenization_path_logprob_std": 56.57334755684088, "trajectory_balance_rmse": 945.2599610093673} {"canonical_length": 293, "canonical_nll": 1.6726956416315593, "generation_rouge_l_f1": 0.20793950850661627, "generation_token_f1": 0.4612476370510397, "prompt_id": "c754674f366c507b", "reading_canonical_log_prob": -490.0998229980469, "reading_distances": [27, 30, 36, 32], "reading_log_probs": [-480.6893615722656, -481.2672119140625, -482.287109375, -484.14398193359375], "reading_logprob_std": 1.5161652057773707, "tokenization_path_logprob_std": 95.04529558222261, "trajectory_balance_rmse": 6014.636086555779} {"canonical_length": 15, "canonical_nll": 0.8749835968017579, "generation_rouge_l_f1": 0.12, "generation_token_f1": 0.16, "prompt_id": "f7c03b54285f64b4", "reading_canonical_log_prob": -13.124753952026367, "reading_distances": [26, 26, 29, 25], "reading_log_probs": [-16.201669692993164, -18.66817283630371, -16.554637908935547, -15.244844436645508], "reading_logprob_std": 1.444111752652135, "tokenization_path_logprob_std": 78.68724624213374, "trajectory_balance_rmse": 264.8200650770313} {"canonical_length": 33, "canonical_nll": 0.33666662736372516, "generation_rouge_l_f1": 0.14705882352941177, "generation_token_f1": 0.16176470588235292, "prompt_id": "278cabb1c777cfb5", "reading_canonical_log_prob": -11.10999870300293, "reading_distances": [58, 44, 50, 56], "reading_log_probs": [-18.283784866333008, -17.47732925415039, -15.839786529541016, -18.215858459472656], "reading_logprob_std": 1.1365445401030239, "tokenization_path_logprob_std": 71.43060868714151, "trajectory_balance_rmse": 432.94587520815253} {"canonical_length": 82, "canonical_nll": 0.4957418674375953, "generation_rouge_l_f1": 0.396551724137931, "generation_token_f1": 0.4310344827586207, "prompt_id": "26496139eab8348f", "reading_canonical_log_prob": -40.65083312988281, "reading_distances": [17, 20, 18, 17], "reading_log_probs": [-40.81310272216797, -41.82439041137695, -40.80095291137695, -41.4398078918457], "reading_logprob_std": 0.501586939519103, "tokenization_path_logprob_std": 17.661015079839917, "trajectory_balance_rmse": 827.6748074573678} {"canonical_length": 130, "canonical_nll": 1.776717083270733, "generation_rouge_l_f1": 0.1818181818181818, "generation_token_f1": 0.3046683046683047, "prompt_id": "5e25b104005e6f72", "reading_canonical_log_prob": -230.9732208251953, "reading_distances": [25, 23, 23, 19], "reading_log_probs": [-231.45947265625, -228.47354125976562, -221.25318908691406, -218.3441162109375], "reading_logprob_std": 6.112132931087283, "tokenization_path_logprob_std": 247.5822141620813, "trajectory_balance_rmse": 1698.2731821451525} {"canonical_length": 107, "canonical_nll": 2.4201138220100757, "generation_rouge_l_f1": 0.18181818181818182, "generation_token_f1": 0.3409090909090909, "prompt_id": "14f8273006e150ca", "reading_canonical_log_prob": -258.9521789550781, "reading_distances": [25, 25, 34, 30], "reading_log_probs": [-262.4158935546875, -260.07867431640625, -261.4510498046875, -261.56103515625], "reading_logprob_std": 0.9668335356970269, "tokenization_path_logprob_std": 85.8302289290649, "trajectory_balance_rmse": 1946.1529901439912} {"canonical_length": 185, "canonical_nll": 1.5776649268897804, "generation_rouge_l_f1": 0.33663366336633666, "generation_token_f1": 0.46864686468646866, "prompt_id": "c0e6a5cad32d62b6", "reading_canonical_log_prob": -291.8680114746094, "reading_distances": [13, 17, 17, 12], "reading_log_probs": [-291.158935546875, -292.4312438964844, -293.11468505859375, -293.4001159667969], "reading_logprob_std": 0.9980769040914698, "tokenization_path_logprob_std": 198.14142311417163, "trajectory_balance_rmse": 3225.6033483051733} {"canonical_length": 16, "canonical_nll": 1.1023547649383545, "generation_rouge_l_f1": 0.4, "generation_token_f1": 0.5142857142857143, "prompt_id": "937bfa51bfbd0542", "reading_canonical_log_prob": -17.637676239013672, "reading_distances": [31, 34, 37, 39], "reading_log_probs": [-19.697582244873047, -18.692058563232422, -18.246824264526367, -17.943723678588867], "reading_logprob_std": 0.7660453156701688, "tokenization_path_logprob_std": 32.011280416719046, "trajectory_balance_rmse": 410.5773546706067} {"canonical_length": 75, "canonical_nll": 0.4197184499104818, "generation_rouge_l_f1": 0.23870967741935484, "generation_token_f1": 0.2967741935483871, "prompt_id": "19b52394b9bc5537", "reading_canonical_log_prob": -31.478883743286133, "reading_distances": [13, 13, 17, 12], "reading_log_probs": [-30.694019317626953, -30.27313804626465, -30.558565139770508, -30.799560546875], "reading_logprob_std": 0.22790639362375334, "tokenization_path_logprob_std": 57.056255756779294, "trajectory_balance_rmse": 566.6478875905483} {"canonical_length": 339, "canonical_nll": 1.413351143355918, "generation_rouge_l_f1": 0.2105263157894737, "generation_token_f1": 0.39564428312159705, "prompt_id": "c03437022c1ce69c", "reading_canonical_log_prob": -479.12603759765625, "reading_distances": [29, 29, 24, 28], "reading_log_probs": [-474.66375732421875, -474.5805358886719, -479.37445068359375, -472.06591796875], "reading_logprob_std": 3.0504903955255145, "tokenization_path_logprob_std": 164.13938668903742, "trajectory_balance_rmse": 5660.25981837899} {"canonical_length": 22, "canonical_nll": 0.634553995999423, "generation_rouge_l_f1": 0.3829787234042553, "generation_token_f1": 0.3829787234042553, "prompt_id": "01ef5d8df6df3761", "reading_canonical_log_prob": -13.960187911987305, "reading_distances": [21, 20, 21, 26], "reading_log_probs": [-13.674056053161621, -14.014665603637695, -14.079109191894531, -13.239121437072754], "reading_logprob_std": 0.38518958426945665, "tokenization_path_logprob_std": 43.484934025425105, "trajectory_balance_rmse": 334.32232704929015} {"canonical_length": 110, "canonical_nll": 1.5560293024236507, "generation_rouge_l_f1": 0.18229166666666666, "generation_token_f1": 0.3020833333333333, "prompt_id": "1275798cfa3be44f", "reading_canonical_log_prob": -171.16322326660156, "reading_distances": [36, 34, 37, 38], "reading_log_probs": [-166.80975341796875, -167.87091064453125, -171.83741760253906, -168.0818328857422], "reading_logprob_std": 2.1966590197659213, "tokenization_path_logprob_std": 93.91169755686161, "trajectory_balance_rmse": 2478.0643749091996} {"canonical_length": 4, "canonical_nll": 0.26717281341552734, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "1f2fdcd7eec76cc5", "reading_canonical_log_prob": -1.0686912536621094, "reading_distances": [29, 29, 25, 33], "reading_log_probs": [-6.038290023803711, -1.7632970809936523, -2.436235189437866, -3.695237874984741], "reading_logprob_std": 1.882165989722944, "tokenization_path_logprob_std": 29.62337931806994, "trajectory_balance_rmse": 50.36777356336583} {"canonical_length": 128, "canonical_nll": 1.2076599597930908, "generation_rouge_l_f1": 0.18857142857142858, "generation_token_f1": 0.2514285714285714, "prompt_id": "b8b97c6b0ef87a98", "reading_canonical_log_prob": -154.58047485351562, "reading_distances": [32, 28, 24, 26], "reading_log_probs": [-146.80178833007812, -150.42376708984375, -148.95738220214844, -146.91598510742188], "reading_logprob_std": 1.7416581456301659, "tokenization_path_logprob_std": 60.535091859800445, "trajectory_balance_rmse": 1625.1524673955732} {"canonical_length": 168, "canonical_nll": 1.063151586623419, "generation_rouge_l_f1": 0.3037249283667622, "generation_token_f1": 0.5329512893982808, "prompt_id": "7ec56cfda55797c0", "reading_canonical_log_prob": -178.60946655273438, "reading_distances": [27, 35, 24, 27], "reading_log_probs": [-175.2755889892578, -176.3303985595703, -179.032470703125, -177.76382446289062], "reading_logprob_std": 1.6427483623996781, "tokenization_path_logprob_std": 54.44057822128664, "trajectory_balance_rmse": 3194.926553542849} {"canonical_length": 32, "canonical_nll": 1.4820283651351929, "generation_rouge_l_f1": 0.3529411764705882, "generation_token_f1": 0.4313725490196078, "prompt_id": "c6a2889afcb5222e", "reading_canonical_log_prob": -47.42490768432617, "reading_distances": [15, 16, 13, 9], "reading_log_probs": [-48.239349365234375, -49.25754928588867, -49.928558349609375, -48.46112060546875], "reading_logprob_std": 0.7733728021803805, "tokenization_path_logprob_std": 41.77500662840144, "trajectory_balance_rmse": 487.291570496877} {"canonical_length": 32, "canonical_nll": 1.7023060321807861, "generation_rouge_l_f1": 0.22580645161290322, "generation_token_f1": 0.2580645161290323, "prompt_id": "5371fa87b29b339c", "reading_canonical_log_prob": -54.473793029785156, "reading_distances": [12, 17, 19, 16], "reading_log_probs": [-55.06614685058594, -55.465431213378906, -55.60004425048828, -55.80701446533203], "reading_logprob_std": 0.3123827401575652, "tokenization_path_logprob_std": 103.80894927483413, "trajectory_balance_rmse": 748.411584714062} {"canonical_length": 63, "canonical_nll": 1.2492328219943576, "generation_rouge_l_f1": 0.39370078740157477, "generation_token_f1": 0.5354330708661418, "prompt_id": "c89f55288eb64d0c", "reading_canonical_log_prob": -78.70166778564453, "reading_distances": [22, 29, 25, 31], "reading_log_probs": [-82.07708740234375, -81.32969665527344, -80.94436645507812, -81.17529296875], "reading_logprob_std": 0.4899430060351035, "tokenization_path_logprob_std": 54.416670250779944, "trajectory_balance_rmse": 1474.9553099612137} {"canonical_length": 221, "canonical_nll": 1.410556430730345, "generation_rouge_l_f1": 0.3006263048016702, "generation_token_f1": 0.5010438413361169, "prompt_id": "0e1e524655779cac", "reading_canonical_log_prob": -311.73297119140625, "reading_distances": [18, 24, 18, 17], "reading_log_probs": [-310.16546630859375, -307.5260314941406, -306.1811218261719, -307.3338317871094], "reading_logprob_std": 1.6840980314409189, "tokenization_path_logprob_std": 143.31276036428932, "trajectory_balance_rmse": 4507.86143328391} {"canonical_length": 9, "canonical_nll": 0.29438016149732804, "generation_rouge_l_f1": 0.9333333333333333, "generation_token_f1": 0.9333333333333333, "prompt_id": "5f85bd9588d13639", "reading_canonical_log_prob": -2.649421453475952, "reading_distances": [21, 13, 20, 14], "reading_log_probs": [-2.634016990661621, -2.408576011657715, -2.318901538848877, -2.3612236976623535], "reading_logprob_std": 0.14041975215155963, "tokenization_path_logprob_std": 37.34916213878021, "trajectory_balance_rmse": 204.58374306425287} {"canonical_length": 26, "canonical_nll": 0.9198737511268029, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "570dbf2f8a32567a", "reading_canonical_log_prob": -23.916717529296875, "reading_distances": [20, 14, 21, 23], "reading_log_probs": [-23.935565948486328, -24.033432006835938, -22.872285842895508, -22.271778106689453], "reading_logprob_std": 0.8524780911168839, "tokenization_path_logprob_std": 101.16587512570419, "trajectory_balance_rmse": 511.0211147803748} {"canonical_length": 26, "canonical_nll": 0.9961905846228967, "generation_rouge_l_f1": 0.13953488372093026, "generation_token_f1": 0.14617940199335547, "prompt_id": "b4dad9a8bbd784f6", "reading_canonical_log_prob": -25.900955200195312, "reading_distances": [40, 29, 38, 37], "reading_log_probs": [-25.56890106201172, -26.48453140258789, -25.473203659057617, -25.246421813964844], "reading_logprob_std": 0.5445685557144238, "tokenization_path_logprob_std": 44.44294265000086, "trajectory_balance_rmse": 590.0873904489132} {"canonical_length": 10, "canonical_nll": 1.7797101974487304, "generation_rouge_l_f1": 0.26666666666666666, "generation_token_f1": 0.39999999999999997, "prompt_id": "9e76d45f4e1b2541", "reading_canonical_log_prob": -17.797101974487305, "reading_distances": [28, 41, 31, 40], "reading_log_probs": [-17.189395904541016, -16.90312957763672, -17.28247833251953, -18.1962890625], "reading_logprob_std": 0.5594389720375161, "tokenization_path_logprob_std": 33.03270806243806, "trajectory_balance_rmse": 310.5820483797201} {"canonical_length": 256, "canonical_nll": 1.4008870124816895, "generation_rouge_l_f1": 0.24230769230769228, "generation_token_f1": 0.48461538461538456, "prompt_id": "6a9973c4375e58ef", "reading_canonical_log_prob": -358.6270751953125, "reading_distances": [32, 35, 29, 33], "reading_log_probs": [-348.552978515625, -348.194580078125, -353.41961669921875, -347.31866455078125], "reading_logprob_std": 2.7481205764752548, "tokenization_path_logprob_std": 290.6392409087276, "trajectory_balance_rmse": 5748.649654048954} {"canonical_length": 277, "canonical_nll": 1.0830576101365073, "generation_rouge_l_f1": 0.24482109227871937, "generation_token_f1": 0.527306967984934, "prompt_id": "4a7ad2175af88b68", "reading_canonical_log_prob": -300.0069580078125, "reading_distances": [30, 31, 28, 36], "reading_log_probs": [-305.9254150390625, -302.12567138671875, -305.39813232421875, -305.7124938964844], "reading_logprob_std": 1.7896585754457577, "tokenization_path_logprob_std": 300.58386281028504, "trajectory_balance_rmse": 5222.7695218994995} {"canonical_length": 24, "canonical_nll": 0.1775296926498413, "generation_rouge_l_f1": 0.4827586206896552, "generation_token_f1": 0.4827586206896552, "prompt_id": "7271cafab8ca2bf4", "reading_canonical_log_prob": -4.260712623596191, "reading_distances": [28, 34, 34, 26], "reading_log_probs": [-6.021023750305176, -6.958114147186279, -10.205879211425781, -8.88692855834961], "reading_logprob_std": 1.8844207284473824, "tokenization_path_logprob_std": 78.66983048043217, "trajectory_balance_rmse": 491.0727938818214} {"canonical_length": 11, "canonical_nll": 0.5213183489712802, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "7798c57a62239065", "reading_canonical_log_prob": -5.734501838684082, "reading_distances": [14, 14, 17, 15], "reading_log_probs": [-6.727231979370117, -6.864659309387207, -6.284961700439453, -6.445993423461914], "reading_logprob_std": 0.26312957100816653, "tokenization_path_logprob_std": 50.341434429314006, "trajectory_balance_rmse": 240.468092031072} {"canonical_length": 97, "canonical_nll": 1.2076825014094716, "generation_rouge_l_f1": 0.2670299727520436, "generation_token_f1": 0.3978201634877384, "prompt_id": "fd06dfbeb5280c45", "reading_canonical_log_prob": -117.14520263671875, "reading_distances": [39, 44, 39, 45], "reading_log_probs": [-115.93916320800781, -115.00593566894531, -113.66702270507812, -114.09333801269531], "reading_logprob_std": 1.0102752885922561, "tokenization_path_logprob_std": 252.1504042212195, "trajectory_balance_rmse": 1969.083409691522} {"canonical_length": 210, "canonical_nll": 1.0665937151227678, "generation_rouge_l_f1": 0.2673031026252984, "generation_token_f1": 0.5011933174224343, "prompt_id": "3efd48165ac7ab7c", "reading_canonical_log_prob": -223.98468017578125, "reading_distances": [16, 19, 17, 19], "reading_log_probs": [-223.0159912109375, -233.3258819580078, -224.05194091796875, -226.6521759033203], "reading_logprob_std": 4.635860537040587, "tokenization_path_logprob_std": 116.1409252737791, "trajectory_balance_rmse": 2143.1391134160217} {"canonical_length": 69, "canonical_nll": 1.3911440199699954, "generation_rouge_l_f1": 0.36220472440944884, "generation_token_f1": 0.4409448818897638, "prompt_id": "18e11e5d4d3c48a7", "reading_canonical_log_prob": -95.98893737792969, "reading_distances": [32, 40, 35, 39], "reading_log_probs": [-98.43429565429688, -100.90106201171875, -98.35733032226562, -95.65235900878906], "reading_logprob_std": 2.1441073385693756, "tokenization_path_logprob_std": 100.59671380141694, "trajectory_balance_rmse": 1765.598521587639} {"canonical_length": 17, "canonical_nll": 0.30826882755055146, "generation_rouge_l_f1": 0.7692307692307693, "generation_token_f1": 1.0, "prompt_id": "b05a564949c24a61", "reading_canonical_log_prob": -5.240570068359375, "reading_distances": [70, 64, 51, 67], "reading_log_probs": [-6.014692306518555, -5.891101837158203, -4.725162506103516, -5.869242191314697], "reading_logprob_std": 0.6033329746926339, "tokenization_path_logprob_std": 52.00950771464863, "trajectory_balance_rmse": 140.66013986000726} {"canonical_length": 129, "canonical_nll": 0.711096120435138, "generation_rouge_l_f1": 0.424, "generation_token_f1": 0.568, "prompt_id": "977aa631a6e3fb74", "reading_canonical_log_prob": -91.73139953613281, "reading_distances": [9, 15, 14, 11], "reading_log_probs": [-92.69910430908203, -91.55966186523438, -90.94891357421875, -91.7150650024414], "reading_logprob_std": 0.7253724140775587, "tokenization_path_logprob_std": 87.48746692935767, "trajectory_balance_rmse": 1313.4756988791505} {"canonical_length": 85, "canonical_nll": 1.189377010569853, "generation_rouge_l_f1": 0.20057306590257878, "generation_token_f1": 0.32091690544412604, "prompt_id": "308d6723913890bf", "reading_canonical_log_prob": -101.0970458984375, "reading_distances": [24, 23, 23, 21], "reading_log_probs": [-101.1944808959961, -99.67666625976562, -100.44032287597656, -100.45516967773438], "reading_logprob_std": 0.619714725589998, "tokenization_path_logprob_std": 57.257203600705886, "trajectory_balance_rmse": 1745.9556685434923} {"canonical_length": 189, "canonical_nll": 1.15555988291584, "generation_rouge_l_f1": 0.30973451327433627, "generation_token_f1": 0.5619469026548672, "prompt_id": "ff140f7a02786106", "reading_canonical_log_prob": -218.40081787109375, "reading_distances": [15, 15, 20, 21], "reading_log_probs": [-219.60337829589844, -218.13951110839844, -216.2935333251953, -213.99972534179688], "reading_logprob_std": 2.420501369153333, "tokenization_path_logprob_std": 182.23706737646586, "trajectory_balance_rmse": 3362.104007328448} {"canonical_length": 27, "canonical_nll": 1.8048646714952257, "generation_rouge_l_f1": 0.35294117647058826, "generation_token_f1": 0.35294117647058826, "prompt_id": "536f5c5bf3146493", "reading_canonical_log_prob": -48.731346130371094, "reading_distances": [31, 42, 32, 33], "reading_log_probs": [-55.8297119140625, -53.769874572753906, -53.192481994628906, -52.61957550048828], "reading_logprob_std": 1.3990377440202015, "tokenization_path_logprob_std": 51.78893640132759, "trajectory_balance_rmse": 742.3532772922003} {"canonical_length": 210, "canonical_nll": 1.533834984188988, "generation_rouge_l_f1": 0.21765913757700203, "generation_token_f1": 0.44763860369609854, "prompt_id": "18e930a2b5a84e85", "reading_canonical_log_prob": -322.1053466796875, "reading_distances": [14, 19, 17, 14], "reading_log_probs": [-317.4683837890625, -320.5128479003906, -317.0699462890625, -319.9946594238281], "reading_logprob_std": 1.743693887330863, "tokenization_path_logprob_std": 209.26240219557465, "trajectory_balance_rmse": 4164.698856383523} {"canonical_length": 21, "canonical_nll": 1.564734867640904, "generation_rouge_l_f1": 0.07865168539325844, "generation_token_f1": 0.07865168539325844, "prompt_id": "2631f3bcf5083818", "reading_canonical_log_prob": -32.859432220458984, "reading_distances": [32, 21, 35, 28], "reading_log_probs": [-32.417152404785156, -29.839927673339844, -29.970806121826172, -32.547691345214844], "reading_logprob_std": 1.4897759505210888, "tokenization_path_logprob_std": 60.08287254193489, "trajectory_balance_rmse": 282.13600239163264} {"canonical_length": 169, "canonical_nll": 0.8971237995215422, "generation_rouge_l_f1": 0.2702702702702703, "generation_token_f1": 0.45700245700245706, "prompt_id": "62ad42126bc04505", "reading_canonical_log_prob": -151.61392211914062, "reading_distances": [43, 45, 43, 38], "reading_log_probs": [-151.6907958984375, -149.5548553466797, -151.73971557617188, -149.6759033203125], "reading_logprob_std": 1.2135353015971193, "tokenization_path_logprob_std": 161.64360716654502, "trajectory_balance_rmse": 2873.857977940864} {"canonical_length": 221, "canonical_nll": 1.2787036464225112, "generation_rouge_l_f1": 0.3136094674556213, "generation_token_f1": 0.5029585798816568, "prompt_id": "44a0d1ef8b9a47d0", "reading_canonical_log_prob": -282.593505859375, "reading_distances": [15, 7, 13, 19], "reading_log_probs": [-279.06536865234375, -280.7912292480469, -281.04144287109375, -279.02203369140625], "reading_logprob_std": 1.0861251939214136, "tokenization_path_logprob_std": 178.81701220570767, "trajectory_balance_rmse": 4942.668836323909} {"canonical_length": 84, "canonical_nll": 1.273427509126209, "generation_rouge_l_f1": 0.024390243902439022, "generation_token_f1": 0.024390243902439022, "prompt_id": "1bc849b3b651f6ed", "reading_canonical_log_prob": -106.96791076660156, "reading_distances": [23, 20, 21, 26], "reading_log_probs": [-106.39864349365234, -108.25559997558594, -106.25630187988281, -106.47367858886719], "reading_logprob_std": 0.9440104216345767, "tokenization_path_logprob_std": 157.24926011629609, "trajectory_balance_rmse": 1953.0999191125231} {"canonical_length": 225, "canonical_nll": 1.2959532335069444, "generation_rouge_l_f1": 0.21501014198782964, "generation_token_f1": 0.49492900608519264, "prompt_id": "a8febb0397bb60c3", "reading_canonical_log_prob": -291.5894775390625, "reading_distances": [16, 15, 18, 12], "reading_log_probs": [-285.8533630371094, -285.89959716796875, -284.6490173339844, -284.96453857421875], "reading_logprob_std": 0.6311651422355017, "tokenization_path_logprob_std": 40.06545071415317, "trajectory_balance_rmse": 4305.293361866042} {"canonical_length": 7, "canonical_nll": 2.1484998975481306, "generation_rouge_l_f1": 0.5454545454545454, "generation_token_f1": 0.5454545454545454, "prompt_id": "37af528e072ecf7b", "reading_canonical_log_prob": -15.039499282836914, "reading_distances": [9, 15, 22, 20], "reading_log_probs": [-14.580262184143066, -14.729242324829102, -14.7530517578125, -13.390824317932129], "reading_logprob_std": 0.6528403968373006, "tokenization_path_logprob_std": 50.80001230811874, "trajectory_balance_rmse": 130.17280667095784} {"canonical_length": 252, "canonical_nll": 1.496034652467758, "generation_rouge_l_f1": 0.18540433925049307, "generation_token_f1": 0.43786982248520706, "prompt_id": "192d126d4e011ab4", "reading_canonical_log_prob": -377.000732421875, "reading_distances": [21, 25, 16, 20], "reading_log_probs": [-372.01666259765625, -370.71343994140625, -368.94146728515625, -371.56494140625], "reading_logprob_std": 1.3572897578056093, "tokenization_path_logprob_std": 157.35108627969012, "trajectory_balance_rmse": 5756.772859554134} {"canonical_length": 182, "canonical_nll": 1.3578937237079327, "generation_rouge_l_f1": 0.12093023255813953, "generation_token_f1": 0.19069767441860463, "prompt_id": "55712d2518889f49", "reading_canonical_log_prob": -247.13665771484375, "reading_distances": [23, 25, 23, 19], "reading_log_probs": [-263.96136474609375, -260.00921630859375, -257.49139404296875, -254.48147583007812], "reading_logprob_std": 4.013552493725823, "tokenization_path_logprob_std": 94.00800976927636, "trajectory_balance_rmse": 3435.9430269160193} {"canonical_length": 12, "canonical_nll": 1.010965903600057, "generation_rouge_l_f1": 0.5263157894736842, "generation_token_f1": 0.5263157894736842, "prompt_id": "587eb61a8e257f2b", "reading_canonical_log_prob": -12.131590843200684, "reading_distances": [21, 25, 23, 29], "reading_log_probs": [-13.994817733764648, -13.268508911132812, -12.142069816589355, -13.2164888381958], "reading_logprob_std": 0.7633229207332929, "tokenization_path_logprob_std": 41.457278890415665, "trajectory_balance_rmse": 354.6476132747488} {"canonical_length": 109, "canonical_nll": 0.8577847262041285, "generation_rouge_l_f1": 0.09495548961424331, "generation_token_f1": 0.142433234421365, "prompt_id": "fff5e81d18ab40f8", "reading_canonical_log_prob": -93.49853515625, "reading_distances": [15, 15, 14, 16], "reading_log_probs": [-93.00985717773438, -93.57271575927734, -93.92122650146484, -93.0074691772461], "reading_logprob_std": 0.44938155772510746, "tokenization_path_logprob_std": 148.9729144384238, "trajectory_balance_rmse": 1253.1789552259013} {"canonical_length": 15, "canonical_nll": 0.48791065216064455, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.75, "prompt_id": "e02d75efb09e53b0", "reading_canonical_log_prob": -7.318659782409668, "reading_distances": [25, 28, 20, 18], "reading_log_probs": [-7.671755313873291, -7.888486385345459, -7.717013359069824, -8.003739356994629], "reading_logprob_std": 0.1538755170540297, "tokenization_path_logprob_std": 41.24273961821525, "trajectory_balance_rmse": 251.8055928119506} {"canonical_length": 27, "canonical_nll": 1.352319222909433, "generation_rouge_l_f1": 0.183206106870229, "generation_token_f1": 0.21374045801526717, "prompt_id": "93316a155b82ba8e", "reading_canonical_log_prob": -36.51261901855469, "reading_distances": [27, 27, 28, 24], "reading_log_probs": [-39.04011535644531, -37.559391021728516, -36.356449127197266, -38.072662353515625], "reading_logprob_std": 1.1175299697006498, "tokenization_path_logprob_std": 125.17791626109887, "trajectory_balance_rmse": 552.4954424377042} {"canonical_length": 110, "canonical_nll": 1.3837488347833806, "generation_rouge_l_f1": 0.4491525423728814, "generation_token_f1": 0.5847457627118645, "prompt_id": "37f1255c888facef", "reading_canonical_log_prob": -152.21237182617188, "reading_distances": [14, 19, 18, 12], "reading_log_probs": [-147.97471618652344, -147.67105102539062, -146.6854248046875, -149.05020141601562], "reading_logprob_std": 0.9736887141433365, "tokenization_path_logprob_std": 202.70514538235736, "trajectory_balance_rmse": 2438.9864616738137} {"canonical_length": 150, "canonical_nll": 0.9281134033203124, "generation_rouge_l_f1": 0.43646408839779005, "generation_token_f1": 0.56353591160221, "prompt_id": "cf1ad5289e6ad557", "reading_canonical_log_prob": -139.21701049804688, "reading_distances": [19, 21, 21, 19], "reading_log_probs": [-137.1703338623047, -138.66917419433594, -139.09912109375, -139.1469268798828], "reading_logprob_std": 0.9259702047177063, "tokenization_path_logprob_std": 106.53861152305635, "trajectory_balance_rmse": 2788.0337919632802} {"canonical_length": 314, "canonical_nll": 1.0168893413179239, "generation_rouge_l_f1": 0.21490467937608318, "generation_token_f1": 0.4402079722703639, "prompt_id": "ff9bcfc5d56ada93", "reading_canonical_log_prob": -319.3032531738281, "reading_distances": [41, 34, 36, 40], "reading_log_probs": [-329.9554443359375, -332.5927429199219, -332.07110595703125, -333.74700927734375], "reading_logprob_std": 1.5869367899856677, "tokenization_path_logprob_std": 167.5155390095283, "trajectory_balance_rmse": 5364.279676089038} {"canonical_length": 12, "canonical_nll": 1.9219449361165364, "generation_rouge_l_f1": 0.10526315789473685, "generation_token_f1": 0.10526315789473685, "prompt_id": "6f26ffa7e9fb10c0", "reading_canonical_log_prob": -23.063339233398438, "reading_distances": [22, 21, 20, 22], "reading_log_probs": [-22.783584594726562, -22.843706130981445, -23.27399444580078, -24.33511734008789], "reading_logprob_std": 0.718030147137489, "tokenization_path_logprob_std": 36.12926580819517, "trajectory_balance_rmse": 244.41191575229522} {"canonical_length": 41, "canonical_nll": 1.0523182938738567, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "07e67bead15cc6c8", "reading_canonical_log_prob": -43.145050048828125, "reading_distances": [31, 38, 35, 24], "reading_log_probs": [-41.5130615234375, -40.98936080932617, -42.692474365234375, -44.53265380859375], "reading_logprob_std": 1.5712255329484734, "tokenization_path_logprob_std": 106.71447879218323, "trajectory_balance_rmse": 754.1362742977606} {"canonical_length": 413, "canonical_nll": 1.1770071255958687, "generation_rouge_l_f1": 0.22488755622188908, "generation_token_f1": 0.4917541229385307, "prompt_id": "52cb3482c2bf1893", "reading_canonical_log_prob": -486.10394287109375, "reading_distances": [26, 24, 22, 33], "reading_log_probs": [-482.7247619628906, -478.3395080566406, -480.363525390625, -483.5552673339844], "reading_logprob_std": 2.3626192524464464, "tokenization_path_logprob_std": 71.20720641823877, "trajectory_balance_rmse": 7240.761782731899} {"canonical_length": 265, "canonical_nll": 1.6354826153449293, "generation_rouge_l_f1": 0.23228346456692914, "generation_token_f1": 0.43700787401574803, "prompt_id": "3b670d936478ce43", "reading_canonical_log_prob": -433.40289306640625, "reading_distances": [31, 26, 30, 29], "reading_log_probs": [-432.56500244140625, -431.429443359375, -431.26885986328125, -432.8723449707031], "reading_logprob_std": 0.8032670815095018, "tokenization_path_logprob_std": 371.26228230307856, "trajectory_balance_rmse": 6108.285701062601} {"canonical_length": 78, "canonical_nll": 2.2808360564403043, "generation_rouge_l_f1": 0.15568862275449102, "generation_token_f1": 0.2634730538922156, "prompt_id": "d940482038c5842a", "reading_canonical_log_prob": -177.90521240234375, "reading_distances": [20, 30, 28, 26], "reading_log_probs": [-174.30856323242188, -175.6283416748047, -171.79360961914062, -175.0370330810547], "reading_logprob_std": 1.6875052266431814, "tokenization_path_logprob_std": 170.36835348021435, "trajectory_balance_rmse": 1844.1050745046325} {"canonical_length": 291, "canonical_nll": 0.9770178516296177, "generation_rouge_l_f1": 0.23157894736842105, "generation_token_f1": 0.4456140350877193, "prompt_id": "07d10dd03fee840c", "reading_canonical_log_prob": -284.31219482421875, "reading_distances": [26, 27, 19, 24], "reading_log_probs": [-288.045166015625, -288.7471923828125, -286.6870422363281, -285.5777587890625], "reading_logprob_std": 1.4126062597034101, "tokenization_path_logprob_std": 275.85596923569534, "trajectory_balance_rmse": 5590.559514281647} {"canonical_length": 4, "canonical_nll": 2.054872989654541, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "7ef3a945dcbdd086", "reading_canonical_log_prob": -8.219491958618164, "reading_distances": [21, 23, 17, 24], "reading_log_probs": [-7.604205131530762, -6.981173992156982, -7.457206726074219, -7.182088851928711], "reading_logprob_std": 0.2784813390895621, "tokenization_path_logprob_std": 16.139350401225826, "trajectory_balance_rmse": 47.80776042069507} {"canonical_length": 156, "canonical_nll": 1.0738360087076824, "generation_rouge_l_f1": 0.23383084577114427, "generation_token_f1": 0.40298507462686567, "prompt_id": "ef63cfbdbb1633a2", "reading_canonical_log_prob": -167.51841735839844, "reading_distances": [25, 27, 34, 30], "reading_log_probs": [-168.64439392089844, -164.11685180664062, -176.8388214111328, -170.55203247070312], "reading_logprob_std": 5.2762845221491155, "tokenization_path_logprob_std": 194.14182971883739, "trajectory_balance_rmse": 2780.670047214008} {"canonical_length": 152, "canonical_nll": 1.5167518414949115, "generation_rouge_l_f1": 0.2925170068027211, "generation_token_f1": 0.5374149659863947, "prompt_id": "1d1550150b11ce18", "reading_canonical_log_prob": -230.54627990722656, "reading_distances": [24, 30, 29, 27], "reading_log_probs": [-226.3485107421875, -226.3946533203125, -226.73385620117188, -225.191162109375], "reading_logprob_std": 0.6728936808169721, "tokenization_path_logprob_std": 240.6376416263008, "trajectory_balance_rmse": 3697.5372930168455} {"canonical_length": 436, "canonical_nll": 0.8480736268769711, "generation_rouge_l_f1": 0.28471001757469244, "generation_token_f1": 0.4288224956063269, "prompt_id": "becce7dcacc67b08", "reading_canonical_log_prob": -369.7601013183594, "reading_distances": [22, 25, 20, 29], "reading_log_probs": [-373.7872009277344, -368.15130615234375, -366.7086486816406, -367.343994140625], "reading_logprob_std": 3.247060242687472, "tokenization_path_logprob_std": 81.44624068819134, "trajectory_balance_rmse": 5537.769869968663} {"canonical_length": 289, "canonical_nll": 1.0858204983509947, "generation_rouge_l_f1": 0.19761499148211245, "generation_token_f1": 0.3713798977853492, "prompt_id": "11042f17f722f10a", "reading_canonical_log_prob": -313.8021240234375, "reading_distances": [23, 29, 28, 26], "reading_log_probs": [-313.371337890625, -309.4840087890625, -310.24285888671875, -312.52392578125], "reading_logprob_std": 1.8402226013964085, "tokenization_path_logprob_std": 178.4695815100559, "trajectory_balance_rmse": 6152.576678585444} {"canonical_length": 31, "canonical_nll": 1.0402384111958165, "generation_rouge_l_f1": 0.13821138211382114, "generation_token_f1": 0.19512195121951217, "prompt_id": "b29cd52aed559d3c", "reading_canonical_log_prob": -32.24739074707031, "reading_distances": [21, 19, 19, 18], "reading_log_probs": [-29.783916473388672, -28.260356903076172, -29.177204132080078, -29.557409286499023], "reading_logprob_std": 0.6713263501718943, "tokenization_path_logprob_std": 43.18759670394244, "trajectory_balance_rmse": 509.18398543991907} {"canonical_length": 192, "canonical_nll": 0.8547577857971191, "generation_rouge_l_f1": 0.24830699774266365, "generation_token_f1": 0.44243792325056436, "prompt_id": "3d6b176b83ba1bb0", "reading_canonical_log_prob": -164.11349487304688, "reading_distances": [20, 22, 20, 16], "reading_log_probs": [-168.1961669921875, -166.70620727539062, -165.94873046875, -165.94400024414062], "reading_logprob_std": 1.0605799903699142, "tokenization_path_logprob_std": 65.4506726596667, "trajectory_balance_rmse": 2180.1238763657757} {"canonical_length": 146, "canonical_nll": 1.9734486619087115, "generation_rouge_l_f1": 0.2085308056872038, "generation_token_f1": 0.4075829383886256, "prompt_id": "ed74cc90058602c2", "reading_canonical_log_prob": -288.1235046386719, "reading_distances": [28, 29, 22, 30], "reading_log_probs": [-285.28985595703125, -286.74267578125, -285.4679870605469, -286.05548095703125], "reading_logprob_std": 0.6564306838142988, "tokenization_path_logprob_std": 226.12877681339265, "trajectory_balance_rmse": 3392.0248267093916} {"canonical_length": 217, "canonical_nll": 1.563913090437788, "generation_rouge_l_f1": 0.23908045977011494, "generation_token_f1": 0.5195402298850574, "prompt_id": "b2e99fb7bff1b5b7", "reading_canonical_log_prob": -339.369140625, "reading_distances": [26, 37, 39, 36], "reading_log_probs": [-338.5223388671875, -338.3908386230469, -341.3455810546875, -335.9337158203125], "reading_logprob_std": 2.212562591626747, "tokenization_path_logprob_std": 253.52456603315747, "trajectory_balance_rmse": 4888.76315965228} {"canonical_length": 318, "canonical_nll": 0.757380503528523, "generation_rouge_l_f1": 0.3227953410981697, "generation_token_f1": 0.5557404326123129, "prompt_id": "5a32b0efbf18f769", "reading_canonical_log_prob": -240.8470001220703, "reading_distances": [26, 29, 29, 25], "reading_log_probs": [-236.14285278320312, -236.27322387695312, -238.6694793701172, -236.9213104248047], "reading_logprob_std": 1.1627906824723644, "tokenization_path_logprob_std": 176.9984875850574, "trajectory_balance_rmse": 5083.258981369656} {"canonical_length": 25, "canonical_nll": 0.4982911682128906, "generation_rouge_l_f1": 0.14012738853503184, "generation_token_f1": 0.15286624203821658, "prompt_id": "3697c04353d3a1f1", "reading_canonical_log_prob": -12.457279205322266, "reading_distances": [29, 29, 32, 26], "reading_log_probs": [-15.144577980041504, -13.332319259643555, -15.439421653747559, -14.94502067565918], "reading_logprob_std": 0.9441114103972584, "tokenization_path_logprob_std": 56.393940252992664, "trajectory_balance_rmse": 292.90881479308564} {"canonical_length": 83, "canonical_nll": 1.7536893178181476, "generation_rouge_l_f1": 0.29770992366412213, "generation_token_f1": 0.5038167938931298, "prompt_id": "4983a8fb02a849d4", "reading_canonical_log_prob": -145.55621337890625, "reading_distances": [18, 19, 13, 10], "reading_log_probs": [-147.10789489746094, -146.95594787597656, -146.6411895751953, -145.88278198242188], "reading_logprob_std": 0.5452631864170941, "tokenization_path_logprob_std": 142.45935611683498, "trajectory_balance_rmse": 1409.893293814615} {"canonical_length": 244, "canonical_nll": 1.7122690169537653, "generation_rouge_l_f1": 0.2011605415860735, "generation_token_f1": 0.41392649903288203, "prompt_id": "fe89f27185bbc3b2", "reading_canonical_log_prob": -417.79364013671875, "reading_distances": [43, 39, 29, 41], "reading_log_probs": [-415.8414001464844, -415.1060791015625, -416.72711181640625, -413.8946533203125], "reading_logprob_std": 1.1983713261727214, "tokenization_path_logprob_std": 82.77840886473285, "trajectory_balance_rmse": 6111.60868942679} {"canonical_length": 87, "canonical_nll": 1.6004866676769038, "generation_rouge_l_f1": 0.1941747572815534, "generation_token_f1": 0.3689320388349514, "prompt_id": "6b767f6a353fd065", "reading_canonical_log_prob": -139.24234008789062, "reading_distances": [20, 17, 23, 18], "reading_log_probs": [-135.0018310546875, -134.64610290527344, -132.32162475585938, -143.54913330078125], "reading_logprob_std": 4.925194759460245, "tokenization_path_logprob_std": 123.84554618849128, "trajectory_balance_rmse": 1609.4436469784107} {"canonical_length": 136, "canonical_nll": 1.3681687747730928, "generation_rouge_l_f1": 0.2524271844660194, "generation_token_f1": 0.48543689320388356, "prompt_id": "e77b6aefa4b6596d", "reading_canonical_log_prob": -186.07095336914062, "reading_distances": [31, 25, 29, 28], "reading_log_probs": [-183.77133178710938, -185.8123779296875, -182.9501495361328, -184.42105102539062], "reading_logprob_std": 1.2094777749314154, "tokenization_path_logprob_std": 116.47172611627809, "trajectory_balance_rmse": 2486.515612486676} {"canonical_length": 9, "canonical_nll": 1.7471996943155925, "generation_rouge_l_f1": 0.4444444444444445, "generation_token_f1": 0.6666666666666665, "prompt_id": "383239282305a169", "reading_canonical_log_prob": -15.724797248840332, "reading_distances": [31, 34, 35, 32], "reading_log_probs": [-15.158621788024902, -13.9632568359375, -15.20102310180664, -13.924062728881836], "reading_logprob_std": 0.7140880321298705, "tokenization_path_logprob_std": 24.782642801747716, "trajectory_balance_rmse": 131.99423539874505} {"canonical_length": 10, "canonical_nll": 0.6512556076049805, "generation_rouge_l_f1": 0.6153846153846153, "generation_token_f1": 0.6153846153846153, "prompt_id": "29653e40ee0ea781", "reading_canonical_log_prob": -6.512556076049805, "reading_distances": [30, 24, 20, 25], "reading_log_probs": [-7.101836204528809, -6.967477321624756, -7.886247158050537, -6.5058979988098145], "reading_logprob_std": 0.5738069131865696, "tokenization_path_logprob_std": 29.83777308276869, "trajectory_balance_rmse": 161.50701539935233} {"canonical_length": 251, "canonical_nll": 1.6082768535234064, "generation_rouge_l_f1": 0.22664015904572565, "generation_token_f1": 0.4532803180914513, "prompt_id": "b8e0545bd0cc139e", "reading_canonical_log_prob": -403.677490234375, "reading_distances": [48, 52, 34, 45], "reading_log_probs": [-399.3575134277344, -398.8033447265625, -396.98358154296875, -398.4327392578125], "reading_logprob_std": 1.014347382496989, "tokenization_path_logprob_std": 96.91551205607959, "trajectory_balance_rmse": 5683.178532268937} {"canonical_length": 430, "canonical_nll": 1.2545834563499274, "generation_rouge_l_f1": 0.2169811320754717, "generation_token_f1": 0.41823899371069184, "prompt_id": "a7d20d598bd1e5dd", "reading_canonical_log_prob": -539.4708862304688, "reading_distances": [36, 35, 38, 29], "reading_log_probs": [-611.5587158203125, -616.9873046875, -622.4159545898438, -623.7526245117188], "reading_logprob_std": 5.575744333929378, "tokenization_path_logprob_std": 318.82592101553223, "trajectory_balance_rmse": 7622.422827949734} {"canonical_length": 47, "canonical_nll": 2.815251776512633, "generation_rouge_l_f1": 0.20512820512820512, "generation_token_f1": 0.282051282051282, "prompt_id": "ecbe9c8ff4f546fd", "reading_canonical_log_prob": -132.31683349609375, "reading_distances": [12, 14, 10, 13], "reading_log_probs": [-133.0715789794922, -131.8075408935547, -135.6358642578125, -138.54617309570312], "reading_logprob_std": 2.9816054578814395, "tokenization_path_logprob_std": 86.96793849376759, "trajectory_balance_rmse": 861.4994773264452} {"canonical_length": 12, "canonical_nll": 1.0094889005025227, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "641a36ce0ce7d08c", "reading_canonical_log_prob": -12.113866806030273, "reading_distances": [25, 23, 26, 29], "reading_log_probs": [-13.082592010498047, -13.29879379272461, -16.351449966430664, -11.313347816467285], "reading_logprob_std": 2.0917612052562546, "tokenization_path_logprob_std": 46.888622873210615, "trajectory_balance_rmse": 305.3547820990419} {"canonical_length": 105, "canonical_nll": 1.7101242792038691, "generation_rouge_l_f1": 0.3243243243243243, "generation_token_f1": 0.42342342342342343, "prompt_id": "32d37daf5bb72f67", "reading_canonical_log_prob": -179.56304931640625, "reading_distances": [19, 20, 24, 20], "reading_log_probs": [-175.56715393066406, -177.53598022460938, -176.8375244140625, -175.83082580566406], "reading_logprob_std": 0.911430889607709, "tokenization_path_logprob_std": 124.45097957165919, "trajectory_balance_rmse": 2470.5516013157066} {"canonical_length": 19, "canonical_nll": 2.3582789772435238, "generation_rouge_l_f1": 0.3448275862068966, "generation_token_f1": 0.3448275862068966, "prompt_id": "b21a0da1ea5ee3b7", "reading_canonical_log_prob": -44.80730056762695, "reading_distances": [20, 26, 23, 24], "reading_log_probs": [-44.29543685913086, -46.60808563232422, -43.538978576660156, -44.351806640625], "reading_logprob_std": 1.3258536087685795, "tokenization_path_logprob_std": 64.06865637182368, "trajectory_balance_rmse": 410.7237113902034} {"canonical_length": 281, "canonical_nll": 1.5477923736029249, "generation_rouge_l_f1": 0.1741424802110818, "generation_token_f1": 0.3218997361477573, "prompt_id": "11a80e6e30ed7f8f", "reading_canonical_log_prob": -434.9296569824219, "reading_distances": [19, 17, 19, 21], "reading_log_probs": [-431.9406433105469, -434.6150207519531, -432.07965087890625, -432.7742919921875], "reading_logprob_std": 1.2303623099182213, "tokenization_path_logprob_std": 69.09189945769586, "trajectory_balance_rmse": 6154.33942266091} {"canonical_length": 119, "canonical_nll": 1.8849586038028492, "generation_rouge_l_f1": 0.1958762886597938, "generation_token_f1": 0.3814432989690722, "prompt_id": "e2dff42d7f8d6988", "reading_canonical_log_prob": -224.31007385253906, "reading_distances": [23, 21, 26, 21], "reading_log_probs": [-220.82803344726562, -218.20733642578125, -220.72998046875, -220.3931884765625], "reading_logprob_std": 1.2356465384226525, "tokenization_path_logprob_std": 155.28878331329216, "trajectory_balance_rmse": 2453.4682035610645} {"canonical_length": 338, "canonical_nll": 1.677577272674741, "generation_rouge_l_f1": 0.17869415807560138, "generation_token_f1": 0.37457044673539525, "prompt_id": "bd63a203fd3cb2b2", "reading_canonical_log_prob": -567.0211181640625, "reading_distances": [29, 29, 31, 30], "reading_log_probs": [-558.5277099609375, -563.0304565429688, -562.7131958007812, -562.853759765625], "reading_logprob_std": 2.172927360498448, "tokenization_path_logprob_std": 204.29995183277387, "trajectory_balance_rmse": 5745.087072636958} {"canonical_length": 17, "canonical_nll": 0.4457208689521341, "generation_rouge_l_f1": 0.10894941634241245, "generation_token_f1": 0.11673151750972763, "prompt_id": "f5560d8544cd62d8", "reading_canonical_log_prob": -7.577254772186279, "reading_distances": [25, 21, 22, 18], "reading_log_probs": [-8.281862258911133, -8.64997673034668, -8.17837142944336, -8.150514602661133], "reading_logprob_std": 0.2302391514449754, "tokenization_path_logprob_std": 47.56329967337111, "trajectory_balance_rmse": 395.64629777769824} {"canonical_length": 143, "canonical_nll": 1.5376576243580637, "generation_rouge_l_f1": 0.23152709359605916, "generation_token_f1": 0.45320197044334976, "prompt_id": "d5ad06fe63008db3", "reading_canonical_log_prob": -219.88504028320312, "reading_distances": [21, 24, 21, 26], "reading_log_probs": [-219.29782104492188, -220.20761108398438, -220.38169860839844, -220.03775024414062], "reading_logprob_std": 0.4767479554218002, "tokenization_path_logprob_std": 276.264783424491, "trajectory_balance_rmse": 3268.2434081599727} {"canonical_length": 68, "canonical_nll": 1.2559634937959558, "generation_rouge_l_f1": 0.15211267605633802, "generation_token_f1": 0.20845070422535208, "prompt_id": "32cf7dbccdbc97bc", "reading_canonical_log_prob": -85.405517578125, "reading_distances": [16, 20, 16, 23], "reading_log_probs": [-87.62645721435547, -85.09249877929688, -84.99852752685547, -88.10501098632812], "reading_logprob_std": 1.6403829172760396, "tokenization_path_logprob_std": 21.33843744280007, "trajectory_balance_rmse": 1716.996141985756} {"canonical_length": 47, "canonical_nll": 1.8424656644780586, "generation_rouge_l_f1": 0.13427561837455831, "generation_token_f1": 0.15547703180212014, "prompt_id": "c180433ec97360b0", "reading_canonical_log_prob": -86.59588623046875, "reading_distances": [32, 24, 28, 36], "reading_log_probs": [-86.86024475097656, -85.36332702636719, -85.79261016845703, -85.74296569824219], "reading_logprob_std": 0.6428966265018066, "tokenization_path_logprob_std": 12.896359741407494, "trajectory_balance_rmse": 912.5405394379383} {"canonical_length": 122, "canonical_nll": 0.9395493053999103, "generation_rouge_l_f1": 0.3313953488372093, "generation_token_f1": 0.46511627906976744, "prompt_id": "aaf9ee6d37011f66", "reading_canonical_log_prob": -114.62501525878906, "reading_distances": [35, 34, 46, 32], "reading_log_probs": [-120.52153015136719, -117.93672180175781, -118.39765167236328, -121.50257873535156], "reading_logprob_std": 1.7010514984580387, "tokenization_path_logprob_std": 114.34691030258381, "trajectory_balance_rmse": 2042.0445665393097} {"canonical_length": 19, "canonical_nll": 0.3881208520186575, "generation_rouge_l_f1": 0.08764940239043825, "generation_token_f1": 0.08764940239043825, "prompt_id": "f459f20b3852c4a3", "reading_canonical_log_prob": -7.374296188354492, "reading_distances": [15, 21, 20, 22], "reading_log_probs": [-5.517940521240234, -5.88959264755249, -5.383396148681641, -9.83962345123291], "reading_logprob_std": 2.1320981617502657, "tokenization_path_logprob_std": 78.31410171699844, "trajectory_balance_rmse": 249.91404686214} {"canonical_length": 19, "canonical_nll": 1.6469656291760897, "generation_rouge_l_f1": 0.3636363636363636, "generation_token_f1": 0.3636363636363636, "prompt_id": "a05d7928d7de0d6c", "reading_canonical_log_prob": -31.292346954345703, "reading_distances": [26, 28, 28, 20], "reading_log_probs": [-31.330232620239258, -30.620243072509766, -32.31124496459961, -30.039329528808594], "reading_logprob_std": 0.9785809820359841, "tokenization_path_logprob_std": 18.868120430806858, "trajectory_balance_rmse": 365.8144335756989} {"canonical_length": 15, "canonical_nll": 0.9656520843505859, "generation_rouge_l_f1": 0.8181818181818181, "generation_token_f1": 0.8181818181818181, "prompt_id": "2c22dccbca7a08c7", "reading_canonical_log_prob": -14.484781265258789, "reading_distances": [26, 28, 26, 28], "reading_log_probs": [-15.840825080871582, -14.427696228027344, -14.602469444274902, -14.743005752563477], "reading_logprob_std": 0.638055727866997, "tokenization_path_logprob_std": 40.71807578715634, "trajectory_balance_rmse": 344.52115766091396} {"canonical_length": 166, "canonical_nll": 1.7755950560052711, "generation_rouge_l_f1": 0.26628895184135976, "generation_token_f1": 0.43626062322946174, "prompt_id": "6e818f0ad7674362", "reading_canonical_log_prob": -294.748779296875, "reading_distances": [25, 24, 28, 23], "reading_log_probs": [-287.08941650390625, -288.7917175292969, -287.97662353515625, -287.88177490234375], "reading_logprob_std": 0.6960703386210766, "tokenization_path_logprob_std": 236.15094418921672, "trajectory_balance_rmse": 3105.7415132906945} {"canonical_length": 70, "canonical_nll": 1.0135868617466517, "generation_rouge_l_f1": 0.2418604651162791, "generation_token_f1": 0.45581395348837206, "prompt_id": "8ebd94ecdb15c9f4", "reading_canonical_log_prob": -70.95108032226562, "reading_distances": [22, 16, 20, 23], "reading_log_probs": [-69.08013153076172, -68.64891052246094, -69.45645904541016, -68.02474212646484], "reading_logprob_std": 0.6146106416840655, "tokenization_path_logprob_std": 86.43572507000268, "trajectory_balance_rmse": 1712.3985589490649} {"canonical_length": 3, "canonical_nll": 0.1397848924001058, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "2148a4bb4107090b", "reading_canonical_log_prob": -0.4193546772003174, "reading_distances": [15, 22, 21, 16], "reading_log_probs": [-0.47013378143310547, -0.4757497310638428, -0.4192090630531311, -0.45301640033721924], "reading_logprob_std": 0.025453320161916232, "tokenization_path_logprob_std": 21.96809815752859, "trajectory_balance_rmse": 49.25924847358943} {"canonical_length": 368, "canonical_nll": 0.9763155398161515, "generation_rouge_l_f1": 0.208398133748056, "generation_token_f1": 0.43856920684292383, "prompt_id": "5d2176dd2ade1ce9", "reading_canonical_log_prob": -359.28411865234375, "reading_distances": [25, 27, 20, 22], "reading_log_probs": [-349.9385986328125, -349.8082275390625, -352.1856384277344, -349.1272277832031], "reading_logprob_std": 1.3289791191387637, "tokenization_path_logprob_std": 196.88590799111287, "trajectory_balance_rmse": 7106.196416270225} {"canonical_length": 12, "canonical_nll": 0.8181119759877523, "generation_rouge_l_f1": 0.06756756756756757, "generation_token_f1": 0.06756756756756757, "prompt_id": "9d6ee475e283101d", "reading_canonical_log_prob": -9.817343711853027, "reading_distances": [30, 28, 33, 33], "reading_log_probs": [-9.499765396118164, -9.179056167602539, -9.413002014160156, -9.796785354614258], "reading_logprob_std": 0.2553123592560571, "tokenization_path_logprob_std": 25.753463857114006, "trajectory_balance_rmse": 282.26798046349353} {"canonical_length": 12, "canonical_nll": 0.5357809066772461, "generation_rouge_l_f1": 0.5, "generation_token_f1": 0.5, "prompt_id": "a41c0de97fa957ce", "reading_canonical_log_prob": -6.429370880126953, "reading_distances": [20, 17, 24, 23], "reading_log_probs": [-8.50201416015625, -8.319045066833496, -8.293261528015137, -8.137592315673828], "reading_logprob_std": 0.149354563238383, "tokenization_path_logprob_std": 14.359702015657339, "trajectory_balance_rmse": 203.2354632333585} {"canonical_length": 9, "canonical_nll": 0.0833393997616238, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "916ba9265d9ad9c6", "reading_canonical_log_prob": -0.7500545978546143, "reading_distances": [28, 36, 39, 32], "reading_log_probs": [-0.9220675826072693, -0.8330195546150208, -0.9183531999588013, -0.8496506214141846], "reading_logprob_std": 0.046067009078645284, "tokenization_path_logprob_std": 21.600628805669146, "trajectory_balance_rmse": 182.45679863566127} {"canonical_length": 99, "canonical_nll": 0.7863521383266256, "generation_rouge_l_f1": 0.5638297872340425, "generation_token_f1": 0.6595744680851063, "prompt_id": "5baddab4b656da75", "reading_canonical_log_prob": -77.84886169433594, "reading_distances": [20, 25, 26, 30], "reading_log_probs": [-81.99638366699219, -78.81002807617188, -81.98161315917969, -80.4096450805664], "reading_logprob_std": 1.5209544591141424, "tokenization_path_logprob_std": 149.70670034915713, "trajectory_balance_rmse": 1843.6634782105195} {"canonical_length": 54, "canonical_nll": 0.7759371157045718, "generation_rouge_l_f1": 0.6929133858267716, "generation_token_f1": 0.6929133858267716, "prompt_id": "22a09297862230a9", "reading_canonical_log_prob": -41.900604248046875, "reading_distances": [31, 27, 27, 34], "reading_log_probs": [-42.297882080078125, -43.79555892944336, -43.125911712646484, -44.16402053833008], "reading_logprob_std": 0.8202135417607317, "tokenization_path_logprob_std": 85.71758189549455, "trajectory_balance_rmse": 834.2546949526534} {"canonical_length": 40, "canonical_nll": 0.13160147666931152, "generation_rouge_l_f1": 0.21951219512195125, "generation_token_f1": 0.22764227642276422, "prompt_id": "c6508f3625213115", "reading_canonical_log_prob": -5.264059066772461, "reading_distances": [10, 15, 15, 12], "reading_log_probs": [-5.189925193786621, -4.587841987609863, -4.529198169708252, -4.413151264190674], "reading_logprob_std": 0.3475947060052452, "tokenization_path_logprob_std": 33.46198692058252, "trajectory_balance_rmse": 229.58332356074024} {"canonical_length": 272, "canonical_nll": 1.5286862990435433, "generation_rouge_l_f1": 0.2900188323917137, "generation_token_f1": 0.5009416195856873, "prompt_id": "a4b74ba3880275a6", "reading_canonical_log_prob": -415.80267333984375, "reading_distances": [20, 15, 19, 14], "reading_log_probs": [-420.9219055175781, -419.7160339355469, -424.10614013671875, -418.24664306640625], "reading_logprob_std": 2.4919261871361127, "tokenization_path_logprob_std": 229.4442582407607, "trajectory_balance_rmse": 5814.271764944128} {"canonical_length": 17, "canonical_nll": 0.45746203029856963, "generation_rouge_l_f1": 0.18633540372670807, "generation_token_f1": 0.18633540372670807, "prompt_id": "5f38035dc4f5627c", "reading_canonical_log_prob": -7.776854515075684, "reading_distances": [20, 17, 20, 22], "reading_log_probs": [-8.794037818908691, -8.893264770507812, -8.597291946411133, -8.696977615356445], "reading_logprob_std": 0.12716177421442956, "tokenization_path_logprob_std": 31.700468913524436, "trajectory_balance_rmse": 401.15528707852593} {"canonical_length": 34, "canonical_nll": 0.7661750456866097, "generation_rouge_l_f1": 0.5423728813559322, "generation_token_f1": 0.7796610169491525, "prompt_id": "0e027c7a776ba5aa", "reading_canonical_log_prob": -26.049951553344727, "reading_distances": [18, 25, 24, 27], "reading_log_probs": [-26.161964416503906, -26.94962501525879, -26.60793685913086, -25.726547241210938], "reading_logprob_std": 0.5321658929114302, "tokenization_path_logprob_std": 69.70339949678176, "trajectory_balance_rmse": 540.1772359054349} {"canonical_length": 23, "canonical_nll": 2.3078878651494565, "generation_rouge_l_f1": 0.07692307692307691, "generation_token_f1": 0.15384615384615383, "prompt_id": "9d315888010aca18", "reading_canonical_log_prob": -53.0814208984375, "reading_distances": [19, 24, 26, 19], "reading_log_probs": [-51.161407470703125, -50.902740478515625, -50.383583068847656, -51.083106994628906], "reading_logprob_std": 0.34993339799774814, "tokenization_path_logprob_std": 62.175926011217584, "trajectory_balance_rmse": 411.67286036321536} {"canonical_length": 390, "canonical_nll": 0.9762477776943109, "generation_rouge_l_f1": 0.31726283048211507, "generation_token_f1": 0.4852255054432348, "prompt_id": "b3017655183fdef3", "reading_canonical_log_prob": -380.73663330078125, "reading_distances": [23, 22, 20, 23], "reading_log_probs": [-382.2895812988281, -381.5971374511719, -379.7716979980469, -381.6863708496094], "reading_logprob_std": 1.0873990975982262, "tokenization_path_logprob_std": 239.60806771954003, "trajectory_balance_rmse": 7352.992595769447} {"canonical_length": 24, "canonical_nll": 0.9930164019266764, "generation_rouge_l_f1": 0.6285714285714286, "generation_token_f1": 0.6285714285714286, "prompt_id": "70f7bc293fb9b6e2", "reading_canonical_log_prob": -23.832393646240234, "reading_distances": [36, 38, 32, 32], "reading_log_probs": [-28.337631225585938, -27.41819190979004, -29.376026153564453, -25.173969268798828], "reading_logprob_std": 1.790235596924426, "tokenization_path_logprob_std": 46.742727351551146, "trajectory_balance_rmse": 393.9968424561193} {"canonical_length": 85, "canonical_nll": 1.397571070053998, "generation_rouge_l_f1": 0.1732283464566929, "generation_token_f1": 0.2362204724409449, "prompt_id": "6a1786b00d1d1907", "reading_canonical_log_prob": -118.79354095458984, "reading_distances": [30, 26, 28, 25], "reading_log_probs": [-118.50128173828125, -119.39834594726562, -118.09020233154297, -120.84490966796875], "reading_logprob_std": 1.2199214856315144, "tokenization_path_logprob_std": 118.9694406096729, "trajectory_balance_rmse": 1294.6900388769639} {"canonical_length": 108, "canonical_nll": 1.4866823267053675, "generation_rouge_l_f1": 0.24299065420560748, "generation_token_f1": 0.411214953271028, "prompt_id": "7988aedbe67fdb66", "reading_canonical_log_prob": -160.5616912841797, "reading_distances": [15, 15, 17, 16], "reading_log_probs": [-158.87014770507812, -159.56317138671875, -161.57156372070312, -159.23773193359375], "reading_logprob_std": 1.2075929879041423, "tokenization_path_logprob_std": 56.92892226368006, "trajectory_balance_rmse": 2171.719268760407} {"canonical_length": 98, "canonical_nll": 2.4168431807537467, "generation_rouge_l_f1": 0.23626373626373626, "generation_token_f1": 0.30769230769230765, "prompt_id": "0b1fe757c0710416", "reading_canonical_log_prob": -236.8506317138672, "reading_distances": [12, 14, 16, 18], "reading_log_probs": [-230.12998962402344, -228.2342529296875, -229.90621948242188, -228.86160278320312], "reading_logprob_std": 0.8912986737034096, "tokenization_path_logprob_std": 132.65974699914483, "trajectory_balance_rmse": 2019.5145139581118} {"canonical_length": 13, "canonical_nll": 0.8304649499746469, "generation_rouge_l_f1": 0.05223880597014925, "generation_token_f1": 0.05970149253731343, "prompt_id": "78fa2b118c48dd30", "reading_canonical_log_prob": -10.79604434967041, "reading_distances": [22, 33, 29, 27], "reading_log_probs": [-10.920412063598633, -9.74968433380127, -8.207927703857422, -9.574189186096191], "reading_logprob_std": 1.1111170560056984, "tokenization_path_logprob_std": 30.734271094636874, "trajectory_balance_rmse": 246.6628377324799} {"canonical_length": 424, "canonical_nll": 1.1185975704552993, "generation_rouge_l_f1": 0.1895734597156398, "generation_token_f1": 0.43601895734597157, "prompt_id": "995817411c15ddee", "reading_canonical_log_prob": -474.2853698730469, "reading_distances": [18, 16, 10, 10], "reading_log_probs": [-476.6695861816406, -477.1690673828125, -477.2650146484375, -478.5792236328125], "reading_logprob_std": 0.8152516741330729, "tokenization_path_logprob_std": 155.64563131168234, "trajectory_balance_rmse": 7538.532836734398} {"canonical_length": 64, "canonical_nll": 1.3312032222747803, "generation_rouge_l_f1": 0.31213872832369943, "generation_token_f1": 0.4971098265895954, "prompt_id": "fd32715308ec2a93", "reading_canonical_log_prob": -85.19700622558594, "reading_distances": [37, 47, 42, 46], "reading_log_probs": [-86.15689086914062, -86.46619415283203, -88.19081115722656, -87.46651458740234], "reading_logprob_std": 0.933057913875498, "tokenization_path_logprob_std": 184.90064546673784, "trajectory_balance_rmse": 1490.1428863534193} {"canonical_length": 333, "canonical_nll": 1.1993611653645833, "generation_rouge_l_f1": 0.20408163265306123, "generation_token_f1": 0.3487940630797774, "prompt_id": "6c1da32736122296", "reading_canonical_log_prob": -399.38726806640625, "reading_distances": [35, 35, 36, 41], "reading_log_probs": [-397.9328308105469, -400.0482177734375, -395.192138671875, -400.50628662109375], "reading_logprob_std": 2.4262527682724024, "tokenization_path_logprob_std": 147.09143783332968, "trajectory_balance_rmse": 5192.523946340084} {"canonical_length": 53, "canonical_nll": 0.7737082715304393, "generation_rouge_l_f1": 0.25065274151436034, "generation_token_f1": 0.2819843342036553, "prompt_id": "ffb0593302e5d193", "reading_canonical_log_prob": -41.00653839111328, "reading_distances": [40, 37, 35, 40], "reading_log_probs": [-41.995243072509766, -42.60276412963867, -42.210391998291016, -41.53102493286133], "reading_logprob_std": 0.4467465520531281, "tokenization_path_logprob_std": 95.71537648837482, "trajectory_balance_rmse": 870.8193829242371} {"canonical_length": 142, "canonical_nll": 1.3903318593199825, "generation_rouge_l_f1": 0.26329113924050634, "generation_token_f1": 0.4354430379746836, "prompt_id": "780311bcf262bb9d", "reading_canonical_log_prob": -197.4271240234375, "reading_distances": [35, 31, 31, 37], "reading_log_probs": [-199.3174285888672, -197.60191345214844, -196.0614776611328, -198.7965087890625], "reading_logprob_std": 1.4461431977826729, "tokenization_path_logprob_std": 161.22685417012985, "trajectory_balance_rmse": 3151.117692180357} {"canonical_length": 32, "canonical_nll": 1.5965638160705566, "generation_rouge_l_f1": 0.40625, "generation_token_f1": 0.40625, "prompt_id": "ddc773449a2312dd", "reading_canonical_log_prob": -51.09004211425781, "reading_distances": [34, 28, 33, 31], "reading_log_probs": [-49.31049728393555, -50.21403503417969, -51.18370819091797, -50.349632263183594], "reading_logprob_std": 0.7669982815000994, "tokenization_path_logprob_std": 101.62641825073887, "trajectory_balance_rmse": 806.1435783049377} {"canonical_length": 259, "canonical_nll": 0.7333495386779079, "generation_rouge_l_f1": 0.35673624288425043, "generation_token_f1": 0.6261859582542694, "prompt_id": "fa8dcc0f054c9984", "reading_canonical_log_prob": -189.93753051757812, "reading_distances": [21, 28, 27, 21], "reading_log_probs": [-185.681640625, -185.83154296875, -186.049072265625, -187.26431274414062], "reading_logprob_std": 0.7210689636917492, "tokenization_path_logprob_std": 161.81753490119615, "trajectory_balance_rmse": 3046.890576847052} {"canonical_length": 79, "canonical_nll": 1.0306017911886867, "generation_rouge_l_f1": 0.24472573839662448, "generation_token_f1": 0.3881856540084388, "prompt_id": "2d9359e9dd847731", "reading_canonical_log_prob": -81.41754150390625, "reading_distances": [21, 13, 17, 16], "reading_log_probs": [-81.84883117675781, -81.46113586425781, -82.6719970703125, -83.8575439453125], "reading_logprob_std": 1.0597686333305447, "tokenization_path_logprob_std": 155.05541984884937, "trajectory_balance_rmse": 1603.6638848056737} {"canonical_length": 91, "canonical_nll": 1.847825438111693, "generation_rouge_l_f1": 0.17142857142857143, "generation_token_f1": 0.2857142857142857, "prompt_id": "5cb4ccbc96ab0eb7", "reading_canonical_log_prob": -168.15211486816406, "reading_distances": [27, 35, 32, 31], "reading_log_probs": [-166.26589965820312, -168.11898803710938, -166.31805419921875, -167.426025390625], "reading_logprob_std": 0.9006344259317801, "tokenization_path_logprob_std": 111.77964646516368, "trajectory_balance_rmse": 2050.253650720095} {"canonical_length": 76, "canonical_nll": 0.4875574613872327, "generation_rouge_l_f1": 0.4507042253521127, "generation_token_f1": 0.6338028169014086, "prompt_id": "23716ea232c0bbd4", "reading_canonical_log_prob": -37.05436706542969, "reading_distances": [4, 4, 2, 5], "reading_log_probs": [-37.06806945800781, -36.42534637451172, -37.74539566040039, -36.33816909790039], "reading_logprob_std": 0.6541556182801198, "tokenization_path_logprob_std": 45.96706620280576, "trajectory_balance_rmse": 996.9659823623934} {"canonical_length": 7, "canonical_nll": 1.964364460536412, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.6666666666666666, "prompt_id": "d26a8398a22a7919", "reading_canonical_log_prob": -13.750551223754883, "reading_distances": [39, 35, 40, 40], "reading_log_probs": [-13.922740936279297, -13.06680679321289, -13.49702262878418, -13.503003120422363], "reading_logprob_std": 0.3494552708610991, "tokenization_path_logprob_std": 27.621677003920283, "trajectory_balance_rmse": 194.54893707785567} {"canonical_length": 177, "canonical_nll": 2.0091473423155013, "generation_rouge_l_f1": 0.2027027027027027, "generation_token_f1": 0.33333333333333337, "prompt_id": "cf319e0486bfdfdc", "reading_canonical_log_prob": -355.61907958984375, "reading_distances": [31, 27, 24, 18], "reading_log_probs": [-352.14208984375, -351.69451904296875, -350.47900390625, -351.4945068359375], "reading_logprob_std": 0.7032227207641685, "tokenization_path_logprob_std": 184.20158210694126, "trajectory_balance_rmse": 4472.344368775833} {"canonical_length": 8, "canonical_nll": 1.4724533557891846, "generation_rouge_l_f1": 0.4444444444444444, "generation_token_f1": 0.4444444444444444, "prompt_id": "2145da07c0ca4001", "reading_canonical_log_prob": -11.779626846313477, "reading_distances": [22, 22, 21, 25], "reading_log_probs": [-16.067649841308594, -16.51141357421875, -16.492712020874023, -15.496293067932129], "reading_logprob_std": 0.4767706287959099, "tokenization_path_logprob_std": 29.527953689062763, "trajectory_balance_rmse": 225.5002807494693} {"canonical_length": 272, "canonical_nll": 1.2857202642104204, "generation_rouge_l_f1": 0.22556390977443608, "generation_token_f1": 0.462406015037594, "prompt_id": "924cae405c1c753d", "reading_canonical_log_prob": -349.7159118652344, "reading_distances": [27, 22, 24, 21], "reading_log_probs": [-353.166015625, -352.43914794921875, -351.3874816894531, -354.3938293457031], "reading_logprob_std": 1.263723124497139, "tokenization_path_logprob_std": 83.1334184021605, "trajectory_balance_rmse": 5891.2213232095555} {"canonical_length": 147, "canonical_nll": 1.503115907007334, "generation_rouge_l_f1": 0.24937655860349123, "generation_token_f1": 0.4339152119700748, "prompt_id": "eb989912613f5027", "reading_canonical_log_prob": -220.95803833007812, "reading_distances": [30, 30, 34, 36], "reading_log_probs": [-217.25146484375, -219.03457641601562, -218.20599365234375, -219.1919708251953], "reading_logprob_std": 0.8916113209281803, "tokenization_path_logprob_std": 138.19895678907378, "trajectory_balance_rmse": 3098.2341893777466} {"canonical_length": 7, "canonical_nll": 1.1553238459995814, "generation_rouge_l_f1": 0.06756756756756757, "generation_token_f1": 0.06756756756756757, "prompt_id": "1a61687e59394e25", "reading_canonical_log_prob": -8.08726692199707, "reading_distances": [19, 19, 20, 19], "reading_log_probs": [-9.07168960571289, -9.129236221313477, -9.046019554138184, -8.876445770263672], "reading_logprob_std": 0.1086561908847881, "tokenization_path_logprob_std": 14.342806620600822, "trajectory_balance_rmse": 66.11201887338123} {"canonical_length": 12, "canonical_nll": 2.0162086486816406, "generation_rouge_l_f1": 0.36363636363636365, "generation_token_f1": 0.36363636363636365, "prompt_id": "687a259eb6971a8a", "reading_canonical_log_prob": -24.194503784179688, "reading_distances": [22, 21, 20, 22], "reading_log_probs": [-22.9938907623291, -23.59669303894043, -21.627578735351562, -21.13516616821289], "reading_logprob_std": 1.1497837502733512, "tokenization_path_logprob_std": 28.709474700151524, "trajectory_balance_rmse": 233.8894192461014} {"canonical_length": 177, "canonical_nll": 1.3282427599201094, "generation_rouge_l_f1": 0.24697336561743344, "generation_token_f1": 0.4261501210653753, "prompt_id": "ae453e9c5b7b5add", "reading_canonical_log_prob": -235.09896850585938, "reading_distances": [27, 28, 24, 25], "reading_log_probs": [-228.02679443359375, -226.7686767578125, -231.20272827148438, -230.3594970703125], "reading_logprob_std": 2.048917592724565, "tokenization_path_logprob_std": 183.76261832687513, "trajectory_balance_rmse": 3230.6714750450087} {"canonical_length": 5, "canonical_nll": 1.71339111328125, "generation_rouge_l_f1": 0.09523809523809525, "generation_token_f1": 0.1904761904761905, "prompt_id": "ffdd99b122d4c9bc", "reading_canonical_log_prob": -8.56695556640625, "reading_distances": [42, 43, 43, 40], "reading_log_probs": [-9.469666481018066, -9.920721054077148, -11.56623649597168, -11.024322509765625], "reading_logprob_std": 0.9676147405759368, "tokenization_path_logprob_std": 33.05879001837731, "trajectory_balance_rmse": 72.40483147928005} {"canonical_length": 350, "canonical_nll": 1.013180890764509, "generation_rouge_l_f1": 0.27177700348432055, "generation_token_f1": 0.5505226480836237, "prompt_id": "56f7657e17072dd3", "reading_canonical_log_prob": -354.6133117675781, "reading_distances": [16, 12, 11, 15], "reading_log_probs": [-351.22430419921875, -351.1996154785156, -351.090576171875, -349.630126953125], "reading_logprob_std": 0.772872862143965, "tokenization_path_logprob_std": 332.01670491975875, "trajectory_balance_rmse": 7118.239643267372} {"canonical_length": 172, "canonical_nll": 1.1590792633766351, "generation_rouge_l_f1": 0.3404255319148936, "generation_token_f1": 0.5673758865248227, "prompt_id": "f0c44899ac8f3454", "reading_canonical_log_prob": -199.36163330078125, "reading_distances": [26, 32, 27, 32], "reading_log_probs": [-195.64456176757812, -199.17465209960938, -197.09573364257812, -196.36212158203125], "reading_logprob_std": 1.5235032591881457, "tokenization_path_logprob_std": 251.88152893236455, "trajectory_balance_rmse": 3630.952665218525} {"canonical_length": 316, "canonical_nll": 1.137688938575455, "generation_rouge_l_f1": 0.236331569664903, "generation_token_f1": 0.46913580246913583, "prompt_id": "fb86a83720d2950d", "reading_canonical_log_prob": -359.50970458984375, "reading_distances": [25, 22, 27, 26], "reading_log_probs": [-360.1856994628906, -358.58349609375, -359.62896728515625, -360.8379211425781], "reading_logprob_std": 0.9547866784210729, "tokenization_path_logprob_std": 177.28047652366942, "trajectory_balance_rmse": 5742.336294893047} {"canonical_length": 60, "canonical_nll": 1.5838114420572917, "generation_rouge_l_f1": 0.13664596273291924, "generation_token_f1": 0.21739130434782608, "prompt_id": "4f164749bc77ca14", "reading_canonical_log_prob": -95.0286865234375, "reading_distances": [26, 25, 19, 23], "reading_log_probs": [-94.6139144897461, -93.10908508300781, -95.986572265625, -96.5623779296875], "reading_logprob_std": 1.5406147805252715, "tokenization_path_logprob_std": 67.4563015516497, "trajectory_balance_rmse": 1445.2506359951012} {"canonical_length": 161, "canonical_nll": 1.3852636680839965, "generation_rouge_l_f1": 0.2932692307692308, "generation_token_f1": 0.49038461538461536, "prompt_id": "8d15f97380395bd9", "reading_canonical_log_prob": -223.02745056152344, "reading_distances": [26, 31, 27, 30], "reading_log_probs": [-220.610107421875, -220.52481079101562, -220.07623291015625, -218.85592651367188], "reading_logprob_std": 0.8085477060685913, "tokenization_path_logprob_std": 219.80981560926924, "trajectory_balance_rmse": 2966.353734315686} {"canonical_length": 9, "canonical_nll": 0.570008807712131, "generation_rouge_l_f1": 0.8571428571428571, "generation_token_f1": 0.8571428571428571, "prompt_id": "a75869a5f1810362", "reading_canonical_log_prob": -5.13007926940918, "reading_distances": [44, 44, 46, 48], "reading_log_probs": [-6.6368560791015625, -7.115662574768066, -6.589241981506348, -6.921036720275879], "reading_logprob_std": 0.24788556034362336, "tokenization_path_logprob_std": 14.599022198883182, "trajectory_balance_rmse": 143.24641475447618} {"canonical_length": 228, "canonical_nll": 1.354229308011239, "generation_rouge_l_f1": 0.20491803278688525, "generation_token_f1": 0.5081967213114753, "prompt_id": "f3ee3068b8bfca4b", "reading_canonical_log_prob": -308.7642822265625, "reading_distances": [22, 24, 28, 26], "reading_log_probs": [-310.1556396484375, -309.9583740234375, -309.65869140625, -309.73077392578125], "reading_logprob_std": 0.22605116117587953, "tokenization_path_logprob_std": 90.45254077899529, "trajectory_balance_rmse": 4785.915665841034} {"canonical_length": 126, "canonical_nll": 1.3460017613002233, "generation_rouge_l_f1": 0.0851063829787234, "generation_token_f1": 0.1276595744680851, "prompt_id": "6a31e531aca13031", "reading_canonical_log_prob": -169.59622192382812, "reading_distances": [22, 30, 22, 21], "reading_log_probs": [-166.8680419921875, -166.22023010253906, -165.41412353515625, -165.4541778564453], "reading_logprob_std": 0.6934694843551109, "tokenization_path_logprob_std": 159.29621568044155, "trajectory_balance_rmse": 2305.4920532423444} {"canonical_length": 11, "canonical_nll": 0.4934746135364879, "generation_rouge_l_f1": 0.23529411764705882, "generation_token_f1": 0.23529411764705882, "prompt_id": "01a0ba7b191885ca", "reading_canonical_log_prob": -5.428220748901367, "reading_distances": [12, 6, 7, 9], "reading_log_probs": [-4.819305419921875, -5.47146463394165, -5.870434761047363, -5.111804008483887], "reading_logprob_std": 0.4545871298111258, "tokenization_path_logprob_std": 23.745486939566106, "trajectory_balance_rmse": 165.0590752711718} {"canonical_length": 9, "canonical_nll": 1.7196371290418837, "generation_rouge_l_f1": 0.056074766355140186, "generation_token_f1": 0.06542056074766356, "prompt_id": "8b612a049e9d2b42", "reading_canonical_log_prob": -15.476734161376953, "reading_distances": [16, 11, 14, 13], "reading_log_probs": [-13.951887130737305, -13.259292602539062, -13.665668487548828, -12.280096054077148], "reading_logprob_std": 0.7303122565629689, "tokenization_path_logprob_std": 35.91213731877897, "trajectory_balance_rmse": 157.73704665388487} {"canonical_length": 168, "canonical_nll": 0.5810379754929316, "generation_rouge_l_f1": 0.3076923076923077, "generation_token_f1": 0.48484848484848475, "prompt_id": "1792c7fe245165d5", "reading_canonical_log_prob": -97.6143798828125, "reading_distances": [53, 51, 52, 52], "reading_log_probs": [-101.72811889648438, -101.00890350341797, -101.63232421875, -101.30281066894531], "reading_logprob_std": 0.3279924389469751, "tokenization_path_logprob_std": 79.1527213222046, "trajectory_balance_rmse": 2280.441990302855} {"canonical_length": 240, "canonical_nll": 1.443689473470052, "generation_rouge_l_f1": 0.2625482625482626, "generation_token_f1": 0.5366795366795366, "prompt_id": "9148af04e8535f5f", "reading_canonical_log_prob": -346.4854736328125, "reading_distances": [24, 22, 23, 22], "reading_log_probs": [-340.68701171875, -341.72955322265625, -340.98736572265625, -341.443603515625], "reading_logprob_std": 0.46460546936519753, "tokenization_path_logprob_std": 188.84699496263485, "trajectory_balance_rmse": 4296.255496479061} {"canonical_length": 11, "canonical_nll": 0.847213398326527, "generation_rouge_l_f1": 0.10050251256281405, "generation_token_f1": 0.10050251256281405, "prompt_id": "4f964e48306df1f7", "reading_canonical_log_prob": -9.319347381591797, "reading_distances": [19, 16, 11, 12], "reading_log_probs": [-8.847152709960938, -8.744010925292969, -9.280438423156738, -8.291763305664062], "reading_logprob_std": 0.4058522909962521, "tokenization_path_logprob_std": 31.442404730489738, "trajectory_balance_rmse": 169.32983509261317} {"canonical_length": 13, "canonical_nll": 1.4908511822040265, "generation_rouge_l_f1": 0.4444444444444444, "generation_token_f1": 0.4444444444444444, "prompt_id": "65511a559eb975a4", "reading_canonical_log_prob": -19.381065368652344, "reading_distances": [22, 21, 18, 18], "reading_log_probs": [-19.148059844970703, -20.593528747558594, -21.2294921875, -21.309860229492188], "reading_logprob_std": 1.0007980631525113, "tokenization_path_logprob_std": 3.443454718662707, "trajectory_balance_rmse": 329.50103059166594} {"canonical_length": 273, "canonical_nll": 1.1518345647679145, "generation_rouge_l_f1": 0.20996441281138786, "generation_token_f1": 0.42348754448398573, "prompt_id": "dfd43bd8f8a6af69", "reading_canonical_log_prob": -314.4508361816406, "reading_distances": [16, 9, 16, 19], "reading_log_probs": [-312.3931884765625, -315.7712707519531, -314.13397216796875, -313.85821533203125], "reading_logprob_std": 1.3845769790864793, "tokenization_path_logprob_std": 287.0530631487271, "trajectory_balance_rmse": 4570.628439625579} {"canonical_length": 137, "canonical_nll": 1.9382497968464871, "generation_rouge_l_f1": 0.1687979539641944, "generation_token_f1": 0.30179028132992325, "prompt_id": "9ebcabae9c8c462a", "reading_canonical_log_prob": -265.54022216796875, "reading_distances": [13, 16, 16, 21], "reading_log_probs": [-259.3741760253906, -258.0743408203125, -256.21527099609375, -255.3578338623047], "reading_logprob_std": 1.8113072171336508, "tokenization_path_logprob_std": 359.2004057783103, "trajectory_balance_rmse": 3210.047797824224} {"canonical_length": 224, "canonical_nll": 1.0735630307878767, "generation_rouge_l_f1": 0.2922636103151862, "generation_token_f1": 0.42406876790830944, "prompt_id": "4e90bb5af9e629db", "reading_canonical_log_prob": -240.47811889648438, "reading_distances": [57, 54, 44, 47], "reading_log_probs": [-236.94668579101562, -240.2992401123047, -238.62550354003906, -244.24818420410156], "reading_logprob_std": 3.1275658798207298, "tokenization_path_logprob_std": 84.06101306926139, "trajectory_balance_rmse": 4289.900842940129} {"canonical_length": 17, "canonical_nll": 0.870783805847168, "generation_rouge_l_f1": 0.0979020979020979, "generation_token_f1": 0.10489510489510488, "prompt_id": "3779413f4fba856c", "reading_canonical_log_prob": -14.803324699401855, "reading_distances": [12, 16, 21, 18], "reading_log_probs": [-15.415830612182617, -15.639235496520996, -15.27034854888916, -15.253957748413086], "reading_logprob_std": 0.1784336428413928, "tokenization_path_logprob_std": 72.22581186821887, "trajectory_balance_rmse": 429.7738252371942} {"canonical_length": 125, "canonical_nll": 1.244675048828125, "generation_rouge_l_f1": 0.22085889570552147, "generation_token_f1": 0.30674846625766866, "prompt_id": "f749a091b4c13176", "reading_canonical_log_prob": -155.58438110351562, "reading_distances": [18, 22, 15, 19], "reading_log_probs": [-153.96231079101562, -156.03228759765625, -154.03602600097656, -153.04824829101562], "reading_logprob_std": 1.2580078138458317, "tokenization_path_logprob_std": 61.92329205718775, "trajectory_balance_rmse": 2616.7402484630406} {"canonical_length": 45, "canonical_nll": 1.5377912733289931, "generation_rouge_l_f1": 0.32608695652173914, "generation_token_f1": 0.43478260869565216, "prompt_id": "6378420685c9b069", "reading_canonical_log_prob": -69.20060729980469, "reading_distances": [23, 19, 22, 21], "reading_log_probs": [-69.50959014892578, -68.77755737304688, -69.55258178710938, -69.08380126953125], "reading_logprob_std": 0.36891956846066115, "tokenization_path_logprob_std": 148.52607313743783, "trajectory_balance_rmse": 1186.9190555019936} {"canonical_length": 119, "canonical_nll": 0.8767478365858062, "generation_rouge_l_f1": 0.25753424657534246, "generation_token_f1": 0.4000000000000001, "prompt_id": "634524f69adfb3f0", "reading_canonical_log_prob": -104.33299255371094, "reading_distances": [28, 25, 28, 25], "reading_log_probs": [-103.46019744873047, -104.16619110107422, -104.56430053710938, -104.48226928710938], "reading_logprob_std": 0.502263806716551, "tokenization_path_logprob_std": 120.32541885859449, "trajectory_balance_rmse": 1937.9360699364938} {"canonical_length": 162, "canonical_nll": 1.6887726960358795, "generation_rouge_l_f1": 0.2549450549450549, "generation_token_f1": 0.3516483516483516, "prompt_id": "c5c8a284983e4726", "reading_canonical_log_prob": -273.5811767578125, "reading_distances": [20, 22, 23, 19], "reading_log_probs": [-269.11138916015625, -269.31488037109375, -270.16986083984375, -272.2384033203125], "reading_logprob_std": 1.4287767074641002, "tokenization_path_logprob_std": 24.114497906792405, "trajectory_balance_rmse": 2813.1698961226357} {"canonical_length": 88, "canonical_nll": 1.3985320871526545, "generation_rouge_l_f1": 0.21700879765395897, "generation_token_f1": 0.3225806451612903, "prompt_id": "626ba72af7ac67f2", "reading_canonical_log_prob": -123.0708236694336, "reading_distances": [31, 26, 29, 27], "reading_log_probs": [-113.87664794921875, -117.45234680175781, -114.32601928710938, -117.25650024414062], "reading_logprob_std": 1.888803780645231, "tokenization_path_logprob_std": 57.86549838902636, "trajectory_balance_rmse": 1368.6106211995925} {"canonical_length": 297, "canonical_nll": 0.8862922231757681, "generation_rouge_l_f1": 0.3249097472924188, "generation_token_f1": 0.5415162454873647, "prompt_id": "c0257d516383b045", "reading_canonical_log_prob": -263.2287902832031, "reading_distances": [26, 29, 34, 23], "reading_log_probs": [-260.42584228515625, -259.8076477050781, -259.9652404785156, -261.8968505859375], "reading_logprob_std": 0.9521425729554678, "tokenization_path_logprob_std": 243.16508026437063, "trajectory_balance_rmse": 6008.848343253322} {"canonical_length": 38, "canonical_nll": 0.512853271082828, "generation_rouge_l_f1": 0.1415929203539823, "generation_token_f1": 0.22123893805309736, "prompt_id": "dae9d23ce12cb989", "reading_canonical_log_prob": -19.48842430114746, "reading_distances": [21, 14, 18, 13], "reading_log_probs": [-18.72380256652832, -16.027481079101562, -17.040700912475586, -19.966636657714844], "reading_logprob_std": 1.750054246973056, "tokenization_path_logprob_std": 37.88769255296836, "trajectory_balance_rmse": 285.18822450717016} {"canonical_length": 10, "canonical_nll": 0.7181251049041748, "generation_rouge_l_f1": 0.7058823529411765, "generation_token_f1": 0.823529411764706, "prompt_id": "cde09f27bb3893ac", "reading_canonical_log_prob": -7.181251049041748, "reading_distances": [29, 28, 22, 26], "reading_log_probs": [-8.145301818847656, -8.614970207214355, -7.561457633972168, -8.833459854125977], "reading_logprob_std": 0.5635187089932605, "tokenization_path_logprob_std": 35.51880328029828, "trajectory_balance_rmse": 207.28975283856164} {"canonical_length": 174, "canonical_nll": 1.6141473178205819, "generation_rouge_l_f1": 0.15942028985507248, "generation_token_f1": 0.3961352657004831, "prompt_id": "64bb51344f196fcb", "reading_canonical_log_prob": -280.86163330078125, "reading_distances": [27, 18, 20, 18], "reading_log_probs": [-283.396484375, -281.8294677734375, -279.35723876953125, -286.6257629394531], "reading_logprob_std": 3.043395763969174, "tokenization_path_logprob_std": 152.97778878753724, "trajectory_balance_rmse": 3304.7023055634636} {"canonical_length": 312, "canonical_nll": 1.4038214072203026, "generation_rouge_l_f1": 0.19064124783362218, "generation_token_f1": 0.36741767764298094, "prompt_id": "bc3436553628cc88", "reading_canonical_log_prob": -437.9922790527344, "reading_distances": [37, 30, 30, 31], "reading_log_probs": [-437.03204345703125, -436.1090087890625, -438.211181640625, -438.547607421875], "reading_logprob_std": 1.118718399252203, "tokenization_path_logprob_std": 416.49093662916164, "trajectory_balance_rmse": 7091.214800555561} {"canonical_length": 39, "canonical_nll": 1.318632566011869, "generation_rouge_l_f1": 0.3, "generation_token_f1": 0.35000000000000003, "prompt_id": "b3b9c2d52d21969e", "reading_canonical_log_prob": -51.42667007446289, "reading_distances": [27, 30, 29, 26], "reading_log_probs": [-48.891273498535156, -47.86377716064453, -49.14004135131836, -50.27973175048828], "reading_logprob_std": 0.9920530713675402, "tokenization_path_logprob_std": 33.402017208559506, "trajectory_balance_rmse": 899.844025121103} {"canonical_length": 21, "canonical_nll": 0.5521353767031715, "generation_rouge_l_f1": 0.25, "generation_token_f1": 0.375, "prompt_id": "cc11564e8f134b4b", "reading_canonical_log_prob": -11.594842910766602, "reading_distances": [15, 14, 14, 14], "reading_log_probs": [-10.59696102142334, -12.047910690307617, -12.984481811523438, -12.737483978271484], "reading_logprob_std": 1.0724206927860847, "tokenization_path_logprob_std": 17.82047992652392, "trajectory_balance_rmse": 168.63438094233152} {"canonical_length": 20, "canonical_nll": 1.2573779106140137, "generation_rouge_l_f1": 0.0892193308550186, "generation_token_f1": 0.11152416356877323, "prompt_id": "e07efb6f1797853d", "reading_canonical_log_prob": -25.147558212280273, "reading_distances": [27, 28, 25, 24], "reading_log_probs": [-28.63915252685547, -28.992660522460938, -30.881546020507812, -30.845624923706055], "reading_logprob_std": 1.1910944807054504, "tokenization_path_logprob_std": 58.017527696466196, "trajectory_balance_rmse": 349.5695770684603} {"canonical_length": 10, "canonical_nll": 0.11836612224578857, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "302fe191c6bb7b5b", "reading_canonical_log_prob": -1.1836612224578857, "reading_distances": [24, 30, 26, 22], "reading_log_probs": [-0.8575202822685242, -0.8681577444076538, -1.2440953254699707, -1.1052374839782715], "reading_logprob_std": 0.1887976428974201, "tokenization_path_logprob_std": 28.404102099575883, "trajectory_balance_rmse": 164.25079290678212} {"canonical_length": 22, "canonical_nll": 1.8060436248779297, "generation_rouge_l_f1": 0.34782608695652173, "generation_token_f1": 0.391304347826087, "prompt_id": "de0e448f061c6724", "reading_canonical_log_prob": -39.73295974731445, "reading_distances": [32, 25, 34, 28], "reading_log_probs": [-42.4980354309082, -41.45524597167969, -42.13498306274414, -40.786991119384766], "reading_logprob_std": 0.7567792971011322, "tokenization_path_logprob_std": 28.604024741049503, "trajectory_balance_rmse": 622.5026867948183} {"canonical_length": 79, "canonical_nll": 1.291098775742929, "generation_rouge_l_f1": 0.2466666666666667, "generation_token_f1": 0.3466666666666667, "prompt_id": "f979fa68d4ceeb01", "reading_canonical_log_prob": -101.9968032836914, "reading_distances": [26, 27, 25, 24], "reading_log_probs": [-105.24395751953125, -107.15424346923828, -107.32611083984375, -104.58155822753906], "reading_logprob_std": 1.3724717504261095, "tokenization_path_logprob_std": 59.26149098577346, "trajectory_balance_rmse": 1602.1878246017513} {"canonical_length": 193, "canonical_nll": 1.4033079789710168, "generation_rouge_l_f1": 0.2203023758099352, "generation_token_f1": 0.41900647948164144, "prompt_id": "e2f73d0c042e6f14", "reading_canonical_log_prob": -270.83843994140625, "reading_distances": [12, 13, 12, 16], "reading_log_probs": [-262.76055908203125, -264.4427185058594, -263.8892822265625, -264.2085876464844], "reading_logprob_std": 0.7451805102004698, "tokenization_path_logprob_std": 271.31116281423147, "trajectory_balance_rmse": 4252.192791233995} {"canonical_length": 216, "canonical_nll": 1.7809889051649306, "generation_rouge_l_f1": 0.23651452282157676, "generation_token_f1": 0.4481327800829876, "prompt_id": "409538e18660e490", "reading_canonical_log_prob": -384.693603515625, "reading_distances": [28, 38, 30, 31], "reading_log_probs": [-386.9240417480469, -384.5948486328125, -385.8453063964844, -385.9036560058594], "reading_logprob_std": 0.9535033484038725, "tokenization_path_logprob_std": 76.66865192107072, "trajectory_balance_rmse": 4883.822025245949} {"canonical_length": 64, "canonical_nll": 1.7885913848876953, "generation_rouge_l_f1": 0.25581395348837205, "generation_token_f1": 0.39534883720930236, "prompt_id": "cc38da16fec412a8", "reading_canonical_log_prob": -114.4698486328125, "reading_distances": [32, 31, 26, 30], "reading_log_probs": [-112.64259338378906, -113.5364761352539, -111.07972717285156, -113.8516845703125], "reading_logprob_std": 1.2423834049627034, "tokenization_path_logprob_std": 57.29357365934133, "trajectory_balance_rmse": 1107.2083718237132} {"canonical_length": 53, "canonical_nll": 0.7786579491957178, "generation_rouge_l_f1": 0.30588235294117644, "generation_token_f1": 0.36470588235294116, "prompt_id": "0e93241c78cc22bd", "reading_canonical_log_prob": -41.26887130737305, "reading_distances": [32, 28, 26, 35], "reading_log_probs": [-42.59223175048828, -41.6454963684082, -41.97443389892578, -41.23818588256836], "reading_logprob_std": 0.5721003083636134, "tokenization_path_logprob_std": 152.72463842629068, "trajectory_balance_rmse": 923.8504657709758} {"canonical_length": 19, "canonical_nll": 0.2848285373888518, "generation_rouge_l_f1": 0.31111111111111106, "generation_token_f1": 0.3333333333333333, "prompt_id": "32e3b4bcc84e208d", "reading_canonical_log_prob": -5.411742210388184, "reading_distances": [31, 28, 36, 29], "reading_log_probs": [-4.7614898681640625, -5.397385120391846, -6.2151875495910645, -5.107133865356445], "reading_logprob_std": 0.6203425844501089, "tokenization_path_logprob_std": 28.24716839153195, "trajectory_balance_rmse": 381.72480201070874} {"canonical_length": 28, "canonical_nll": 1.6203459330967493, "generation_rouge_l_f1": 0.20588235294117646, "generation_token_f1": 0.23529411764705885, "prompt_id": "08748791c0916706", "reading_canonical_log_prob": -45.369686126708984, "reading_distances": [22, 19, 23, 21], "reading_log_probs": [-47.462188720703125, -46.075782775878906, -45.8140869140625, -43.9759521484375], "reading_logprob_std": 1.4331991967660578, "tokenization_path_logprob_std": 133.68565147779373, "trajectory_balance_rmse": 918.0164757359339} {"canonical_length": 268, "canonical_nll": 1.4962980355789413, "generation_rouge_l_f1": 0.2283464566929134, "generation_token_f1": 0.3779527559055118, "prompt_id": "340800ab7642e87c", "reading_canonical_log_prob": -401.00787353515625, "reading_distances": [20, 20, 25, 20], "reading_log_probs": [-401.85809326171875, -403.4490051269531, -404.024658203125, -400.43731689453125], "reading_logprob_std": 1.6205521135195033, "tokenization_path_logprob_std": 260.3987808908154, "trajectory_balance_rmse": 4883.279696214937} {"canonical_length": 335, "canonical_nll": 1.205422654792444, "generation_rouge_l_f1": 0.23255813953488372, "generation_token_f1": 0.4551495016611295, "prompt_id": "ba9377f6ac227fec", "reading_canonical_log_prob": -403.81658935546875, "reading_distances": [21, 25, 30, 26], "reading_log_probs": [-400.6906433105469, -400.410400390625, -399.77142333984375, -400.1044921875], "reading_logprob_std": 0.39579886992408975, "tokenization_path_logprob_std": 80.23276984352762, "trajectory_balance_rmse": 7027.270631178276} {"canonical_length": 12, "canonical_nll": 0.17225046952565512, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.6666666666666666, "prompt_id": "39897df6fcbea057", "reading_canonical_log_prob": -2.0670056343078613, "reading_distances": [22, 20, 26, 18], "reading_log_probs": [-2.009881019592285, -2.2380683422088623, -2.780095100402832, -2.6541008949279785], "reading_logprob_std": 0.35859320773368225, "tokenization_path_logprob_std": 57.811324643805484, "trajectory_balance_rmse": 219.0630291452877} {"canonical_length": 55, "canonical_nll": 0.5862547440962358, "generation_rouge_l_f1": 0.3, "generation_token_f1": 0.325, "prompt_id": "d9f80153d939b659", "reading_canonical_log_prob": -32.24401092529297, "reading_distances": [21, 17, 22, 21], "reading_log_probs": [-32.93769836425781, -32.29958724975586, -33.70766830444336, -33.5799560546875], "reading_logprob_std": 0.6487732909182001, "tokenization_path_logprob_std": 63.427152142043894, "trajectory_balance_rmse": 1085.3988555850144} {"canonical_length": 105, "canonical_nll": 1.5128063383556547, "generation_rouge_l_f1": 0.20765027322404375, "generation_token_f1": 0.30601092896174864, "prompt_id": "ea105e2e33c793e5", "reading_canonical_log_prob": -158.84466552734375, "reading_distances": [34, 44, 32, 30], "reading_log_probs": [-162.98458862304688, -163.98703002929688, -162.7288818359375, -162.21585083007812], "reading_logprob_std": 0.7441127811835412, "tokenization_path_logprob_std": 48.83586503127592, "trajectory_balance_rmse": 2230.848198596668} {"canonical_length": 311, "canonical_nll": 1.4211414987251307, "generation_rouge_l_f1": 0.24727272727272728, "generation_token_f1": 0.48727272727272725, "prompt_id": "2fb44fa5d32cf514", "reading_canonical_log_prob": -441.9750061035156, "reading_distances": [33, 30, 30, 31], "reading_log_probs": [-449.81292724609375, -446.33880615234375, -449.18548583984375, -451.40142822265625], "reading_logprob_std": 2.114053443798602, "tokenization_path_logprob_std": 154.88024113781924, "trajectory_balance_rmse": 6138.569298895236} {"canonical_length": 134, "canonical_nll": 1.0627596271571829, "generation_rouge_l_f1": 0.2061855670103093, "generation_token_f1": 0.31958762886597936, "prompt_id": "f67931e5043d993b", "reading_canonical_log_prob": -142.4097900390625, "reading_distances": [27, 26, 23, 29], "reading_log_probs": [-141.5283203125, -141.3017120361328, -141.0458984375, -140.51962280273438], "reading_logprob_std": 0.4335534427452575, "tokenization_path_logprob_std": 211.52939171251205, "trajectory_balance_rmse": 2626.4202699210437} {"canonical_length": 42, "canonical_nll": 1.4288806915283203, "generation_rouge_l_f1": 0.1423948220064725, "generation_token_f1": 0.20711974110032363, "prompt_id": "f4dd2f85181ef8ef", "reading_canonical_log_prob": -60.01298904418945, "reading_distances": [23, 21, 17, 16], "reading_log_probs": [-58.77299499511719, -57.809226989746094, -58.13364028930664, -58.46953201293945], "reading_logprob_std": 0.4167116261239537, "tokenization_path_logprob_std": 42.01284088078276, "trajectory_balance_rmse": 958.1431306380765} {"canonical_length": 386, "canonical_nll": 0.8996692716766516, "generation_rouge_l_f1": 0.24555735056542807, "generation_token_f1": 0.46849757673667203, "prompt_id": "044b1de06d0cdd9b", "reading_canonical_log_prob": -347.2723388671875, "reading_distances": [44, 47, 39, 43], "reading_log_probs": [-345.0567626953125, -344.87774658203125, -344.7032775878906, -344.22454833984375], "reading_logprob_std": 0.3577555368637983, "tokenization_path_logprob_std": 313.62024718856554, "trajectory_balance_rmse": 6468.216978276122} {"canonical_length": 120, "canonical_nll": 0.798577880859375, "generation_rouge_l_f1": 0.3795620437956204, "generation_token_f1": 0.5182481751824818, "prompt_id": "c7dd76b068a7ae00", "reading_canonical_log_prob": -95.829345703125, "reading_distances": [41, 43, 39, 47], "reading_log_probs": [-97.3451919555664, -95.49507904052734, -95.92465209960938, -95.70209503173828], "reading_logprob_std": 0.8375330373591106, "tokenization_path_logprob_std": 211.89753778511565, "trajectory_balance_rmse": 2094.0412206161086} {"canonical_length": 35, "canonical_nll": 2.4727199009486607, "generation_rouge_l_f1": 0.2912621359223301, "generation_token_f1": 0.36893203883495146, "prompt_id": "2fa90c11cfdc9fd0", "reading_canonical_log_prob": -86.54519653320312, "reading_distances": [20, 19, 24, 27], "reading_log_probs": [-81.86204528808594, -89.11830139160156, -85.81546020507812, -84.47007751464844], "reading_logprob_std": 3.019511190268401, "tokenization_path_logprob_std": 30.29320936259115, "trajectory_balance_rmse": 673.2209320827371} {"canonical_length": 33, "canonical_nll": 1.2832851987896543, "generation_rouge_l_f1": 0.23076923076923078, "generation_token_f1": 0.25641025641025644, "prompt_id": "acac7a68437d400e", "reading_canonical_log_prob": -42.348411560058594, "reading_distances": [24, 27, 34, 27], "reading_log_probs": [-42.44937515258789, -42.51808547973633, -42.5555305480957, -43.07792282104492], "reading_logprob_std": 0.2884983993716429, "tokenization_path_logprob_std": 62.22116793366588, "trajectory_balance_rmse": 520.0230128303933} {"canonical_length": 41, "canonical_nll": 0.4243856290491616, "generation_rouge_l_f1": 0.41666666666666663, "generation_token_f1": 0.41666666666666663, "prompt_id": "98a669e9058b6ddc", "reading_canonical_log_prob": -17.399810791015625, "reading_distances": [12, 12, 9, 12], "reading_log_probs": [-21.35947036743164, -22.118759155273438, -21.042945861816406, -23.3232479095459], "reading_logprob_std": 1.014111856557934, "tokenization_path_logprob_std": 59.180952762833975, "trajectory_balance_rmse": 578.6672775348385} {"canonical_length": 133, "canonical_nll": 1.401227449115954, "generation_rouge_l_f1": 0.19895287958115185, "generation_token_f1": 0.35602094240837695, "prompt_id": "610adc3864696d51", "reading_canonical_log_prob": -186.36325073242188, "reading_distances": [21, 19, 21, 21], "reading_log_probs": [-183.95108032226562, -186.85256958007812, -186.40536499023438, -182.63510131835938], "reading_logprob_std": 2.007815407405782, "tokenization_path_logprob_std": 212.72646542420517, "trajectory_balance_rmse": 3318.7170642825736} {"canonical_length": 34, "canonical_nll": 0.6603871513815487, "generation_rouge_l_f1": 0.1897810218978102, "generation_token_f1": 0.2116788321167883, "prompt_id": "74b66fda46123ad6", "reading_canonical_log_prob": -22.453163146972656, "reading_distances": [38, 42, 44, 39], "reading_log_probs": [-24.716144561767578, -22.950214385986328, -24.329633712768555, -25.111970901489258], "reading_logprob_std": 0.9404174118275581, "tokenization_path_logprob_std": 91.22169838853542, "trajectory_balance_rmse": 673.727789044979} {"canonical_length": 23, "canonical_nll": 0.6694022883539614, "generation_rouge_l_f1": 0.3620689655172414, "generation_token_f1": 0.3620689655172414, "prompt_id": "9ce2e67b60ca246c", "reading_canonical_log_prob": -15.396252632141113, "reading_distances": [19, 18, 17, 19], "reading_log_probs": [-15.950736999511719, -16.236652374267578, -15.71798324584961, -15.592958450317383], "reading_logprob_std": 0.28327226797307564, "tokenization_path_logprob_std": 75.52591096545773, "trajectory_balance_rmse": 416.613555784896} {"canonical_length": 9, "canonical_nll": 1.0853868060641818, "generation_rouge_l_f1": 0.625, "generation_token_f1": 0.625, "prompt_id": "f37797ee8ed6a36e", "reading_canonical_log_prob": -9.768481254577637, "reading_distances": [42, 32, 42, 48], "reading_log_probs": [-7.917611598968506, -7.824918746948242, -9.683032035827637, -7.5605692863464355], "reading_logprob_std": 0.9695400220063128, "tokenization_path_logprob_std": 17.326060293959, "trajectory_balance_rmse": 170.6507389642331} {"canonical_length": 25, "canonical_nll": 1.1101441955566407, "generation_rouge_l_f1": 0.21848739495798317, "generation_token_f1": 0.28571428571428575, "prompt_id": "b9be7345d7f0f09d", "reading_canonical_log_prob": -27.753604888916016, "reading_distances": [29, 23, 24, 28], "reading_log_probs": [-26.62742042541504, -26.811073303222656, -27.436336517333984, -27.915250778198242], "reading_logprob_std": 0.5906283019848154, "tokenization_path_logprob_std": 13.10122834705311, "trajectory_balance_rmse": 568.834440492352} {"canonical_length": 102, "canonical_nll": 0.7686642665489047, "generation_rouge_l_f1": 0.4577464788732394, "generation_token_f1": 0.5492957746478874, "prompt_id": "101280c93f2c70da", "reading_canonical_log_prob": -78.40375518798828, "reading_distances": [18, 16, 10, 18], "reading_log_probs": [-78.68896484375, -78.44195556640625, -79.70085144042969, -80.68743896484375], "reading_logprob_std": 1.0279103534978316, "tokenization_path_logprob_std": 91.59847362418977, "trajectory_balance_rmse": 871.1440763670487} {"canonical_length": 309, "canonical_nll": 1.1452918191557948, "generation_rouge_l_f1": 0.30155979202772965, "generation_token_f1": 0.5753899480069324, "prompt_id": "bc039a3de3397314", "reading_canonical_log_prob": -353.8951721191406, "reading_distances": [35, 38, 27, 35], "reading_log_probs": [-355.9642028808594, -356.1975402832031, -354.29290771484375, -353.69384765625], "reading_logprob_std": 1.2334626621057427, "tokenization_path_logprob_std": 282.41480552970756, "trajectory_balance_rmse": 6777.617556686084} {"canonical_length": 144, "canonical_nll": 1.0635586844550238, "generation_rouge_l_f1": 0.22950819672131145, "generation_token_f1": 0.3879781420765027, "prompt_id": "3bf14cf65dcd8fce", "reading_canonical_log_prob": -153.15245056152344, "reading_distances": [35, 36, 35, 39], "reading_log_probs": [-151.59750366210938, -150.6301727294922, -151.19769287109375, -152.67169189453125], "reading_logprob_std": 0.8617816280785291, "tokenization_path_logprob_std": 160.89045309085867, "trajectory_balance_rmse": 2213.5959968633374} {"canonical_length": 38, "canonical_nll": 1.571286352057206, "generation_rouge_l_f1": 0.1044176706827309, "generation_token_f1": 0.15261044176706826, "prompt_id": "ab60c5be348f9273", "reading_canonical_log_prob": -59.70888137817383, "reading_distances": [15, 20, 25, 22], "reading_log_probs": [-59.69869613647461, -60.483551025390625, -60.09938049316406, -60.21754455566406], "reading_logprob_std": 0.3263507849283931, "tokenization_path_logprob_std": 101.97064826260012, "trajectory_balance_rmse": 928.4015989362875} {"canonical_length": 12, "canonical_nll": 2.717132886250814, "generation_rouge_l_f1": 0.3529411764705882, "generation_token_f1": 0.3529411764705882, "prompt_id": "8392d28ac48cc769", "reading_canonical_log_prob": -32.605594635009766, "reading_distances": [38, 35, 33, 32], "reading_log_probs": [-39.05052185058594, -39.579254150390625, -37.167903900146484, -36.319644927978516], "reading_logprob_std": 1.5394994466452778, "tokenization_path_logprob_std": 64.88559155719842, "trajectory_balance_rmse": 354.8683974908171} {"canonical_length": 90, "canonical_nll": 1.861128913031684, "generation_rouge_l_f1": 0.17079889807162535, "generation_token_f1": 0.2865013774104683, "prompt_id": "2a39144255d482e3", "reading_canonical_log_prob": -167.50160217285156, "reading_distances": [29, 29, 25, 23], "reading_log_probs": [-169.9988250732422, -167.5135498046875, -169.51675415039062, -170.8020782470703], "reading_logprob_std": 1.400401912455955, "tokenization_path_logprob_std": 151.5190030929711, "trajectory_balance_rmse": 1847.4094995257753} {"canonical_length": 292, "canonical_nll": 1.411132185426477, "generation_rouge_l_f1": 0.20503597122302156, "generation_token_f1": 0.4748201438848921, "prompt_id": "7938846720b7f660", "reading_canonical_log_prob": -412.05059814453125, "reading_distances": [20, 13, 22, 17], "reading_log_probs": [-412.8231201171875, -412.23040771484375, -411.6632080078125, -411.6356201171875], "reading_logprob_std": 0.5614793873522894, "tokenization_path_logprob_std": 257.55970964985295, "trajectory_balance_rmse": 5702.113579409293} {"canonical_length": 196, "canonical_nll": 1.5563721948740434, "generation_rouge_l_f1": 0.30845771144278605, "generation_token_f1": 0.5771144278606964, "prompt_id": "2a7c37e2ef87098f", "reading_canonical_log_prob": -305.0489501953125, "reading_distances": [27, 30, 28, 31], "reading_log_probs": [-306.3164978027344, -305.89404296875, -307.7502136230469, -306.35333251953125], "reading_logprob_std": 0.8084429037364244, "tokenization_path_logprob_std": 151.2509700236334, "trajectory_balance_rmse": 4446.876013720555} {"canonical_length": 284, "canonical_nll": 1.2828753834039393, "generation_rouge_l_f1": 0.27106227106227104, "generation_token_f1": 0.45421245421245415, "prompt_id": "f9f1a4c90d1c6936", "reading_canonical_log_prob": -364.33660888671875, "reading_distances": [27, 25, 28, 28], "reading_log_probs": [-360.15533447265625, -360.867919921875, -362.317626953125, -361.71368408203125], "reading_logprob_std": 0.9483957850485674, "tokenization_path_logprob_std": 235.870745453009, "trajectory_balance_rmse": 6217.953183323285} {"canonical_length": 149, "canonical_nll": 1.2524764297792577, "generation_rouge_l_f1": 0.25298329355608595, "generation_token_f1": 0.45346062052505964, "prompt_id": "4c7fd497c60d26bd", "reading_canonical_log_prob": -186.61898803710938, "reading_distances": [20, 22, 27, 23], "reading_log_probs": [-197.31301879882812, -187.3677215576172, -192.38246154785156, -190.70590209960938], "reading_logprob_std": 4.143018637643854, "tokenization_path_logprob_std": 124.56930643859448, "trajectory_balance_rmse": 2962.205215347789} {"canonical_length": 25, "canonical_nll": 1.15939208984375, "generation_rouge_l_f1": 0.2, "generation_token_f1": 0.3333333333333333, "prompt_id": "c2e4f5be5a2ad0b2", "reading_canonical_log_prob": -28.98480224609375, "reading_distances": [13, 17, 20, 16], "reading_log_probs": [-26.753721237182617, -27.23965835571289, -28.08656883239746, -26.58404541015625], "reading_logprob_std": 0.6736758497103509, "tokenization_path_logprob_std": 109.91110305823126, "trajectory_balance_rmse": 493.5985604758149} {"canonical_length": 14, "canonical_nll": 1.7279578617640905, "generation_rouge_l_f1": 0.030188679245283016, "generation_token_f1": 0.030188679245283016, "prompt_id": "619e2fb6adbffc57", "reading_canonical_log_prob": -24.191410064697266, "reading_distances": [36, 30, 30, 30], "reading_log_probs": [-25.156005859375, -24.583843231201172, -24.371280670166016, -24.234603881835938], "reading_logprob_std": 0.40599433594549234, "tokenization_path_logprob_std": 67.9210803197652, "trajectory_balance_rmse": 374.5881639677128} {"canonical_length": 136, "canonical_nll": 1.3704524320714615, "generation_rouge_l_f1": 0.25, "generation_token_f1": 0.5270270270270271, "prompt_id": "237ba288f6603bf1", "reading_canonical_log_prob": -186.38153076171875, "reading_distances": [18, 17, 18, 21], "reading_log_probs": [-184.11611938476562, -180.55941772460938, -181.1764373779297, -180.18276977539062], "reading_logprob_std": 1.7858949293657973, "tokenization_path_logprob_std": 201.28164582207907, "trajectory_balance_rmse": 2697.314992560849} {"canonical_length": 9, "canonical_nll": 0.9875580469767252, "generation_rouge_l_f1": 0.26666666666666666, "generation_token_f1": 0.26666666666666666, "prompt_id": "52afc3f6a37f4218", "reading_canonical_log_prob": -8.888022422790527, "reading_distances": [27, 26, 24, 22], "reading_log_probs": [-6.995357036590576, -6.584191799163818, -6.7546491622924805, -6.426525115966797], "reading_logprob_std": 0.24360945051285698, "tokenization_path_logprob_std": 28.840099263108314, "trajectory_balance_rmse": 218.2719776861845} {"canonical_length": 115, "canonical_nll": 1.606640625, "generation_rouge_l_f1": 0.1958762886597938, "generation_token_f1": 0.30412371134020616, "prompt_id": "0096b48ae77be6ff", "reading_canonical_log_prob": -184.763671875, "reading_distances": [35, 31, 30, 25], "reading_log_probs": [-191.29510498046875, -193.6297149658203, -192.07933044433594, -195.79238891601562], "reading_logprob_std": 1.9823939684178349, "tokenization_path_logprob_std": 137.71462877026218, "trajectory_balance_rmse": 2841.750584237005} {"canonical_length": 252, "canonical_nll": 1.047558860173301, "generation_rouge_l_f1": 0.30181086519114686, "generation_token_f1": 0.47887323943661975, "prompt_id": "2061c66b7116786e", "reading_canonical_log_prob": -263.9848327636719, "reading_distances": [31, 38, 33, 29], "reading_log_probs": [-259.48046875, -261.7440185546875, -258.05645751953125, -261.67376708984375], "reading_logprob_std": 1.7946678776808311, "tokenization_path_logprob_std": 246.76288931650706, "trajectory_balance_rmse": 5036.6757096268} {"canonical_length": 94, "canonical_nll": 1.3526478219539562, "generation_rouge_l_f1": 0.27891156462585037, "generation_token_f1": 0.39455782312925164, "prompt_id": "5622ae2789dbdaf4", "reading_canonical_log_prob": -127.14889526367188, "reading_distances": [18, 18, 16, 17], "reading_log_probs": [-127.11311340332031, -126.13348388671875, -127.19033813476562, -124.65143585205078], "reading_logprob_std": 1.1826850135468843, "tokenization_path_logprob_std": 77.4427289493034, "trajectory_balance_rmse": 1843.5767566388695} {"canonical_length": 279, "canonical_nll": 1.2949815975722445, "generation_rouge_l_f1": 0.23853211009174313, "generation_token_f1": 0.4660550458715596, "prompt_id": "531c2e72ca7ad121", "reading_canonical_log_prob": -361.29986572265625, "reading_distances": [27, 16, 22, 25], "reading_log_probs": [-356.41253662109375, -357.29937744140625, -356.1096496582031, -357.78802490234375], "reading_logprob_std": 0.7768186813358231, "tokenization_path_logprob_std": 163.90065449944882, "trajectory_balance_rmse": 5625.845053239874} {"canonical_length": 33, "canonical_nll": 0.6575829477021189, "generation_rouge_l_f1": 0.5517241379310344, "generation_token_f1": 0.689655172413793, "prompt_id": "50acd3d956ee080e", "reading_canonical_log_prob": -21.700237274169922, "reading_distances": [26, 23, 37, 36], "reading_log_probs": [-20.629457473754883, -20.713003158569336, -21.004554748535156, -19.39291000366211], "reading_logprob_std": 0.7130786856834808, "tokenization_path_logprob_std": 54.92993224587239, "trajectory_balance_rmse": 689.461081613923} {"canonical_length": 26, "canonical_nll": 1.1737882907573993, "generation_rouge_l_f1": 0.1518987341772152, "generation_token_f1": 0.16877637130801687, "prompt_id": "3fcc8e674911ba73", "reading_canonical_log_prob": -30.518495559692383, "reading_distances": [16, 15, 15, 15], "reading_log_probs": [-29.72910499572754, -30.817277908325195, -30.788366317749023, -28.405553817749023], "reading_logprob_std": 1.1384560793775944, "tokenization_path_logprob_std": 39.13058721836293, "trajectory_balance_rmse": 520.1338600075715} {"canonical_length": 347, "canonical_nll": 0.9530626456400847, "generation_rouge_l_f1": 0.2, "generation_token_f1": 0.4262295081967213, "prompt_id": "915d77693c098234", "reading_canonical_log_prob": -330.7127380371094, "reading_distances": [17, 13, 11, 17], "reading_log_probs": [-319.58831787109375, -325.4100341796875, -328.33367919921875, -324.4581298828125], "reading_logprob_std": 3.6350435469792615, "tokenization_path_logprob_std": 159.34747644427276, "trajectory_balance_rmse": 5628.97321254029} {"canonical_length": 16, "canonical_nll": 0.31243908405303955, "generation_rouge_l_f1": 0.4827586206896552, "generation_token_f1": 0.896551724137931, "prompt_id": "d4ba094ba07a6bba", "reading_canonical_log_prob": -4.999025344848633, "reading_distances": [21, 21, 28, 21], "reading_log_probs": [-3.9427947998046875, -3.7457008361816406, -3.9266929626464844, -3.9354684352874756], "reading_logprob_std": 0.09487091041088541, "tokenization_path_logprob_std": 40.38669367716469, "trajectory_balance_rmse": 250.56421055300578} {"canonical_length": 35, "canonical_nll": 1.0268731253487724, "generation_rouge_l_f1": 0.16, "generation_token_f1": 0.22, "prompt_id": "40f13ac2314179d4", "reading_canonical_log_prob": -35.94055938720703, "reading_distances": [20, 25, 28, 30], "reading_log_probs": [-34.30855178833008, -34.762752532958984, -34.70677185058594, -34.6610107421875], "reading_logprob_std": 0.20507806570283094, "tokenization_path_logprob_std": 107.14464235237442, "trajectory_balance_rmse": 786.0713130303907} {"canonical_length": 228, "canonical_nll": 1.1226878919099506, "generation_rouge_l_f1": 0.2251082251082251, "generation_token_f1": 0.43722943722943725, "prompt_id": "c60f4d0722caa17f", "reading_canonical_log_prob": -255.97283935546875, "reading_distances": [62, 53, 67, 59], "reading_log_probs": [-246.68569946289062, -244.76718139648438, -246.19517517089844, -247.64212036132812], "reading_logprob_std": 1.1984071804753795, "tokenization_path_logprob_std": 189.52792760516294, "trajectory_balance_rmse": 4646.086410697275} {"canonical_length": 20, "canonical_nll": 1.811983871459961, "generation_rouge_l_f1": 0.13930348258706468, "generation_token_f1": 0.1890547263681592, "prompt_id": "2ca8b957d3b2b5bc", "reading_canonical_log_prob": -36.23967742919922, "reading_distances": [24, 23, 30, 30], "reading_log_probs": [-38.34539031982422, -38.879451751708984, -37.774742126464844, -36.93472671508789], "reading_logprob_std": 0.8321056370666714, "tokenization_path_logprob_std": 57.00503702788757, "trajectory_balance_rmse": 470.5114746531338} {"canonical_length": 159, "canonical_nll": 1.3193582978638463, "generation_rouge_l_f1": 0.3225806451612903, "generation_token_f1": 0.5483870967741935, "prompt_id": "762bfb8c99f50dd6", "reading_canonical_log_prob": -209.77796936035156, "reading_distances": [22, 20, 31, 20], "reading_log_probs": [-205.04579162597656, -208.69667053222656, -204.833984375, -206.21524047851562], "reading_logprob_std": 1.7731116152722666, "tokenization_path_logprob_std": 104.06532077123678, "trajectory_balance_rmse": 2924.85239518443} {"canonical_length": 231, "canonical_nll": 1.0926952279491342, "generation_rouge_l_f1": 0.2620967741935484, "generation_token_f1": 0.5201612903225806, "prompt_id": "9f2ef447478c22f4", "reading_canonical_log_prob": -252.41259765625, "reading_distances": [29, 27, 19, 24], "reading_log_probs": [-251.8529510498047, -250.7333526611328, -247.2368927001953, -249.5673828125], "reading_logprob_std": 1.9748674958968966, "tokenization_path_logprob_std": 208.72894125210217, "trajectory_balance_rmse": 4785.225151004353} {"canonical_length": 361, "canonical_nll": 1.1883509416659452, "generation_rouge_l_f1": 0.2727272727272727, "generation_token_f1": 0.4882154882154882, "prompt_id": "dc0376ea25a8c1bc", "reading_canonical_log_prob": -428.99468994140625, "reading_distances": [27, 26, 30, 30], "reading_log_probs": [-428.608642578125, -430.9165344238281, -425.0816650390625, -425.651611328125], "reading_logprob_std": 2.717227813099078, "tokenization_path_logprob_std": 170.0878908849639, "trajectory_balance_rmse": 5916.617774745353} {"canonical_length": 18, "canonical_nll": 3.179303910997179, "generation_rouge_l_f1": 0.14285714285714288, "generation_token_f1": 0.21428571428571427, "prompt_id": "9cf805580809e94d", "reading_canonical_log_prob": -57.22747039794922, "reading_distances": [31, 22, 23, 20], "reading_log_probs": [-59.96025466918945, -58.966064453125, -59.10604476928711, -59.49648666381836], "reading_logprob_std": 0.4459593203729616, "tokenization_path_logprob_std": 31.516204602686948, "trajectory_balance_rmse": 395.98333266548514} {"canonical_length": 280, "canonical_nll": 1.6031223842075892, "generation_rouge_l_f1": 0.2715596330275229, "generation_token_f1": 0.45504587155963305, "prompt_id": "d3ad116b6d046506", "reading_canonical_log_prob": -448.874267578125, "reading_distances": [24, 22, 22, 24], "reading_log_probs": [-446.79351806640625, -445.59002685546875, -446.5399169921875, -447.75921630859375], "reading_logprob_std": 0.891611026596697, "tokenization_path_logprob_std": 153.73078159149716, "trajectory_balance_rmse": 6604.989819992824} {"canonical_length": 141, "canonical_nll": 1.322394404850953, "generation_rouge_l_f1": 0.31932773109243695, "generation_token_f1": 0.48179271708683474, "prompt_id": "d4c38137aba16167", "reading_canonical_log_prob": -186.45761108398438, "reading_distances": [12, 14, 9, 16], "reading_log_probs": [-196.19644165039062, -205.7933807373047, -197.50238037109375, -200.80972290039062], "reading_logprob_std": 4.277899862333277, "tokenization_path_logprob_std": 36.140888862537224, "trajectory_balance_rmse": 1365.2918338077745} {"canonical_length": 41, "canonical_nll": 1.6375881288109757, "generation_rouge_l_f1": 0.2121212121212121, "generation_token_f1": 0.24242424242424243, "prompt_id": "84547b634d24f63a", "reading_canonical_log_prob": -67.14111328125, "reading_distances": [24, 28, 29, 24], "reading_log_probs": [-66.68360137939453, -65.96870422363281, -66.24359893798828, -67.23016357421875], "reading_logprob_std": 0.55102683715591, "tokenization_path_logprob_std": 158.88333445464198, "trajectory_balance_rmse": 998.7466794445506} {"canonical_length": 6, "canonical_nll": 0.07956103980541229, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "58c9afa198f60472", "reading_canonical_log_prob": -0.47736623883247375, "reading_distances": [43, 40, 52, 40], "reading_log_probs": [-1.8897755146026611, -1.8987720012664795, -5.898736953735352, -2.9485011100769043], "reading_logprob_std": 1.8929319190039504, "tokenization_path_logprob_std": 11.903931731179933, "trajectory_balance_rmse": 60.22500620546513} {"canonical_length": 238, "canonical_nll": 1.60134220924698, "generation_rouge_l_f1": 0.22924901185770752, "generation_token_f1": 0.533596837944664, "prompt_id": "74608c6aad72d41d", "reading_canonical_log_prob": -381.11944580078125, "reading_distances": [37, 39, 36, 32], "reading_log_probs": [-381.1774597167969, -380.33642578125, -381.2998962402344, -380.8819274902344], "reading_logprob_std": 0.4291671581647911, "tokenization_path_logprob_std": 204.1719055068716, "trajectory_balance_rmse": 5135.900342591525} {"canonical_length": 296, "canonical_nll": 1.4814306723104942, "generation_rouge_l_f1": 0.2681564245810056, "generation_token_f1": 0.5400372439478585, "prompt_id": "5c242c11a7bd0e44", "reading_canonical_log_prob": -438.50347900390625, "reading_distances": [26, 35, 28, 36], "reading_log_probs": [-435.72735595703125, -433.17645263671875, -433.6983642578125, -434.03399658203125], "reading_logprob_std": 1.1034719243300375, "tokenization_path_logprob_std": 233.22232978673787, "trajectory_balance_rmse": 6797.8478563906065} {"canonical_length": 42, "canonical_nll": 0.6037926446823847, "generation_rouge_l_f1": 0.5376344086021505, "generation_token_f1": 0.6021505376344085, "prompt_id": "c3513f19426dd701", "reading_canonical_log_prob": -25.359291076660156, "reading_distances": [23, 27, 26, 33], "reading_log_probs": [-25.075641632080078, -25.782257080078125, -23.50450325012207, -26.257530212402344], "reading_logprob_std": 1.2026999604410276, "tokenization_path_logprob_std": 28.595581488025683, "trajectory_balance_rmse": 340.94463735104426} {"canonical_length": 356, "canonical_nll": 1.130577687467082, "generation_rouge_l_f1": 0.1870748299319728, "generation_token_f1": 0.391156462585034, "prompt_id": "80a15e1bafbe7e14", "reading_canonical_log_prob": -402.48565673828125, "reading_distances": [20, 20, 17, 16], "reading_log_probs": [-398.1983642578125, -399.83203125, -398.9786071777344, -398.89691162109375], "reading_logprob_std": 0.669270174995456, "tokenization_path_logprob_std": 285.14307080084507, "trajectory_balance_rmse": 6981.912673177747} {"canonical_length": 51, "canonical_nll": 1.6726762360217524, "generation_rouge_l_f1": 0.14035087719298248, "generation_token_f1": 0.1754385964912281, "prompt_id": "74f44ff0ad2c7451", "reading_canonical_log_prob": -85.30648803710938, "reading_distances": [23, 25, 25, 22], "reading_log_probs": [-81.55343627929688, -80.3666763305664, -79.83793640136719, -80.77063751220703], "reading_logprob_std": 0.7232331820840392, "tokenization_path_logprob_std": 80.84666715811365, "trajectory_balance_rmse": 1181.0766926002116} {"canonical_length": 21, "canonical_nll": 1.2616047632126581, "generation_rouge_l_f1": 0.22857142857142854, "generation_token_f1": 0.2857142857142857, "prompt_id": "0b30260636e7066b", "reading_canonical_log_prob": -26.49370002746582, "reading_distances": [32, 34, 23, 31], "reading_log_probs": [-32.06499481201172, -29.671218872070312, -26.748397827148438, -28.49531364440918], "reading_logprob_std": 2.2307760463216733, "tokenization_path_logprob_std": 83.31951324695919, "trajectory_balance_rmse": 544.337942208327} {"canonical_length": 142, "canonical_nll": 1.1713465301083847, "generation_rouge_l_f1": 0.27956989247311825, "generation_token_f1": 0.5017921146953406, "prompt_id": "270a3b8ee410411a", "reading_canonical_log_prob": -166.33120727539062, "reading_distances": [24, 31, 31, 29], "reading_log_probs": [-162.71768188476562, -163.01797485351562, -163.65122985839844, -161.2497100830078], "reading_logprob_std": 1.017009436547957, "tokenization_path_logprob_std": 107.6901988245299, "trajectory_balance_rmse": 2420.4738383919166} {"canonical_length": 2, "canonical_nll": 1.1362297534942627, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "aa51f5831004ecdf", "reading_canonical_log_prob": -2.2724595069885254, "reading_distances": [35, 33, 39, 28], "reading_log_probs": [-1.7995291948318481, -1.9739797115325928, -1.9718760251998901, -2.0926544666290283], "reading_logprob_std": 0.12066979482202499, "tokenization_path_logprob_std": 25.92881243364339, "trajectory_balance_rmse": 44.30613559162145} {"canonical_length": 53, "canonical_nll": 0.7795804581552181, "generation_rouge_l_f1": 0.3177570093457944, "generation_token_f1": 0.42990654205607476, "prompt_id": "d23e5c49d3b87636", "reading_canonical_log_prob": -41.31776428222656, "reading_distances": [20, 16, 18, 13], "reading_log_probs": [-39.50801086425781, -40.39365005493164, -39.56306838989258, -40.36578369140625], "reading_logprob_std": 0.4880365954499054, "tokenization_path_logprob_std": 113.26908849489367, "trajectory_balance_rmse": 826.5733746942544} {"canonical_length": 78, "canonical_nll": 1.7084988325070112, "generation_rouge_l_f1": 0.23834196891191708, "generation_token_f1": 0.4145077720207254, "prompt_id": "1cf9ae29e6e4ee3d", "reading_canonical_log_prob": -133.26290893554688, "reading_distances": [19, 22, 27, 21], "reading_log_probs": [-135.43911743164062, -134.06932067871094, -134.4582977294922, -135.29141235351562], "reading_logprob_std": 0.6582209866365991, "tokenization_path_logprob_std": 86.94584027910904, "trajectory_balance_rmse": 1594.9737747139795} {"canonical_length": 113, "canonical_nll": 1.3614597826932384, "generation_rouge_l_f1": 0.24164524421593833, "generation_token_f1": 0.3341902313624679, "prompt_id": "4b26a788bce312c3", "reading_canonical_log_prob": -153.84495544433594, "reading_distances": [27, 37, 31, 34], "reading_log_probs": [-157.1750030517578, -156.27212524414062, -157.68222045898438, -156.77122497558594], "reading_logprob_std": 0.5988096801227619, "tokenization_path_logprob_std": 207.6550118064754, "trajectory_balance_rmse": 2303.1801064709366} {"canonical_length": 320, "canonical_nll": 1.5824939727783203, "generation_rouge_l_f1": 0.271304347826087, "generation_token_f1": 0.5043478260869566, "prompt_id": "c59752dc85c30f5b", "reading_canonical_log_prob": -506.3980712890625, "reading_distances": [35, 38, 41, 35], "reading_log_probs": [-507.405517578125, -500.2492370605469, -510.4272766113281, -503.19281005859375], "reading_logprob_std": 4.497081465988604, "tokenization_path_logprob_std": 253.62959427524248, "trajectory_balance_rmse": 7640.382972034167} {"canonical_length": 9, "canonical_nll": 1.7981541951497395, "generation_rouge_l_f1": 0.5454545454545454, "generation_token_f1": 0.5454545454545454, "prompt_id": "b4e868a234198084", "reading_canonical_log_prob": -16.183387756347656, "reading_distances": [25, 32, 30, 31], "reading_log_probs": [-18.777362823486328, -21.90258026123047, -19.71558952331543, -20.232927322387695], "reading_logprob_std": 1.3103507547655469, "tokenization_path_logprob_std": 37.21149273732877, "trajectory_balance_rmse": 262.56656058534446} {"canonical_length": 68, "canonical_nll": 1.8553302989286535, "generation_rouge_l_f1": 0.2079207920792079, "generation_token_f1": 0.3465346534653465, "prompt_id": "466e7627e6013e86", "reading_canonical_log_prob": -126.16246032714844, "reading_distances": [33, 36, 37, 37], "reading_log_probs": [-118.43620300292969, -119.19046020507812, -121.74092102050781, -120.3597640991211], "reading_logprob_std": 1.4425055979856751, "tokenization_path_logprob_std": 117.46447919280929, "trajectory_balance_rmse": 1517.375220323189} {"canonical_length": 14, "canonical_nll": 0.9442655699593681, "generation_rouge_l_f1": 0.4166666666666667, "generation_token_f1": 0.4166666666666667, "prompt_id": "ca45cac53e43d02f", "reading_canonical_log_prob": -13.219717979431152, "reading_distances": [24, 24, 23, 29], "reading_log_probs": [-13.131240844726562, -12.995945930480957, -12.777958869934082, -13.756982803344727], "reading_logprob_std": 0.4203023920270856, "tokenization_path_logprob_std": 49.2538708220235, "trajectory_balance_rmse": 190.06701547110575} {"canonical_length": 20, "canonical_nll": 2.3555875778198243, "generation_rouge_l_f1": 0.12612612612612611, "generation_token_f1": 0.14414414414414414, "prompt_id": "685a9c22a80a27a9", "reading_canonical_log_prob": -47.111751556396484, "reading_distances": [24, 21, 20, 16], "reading_log_probs": [-44.495357513427734, -41.33784866333008, -43.87603759765625, -39.18800354003906], "reading_logprob_std": 2.442050469400892, "tokenization_path_logprob_std": 45.650433016689604, "trajectory_balance_rmse": 427.1723925848645} {"canonical_length": 12, "canonical_nll": 1.0480697949727376, "generation_rouge_l_f1": 0.7000000000000001, "generation_token_f1": 0.7000000000000001, "prompt_id": "426d94bf331d4f8d", "reading_canonical_log_prob": -12.576837539672852, "reading_distances": [23, 26, 23, 31], "reading_log_probs": [-11.434111595153809, -9.89787769317627, -11.889115333557129, -9.488027572631836], "reading_logprob_std": 1.163776643937977, "tokenization_path_logprob_std": 48.334400383256444, "trajectory_balance_rmse": 152.33984519182945} {"canonical_length": 123, "canonical_nll": 1.1438648254890753, "generation_rouge_l_f1": 0.21316614420062696, "generation_token_f1": 0.36990595611285265, "prompt_id": "efba508a45047679", "reading_canonical_log_prob": -140.69537353515625, "reading_distances": [29, 34, 36, 30], "reading_log_probs": [-139.26980590820312, -143.13882446289062, -143.47406005859375, -142.38624572753906], "reading_logprob_std": 1.919621324669558, "tokenization_path_logprob_std": 105.53937568545605, "trajectory_balance_rmse": 2385.8158222718803} {"canonical_length": 138, "canonical_nll": 1.1184588446133379, "generation_rouge_l_f1": 0.26763990267639903, "generation_token_f1": 0.4330900243309002, "prompt_id": "79eb49a82dc5be50", "reading_canonical_log_prob": -154.34732055664062, "reading_distances": [20, 28, 20, 15], "reading_log_probs": [-156.42242431640625, -153.34689331054688, -155.45437622070312, -156.47341918945312], "reading_logprob_std": 1.4621247501753474, "tokenization_path_logprob_std": 192.59466894706216, "trajectory_balance_rmse": 2747.733760124839} {"canonical_length": 355, "canonical_nll": 1.1860569913622359, "generation_rouge_l_f1": 0.27586206896551724, "generation_token_f1": 0.48589341692789967, "prompt_id": "144d91c0bdbf405c", "reading_canonical_log_prob": -421.05023193359375, "reading_distances": [30, 28, 30, 35], "reading_log_probs": [-419.6558837890625, -418.8951721191406, -417.9324951171875, -419.0233154296875], "reading_logprob_std": 0.711919363450826, "tokenization_path_logprob_std": 250.4428548360177, "trajectory_balance_rmse": 6080.382614490309} {"canonical_length": 194, "canonical_nll": 1.1730552751993395, "generation_rouge_l_f1": 0.3059866962305986, "generation_token_f1": 0.5454545454545454, "prompt_id": "6bebc48cdc1b16ab", "reading_canonical_log_prob": -227.57272338867188, "reading_distances": [58, 61, 53, 54], "reading_log_probs": [-225.32192993164062, -223.59185791015625, -227.3280792236328, -223.708984375], "reading_logprob_std": 1.7485778360669375, "tokenization_path_logprob_std": 124.27286676133232, "trajectory_balance_rmse": 3515.3094445692677} {"canonical_length": 136, "canonical_nll": 2.7020797729492188, "generation_rouge_l_f1": 0.2116788321167883, "generation_token_f1": 0.35036496350364965, "prompt_id": "f19548b23e650d6d", "reading_canonical_log_prob": -367.48284912109375, "reading_distances": [20, 12, 22, 19], "reading_log_probs": [-392.81890869140625, -375.8885498046875, -383.1007080078125, -382.5884094238281], "reading_logprob_std": 6.969632860523465, "tokenization_path_logprob_std": 172.69661222184652, "trajectory_balance_rmse": 2843.055967162383} {"canonical_length": 283, "canonical_nll": 1.0796370759027165, "generation_rouge_l_f1": 0.24772313296903462, "generation_token_f1": 0.4808743169398907, "prompt_id": "b85ad6ca4cfd648f", "reading_canonical_log_prob": -305.53729248046875, "reading_distances": [25, 23, 24, 23], "reading_log_probs": [-300.927978515625, -298.7597351074219, -300.66717529296875, -299.449462890625], "reading_logprob_std": 1.0227488378025056, "tokenization_path_logprob_std": 274.0285621086662, "trajectory_balance_rmse": 5165.749730961689} {"canonical_length": 146, "canonical_nll": 1.4937653214964148, "generation_rouge_l_f1": 0.22622107969151672, "generation_token_f1": 0.47300771208226217, "prompt_id": "6b7b364a0fc8e714", "reading_canonical_log_prob": -218.08973693847656, "reading_distances": [33, 22, 34, 27], "reading_log_probs": [-210.86073303222656, -217.78460693359375, -210.28814697265625, -211.05064392089844], "reading_logprob_std": 3.5405804860885253, "tokenization_path_logprob_std": 90.1888728740046, "trajectory_balance_rmse": 3026.682386276933} {"canonical_length": 112, "canonical_nll": 2.6521699087960378, "generation_rouge_l_f1": 0.13917525773195877, "generation_token_f1": 0.24742268041237114, "prompt_id": "1d91ffc1f9553c07", "reading_canonical_log_prob": -297.04302978515625, "reading_distances": [12, 16, 19, 16], "reading_log_probs": [-294.4949035644531, -295.1648254394531, -295.60479736328125, -294.0758972167969], "reading_logprob_std": 0.6814874362030416, "tokenization_path_logprob_std": 150.75534980586295, "trajectory_balance_rmse": 2258.9643970585344} {"canonical_length": 37, "canonical_nll": 0.9031132878483953, "generation_rouge_l_f1": 0.5526315789473685, "generation_token_f1": 0.6052631578947367, "prompt_id": "1ac7a2e92a59b9ce", "reading_canonical_log_prob": -33.415191650390625, "reading_distances": [13, 20, 20, 13], "reading_log_probs": [-34.09454345703125, -34.30622100830078, -34.025630950927734, -34.61072540283203], "reading_logprob_std": 0.2629621874931602, "tokenization_path_logprob_std": 101.04206198680943, "trajectory_balance_rmse": 867.6773406799879} {"canonical_length": 43, "canonical_nll": 2.161134586777798, "generation_rouge_l_f1": 0.19095477386934673, "generation_token_f1": 0.2814070351758794, "prompt_id": "66c94013abca2b3a", "reading_canonical_log_prob": -92.92878723144531, "reading_distances": [36, 24, 28, 26], "reading_log_probs": [-81.61709594726562, -84.62297058105469, -88.02487182617188, -84.34203338623047], "reading_logprob_std": 2.625758855999176, "tokenization_path_logprob_std": 125.94345264772777, "trajectory_balance_rmse": 841.0940267070879} {"canonical_length": 15, "canonical_nll": 0.266204563776652, "generation_rouge_l_f1": 0.8571428571428571, "generation_token_f1": 1.0, "prompt_id": "785090c8f7a0ea91", "reading_canonical_log_prob": -3.9930684566497803, "reading_distances": [17, 16, 18, 13], "reading_log_probs": [-3.665501594543457, -3.6819565296173096, -3.8946194648742676, -3.65236759185791], "reading_logprob_std": 0.11464629106863236, "tokenization_path_logprob_std": 43.85347182188314, "trajectory_balance_rmse": 347.48899769168366} {"canonical_length": 217, "canonical_nll": 1.1358726958525345, "generation_rouge_l_f1": 0.20087336244541484, "generation_token_f1": 0.34934497816593885, "prompt_id": "30ed32c8494397b3", "reading_canonical_log_prob": -246.484375, "reading_distances": [23, 23, 22, 15], "reading_log_probs": [-236.50946044921875, -238.12713623046875, -239.52456665039062, -239.29737854003906], "reading_logprob_std": 1.3800398816045036, "tokenization_path_logprob_std": 92.91253355752602, "trajectory_balance_rmse": 3539.108915380093} {"canonical_length": 16, "canonical_nll": 1.175022840499878, "generation_rouge_l_f1": 0.23529411764705882, "generation_token_f1": 0.23529411764705882, "prompt_id": "e8d7fb5246616f5e", "reading_canonical_log_prob": -18.800365447998047, "reading_distances": [79, 78, 81, 78], "reading_log_probs": [-16.49732208251953, -15.919503211975098, -16.145282745361328, -16.676776885986328], "reading_logprob_std": 0.34119107612360666, "tokenization_path_logprob_std": 32.686827645692595, "trajectory_balance_rmse": 302.5511978406031} {"canonical_length": 6, "canonical_nll": 1.9448076883951824, "generation_rouge_l_f1": 0.4000000000000001, "generation_token_f1": 0.4000000000000001, "prompt_id": "96f83a49dbeabc97", "reading_canonical_log_prob": -11.668846130371094, "reading_distances": [31, 28, 32, 30], "reading_log_probs": [-11.436202049255371, -10.767417907714844, -10.639020919799805, -10.727275848388672], "reading_logprob_std": 0.3664278513869034, "tokenization_path_logprob_std": 8.930217935263412, "trajectory_balance_rmse": 112.19167405582606} {"canonical_length": 80, "canonical_nll": 1.2241522789001464, "generation_rouge_l_f1": 0.026666666666666665, "generation_token_f1": 0.026666666666666665, "prompt_id": "dea31d77a681f384", "reading_canonical_log_prob": -97.93218231201172, "reading_distances": [13, 18, 15, 14], "reading_log_probs": [-98.31733703613281, -97.98738098144531, -98.42787170410156, -98.00480651855469], "reading_logprob_std": 0.2221268617586844, "tokenization_path_logprob_std": 110.96263128129841, "trajectory_balance_rmse": 1446.7407749641006} {"canonical_length": 9, "canonical_nll": 1.4648265838623047, "generation_rouge_l_f1": 0.16666666666666666, "generation_token_f1": 0.16666666666666666, "prompt_id": "428bb97daf254741", "reading_canonical_log_prob": -13.183439254760742, "reading_distances": [21, 22, 25, 22], "reading_log_probs": [-12.46969223022461, -11.106855392456055, -11.48387336730957, -11.862947463989258], "reading_logprob_std": 0.5812929004228149, "tokenization_path_logprob_std": 75.63211881295689, "trajectory_balance_rmse": 274.2523610156866} {"canonical_length": 249, "canonical_nll": 1.4248204978115588, "generation_rouge_l_f1": 0.24081632653061225, "generation_token_f1": 0.4857142857142857, "prompt_id": "ed06fac728bea8f5", "reading_canonical_log_prob": -354.7803039550781, "reading_distances": [21, 18, 19, 17], "reading_log_probs": [-349.03985595703125, -351.5741271972656, -351.8782958984375, -351.8348388671875], "reading_logprob_std": 1.3678931237273253, "tokenization_path_logprob_std": 180.55587366651568, "trajectory_balance_rmse": 5426.313596975302} {"canonical_length": 42, "canonical_nll": 1.8166549319312686, "generation_rouge_l_f1": 0.18666666666666668, "generation_token_f1": 0.3466666666666666, "prompt_id": "2455341860ece852", "reading_canonical_log_prob": -76.29950714111328, "reading_distances": [18, 16, 24, 16], "reading_log_probs": [-78.43753051757812, -77.84719848632812, -79.73091125488281, -77.35294342041016], "reading_logprob_std": 1.0265251943848368, "tokenization_path_logprob_std": 96.55578306334401, "trajectory_balance_rmse": 883.1137218886422} {"canonical_length": 56, "canonical_nll": 1.8190443856375558, "generation_rouge_l_f1": 0.2732919254658386, "generation_token_f1": 0.34782608695652173, "prompt_id": "2b713b1aba655ea9", "reading_canonical_log_prob": -101.86648559570312, "reading_distances": [19, 18, 10, 18], "reading_log_probs": [-104.53327941894531, -103.55364990234375, -103.59211730957031, -103.48985290527344], "reading_logprob_std": 0.4958332046493199, "tokenization_path_logprob_std": 113.1745167141094, "trajectory_balance_rmse": 1179.2593241260283} {"canonical_length": 21, "canonical_nll": 2.3044855026971724, "generation_rouge_l_f1": 0.07142857142857144, "generation_token_f1": 0.07142857142857144, "prompt_id": "12ac8fa81a314c1b", "reading_canonical_log_prob": -48.394195556640625, "reading_distances": [27, 26, 32, 32], "reading_log_probs": [-47.3094596862793, -46.29592514038086, -47.683799743652344, -47.18021011352539], "reading_logprob_std": 0.5877816027266761, "tokenization_path_logprob_std": 40.05158634875802, "trajectory_balance_rmse": 433.8676838155638} {"canonical_length": 174, "canonical_nll": 1.2956695556640625, "generation_rouge_l_f1": 0.19867549668874174, "generation_token_f1": 0.4282560706401766, "prompt_id": "46239be323c13eb4", "reading_canonical_log_prob": -225.44650268554688, "reading_distances": [21, 15, 22, 21], "reading_log_probs": [-225.51229858398438, -230.29885864257812, -227.4163360595703, -231.48971557617188], "reading_logprob_std": 2.7170070415434466, "tokenization_path_logprob_std": 179.73778798752755, "trajectory_balance_rmse": 2916.515711804572} {"canonical_length": 298, "canonical_nll": 1.2775200965420512, "generation_rouge_l_f1": 0.16458852867830423, "generation_token_f1": 0.2543640897755611, "prompt_id": "3fde994857b8ea58", "reading_canonical_log_prob": -380.70098876953125, "reading_distances": [39, 36, 35, 35], "reading_log_probs": [-376.38909912109375, -373.135986328125, -376.3741455078125, -377.53228759765625], "reading_logprob_std": 1.8939438387708452, "tokenization_path_logprob_std": 153.45858625637482, "trajectory_balance_rmse": 6300.737029348107} {"canonical_length": 111, "canonical_nll": 1.0988902873821087, "generation_rouge_l_f1": 0.3146067415730337, "generation_token_f1": 0.46067415730337075, "prompt_id": "366758412bdbb007", "reading_canonical_log_prob": -121.97682189941406, "reading_distances": [39, 36, 33, 39], "reading_log_probs": [-120.72169494628906, -119.68374633789062, -118.57792663574219, -120.02200317382812], "reading_logprob_std": 0.8937402870334292, "tokenization_path_logprob_std": 76.90956696594509, "trajectory_balance_rmse": 2271.7748344504325} {"canonical_length": 17, "canonical_nll": 1.3942424549775965, "generation_rouge_l_f1": 0.39999999999999997, "generation_token_f1": 0.5142857142857142, "prompt_id": "f77e57c7031a2b6c", "reading_canonical_log_prob": -23.70212173461914, "reading_distances": [10, 15, 16, 13], "reading_log_probs": [-22.44292640686035, -20.86139488220215, -21.656879425048828, -22.04491424560547], "reading_logprob_std": 0.6746363381501983, "tokenization_path_logprob_std": 23.159543307894324, "trajectory_balance_rmse": 491.4064653354133} {"canonical_length": 8, "canonical_nll": 1.3218789100646973, "generation_rouge_l_f1": 0.2857142857142857, "generation_token_f1": 0.2857142857142857, "prompt_id": "1f37960a14b0dc02", "reading_canonical_log_prob": -10.575031280517578, "reading_distances": [24, 23, 21, 30], "reading_log_probs": [-11.546465873718262, -11.133810043334961, -10.654890060424805, -10.461090087890625], "reading_logprob_std": 0.48842500102353853, "tokenization_path_logprob_std": 66.29229006561488, "trajectory_balance_rmse": 196.2179608649815} {"canonical_length": 28, "canonical_nll": 2.3133773803710938, "generation_rouge_l_f1": 0.2222222222222222, "generation_token_f1": 0.3888888888888889, "prompt_id": "8d8a8351f2dceabb", "reading_canonical_log_prob": -64.77456665039062, "reading_distances": [20, 20, 21, 19], "reading_log_probs": [-67.03273010253906, -69.35870361328125, -67.48411560058594, -68.30537414550781], "reading_logprob_std": 1.0219088871344826, "tokenization_path_logprob_std": 83.40962823409532, "trajectory_balance_rmse": 631.9003276775554} {"canonical_length": 2, "canonical_nll": 0.362510085105896, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "7a9b64f9112cd0c5", "reading_canonical_log_prob": -0.725020170211792, "reading_distances": [45, 42, 52, 41], "reading_log_probs": [-1.6457102298736572, -1.5175540447235107, -1.9324450492858887, -1.7052642107009888], "reading_logprob_std": 0.17348590879045372, "tokenization_path_logprob_std": 0.0, "trajectory_balance_rmse": 0.7287560701370239} {"canonical_length": 5, "canonical_nll": 1.091444206237793, "generation_rouge_l_f1": 0.8, "generation_token_f1": 0.8, "prompt_id": "2797e2b7991dcf3e", "reading_canonical_log_prob": -5.457221031188965, "reading_distances": [28, 20, 20, 23], "reading_log_probs": [-6.577447891235352, -5.56296968460083, -6.106876373291016, -8.254243850708008], "reading_logprob_std": 1.162333374480212, "tokenization_path_logprob_std": 48.58727696259381, "trajectory_balance_rmse": 76.21185233977315} {"canonical_length": 313, "canonical_nll": 1.7101259079223243, "generation_rouge_l_f1": 0.18620689655172412, "generation_token_f1": 0.4413793103448276, "prompt_id": "d353ae8508e0bb05", "reading_canonical_log_prob": -535.2694091796875, "reading_distances": [42, 41, 40, 45], "reading_log_probs": [-540.1325073242188, -537.4656982421875, -544.1277465820312, -539.9944458007812], "reading_logprob_std": 2.7530966812726687, "tokenization_path_logprob_std": 264.6881142539336, "trajectory_balance_rmse": 6401.62395516861} {"canonical_length": 7, "canonical_nll": 0.009918037269796644, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "1099a8369d87bee2", "reading_canonical_log_prob": -0.06942626088857651, "reading_distances": [20, 19, 25, 19], "reading_log_probs": [-0.1541966199874878, -0.18174374103546143, -0.1572979837656021, -0.19151045382022858], "reading_logprob_std": 0.018312703616544964, "tokenization_path_logprob_std": 26.41102643795086, "trajectory_balance_rmse": 171.9657817651731} {"canonical_length": 264, "canonical_nll": 1.3153244943329783, "generation_rouge_l_f1": 0.2628571428571429, "generation_token_f1": 0.5295238095238095, "prompt_id": "d8d494771b6c2d81", "reading_canonical_log_prob": -347.24566650390625, "reading_distances": [25, 25, 31, 25], "reading_log_probs": [-345.30584716796875, -345.91729736328125, -346.3541259765625, -346.65972900390625], "reading_logprob_std": 0.5874502118108915, "tokenization_path_logprob_std": 326.3741647666002, "trajectory_balance_rmse": 6062.600471912361} {"canonical_length": 372, "canonical_nll": 1.2328384563487063, "generation_rouge_l_f1": 0.21324717285945072, "generation_token_f1": 0.44911147011308566, "prompt_id": "27da9cdd8cf02e41", "reading_canonical_log_prob": -458.61590576171875, "reading_distances": [33, 34, 34, 31], "reading_log_probs": [-464.0047607421875, -459.74066162109375, -462.47021484375, -464.41033935546875], "reading_logprob_std": 2.1158587795439323, "tokenization_path_logprob_std": 207.7236223127163, "trajectory_balance_rmse": 7181.036836423241} {"canonical_length": 51, "canonical_nll": 0.973027696796492, "generation_rouge_l_f1": 0.288659793814433, "generation_token_f1": 0.32989690721649484, "prompt_id": "47574188e8957081", "reading_canonical_log_prob": -49.624412536621094, "reading_distances": [33, 28, 28, 33], "reading_log_probs": [-46.88463592529297, -45.28669357299805, -44.61842346191406, -45.77947998046875], "reading_logprob_std": 0.9551607184503229, "tokenization_path_logprob_std": 122.92351908049194, "trajectory_balance_rmse": 1023.7136667124682} {"canonical_length": 238, "canonical_nll": 2.0760285193178833, "generation_rouge_l_f1": 0.1690721649484536, "generation_token_f1": 0.3505154639175258, "prompt_id": "b7c16abbc7061a65", "reading_canonical_log_prob": -494.09478759765625, "reading_distances": [27, 30, 34, 27], "reading_log_probs": [-496.33233642578125, -499.4886779785156, -497.582275390625, -506.1662292480469], "reading_logprob_std": 4.379294207440301, "tokenization_path_logprob_std": 165.04920093384501, "trajectory_balance_rmse": 5233.410149661917} {"canonical_length": 334, "canonical_nll": 1.317577521958037, "generation_rouge_l_f1": 0.23627287853577372, "generation_token_f1": 0.47920133111480867, "prompt_id": "01fef5f51abd900f", "reading_canonical_log_prob": -440.0708923339844, "reading_distances": [20, 31, 34, 33], "reading_log_probs": [-437.23089599609375, -440.95965576171875, -439.163330078125, -437.8399658203125], "reading_logprob_std": 1.6512485471308052, "tokenization_path_logprob_std": 278.4619778366982, "trajectory_balance_rmse": 7002.220841467283} {"canonical_length": 60, "canonical_nll": 1.06620241800944, "generation_rouge_l_f1": 0.24264705882352944, "generation_token_f1": 0.27941176470588236, "prompt_id": "1d0839cb7b1b8564", "reading_canonical_log_prob": -63.972145080566406, "reading_distances": [29, 34, 29, 30], "reading_log_probs": [-64.24971008300781, -63.79420471191406, -65.50567626953125, -65.70037841796875], "reading_logprob_std": 0.9349650991541905, "tokenization_path_logprob_std": 65.23138546411354, "trajectory_balance_rmse": 1036.2559579505855} {"canonical_length": 198, "canonical_nll": 1.460035381895123, "generation_rouge_l_f1": 0.33181818181818185, "generation_token_f1": 0.5454545454545454, "prompt_id": "b900859423a57fd5", "reading_canonical_log_prob": -289.0870056152344, "reading_distances": [19, 22, 22, 25], "reading_log_probs": [-290.2339782714844, -293.153076171875, -284.02520751953125, -292.61785888671875], "reading_logprob_std": 4.185201655683817, "tokenization_path_logprob_std": 268.8287312122484, "trajectory_balance_rmse": 4279.637372658227} {"canonical_length": 48, "canonical_nll": 1.4787483215332031, "generation_rouge_l_f1": 0.20481927710843376, "generation_token_f1": 0.3132530120481928, "prompt_id": "5fa2bbf4741331df", "reading_canonical_log_prob": -70.97991943359375, "reading_distances": [18, 16, 18, 16], "reading_log_probs": [-70.0316390991211, -70.79257202148438, -68.52934265136719, -71.96533966064453], "reading_logprob_std": 1.439872041649638, "tokenization_path_logprob_std": 30.332945502541325, "trajectory_balance_rmse": 1138.930692110771} {"canonical_length": 501, "canonical_nll": 0.5961191630411052, "generation_rouge_l_f1": 0.253922967189729, "generation_token_f1": 0.4536376604850214, "prompt_id": "97c9524265f7d241", "reading_canonical_log_prob": -298.65570068359375, "reading_distances": [22, 18, 21, 19], "reading_log_probs": [-292.30072021484375, -293.390625, -291.1795959472656, -291.439697265625], "reading_logprob_std": 0.9978557751391128, "tokenization_path_logprob_std": 158.27668815903877, "trajectory_balance_rmse": 6551.790405475161} {"canonical_length": 11, "canonical_nll": 1.6292499195445667, "generation_rouge_l_f1": 0.5, "generation_token_f1": 0.5, "prompt_id": "990e2b768a18fc67", "reading_canonical_log_prob": -17.921749114990234, "reading_distances": [25, 26, 23, 31], "reading_log_probs": [-18.772327423095703, -18.52017593383789, -17.64824867248535, -23.54100799560547], "reading_logprob_std": 2.657710041801695, "tokenization_path_logprob_std": 54.854711036898394, "trajectory_balance_rmse": 215.85109716667483} {"canonical_length": 30, "canonical_nll": 1.5358629862467448, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "4a59564b57b1ad24", "reading_canonical_log_prob": -46.075889587402344, "reading_distances": [31, 30, 32, 32], "reading_log_probs": [-45.41375732421875, -45.01319122314453, -45.277992248535156, -43.99087142944336], "reading_logprob_std": 0.643909088642476, "tokenization_path_logprob_std": 33.647270383499624, "trajectory_balance_rmse": 602.1617202091963} {"canonical_length": 118, "canonical_nll": 1.5931675797801907, "generation_rouge_l_f1": 0.1818181818181818, "generation_token_f1": 0.2754820936639118, "prompt_id": "fc01869b83f5b227", "reading_canonical_log_prob": -187.9937744140625, "reading_distances": [26, 26, 28, 33], "reading_log_probs": [-190.59786987304688, -188.42282104492188, -191.74032592773438, -188.8287353515625], "reading_logprob_std": 1.549565663656567, "tokenization_path_logprob_std": 175.92283314509956, "trajectory_balance_rmse": 2591.648020887863} {"canonical_length": 10, "canonical_nll": 0.430483341217041, "generation_rouge_l_f1": 0.30769230769230765, "generation_token_f1": 0.30769230769230765, "prompt_id": "e35706a135002934", "reading_canonical_log_prob": -4.30483341217041, "reading_distances": [32, 30, 33, 36], "reading_log_probs": [-5.23783540725708, -5.3427839279174805, -5.133406162261963, -5.604426383972168], "reading_logprob_std": 0.20216834491522476, "tokenization_path_logprob_std": 11.407708703084612, "trajectory_balance_rmse": 214.7079696666081} {"canonical_length": 123, "canonical_nll": 1.0271389348720148, "generation_rouge_l_f1": 0.25, "generation_token_f1": 0.359375, "prompt_id": "5df9638e97e9c3b6", "reading_canonical_log_prob": -126.33808898925781, "reading_distances": [11, 11, 10, 14], "reading_log_probs": [-124.443603515625, -124.32920837402344, -125.0005111694336, -124.62088775634766], "reading_logprob_std": 0.293612602798141, "tokenization_path_logprob_std": 103.82433083361214, "trajectory_balance_rmse": 2219.796491065861} {"canonical_length": 25, "canonical_nll": 0.6157456970214844, "generation_rouge_l_f1": 0.12318840579710144, "generation_token_f1": 0.15942028985507245, "prompt_id": "6f6b5ba97837e4b7", "reading_canonical_log_prob": -15.39364242553711, "reading_distances": [27, 27, 33, 26], "reading_log_probs": [-15.260246276855469, -13.415761947631836, -13.34581470489502, -14.557987213134766], "reading_logprob_std": 0.9282273356915477, "tokenization_path_logprob_std": 55.07247096741338, "trajectory_balance_rmse": 449.3570482778069} {"canonical_length": 211, "canonical_nll": 0.759187598929021, "generation_rouge_l_f1": 0.439918533604888, "generation_token_f1": 0.5824847250509165, "prompt_id": "111777e66227bbb0", "reading_canonical_log_prob": -160.18858337402344, "reading_distances": [40, 43, 44, 36], "reading_log_probs": [-157.25704956054688, -158.25140380859375, -158.15731811523438, -156.07911682128906], "reading_logprob_std": 1.009676221636073, "tokenization_path_logprob_std": 102.16731281236467, "trajectory_balance_rmse": 3580.6395883711357} {"canonical_length": 264, "canonical_nll": 1.176661982680812, "generation_rouge_l_f1": 0.24427480916030533, "generation_token_f1": 0.4732824427480916, "prompt_id": "7eae8a9735d3d72c", "reading_canonical_log_prob": -310.6387634277344, "reading_distances": [31, 30, 35, 31], "reading_log_probs": [-312.3789367675781, -311.61456298828125, -313.96258544921875, -313.5686340332031], "reading_logprob_std": 1.0799070503606647, "tokenization_path_logprob_std": 113.52481088771664, "trajectory_balance_rmse": 5697.73813792927} {"canonical_length": 6, "canonical_nll": 0.009781024729212126, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "88814d5eacafd109", "reading_canonical_log_prob": -0.05868614837527275, "reading_distances": [38, 42, 36, 41], "reading_log_probs": [-0.165853351354599, -0.19363972544670105, -0.10987163335084915, -0.16419105231761932], "reading_logprob_std": 0.0350520035723781, "tokenization_path_logprob_std": 18.26330178749758, "trajectory_balance_rmse": 114.23424204136073} {"canonical_length": 150, "canonical_nll": 1.300060017903646, "generation_rouge_l_f1": 0.2160493827160494, "generation_token_f1": 0.42592592592592593, "prompt_id": "c6b7c5355ef042f7", "reading_canonical_log_prob": -195.00900268554688, "reading_distances": [40, 40, 38, 41], "reading_log_probs": [-192.18777465820312, -190.32493591308594, -189.90835571289062, -192.63577270507812], "reading_logprob_std": 1.348422855297247, "tokenization_path_logprob_std": 87.15455850281629, "trajectory_balance_rmse": 2959.033102240802} {"canonical_length": 21, "canonical_nll": 1.3982360476539248, "generation_rouge_l_f1": 0.43902439024390244, "generation_token_f1": 0.43902439024390244, "prompt_id": "62bbc871380d57bb", "reading_canonical_log_prob": -29.362957000732422, "reading_distances": [13, 17, 17, 13], "reading_log_probs": [-29.272024154663086, -29.90980339050293, -29.186969757080078, -28.491907119750977], "reading_logprob_std": 0.5801299824469791, "tokenization_path_logprob_std": 38.18210803296301, "trajectory_balance_rmse": 418.8650721011029} {"canonical_length": 28, "canonical_nll": 2.105405535016741, "generation_rouge_l_f1": 0.17391304347826086, "generation_token_f1": 0.20289855072463767, "prompt_id": "085c601e3edc073f", "reading_canonical_log_prob": -58.95135498046875, "reading_distances": [15, 11, 13, 11], "reading_log_probs": [-58.97429656982422, -59.53187561035156, -61.0660514831543, -59.56233596801758], "reading_logprob_std": 0.8966561553091568, "tokenization_path_logprob_std": 59.683730704323445, "trajectory_balance_rmse": 617.7584313606689} {"canonical_length": 314, "canonical_nll": 1.479877739195611, "generation_rouge_l_f1": 0.2827586206896551, "generation_token_f1": 0.5413793103448276, "prompt_id": "575d828d632e226f", "reading_canonical_log_prob": -464.6816101074219, "reading_distances": [37, 32, 28, 34], "reading_log_probs": [-467.91143798828125, -469.8373718261719, -467.74786376953125, -466.67486572265625], "reading_logprob_std": 1.3160516144647074, "tokenization_path_logprob_std": 270.1364038576585, "trajectory_balance_rmse": 6907.70725636027} {"canonical_length": 382, "canonical_nll": 1.0266718839475621, "generation_rouge_l_f1": 0.21565217391304348, "generation_token_f1": 0.4000000000000001, "prompt_id": "acf1157605893744", "reading_canonical_log_prob": -392.18865966796875, "reading_distances": [20, 17, 14, 14], "reading_log_probs": [-390.15435791015625, -390.1748046875, -389.4940185546875, -390.9996337890625], "reading_logprob_std": 0.6165527475367678, "tokenization_path_logprob_std": 240.33618287920476, "trajectory_balance_rmse": 7541.802591620688} {"canonical_length": 19, "canonical_nll": 2.111451399953742, "generation_rouge_l_f1": 0.28571428571428575, "generation_token_f1": 0.28571428571428575, "prompt_id": "c0621d0076f956e9", "reading_canonical_log_prob": -40.117576599121094, "reading_distances": [34, 34, 32, 34], "reading_log_probs": [-44.05619812011719, -42.973655700683594, -44.58224105834961, -42.56359100341797], "reading_logprob_std": 0.9357324769342937, "tokenization_path_logprob_std": 34.84939355931958, "trajectory_balance_rmse": 327.7569236024135} {"canonical_length": 252, "canonical_nll": 1.1510036407955109, "generation_rouge_l_f1": 0.3898635477582846, "generation_token_f1": 0.5964912280701754, "prompt_id": "71ec6348f5f044b8", "reading_canonical_log_prob": -290.05291748046875, "reading_distances": [19, 20, 17, 19], "reading_log_probs": [-291.82476806640625, -292.5318908691406, -290.58660888671875, -294.485595703125], "reading_logprob_std": 1.6308547679232626, "tokenization_path_logprob_std": 153.75416596119854, "trajectory_balance_rmse": 5507.603190266973} {"canonical_length": 13, "canonical_nll": 0.42793057515070987, "generation_rouge_l_f1": 0.5882352941176471, "generation_token_f1": 0.5882352941176471, "prompt_id": "73a9260fbda0ad11", "reading_canonical_log_prob": -5.5630974769592285, "reading_distances": [40, 30, 36, 38], "reading_log_probs": [-4.210536479949951, -3.5122108459472656, -4.413846015930176, -4.764664649963379], "reading_logprob_std": 0.5276292663301922, "tokenization_path_logprob_std": 40.00843692579128, "trajectory_balance_rmse": 158.83138005532066} {"canonical_length": 142, "canonical_nll": 1.0070636373170665, "generation_rouge_l_f1": 0.3076923076923077, "generation_token_f1": 0.48806366047745364, "prompt_id": "0e246b28f0fca81c", "reading_canonical_log_prob": -143.00303649902344, "reading_distances": [10, 7, 10, 8], "reading_log_probs": [-141.7115936279297, -139.4983673095703, -142.281494140625, -143.57833862304688], "reading_logprob_std": 1.7024898603604142, "tokenization_path_logprob_std": 33.80766055705068, "trajectory_balance_rmse": 2744.8639704483103} {"canonical_length": 38, "canonical_nll": 1.2875940423262746, "generation_rouge_l_f1": 0.0234375, "generation_token_f1": 0.03125, "prompt_id": "7dd4910bb043bf96", "reading_canonical_log_prob": -48.92857360839844, "reading_distances": [14, 9, 11, 12], "reading_log_probs": [-49.74897766113281, -49.76128005981445, -49.9659309387207, -50.711524963378906], "reading_logprob_std": 0.45409940095057183, "tokenization_path_logprob_std": 91.81896706220991, "trajectory_balance_rmse": 541.3026128424209} {"canonical_length": 187, "canonical_nll": 1.2820383979674967, "generation_rouge_l_f1": 0.17452830188679244, "generation_token_f1": 0.3537735849056604, "prompt_id": "b7fe1fa68d367af3", "reading_canonical_log_prob": -239.74118041992188, "reading_distances": [17, 13, 15, 15], "reading_log_probs": [-232.06573486328125, -231.33834838867188, -233.3111572265625, -231.26803588867188], "reading_logprob_std": 0.9481453369762703, "tokenization_path_logprob_std": 213.92870669914478, "trajectory_balance_rmse": 3825.3432413946334} {"canonical_length": 96, "canonical_nll": 1.3833699226379395, "generation_rouge_l_f1": 0.2801120448179272, "generation_token_f1": 0.43137254901960786, "prompt_id": "b435d26419c9a382", "reading_canonical_log_prob": -132.8035125732422, "reading_distances": [41, 46, 48, 52], "reading_log_probs": [-132.92385864257812, -134.17446899414062, -135.55203247070312, -136.2003173828125], "reading_logprob_std": 1.4614069120727116, "tokenization_path_logprob_std": 181.35892135680547, "trajectory_balance_rmse": 2193.6809744989264} {"canonical_length": 113, "canonical_nll": 1.3090238317979122, "generation_rouge_l_f1": 0.18750000000000003, "generation_token_f1": 0.33522727272727276, "prompt_id": "fc2765baa4652639", "reading_canonical_log_prob": -147.91969299316406, "reading_distances": [19, 26, 24, 21], "reading_log_probs": [-151.058837890625, -149.72528076171875, -150.181396484375, -149.21792602539062], "reading_logprob_std": 0.7816094633982147, "tokenization_path_logprob_std": 222.90318582022974, "trajectory_balance_rmse": 2184.02634069926} {"canonical_length": 216, "canonical_nll": 1.46551513671875, "generation_rouge_l_f1": 0.21894736842105264, "generation_token_f1": 0.4589473684210526, "prompt_id": "6389a5bbf9eaba60", "reading_canonical_log_prob": -316.55126953125, "reading_distances": [34, 29, 31, 26], "reading_log_probs": [-307.6491394042969, -307.438232421875, -306.6986083984375, -309.90899658203125], "reading_logprob_std": 1.384836239795425, "tokenization_path_logprob_std": 154.8735010288407, "trajectory_balance_rmse": 4470.359544568413} {"canonical_length": 13, "canonical_nll": 0.7279148101806641, "generation_rouge_l_f1": 0.14285714285714288, "generation_token_f1": 0.14285714285714288, "prompt_id": "286cf1906a688fcb", "reading_canonical_log_prob": -9.462892532348633, "reading_distances": [25, 25, 24, 24], "reading_log_probs": [-9.846879005432129, -9.81259536743164, -10.5492582321167, -7.949775218963623], "reading_logprob_std": 1.112938967327855, "tokenization_path_logprob_std": 19.72358206578916, "trajectory_balance_rmse": 255.15351636412086} {"canonical_length": 93, "canonical_nll": 1.3700357252551663, "generation_rouge_l_f1": 0.2972972972972973, "generation_token_f1": 0.4459459459459459, "prompt_id": "33333486aa163f12", "reading_canonical_log_prob": -127.41332244873047, "reading_distances": [34, 26, 32, 28], "reading_log_probs": [-123.90121459960938, -124.4744873046875, -124.35133361816406, -126.28787994384766], "reading_logprob_std": 1.0520294496996205, "tokenization_path_logprob_std": 170.72318961042686, "trajectory_balance_rmse": 2160.45027367062} {"canonical_length": 60, "canonical_nll": 2.0774256388346353, "generation_rouge_l_f1": 0.2006269592476489, "generation_token_f1": 0.2507836990595611, "prompt_id": "c5556f6cb2c9f8dd", "reading_canonical_log_prob": -124.64553833007812, "reading_distances": [25, 23, 27, 23], "reading_log_probs": [-123.06668090820312, -121.26040649414062, -119.03138732910156, -123.58175659179688], "reading_logprob_std": 2.059016302937463, "tokenization_path_logprob_std": 36.40080009703859, "trajectory_balance_rmse": 351.5780358830619} {"canonical_length": 11, "canonical_nll": 3.0162748856977983, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "b1950a5b21116b51", "reading_canonical_log_prob": -33.17902374267578, "reading_distances": [17, 18, 12, 19], "reading_log_probs": [-34.03166961669922, -34.022701263427734, -35.783050537109375, -35.255149841308594], "reading_logprob_std": 0.8879169268033241, "tokenization_path_logprob_std": 30.846313803268842, "trajectory_balance_rmse": 228.03452537508335} {"canonical_length": 263, "canonical_nll": 1.4458803822332462, "generation_rouge_l_f1": 0.22433460076045628, "generation_token_f1": 0.4524714828897338, "prompt_id": "355011c0ec753a64", "reading_canonical_log_prob": -380.26654052734375, "reading_distances": [18, 21, 20, 18], "reading_log_probs": [-376.89544677734375, -379.2873229980469, -378.38067626953125, -377.75164794921875], "reading_logprob_std": 1.009787177248472, "tokenization_path_logprob_std": 244.24127194112677, "trajectory_balance_rmse": 6066.644786755493} {"canonical_length": 29, "canonical_nll": 0.969592127306708, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "c2e4ff98064c8812", "reading_canonical_log_prob": -28.11817169189453, "reading_distances": [25, 22, 22, 29], "reading_log_probs": [-29.506065368652344, -26.068017959594727, -25.242080688476562, -25.95062828063965], "reading_logprob_std": 1.9113882526718886, "tokenization_path_logprob_std": 94.39799348279946, "trajectory_balance_rmse": 622.0162374053358} {"canonical_length": 57, "canonical_nll": 1.8815974185341282, "generation_rouge_l_f1": 0.1128526645768025, "generation_token_f1": 0.1755485893416928, "prompt_id": "e95bb93274990ecf", "reading_canonical_log_prob": -107.25105285644531, "reading_distances": [28, 30, 27, 28], "reading_log_probs": [-107.62761688232422, -106.88333892822266, -108.02487182617188, -106.54570770263672], "reading_logprob_std": 0.6762215700992305, "tokenization_path_logprob_std": 141.90807442975682, "trajectory_balance_rmse": 1099.8834146944655} {"canonical_length": 40, "canonical_nll": 1.8821361541748047, "generation_rouge_l_f1": 0.18300653594771243, "generation_token_f1": 0.22222222222222224, "prompt_id": "0eb7b4a395752869", "reading_canonical_log_prob": -75.28544616699219, "reading_distances": [30, 29, 36, 29], "reading_log_probs": [-83.20348358154297, -80.68170928955078, -81.17633056640625, -81.70179748535156], "reading_logprob_std": 1.0910660564963242, "tokenization_path_logprob_std": 82.47905492326433, "trajectory_balance_rmse": 718.8044958879466} {"canonical_length": 124, "canonical_nll": 1.8580111842001639, "generation_rouge_l_f1": 0.26344086021505375, "generation_token_f1": 0.3333333333333333, "prompt_id": "5d63d37b660c138f", "reading_canonical_log_prob": -230.3933868408203, "reading_distances": [38, 28, 37, 39], "reading_log_probs": [-233.74319458007812, -233.19564819335938, -233.7352752685547, -233.189453125], "reading_logprob_std": 0.31565499781225576, "tokenization_path_logprob_std": 59.65156264609001, "trajectory_balance_rmse": 2754.9869841704894} {"canonical_length": 241, "canonical_nll": 1.1519209359196707, "generation_rouge_l_f1": 0.29133858267716534, "generation_token_f1": 0.5511811023622047, "prompt_id": "17f90ddfcc762d15", "reading_canonical_log_prob": -277.6129455566406, "reading_distances": [23, 25, 31, 31], "reading_log_probs": [-272.620849609375, -274.12335205078125, -273.4149169921875, -271.5661926269531], "reading_logprob_std": 1.0976901610191363, "tokenization_path_logprob_std": 95.09268714341316, "trajectory_balance_rmse": 5097.9280604558135} {"canonical_length": 61, "canonical_nll": 1.4732525934938525, "generation_rouge_l_f1": 0.2517482517482518, "generation_token_f1": 0.36363636363636365, "prompt_id": "9e46c646f8e1ae11", "reading_canonical_log_prob": -89.868408203125, "reading_distances": [21, 19, 23, 26], "reading_log_probs": [-84.10353088378906, -84.5601577758789, -84.13331604003906, -84.2031478881836], "reading_logprob_std": 0.21091966007212284, "tokenization_path_logprob_std": 66.260411061962, "trajectory_balance_rmse": 1287.1601681983375} {"canonical_length": 317, "canonical_nll": 1.1187199252637028, "generation_rouge_l_f1": 0.26870748299319724, "generation_token_f1": 0.5034013605442177, "prompt_id": "c405755d9e3e8c9b", "reading_canonical_log_prob": -354.63421630859375, "reading_distances": [23, 26, 23, 21], "reading_log_probs": [-347.67169189453125, -347.43194580078125, -347.7266845703125, -346.203857421875], "reading_logprob_std": 0.7146744571052194, "tokenization_path_logprob_std": 224.43360746706358, "trajectory_balance_rmse": 5953.01722731849} {"canonical_length": 225, "canonical_nll": 0.5490644327799479, "generation_rouge_l_f1": 0.26991150442477874, "generation_token_f1": 0.43805309734513276, "prompt_id": "d5c04061128f24af", "reading_canonical_log_prob": -123.53949737548828, "reading_distances": [27, 26, 23, 22], "reading_log_probs": [-122.93888854980469, -121.61294555664062, -121.75452423095703, -123.2496337890625], "reading_logprob_std": 0.8262138416954325, "tokenization_path_logprob_std": 189.80513562996586, "trajectory_balance_rmse": 3261.5553217111988} {"canonical_length": 61, "canonical_nll": 1.2240460505251025, "generation_rouge_l_f1": 0.3875, "generation_token_f1": 0.4125, "prompt_id": "1419b413999c869f", "reading_canonical_log_prob": -74.66680908203125, "reading_distances": [17, 13, 12, 12], "reading_log_probs": [-72.9385757446289, -75.48725128173828, -76.11433410644531, -72.87046813964844], "reading_logprob_std": 1.6918747024202159, "tokenization_path_logprob_std": 26.254537476168665, "trajectory_balance_rmse": 729.3880778791998} {"canonical_length": 116, "canonical_nll": 1.3722185595282193, "generation_rouge_l_f1": 0.2737430167597765, "generation_token_f1": 0.41899441340782123, "prompt_id": "3501affcfd13e96c", "reading_canonical_log_prob": -159.17735290527344, "reading_distances": [24, 21, 25, 18], "reading_log_probs": [-161.6546173095703, -159.71054077148438, -160.72872924804688, -163.41305541992188], "reading_logprob_std": 1.572676496959261, "tokenization_path_logprob_std": 95.65402623328957, "trajectory_balance_rmse": 2392.0453879480247} {"canonical_length": 26, "canonical_nll": 2.1516654674823465, "generation_rouge_l_f1": 0.3636363636363637, "generation_token_f1": 0.4545454545454546, "prompt_id": "6fdd3868e470d9e0", "reading_canonical_log_prob": -55.943302154541016, "reading_distances": [57, 60, 47, 39], "reading_log_probs": [-56.105255126953125, -56.61393737792969, -51.58742141723633, -55.90408706665039], "reading_logprob_std": 2.329402644205121, "tokenization_path_logprob_std": 54.90570274515649, "trajectory_balance_rmse": 553.0647186613625} {"canonical_length": 304, "canonical_nll": 0.793209477474815, "generation_rouge_l_f1": 0.31157894736842107, "generation_token_f1": 0.5642105263157894, "prompt_id": "1d8b10c0b2c4c1d8", "reading_canonical_log_prob": -241.13568115234375, "reading_distances": [27, 19, 19, 23], "reading_log_probs": [-238.49905395507812, -239.62242126464844, -236.96388244628906, -238.38137817382812], "reading_logprob_std": 1.089719535885638, "tokenization_path_logprob_std": 147.99575273404344, "trajectory_balance_rmse": 4717.719459153067} {"canonical_length": 31, "canonical_nll": 0.7305673783825289, "generation_rouge_l_f1": 0.3010752688172043, "generation_token_f1": 0.38709677419354843, "prompt_id": "6f83d657c17ec17d", "reading_canonical_log_prob": -22.6475887298584, "reading_distances": [16, 19, 23, 24], "reading_log_probs": [-23.950714111328125, -21.952045440673828, -22.73280906677246, -23.34832000732422], "reading_logprob_std": 0.8553204093788551, "tokenization_path_logprob_std": 73.76735408226365, "trajectory_balance_rmse": 493.4662616565276} {"canonical_length": 39, "canonical_nll": 1.3396399082281651, "generation_rouge_l_f1": 0.09756097560975609, "generation_token_f1": 0.09756097560975609, "prompt_id": "6587efe9dc00b7ac", "reading_canonical_log_prob": -52.24595642089844, "reading_distances": [24, 22, 21, 23], "reading_log_probs": [-51.69651412963867, -54.99188232421875, -49.640174865722656, -50.82869338989258], "reading_logprob_std": 2.295404587594885, "tokenization_path_logprob_std": 123.8675240612578, "trajectory_balance_rmse": 911.2840378923622} {"canonical_length": 15, "canonical_nll": 0.8873584747314454, "generation_rouge_l_f1": 0.64, "generation_token_f1": 0.64, "prompt_id": "ea95653fa930225e", "reading_canonical_log_prob": -13.31037712097168, "reading_distances": [15, 23, 15, 16], "reading_log_probs": [-14.625401496887207, -15.911840438842773, -14.50071907043457, -15.800281524658203], "reading_logprob_std": 0.7496324331875969, "tokenization_path_logprob_std": 61.84505714081093, "trajectory_balance_rmse": 427.1818548046486} {"canonical_length": 218, "canonical_nll": 1.5391285747563073, "generation_rouge_l_f1": 0.2652631578947368, "generation_token_f1": 0.46736842105263166, "prompt_id": "dbfaa5f0bb7fb12c", "reading_canonical_log_prob": -335.530029296875, "reading_distances": [38, 42, 40, 45], "reading_log_probs": [-329.4482421875, -331.05755615234375, -327.86651611328125, -331.3616943359375], "reading_logprob_std": 1.6136024734894399, "tokenization_path_logprob_std": 106.87683632855077, "trajectory_balance_rmse": 4858.11628934159} {"canonical_length": 236, "canonical_nll": 1.2341282731395657, "generation_rouge_l_f1": 0.23886639676113358, "generation_token_f1": 0.5465587044534413, "prompt_id": "6ef66ee37f69a6e9", "reading_canonical_log_prob": -291.2542724609375, "reading_distances": [37, 45, 46, 48], "reading_log_probs": [-287.14093017578125, -291.91510009765625, -288.8382873535156, -287.34912109375], "reading_logprob_std": 2.2032101648486737, "tokenization_path_logprob_std": 265.25570756833207, "trajectory_balance_rmse": 5495.540820394527} {"canonical_length": 254, "canonical_nll": 0.34400951956200787, "generation_rouge_l_f1": 0.5759162303664922, "generation_token_f1": 0.7050610820244329, "prompt_id": "e2bfc09d77b30452", "reading_canonical_log_prob": -87.37841796875, "reading_distances": [21, 26, 24, 23], "reading_log_probs": [-86.4903564453125, -89.51951599121094, -87.3128433227539, -86.94532775878906], "reading_logprob_std": 1.3444394161870485, "tokenization_path_logprob_std": 93.73176695212193, "trajectory_balance_rmse": 3330.70372644375} {"canonical_length": 315, "canonical_nll": 0.906769283234127, "generation_rouge_l_f1": 0.25828970331588136, "generation_token_f1": 0.5340314136125655, "prompt_id": "3024b6f11a619ed8", "reading_canonical_log_prob": -285.63232421875, "reading_distances": [16, 13, 15, 12], "reading_log_probs": [-283.4585876464844, -284.5674743652344, -283.9057922363281, -284.1115417480469], "reading_logprob_std": 0.4604347050746903, "tokenization_path_logprob_std": 84.02225311404577, "trajectory_balance_rmse": 5591.07918163053} {"canonical_length": 147, "canonical_nll": 1.5006153340242347, "generation_rouge_l_f1": 0.22596153846153846, "generation_token_f1": 0.4663461538461538, "prompt_id": "2027b774c4dd5ef1", "reading_canonical_log_prob": -220.5904541015625, "reading_distances": [79, 81, 68, 82], "reading_log_probs": [-222.37515258789062, -220.71401977539062, -224.32040405273438, -228.43917846679688], "reading_logprob_std": 3.3287154977496356, "tokenization_path_logprob_std": 51.6041134143849, "trajectory_balance_rmse": 3095.8788542168486} {"canonical_length": 320, "canonical_nll": 1.0398348808288573, "generation_rouge_l_f1": 0.20599250936329588, "generation_token_f1": 0.37078651685393255, "prompt_id": "ea4a6ac8af4f585f", "reading_canonical_log_prob": -332.7471618652344, "reading_distances": [29, 33, 37, 34], "reading_log_probs": [-331.9560852050781, -333.28582763671875, -333.6455078125, -331.1380920410156], "reading_logprob_std": 1.166217116950338, "tokenization_path_logprob_std": 121.99594050894679, "trajectory_balance_rmse": 6329.69147256385} {"canonical_length": 142, "canonical_nll": 1.3429311832911532, "generation_rouge_l_f1": 0.27548209366391185, "generation_token_f1": 0.4683195592286502, "prompt_id": "e9a730206e7c88d2", "reading_canonical_log_prob": -190.69622802734375, "reading_distances": [33, 22, 21, 22], "reading_log_probs": [-192.61793518066406, -195.6368408203125, -193.4508056640625, -194.5575714111328], "reading_logprob_std": 1.3146024681313566, "tokenization_path_logprob_std": 49.31288444891345, "trajectory_balance_rmse": 2692.5571239542455} {"canonical_length": 324, "canonical_nll": 0.9404419322072723, "generation_rouge_l_f1": 0.25187969924812026, "generation_token_f1": 0.518796992481203, "prompt_id": "21922a55d95c3a47", "reading_canonical_log_prob": -304.70318603515625, "reading_distances": [18, 19, 18, 19], "reading_log_probs": [-299.9747314453125, -297.0260009765625, -297.589111328125, -298.630615234375], "reading_logprob_std": 1.296452007098453, "tokenization_path_logprob_std": 98.51915627094309, "trajectory_balance_rmse": 6371.085641188709} {"canonical_length": 17, "canonical_nll": 0.6297670252182904, "generation_rouge_l_f1": 0.39999999999999997, "generation_token_f1": 0.39999999999999997, "prompt_id": "732a8e545760e4c1", "reading_canonical_log_prob": -10.706039428710938, "reading_distances": [24, 22, 16, 17], "reading_log_probs": [-12.819724082946777, -12.266725540161133, -12.380487442016602, -11.607575416564941], "reading_logprob_std": 0.5010700432830698, "tokenization_path_logprob_std": 50.599669567342936, "trajectory_balance_rmse": 321.86676449098707} {"canonical_length": 8, "canonical_nll": 0.510654866695404, "generation_rouge_l_f1": 0.875, "generation_token_f1": 0.875, "prompt_id": "530dfa24f61035c2", "reading_canonical_log_prob": -4.085238933563232, "reading_distances": [34, 32, 26, 35], "reading_log_probs": [-3.8357393741607666, -4.051420211791992, -3.71211838722229, -4.152739524841309], "reading_logprob_std": 0.2003804269246851, "tokenization_path_logprob_std": 33.843610270805485, "trajectory_balance_rmse": 109.96471797441878} {"canonical_length": 8, "canonical_nll": 1.9349392652511597, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.6666666666666666, "prompt_id": "ab024353313a6e65", "reading_canonical_log_prob": -15.479514122009277, "reading_distances": [27, 29, 36, 33], "reading_log_probs": [-14.094955444335938, -16.292861938476562, -14.137304306030273, -14.521172523498535], "reading_logprob_std": 1.0387056962059051, "tokenization_path_logprob_std": 10.261871023520689, "trajectory_balance_rmse": 145.09825680649814} {"canonical_length": 12, "canonical_nll": 0.6092309951782227, "generation_rouge_l_f1": 0.2432432432432432, "generation_token_f1": 0.2432432432432432, "prompt_id": "2321d8df80c63fef", "reading_canonical_log_prob": -7.310771942138672, "reading_distances": [20, 20, 20, 16], "reading_log_probs": [-8.673754692077637, -8.578362464904785, -8.58511734008789, -8.307903289794922], "reading_logprob_std": 0.15833638483322449, "tokenization_path_logprob_std": 49.25419774713702, "trajectory_balance_rmse": 235.76885561905763} {"canonical_length": 16, "canonical_nll": 0.24939537048339844, "generation_rouge_l_f1": 0.5263157894736842, "generation_token_f1": 0.5263157894736842, "prompt_id": "f3f8000f2b6cf3b2", "reading_canonical_log_prob": -3.990325927734375, "reading_distances": [5, 7, 4, 12], "reading_log_probs": [-4.227663993835449, -3.998345136642456, -4.16574764251709, -4.553372859954834], "reading_logprob_std": 0.23252951785843967, "tokenization_path_logprob_std": 16.41573508775944, "trajectory_balance_rmse": 172.87808417818283} {"canonical_length": 8, "canonical_nll": 0.061581727117300034, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "30a8ec8b58b46182", "reading_canonical_log_prob": -0.49265381693840027, "reading_distances": [28, 24, 26, 26], "reading_log_probs": [-0.6701261401176453, -0.8893089294433594, -0.7907178401947021, -0.860571563243866], "reading_logprob_std": 0.0975869245736633, "tokenization_path_logprob_std": 24.721105925826823, "trajectory_balance_rmse": 51.94237693589389} {"canonical_length": 267, "canonical_nll": 1.4588017267234317, "generation_rouge_l_f1": 0.2234848484848485, "generation_token_f1": 0.5189393939393939, "prompt_id": "f1f8d8a792fc708b", "reading_canonical_log_prob": -389.50006103515625, "reading_distances": [24, 28, 29, 30], "reading_log_probs": [-387.173095703125, -392.2478942871094, -388.116455078125, -386.8082275390625], "reading_logprob_std": 2.5024464682396035, "tokenization_path_logprob_std": 216.88056248774217, "trajectory_balance_rmse": 5526.762111621209} {"canonical_length": 194, "canonical_nll": 0.991086507580944, "generation_rouge_l_f1": 0.22964509394572025, "generation_token_f1": 0.4342379958246347, "prompt_id": "1d19c88736fcd9d4", "reading_canonical_log_prob": -192.27078247070312, "reading_distances": [52, 48, 29, 47], "reading_log_probs": [-189.19781494140625, -188.45346069335938, -189.10772705078125, -188.73257446289062], "reading_logprob_std": 0.3446418677977189, "tokenization_path_logprob_std": 135.17902603673593, "trajectory_balance_rmse": 3430.7919176097935} {"canonical_length": 15, "canonical_nll": 2.691705067952474, "generation_rouge_l_f1": 0.09090909090909091, "generation_token_f1": 0.09090909090909091, "prompt_id": "1dc2cb66ddb80668", "reading_canonical_log_prob": -40.37557601928711, "reading_distances": [15, 15, 15, 13], "reading_log_probs": [-40.426753997802734, -40.626502990722656, -41.45264434814453, -41.418678283691406], "reading_logprob_std": 0.5313088705263718, "tokenization_path_logprob_std": 39.16563914108045, "trajectory_balance_rmse": 310.8796315487174} {"canonical_length": 16, "canonical_nll": 1.772430181503296, "generation_rouge_l_f1": 0.06766917293233084, "generation_token_f1": 0.07518796992481203, "prompt_id": "bf66528e05d6425b", "reading_canonical_log_prob": -28.358882904052734, "reading_distances": [29, 37, 31, 36], "reading_log_probs": [-26.97255516052246, -28.613712310791016, -28.34761619567871, -27.734214782714844], "reading_logprob_std": 0.7294346760880905, "tokenization_path_logprob_std": 85.74815425347352, "trajectory_balance_rmse": 537.858671510402} {"canonical_length": 118, "canonical_nll": 1.283430843029992, "generation_rouge_l_f1": 0.23076923076923075, "generation_token_f1": 0.43589743589743596, "prompt_id": "471e2be2592f5add", "reading_canonical_log_prob": -151.44483947753906, "reading_distances": [20, 22, 21, 20], "reading_log_probs": [-150.65432739257812, -146.41412353515625, -150.39736938476562, -148.82858276367188], "reading_logprob_std": 1.9479829727903095, "tokenization_path_logprob_std": 119.96022308204782, "trajectory_balance_rmse": 2617.7830220808296} {"canonical_length": 19, "canonical_nll": 3.993524250231291, "generation_rouge_l_f1": 0.2448979591836735, "generation_token_f1": 0.28571428571428575, "prompt_id": "57103ee2e4cc89e4", "reading_canonical_log_prob": -75.87696075439453, "reading_distances": [36, 32, 40, 39], "reading_log_probs": [-81.11756134033203, -78.87922668457031, -77.52481842041016, -78.98501586914062], "reading_logprob_std": 1.4844607211994534, "tokenization_path_logprob_std": 22.233574805536737, "trajectory_balance_rmse": 261.9394444915801} {"canonical_length": 23, "canonical_nll": 1.8591724893321162, "generation_rouge_l_f1": 0.16, "generation_token_f1": 0.32, "prompt_id": "aa0821e4d154fd43", "reading_canonical_log_prob": -42.76096725463867, "reading_distances": [20, 22, 25, 21], "reading_log_probs": [-47.825199127197266, -48.23749542236328, -43.70213317871094, -44.159271240234375], "reading_logprob_std": 2.380810203325906, "tokenization_path_logprob_std": 33.757315893290134, "trajectory_balance_rmse": 509.1289966952728} {"canonical_length": 24, "canonical_nll": 1.6453315416971843, "generation_rouge_l_f1": 0.40816326530612246, "generation_token_f1": 0.44897959183673464, "prompt_id": "2aaa42d342162266", "reading_canonical_log_prob": -39.48795700073242, "reading_distances": [31, 32, 27, 28], "reading_log_probs": [-36.01816177368164, -34.41838836669922, -37.80903244018555, -36.953529357910156], "reading_logprob_std": 1.451915338946754, "tokenization_path_logprob_std": 97.02217587519436, "trajectory_balance_rmse": 692.1903055963355} {"canonical_length": 10, "canonical_nll": 0.5168148040771484, "generation_rouge_l_f1": 0.375, "generation_token_f1": 0.5, "prompt_id": "b5604eaaebf6e52c", "reading_canonical_log_prob": -5.168148040771484, "reading_distances": [19, 20, 12, 15], "reading_log_probs": [-6.887319564819336, -6.517348766326904, -5.691925525665283, -7.4340009689331055], "reading_logprob_std": 0.7315000281949119, "tokenization_path_logprob_std": 48.1700916583256, "trajectory_balance_rmse": 232.37478835960655} {"canonical_length": 223, "canonical_nll": 1.282119272001121, "generation_rouge_l_f1": 0.14893617021276595, "generation_token_f1": 0.2723404255319149, "prompt_id": "5dbcfc1c200ae9f7", "reading_canonical_log_prob": -285.91259765625, "reading_distances": [23, 24, 25, 24], "reading_log_probs": [-281.9761962890625, -275.79644775390625, -278.7466125488281, -278.81976318359375], "reading_logprob_std": 2.5237510535279415, "tokenization_path_logprob_std": 239.74290627787113, "trajectory_balance_rmse": 4905.485044153399} {"canonical_length": 163, "canonical_nll": 1.2581411724441622, "generation_rouge_l_f1": 0.28478964401294493, "generation_token_f1": 0.4530744336569579, "prompt_id": "d526cee0379fafdd", "reading_canonical_log_prob": -205.07701110839844, "reading_distances": [34, 29, 22, 30], "reading_log_probs": [-205.64688110351562, -204.7091522216797, -204.6436767578125, -204.9378204345703], "reading_logprob_std": 0.4593127237010302, "tokenization_path_logprob_std": 128.96559167260202, "trajectory_balance_rmse": 4026.9666611536204} {"canonical_length": 85, "canonical_nll": 1.5544534122242648, "generation_rouge_l_f1": 0.34285714285714286, "generation_token_f1": 0.5485714285714285, "prompt_id": "b641694c0de50f89", "reading_canonical_log_prob": -132.1285400390625, "reading_distances": [35, 25, 31, 37], "reading_log_probs": [-132.748291015625, -128.32986450195312, -131.18890380859375, -132.8594512939453], "reading_logprob_std": 2.110460634742413, "tokenization_path_logprob_std": 61.68542331420921, "trajectory_balance_rmse": 1753.9648289198105} {"canonical_length": 13, "canonical_nll": 0.5337449587308444, "generation_rouge_l_f1": 0.23404255319148937, "generation_token_f1": 0.2553191489361702, "prompt_id": "2e5bfbd760499e1d", "reading_canonical_log_prob": -6.938684463500977, "reading_distances": [22, 24, 25, 19], "reading_log_probs": [-6.642521858215332, -6.533188819885254, -6.4609198570251465, -6.986615180969238], "reading_logprob_std": 0.2328280940246855, "tokenization_path_logprob_std": 16.75288625939645, "trajectory_balance_rmse": 351.9374380085821} {"canonical_length": 363, "canonical_nll": 1.4022109186682163, "generation_rouge_l_f1": 0.21788617886178863, "generation_token_f1": 0.44878048780487806, "prompt_id": "1d0a6d1714584f14", "reading_canonical_log_prob": -509.0025634765625, "reading_distances": [36, 39, 33, 28], "reading_log_probs": [-505.4385986328125, -500.99371337890625, -502.1786193847656, -505.9722595214844], "reading_logprob_std": 2.4367138558861767, "tokenization_path_logprob_std": 117.17440010161674, "trajectory_balance_rmse": 6946.402072285167} {"canonical_length": 37, "canonical_nll": 1.8368373561549831, "generation_rouge_l_f1": 0.05405405405405406, "generation_token_f1": 0.05405405405405406, "prompt_id": "c7b4cbd5453e19a7", "reading_canonical_log_prob": -67.96298217773438, "reading_distances": [24, 21, 16, 21], "reading_log_probs": [-66.32365417480469, -65.06681823730469, -63.374393463134766, -63.25261306762695], "reading_logprob_std": 1.4685463697084464, "tokenization_path_logprob_std": 51.45200782603316, "trajectory_balance_rmse": 748.2375251705157} {"canonical_length": 10, "canonical_nll": 0.27678885459899905, "generation_rouge_l_f1": 0.9333333333333333, "generation_token_f1": 0.9333333333333333, "prompt_id": "88beef2f88591814", "reading_canonical_log_prob": -2.7678885459899902, "reading_distances": [62, 63, 57, 70], "reading_log_probs": [-5.910678386688232, -7.9736433029174805, -5.225920677185059, -7.696025371551514], "reading_logprob_std": 1.342902632595596, "tokenization_path_logprob_std": 24.412576636803706, "trajectory_balance_rmse": 270.1810941859054} {"canonical_length": 65, "canonical_nll": 0.8721531794621394, "generation_rouge_l_f1": 0.594059405940594, "generation_token_f1": 0.6138613861386139, "prompt_id": "cfc50376a0f13c13", "reading_canonical_log_prob": -56.68995666503906, "reading_distances": [28, 24, 30, 17], "reading_log_probs": [-56.874717712402344, -57.810691833496094, -57.47422409057617, -57.2532844543457], "reading_logprob_std": 0.39279936377117886, "tokenization_path_logprob_std": 56.004689909227686, "trajectory_balance_rmse": 671.3743262961001} {"canonical_length": 90, "canonical_nll": 1.6224707709418402, "generation_rouge_l_f1": 0.25170068027210885, "generation_token_f1": 0.37414965986394555, "prompt_id": "6a7a68d7cf446159", "reading_canonical_log_prob": -146.02236938476562, "reading_distances": [21, 18, 14, 18], "reading_log_probs": [-143.5255584716797, -145.56893920898438, -147.21688842773438, -142.9040069580078], "reading_logprob_std": 1.9707484844593701, "tokenization_path_logprob_std": 82.72006246987463, "trajectory_balance_rmse": 2090.801564730079} {"canonical_length": 216, "canonical_nll": 1.586835507993345, "generation_rouge_l_f1": 0.21161825726141076, "generation_token_f1": 0.4605809128630705, "prompt_id": "a929bd728f13cca9", "reading_canonical_log_prob": -342.7564697265625, "reading_distances": [35, 36, 29, 38], "reading_log_probs": [-328.8062744140625, -332.1243896484375, -332.31585693359375, -331.79864501953125], "reading_logprob_std": 1.6505458051570587, "tokenization_path_logprob_std": 264.4234137830416, "trajectory_balance_rmse": 5523.023448934245} {"canonical_length": 321, "canonical_nll": 0.9200441354531737, "generation_rouge_l_f1": 0.34551495016611294, "generation_token_f1": 0.548172757475083, "prompt_id": "167bbbea26df8184", "reading_canonical_log_prob": -295.33416748046875, "reading_distances": [31, 34, 34, 37], "reading_log_probs": [-299.31085205078125, -298.1556091308594, -296.81732177734375, -297.1883544921875], "reading_logprob_std": 1.115105942908258, "tokenization_path_logprob_std": 247.42313672984574, "trajectory_balance_rmse": 5950.511069650863} {"canonical_length": 195, "canonical_nll": 1.6957128280248397, "generation_rouge_l_f1": 0.24752475247524752, "generation_token_f1": 0.4752475247524753, "prompt_id": "00b0a2b88466f41c", "reading_canonical_log_prob": -330.66400146484375, "reading_distances": [21, 26, 16, 25], "reading_log_probs": [-324.8770751953125, -323.48968505859375, -323.587890625, -325.03631591796875], "reading_logprob_std": 0.8221847437014087, "tokenization_path_logprob_std": 77.98909775678868, "trajectory_balance_rmse": 4049.615612268448} {"canonical_length": 89, "canonical_nll": 1.37132923254806, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "a26fac121253e753", "reading_canonical_log_prob": -122.04830169677734, "reading_distances": [29, 27, 29, 31], "reading_log_probs": [-121.51995849609375, -120.49949645996094, -120.44446563720703, -119.32638549804688], "reading_logprob_std": 0.896247318567133, "tokenization_path_logprob_std": 93.39323262275984, "trajectory_balance_rmse": 2094.0471562006915} {"canonical_length": 223, "canonical_nll": 0.9780413024628644, "generation_rouge_l_f1": 0.23234624145785876, "generation_token_f1": 0.4464692482915718, "prompt_id": "109439f51e6651d1", "reading_canonical_log_prob": -218.10321044921875, "reading_distances": [38, 33, 38, 41], "reading_log_probs": [-218.68594360351562, -218.71292114257812, -219.99966430664062, -219.18658447265625], "reading_logprob_std": 0.6136227729803633, "tokenization_path_logprob_std": 238.4324956480836, "trajectory_balance_rmse": 3566.853054012249} {"canonical_length": 174, "canonical_nll": 1.0030221171762752, "generation_rouge_l_f1": 0.2272727272727273, "generation_token_f1": 0.4545454545454546, "prompt_id": "25f04ac4bbf2d3c3", "reading_canonical_log_prob": -174.52584838867188, "reading_distances": [22, 20, 16, 20], "reading_log_probs": [-173.60855102539062, -172.49595642089844, -171.77842712402344, -172.09397888183594], "reading_logprob_std": 0.798809331232828, "tokenization_path_logprob_std": 101.53852236879659, "trajectory_balance_rmse": 3465.3892851763003} {"canonical_length": 13, "canonical_nll": 0.9697936131403997, "generation_rouge_l_f1": 0.11267605633802819, "generation_token_f1": 0.11267605633802819, "prompt_id": "68cd67de83285a28", "reading_canonical_log_prob": -12.607316970825195, "reading_distances": [10, 9, 11, 10], "reading_log_probs": [-11.884984970092773, -10.580156326293945, -11.130905151367188, -10.620059967041016], "reading_logprob_std": 0.6080795450464149, "tokenization_path_logprob_std": 19.834944066182086, "trajectory_balance_rmse": 185.92866344262768} {"canonical_length": 12, "canonical_nll": 0.34983237584431964, "generation_rouge_l_f1": 0.15384615384615385, "generation_token_f1": 0.15384615384615385, "prompt_id": "41752c3a8cc3cb46", "reading_canonical_log_prob": -4.197988510131836, "reading_distances": [20, 21, 21, 19], "reading_log_probs": [-4.459625244140625, -3.9731156826019287, -4.110177040100098, -4.188303470611572], "reading_logprob_std": 0.20486101885157482, "tokenization_path_logprob_std": 37.71925842470772, "trajectory_balance_rmse": 217.39776189126448} {"canonical_length": 33, "canonical_nll": 0.4923618201053504, "generation_rouge_l_f1": 0.23602484472049687, "generation_token_f1": 0.36024844720496896, "prompt_id": "6595c3b5c5d331ea", "reading_canonical_log_prob": -16.247940063476562, "reading_distances": [24, 21, 22, 23], "reading_log_probs": [-17.417272567749023, -17.291532516479492, -19.381656646728516, -19.369564056396484], "reading_logprob_std": 1.168083816141014, "tokenization_path_logprob_std": 57.607946967120974, "trajectory_balance_rmse": 353.09399246965313} {"canonical_length": 7, "canonical_nll": 0.3409247398376465, "generation_rouge_l_f1": 0.5714285714285715, "generation_token_f1": 0.8571428571428571, "prompt_id": "77f51b75ebf88292", "reading_canonical_log_prob": -2.3864731788635254, "reading_distances": [19, 22, 19, 26], "reading_log_probs": [-2.4721953868865967, -2.09450626373291, -2.758441686630249, -2.2354671955108643], "reading_logprob_std": 0.290804745904177, "tokenization_path_logprob_std": 38.8155191476144, "trajectory_balance_rmse": 136.51717285542276} {"canonical_length": 250, "canonical_nll": 1.035574462890625, "generation_rouge_l_f1": 0.21073558648111332, "generation_token_f1": 0.489065606361829, "prompt_id": "3d42fc04af88620e", "reading_canonical_log_prob": -258.89361572265625, "reading_distances": [29, 25, 33, 29], "reading_log_probs": [-254.6462860107422, -255.31082153320312, -254.44224548339844, -256.53521728515625], "reading_logprob_std": 0.9436366546292769, "tokenization_path_logprob_std": 77.1869580016882, "trajectory_balance_rmse": 4737.446180076861} {"canonical_length": 311, "canonical_nll": 0.6151130897055869, "generation_rouge_l_f1": 0.3700787401574803, "generation_token_f1": 0.5354330708661417, "prompt_id": "391687d0204ea44e", "reading_canonical_log_prob": -191.3001708984375, "reading_distances": [21, 24, 23, 26], "reading_log_probs": [-197.38470458984375, -201.5877227783203, -195.83529663085938, -199.49090576171875], "reading_logprob_std": 2.5058703842703896, "tokenization_path_logprob_std": 129.08507838074232, "trajectory_balance_rmse": 4150.221315705343} {"canonical_length": 45, "canonical_nll": 1.562620374891493, "generation_rouge_l_f1": 0.0816326530612245, "generation_token_f1": 0.0816326530612245, "prompt_id": "9043379228e84740", "reading_canonical_log_prob": -70.31791687011719, "reading_distances": [32, 31, 36, 33], "reading_log_probs": [-66.90267944335938, -71.44235229492188, -68.57998657226562, -69.44306182861328], "reading_logprob_std": 1.88879878635936, "tokenization_path_logprob_std": 63.04509107415807, "trajectory_balance_rmse": 985.1404154949646} {"canonical_length": 14, "canonical_nll": 1.3768585750034876, "generation_rouge_l_f1": 0.20408163265306123, "generation_token_f1": 0.20408163265306123, "prompt_id": "82a1a337beb39341", "reading_canonical_log_prob": -19.276020050048828, "reading_distances": [26, 23, 25, 34], "reading_log_probs": [-16.526836395263672, -14.736900329589844, -16.173124313354492, -15.962465286254883], "reading_logprob_std": 0.7776365970585152, "tokenization_path_logprob_std": 54.6365154634019, "trajectory_balance_rmse": 343.05807710551835} {"canonical_length": 15, "canonical_nll": 0.9899727503458658, "generation_rouge_l_f1": 0.14285714285714288, "generation_token_f1": 0.163265306122449, "prompt_id": "32a6cf563f35ecba", "reading_canonical_log_prob": -14.849591255187988, "reading_distances": [33, 25, 30, 33], "reading_log_probs": [-18.573068618774414, -16.805150985717773, -13.604249000549316, -11.05135726928711], "reading_logprob_std": 3.344896792406016, "tokenization_path_logprob_std": 29.62097924147796, "trajectory_balance_rmse": 185.8555518871177} {"canonical_length": 147, "canonical_nll": 1.4855950803172833, "generation_rouge_l_f1": 0.1878172588832487, "generation_token_f1": 0.3604060913705584, "prompt_id": "2655c987f350c6c7", "reading_canonical_log_prob": -218.38247680664062, "reading_distances": [14, 17, 14, 15], "reading_log_probs": [-214.70423889160156, -217.01480102539062, -215.82177734375, -216.6112518310547], "reading_logprob_std": 1.0179110693261868, "tokenization_path_logprob_std": 58.293365616054444, "trajectory_balance_rmse": 3044.0968089839193} {"canonical_length": 392, "canonical_nll": 0.9853455679757255, "generation_rouge_l_f1": 0.30823117338003503, "generation_token_f1": 0.4903677758318739, "prompt_id": "dd04ff95cd33338f", "reading_canonical_log_prob": -386.2554626464844, "reading_distances": [25, 30, 24, 23], "reading_log_probs": [-385.9345703125, -393.2449951171875, -383.65533447265625, -384.2698974609375], "reading_logprob_std": 4.418710402221526, "tokenization_path_logprob_std": 72.54508209532625, "trajectory_balance_rmse": 4198.385510015844} {"canonical_length": 197, "canonical_nll": 0.6290144169996232, "generation_rouge_l_f1": 0.43209876543209874, "generation_token_f1": 0.5967078189300411, "prompt_id": "f51cb0e119af1f71", "reading_canonical_log_prob": -123.91584014892578, "reading_distances": [38, 44, 38, 39], "reading_log_probs": [-123.0752944946289, -120.14793395996094, -120.36431884765625, -121.34571075439453], "reading_logprob_std": 1.3340074436452558, "tokenization_path_logprob_std": 196.6581369877964, "trajectory_balance_rmse": 3561.238550418308} {"canonical_length": 11, "canonical_nll": 2.1895389556884766, "generation_rouge_l_f1": 0.18181818181818182, "generation_token_f1": 0.18181818181818182, "prompt_id": "40f09143e7b23121", "reading_canonical_log_prob": -24.084928512573242, "reading_distances": [42, 47, 49, 42], "reading_log_probs": [-23.10079574584961, -22.510528564453125, -22.145984649658203, -22.073566436767578], "reading_logprob_std": 0.46942803785430304, "tokenization_path_logprob_std": 88.04197593493099, "trajectory_balance_rmse": 297.04967451879884} {"canonical_length": 98, "canonical_nll": 1.2361348989058514, "generation_rouge_l_f1": 0.2228571428571429, "generation_token_f1": 0.3542857142857143, "prompt_id": "09e3575b08c40649", "reading_canonical_log_prob": -121.14122009277344, "reading_distances": [16, 17, 13, 13], "reading_log_probs": [-122.45222473144531, -123.59642028808594, -120.93254852294922, -121.57769775390625], "reading_logprob_std": 1.1536560603233945, "tokenization_path_logprob_std": 65.11828472107031, "trajectory_balance_rmse": 1962.5086297714943} {"canonical_length": 11, "canonical_nll": 0.6614182645624335, "generation_rouge_l_f1": 0.06866952789699571, "generation_token_f1": 0.07725321888412016, "prompt_id": "fc90d5db5297a0af", "reading_canonical_log_prob": -7.275600910186768, "reading_distances": [18, 22, 20, 20], "reading_log_probs": [-7.155453681945801, -7.884045124053955, -7.211928367614746, -7.226840972900391], "reading_logprob_std": 0.344360717044488, "tokenization_path_logprob_std": 48.637291827212444, "trajectory_balance_rmse": 251.0549380630029} {"canonical_length": 244, "canonical_nll": 0.9760984201900295, "generation_rouge_l_f1": 0.30957230142566194, "generation_token_f1": 0.5784114052953157, "prompt_id": "e15ab7087e38fd8b", "reading_canonical_log_prob": -238.1680145263672, "reading_distances": [21, 16, 16, 15], "reading_log_probs": [-238.54971313476562, -236.40234375, -235.38604736328125, -235.89715576171875], "reading_logprob_std": 1.3906036473929362, "tokenization_path_logprob_std": 117.4273180476913, "trajectory_balance_rmse": 4478.088987334121} {"canonical_length": 83, "canonical_nll": 0.9593730145190136, "generation_rouge_l_f1": 0.40462427745664736, "generation_token_f1": 0.5664739884393065, "prompt_id": "514750b55e58033f", "reading_canonical_log_prob": -79.62796020507812, "reading_distances": [17, 18, 16, 13], "reading_log_probs": [-82.34839630126953, -77.67921447753906, -81.91212463378906, -79.92717742919922], "reading_logprob_std": 2.136289109751235, "tokenization_path_logprob_std": 165.7641469444571, "trajectory_balance_rmse": 1735.1129594266254} {"canonical_length": 145, "canonical_nll": 1.2648324900660022, "generation_rouge_l_f1": 0.2014388489208633, "generation_token_f1": 0.32613908872901676, "prompt_id": "612827ad6442cf24", "reading_canonical_log_prob": -183.4007110595703, "reading_distances": [27, 26, 28, 29], "reading_log_probs": [-179.55960083007812, -178.14564514160156, -178.06553649902344, -178.8070526123047], "reading_logprob_std": 0.6947152364789742, "tokenization_path_logprob_std": 235.59987814758736, "trajectory_balance_rmse": 3488.3399692478743} {"canonical_length": 184, "canonical_nll": 0.8833759142004926, "generation_rouge_l_f1": 0.20956719817767652, "generation_token_f1": 0.36902050113895224, "prompt_id": "637d5437d0c071a7", "reading_canonical_log_prob": -162.54116821289062, "reading_distances": [249, 215, 215, 227], "reading_log_probs": [-165.88510131835938, -169.17958068847656, -160.71510314941406, -166.626220703125], "reading_logprob_std": 3.5501182056303735, "tokenization_path_logprob_std": 97.03662497150368, "trajectory_balance_rmse": 2885.8674915036736} {"canonical_length": 65, "canonical_nll": 1.0544475848858172, "generation_rouge_l_f1": 0.17391304347826086, "generation_token_f1": 0.22073578595317725, "prompt_id": "c97d2c584d8df586", "reading_canonical_log_prob": -68.53909301757812, "reading_distances": [15, 19, 16, 17], "reading_log_probs": [-67.51669311523438, -67.60881805419922, -68.50872802734375, -68.12246704101562], "reading_logprob_std": 0.46390131056806744, "tokenization_path_logprob_std": 62.466243046787504, "trajectory_balance_rmse": 937.8246589220653} {"canonical_length": 17, "canonical_nll": 0.8698446610394646, "generation_rouge_l_f1": 0.689655172413793, "generation_token_f1": 0.7586206896551724, "prompt_id": "a17a07dc87573da6", "reading_canonical_log_prob": -14.787359237670898, "reading_distances": [29, 25, 21, 24], "reading_log_probs": [-15.9412841796875, -16.766632080078125, -16.683027267456055, -16.373075485229492], "reading_logprob_std": 0.3736946571981312, "tokenization_path_logprob_std": 91.04825019196628, "trajectory_balance_rmse": 475.2879118399879} {"canonical_length": 13, "canonical_nll": 0.6037716498741736, "generation_rouge_l_f1": 0.1728395061728395, "generation_token_f1": 0.1728395061728395, "prompt_id": "4b95b246a8b6c8d6", "reading_canonical_log_prob": -7.849031448364258, "reading_distances": [10, 12, 14, 12], "reading_log_probs": [-6.514091968536377, -5.2419867515563965, -5.545042991638184, -7.1116437911987305], "reading_logprob_std": 0.8639094464112259, "tokenization_path_logprob_std": 32.50183147423662, "trajectory_balance_rmse": 337.2600109532138} {"canonical_length": 99, "canonical_nll": 1.1307385377209596, "generation_rouge_l_f1": 0.29906542056074764, "generation_token_f1": 0.48598130841121495, "prompt_id": "0c93e5e51c22b30d", "reading_canonical_log_prob": -111.943115234375, "reading_distances": [17, 18, 18, 15], "reading_log_probs": [-113.1279296875, -115.8648681640625, -115.95161437988281, -115.8877944946289], "reading_logprob_std": 1.3872335996645795, "tokenization_path_logprob_std": 118.83264983524363, "trajectory_balance_rmse": 1941.01367341456} {"canonical_length": 306, "canonical_nll": 1.1825675216375613, "generation_rouge_l_f1": 0.22491349480968859, "generation_token_f1": 0.46020761245674746, "prompt_id": "6c10022ee752e034", "reading_canonical_log_prob": -361.86566162109375, "reading_distances": [29, 34, 30, 36], "reading_log_probs": [-363.72802734375, -364.0471496582031, -361.7369689941406, -363.8717041015625], "reading_logprob_std": 1.0805710387213219, "tokenization_path_logprob_std": 224.89226646370273, "trajectory_balance_rmse": 6352.181968307406} {"canonical_length": 69, "canonical_nll": 1.1186459306357563, "generation_rouge_l_f1": 0.35443037974683544, "generation_token_f1": 0.4683544303797469, "prompt_id": "7884c89879ae1c04", "reading_canonical_log_prob": -77.18656921386719, "reading_distances": [35, 28, 26, 37], "reading_log_probs": [-77.60600280761719, -78.00206756591797, -78.18733215332031, -80.06536865234375], "reading_logprob_std": 1.0939929498328, "tokenization_path_logprob_std": 134.40545027542214, "trajectory_balance_rmse": 1287.2671371005051} {"canonical_length": 46, "canonical_nll": 1.0980893010678499, "generation_rouge_l_f1": 0.47619047619047616, "generation_token_f1": 0.47619047619047616, "prompt_id": "803eb754dd6450de", "reading_canonical_log_prob": -50.512107849121094, "reading_distances": [14, 15, 17, 13], "reading_log_probs": [-46.794677734375, -49.65126037597656, -51.48908615112305, -51.20814514160156], "reading_logprob_std": 2.151678626294758, "tokenization_path_logprob_std": 61.5316530148249, "trajectory_balance_rmse": 1079.8328561770174} {"canonical_length": 12, "canonical_nll": 0.9836553732554117, "generation_rouge_l_f1": 0.5, "generation_token_f1": 0.5, "prompt_id": "7f58078c29e97e0d", "reading_canonical_log_prob": -11.803864479064941, "reading_distances": [39, 46, 39, 39], "reading_log_probs": [-22.824199676513672, -14.851890563964844, -14.007769584655762, -25.353126525878906], "reading_logprob_std": 5.681756214993111, "tokenization_path_logprob_std": 17.42210067935256, "trajectory_balance_rmse": 180.9687795417458} {"canonical_length": 135, "canonical_nll": 1.3938969364872684, "generation_rouge_l_f1": 0.23057644110275688, "generation_token_f1": 0.3759398496240602, "prompt_id": "979bfe699ee39dca", "reading_canonical_log_prob": -188.17608642578125, "reading_distances": [22, 21, 20, 29], "reading_log_probs": [-187.11842346191406, -188.54685974121094, -187.26914978027344, -188.0545654296875], "reading_logprob_std": 0.6727607664607427, "tokenization_path_logprob_std": 199.56727076643014, "trajectory_balance_rmse": 2485.5983567516787} {"canonical_length": 13, "canonical_nll": 1.8905409299410307, "generation_rouge_l_f1": 0.10062893081761007, "generation_token_f1": 0.12578616352201258, "prompt_id": "f30d075da952702b", "reading_canonical_log_prob": -24.5770320892334, "reading_distances": [34, 26, 31, 29], "reading_log_probs": [-26.03642463684082, -21.07926368713379, -24.069393157958984, -21.116687774658203], "reading_logprob_std": 2.4205235695424907, "tokenization_path_logprob_std": 66.18216610610898, "trajectory_balance_rmse": 188.5126757464062} {"canonical_length": 117, "canonical_nll": 1.1489698621961806, "generation_rouge_l_f1": 0.28795811518324604, "generation_token_f1": 0.38219895287958117, "prompt_id": "ad627413d7298d7b", "reading_canonical_log_prob": -134.42947387695312, "reading_distances": [24, 17, 22, 29], "reading_log_probs": [-133.42312622070312, -132.1106414794922, -131.94207763671875, -133.41445922851562], "reading_logprob_std": 0.8068693752403293, "tokenization_path_logprob_std": 118.3019156504209, "trajectory_balance_rmse": 2194.4605619955614} {"canonical_length": 314, "canonical_nll": 1.468189506773736, "generation_rouge_l_f1": 0.2809611829944547, "generation_token_f1": 0.47689463955637706, "prompt_id": "802f028e684a8238", "reading_canonical_log_prob": -461.0115051269531, "reading_distances": [24, 22, 23, 19], "reading_log_probs": [-451.74676513671875, -451.0020446777344, -451.75811767578125, -452.1556701660156], "reading_logprob_std": 0.4815328729897054, "tokenization_path_logprob_std": 291.19108139901164, "trajectory_balance_rmse": 5767.775219974997} {"canonical_length": 190, "canonical_nll": 1.4180712248149672, "generation_rouge_l_f1": 0.23605150214592274, "generation_token_f1": 0.4291845493562232, "prompt_id": "2e2423f516c40a85", "reading_canonical_log_prob": -269.43353271484375, "reading_distances": [8, 16, 21, 16], "reading_log_probs": [-267.77197265625, -262.73138427734375, -264.94573974609375, -263.77813720703125], "reading_logprob_std": 2.173863741330724, "tokenization_path_logprob_std": 158.9897549480472, "trajectory_balance_rmse": 4039.6791920174887} {"canonical_length": 297, "canonical_nll": 1.4888424857296927, "generation_rouge_l_f1": 0.27177700348432055, "generation_token_f1": 0.505226480836237, "prompt_id": "8d1b09dca5de1a98", "reading_canonical_log_prob": -442.18621826171875, "reading_distances": [26, 21, 28, 23], "reading_log_probs": [-437.6494140625, -436.8204345703125, -437.5051574707031, -436.2575988769531], "reading_logprob_std": 0.6446703541566375, "tokenization_path_logprob_std": 194.3881894403071, "trajectory_balance_rmse": 7102.1536053704285} {"canonical_length": 82, "canonical_nll": 0.669726022859899, "generation_rouge_l_f1": 0.34334763948497854, "generation_token_f1": 0.3776824034334764, "prompt_id": "0609bdeeadeb9baf", "reading_canonical_log_prob": -54.91753387451172, "reading_distances": [35, 36, 39, 34], "reading_log_probs": [-53.93333053588867, -53.78414535522461, -53.709014892578125, -53.11726379394531], "reading_logprob_std": 0.3581298204090442, "tokenization_path_logprob_std": 91.63608831058171, "trajectory_balance_rmse": 1006.8999194559948} {"canonical_length": 222, "canonical_nll": 1.0547965109885276, "generation_rouge_l_f1": 0.34468937875751504, "generation_token_f1": 0.5771543086172345, "prompt_id": "e9f992cd65ad5461", "reading_canonical_log_prob": -234.16482543945312, "reading_distances": [25, 32, 29, 24], "reading_log_probs": [-231.79977416992188, -233.1499481201172, -232.66465759277344, -232.8268280029297], "reading_logprob_std": 0.5767759064255786, "tokenization_path_logprob_std": 115.24707468816527, "trajectory_balance_rmse": 3045.137264709065} {"canonical_length": 6, "canonical_nll": 1.356082280476888, "generation_rouge_l_f1": 0.5, "generation_token_f1": 0.5, "prompt_id": "15eeee68ceb20913", "reading_canonical_log_prob": -8.136493682861328, "reading_distances": [32, 38, 36, 31], "reading_log_probs": [-8.306304931640625, -8.167472839355469, -6.996826171875, -7.469419479370117], "reading_logprob_std": 0.6134209253649261, "tokenization_path_logprob_std": 50.05968868236064, "trajectory_balance_rmse": 157.50248074324682} {"canonical_length": 105, "canonical_nll": 1.3149650937034971, "generation_rouge_l_f1": 0.20498614958448755, "generation_token_f1": 0.3268698060941828, "prompt_id": "728a7ca9e481ba81", "reading_canonical_log_prob": -138.0713348388672, "reading_distances": [29, 30, 30, 24], "reading_log_probs": [-132.90243530273438, -129.18618774414062, -132.26675415039062, -128.71807861328125], "reading_logprob_std": 2.1218219560509524, "tokenization_path_logprob_std": 121.305630539253, "trajectory_balance_rmse": 2028.9421084567139} {"canonical_length": 123, "canonical_nll": 1.641124694327998, "generation_rouge_l_f1": 0.08533333333333332, "generation_token_f1": 0.11733333333333333, "prompt_id": "92cf4663b97aed32", "reading_canonical_log_prob": -201.85833740234375, "reading_distances": [26, 24, 30, 22], "reading_log_probs": [-199.90257263183594, -198.34194946289062, -199.8819580078125, -199.48629760742188], "reading_logprob_std": 0.7329711613747865, "tokenization_path_logprob_std": 158.72762204543355, "trajectory_balance_rmse": 1798.2915341034125} {"canonical_length": 66, "canonical_nll": 0.8999933762983843, "generation_rouge_l_f1": 0.4489795918367347, "generation_token_f1": 0.4489795918367347, "prompt_id": "3226fde653ee5231", "reading_canonical_log_prob": -59.39956283569336, "reading_distances": [8, 13, 12, 12], "reading_log_probs": [-61.45463180541992, -62.83794403076172, -62.69264602661133, -62.635643005371094], "reading_logprob_std": 0.639420697348848, "tokenization_path_logprob_std": 72.74442171654549, "trajectory_balance_rmse": 1468.2723113306658} {"canonical_length": 259, "canonical_nll": 1.19167254238055, "generation_rouge_l_f1": 0.30446194225721784, "generation_token_f1": 0.4724409448818898, "prompt_id": "d40a7f2ec7ebb044", "reading_canonical_log_prob": -308.6431884765625, "reading_distances": [22, 24, 26, 20], "reading_log_probs": [-309.6195373535156, -311.3135070800781, -310.56219482421875, -311.7846374511719], "reading_logprob_std": 0.9454538100394453, "tokenization_path_logprob_std": 66.54464560066943, "trajectory_balance_rmse": 4566.00708939794} {"canonical_length": 13, "canonical_nll": 1.1925087708693285, "generation_rouge_l_f1": 0.3571428571428571, "generation_token_f1": 0.3571428571428571, "prompt_id": "6844823da97b3fe4", "reading_canonical_log_prob": -15.50261402130127, "reading_distances": [20, 24, 22, 21], "reading_log_probs": [-14.687236785888672, -14.415993690490723, -15.467086791992188, -14.746648788452148], "reading_logprob_std": 0.4489273376288122, "tokenization_path_logprob_std": 89.69584883340285, "trajectory_balance_rmse": 311.5818673923069} {"canonical_length": 118, "canonical_nll": 0.8830263816704185, "generation_rouge_l_f1": 0.17391304347826086, "generation_token_f1": 0.21739130434782608, "prompt_id": "49597fd7a90f1c14", "reading_canonical_log_prob": -104.19711303710938, "reading_distances": [24, 25, 31, 24], "reading_log_probs": [-108.40509033203125, -107.09007263183594, -108.28787231445312, -108.46484375], "reading_logprob_std": 0.6520877166485021, "tokenization_path_logprob_std": 148.46823015801064, "trajectory_balance_rmse": 1839.016945572978} {"canonical_length": 258, "canonical_nll": 1.635764898255814, "generation_rouge_l_f1": 0.16666666666666666, "generation_token_f1": 0.2992424242424242, "prompt_id": "06dc86d1f57bb10a", "reading_canonical_log_prob": -422.02734375, "reading_distances": [29, 34, 39, 30], "reading_log_probs": [-422.15826416015625, -418.084716796875, -419.2513427734375, -426.1080627441406], "reading_logprob_std": 3.5753157673931013, "tokenization_path_logprob_std": 105.31589138445219, "trajectory_balance_rmse": 5213.1993874625705} {"canonical_length": 250, "canonical_nll": 1.3313876953125, "generation_rouge_l_f1": 0.21673003802281368, "generation_token_f1": 0.5019011406844107, "prompt_id": "104414445caa5ddc", "reading_canonical_log_prob": -332.846923828125, "reading_distances": [37, 32, 36, 35], "reading_log_probs": [-328.0399169921875, -326.8040771484375, -327.93292236328125, -326.789794921875], "reading_logprob_std": 0.6881611372021597, "tokenization_path_logprob_std": 294.02833507221544, "trajectory_balance_rmse": 5484.472427878342} {"canonical_length": 274, "canonical_nll": 0.9401295237297559, "generation_rouge_l_f1": 0.39468690702087283, "generation_token_f1": 0.6375711574952562, "prompt_id": "96cee7006ba2cc41", "reading_canonical_log_prob": -257.5954895019531, "reading_distances": [20, 20, 21, 18], "reading_log_probs": [-260.7330322265625, -263.3212585449219, -262.8885803222656, -261.10955810546875], "reading_logprob_std": 1.282277405358613, "tokenization_path_logprob_std": 74.09067047296662, "trajectory_balance_rmse": 4546.316838443343} {"canonical_length": 150, "canonical_nll": 1.9738956705729167, "generation_rouge_l_f1": 0.33980582524271846, "generation_token_f1": 0.5048543689320388, "prompt_id": "c61662a19e060e22", "reading_canonical_log_prob": -296.0843505859375, "reading_distances": [28, 29, 33, 33], "reading_log_probs": [-290.81329345703125, -289.40692138671875, -292.892578125, -290.1212463378906], "reading_logprob_std": 1.5033456924737028, "tokenization_path_logprob_std": 206.13357363669923, "trajectory_balance_rmse": 3142.900130344835} {"canonical_length": 80, "canonical_nll": 1.1909406661987305, "generation_rouge_l_f1": 0.2292263610315186, "generation_token_f1": 0.30945558739255014, "prompt_id": "448e404116b3a283", "reading_canonical_log_prob": -95.27525329589844, "reading_distances": [18, 27, 26, 24], "reading_log_probs": [-93.24418640136719, -93.45336151123047, -91.89608764648438, -93.626220703125], "reading_logprob_std": 0.7882161965976711, "tokenization_path_logprob_std": 80.95635958319701, "trajectory_balance_rmse": 1330.1502983465653} {"canonical_length": 40, "canonical_nll": 2.2998983383178713, "generation_rouge_l_f1": 0.2631578947368421, "generation_token_f1": 0.34210526315789475, "prompt_id": "62e94a0db3b86cbd", "reading_canonical_log_prob": -91.99593353271484, "reading_distances": [16, 17, 15, 16], "reading_log_probs": [-89.26680755615234, -88.97003936767578, -90.32908630371094, -88.20744323730469], "reading_logprob_std": 0.8788577335044574, "tokenization_path_logprob_std": 64.01060890239243, "trajectory_balance_rmse": 1060.8243569692481} {"canonical_length": 203, "canonical_nll": 1.0237950667959128, "generation_rouge_l_f1": 0.24019607843137256, "generation_token_f1": 0.5049019607843136, "prompt_id": "6996ebbe576a8626", "reading_canonical_log_prob": -207.8303985595703, "reading_distances": [33, 32, 37, 29], "reading_log_probs": [-204.80978393554688, -205.712646484375, -207.19769287109375, -205.98255920410156], "reading_logprob_std": 0.9852003676371722, "tokenization_path_logprob_std": 307.86695233158895, "trajectory_balance_rmse": 4020.0048295966794} {"canonical_length": 97, "canonical_nll": 1.3372256878725033, "generation_rouge_l_f1": 0.24083769633507854, "generation_token_f1": 0.450261780104712, "prompt_id": "539a3dd889cf333b", "reading_canonical_log_prob": -129.7108917236328, "reading_distances": [39, 25, 33, 33], "reading_log_probs": [-132.83016967773438, -130.81558227539062, -133.78759765625, -134.39151000976562], "reading_logprob_std": 1.565190592345969, "tokenization_path_logprob_std": 164.02022004102685, "trajectory_balance_rmse": 2453.9114759446597} {"canonical_length": 83, "canonical_nll": 1.8802504941641567, "generation_rouge_l_f1": 0.21114369501466274, "generation_token_f1": 0.30498533724340177, "prompt_id": "22dd62a830e20aab", "reading_canonical_log_prob": -156.060791015625, "reading_distances": [24, 37, 29, 27], "reading_log_probs": [-151.1122589111328, -152.86375427246094, -150.13009643554688, -152.41432189941406], "reading_logprob_std": 1.2456694234994832, "tokenization_path_logprob_std": 102.4734233017574, "trajectory_balance_rmse": 1732.926288535758} {"canonical_length": 26, "canonical_nll": 1.2631532228910005, "generation_rouge_l_f1": 0.13245033112582782, "generation_token_f1": 0.15894039735099338, "prompt_id": "23f859fe909ddbe9", "reading_canonical_log_prob": -32.841983795166016, "reading_distances": [17, 14, 14, 16], "reading_log_probs": [-34.40424346923828, -32.99744415283203, -31.778751373291016, -34.243141174316406], "reading_logprob_std": 1.225031879299139, "tokenization_path_logprob_std": 93.2686438088084, "trajectory_balance_rmse": 517.4816105530134} {"canonical_length": 97, "canonical_nll": 0.6691526432627255, "generation_rouge_l_f1": 0.29283489096573206, "generation_token_f1": 0.39875389408099693, "prompt_id": "9658518adf176971", "reading_canonical_log_prob": -64.90780639648438, "reading_distances": [16, 13, 18, 14], "reading_log_probs": [-60.82337951660156, -60.813079833984375, -61.68458557128906, -62.862674713134766], "reading_logprob_std": 0.9681925727709104, "tokenization_path_logprob_std": 75.49273448949597, "trajectory_balance_rmse": 1110.7191302404892} {"canonical_length": 243, "canonical_nll": 1.3242909623762218, "generation_rouge_l_f1": 0.2454728370221328, "generation_token_f1": 0.4627766599597586, "prompt_id": "c496b50e51f7e82e", "reading_canonical_log_prob": -321.8027038574219, "reading_distances": [30, 37, 26, 33], "reading_log_probs": [-320.2013854980469, -321.2625732421875, -321.2115478515625, -320.8290710449219], "reading_logprob_std": 0.48967274851602266, "tokenization_path_logprob_std": 113.3403324946475, "trajectory_balance_rmse": 5375.325723124504} {"canonical_length": 157, "canonical_nll": 1.0840162532344746, "generation_rouge_l_f1": 0.2184466019417476, "generation_token_f1": 0.43203883495145634, "prompt_id": "9433be6b41e84522", "reading_canonical_log_prob": -170.1905517578125, "reading_distances": [41, 34, 38, 35], "reading_log_probs": [-168.9231414794922, -169.1857452392578, -172.267333984375, -166.63597106933594], "reading_logprob_std": 2.3128899404922456, "tokenization_path_logprob_std": 192.92585440234865, "trajectory_balance_rmse": 3401.9566223755546} {"canonical_length": 14, "canonical_nll": 1.4616564341953822, "generation_rouge_l_f1": 0.3636363636363637, "generation_token_f1": 0.3636363636363637, "prompt_id": "22dd55d64faca0d8", "reading_canonical_log_prob": -20.46319007873535, "reading_distances": [40, 42, 49, 49], "reading_log_probs": [-22.246826171875, -22.314109802246094, -23.71847152709961, -21.76820182800293], "reading_logprob_std": 0.8402949715629388, "tokenization_path_logprob_std": 42.0135987911196, "trajectory_balance_rmse": 403.04655846840336} {"canonical_length": 52, "canonical_nll": 1.2383810190054088, "generation_rouge_l_f1": 0.07692307692307693, "generation_token_f1": 0.07692307692307693, "prompt_id": "9ce142b4fe68303a", "reading_canonical_log_prob": -64.39581298828125, "reading_distances": [31, 27, 33, 38], "reading_log_probs": [-65.4314193725586, -65.34674072265625, -65.54010009765625, -63.5997314453125], "reading_logprob_std": 0.9232425830824001, "tokenization_path_logprob_std": 87.52489162195961, "trajectory_balance_rmse": 1213.7791475907009} {"canonical_length": 28, "canonical_nll": 1.080104214804513, "generation_rouge_l_f1": 0.6779661016949152, "generation_token_f1": 0.711864406779661, "prompt_id": "e62a412929959b77", "reading_canonical_log_prob": -30.242918014526367, "reading_distances": [27, 24, 28, 24], "reading_log_probs": [-31.355655670166016, -30.06391143798828, -29.95779037475586, -31.685930252075195], "reading_logprob_std": 0.8831942969427622, "tokenization_path_logprob_std": 45.76775512626289, "trajectory_balance_rmse": 764.6174766252321} {"canonical_length": 21, "canonical_nll": 1.161252702985491, "generation_rouge_l_f1": 0.4210526315789474, "generation_token_f1": 0.4210526315789474, "prompt_id": "273ee8648fccd69e", "reading_canonical_log_prob": -24.386306762695312, "reading_distances": [31, 32, 30, 28], "reading_log_probs": [-22.71640396118164, -22.03042984008789, -21.939754486083984, -22.583595275878906], "reading_logprob_std": 0.3894577459593654, "tokenization_path_logprob_std": 39.98684707989829, "trajectory_balance_rmse": 399.7812756568589} {"canonical_length": 109, "canonical_nll": 1.2035640051605505, "generation_rouge_l_f1": 0.26573426573426573, "generation_token_f1": 0.3916083916083916, "prompt_id": "0ffd3304b685d298", "reading_canonical_log_prob": -131.1884765625, "reading_distances": [22, 24, 24, 19], "reading_log_probs": [-132.64260864257812, -132.66470336914062, -132.06832885742188, -135.755859375], "reading_logprob_std": 1.6716111215522134, "tokenization_path_logprob_std": 146.8145504234221, "trajectory_balance_rmse": 2229.3422474190047} {"canonical_length": 288, "canonical_nll": 1.1722164154052734, "generation_rouge_l_f1": 0.38076923076923075, "generation_token_f1": 0.6076923076923078, "prompt_id": "6b90af69594f1bfc", "reading_canonical_log_prob": -337.59832763671875, "reading_distances": [30, 33, 33, 33], "reading_log_probs": [-334.73480224609375, -335.5665283203125, -335.9382629394531, -335.2958984375], "reading_logprob_std": 0.5065384283641254, "tokenization_path_logprob_std": 157.78301963284716, "trajectory_balance_rmse": 6715.335303036489} {"canonical_length": 256, "canonical_nll": 1.211787462234497, "generation_rouge_l_f1": 0.2958579881656805, "generation_token_f1": 0.5404339250493096, "prompt_id": "7021b5d065291a9c", "reading_canonical_log_prob": -310.21759033203125, "reading_distances": [18, 24, 19, 23], "reading_log_probs": [-304.4952087402344, -304.8726806640625, -305.2147216796875, -304.4881591796875], "reading_logprob_std": 0.3479673732723696, "tokenization_path_logprob_std": 167.89815062183504, "trajectory_balance_rmse": 5387.454450494915} {"canonical_length": 68, "canonical_nll": 0.7742023468017578, "generation_rouge_l_f1": 0.35748792270531404, "generation_token_f1": 0.5314009661835749, "prompt_id": "56589cd90846fa45", "reading_canonical_log_prob": -52.64575958251953, "reading_distances": [39, 45, 32, 27], "reading_log_probs": [-55.682350158691406, -52.53937911987305, -54.91178894042969, -51.55712890625], "reading_logprob_std": 1.9437157222549453, "tokenization_path_logprob_std": 48.86198611890857, "trajectory_balance_rmse": 1039.5170964392491} {"canonical_length": 187, "canonical_nll": 1.6216130384149399, "generation_rouge_l_f1": 0.23157894736842108, "generation_token_f1": 0.4894736842105263, "prompt_id": "b2517684dcd5d425", "reading_canonical_log_prob": -303.24163818359375, "reading_distances": [47, 36, 38, 36], "reading_log_probs": [-296.4006042480469, -299.0533447265625, -298.49420166015625, -297.2216796875], "reading_logprob_std": 1.2035121174624184, "tokenization_path_logprob_std": 112.09443019761244, "trajectory_balance_rmse": 4366.891832861035} {"canonical_length": 2, "canonical_nll": 3.6942460536956787, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "44fca97732ef92a8", "reading_canonical_log_prob": -7.388492107391357, "reading_distances": [22, 24, 22, 18], "reading_log_probs": [-10.127691268920898, -11.858540534973145, -12.586313247680664, -7.959505558013916], "reading_logprob_std": 2.0591496376596794, "tokenization_path_logprob_std": 19.52301104031638, "trajectory_balance_rmse": 30.08054177696105} {"canonical_length": 14, "canonical_nll": 0.2599175827843802, "generation_rouge_l_f1": 0.47058823529411764, "generation_token_f1": 0.47058823529411764, "prompt_id": "907693271b754ffe", "reading_canonical_log_prob": -3.6388461589813232, "reading_distances": [5, 6, 7, 8], "reading_log_probs": [-3.5690250396728516, -4.293575286865234, -4.017091751098633, -4.003953456878662], "reading_logprob_std": 0.2993598909085562, "tokenization_path_logprob_std": 23.583131995805335, "trajectory_balance_rmse": 195.53866089546034} {"canonical_length": 220, "canonical_nll": 1.1233068292791193, "generation_rouge_l_f1": 0.18448637316561844, "generation_token_f1": 0.389937106918239, "prompt_id": "3d885c9990d3f7a8", "reading_canonical_log_prob": -247.12750244140625, "reading_distances": [43, 52, 46, 50], "reading_log_probs": [-243.62628173828125, -248.38674926757812, -246.0413818359375, -255.27340698242188], "reading_logprob_std": 5.019191312751689, "tokenization_path_logprob_std": 94.19643854336825, "trajectory_balance_rmse": 3985.875564551138} {"canonical_length": 12, "canonical_nll": 1.0699489911397297, "generation_rouge_l_f1": 0.16666666666666669, "generation_token_f1": 0.16666666666666669, "prompt_id": "946f59f054cd3b7a", "reading_canonical_log_prob": -12.839387893676758, "reading_distances": [27, 25, 33, 27], "reading_log_probs": [-14.883477210998535, -14.080188751220703, -13.4861421585083, -13.827311515808105], "reading_logprob_std": 0.5948766227189961, "tokenization_path_logprob_std": 47.8516931053345, "trajectory_balance_rmse": 280.38101680142853} {"canonical_length": 114, "canonical_nll": 1.615160624186198, "generation_rouge_l_f1": 0.2521739130434783, "generation_token_f1": 0.408695652173913, "prompt_id": "1e094a0368e082c2", "reading_canonical_log_prob": -184.12831115722656, "reading_distances": [30, 31, 29, 21], "reading_log_probs": [-181.99215698242188, -177.19155883789062, -179.6817169189453, -180.57130432128906], "reading_logprob_std": 2.0169620444380674, "tokenization_path_logprob_std": 186.82784933067904, "trajectory_balance_rmse": 2554.761026904676} {"canonical_length": 8, "canonical_nll": 0.08203766494989395, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "f01addec53ef2695", "reading_canonical_log_prob": -0.6563013195991516, "reading_distances": [24, 25, 23, 24], "reading_log_probs": [-1.8743846416473389, -0.9878767728805542, -1.0376921892166138, -2.2497713565826416], "reading_logprob_std": 0.6252240312569307, "tokenization_path_logprob_std": 42.05179437130348, "trajectory_balance_rmse": 144.37750557733207} {"canonical_length": 43, "canonical_nll": 1.154495505399482, "generation_rouge_l_f1": 0.3384615384615385, "generation_token_f1": 0.4615384615384615, "prompt_id": "bd70996cafa5e523", "reading_canonical_log_prob": -49.643306732177734, "reading_distances": [32, 29, 25, 18], "reading_log_probs": [-57.35987091064453, -61.764122009277344, -65.92581176757812, -62.80089569091797], "reading_logprob_std": 3.541859010008016, "tokenization_path_logprob_std": 110.00063581847355, "trajectory_balance_rmse": 811.4879282172959} {"canonical_length": 280, "canonical_nll": 1.0796643938337054, "generation_rouge_l_f1": 0.20599250936329586, "generation_token_f1": 0.4943820224719101, "prompt_id": "40b3b0fca502cfb8", "reading_canonical_log_prob": -302.3060302734375, "reading_distances": [29, 29, 26, 34], "reading_log_probs": [-292.8890380859375, -292.83721923828125, -295.82171630859375, -295.3175354003906], "reading_logprob_std": 1.5762369579646036, "tokenization_path_logprob_std": 123.75991937043871, "trajectory_balance_rmse": 5254.2351063335955} {"canonical_length": 74, "canonical_nll": 1.2828319652660474, "generation_rouge_l_f1": 0.3232323232323232, "generation_token_f1": 0.4444444444444444, "prompt_id": "88f331fe7934b488", "reading_canonical_log_prob": -94.9295654296875, "reading_distances": [15, 10, 13, 16], "reading_log_probs": [-97.40843200683594, -96.65602111816406, -95.74092102050781, -96.91868591308594], "reading_logprob_std": 0.7000057973969318, "tokenization_path_logprob_std": 98.58359307636424, "trajectory_balance_rmse": 1453.23257077914} {"canonical_length": 33, "canonical_nll": 0.582313364202326, "generation_rouge_l_f1": 0.24324324324324326, "generation_token_f1": 0.3783783783783784, "prompt_id": "8b4c34e2f1db7af4", "reading_canonical_log_prob": -19.216341018676758, "reading_distances": [18, 17, 20, 21], "reading_log_probs": [-18.82097625732422, -18.37464714050293, -20.15233612060547, -18.75749969482422], "reading_logprob_std": 0.7761057103191609, "tokenization_path_logprob_std": 72.51244888273894, "trajectory_balance_rmse": 841.8150083529332} {"canonical_length": 155, "canonical_nll": 1.9561338363155243, "generation_rouge_l_f1": 0.14536340852130328, "generation_token_f1": 0.2556390977443609, "prompt_id": "7a99289ed340c72d", "reading_canonical_log_prob": -303.20074462890625, "reading_distances": [13, 21, 17, 17], "reading_log_probs": [-296.6979675292969, -296.7983703613281, -297.5338134765625, -299.06591796875], "reading_logprob_std": 1.0933854673519743, "tokenization_path_logprob_std": 91.12906352816957, "trajectory_balance_rmse": 2392.1069539635732} {"canonical_length": 21, "canonical_nll": 1.0665113358270555, "generation_rouge_l_f1": 0.4117647058823529, "generation_token_f1": 0.5294117647058824, "prompt_id": "321ea8e8622b4791", "reading_canonical_log_prob": -22.396738052368164, "reading_distances": [26, 33, 35, 27], "reading_log_probs": [-21.192201614379883, -22.211641311645508, -19.55821418762207, -21.986772537231445], "reading_logprob_std": 1.2017228009636836, "tokenization_path_logprob_std": 62.09644986437019, "trajectory_balance_rmse": 450.73912651984256} {"canonical_length": 176, "canonical_nll": 1.8587778264825994, "generation_rouge_l_f1": 0.2518891687657431, "generation_token_f1": 0.4534005037783375, "prompt_id": "e9ba47f9f768bdc0", "reading_canonical_log_prob": -327.1448974609375, "reading_distances": [25, 29, 29, 20], "reading_log_probs": [-323.9347839355469, -324.2210693359375, -327.072265625, -324.95379638671875], "reading_logprob_std": 1.4176934959359182, "tokenization_path_logprob_std": 124.29271741747523, "trajectory_balance_rmse": 4344.314252392165} {"canonical_length": 7, "canonical_nll": 0.6362960679190499, "generation_rouge_l_f1": 0.044444444444444446, "generation_token_f1": 0.044444444444444446, "prompt_id": "cd4e165565532ef7", "reading_canonical_log_prob": -4.45407247543335, "reading_distances": [31, 32, 36, 37], "reading_log_probs": [-9.267032623291016, -7.722925186157227, -5.880208492279053, -7.190138339996338], "reading_logprob_std": 1.4013005322163805, "tokenization_path_logprob_std": 12.739046649476142, "trajectory_balance_rmse": 154.3740809160113} {"canonical_length": 270, "canonical_nll": 0.9599570945457175, "generation_rouge_l_f1": 0.28790786948176583, "generation_token_f1": 0.5412667946257197, "prompt_id": "df9cf174d34cff3e", "reading_canonical_log_prob": -259.18841552734375, "reading_distances": [32, 36, 42, 30], "reading_log_probs": [-262.5896301269531, -263.06427001953125, -260.2743225097656, -263.028564453125], "reading_logprob_std": 1.3275769409531042, "tokenization_path_logprob_std": 193.2860820845203, "trajectory_balance_rmse": 4065.1717897866893} {"canonical_length": 143, "canonical_nll": 1.5482305780157342, "generation_rouge_l_f1": 0.4566929133858268, "generation_token_f1": 0.5826771653543307, "prompt_id": "1fba2562017bea7b", "reading_canonical_log_prob": -221.39697265625, "reading_distances": [29, 34, 33, 29], "reading_log_probs": [-216.2173614501953, -221.04275512695312, -218.66876220703125, -218.32998657226562], "reading_logprob_std": 1.9762485348627088, "tokenization_path_logprob_std": 133.92525411302466, "trajectory_balance_rmse": 3118.526446865287} {"canonical_length": 28, "canonical_nll": 1.2457686832972936, "generation_rouge_l_f1": 0.39999999999999997, "generation_token_f1": 0.39999999999999997, "prompt_id": "56997da805bc232c", "reading_canonical_log_prob": -34.88152313232422, "reading_distances": [25, 26, 29, 36], "reading_log_probs": [-35.677825927734375, -34.322265625, -34.88253402709961, -34.59432601928711], "reading_logprob_std": 0.5855899012582638, "tokenization_path_logprob_std": 28.800436366826478, "trajectory_balance_rmse": 617.545627989884} {"canonical_length": 110, "canonical_nll": 1.507383866743608, "generation_rouge_l_f1": 0.24932249322493227, "generation_token_f1": 0.37940379403794033, "prompt_id": "02d66047fd43a021", "reading_canonical_log_prob": -165.81222534179688, "reading_distances": [22, 19, 16, 18], "reading_log_probs": [-167.0361785888672, -163.15670776367188, -168.7765350341797, -168.51303100585938], "reading_logprob_std": 2.5916905534795296, "tokenization_path_logprob_std": 142.39996883393331, "trajectory_balance_rmse": 2101.9042936289557} {"canonical_length": 79, "canonical_nll": 1.5843820692617683, "generation_rouge_l_f1": 0.16352201257861634, "generation_token_f1": 0.26415094339622647, "prompt_id": "83a9bce5dad053f6", "reading_canonical_log_prob": -125.16618347167969, "reading_distances": [20, 21, 22, 24], "reading_log_probs": [-133.34173583984375, -134.03610229492188, -131.66510009765625, -130.8136444091797], "reading_logprob_std": 1.4836711647681846, "tokenization_path_logprob_std": 123.0312682437536, "trajectory_balance_rmse": 1400.5492784498404} {"canonical_length": 21, "canonical_nll": 6.060200645810082, "generation_rouge_l_f1": 0.01593625498007968, "generation_token_f1": 0.02390438247011952, "prompt_id": "35f4df3b6373154b", "reading_canonical_log_prob": -127.26421356201172, "reading_distances": [20, 17, 23, 20], "reading_log_probs": [-124.20653533935547, -120.9709243774414, -123.9778060913086, -121.89805603027344], "reading_logprob_std": 1.5831625297032847, "tokenization_path_logprob_std": 23.354643975589482, "trajectory_balance_rmse": 243.21134258953683} {"canonical_length": 228, "canonical_nll": 1.2239418699030291, "generation_rouge_l_f1": 0.19120458891013384, "generation_token_f1": 0.37858508604206503, "prompt_id": "7c0757fa3b0cff4b", "reading_canonical_log_prob": -279.0587463378906, "reading_distances": [21, 29, 32, 26], "reading_log_probs": [-271.5189208984375, -270.1128234863281, -271.4754638671875, -270.53240966796875], "reading_logprob_std": 0.6996663161693087, "tokenization_path_logprob_std": 218.0442885332936, "trajectory_balance_rmse": 3070.9019591856145} {"canonical_length": 211, "canonical_nll": 1.6017629614373519, "generation_rouge_l_f1": 0.2439024390243902, "generation_token_f1": 0.4700665188470066, "prompt_id": "5fa3ea0df830456f", "reading_canonical_log_prob": -337.97198486328125, "reading_distances": [20, 27, 23, 19], "reading_log_probs": [-343.8004150390625, -340.23797607421875, -339.61578369140625, -341.9742431640625], "reading_logprob_std": 1.881955819170838, "tokenization_path_logprob_std": 117.99488449457857, "trajectory_balance_rmse": 4415.255233069231} {"canonical_length": 145, "canonical_nll": 1.1429397056842672, "generation_rouge_l_f1": 0.31527093596059114, "generation_token_f1": 0.4236453201970443, "prompt_id": "70669892361b488c", "reading_canonical_log_prob": -165.72625732421875, "reading_distances": [35, 36, 29, 32], "reading_log_probs": [-167.158203125, -164.90817260742188, -164.4776611328125, -164.95297241210938], "reading_logprob_std": 1.208451252757247, "tokenization_path_logprob_std": 160.40257551270338, "trajectory_balance_rmse": 2399.5536648841344} {"canonical_length": 12, "canonical_nll": 0.9963777860005697, "generation_rouge_l_f1": 0.6666666666666667, "generation_token_f1": 0.8333333333333333, "prompt_id": "e90b400c44335d9a", "reading_canonical_log_prob": -11.956533432006836, "reading_distances": [19, 18, 21, 25], "reading_log_probs": [-12.082969665527344, -11.943670272827148, -12.373205184936523, -13.486408233642578], "reading_logprob_std": 0.6998239347696029, "tokenization_path_logprob_std": 43.79911715671468, "trajectory_balance_rmse": 210.61547285682923} {"canonical_length": 232, "canonical_nll": 1.3677401049383755, "generation_rouge_l_f1": 0.2721649484536082, "generation_token_f1": 0.5154639175257731, "prompt_id": "9f0a3bc3aaf621a1", "reading_canonical_log_prob": -317.3157043457031, "reading_distances": [20, 31, 25, 26], "reading_log_probs": [-312.1673583984375, -311.96527099609375, -312.3419189453125, -313.77557373046875], "reading_logprob_std": 0.8232098237807415, "tokenization_path_logprob_std": 239.61634328199946, "trajectory_balance_rmse": 5033.9290593523665} {"canonical_length": 93, "canonical_nll": 0.5662157612462198, "generation_rouge_l_f1": 0.19771863117870722, "generation_token_f1": 0.3269961977186312, "prompt_id": "f2be334fa1e04283", "reading_canonical_log_prob": -52.65806579589844, "reading_distances": [25, 15, 23, 24], "reading_log_probs": [-47.58871841430664, -50.75996398925781, -48.691734313964844, -49.16797637939453], "reading_logprob_std": 1.3167606507526062, "tokenization_path_logprob_std": 54.081239283980665, "trajectory_balance_rmse": 1692.9733538537935} {"canonical_length": 54, "canonical_nll": 1.552025971589265, "generation_rouge_l_f1": 0.1937984496124031, "generation_token_f1": 0.24031007751937983, "prompt_id": "1a42eb844c9543a9", "reading_canonical_log_prob": -83.80940246582031, "reading_distances": [45, 31, 40, 36], "reading_log_probs": [-83.78993225097656, -82.64151000976562, -82.57119750976562, -84.02308654785156], "reading_logprob_std": 0.7572003420965507, "tokenization_path_logprob_std": 30.301408506939694, "trajectory_balance_rmse": 1099.3789745897004} {"canonical_length": 40, "canonical_nll": 0.564995288848877, "generation_rouge_l_f1": 0.15436241610738255, "generation_token_f1": 0.21476510067114093, "prompt_id": "3afed29700596cb5", "reading_canonical_log_prob": -22.599811553955078, "reading_distances": [11, 9, 16, 14], "reading_log_probs": [-24.417789459228516, -22.684001922607422, -24.596052169799805, -25.64957046508789], "reading_logprob_std": 1.2286684862404131, "tokenization_path_logprob_std": 50.54221722961402, "trajectory_balance_rmse": 485.2078927392227} {"canonical_length": 28, "canonical_nll": 0.5116853032793317, "generation_rouge_l_f1": 0.523076923076923, "generation_token_f1": 0.6153846153846154, "prompt_id": "e24f8c298fcc6b27", "reading_canonical_log_prob": -14.327188491821289, "reading_distances": [13, 12, 13, 12], "reading_log_probs": [-14.680688858032227, -14.684636116027832, -14.126380920410156, -14.948628425598145], "reading_logprob_std": 0.345988468637106, "tokenization_path_logprob_std": 23.421124452881475, "trajectory_balance_rmse": 566.1812412816167} {"canonical_length": 232, "canonical_nll": 1.4885635375976562, "generation_rouge_l_f1": 0.21513944223107573, "generation_token_f1": 0.41832669322709165, "prompt_id": "d2492eb6d00f3204", "reading_canonical_log_prob": -345.34674072265625, "reading_distances": [31, 27, 20, 34], "reading_log_probs": [-346.8710632324219, -346.39306640625, -345.5356140136719, -345.73455810546875], "reading_logprob_std": 0.6131874390346012, "tokenization_path_logprob_std": 231.58355913898194, "trajectory_balance_rmse": 5623.18062083632} {"canonical_length": 83, "canonical_nll": 1.7403696818524097, "generation_rouge_l_f1": 0.2428115015974441, "generation_token_f1": 0.37699680511182115, "prompt_id": "d1d32462252d2ba1", "reading_canonical_log_prob": -144.45068359375, "reading_distances": [18, 20, 14, 14], "reading_log_probs": [-145.18673706054688, -143.95846557617188, -144.51095581054688, -145.5892791748047], "reading_logprob_std": 0.7219736272538637, "tokenization_path_logprob_std": 114.07548225678525, "trajectory_balance_rmse": 1652.0475907880784} {"canonical_length": 171, "canonical_nll": 1.1760759855571545, "generation_rouge_l_f1": 0.2938388625592417, "generation_token_f1": 0.5260663507109005, "prompt_id": "981d7ddcc229e43f", "reading_canonical_log_prob": -201.10899353027344, "reading_distances": [22, 19, 20, 21], "reading_log_probs": [-184.6224365234375, -185.98989868164062, -187.881103515625, -186.32830810546875], "reading_logprob_std": 1.3385692430880427, "tokenization_path_logprob_std": 86.92001154838348, "trajectory_balance_rmse": 2972.167195070238} {"canonical_length": 14, "canonical_nll": 1.381448745727539, "generation_rouge_l_f1": 0.33333333333333337, "generation_token_f1": 0.4999999999999999, "prompt_id": "a353a5c04a14d88b", "reading_canonical_log_prob": -19.340282440185547, "reading_distances": [35, 31, 31, 35], "reading_log_probs": [-17.757797241210938, -16.709972381591797, -18.383581161499023, -16.562942504882812], "reading_logprob_std": 0.8686449957186981, "tokenization_path_logprob_std": 36.388358146545464, "trajectory_balance_rmse": 323.4814403242408} {"canonical_length": 37, "canonical_nll": 0.6526764534615181, "generation_rouge_l_f1": 0.2027972027972028, "generation_token_f1": 0.2097902097902098, "prompt_id": "104ca723f309949c", "reading_canonical_log_prob": -24.149028778076172, "reading_distances": [17, 15, 21, 14], "reading_log_probs": [-25.698776245117188, -24.94137954711914, -26.158668518066406, -25.285362243652344], "reading_logprob_std": 0.5258994026664379, "tokenization_path_logprob_std": 65.39114851345862, "trajectory_balance_rmse": 565.1096010760021} {"canonical_length": 270, "canonical_nll": 0.9537276656539352, "generation_rouge_l_f1": 0.2162162162162162, "generation_token_f1": 0.3742203742203742, "prompt_id": "5768cefb5b534d32", "reading_canonical_log_prob": -257.5064697265625, "reading_distances": [22, 20, 25, 20], "reading_log_probs": [-256.13726806640625, -252.60032653808594, -257.0288391113281, -256.5594177246094], "reading_logprob_std": 2.0205104080755927, "tokenization_path_logprob_std": 165.4041203449967, "trajectory_balance_rmse": 4313.218612768265} {"canonical_length": 460, "canonical_nll": 0.7183066990064538, "generation_rouge_l_f1": 0.32033426183844016, "generation_token_f1": 0.5125348189415042, "prompt_id": "2c20880d9a9f5257", "reading_canonical_log_prob": -330.42108154296875, "reading_distances": [31, 27, 35, 21], "reading_log_probs": [-328.0767822265625, -327.9901123046875, -328.7975769042969, -327.0045166015625], "reading_logprob_std": 0.7368442526662459, "tokenization_path_logprob_std": 190.00409137666645, "trajectory_balance_rmse": 7607.730241403875} {"canonical_length": 43, "canonical_nll": 1.3259764383005541, "generation_rouge_l_f1": 0.18719211822660098, "generation_token_f1": 0.25615763546798026, "prompt_id": "1a2898fe9876d2bc", "reading_canonical_log_prob": -57.01698684692383, "reading_distances": [19, 17, 12, 16], "reading_log_probs": [-57.42662811279297, -60.50853729248047, -63.70857620239258, -59.866233825683594], "reading_logprob_std": 2.5872944320096756, "tokenization_path_logprob_std": 87.65874490456949, "trajectory_balance_rmse": 922.9133753079257} {"canonical_length": 85, "canonical_nll": 0.949482816808364, "generation_rouge_l_f1": 0.3541666666666667, "generation_token_f1": 0.5416666666666666, "prompt_id": "52568183284a12f5", "reading_canonical_log_prob": -80.70603942871094, "reading_distances": [17, 20, 15, 17], "reading_log_probs": [-80.12113952636719, -80.78285217285156, -80.52531433105469, -79.81240844726562], "reading_logprob_std": 0.4294237749679137, "tokenization_path_logprob_std": 133.32976050056746, "trajectory_balance_rmse": 1740.5109143984719} {"canonical_length": 202, "canonical_nll": 1.6935871992961016, "generation_rouge_l_f1": 0.20055710306406685, "generation_token_f1": 0.4958217270194986, "prompt_id": "ea255680cda4b618", "reading_canonical_log_prob": -342.1046142578125, "reading_distances": [29, 17, 25, 26], "reading_log_probs": [-340.3755798339844, -338.9686279296875, -343.9989929199219, -344.06585693359375], "reading_logprob_std": 2.582272374186029, "tokenization_path_logprob_std": 285.1892779104531, "trajectory_balance_rmse": 4166.05245475968} {"canonical_length": 19, "canonical_nll": 1.237041573775442, "generation_rouge_l_f1": 0.3448275862068966, "generation_token_f1": 0.41379310344827586, "prompt_id": "559aadd3a6bc5deb", "reading_canonical_log_prob": -23.5037899017334, "reading_distances": [32, 35, 35, 39], "reading_log_probs": [-25.700468063354492, -24.02971839904785, -25.337221145629883, -23.934322357177734], "reading_logprob_std": 0.900435647674147, "tokenization_path_logprob_std": 86.80562738112505, "trajectory_balance_rmse": 412.592547765351} {"canonical_length": 24, "canonical_nll": 0.8970523675282797, "generation_rouge_l_f1": 0.6486486486486486, "generation_token_f1": 0.6486486486486486, "prompt_id": "4196e10bb50daa8f", "reading_canonical_log_prob": -21.52925682067871, "reading_distances": [25, 23, 20, 18], "reading_log_probs": [-22.943410873413086, -22.947113037109375, -22.52713966369629, -23.47848892211914], "reading_logprob_std": 0.3898084274475719, "tokenization_path_logprob_std": 15.681215201724095, "trajectory_balance_rmse": 501.0526895929227} {"canonical_length": 293, "canonical_nll": 1.3572196049088097, "generation_rouge_l_f1": 0.2546816479400749, "generation_token_f1": 0.50187265917603, "prompt_id": "0d2dbbf4ae30ce19", "reading_canonical_log_prob": -397.66534423828125, "reading_distances": [18, 13, 20, 21], "reading_log_probs": [-392.78033447265625, -392.78912353515625, -389.2394104003906, -390.9920654296875], "reading_logprob_std": 1.698965556068563, "tokenization_path_logprob_std": 182.33046540712502, "trajectory_balance_rmse": 6870.997590422883} {"canonical_length": 227, "canonical_nll": 1.125223302631126, "generation_rouge_l_f1": 0.22391857506361323, "generation_token_f1": 0.3867684478371501, "prompt_id": "cd0a37baf57f0d68", "reading_canonical_log_prob": -255.42568969726562, "reading_distances": [23, 29, 26, 27], "reading_log_probs": [-248.6756134033203, -248.650390625, -253.0152587890625, -252.08326721191406], "reading_logprob_std": 2.275788998572214, "tokenization_path_logprob_std": 273.36529061177333, "trajectory_balance_rmse": 4293.497773012742} {"canonical_length": 27, "canonical_nll": 1.5475617867928964, "generation_rouge_l_f1": 0.3947368421052632, "generation_token_f1": 0.47368421052631576, "prompt_id": "f0dd5c7b12d6ff90", "reading_canonical_log_prob": -41.7841682434082, "reading_distances": [23, 23, 19, 24], "reading_log_probs": [-52.585777282714844, -50.7724494934082, -51.717628479003906, -49.934165954589844], "reading_logprob_std": 1.1492643964203146, "tokenization_path_logprob_std": 103.25293259760988, "trajectory_balance_rmse": 563.2551975738955} {"canonical_length": 29, "canonical_nll": 1.0161658977640087, "generation_rouge_l_f1": 0.3125, "generation_token_f1": 0.43750000000000006, "prompt_id": "c737e978a37038bd", "reading_canonical_log_prob": -29.46881103515625, "reading_distances": [23, 22, 27, 26], "reading_log_probs": [-29.368560791015625, -29.159374237060547, -28.873138427734375, -29.126338958740234], "reading_logprob_std": 0.20310229883024988, "tokenization_path_logprob_std": 50.37050405855406, "trajectory_balance_rmse": 471.75810012668893} {"canonical_length": 90, "canonical_nll": 1.2743579440646702, "generation_rouge_l_f1": 0.3414634146341463, "generation_token_f1": 0.5609756097560976, "prompt_id": "6fe0c1612b081574", "reading_canonical_log_prob": -114.69221496582031, "reading_distances": [20, 25, 24, 27], "reading_log_probs": [-117.41058349609375, -117.42913818359375, -115.67716979980469, -115.64852142333984], "reading_logprob_std": 1.0145089149154947, "tokenization_path_logprob_std": 191.96050278810793, "trajectory_balance_rmse": 2008.3408555257631} {"canonical_length": 276, "canonical_nll": 1.3440602620442708, "generation_rouge_l_f1": 0.2727272727272727, "generation_token_f1": 0.5909090909090908, "prompt_id": "a45d6c8378dedb7b", "reading_canonical_log_prob": -370.96063232421875, "reading_distances": [13, 14, 15, 14], "reading_log_probs": [-369.58209228515625, -368.10198974609375, -370.6407165527344, -371.5604248046875], "reading_logprob_std": 1.4854000175557698, "tokenization_path_logprob_std": 203.18507743395682, "trajectory_balance_rmse": 5592.8423155187} {"canonical_length": 103, "canonical_nll": 1.156596433769152, "generation_rouge_l_f1": 0.27397260273972607, "generation_token_f1": 0.3616438356164383, "prompt_id": "2e7570c8d7724f9e", "reading_canonical_log_prob": -119.12943267822266, "reading_distances": [25, 22, 17, 25], "reading_log_probs": [-112.95989227294922, -116.24534606933594, -113.17173767089844, -114.70576477050781], "reading_logprob_std": 1.529100340165703, "tokenization_path_logprob_std": 79.07114862936334, "trajectory_balance_rmse": 1489.0050996486943} {"canonical_length": 30, "canonical_nll": 1.7119229634602864, "generation_rouge_l_f1": 0.12186379928315415, "generation_token_f1": 0.12903225806451613, "prompt_id": "7aa76ccd6b778b98", "reading_canonical_log_prob": -51.357688903808594, "reading_distances": [40, 36, 27, 35], "reading_log_probs": [-49.99618911743164, -50.06379699707031, -50.09930419921875, -50.639007568359375], "reading_logprob_std": 0.2960611980265754, "tokenization_path_logprob_std": 80.3265436352498, "trajectory_balance_rmse": 710.1703970989569} {"canonical_length": 252, "canonical_nll": 1.4483062501937625, "generation_rouge_l_f1": 0.3002207505518764, "generation_token_f1": 0.5518763796909493, "prompt_id": "0667a89ba9993054", "reading_canonical_log_prob": -364.9731750488281, "reading_distances": [28, 27, 25, 26], "reading_log_probs": [-365.57501220703125, -364.3843994140625, -366.8524169921875, -363.2511901855469], "reading_logprob_std": 1.5490206322558204, "tokenization_path_logprob_std": 99.42511863788944, "trajectory_balance_rmse": 4725.290442011358} {"canonical_length": 248, "canonical_nll": 1.0412671489100302, "generation_rouge_l_f1": 0.31067961165048547, "generation_token_f1": 0.5475728155339806, "prompt_id": "45c2d1c36166f6b6", "reading_canonical_log_prob": -258.2342529296875, "reading_distances": [24, 24, 22, 32], "reading_log_probs": [-255.62521362304688, -252.7339324951172, -255.994140625, -251.46539306640625], "reading_logprob_std": 2.208834997209304, "tokenization_path_logprob_std": 94.08323337715325, "trajectory_balance_rmse": 4268.172382220996} {"canonical_length": 20, "canonical_nll": 1.2761119842529296, "generation_rouge_l_f1": 0.5, "generation_token_f1": 0.5, "prompt_id": "e5f7552dff198384", "reading_canonical_log_prob": -25.522239685058594, "reading_distances": [29, 20, 21, 27], "reading_log_probs": [-23.798418045043945, -22.999069213867188, -24.350189208984375, -23.518373489379883], "reading_logprob_std": 0.5633940663195759, "tokenization_path_logprob_std": 61.07831024581836, "trajectory_balance_rmse": 477.40164031347797} {"canonical_length": 117, "canonical_nll": 1.721433851453993, "generation_rouge_l_f1": 0.2033898305084746, "generation_token_f1": 0.3050847457627119, "prompt_id": "4a930bccec9843cd", "reading_canonical_log_prob": -201.4077606201172, "reading_distances": [34, 32, 26, 27], "reading_log_probs": [-210.4852294921875, -198.33279418945312, -202.7049102783203, -203.6309814453125], "reading_logprob_std": 5.026923862594978, "tokenization_path_logprob_std": 217.67092419519514, "trajectory_balance_rmse": 2683.4872317313334} {"canonical_length": 49, "canonical_nll": 1.0282769495127153, "generation_rouge_l_f1": 0.21897810218978103, "generation_token_f1": 0.32116788321167883, "prompt_id": "35925631dfc295f8", "reading_canonical_log_prob": -50.38557052612305, "reading_distances": [33, 29, 36, 28], "reading_log_probs": [-52.004295349121094, -53.63161849975586, -53.12183380126953, -52.125396728515625], "reading_logprob_std": 0.787041383687493, "tokenization_path_logprob_std": 147.0525947286237, "trajectory_balance_rmse": 943.6836058569345} {"canonical_length": 295, "canonical_nll": 1.143294201867055, "generation_rouge_l_f1": 0.24416517055655299, "generation_token_f1": 0.5637342908438061, "prompt_id": "f6597a51ed7cc783", "reading_canonical_log_prob": -337.27178955078125, "reading_distances": [21, 21, 19, 23], "reading_log_probs": [-335.1074523925781, -334.4256896972656, -335.128173828125, -333.12701416015625], "reading_logprob_std": 0.9386181094455447, "tokenization_path_logprob_std": 325.1467144225239, "trajectory_balance_rmse": 5583.685588217544} {"canonical_length": 213, "canonical_nll": 1.2665585799955985, "generation_rouge_l_f1": 0.2597402597402597, "generation_token_f1": 0.4935064935064935, "prompt_id": "87270c0f1888627a", "reading_canonical_log_prob": -269.7769775390625, "reading_distances": [13, 15, 9, 10], "reading_log_probs": [-266.392822265625, -267.8521728515625, -266.296142578125, -267.9423522949219], "reading_logprob_std": 0.8981213439674615, "tokenization_path_logprob_std": 421.2123528103071, "trajectory_balance_rmse": 4596.7345786791775} {"canonical_length": 14, "canonical_nll": 1.1055423191615514, "generation_rouge_l_f1": 0.1258741258741259, "generation_token_f1": 0.13986013986013984, "prompt_id": "a53e27e32808bf25", "reading_canonical_log_prob": -15.477592468261719, "reading_distances": [25, 20, 27, 27], "reading_log_probs": [-13.810848236083984, -13.753120422363281, -14.15335750579834, -13.838654518127441], "reading_logprob_std": 0.1798059288469325, "tokenization_path_logprob_std": 27.34932241365552, "trajectory_balance_rmse": 258.3474676045704} {"canonical_length": 128, "canonical_nll": 1.2707855701446533, "generation_rouge_l_f1": 0.2980769230769231, "generation_token_f1": 0.4903846153846153, "prompt_id": "1e1844cf1025dc97", "reading_canonical_log_prob": -162.66055297851562, "reading_distances": [15, 16, 18, 16], "reading_log_probs": [-159.1000518798828, -159.71139526367188, -159.19439697265625, -158.72560119628906], "reading_logprob_std": 0.4063764367366027, "tokenization_path_logprob_std": 100.1231014473141, "trajectory_balance_rmse": 2785.9649555386077} {"canonical_length": 155, "canonical_nll": 1.5788470852759577, "generation_rouge_l_f1": 0.2016348773841962, "generation_token_f1": 0.39782016348773847, "prompt_id": "87981277aeac45ee", "reading_canonical_log_prob": -244.72129821777344, "reading_distances": [21, 21, 23, 18], "reading_log_probs": [-240.85476684570312, -240.31460571289062, -242.38706970214844, -240.7963104248047], "reading_logprob_std": 0.8991108914330233, "tokenization_path_logprob_std": 47.26504294402239, "trajectory_balance_rmse": 3040.9430927146454} {"canonical_length": 26, "canonical_nll": 0.1810635786790114, "generation_rouge_l_f1": 0.051470588235294115, "generation_token_f1": 0.0661764705882353, "prompt_id": "40b80040395896ae", "reading_canonical_log_prob": -4.707653045654297, "reading_distances": [39, 46, 45, 47], "reading_log_probs": [-3.0892333984375, -4.905604362487793, -4.231758117675781, -6.26508092880249], "reading_logprob_std": 1.3268766803881709, "tokenization_path_logprob_std": 98.07931191691814, "trajectory_balance_rmse": 318.2967736192909} {"canonical_length": 356, "canonical_nll": 1.440365737743592, "generation_rouge_l_f1": 0.20100502512562812, "generation_token_f1": 0.46566164154103856, "prompt_id": "63d17b19976fa874", "reading_canonical_log_prob": -512.7702026367188, "reading_distances": [29, 29, 27, 20], "reading_log_probs": [-502.9160461425781, -499.65936279296875, -497.60968017578125, -499.81072998046875], "reading_logprob_std": 2.188515969388616, "tokenization_path_logprob_std": 197.72814473164166, "trajectory_balance_rmse": 7791.023192161525} {"canonical_length": 19, "canonical_nll": 0.6405893626965975, "generation_rouge_l_f1": 0.09486166007905138, "generation_token_f1": 0.10276679841897234, "prompt_id": "53bd5d756df134c2", "reading_canonical_log_prob": -12.171197891235352, "reading_distances": [27, 21, 25, 25], "reading_log_probs": [-10.027313232421875, -9.472558975219727, -10.865388870239258, -11.505459785461426], "reading_logprob_std": 0.8980252358232315, "tokenization_path_logprob_std": 83.51627398319782, "trajectory_balance_rmse": 243.18247771045222} {"canonical_length": 114, "canonical_nll": 1.0437221861722177, "generation_rouge_l_f1": 0.18285714285714286, "generation_token_f1": 0.3142857142857143, "prompt_id": "1cf0a058de57af94", "reading_canonical_log_prob": -118.98432922363281, "reading_distances": [38, 28, 33, 36], "reading_log_probs": [-126.0147705078125, -124.83100891113281, -125.31835174560547, -128.39675903320312], "reading_logprob_std": 1.5808444037104261, "tokenization_path_logprob_std": 74.349161249052, "trajectory_balance_rmse": 1535.0130437503817} {"canonical_length": 257, "canonical_nll": 1.4326574422042193, "generation_rouge_l_f1": 0.26175869120654394, "generation_token_f1": 0.5848670756646218, "prompt_id": "a5e93d67499f0604", "reading_canonical_log_prob": -368.1929626464844, "reading_distances": [15, 19, 20, 19], "reading_log_probs": [-367.018798828125, -368.6105651855469, -365.80072021484375, -365.75286865234375], "reading_logprob_std": 1.3442458499721235, "tokenization_path_logprob_std": 278.28349849128824, "trajectory_balance_rmse": 4537.65185504512} {"canonical_length": 118, "canonical_nll": 1.314995587882349, "generation_rouge_l_f1": 0.39183673469387753, "generation_token_f1": 0.5714285714285714, "prompt_id": "9493ef83593c3f2d", "reading_canonical_log_prob": -155.1694793701172, "reading_distances": [32, 36, 37, 32], "reading_log_probs": [-155.14437866210938, -151.50958251953125, -152.7915802001953, -151.24217224121094], "reading_logprob_std": 1.7816263610657, "tokenization_path_logprob_std": 44.08997569116952, "trajectory_balance_rmse": 2496.3591319325883} {"canonical_length": 22, "canonical_nll": 2.06446664983576, "generation_rouge_l_f1": 0.20689655172413793, "generation_token_f1": 0.20689655172413793, "prompt_id": "be453082a5e24aed", "reading_canonical_log_prob": -45.41826629638672, "reading_distances": [35, 24, 25, 37], "reading_log_probs": [-37.558292388916016, -36.767093658447266, -38.07545471191406, -39.85659408569336], "reading_logprob_std": 1.3103727839131365, "tokenization_path_logprob_std": 40.13422321466996, "trajectory_balance_rmse": 417.17828297023647} {"canonical_length": 280, "canonical_nll": 1.0929275512695313, "generation_rouge_l_f1": 0.3303411131059246, "generation_token_f1": 0.5888689407540396, "prompt_id": "6b6295e1afd98591", "reading_canonical_log_prob": -306.01971435546875, "reading_distances": [14, 17, 17, 20], "reading_log_probs": [-305.67059326171875, -305.29388427734375, -305.94842529296875, -305.43231201171875], "reading_logprob_std": 0.28720453635883947, "tokenization_path_logprob_std": 229.43630212871878, "trajectory_balance_rmse": 5796.517054048908} {"canonical_length": 200, "canonical_nll": 1.351474609375, "generation_rouge_l_f1": 0.2527233115468409, "generation_token_f1": 0.4008714596949891, "prompt_id": "900ff807cdf33dd9", "reading_canonical_log_prob": -270.294921875, "reading_distances": [23, 20, 27, 21], "reading_log_probs": [-264.5436706542969, -268.28594970703125, -269.4772644042969, -261.84912109375], "reading_logprob_std": 3.495785090692972, "tokenization_path_logprob_std": 178.97693593468603, "trajectory_balance_rmse": 3505.853466502191} {"canonical_length": 120, "canonical_nll": 1.4885658264160155, "generation_rouge_l_f1": 0.21666666666666667, "generation_token_f1": 0.36666666666666664, "prompt_id": "6b5f709d21a7a023", "reading_canonical_log_prob": -178.62789916992188, "reading_distances": [13, 19, 16, 18], "reading_log_probs": [-173.1082763671875, -174.34310913085938, -173.29669189453125, -172.56253051757812], "reading_logprob_std": 0.7451282359162795, "tokenization_path_logprob_std": 105.55803160887179, "trajectory_balance_rmse": 2274.5988758539916} {"canonical_length": 52, "canonical_nll": 1.1189283224252553, "generation_rouge_l_f1": 0.17704918032786884, "generation_token_f1": 0.24262295081967214, "prompt_id": "e072285f618dcbd9", "reading_canonical_log_prob": -58.18427276611328, "reading_distances": [28, 34, 21, 27], "reading_log_probs": [-55.603641510009766, -55.632354736328125, -54.31549072265625, -52.98878860473633], "reading_logprob_std": 1.2576540279702804, "tokenization_path_logprob_std": 21.225548419256686, "trajectory_balance_rmse": 1207.0962923848215} {"canonical_length": 147, "canonical_nll": 1.4179924166932398, "generation_rouge_l_f1": 0.18092909535452323, "generation_token_f1": 0.3276283618581907, "prompt_id": "21874077bcbf7b19", "reading_canonical_log_prob": -208.44488525390625, "reading_distances": [27, 20, 22, 23], "reading_log_probs": [-209.04823303222656, -209.33547973632812, -210.91433715820312, -208.4847412109375], "reading_logprob_std": 1.040907519700428, "tokenization_path_logprob_std": 166.90298443934947, "trajectory_balance_rmse": 3996.284186823474} {"canonical_length": 108, "canonical_nll": 1.1147539350721571, "generation_rouge_l_f1": 0.2821576763485477, "generation_token_f1": 0.5145228215767635, "prompt_id": "8ba54ff004576b32", "reading_canonical_log_prob": -120.39342498779297, "reading_distances": [20, 23, 22, 21], "reading_log_probs": [-118.81707000732422, -117.98959350585938, -118.7320556640625, -118.69094848632812], "reading_logprob_std": 0.38217470940966425, "tokenization_path_logprob_std": 151.7891335435742, "trajectory_balance_rmse": 2639.7451535063096} {"canonical_length": 8, "canonical_nll": 0.03159480541944504, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "bccef35e52f5dc4b", "reading_canonical_log_prob": -0.2527584433555603, "reading_distances": [34, 34, 34, 34], "reading_log_probs": [-0.7411847114562988, -0.5173240900039673, -1.0127032995224, -0.49007511138916016], "reading_logprob_std": 0.24258593732228892, "tokenization_path_logprob_std": 33.45070554670646, "trajectory_balance_rmse": 149.2417705296139} {"canonical_length": 53, "canonical_nll": 1.0612189814729511, "generation_rouge_l_f1": 0.1458966565349544, "generation_token_f1": 0.20668693009118544, "prompt_id": "492f8c42a0de57b2", "reading_canonical_log_prob": -56.244606018066406, "reading_distances": [47, 39, 37, 45], "reading_log_probs": [-52.01912307739258, -52.30769348144531, -52.614097595214844, -53.288116455078125], "reading_logprob_std": 0.5444435889952286, "tokenization_path_logprob_std": 108.0236369407407, "trajectory_balance_rmse": 970.704106401859} {"canonical_length": 211, "canonical_nll": 1.2109685237938759, "generation_rouge_l_f1": 0.2186788154897494, "generation_token_f1": 0.4054669703872437, "prompt_id": "6aa6f6caad65a47f", "reading_canonical_log_prob": -255.5143585205078, "reading_distances": [21, 21, 29, 27], "reading_log_probs": [-255.61721801757812, -255.58314514160156, -255.8914337158203, -255.41165161132812], "reading_logprob_std": 0.19858992529177916, "tokenization_path_logprob_std": 196.8571704874087, "trajectory_balance_rmse": 3777.917614331037} {"canonical_length": 107, "canonical_nll": 1.4441218866366092, "generation_rouge_l_f1": 0.224, "generation_token_f1": 0.4, "prompt_id": "83dae6df424d2709", "reading_canonical_log_prob": -154.5210418701172, "reading_distances": [31, 32, 29, 25], "reading_log_probs": [-159.11685180664062, -158.16583251953125, -161.0106201171875, -158.07347106933594], "reading_logprob_std": 1.3634402292247236, "tokenization_path_logprob_std": 91.89485805387152, "trajectory_balance_rmse": 2106.074740724876} {"canonical_length": 104, "canonical_nll": 1.2357576810396635, "generation_rouge_l_f1": 0.17506631299734748, "generation_token_f1": 0.36604774535809026, "prompt_id": "d7aad76280327bd4", "reading_canonical_log_prob": -128.518798828125, "reading_distances": [28, 32, 36, 32], "reading_log_probs": [-127.78630065917969, -130.1994171142578, -130.26303100585938, -128.89874267578125], "reading_logprob_std": 1.1815221413474735, "tokenization_path_logprob_std": 84.19383449209327, "trajectory_balance_rmse": 2327.51902172933} {"canonical_length": 362, "canonical_nll": 1.5362574118935601, "generation_rouge_l_f1": 0.1585760517799353, "generation_token_f1": 0.32362459546925565, "prompt_id": "117db47d424e536b", "reading_canonical_log_prob": -556.1251831054688, "reading_distances": [27, 15, 15, 23], "reading_log_probs": [-553.7327880859375, -553.52783203125, -560.105712890625, -552.7039794921875], "reading_logprob_std": 3.4211052256656886, "tokenization_path_logprob_std": 283.5629719986777, "trajectory_balance_rmse": 6961.975594472729} {"canonical_length": 207, "canonical_nll": 1.4622581592504529, "generation_rouge_l_f1": 0.27038626609442057, "generation_token_f1": 0.4678111587982833, "prompt_id": "f593668e2032bc26", "reading_canonical_log_prob": -302.68743896484375, "reading_distances": [32, 31, 31, 33], "reading_log_probs": [-303.28857421875, -304.19622802734375, -306.3243713378906, -304.8464660644531], "reading_logprob_std": 1.2781149544064618, "tokenization_path_logprob_std": 211.58945766217053, "trajectory_balance_rmse": 4442.180137826674} {"canonical_length": 243, "canonical_nll": 1.0275247832875194, "generation_rouge_l_f1": 0.2825112107623319, "generation_token_f1": 0.5201793721973095, "prompt_id": "62a2987fac290e0c", "reading_canonical_log_prob": -249.6885223388672, "reading_distances": [13, 16, 18, 14], "reading_log_probs": [-249.782470703125, -251.67861938476562, -250.45362854003906, -250.8155517578125], "reading_logprob_std": 0.7900192953558558, "tokenization_path_logprob_std": 218.83629106440364, "trajectory_balance_rmse": 4947.786559939245} {"canonical_length": 28, "canonical_nll": 0.5754854338509696, "generation_rouge_l_f1": 0.11518324607329844, "generation_token_f1": 0.13612565445026178, "prompt_id": "0597c4917add4bdd", "reading_canonical_log_prob": -16.11359214782715, "reading_distances": [17, 16, 17, 20], "reading_log_probs": [-14.544628143310547, -18.137609481811523, -15.375150680541992, -19.212453842163086], "reading_logprob_std": 2.2154616509208664, "tokenization_path_logprob_std": 38.58741292664385, "trajectory_balance_rmse": 326.64430188504724} {"canonical_length": 51, "canonical_nll": 1.6696897020526962, "generation_rouge_l_f1": 0.09146341463414634, "generation_token_f1": 0.12195121951219512, "prompt_id": "ea67625f5161437c", "reading_canonical_log_prob": -85.1541748046875, "reading_distances": [19, 21, 20, 20], "reading_log_probs": [-85.61103820800781, -85.87954711914062, -84.80540466308594, -87.53617095947266], "reading_logprob_std": 1.1468270606533268, "tokenization_path_logprob_std": 118.2742562206507, "trajectory_balance_rmse": 1031.0162162707477} {"canonical_length": 98, "canonical_nll": 0.9672938755580357, "generation_rouge_l_f1": 0.3392857142857143, "generation_token_f1": 0.5089285714285714, "prompt_id": "f9bf7abf6deb13c3", "reading_canonical_log_prob": -94.7947998046875, "reading_distances": [17, 20, 14, 24], "reading_log_probs": [-93.13943481445312, -93.9517593383789, -92.62680053710938, -93.16909790039062], "reading_logprob_std": 0.5466341033195488, "tokenization_path_logprob_std": 107.37521029265982, "trajectory_balance_rmse": 1995.0577606597594} {"canonical_length": 144, "canonical_nll": 1.4973222944471571, "generation_rouge_l_f1": 0.10370370370370369, "generation_token_f1": 0.11851851851851852, "prompt_id": "dd61adf3859cf501", "reading_canonical_log_prob": -215.61441040039062, "reading_distances": [40, 41, 39, 36], "reading_log_probs": [-234.81521606445312, -228.5115203857422, -229.44993591308594, -226.11264038085938], "reading_logprob_std": 3.6745796164361515, "tokenization_path_logprob_std": 195.15154524308494, "trajectory_balance_rmse": 2507.1972418798396} {"canonical_length": 38, "canonical_nll": 2.385584178723787, "generation_rouge_l_f1": 0.3185840707964602, "generation_token_f1": 0.3893805309734513, "prompt_id": "932b65e0c79cc6cd", "reading_canonical_log_prob": -90.6521987915039, "reading_distances": [31, 23, 29, 24], "reading_log_probs": [-94.03822326660156, -89.54468536376953, -97.20365905761719, -92.93038940429688], "reading_logprob_std": 3.1599428518142227, "tokenization_path_logprob_std": 45.68581380858616, "trajectory_balance_rmse": 616.4266744158601} {"canonical_length": 246, "canonical_nll": 1.3003019007240855, "generation_rouge_l_f1": 0.24705882352941178, "generation_token_f1": 0.40784313725490196, "prompt_id": "5a38778847ffa7e8", "reading_canonical_log_prob": -319.874267578125, "reading_distances": [28, 29, 34, 30], "reading_log_probs": [-324.1390686035156, -324.836181640625, -321.1090087890625, -320.22735595703125], "reading_logprob_std": 2.2523899946823898, "tokenization_path_logprob_std": 237.7961937821305, "trajectory_balance_rmse": 5428.537047909176} {"canonical_length": 163, "canonical_nll": 0.7965869552518693, "generation_rouge_l_f1": 0.2492753623188406, "generation_token_f1": 0.41739130434782606, "prompt_id": "bea3ed585e39be34", "reading_canonical_log_prob": -129.8436737060547, "reading_distances": [19, 15, 16, 14], "reading_log_probs": [-129.16468811035156, -130.1184539794922, -131.10092163085938, -130.43939208984375], "reading_logprob_std": 0.805678032503483, "tokenization_path_logprob_std": 260.1584118823323, "trajectory_balance_rmse": 2891.0841998572064} {"canonical_length": 256, "canonical_nll": 0.6404746770858765, "generation_rouge_l_f1": 0.4498977505112475, "generation_token_f1": 0.6380368098159509, "prompt_id": "e77970903b119833", "reading_canonical_log_prob": -163.96151733398438, "reading_distances": [21, 26, 26, 33], "reading_log_probs": [-164.49420166015625, -168.48788452148438, -166.936279296875, -171.67933654785156], "reading_logprob_std": 3.008721298584664, "tokenization_path_logprob_std": 174.76435189342612, "trajectory_balance_rmse": 2340.8867589921506} {"canonical_length": 28, "canonical_nll": 1.1159087589808874, "generation_rouge_l_f1": 0.10526315789473684, "generation_token_f1": 0.14354066985645936, "prompt_id": "4ea6cdd31bfaf581", "reading_canonical_log_prob": -31.245445251464844, "reading_distances": [24, 17, 19, 19], "reading_log_probs": [-32.55073547363281, -31.913488388061523, -32.46488952636719, -34.010536193847656], "reading_logprob_std": 0.8960630489927176, "tokenization_path_logprob_std": 6.858681595677441, "trajectory_balance_rmse": 250.81598334576023} {"canonical_length": 92, "canonical_nll": 0.7605702773384426, "generation_rouge_l_f1": 0.43396226415094336, "generation_token_f1": 0.5943396226415095, "prompt_id": "c7f44052e9719f7b", "reading_canonical_log_prob": -69.97246551513672, "reading_distances": [30, 36, 40, 40], "reading_log_probs": [-68.69145202636719, -70.69593811035156, -67.96772766113281, -70.85458374023438], "reading_logprob_std": 1.4440429051166221, "tokenization_path_logprob_std": 167.48811481396103, "trajectory_balance_rmse": 1795.877365744755} {"canonical_length": 22, "canonical_nll": 1.9785177057439631, "generation_rouge_l_f1": 0.37209302325581395, "generation_token_f1": 0.46511627906976744, "prompt_id": "4d4bec1ba05bf153", "reading_canonical_log_prob": -43.52738952636719, "reading_distances": [41, 31, 45, 38], "reading_log_probs": [-42.37485122680664, -47.089752197265625, -44.75699996948242, -43.64701843261719], "reading_logprob_std": 2.001031836654661, "tokenization_path_logprob_std": 83.28436707002169, "trajectory_balance_rmse": 672.794106040133} {"canonical_length": 56, "canonical_nll": 1.9135830742972237, "generation_rouge_l_f1": 0.3235294117647059, "generation_token_f1": 0.3235294117647059, "prompt_id": "2dc7b73ea6e1f7f9", "reading_canonical_log_prob": -107.16065216064453, "reading_distances": [45, 47, 42, 49], "reading_log_probs": [-110.58990478515625, -117.40584564208984, -113.86617279052734, -112.2822265625], "reading_logprob_std": 2.9060938392226565, "tokenization_path_logprob_std": 128.1136266998549, "trajectory_balance_rmse": 1165.868409072356} {"canonical_length": 14, "canonical_nll": 1.7685202189854212, "generation_rouge_l_f1": 0.14285714285714288, "generation_token_f1": 0.21428571428571427, "prompt_id": "cc99639a6604cf81", "reading_canonical_log_prob": -24.7592830657959, "reading_distances": [37, 35, 37, 33], "reading_log_probs": [-28.893600463867188, -24.862056732177734, -25.170921325683594, -27.302200317382812], "reading_logprob_std": 1.898163310646633, "tokenization_path_logprob_std": 23.63515463445906, "trajectory_balance_rmse": 309.6691902834383} {"canonical_length": 41, "canonical_nll": 0.9762666283584223, "generation_rouge_l_f1": 0.1971830985915493, "generation_token_f1": 0.2535211267605634, "prompt_id": "34af9dd76e319507", "reading_canonical_log_prob": -40.02693176269531, "reading_distances": [29, 35, 37, 35], "reading_log_probs": [-40.447357177734375, -39.54676055908203, -42.151611328125, -40.80850601196289], "reading_logprob_std": 1.0811708917599672, "tokenization_path_logprob_std": 104.41586052003053, "trajectory_balance_rmse": 812.0666949912992} {"canonical_length": 319, "canonical_nll": 1.2870479571781936, "generation_rouge_l_f1": 0.31649831649831656, "generation_token_f1": 0.5252525252525252, "prompt_id": "761d4b8b4ce8e2b9", "reading_canonical_log_prob": -410.56829833984375, "reading_distances": [29, 23, 27, 31], "reading_log_probs": [-410.16497802734375, -408.90582275390625, -407.9863586425781, -406.8642883300781], "reading_logprob_std": 1.3993666944320133, "tokenization_path_logprob_std": 218.9758400888635, "trajectory_balance_rmse": 6867.746510174096} {"canonical_length": 30, "canonical_nll": 0.5538286209106446, "generation_rouge_l_f1": 0.42424242424242425, "generation_token_f1": 0.42424242424242425, "prompt_id": "667ddd428394372e", "reading_canonical_log_prob": -16.614858627319336, "reading_distances": [32, 33, 31, 34], "reading_log_probs": [-17.303810119628906, -16.594478607177734, -15.345611572265625, -17.093408584594727], "reading_logprob_std": 0.877750721112335, "tokenization_path_logprob_std": 71.32250537570293, "trajectory_balance_rmse": 621.9593862176134} {"canonical_length": 15, "canonical_nll": 0.8659201939900716, "generation_rouge_l_f1": 0.5714285714285715, "generation_token_f1": 0.5714285714285715, "prompt_id": "eaab92bbea496d12", "reading_canonical_log_prob": -12.988802909851074, "reading_distances": [28, 24, 20, 23], "reading_log_probs": [-14.020936965942383, -13.417543411254883, -13.34017562866211, -13.193440437316895], "reading_logprob_std": 0.36400657253127167, "tokenization_path_logprob_std": 41.50651857767398, "trajectory_balance_rmse": 328.66807199461107} {"canonical_length": 77, "canonical_nll": 0.9417536351587865, "generation_rouge_l_f1": 0.6888888888888889, "generation_token_f1": 0.7555555555555555, "prompt_id": "7cb40862b1b15473", "reading_canonical_log_prob": -72.51502990722656, "reading_distances": [33, 25, 25, 25], "reading_log_probs": [-69.24974060058594, -73.54647064208984, -70.6950912475586, -68.67752838134766], "reading_logprob_std": 2.1753520304237344, "tokenization_path_logprob_std": 116.43587102509619, "trajectory_balance_rmse": 774.9773815988888} {"canonical_length": 37, "canonical_nll": 1.314212283572635, "generation_rouge_l_f1": 0.14414414414414414, "generation_token_f1": 0.2342342342342342, "prompt_id": "b218ec9e88427665", "reading_canonical_log_prob": -48.6258544921875, "reading_distances": [28, 35, 29, 27], "reading_log_probs": [-54.89044189453125, -52.3382568359375, -53.613525390625, -55.496238708496094], "reading_logprob_std": 1.4040098213788696, "tokenization_path_logprob_std": 94.87182468914348, "trajectory_balance_rmse": 926.7518632388542} {"canonical_length": 68, "canonical_nll": 1.558898364796358, "generation_rouge_l_f1": 0.21052631578947373, "generation_token_f1": 0.29824561403508776, "prompt_id": "605c5bbf05c7cb07", "reading_canonical_log_prob": -106.00508880615234, "reading_distances": [34, 31, 35, 28], "reading_log_probs": [-99.96986389160156, -99.34304809570312, -99.68578338623047, -99.20158386230469], "reading_logprob_std": 0.3459025352381215, "tokenization_path_logprob_std": 75.37164712154772, "trajectory_balance_rmse": 1843.9640047489688} {"canonical_length": 163, "canonical_nll": 2.177920288834835, "generation_rouge_l_f1": 0.17266187050359713, "generation_token_f1": 0.36450839328537166, "prompt_id": "371fd9b264246fc1", "reading_canonical_log_prob": -355.0010070800781, "reading_distances": [18, 15, 18, 17], "reading_log_probs": [-352.3495178222656, -348.92486572265625, -353.4208068847656, -349.37884521484375], "reading_logprob_std": 2.207144392891555, "tokenization_path_logprob_std": 169.78602535256917, "trajectory_balance_rmse": 3356.0157720642296} {"canonical_length": 9, "canonical_nll": 1.7685320112440321, "generation_rouge_l_f1": 0.3157894736842105, "generation_token_f1": 0.3157894736842105, "prompt_id": "ed3c13f24a96d3b3", "reading_canonical_log_prob": -15.916788101196289, "reading_distances": [28, 24, 28, 29], "reading_log_probs": [-17.830772399902344, -15.726951599121094, -15.074813842773438, -15.286107063293457], "reading_logprob_std": 1.2636245351109272, "tokenization_path_logprob_std": 28.918727420143906, "trajectory_balance_rmse": 164.90737664573246} {"canonical_length": 27, "canonical_nll": 0.9431981687192563, "generation_rouge_l_f1": 0.4745762711864407, "generation_token_f1": 0.4745762711864407, "prompt_id": "9e311e757597dd6a", "reading_canonical_log_prob": -25.466350555419922, "reading_distances": [32, 34, 33, 34], "reading_log_probs": [-28.455894470214844, -28.02688980102539, -27.215709686279297, -27.126331329345703], "reading_logprob_std": 0.6433537570297228, "tokenization_path_logprob_std": 55.41088524174891, "trajectory_balance_rmse": 625.9073868899719} {"canonical_length": 23, "canonical_nll": 1.0604539124862007, "generation_rouge_l_f1": 0.08695652173913042, "generation_token_f1": 0.08695652173913042, "prompt_id": "6d2468c671bba59a", "reading_canonical_log_prob": -24.390439987182617, "reading_distances": [38, 41, 45, 45], "reading_log_probs": [-20.79705810546875, -22.40316390991211, -21.39739990234375, -21.757230758666992], "reading_logprob_std": 0.6720730845706215, "tokenization_path_logprob_std": 55.31681903362679, "trajectory_balance_rmse": 442.14063994480694} {"canonical_length": 34, "canonical_nll": 1.661345874561983, "generation_rouge_l_f1": 0.25503355704697983, "generation_token_f1": 0.32214765100671144, "prompt_id": "d8ebdfdb02245d42", "reading_canonical_log_prob": -56.48575973510742, "reading_distances": [20, 20, 21, 20], "reading_log_probs": [-57.42917251586914, -57.380088806152344, -57.70338439941406, -59.44419479370117], "reading_logprob_std": 0.9803651762273593, "tokenization_path_logprob_std": 87.11234712953014, "trajectory_balance_rmse": 912.8679288644518} {"canonical_length": 282, "canonical_nll": 1.201418288210605, "generation_rouge_l_f1": 0.28467153284671537, "generation_token_f1": 0.5072992700729928, "prompt_id": "0659b1d0d1c9b272", "reading_canonical_log_prob": -338.7999572753906, "reading_distances": [22, 21, 19, 21], "reading_log_probs": [-333.9881286621094, -335.52117919921875, -335.0929260253906, -334.25872802734375], "reading_logprob_std": 0.7139745106279839, "tokenization_path_logprob_std": 89.85483022283098, "trajectory_balance_rmse": 5057.457634221073} {"canonical_length": 126, "canonical_nll": 1.5847654796781994, "generation_rouge_l_f1": 0.308300395256917, "generation_token_f1": 0.5059288537549407, "prompt_id": "11e44461a25c9604", "reading_canonical_log_prob": -199.68045043945312, "reading_distances": [13, 16, 12, 13], "reading_log_probs": [-197.6277313232422, -199.9305419921875, -197.73831176757812, -199.1090850830078], "reading_logprob_std": 1.1131508682891125, "tokenization_path_logprob_std": 159.29535143006552, "trajectory_balance_rmse": 3272.962603955297} {"canonical_length": 4, "canonical_nll": 0.657460629940033, "generation_rouge_l_f1": 0.5, "generation_token_f1": 0.5, "prompt_id": "72cbdeacc12e1d9a", "reading_canonical_log_prob": -2.629842519760132, "reading_distances": [20, 24, 19, 21], "reading_log_probs": [-2.5642807483673096, -2.5238795280456543, -3.010279893875122, -2.665376663208008], "reading_logprob_std": 0.22104540899580294, "tokenization_path_logprob_std": 27.61598673639671, "trajectory_balance_rmse": 73.00443253377438} {"canonical_length": 9, "canonical_nll": 0.8841730753580729, "generation_rouge_l_f1": 0.875, "generation_token_f1": 0.875, "prompt_id": "e46df451982e05b1", "reading_canonical_log_prob": -7.957557678222656, "reading_distances": [16, 22, 15, 22], "reading_log_probs": [-10.103239059448242, -9.758742332458496, -11.25351333618164, -11.879003524780273], "reading_logprob_std": 0.9881056875079862, "tokenization_path_logprob_std": 29.40263114301615, "trajectory_balance_rmse": 118.79162125238484} {"canonical_length": 112, "canonical_nll": 1.2069222586495536, "generation_rouge_l_f1": 0.28571428571428575, "generation_token_f1": 0.5051020408163266, "prompt_id": "ad4b8c7e2797f458", "reading_canonical_log_prob": -135.17529296875, "reading_distances": [36, 35, 37, 30], "reading_log_probs": [-117.72935485839844, -116.63903045654297, -116.89141082763672, -121.61051177978516], "reading_logprob_std": 2.309465536006587, "tokenization_path_logprob_std": 109.88716852497045, "trajectory_balance_rmse": 1875.3884953510033} {"canonical_length": 229, "canonical_nll": 1.2484941107737446, "generation_rouge_l_f1": 0.25503355704697983, "generation_token_f1": 0.4697986577181208, "prompt_id": "9d2ca51ff7ce1141", "reading_canonical_log_prob": -285.9051513671875, "reading_distances": [29, 33, 32, 26], "reading_log_probs": [-282.7136535644531, -283.9285888671875, -281.1611328125, -280.5828857421875], "reading_logprob_std": 1.5169451780102947, "tokenization_path_logprob_std": 165.21464113171263, "trajectory_balance_rmse": 5135.2424490486665} {"canonical_length": 3, "canonical_nll": 1.1515288352966309, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.6666666666666666, "prompt_id": "f22d1f7967fa63cf", "reading_canonical_log_prob": -3.4545865058898926, "reading_distances": [24, 31, 24, 32], "reading_log_probs": [-3.9289588928222656, -4.345743179321289, -4.002487659454346, -4.110516548156738], "reading_logprob_std": 0.18186647206304712, "tokenization_path_logprob_std": 36.1417792863797, "trajectory_balance_rmse": 57.729030776289335} {"canonical_length": 341, "canonical_nll": 1.4077369488579088, "generation_rouge_l_f1": 0.2195540308747856, "generation_token_f1": 0.41852487135506006, "prompt_id": "a546d753c3fe436f", "reading_canonical_log_prob": -480.0382995605469, "reading_distances": [20, 22, 25, 22], "reading_log_probs": [-480.0500793457031, -480.10223388671875, -478.2117614746094, -476.75042724609375], "reading_logprob_std": 1.6128088359108055, "tokenization_path_logprob_std": 152.45175187815082, "trajectory_balance_rmse": 7076.293184552167} {"canonical_length": 175, "canonical_nll": 1.3976236397879465, "generation_rouge_l_f1": 0.1862745098039216, "generation_token_f1": 0.3725490196078432, "prompt_id": "55985253cc61ade7", "reading_canonical_log_prob": -244.58413696289062, "reading_distances": [25, 30, 26, 24], "reading_log_probs": [-244.96560668945312, -244.8601531982422, -246.50384521484375, -245.52969360351562], "reading_logprob_std": 0.7524665777742594, "tokenization_path_logprob_std": 134.40676848341045, "trajectory_balance_rmse": 3248.206618228027} {"canonical_length": 134, "canonical_nll": 1.5210840025944496, "generation_rouge_l_f1": 0.25210084033613445, "generation_token_f1": 0.45378151260504207, "prompt_id": "e4db8e939e6495ed", "reading_canonical_log_prob": -203.82525634765625, "reading_distances": [17, 15, 20, 15], "reading_log_probs": [-204.13259887695312, -202.39950561523438, -201.73593139648438, -203.028564453125], "reading_logprob_std": 1.019536874402087, "tokenization_path_logprob_std": 104.34171969043265, "trajectory_balance_rmse": 2942.4174161200444} {"canonical_length": 36, "canonical_nll": 0.7607228491041396, "generation_rouge_l_f1": 0.2436548223350254, "generation_token_f1": 0.3147208121827411, "prompt_id": "8ba984bcebc3168b", "reading_canonical_log_prob": -27.386022567749023, "reading_distances": [16, 17, 15, 15], "reading_log_probs": [-31.4309024810791, -32.09177780151367, -27.0765438079834, -29.374338150024414], "reading_logprob_std": 2.2628072829738333, "tokenization_path_logprob_std": 62.16713808351451, "trajectory_balance_rmse": 718.8265753293463} {"canonical_length": 4, "canonical_nll": 2.04276180267334, "generation_rouge_l_f1": 0.01801801801801802, "generation_token_f1": 0.01801801801801802, "prompt_id": "fe48fe173961fdd5", "reading_canonical_log_prob": -8.17104721069336, "reading_distances": [18, 27, 24, 23], "reading_log_probs": [-7.12103796005249, -7.501319885253906, -7.161228656768799, -6.7926025390625], "reading_logprob_std": 0.28981703935969483, "tokenization_path_logprob_std": 17.92793209697339, "trajectory_balance_rmse": 38.65574552767081} {"canonical_length": 267, "canonical_nll": 1.734784758492802, "generation_rouge_l_f1": 0.15298507462686567, "generation_token_f1": 0.30597014925373134, "prompt_id": "51d00b55119644b0", "reading_canonical_log_prob": -463.1875305175781, "reading_distances": [25, 28, 26, 25], "reading_log_probs": [-465.73626708984375, -469.1628723144531, -465.1300964355469, -464.2211608886719], "reading_logprob_std": 2.1585985029226413, "tokenization_path_logprob_std": 186.9609600601288, "trajectory_balance_rmse": 5630.207559784664} {"canonical_length": 7, "canonical_nll": 0.14705043179648264, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "e708a1adc69b9f0e", "reading_canonical_log_prob": -1.0293530225753784, "reading_distances": [22, 28, 24, 25], "reading_log_probs": [-0.8071825504302979, -0.7038798332214355, -0.7188054323196411, -0.6147854924201965], "reading_logprob_std": 0.07878203966596341, "tokenization_path_logprob_std": 16.032472238674057, "trajectory_balance_rmse": 118.46711616313578} {"canonical_length": 22, "canonical_nll": 1.2191774194890803, "generation_rouge_l_f1": 0.11255411255411255, "generation_token_f1": 0.13852813852813853, "prompt_id": "d7a89d7235636e3d", "reading_canonical_log_prob": -26.821903228759766, "reading_distances": [14, 16, 15, 14], "reading_log_probs": [-25.929935455322266, -25.96540641784668, -26.284461975097656, -25.821577072143555], "reading_logprob_std": 0.19904659092209864, "tokenization_path_logprob_std": 48.38111748097219, "trajectory_balance_rmse": 548.2076879504075} {"canonical_length": 12, "canonical_nll": 0.6246183315912882, "generation_rouge_l_f1": 0.9, "generation_token_f1": 0.9, "prompt_id": "17e2c9ee9612ad40", "reading_canonical_log_prob": -7.495419979095459, "reading_distances": [21, 22, 23, 29], "reading_log_probs": [-6.861808776855469, -7.504889011383057, -7.997581958770752, -7.954535007476807], "reading_logprob_std": 0.5279176408391717, "tokenization_path_logprob_std": 25.515072766803762, "trajectory_balance_rmse": 255.71332393088667} {"canonical_length": 324, "canonical_nll": 1.1784976912133487, "generation_rouge_l_f1": 0.2285714285714286, "generation_token_f1": 0.45333333333333337, "prompt_id": "36541d1f10617d25", "reading_canonical_log_prob": -381.833251953125, "reading_distances": [11, 11, 15, 17], "reading_log_probs": [-385.27349853515625, -389.72711181640625, -385.14886474609375, -387.0859375], "reading_logprob_std": 2.1374349639615025, "tokenization_path_logprob_std": 184.14435464202592, "trajectory_balance_rmse": 6319.777938436011} {"canonical_length": 315, "canonical_nll": 1.1472240629650297, "generation_rouge_l_f1": 0.22996515679442509, "generation_token_f1": 0.5261324041811847, "prompt_id": "e15e866d282a743f", "reading_canonical_log_prob": -361.3755798339844, "reading_distances": [24, 27, 30, 33], "reading_log_probs": [-355.8502197265625, -356.7173156738281, -357.56988525390625, -357.609375], "reading_logprob_std": 0.8330634406660983, "tokenization_path_logprob_std": 234.29226191122717, "trajectory_balance_rmse": 5742.935281610042} {"canonical_length": 10, "canonical_nll": 1.7977611541748046, "generation_rouge_l_f1": 0.26666666666666666, "generation_token_f1": 0.26666666666666666, "prompt_id": "718643fa581e9e44", "reading_canonical_log_prob": -17.977611541748047, "reading_distances": [18, 21, 22, 21], "reading_log_probs": [-16.271677017211914, -16.633338928222656, -17.109058380126953, -16.26926612854004], "reading_logprob_std": 0.3975048976041953, "tokenization_path_logprob_std": 24.697656636296102, "trajectory_balance_rmse": 137.76940743884924} {"canonical_length": 159, "canonical_nll": 0.9692782036163522, "generation_rouge_l_f1": 0.3761904761904762, "generation_token_f1": 0.5714285714285715, "prompt_id": "482bbea97f7ae0fa", "reading_canonical_log_prob": -154.115234375, "reading_distances": [12, 10, 15, 12], "reading_log_probs": [-154.60284423828125, -155.20297241210938, -154.31829833984375, -155.66329956054688], "reading_logprob_std": 0.6034113622619534, "tokenization_path_logprob_std": 170.01160671724278, "trajectory_balance_rmse": 3169.744349593685} {"canonical_length": 11, "canonical_nll": 0.8545420386574485, "generation_rouge_l_f1": 0.08780487804878051, "generation_token_f1": 0.08780487804878051, "prompt_id": "f9bf285092a25c3b", "reading_canonical_log_prob": -9.399962425231934, "reading_distances": [47, 46, 43, 42], "reading_log_probs": [-11.638683319091797, -11.46264362335205, -12.19486141204834, -11.73933219909668], "reading_logprob_std": 0.3123380332922423, "tokenization_path_logprob_std": 16.70995726946322, "trajectory_balance_rmse": 220.0073259910813} {"canonical_length": 150, "canonical_nll": 1.019150390625, "generation_rouge_l_f1": 0.36453201970443355, "generation_token_f1": 0.5221674876847291, "prompt_id": "9d4eb4020bd4213f", "reading_canonical_log_prob": -152.87255859375, "reading_distances": [31, 18, 28, 24], "reading_log_probs": [-150.65969848632812, -149.28817749023438, -149.23025512695312, -152.33538818359375], "reading_logprob_std": 1.4623899796578317, "tokenization_path_logprob_std": 88.91500923649983, "trajectory_balance_rmse": 2761.9060287799784} {"canonical_length": 134, "canonical_nll": 1.0508746246793377, "generation_rouge_l_f1": 0.311614730878187, "generation_token_f1": 0.43059490084985835, "prompt_id": "13778e3d12464b89", "reading_canonical_log_prob": -140.81719970703125, "reading_distances": [18, 19, 21, 19], "reading_log_probs": [-140.32525634765625, -139.8244171142578, -140.24876403808594, -140.214599609375], "reading_logprob_std": 0.22405650781090014, "tokenization_path_logprob_std": 88.10620724858234, "trajectory_balance_rmse": 1560.4403579787386} {"canonical_length": 395, "canonical_nll": 1.3448484782931172, "generation_rouge_l_f1": 0.21895424836601307, "generation_token_f1": 0.3790849673202615, "prompt_id": "513a1cb695ee9eb0", "reading_canonical_log_prob": -531.2151489257812, "reading_distances": [25, 24, 16, 27], "reading_log_probs": [-534.2236328125, -536.2386474609375, -531.8087768554688, -533.8405151367188], "reading_logprob_std": 1.8152443625510093, "tokenization_path_logprob_std": 410.66326178109165, "trajectory_balance_rmse": 5720.990226225664} {"canonical_length": 193, "canonical_nll": 1.1211751829157222, "generation_rouge_l_f1": 0.30869565217391304, "generation_token_f1": 0.5347826086956522, "prompt_id": "444facd57b3cdbe1", "reading_canonical_log_prob": -216.38681030273438, "reading_distances": [24, 22, 22, 14], "reading_log_probs": [-220.69027709960938, -217.28382873535156, -221.3335723876953, -223.21043395996094], "reading_logprob_std": 2.4734690953247287, "tokenization_path_logprob_std": 213.2534357945614, "trajectory_balance_rmse": 3813.2369789965524} {"canonical_length": 226, "canonical_nll": 1.1711828181173949, "generation_rouge_l_f1": 0.30165289256198347, "generation_token_f1": 0.5413223140495868, "prompt_id": "5870502ae5fa3ae7", "reading_canonical_log_prob": -264.68731689453125, "reading_distances": [27, 20, 23, 23], "reading_log_probs": [-261.296875, -261.9423522949219, -261.6834716796875, -262.3160705566406], "reading_logprob_std": 0.429313731911138, "tokenization_path_logprob_std": 194.54005786928144, "trajectory_balance_rmse": 2741.292174231028} {"canonical_length": 17, "canonical_nll": 1.0182896782370174, "generation_rouge_l_f1": 0.14545454545454545, "generation_token_f1": 0.2, "prompt_id": "8c61a7a56532124f", "reading_canonical_log_prob": -17.310924530029297, "reading_distances": [29, 40, 31, 29], "reading_log_probs": [-16.51970863342285, -16.484289169311523, -15.831768035888672, -15.330037117004395], "reading_logprob_std": 0.5700621797908052, "tokenization_path_logprob_std": 75.22673693737808, "trajectory_balance_rmse": 372.73560858809276} {"canonical_length": 12, "canonical_nll": 0.5646488666534424, "generation_rouge_l_f1": 0.09836065573770492, "generation_token_f1": 0.09836065573770492, "prompt_id": "fb0ad86c22711d8d", "reading_canonical_log_prob": -6.775786399841309, "reading_distances": [9, 12, 14, 16], "reading_log_probs": [-6.98173713684082, -5.441219806671143, -5.879781723022461, -6.44910192489624], "reading_logprob_std": 0.6710370120550482, "tokenization_path_logprob_std": 40.55084659282461, "trajectory_balance_rmse": 180.13345260152207} {"canonical_length": 219, "canonical_nll": 1.422999273152112, "generation_rouge_l_f1": 0.21632653061224488, "generation_token_f1": 0.43265306122448977, "prompt_id": "a18ee8ea214b4b09", "reading_canonical_log_prob": -311.6368408203125, "reading_distances": [36, 42, 52, 45], "reading_log_probs": [-311.8482666015625, -312.3266906738281, -307.3423156738281, -308.5465087890625], "reading_logprob_std": 2.4497948602857775, "tokenization_path_logprob_std": 152.37379819996113, "trajectory_balance_rmse": 4854.707527133664} {"canonical_length": 283, "canonical_nll": 1.405998742201303, "generation_rouge_l_f1": 0.20326678765880216, "generation_token_f1": 0.4065335753176043, "prompt_id": "6933d659ab69515c", "reading_canonical_log_prob": -397.89764404296875, "reading_distances": [32, 27, 25, 23], "reading_log_probs": [-393.895751953125, -393.5091247558594, -393.3498840332031, -386.06842041015625], "reading_logprob_std": 3.7652326479790865, "tokenization_path_logprob_std": 348.92923497486964, "trajectory_balance_rmse": 5321.067172377894} {"canonical_length": 49, "canonical_nll": 1.311746714066486, "generation_rouge_l_f1": 0.2900763358778626, "generation_token_f1": 0.42748091603053434, "prompt_id": "e748f45af7778fcd", "reading_canonical_log_prob": -64.27558898925781, "reading_distances": [21, 31, 38, 33], "reading_log_probs": [-65.06024932861328, -65.07095336914062, -66.55400848388672, -65.01127624511719], "reading_logprob_std": 0.7537056203849887, "tokenization_path_logprob_std": 57.896892026066475, "trajectory_balance_rmse": 859.3548615242655} {"canonical_length": 8, "canonical_nll": 1.1972754001617432, "generation_rouge_l_f1": 0.631578947368421, "generation_token_f1": 0.631578947368421, "prompt_id": "3f21812c287e05df", "reading_canonical_log_prob": -9.578203201293945, "reading_distances": [41, 27, 42, 32], "reading_log_probs": [-12.2911376953125, -12.480477333068848, -11.652738571166992, -12.022106170654297], "reading_logprob_std": 0.3591040825715286, "tokenization_path_logprob_std": 20.758783871917686, "trajectory_balance_rmse": 186.6215756576284} {"canonical_length": 38, "canonical_nll": 1.5513419101112766, "generation_rouge_l_f1": 0.12274368231046931, "generation_token_f1": 0.18050541516245489, "prompt_id": "cc2b221229d6b7ba", "reading_canonical_log_prob": -58.950992584228516, "reading_distances": [27, 19, 21, 26], "reading_log_probs": [-61.14752960205078, -60.3753662109375, -60.979408264160156, -60.10136413574219], "reading_logprob_std": 0.4941211643285163, "tokenization_path_logprob_std": 209.02452396612338, "trajectory_balance_rmse": 942.975060794263} {"canonical_length": 57, "canonical_nll": 1.0049971195689418, "generation_rouge_l_f1": 0.27, "generation_token_f1": 0.38, "prompt_id": "6c3587030700c60e", "reading_canonical_log_prob": -57.28483581542969, "reading_distances": [33, 25, 30, 23], "reading_log_probs": [-58.346656799316406, -58.3870735168457, -58.07143783569336, -57.93963623046875], "reading_logprob_std": 0.2160714395856372, "tokenization_path_logprob_std": 57.660680733589714, "trajectory_balance_rmse": 1009.731727089045} {"canonical_length": 108, "canonical_nll": 1.6441766244393807, "generation_rouge_l_f1": 0.22058823529411764, "generation_token_f1": 0.4117647058823529, "prompt_id": "3f751ca73a73cf7d", "reading_canonical_log_prob": -177.57107543945312, "reading_distances": [26, 21, 13, 23], "reading_log_probs": [-175.56222534179688, -177.3455047607422, -174.0823974609375, -177.41419982910156], "reading_logprob_std": 1.5956524482582177, "tokenization_path_logprob_std": 21.025864109266113, "trajectory_balance_rmse": 1743.6676274542372} {"canonical_length": 77, "canonical_nll": 0.6931510033545556, "generation_rouge_l_f1": 0.2275449101796407, "generation_token_f1": 0.3233532934131737, "prompt_id": "7c7baba7466fe6f6", "reading_canonical_log_prob": -53.37262725830078, "reading_distances": [8, 7, 7, 11], "reading_log_probs": [-54.37445831298828, -53.923152923583984, -53.69799041748047, -55.1357421875], "reading_logprob_std": 0.6343657945186412, "tokenization_path_logprob_std": 117.13606532296639, "trajectory_balance_rmse": 1458.386421321048} {"canonical_length": 294, "canonical_nll": 0.7925309161750638, "generation_rouge_l_f1": 0.2304964539007092, "generation_token_f1": 0.5, "prompt_id": "fd3d2e71f4e4243f", "reading_canonical_log_prob": -233.00408935546875, "reading_distances": [15, 19, 17, 20], "reading_log_probs": [-227.686767578125, -226.8975830078125, -226.6248779296875, -226.93478393554688], "reading_logprob_std": 0.45531079232672483, "tokenization_path_logprob_std": 200.17535526874093, "trajectory_balance_rmse": 4796.894533104741} {"canonical_length": 399, "canonical_nll": 0.8975491248874139, "generation_rouge_l_f1": 0.2119815668202765, "generation_token_f1": 0.4178187403993855, "prompt_id": "5bc59064e72854cb", "reading_canonical_log_prob": -358.1221008300781, "reading_distances": [22, 20, 21, 24], "reading_log_probs": [-355.4468994140625, -356.7212219238281, -355.63250732421875, -356.328857421875], "reading_logprob_std": 0.5958433045590315, "tokenization_path_logprob_std": 92.80339935074124, "trajectory_balance_rmse": 7525.628717016181} {"canonical_length": 280, "canonical_nll": 1.1691848754882812, "generation_rouge_l_f1": 0.27522935779816515, "generation_token_f1": 0.5394495412844037, "prompt_id": "cd25f5705dfa5cbb", "reading_canonical_log_prob": -327.37176513671875, "reading_distances": [22, 18, 25, 28], "reading_log_probs": [-328.6999816894531, -326.33966064453125, -331.1470947265625, -328.3005676269531], "reading_logprob_std": 1.9743839634261364, "tokenization_path_logprob_std": 149.72231430103173, "trajectory_balance_rmse": 5985.243138237331} {"canonical_length": 365, "canonical_nll": 0.8526720856967037, "generation_rouge_l_f1": 0.1930379746835443, "generation_token_f1": 0.43670886075949367, "prompt_id": "ee3200ffbd24e2bc", "reading_canonical_log_prob": -311.2253112792969, "reading_distances": [26, 28, 28, 31], "reading_log_probs": [-312.9798583984375, -308.94854736328125, -313.3254699707031, -313.021728515625], "reading_logprob_std": 2.0859287342107313, "tokenization_path_logprob_std": 198.7050978860685, "trajectory_balance_rmse": 5071.36867774499} {"canonical_length": 115, "canonical_nll": 2.150439585810122, "generation_rouge_l_f1": 0.21761658031088082, "generation_token_f1": 0.2694300518134715, "prompt_id": "4180e393ac001d5f", "reading_canonical_log_prob": -247.30055236816406, "reading_distances": [23, 25, 20, 17], "reading_log_probs": [-251.3296661376953, -253.06761169433594, -252.26840209960938, -249.98800659179688], "reading_logprob_std": 1.3236514375354849, "tokenization_path_logprob_std": 113.58050904503202, "trajectory_balance_rmse": 2546.8704702984483} {"canonical_length": 9, "canonical_nll": 0.9127914640638564, "generation_rouge_l_f1": 0.7692307692307692, "generation_token_f1": 0.7692307692307692, "prompt_id": "79896777ff1e0459", "reading_canonical_log_prob": -8.215123176574707, "reading_distances": [36, 51, 52, 45], "reading_log_probs": [-5.967486381530762, -6.928163528442383, -6.611757278442383, -5.706564426422119], "reading_logprob_std": 0.5640520969074008, "tokenization_path_logprob_std": 37.3201418165549, "trajectory_balance_rmse": 179.92082441577006} {"canonical_length": 147, "canonical_nll": 1.4004643239131591, "generation_rouge_l_f1": 0.16786570743405277, "generation_token_f1": 0.3405275779376499, "prompt_id": "4be1dda9c6193970", "reading_canonical_log_prob": -205.86825561523438, "reading_distances": [43, 40, 45, 43], "reading_log_probs": [-204.63571166992188, -204.0247802734375, -204.33187866210938, -203.5728759765625], "reading_logprob_std": 0.4536686905525295, "tokenization_path_logprob_std": 208.80520515878703, "trajectory_balance_rmse": 3886.6576810924903} {"canonical_length": 9, "canonical_nll": 1.494913313123915, "generation_rouge_l_f1": 0.05263157894736842, "generation_token_f1": 0.05263157894736842, "prompt_id": "e275be5afb765439", "reading_canonical_log_prob": -13.454219818115234, "reading_distances": [32, 31, 30, 24], "reading_log_probs": [-23.293527603149414, -27.18466567993164, -21.39736557006836, -26.801816940307617], "reading_logprob_std": 2.79720083884224, "tokenization_path_logprob_std": 33.020381243339024, "trajectory_balance_rmse": 122.42285646919059} {"canonical_length": 8, "canonical_nll": 1.4943230152130127, "generation_rouge_l_f1": 0.05555555555555555, "generation_token_f1": 0.05555555555555555, "prompt_id": "b4100d452af762d4", "reading_canonical_log_prob": -11.954584121704102, "reading_distances": [17, 18, 23, 23], "reading_log_probs": [-16.911819458007812, -20.667675018310547, -17.500505447387695, -19.692846298217773], "reading_logprob_std": 1.7789211384205372, "tokenization_path_logprob_std": 10.247899713873384, "trajectory_balance_rmse": 111.10984256689403} {"canonical_length": 65, "canonical_nll": 0.48361329298753003, "generation_rouge_l_f1": 0.2642857142857143, "generation_token_f1": 0.3428571428571429, "prompt_id": "660009a4d953661a", "reading_canonical_log_prob": -31.434864044189453, "reading_distances": [24, 31, 30, 34], "reading_log_probs": [-31.215421676635742, -32.20426940917969, -31.975692749023438, -31.898502349853516], "reading_logprob_std": 0.42564953205382255, "tokenization_path_logprob_std": 59.1629290002315, "trajectory_balance_rmse": 875.4747807540131} {"canonical_length": 245, "canonical_nll": 0.9647109673947704, "generation_rouge_l_f1": 0.3537735849056604, "generation_token_f1": 0.5896226415094339, "prompt_id": "9c1c5be112145501", "reading_canonical_log_prob": -236.35418701171875, "reading_distances": [26, 30, 16, 16], "reading_log_probs": [-235.21389770507812, -234.95599365234375, -235.1016845703125, -236.26983642578125], "reading_logprob_std": 0.5990339656646148, "tokenization_path_logprob_std": 54.08160014974333, "trajectory_balance_rmse": 4114.285681902084} {"canonical_length": 41, "canonical_nll": 0.582754786421613, "generation_rouge_l_f1": 0.18518518518518515, "generation_token_f1": 0.21296296296296297, "prompt_id": "8a1926c471e546ee", "reading_canonical_log_prob": -23.892946243286133, "reading_distances": [21, 15, 27, 24], "reading_log_probs": [-20.85249137878418, -21.306346893310547, -21.46192169189453, -20.847774505615234], "reading_logprob_std": 0.31478581377158543, "tokenization_path_logprob_std": 37.92745030528848, "trajectory_balance_rmse": 463.8685712207956} {"canonical_length": 311, "canonical_nll": 1.1778171944081592, "generation_rouge_l_f1": 0.21143847487001732, "generation_token_f1": 0.44714038128249567, "prompt_id": "a1e93e4a3a2191cd", "reading_canonical_log_prob": -366.3011474609375, "reading_distances": [37, 35, 36, 42], "reading_log_probs": [-366.35662841796875, -365.5750732421875, -364.3403015136719, -362.3209228515625], "reading_logprob_std": 1.7596251557228797, "tokenization_path_logprob_std": 235.24074829515357, "trajectory_balance_rmse": 6004.23623726763} {"canonical_length": 130, "canonical_nll": 1.4980806790865384, "generation_rouge_l_f1": 0.24870466321243523, "generation_token_f1": 0.42487046632124353, "prompt_id": "c4f0abd89b25e9b2", "reading_canonical_log_prob": -194.75048828125, "reading_distances": [22, 21, 18, 21], "reading_log_probs": [-195.3485107421875, -194.74325561523438, -195.6532745361328, -195.66748046875], "reading_logprob_std": 0.43238526898637125, "tokenization_path_logprob_std": 200.8001502347313, "trajectory_balance_rmse": 3005.38205026707} {"canonical_length": 226, "canonical_nll": 1.12365945461577, "generation_rouge_l_f1": 0.3393665158371041, "generation_token_f1": 0.5113122171945702, "prompt_id": "8d709bdbfc4890e5", "reading_canonical_log_prob": -253.94703674316406, "reading_distances": [23, 27, 30, 20], "reading_log_probs": [-250.3424835205078, -248.5537567138672, -249.36105346679688, -249.34909057617188], "reading_logprob_std": 0.732234319938505, "tokenization_path_logprob_std": 107.28295496451976, "trajectory_balance_rmse": 3660.7842525211195} {"canonical_length": 410, "canonical_nll": 1.1748168200981326, "generation_rouge_l_f1": 0.2660098522167488, "generation_token_f1": 0.44006568144499175, "prompt_id": "fbcfdeca84611d04", "reading_canonical_log_prob": -481.6748962402344, "reading_distances": [30, 19, 29, 28], "reading_log_probs": [-484.8677978515625, -482.2789306640625, -486.6771240234375, -484.47076416015625], "reading_logprob_std": 1.8062344119328528, "tokenization_path_logprob_std": 162.948260583994, "trajectory_balance_rmse": 7289.437461350554} {"canonical_length": 45, "canonical_nll": 1.299867418077257, "generation_rouge_l_f1": 0.36363636363636365, "generation_token_f1": 0.4772727272727273, "prompt_id": "8c6898d598b2732f", "reading_canonical_log_prob": -58.49403381347656, "reading_distances": [31, 31, 33, 26], "reading_log_probs": [-62.30827331542969, -63.41675567626953, -65.12826538085938, -62.46592712402344], "reading_logprob_std": 1.295097438909116, "tokenization_path_logprob_std": 11.96527270233809, "trajectory_balance_rmse": 909.9362194517627} {"canonical_length": 93, "canonical_nll": 0.5848072831348706, "generation_rouge_l_f1": 0.5490196078431372, "generation_token_f1": 0.588235294117647, "prompt_id": "f5259298cfe12698", "reading_canonical_log_prob": -54.38707733154297, "reading_distances": [19, 16, 17, 21], "reading_log_probs": [-53.4942512512207, -53.582027435302734, -53.6396484375, -54.14229965209961], "reading_logprob_std": 0.2913610461407534, "tokenization_path_logprob_std": 66.92815653805424, "trajectory_balance_rmse": 1164.947978357167} {"canonical_length": 183, "canonical_nll": 0.41351405909804045, "generation_rouge_l_f1": 0.21026894865525672, "generation_token_f1": 0.3422982885085575, "prompt_id": "af8c4385ce2b0480", "reading_canonical_log_prob": -75.6730728149414, "reading_distances": [31, 31, 37, 35], "reading_log_probs": [-73.77691650390625, -74.14875030517578, -74.34087371826172, -69.93710327148438], "reading_logprob_std": 2.0890298798213967, "tokenization_path_logprob_std": 27.627055597547873, "trajectory_balance_rmse": 2589.4241082320364} {"canonical_length": 164, "canonical_nll": 1.6891669296636813, "generation_rouge_l_f1": 0.2513368983957219, "generation_token_f1": 0.5026737967914437, "prompt_id": "8db3810d387bcbf4", "reading_canonical_log_prob": -277.02337646484375, "reading_distances": [24, 22, 19, 25], "reading_log_probs": [-281.66156005859375, -278.51470947265625, -277.58099365234375, -279.866455078125], "reading_logprob_std": 1.7724387163988427, "tokenization_path_logprob_std": 312.3598899868231, "trajectory_balance_rmse": 3697.121954117375} {"canonical_length": 170, "canonical_nll": 0.7405587589039522, "generation_rouge_l_f1": 0.4085510688836105, "generation_token_f1": 0.5605700712589072, "prompt_id": "d92a9462a284aee7", "reading_canonical_log_prob": -125.89498901367188, "reading_distances": [12, 16, 13, 14], "reading_log_probs": [-126.1617202758789, -128.24366760253906, -127.22948455810547, -126.21102142333984], "reading_logprob_std": 0.9863459810006469, "tokenization_path_logprob_std": 135.8239991122816, "trajectory_balance_rmse": 3291.284784442682} {"canonical_length": 76, "canonical_nll": 0.8747148011860094, "generation_rouge_l_f1": 0.2889733840304183, "generation_token_f1": 0.43346007604562736, "prompt_id": "32c20ba8cc10f798", "reading_canonical_log_prob": -66.47832489013672, "reading_distances": [33, 33, 28, 28], "reading_log_probs": [-65.14191436767578, -65.65737915039062, -65.63616943359375, -65.23223876953125], "reading_logprob_std": 0.2680958607894471, "tokenization_path_logprob_std": 95.26978692407518, "trajectory_balance_rmse": 1724.6743558923058} {"canonical_length": 21, "canonical_nll": 0.44541368030366446, "generation_rouge_l_f1": 0.18181818181818182, "generation_token_f1": 0.18181818181818182, "prompt_id": "2e575375b1e838f0", "reading_canonical_log_prob": -9.353687286376953, "reading_distances": [25, 23, 27, 29], "reading_log_probs": [-11.141583442687988, -10.871402740478516, -10.545482635498047, -10.61781120300293], "reading_logprob_std": 0.2705603980060992, "tokenization_path_logprob_std": 81.9136937151435, "trajectory_balance_rmse": 243.3745187558932} {"canonical_length": 9, "canonical_nll": 0.03815457555982801, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "f19b5cf2d17aa856", "reading_canonical_log_prob": -0.34339118003845215, "reading_distances": [37, 36, 30, 42], "reading_log_probs": [-1.0180203914642334, -0.9711762070655823, -1.1289352178573608, -0.6878652572631836], "reading_logprob_std": 0.18779301523831862, "tokenization_path_logprob_std": 22.44929494896359, "trajectory_balance_rmse": 186.12216309979496} {"canonical_length": 3, "canonical_nll": 1.5933200518290203, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.6666666666666666, "prompt_id": "745badf44c39e597", "reading_canonical_log_prob": -4.7799601554870605, "reading_distances": [24, 26, 19, 20], "reading_log_probs": [-8.782699584960938, -5.475607395172119, -4.968521595001221, -5.5124993324279785], "reading_logprob_std": 1.749605455169521, "tokenization_path_logprob_std": 7.845056116080439, "trajectory_balance_rmse": 41.904023480091205} {"canonical_length": 272, "canonical_nll": 1.3339755114387064, "generation_rouge_l_f1": 0.2552380952380952, "generation_token_f1": 0.5142857142857143, "prompt_id": "8f529e5554e9c902", "reading_canonical_log_prob": -362.8413391113281, "reading_distances": [21, 22, 18, 19], "reading_log_probs": [-356.90032958984375, -356.564453125, -360.80767822265625, -356.1621398925781], "reading_logprob_std": 2.153929585465316, "tokenization_path_logprob_std": 46.94851311044587, "trajectory_balance_rmse": 5183.567007089903} {"canonical_length": 324, "canonical_nll": 0.9672510594497492, "generation_rouge_l_f1": 0.20626151012891344, "generation_token_f1": 0.45672191528545125, "prompt_id": "642f20171ea337c8", "reading_canonical_log_prob": -313.38934326171875, "reading_distances": [26, 21, 25, 27], "reading_log_probs": [-313.82879638671875, -314.4273681640625, -312.168701171875, -310.7437744140625], "reading_logprob_std": 1.6666438488900848, "tokenization_path_logprob_std": 248.14248509952225, "trajectory_balance_rmse": 4853.996621929493} {"canonical_length": 14, "canonical_nll": 2.017937115260533, "generation_rouge_l_f1": 0.4210526315789474, "generation_token_f1": 0.4210526315789474, "prompt_id": "c076cc444fb03f81", "reading_canonical_log_prob": -28.25111961364746, "reading_distances": [196, 181, 192, 186], "reading_log_probs": [-29.030231475830078, -27.50670051574707, -28.528343200683594, -27.365917205810547], "reading_logprob_std": 0.8040412679111064, "tokenization_path_logprob_std": 38.95483445243701, "trajectory_balance_rmse": 400.9417446473649} {"canonical_length": 183, "canonical_nll": 1.2003448986616292, "generation_rouge_l_f1": 0.21621621621621623, "generation_token_f1": 0.3027027027027027, "prompt_id": "8082ddc02c38399e", "reading_canonical_log_prob": -219.66311645507812, "reading_distances": [18, 26, 25, 25], "reading_log_probs": [-221.03704833984375, -218.0113067626953, -217.81826782226562, -219.53643798828125], "reading_logprob_std": 1.5023007229581617, "tokenization_path_logprob_std": 164.7368007391342, "trajectory_balance_rmse": 1630.4153543560437} {"canonical_length": 7, "canonical_nll": 0.1486891508102417, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "3af881f54016c0d1", "reading_canonical_log_prob": -1.040824055671692, "reading_distances": [35, 38, 36, 40], "reading_log_probs": [-1.733575463294983, -1.4506644010543823, -1.521649718284607, -1.803463101387024], "reading_logprob_std": 0.16801787940789176, "tokenization_path_logprob_std": 11.34627178517453, "trajectory_balance_rmse": 71.33485159125195} {"canonical_length": 241, "canonical_nll": 0.639475019146298, "generation_rouge_l_f1": 0.31225296442687744, "generation_token_f1": 0.5652173913043478, "prompt_id": "edd1c37dcb8ffb27", "reading_canonical_log_prob": -154.1134796142578, "reading_distances": [14, 14, 15, 16], "reading_log_probs": [-156.55630493164062, -155.94691467285156, -156.74114990234375, -153.41932678222656], "reading_logprob_std": 1.535691349375156, "tokenization_path_logprob_std": 59.552751163452236, "trajectory_balance_rmse": 3540.104507828834} {"canonical_length": 277, "canonical_nll": 1.6497826972162681, "generation_rouge_l_f1": 0.22957198443579765, "generation_token_f1": 0.45136186770428016, "prompt_id": "b7a747a09567587c", "reading_canonical_log_prob": -456.98980712890625, "reading_distances": [36, 29, 37, 27], "reading_log_probs": [-455.4678039550781, -456.9346923828125, -457.0806884765625, -457.35833740234375], "reading_logprob_std": 0.8468152957919419, "tokenization_path_logprob_std": 301.01195205386836, "trajectory_balance_rmse": 5433.545129753768} {"canonical_length": 203, "canonical_nll": 1.5217515165582667, "generation_rouge_l_f1": 0.2241379310344828, "generation_token_f1": 0.5301724137931034, "prompt_id": "85ba85bd1384cec6", "reading_canonical_log_prob": -308.9155578613281, "reading_distances": [19, 22, 16, 16], "reading_log_probs": [-302.5748291015625, -298.84033203125, -300.0984802246094, -303.02618408203125], "reading_logprob_std": 1.9991297604874556, "tokenization_path_logprob_std": 171.16230220023866, "trajectory_balance_rmse": 4531.667769086949} {"canonical_length": 25, "canonical_nll": 0.4987751770019531, "generation_rouge_l_f1": 0.42857142857142855, "generation_token_f1": 0.42857142857142855, "prompt_id": "ff4cb1405a9ba9bb", "reading_canonical_log_prob": -12.469379425048828, "reading_distances": [16, 15, 16, 22], "reading_log_probs": [-13.052169799804688, -13.100275039672852, -13.314033508300781, -13.13054084777832], "reading_logprob_std": 0.11449417583682003, "tokenization_path_logprob_std": 80.44209143797495, "trajectory_balance_rmse": 360.5670452827953} {"canonical_length": 103, "canonical_nll": 1.427320758115898, "generation_rouge_l_f1": 0.19839142091152817, "generation_token_f1": 0.34316353887399464, "prompt_id": "ea60f11b47ae285e", "reading_canonical_log_prob": -147.0140380859375, "reading_distances": [22, 29, 23, 29], "reading_log_probs": [-145.3019561767578, -143.90025329589844, -144.89125061035156, -143.69921875], "reading_logprob_std": 0.7716674483225653, "tokenization_path_logprob_std": 167.0961717588593, "trajectory_balance_rmse": 2473.20618651215} {"canonical_length": 22, "canonical_nll": 0.4856817939064719, "generation_rouge_l_f1": 0.3636363636363637, "generation_token_f1": 0.4242424242424242, "prompt_id": "6245bcac7beebc0c", "reading_canonical_log_prob": -10.684999465942383, "reading_distances": [25, 30, 21, 25], "reading_log_probs": [-10.125629425048828, -11.081083297729492, -9.150382995605469, -9.352815628051758], "reading_logprob_std": 0.8763908294413022, "tokenization_path_logprob_std": 68.71845677725408, "trajectory_balance_rmse": 476.0042646679242} {"canonical_length": 156, "canonical_nll": 1.3187224559294872, "generation_rouge_l_f1": 0.18097447795823665, "generation_token_f1": 0.3897911832946636, "prompt_id": "5016da9f4e182301", "reading_canonical_log_prob": -205.720703125, "reading_distances": [14, 16, 20, 18], "reading_log_probs": [-206.6370086669922, -205.89712524414062, -207.40054321289062, -209.92239379882812], "reading_logprob_std": 1.7499273996985607, "tokenization_path_logprob_std": 96.5371109732105, "trajectory_balance_rmse": 3690.621640905011} {"canonical_length": 103, "canonical_nll": 1.5181232933859223, "generation_rouge_l_f1": 0.18716577540106952, "generation_token_f1": 0.3422459893048128, "prompt_id": "d6b153dd541c60bc", "reading_canonical_log_prob": -156.36669921875, "reading_distances": [30, 33, 31, 34], "reading_log_probs": [-151.86099243164062, -149.76773071289062, -151.71588134765625, -152.2942352294922], "reading_logprob_std": 1.1218849596605287, "tokenization_path_logprob_std": 142.56238985894888, "trajectory_balance_rmse": 1699.6192720843735} {"canonical_length": 50, "canonical_nll": 0.9319099426269531, "generation_rouge_l_f1": 0.14432989690721654, "generation_token_f1": 0.22680412371134018, "prompt_id": "f50b16ccb1ed30ed", "reading_canonical_log_prob": -46.595497131347656, "reading_distances": [14, 17, 13, 16], "reading_log_probs": [-55.44321823120117, -53.5366325378418, -50.87786865234375, -52.795684814453125], "reading_logprob_std": 1.8881862270303507, "tokenization_path_logprob_std": 51.31034656701577, "trajectory_balance_rmse": 942.772271792802} {"canonical_length": 26, "canonical_nll": 0.45885724287766677, "generation_rouge_l_f1": 0.7499999999999999, "generation_token_f1": 0.7499999999999999, "prompt_id": "ac38f5f5447a0903", "reading_canonical_log_prob": -11.930288314819336, "reading_distances": [35, 43, 36, 36], "reading_log_probs": [-10.286096572875977, -10.50002670288086, -11.333211898803711, -10.412653923034668], "reading_logprob_std": 0.47499929204428737, "tokenization_path_logprob_std": 98.15828822061303, "trajectory_balance_rmse": 553.4013787395575} {"canonical_length": 10, "canonical_nll": 1.4544219970703125, "generation_rouge_l_f1": 0.25, "generation_token_f1": 0.25, "prompt_id": "0027f65734731d27", "reading_canonical_log_prob": -14.544219970703125, "reading_distances": [25, 23, 18, 21], "reading_log_probs": [-22.535499572753906, -18.0721435546875, -17.037031173706055, -19.380264282226562], "reading_logprob_std": 2.387180873922762, "tokenization_path_logprob_std": 32.03469567735084, "trajectory_balance_rmse": 260.4477346465763} {"canonical_length": 146, "canonical_nll": 1.3672565825997967, "generation_rouge_l_f1": 0.26011560693641617, "generation_token_f1": 0.44508670520231214, "prompt_id": "488c8f3398b00ac7", "reading_canonical_log_prob": -199.6194610595703, "reading_distances": [15, 14, 19, 12], "reading_log_probs": [-197.57583618164062, -197.7728271484375, -198.51641845703125, -199.02976989746094], "reading_logprob_std": 0.6729157028501435, "tokenization_path_logprob_std": 144.28639275452812, "trajectory_balance_rmse": 2421.4941759843487} {"canonical_length": 4, "canonical_nll": 4.031225204467773, "generation_rouge_l_f1": 0.6, "generation_token_f1": 0.6, "prompt_id": "91adc07e055d938f", "reading_canonical_log_prob": -16.124900817871094, "reading_distances": [31, 31, 34, 30], "reading_log_probs": [-17.693660736083984, -18.126163482666016, -19.290386199951172, -19.52568817138672], "reading_logprob_std": 0.8879924110039769, "tokenization_path_logprob_std": 4.520454022551729, "trajectory_balance_rmse": 27.707354060340528} {"canonical_length": 263, "canonical_nll": 1.2497382218393536, "generation_rouge_l_f1": 0.2840690978886756, "generation_token_f1": 0.491362763915547, "prompt_id": "5f3070e7c5d93f6b", "reading_canonical_log_prob": -328.68115234375, "reading_distances": [19, 21, 16, 20], "reading_log_probs": [-328.32275390625, -328.28759765625, -328.7700500488281, -329.2760009765625], "reading_logprob_std": 0.4632923441812149, "tokenization_path_logprob_std": 188.55024445644437, "trajectory_balance_rmse": 5740.935414333508} {"canonical_length": 17, "canonical_nll": 0.09652092877556295, "generation_rouge_l_f1": 0.2033898305084746, "generation_token_f1": 0.2033898305084746, "prompt_id": "f969cf0286bd838d", "reading_canonical_log_prob": -1.6408557891845703, "reading_distances": [18, 22, 21, 15], "reading_log_probs": [-2.3815081119537354, -1.9306964874267578, -2.2510037422180176, -2.0311663150787354], "reading_logprob_std": 0.20494334665908762, "tokenization_path_logprob_std": 38.63180542981305, "trajectory_balance_rmse": 241.05530545931822} {"canonical_length": 291, "canonical_nll": 0.6188577868274807, "generation_rouge_l_f1": 0.2914572864321608, "generation_token_f1": 0.507537688442211, "prompt_id": "60292acfc5a172b6", "reading_canonical_log_prob": -180.08761596679688, "reading_distances": [25, 36, 33, 23], "reading_log_probs": [-182.9078369140625, -184.13742065429688, -183.49798583984375, -183.47593688964844], "reading_logprob_std": 0.5024783013675937, "tokenization_path_logprob_std": 218.52518110764004, "trajectory_balance_rmse": 4834.537963756653} {"canonical_length": 174, "canonical_nll": 0.8342851660717493, "generation_rouge_l_f1": 0.3289817232375979, "generation_token_f1": 0.5274151436031331, "prompt_id": "2b8f8cb7903f03af", "reading_canonical_log_prob": -145.16561889648438, "reading_distances": [37, 33, 41, 42], "reading_log_probs": [-145.64300537109375, -157.74789428710938, -155.70811462402344, -157.9268798828125], "reading_logprob_std": 5.829836969941496, "tokenization_path_logprob_std": 279.3128072203535, "trajectory_balance_rmse": 3329.632862187353} {"canonical_length": 71, "canonical_nll": 1.155150614993673, "generation_rouge_l_f1": 0.19801980198019803, "generation_token_f1": 0.297029702970297, "prompt_id": "2a8ed34f73d4ba31", "reading_canonical_log_prob": -82.01569366455078, "reading_distances": [19, 14, 21, 19], "reading_log_probs": [-82.21318817138672, -81.40534973144531, -81.22270202636719, -80.8830337524414], "reading_logprob_std": 0.5645444497680698, "tokenization_path_logprob_std": 24.462158667298464, "trajectory_balance_rmse": 1110.5122294655368} {"canonical_length": 370, "canonical_nll": 0.7115692963471284, "generation_rouge_l_f1": 0.3968503937007874, "generation_token_f1": 0.5795275590551181, "prompt_id": "a2f2bd1ddbd7ac6c", "reading_canonical_log_prob": -263.2806396484375, "reading_distances": [22, 19, 21, 22], "reading_log_probs": [-263.154541015625, -260.7436218261719, -263.4091491699219, -262.279541015625], "reading_logprob_std": 1.2035738451715043, "tokenization_path_logprob_std": 186.72175327455452, "trajectory_balance_rmse": 6789.490929987274} {"canonical_length": 292, "canonical_nll": 1.4654266148397368, "generation_rouge_l_f1": 0.18809980806142035, "generation_token_f1": 0.3877159309021113, "prompt_id": "439bf3be7d027444", "reading_canonical_log_prob": -427.9045715332031, "reading_distances": [22, 23, 22, 22], "reading_log_probs": [-418.3414611816406, -417.970703125, -420.36236572265625, -418.2410888671875], "reading_logprob_std": 1.1001729117932515, "tokenization_path_logprob_std": 158.14842687840857, "trajectory_balance_rmse": 5389.074860425455} {"canonical_length": 152, "canonical_nll": 1.1388329957660877, "generation_rouge_l_f1": 0.20338983050847456, "generation_token_f1": 0.2994350282485876, "prompt_id": "53583cdc2f7841b7", "reading_canonical_log_prob": -173.1026153564453, "reading_distances": [23, 21, 25, 25], "reading_log_probs": [-162.65133666992188, -165.5860137939453, -157.03778076171875, -159.22579956054688], "reading_logprob_std": 3.76575228845991, "tokenization_path_logprob_std": 125.73478352432248, "trajectory_balance_rmse": 2469.1628807673146} {"canonical_length": 73, "canonical_nll": 1.1078145275377247, "generation_rouge_l_f1": 0.21951219512195122, "generation_token_f1": 0.2926829268292683, "prompt_id": "11b8b10c9b5e1733", "reading_canonical_log_prob": -80.8704605102539, "reading_distances": [17, 16, 12, 20], "reading_log_probs": [-80.19597625732422, -80.12438201904297, -80.48878479003906, -79.83629608154297], "reading_logprob_std": 0.267979601562866, "tokenization_path_logprob_std": 123.78069444078817, "trajectory_balance_rmse": 1644.0125946129958} {"canonical_length": 17, "canonical_nll": 0.771208594827091, "generation_rouge_l_f1": 0.17886178861788618, "generation_token_f1": 0.17886178861788618, "prompt_id": "d62e67048e61fa64", "reading_canonical_log_prob": -13.110546112060547, "reading_distances": [15, 9, 17, 13], "reading_log_probs": [-18.371685028076172, -15.55967903137207, -19.748931884765625, -17.202131271362305], "reading_logprob_std": 1.7773042511612644, "tokenization_path_logprob_std": 70.24314114055308, "trajectory_balance_rmse": 259.2500702760941} {"canonical_length": 17, "canonical_nll": 1.311509973862592, "generation_rouge_l_f1": 0.13333333333333333, "generation_token_f1": 0.13333333333333333, "prompt_id": "15d2843edae95663", "reading_canonical_log_prob": -22.295669555664062, "reading_distances": [21, 22, 23, 26], "reading_log_probs": [-23.849464416503906, -24.72140884399414, -23.20176124572754, -23.85357666015625], "reading_logprob_std": 0.6236417407268678, "tokenization_path_logprob_std": 28.428757189212998, "trajectory_balance_rmse": 325.3222039264885} {"canonical_length": 243, "canonical_nll": 1.4009715892650463, "generation_rouge_l_f1": 0.23721881390593044, "generation_token_f1": 0.46625766871165647, "prompt_id": "960951e022fd8e15", "reading_canonical_log_prob": -340.43609619140625, "reading_distances": [33, 34, 33, 30], "reading_log_probs": [-347.864990234375, -345.257080078125, -345.08544921875, -346.5975341796875], "reading_logprob_std": 1.2989150527317164, "tokenization_path_logprob_std": 241.1285664092128, "trajectory_balance_rmse": 5620.34682196793} {"canonical_length": 192, "canonical_nll": 1.2836150328318279, "generation_rouge_l_f1": 0.27250608272506077, "generation_token_f1": 0.559610705596107, "prompt_id": "e17ce1bfa59fa95a", "reading_canonical_log_prob": -246.45408630371094, "reading_distances": [18, 25, 15, 25], "reading_log_probs": [-246.15773010253906, -247.045654296875, -246.43960571289062, -245.98397827148438], "reading_logprob_std": 0.46548363449518404, "tokenization_path_logprob_std": 167.68487469634314, "trajectory_balance_rmse": 4077.2383173523663} {"canonical_length": 37, "canonical_nll": 1.2289584391825907, "generation_rouge_l_f1": 0.375, "generation_token_f1": 0.4166666666666667, "prompt_id": "305cbe322d895404", "reading_canonical_log_prob": -45.47146224975586, "reading_distances": [13, 24, 14, 13], "reading_log_probs": [-44.24565124511719, -43.482269287109375, -46.016075134277344, -44.78984832763672], "reading_logprob_std": 1.0664141794442348, "tokenization_path_logprob_std": 30.093353838659233, "trajectory_balance_rmse": 525.4119982936194} {"canonical_length": 25, "canonical_nll": 0.9751921081542969, "generation_rouge_l_f1": 0.13688212927756654, "generation_token_f1": 0.1520912547528517, "prompt_id": "fee5a1b64fc319ac", "reading_canonical_log_prob": -24.379802703857422, "reading_distances": [27, 26, 24, 24], "reading_log_probs": [-26.803300857543945, -25.85204315185547, -26.33652114868164, -25.141292572021484], "reading_logprob_std": 0.7102524625670497, "tokenization_path_logprob_std": 51.142959975426336, "trajectory_balance_rmse": 561.256851316408} {"canonical_length": 12, "canonical_nll": 1.2308882077534993, "generation_rouge_l_f1": 0.2790697674418604, "generation_token_f1": 0.2790697674418604, "prompt_id": "dce2143d59fcc65f", "reading_canonical_log_prob": -14.770658493041992, "reading_distances": [11, 11, 13, 20], "reading_log_probs": [-16.50457763671875, -15.701789855957031, -18.60441780090332, -19.471092224121094], "reading_logprob_std": 1.7615811935735486, "tokenization_path_logprob_std": 34.572627435291245, "trajectory_balance_rmse": 251.1971499263943} {"canonical_length": 30, "canonical_nll": 0.34720195134480797, "generation_rouge_l_f1": 0.12203389830508475, "generation_token_f1": 0.14915254237288136, "prompt_id": "bc27bf2b83eeb772", "reading_canonical_log_prob": -10.416058540344238, "reading_distances": [35, 42, 39, 37], "reading_log_probs": [-10.8809814453125, -10.667499542236328, -10.698580741882324, -11.395782470703125], "reading_logprob_std": 0.3368131221259062, "tokenization_path_logprob_std": 35.32866434374853, "trajectory_balance_rmse": 367.3244499769562} {"canonical_length": 171, "canonical_nll": 2.107825563665022, "generation_rouge_l_f1": 0.25123152709359603, "generation_token_f1": 0.42857142857142855, "prompt_id": "8e8e4001993842cc", "reading_canonical_log_prob": -360.43817138671875, "reading_distances": [23, 25, 30, 23], "reading_log_probs": [-353.87506103515625, -354.17144775390625, -353.1154479980469, -353.8221740722656], "reading_logprob_std": 0.4476087779415764, "tokenization_path_logprob_std": 178.12114458611748, "trajectory_balance_rmse": 2653.7832972216197} {"canonical_length": 93, "canonical_nll": 1.8188527425130208, "generation_rouge_l_f1": 0.2153846153846154, "generation_token_f1": 0.3692307692307692, "prompt_id": "ae4b947d6478d1ef", "reading_canonical_log_prob": -169.15330505371094, "reading_distances": [21, 24, 18, 27], "reading_log_probs": [-164.0448760986328, -165.95127868652344, -163.18402099609375, -165.1679229736328], "reading_logprob_std": 1.2194228475244244, "tokenization_path_logprob_std": 38.44012623010505, "trajectory_balance_rmse": 1874.0001235267018} {"canonical_length": 153, "canonical_nll": 2.064385906543607, "generation_rouge_l_f1": 0.3179916317991632, "generation_token_f1": 0.4435146443514644, "prompt_id": "e8d78f3e32f3fc0a", "reading_canonical_log_prob": -315.8510437011719, "reading_distances": [18, 20, 23, 24], "reading_log_probs": [-314.2413330078125, -315.5502624511719, -316.7254638671875, -312.2743835449219], "reading_logprob_std": 1.9078279450646893, "tokenization_path_logprob_std": 111.59312316679669, "trajectory_balance_rmse": 3913.1642780307766} {"canonical_length": 14, "canonical_nll": 1.368807247706822, "generation_rouge_l_f1": 0.052830188679245285, "generation_token_f1": 0.052830188679245285, "prompt_id": "0652d72e6ba6a111", "reading_canonical_log_prob": -19.163301467895508, "reading_distances": [35, 27, 25, 28], "reading_log_probs": [-19.317676544189453, -19.072999954223633, -19.849533081054688, -20.236970901489258], "reading_logprob_std": 0.5240690359019494, "tokenization_path_logprob_std": 27.208607717225032, "trajectory_balance_rmse": 187.83234811373447} {"canonical_length": 35, "canonical_nll": 1.583516148158482, "generation_rouge_l_f1": 0.047619047619047616, "generation_token_f1": 0.047619047619047616, "prompt_id": "abb413900ce0e656", "reading_canonical_log_prob": -55.423065185546875, "reading_distances": [28, 24, 23, 27], "reading_log_probs": [-53.286041259765625, -53.40044403076172, -53.250308990478516, -54.335838317871094], "reading_logprob_std": 0.5157773724836752, "tokenization_path_logprob_std": 63.982702395088936, "trajectory_balance_rmse": 853.000730522139} {"canonical_length": 294, "canonical_nll": 0.8908803511639031, "generation_rouge_l_f1": 0.20279720279720279, "generation_token_f1": 0.4440559440559441, "prompt_id": "7607920df7068b2c", "reading_canonical_log_prob": -261.9188232421875, "reading_distances": [14, 8, 13, 9], "reading_log_probs": [-257.2626953125, -260.0899658203125, -257.3342590332031, -259.29107666015625], "reading_logprob_std": 1.4193363999265438, "tokenization_path_logprob_std": 173.4639278456436, "trajectory_balance_rmse": 5303.466567333457} {"canonical_length": 50, "canonical_nll": 2.67603515625, "generation_rouge_l_f1": 0.24242424242424243, "generation_token_f1": 0.33333333333333337, "prompt_id": "6e3d23fca9ec801f", "reading_canonical_log_prob": -133.8017578125, "reading_distances": [28, 32, 27, 24], "reading_log_probs": [-132.17459106445312, -128.4056396484375, -130.58322143554688, -130.53610229492188], "reading_logprob_std": 1.5466375157364318, "tokenization_path_logprob_std": 94.42971032055625, "trajectory_balance_rmse": 1237.5135704550025} {"canonical_length": 18, "canonical_nll": 1.9330482482910156, "generation_rouge_l_f1": 0.5384615384615384, "generation_token_f1": 0.5384615384615384, "prompt_id": "4afc9f7951fa42e2", "reading_canonical_log_prob": -34.79486846923828, "reading_distances": [18, 20, 24, 21], "reading_log_probs": [-36.49995040893555, -35.033966064453125, -35.35859680175781, -34.88606262207031], "reading_logprob_std": 0.7306949339682599, "tokenization_path_logprob_std": 44.461522301796954, "trajectory_balance_rmse": 390.64450250594876} {"canonical_length": 501, "canonical_nll": 0.6082255045572916, "generation_rouge_l_f1": 0.3197969543147208, "generation_token_f1": 0.47969543147208127, "prompt_id": "031d7ff67f65087b", "reading_canonical_log_prob": -304.7209777832031, "reading_distances": [34, 27, 22, 24], "reading_log_probs": [-303.9725036621094, -302.5553283691406, -301.9700622558594, -304.5537109375], "reading_logprob_std": 1.2030263614614405, "tokenization_path_logprob_std": 176.67600991723478, "trajectory_balance_rmse": 5874.428323100427} {"canonical_length": 128, "canonical_nll": 1.0672527551651, "generation_rouge_l_f1": 0.22598870056497172, "generation_token_f1": 0.4463276836158192, "prompt_id": "81ff4f6d7865cbd6", "reading_canonical_log_prob": -136.6083526611328, "reading_distances": [20, 19, 18, 19], "reading_log_probs": [-145.0458984375, -141.67129516601562, -142.3811492919922, -141.73141479492188], "reading_logprob_std": 1.5917573944109857, "tokenization_path_logprob_std": 104.03313327484855, "trajectory_balance_rmse": 2185.899794359385} {"canonical_length": 112, "canonical_nll": 1.3767035348074776, "generation_rouge_l_f1": 0.20833333333333334, "generation_token_f1": 0.32291666666666663, "prompt_id": "fa3e1b2e7d03f031", "reading_canonical_log_prob": -154.1907958984375, "reading_distances": [26, 29, 30, 28], "reading_log_probs": [-156.82485961914062, -157.97512817382812, -158.40957641601562, -155.08514404296875], "reading_logprob_std": 1.4847463171334627, "tokenization_path_logprob_std": 164.8080458568391, "trajectory_balance_rmse": 2600.7310762480734} {"canonical_length": 136, "canonical_nll": 1.2042905022116268, "generation_rouge_l_f1": 0.23737373737373738, "generation_token_f1": 0.45959595959595967, "prompt_id": "c4484ce7220f53d7", "reading_canonical_log_prob": -163.78350830078125, "reading_distances": [19, 15, 13, 15], "reading_log_probs": [-162.7455596923828, -163.06903076171875, -166.50115966796875, -169.64089965820312], "reading_logprob_std": 3.2478430635003726, "tokenization_path_logprob_std": 95.72942221217072, "trajectory_balance_rmse": 2679.155614646568} {"canonical_length": 5, "canonical_nll": 0.020591533184051512, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "e265516b7f250f1e", "reading_canonical_log_prob": -0.10295766592025757, "reading_distances": [36, 36, 40, 41], "reading_log_probs": [-0.594264805316925, -0.5048083066940308, -0.73633873462677, -1.0964372158050537], "reading_logprob_std": 0.26039499126205484, "tokenization_path_logprob_std": 38.63842634367148, "trajectory_balance_rmse": 113.79981281994878} {"canonical_length": 20, "canonical_nll": 0.4802659511566162, "generation_rouge_l_f1": 0.4383561643835616, "generation_token_f1": 0.4383561643835616, "prompt_id": "d3f0dad3a37f19c3", "reading_canonical_log_prob": -9.605319023132324, "reading_distances": [16, 14, 16, 17], "reading_log_probs": [-9.169344902038574, -9.221893310546875, -8.181110382080078, -9.212918281555176], "reading_logprob_std": 0.5106535549200153, "tokenization_path_logprob_std": 43.53692326987888, "trajectory_balance_rmse": 489.1180927841053} {"canonical_length": 23, "canonical_nll": 1.5427508146866509, "generation_rouge_l_f1": 0.23529411764705885, "generation_token_f1": 0.29411764705882354, "prompt_id": "448ac2a15956c368", "reading_canonical_log_prob": -35.48326873779297, "reading_distances": [29, 35, 41, 31], "reading_log_probs": [-36.937164306640625, -37.15310287475586, -36.804603576660156, -36.730648040771484], "reading_logprob_std": 0.18534653337727736, "tokenization_path_logprob_std": 26.188326161417763, "trajectory_balance_rmse": 422.5050698754597} {"canonical_length": 165, "canonical_nll": 1.4529784231474905, "generation_rouge_l_f1": 0.3113207547169811, "generation_token_f1": 0.4622641509433961, "prompt_id": "e00cf91341eb5691", "reading_canonical_log_prob": -239.74143981933594, "reading_distances": [14, 14, 9, 17], "reading_log_probs": [-238.36624145507812, -238.24026489257812, -237.85458374023438, -238.7657012939453], "reading_logprob_std": 0.3755219072794278, "tokenization_path_logprob_std": 95.22494001500998, "trajectory_balance_rmse": 3831.385357616238} {"canonical_length": 224, "canonical_nll": 1.0903491292681013, "generation_rouge_l_f1": 0.30962343096234307, "generation_token_f1": 0.5690376569037656, "prompt_id": "bf8645b4ff7492c2", "reading_canonical_log_prob": -244.2382049560547, "reading_distances": [52, 55, 60, 48], "reading_log_probs": [-245.08837890625, -247.07781982421875, -245.47817993164062, -242.59652709960938], "reading_logprob_std": 1.8543619914870106, "tokenization_path_logprob_std": 156.20911091592484, "trajectory_balance_rmse": 4590.732821362232} {"canonical_length": 261, "canonical_nll": 1.3070147868773947, "generation_rouge_l_f1": 0.22878228782287824, "generation_token_f1": 0.45018450184501846, "prompt_id": "93af13c34520a089", "reading_canonical_log_prob": -341.130859375, "reading_distances": [25, 28, 26, 22], "reading_log_probs": [-335.287109375, -335.1527099609375, -335.8579406738281, -336.19146728515625], "reading_logprob_std": 0.48728687147738975, "tokenization_path_logprob_std": 313.8366695375035, "trajectory_balance_rmse": 5831.741630128611} {"canonical_length": 107, "canonical_nll": 1.171872718311916, "generation_rouge_l_f1": 0.2589531680440771, "generation_token_f1": 0.39669421487603307, "prompt_id": "8f7942ab9e010ef0", "reading_canonical_log_prob": -125.390380859375, "reading_distances": [34, 31, 34, 35], "reading_log_probs": [-126.64947509765625, -126.02047729492188, -125.69430541992188, -125.80816650390625], "reading_logprob_std": 0.4262448249435858, "tokenization_path_logprob_std": 113.4382411561428, "trajectory_balance_rmse": 1828.0393032624347} {"canonical_length": 39, "canonical_nll": 2.3091845390124197, "generation_rouge_l_f1": 0.13333333333333333, "generation_token_f1": 0.17777777777777778, "prompt_id": "ba9caab6c5bf7b20", "reading_canonical_log_prob": -90.05819702148438, "reading_distances": [38, 37, 40, 36], "reading_log_probs": [-110.8779296875, -103.74803161621094, -110.67937469482422, -110.20503234863281], "reading_logprob_std": 3.4313388715153854, "tokenization_path_logprob_std": 31.82614161607104, "trajectory_balance_rmse": 742.2760544350881} {"canonical_length": 292, "canonical_nll": 1.4124218666390196, "generation_rouge_l_f1": 0.15495495495495495, "generation_token_f1": 0.27747747747747753, "prompt_id": "90e4d685d3873c08", "reading_canonical_log_prob": -412.42718505859375, "reading_distances": [30, 30, 22, 20], "reading_log_probs": [-412.3277587890625, -411.4716796875, -411.6734924316406, -411.9198913574219], "reading_logprob_std": 0.3685134836697685, "tokenization_path_logprob_std": 303.0814069338592, "trajectory_balance_rmse": 6600.270950993145} {"canonical_length": 15, "canonical_nll": 1.8004149119059245, "generation_rouge_l_f1": 0.0980392156862745, "generation_token_f1": 0.0980392156862745, "prompt_id": "ec455005996bdab4", "reading_canonical_log_prob": -27.006223678588867, "reading_distances": [42, 34, 48, 44], "reading_log_probs": [-27.363750457763672, -26.743694305419922, -27.822935104370117, -27.07596778869629], "reading_logprob_std": 0.4574627103970879, "tokenization_path_logprob_std": 69.39878988836114, "trajectory_balance_rmse": 277.0483507701827} {"canonical_length": 243, "canonical_nll": 1.711204748585391, "generation_rouge_l_f1": 0.16289592760180993, "generation_token_f1": 0.3212669683257919, "prompt_id": "ab90737e4575d4fd", "reading_canonical_log_prob": -415.82275390625, "reading_distances": [11, 17, 14, 17], "reading_log_probs": [-417.6535949707031, -419.82000732421875, -420.46368408203125, -420.172607421875], "reading_logprob_std": 1.2766733870927947, "tokenization_path_logprob_std": 262.6244061655627, "trajectory_balance_rmse": 4981.5242577080035} {"canonical_length": 240, "canonical_nll": 1.3771991729736328, "generation_rouge_l_f1": 0.19639278557114226, "generation_token_f1": 0.43687374749499, "prompt_id": "ec5a98de132967a0", "reading_canonical_log_prob": -330.5278015136719, "reading_distances": [33, 32, 29, 37], "reading_log_probs": [-333.9359130859375, -327.74273681640625, -330.17449951171875, -332.9500427246094], "reading_logprob_std": 2.801915260223669, "tokenization_path_logprob_std": 219.9080579301288, "trajectory_balance_rmse": 5143.225980422216} {"canonical_length": 37, "canonical_nll": 1.1346711854676943, "generation_rouge_l_f1": 0.06060606060606061, "generation_token_f1": 0.06060606060606061, "prompt_id": "71251a192c44eeb3", "reading_canonical_log_prob": -41.98283386230469, "reading_distances": [22, 24, 25, 22], "reading_log_probs": [-45.079811096191406, -46.99972152709961, -45.115013122558594, -44.61248779296875], "reading_logprob_std": 1.0570883238683157, "tokenization_path_logprob_std": 90.03957464343486, "trajectory_balance_rmse": 677.6293179154158} {"canonical_length": 221, "canonical_nll": 1.798016578363617, "generation_rouge_l_f1": 0.2800875273522976, "generation_token_f1": 0.4638949671772429, "prompt_id": "646766dc50567beb", "reading_canonical_log_prob": -397.3616638183594, "reading_distances": [44, 41, 39, 54], "reading_log_probs": [-395.8531799316406, -391.9460144042969, -394.7645263671875, -395.06878662109375], "reading_logprob_std": 1.7042758044634305, "tokenization_path_logprob_std": 185.396430366448, "trajectory_balance_rmse": 3990.0677223871653} {"canonical_length": 87, "canonical_nll": 1.1655925882273708, "generation_rouge_l_f1": 0.24175824175824173, "generation_token_f1": 0.36813186813186816, "prompt_id": "3175f6135d7796a6", "reading_canonical_log_prob": -101.40655517578125, "reading_distances": [40, 40, 42, 47], "reading_log_probs": [-98.45283508300781, -98.58509063720703, -98.75521850585938, -100.05352020263672], "reading_logprob_std": 0.7383504302615231, "tokenization_path_logprob_std": 35.81363186764298, "trajectory_balance_rmse": 1447.406386514707} {"canonical_length": 315, "canonical_nll": 1.0397406684027777, "generation_rouge_l_f1": 0.2149046793760832, "generation_token_f1": 0.4506065857885615, "prompt_id": "364270b26598f5f6", "reading_canonical_log_prob": -327.518310546875, "reading_distances": [25, 17, 25, 26], "reading_log_probs": [-328.94219970703125, -327.32269287109375, -328.42340087890625, -328.1016540527344], "reading_logprob_std": 0.678253349530169, "tokenization_path_logprob_std": 220.8437581369321, "trajectory_balance_rmse": 5327.8344505582945} {"canonical_length": 132, "canonical_nll": 1.7140123771898674, "generation_rouge_l_f1": 0.17690417690417692, "generation_token_f1": 0.33415233415233414, "prompt_id": "3b49b131c410b62c", "reading_canonical_log_prob": -226.2496337890625, "reading_distances": [18, 13, 10, 23], "reading_log_probs": [-219.26744079589844, -222.80430603027344, -223.58895874023438, -221.15054321289062], "reading_logprob_std": 1.9154522824829112, "tokenization_path_logprob_std": 233.76308030982415, "trajectory_balance_rmse": 2560.890649870239} {"canonical_length": 9, "canonical_nll": 0.48560497495863175, "generation_rouge_l_f1": 0.8571428571428571, "generation_token_f1": 0.8571428571428571, "prompt_id": "300deaab60201ff0", "reading_canonical_log_prob": -4.3704447746276855, "reading_distances": [31, 19, 23, 20], "reading_log_probs": [-4.104403495788574, -4.2986040115356445, -4.160574913024902, -4.40388822555542], "reading_logprob_std": 0.13536935202324224, "tokenization_path_logprob_std": 68.2507787364319, "trajectory_balance_rmse": 185.54460713031582} {"canonical_length": 239, "canonical_nll": 1.4076402736009415, "generation_rouge_l_f1": 0.3010309278350516, "generation_token_f1": 0.5814432989690721, "prompt_id": "166e7bed707f11c1", "reading_canonical_log_prob": -336.426025390625, "reading_distances": [21, 26, 23, 26], "reading_log_probs": [-328.9208984375, -328.61407470703125, -330.1609191894531, -329.0933837890625], "reading_logprob_std": 0.6722871623148724, "tokenization_path_logprob_std": 292.4057446902786, "trajectory_balance_rmse": 5423.4366812460075} {"canonical_length": 278, "canonical_nll": 1.2972008135678956, "generation_rouge_l_f1": 0.24299065420560748, "generation_token_f1": 0.5233644859813084, "prompt_id": "af085ada31a636ce", "reading_canonical_log_prob": -360.621826171875, "reading_distances": [29, 31, 32, 32], "reading_log_probs": [-358.36334228515625, -358.79193115234375, -358.14483642578125, -360.13995361328125], "reading_logprob_std": 0.894621417652706, "tokenization_path_logprob_std": 77.65925223356798, "trajectory_balance_rmse": 5432.754847893278} {"canonical_length": 157, "canonical_nll": 1.5428710743120522, "generation_rouge_l_f1": 0.21052631578947367, "generation_token_f1": 0.4066985645933014, "prompt_id": "f0f7445e4deb0e6e", "reading_canonical_log_prob": -242.2307586669922, "reading_distances": [28, 29, 13, 35], "reading_log_probs": [-234.82244873046875, -236.25399780273438, -236.72564697265625, -239.2828369140625], "reading_logprob_std": 1.8597064898782, "tokenization_path_logprob_std": 151.84823584310828, "trajectory_balance_rmse": 3875.747737132316} {"canonical_length": 20, "canonical_nll": 0.7096134662628174, "generation_rouge_l_f1": 0.7428571428571428, "generation_token_f1": 0.9142857142857143, "prompt_id": "b5d6efcaf5d67694", "reading_canonical_log_prob": -14.192269325256348, "reading_distances": [41, 35, 40, 28], "reading_log_probs": [-18.349180221557617, -19.45342254638672, -21.02759552001953, -19.433725357055664], "reading_logprob_std": 1.102585159238704, "tokenization_path_logprob_std": 64.43296377975776, "trajectory_balance_rmse": 470.1450655733296} {"canonical_length": 91, "canonical_nll": 1.8951546805245536, "generation_rouge_l_f1": 0.25609756097560976, "generation_token_f1": 0.36585365853658536, "prompt_id": "b7652a39606177d7", "reading_canonical_log_prob": -172.45907592773438, "reading_distances": [24, 13, 22, 24], "reading_log_probs": [-172.7947540283203, -172.04214477539062, -171.18441772460938, -170.79931640625], "reading_logprob_std": 0.8930283719085295, "tokenization_path_logprob_std": 150.7892720612662, "trajectory_balance_rmse": 1888.399422716901} {"canonical_length": 18, "canonical_nll": 0.6760937372843424, "generation_rouge_l_f1": 0.08396946564885495, "generation_token_f1": 0.0916030534351145, "prompt_id": "f3723ab195d4cd31", "reading_canonical_log_prob": -12.169687271118164, "reading_distances": [25, 26, 22, 20], "reading_log_probs": [-10.172006607055664, -11.177799224853516, -10.633256912231445, -11.11070442199707], "reading_logprob_std": 0.4685534822910393, "tokenization_path_logprob_std": 74.42320249474147, "trajectory_balance_rmse": 377.37811322723013} {"canonical_length": 7, "canonical_nll": 0.9375054495675224, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "ff3fea8e176d412d", "reading_canonical_log_prob": -6.562538146972656, "reading_distances": [15, 23, 26, 19], "reading_log_probs": [-6.680807113647461, -5.4517822265625, -5.356389045715332, -7.0118818283081055], "reading_logprob_std": 0.8444851902176163, "tokenization_path_logprob_std": 31.266564158822305, "trajectory_balance_rmse": 134.1563469068038} {"canonical_length": 69, "canonical_nll": 1.2710395481275476, "generation_rouge_l_f1": 0.4897959183673469, "generation_token_f1": 0.5714285714285714, "prompt_id": "f5b5083636173713", "reading_canonical_log_prob": -87.70172882080078, "reading_distances": [16, 25, 19, 24], "reading_log_probs": [-92.37754821777344, -95.5108413696289, -94.44530487060547, -91.71171569824219], "reading_logprob_std": 1.769597650970961, "tokenization_path_logprob_std": 116.51369590685492, "trajectory_balance_rmse": 1079.4934638764291} {"canonical_length": 291, "canonical_nll": 0.9859747083736039, "generation_rouge_l_f1": 0.2365988909426987, "generation_token_f1": 0.44731977818853974, "prompt_id": "640dfdad1fff517f", "reading_canonical_log_prob": -286.91864013671875, "reading_distances": [22, 25, 24, 22], "reading_log_probs": [-283.4399719238281, -284.49957275390625, -281.4156494140625, -281.7461853027344], "reading_logprob_std": 1.4517382757820054, "tokenization_path_logprob_std": 344.90395197875125, "trajectory_balance_rmse": 6225.5981846110435} {"canonical_length": 17, "canonical_nll": 2.030874588910271, "generation_rouge_l_f1": 0.17391304347826086, "generation_token_f1": 0.17391304347826086, "prompt_id": "f39da6ed7768e21d", "reading_canonical_log_prob": -34.52486801147461, "reading_distances": [34, 32, 27, 33], "reading_log_probs": [-35.750789642333984, -36.4781379699707, -36.178619384765625, -35.880958557128906], "reading_logprob_std": 0.32454512101794675, "tokenization_path_logprob_std": 34.959937733152174, "trajectory_balance_rmse": 385.3781012353013} {"canonical_length": 16, "canonical_nll": 1.8595256805419922, "generation_rouge_l_f1": 0.19354838709677422, "generation_token_f1": 0.2580645161290323, "prompt_id": "968e82be4a52a789", "reading_canonical_log_prob": -29.752410888671875, "reading_distances": [34, 29, 30, 27], "reading_log_probs": [-29.086626052856445, -30.47357940673828, -30.877464294433594, -30.388593673706055], "reading_logprob_std": 0.7764878707046989, "tokenization_path_logprob_std": 40.345509123242266, "trajectory_balance_rmse": 415.4887545892089} {"canonical_length": 15, "canonical_nll": 2.0043294270833334, "generation_rouge_l_f1": 0.17391304347826086, "generation_token_f1": 0.2608695652173913, "prompt_id": "e36a9df5c4b484e7", "reading_canonical_log_prob": -30.06494140625, "reading_distances": [23, 25, 23, 26], "reading_log_probs": [-29.83022117614746, -29.355701446533203, -29.327768325805664, -31.793930053710938], "reading_logprob_std": 1.167671224585841, "tokenization_path_logprob_std": 82.41656148196513, "trajectory_balance_rmse": 332.77784130894435} {"canonical_length": 243, "canonical_nll": 1.164171760464892, "generation_rouge_l_f1": 0.25482625482625487, "generation_token_f1": 0.5212355212355213, "prompt_id": "87e83988cc685281", "reading_canonical_log_prob": -282.89373779296875, "reading_distances": [28, 33, 37, 39], "reading_log_probs": [-285.79150390625, -286.52056884765625, -284.86456298828125, -285.8450927734375], "reading_logprob_std": 0.6802994998273282, "tokenization_path_logprob_std": 207.81458607150066, "trajectory_balance_rmse": 5200.956889664365} {"canonical_length": 106, "canonical_nll": 0.7702841128943101, "generation_rouge_l_f1": 0.1871345029239766, "generation_token_f1": 0.34502923976608185, "prompt_id": "753de9bc3063e720", "reading_canonical_log_prob": -81.65011596679688, "reading_distances": [21, 16, 16, 21], "reading_log_probs": [-79.1396713256836, -79.23432922363281, -81.06864929199219, -77.89938354492188], "reading_logprob_std": 1.3057336808956337, "tokenization_path_logprob_std": 92.70107393145938, "trajectory_balance_rmse": 2373.002229037035} {"canonical_length": 18, "canonical_nll": 1.8580053117540147, "generation_rouge_l_f1": 0.21052631578947367, "generation_token_f1": 0.21052631578947367, "prompt_id": "52785eaaf4af8786", "reading_canonical_log_prob": -33.444095611572266, "reading_distances": [20, 24, 24, 27], "reading_log_probs": [-30.448745727539062, -27.599597930908203, -29.629037857055664, -34.723594665527344], "reading_logprob_std": 2.9984490791914573, "tokenization_path_logprob_std": 67.2844746814731, "trajectory_balance_rmse": 371.28184204063643} {"canonical_length": 293, "canonical_nll": 1.7264977152437073, "generation_rouge_l_f1": 0.21946902654867256, "generation_token_f1": 0.40707964601769914, "prompt_id": "9296c91d93887868", "reading_canonical_log_prob": -505.86383056640625, "reading_distances": [19, 19, 23, 17], "reading_log_probs": [-501.1683654785156, -503.56243896484375, -500.09881591796875, -499.17767333984375], "reading_logprob_std": 1.8909816048293409, "tokenization_path_logprob_std": 207.08705517056654, "trajectory_balance_rmse": 5968.643076070966} {"canonical_length": 130, "canonical_nll": 1.7186668982872597, "generation_rouge_l_f1": 0.285, "generation_token_f1": 0.41000000000000003, "prompt_id": "dd7d0916ee5ef6b2", "reading_canonical_log_prob": -223.42669677734375, "reading_distances": [54, 60, 62, 57], "reading_log_probs": [-214.93374633789062, -214.87387084960938, -217.51516723632812, -214.7408447265625], "reading_logprob_std": 1.3352758111072789, "tokenization_path_logprob_std": 108.06891819778289, "trajectory_balance_rmse": 2680.936886251898} {"canonical_length": 172, "canonical_nll": 1.8507844791855923, "generation_rouge_l_f1": 0.20476190476190476, "generation_token_f1": 0.4238095238095238, "prompt_id": "0d595f7cde9d737e", "reading_canonical_log_prob": -318.3349304199219, "reading_distances": [17, 18, 19, 20], "reading_log_probs": [-314.00445556640625, -317.53863525390625, -312.356689453125, -315.89959716796875], "reading_logprob_std": 2.2525601013601526, "tokenization_path_logprob_std": 263.4166632748651, "trajectory_balance_rmse": 3459.577766723143} {"canonical_length": 154, "canonical_nll": 0.861897258015422, "generation_rouge_l_f1": 0.20705882352941174, "generation_token_f1": 0.4611764705882353, "prompt_id": "8c2391cc4e076841", "reading_canonical_log_prob": -132.732177734375, "reading_distances": [233, 246, 247, 230], "reading_log_probs": [-122.3823013305664, -126.54380798339844, -124.10609436035156, -125.86530303955078], "reading_logprob_std": 1.8690133934706727, "tokenization_path_logprob_std": 122.83289616923678, "trajectory_balance_rmse": 2599.583279821414} {"canonical_length": 395, "canonical_nll": 0.7882066171380538, "generation_rouge_l_f1": 0.33333333333333337, "generation_token_f1": 0.48263888888888884, "prompt_id": "a3d283a16147427b", "reading_canonical_log_prob": -311.34161376953125, "reading_distances": [14, 17, 17, 14], "reading_log_probs": [-304.4168701171875, -301.623046875, -301.7160949707031, -303.4319763183594], "reading_logprob_std": 1.3630475344433741, "tokenization_path_logprob_std": 255.77389405178332, "trajectory_balance_rmse": 5746.876864951876} {"canonical_length": 11, "canonical_nll": 1.154440533031117, "generation_rouge_l_f1": 0.23529411764705885, "generation_token_f1": 0.23529411764705885, "prompt_id": "977c41664864ae25", "reading_canonical_log_prob": -12.698845863342285, "reading_distances": [30, 32, 35, 27], "reading_log_probs": [-13.329696655273438, -12.701028823852539, -12.475648880004883, -12.160825729370117], "reading_logprob_std": 0.49435220277647746, "tokenization_path_logprob_std": 24.55851614144281, "trajectory_balance_rmse": 256.17954590049493} {"canonical_length": 20, "canonical_nll": 0.9121614456176758, "generation_rouge_l_f1": 0.11702127659574467, "generation_token_f1": 0.14893617021276595, "prompt_id": "4c8a56abb260b4fb", "reading_canonical_log_prob": -18.243228912353516, "reading_distances": [28, 29, 25, 28], "reading_log_probs": [-19.864978790283203, -19.92441177368164, -19.943157196044922, -19.627479553222656], "reading_logprob_std": 0.14555140524671667, "tokenization_path_logprob_std": 54.131696048198854, "trajectory_balance_rmse": 239.34118924581148} {"canonical_length": 244, "canonical_nll": 0.7373763537797772, "generation_rouge_l_f1": 0.32806324110671936, "generation_token_f1": 0.5889328063241106, "prompt_id": "7a98c8dec132eec9", "reading_canonical_log_prob": -179.91983032226562, "reading_distances": [32, 35, 31, 33], "reading_log_probs": [-186.6413116455078, -183.29129028320312, -186.2199249267578, -186.89520263671875], "reading_logprob_std": 1.6704728782312916, "tokenization_path_logprob_std": 198.64092667302015, "trajectory_balance_rmse": 4532.754962010436} {"canonical_length": 419, "canonical_nll": 1.2437789297900508, "generation_rouge_l_f1": 0.2096069868995633, "generation_token_f1": 0.37554585152838427, "prompt_id": "7f79088a335cbb01", "reading_canonical_log_prob": -521.1433715820312, "reading_distances": [23, 13, 20, 16], "reading_log_probs": [-502.74700927734375, -508.5025634765625, -503.8047180175781, -506.973876953125], "reading_logprob_std": 2.6857923267643216, "tokenization_path_logprob_std": 73.30847256420195, "trajectory_balance_rmse": 7488.522904966749} {"canonical_length": 12, "canonical_nll": 2.767210006713867, "generation_rouge_l_f1": 0.4615384615384615, "generation_token_f1": 0.4615384615384615, "prompt_id": "48b64f1a7aabf6e4", "reading_canonical_log_prob": -33.206520080566406, "reading_distances": [30, 23, 21, 27], "reading_log_probs": [-27.82781410217285, -29.40135383605957, -30.315139770507812, -28.89704132080078], "reading_logprob_std": 1.0370793195169405, "tokenization_path_logprob_std": 46.6210419299777, "trajectory_balance_rmse": 315.7835418041327} {"canonical_length": 319, "canonical_nll": 1.4971939134747256, "generation_rouge_l_f1": 0.2017094017094017, "generation_token_f1": 0.46837606837606843, "prompt_id": "735c6c6b04d8c3fe", "reading_canonical_log_prob": -477.6048583984375, "reading_distances": [35, 38, 37, 39], "reading_log_probs": [-475.6952209472656, -477.3155517578125, -477.24322509765625, -480.56744384765625], "reading_logprob_std": 2.049203267612688, "tokenization_path_logprob_std": 265.9016040435056, "trajectory_balance_rmse": 6437.518122305668} {"canonical_length": 57, "canonical_nll": 1.2820094593784266, "generation_rouge_l_f1": 0.391304347826087, "generation_token_f1": 0.5217391304347826, "prompt_id": "eab393eed1fb7651", "reading_canonical_log_prob": -73.07453918457031, "reading_distances": [35, 35, 37, 35], "reading_log_probs": [-72.7151107788086, -73.57806396484375, -71.80734252929688, -70.62895202636719], "reading_logprob_std": 1.2630026900823403, "tokenization_path_logprob_std": 53.732827876494234, "trajectory_balance_rmse": 1170.5505385289462} {"canonical_length": 20, "canonical_nll": 0.010464628040790559, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "691dfdc26548dcb9", "reading_canonical_log_prob": -0.20929256081581116, "reading_distances": [30, 31, 33, 31], "reading_log_probs": [-0.6950953006744385, -0.5211320519447327, -0.5489877462387085, -0.4841460585594177], "reading_logprob_std": 0.09240245306677092, "tokenization_path_logprob_std": 37.097378677144455, "trajectory_balance_rmse": 283.3942378653698} {"canonical_length": 230, "canonical_nll": 1.889743705417799, "generation_rouge_l_f1": 0.2510822510822511, "generation_token_f1": 0.4458874458874459, "prompt_id": "2b073d802d00719d", "reading_canonical_log_prob": -434.64105224609375, "reading_distances": [24, 28, 25, 21], "reading_log_probs": [-432.193359375, -429.2762756347656, -431.6474304199219, -429.05169677734375], "reading_logprob_std": 1.6095574875988659, "tokenization_path_logprob_std": 84.72156495814305, "trajectory_balance_rmse": 4147.236567947182} {"canonical_length": 11, "canonical_nll": 4.171330885453657, "generation_rouge_l_f1": 0.048780487804878044, "generation_token_f1": 0.048780487804878044, "prompt_id": "11a956113e5f3fb8", "reading_canonical_log_prob": -45.884639739990234, "reading_distances": [29, 30, 27, 26], "reading_log_probs": [-41.75739288330078, -42.07584762573242, -41.94932174682617, -41.206298828125], "reading_logprob_std": 0.38364064527130054, "tokenization_path_logprob_std": 0.0, "trajectory_balance_rmse": 45.864356994628906} {"canonical_length": 162, "canonical_nll": 1.1622837207935475, "generation_rouge_l_f1": 0.41566265060240964, "generation_token_f1": 0.5240963855421686, "prompt_id": "b48c6b78ff2012e6", "reading_canonical_log_prob": -188.2899627685547, "reading_distances": [30, 20, 32, 22], "reading_log_probs": [-184.46334838867188, -183.6825714111328, -184.6064453125, -185.53663635253906], "reading_logprob_std": 0.7603941760250853, "tokenization_path_logprob_std": 195.6148657267347, "trajectory_balance_rmse": 2903.4032410446166} {"canonical_length": 142, "canonical_nll": 0.8458576471033231, "generation_rouge_l_f1": 0.3444730077120823, "generation_token_f1": 0.493573264781491, "prompt_id": "c057418a3f03722b", "reading_canonical_log_prob": -120.11178588867188, "reading_distances": [34, 36, 34, 30], "reading_log_probs": [-121.26427459716797, -118.49897766113281, -116.86270904541016, -117.93096923828125], "reading_logprob_std": 1.876893698316264, "tokenization_path_logprob_std": 182.9015423470026, "trajectory_balance_rmse": 2680.87624834314} {"canonical_length": 239, "canonical_nll": 1.2798364870717835, "generation_rouge_l_f1": 0.2833675564681725, "generation_token_f1": 0.5503080082135524, "prompt_id": "14f504daadaa6ae8", "reading_canonical_log_prob": -305.88092041015625, "reading_distances": [23, 17, 20, 23], "reading_log_probs": [-304.1044006347656, -307.3035888671875, -304.90142822265625, -304.6549987792969], "reading_logprob_std": 1.4147835324838058, "tokenization_path_logprob_std": 269.60874174801415, "trajectory_balance_rmse": 5650.033940185236} {"canonical_length": 97, "canonical_nll": 1.4918186148417365, "generation_rouge_l_f1": 0.17575757575757575, "generation_token_f1": 0.3393939393939393, "prompt_id": "20abb31273f6c931", "reading_canonical_log_prob": -144.70640563964844, "reading_distances": [20, 23, 30, 29], "reading_log_probs": [-137.89413452148438, -138.1103515625, -138.26531982421875, -140.87925720214844], "reading_logprob_std": 1.4029435714347633, "tokenization_path_logprob_std": 205.4672378501197, "trajectory_balance_rmse": 2294.2536204427006} {"canonical_length": 5, "canonical_nll": 0.6961849212646485, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "609097a0b2afbf5d", "reading_canonical_log_prob": -3.480924606323242, "reading_distances": [26, 26, 23, 33], "reading_log_probs": [-4.186622619628906, -3.433908462524414, -4.0320000648498535, -3.171807289123535], "reading_logprob_std": 0.48189548426523726, "tokenization_path_logprob_std": 33.965551737681494, "trajectory_balance_rmse": 77.76300492911449} {"canonical_length": 70, "canonical_nll": 2.2454068865094867, "generation_rouge_l_f1": 0.07669616519174043, "generation_token_f1": 0.11209439528023599, "prompt_id": "b13cf3764c885d75", "reading_canonical_log_prob": -157.17848205566406, "reading_distances": [30, 22, 25, 35], "reading_log_probs": [-155.51583862304688, -156.41275024414062, -157.729736328125, -155.16111755371094], "reading_logprob_std": 1.1449330288662882, "tokenization_path_logprob_std": 82.2179822117681, "trajectory_balance_rmse": 1462.226675195208} {"canonical_length": 42, "canonical_nll": 0.9043478284563337, "generation_rouge_l_f1": 0.25477707006369427, "generation_token_f1": 0.28025477707006374, "prompt_id": "d289e388118f2de0", "reading_canonical_log_prob": -37.982608795166016, "reading_distances": [21, 22, 26, 20], "reading_log_probs": [-36.435523986816406, -37.77649688720703, -37.934226989746094, -35.747314453125], "reading_logprob_std": 1.0584185510980055, "tokenization_path_logprob_std": 101.72921899479924, "trajectory_balance_rmse": 766.8743224519022} {"canonical_length": 10, "canonical_nll": 2.036913299560547, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "58e5e1069ca06eb2", "reading_canonical_log_prob": -20.36913299560547, "reading_distances": [23, 12, 20, 22], "reading_log_probs": [-24.71276092529297, -25.227771759033203, -20.94965934753418, -21.82760238647461], "reading_logprob_std": 2.1091959562956775, "tokenization_path_logprob_std": 58.07393795404802, "trajectory_balance_rmse": 279.89128961182513} {"canonical_length": 113, "canonical_nll": 1.5592662169870022, "generation_rouge_l_f1": 0.2479784366576819, "generation_token_f1": 0.33962264150943394, "prompt_id": "ec3868510227235c", "reading_canonical_log_prob": -176.19708251953125, "reading_distances": [34, 28, 26, 26], "reading_log_probs": [-178.00341796875, -177.78167724609375, -178.21751403808594, -177.44288635253906], "reading_logprob_std": 0.33090567611060423, "tokenization_path_logprob_std": 135.533254029986, "trajectory_balance_rmse": 2383.7837274795183} {"canonical_length": 16, "canonical_nll": 2.73439621925354, "generation_rouge_l_f1": 0.24, "generation_token_f1": 0.32, "prompt_id": "d625c7ada7581821", "reading_canonical_log_prob": -43.75033950805664, "reading_distances": [11, 10, 11, 10], "reading_log_probs": [-39.761680603027344, -42.93360900878906, -41.97304153442383, -45.26948165893555], "reading_logprob_std": 2.2827729456491057, "tokenization_path_logprob_std": 33.0213219470925, "trajectory_balance_rmse": 398.1414776206241} {"canonical_length": 7, "canonical_nll": 1.9002629688807897, "generation_rouge_l_f1": 0.5714285714285715, "generation_token_f1": 0.5714285714285715, "prompt_id": "34ca0227975643ca", "reading_canonical_log_prob": -13.301840782165527, "reading_distances": [23, 19, 21, 16], "reading_log_probs": [-11.89370059967041, -10.948285102844238, -11.909677505493164, -12.015280723571777], "reading_logprob_std": 0.49856080336955877, "tokenization_path_logprob_std": 18.180188304524723, "trajectory_balance_rmse": 79.50046030326718} {"canonical_length": 92, "canonical_nll": 0.47230480028235394, "generation_rouge_l_f1": 0.40752351097178674, "generation_token_f1": 0.47021943573667707, "prompt_id": "03d8c4a193a46caa", "reading_canonical_log_prob": -43.45204162597656, "reading_distances": [29, 29, 32, 23], "reading_log_probs": [-44.224884033203125, -43.22127914428711, -43.53362274169922, -44.357177734375], "reading_logprob_std": 0.5453312086343133, "tokenization_path_logprob_std": 146.20976873841647, "trajectory_balance_rmse": 1343.0945162088572} {"canonical_length": 22, "canonical_nll": 1.0366160652854226, "generation_rouge_l_f1": 0.3888888888888889, "generation_token_f1": 0.5, "prompt_id": "cfb9b15522adbd99", "reading_canonical_log_prob": -22.805553436279297, "reading_distances": [22, 14, 16, 16], "reading_log_probs": [-22.51516342163086, -22.931427001953125, -22.82822036743164, -23.02426528930664], "reading_logprob_std": 0.22139122605483436, "tokenization_path_logprob_std": 30.161303772065356, "trajectory_balance_rmse": 556.8583396090701} {"canonical_length": 115, "canonical_nll": 1.290162990404212, "generation_rouge_l_f1": 0.2819383259911894, "generation_token_f1": 0.4229074889867841, "prompt_id": "7e152e3b35975991", "reading_canonical_log_prob": -148.36874389648438, "reading_distances": [25, 29, 33, 23], "reading_log_probs": [-150.9091796875, -150.00555419921875, -151.75762939453125, -150.65281677246094], "reading_logprob_std": 0.7252272707730764, "tokenization_path_logprob_std": 91.53364032204253, "trajectory_balance_rmse": 3060.9363814415133} {"canonical_length": 10, "canonical_nll": 3.7861976623535156, "generation_rouge_l_f1": 0.09090909090909091, "generation_token_f1": 0.09090909090909091, "prompt_id": "6df3b4a03782fc55", "reading_canonical_log_prob": -37.861976623535156, "reading_distances": [26, 26, 24, 27], "reading_log_probs": [-30.611812591552734, -31.667814254760742, -33.87354278564453, -32.28583908081055], "reading_logprob_std": 1.3639521054151558, "tokenization_path_logprob_std": 43.727411363748224, "trajectory_balance_rmse": 273.9140051953819} {"canonical_length": 150, "canonical_nll": 1.5881563313802083, "generation_rouge_l_f1": 0.19902912621359223, "generation_token_f1": 0.3932038834951456, "prompt_id": "68f61446f4cfbca0", "reading_canonical_log_prob": -238.22344970703125, "reading_distances": [20, 26, 25, 32], "reading_log_probs": [-238.5535888671875, -237.69900512695312, -237.62767028808594, -238.57723999023438], "reading_logprob_std": 0.5217171738097923, "tokenization_path_logprob_std": 164.5521693485774, "trajectory_balance_rmse": 2884.9793806025655} {"canonical_length": 206, "canonical_nll": 1.3924350182986953, "generation_rouge_l_f1": 0.26106194690265483, "generation_token_f1": 0.41150442477876104, "prompt_id": "6a128630a7f46740", "reading_canonical_log_prob": -286.84161376953125, "reading_distances": [26, 28, 25, 28], "reading_log_probs": [-281.1146240234375, -281.98065185546875, -281.49810791015625, -282.4183349609375], "reading_logprob_std": 0.5677410861433577, "tokenization_path_logprob_std": 137.87014378432366, "trajectory_balance_rmse": 4201.628966900261} {"canonical_length": 324, "canonical_nll": 1.249506820867091, "generation_rouge_l_f1": 0.1946902654867257, "generation_token_f1": 0.37522123893805315, "prompt_id": "fc8ef010d8778b0f", "reading_canonical_log_prob": -404.8402099609375, "reading_distances": [41, 46, 39, 39], "reading_log_probs": [-402.43463134765625, -403.4171142578125, -404.01983642578125, -402.49725341796875], "reading_logprob_std": 0.7642941364570189, "tokenization_path_logprob_std": 315.48522865205484, "trajectory_balance_rmse": 6969.30806021294} {"canonical_length": 54, "canonical_nll": 0.37081146240234375, "generation_rouge_l_f1": 0.254071661237785, "generation_token_f1": 0.28013029315960913, "prompt_id": "68eb421daa21c33e", "reading_canonical_log_prob": -20.023818969726562, "reading_distances": [20, 21, 24, 17], "reading_log_probs": [-20.4298095703125, -20.121658325195312, -21.890277862548828, -23.877164840698242], "reading_logprob_std": 1.7149282614389576, "tokenization_path_logprob_std": 38.03645064552362, "trajectory_balance_rmse": 718.5250704865508} {"canonical_length": 133, "canonical_nll": 0.9410959688344396, "generation_rouge_l_f1": 0.34385964912280703, "generation_token_f1": 0.617543859649123, "prompt_id": "a12c87820d75347d", "reading_canonical_log_prob": -125.16576385498047, "reading_distances": [22, 28, 23, 20], "reading_log_probs": [-125.78106689453125, -129.37481689453125, -128.2703857421875, -126.85435485839844], "reading_logprob_std": 1.5769520958396652, "tokenization_path_logprob_std": 54.948418637614864, "trajectory_balance_rmse": 2498.840055278296} {"canonical_length": 220, "canonical_nll": 0.0904126600785689, "generation_rouge_l_f1": 0.7453083109919572, "generation_token_f1": 0.8310991957104558, "prompt_id": "c386599fdaa650e0", "reading_canonical_log_prob": -19.890785217285156, "reading_distances": [14, 10, 11, 12], "reading_log_probs": [-22.759851455688477, -22.223711013793945, -22.041616439819336, -23.4577579498291], "reading_logprob_std": 0.6358634966589761, "tokenization_path_logprob_std": 165.6673033208484, "trajectory_balance_rmse": 2368.7497852298566} {"canonical_length": 13, "canonical_nll": 0.8073491316575271, "generation_rouge_l_f1": 0.5217391304347826, "generation_token_f1": 0.5217391304347826, "prompt_id": "28aa44aad6506016", "reading_canonical_log_prob": -10.495538711547852, "reading_distances": [24, 28, 30, 33], "reading_log_probs": [-14.846817016601562, -11.304287910461426, -10.764443397521973, -12.517788887023926], "reading_logprob_std": 1.8137873815535994, "tokenization_path_logprob_std": 42.59641547998026, "trajectory_balance_rmse": 286.96703914368607} {"canonical_length": 150, "canonical_nll": 1.3576583862304688, "generation_rouge_l_f1": 0.28240740740740744, "generation_token_f1": 0.4768518518518518, "prompt_id": "764869f6f55c420d", "reading_canonical_log_prob": -203.6487579345703, "reading_distances": [11, 13, 15, 13], "reading_log_probs": [-202.71209716796875, -197.8790740966797, -199.8433074951172, -198.82455444335938], "reading_logprob_std": 2.091471936625339, "tokenization_path_logprob_std": 103.07599713070802, "trajectory_balance_rmse": 2441.8889895982993} {"canonical_length": 266, "canonical_nll": 0.7701445271198014, "generation_rouge_l_f1": 0.27833001988071576, "generation_token_f1": 0.5566600397614315, "prompt_id": "8fb6a1df0bc7d753", "reading_canonical_log_prob": -204.8584442138672, "reading_distances": [44, 32, 42, 44], "reading_log_probs": [-212.4302978515625, -211.6080780029297, -211.80064392089844, -215.13107299804688], "reading_logprob_std": 1.6306192369274384, "tokenization_path_logprob_std": 189.51353969742937, "trajectory_balance_rmse": 4411.17540403657} {"canonical_length": 81, "canonical_nll": 1.6242283950617284, "generation_rouge_l_f1": 0.1543026706231454, "generation_token_f1": 0.26706231454005935, "prompt_id": "f5aeb1160ca2b326", "reading_canonical_log_prob": -131.5625, "reading_distances": [52, 48, 48, 53], "reading_log_probs": [-124.90504455566406, -125.18061065673828, -126.69248962402344, -125.7908706665039], "reading_logprob_std": 0.7919751933670922, "tokenization_path_logprob_std": 119.54023248740941, "trajectory_balance_rmse": 1748.2314752222064} {"canonical_length": 29, "canonical_nll": 0.8218097029061153, "generation_rouge_l_f1": 0.4946236559139785, "generation_token_f1": 0.5376344086021505, "prompt_id": "7d27494a4e6427f6", "reading_canonical_log_prob": -23.832481384277344, "reading_distances": [38, 36, 37, 27], "reading_log_probs": [-25.249713897705078, -26.01763343811035, -25.62799072265625, -25.34082794189453], "reading_logprob_std": 0.34562083438708946, "tokenization_path_logprob_std": 36.589212603956256, "trajectory_balance_rmse": 528.8309467890015} {"canonical_length": 13, "canonical_nll": 0.4052589856661283, "generation_rouge_l_f1": 0.8421052631578948, "generation_token_f1": 0.8421052631578948, "prompt_id": "b015ce761d144f7f", "reading_canonical_log_prob": -5.268366813659668, "reading_distances": [29, 23, 31, 25], "reading_log_probs": [-6.223487377166748, -6.3031158447265625, -5.840139865875244, -6.682954788208008], "reading_logprob_std": 0.3456114970060906, "tokenization_path_logprob_std": 52.05422684333858, "trajectory_balance_rmse": 188.1819391160583} {"canonical_length": 246, "canonical_nll": 1.3068274521246188, "generation_rouge_l_f1": 0.2058252427184466, "generation_token_f1": 0.4776699029126214, "prompt_id": "e7d47f51d6085f91", "reading_canonical_log_prob": -321.47955322265625, "reading_distances": [24, 16, 22, 18], "reading_log_probs": [-320.6548767089844, -320.7974853515625, -321.65289306640625, -323.1011657714844], "reading_logprob_std": 1.1231230162641237, "tokenization_path_logprob_std": 132.69288004578223, "trajectory_balance_rmse": 5470.330889922043} {"canonical_length": 205, "canonical_nll": 1.5784953791920733, "generation_rouge_l_f1": 0.2558139534883721, "generation_token_f1": 0.5162790697674419, "prompt_id": "321f893720a9d001", "reading_canonical_log_prob": -323.591552734375, "reading_distances": [29, 23, 37, 23], "reading_log_probs": [-323.7889404296875, -324.05743408203125, -322.89697265625, -325.2220458984375], "reading_logprob_std": 0.9587510808310598, "tokenization_path_logprob_std": 406.8368035330538, "trajectory_balance_rmse": 4672.616646023222} {"canonical_length": 159, "canonical_nll": 1.2186801358588837, "generation_rouge_l_f1": 0.3355263157894737, "generation_token_f1": 0.5986842105263157, "prompt_id": "46617b87eb38bf31", "reading_canonical_log_prob": -193.7701416015625, "reading_distances": [34, 38, 37, 34], "reading_log_probs": [-193.72982788085938, -193.12474060058594, -191.49188232421875, -195.7216796875], "reading_logprob_std": 1.7474626619072013, "tokenization_path_logprob_std": 104.9765202819644, "trajectory_balance_rmse": 1898.2238906772627} {"canonical_length": 4, "canonical_nll": 0.5853859782218933, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "03d95aafdc590416", "reading_canonical_log_prob": -2.3415439128875732, "reading_distances": [38, 34, 23, 35], "reading_log_probs": [-3.333949327468872, -3.049428939819336, -2.6593761444091797, -3.368152141571045], "reading_logprob_std": 0.32828591952226766, "tokenization_path_logprob_std": 19.069537183240048, "trajectory_balance_rmse": 67.34425040857543} {"canonical_length": 153, "canonical_nll": 0.6374614441316891, "generation_rouge_l_f1": 0.25000000000000006, "generation_token_f1": 0.4285714285714286, "prompt_id": "3e79c37d02563d09", "reading_canonical_log_prob": -97.53160095214844, "reading_distances": [39, 36, 39, 30], "reading_log_probs": [-100.02896118164062, -99.77920532226562, -98.65338134765625, -104.23667907714844], "reading_logprob_std": 2.4489666730847977, "tokenization_path_logprob_std": 120.50398455903564, "trajectory_balance_rmse": 2121.914525216687} {"canonical_length": 385, "canonical_nll": 0.9461356819450081, "generation_rouge_l_f1": 0.24263431542461003, "generation_token_f1": 0.4055459272097054, "prompt_id": "8a6a50a6d59a29d8", "reading_canonical_log_prob": -364.2622375488281, "reading_distances": [35, 25, 33, 31], "reading_log_probs": [-371.3831787109375, -364.5310363769531, -364.234130859375, -365.4145202636719], "reading_logprob_std": 3.3658535225043944, "tokenization_path_logprob_std": 174.0101366404115, "trajectory_balance_rmse": 6659.416368973995} {"canonical_length": 28, "canonical_nll": 0.8216633115495954, "generation_rouge_l_f1": 0.17467248908296945, "generation_token_f1": 0.1834061135371179, "prompt_id": "280d0259825e43a6", "reading_canonical_log_prob": -23.006572723388672, "reading_distances": [32, 33, 36, 36], "reading_log_probs": [-22.350412368774414, -21.81332015991211, -21.49337387084961, -21.521379470825195], "reading_logprob_std": 0.3977706501121044, "tokenization_path_logprob_std": 93.38927726437817, "trajectory_balance_rmse": 409.5944554605702} {"canonical_length": 245, "canonical_nll": 1.648792500398597, "generation_rouge_l_f1": 0.15789473684210525, "generation_token_f1": 0.3157894736842105, "prompt_id": "fcecb04380951f2a", "reading_canonical_log_prob": -403.95416259765625, "reading_distances": [26, 23, 25, 26], "reading_log_probs": [-397.1983642578125, -396.30859375, -396.4659729003906, -397.3916931152344], "reading_logprob_std": 0.5338769405826245, "tokenization_path_logprob_std": 170.9156146924612, "trajectory_balance_rmse": 5018.149988540848} {"canonical_length": 126, "canonical_nll": 1.413804190499442, "generation_rouge_l_f1": 0.20062695924764892, "generation_token_f1": 0.3448275862068966, "prompt_id": "1ae061f91192abf8", "reading_canonical_log_prob": -178.1393280029297, "reading_distances": [29, 32, 30, 21], "reading_log_probs": [-182.6238250732422, -189.42514038085938, -185.21908569335938, -190.680908203125], "reading_logprob_std": 3.730607749277836, "tokenization_path_logprob_std": 165.72117654673696, "trajectory_balance_rmse": 2328.081663236526} {"canonical_length": 205, "canonical_nll": 1.267571426019436, "generation_rouge_l_f1": 0.23799582463465555, "generation_token_f1": 0.4133611691022964, "prompt_id": "eedaf90695c8e20b", "reading_canonical_log_prob": -259.8521423339844, "reading_distances": [16, 17, 16, 21], "reading_log_probs": [-257.5074768066406, -256.5106201171875, -255.2281494140625, -256.9125061035156], "reading_logprob_std": 0.9655027907403068, "tokenization_path_logprob_std": 131.27772113573536, "trajectory_balance_rmse": 4609.619264173068} {"canonical_length": 119, "canonical_nll": 1.2714355212299764, "generation_rouge_l_f1": 0.2038216560509554, "generation_token_f1": 0.356687898089172, "prompt_id": "a0b38c7913178787", "reading_canonical_log_prob": -151.3008270263672, "reading_distances": [48, 44, 45, 41], "reading_log_probs": [-131.05679321289062, -129.95068359375, -132.2034912109375, -133.37319946289062], "reading_logprob_std": 1.4736888843020164, "tokenization_path_logprob_std": 88.45122398278033, "trajectory_balance_rmse": 1465.8023018332692} {"canonical_length": 68, "canonical_nll": 1.6535907072179459, "generation_rouge_l_f1": 0.3129251700680272, "generation_token_f1": 0.44897959183673464, "prompt_id": "4101000b96f3bbde", "reading_canonical_log_prob": -112.44416809082031, "reading_distances": [18, 19, 23, 15], "reading_log_probs": [-109.09098815917969, -111.97583770751953, -109.65692901611328, -111.6688003540039], "reading_logprob_std": 1.4377933734733621, "tokenization_path_logprob_std": 58.85840188283317, "trajectory_balance_rmse": 1140.456893420175} {"canonical_length": 138, "canonical_nll": 1.196895212366961, "generation_rouge_l_f1": 0.2194513715710723, "generation_token_f1": 0.4239401496259352, "prompt_id": "5539256ae94ad40f", "reading_canonical_log_prob": -165.17153930664062, "reading_distances": [63, 57, 65, 62], "reading_log_probs": [-163.39291381835938, -160.53912353515625, -161.59463500976562, -161.745361328125], "reading_logprob_std": 1.1791305897172633, "tokenization_path_logprob_std": 247.4581956484089, "trajectory_balance_rmse": 2542.3616913842457} {"canonical_length": 86, "canonical_nll": 1.279618728992551, "generation_rouge_l_f1": 0.20245398773006135, "generation_token_f1": 0.25153374233128833, "prompt_id": "3b0f78eeffb87a9e", "reading_canonical_log_prob": -110.04721069335938, "reading_distances": [32, 33, 33, 35], "reading_log_probs": [-105.60845947265625, -106.71723937988281, -108.64678955078125, -106.58792114257812], "reading_logprob_std": 1.27144975978753, "tokenization_path_logprob_std": 53.507284221416825, "trajectory_balance_rmse": 1641.0900167302707} {"canonical_length": 302, "canonical_nll": 1.392162802993067, "generation_rouge_l_f1": 0.3665338645418327, "generation_token_f1": 0.5657370517928287, "prompt_id": "3a8d8459792a6678", "reading_canonical_log_prob": -420.43316650390625, "reading_distances": [14, 16, 18, 17], "reading_log_probs": [-416.844482421875, -415.97760009765625, -412.36749267578125, -413.6651611328125], "reading_logprob_std": 2.060889755213671, "tokenization_path_logprob_std": 118.72993128490124, "trajectory_balance_rmse": 6106.812746792326} {"canonical_length": 289, "canonical_nll": 1.2529937849737782, "generation_rouge_l_f1": 0.24489795918367346, "generation_token_f1": 0.49387755102040815, "prompt_id": "a788a16b80a319d9", "reading_canonical_log_prob": -362.1152038574219, "reading_distances": [12, 20, 16, 20], "reading_log_probs": [-360.86065673828125, -360.20623779296875, -360.52301025390625, -358.799072265625], "reading_logprob_std": 0.9057603455607655, "tokenization_path_logprob_std": 202.83266044983165, "trajectory_balance_rmse": 5469.883832687723} {"canonical_length": 252, "canonical_nll": 1.2423641870892237, "generation_rouge_l_f1": 0.19209039548022602, "generation_token_f1": 0.4067796610169492, "prompt_id": "79833cd8d7978d26", "reading_canonical_log_prob": -313.0757751464844, "reading_distances": [30, 30, 26, 31], "reading_log_probs": [-309.175048828125, -307.5733947753906, -309.8827819824219, -308.93658447265625], "reading_logprob_std": 0.9665183358054652, "tokenization_path_logprob_std": 189.66377100093442, "trajectory_balance_rmse": 4223.427946827411} {"canonical_length": 26, "canonical_nll": 1.7122298020582933, "generation_rouge_l_f1": 0.08058608058608058, "generation_token_f1": 0.12454212454212454, "prompt_id": "7c29635f759f46ec", "reading_canonical_log_prob": -44.517974853515625, "reading_distances": [18, 22, 22, 16], "reading_log_probs": [-40.36379623413086, -42.186614990234375, -44.20047378540039, -43.61375427246094], "reading_logprob_std": 1.708839364050207, "tokenization_path_logprob_std": 28.84163459596364, "trajectory_balance_rmse": 532.2177519187615} {"canonical_length": 346, "canonical_nll": 1.3034683844946713, "generation_rouge_l_f1": 0.28998505231689087, "generation_token_f1": 0.46337817638266066, "prompt_id": "88945c2caccedba1", "reading_canonical_log_prob": -451.00006103515625, "reading_distances": [39, 36, 53, 44], "reading_log_probs": [-438.76416015625, -443.7484130859375, -440.01519775390625, -447.74542236328125], "reading_logprob_std": 4.049077009632029, "tokenization_path_logprob_std": 116.71682525666593, "trajectory_balance_rmse": 7207.054286137991} {"canonical_length": 17, "canonical_nll": 1.3128837136661304, "generation_rouge_l_f1": 0.3636363636363636, "generation_token_f1": 0.3636363636363636, "prompt_id": "f3be16dc1ce038fa", "reading_canonical_log_prob": -22.31902313232422, "reading_distances": [34, 49, 48, 47], "reading_log_probs": [-24.62454605102539, -22.4200382232666, -22.49419403076172, -23.608606338500977], "reading_logprob_std": 1.044449457543433, "tokenization_path_logprob_std": 76.31051306988105, "trajectory_balance_rmse": 467.6982195027528} {"canonical_length": 56, "canonical_nll": 1.4174889155796595, "generation_rouge_l_f1": 0.3287671232876712, "generation_token_f1": 0.3835616438356164, "prompt_id": "08f613a2192f5ff2", "reading_canonical_log_prob": -79.37937927246094, "reading_distances": [23, 24, 20, 20], "reading_log_probs": [-79.68122100830078, -78.97871398925781, -79.53582000732422, -79.29888153076172], "reading_logprob_std": 0.3068462224318951, "tokenization_path_logprob_std": 48.88420632169279, "trajectory_balance_rmse": 1146.0813894752355} {"canonical_length": 180, "canonical_nll": 1.7000878228081597, "generation_rouge_l_f1": 0.19806763285024154, "generation_token_f1": 0.4251207729468599, "prompt_id": "ddaf5127b3985530", "reading_canonical_log_prob": -306.01580810546875, "reading_distances": [26, 27, 31, 27], "reading_log_probs": [-307.77001953125, -308.65374755859375, -308.925537109375, -310.82958984375], "reading_logprob_std": 1.2881104579275817, "tokenization_path_logprob_std": 86.09295905362026, "trajectory_balance_rmse": 3489.404104098766} {"canonical_length": 184, "canonical_nll": 0.9306760870892069, "generation_rouge_l_f1": 0.2756892230576441, "generation_token_f1": 0.5012531328320801, "prompt_id": "6fc1da282516cfc1", "reading_canonical_log_prob": -171.24440002441406, "reading_distances": [18, 12, 9, 14], "reading_log_probs": [-170.217529296875, -169.80055236816406, -167.22918701171875, -169.66952514648438], "reading_logprob_std": 1.3536588092497999, "tokenization_path_logprob_std": 164.12736068800334, "trajectory_balance_rmse": 3394.2159892242844} {"canonical_length": 80, "canonical_nll": 1.2812549591064453, "generation_rouge_l_f1": 0.13095238095238096, "generation_token_f1": 0.2619047619047619, "prompt_id": "9d2634370eff9e7e", "reading_canonical_log_prob": -102.50039672851562, "reading_distances": [39, 26, 33, 34], "reading_log_probs": [-116.16738891601562, -116.11980438232422, -113.15914916992188, -117.35453796386719], "reading_logprob_std": 1.7877459494224537, "tokenization_path_logprob_std": 141.4072001467576, "trajectory_balance_rmse": 1742.776214806779} {"canonical_length": 132, "canonical_nll": 0.7485512820157137, "generation_rouge_l_f1": 0.41750841750841755, "generation_token_f1": 0.6262626262626264, "prompt_id": "8a81f9867fa3fdd7", "reading_canonical_log_prob": -98.80876922607422, "reading_distances": [33, 32, 31, 34], "reading_log_probs": [-93.563720703125, -95.30810546875, -92.84945678710938, -93.89273071289062], "reading_logprob_std": 1.0327131847010544, "tokenization_path_logprob_std": 51.74831145074286, "trajectory_balance_rmse": 1713.7195079348346} {"canonical_length": 33, "canonical_nll": 1.09237104473692, "generation_rouge_l_f1": 0.8363636363636363, "generation_token_f1": 0.8363636363636363, "prompt_id": "32094cae1c9e7173", "reading_canonical_log_prob": -36.04824447631836, "reading_distances": [44, 46, 39, 44], "reading_log_probs": [-33.176475524902344, -32.32570266723633, -33.43183517456055, -33.007957458496094], "reading_logprob_std": 0.4731185886269553, "tokenization_path_logprob_std": 51.02296303645874, "trajectory_balance_rmse": 596.7929839579423} {"canonical_length": 23, "canonical_nll": 1.8933463718580164, "generation_rouge_l_f1": 0.42857142857142855, "generation_token_f1": 0.42857142857142855, "prompt_id": "896dee39b103c9ac", "reading_canonical_log_prob": -43.546966552734375, "reading_distances": [27, 29, 30, 26], "reading_log_probs": [-42.68780517578125, -54.05934143066406, -56.071739196777344, -52.63809585571289], "reading_logprob_std": 5.953356207752604, "tokenization_path_logprob_std": 91.64810023608727, "trajectory_balance_rmse": 596.1157148179175} {"canonical_length": 105, "canonical_nll": 1.7542137509300595, "generation_rouge_l_f1": 0.2458100558659218, "generation_token_f1": 0.4134078212290503, "prompt_id": "1057782952469d89", "reading_canonical_log_prob": -184.19244384765625, "reading_distances": [11, 10, 10, 9], "reading_log_probs": [-183.9508514404297, -181.6235809326172, -182.12417602539062, -180.3995361328125], "reading_logprob_std": 1.4744500494666866, "tokenization_path_logprob_std": 165.26909361938408, "trajectory_balance_rmse": 2503.4107657911914} {"canonical_length": 25, "canonical_nll": 0.7093760681152343, "generation_rouge_l_f1": 0.2068965517241379, "generation_token_f1": 0.21839080459770116, "prompt_id": "03db11358469fd8f", "reading_canonical_log_prob": -17.73440170288086, "reading_distances": [19, 20, 25, 24], "reading_log_probs": [-19.116533279418945, -18.409976959228516, -18.60161018371582, -18.200830459594727], "reading_logprob_std": 0.39200012974345266, "tokenization_path_logprob_std": 33.605541697978374, "trajectory_balance_rmse": 234.92766463730553} {"canonical_length": 240, "canonical_nll": 1.1902088165283202, "generation_rouge_l_f1": 0.21588594704684316, "generation_token_f1": 0.505091649694501, "prompt_id": "85339fcc838b71e5", "reading_canonical_log_prob": -285.6501159667969, "reading_distances": [21, 19, 22, 25], "reading_log_probs": [-283.25958251953125, -285.2799072265625, -283.9803771972656, -283.331787109375], "reading_logprob_std": 0.9359081944930352, "tokenization_path_logprob_std": 218.18400093928244, "trajectory_balance_rmse": 5527.035953909315} {"canonical_length": 28, "canonical_nll": 0.47488512311662945, "generation_rouge_l_f1": 0.20491803278688525, "generation_token_f1": 0.20491803278688525, "prompt_id": "5af4615607f58b8b", "reading_canonical_log_prob": -13.296783447265625, "reading_distances": [47, 45, 38, 39], "reading_log_probs": [-13.79415512084961, -13.435526847839355, -13.5853271484375, -13.627345085144043], "reading_logprob_std": 0.14749260475902093, "tokenization_path_logprob_std": 88.01348039761564, "trajectory_balance_rmse": 614.5365719708515} {"canonical_length": 65, "canonical_nll": 1.8040534386268028, "generation_rouge_l_f1": 0.18354430379746833, "generation_token_f1": 0.2784810126582279, "prompt_id": "13f18463e2c1c5f7", "reading_canonical_log_prob": -117.26347351074219, "reading_distances": [25, 18, 20, 18], "reading_log_probs": [-113.1871337890625, -113.62263488769531, -117.25817108154297, -112.56597900390625], "reading_logprob_std": 2.1114598164492486, "tokenization_path_logprob_std": 34.67333339701434, "trajectory_balance_rmse": 1344.0905060375635} {"canonical_length": 64, "canonical_nll": 1.898523211479187, "generation_rouge_l_f1": 0.16828478964401294, "generation_token_f1": 0.27831715210355984, "prompt_id": "d02d88b60737f21f", "reading_canonical_log_prob": -121.50548553466797, "reading_distances": [21, 23, 16, 19], "reading_log_probs": [-114.79420471191406, -114.800048828125, -120.04203033447266, -118.35848236083984], "reading_logprob_std": 2.6334219844192606, "tokenization_path_logprob_std": 58.69498384311468, "trajectory_balance_rmse": 1577.2049160239762} {"canonical_length": 236, "canonical_nll": 1.4393499342061706, "generation_rouge_l_f1": 0.2048192771084337, "generation_token_f1": 0.4457831325301205, "prompt_id": "559f28b026670bdd", "reading_canonical_log_prob": -339.68658447265625, "reading_distances": [19, 13, 16, 14], "reading_log_probs": [-338.28387451171875, -338.24798583984375, -339.69464111328125, -338.898193359375], "reading_logprob_std": 0.6781622452604711, "tokenization_path_logprob_std": 53.800614089575035, "trajectory_balance_rmse": 4999.791423094487} {"canonical_length": 50, "canonical_nll": 0.7841214752197265, "generation_rouge_l_f1": 0.45945945945945943, "generation_token_f1": 0.5270270270270271, "prompt_id": "44cafb7b8f27398a", "reading_canonical_log_prob": -39.20607376098633, "reading_distances": [30, 23, 28, 26], "reading_log_probs": [-37.149391174316406, -37.24945068359375, -36.707828521728516, -37.21450424194336], "reading_logprob_std": 0.2517481046926163, "tokenization_path_logprob_std": 92.39931531733045, "trajectory_balance_rmse": 1039.0175597284895} {"canonical_length": 274, "canonical_nll": 1.0917474231580748, "generation_rouge_l_f1": 0.23177570093457944, "generation_token_f1": 0.4897196261682243, "prompt_id": "55f3c7490c4e7162", "reading_canonical_log_prob": -299.1387939453125, "reading_distances": [14, 17, 22, 21], "reading_log_probs": [-294.0793151855469, -295.8141174316406, -292.7499084472656, -294.7657470703125], "reading_logprob_std": 1.2845269009076732, "tokenization_path_logprob_std": 234.65915073581962, "trajectory_balance_rmse": 4786.2494268756345} {"canonical_length": 308, "canonical_nll": 1.120258479923397, "generation_rouge_l_f1": 0.24, "generation_token_f1": 0.4417391304347827, "prompt_id": "41cd8f9fa13781b8", "reading_canonical_log_prob": -345.03961181640625, "reading_distances": [28, 27, 31, 17], "reading_log_probs": [-349.159912109375, -349.2354431152344, -347.6207580566406, -351.3995361328125], "reading_logprob_std": 1.5534983658769077, "tokenization_path_logprob_std": 138.48999353861902, "trajectory_balance_rmse": 6359.31597608404} {"canonical_length": 373, "canonical_nll": 1.1766061245915718, "generation_rouge_l_f1": 0.21717990275526738, "generation_token_f1": 0.4278768233387358, "prompt_id": "28de6d5b6a032c7b", "reading_canonical_log_prob": -438.87408447265625, "reading_distances": [26, 21, 23, 25], "reading_log_probs": [-434.99456787109375, -434.21630859375, -434.1603698730469, -433.9798889160156], "reading_logprob_std": 0.44932764260803326, "tokenization_path_logprob_std": 302.7830897284959, "trajectory_balance_rmse": 6195.968298758657} {"canonical_length": 132, "canonical_nll": 1.6542298288056345, "generation_rouge_l_f1": 0.22222222222222224, "generation_token_f1": 0.42145593869731807, "prompt_id": "dc17c375b90b9957", "reading_canonical_log_prob": -218.35833740234375, "reading_distances": [23, 24, 22, 18], "reading_log_probs": [-215.08718872070312, -212.76220703125, -214.8145751953125, -215.6270751953125], "reading_logprob_std": 1.2533680067555468, "tokenization_path_logprob_std": 232.16366755528566, "trajectory_balance_rmse": 2791.048203321864} {"canonical_length": 24, "canonical_nll": 0.42020905017852783, "generation_rouge_l_f1": 0.37142857142857144, "generation_token_f1": 0.45714285714285713, "prompt_id": "c1a810543bd499fe", "reading_canonical_log_prob": -10.085017204284668, "reading_distances": [28, 25, 24, 27], "reading_log_probs": [-10.70538330078125, -8.978310585021973, -9.077335357666016, -9.013132095336914], "reading_logprob_std": 0.8422279439578296, "tokenization_path_logprob_std": 56.689061423436456, "trajectory_balance_rmse": 386.1433585326118} {"canonical_length": 15, "canonical_nll": 1.1582764943440755, "generation_rouge_l_f1": 0.34782608695652173, "generation_token_f1": 0.34782608695652173, "prompt_id": "f21bbbf6325c37df", "reading_canonical_log_prob": -17.374147415161133, "reading_distances": [42, 39, 38, 36], "reading_log_probs": [-18.432525634765625, -16.917194366455078, -18.351619720458984, -17.79190444946289], "reading_logprob_std": 0.698160985068339, "tokenization_path_logprob_std": 61.28016792667423, "trajectory_balance_rmse": 314.0290123996709} {"canonical_length": 141, "canonical_nll": 1.592248578443595, "generation_rouge_l_f1": 0.25, "generation_token_f1": 0.4583333333333333, "prompt_id": "ff640b48674add94", "reading_canonical_log_prob": -224.50704956054688, "reading_distances": [48, 54, 50, 41], "reading_log_probs": [-226.68545532226562, -228.5507354736328, -229.67630004882812, -225.452392578125], "reading_logprob_std": 1.885313514702058, "tokenization_path_logprob_std": 127.42350542478205, "trajectory_balance_rmse": 3117.3811411614906} {"canonical_length": 133, "canonical_nll": 2.0051141036184212, "generation_rouge_l_f1": 0.20565552699228792, "generation_token_f1": 0.35475578406169667, "prompt_id": "294ad70d174d121a", "reading_canonical_log_prob": -266.68017578125, "reading_distances": [29, 22, 27, 28], "reading_log_probs": [-264.48101806640625, -266.2833557128906, -265.94598388671875, -267.7378234863281], "reading_logprob_std": 1.3367034323801434, "tokenization_path_logprob_std": 299.7930302204649, "trajectory_balance_rmse": 3326.9281361624826} {"canonical_length": 49, "canonical_nll": 2.5109589245854593, "generation_rouge_l_f1": 0.27906976744186046, "generation_token_f1": 0.44186046511627913, "prompt_id": "fdd72207f77ed671", "reading_canonical_log_prob": -123.0369873046875, "reading_distances": [19, 23, 20, 23], "reading_log_probs": [-123.71890258789062, -123.46842956542969, -124.41558837890625, -124.35541534423828], "reading_logprob_std": 0.4691064821581241, "tokenization_path_logprob_std": 90.02403003364984, "trajectory_balance_rmse": 957.2850520144733} {"canonical_length": 9, "canonical_nll": 1.5828147464328342, "generation_rouge_l_f1": 0.3157894736842105, "generation_token_f1": 0.4210526315789474, "prompt_id": "bead008713b25f77", "reading_canonical_log_prob": -14.245332717895508, "reading_distances": [36, 42, 33, 40], "reading_log_probs": [-16.777360916137695, -15.122337341308594, -13.920108795166016, -15.919974327087402], "reading_logprob_std": 1.2151524205879884, "tokenization_path_logprob_std": 70.42332085087997, "trajectory_balance_rmse": 227.25741342640217} {"canonical_length": 353, "canonical_nll": 1.0142357153527797, "generation_rouge_l_f1": 0.1942675159235669, "generation_token_f1": 0.3853503184713376, "prompt_id": "5e76d06be1ed709a", "reading_canonical_log_prob": -358.02520751953125, "reading_distances": [31, 31, 23, 25], "reading_log_probs": [-358.4445495605469, -360.54364013671875, -357.90887451171875, -359.10333251953125], "reading_logprob_std": 1.139090640788965, "tokenization_path_logprob_std": 320.2222753412129, "trajectory_balance_rmse": 6103.795672401098} {"canonical_length": 238, "canonical_nll": 1.3772690716911764, "generation_rouge_l_f1": 0.2593984962406015, "generation_token_f1": 0.4661654135338346, "prompt_id": "f6f871799dc3c062", "reading_canonical_log_prob": -327.7900390625, "reading_distances": [30, 29, 31, 29], "reading_log_probs": [-318.19158935546875, -319.0177917480469, -319.5146484375, -313.51422119140625], "reading_logprob_std": 2.7515473108730526, "tokenization_path_logprob_std": 183.7092795535463, "trajectory_balance_rmse": 3832.8786652592407} {"canonical_length": 108, "canonical_nll": 0.8039201100667318, "generation_rouge_l_f1": 0.34104046242774566, "generation_token_f1": 0.41040462427745666, "prompt_id": "e5243c35b3e882de", "reading_canonical_log_prob": -86.82337188720703, "reading_distances": [26, 27, 23, 32], "reading_log_probs": [-84.58224487304688, -84.4051513671875, -86.01921081542969, -85.44315338134766], "reading_logprob_std": 0.7556385236050345, "tokenization_path_logprob_std": 90.68055039262956, "trajectory_balance_rmse": 1489.4199516538754} {"canonical_length": 147, "canonical_nll": 1.6102431939572703, "generation_rouge_l_f1": 0.1897810218978102, "generation_token_f1": 0.36982968369829683, "prompt_id": "26ddb945cf9f0d7d", "reading_canonical_log_prob": -236.70574951171875, "reading_distances": [33, 39, 29, 28], "reading_log_probs": [-232.99920654296875, -236.40528869628906, -234.60906982421875, -234.7353973388672], "reading_logprob_std": 1.3915911732670647, "tokenization_path_logprob_std": 249.092303416829, "trajectory_balance_rmse": 3572.8126980868246} {"canonical_length": 48, "canonical_nll": 1.6852037111918132, "generation_rouge_l_f1": 0.11838006230529595, "generation_token_f1": 0.17445482866043616, "prompt_id": "e4581a40938db0d9", "reading_canonical_log_prob": -80.88977813720703, "reading_distances": [19, 20, 15, 22], "reading_log_probs": [-81.94229888916016, -77.16671752929688, -76.29512023925781, -80.92768859863281], "reading_logprob_std": 2.7702513859289946, "tokenization_path_logprob_std": 68.71647800898012, "trajectory_balance_rmse": 994.320211404725} {"canonical_length": 10, "canonical_nll": 0.023759551346302032, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "267c9641a2f8cfb9", "reading_canonical_log_prob": -0.23759551346302032, "reading_distances": [22, 25, 25, 23], "reading_log_probs": [-0.6113056540489197, -0.4354180693626404, -1.0581319332122803, -0.42010366916656494], "reading_logprob_std": 0.2975224806668399, "tokenization_path_logprob_std": 54.23883931999697, "trajectory_balance_rmse": 189.17333367856804} {"canonical_length": 63, "canonical_nll": 0.8005589076450893, "generation_rouge_l_f1": 0.3146067415730337, "generation_token_f1": 0.3539325842696629, "prompt_id": "df9c5775b476f764", "reading_canonical_log_prob": -50.435211181640625, "reading_distances": [29, 29, 30, 25], "reading_log_probs": [-47.546897888183594, -47.843292236328125, -46.19270706176758, -48.205360412597656], "reading_logprob_std": 0.8785192842397186, "tokenization_path_logprob_std": 88.93311000457297, "trajectory_balance_rmse": 903.5369100844501} {"canonical_length": 106, "canonical_nll": 1.9164995877247937, "generation_rouge_l_f1": 0.16300940438871472, "generation_token_f1": 0.31974921630094044, "prompt_id": "518ad8fe3c294839", "reading_canonical_log_prob": -203.14895629882812, "reading_distances": [18, 25, 21, 17], "reading_log_probs": [-204.57736206054688, -208.0251007080078, -203.7285614013672, -210.17233276367188], "reading_logprob_std": 3.006993326578948, "tokenization_path_logprob_std": 24.626478685219013, "trajectory_balance_rmse": 2272.247926428619} {"canonical_length": 203, "canonical_nll": 0.6539287097348369, "generation_rouge_l_f1": 0.37410071942446044, "generation_token_f1": 0.5755395683453237, "prompt_id": "6369662548b008b0", "reading_canonical_log_prob": -132.74752807617188, "reading_distances": [23, 23, 26, 23], "reading_log_probs": [-135.35145568847656, -135.66273498535156, -133.3759765625, -134.0086669921875], "reading_logprob_std": 1.0865847202367727, "tokenization_path_logprob_std": 146.7912443475534, "trajectory_balance_rmse": 1981.357852561433} {"canonical_length": 68, "canonical_nll": 0.41021279727711396, "generation_rouge_l_f1": 0.6724137931034484, "generation_token_f1": 0.689655172413793, "prompt_id": "3f0899fb36d7151b", "reading_canonical_log_prob": -27.89447021484375, "reading_distances": [32, 35, 29, 37], "reading_log_probs": [-28.087589263916016, -28.696044921875, -28.356847763061523, -28.709049224853516], "reading_logprob_std": 0.2983566942777748, "tokenization_path_logprob_std": 73.58017108151827, "trajectory_balance_rmse": 1307.9047454194413} {"canonical_length": 59, "canonical_nll": 0.9609275429935779, "generation_rouge_l_f1": 0.39285714285714285, "generation_token_f1": 0.5, "prompt_id": "ca44a0ae267d5905", "reading_canonical_log_prob": -56.694725036621094, "reading_distances": [16, 19, 18, 20], "reading_log_probs": [-62.47321701049805, -62.742462158203125, -61.89308547973633, -62.5422477722168], "reading_logprob_std": 0.36477721658429224, "tokenization_path_logprob_std": 58.36792934211417, "trajectory_balance_rmse": 1122.280549533668} {"canonical_length": 144, "canonical_nll": 2.568219714694553, "generation_rouge_l_f1": 0.16853932584269662, "generation_token_f1": 0.2808988764044944, "prompt_id": "b3b2b51b9f629425", "reading_canonical_log_prob": -369.8236389160156, "reading_distances": [9, 12, 15, 11], "reading_log_probs": [-381.6942138671875, -394.1546630859375, -386.4756164550781, -379.5089111328125], "reading_logprob_std": 6.486530556725883, "tokenization_path_logprob_std": 127.03967785397391, "trajectory_balance_rmse": 2870.915759212965} {"canonical_length": 120, "canonical_nll": 1.1079999287923177, "generation_rouge_l_f1": 0.21932114882506526, "generation_token_f1": 0.3812010443864229, "prompt_id": "4412ea6c558d478d", "reading_canonical_log_prob": -132.95999145507812, "reading_distances": [24, 26, 31, 24], "reading_log_probs": [-137.083984375, -135.72216796875, -134.32669067382812, -135.16952514648438], "reading_logprob_std": 1.157781914353344, "tokenization_path_logprob_std": 74.62344038603939, "trajectory_balance_rmse": 2346.848633769827} {"canonical_length": 30, "canonical_nll": 0.9974594751993815, "generation_rouge_l_f1": 0.08609271523178808, "generation_token_f1": 0.12582781456953643, "prompt_id": "7938d8548e06b8bf", "reading_canonical_log_prob": -29.923784255981445, "reading_distances": [8, 9, 6, 8], "reading_log_probs": [-30.857940673828125, -28.73149871826172, -31.53868293762207, -31.782150268554688], "reading_logprob_std": 1.3870040632917877, "tokenization_path_logprob_std": 78.06722708344438, "trajectory_balance_rmse": 343.6944301269832} {"canonical_length": 332, "canonical_nll": 1.1378798886954067, "generation_rouge_l_f1": 0.21390374331550804, "generation_token_f1": 0.4563279857397505, "prompt_id": "8829bdef994324cc", "reading_canonical_log_prob": -377.776123046875, "reading_distances": [34, 33, 33, 35], "reading_log_probs": [-374.12469482421875, -375.44451904296875, -377.7826232910156, -375.89459228515625], "reading_logprob_std": 1.5135188616815387, "tokenization_path_logprob_std": 392.406738425449, "trajectory_balance_rmse": 6736.8062251152005} {"canonical_length": 14, "canonical_nll": 0.6665353775024414, "generation_rouge_l_f1": 0.0606060606060606, "generation_token_f1": 0.06926406926406926, "prompt_id": "75996f4d9b7d5f4d", "reading_canonical_log_prob": -9.33149528503418, "reading_distances": [10, 18, 19, 23], "reading_log_probs": [-8.525466918945312, -8.833405494689941, -9.100992202758789, -8.987569808959961], "reading_logprob_std": 0.24963782453196537, "tokenization_path_logprob_std": 44.87913003290517, "trajectory_balance_rmse": 239.74896488888345} {"canonical_length": 164, "canonical_nll": 1.2797425432903011, "generation_rouge_l_f1": 0.15609756097560976, "generation_token_f1": 0.3317073170731707, "prompt_id": "f5d8be75fb333542", "reading_canonical_log_prob": -209.87777709960938, "reading_distances": [18, 26, 19, 19], "reading_log_probs": [-207.24917602539062, -207.47308349609375, -206.58334350585938, -204.05665588378906], "reading_logprob_std": 1.5687951208357633, "tokenization_path_logprob_std": 206.27852066654424, "trajectory_balance_rmse": 3720.5532815306497} {"canonical_length": 45, "canonical_nll": 1.2917508443196615, "generation_rouge_l_f1": 0.29473684210526313, "generation_token_f1": 0.42105263157894735, "prompt_id": "61f6f6b3601f9e9d", "reading_canonical_log_prob": -58.128787994384766, "reading_distances": [19, 15, 27, 22], "reading_log_probs": [-58.76026916503906, -57.208927154541016, -59.39255142211914, -59.878936767578125], "reading_logprob_std": 1.1615936277039411, "tokenization_path_logprob_std": 81.24775040797213, "trajectory_balance_rmse": 886.854547688578} {"canonical_length": 26, "canonical_nll": 1.5982011648324819, "generation_rouge_l_f1": 0.30508474576271183, "generation_token_f1": 0.3728813559322034, "prompt_id": "4478590e9673bdfe", "reading_canonical_log_prob": -41.55323028564453, "reading_distances": [35, 29, 29, 31], "reading_log_probs": [-38.30419158935547, -39.2452392578125, -38.38776397705078, -39.03615951538086], "reading_logprob_std": 0.4679500572499202, "tokenization_path_logprob_std": 96.06367513159901, "trajectory_balance_rmse": 837.824788555943} {"canonical_length": 90, "canonical_nll": 0.848294915093316, "generation_rouge_l_f1": 0.14285714285714288, "generation_token_f1": 0.27976190476190477, "prompt_id": "b3f4ac8bd91cf595", "reading_canonical_log_prob": -76.34654235839844, "reading_distances": [30, 32, 23, 32], "reading_log_probs": [-82.42522430419922, -83.2071533203125, -85.16868591308594, -81.89823150634766], "reading_logprob_std": 1.4338896531463508, "tokenization_path_logprob_std": 61.57529350653759, "trajectory_balance_rmse": 1690.6569104258076} {"canonical_length": 116, "canonical_nll": 1.015546206770272, "generation_rouge_l_f1": 0.3089430894308943, "generation_token_f1": 0.5447154471544715, "prompt_id": "47d49f308f57d626", "reading_canonical_log_prob": -117.80335998535156, "reading_distances": [32, 35, 39, 36], "reading_log_probs": [-120.18972778320312, -122.98876190185547, -123.35395050048828, -120.13369750976562], "reading_logprob_std": 1.7441526479763223, "tokenization_path_logprob_std": 98.8302505399285, "trajectory_balance_rmse": 2717.6050819730067} {"canonical_length": 264, "canonical_nll": 1.228855017459754, "generation_rouge_l_f1": 0.26815642458100564, "generation_token_f1": 0.48417132216014896, "prompt_id": "d8021c0fb98991e7", "reading_canonical_log_prob": -324.417724609375, "reading_distances": [15, 24, 20, 24], "reading_log_probs": [-330.413330078125, -322.6012878417969, -325.5276184082031, -325.8447265625], "reading_logprob_std": 3.226894577958917, "tokenization_path_logprob_std": 212.2955675850639, "trajectory_balance_rmse": 5036.517803549265} {"canonical_length": 165, "canonical_nll": 1.4854031649502841, "generation_rouge_l_f1": 0.21957040572792363, "generation_token_f1": 0.42482100238663484, "prompt_id": "fb0e2e105f1d060e", "reading_canonical_log_prob": -245.09152221679688, "reading_distances": [35, 27, 23, 27], "reading_log_probs": [-245.63644409179688, -241.21038818359375, -239.5423126220703, -243.48374938964844], "reading_logprob_std": 2.6590736920726146, "tokenization_path_logprob_std": 160.32784296165613, "trajectory_balance_rmse": 3093.05805187999} {"canonical_length": 12, "canonical_nll": 0.8618132273356119, "generation_rouge_l_f1": 0.5263157894736842, "generation_token_f1": 0.5263157894736842, "prompt_id": "d8467aeeb0933718", "reading_canonical_log_prob": -10.341758728027344, "reading_distances": [41, 42, 45, 36], "reading_log_probs": [-11.05589485168457, -12.077396392822266, -11.484644889831543, -10.802781105041504], "reading_logprob_std": 0.5576957996709765, "tokenization_path_logprob_std": 32.18676415116405, "trajectory_balance_rmse": 230.61417093872026} {"canonical_length": 62, "canonical_nll": 0.41137941422001006, "generation_rouge_l_f1": 0.7563025210084034, "generation_token_f1": 0.8403361344537815, "prompt_id": "b80e624242a44cf6", "reading_canonical_log_prob": -25.505523681640625, "reading_distances": [23, 29, 22, 21], "reading_log_probs": [-27.280033111572266, -26.34937286376953, -26.449981689453125, -26.449127197265625], "reading_logprob_std": 0.4345104804686907, "tokenization_path_logprob_std": 133.2089001699391, "trajectory_balance_rmse": 1292.177988511207} {"canonical_length": 225, "canonical_nll": 1.2076778157552084, "generation_rouge_l_f1": 0.17586912065439672, "generation_token_f1": 0.3599182004089979, "prompt_id": "aad9f0c226359f45", "reading_canonical_log_prob": -271.7275085449219, "reading_distances": [13, 20, 17, 19], "reading_log_probs": [-271.8026123046875, -272.33929443359375, -270.1709899902344, -269.59619140625], "reading_logprob_std": 1.3030433625246756, "tokenization_path_logprob_std": 266.19765198631666, "trajectory_balance_rmse": 4374.140958741569} {"canonical_length": 221, "canonical_nll": 0.9736601540405826, "generation_rouge_l_f1": 0.3168724279835391, "generation_token_f1": 0.5061728395061729, "prompt_id": "2fd9ba0f00f7f678", "reading_canonical_log_prob": -215.17889404296875, "reading_distances": [36, 26, 21, 28], "reading_log_probs": [-220.72756958007812, -221.27059936523438, -218.8629150390625, -216.75265502929688], "reading_logprob_std": 2.046006796519524, "tokenization_path_logprob_std": 268.40160378169065, "trajectory_balance_rmse": 4123.350268263784} {"canonical_length": 8, "canonical_nll": 1.4194707870483398, "generation_rouge_l_f1": 0.07407407407407407, "generation_token_f1": 0.08641975308641975, "prompt_id": "b1a346c0437945a9", "reading_canonical_log_prob": -11.355766296386719, "reading_distances": [28, 31, 24, 31], "reading_log_probs": [-11.69597339630127, -10.643461227416992, -11.275446891784668, -11.416208267211914], "reading_logprob_std": 0.4452761845561937, "tokenization_path_logprob_std": 48.30894713998304, "trajectory_balance_rmse": 158.5362201680607} {"canonical_length": 21, "canonical_nll": 2.3050535292852494, "generation_rouge_l_f1": 0.12435233160621761, "generation_token_f1": 0.1398963730569948, "prompt_id": "c2ce9d8e95420539", "reading_canonical_log_prob": -48.406124114990234, "reading_distances": [19, 16, 19, 22], "reading_log_probs": [-59.74681091308594, -54.19070053100586, -57.070526123046875, -57.667724609375], "reading_logprob_std": 2.2930184006336973, "tokenization_path_logprob_std": 47.26373294750342, "trajectory_balance_rmse": 321.3822262727576} {"canonical_length": 86, "canonical_nll": 1.2334686989007995, "generation_rouge_l_f1": 0.23529411764705882, "generation_token_f1": 0.34117647058823525, "prompt_id": "ecdc1eebf63862d2", "reading_canonical_log_prob": -106.07830810546875, "reading_distances": [32, 35, 28, 36], "reading_log_probs": [-105.76502990722656, -106.02392578125, -105.71057891845703, -107.11508178710938], "reading_logprob_std": 0.6553668219489308, "tokenization_path_logprob_std": 143.84871419356583, "trajectory_balance_rmse": 2027.5288187686813} {"canonical_length": 149, "canonical_nll": 0.9228000512859166, "generation_rouge_l_f1": 0.2934782608695652, "generation_token_f1": 0.483695652173913, "prompt_id": "a473b04dd30c4a84", "reading_canonical_log_prob": -137.49720764160156, "reading_distances": [50, 50, 49, 47], "reading_log_probs": [-134.18087768554688, -136.0030059814453, -133.03721618652344, -135.8372802734375], "reading_logprob_std": 1.4152595666819199, "tokenization_path_logprob_std": 78.9817906074487, "trajectory_balance_rmse": 3411.7891594897956} {"canonical_length": 147, "canonical_nll": 1.4214095550329506, "generation_rouge_l_f1": 0.2210796915167095, "generation_token_f1": 0.3701799485861182, "prompt_id": "30cb3266dd228382", "reading_canonical_log_prob": -208.94720458984375, "reading_distances": [17, 20, 16, 12], "reading_log_probs": [-207.96856689453125, -207.01742553710938, -207.62271118164062, -209.14901733398438], "reading_logprob_std": 0.8970978238466746, "tokenization_path_logprob_std": 246.78578533182576, "trajectory_balance_rmse": 2513.8724433805455} {"canonical_length": 151, "canonical_nll": 1.1104578687655215, "generation_rouge_l_f1": 0.2961038961038961, "generation_token_f1": 0.5246753246753247, "prompt_id": "d49ee101a0892105", "reading_canonical_log_prob": -167.67913818359375, "reading_distances": [21, 17, 24, 23], "reading_log_probs": [-163.16835021972656, -164.79940795898438, -165.55999755859375, -166.40011596679688], "reading_logprob_std": 1.3745059049027704, "tokenization_path_logprob_std": 175.19574591382815, "trajectory_balance_rmse": 2521.5634874272555} {"canonical_length": 18, "canonical_nll": 0.750913937886556, "generation_rouge_l_f1": 0.3018867924528302, "generation_token_f1": 0.3018867924528302, "prompt_id": "c17572e82dc23485", "reading_canonical_log_prob": -13.516450881958008, "reading_distances": [26, 29, 24, 21], "reading_log_probs": [-13.326277732849121, -13.320980072021484, -13.975603103637695, -13.408895492553711], "reading_logprob_std": 0.31436379505927564, "tokenization_path_logprob_std": 54.31908315177698, "trajectory_balance_rmse": 483.6309803465994} {"canonical_length": 267, "canonical_nll": 1.3318249420280315, "generation_rouge_l_f1": 0.23391812865497075, "generation_token_f1": 0.530214424951267, "prompt_id": "50c4d271f8ba07ce", "reading_canonical_log_prob": -355.5972595214844, "reading_distances": [13, 16, 12, 13], "reading_log_probs": [-352.76373291015625, -353.45977783203125, -354.74298095703125, -351.072021484375], "reading_logprob_std": 1.529916201357459, "tokenization_path_logprob_std": 137.83498385157372, "trajectory_balance_rmse": 5174.815164194737} {"canonical_length": 27, "canonical_nll": 1.956207557960793, "generation_rouge_l_f1": 0.28070175438596495, "generation_token_f1": 0.31578947368421056, "prompt_id": "71a578aa99112e47", "reading_canonical_log_prob": -52.817604064941406, "reading_distances": [10, 13, 11, 16], "reading_log_probs": [-50.693206787109375, -50.802982330322266, -51.58185577392578, -51.54026794433594], "reading_logprob_std": 0.4718071428994959, "tokenization_path_logprob_std": 39.67699424547206, "trajectory_balance_rmse": 715.9466726864373} {"canonical_length": 260, "canonical_nll": 1.1082855224609376, "generation_rouge_l_f1": 0.25047438330170774, "generation_token_f1": 0.48956356736242884, "prompt_id": "c22b91c04ec53dee", "reading_canonical_log_prob": -288.15423583984375, "reading_distances": [38, 39, 38, 43], "reading_log_probs": [-271.052978515625, -271.27557373046875, -271.63385009765625, -276.02734375], "reading_logprob_std": 2.3654058259353614, "tokenization_path_logprob_std": 143.81101379575702, "trajectory_balance_rmse": 5252.312358557035} {"canonical_length": 42, "canonical_nll": 1.7210070292154949, "generation_rouge_l_f1": 0.14814814814814814, "generation_token_f1": 0.18518518518518517, "prompt_id": "ce56aa7776fbfa4a", "reading_canonical_log_prob": -72.28229522705078, "reading_distances": [18, 19, 16, 16], "reading_log_probs": [-74.50456237792969, -72.71733856201172, -71.8847885131836, -71.6365737915039], "reading_logprob_std": 1.2976114346332863, "tokenization_path_logprob_std": 65.70365251712278, "trajectory_balance_rmse": 869.7442580253269} {"canonical_length": 92, "canonical_nll": 1.3511785424273948, "generation_rouge_l_f1": 0.27491408934707906, "generation_token_f1": 0.28865979381443296, "prompt_id": "61cafca95080c624", "reading_canonical_log_prob": -124.30842590332031, "reading_distances": [28, 27, 23, 27], "reading_log_probs": [-120.55134582519531, -123.2254638671875, -126.23487854003906, -126.31021118164062], "reading_logprob_std": 2.756747771241867, "tokenization_path_logprob_std": 19.44009491709885, "trajectory_balance_rmse": 894.5734938830199} {"canonical_length": 9, "canonical_nll": 2.4150043063693576, "generation_rouge_l_f1": 0.6153846153846153, "generation_token_f1": 0.7692307692307692, "prompt_id": "33cbcf32b9a54c98", "reading_canonical_log_prob": -21.73503875732422, "reading_distances": [24, 22, 17, 21], "reading_log_probs": [-19.695964813232422, -18.658721923828125, -19.131589889526367, -18.621719360351562], "reading_logprob_std": 0.5027703599699386, "tokenization_path_logprob_std": 22.12869897700022, "trajectory_balance_rmse": 161.68913683429278} {"canonical_length": 13, "canonical_nll": 0.5151723348177396, "generation_rouge_l_f1": 0.208, "generation_token_f1": 0.208, "prompt_id": "5e0db48c41c73148", "reading_canonical_log_prob": -6.697240352630615, "reading_distances": [39, 37, 34, 28], "reading_log_probs": [-6.578372955322266, -6.6194167137146, -6.439378261566162, -6.872227668762207], "reading_logprob_std": 0.18051789850322494, "tokenization_path_logprob_std": 31.979032363164592, "trajectory_balance_rmse": 261.21885196308415} {"canonical_length": 45, "canonical_nll": 1.580251736111111, "generation_rouge_l_f1": 0.28767123287671237, "generation_token_f1": 0.3424657534246575, "prompt_id": "c70b50c5f56eb753", "reading_canonical_log_prob": -71.111328125, "reading_distances": [26, 33, 32, 25], "reading_log_probs": [-82.6158447265625, -82.40104675292969, -87.10693359375, -83.93958282470703], "reading_logprob_std": 2.1701166449420506, "tokenization_path_logprob_std": 108.51627632022434, "trajectory_balance_rmse": 883.0927512253505} {"canonical_length": 28, "canonical_nll": 0.9599430220467704, "generation_rouge_l_f1": 0.14285714285714285, "generation_token_f1": 0.2857142857142857, "prompt_id": "a944937cb2c4cc95", "reading_canonical_log_prob": -26.87840461730957, "reading_distances": [16, 13, 12, 14], "reading_log_probs": [-22.868595123291016, -23.525896072387695, -23.722816467285156, -23.164691925048828], "reading_logprob_std": 0.3797105301912715, "tokenization_path_logprob_std": 68.9627836377673, "trajectory_balance_rmse": 624.1390493441381} {"canonical_length": 158, "canonical_nll": 1.2741001950034612, "generation_rouge_l_f1": 0.23471882640586797, "generation_token_f1": 0.38141809290953543, "prompt_id": "b725cba739a4a837", "reading_canonical_log_prob": -201.30783081054688, "reading_distances": [35, 37, 30, 31], "reading_log_probs": [-197.1116943359375, -200.33297729492188, -198.56893920898438, -199.4112548828125], "reading_logprob_std": 1.368061386860006, "tokenization_path_logprob_std": 138.93068772655434, "trajectory_balance_rmse": 2848.592377625262} {"canonical_length": 501, "canonical_nll": 0.6368123128742516, "generation_rouge_l_f1": 0.18262150220913106, "generation_token_f1": 0.35346097201767296, "prompt_id": "c3741a95aac9caea", "reading_canonical_log_prob": -319.04296875, "reading_distances": [33, 35, 25, 30], "reading_log_probs": [-317.8084716796875, -321.2020263671875, -319.99420166015625, -320.89971923828125], "reading_logprob_std": 1.533504500604226, "tokenization_path_logprob_std": 113.81983311136932, "trajectory_balance_rmse": 5120.940996306196} {"canonical_length": 14, "canonical_nll": 0.4622558185032436, "generation_rouge_l_f1": 0.7586206896551724, "generation_token_f1": 0.7586206896551724, "prompt_id": "77d6fb3ebd2588d3", "reading_canonical_log_prob": -6.47158145904541, "reading_distances": [23, 18, 19, 26], "reading_log_probs": [-7.312839031219482, -7.448579788208008, -6.463384628295898, -7.071572303771973], "reading_logprob_std": 0.43597059959773815, "tokenization_path_logprob_std": 43.46037415881703, "trajectory_balance_rmse": 211.91895834791632} {"canonical_length": 242, "canonical_nll": 1.4133284387509686, "generation_rouge_l_f1": 0.21386138613861383, "generation_token_f1": 0.38811881188118813, "prompt_id": "8d10c80bdd9627ab", "reading_canonical_log_prob": -342.0254821777344, "reading_distances": [16, 20, 18, 18], "reading_log_probs": [-342.0425720214844, -342.97344970703125, -340.8661804199219, -341.355712890625], "reading_logprob_std": 0.9137607375905438, "tokenization_path_logprob_std": 364.1313210766011, "trajectory_balance_rmse": 5266.360435027904} {"canonical_length": 233, "canonical_nll": 1.3066605334629828, "generation_rouge_l_f1": 0.19784946236559137, "generation_token_f1": 0.4559139784946236, "prompt_id": "a741547771fa6f03", "reading_canonical_log_prob": -304.451904296875, "reading_distances": [22, 28, 26, 20], "reading_log_probs": [-299.6082458496094, -300.7134094238281, -301.3382263183594, -299.93646240234375], "reading_logprob_std": 0.7789378428161992, "tokenization_path_logprob_std": 135.7426583652883, "trajectory_balance_rmse": 4865.634563002624} {"canonical_length": 177, "canonical_nll": 0.9092112395722988, "generation_rouge_l_f1": 0.22737819025522044, "generation_token_f1": 0.4176334106728538, "prompt_id": "51a6d546f1967143", "reading_canonical_log_prob": -160.93038940429688, "reading_distances": [27, 33, 28, 28], "reading_log_probs": [-157.18215942382812, -157.54629516601562, -158.44827270507812, -159.16542053222656], "reading_logprob_std": 0.8952836488252645, "tokenization_path_logprob_std": 238.6225515100916, "trajectory_balance_rmse": 3053.9395690626657} {"canonical_length": 163, "canonical_nll": 1.2388554672522047, "generation_rouge_l_f1": 0.27338129496402874, "generation_token_f1": 0.5179856115107914, "prompt_id": "715a0325621f71c4", "reading_canonical_log_prob": -201.93344116210938, "reading_distances": [46, 42, 32, 36], "reading_log_probs": [-183.95513916015625, -187.19871520996094, -192.43687438964844, -188.24050903320312], "reading_logprob_std": 3.499501645586574, "tokenization_path_logprob_std": 183.89656358498755, "trajectory_balance_rmse": 2722.5470946345986} {"canonical_length": 264, "canonical_nll": 1.2865608677719578, "generation_rouge_l_f1": 0.36399217221135033, "generation_token_f1": 0.5362035225048924, "prompt_id": "0857b1f32ec14a13", "reading_canonical_log_prob": -339.6520690917969, "reading_distances": [17, 23, 21, 17], "reading_log_probs": [-353.4891052246094, -354.829833984375, -348.19189453125, -350.59674072265625], "reading_logprob_std": 2.9719293831245017, "tokenization_path_logprob_std": 92.6468756900304, "trajectory_balance_rmse": 4317.445099831981} {"canonical_length": 2, "canonical_nll": 3.3173654079437256, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "7c6d4e136a035811", "reading_canonical_log_prob": -6.634730815887451, "reading_distances": [65, 71, 83, 67], "reading_log_probs": [-6.851675033569336, -7.99647331237793, -6.874696254730225, -6.78046989440918], "reading_logprob_std": 0.5818139689930573, "tokenization_path_logprob_std": 0.0, "trajectory_balance_rmse": 6.7354207038879395} {"canonical_length": 13, "canonical_nll": 0.952919446505033, "generation_rouge_l_f1": 0.5454545454545454, "generation_token_f1": 0.7272727272727272, "prompt_id": "d3b27b0b1b980505", "reading_canonical_log_prob": -12.38795280456543, "reading_distances": [31, 27, 34, 37], "reading_log_probs": [-11.067546844482422, -11.23595905303955, -11.605216979980469, -10.770669937133789], "reading_logprob_std": 0.34819796685071674, "tokenization_path_logprob_std": 16.74886124266492, "trajectory_balance_rmse": 225.3051198409227} {"canonical_length": 131, "canonical_nll": 0.4682924401669102, "generation_rouge_l_f1": 0.28571428571428564, "generation_token_f1": 0.4314868804664724, "prompt_id": "37ec341c729bc100", "reading_canonical_log_prob": -61.346309661865234, "reading_distances": [9, 8, 14, 7], "reading_log_probs": [-58.71323013305664, -60.411338806152344, -62.64916229248047, -59.2021484375], "reading_logprob_std": 1.7551371256419446, "tokenization_path_logprob_std": 105.73058170920423, "trajectory_balance_rmse": 1285.1550187252558} {"canonical_length": 12, "canonical_nll": 3.6468734741210938, "generation_rouge_l_f1": 0.04580152671755726, "generation_token_f1": 0.04580152671755726, "prompt_id": "b0f40cc1f3136930", "reading_canonical_log_prob": -43.762481689453125, "reading_distances": [10, 12, 9, 13], "reading_log_probs": [-43.89530944824219, -44.30292510986328, -42.33062744140625, -44.0191650390625], "reading_logprob_std": 0.8874757089151338, "tokenization_path_logprob_std": 43.83592593742948, "trajectory_balance_rmse": 296.3039506328279} {"canonical_length": 264, "canonical_nll": 1.2548405040394177, "generation_rouge_l_f1": 0.2596899224806202, "generation_token_f1": 0.4728682170542635, "prompt_id": "dc37089a4982f612", "reading_canonical_log_prob": -331.27789306640625, "reading_distances": [16, 16, 14, 14], "reading_log_probs": [-337.8214111328125, -337.45806884765625, -331.60040283203125, -333.8057861328125], "reading_logprob_std": 2.9926763623652444, "tokenization_path_logprob_std": 238.48718518037703, "trajectory_balance_rmse": 5220.196116203433} {"canonical_length": 3, "canonical_nll": 1.0865803559621174, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.6666666666666666, "prompt_id": "9ba6dc4bdecfc908", "reading_canonical_log_prob": -3.2597410678863525, "reading_distances": [24, 25, 31, 15], "reading_log_probs": [-3.5126760005950928, -3.470046281814575, -3.3007891178131104, -3.022104024887085], "reading_logprob_std": 0.22254959025925167, "tokenization_path_logprob_std": 13.068800568580627, "trajectory_balance_rmse": 14.519914710103341} {"canonical_length": 261, "canonical_nll": 0.9789902259563578, "generation_rouge_l_f1": 0.39925373134328357, "generation_token_f1": 0.5447761194029851, "prompt_id": "5cf08716fd03e235", "reading_canonical_log_prob": -255.51644897460938, "reading_distances": [30, 30, 30, 31], "reading_log_probs": [-252.88095092773438, -252.89114379882812, -253.10568237304688, -255.33493041992188], "reading_logprob_std": 1.192346781525507, "tokenization_path_logprob_std": 94.50879491477421, "trajectory_balance_rmse": 4304.180303495873} {"canonical_length": 102, "canonical_nll": 0.8467885933670343, "generation_rouge_l_f1": 0.3636363636363636, "generation_token_f1": 0.5252525252525253, "prompt_id": "8cfc1971edf74b1e", "reading_canonical_log_prob": -86.3724365234375, "reading_distances": [38, 31, 32, 39], "reading_log_probs": [-78.44758605957031, -82.65608215332031, -78.23260498046875, -77.99382019042969], "reading_logprob_std": 2.2234435527050382, "tokenization_path_logprob_std": 97.2220186041486, "trajectory_balance_rmse": 1840.8298999523622} {"canonical_length": 42, "canonical_nll": 2.9068607148670016, "generation_rouge_l_f1": 0.4166666666666667, "generation_token_f1": 0.4444444444444445, "prompt_id": "73cb687de2d9ad9d", "reading_canonical_log_prob": -122.08815002441406, "reading_distances": [24, 27, 23, 23], "reading_log_probs": [-118.33857727050781, -119.60458374023438, -120.36018371582031, -118.5568618774414], "reading_logprob_std": 0.9424230847659322, "tokenization_path_logprob_std": 59.033621991724786, "trajectory_balance_rmse": 834.8919665648036} {"canonical_length": 50, "canonical_nll": 0.9468081665039062, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "5b7c88eabb38a156", "reading_canonical_log_prob": -47.34040832519531, "reading_distances": [29, 26, 30, 26], "reading_log_probs": [-45.306602478027344, -45.47959899902344, -46.31281280517578, -45.07713317871094], "reading_logprob_std": 0.538374970057167, "tokenization_path_logprob_std": 107.07074298502312, "trajectory_balance_rmse": 939.9670955704188} {"canonical_length": 297, "canonical_nll": 1.9065574363425926, "generation_rouge_l_f1": 0.2656546489563567, "generation_token_f1": 0.48956356736242884, "prompt_id": "7564700da05daae9", "reading_canonical_log_prob": -566.24755859375, "reading_distances": [25, 22, 21, 30], "reading_log_probs": [-557.8652954101562, -561.114013671875, -559.1158447265625, -554.82763671875], "reading_logprob_std": 2.633839980448772, "tokenization_path_logprob_std": 298.74359941635635, "trajectory_balance_rmse": 5895.579978719339} {"canonical_length": 240, "canonical_nll": 1.3649037679036458, "generation_rouge_l_f1": 0.2362204724409449, "generation_token_f1": 0.48031496062992124, "prompt_id": "ca2d7d9e3b473a6f", "reading_canonical_log_prob": -327.576904296875, "reading_distances": [17, 14, 21, 13], "reading_log_probs": [-329.1158142089844, -330.7947082519531, -330.155517578125, -330.66302490234375], "reading_logprob_std": 0.7625067288868556, "tokenization_path_logprob_std": 155.67113715183322, "trajectory_balance_rmse": 5400.9861115332} {"canonical_length": 282, "canonical_nll": 1.195404160952737, "generation_rouge_l_f1": 0.2840690978886757, "generation_token_f1": 0.4261036468330135, "prompt_id": "9e92f8d560ad3563", "reading_canonical_log_prob": -337.1039733886719, "reading_distances": [22, 25, 24, 15], "reading_log_probs": [-337.27264404296875, -334.27105712890625, -337.3348388671875, -340.1422424316406], "reading_logprob_std": 2.397691203423149, "tokenization_path_logprob_std": 160.28682506239397, "trajectory_balance_rmse": 4037.0846666236275} {"canonical_length": 16, "canonical_nll": 1.866213083267212, "generation_rouge_l_f1": 0.26666666666666666, "generation_token_f1": 0.3333333333333333, "prompt_id": "8efbf41d41230763", "reading_canonical_log_prob": -29.85940933227539, "reading_distances": [37, 36, 42, 28], "reading_log_probs": [-28.128490447998047, -26.62262535095215, -26.638423919677734, -27.911518096923828], "reading_logprob_std": 0.8071177031048187, "tokenization_path_logprob_std": 45.05154358791447, "trajectory_balance_rmse": 410.1008351255814} {"canonical_length": 39, "canonical_nll": 1.737320337540064, "generation_rouge_l_f1": 0.35897435897435903, "generation_token_f1": 0.4615384615384615, "prompt_id": "3f9b43f9811260d6", "reading_canonical_log_prob": -67.7554931640625, "reading_distances": [32, 31, 28, 35], "reading_log_probs": [-72.90729522705078, -74.18031311035156, -72.47163391113281, -72.18046569824219], "reading_logprob_std": 0.8823458167058807, "tokenization_path_logprob_std": 88.28880312269447, "trajectory_balance_rmse": 702.3076011033224} {"canonical_length": 281, "canonical_nll": 0.9498597277441059, "generation_rouge_l_f1": 0.2548672566371681, "generation_token_f1": 0.4814159292035398, "prompt_id": "9e7206bd2d909c4b", "reading_canonical_log_prob": -266.91058349609375, "reading_distances": [28, 37, 33, 32], "reading_log_probs": [-267.81396484375, -269.07867431640625, -268.5233459472656, -267.600830078125], "reading_logprob_std": 0.67648520232518, "tokenization_path_logprob_std": 59.25825457004653, "trajectory_balance_rmse": 4178.8908146335825} {"canonical_length": 193, "canonical_nll": 1.3485180158071568, "generation_rouge_l_f1": 0.29247311827956984, "generation_token_f1": 0.4817204301075269, "prompt_id": "59cc1c6a894af7a8", "reading_canonical_log_prob": -260.26397705078125, "reading_distances": [20, 23, 27, 23], "reading_log_probs": [-253.94537353515625, -253.86788940429688, -255.4477081298828, -257.39990234375], "reading_logprob_std": 1.657779317171122, "tokenization_path_logprob_std": 133.3596286568467, "trajectory_balance_rmse": 3492.0745918773605} {"canonical_length": 85, "canonical_nll": 1.1176860136144302, "generation_rouge_l_f1": 0.24719101123595508, "generation_token_f1": 0.33707865168539325, "prompt_id": "bb9e06240071dbbd", "reading_canonical_log_prob": -95.00331115722656, "reading_distances": [25, 25, 20, 24], "reading_log_probs": [-98.28909301757812, -98.4256362915039, -99.65955352783203, -98.56622314453125], "reading_logprob_std": 0.62658431991002, "tokenization_path_logprob_std": 34.50595277439043, "trajectory_balance_rmse": 1632.880215654798} {"canonical_length": 59, "canonical_nll": 0.9659504647982322, "generation_rouge_l_f1": 0.19649122807017544, "generation_token_f1": 0.2666666666666666, "prompt_id": "fbd2f42b1f0643a2", "reading_canonical_log_prob": -56.9910774230957, "reading_distances": [17, 17, 13, 17], "reading_log_probs": [-55.34618377685547, -54.79632568359375, -55.79488754272461, -56.051063537597656], "reading_logprob_std": 0.5505795036796023, "tokenization_path_logprob_std": 60.49667494931418, "trajectory_balance_rmse": 893.2206101420549} {"canonical_length": 73, "canonical_nll": 0.7768445472194724, "generation_rouge_l_f1": 0.3668122270742358, "generation_token_f1": 0.48034934497816595, "prompt_id": "1a0ff319dcddb777", "reading_canonical_log_prob": -56.709651947021484, "reading_distances": [48, 48, 55, 41], "reading_log_probs": [-55.48294448852539, -55.69984817504883, -55.581443786621094, -54.32569885253906], "reading_logprob_std": 0.6373884269807464, "tokenization_path_logprob_std": 104.1348946564496, "trajectory_balance_rmse": 641.1947130565869} {"canonical_length": 15, "canonical_nll": 0.7931293487548828, "generation_rouge_l_f1": 0.3636363636363637, "generation_token_f1": 0.5454545454545455, "prompt_id": "9f8d22c249bf7888", "reading_canonical_log_prob": -11.896940231323242, "reading_distances": [20, 21, 24, 21], "reading_log_probs": [-11.846086502075195, -10.981399536132812, -11.534513473510742, -11.95591926574707], "reading_logprob_std": 0.43684771469727307, "tokenization_path_logprob_std": 34.866696981551144, "trajectory_balance_rmse": 345.3132562683708} {"canonical_length": 168, "canonical_nll": 1.1580198378789992, "generation_rouge_l_f1": 0.21836228287841192, "generation_token_f1": 0.37717121588089336, "prompt_id": "9edad123b1e31f1d", "reading_canonical_log_prob": -194.54733276367188, "reading_distances": [28, 29, 28, 31], "reading_log_probs": [-191.72946166992188, -192.65826416015625, -192.55258178710938, -186.05279541015625], "reading_logprob_std": 3.1577332288522335, "tokenization_path_logprob_std": 234.61946321417747, "trajectory_balance_rmse": 3191.9894667547296} {"canonical_length": 82, "canonical_nll": 1.1068695812690548, "generation_rouge_l_f1": 0.15105740181268884, "generation_token_f1": 0.26586102719033233, "prompt_id": "e027ef71856c6f38", "reading_canonical_log_prob": -90.7633056640625, "reading_distances": [29, 27, 25, 26], "reading_log_probs": [-96.96324157714844, -94.87476348876953, -100.04627227783203, -98.7273941040039], "reading_logprob_std": 2.2417582127874356, "tokenization_path_logprob_std": 86.29137734939951, "trajectory_balance_rmse": 2073.852317347237} {"canonical_length": 27, "canonical_nll": 2.01760412145544, "generation_rouge_l_f1": 0.17910447761194032, "generation_token_f1": 0.17910447761194032, "prompt_id": "59a16e60dba35dc9", "reading_canonical_log_prob": -54.475311279296875, "reading_distances": [15, 14, 18, 15], "reading_log_probs": [-52.980201721191406, -54.22726058959961, -52.85799026489258, -54.97353744506836], "reading_logprob_std": 1.0186121689211751, "tokenization_path_logprob_std": 67.76907699781624, "trajectory_balance_rmse": 640.196339978752} {"canonical_length": 16, "canonical_nll": 0.31080782413482666, "generation_rouge_l_f1": 0.2962962962962963, "generation_token_f1": 0.4444444444444445, "prompt_id": "ca87800b07a107f6", "reading_canonical_log_prob": -4.972925186157227, "reading_distances": [27, 32, 30, 26], "reading_log_probs": [-4.695237159729004, -6.146726608276367, -6.2385759353637695, -5.338754653930664], "reading_logprob_std": 0.7287983418718141, "tokenization_path_logprob_std": 9.247036534550949, "trajectory_balance_rmse": 237.6831398988236} {"canonical_length": 100, "canonical_nll": 1.419356689453125, "generation_rouge_l_f1": 0.20786516853932585, "generation_token_f1": 0.34831460674157305, "prompt_id": "911e3e27940b6a5a", "reading_canonical_log_prob": -141.9356689453125, "reading_distances": [29, 27, 23, 30], "reading_log_probs": [-138.00213623046875, -138.51734924316406, -139.39822387695312, -137.22317504882812], "reading_logprob_std": 0.9130055703593235, "tokenization_path_logprob_std": 225.72602648706558, "trajectory_balance_rmse": 1977.5562068786462} {"canonical_length": 190, "canonical_nll": 2.1680641575863486, "generation_rouge_l_f1": 0.22488038277511962, "generation_token_f1": 0.4354066985645933, "prompt_id": "1e23f4d5f326fadb", "reading_canonical_log_prob": -411.93218994140625, "reading_distances": [23, 25, 31, 29], "reading_log_probs": [-402.4700927734375, -405.216796875, -404.83203125, -410.228759765625], "reading_logprob_std": 3.2623169879551637, "tokenization_path_logprob_std": 208.61777388798816, "trajectory_balance_rmse": 4176.198494500893} {"canonical_length": 151, "canonical_nll": 1.457312577607616, "generation_rouge_l_f1": 0.19662921348314608, "generation_token_f1": 0.5393258426966292, "prompt_id": "75c88d19daf68f00", "reading_canonical_log_prob": -220.05419921875, "reading_distances": [18, 19, 24, 20], "reading_log_probs": [-218.08970642089844, -217.23402404785156, -217.43833923339844, -218.8262481689453], "reading_logprob_std": 0.7189215791331379, "tokenization_path_logprob_std": 184.12813005728148, "trajectory_balance_rmse": 3588.309067648761} {"canonical_length": 165, "canonical_nll": 1.3062341863458806, "generation_rouge_l_f1": 0.2252559726962457, "generation_token_f1": 0.49146757679180886, "prompt_id": "0efac862acbbc3cf", "reading_canonical_log_prob": -215.5286407470703, "reading_distances": [28, 33, 24, 31], "reading_log_probs": [-219.0980224609375, -218.5521697998047, -219.00404357910156, -217.68798828125], "reading_logprob_std": 0.6440745752767169, "tokenization_path_logprob_std": 193.7108651950499, "trajectory_balance_rmse": 3748.5722067216775} {"canonical_length": 225, "canonical_nll": 0.9496424018012153, "generation_rouge_l_f1": 0.25806451612903225, "generation_token_f1": 0.3988269794721408, "prompt_id": "867d14d7a7b70c9c", "reading_canonical_log_prob": -213.66954040527344, "reading_distances": [14, 16, 20, 20], "reading_log_probs": [-211.70782470703125, -207.41986083984375, -203.9851531982422, -206.67369079589844], "reading_logprob_std": 3.200924532146325, "tokenization_path_logprob_std": 40.608874919562346, "trajectory_balance_rmse": 3587.414360727514} {"canonical_length": 263, "canonical_nll": 1.523801158136288, "generation_rouge_l_f1": 0.11854684512428298, "generation_token_f1": 0.20650095602294455, "prompt_id": "67b31fbc0903ce67", "reading_canonical_log_prob": -400.75970458984375, "reading_distances": [27, 20, 20, 14], "reading_log_probs": [-399.9112854003906, -399.47515869140625, -396.00738525390625, -398.5181579589844], "reading_logprob_std": 1.7468240113670412, "tokenization_path_logprob_std": 174.1414008676488, "trajectory_balance_rmse": 5227.0382507065415} {"canonical_length": 38, "canonical_nll": 0.4554180346037212, "generation_rouge_l_f1": 0.2786885245901639, "generation_token_f1": 0.32786885245901637, "prompt_id": "c50601d0ada030f4", "reading_canonical_log_prob": -17.305885314941406, "reading_distances": [28, 26, 24, 26], "reading_log_probs": [-15.461400985717773, -15.874610900878906, -16.074321746826172, -18.090574264526367], "reading_logprob_std": 1.171701707259411, "tokenization_path_logprob_std": 46.89814941419548, "trajectory_balance_rmse": 240.13234176326782} {"canonical_length": 35, "canonical_nll": 0.4257843290056501, "generation_rouge_l_f1": 0.6176470588235294, "generation_token_f1": 0.6176470588235294, "prompt_id": "d0d4fbc6a94b1aa0", "reading_canonical_log_prob": -14.902451515197754, "reading_distances": [21, 19, 20, 21], "reading_log_probs": [-11.950883865356445, -11.782159805297852, -11.559150695800781, -10.954231262207031], "reading_logprob_std": 0.43554232022271516, "tokenization_path_logprob_std": 78.95759966443123, "trajectory_balance_rmse": 575.8009243736956} {"canonical_length": 74, "canonical_nll": 0.4650790755813186, "generation_rouge_l_f1": 0.32142857142857145, "generation_token_f1": 0.5357142857142857, "prompt_id": "a83eef42db047836", "reading_canonical_log_prob": -34.41585159301758, "reading_distances": [20, 25, 26, 29], "reading_log_probs": [-35.19379806518555, -37.38697814941406, -34.122737884521484, -33.403480529785156], "reading_logprob_std": 1.7369344308508092, "tokenization_path_logprob_std": 76.92837899782427, "trajectory_balance_rmse": 1170.8165920364818} {"canonical_length": 107, "canonical_nll": 1.1787382464542566, "generation_rouge_l_f1": 0.2835820895522388, "generation_token_f1": 0.4552238805970149, "prompt_id": "8e875ab98664b402", "reading_canonical_log_prob": -126.12499237060547, "reading_distances": [18, 17, 18, 18], "reading_log_probs": [-127.5772705078125, -127.33575439453125, -124.4631576538086, -127.94917297363281], "reading_logprob_std": 1.5988208556770993, "tokenization_path_logprob_std": 103.21104350034172, "trajectory_balance_rmse": 1992.434500985658} {"canonical_length": 31, "canonical_nll": 0.7083662709882182, "generation_rouge_l_f1": 0.40366972477064217, "generation_token_f1": 0.47706422018348615, "prompt_id": "54f5b26b0d406b0a", "reading_canonical_log_prob": -21.959354400634766, "reading_distances": [29, 31, 27, 34], "reading_log_probs": [-28.29547691345215, -26.071664810180664, -25.576099395751953, -27.496417999267578], "reading_logprob_std": 1.2563830284977877, "tokenization_path_logprob_std": 69.9058423595817, "trajectory_balance_rmse": 627.4071289054853} {"canonical_length": 262, "canonical_nll": 1.244416564475489, "generation_rouge_l_f1": 0.2518796992481203, "generation_token_f1": 0.48120300751879697, "prompt_id": "652a0c6a273ec7b9", "reading_canonical_log_prob": -326.0371398925781, "reading_distances": [16, 17, 13, 13], "reading_log_probs": [-326.15325927734375, -324.3382568359375, -324.9853515625, -324.2265930175781], "reading_logprob_std": 0.8839871837395331, "tokenization_path_logprob_std": 223.348499705725, "trajectory_balance_rmse": 4186.261759002378} {"canonical_length": 304, "canonical_nll": 1.6498744362278988, "generation_rouge_l_f1": 0.15042735042735042, "generation_token_f1": 0.31111111111111117, "prompt_id": "13a435b70dd47ea2", "reading_canonical_log_prob": -501.56182861328125, "reading_distances": [23, 14, 17, 11], "reading_log_probs": [-497.88531494140625, -499.2666015625, -499.8634338378906, -497.415771484375], "reading_logprob_std": 1.1479768397830667, "tokenization_path_logprob_std": 469.7786681584241, "trajectory_balance_rmse": 6659.465881435511} {"canonical_length": 15, "canonical_nll": 0.39986708958943684, "generation_rouge_l_f1": 0.3529411764705882, "generation_token_f1": 0.4705882352941177, "prompt_id": "1f8b04daa461e185", "reading_canonical_log_prob": -5.998006343841553, "reading_distances": [24, 18, 25, 23], "reading_log_probs": [-8.17667293548584, -6.106766223907471, -6.6992692947387695, -6.560003757476807], "reading_logprob_std": 0.8970662175532564, "tokenization_path_logprob_std": 23.514733468292672, "trajectory_balance_rmse": 199.86295901449628} {"canonical_length": 18, "canonical_nll": 0.30403802129957413, "generation_rouge_l_f1": 0.27826086956521745, "generation_token_f1": 0.27826086956521745, "prompt_id": "713c6185b310a4ca", "reading_canonical_log_prob": -5.472684383392334, "reading_distances": [14, 27, 24, 30], "reading_log_probs": [-6.086584568023682, -8.093681335449219, -6.369315147399902, -6.34406852722168], "reading_logprob_std": 0.9224014457111728, "tokenization_path_logprob_std": 61.147405214450444, "trajectory_balance_rmse": 389.4849542257629} {"canonical_length": 227, "canonical_nll": 0.8476707021570415, "generation_rouge_l_f1": 0.25523012552301255, "generation_token_f1": 0.49790794979079495, "prompt_id": "a7de6528410fa4b7", "reading_canonical_log_prob": -192.42124938964844, "reading_distances": [21, 28, 25, 25], "reading_log_probs": [-193.06741333007812, -189.2763671875, -190.922119140625, -189.1007537841797], "reading_logprob_std": 1.843127262919933, "tokenization_path_logprob_std": 96.0303839057831, "trajectory_balance_rmse": 3190.1373989201247} {"canonical_length": 154, "canonical_nll": 1.664507581042005, "generation_rouge_l_f1": 0.1990521327014218, "generation_token_f1": 0.3459715639810426, "prompt_id": "b7245c913ba925d6", "reading_canonical_log_prob": -256.33416748046875, "reading_distances": [22, 28, 21, 28], "reading_log_probs": [-254.87564086914062, -253.62918090820312, -254.8824462890625, -254.62144470214844], "reading_logprob_std": 0.5945381834724849, "tokenization_path_logprob_std": 154.38704948766932, "trajectory_balance_rmse": 2984.2018751748647} {"canonical_length": 9, "canonical_nll": 0.5288456281026205, "generation_rouge_l_f1": 0.8, "generation_token_f1": 0.8, "prompt_id": "402312ec99016f32", "reading_canonical_log_prob": -4.759610652923584, "reading_distances": [23, 21, 22, 20], "reading_log_probs": [-5.331528663635254, -5.399919509887695, -5.571817398071289, -5.896125316619873], "reading_logprob_std": 0.2520135705571323, "tokenization_path_logprob_std": 35.70092377949907, "trajectory_balance_rmse": 160.58018439905769} {"canonical_length": 290, "canonical_nll": 1.4732495538119612, "generation_rouge_l_f1": 0.25569176882662, "generation_token_f1": 0.5078809106830122, "prompt_id": "083ca3a9f1344c38", "reading_canonical_log_prob": -427.24237060546875, "reading_distances": [30, 35, 32, 31], "reading_log_probs": [-426.07733154296875, -426.12762451171875, -428.9877624511719, -430.0037536621094], "reading_logprob_std": 2.0026426806371767, "tokenization_path_logprob_std": 223.20241628050766, "trajectory_balance_rmse": 5724.805544776719} {"canonical_length": 58, "canonical_nll": 1.511004480822333, "generation_rouge_l_f1": 0.23333333333333336, "generation_token_f1": 0.35555555555555557, "prompt_id": "c06dd3bfbe8f1823", "reading_canonical_log_prob": -87.63825988769531, "reading_distances": [35, 29, 32, 34], "reading_log_probs": [-90.70014953613281, -92.99577331542969, -91.07005310058594, -91.72746276855469], "reading_logprob_std": 1.0087652155518514, "tokenization_path_logprob_std": 136.0462928729133, "trajectory_balance_rmse": 1124.6670143658844} {"canonical_length": 117, "canonical_nll": 0.8293024046808226, "generation_rouge_l_f1": 0.2598870056497175, "generation_token_f1": 0.4180790960451977, "prompt_id": "22135eb61f4e8a41", "reading_canonical_log_prob": -97.02838134765625, "reading_distances": [23, 16, 22, 21], "reading_log_probs": [-102.74291229248047, -100.20219421386719, -100.399169921875, -100.79627990722656], "reading_logprob_std": 1.1650174490520304, "tokenization_path_logprob_std": 123.28611498872145, "trajectory_balance_rmse": 2233.2553222120537} {"canonical_length": 21, "canonical_nll": 1.6462024507068453, "generation_rouge_l_f1": 0.1142857142857143, "generation_token_f1": 0.1714285714285714, "prompt_id": "b36f1ed6b55c0058", "reading_canonical_log_prob": -34.57025146484375, "reading_distances": [27, 26, 24, 28], "reading_log_probs": [-42.569305419921875, -44.10881042480469, -40.68760299682617, -44.37492370605469], "reading_logprob_std": 1.6966375477391344, "tokenization_path_logprob_std": 89.54475689799541, "trajectory_balance_rmse": 453.1346509212005} {"canonical_length": 25, "canonical_nll": 2.0669076538085935, "generation_rouge_l_f1": 0.25, "generation_token_f1": 0.3125, "prompt_id": "017529d62a606652", "reading_canonical_log_prob": -51.672691345214844, "reading_distances": [25, 19, 26, 27], "reading_log_probs": [-50.067604064941406, -49.7151985168457, -50.30529022216797, -48.8572998046875], "reading_logprob_std": 0.6341921010350244, "tokenization_path_logprob_std": 99.01286817960435, "trajectory_balance_rmse": 495.2055310932424} {"canonical_length": 231, "canonical_nll": 1.6730787929518398, "generation_rouge_l_f1": 0.1478599221789883, "generation_token_f1": 0.24513618677042798, "prompt_id": "47cd1ce3b4c7133a", "reading_canonical_log_prob": -386.481201171875, "reading_distances": [15, 12, 12, 9], "reading_log_probs": [-382.81549072265625, -382.5191650390625, -385.5066223144531, -385.81793212890625], "reading_logprob_std": 1.7380144715302857, "tokenization_path_logprob_std": 96.81206827578443, "trajectory_balance_rmse": 5065.826922733184} {"canonical_length": 34, "canonical_nll": 0.6922250074498794, "generation_rouge_l_f1": 0.2007722007722008, "generation_token_f1": 0.2007722007722008, "prompt_id": "549cddbffb2c458b", "reading_canonical_log_prob": -23.5356502532959, "reading_distances": [11, 13, 17, 10], "reading_log_probs": [-23.30674171447754, -23.780071258544922, -23.816333770751953, -23.87897491455078], "reading_logprob_std": 0.2623922909548477, "tokenization_path_logprob_std": 107.76386468669723, "trajectory_balance_rmse": 480.62116525295403} {"canonical_length": 356, "canonical_nll": 1.4386804130639923, "generation_rouge_l_f1": 0.22260273972602737, "generation_token_f1": 0.46917808219178087, "prompt_id": "6b4ba85e93ef4f52", "reading_canonical_log_prob": -512.1702270507812, "reading_distances": [29, 24, 24, 28], "reading_log_probs": [-504.572998046875, -500.0142517089844, -501.81927490234375, -505.1591796875], "reading_logprob_std": 2.4081826628592746, "tokenization_path_logprob_std": 254.05496724837877, "trajectory_balance_rmse": 7479.349909664867} {"canonical_length": 420, "canonical_nll": 1.3442022414434525, "generation_rouge_l_f1": 0.18096723868954756, "generation_token_f1": 0.34945397815912643, "prompt_id": "71f47b6572a134fd", "reading_canonical_log_prob": -564.56494140625, "reading_distances": [28, 37, 36, 36], "reading_log_probs": [-551.639404296875, -555.142333984375, -555.7252807617188, -555.2322998046875], "reading_logprob_std": 1.8811510667148177, "tokenization_path_logprob_std": 250.3189485415398, "trajectory_balance_rmse": 7384.99505270323} {"canonical_length": 144, "canonical_nll": 1.2272539138793945, "generation_rouge_l_f1": 0.2594594594594595, "generation_token_f1": 0.46486486486486484, "prompt_id": "c1009daeb59a17f6", "reading_canonical_log_prob": -176.7245635986328, "reading_distances": [18, 17, 20, 19], "reading_log_probs": [-178.82736206054688, -177.480712890625, -179.2262725830078, -179.17344665527344], "reading_logprob_std": 0.8168784566252717, "tokenization_path_logprob_std": 106.638818705216, "trajectory_balance_rmse": 2613.3838872153146} {"canonical_length": 19, "canonical_nll": 2.3337729604620683, "generation_rouge_l_f1": 0.15384615384615383, "generation_token_f1": 0.30769230769230765, "prompt_id": "231e10798735d9e9", "reading_canonical_log_prob": -44.3416862487793, "reading_distances": [17, 20, 16, 21], "reading_log_probs": [-43.138633728027344, -44.076416015625, -43.463687896728516, -43.14968490600586], "reading_logprob_std": 0.4395150865311429, "tokenization_path_logprob_std": 17.22949662386022, "trajectory_balance_rmse": 480.67889065824534} {"canonical_length": 133, "canonical_nll": 1.7821867519751526, "generation_rouge_l_f1": 0.112, "generation_token_f1": 0.18666666666666665, "prompt_id": "948eaf298aee4503", "reading_canonical_log_prob": -237.0308380126953, "reading_distances": [24, 19, 19, 19], "reading_log_probs": [-231.24522399902344, -229.843505859375, -231.17222595214844, -228.00135803222656], "reading_logprob_std": 1.5194905378669772, "tokenization_path_logprob_std": 192.86743451762447, "trajectory_balance_rmse": 2699.322608340481} {"canonical_length": 43, "canonical_nll": 0.44529240630393807, "generation_rouge_l_f1": 0.39560439560439553, "generation_token_f1": 0.6373626373626373, "prompt_id": "cc672b4406958c5b", "reading_canonical_log_prob": -19.147573471069336, "reading_distances": [44, 54, 40, 46], "reading_log_probs": [-28.8731689453125, -26.476943969726562, -23.05603790283203, -23.37124252319336], "reading_logprob_std": 2.7583122224716186, "tokenization_path_logprob_std": 34.772024094384, "trajectory_balance_rmse": 769.733106408078} {"canonical_length": 189, "canonical_nll": 1.0433454564008764, "generation_rouge_l_f1": 0.2300469483568075, "generation_token_f1": 0.403755868544601, "prompt_id": "f0c1c49e728adae3", "reading_canonical_log_prob": -197.19229125976562, "reading_distances": [17, 19, 17, 21], "reading_log_probs": [-182.5433349609375, -182.78067016601562, -183.63113403320312, -188.06314086914062], "reading_logprob_std": 2.58163969532911, "tokenization_path_logprob_std": 52.60950565709007, "trajectory_balance_rmse": 3571.882858434944} {"canonical_length": 202, "canonical_nll": 1.5852760843711324, "generation_rouge_l_f1": 0.2391304347826087, "generation_token_f1": 0.48695652173913045, "prompt_id": "91ba22c36f60a796", "reading_canonical_log_prob": -320.22576904296875, "reading_distances": [28, 31, 29, 23], "reading_log_probs": [-311.80364990234375, -318.7489318847656, -314.54815673828125, -314.343505859375], "reading_logprob_std": 2.87686606322858, "tokenization_path_logprob_std": 180.64230243656448, "trajectory_balance_rmse": 4550.786683517908} {"canonical_length": 91, "canonical_nll": 1.214677244752318, "generation_rouge_l_f1": 0.32653061224489793, "generation_token_f1": 0.41836734693877553, "prompt_id": "f1222deb29a5be25", "reading_canonical_log_prob": -110.53562927246094, "reading_distances": [13, 13, 9, 15], "reading_log_probs": [-108.60801696777344, -109.38490295410156, -110.20150756835938, -108.79386901855469], "reading_logprob_std": 0.7173397388139373, "tokenization_path_logprob_std": 92.80587198101698, "trajectory_balance_rmse": 2145.7794296274374} {"canonical_length": 9, "canonical_nll": 0.3059052626291911, "generation_rouge_l_f1": 0.9333333333333333, "generation_token_f1": 0.9333333333333333, "prompt_id": "6b02223ad9dd90ef", "reading_canonical_log_prob": -2.7531473636627197, "reading_distances": [25, 29, 27, 25], "reading_log_probs": [-2.7718939781188965, -2.9338936805725098, -2.8734748363494873, -2.5533769130706787], "reading_logprob_std": 0.16713727133853779, "tokenization_path_logprob_std": 47.668081500547075, "trajectory_balance_rmse": 169.57232961544872} {"canonical_length": 170, "canonical_nll": 1.2742337395163144, "generation_rouge_l_f1": 0.2285714285714286, "generation_token_f1": 0.4666666666666667, "prompt_id": "a027618ad6b82f3e", "reading_canonical_log_prob": -216.61973571777344, "reading_distances": [23, 19, 18, 19], "reading_log_probs": [-218.64349365234375, -221.51490783691406, -221.27511596679688, -219.72601318359375], "reading_logprob_std": 1.3539961723049325, "tokenization_path_logprob_std": 162.08600787284766, "trajectory_balance_rmse": 3632.3243221642037} {"canonical_length": 38, "canonical_nll": 1.8387019508763363, "generation_rouge_l_f1": 0.05405405405405406, "generation_token_f1": 0.05405405405405406, "prompt_id": "252eb77faa4d024a", "reading_canonical_log_prob": -69.87067413330078, "reading_distances": [20, 24, 22, 13], "reading_log_probs": [-69.27006530761719, -69.79501342773438, -71.48418426513672, -69.9148941040039], "reading_logprob_std": 0.9541155838355013, "tokenization_path_logprob_std": 39.42241864631445, "trajectory_balance_rmse": 675.7818948014576} {"canonical_length": 273, "canonical_nll": 0.7936302995507097, "generation_rouge_l_f1": 0.368763557483731, "generation_token_f1": 0.6160520607375272, "prompt_id": "75be9357ba7c860a", "reading_canonical_log_prob": -216.66107177734375, "reading_distances": [24, 18, 17, 16], "reading_log_probs": [-218.37034606933594, -218.18836975097656, -211.6189422607422, -207.9134063720703], "reading_logprob_std": 5.143162149956439, "tokenization_path_logprob_std": 149.39450086142062, "trajectory_balance_rmse": 4612.643287768291} {"canonical_length": 347, "canonical_nll": 1.1265315075085556, "generation_rouge_l_f1": 0.2833876221498371, "generation_token_f1": 0.5081433224755699, "prompt_id": "8d16eaea1c875841", "reading_canonical_log_prob": -390.90643310546875, "reading_distances": [41, 42, 47, 53], "reading_log_probs": [-394.2408142089844, -394.49005126953125, -394.002685546875, -395.91192626953125], "reading_logprob_std": 0.8571218258055793, "tokenization_path_logprob_std": 149.68461060731724, "trajectory_balance_rmse": 7437.863673546139} {"canonical_length": 355, "canonical_nll": 0.974138974471831, "generation_rouge_l_f1": 0.24715447154471548, "generation_token_f1": 0.49756097560975615, "prompt_id": "f4746ccf1632be92", "reading_canonical_log_prob": -345.8193359375, "reading_distances": [23, 26, 23, 20], "reading_log_probs": [-337.82513427734375, -340.7779541015625, -340.6610107421875, -336.7880554199219], "reading_logprob_std": 2.0159700510752994, "tokenization_path_logprob_std": 296.2945171171791, "trajectory_balance_rmse": 7125.85349733856} {"canonical_length": 450, "canonical_nll": 1.0234604220920138, "generation_rouge_l_f1": 0.1340941512125535, "generation_token_f1": 0.24536376604850216, "prompt_id": "06190df6c1e1573b", "reading_canonical_log_prob": -460.55718994140625, "reading_distances": [16, 15, 17, 12], "reading_log_probs": [-459.02398681640625, -455.5793151855469, -454.6773376464844, -458.25885009765625], "reading_logprob_std": 2.0849711851185306, "tokenization_path_logprob_std": 294.64789896276955, "trajectory_balance_rmse": 7389.388028236224} {"canonical_length": 102, "canonical_nll": 0.9111513624004289, "generation_rouge_l_f1": 0.3349753694581281, "generation_token_f1": 0.4630541871921182, "prompt_id": "56fa8783ee42ce19", "reading_canonical_log_prob": -92.93743896484375, "reading_distances": [13, 19, 16, 19], "reading_log_probs": [-91.43199157714844, -92.59733581542969, -89.44583129882812, -90.91219329833984], "reading_logprob_std": 1.3068713794848958, "tokenization_path_logprob_std": 65.13308073877556, "trajectory_balance_rmse": 1866.9211586561466} {"canonical_length": 202, "canonical_nll": 2.2309006794844524, "generation_rouge_l_f1": 0.2295805739514349, "generation_token_f1": 0.41501103752759383, "prompt_id": "11f8f65e51ee2cf5", "reading_canonical_log_prob": -450.6419372558594, "reading_distances": [22, 23, 23, 20], "reading_log_probs": [-447.5193176269531, -445.9664306640625, -442.14990234375, -449.71368408203125], "reading_logprob_std": 3.1868991371546658, "tokenization_path_logprob_std": 191.06038958500835, "trajectory_balance_rmse": 3476.187662949292} {"canonical_length": 11, "canonical_nll": 2.9670639038085938, "generation_rouge_l_f1": 0.3636363636363636, "generation_token_f1": 0.3636363636363636, "prompt_id": "8980941d8fb125ec", "reading_canonical_log_prob": -32.63770294189453, "reading_distances": [15, 18, 16, 21], "reading_log_probs": [-34.38922882080078, -31.837669372558594, -32.9660758972168, -32.7801399230957], "reading_logprob_std": 1.0536093185067323, "tokenization_path_logprob_std": 35.82325842532562, "trajectory_balance_rmse": 227.55470671161723} {"canonical_length": 69, "canonical_nll": 0.47232973402824957, "generation_rouge_l_f1": 0.1376518218623482, "generation_token_f1": 0.2348178137651822, "prompt_id": "2d0e3ff402d06b28", "reading_canonical_log_prob": -32.59075164794922, "reading_distances": [16, 14, 13, 14], "reading_log_probs": [-37.048728942871094, -36.769134521484375, -36.659889221191406, -34.01317596435547], "reading_logprob_std": 1.415870886979719, "tokenization_path_logprob_std": 154.194708278498, "trajectory_balance_rmse": 988.0142290593999} {"canonical_length": 53, "canonical_nll": 0.5143101170377912, "generation_rouge_l_f1": 0.225, "generation_token_f1": 0.3875, "prompt_id": "54b6e30181a1bec1", "reading_canonical_log_prob": -27.25843620300293, "reading_distances": [19, 24, 19, 16], "reading_log_probs": [-21.824352264404297, -22.148223876953125, -24.214492797851562, -23.9302978515625], "reading_logprob_std": 1.2171925716945518, "tokenization_path_logprob_std": 29.856901818859935, "trajectory_balance_rmse": 418.79480248878195} {"canonical_length": 9, "canonical_nll": 0.480759408738878, "generation_rouge_l_f1": 0.7999999999999999, "generation_token_f1": 0.7999999999999999, "prompt_id": "2996ca79ae58108b", "reading_canonical_log_prob": -4.326834678649902, "reading_distances": [30, 28, 26, 36], "reading_log_probs": [-7.421718120574951, -7.052749156951904, -5.829689979553223, -7.74497127532959], "reading_logprob_std": 0.8375824314916822, "tokenization_path_logprob_std": 38.949131983600246, "trajectory_balance_rmse": 135.18033284299366} {"canonical_length": 11, "canonical_nll": 1.722537474198775, "generation_rouge_l_f1": 0.23529411764705882, "generation_token_f1": 0.23529411764705882, "prompt_id": "e6cc80b8ba91ddd7", "reading_canonical_log_prob": -18.947912216186523, "reading_distances": [21, 22, 27, 25], "reading_log_probs": [-18.08384132385254, -21.2816162109375, -18.7596435546875, -20.884906768798828], "reading_logprob_std": 1.569578199712542, "tokenization_path_logprob_std": 31.659333403808027, "trajectory_balance_rmse": 219.618213401685} {"canonical_length": 218, "canonical_nll": 1.8488175978354358, "generation_rouge_l_f1": 0.2157676348547718, "generation_token_f1": 0.45228215767634855, "prompt_id": "1826061f4f79f3ac", "reading_canonical_log_prob": -403.042236328125, "reading_distances": [14, 16, 13, 20], "reading_log_probs": [-397.34979248046875, -399.7841796875, -400.69268798828125, -399.0577697753906], "reading_logprob_std": 1.4155209331349745, "tokenization_path_logprob_std": 117.55032770203331, "trajectory_balance_rmse": 4497.293390870381} {"canonical_length": 182, "canonical_nll": 1.7902919224330358, "generation_rouge_l_f1": 0.24713958810068654, "generation_token_f1": 0.41189931350114417, "prompt_id": "d48fbaa3dac1814d", "reading_canonical_log_prob": -325.8331298828125, "reading_distances": [29, 25, 30, 24], "reading_log_probs": [-321.96514892578125, -320.9207763671875, -320.0603332519531, -323.740234375], "reading_logprob_std": 1.583807207127996, "tokenization_path_logprob_std": 35.38651826362876, "trajectory_balance_rmse": 3548.1183468733325} {"canonical_length": 105, "canonical_nll": 0.5206334795270647, "generation_rouge_l_f1": 0.44787644787644787, "generation_token_f1": 0.5328185328185329, "prompt_id": "7fbf1f51feeeeb73", "reading_canonical_log_prob": -54.6665153503418, "reading_distances": [17, 20, 17, 20], "reading_log_probs": [-57.36465072631836, -56.52960968017578, -55.97285461425781, -59.02253723144531], "reading_logprob_std": 1.3294173433955767, "tokenization_path_logprob_std": 56.82561705525166, "trajectory_balance_rmse": 870.6767349893048} {"canonical_length": 18, "canonical_nll": 0.2816910213894314, "generation_rouge_l_f1": 0.13145539906103287, "generation_token_f1": 0.1408450704225352, "prompt_id": "8cc0717e876412c6", "reading_canonical_log_prob": -5.070438385009766, "reading_distances": [20, 21, 29, 26], "reading_log_probs": [-6.184961318969727, -7.488768577575684, -6.221957206726074, -7.127946376800537], "reading_logprob_std": 0.6548742225785892, "tokenization_path_logprob_std": 77.00590861178352, "trajectory_balance_rmse": 367.96120859048597} {"canonical_length": 152, "canonical_nll": 2.0505218505859375, "generation_rouge_l_f1": 0.25280898876404495, "generation_token_f1": 0.4213483146067416, "prompt_id": "3f19d2fd29921f16", "reading_canonical_log_prob": -311.6793212890625, "reading_distances": [19, 17, 16, 12], "reading_log_probs": [-306.0919189453125, -307.4573059082031, -315.20037841796875, -305.87353515625], "reading_logprob_std": 4.418988514710098, "tokenization_path_logprob_std": 283.3571436876992, "trajectory_balance_rmse": 2991.8812300783643} {"canonical_length": 257, "canonical_nll": 1.9246051951605059, "generation_rouge_l_f1": 0.1872340425531915, "generation_token_f1": 0.37872340425531914, "prompt_id": "774504b35575be36", "reading_canonical_log_prob": -494.62353515625, "reading_distances": [31, 24, 28, 29], "reading_log_probs": [-494.65057373046875, -491.76934814453125, -494.0436706542969, -492.8918762207031], "reading_logprob_std": 1.2754754580567813, "tokenization_path_logprob_std": 369.8902797561736, "trajectory_balance_rmse": 5381.434763016854} {"canonical_length": 9, "canonical_nll": 0.23939945962693956, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "18b528d0286abbc1", "reading_canonical_log_prob": -2.154595136642456, "reading_distances": [34, 30, 33, 31], "reading_log_probs": [-2.9482057094573975, -2.1813201904296875, -3.684211254119873, -4.432705879211426], "reading_logprob_std": 0.9670066178604135, "tokenization_path_logprob_std": 19.32393873735609, "trajectory_balance_rmse": 195.68361093682614} {"canonical_length": 4, "canonical_nll": 0.6442370414733887, "generation_rouge_l_f1": 0.8, "generation_token_f1": 0.8, "prompt_id": "ad4fd52fcf4632b3", "reading_canonical_log_prob": -2.5769481658935547, "reading_distances": [22, 29, 20, 25], "reading_log_probs": [-3.0532219409942627, -2.911543369293213, -2.7451205253601074, -2.8584728240966797], "reading_logprob_std": 0.12789589122489226, "tokenization_path_logprob_std": 37.53986910788237, "trajectory_balance_rmse": 100.36615580554513} {"canonical_length": 349, "canonical_nll": 1.442224890591422, "generation_rouge_l_f1": 0.20608108108108106, "generation_token_f1": 0.42229729729729737, "prompt_id": "ba9050b431bd2ba0", "reading_canonical_log_prob": -503.33648681640625, "reading_distances": [18, 15, 21, 19], "reading_log_probs": [-497.27978515625, -496.2735595703125, -495.0725402832031, -496.1505126953125], "reading_logprob_std": 0.9027406665788954, "tokenization_path_logprob_std": 167.82125624124396, "trajectory_balance_rmse": 6095.92296984635} {"canonical_length": 35, "canonical_nll": 0.6334677015032087, "generation_rouge_l_f1": 0.1729323308270677, "generation_token_f1": 0.19548872180451127, "prompt_id": "da7f7e74060fc4e6", "reading_canonical_log_prob": -22.171369552612305, "reading_distances": [27, 33, 26, 26], "reading_log_probs": [-21.22384262084961, -21.607803344726562, -22.30556869506836, -21.156963348388672], "reading_logprob_std": 0.5268979505137001, "tokenization_path_logprob_std": 67.9270033611413, "trajectory_balance_rmse": 705.6970193587473} {"canonical_length": 366, "canonical_nll": 1.0977599764130805, "generation_rouge_l_f1": 0.22118380062305293, "generation_token_f1": 0.470404984423676, "prompt_id": "378c4d815b07ba59", "reading_canonical_log_prob": -401.7801513671875, "reading_distances": [29, 25, 29, 34], "reading_log_probs": [-399.0841064453125, -397.83172607421875, -396.914794921875, -399.3478088378906], "reading_logprob_std": 1.132943600811552, "tokenization_path_logprob_std": 317.22211934359234, "trajectory_balance_rmse": 6772.4730797253505} {"canonical_length": 158, "canonical_nll": 0.561008839667598, "generation_rouge_l_f1": 0.3083700440528635, "generation_token_f1": 0.49339207048458145, "prompt_id": "e2790ed28464c480", "reading_canonical_log_prob": -88.63939666748047, "reading_distances": [14, 24, 22, 19], "reading_log_probs": [-89.76301574707031, -87.90988159179688, -87.84623718261719, -87.08194732666016], "reading_logprob_std": 1.139076339248703, "tokenization_path_logprob_std": 169.02481609618255, "trajectory_balance_rmse": 1969.9224590858062} {"canonical_length": 253, "canonical_nll": 1.158448954344738, "generation_rouge_l_f1": 0.2749003984063745, "generation_token_f1": 0.5378486055776892, "prompt_id": "2481a184df55ff40", "reading_canonical_log_prob": -293.08758544921875, "reading_distances": [47, 39, 37, 41], "reading_log_probs": [-306.48956298828125, -299.3052978515625, -299.847412109375, -299.28607177734375], "reading_logprob_std": 3.5146298994073777, "tokenization_path_logprob_std": 260.3757057593574, "trajectory_balance_rmse": 4468.630216120267} {"canonical_length": 342, "canonical_nll": 1.4872753299467745, "generation_rouge_l_f1": 0.21105527638190955, "generation_token_f1": 0.44556113902847566, "prompt_id": "130ff8d17ea30994", "reading_canonical_log_prob": -508.6481628417969, "reading_distances": [18, 20, 17, 21], "reading_log_probs": [-509.757080078125, -510.2492980957031, -507.564697265625, -507.034423828125], "reading_logprob_std": 1.5886410891269154, "tokenization_path_logprob_std": 257.52704898651234, "trajectory_balance_rmse": 7178.048645619785} {"canonical_length": 16, "canonical_nll": 0.8556859493255615, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.6666666666666666, "prompt_id": "16bf564574acdadf", "reading_canonical_log_prob": -13.690975189208984, "reading_distances": [27, 28, 22, 31], "reading_log_probs": [-14.91183090209961, -15.223549842834473, -13.500574111938477, -14.500219345092773], "reading_logprob_std": 0.7499656364371146, "tokenization_path_logprob_std": 25.29079426800789, "trajectory_balance_rmse": 213.42134695878443} {"canonical_length": 50, "canonical_nll": 0.46981788635253907, "generation_rouge_l_f1": 0.34545454545454546, "generation_token_f1": 0.41818181818181815, "prompt_id": "effedc7b02b1c679", "reading_canonical_log_prob": -23.490894317626953, "reading_distances": [29, 29, 27, 26], "reading_log_probs": [-27.084606170654297, -28.267440795898438, -26.337169647216797, -26.52619171142578], "reading_logprob_std": 0.8690612865753335, "tokenization_path_logprob_std": 67.79014749404415, "trajectory_balance_rmse": 766.3922550740671} {"canonical_length": 176, "canonical_nll": 1.420087294145064, "generation_rouge_l_f1": 0.15384615384615385, "generation_token_f1": 0.37556561085972845, "prompt_id": "33cc463f90c8e3a1", "reading_canonical_log_prob": -249.93536376953125, "reading_distances": [25, 17, 25, 28], "reading_log_probs": [-239.3154296875, -241.2850341796875, -241.1914520263672, -240.0390167236328], "reading_logprob_std": 0.9492015994874715, "tokenization_path_logprob_std": 60.144702509851186, "trajectory_balance_rmse": 3403.92438965211} {"canonical_length": 283, "canonical_nll": 1.2275507087842314, "generation_rouge_l_f1": 0.2654545454545454, "generation_token_f1": 0.5309090909090908, "prompt_id": "e32f1e4a8a806d4a", "reading_canonical_log_prob": -347.3968505859375, "reading_distances": [15, 14, 15, 15], "reading_log_probs": [-344.11737060546875, -343.65960693359375, -343.43707275390625, -344.08575439453125], "reading_logprob_std": 0.33232305083553976, "tokenization_path_logprob_std": 363.12490987210293, "trajectory_balance_rmse": 5914.6471578681085} {"canonical_length": 193, "canonical_nll": 1.2853317656047603, "generation_rouge_l_f1": 0.2788671023965142, "generation_token_f1": 0.47930283224400877, "prompt_id": "f16b4d6e1a4f11c6", "reading_canonical_log_prob": -248.06903076171875, "reading_distances": [11, 8, 9, 13], "reading_log_probs": [-247.88426208496094, -247.89263916015625, -246.74220275878906, -246.3785400390625], "reading_logprob_std": 0.7810153218298619, "tokenization_path_logprob_std": 199.13171658340391, "trajectory_balance_rmse": 3971.8440599990768} {"canonical_length": 11, "canonical_nll": 0.9207068356600675, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "7cf86e055cecbc26", "reading_canonical_log_prob": -10.127775192260742, "reading_distances": [9, 16, 13, 16], "reading_log_probs": [-10.609231948852539, -10.164236068725586, -11.436076164245605, -10.685770034790039], "reading_logprob_std": 0.5275797698522152, "tokenization_path_logprob_std": 35.82659679596727, "trajectory_balance_rmse": 192.04524510262462} {"canonical_length": 251, "canonical_nll": 1.559098201918887, "generation_rouge_l_f1": 0.19123505976095614, "generation_token_f1": 0.4143426294820716, "prompt_id": "279ec9f263f853e8", "reading_canonical_log_prob": -391.3336486816406, "reading_distances": [29, 20, 27, 24], "reading_log_probs": [-386.69342041015625, -384.72540283203125, -384.9854431152344, -384.1938171386719], "reading_logprob_std": 1.0807085687807747, "tokenization_path_logprob_std": 166.32836405395128, "trajectory_balance_rmse": 5046.541706266144} {"canonical_length": 102, "canonical_nll": 1.0562150244619333, "generation_rouge_l_f1": 0.21022727272727273, "generation_token_f1": 0.3522727272727273, "prompt_id": "cf98e6698c25bbf0", "reading_canonical_log_prob": -107.73393249511719, "reading_distances": [110, 109, 100, 109], "reading_log_probs": [-98.34616088867188, -100.89379119873047, -100.27513885498047, -98.74186706542969], "reading_logprob_std": 1.2156065066493733, "tokenization_path_logprob_std": 133.12485254501834, "trajectory_balance_rmse": 1814.256670847892} {"canonical_length": 125, "canonical_nll": 1.36229248046875, "generation_rouge_l_f1": 0.21938775510204084, "generation_token_f1": 0.41836734693877553, "prompt_id": "588a316f7d854f31", "reading_canonical_log_prob": -170.28656005859375, "reading_distances": [34, 25, 33, 35], "reading_log_probs": [-169.44998168945312, -169.09080505371094, -169.97698974609375, -169.82135009765625], "reading_logprob_std": 0.39664246205398407, "tokenization_path_logprob_std": 98.1889097447987, "trajectory_balance_rmse": 2962.9816949516517} {"canonical_length": 116, "canonical_nll": 1.7604965999208648, "generation_rouge_l_f1": 0.17021276595744683, "generation_token_f1": 0.3617021276595745, "prompt_id": "6bdc8fa7262e4a79", "reading_canonical_log_prob": -204.2176055908203, "reading_distances": [18, 15, 18, 11], "reading_log_probs": [-204.9720001220703, -207.02792358398438, -204.99928283691406, -207.07681274414062], "reading_logprob_std": 1.1934440968572164, "tokenization_path_logprob_std": 168.25313638429776, "trajectory_balance_rmse": 2328.117596477749} {"canonical_length": 7, "canonical_nll": 2.316887174333845, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "73284d596fde1232", "reading_canonical_log_prob": -16.218210220336914, "reading_distances": [26, 25, 31, 27], "reading_log_probs": [-16.094139099121094, -15.313253402709961, -15.440194129943848, -15.579708099365234], "reading_logprob_std": 0.3426176134246428, "tokenization_path_logprob_std": 29.34834201646861, "trajectory_balance_rmse": 186.05935075596454} {"canonical_length": 47, "canonical_nll": 1.0925413902769698, "generation_rouge_l_f1": 0.17582417582417584, "generation_token_f1": 0.26373626373626374, "prompt_id": "08b3c472b8ed1840", "reading_canonical_log_prob": -51.34944534301758, "reading_distances": [37, 51, 48, 51], "reading_log_probs": [-47.930023193359375, -48.34733200073242, -45.52538299560547, -45.89996337890625], "reading_logprob_std": 1.4192399313779605, "tokenization_path_logprob_std": 69.87182955361247, "trajectory_balance_rmse": 654.6174828286099} {"canonical_length": 149, "canonical_nll": 0.7850997719988727, "generation_rouge_l_f1": 0.2782152230971129, "generation_token_f1": 0.47244094488188976, "prompt_id": "51fa8b33eba8ebe6", "reading_canonical_log_prob": -116.97986602783203, "reading_distances": [40, 36, 42, 45], "reading_log_probs": [-118.06629180908203, -120.42774200439453, -119.80441284179688, -118.24595642089844], "reading_logprob_std": 1.1621569427570857, "tokenization_path_logprob_std": 74.88159631793859, "trajectory_balance_rmse": 2196.976753665944} {"canonical_length": 25, "canonical_nll": 0.8832373046875, "generation_rouge_l_f1": 0.25263157894736843, "generation_token_f1": 0.2947368421052632, "prompt_id": "6e11957de8d37c71", "reading_canonical_log_prob": -22.0809326171875, "reading_distances": [26, 22, 27, 19], "reading_log_probs": [-21.22488021850586, -20.433544158935547, -21.537425994873047, -21.681427001953125], "reading_logprob_std": 0.5574372591472916, "tokenization_path_logprob_std": 36.57510717147748, "trajectory_balance_rmse": 478.1283392457426} {"canonical_length": 124, "canonical_nll": 1.0723624690886466, "generation_rouge_l_f1": 0.46428571428571425, "generation_token_f1": 0.6071428571428571, "prompt_id": "613e08997d787f9d", "reading_canonical_log_prob": -132.9729461669922, "reading_distances": [21, 18, 24, 20], "reading_log_probs": [-131.126220703125, -133.53323364257812, -132.35337829589844, -132.33682250976562], "reading_logprob_std": 0.9827222443421567, "tokenization_path_logprob_std": 185.9240458797834, "trajectory_balance_rmse": 2597.180749791538} {"canonical_length": 7, "canonical_nll": 0.16220545768737793, "generation_rouge_l_f1": 0.9090909090909091, "generation_token_f1": 0.9090909090909091, "prompt_id": "82a7bb73a8e6bdca", "reading_canonical_log_prob": -1.1354382038116455, "reading_distances": [33, 34, 31, 40], "reading_log_probs": [-3.6907613277435303, -2.291219711303711, -2.4992241859436035, -2.750598669052124], "reading_logprob_std": 0.6177828191932101, "tokenization_path_logprob_std": 19.74447047580133, "trajectory_balance_rmse": 113.28275640611996} {"canonical_length": 500, "canonical_nll": 0.8190447387695312, "generation_rouge_l_f1": 0.24055944055944056, "generation_token_f1": 0.3944055944055944, "prompt_id": "04c1fbf027769e8a", "reading_canonical_log_prob": -409.5223693847656, "reading_distances": [24, 24, 28, 22], "reading_log_probs": [-411.60321044921875, -411.08538818359375, -414.21636962890625, -408.4599609375], "reading_logprob_std": 2.3595358362045524, "tokenization_path_logprob_std": 165.08310812875882, "trajectory_balance_rmse": 7853.435986638473} {"canonical_length": 47, "canonical_nll": 1.6428704769053357, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "5950f3bf48416566", "reading_canonical_log_prob": -77.21491241455078, "reading_distances": [32, 35, 36, 48], "reading_log_probs": [-80.05811309814453, -77.78680419921875, -77.41847229003906, -78.23926544189453], "reading_logprob_std": 1.1707848320812324, "tokenization_path_logprob_std": 62.61398630615686, "trajectory_balance_rmse": 1073.1988826925765} {"canonical_length": 310, "canonical_nll": 1.1623277233492944, "generation_rouge_l_f1": 0.3250883392226149, "generation_token_f1": 0.5547703180212014, "prompt_id": "f93b23d1ff15b3d1", "reading_canonical_log_prob": -360.32159423828125, "reading_distances": [22, 25, 23, 24], "reading_log_probs": [-354.5535888671875, -353.3089904785156, -355.01214599609375, -354.85882568359375], "reading_logprob_std": 0.7734502947737202, "tokenization_path_logprob_std": 233.7896529794203, "trajectory_balance_rmse": 5582.534325392437} {"canonical_length": 228, "canonical_nll": 1.4144068935461211, "generation_rouge_l_f1": 0.20883534136546186, "generation_token_f1": 0.465863453815261, "prompt_id": "cb117bc3f0e8e883", "reading_canonical_log_prob": -322.4847717285156, "reading_distances": [25, 19, 19, 23], "reading_log_probs": [-313.8748779296875, -310.48724365234375, -314.3749084472656, -314.1370849609375], "reading_logprob_std": 1.8322729097358432, "tokenization_path_logprob_std": 311.19562095853007, "trajectory_balance_rmse": 4565.621663127492} {"canonical_length": 318, "canonical_nll": 1.8762644641804245, "generation_rouge_l_f1": 0.2298456260720412, "generation_token_f1": 0.5454545454545454, "prompt_id": "0ff4ea021b8e685b", "reading_canonical_log_prob": -596.652099609375, "reading_distances": [16, 20, 14, 14], "reading_log_probs": [-598.3717041015625, -601.5073852539062, -600.99853515625, -599.5221557617188], "reading_logprob_std": 1.4270026096835229, "tokenization_path_logprob_std": 172.03171958587842, "trajectory_balance_rmse": 6241.777472629779} {"canonical_length": 70, "canonical_nll": 1.389593723842076, "generation_rouge_l_f1": 0.2288557213930348, "generation_token_f1": 0.41791044776119396, "prompt_id": "d8931b93e350f362", "reading_canonical_log_prob": -97.27156066894531, "reading_distances": [14, 16, 14, 11], "reading_log_probs": [-97.86216735839844, -97.38020324707031, -98.56742095947266, -99.30906677246094], "reading_logprob_std": 0.8417852491876081, "tokenization_path_logprob_std": 127.89732383415756, "trajectory_balance_rmse": 1494.19126513761} {"canonical_length": 43, "canonical_nll": 1.8329662943995275, "generation_rouge_l_f1": 0.25925925925925924, "generation_token_f1": 0.37037037037037035, "prompt_id": "cfd710328af17995", "reading_canonical_log_prob": -78.81755065917969, "reading_distances": [32, 31, 27, 30], "reading_log_probs": [-77.44320678710938, -79.13057708740234, -76.35601806640625, -76.2659912109375], "reading_logprob_std": 1.3331406901516063, "tokenization_path_logprob_std": 106.42391110972586, "trajectory_balance_rmse": 860.889426150719} {"canonical_length": 237, "canonical_nll": 0.7597214582097178, "generation_rouge_l_f1": 0.22823984526112187, "generation_token_f1": 0.39458413926499036, "prompt_id": "3f921f925896c713", "reading_canonical_log_prob": -180.05398559570312, "reading_distances": [37, 35, 42, 39], "reading_log_probs": [-186.72100830078125, -181.22439575195312, -185.83181762695312, -182.32626342773438], "reading_logprob_std": 2.662213224279594, "tokenization_path_logprob_std": 130.2198335798216, "trajectory_balance_rmse": 2288.0940245483766} {"canonical_length": 242, "canonical_nll": 1.019006776415612, "generation_rouge_l_f1": 0.3505154639175258, "generation_token_f1": 0.5567010309278352, "prompt_id": "74c9b8835b365ad8", "reading_canonical_log_prob": -246.59963989257812, "reading_distances": [21, 15, 17, 19], "reading_log_probs": [-245.36419677734375, -243.81558227539062, -244.99237060546875, -245.12506103515625], "reading_logprob_std": 0.6898562427031654, "tokenization_path_logprob_std": 118.96790076431074, "trajectory_balance_rmse": 4458.070651267828} {"canonical_length": 243, "canonical_nll": 1.2612726658950617, "generation_rouge_l_f1": 0.2105263157894737, "generation_token_f1": 0.4873294346978557, "prompt_id": "d4cbdb2804902b19", "reading_canonical_log_prob": -306.4892578125, "reading_distances": [26, 16, 23, 24], "reading_log_probs": [-304.4743347167969, -305.6207275390625, -308.25872802734375, -306.9832458496094], "reading_logprob_std": 1.6424788637064769, "tokenization_path_logprob_std": 101.32064212521222, "trajectory_balance_rmse": 4328.371378193163} {"canonical_length": 140, "canonical_nll": 1.0835233415876115, "generation_rouge_l_f1": 0.23999999999999996, "generation_token_f1": 0.4184615384615385, "prompt_id": "ad65facd3056e4a9", "reading_canonical_log_prob": -151.69326782226562, "reading_distances": [18, 13, 19, 17], "reading_log_probs": [-151.22994995117188, -147.9295654296875, -151.34637451171875, -150.8568572998047], "reading_logprob_std": 1.6209189752298414, "tokenization_path_logprob_std": 123.5655239346293, "trajectory_balance_rmse": 1827.3797232129664} {"canonical_length": 245, "canonical_nll": 1.1685346330915178, "generation_rouge_l_f1": 0.2505307855626327, "generation_token_f1": 0.5265392781316348, "prompt_id": "4eba5e8b52e3bb5e", "reading_canonical_log_prob": -286.2909851074219, "reading_distances": [19, 24, 22, 25], "reading_log_probs": [-287.60198974609375, -286.77813720703125, -286.43316650390625, -289.6905517578125], "reading_logprob_std": 1.4611266785251398, "tokenization_path_logprob_std": 250.3641563156488, "trajectory_balance_rmse": 4581.224383736867} {"canonical_length": 118, "canonical_nll": 1.7025749077231198, "generation_rouge_l_f1": 0.18528610354223432, "generation_token_f1": 0.34332425068119893, "prompt_id": "4873d3597b3d71b4", "reading_canonical_log_prob": -200.90383911132812, "reading_distances": [21, 22, 20, 20], "reading_log_probs": [-203.19940185546875, -203.83328247070312, -202.73455810546875, -201.1185302734375], "reading_logprob_std": 1.1596167313817904, "tokenization_path_logprob_std": 147.9880200912751, "trajectory_balance_rmse": 2879.9010646971083} {"canonical_length": 38, "canonical_nll": 1.4462366606059827, "generation_rouge_l_f1": 0.17391304347826086, "generation_token_f1": 0.17391304347826086, "prompt_id": "2089d8bccf6a73c9", "reading_canonical_log_prob": -54.956993103027344, "reading_distances": [12, 10, 9, 11], "reading_log_probs": [-60.06427764892578, -61.848472595214844, -56.59294128417969, -62.36780548095703], "reading_logprob_std": 2.610541781814963, "tokenization_path_logprob_std": 113.32909901683188, "trajectory_balance_rmse": 714.0188109382657} {"canonical_length": 277, "canonical_nll": 1.607526648345837, "generation_rouge_l_f1": 0.2191235059760956, "generation_token_f1": 0.46215139442231074, "prompt_id": "9b806d8ec328a266", "reading_canonical_log_prob": -445.2848815917969, "reading_distances": [31, 23, 30, 30], "reading_log_probs": [-449.3019104003906, -445.4395751953125, -448.97296142578125, -449.5451354980469], "reading_logprob_std": 1.9311660107829478, "tokenization_path_logprob_std": 256.84132083699905, "trajectory_balance_rmse": 6084.962929004433} {"canonical_length": 400, "canonical_nll": 0.9044097900390625, "generation_rouge_l_f1": 0.2895622895622895, "generation_token_f1": 0.4781144781144781, "prompt_id": "ce03043d2a4264da", "reading_canonical_log_prob": -361.763916015625, "reading_distances": [20, 23, 24, 18], "reading_log_probs": [-361.4698791503906, -366.61785888671875, -364.9577331542969, -361.72723388671875], "reading_logprob_std": 2.514017505813113, "tokenization_path_logprob_std": 184.15540527792163, "trajectory_balance_rmse": 6395.783227272782} {"canonical_length": 12, "canonical_nll": 0.8089162508646647, "generation_rouge_l_f1": 0.45454545454545453, "generation_token_f1": 0.45454545454545453, "prompt_id": "59741a0906fe97c9", "reading_canonical_log_prob": -9.706995010375977, "reading_distances": [27, 27, 30, 23], "reading_log_probs": [-9.699399948120117, -9.646238327026367, -9.507270812988281, -9.822876930236816], "reading_logprob_std": 0.13073724124595315, "tokenization_path_logprob_std": 45.18752047890628, "trajectory_balance_rmse": 200.13084122992313} {"canonical_length": 103, "canonical_nll": 0.5216193338042324, "generation_rouge_l_f1": 0.3181818181818182, "generation_token_f1": 0.4621212121212121, "prompt_id": "995c665a4ec8066d", "reading_canonical_log_prob": -53.72679138183594, "reading_distances": [31, 27, 34, 26], "reading_log_probs": [-54.23688888549805, -57.20672607421875, -55.90721893310547, -56.70878601074219], "reading_logprob_std": 1.3006211237267074, "tokenization_path_logprob_std": 152.66141973596532, "trajectory_balance_rmse": 1358.4133289162094} {"canonical_length": 63, "canonical_nll": 0.8597298879472036, "generation_rouge_l_f1": 0.6140350877192983, "generation_token_f1": 0.6842105263157895, "prompt_id": "e7b51e88b3b34da7", "reading_canonical_log_prob": -54.16298294067383, "reading_distances": [29, 30, 26, 28], "reading_log_probs": [-57.418304443359375, -57.104469299316406, -57.844661712646484, -56.53841018676758], "reading_logprob_std": 0.5499308818695313, "tokenization_path_logprob_std": 93.8759551980857, "trajectory_balance_rmse": 1021.6150258779603} {"canonical_length": 90, "canonical_nll": 0.9921405368381077, "generation_rouge_l_f1": 0.42528735632183906, "generation_token_f1": 0.5862068965517242, "prompt_id": "6b580b925c517c44", "reading_canonical_log_prob": -89.29264831542969, "reading_distances": [26, 23, 27, 20], "reading_log_probs": [-89.58888244628906, -89.56510925292969, -88.2904052734375, -87.2468032836914], "reading_logprob_std": 1.1276990161517058, "tokenization_path_logprob_std": 168.1434150620166, "trajectory_balance_rmse": 1600.5597984116164} {"canonical_length": 169, "canonical_nll": 0.7823005089393029, "generation_rouge_l_f1": 0.2783018867924528, "generation_token_f1": 0.5, "prompt_id": "58b4edff73083c67", "reading_canonical_log_prob": -132.2087860107422, "reading_distances": [54, 51, 52, 62], "reading_log_probs": [-132.5461883544922, -136.5313262939453, -129.29803466796875, -133.27789306640625], "reading_logprob_std": 2.9680496814304553, "tokenization_path_logprob_std": 42.4013636856056, "trajectory_balance_rmse": 2412.317404799786} {"canonical_length": 6, "canonical_nll": 0.9546105861663818, "generation_rouge_l_f1": 0.1, "generation_token_f1": 0.1, "prompt_id": "dc845859825dc504", "reading_canonical_log_prob": -5.727663516998291, "reading_distances": [33, 29, 30, 35], "reading_log_probs": [-6.712133884429932, -9.933467864990234, -7.472529411315918, -7.179350852966309], "reading_logprob_std": 1.4405096876649108, "tokenization_path_logprob_std": 17.450638939873848, "trajectory_balance_rmse": 113.95418338707742} {"canonical_length": 55, "canonical_nll": 0.4140830300071023, "generation_rouge_l_f1": 0.6, "generation_token_f1": 0.6307692307692307, "prompt_id": "f074eb196b10e4a0", "reading_canonical_log_prob": -22.774566650390625, "reading_distances": [30, 31, 27, 29], "reading_log_probs": [-23.930377960205078, -24.74694061279297, -26.22477912902832, -26.730588912963867], "reading_logprob_std": 1.2957283494047418, "tokenization_path_logprob_std": 36.51166793196723, "trajectory_balance_rmse": 367.9373838534712} {"canonical_length": 39, "canonical_nll": 2.1073207366160855, "generation_rouge_l_f1": 0.1342281879194631, "generation_token_f1": 0.25503355704697983, "prompt_id": "b1076398ceba44c5", "reading_canonical_log_prob": -82.18550872802734, "reading_distances": [9, 12, 15, 13], "reading_log_probs": [-84.21568298339844, -84.74853515625, -80.41368103027344, -84.68357849121094], "reading_logprob_std": 2.0813713525113595, "tokenization_path_logprob_std": 61.37309797826073, "trajectory_balance_rmse": 827.8697760812271} {"canonical_length": 278, "canonical_nll": 1.7596909776866008, "generation_rouge_l_f1": 0.18315018315018317, "generation_token_f1": 0.41025641025641024, "prompt_id": "43d587270196a951", "reading_canonical_log_prob": -489.194091796875, "reading_distances": [23, 24, 29, 22], "reading_log_probs": [-488.9783020019531, -484.1639404296875, -487.82049560546875, -486.55743408203125], "reading_logprob_std": 2.063042383537916, "tokenization_path_logprob_std": 256.02888710094544, "trajectory_balance_rmse": 5847.782526094099} {"canonical_length": 77, "canonical_nll": 1.3779021424132507, "generation_rouge_l_f1": 0.26905829596412556, "generation_token_f1": 0.34080717488789236, "prompt_id": "9ec5781f60b3ed7d", "reading_canonical_log_prob": -106.09846496582031, "reading_distances": [14, 16, 9, 14], "reading_log_probs": [-107.12208557128906, -107.12259674072266, -105.38967895507812, -106.51438903808594], "reading_logprob_std": 0.8169262234040241, "tokenization_path_logprob_std": 217.3575380591067, "trajectory_balance_rmse": 1550.644308595682} {"canonical_length": 50, "canonical_nll": 1.046190948486328, "generation_rouge_l_f1": 0.44961240310077516, "generation_token_f1": 0.49612403100775193, "prompt_id": "9c285d8b80679a4e", "reading_canonical_log_prob": -52.309547424316406, "reading_distances": [21, 24, 26, 20], "reading_log_probs": [-51.27476501464844, -51.991031646728516, -51.980796813964844, -51.40406036376953], "reading_logprob_std": 0.3769948108601488, "tokenization_path_logprob_std": 27.53776852500902, "trajectory_balance_rmse": 984.6160887041655} {"canonical_length": 260, "canonical_nll": 1.4215712327223557, "generation_rouge_l_f1": 0.22444889779559118, "generation_token_f1": 0.41683366733466926, "prompt_id": "46fcd75d760af9c5", "reading_canonical_log_prob": -369.6085205078125, "reading_distances": [24, 20, 34, 25], "reading_log_probs": [-368.6488037109375, -367.825927734375, -366.61541748046875, -365.1820068359375], "reading_logprob_std": 1.5094365360247177, "tokenization_path_logprob_std": 153.79658816121355, "trajectory_balance_rmse": 4918.441625009307} {"canonical_length": 328, "canonical_nll": 1.079164272401391, "generation_rouge_l_f1": 0.1658206429780034, "generation_token_f1": 0.31472081218274117, "prompt_id": "5feba3faef17f1e8", "reading_canonical_log_prob": -353.96588134765625, "reading_distances": [27, 24, 28, 26], "reading_log_probs": [-351.27496337890625, -348.62310791015625, -349.7956848144531, -350.10418701171875], "reading_logprob_std": 1.0899168663881313, "tokenization_path_logprob_std": 426.3031154926352, "trajectory_balance_rmse": 6743.325314065848} {"canonical_length": 232, "canonical_nll": 1.1630151682886585, "generation_rouge_l_f1": 0.2942430703624733, "generation_token_f1": 0.5671641791044776, "prompt_id": "49ec1dbe6be4038c", "reading_canonical_log_prob": -269.81951904296875, "reading_distances": [17, 24, 23, 16], "reading_log_probs": [-268.2218933105469, -268.2464904785156, -268.37542724609375, -269.28302001953125], "reading_logprob_std": 0.5053802532404137, "tokenization_path_logprob_std": 250.07425794244256, "trajectory_balance_rmse": 4888.61464865067} {"canonical_length": 368, "canonical_nll": 1.1749702121900476, "generation_rouge_l_f1": 0.20249221183800623, "generation_token_f1": 0.39875389408099693, "prompt_id": "af5752cd92c29a4d", "reading_canonical_log_prob": -432.3890380859375, "reading_distances": [17, 20, 19, 15], "reading_log_probs": [-421.80718994140625, -418.50421142578125, -419.04931640625, -421.6619567871094], "reading_logprob_std": 1.723151452671353, "tokenization_path_logprob_std": 371.6902901884609, "trajectory_balance_rmse": 6610.472498000142} {"canonical_length": 152, "canonical_nll": 1.513887606169048, "generation_rouge_l_f1": 0.19184652278177458, "generation_token_f1": 0.3597122302158274, "prompt_id": "a82619f7741e0303", "reading_canonical_log_prob": -230.1109161376953, "reading_distances": [26, 16, 20, 19], "reading_log_probs": [-229.28028869628906, -231.52810668945312, -230.89569091796875, -229.2823486328125], "reading_logprob_std": 1.1441323713335472, "tokenization_path_logprob_std": 131.9058255857597, "trajectory_balance_rmse": 3592.3128365485964} {"canonical_length": 419, "canonical_nll": 1.4105845157741648, "generation_rouge_l_f1": 0.2785515320334262, "generation_token_f1": 0.5376044568245125, "prompt_id": "9294062be51005dc", "reading_canonical_log_prob": -591.034912109375, "reading_distances": [19, 22, 20, 26], "reading_log_probs": [-588.7274169921875, -587.5665283203125, -588.294921875, -586.2298583984375], "reading_logprob_std": 1.093703352410257, "tokenization_path_logprob_std": 204.076718205665, "trajectory_balance_rmse": 6667.560764388743} {"canonical_length": 9, "canonical_nll": 0.11438886324564616, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "3e7c9dafc0fa7462", "reading_canonical_log_prob": -1.0294997692108154, "reading_distances": [23, 32, 26, 29], "reading_log_probs": [-1.246906042098999, -0.9157506227493286, -1.2789208889007568, -0.8428005576133728], "reading_logprob_std": 0.22386852584233002, "tokenization_path_logprob_std": 19.39167016087216, "trajectory_balance_rmse": 186.03521318682422} {"canonical_length": 12, "canonical_nll": 0.8835573196411133, "generation_rouge_l_f1": 0.75, "generation_token_f1": 0.75, "prompt_id": "c34fdb5fae4b8ff6", "reading_canonical_log_prob": -10.60268783569336, "reading_distances": [54, 44, 42, 59], "reading_log_probs": [-8.986230850219727, -7.445176124572754, -8.56680679321289, -9.496129989624023], "reading_logprob_std": 0.8726813114472186, "tokenization_path_logprob_std": 87.15663426056804, "trajectory_balance_rmse": 281.07425629548277} {"canonical_length": 122, "canonical_nll": 1.7099884533491292, "generation_rouge_l_f1": 0.1776504297994269, "generation_token_f1": 0.3553008595988538, "prompt_id": "d51b2c6c349d91e4", "reading_canonical_log_prob": -208.61859130859375, "reading_distances": [20, 25, 30, 31], "reading_log_probs": [-207.38389587402344, -205.2742462158203, -208.25155639648438, -206.906982421875], "reading_logprob_std": 1.250632690117235, "tokenization_path_logprob_std": 49.45149642503558, "trajectory_balance_rmse": 2345.7427926816204} {"canonical_length": 311, "canonical_nll": 1.2886470291775523, "generation_rouge_l_f1": 0.21516754850088182, "generation_token_f1": 0.4126984126984127, "prompt_id": "fc10a987783bea33", "reading_canonical_log_prob": -400.76922607421875, "reading_distances": [27, 25, 26, 25], "reading_log_probs": [-402.906494140625, -404.40545654296875, -401.04827880859375, -402.5534973144531], "reading_logprob_std": 1.378118806494701, "tokenization_path_logprob_std": 438.99951309377684, "trajectory_balance_rmse": 6844.874326697267} {"canonical_length": 31, "canonical_nll": 1.2154479488249748, "generation_rouge_l_f1": 0.11406844106463879, "generation_token_f1": 0.12167300380228137, "prompt_id": "7386d29fb448f53c", "reading_canonical_log_prob": -37.67888641357422, "reading_distances": [15, 17, 12, 17], "reading_log_probs": [-34.077056884765625, -33.22654724121094, -34.191043853759766, -34.418052673339844], "reading_logprob_std": 0.5207454372726453, "tokenization_path_logprob_std": 39.015130545871706, "trajectory_balance_rmse": 350.643029952321} {"canonical_length": 253, "canonical_nll": 0.9064008991708868, "generation_rouge_l_f1": 0.17570093457943925, "generation_token_f1": 0.34392523364485983, "prompt_id": "d5d2ea5bf0257ac3", "reading_canonical_log_prob": -229.31942749023438, "reading_distances": [34, 33, 30, 34], "reading_log_probs": [-232.514892578125, -231.76658630371094, -230.18692016601562, -230.71963500976562], "reading_logprob_std": 1.043935055125876, "tokenization_path_logprob_std": 154.29336569639796, "trajectory_balance_rmse": 5483.2343173041} {"canonical_length": 379, "canonical_nll": 1.1418329002360241, "generation_rouge_l_f1": 0.18040435458786935, "generation_token_f1": 0.3825816485225506, "prompt_id": "0a54bd903d89ace0", "reading_canonical_log_prob": -432.7546691894531, "reading_distances": [21, 18, 22, 20], "reading_log_probs": [-426.7755126953125, -428.078369140625, -427.19622802734375, -425.464111328125], "reading_logprob_std": 1.0880776231206484, "tokenization_path_logprob_std": 200.85703271929233, "trajectory_balance_rmse": 6767.974809517327} {"canonical_length": 128, "canonical_nll": 1.0017837285995483, "generation_rouge_l_f1": 0.22499999999999998, "generation_token_f1": 0.365, "prompt_id": "49a85c182167a715", "reading_canonical_log_prob": -128.2283172607422, "reading_distances": [15, 10, 18, 18], "reading_log_probs": [-126.47979736328125, -127.120361328125, -125.82597351074219, -126.70372009277344], "reading_logprob_std": 0.5406337342292884, "tokenization_path_logprob_std": 57.68283038128173, "trajectory_balance_rmse": 2739.24973243529} {"canonical_length": 127, "canonical_nll": 1.230965681902067, "generation_rouge_l_f1": 0.1784776902887139, "generation_token_f1": 0.36220472440944884, "prompt_id": "50a15e61cf0ab437", "reading_canonical_log_prob": -156.3326416015625, "reading_distances": [35, 28, 41, 28], "reading_log_probs": [-172.15240478515625, -176.19223022460938, -176.32833862304688, -173.81517028808594], "reading_logprob_std": 2.010562869679905, "tokenization_path_logprob_std": 131.45523130517853, "trajectory_balance_rmse": 2819.6251439795183} {"canonical_length": 14, "canonical_nll": 0.7944326400756836, "generation_rouge_l_f1": 0.42857142857142855, "generation_token_f1": 0.42857142857142855, "prompt_id": "394139a0ae5c1d75", "reading_canonical_log_prob": -11.12205696105957, "reading_distances": [20, 29, 26, 29], "reading_log_probs": [-8.69244384765625, -8.740859985351562, -8.836270332336426, -8.646268844604492], "reading_logprob_std": 0.08129857243087069, "tokenization_path_logprob_std": 44.09825096661274, "trajectory_balance_rmse": 240.21356069056685} {"canonical_length": 58, "canonical_nll": 0.35715142611799566, "generation_rouge_l_f1": 0.495575221238938, "generation_token_f1": 0.5132743362831859, "prompt_id": "ee3925e13ca1cdfe", "reading_canonical_log_prob": -20.71478271484375, "reading_distances": [21, 20, 14, 22], "reading_log_probs": [-21.12732696533203, -20.952983856201172, -21.44841766357422, -21.120084762573242], "reading_logprob_std": 0.20710848002478224, "tokenization_path_logprob_std": 16.574967694027418, "trajectory_balance_rmse": 877.9412647481893} {"canonical_length": 293, "canonical_nll": 1.182739674434727, "generation_rouge_l_f1": 0.22735346358792186, "generation_token_f1": 0.4369449378330373, "prompt_id": "e5cf7a10fe204155", "reading_canonical_log_prob": -346.542724609375, "reading_distances": [28, 24, 27, 30], "reading_log_probs": [-354.13897705078125, -351.0758056640625, -356.9998474121094, -349.62530517578125], "reading_logprob_std": 3.2853580152244954, "tokenization_path_logprob_std": 125.38209829809007, "trajectory_balance_rmse": 5447.735360187862} {"canonical_length": 302, "canonical_nll": 1.2883066341576987, "generation_rouge_l_f1": 0.2398589065255732, "generation_token_f1": 0.4656084656084655, "prompt_id": "a6e94611dc8387e2", "reading_canonical_log_prob": -389.068603515625, "reading_distances": [32, 34, 29, 28], "reading_log_probs": [-383.1708984375, -384.09039306640625, -383.55340576171875, -387.0767517089844], "reading_logprob_std": 1.7764227156386356, "tokenization_path_logprob_std": 253.21776997540053, "trajectory_balance_rmse": 6335.722736454555} {"canonical_length": 195, "canonical_nll": 1.7678602952223557, "generation_rouge_l_f1": 0.27499999999999997, "generation_token_f1": 0.5549999999999999, "prompt_id": "747395c5d51f280d", "reading_canonical_log_prob": -344.7327575683594, "reading_distances": [21, 17, 16, 15], "reading_log_probs": [-354.6226806640625, -352.4320068359375, -351.6407470703125, -351.5378723144531], "reading_logprob_std": 1.433037976560824, "tokenization_path_logprob_std": 128.69787253524825, "trajectory_balance_rmse": 3720.479606652555} {"canonical_length": 35, "canonical_nll": 0.5849100385393415, "generation_rouge_l_f1": 0.15267175572519082, "generation_token_f1": 0.16793893129770995, "prompt_id": "c31ebbb64961b272", "reading_canonical_log_prob": -20.471851348876953, "reading_distances": [16, 17, 15, 20], "reading_log_probs": [-20.994831085205078, -20.950910568237305, -20.644161224365234, -21.176721572875977], "reading_logprob_std": 0.22111249158235124, "tokenization_path_logprob_std": 29.806532269392637, "trajectory_balance_rmse": 378.84455977926314} {"canonical_length": 111, "canonical_nll": 1.2952633419552364, "generation_rouge_l_f1": 0.29490616621983917, "generation_token_f1": 0.45040214477211793, "prompt_id": "0a9b1baeb01418e9", "reading_canonical_log_prob": -143.77423095703125, "reading_distances": [25, 28, 28, 22], "reading_log_probs": [-143.76690673828125, -147.87747192382812, -146.24606323242188, -142.40753173828125], "reading_logprob_std": 2.4530071246169896, "tokenization_path_logprob_std": 81.46541989300401, "trajectory_balance_rmse": 2204.8251751468642} {"canonical_length": 86, "canonical_nll": 1.0398376819699309, "generation_rouge_l_f1": 0.2631578947368421, "generation_token_f1": 0.4736842105263158, "prompt_id": "ffff4b859d3e455e", "reading_canonical_log_prob": -89.42604064941406, "reading_distances": [18, 19, 17, 19], "reading_log_probs": [-87.40272521972656, -88.13626098632812, -87.79386138916016, -87.97616577148438], "reading_logprob_std": 0.315700166309546, "tokenization_path_logprob_std": 72.77001589902885, "trajectory_balance_rmse": 2053.6367298264} {"canonical_length": 12, "canonical_nll": 0.5787300268809, "generation_rouge_l_f1": 0.2142857142857143, "generation_token_f1": 0.2142857142857143, "prompt_id": "a9db5a2fccb43cd0", "reading_canonical_log_prob": -6.944760322570801, "reading_distances": [21, 28, 23, 23], "reading_log_probs": [-8.836564064025879, -11.099959373474121, -8.214652061462402, -9.961647033691406], "reading_logprob_std": 1.2730635377911401, "tokenization_path_logprob_std": 82.13377269233231, "trajectory_balance_rmse": 326.33056082855757} {"canonical_length": 115, "canonical_nll": 1.822806980298913, "generation_rouge_l_f1": 0.023529411764705885, "generation_token_f1": 0.023529411764705885, "prompt_id": "6d2c7ff3221f2284", "reading_canonical_log_prob": -209.622802734375, "reading_distances": [11, 13, 15, 13], "reading_log_probs": [-208.41140747070312, -210.00222778320312, -207.20506286621094, -208.23587036132812], "reading_logprob_std": 1.155548356657, "tokenization_path_logprob_std": 152.99176402140833, "trajectory_balance_rmse": 1643.867662687464} {"canonical_length": 16, "canonical_nll": 1.1755231618881226, "generation_rouge_l_f1": 0.5641025641025641, "generation_token_f1": 0.6153846153846154, "prompt_id": "686af343746a5e07", "reading_canonical_log_prob": -18.80837059020996, "reading_distances": [22, 25, 24, 26], "reading_log_probs": [-15.664989471435547, -15.076953887939453, -15.261114120483398, -14.786592483520508], "reading_logprob_std": 0.3678629653502385, "tokenization_path_logprob_std": 67.60979669207119, "trajectory_balance_rmse": 400.1525936967073} {"canonical_length": 8, "canonical_nll": 3.5133328437805176, "generation_rouge_l_f1": 0.014925373134328361, "generation_token_f1": 0.029850746268656723, "prompt_id": "90063df346a3b4e5", "reading_canonical_log_prob": -28.10666275024414, "reading_distances": [19, 18, 22, 26], "reading_log_probs": [-25.553030014038086, -25.20770835876465, -25.322128295898438, -25.53841209411621], "reading_logprob_std": 0.16882236635592104, "tokenization_path_logprob_std": 30.205760456785082, "trajectory_balance_rmse": 248.2452996121354} {"canonical_length": 276, "canonical_nll": 0.8297838404558707, "generation_rouge_l_f1": 0.23018867924528302, "generation_token_f1": 0.4792452830188679, "prompt_id": "73f47d65b9b8ef25", "reading_canonical_log_prob": -229.0203399658203, "reading_distances": [25, 20, 25, 29], "reading_log_probs": [-237.6953125, -235.81387329101562, -237.90496826171875, -234.62396240234375], "reading_logprob_std": 1.5697803988305592, "tokenization_path_logprob_std": 302.3610158519646, "trajectory_balance_rmse": 4794.398813639503} {"canonical_length": 73, "canonical_nll": 0.7957389518006207, "generation_rouge_l_f1": 0.4457831325301205, "generation_token_f1": 0.5301204819277109, "prompt_id": "96bcd59d4201d57e", "reading_canonical_log_prob": -58.08894348144531, "reading_distances": [19, 20, 18, 20], "reading_log_probs": [-57.20851516723633, -59.57213592529297, -56.50648498535156, -57.5053825378418], "reading_logprob_std": 1.3176734949143987, "tokenization_path_logprob_std": 104.65964664874343, "trajectory_balance_rmse": 1167.7522819725318} {"canonical_length": 159, "canonical_nll": 1.4002930263303361, "generation_rouge_l_f1": 0.25581395348837205, "generation_token_f1": 0.45348837209302323, "prompt_id": "3d7e3fa66ce5f5da", "reading_canonical_log_prob": -222.64659118652344, "reading_distances": [24, 22, 25, 26], "reading_log_probs": [-217.78546142578125, -220.909912109375, -219.50741577148438, -219.36366271972656], "reading_logprob_std": 1.2779079994200304, "tokenization_path_logprob_std": 77.24153482107303, "trajectory_balance_rmse": 3656.1282894612787} {"canonical_length": 40, "canonical_nll": 0.9524213790893554, "generation_rouge_l_f1": 0.1276595744680851, "generation_token_f1": 0.1276595744680851, "prompt_id": "d99d42ddbcbf4292", "reading_canonical_log_prob": -38.09685516357422, "reading_distances": [18, 15, 21, 17], "reading_log_probs": [-37.169429779052734, -37.06711196899414, -37.166561126708984, -38.18336868286133], "reading_logprob_std": 0.5266534325776847, "tokenization_path_logprob_std": 82.6385888868028, "trajectory_balance_rmse": 638.0484764640538} {"canonical_length": 15, "canonical_nll": 0.19843521118164062, "generation_rouge_l_f1": 0.27586206896551724, "generation_token_f1": 0.27586206896551724, "prompt_id": "4bb936b473e4256b", "reading_canonical_log_prob": -2.9765281677246094, "reading_distances": [17, 17, 23, 18], "reading_log_probs": [-6.531533718109131, -6.416877269744873, -3.5943922996520996, -3.1093480587005615], "reading_logprob_std": 1.8141283330344498, "tokenization_path_logprob_std": 33.54145741174852, "trajectory_balance_rmse": 201.15641457682386} {"canonical_length": 378, "canonical_nll": 1.3484156371424438, "generation_rouge_l_f1": 0.29354838709677417, "generation_token_f1": 0.5290322580645161, "prompt_id": "d413361106deda8e", "reading_canonical_log_prob": -509.70111083984375, "reading_distances": [24, 24, 35, 20], "reading_log_probs": [-510.6464538574219, -508.00244140625, -507.89544677734375, -508.7945556640625], "reading_logprob_std": 1.2726500682100605, "tokenization_path_logprob_std": 91.84193324814241, "trajectory_balance_rmse": 7574.3124031980005} {"canonical_length": 370, "canonical_nll": 0.5875781909839527, "generation_rouge_l_f1": 0.3442340791738382, "generation_token_f1": 0.47848537005163516, "prompt_id": "2807bad41faf5ef6", "reading_canonical_log_prob": -217.4039306640625, "reading_distances": [20, 12, 12, 15], "reading_log_probs": [-212.67198181152344, -212.88827514648438, -214.01670837402344, -211.7828369140625], "reading_logprob_std": 0.9188393003161038, "tokenization_path_logprob_std": 161.4421796729645, "trajectory_balance_rmse": 5402.683023227783} {"canonical_length": 76, "canonical_nll": 1.5427467948512028, "generation_rouge_l_f1": 0.18181818181818185, "generation_token_f1": 0.29467084639498436, "prompt_id": "f47b46ea69a09ffc", "reading_canonical_log_prob": -117.2487564086914, "reading_distances": [26, 30, 23, 20], "reading_log_probs": [-112.14834594726562, -113.27227783203125, -112.69046020507812, -113.585693359375], "reading_logprob_std": 0.6364787098808952, "tokenization_path_logprob_std": 64.0973632445446, "trajectory_balance_rmse": 1614.56928816735} {"canonical_length": 148, "canonical_nll": 1.406192057841533, "generation_rouge_l_f1": 0.18105263157894738, "generation_token_f1": 0.2905263157894737, "prompt_id": "abdba20a44568a50", "reading_canonical_log_prob": -208.11642456054688, "reading_distances": [31, 31, 31, 29], "reading_log_probs": [-202.6143798828125, -201.12684631347656, -203.2043914794922, -203.33367919921875], "reading_logprob_std": 1.0116521935098917, "tokenization_path_logprob_std": 187.83824820549557, "trajectory_balance_rmse": 3404.647303163334} {"canonical_length": 19, "canonical_nll": 0.530234136079487, "generation_rouge_l_f1": 0.029197080291970802, "generation_token_f1": 0.029197080291970802, "prompt_id": "f10790becafaebfa", "reading_canonical_log_prob": -10.074448585510254, "reading_distances": [31, 29, 23, 29], "reading_log_probs": [-12.288536071777344, -13.482741355895996, -10.84200382232666, -11.01833724975586], "reading_logprob_std": 1.2318638538806985, "tokenization_path_logprob_std": 85.90247465704054, "trajectory_balance_rmse": 282.147864745884} {"canonical_length": 7, "canonical_nll": 0.0938552703176226, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "3eb075dbec50f70d", "reading_canonical_log_prob": -0.6569868922233582, "reading_distances": [65, 69, 68, 66], "reading_log_probs": [-0.7625159025192261, -0.695590615272522, -0.7762656211853027, -0.6177713871002197], "reading_logprob_std": 0.0726313377446392, "tokenization_path_logprob_std": 50.35346238836162, "trajectory_balance_rmse": 150.27588690518633} {"canonical_length": 5, "canonical_nll": 0.8875579833984375, "generation_rouge_l_f1": 0.8571428571428571, "generation_token_f1": 0.8571428571428571, "prompt_id": "56ce6de59318b6ea", "reading_canonical_log_prob": -4.4377899169921875, "reading_distances": [33, 24, 33, 24], "reading_log_probs": [-8.081525802612305, -6.929797649383545, -8.296419143676758, -7.510912895202637], "reading_logprob_std": 0.6137742039392807, "tokenization_path_logprob_std": 46.6792364479959, "trajectory_balance_rmse": 90.81329589947234} {"canonical_length": 128, "canonical_nll": 1.2590594291687012, "generation_rouge_l_f1": 0.19387755102040816, "generation_token_f1": 0.36224489795918363, "prompt_id": "7b4da759d302cb91", "reading_canonical_log_prob": -161.15960693359375, "reading_distances": [51, 42, 40, 46], "reading_log_probs": [-154.7854461669922, -154.47557067871094, -156.1793212890625, -156.4002685546875], "reading_logprob_std": 0.9705070740002895, "tokenization_path_logprob_std": 56.46981931593848, "trajectory_balance_rmse": 2436.922021195478} {"canonical_length": 179, "canonical_nll": 1.882998162807699, "generation_rouge_l_f1": 0.2305630026809651, "generation_token_f1": 0.4932975871313673, "prompt_id": "e1ce0d3f6081406f", "reading_canonical_log_prob": -337.0566711425781, "reading_distances": [15, 18, 20, 21], "reading_log_probs": [-331.3428955078125, -331.0111083984375, -330.56610107421875, -331.30487060546875], "reading_logprob_std": 0.3588228839320141, "tokenization_path_logprob_std": 229.80122752585268, "trajectory_balance_rmse": 4119.729636821002} {"canonical_length": 193, "canonical_nll": 1.034354017188512, "generation_rouge_l_f1": 0.13192612137203164, "generation_token_f1": 0.24274406332453824, "prompt_id": "50afdbd8ae3129b4", "reading_canonical_log_prob": -199.6303253173828, "reading_distances": [20, 20, 21, 18], "reading_log_probs": [-192.19027709960938, -194.31578063964844, -192.28375244140625, -191.62437438964844], "reading_logprob_std": 1.1780752109378156, "tokenization_path_logprob_std": 38.54677197649151, "trajectory_balance_rmse": 2576.3813659058455} {"canonical_length": 308, "canonical_nll": 0.8768647429230926, "generation_rouge_l_f1": 0.19322033898305083, "generation_token_f1": 0.376271186440678, "prompt_id": "517e8812395e8d56", "reading_canonical_log_prob": -270.0743408203125, "reading_distances": [13, 15, 16, 16], "reading_log_probs": [-267.74652099609375, -268.1999816894531, -266.36181640625, -267.13092041015625], "reading_logprob_std": 0.7966210976708975, "tokenization_path_logprob_std": 276.8924632060853, "trajectory_balance_rmse": 4829.350314408283} {"canonical_length": 9, "canonical_nll": 1.456480450100369, "generation_rouge_l_f1": 0.13333333333333333, "generation_token_f1": 0.26666666666666666, "prompt_id": "7febc4a2c9567c65", "reading_canonical_log_prob": -13.10832405090332, "reading_distances": [33, 34, 34, 39], "reading_log_probs": [-15.111898422241211, -14.633177757263184, -15.31112003326416, -18.302663803100586], "reading_logprob_std": 1.666431170148205, "tokenization_path_logprob_std": 50.77892597161043, "trajectory_balance_rmse": 114.12726251572293} {"canonical_length": 13, "canonical_nll": 0.8237968591543344, "generation_rouge_l_f1": 0.4761904761904762, "generation_token_f1": 0.761904761904762, "prompt_id": "a3b4630e36098b06", "reading_canonical_log_prob": -10.709359169006348, "reading_distances": [39, 44, 46, 44], "reading_log_probs": [-11.17139720916748, -11.162092208862305, -12.63398551940918, -10.94910717010498], "reading_logprob_std": 0.7767084183738611, "tokenization_path_logprob_std": 61.16837018705772, "trajectory_balance_rmse": 351.8618469492327} {"canonical_length": 67, "canonical_nll": 1.0170733323737757, "generation_rouge_l_f1": 0.16617210682492584, "generation_token_f1": 0.2611275964391691, "prompt_id": "0d478580ddc19eb6", "reading_canonical_log_prob": -68.14391326904297, "reading_distances": [26, 22, 27, 24], "reading_log_probs": [-63.39246368408203, -64.15831756591797, -64.40065002441406, -64.44104766845703], "reading_logprob_std": 0.4867231877917323, "tokenization_path_logprob_std": 123.58039253224551, "trajectory_balance_rmse": 1432.4717031485475} {"canonical_length": 5, "canonical_nll": 3.3577220916748045, "generation_rouge_l_f1": 0.4444444444444444, "generation_token_f1": 0.4444444444444444, "prompt_id": "7c79a39b53212c05", "reading_canonical_log_prob": -16.788610458374023, "reading_distances": [33, 32, 33, 33], "reading_log_probs": [-16.596296310424805, -16.187044143676758, -16.18910789489746, -15.927654266357422], "reading_logprob_std": 0.276287484971699, "tokenization_path_logprob_std": 24.060722421692233, "trajectory_balance_rmse": 110.710014023795} {"canonical_length": 89, "canonical_nll": 0.9488767130991046, "generation_rouge_l_f1": 0.33519553072625696, "generation_token_f1": 0.5251396648044693, "prompt_id": "1791e9720c6a20af", "reading_canonical_log_prob": -84.45002746582031, "reading_distances": [25, 25, 25, 20], "reading_log_probs": [-87.19559478759766, -84.73612976074219, -84.33373260498047, -85.67295837402344], "reading_logprob_std": 1.2711786525589177, "tokenization_path_logprob_std": 139.68286951788824, "trajectory_balance_rmse": 2149.6212686153253} {"canonical_length": 14, "canonical_nll": 1.1966077259608678, "generation_rouge_l_f1": 0.12371134020618557, "generation_token_f1": 0.12371134020618557, "prompt_id": "a58d338f47d878b7", "reading_canonical_log_prob": -16.75250816345215, "reading_distances": [9, 9, 9, 8], "reading_log_probs": [-17.915006637573242, -14.43565845489502, -18.378339767456055, -17.978078842163086], "reading_logprob_std": 1.8388902090076844, "tokenization_path_logprob_std": 57.76994006633714, "trajectory_balance_rmse": 280.3813203826576} {"canonical_length": 5, "canonical_nll": 0.8862531661987305, "generation_rouge_l_f1": 0.25, "generation_token_f1": 0.375, "prompt_id": "c979ade41611b06d", "reading_canonical_log_prob": -4.431265830993652, "reading_distances": [30, 29, 36, 38], "reading_log_probs": [-7.244166374206543, -3.695765972137451, -5.309635162353516, -6.479771137237549], "reading_logprob_std": 1.544946957651756, "tokenization_path_logprob_std": 21.9665711585504, "trajectory_balance_rmse": 93.02012892702619} {"canonical_length": 197, "canonical_nll": 1.106795335179053, "generation_rouge_l_f1": 0.2652173913043478, "generation_token_f1": 0.47391304347826085, "prompt_id": "58bae05f46b8fa96", "reading_canonical_log_prob": -218.03868103027344, "reading_distances": [21, 21, 21, 18], "reading_log_probs": [-213.8917694091797, -212.41371154785156, -209.7967529296875, -214.8683319091797], "reading_logprob_std": 2.2079786419145435, "tokenization_path_logprob_std": 115.56617861376056, "trajectory_balance_rmse": 4002.6015164626983} {"canonical_length": 91, "canonical_nll": 1.0542999099899124, "generation_rouge_l_f1": 0.2948717948717948, "generation_token_f1": 0.5128205128205128, "prompt_id": "005445d3abeec9e6", "reading_canonical_log_prob": -95.94129180908203, "reading_distances": [18, 23, 20, 21], "reading_log_probs": [-93.65573120117188, -93.57624053955078, -92.81547546386719, -93.6961669921875], "reading_logprob_std": 0.4166080237095898, "tokenization_path_logprob_std": 144.33144252144294, "trajectory_balance_rmse": 1983.8256640379411} {"canonical_length": 48, "canonical_nll": 1.2239400545756023, "generation_rouge_l_f1": 0.21714285714285717, "generation_token_f1": 0.2514285714285714, "prompt_id": "3ba7853ae081fbb0", "reading_canonical_log_prob": -58.749122619628906, "reading_distances": [23, 12, 18, 19], "reading_log_probs": [-60.78784942626953, -58.59186553955078, -58.492515563964844, -59.556434631347656], "reading_logprob_std": 1.0676872552298462, "tokenization_path_logprob_std": 54.75988952116596, "trajectory_balance_rmse": 873.9023828101215} {"canonical_length": 60, "canonical_nll": 1.004994519551595, "generation_rouge_l_f1": 0.4247787610619469, "generation_token_f1": 0.5486725663716814, "prompt_id": "19c4e3cd0eb041a5", "reading_canonical_log_prob": -60.2996711730957, "reading_distances": [36, 31, 31, 26], "reading_log_probs": [-63.80735778808594, -60.551605224609375, -61.48396301269531, -60.48752975463867], "reading_logprob_std": 1.5514944987864883, "tokenization_path_logprob_std": 163.3986038100577, "trajectory_balance_rmse": 1527.5544063111274} {"canonical_length": 139, "canonical_nll": 1.1757491955654227, "generation_rouge_l_f1": 0.3185185185185185, "generation_token_f1": 0.4962962962962963, "prompt_id": "9a7d94fcb2412368", "reading_canonical_log_prob": -163.42913818359375, "reading_distances": [22, 32, 29, 23], "reading_log_probs": [-161.75912475585938, -162.72235107421875, -163.50271606445312, -160.7386474609375], "reading_logprob_std": 1.1969897761156347, "tokenization_path_logprob_std": 133.13710971297292, "trajectory_balance_rmse": 3186.485137101301} {"canonical_length": 27, "canonical_nll": 1.1895482098614727, "generation_rouge_l_f1": 0.16, "generation_token_f1": 0.18285714285714288, "prompt_id": "2a25293cb727ba1d", "reading_canonical_log_prob": -32.117801666259766, "reading_distances": [31, 27, 32, 29], "reading_log_probs": [-32.70962142944336, -32.17475891113281, -35.08118438720703, -35.14033126831055], "reading_logprob_std": 1.5562820519877454, "tokenization_path_logprob_std": 50.81281578401132, "trajectory_balance_rmse": 676.5454175504458} {"canonical_length": 33, "canonical_nll": 1.5003450567072087, "generation_rouge_l_f1": 0.17142857142857143, "generation_token_f1": 0.17142857142857143, "prompt_id": "cc6712ec80ac0d9e", "reading_canonical_log_prob": -49.51138687133789, "reading_distances": [29, 23, 29, 25], "reading_log_probs": [-52.551300048828125, -52.12965774536133, -51.12934875488281, -50.163841247558594], "reading_logprob_std": 1.0683688723294444, "tokenization_path_logprob_std": 66.08251670268157, "trajectory_balance_rmse": 669.9761840143199} {"canonical_length": 48, "canonical_nll": 0.6088259220123291, "generation_rouge_l_f1": 0.1273885350318471, "generation_token_f1": 0.17197452229299362, "prompt_id": "aa6e8f58031f9e55", "reading_canonical_log_prob": -29.223644256591797, "reading_distances": [30, 38, 36, 39], "reading_log_probs": [-30.548891067504883, -34.013240814208984, -30.642505645751953, -32.10600280761719], "reading_logprob_std": 1.6221466271680873, "tokenization_path_logprob_std": 114.72786917038681, "trajectory_balance_rmse": 1006.5429447679974} {"canonical_length": 259, "canonical_nll": 1.4717039601683157, "generation_rouge_l_f1": 0.31297709923664124, "generation_token_f1": 0.549618320610687, "prompt_id": "ef58deb0251c5bd3", "reading_canonical_log_prob": -381.17132568359375, "reading_distances": [24, 25, 18, 23], "reading_log_probs": [-381.5735778808594, -379.6847229003906, -382.0949401855469, -380.87103271484375], "reading_logprob_std": 1.0427363574852497, "tokenization_path_logprob_std": 207.29941990333649, "trajectory_balance_rmse": 5158.86528619488} {"canonical_length": 136, "canonical_nll": 1.5085150774787455, "generation_rouge_l_f1": 0.13861386138613863, "generation_token_f1": 0.2574257425742574, "prompt_id": "209efb2fe475a408", "reading_canonical_log_prob": -205.15805053710938, "reading_distances": [25, 28, 23, 24], "reading_log_probs": [-203.8952178955078, -202.9868927001953, -205.1484375, -207.39764404296875], "reading_logprob_std": 1.9115555698221043, "tokenization_path_logprob_std": 75.85573053651218, "trajectory_balance_rmse": 3338.7722159163486} {"canonical_length": 16, "canonical_nll": 0.919676661491394, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.6666666666666666, "prompt_id": "aa1f86a1c5954cf7", "reading_canonical_log_prob": -14.714826583862305, "reading_distances": [28, 29, 27, 28], "reading_log_probs": [-14.131125450134277, -16.072542190551758, -14.470837593078613, -13.86727523803711], "reading_logprob_std": 0.9894050274577431, "tokenization_path_logprob_std": 20.438463790648026, "trajectory_balance_rmse": 156.3657465055499} {"canonical_length": 8, "canonical_nll": 0.016416652128100395, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "878c5da5498115bc", "reading_canonical_log_prob": -0.13133321702480316, "reading_distances": [32, 33, 32, 30], "reading_log_probs": [-0.1493368148803711, -0.23883648216724396, -0.2706390917301178, -0.2567738890647888], "reading_logprob_std": 0.05461422136897201, "tokenization_path_logprob_std": 63.243207118130115, "trajectory_balance_rmse": 142.03563169798574} {"canonical_length": 30, "canonical_nll": 0.06074288288752238, "generation_rouge_l_f1": 0.9056603773584906, "generation_token_f1": 0.9056603773584906, "prompt_id": "3e49b0b70cea783a", "reading_canonical_log_prob": -1.8222864866256714, "reading_distances": [41, 34, 38, 39], "reading_log_probs": [-3.693582057952881, -2.132981777191162, -2.743701219558716, -2.3720450401306152], "reading_logprob_std": 0.6863226118000564, "tokenization_path_logprob_std": 94.2194071128941, "trajectory_balance_rmse": 559.2032564503912} {"canonical_length": 406, "canonical_nll": 1.25861497700508, "generation_rouge_l_f1": 0.192846034214619, "generation_token_f1": 0.39191290824261277, "prompt_id": "50289ab6cb9cf49e", "reading_canonical_log_prob": -510.9976806640625, "reading_distances": [25, 21, 20, 23], "reading_log_probs": [-510.5916748046875, -509.2821044921875, -509.9855651855469, -509.65087890625], "reading_logprob_std": 0.5560505535644026, "tokenization_path_logprob_std": 199.06925987790842, "trajectory_balance_rmse": 6756.271148028426} {"canonical_length": 168, "canonical_nll": 0.2890401567731585, "generation_rouge_l_f1": 0.2602409638554217, "generation_token_f1": 0.48192771084337355, "prompt_id": "97290896fed51229", "reading_canonical_log_prob": -48.558746337890625, "reading_distances": [30, 30, 34, 39], "reading_log_probs": [-51.26647186279297, -49.69440460205078, -48.196388244628906, -49.70482635498047], "reading_logprob_std": 1.2534981963717369, "tokenization_path_logprob_std": 76.06404725040476, "trajectory_balance_rmse": 1399.6730475444374} {"canonical_length": 36, "canonical_nll": 1.074567159016927, "generation_rouge_l_f1": 0.1357142857142857, "generation_token_f1": 0.1714285714285714, "prompt_id": "11c3ac95688ce9b6", "reading_canonical_log_prob": -38.684417724609375, "reading_distances": [23, 21, 17, 18], "reading_log_probs": [-39.61922836303711, -40.62434005737305, -40.044151306152344, -41.175262451171875], "reading_logprob_std": 0.6789451378507095, "tokenization_path_logprob_std": 43.79109314081845, "trajectory_balance_rmse": 615.9720683684902} {"canonical_length": 12, "canonical_nll": 0.4887211322784424, "generation_rouge_l_f1": 0.1818181818181818, "generation_token_f1": 0.1818181818181818, "prompt_id": "b6d0ecb07aebae2f", "reading_canonical_log_prob": -5.864653587341309, "reading_distances": [26, 27, 26, 27], "reading_log_probs": [-5.860984802246094, -6.403704643249512, -7.326418876647949, -6.251838684082031], "reading_logprob_std": 0.6207519650598478, "tokenization_path_logprob_std": 49.62412248278326, "trajectory_balance_rmse": 246.3246154210215} {"canonical_length": 27, "canonical_nll": 1.260637777822989, "generation_rouge_l_f1": 0.12173913043478259, "generation_token_f1": 0.19130434782608693, "prompt_id": "a7179a571daab3f1", "reading_canonical_log_prob": -34.0372200012207, "reading_distances": [28, 19, 18, 19], "reading_log_probs": [-35.85374450683594, -38.175376892089844, -38.190093994140625, -36.021793365478516], "reading_logprob_std": 1.2979603388079266, "tokenization_path_logprob_std": 55.993962836528446, "trajectory_balance_rmse": 400.33313810099503} {"canonical_length": 226, "canonical_nll": 0.995068980529245, "generation_rouge_l_f1": 0.2469135802469136, "generation_token_f1": 0.32592592592592595, "prompt_id": "a32be7c1cc7860a7", "reading_canonical_log_prob": -224.88558959960938, "reading_distances": [38, 29, 34, 38], "reading_log_probs": [-232.52459716796875, -228.02200317382812, -228.9822998046875, -229.06756591796875], "reading_logprob_std": 1.9747347866381517, "tokenization_path_logprob_std": 186.36127169985244, "trajectory_balance_rmse": 3435.378067702152} {"canonical_length": 317, "canonical_nll": 1.1691513302198344, "generation_rouge_l_f1": 0.2141623488773748, "generation_token_f1": 0.46632124352331605, "prompt_id": "7ebfd1446f72fed6", "reading_canonical_log_prob": -370.6209716796875, "reading_distances": [25, 29, 19, 24], "reading_log_probs": [-365.3528747558594, -364.8117370605469, -365.9241027832031, -366.3398742675781], "reading_logprob_std": 0.6670036389404377, "tokenization_path_logprob_std": 291.30372940764477, "trajectory_balance_rmse": 7178.837040357724} {"canonical_length": 7, "canonical_nll": 1.0483621869768416, "generation_rouge_l_f1": 0.5000000000000001, "generation_token_f1": 0.5000000000000001, "prompt_id": "05c61d7a9c151e3a", "reading_canonical_log_prob": -7.338535308837891, "reading_distances": [48, 57, 44, 50], "reading_log_probs": [-4.97114896774292, -5.225531578063965, -5.218755722045898, -4.0474629402160645], "reading_logprob_std": 0.5581990313447647, "tokenization_path_logprob_std": 55.863702731006164, "trajectory_balance_rmse": 200.31028300745854} {"canonical_length": 20, "canonical_nll": 1.6434555053710938, "generation_rouge_l_f1": 0.625, "generation_token_f1": 0.7499999999999999, "prompt_id": "3b1c1940eb0db061", "reading_canonical_log_prob": -32.869110107421875, "reading_distances": [19, 25, 18, 21], "reading_log_probs": [-32.14609909057617, -30.649106979370117, -30.68578338623047, -31.47515869140625], "reading_logprob_std": 0.7147545218936774, "tokenization_path_logprob_std": 59.07191424401351, "trajectory_balance_rmse": 507.77198737342775} {"canonical_length": 191, "canonical_nll": 1.4654157548674738, "generation_rouge_l_f1": 0.2118226600985222, "generation_token_f1": 0.3694581280788178, "prompt_id": "13ae7fb963a43898", "reading_canonical_log_prob": -279.8944091796875, "reading_distances": [42, 40, 46, 36], "reading_log_probs": [-286.0369567871094, -287.3468933105469, -287.07763671875, -286.7315673828125], "reading_logprob_std": 0.5665922541874485, "tokenization_path_logprob_std": 61.041166824937825, "trajectory_balance_rmse": 3113.864409251687} {"canonical_length": 346, "canonical_nll": 1.8052539274182622, "generation_rouge_l_f1": 0.16271186440677965, "generation_token_f1": 0.3627118644067797, "prompt_id": "6aeff1f7dd4bd317", "reading_canonical_log_prob": -624.6178588867188, "reading_distances": [22, 24, 20, 26], "reading_log_probs": [-615.2203979492188, -615.5869140625, -614.2637939453125, -613.7327880859375], "reading_logprob_std": 0.8530731500032013, "tokenization_path_logprob_std": 510.7287886566689, "trajectory_balance_rmse": 6376.711855282287} {"canonical_length": 35, "canonical_nll": 2.8228057861328124, "generation_rouge_l_f1": 0.35443037974683544, "generation_token_f1": 0.4050632911392405, "prompt_id": "a5fa29779141ee07", "reading_canonical_log_prob": -98.79820251464844, "reading_distances": [15, 16, 16, 13], "reading_log_probs": [-98.57360076904297, -98.08596801757812, -97.08136749267578, -98.19329833984375], "reading_logprob_std": 0.6368089725766103, "tokenization_path_logprob_std": 105.84485233428053, "trajectory_balance_rmse": 737.9977005941114} {"canonical_length": 44, "canonical_nll": 1.5757226076993076, "generation_rouge_l_f1": 0.09302325581395349, "generation_token_f1": 0.09302325581395349, "prompt_id": "635690e839a377b9", "reading_canonical_log_prob": -69.33179473876953, "reading_distances": [32, 30, 23, 35], "reading_log_probs": [-69.55438232421875, -71.56907653808594, -70.0868911743164, -69.69013214111328], "reading_logprob_std": 0.9240179333362974, "tokenization_path_logprob_std": 100.72674694826976, "trajectory_balance_rmse": 1318.0226185181104} {"canonical_length": 124, "canonical_nll": 1.6381102531186995, "generation_rouge_l_f1": 0.14646464646464646, "generation_token_f1": 0.2525252525252525, "prompt_id": "908852670070fd4b", "reading_canonical_log_prob": -203.12567138671875, "reading_distances": [21, 20, 23, 20], "reading_log_probs": [-198.59024047851562, -194.73159790039062, -196.0224151611328, -196.17120361328125], "reading_logprob_std": 1.6097463716991842, "tokenization_path_logprob_std": 104.87844396461811, "trajectory_balance_rmse": 2734.880467522949} {"canonical_length": 14, "canonical_nll": 2.703188214983259, "generation_rouge_l_f1": 0.28571428571428575, "generation_token_f1": 0.28571428571428575, "prompt_id": "410cbaa296c6e2c6", "reading_canonical_log_prob": -37.844635009765625, "reading_distances": [25, 24, 23, 21], "reading_log_probs": [-36.50288772583008, -36.4259033203125, -35.98847961425781, -36.381465911865234], "reading_logprob_std": 0.22968057036640918, "tokenization_path_logprob_std": 31.713097483906125, "trajectory_balance_rmse": 372.7203477385396} {"canonical_length": 24, "canonical_nll": 1.4215577443440754, "generation_rouge_l_f1": 0.17391304347826086, "generation_token_f1": 0.17391304347826086, "prompt_id": "b9eb5c3d26d069d5", "reading_canonical_log_prob": -34.11738586425781, "reading_distances": [15, 12, 16, 17], "reading_log_probs": [-36.35284423828125, -35.248207092285156, -37.981101989746094, -33.853816986083984], "reading_logprob_std": 1.7455682820368805, "tokenization_path_logprob_std": 86.27645899588319, "trajectory_balance_rmse": 500.38871366284513} {"canonical_length": 120, "canonical_nll": 0.8014169057210286, "generation_rouge_l_f1": 0.29444444444444445, "generation_token_f1": 0.4166666666666667, "prompt_id": "fd96f68afab8f1c0", "reading_canonical_log_prob": -96.17002868652344, "reading_distances": [32, 25, 24, 30], "reading_log_probs": [-94.86656951904297, -94.97254943847656, -94.46875, -94.4085693359375], "reading_logprob_std": 0.2820705324582913, "tokenization_path_logprob_std": 51.387258524749946, "trajectory_balance_rmse": 1410.3540600000142} {"canonical_length": 92, "canonical_nll": 2.5277262148649795, "generation_rouge_l_f1": 0.2028985507246377, "generation_token_f1": 0.39855072463768115, "prompt_id": "930a5444ff77055e", "reading_canonical_log_prob": -232.55081176757812, "reading_distances": [46, 50, 45, 43], "reading_log_probs": [-228.02593994140625, -229.85594177246094, -226.94203186035156, -232.02523803710938], "reading_logprob_std": 2.227738969285329, "tokenization_path_logprob_std": 236.4481754699154, "trajectory_balance_rmse": 2251.206961116194} {"canonical_length": 88, "canonical_nll": 0.7759366468949751, "generation_rouge_l_f1": 0.24464831804281348, "generation_token_f1": 0.37920489296636084, "prompt_id": "6881f74e519464d9", "reading_canonical_log_prob": -68.28242492675781, "reading_distances": [14, 19, 18, 17], "reading_log_probs": [-72.92509460449219, -71.8809814453125, -71.72774505615234, -72.30777740478516], "reading_logprob_std": 0.5359549470012592, "tokenization_path_logprob_std": 212.24346791313408, "trajectory_balance_rmse": 1686.4954099552854} {"canonical_length": 242, "canonical_nll": 1.394437805680204, "generation_rouge_l_f1": 0.203921568627451, "generation_token_f1": 0.5019607843137254, "prompt_id": "1afba92bdf32d325", "reading_canonical_log_prob": -337.4539489746094, "reading_distances": [32, 28, 26, 33], "reading_log_probs": [-330.1142272949219, -330.23394775390625, -331.810302734375, -331.2677001953125], "reading_logprob_std": 0.8200336002521578, "tokenization_path_logprob_std": 371.39817843099195, "trajectory_balance_rmse": 4982.173201462332} {"canonical_length": 114, "canonical_nll": 1.3869130987870066, "generation_rouge_l_f1": 0.2920110192837465, "generation_token_f1": 0.4242424242424243, "prompt_id": "19d150b4a3d84266", "reading_canonical_log_prob": -158.10809326171875, "reading_distances": [35, 36, 33, 35], "reading_log_probs": [-159.261474609375, -155.80587768554688, -158.57989501953125, -154.70660400390625], "reading_logprob_std": 2.1805676852206037, "tokenization_path_logprob_std": 60.28034458313481, "trajectory_balance_rmse": 2350.17050959769} {"canonical_length": 6, "canonical_nll": 2.1018800735473633, "generation_rouge_l_f1": 0.6153846153846154, "generation_token_f1": 0.6153846153846154, "prompt_id": "dee72943ee9496d3", "reading_canonical_log_prob": -12.61128044128418, "reading_distances": [38, 45, 40, 41], "reading_log_probs": [-8.008638381958008, -6.15672492980957, -8.801868438720703, -9.450928688049316], "reading_logprob_std": 1.4262082735654689, "tokenization_path_logprob_std": 21.10213484904655, "trajectory_balance_rmse": 136.48018095412505} {"canonical_length": 40, "canonical_nll": 0.7648662567138672, "generation_rouge_l_f1": 0.18320610687022898, "generation_token_f1": 0.22900763358778625, "prompt_id": "f1a8da244ce8319b", "reading_canonical_log_prob": -30.594650268554688, "reading_distances": [21, 23, 19, 25], "reading_log_probs": [-33.025184631347656, -32.16594696044922, -33.854000091552734, -31.787940979003906], "reading_logprob_std": 0.9227228027877467, "tokenization_path_logprob_std": 26.66701863684155, "trajectory_balance_rmse": 621.3260413961806} {"canonical_length": 286, "canonical_nll": 1.478665118450885, "generation_rouge_l_f1": 0.30280373831775703, "generation_token_f1": 0.5271028037383177, "prompt_id": "201696a6ff512973", "reading_canonical_log_prob": -422.8982238769531, "reading_distances": [23, 21, 20, 17], "reading_log_probs": [-418.1843566894531, -422.8741455078125, -418.82061767578125, -418.3397216796875], "reading_logprob_std": 2.2294699254325967, "tokenization_path_logprob_std": 196.08295279125963, "trajectory_balance_rmse": 4384.432953045076} {"canonical_length": 202, "canonical_nll": 1.9481542606164914, "generation_rouge_l_f1": 0.25411764705882356, "generation_token_f1": 0.41882352941176476, "prompt_id": "8d88867a9f02da1b", "reading_canonical_log_prob": -393.52716064453125, "reading_distances": [25, 29, 29, 32], "reading_log_probs": [-395.546630859375, -393.4971618652344, -391.48370361328125, -392.720703125], "reading_logprob_std": 1.7049105212234614, "tokenization_path_logprob_std": 71.10655568316231, "trajectory_balance_rmse": 3343.825598312509} {"canonical_length": 71, "canonical_nll": 0.6919797440649758, "generation_rouge_l_f1": 0.24124513618677043, "generation_token_f1": 0.32684824902723736, "prompt_id": "de7efec5b0e0f3a3", "reading_canonical_log_prob": -49.13056182861328, "reading_distances": [21, 25, 25, 22], "reading_log_probs": [-49.78679656982422, -48.71100616455078, -48.86865997314453, -47.347251892089844], "reading_logprob_std": 1.0062736965748575, "tokenization_path_logprob_std": 31.73992301146958, "trajectory_balance_rmse": 849.8071898472906} {"canonical_length": 8, "canonical_nll": 3.045827865600586, "generation_rouge_l_f1": 0.25, "generation_token_f1": 0.25, "prompt_id": "5d9fef90efd251af", "reading_canonical_log_prob": -24.366622924804688, "reading_distances": [11, 13, 11, 15], "reading_log_probs": [-21.468738555908203, -22.605127334594727, -23.40018081665039, -21.869388580322266], "reading_logprob_std": 0.8514260356228944, "tokenization_path_logprob_std": 27.05788864193112, "trajectory_balance_rmse": 200.78283667119183} {"canonical_length": 35, "canonical_nll": 1.7170892987932478, "generation_rouge_l_f1": 0.2153846153846154, "generation_token_f1": 0.27692307692307694, "prompt_id": "207d314006c9124a", "reading_canonical_log_prob": -60.09812545776367, "reading_distances": [27, 21, 21, 30], "reading_log_probs": [-64.23787689208984, -64.00872802734375, -64.14458465576172, -60.485435485839844], "reading_logprob_std": 1.8249075126143404, "tokenization_path_logprob_std": 55.62175512728529, "trajectory_balance_rmse": 839.368500857044} {"canonical_length": 243, "canonical_nll": 0.7182791753070344, "generation_rouge_l_f1": 0.38257575757575757, "generation_token_f1": 0.5984848484848485, "prompt_id": "182ccb8ddd147837", "reading_canonical_log_prob": -174.54183959960938, "reading_distances": [16, 6, 14, 12], "reading_log_probs": [-173.74688720703125, -175.1151123046875, -172.03701782226562, -174.7436981201172], "reading_logprob_std": 1.3762284694427345, "tokenization_path_logprob_std": 125.81181379072828, "trajectory_balance_rmse": 1708.8674608354604} {"canonical_length": 8, "canonical_nll": 0.8601270914077759, "generation_rouge_l_f1": 0.0380952380952381, "generation_token_f1": 0.05714285714285715, "prompt_id": "e6557b0f23801243", "reading_canonical_log_prob": -6.881016731262207, "reading_distances": [8, 6, 8, 9], "reading_log_probs": [-6.291647911071777, -6.388086318969727, -6.366341590881348, -6.537852764129639], "reading_logprob_std": 0.1032049925452207, "tokenization_path_logprob_std": 31.354072510626953, "trajectory_balance_rmse": 219.20133756889624} {"canonical_length": 56, "canonical_nll": 0.6542481013706752, "generation_rouge_l_f1": 0.19847328244274812, "generation_token_f1": 0.2595419847328244, "prompt_id": "f461f97cdf966620", "reading_canonical_log_prob": -36.63789367675781, "reading_distances": [24, 22, 20, 22], "reading_log_probs": [-41.097251892089844, -41.956817626953125, -42.10588073730469, -41.47334289550781], "reading_logprob_std": 0.46131167978938636, "tokenization_path_logprob_std": 20.463090025127585, "trajectory_balance_rmse": 938.6650880560811} {"canonical_length": 319, "canonical_nll": 1.0529101141567887, "generation_rouge_l_f1": 0.22564102564102564, "generation_token_f1": 0.4923076923076923, "prompt_id": "e632ea14cc0c360e", "reading_canonical_log_prob": -335.8783264160156, "reading_distances": [12, 15, 14, 15], "reading_log_probs": [-332.27899169921875, -334.90203857421875, -330.8017578125, -333.7598876953125], "reading_logprob_std": 1.7823905102992876, "tokenization_path_logprob_std": 236.07546062217006, "trajectory_balance_rmse": 5660.660205378029} {"canonical_length": 123, "canonical_nll": 1.3321417831793063, "generation_rouge_l_f1": 0.2421052631578947, "generation_token_f1": 0.431578947368421, "prompt_id": "fbb4ad27a6d06f7c", "reading_canonical_log_prob": -163.8534393310547, "reading_distances": [43, 40, 44, 45], "reading_log_probs": [-160.27603149414062, -157.2282257080078, -160.94168090820312, -157.85528564453125], "reading_logprob_std": 1.8097193413228008, "tokenization_path_logprob_std": 179.09796448133406, "trajectory_balance_rmse": 2392.2164803027795} {"canonical_length": 285, "canonical_nll": 1.0052019085800439, "generation_rouge_l_f1": 0.24609375000000003, "generation_token_f1": 0.46874999999999994, "prompt_id": "db0aefd45e9cb0fd", "reading_canonical_log_prob": -286.4825439453125, "reading_distances": [23, 26, 25, 26], "reading_log_probs": [-284.3802795410156, -286.1973876953125, -282.7650146484375, -284.27056884765625], "reading_logprob_std": 1.404857887392516, "tokenization_path_logprob_std": 172.01289792526396, "trajectory_balance_rmse": 4713.302983748882} {"canonical_length": 32, "canonical_nll": 0.16433554887771606, "generation_rouge_l_f1": 0.7804878048780488, "generation_token_f1": 0.7804878048780488, "prompt_id": "5f8ab685b6a4b0df", "reading_canonical_log_prob": -5.258737564086914, "reading_distances": [27, 24, 32, 28], "reading_log_probs": [-6.3167314529418945, -5.6602349281311035, -6.133044719696045, -6.789584159851074], "reading_logprob_std": 0.46711356650495667, "tokenization_path_logprob_std": 100.88270141763506, "trajectory_balance_rmse": 388.62307390528446} {"canonical_length": 12, "canonical_nll": 1.242853005727132, "generation_rouge_l_f1": 0.28571428571428564, "generation_token_f1": 0.28571428571428564, "prompt_id": "6c7141219130f618", "reading_canonical_log_prob": -14.914236068725586, "reading_distances": [22, 20, 26, 20], "reading_log_probs": [-22.661090850830078, -17.45524024963379, -20.947246551513672, -18.843547821044922], "reading_logprob_std": 2.2941743580628753, "tokenization_path_logprob_std": 15.356831853638177, "trajectory_balance_rmse": 305.7885582131583} {"canonical_length": 8, "canonical_nll": 0.09430202096700668, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "1d180306e638dc8c", "reading_canonical_log_prob": -0.7544161677360535, "reading_distances": [24, 26, 26, 24], "reading_log_probs": [-1.1512701511383057, -1.461598515510559, -1.388035774230957, -1.0312979221343994], "reading_logprob_std": 0.20095334500564332, "tokenization_path_logprob_std": 19.83166714066437, "trajectory_balance_rmse": 123.56868881844169} {"canonical_length": 162, "canonical_nll": 1.265681325653453, "generation_rouge_l_f1": 0.3036649214659686, "generation_token_f1": 0.4607329842931937, "prompt_id": "042d56a64e3fa50f", "reading_canonical_log_prob": -205.04037475585938, "reading_distances": [29, 27, 31, 21], "reading_log_probs": [-205.71954345703125, -210.4783935546875, -204.45974731445312, -207.01734924316406], "reading_logprob_std": 2.592654661868466, "tokenization_path_logprob_std": 81.22475146197594, "trajectory_balance_rmse": 3367.6428875216743} {"canonical_length": 278, "canonical_nll": 1.4532048067600607, "generation_rouge_l_f1": 0.21973929236499068, "generation_token_f1": 0.46927374301675984, "prompt_id": "519601f060a2cff2", "reading_canonical_log_prob": -403.9909362792969, "reading_distances": [22, 21, 30, 27], "reading_log_probs": [-399.49395751953125, -400.0300598144531, -400.281982421875, -401.29278564453125], "reading_logprob_std": 0.7540908553590863, "tokenization_path_logprob_std": 122.14854848295691, "trajectory_balance_rmse": 5836.294907708362} {"canonical_length": 41, "canonical_nll": 1.4728676400533536, "generation_rouge_l_f1": 0.1793103448275862, "generation_token_f1": 0.2896551724137931, "prompt_id": "acf7f177082b10c2", "reading_canonical_log_prob": -60.3875732421875, "reading_distances": [10, 11, 15, 17], "reading_log_probs": [-59.96269989013672, -60.053218841552734, -59.31158447265625, -59.89729690551758], "reading_logprob_std": 0.3358837545834726, "tokenization_path_logprob_std": 160.8278202836844, "trajectory_balance_rmse": 907.6223644418471} {"canonical_length": 10, "canonical_nll": 0.9611501693725586, "generation_rouge_l_f1": 0.4545454545454546, "generation_token_f1": 0.6363636363636364, "prompt_id": "78f8a9aafe333ebf", "reading_canonical_log_prob": -9.611501693725586, "reading_distances": [23, 24, 29, 25], "reading_log_probs": [-9.136781692504883, -9.698192596435547, -9.737598419189453, -9.691309928894043], "reading_logprob_std": 0.286851779580545, "tokenization_path_logprob_std": 45.82620599502683, "trajectory_balance_rmse": 291.3493380517692} {"canonical_length": 303, "canonical_nll": 1.0769082248801052, "generation_rouge_l_f1": 0.17957746478873238, "generation_token_f1": 0.38028169014084506, "prompt_id": "c2a3bc6c2ee0dd9e", "reading_canonical_log_prob": -326.3031921386719, "reading_distances": [30, 28, 31, 32], "reading_log_probs": [-323.30218505859375, -324.785888671875, -324.4504699707031, -324.47735595703125], "reading_logprob_std": 0.6525196222759506, "tokenization_path_logprob_std": 16.94201606049953, "trajectory_balance_rmse": 6113.541296557338} {"canonical_length": 335, "canonical_nll": 1.2678651724288712, "generation_rouge_l_f1": 0.26306913996627324, "generation_token_f1": 0.4957841483979764, "prompt_id": "994646dcabc2bac8", "reading_canonical_log_prob": -424.7348327636719, "reading_distances": [20, 18, 23, 20], "reading_log_probs": [-422.09552001953125, -421.15673828125, -419.5678405761719, -420.4600830078125], "reading_logprob_std": 1.070480802561878, "tokenization_path_logprob_std": 287.94161900766255, "trajectory_balance_rmse": 7017.336651275223} {"canonical_length": 10, "canonical_nll": 1.058049774169922, "generation_rouge_l_f1": 0.7000000000000001, "generation_token_f1": 0.7000000000000001, "prompt_id": "bb2e66b420114e36", "reading_canonical_log_prob": -10.580497741699219, "reading_distances": [25, 24, 29, 26], "reading_log_probs": [-10.288240432739258, -10.72592830657959, -10.111196517944336, -10.151394844055176], "reading_logprob_std": 0.2815495389737804, "tokenization_path_logprob_std": 14.994420089857787, "trajectory_balance_rmse": 207.00919373582732} {"canonical_length": 282, "canonical_nll": 0.9263271034186613, "generation_rouge_l_f1": 0.3088803088803089, "generation_token_f1": 0.5714285714285715, "prompt_id": "57dc4fcf5ed48b59", "reading_canonical_log_prob": -261.2242431640625, "reading_distances": [20, 17, 18, 18], "reading_log_probs": [-262.3489074707031, -262.9910888671875, -262.3102722167969, -262.46746826171875], "reading_logprob_std": 0.31495326855224676, "tokenization_path_logprob_std": 254.36424693952856, "trajectory_balance_rmse": 5724.934438691443} {"canonical_length": 15, "canonical_nll": 0.968563461303711, "generation_rouge_l_f1": 0.6, "generation_token_f1": 0.6, "prompt_id": "3b7a47e13ee4a0c9", "reading_canonical_log_prob": -14.528451919555664, "reading_distances": [16, 17, 24, 21], "reading_log_probs": [-14.170165061950684, -13.405583381652832, -13.817342758178711, -13.278301239013672], "reading_logprob_std": 0.40628418727181376, "tokenization_path_logprob_std": 25.95325343173494, "trajectory_balance_rmse": 234.61501580434282} {"canonical_length": 38, "canonical_nll": 1.3296249791195518, "generation_rouge_l_f1": 0.20338983050847456, "generation_token_f1": 0.3050847457627119, "prompt_id": "4ae552c8baec5eb5", "reading_canonical_log_prob": -50.52574920654297, "reading_distances": [40, 48, 42, 38], "reading_log_probs": [-60.90105438232422, -55.876712799072266, -58.667720794677734, -61.37510299682617], "reading_logprob_std": 2.5134380195032975, "tokenization_path_logprob_std": 86.02249608428225, "trajectory_balance_rmse": 760.0197477382115} {"canonical_length": 27, "canonical_nll": 1.148967460349754, "generation_rouge_l_f1": 0.3111111111111111, "generation_token_f1": 0.39999999999999997, "prompt_id": "17ccb345d2ac2714", "reading_canonical_log_prob": -31.02212142944336, "reading_distances": [33, 36, 31, 36], "reading_log_probs": [-27.009048461914062, -26.54339599609375, -28.021343231201172, -29.396434783935547], "reading_logprob_std": 1.2634681602147964, "tokenization_path_logprob_std": 86.01214115398491, "trajectory_balance_rmse": 628.3988343985652} {"canonical_length": 3, "canonical_nll": 2.723395347595215, "generation_rouge_l_f1": 0.06349206349206349, "generation_token_f1": 0.06349206349206349, "prompt_id": "5e428837a52a1f42", "reading_canonical_log_prob": -8.170186042785645, "reading_distances": [38, 45, 47, 50], "reading_log_probs": [-8.620492935180664, -9.605361938476562, -8.531559944152832, -8.426810264587402], "reading_logprob_std": 0.5453130808569188, "tokenization_path_logprob_std": 26.432720215319044, "trajectory_balance_rmse": 53.671368506692765} {"canonical_length": 22, "canonical_nll": 1.2297479456121272, "generation_rouge_l_f1": 0.5714285714285714, "generation_token_f1": 0.5714285714285714, "prompt_id": "6a0a62922b0dc5d9", "reading_canonical_log_prob": -27.054454803466797, "reading_distances": [25, 21, 22, 14], "reading_log_probs": [-30.400941848754883, -29.877105712890625, -27.163619995117188, -30.492218017578125], "reading_logprob_std": 1.57013590450375, "tokenization_path_logprob_std": 68.35830706389798, "trajectory_balance_rmse": 623.3651259174227} {"canonical_length": 138, "canonical_nll": 1.7378572104633718, "generation_rouge_l_f1": 0.3181818181818182, "generation_token_f1": 0.44886363636363635, "prompt_id": "b8e84fe41fdcb18b", "reading_canonical_log_prob": -239.8242950439453, "reading_distances": [64, 67, 61, 50], "reading_log_probs": [-228.82928466796875, -226.3885498046875, -228.7061309814453, -227.4088134765625], "reading_logprob_std": 1.1577724693148657, "tokenization_path_logprob_std": 172.0220741812067, "trajectory_balance_rmse": 2273.8059106556957} {"canonical_length": 16, "canonical_nll": 0.7103255987167358, "generation_rouge_l_f1": 0.3870967741935484, "generation_token_f1": 0.45161290322580644, "prompt_id": "b63558947632bbd1", "reading_canonical_log_prob": -11.365209579467773, "reading_distances": [27, 27, 22, 28], "reading_log_probs": [-8.910209655761719, -10.537578582763672, -9.285980224609375, -9.005202293395996], "reading_logprob_std": 0.7523335106628543, "tokenization_path_logprob_std": 49.703759440518944, "trajectory_balance_rmse": 276.6123313864731} {"canonical_length": 145, "canonical_nll": 1.2663471090382543, "generation_rouge_l_f1": 0.324455205811138, "generation_token_f1": 0.5326876513317191, "prompt_id": "f6960da157e7277f", "reading_canonical_log_prob": -183.62033081054688, "reading_distances": [18, 13, 15, 21], "reading_log_probs": [-181.0616455078125, -181.63687133789062, -182.03318786621094, -180.82022094726562], "reading_logprob_std": 0.5498741163340806, "tokenization_path_logprob_std": 176.46268395576232, "trajectory_balance_rmse": 3058.076658457461} {"canonical_length": 147, "canonical_nll": 1.401932074099171, "generation_rouge_l_f1": 0.33582089552238803, "generation_token_f1": 0.5149253731343284, "prompt_id": "92be2f9693f01c35", "reading_canonical_log_prob": -206.08401489257812, "reading_distances": [70, 65, 62, 72], "reading_log_probs": [-194.93011474609375, -193.38975524902344, -193.49984741210938, -197.40682983398438], "reading_logprob_std": 1.8700715183252308, "tokenization_path_logprob_std": 128.54236101925022, "trajectory_balance_rmse": 2454.5483655558055} {"canonical_length": 21, "canonical_nll": 0.280702999659947, "generation_rouge_l_f1": 0.3140495867768595, "generation_token_f1": 0.3305785123966942, "prompt_id": "45f3d6045c73150a", "reading_canonical_log_prob": -5.894762992858887, "reading_distances": [33, 34, 22, 17], "reading_log_probs": [-5.88525390625, -6.004828453063965, -6.156470775604248, -5.88098669052124], "reading_logprob_std": 0.1297749747535074, "tokenization_path_logprob_std": 84.01144216532245, "trajectory_balance_rmse": 324.4976855472152} {"canonical_length": 98, "canonical_nll": 1.1597463646713568, "generation_rouge_l_f1": 0.2681992337164751, "generation_token_f1": 0.4521072796934866, "prompt_id": "949e9bf0a5c54d25", "reading_canonical_log_prob": -113.65514373779297, "reading_distances": [17, 15, 17, 14], "reading_log_probs": [-113.39680480957031, -112.61001586914062, -111.95274353027344, -112.9738998413086], "reading_logprob_std": 0.6117173306584677, "tokenization_path_logprob_std": 128.45823067390359, "trajectory_balance_rmse": 1998.5186756905568} {"canonical_length": 14, "canonical_nll": 0.7966723442077637, "generation_rouge_l_f1": 0.1935483870967742, "generation_token_f1": 0.20967741935483872, "prompt_id": "5602e1c51a5b2686", "reading_canonical_log_prob": -11.153412818908691, "reading_distances": [19, 17, 19, 20], "reading_log_probs": [-10.644250869750977, -10.107965469360352, -10.727447509765625, -9.958454132080078], "reading_logprob_std": 0.3832213601242922, "tokenization_path_logprob_std": 71.00598006629897, "trajectory_balance_rmse": 341.6594266892135} {"canonical_length": 179, "canonical_nll": 1.2977845602195357, "generation_rouge_l_f1": 0.23608017817371935, "generation_token_f1": 0.4454342984409799, "prompt_id": "94a9111b40eefe1b", "reading_canonical_log_prob": -232.30343627929688, "reading_distances": [37, 36, 37, 38], "reading_log_probs": [-234.16317749023438, -233.5531005859375, -232.7850799560547, -231.68319702148438], "reading_logprob_std": 1.0693530286278803, "tokenization_path_logprob_std": 128.2077587537822, "trajectory_balance_rmse": 3806.1672147246722} {"canonical_length": 137, "canonical_nll": 1.6353788723910812, "generation_rouge_l_f1": 0.2595419847328244, "generation_token_f1": 0.47328244274809156, "prompt_id": "726f150c14796002", "reading_canonical_log_prob": -224.04690551757812, "reading_distances": [17, 16, 13, 22], "reading_log_probs": [-227.45108032226562, -227.65951538085938, -221.12680053710938, -227.483154296875], "reading_logprob_std": 3.2035356547547136, "tokenization_path_logprob_std": 191.5890322172029, "trajectory_balance_rmse": 3410.338644028269} {"canonical_length": 57, "canonical_nll": 0.362862369470429, "generation_rouge_l_f1": 0.31428571428571433, "generation_token_f1": 0.4476190476190476, "prompt_id": "39559806a2fa846b", "reading_canonical_log_prob": -20.683155059814453, "reading_distances": [27, 21, 21, 22], "reading_log_probs": [-23.499156951904297, -23.895286560058594, -27.30146026611328, -25.51251220703125], "reading_logprob_std": 1.7341136499228802, "tokenization_path_logprob_std": 19.996604784316983, "trajectory_balance_rmse": 627.0783575684699} {"canonical_length": 13, "canonical_nll": 0.803038450387808, "generation_rouge_l_f1": 0.19999999999999998, "generation_token_f1": 0.25, "prompt_id": "3dc6a3c96d3d9987", "reading_canonical_log_prob": -10.439499855041504, "reading_distances": [24, 34, 30, 30], "reading_log_probs": [-13.439164161682129, -16.948448181152344, -12.028695106506348, -13.826559066772461], "reading_logprob_std": 2.074388441505181, "tokenization_path_logprob_std": 30.345469578568586, "trajectory_balance_rmse": 224.7504515479306} {"canonical_length": 14, "canonical_nll": 0.7169483729771206, "generation_rouge_l_f1": 0.64, "generation_token_f1": 0.64, "prompt_id": "8e74067a85468916", "reading_canonical_log_prob": -10.037277221679688, "reading_distances": [17, 17, 17, 23], "reading_log_probs": [-9.351521492004395, -9.847681045532227, -9.944787979125977, -9.707306861877441], "reading_logprob_std": 0.2598484503437471, "tokenization_path_logprob_std": 44.66805080478961, "trajectory_balance_rmse": 337.6250053815013} {"canonical_length": 27, "canonical_nll": 1.352132585313585, "generation_rouge_l_f1": 0.14492753623188406, "generation_token_f1": 0.2173913043478261, "prompt_id": "3090724d0b15fbad", "reading_canonical_log_prob": -36.5075798034668, "reading_distances": [25, 17, 19, 26], "reading_log_probs": [-39.337371826171875, -36.207855224609375, -37.36168670654297, -35.00676727294922], "reading_logprob_std": 1.8432518037156724, "tokenization_path_logprob_std": 48.33284302150793, "trajectory_balance_rmse": 531.5622442865697} {"canonical_length": 273, "canonical_nll": 1.5966072501717032, "generation_rouge_l_f1": 0.22846441947565543, "generation_token_f1": 0.50561797752809, "prompt_id": "d410f93c21deda57", "reading_canonical_log_prob": -435.873779296875, "reading_distances": [22, 17, 18, 13], "reading_log_probs": [-434.5284118652344, -434.02227783203125, -433.055908203125, -432.764892578125], "reading_logprob_std": 0.8233072027351116, "tokenization_path_logprob_std": 351.0341752293029, "trajectory_balance_rmse": 6308.8005455059865} {"canonical_length": 354, "canonical_nll": 1.1743598550052967, "generation_rouge_l_f1": 0.23011844331641287, "generation_token_f1": 0.4196277495769882, "prompt_id": "710012fa888db2a9", "reading_canonical_log_prob": -415.723388671875, "reading_distances": [33, 26, 36, 26], "reading_log_probs": [-416.90765380859375, -411.4141845703125, -412.1644287109375, -412.4224853515625], "reading_logprob_std": 2.4906367099717825, "tokenization_path_logprob_std": 201.8724019899594, "trajectory_balance_rmse": 6662.60336523329} {"canonical_length": 191, "canonical_nll": 1.438396353996237, "generation_rouge_l_f1": 0.31304347826086953, "generation_token_f1": 0.5347826086956522, "prompt_id": "b61ed23f42bef95d", "reading_canonical_log_prob": -274.73370361328125, "reading_distances": [20, 19, 23, 25], "reading_log_probs": [-271.200439453125, -269.5, -272.65032958984375, -270.38623046875], "reading_logprob_std": 1.3383054525237863, "tokenization_path_logprob_std": 174.06817325013543, "trajectory_balance_rmse": 4307.627983105979} {"canonical_length": 255, "canonical_nll": 1.0399167528339461, "generation_rouge_l_f1": 0.2918287937743191, "generation_token_f1": 0.5252918287937743, "prompt_id": "0cd98f075a94d815", "reading_canonical_log_prob": -265.17877197265625, "reading_distances": [17, 29, 27, 20], "reading_log_probs": [-263.966796875, -259.01983642578125, -261.77044677734375, -262.66424560546875], "reading_logprob_std": 2.094426396566961, "tokenization_path_logprob_std": 308.84553126575076, "trajectory_balance_rmse": 5072.441603281524} {"canonical_length": 315, "canonical_nll": 1.3211071196056547, "generation_rouge_l_f1": 0.22260869565217392, "generation_token_f1": 0.4173913043478261, "prompt_id": "3f6acc98eba7e06b", "reading_canonical_log_prob": -416.14874267578125, "reading_distances": [22, 23, 25, 22], "reading_log_probs": [-411.37591552734375, -412.76708984375, -412.7370300292969, -414.0481872558594], "reading_logprob_std": 1.0912638812253717, "tokenization_path_logprob_std": 118.77510743490606, "trajectory_balance_rmse": 7351.457066487538} {"canonical_length": 44, "canonical_nll": 0.9040057442405007, "generation_rouge_l_f1": 0.17869415807560135, "generation_token_f1": 0.1924398625429553, "prompt_id": "43a708261a941980", "reading_canonical_log_prob": -39.77625274658203, "reading_distances": [15, 12, 13, 12], "reading_log_probs": [-38.18357849121094, -38.662940979003906, -39.02091979980469, -38.7614631652832], "reading_logprob_std": 0.35000628099561204, "tokenization_path_logprob_std": 73.7069126274902, "trajectory_balance_rmse": 940.0016731838597} {"canonical_length": 186, "canonical_nll": 1.573723905829973, "generation_rouge_l_f1": 0.25688073394495414, "generation_token_f1": 0.47706422018348627, "prompt_id": "6a69846ab5b6aa8a", "reading_canonical_log_prob": -292.712646484375, "reading_distances": [19, 25, 20, 24], "reading_log_probs": [-281.3458251953125, -285.7943420410156, -286.8399658203125, -283.61553955078125], "reading_logprob_std": 2.4386458116498217, "tokenization_path_logprob_std": 211.76108060373332, "trajectory_balance_rmse": 3965.4265887915676} {"canonical_length": 99, "canonical_nll": 1.1728897865372474, "generation_rouge_l_f1": 0.1657142857142857, "generation_token_f1": 0.2971428571428571, "prompt_id": "692adc91823ed2ff", "reading_canonical_log_prob": -116.1160888671875, "reading_distances": [29, 20, 22, 31], "reading_log_probs": [-110.99884033203125, -111.5439453125, -110.24662017822266, -113.10429382324219], "reading_logprob_std": 1.2103700413379885, "tokenization_path_logprob_std": 77.8120198878129, "trajectory_balance_rmse": 2209.3464141049817} {"canonical_length": 11, "canonical_nll": 1.902406172318892, "generation_rouge_l_f1": 0.14285714285714288, "generation_token_f1": 0.14285714285714288, "prompt_id": "9477d8c348a14e9c", "reading_canonical_log_prob": -20.926467895507812, "reading_distances": [24, 27, 24, 25], "reading_log_probs": [-22.106246948242188, -21.792232513427734, -21.628765106201172, -21.940637588500977], "reading_logprob_std": 0.20413037899288106, "tokenization_path_logprob_std": 55.27884999465011, "trajectory_balance_rmse": 233.9141254442547} {"canonical_length": 53, "canonical_nll": 0.7081817051149765, "generation_rouge_l_f1": 0.5656565656565656, "generation_token_f1": 0.6464646464646464, "prompt_id": "7e54ccf509187c23", "reading_canonical_log_prob": -37.53363037109375, "reading_distances": [14, 14, 14, 14], "reading_log_probs": [-48.16831588745117, -45.72525405883789, -39.46327209472656, -47.92481231689453], "reading_logprob_std": 4.056412311609869, "tokenization_path_logprob_std": 86.66946008881239, "trajectory_balance_rmse": 828.9896949741969} {"canonical_length": 65, "canonical_nll": 1.2521559495192307, "generation_rouge_l_f1": 0.15946843853820597, "generation_token_f1": 0.19269102990033224, "prompt_id": "ba0c8c0d4edf2258", "reading_canonical_log_prob": -81.39013671875, "reading_distances": [25, 25, 22, 20], "reading_log_probs": [-79.53828430175781, -82.13751220703125, -79.50627899169922, -80.32471466064453], "reading_logprob_std": 1.233387708530756, "tokenization_path_logprob_std": 134.39757664421006, "trajectory_balance_rmse": 1549.9692477736012} {"canonical_length": 181, "canonical_nll": 1.5971855037120166, "generation_rouge_l_f1": 0.2657004830917874, "generation_token_f1": 0.4251207729468599, "prompt_id": "1f441ed2a6f2f793", "reading_canonical_log_prob": -289.090576171875, "reading_distances": [21, 19, 24, 17], "reading_log_probs": [-284.23223876953125, -286.6303405761719, -285.90576171875, -284.8489990234375], "reading_logprob_std": 1.0703155371858961, "tokenization_path_logprob_std": 223.78992240674188, "trajectory_balance_rmse": 3283.579061041397} {"canonical_length": 234, "canonical_nll": 1.4327150002504005, "generation_rouge_l_f1": 0.2489451476793249, "generation_token_f1": 0.4978902953586498, "prompt_id": "a1958dd73d8cbad2", "reading_canonical_log_prob": -335.25531005859375, "reading_distances": [16, 12, 15, 15], "reading_log_probs": [-334.5617980957031, -332.19287109375, -342.39630126953125, -334.33099365234375], "reading_logprob_std": 4.479353266189343, "tokenization_path_logprob_std": 176.7030705714341, "trajectory_balance_rmse": 5018.592157887217} {"canonical_length": 248, "canonical_nll": 0.846390139672064, "generation_rouge_l_f1": 0.31092436974789917, "generation_token_f1": 0.5756302521008403, "prompt_id": "3e4f398cc2a68a1a", "reading_canonical_log_prob": -209.90475463867188, "reading_distances": [14, 16, 12, 12], "reading_log_probs": [-213.8927001953125, -213.07284545898438, -211.87326049804688, -214.1427001953125], "reading_logprob_std": 1.0225274346135405, "tokenization_path_logprob_std": 234.86410485873083, "trajectory_balance_rmse": 4358.436568987229} {"canonical_length": 7, "canonical_nll": 0.44296155657087055, "generation_rouge_l_f1": 0.7692307692307692, "generation_token_f1": 0.7692307692307692, "prompt_id": "96fa3fcdd459d2ed", "reading_canonical_log_prob": -3.1007308959960938, "reading_distances": [14, 15, 19, 19], "reading_log_probs": [-5.172305107116699, -5.478574752807617, -4.7561774253845215, -4.468260288238525], "reading_logprob_std": 0.44610648431853744, "tokenization_path_logprob_std": 11.777442146724859, "trajectory_balance_rmse": 103.36301232759605} {"canonical_length": 17, "canonical_nll": 0.5129387799431296, "generation_rouge_l_f1": 0.45454545454545453, "generation_token_f1": 0.45454545454545453, "prompt_id": "27d776f87907a327", "reading_canonical_log_prob": -8.719959259033203, "reading_distances": [23, 20, 24, 22], "reading_log_probs": [-9.151068687438965, -6.896849632263184, -8.021081924438477, -8.531038284301758], "reading_logprob_std": 0.9546961829239523, "tokenization_path_logprob_std": 48.806454262467405, "trajectory_balance_rmse": 315.25999193768087} {"canonical_length": 37, "canonical_nll": 0.9051849777634079, "generation_rouge_l_f1": 0.13043478260869565, "generation_token_f1": 0.18840579710144925, "prompt_id": "38bff471c2090786", "reading_canonical_log_prob": -33.491844177246094, "reading_distances": [40, 34, 37, 46], "reading_log_probs": [-34.578369140625, -36.87699890136719, -29.83116912841797, -33.27106857299805], "reading_logprob_std": 2.944033520221283, "tokenization_path_logprob_std": 31.465400200723778, "trajectory_balance_rmse": 489.10094919985846} {"canonical_length": 60, "canonical_nll": 1.1822522481282551, "generation_rouge_l_f1": 0.24468085106382978, "generation_token_f1": 0.3191489361702128, "prompt_id": "68df2af9448f2f1d", "reading_canonical_log_prob": -70.93513488769531, "reading_distances": [31, 30, 28, 24], "reading_log_probs": [-68.67253112792969, -68.24635314941406, -69.61715698242188, -70.35679626464844], "reading_logprob_std": 0.9482806495685888, "tokenization_path_logprob_std": 105.27933172940232, "trajectory_balance_rmse": 1556.3054723710377} {"canonical_length": 14, "canonical_nll": 0.6577751977103097, "generation_rouge_l_f1": 0.14285714285714288, "generation_token_f1": 0.14285714285714288, "prompt_id": "6da3ca8bcb458aa8", "reading_canonical_log_prob": -9.208852767944336, "reading_distances": [15, 21, 15, 19], "reading_log_probs": [-7.135685920715332, -7.75180721282959, -7.812737941741943, -8.900981903076172], "reading_logprob_std": 0.733874690309457, "tokenization_path_logprob_std": 43.129025970266774, "trajectory_balance_rmse": 279.3132762247678} {"canonical_length": 29, "canonical_nll": 0.9718610171614022, "generation_rouge_l_f1": 0.136, "generation_token_f1": 0.168, "prompt_id": "eba9c966fdb6009a", "reading_canonical_log_prob": -28.183969497680664, "reading_distances": [35, 30, 32, 34], "reading_log_probs": [-25.293556213378906, -24.241212844848633, -25.21417808532715, -26.924543380737305], "reading_logprob_std": 1.1122844420510538, "tokenization_path_logprob_std": 103.39209998835175, "trajectory_balance_rmse": 551.5176653153214} {"canonical_length": 237, "canonical_nll": 1.303765276816324, "generation_rouge_l_f1": 0.27956989247311825, "generation_token_f1": 0.5376344086021505, "prompt_id": "16c0d54d78758bbb", "reading_canonical_log_prob": -308.99237060546875, "reading_distances": [20, 34, 21, 33], "reading_log_probs": [-311.7360534667969, -313.07293701171875, -312.35699462890625, -313.4622802734375], "reading_logprob_std": 0.765860221929629, "tokenization_path_logprob_std": 134.09422299754965, "trajectory_balance_rmse": 4856.8782743229485} {"canonical_length": 40, "canonical_nll": 1.3060476303100585, "generation_rouge_l_f1": 0.3684210526315789, "generation_token_f1": 0.3684210526315789, "prompt_id": "51ab3680c9c02305", "reading_canonical_log_prob": -52.241905212402344, "reading_distances": [23, 18, 18, 19], "reading_log_probs": [-52.33127212524414, -53.7325439453125, -54.404693603515625, -53.26786422729492], "reading_logprob_std": 0.8708203766131933, "tokenization_path_logprob_std": 65.21700890906374, "trajectory_balance_rmse": 664.2806919617099} {"canonical_length": 133, "canonical_nll": 1.0892546230689026, "generation_rouge_l_f1": 0.2619647355163728, "generation_token_f1": 0.37783375314861456, "prompt_id": "a142f5697eaf1704", "reading_canonical_log_prob": -144.87086486816406, "reading_distances": [23, 19, 18, 26], "reading_log_probs": [-147.86819458007812, -146.25546264648438, -146.37918090820312, -152.10610961914062], "reading_logprob_std": 2.7358892300626034, "tokenization_path_logprob_std": 214.27332690183442, "trajectory_balance_rmse": 3116.052118037703} {"canonical_length": 248, "canonical_nll": 1.1314227196478075, "generation_rouge_l_f1": 0.21933085501858737, "generation_token_f1": 0.40148698884758366, "prompt_id": "879233d0df49ade7", "reading_canonical_log_prob": -280.59283447265625, "reading_distances": [21, 27, 28, 22], "reading_log_probs": [-277.9881591796875, -279.18212890625, -277.97039794921875, -278.01800537109375], "reading_logprob_std": 0.5952948812946328, "tokenization_path_logprob_std": 119.86389015792729, "trajectory_balance_rmse": 4470.718189224492} {"canonical_length": 189, "canonical_nll": 1.2009977310422868, "generation_rouge_l_f1": 0.23963133640552994, "generation_token_f1": 0.4746543778801843, "prompt_id": "f6b26f3fe5bcfb2a", "reading_canonical_log_prob": -226.9885711669922, "reading_distances": [8, 13, 18, 16], "reading_log_probs": [-219.47030639648438, -217.1090087890625, -216.89987182617188, -218.29483032226562], "reading_logprob_std": 1.1888518430196304, "tokenization_path_logprob_std": 85.64510270450393, "trajectory_balance_rmse": 2835.4745480512856} {"canonical_length": 12, "canonical_nll": 1.2767664591471355, "generation_rouge_l_f1": 0.33333333333333337, "generation_token_f1": 0.33333333333333337, "prompt_id": "61b8f081f2257b08", "reading_canonical_log_prob": -15.321197509765625, "reading_distances": [43, 37, 38, 42], "reading_log_probs": [-16.376361846923828, -20.579463958740234, -17.473854064941406, -20.211387634277344], "reading_logprob_std": 2.0585667011491995, "tokenization_path_logprob_std": 31.58805646346867, "trajectory_balance_rmse": 254.76020124050814} {"canonical_length": 7, "canonical_nll": 0.010885515383311681, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "756f69ca5eb428f4", "reading_canonical_log_prob": -0.07619860768318176, "reading_distances": [39, 46, 37, 36], "reading_log_probs": [-0.11700284481048584, -0.14483779668807983, -0.14999957382678986, -0.347842276096344], "reading_logprob_std": 0.10627386365445618, "tokenization_path_logprob_std": 19.35054191955814, "trajectory_balance_rmse": 184.79880146301812} {"canonical_length": 59, "canonical_nll": 1.0938288801807468, "generation_rouge_l_f1": 0.32075471698113206, "generation_token_f1": 0.43396226415094336, "prompt_id": "f3fd9530a70f7802", "reading_canonical_log_prob": -64.53590393066406, "reading_distances": [25, 30, 31, 37], "reading_log_probs": [-69.38284301757812, -69.60912322998047, -69.2635269165039, -72.56234741210938], "reading_logprob_std": 1.57844527424325, "tokenization_path_logprob_std": 75.09443582281011, "trajectory_balance_rmse": 1574.9331914384707} {"canonical_length": 84, "canonical_nll": 0.7846827734084356, "generation_rouge_l_f1": 0.35748792270531404, "generation_token_f1": 0.42512077294685985, "prompt_id": "d431ad8fbed7b76c", "reading_canonical_log_prob": -65.9133529663086, "reading_distances": [27, 28, 21, 22], "reading_log_probs": [-64.32395935058594, -65.6812973022461, -64.8928451538086, -66.20796203613281], "reading_logprob_std": 0.8338679170859417, "tokenization_path_logprob_std": 68.36248496162041, "trajectory_balance_rmse": 1291.2112273382597} {"canonical_length": 376, "canonical_nll": 1.4554565105032413, "generation_rouge_l_f1": 0.19682539682539685, "generation_token_f1": 0.3492063492063492, "prompt_id": "455b2b67c74aee94", "reading_canonical_log_prob": -547.2516479492188, "reading_distances": [27, 19, 18, 24], "reading_log_probs": [-534.5470581054688, -537.348388671875, -535.1874389648438, -534.7559814453125], "reading_logprob_std": 1.2870387201207036, "tokenization_path_logprob_std": 129.49234902212157, "trajectory_balance_rmse": 7155.365641802961} {"canonical_length": 307, "canonical_nll": 1.732648520205619, "generation_rouge_l_f1": 0.208695652173913, "generation_token_f1": 0.39652173913043476, "prompt_id": "b3e34a08eb2e7c34", "reading_canonical_log_prob": -531.923095703125, "reading_distances": [25, 26, 28, 27], "reading_log_probs": [-529.6342163085938, -533.8076782226562, -531.1978149414062, -531.2281494140625], "reading_logprob_std": 1.7289091738913016, "tokenization_path_logprob_std": 146.3851673301038, "trajectory_balance_rmse": 6299.625555387997} {"canonical_length": 165, "canonical_nll": 1.358664032907197, "generation_rouge_l_f1": 0.2388758782201405, "generation_token_f1": 0.4637002341920375, "prompt_id": "7461cdb891fcaab7", "reading_canonical_log_prob": -224.1795654296875, "reading_distances": [18, 18, 17, 23], "reading_log_probs": [-222.7976837158203, -221.2247314453125, -222.29940795898438, -223.19387817382812], "reading_logprob_std": 0.8520550176097081, "tokenization_path_logprob_std": 171.96057418471, "trajectory_balance_rmse": 3458.3176388083775} {"canonical_length": 119, "canonical_nll": 1.231740871397387, "generation_rouge_l_f1": 0.16751269035532995, "generation_token_f1": 0.2639593908629442, "prompt_id": "eb07969998e95372", "reading_canonical_log_prob": -146.57716369628906, "reading_distances": [18, 19, 18, 17], "reading_log_probs": [-136.1987762451172, -157.9168701171875, -151.2486572265625, -147.08676147460938], "reading_logprob_std": 9.109520815993541, "tokenization_path_logprob_std": 76.59644494230571, "trajectory_balance_rmse": 2004.0195001597176} {"canonical_length": 134, "canonical_nll": 1.482203469347598, "generation_rouge_l_f1": 0.20670391061452514, "generation_token_f1": 0.4134078212290503, "prompt_id": "b32326c583d0044a", "reading_canonical_log_prob": -198.61526489257812, "reading_distances": [23, 33, 27, 21], "reading_log_probs": [-190.68118286132812, -191.90557861328125, -190.2545928955078, -192.00091552734375], "reading_logprob_std": 0.8759425384514563, "tokenization_path_logprob_std": 101.14916193757107, "trajectory_balance_rmse": 2524.958696165635} {"canonical_length": 324, "canonical_nll": 1.099916057822145, "generation_rouge_l_f1": 0.3157894736842105, "generation_token_f1": 0.555921052631579, "prompt_id": "e86ca4a6dca6aa44", "reading_canonical_log_prob": -356.372802734375, "reading_distances": [38, 31, 34, 32], "reading_log_probs": [-350.09429931640625, -350.0320739746094, -348.49017333984375, -347.7921142578125], "reading_logprob_std": 1.1459825394330958, "tokenization_path_logprob_std": 351.20725057309585, "trajectory_balance_rmse": 6047.746176026132} {"canonical_length": 17, "canonical_nll": 0.02779527096187367, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "39176769401db5cc", "reading_canonical_log_prob": -0.4725196063518524, "reading_distances": [14, 22, 19, 18], "reading_log_probs": [-0.6045798063278198, -0.6401300430297852, -0.5922682881355286, -0.6221718788146973], "reading_logprob_std": 0.020881295065180654, "tokenization_path_logprob_std": 74.49291768283095, "trajectory_balance_rmse": 314.73331160596126} {"canonical_length": 46, "canonical_nll": 1.4643358976944634, "generation_rouge_l_f1": 0.12944983818770228, "generation_token_f1": 0.22006472491909385, "prompt_id": "6614284ebae06e44", "reading_canonical_log_prob": -67.35945129394531, "reading_distances": [15, 22, 21, 18], "reading_log_probs": [-65.78034973144531, -64.84877014160156, -66.68663024902344, -64.65240478515625], "reading_logprob_std": 0.9361181850299172, "tokenization_path_logprob_std": 154.9147688027103, "trajectory_balance_rmse": 918.8591676629808} {"canonical_length": 169, "canonical_nll": 2.032634306236132, "generation_rouge_l_f1": 0.2268041237113402, "generation_token_f1": 0.4278350515463918, "prompt_id": "130451267d979653", "reading_canonical_log_prob": -343.51519775390625, "reading_distances": [29, 24, 16, 21], "reading_log_probs": [-331.8174133300781, -334.6394958496094, -336.423095703125, -336.58306884765625], "reading_logprob_std": 2.214953502593799, "tokenization_path_logprob_std": 77.09543869901216, "trajectory_balance_rmse": 3420.17091714643} {"canonical_length": 89, "canonical_nll": 1.803138132845418, "generation_rouge_l_f1": 0.16111111111111112, "generation_token_f1": 0.31666666666666665, "prompt_id": "13cc5e7d950f14d1", "reading_canonical_log_prob": -160.4792938232422, "reading_distances": [24, 22, 20, 26], "reading_log_probs": [-151.62442016601562, -153.89151000976562, -150.70819091796875, -153.86582946777344], "reading_logprob_std": 1.6100711473478826, "tokenization_path_logprob_std": 93.39590269645237, "trajectory_balance_rmse": 1863.3158890382688} {"canonical_length": 50, "canonical_nll": 1.0993035125732422, "generation_rouge_l_f1": 0.37333333333333335, "generation_token_f1": 0.45333333333333337, "prompt_id": "aa5825f0f273154e", "reading_canonical_log_prob": -54.96517562866211, "reading_distances": [35, 32, 29, 29], "reading_log_probs": [-53.70254135131836, -53.61937713623047, -53.13684844970703, -53.74571990966797], "reading_logprob_std": 0.2811153407972512, "tokenization_path_logprob_std": 30.19347777233307, "trajectory_balance_rmse": 1296.3373842030926} {"canonical_length": 217, "canonical_nll": 1.4553636120211693, "generation_rouge_l_f1": 0.24034334763948495, "generation_token_f1": 0.5278969957081545, "prompt_id": "046e2cee68c73a19", "reading_canonical_log_prob": -315.81390380859375, "reading_distances": [23, 31, 24, 29], "reading_log_probs": [-314.43829345703125, -315.2369079589844, -314.9642333984375, -315.73919677734375], "reading_logprob_std": 0.5426755557083514, "tokenization_path_logprob_std": 151.41100389059565, "trajectory_balance_rmse": 4069.0076513144086} {"canonical_length": 114, "canonical_nll": 1.898213972125137, "generation_rouge_l_f1": 0.15774647887323942, "generation_token_f1": 0.30985915492957744, "prompt_id": "f13ae64aa84fa6f5", "reading_canonical_log_prob": -216.39639282226562, "reading_distances": [25, 29, 25, 26], "reading_log_probs": [-215.08880615234375, -214.66751098632812, -212.8987274169922, -212.63333129882812], "reading_logprob_std": 1.2362648265346758, "tokenization_path_logprob_std": 114.12483714273743, "trajectory_balance_rmse": 2198.872549555711} {"canonical_length": 180, "canonical_nll": 1.2806866115993925, "generation_rouge_l_f1": 0.2618510158013544, "generation_token_f1": 0.47855530474040625, "prompt_id": "55942a6d7a51de21", "reading_canonical_log_prob": -230.52359008789062, "reading_distances": [17, 19, 19, 15], "reading_log_probs": [-225.88763427734375, -225.212890625, -226.65367126464844, -227.8038330078125], "reading_logprob_std": 1.111515952615509, "tokenization_path_logprob_std": 231.18232553711715, "trajectory_balance_rmse": 4054.205722269461} {"canonical_length": 158, "canonical_nll": 1.3298633430577531, "generation_rouge_l_f1": 0.3553299492385787, "generation_token_f1": 0.5177664974619289, "prompt_id": "ab7cc228a99beb55", "reading_canonical_log_prob": -210.118408203125, "reading_distances": [34, 32, 33, 32], "reading_log_probs": [-195.12747192382812, -197.81307983398438, -194.85513305664062, -197.58326721191406], "reading_logprob_std": 1.5695691487021342, "tokenization_path_logprob_std": 97.03622454125102, "trajectory_balance_rmse": 3031.61688981821} {"canonical_length": 11, "canonical_nll": 0.4116860303011807, "generation_rouge_l_f1": 0.339622641509434, "generation_token_f1": 0.37735849056603776, "prompt_id": "c2611a4dd26cfbb1", "reading_canonical_log_prob": -4.528546333312988, "reading_distances": [14, 11, 14, 16], "reading_log_probs": [-4.731342315673828, -4.378006935119629, -4.2671685218811035, -4.82264518737793], "reading_logprob_std": 0.26882182319471243, "tokenization_path_logprob_std": 35.36034816617887, "trajectory_balance_rmse": 180.1053384088694} {"canonical_length": 138, "canonical_nll": 1.4728024247763813, "generation_rouge_l_f1": 0.3481012658227848, "generation_token_f1": 0.5822784810126582, "prompt_id": "6f0ecfb4c606f967", "reading_canonical_log_prob": -203.24673461914062, "reading_distances": [28, 22, 23, 24], "reading_log_probs": [-203.62355041503906, -200.06832885742188, -198.03579711914062, -199.93093872070312], "reading_logprob_std": 2.331658768478595, "tokenization_path_logprob_std": 108.39112589738575, "trajectory_balance_rmse": 2912.484340324242} {"canonical_length": 202, "canonical_nll": 0.7012596508063892, "generation_rouge_l_f1": 0.5317460317460319, "generation_token_f1": 0.6746031746031748, "prompt_id": "5d860ca9ad882bc0", "reading_canonical_log_prob": -141.65444946289062, "reading_distances": [13, 10, 13, 6], "reading_log_probs": [-143.15867614746094, -142.09921264648438, -142.03477478027344, -144.77716064453125], "reading_logprob_std": 1.2813192162653608, "tokenization_path_logprob_std": 238.69160904862846, "trajectory_balance_rmse": 3655.922260087303} {"canonical_length": 253, "canonical_nll": 1.5876106593919836, "generation_rouge_l_f1": 0.2369477911646586, "generation_token_f1": 0.4337349397590362, "prompt_id": "bef65ba5cd28353b", "reading_canonical_log_prob": -401.6654968261719, "reading_distances": [38, 36, 31, 36], "reading_log_probs": [-389.9367370605469, -387.7438659667969, -392.83758544921875, -387.8377380371094], "reading_logprob_std": 2.390653086916869, "tokenization_path_logprob_std": 253.63817036817917, "trajectory_balance_rmse": 4869.043160251482} {"canonical_length": 49, "canonical_nll": 0.7969296513771524, "generation_rouge_l_f1": 0.38461538461538464, "generation_token_f1": 0.4615384615384615, "prompt_id": "ba40c7a321bb21e7", "reading_canonical_log_prob": -39.04955291748047, "reading_distances": [12, 16, 15, 19], "reading_log_probs": [-39.314979553222656, -40.805076599121094, -40.7919807434082, -40.119850158691406], "reading_logprob_std": 0.7054079619012994, "tokenization_path_logprob_std": 105.84781681767456, "trajectory_balance_rmse": 966.0209802961199} {"canonical_length": 15, "canonical_nll": 0.692410659790039, "generation_rouge_l_f1": 0.30136986301369867, "generation_token_f1": 0.3835616438356165, "prompt_id": "a165c584ac95ad42", "reading_canonical_log_prob": -10.386159896850586, "reading_distances": [26, 34, 24, 22], "reading_log_probs": [-11.819499969482422, -13.956489562988281, -13.879636764526367, -10.35946273803711], "reading_logprob_std": 1.7387430892043851, "tokenization_path_logprob_std": 28.229452187266805, "trajectory_balance_rmse": 334.8304019215447} {"canonical_length": 242, "canonical_nll": 1.529809116331999, "generation_rouge_l_f1": 0.19959266802443995, "generation_token_f1": 0.36252545824847254, "prompt_id": "9c4dbe029e6ba3bb", "reading_canonical_log_prob": -370.21380615234375, "reading_distances": [16, 16, 19, 19], "reading_log_probs": [-367.6768798828125, -367.8561096191406, -367.6057434082031, -362.718017578125], "reading_logprob_std": 2.4996672429344624, "tokenization_path_logprob_std": 235.72156771972894, "trajectory_balance_rmse": 4862.727069486222} {"canonical_length": 62, "canonical_nll": 0.6098392240462764, "generation_rouge_l_f1": 0.5378151260504201, "generation_token_f1": 0.6554621848739496, "prompt_id": "dcfe7ed6b29a3c6d", "reading_canonical_log_prob": -37.81003189086914, "reading_distances": [24, 17, 20, 19], "reading_log_probs": [-40.76036071777344, -40.63807678222656, -39.53844451904297, -40.439697265625], "reading_logprob_std": 0.5531470319317988, "tokenization_path_logprob_std": 72.83251498092315, "trajectory_balance_rmse": 1081.72065447287} {"canonical_length": 148, "canonical_nll": 1.5669211310309332, "generation_rouge_l_f1": 0.3095975232198142, "generation_token_f1": 0.5696594427244581, "prompt_id": "c0cc49ea5faf89df", "reading_canonical_log_prob": -231.90432739257812, "reading_distances": [28, 23, 21, 26], "reading_log_probs": [-232.37533569335938, -235.790771484375, -233.05433654785156, -232.27011108398438], "reading_logprob_std": 1.649127330242983, "tokenization_path_logprob_std": 114.24798884913247, "trajectory_balance_rmse": 3462.1873861524887} {"canonical_length": 14, "canonical_nll": 1.6867656707763672, "generation_rouge_l_f1": 0.2380952380952381, "generation_token_f1": 0.2380952380952381, "prompt_id": "adf9198c8ba81554", "reading_canonical_log_prob": -23.61471939086914, "reading_distances": [18, 15, 10, 13], "reading_log_probs": [-23.554401397705078, -23.16842269897461, -23.433345794677734, -22.743907928466797], "reading_logprob_std": 0.3589632394082532, "tokenization_path_logprob_std": 81.9242673301844, "trajectory_balance_rmse": 363.6425156514768} {"canonical_length": 280, "canonical_nll": 1.0845982142857142, "generation_rouge_l_f1": 0.2426470588235294, "generation_token_f1": 0.474264705882353, "prompt_id": "7eceb236d5391917", "reading_canonical_log_prob": -303.6875, "reading_distances": [21, 27, 22, 29], "reading_log_probs": [-301.209716796875, -301.1418762207031, -300.36761474609375, -302.1388854980469], "reading_logprob_std": 0.7250287976947954, "tokenization_path_logprob_std": 168.44762557923414, "trajectory_balance_rmse": 5460.411126845008} {"canonical_length": 7, "canonical_nll": 1.6618393489292689, "generation_rouge_l_f1": 0.25, "generation_token_f1": 0.25, "prompt_id": "8279f5e0aa9b5caf", "reading_canonical_log_prob": -11.632875442504883, "reading_distances": [22, 28, 24, 21], "reading_log_probs": [-11.298225402832031, -11.305927276611328, -11.556180000305176, -11.185710906982422], "reading_logprob_std": 0.15641428838820493, "tokenization_path_logprob_std": 8.765483955483251, "trajectory_balance_rmse": 186.02190948645466} {"canonical_length": 323, "canonical_nll": 0.7662417763157895, "generation_rouge_l_f1": 0.1684981684981685, "generation_token_f1": 0.39194139194139194, "prompt_id": "f4edb7d5c1554a64", "reading_canonical_log_prob": -247.49609375, "reading_distances": [21, 21, 22, 13], "reading_log_probs": [-249.2002716064453, -247.090087890625, -248.18592834472656, -249.27943420410156], "reading_logprob_std": 1.0278575180783218, "tokenization_path_logprob_std": 118.36025031568714, "trajectory_balance_rmse": 6536.295553118669} {"canonical_length": 18, "canonical_nll": 0.7270348866780599, "generation_rouge_l_f1": 0.11320754716981132, "generation_token_f1": 0.22641509433962265, "prompt_id": "39c97bd4effbfbe9", "reading_canonical_log_prob": -13.086627960205078, "reading_distances": [25, 27, 28, 26], "reading_log_probs": [-11.076136589050293, -11.351722717285156, -11.6386079788208, -12.050576210021973], "reading_logprob_std": 0.41656047324299394, "tokenization_path_logprob_std": 76.51551129237029, "trajectory_balance_rmse": 368.01945408497886} {"canonical_length": 32, "canonical_nll": 0.541018009185791, "generation_rouge_l_f1": 0.43243243243243246, "generation_token_f1": 0.5945945945945945, "prompt_id": "e245da9a65e5ed0d", "reading_canonical_log_prob": -17.312576293945312, "reading_distances": [22, 20, 24, 17], "reading_log_probs": [-18.42965316772461, -20.819475173950195, -17.84996795654297, -21.215030670166016], "reading_logprob_std": 1.685815735483512, "tokenization_path_logprob_std": 56.85369410387588, "trajectory_balance_rmse": 360.6994098207999} {"canonical_length": 244, "canonical_nll": 1.2357635498046875, "generation_rouge_l_f1": 0.2589641434262948, "generation_token_f1": 0.48207171314741043, "prompt_id": "1dc2c6d9af2ea2a5", "reading_canonical_log_prob": -301.52630615234375, "reading_distances": [15, 15, 18, 18], "reading_log_probs": [-294.60302734375, -295.32086181640625, -294.49554443359375, -295.9666442871094], "reading_logprob_std": 0.6860884682495216, "tokenization_path_logprob_std": 115.7667860573676, "trajectory_balance_rmse": 4472.152776013101} {"canonical_length": 143, "canonical_nll": 1.2754118592589052, "generation_rouge_l_f1": 0.28217821782178215, "generation_token_f1": 0.43069306930693063, "prompt_id": "8214cf7581d9c6a8", "reading_canonical_log_prob": -182.38389587402344, "reading_distances": [13, 10, 10, 7], "reading_log_probs": [-181.11532592773438, -180.6860809326172, -181.3728790283203, -179.72283935546875], "reading_logprob_std": 0.7252448208241817, "tokenization_path_logprob_std": 79.52474618245223, "trajectory_balance_rmse": 3138.5427257352335} {"canonical_length": 79, "canonical_nll": 0.6230166471457179, "generation_rouge_l_f1": 0.44999999999999996, "generation_token_f1": 0.48999999999999994, "prompt_id": "1669f97fd38169a6", "reading_canonical_log_prob": -49.21831512451172, "reading_distances": [15, 27, 30, 22], "reading_log_probs": [-50.57611846923828, -49.234344482421875, -51.058929443359375, -49.227352142333984], "reading_logprob_std": 0.9370373994292917, "tokenization_path_logprob_std": 67.07168801658268, "trajectory_balance_rmse": 1160.26700118392} {"canonical_length": 354, "canonical_nll": 0.8953088447872528, "generation_rouge_l_f1": 0.25278810408921937, "generation_token_f1": 0.41263940520446096, "prompt_id": "85c649ae8170e371", "reading_canonical_log_prob": -316.9393310546875, "reading_distances": [22, 24, 24, 26], "reading_log_probs": [-314.4084777832031, -314.5444030761719, -315.40338134765625, -316.0382080078125], "reading_logprob_std": 0.7657583999951059, "tokenization_path_logprob_std": 270.81103879573413, "trajectory_balance_rmse": 5346.536610306873} {"canonical_length": 364, "canonical_nll": 1.0192862709799966, "generation_rouge_l_f1": 0.23895253682487727, "generation_token_f1": 0.5040916530278232, "prompt_id": "ff47696a7202ba2a", "reading_canonical_log_prob": -371.02020263671875, "reading_distances": [34, 28, 33, 29], "reading_log_probs": [-372.4769287109375, -372.994384765625, -375.5843505859375, -374.85235595703125], "reading_logprob_std": 1.4793659146876623, "tokenization_path_logprob_std": 175.70953463697865, "trajectory_balance_rmse": 6938.851028783447} {"canonical_length": 41, "canonical_nll": 1.042161895007622, "generation_rouge_l_f1": 0.3220338983050847, "generation_token_f1": 0.3389830508474576, "prompt_id": "3a5ef38cf12c45a8", "reading_canonical_log_prob": -42.7286376953125, "reading_distances": [24, 17, 17, 21], "reading_log_probs": [-43.846923828125, -44.771976470947266, -43.7579460144043, -44.22578048706055], "reading_logprob_std": 0.4612147660222091, "tokenization_path_logprob_std": 61.6153194047166, "trajectory_balance_rmse": 763.5277393173156} {"canonical_length": 230, "canonical_nll": 1.0708855670431385, "generation_rouge_l_f1": 0.26890756302521013, "generation_token_f1": 0.5168067226890756, "prompt_id": "13b43b11a6bf37ad", "reading_canonical_log_prob": -246.30368041992188, "reading_distances": [20, 19, 22, 18], "reading_log_probs": [-239.85108947753906, -239.7677001953125, -237.26641845703125, -240.11837768554688], "reading_logprob_std": 1.3314145835762734, "tokenization_path_logprob_std": 251.47277517799984, "trajectory_balance_rmse": 4283.727737291254} {"canonical_length": 111, "canonical_nll": 1.1741612064945806, "generation_rouge_l_f1": 0.31343283582089554, "generation_token_f1": 0.5149253731343284, "prompt_id": "d6b3da36a5892656", "reading_canonical_log_prob": -130.33189392089844, "reading_distances": [32, 35, 30, 40], "reading_log_probs": [-126.54655456542969, -125.52249908447266, -126.06965637207031, -126.1112060546875], "reading_logprob_std": 0.4196559315865565, "tokenization_path_logprob_std": 174.2594355420597, "trajectory_balance_rmse": 2277.356695908274} {"canonical_length": 184, "canonical_nll": 1.2586872266686482, "generation_rouge_l_f1": 0.23451327433628322, "generation_token_f1": 0.42920353982300885, "prompt_id": "f1036fae25a980bc", "reading_canonical_log_prob": -231.59844970703125, "reading_distances": [17, 8, 10, 14], "reading_log_probs": [-230.79556274414062, -233.1178436279297, -229.48193359375, -230.5125732421875], "reading_logprob_std": 1.5348231915161255, "tokenization_path_logprob_std": 266.9166539667892, "trajectory_balance_rmse": 4485.08195908355} {"canonical_length": 199, "canonical_nll": 1.0186565149968594, "generation_rouge_l_f1": 0.2980561555075594, "generation_token_f1": 0.5529157667386609, "prompt_id": "272e476421ef865b", "reading_canonical_log_prob": -202.712646484375, "reading_distances": [25, 28, 25, 30], "reading_log_probs": [-200.4061279296875, -200.42333984375, -199.265625, -201.9036865234375], "reading_logprob_std": 1.0814659602158938, "tokenization_path_logprob_std": 164.47653922374047, "trajectory_balance_rmse": 3390.9918798604704} {"canonical_length": 202, "canonical_nll": 0.9848993886815439, "generation_rouge_l_f1": 0.259958071278826, "generation_token_f1": 0.5408805031446541, "prompt_id": "548a4dcfac376298", "reading_canonical_log_prob": -198.94967651367188, "reading_distances": [23, 23, 22, 25], "reading_log_probs": [-199.3424835205078, -199.59478759765625, -198.46621704101562, -197.73976135253906], "reading_logprob_std": 0.8486631829351513, "tokenization_path_logprob_std": 135.06238124675374, "trajectory_balance_rmse": 4073.495279824626} {"canonical_length": 415, "canonical_nll": 0.9480128276778991, "generation_rouge_l_f1": 0.2163009404388715, "generation_token_f1": 0.4545454545454546, "prompt_id": "8ea8da1e0ef16ea4", "reading_canonical_log_prob": -393.4253234863281, "reading_distances": [32, 28, 28, 26], "reading_log_probs": [-398.7932434082031, -392.33697509765625, -395.2283630371094, -390.33599853515625], "reading_logprob_std": 3.6766835740097727, "tokenization_path_logprob_std": 73.40520568089646, "trajectory_balance_rmse": 4030.0668083163005} {"canonical_length": 70, "canonical_nll": 1.03131103515625, "generation_rouge_l_f1": 0.1242603550295858, "generation_token_f1": 0.20710059171597633, "prompt_id": "68d1893ed02225a9", "reading_canonical_log_prob": -72.1917724609375, "reading_distances": [12, 15, 21, 12], "reading_log_probs": [-69.28160095214844, -69.17123413085938, -68.66136169433594, -70.50398254394531], "reading_logprob_std": 0.7811601499085469, "tokenization_path_logprob_std": 66.90851847949598, "trajectory_balance_rmse": 1428.0710145296837} {"canonical_length": 207, "canonical_nll": 1.152246300149079, "generation_rouge_l_f1": 0.3036876355748373, "generation_token_f1": 0.5292841648590021, "prompt_id": "06716de5ff9ed39f", "reading_canonical_log_prob": -238.51498413085938, "reading_distances": [27, 29, 30, 30], "reading_log_probs": [-245.71524047851562, -239.8369140625, -240.7710418701172, -246.35440063476562], "reading_logprob_std": 3.3408141237485887, "tokenization_path_logprob_std": 264.4383799606197, "trajectory_balance_rmse": 4705.621491191485} {"canonical_length": 24, "canonical_nll": 0.6488871574401855, "generation_rouge_l_f1": 0.15748031496062992, "generation_token_f1": 0.1811023622047244, "prompt_id": "8d76cf27c40ca6eb", "reading_canonical_log_prob": -15.573291778564453, "reading_distances": [33, 27, 25, 27], "reading_log_probs": [-14.040327072143555, -13.879232406616211, -14.04577350616455, -15.071819305419922], "reading_logprob_std": 0.5471689877162413, "tokenization_path_logprob_std": 91.79376323077784, "trajectory_balance_rmse": 405.8856011487958} {"canonical_length": 37, "canonical_nll": 1.3491366618388407, "generation_rouge_l_f1": 0.14965986394557823, "generation_token_f1": 0.17006802721088435, "prompt_id": "f7669b85c591c347", "reading_canonical_log_prob": -49.91805648803711, "reading_distances": [13, 25, 26, 18], "reading_log_probs": [-50.81999206542969, -49.52336502075195, -49.16354751586914, -48.90736770629883], "reading_logprob_std": 0.8493983949912395, "tokenization_path_logprob_std": 76.23668602640365, "trajectory_balance_rmse": 834.9985237709737} {"canonical_length": 292, "canonical_nll": 1.1536235025484267, "generation_rouge_l_f1": 0.2615384615384615, "generation_token_f1": 0.3487179487179487, "prompt_id": "9875c8d91c22a365", "reading_canonical_log_prob": -336.8580627441406, "reading_distances": [29, 21, 25, 22], "reading_log_probs": [-334.95697021484375, -337.5313720703125, -335.0601501464844, -333.91705322265625], "reading_logprob_std": 1.5328765842083718, "tokenization_path_logprob_std": 194.02476177829894, "trajectory_balance_rmse": 6211.299299470736} {"canonical_length": 293, "canonical_nll": 1.338845106521971, "generation_rouge_l_f1": 0.2285714285714286, "generation_token_f1": 0.4357142857142857, "prompt_id": "967c6a5c34b145df", "reading_canonical_log_prob": -392.2816162109375, "reading_distances": [37, 32, 28, 37], "reading_log_probs": [-383.66693115234375, -388.788818359375, -391.04315185546875, -386.6600341796875], "reading_logprob_std": 3.141476408608149, "tokenization_path_logprob_std": 174.47543584480658, "trajectory_balance_rmse": 6363.183074667283} {"canonical_length": 9, "canonical_nll": 0.051866746611065336, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "abcadc4b6374d8f6", "reading_canonical_log_prob": -0.466800719499588, "reading_distances": [43, 51, 41, 50], "reading_log_probs": [-0.8861374855041504, -1.2207332849502563, -0.8003956079483032, -1.2017403841018677], "reading_logprob_std": 0.21545172129598844, "tokenization_path_logprob_std": 47.69580461002964, "trajectory_balance_rmse": 190.13274446193628} {"canonical_length": 261, "canonical_nll": 1.2864503823934388, "generation_rouge_l_f1": 0.21304347826086956, "generation_token_f1": 0.3913043478260869, "prompt_id": "ba38639bb894f700", "reading_canonical_log_prob": -335.7635498046875, "reading_distances": [21, 30, 29, 25], "reading_log_probs": [-333.59405517578125, -331.1829833984375, -331.20611572265625, -330.8594970703125], "reading_logprob_std": 1.2655253513376525, "tokenization_path_logprob_std": 268.49582322297744, "trajectory_balance_rmse": 4371.881458289123} {"canonical_length": 258, "canonical_nll": 1.1260649215343386, "generation_rouge_l_f1": 0.2771855010660981, "generation_token_f1": 0.40938166311300633, "prompt_id": "49e9fe2e1aacf453", "reading_canonical_log_prob": -290.5247497558594, "reading_distances": [15, 21, 17, 17], "reading_log_probs": [-288.2779846191406, -287.09466552734375, -286.4455261230469, -285.6362609863281], "reading_logprob_std": 1.1157989588202235, "tokenization_path_logprob_std": 128.64313110913812, "trajectory_balance_rmse": 4058.8453294236715} {"canonical_length": 255, "canonical_nll": 1.0794304342830883, "generation_rouge_l_f1": 0.21774193548387097, "generation_token_f1": 0.5, "prompt_id": "6193ec67d5074a6a", "reading_canonical_log_prob": -275.2547607421875, "reading_distances": [21, 21, 17, 15], "reading_log_probs": [-273.4087829589844, -274.4500732421875, -271.8642578125, -276.18182373046875], "reading_logprob_std": 1.8139821560223086, "tokenization_path_logprob_std": 113.28821164718279, "trajectory_balance_rmse": 4252.379837111249} {"canonical_length": 289, "canonical_nll": 0.9342165620269247, "generation_rouge_l_f1": 0.3369565217391304, "generation_token_f1": 0.536231884057971, "prompt_id": "821b32f3ecbd67a9", "reading_canonical_log_prob": -269.98858642578125, "reading_distances": [17, 18, 17, 17], "reading_log_probs": [-273.5680236816406, -271.49200439453125, -272.608154296875, -272.55682373046875], "reading_logprob_std": 0.8483315147612983, "tokenization_path_logprob_std": 218.98316795278663, "trajectory_balance_rmse": 5350.697321624883} {"canonical_length": 22, "canonical_nll": 0.4966075203635476, "generation_rouge_l_f1": 0.6521739130434783, "generation_token_f1": 0.7391304347826089, "prompt_id": "ebadf5b1dd406fa9", "reading_canonical_log_prob": -10.925365447998047, "reading_distances": [30, 36, 37, 43], "reading_log_probs": [-12.165960311889648, -12.1035795211792, -11.86862564086914, -12.632272720336914], "reading_logprob_std": 0.319846736552636, "tokenization_path_logprob_std": 66.30660079952723, "trajectory_balance_rmse": 480.35688068146465} {"canonical_length": 8, "canonical_nll": 0.9595612287521362, "generation_rouge_l_f1": 0.05860805860805861, "generation_token_f1": 0.05860805860805861, "prompt_id": "6edf6d71173f0e03", "reading_canonical_log_prob": -7.67648983001709, "reading_distances": [30, 32, 33, 27], "reading_log_probs": [-16.179685592651367, -17.24233627319336, -16.912181854248047, -17.896739959716797], "reading_logprob_std": 0.7141809837088875, "tokenization_path_logprob_std": 27.975665539730034, "trajectory_balance_rmse": 163.94901951288114} {"canonical_length": 16, "canonical_nll": 1.4420078992843628, "generation_rouge_l_f1": 0.6, "generation_token_f1": 0.6, "prompt_id": "938d726d90109110", "reading_canonical_log_prob": -23.072126388549805, "reading_distances": [25, 24, 32, 23], "reading_log_probs": [-28.93845558166504, -30.82363510131836, -29.468677520751953, -28.214601516723633], "reading_logprob_std": 1.102069067251036, "tokenization_path_logprob_std": 70.91965498524932, "trajectory_balance_rmse": 469.9437883293841} {"canonical_length": 6, "canonical_nll": 0.09590452909469604, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "b4288f8c25b7da9a", "reading_canonical_log_prob": -0.5754271745681763, "reading_distances": [42, 33, 33, 37], "reading_log_probs": [-2.3771584033966064, -0.695468544960022, -1.5016371011734009, -0.9444578289985657], "reading_logprob_std": 0.7455200317245017, "tokenization_path_logprob_std": 16.714815805452492, "trajectory_balance_rmse": 146.70554578739817} {"canonical_length": 266, "canonical_nll": 1.3265882219587053, "generation_rouge_l_f1": 0.2480916030534351, "generation_token_f1": 0.5572519083969466, "prompt_id": "3c7a013fe65f77bc", "reading_canonical_log_prob": -352.8724670410156, "reading_distances": [27, 25, 25, 26], "reading_log_probs": [-347.6961975097656, -346.2840881347656, -346.7928161621094, -348.049072265625], "reading_logprob_std": 0.8107005162815301, "tokenization_path_logprob_std": 84.7189876342465, "trajectory_balance_rmse": 5174.20127962815} {"canonical_length": 28, "canonical_nll": 1.7568749019077845, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "66c2a007594b76e9", "reading_canonical_log_prob": -49.19249725341797, "reading_distances": [61, 62, 51, 74], "reading_log_probs": [-50.9589958190918, -48.17978286743164, -46.36693572998047, -52.920143127441406], "reading_logprob_std": 2.9063020115820883, "tokenization_path_logprob_std": 57.25917167162295, "trajectory_balance_rmse": 566.0864270846702} {"canonical_length": 321, "canonical_nll": 1.529649585949669, "generation_rouge_l_f1": 0.21724137931034485, "generation_token_f1": 0.46551724137931033, "prompt_id": "dd27e1dd582c59a2", "reading_canonical_log_prob": -491.01751708984375, "reading_distances": [22, 25, 24, 24], "reading_log_probs": [-498.98974609375, -497.90716552734375, -497.57476806640625, -496.7040710449219], "reading_logprob_std": 0.9449202628773744, "tokenization_path_logprob_std": 192.7159477210256, "trajectory_balance_rmse": 6501.218736832401} {"canonical_length": 21, "canonical_nll": 0.3905743190220424, "generation_rouge_l_f1": 0.14285714285714285, "generation_token_f1": 0.2857142857142857, "prompt_id": "3c6903de8714c50f", "reading_canonical_log_prob": -8.20206069946289, "reading_distances": [18, 15, 16, 12], "reading_log_probs": [-7.906182289123535, -7.946601867675781, -7.1818647384643555, -7.709996700286865], "reading_logprob_std": 0.3517203781555513, "tokenization_path_logprob_std": 62.36011036162681, "trajectory_balance_rmse": 562.980754943374} {"canonical_length": 45, "canonical_nll": 1.5170905219184028, "generation_rouge_l_f1": 0.21621621621621623, "generation_token_f1": 0.33783783783783783, "prompt_id": "0692a96622cd2aed", "reading_canonical_log_prob": -68.26907348632812, "reading_distances": [21, 27, 23, 25], "reading_log_probs": [-67.97294616699219, -69.94828796386719, -66.4132080078125, -68.1609115600586], "reading_logprob_std": 1.4467222072613597, "tokenization_path_logprob_std": 40.98743269841289, "trajectory_balance_rmse": 1227.7267145555143} {"canonical_length": 322, "canonical_nll": 1.0027113256987579, "generation_rouge_l_f1": 0.27288428324697755, "generation_token_f1": 0.5043177892918826, "prompt_id": "e5a0acf44a48d7ab", "reading_canonical_log_prob": -322.873046875, "reading_distances": [25, 33, 23, 26], "reading_log_probs": [-331.618896484375, -326.18701171875, -327.0458984375, -323.53912353515625], "reading_logprob_std": 3.363367392446236, "tokenization_path_logprob_std": 137.0466757740902, "trajectory_balance_rmse": 5024.3023766366405} {"canonical_length": 191, "canonical_nll": 0.8143866573952879, "generation_rouge_l_f1": 0.2450980392156863, "generation_token_f1": 0.47058823529411764, "prompt_id": "4e165cac00833611", "reading_canonical_log_prob": -155.5478515625, "reading_distances": [18, 19, 18, 15], "reading_log_probs": [-156.45819091796875, -153.04852294921875, -153.3766632080078, -154.0358123779297], "reading_logprob_std": 1.5412781640927617, "tokenization_path_logprob_std": 37.545002781235866, "trajectory_balance_rmse": 2640.2409388198616} {"canonical_length": 296, "canonical_nll": 1.201396014239337, "generation_rouge_l_f1": 0.20138888888888887, "generation_token_f1": 0.4375, "prompt_id": "eadf46baa67f8e23", "reading_canonical_log_prob": -355.61322021484375, "reading_distances": [25, 30, 24, 31], "reading_log_probs": [-353.6736145019531, -356.6078796386719, -353.93670654296875, -354.4770202636719], "reading_logprob_std": 1.3320490616773286, "tokenization_path_logprob_std": 333.2533874512152, "trajectory_balance_rmse": 6562.3440221543} {"canonical_length": 97, "canonical_nll": 1.0504963668351321, "generation_rouge_l_f1": 0.29702970297029707, "generation_token_f1": 0.37623762376237624, "prompt_id": "228080ce5449beea", "reading_canonical_log_prob": -101.89814758300781, "reading_distances": [17, 17, 23, 21], "reading_log_probs": [-100.86985778808594, -100.00326538085938, -99.78926086425781, -100.56484985351562], "reading_logprob_std": 0.4978633752942673, "tokenization_path_logprob_std": 102.04653959299362, "trajectory_balance_rmse": 1081.8889864510782} {"canonical_length": 345, "canonical_nll": 1.1016626330389492, "generation_rouge_l_f1": 0.19528619528619526, "generation_token_f1": 0.39730639730639733, "prompt_id": "0584d61b9b2b3717", "reading_canonical_log_prob": -380.0736083984375, "reading_distances": [39, 24, 40, 33], "reading_log_probs": [-388.9130859375, -392.1385192871094, -389.79730224609375, -392.5997314453125], "reading_logprob_std": 1.787088814666649, "tokenization_path_logprob_std": 221.67507812704514, "trajectory_balance_rmse": 5976.585745589578} {"canonical_length": 16, "canonical_nll": 1.5319828987121582, "generation_rouge_l_f1": 0.48000000000000004, "generation_token_f1": 0.5599999999999999, "prompt_id": "f605a0e57df216ae", "reading_canonical_log_prob": -24.51172637939453, "reading_distances": [34, 36, 38, 30], "reading_log_probs": [-21.919784545898438, -23.37166976928711, -24.519277572631836, -23.315086364746094], "reading_logprob_std": 1.0638955024157684, "tokenization_path_logprob_std": 40.94159738258045, "trajectory_balance_rmse": 288.132843230354} {"canonical_length": 337, "canonical_nll": 1.181376743033661, "generation_rouge_l_f1": 0.2420168067226891, "generation_token_f1": 0.480672268907563, "prompt_id": "ab315c4f854a9130", "reading_canonical_log_prob": -398.12396240234375, "reading_distances": [27, 25, 31, 21], "reading_log_probs": [-397.53802490234375, -397.8529968261719, -398.5288391113281, -397.97149658203125], "reading_logprob_std": 0.41334544553389974, "tokenization_path_logprob_std": 196.15060566913914, "trajectory_balance_rmse": 7481.149858430759} {"canonical_length": 283, "canonical_nll": 1.189863117339333, "generation_rouge_l_f1": 0.23463687150837986, "generation_token_f1": 0.45065176908752325, "prompt_id": "93f170181468154b", "reading_canonical_log_prob": -336.73126220703125, "reading_distances": [27, 25, 27, 26], "reading_log_probs": [-335.3697509765625, -335.52655029296875, -337.0247802734375, -332.587646484375], "reading_logprob_std": 1.8500859965862009, "tokenization_path_logprob_std": 73.29800724845657, "trajectory_balance_rmse": 5649.174298692242} {"canonical_length": 32, "canonical_nll": 0.7093687057495117, "generation_rouge_l_f1": 0.23021582733812954, "generation_token_f1": 0.30215827338129503, "prompt_id": "ce3af5b67c725855", "reading_canonical_log_prob": -22.699798583984375, "reading_distances": [15, 15, 15, 17], "reading_log_probs": [-24.49159049987793, -21.51074981689453, -24.294620513916016, -25.18231201171875], "reading_logprob_std": 1.618116337059005, "tokenization_path_logprob_std": 94.42797452352579, "trajectory_balance_rmse": 805.5726686633366} {"canonical_length": 64, "canonical_nll": 1.5337369441986084, "generation_rouge_l_f1": 0.20408163265306123, "generation_token_f1": 0.3945578231292517, "prompt_id": "5d5d479145243991", "reading_canonical_log_prob": -98.15916442871094, "reading_distances": [23, 19, 19, 21], "reading_log_probs": [-114.15026092529297, -104.03517150878906, -112.07063293457031, -111.29016876220703], "reading_logprob_std": 4.402960779044743, "tokenization_path_logprob_std": 84.38190263472279, "trajectory_balance_rmse": 1279.8103119383034} {"canonical_length": 259, "canonical_nll": 1.7192750437379343, "generation_rouge_l_f1": 0.21631878557874762, "generation_token_f1": 0.4743833017077799, "prompt_id": "a2496cc91dc1f05e", "reading_canonical_log_prob": -445.292236328125, "reading_distances": [14, 15, 15, 16], "reading_log_probs": [-441.7413330078125, -444.60107421875, -442.7327880859375, -442.61407470703125], "reading_logprob_std": 1.2033125819652575, "tokenization_path_logprob_std": 91.8531921052673, "trajectory_balance_rmse": 6029.173924360685} {"canonical_length": 201, "canonical_nll": 0.9589584122842817, "generation_rouge_l_f1": 0.3891402714932127, "generation_token_f1": 0.5610859728506787, "prompt_id": "419c50c5b36f610d", "reading_canonical_log_prob": -192.75064086914062, "reading_distances": [36, 29, 46, 37], "reading_log_probs": [-192.26109313964844, -191.35092163085938, -191.81935119628906, -191.51266479492188], "reading_logprob_std": 0.4003474287829643, "tokenization_path_logprob_std": 198.33025269080952, "trajectory_balance_rmse": 4689.225212467377} {"canonical_length": 120, "canonical_nll": 0.6565858840942382, "generation_rouge_l_f1": 0.18997361477572555, "generation_token_f1": 0.3007915567282322, "prompt_id": "07d185247528da20", "reading_canonical_log_prob": -78.7903060913086, "reading_distances": [20, 20, 15, 20], "reading_log_probs": [-79.0030517578125, -81.15174865722656, -84.60507202148438, -81.85873413085938], "reading_logprob_std": 2.311602866400495, "tokenization_path_logprob_std": 123.91705760508704, "trajectory_balance_rmse": 1820.9759565807235} {"canonical_length": 187, "canonical_nll": 1.2656007654526655, "generation_rouge_l_f1": 0.24464831804281345, "generation_token_f1": 0.5626911314984709, "prompt_id": "f75ec5897fcc8e0d", "reading_canonical_log_prob": -236.66734313964844, "reading_distances": [29, 29, 22, 32], "reading_log_probs": [-230.30337524414062, -231.96316528320312, -232.9620819091797, -231.2828826904297], "reading_logprob_std": 1.120393821749132, "tokenization_path_logprob_std": 102.96151510811308, "trajectory_balance_rmse": 3789.1244341834586} {"canonical_length": 362, "canonical_nll": 1.4783389265366023, "generation_rouge_l_f1": 0.3232, "generation_token_f1": 0.5376000000000001, "prompt_id": "cd40c2a793f29f44", "reading_canonical_log_prob": -535.15869140625, "reading_distances": [25, 21, 17, 17], "reading_log_probs": [-536.4765014648438, -538.0119018554688, -537.87890625, -535.6453247070312], "reading_logprob_std": 1.1409904353732725, "tokenization_path_logprob_std": 329.51398765903264, "trajectory_balance_rmse": 7520.595824377056} {"canonical_length": 131, "canonical_nll": 0.9901260492455868, "generation_rouge_l_f1": 0.187012987012987, "generation_token_f1": 0.31688311688311693, "prompt_id": "79a9080b044df25c", "reading_canonical_log_prob": -129.70651245117188, "reading_distances": [32, 28, 28, 29], "reading_log_probs": [-127.84728240966797, -127.76399230957031, -130.66917419433594, -130.3103790283203], "reading_logprob_std": 1.5569670440239571, "tokenization_path_logprob_std": 145.56051755166055, "trajectory_balance_rmse": 2478.8826968324947} {"canonical_length": 59, "canonical_nll": 1.9180741390939486, "generation_rouge_l_f1": 0.1481481481481481, "generation_token_f1": 0.22839506172839505, "prompt_id": "56223c1e8305751f", "reading_canonical_log_prob": -113.16637420654297, "reading_distances": [20, 19, 18, 21], "reading_log_probs": [-111.59828186035156, -109.80686950683594, -110.30531311035156, -110.64541625976562], "reading_logprob_std": 0.7558745539969403, "tokenization_path_logprob_std": 91.56454189514278, "trajectory_balance_rmse": 1303.3395493354237} {"canonical_length": 24, "canonical_nll": 0.857492208480835, "generation_rouge_l_f1": 0.1727272727272727, "generation_token_f1": 0.2181818181818182, "prompt_id": "5528bdba3c5f6117", "reading_canonical_log_prob": -20.57981300354004, "reading_distances": [47, 41, 48, 44], "reading_log_probs": [-21.80027198791504, -21.576284408569336, -20.999610900878906, -20.484386444091797], "reading_logprob_std": 0.5925251834740703, "tokenization_path_logprob_std": 77.03914209211854, "trajectory_balance_rmse": 611.4519315189235} {"canonical_length": 351, "canonical_nll": 1.4067784496861646, "generation_rouge_l_f1": 0.2305194805194805, "generation_token_f1": 0.4448051948051948, "prompt_id": "daf735e510f876d3", "reading_canonical_log_prob": -493.77923583984375, "reading_distances": [19, 21, 16, 18], "reading_log_probs": [-492.9455261230469, -492.1043395996094, -492.93927001953125, -487.6149597167969], "reading_logprob_std": 2.5547748246437334, "tokenization_path_logprob_std": 532.2467555165491, "trajectory_balance_rmse": 7570.617120762707} {"canonical_length": 73, "canonical_nll": 1.3901957681734267, "generation_rouge_l_f1": 0.18128654970760233, "generation_token_f1": 0.23976608187134502, "prompt_id": "d276e797ca2de2bf", "reading_canonical_log_prob": -101.48429107666016, "reading_distances": [53, 53, 48, 42], "reading_log_probs": [-98.56071472167969, -96.3988037109375, -97.44883728027344, -99.11772918701172], "reading_logprob_std": 1.2076396473845121, "tokenization_path_logprob_std": 154.3683072157179, "trajectory_balance_rmse": 1225.0051833112066} {"canonical_length": 107, "canonical_nll": 1.7503054609922606, "generation_rouge_l_f1": 0.26436781609195403, "generation_token_f1": 0.3448275862068966, "prompt_id": "60ad70d4930701a2", "reading_canonical_log_prob": -187.28268432617188, "reading_distances": [12, 13, 10, 12], "reading_log_probs": [-185.1593017578125, -187.58416748046875, -189.31768798828125, -189.82852172851562], "reading_logprob_std": 2.1070701499430555, "tokenization_path_logprob_std": 119.20390348288167, "trajectory_balance_rmse": 2147.59849053583} {"canonical_length": 11, "canonical_nll": 0.08230139992453835, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "91d144e62a7c6f3d", "reading_canonical_log_prob": -0.9053153991699219, "reading_distances": [35, 35, 25, 26], "reading_log_probs": [-0.7612084746360779, -0.8005045056343079, -0.7397109270095825, -0.6048124432563782], "reading_logprob_std": 0.08497788174780592, "tokenization_path_logprob_std": 38.7420584028913, "trajectory_balance_rmse": 269.44982840438485} {"canonical_length": 31, "canonical_nll": 0.9901541433026714, "generation_rouge_l_f1": 0.1596244131455399, "generation_token_f1": 0.19718309859154928, "prompt_id": "663423620e0283d5", "reading_canonical_log_prob": -30.694778442382812, "reading_distances": [15, 15, 18, 18], "reading_log_probs": [-30.1051082611084, -30.762802124023438, -29.84834098815918, -31.266340255737305], "reading_logprob_std": 0.6420969641904986, "tokenization_path_logprob_std": 35.98334104555607, "trajectory_balance_rmse": 439.5140168818785} {"canonical_length": 51, "canonical_nll": 0.14385188794603535, "generation_rouge_l_f1": 0.8505747126436781, "generation_token_f1": 0.8505747126436781, "prompt_id": "eece35ce9b12a6f2", "reading_canonical_log_prob": -7.336446285247803, "reading_distances": [24, 22, 23, 23], "reading_log_probs": [-6.263705253601074, -5.688540458679199, -5.706374168395996, -6.250761985778809], "reading_logprob_std": 0.32331218021269237, "tokenization_path_logprob_std": 57.651949511607086, "trajectory_balance_rmse": 992.2621757919165} {"canonical_length": 156, "canonical_nll": 1.3770523071289062, "generation_rouge_l_f1": 0.2295918367346939, "generation_token_f1": 0.3877551020408163, "prompt_id": "25055edbaab1f2e6", "reading_canonical_log_prob": -214.82015991210938, "reading_distances": [9, 5, 8, 8], "reading_log_probs": [-214.0050048828125, -214.2106170654297, -215.12548828125, -213.167236328125], "reading_logprob_std": 0.8041558194503684, "tokenization_path_logprob_std": 161.62419769146243, "trajectory_balance_rmse": 3051.895848703425} {"canonical_length": 297, "canonical_nll": 0.9534761062776199, "generation_rouge_l_f1": 0.36551724137931035, "generation_token_f1": 0.6172413793103447, "prompt_id": "51338c6674b54120", "reading_canonical_log_prob": -283.1824035644531, "reading_distances": [28, 22, 22, 29], "reading_log_probs": [-279.02447509765625, -277.47998046875, -279.1164245605469, -278.79046630859375], "reading_logprob_std": 0.7610436726314749, "tokenization_path_logprob_std": 148.5470543590097, "trajectory_balance_rmse": 5037.288169919581} {"canonical_length": 130, "canonical_nll": 2.086354417067308, "generation_rouge_l_f1": 0.15463917525773196, "generation_token_f1": 0.2731958762886598, "prompt_id": "156614ecc1fc9430", "reading_canonical_log_prob": -271.22607421875, "reading_distances": [67, 59, 68, 64], "reading_log_probs": [-271.40362548828125, -269.8304443359375, -272.2895812988281, -275.153076171875], "reading_logprob_std": 2.234121084009494, "tokenization_path_logprob_std": 94.90740293039701, "trajectory_balance_rmse": 3076.862383952883} {"canonical_length": 234, "canonical_nll": 1.4941401033319979, "generation_rouge_l_f1": 0.30165289256198347, "generation_token_f1": 0.5165289256198347, "prompt_id": "39459679cbe7ca40", "reading_canonical_log_prob": -349.6287841796875, "reading_distances": [25, 26, 27, 29], "reading_log_probs": [-349.11981201171875, -348.629638671875, -348.32373046875, -349.31298828125], "reading_logprob_std": 0.4518951145106857, "tokenization_path_logprob_std": 59.16925820469898, "trajectory_balance_rmse": 3968.536720988034} {"canonical_length": 81, "canonical_nll": 1.1196648868513697, "generation_rouge_l_f1": 0.2906574394463668, "generation_token_f1": 0.39446366782006914, "prompt_id": "0fc229d90ecbfd68", "reading_canonical_log_prob": -90.69285583496094, "reading_distances": [24, 26, 32, 30], "reading_log_probs": [-90.06657409667969, -92.08489227294922, -90.18218231201172, -88.82140350341797], "reading_logprob_std": 1.3465946691766966, "tokenization_path_logprob_std": 99.35162034454501, "trajectory_balance_rmse": 1690.037431544717} {"canonical_length": 7, "canonical_nll": 3.3408816201346263, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "fc7a567baaf9e2df", "reading_canonical_log_prob": -23.386171340942383, "reading_distances": [19, 24, 24, 21], "reading_log_probs": [-24.709278106689453, -23.702316284179688, -24.54375457763672, -23.65488624572754], "reading_logprob_std": 0.5517749593236547, "tokenization_path_logprob_std": 49.538271426319056, "trajectory_balance_rmse": 127.86800177411466} {"canonical_length": 3, "canonical_nll": 1.4758820533752441, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "14871a1dde668d87", "reading_canonical_log_prob": -4.427646160125732, "reading_distances": [30, 26, 20, 26], "reading_log_probs": [-4.634616851806641, -4.457536697387695, -4.285074710845947, -4.997621059417725], "reading_logprob_std": 0.30474930925736, "tokenization_path_logprob_std": 26.091434976187998, "trajectory_balance_rmse": 41.33702865112765} {"canonical_length": 3, "canonical_nll": 0.4850122133890788, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "408cd1b200daeec5", "reading_canonical_log_prob": -1.4550366401672363, "reading_distances": [31, 29, 23, 30], "reading_log_probs": [-1.8143466711044312, -1.8907512426376343, -2.252682685852051, -2.1377999782562256], "reading_logprob_std": 0.20571494155464937, "tokenization_path_logprob_std": 21.80531094870678, "trajectory_balance_rmse": 55.321747072033354} {"canonical_length": 129, "canonical_nll": 1.5296540962633236, "generation_rouge_l_f1": 0.19230769230769232, "generation_token_f1": 0.19230769230769232, "prompt_id": "6db695863766c033", "reading_canonical_log_prob": -197.32537841796875, "reading_distances": [9, 13, 11, 12], "reading_log_probs": [-190.21487426757812, -189.812744140625, -191.81781005859375, -190.83065795898438], "reading_logprob_std": 0.8727928586103233, "tokenization_path_logprob_std": 62.360745537994255, "trajectory_balance_rmse": 2389.307082605607} {"canonical_length": 6, "canonical_nll": 0.2996327877044678, "generation_rouge_l_f1": 0.888888888888889, "generation_token_f1": 0.888888888888889, "prompt_id": "66a509bfd96df65b", "reading_canonical_log_prob": -1.7977967262268066, "reading_distances": [16, 17, 13, 19], "reading_log_probs": [-1.8508267402648926, -1.940744161605835, -1.9436194896697998, -1.9307938814163208], "reading_logprob_std": 0.04412307831609203, "tokenization_path_logprob_std": 26.12324821078992, "trajectory_balance_rmse": 96.44651031215784} {"canonical_length": 3, "canonical_nll": 0.39757927258809406, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "2c889c2667571076", "reading_canonical_log_prob": -1.1927378177642822, "reading_distances": [13, 18, 20, 19], "reading_log_probs": [-1.925106167793274, -2.1912777423858643, -2.2799644470214844, -2.4362409114837646], "reading_logprob_std": 0.21415046322696643, "tokenization_path_logprob_std": 49.206613787528504, "trajectory_balance_rmse": 88.33515403559069} {"canonical_length": 15, "canonical_nll": 0.4337038040161133, "generation_rouge_l_f1": 0.11029411764705882, "generation_token_f1": 0.11029411764705882, "prompt_id": "79afdb8eb4cf0c62", "reading_canonical_log_prob": -6.505557060241699, "reading_distances": [23, 24, 26, 24], "reading_log_probs": [-6.84561014175415, -6.536401748657227, -6.502752780914307, -6.8809685707092285], "reading_logprob_std": 0.19944024800283594, "tokenization_path_logprob_std": 71.14750968081475, "trajectory_balance_rmse": 277.8841823379504} {"canonical_length": 145, "canonical_nll": 1.3297797893655712, "generation_rouge_l_f1": 0.2864583333333333, "generation_token_f1": 0.48958333333333326, "prompt_id": "51fd58caa03ae57f", "reading_canonical_log_prob": -192.8180694580078, "reading_distances": [29, 30, 29, 22], "reading_log_probs": [-196.55599975585938, -196.35195922851562, -195.09780883789062, -194.1783447265625], "reading_logprob_std": 1.1166954294484306, "tokenization_path_logprob_std": 108.44496119240567, "trajectory_balance_rmse": 3052.136100751131} {"canonical_length": 19, "canonical_nll": 0.7983434576737253, "generation_rouge_l_f1": 0.2834645669291339, "generation_token_f1": 0.2834645669291339, "prompt_id": "8671c943f4179a67", "reading_canonical_log_prob": -15.168525695800781, "reading_distances": [16, 13, 15, 16], "reading_log_probs": [-16.978546142578125, -17.166048049926758, -17.478775024414062, -17.546756744384766], "reading_logprob_std": 0.26702184248137073, "tokenization_path_logprob_std": 76.8028245461046, "trajectory_balance_rmse": 428.941317547099} {"canonical_length": 289, "canonical_nll": 1.2383022638340722, "generation_rouge_l_f1": 0.22342342342342345, "generation_token_f1": 0.4648648648648649, "prompt_id": "9cb312e17f963300", "reading_canonical_log_prob": -357.8693542480469, "reading_distances": [21, 13, 21, 18], "reading_log_probs": [-363.1939392089844, -362.38482666015625, -363.3451232910156, -361.8157043457031], "reading_logprob_std": 0.7166032851831742, "tokenization_path_logprob_std": 175.41861634413084, "trajectory_balance_rmse": 6361.978282618385} {"canonical_length": 28, "canonical_nll": 1.7890160424368722, "generation_rouge_l_f1": 0.2, "generation_token_f1": 0.25, "prompt_id": "9227ddd5a86a7e6a", "reading_canonical_log_prob": -50.09244918823242, "reading_distances": [25, 25, 27, 24], "reading_log_probs": [-50.8842887878418, -51.31887435913086, -50.848548889160156, -50.991580963134766], "reading_logprob_std": 0.21417258596764288, "tokenization_path_logprob_std": 34.72478047593079, "trajectory_balance_rmse": 610.3940693768104} {"canonical_length": 223, "canonical_nll": 1.6177168362878362, "generation_rouge_l_f1": 0.24786324786324784, "generation_token_f1": 0.5641025641025641, "prompt_id": "3216cdb4c067a3d8", "reading_canonical_log_prob": -360.7508544921875, "reading_distances": [46, 47, 49, 44], "reading_log_probs": [-360.2381591796875, -359.5721740722656, -362.6365966796875, -359.796875], "reading_logprob_std": 1.4111451016180268, "tokenization_path_logprob_std": 205.53364976020697, "trajectory_balance_rmse": 5050.9627125729185} {"canonical_length": 108, "canonical_nll": 0.8266653131555628, "generation_rouge_l_f1": 0.24691358024691357, "generation_token_f1": 0.3518518518518518, "prompt_id": "1ef389c9aee61602", "reading_canonical_log_prob": -89.27985382080078, "reading_distances": [14, 9, 14, 15], "reading_log_probs": [-90.63534545898438, -91.29328918457031, -90.50354766845703, -89.33396911621094], "reading_logprob_std": 0.8151882078994048, "tokenization_path_logprob_std": 61.316567341816736, "trajectory_balance_rmse": 1489.6439287838807} {"canonical_length": 6, "canonical_nll": 3.8975213368733725, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "a730d1195f266d12", "reading_canonical_log_prob": -23.385128021240234, "reading_distances": [32, 30, 25, 29], "reading_log_probs": [-27.317956924438477, -24.59674072265625, -23.834693908691406, -27.38801383972168], "reading_logprob_std": 1.838048111405039, "tokenization_path_logprob_std": 40.24207689464398, "trajectory_balance_rmse": 136.7647298637159} {"canonical_length": 26, "canonical_nll": 1.25830811720628, "generation_rouge_l_f1": 0.26829268292682923, "generation_token_f1": 0.36585365853658536, "prompt_id": "7f409abdb6fbebf8", "reading_canonical_log_prob": -32.71601104736328, "reading_distances": [15, 11, 11, 11], "reading_log_probs": [-35.05149841308594, -34.519981384277344, -33.657588958740234, -35.36092758178711], "reading_logprob_std": 0.7457386716518107, "tokenization_path_logprob_std": 64.40564451636722, "trajectory_balance_rmse": 443.5617119292084} {"canonical_length": 168, "canonical_nll": 1.3072933923630488, "generation_rouge_l_f1": 0.27860696517412936, "generation_token_f1": 0.4626865671641791, "prompt_id": "1c78ad010cb27d62", "reading_canonical_log_prob": -219.6252899169922, "reading_distances": [39, 33, 35, 41], "reading_log_probs": [-217.68118286132812, -215.8860321044922, -218.3154296875, -219.262451171875], "reading_logprob_std": 1.4237354849419817, "tokenization_path_logprob_std": 35.40618197116674, "trajectory_balance_rmse": 3582.576730758606} {"canonical_length": 15, "canonical_nll": 0.46719608306884763, "generation_rouge_l_f1": 0.09395973154362416, "generation_token_f1": 0.09395973154362416, "prompt_id": "779d0398df55911b", "reading_canonical_log_prob": -7.007941246032715, "reading_distances": [6, 5, 8, 7], "reading_log_probs": [-9.233072280883789, -6.4874467849731445, -6.5513505935668945, -7.1195573806762695], "reading_logprob_std": 1.2885244228797026, "tokenization_path_logprob_std": 62.630266716889444, "trajectory_balance_rmse": 208.10738393487568} {"canonical_length": 11, "canonical_nll": 0.574955246665261, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "3e344435d6b1501d", "reading_canonical_log_prob": -6.324507713317871, "reading_distances": [41, 47, 43, 38], "reading_log_probs": [-6.683341026306152, -7.068694114685059, -6.089149475097656, -6.336299896240234], "reading_logprob_std": 0.426125027724153, "tokenization_path_logprob_std": 32.31949768341351, "trajectory_balance_rmse": 149.04989669468532} {"canonical_length": 376, "canonical_nll": 0.5116247217705909, "generation_rouge_l_f1": 0.3670411985018726, "generation_token_f1": 0.5730337078651685, "prompt_id": "e2124bce16d0a6b8", "reading_canonical_log_prob": -192.3708953857422, "reading_distances": [13, 17, 9, 12], "reading_log_probs": [-187.35763549804688, -186.7225341796875, -189.14723205566406, -189.8612060546875], "reading_logprob_std": 1.4751851537930682, "tokenization_path_logprob_std": 178.5421299375498, "trajectory_balance_rmse": 4982.90756857734} {"canonical_length": 240, "canonical_nll": 2.1571131388346356, "generation_rouge_l_f1": 0.2698072805139186, "generation_token_f1": 0.5010706638115632, "prompt_id": "6a9d3b5c30c1cdf4", "reading_canonical_log_prob": -517.7071533203125, "reading_distances": [11, 18, 14, 13], "reading_log_probs": [-517.0715942382812, -523.80029296875, -517.7669677734375, -522.497314453125], "reading_logprob_std": 3.362443511832517, "tokenization_path_logprob_std": 181.12040744527812, "trajectory_balance_rmse": 4322.883392262726} {"canonical_length": 271, "canonical_nll": 1.075284275181619, "generation_rouge_l_f1": 0.26900584795321636, "generation_token_f1": 0.5263157894736843, "prompt_id": "63c15bf241bee263", "reading_canonical_log_prob": -291.40203857421875, "reading_distances": [33, 24, 32, 30], "reading_log_probs": [-288.95306396484375, -289.4009704589844, -290.9361877441406, -289.35638427734375], "reading_logprob_std": 0.8732471232856713, "tokenization_path_logprob_std": 48.71142578369726, "trajectory_balance_rmse": 5810.340489495757} {"canonical_length": 25, "canonical_nll": 1.4295858764648437, "generation_rouge_l_f1": 0.4242424242424242, "generation_token_f1": 0.4242424242424242, "prompt_id": "9932e0e2da85e27d", "reading_canonical_log_prob": -35.739646911621094, "reading_distances": [24, 18, 24, 22], "reading_log_probs": [-34.01403045654297, -34.51017379760742, -34.5756950378418, -34.9410285949707], "reading_logprob_std": 0.38126416506078126, "tokenization_path_logprob_std": 32.82486052453355, "trajectory_balance_rmse": 478.2139812570168} {"canonical_length": 114, "canonical_nll": 0.9612915306760553, "generation_rouge_l_f1": 0.2947368421052632, "generation_token_f1": 0.46315789473684205, "prompt_id": "f5ce2237f02c3721", "reading_canonical_log_prob": -109.58723449707031, "reading_distances": [24, 22, 21, 23], "reading_log_probs": [-107.83770751953125, -110.14353942871094, -108.72091674804688, -106.59217834472656], "reading_logprob_std": 1.4948748634965303, "tokenization_path_logprob_std": 154.40033774741164, "trajectory_balance_rmse": 2339.8998648988136} {"canonical_length": 297, "canonical_nll": 1.4806789822048612, "generation_rouge_l_f1": 0.23529411764705885, "generation_token_f1": 0.43382352941176466, "prompt_id": "3c5a7e9edbe46ab0", "reading_canonical_log_prob": -439.76165771484375, "reading_distances": [13, 21, 18, 16], "reading_log_probs": [-434.80865478515625, -438.645263671875, -439.6202392578125, -438.515625], "reading_logprob_std": 2.117390243079852, "tokenization_path_logprob_std": 325.5576639439319, "trajectory_balance_rmse": 5691.096358375451} {"canonical_length": 129, "canonical_nll": 1.1524246570675871, "generation_rouge_l_f1": 0.3228346456692913, "generation_token_f1": 0.49606299212598426, "prompt_id": "6097f6fb796f153f", "reading_canonical_log_prob": -148.66278076171875, "reading_distances": [26, 25, 32, 21], "reading_log_probs": [-146.172119140625, -144.9640350341797, -145.116455078125, -143.74964904785156], "reading_logprob_std": 0.9919836645773532, "tokenization_path_logprob_std": 197.80766153404363, "trajectory_balance_rmse": 2456.2122355250885} {"canonical_length": 10, "canonical_nll": 0.5037082672119141, "generation_rouge_l_f1": 0.12962962962962962, "generation_token_f1": 0.12962962962962962, "prompt_id": "b2680f937aebb340", "reading_canonical_log_prob": -5.037082672119141, "reading_distances": [7, 10, 5, 6], "reading_log_probs": [-5.961779594421387, -6.960442543029785, -4.020130634307861, -5.879307746887207], "reading_logprob_std": 1.2262733544772262, "tokenization_path_logprob_std": 15.907870779705162, "trajectory_balance_rmse": 47.26978069394718} {"canonical_length": 8, "canonical_nll": 0.5881825089454651, "generation_rouge_l_f1": 0.5, "generation_token_f1": 0.5, "prompt_id": "a6d7ae5badbbf976", "reading_canonical_log_prob": -4.705460071563721, "reading_distances": [29, 35, 25, 32], "reading_log_probs": [-7.481097221374512, -9.322583198547363, -11.251287460327148, -8.110304832458496], "reading_logprob_std": 1.6597338665865298, "tokenization_path_logprob_std": 17.465570911494428, "trajectory_balance_rmse": 89.4652965716019} {"canonical_length": 31, "canonical_nll": 1.8333938352523311, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "5eee9bc9190ed48e", "reading_canonical_log_prob": -56.835208892822266, "reading_distances": [26, 31, 23, 22], "reading_log_probs": [-65.19325256347656, -63.0587158203125, -62.244873046875, -59.911109924316406], "reading_logprob_std": 2.1826290147218326, "tokenization_path_logprob_std": 148.8124577312836, "trajectory_balance_rmse": 781.5178437624464} {"canonical_length": 14, "canonical_nll": 0.6238635608128139, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.6666666666666666, "prompt_id": "89b79b80175b4a32", "reading_canonical_log_prob": -8.734089851379395, "reading_distances": [28, 41, 37, 37], "reading_log_probs": [-9.244420051574707, -9.77789306640625, -10.120908737182617, -9.994056701660156], "reading_logprob_std": 0.38678793578480103, "tokenization_path_logprob_std": 60.24506136902495, "trajectory_balance_rmse": 278.9413799551504} {"canonical_length": 10, "canonical_nll": 0.4826799392700195, "generation_rouge_l_f1": 0.8888888888888888, "generation_token_f1": 0.8888888888888888, "prompt_id": "5a773e9fe3bdab08", "reading_canonical_log_prob": -4.826799392700195, "reading_distances": [19, 16, 20, 21], "reading_log_probs": [-2.9430158138275146, -3.5433497428894043, -3.335418701171875, -3.388929843902588], "reading_logprob_std": 0.2554676170453658, "tokenization_path_logprob_std": 35.218039885451574, "trajectory_balance_rmse": 221.04013151567918} {"canonical_length": 37, "canonical_nll": 1.2344659341348183, "generation_rouge_l_f1": 0.42666666666666664, "generation_token_f1": 0.64, "prompt_id": "bc2e832a6508e078", "reading_canonical_log_prob": -45.67523956298828, "reading_distances": [22, 12, 14, 17], "reading_log_probs": [-45.62974548339844, -44.42524337768555, -45.00444793701172, -45.11122131347656], "reading_logprob_std": 0.4939748528379455, "tokenization_path_logprob_std": 59.735535454301875, "trajectory_balance_rmse": 869.1817408607635} {"canonical_length": 245, "canonical_nll": 0.7386743039500957, "generation_rouge_l_f1": 0.22022471910112357, "generation_token_f1": 0.3595505617977528, "prompt_id": "66a81182d7c90c5b", "reading_canonical_log_prob": -180.97520446777344, "reading_distances": [16, 10, 19, 16], "reading_log_probs": [-173.25465393066406, -175.69924926757812, -175.6621551513672, -172.7720184326172], "reading_logprob_std": 1.5526319616647342, "tokenization_path_logprob_std": 198.15641911289683, "trajectory_balance_rmse": 3652.013446060745} {"canonical_length": 92, "canonical_nll": 1.14598738628885, "generation_rouge_l_f1": 0.2571428571428572, "generation_token_f1": 0.3714285714285714, "prompt_id": "aed52e93b5146883", "reading_canonical_log_prob": -105.43083953857422, "reading_distances": [24, 29, 23, 28], "reading_log_probs": [-106.61454772949219, -106.7353515625, -104.24064636230469, -105.13648986816406], "reading_logprob_std": 1.2047511724115714, "tokenization_path_logprob_std": 161.33539096479925, "trajectory_balance_rmse": 1482.4997780260921} {"canonical_length": 131, "canonical_nll": 0.8249596748643249, "generation_rouge_l_f1": 0.5301204819277109, "generation_token_f1": 0.7148594377510041, "prompt_id": "b1543c03f1b0c7a8", "reading_canonical_log_prob": -108.06971740722656, "reading_distances": [17, 20, 19, 19], "reading_log_probs": [-107.79365539550781, -108.68634796142578, -111.26324462890625, -108.58744049072266], "reading_logprob_std": 1.5076237138760809, "tokenization_path_logprob_std": 171.7207137100423, "trajectory_balance_rmse": 2472.867461270452} {"canonical_length": 501, "canonical_nll": 0.3624273060324663, "generation_rouge_l_f1": 0.07858546168958742, "generation_token_f1": 0.11394891944990175, "prompt_id": "c52f77a538102872", "reading_canonical_log_prob": -181.57608032226562, "reading_distances": [32, 29, 23, 31], "reading_log_probs": [-181.62130737304688, -181.66357421875, -179.63919067382812, -178.83065795898438], "reading_logprob_std": 1.4287397030991573, "tokenization_path_logprob_std": 69.75352351495651, "trajectory_balance_rmse": 2802.8711103897335} {"canonical_length": 67, "canonical_nll": 0.6147500223188258, "generation_rouge_l_f1": 0.17843866171003717, "generation_token_f1": 0.21561338289962828, "prompt_id": "fdb23a8f949c30eb", "reading_canonical_log_prob": -41.18825149536133, "reading_distances": [19, 20, 23, 21], "reading_log_probs": [-45.12982940673828, -42.00225830078125, -41.6392936706543, -43.53736114501953], "reading_logprob_std": 1.5966746724216412, "tokenization_path_logprob_std": 52.12975020716707, "trajectory_balance_rmse": 890.4232374844388} {"canonical_length": 268, "canonical_nll": 1.2867247168697529, "generation_rouge_l_f1": 0.2666666666666667, "generation_token_f1": 0.49019607843137253, "prompt_id": "8aa2f18f3606aa8c", "reading_canonical_log_prob": -344.84222412109375, "reading_distances": [22, 27, 24, 22], "reading_log_probs": [-346.5889587402344, -351.033447265625, -350.1424865722656, -348.58782958984375], "reading_logprob_std": 1.948681992291327, "tokenization_path_logprob_std": 408.86718820402, "trajectory_balance_rmse": 5660.519776407866} {"canonical_length": 123, "canonical_nll": 0.9441438380295668, "generation_rouge_l_f1": 0.14792899408284024, "generation_token_f1": 0.21893491124260356, "prompt_id": "75e528b4a6af3ee2", "reading_canonical_log_prob": -116.12969207763672, "reading_distances": [21, 18, 15, 19], "reading_log_probs": [-115.65788269042969, -114.77926635742188, -115.0094223022461, -113.22198486328125], "reading_logprob_std": 1.0327590784367289, "tokenization_path_logprob_std": 141.06446134662528, "trajectory_balance_rmse": 1495.7775088158746} {"canonical_length": 323, "canonical_nll": 1.5997495857923762, "generation_rouge_l_f1": 0.17421602787456444, "generation_token_f1": 0.3658536585365854, "prompt_id": "af23751b53e953f7", "reading_canonical_log_prob": -516.7191162109375, "reading_distances": [33, 36, 31, 28], "reading_log_probs": [-504.01324462890625, -505.52099609375, -508.083251953125, -511.4410400390625], "reading_logprob_std": 3.2518870801872812, "tokenization_path_logprob_std": 165.05691349791948, "trajectory_balance_rmse": 6110.535661456303} {"canonical_length": 250, "canonical_nll": 1.6489256591796875, "generation_rouge_l_f1": 0.29149797570850206, "generation_token_f1": 0.5101214574898785, "prompt_id": "b8fd0dbaa9eea228", "reading_canonical_log_prob": -412.2314147949219, "reading_distances": [15, 20, 19, 21], "reading_log_probs": [-410.5030822753906, -407.8542785644531, -408.52032470703125, -404.0921630859375], "reading_logprob_std": 2.680999801192313, "tokenization_path_logprob_std": 90.75489806641677, "trajectory_balance_rmse": 5080.723346053622} {"canonical_length": 13, "canonical_nll": 2.2267174353966346, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "44c43dead4708e04", "reading_canonical_log_prob": -28.94732666015625, "reading_distances": [34, 24, 29, 22], "reading_log_probs": [-28.72311782836914, -26.731679916381836, -29.30154800415039, -28.826845169067383], "reading_logprob_std": 1.1376308713356513, "tokenization_path_logprob_std": 31.575472106042277, "trajectory_balance_rmse": 382.5876071527213} {"canonical_length": 138, "canonical_nll": 1.7274050505264946, "generation_rouge_l_f1": 0.09210526315789475, "generation_token_f1": 0.13157894736842107, "prompt_id": "36e5fedd319115db", "reading_canonical_log_prob": -238.38189697265625, "reading_distances": [30, 27, 35, 26], "reading_log_probs": [-229.14279174804688, -230.32179260253906, -228.55181884765625, -229.66098022460938], "reading_logprob_std": 0.7531893483190434, "tokenization_path_logprob_std": 74.41231173240469, "trajectory_balance_rmse": 2917.244816373563} {"canonical_length": 13, "canonical_nll": 0.6226040033193735, "generation_rouge_l_f1": 0.8695652173913043, "generation_token_f1": 0.8695652173913043, "prompt_id": "15b19e1f65328c81", "reading_canonical_log_prob": -8.093852043151855, "reading_distances": [24, 21, 16, 24], "reading_log_probs": [-7.146914482116699, -6.80682373046875, -9.432955741882324, -8.699145317077637], "reading_logprob_std": 1.2505663714820017, "tokenization_path_logprob_std": 39.666429443835874, "trajectory_balance_rmse": 198.77805684464388} {"canonical_length": 20, "canonical_nll": 1.1352127075195313, "generation_rouge_l_f1": 0.2823529411764706, "generation_token_f1": 0.3529411764705882, "prompt_id": "7e1ececb041b4ecb", "reading_canonical_log_prob": -22.704254150390625, "reading_distances": [21, 24, 26, 23], "reading_log_probs": [-23.266311645507812, -22.29587173461914, -23.022987365722656, -21.72024917602539], "reading_logprob_std": 0.7040617129185393, "tokenization_path_logprob_std": 90.26197479878164, "trajectory_balance_rmse": 482.1873198335432} {"canonical_length": 359, "canonical_nll": 1.2971733752067374, "generation_rouge_l_f1": 0.2351046698872786, "generation_token_f1": 0.3800322061191626, "prompt_id": "4db416527bef041c", "reading_canonical_log_prob": -465.68524169921875, "reading_distances": [18, 21, 22, 15], "reading_log_probs": [-458.2396240234375, -459.7532043457031, -459.9827575683594, -460.6829833984375], "reading_logprob_std": 1.0290345380025763, "tokenization_path_logprob_std": 280.9160715978286, "trajectory_balance_rmse": 7496.408710048725} {"canonical_length": 78, "canonical_nll": 0.7825303689027444, "generation_rouge_l_f1": 0.26244343891402716, "generation_token_f1": 0.5158371040723981, "prompt_id": "ba11bfd32f28c7ee", "reading_canonical_log_prob": -61.03736877441406, "reading_distances": [16, 14, 14, 18], "reading_log_probs": [-62.42350769042969, -63.01200866699219, -62.11188888549805, -62.52083969116211], "reading_logprob_std": 0.3732309989221297, "tokenization_path_logprob_std": 100.50964590365608, "trajectory_balance_rmse": 1508.3035375311927} {"canonical_length": 292, "canonical_nll": 1.4748422283015838, "generation_rouge_l_f1": 0.19718309859154926, "generation_token_f1": 0.3697183098591549, "prompt_id": "2bb0a1d6597bcbfc", "reading_canonical_log_prob": -430.6539306640625, "reading_distances": [50, 64, 52, 54], "reading_log_probs": [-435.33697509765625, -435.900146484375, -432.9023132324219, -434.7789611816406], "reading_logprob_std": 1.3013450783132907, "tokenization_path_logprob_std": 174.76242775872245, "trajectory_balance_rmse": 6311.723836930553} {"canonical_length": 2, "canonical_nll": 2.2060608863830566, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "bad903d11c6658df", "reading_canonical_log_prob": -4.412121772766113, "reading_distances": [28, 34, 34, 30], "reading_log_probs": [-5.51483678817749, -5.790849685668945, -4.877680778503418, -4.9307355880737305], "reading_logprob_std": 0.4471962319946472, "tokenization_path_logprob_std": 24.849021175336315, "trajectory_balance_rmse": 41.08478600242773} {"canonical_length": 99, "canonical_nll": 0.8940892075047349, "generation_rouge_l_f1": 0.20233463035019456, "generation_token_f1": 0.3657587548638132, "prompt_id": "7c8a24a51c291f25", "reading_canonical_log_prob": -88.51483154296875, "reading_distances": [10, 13, 11, 10], "reading_log_probs": [-87.20553588867188, -86.69403076171875, -87.55457305908203, -88.24504089355469], "reading_logprob_std": 0.6510854015574072, "tokenization_path_logprob_std": 31.535231920357308, "trajectory_balance_rmse": 1893.5454273330067} {"canonical_length": 108, "canonical_nll": 1.9519223813657407, "generation_rouge_l_f1": 0.20634920634920634, "generation_token_f1": 0.3386243386243386, "prompt_id": "cef16e435e5a8a53", "reading_canonical_log_prob": -210.8076171875, "reading_distances": [20, 21, 18, 23], "reading_log_probs": [-216.9193115234375, -215.27967834472656, -213.64610290527344, -218.04605102539062], "reading_logprob_std": 1.9225149221017444, "tokenization_path_logprob_std": 105.78849769961559, "trajectory_balance_rmse": 2617.3968242554242} {"canonical_length": 38, "canonical_nll": 2.943380456221731, "generation_rouge_l_f1": 0.20253164556962025, "generation_token_f1": 0.27848101265822783, "prompt_id": "289239d155fcb98c", "reading_canonical_log_prob": -111.84845733642578, "reading_distances": [28, 17, 22, 19], "reading_log_probs": [-110.41360473632812, -111.54879760742188, -113.3329086303711, -110.9876480102539], "reading_logprob_std": 1.2628906045780806, "tokenization_path_logprob_std": 76.39328604738881, "trajectory_balance_rmse": 781.8727727456595} {"canonical_length": 8, "canonical_nll": 2.0330278873443604, "generation_rouge_l_f1": 0.6, "generation_token_f1": 0.6, "prompt_id": "c036d970930b8362", "reading_canonical_log_prob": -16.264223098754883, "reading_distances": [24, 36, 31, 26], "reading_log_probs": [-14.589824676513672, -13.510730743408203, -12.71368408203125, -12.394881248474121], "reading_logprob_std": 0.9782706945363993, "tokenization_path_logprob_std": 27.444849746314055, "trajectory_balance_rmse": 109.72429650027142} {"canonical_length": 99, "canonical_nll": 1.1013194382792773, "generation_rouge_l_f1": 0.2175732217573222, "generation_token_f1": 0.401673640167364, "prompt_id": "1d5889fbec769861", "reading_canonical_log_prob": -109.03062438964844, "reading_distances": [22, 16, 23, 24], "reading_log_probs": [-104.96400451660156, -108.31349182128906, -110.67413330078125, -107.7227783203125], "reading_logprob_std": 2.3464074612744796, "tokenization_path_logprob_std": 149.57580360973648, "trajectory_balance_rmse": 1787.6543046152879} {"canonical_length": 185, "canonical_nll": 0.975217747043919, "generation_rouge_l_f1": 0.18352941176470589, "generation_token_f1": 0.3011764705882353, "prompt_id": "c20bcd26ca03079c", "reading_canonical_log_prob": -180.415283203125, "reading_distances": [38, 36, 38, 37], "reading_log_probs": [-169.75238037109375, -169.89215087890625, -171.6974334716797, -169.8297119140625], "reading_logprob_std": 0.9380864924311163, "tokenization_path_logprob_std": 141.40034480872046, "trajectory_balance_rmse": 3264.025196873204} {"canonical_length": 16, "canonical_nll": 0.7735990881919861, "generation_rouge_l_f1": 0.10280373831775699, "generation_token_f1": 0.12149532710280375, "prompt_id": "93426790937338d2", "reading_canonical_log_prob": -12.377585411071777, "reading_distances": [15, 17, 15, 16], "reading_log_probs": [-13.00399112701416, -10.855549812316895, -11.100297927856445, -11.336638450622559], "reading_logprob_std": 0.9732727258822852, "tokenization_path_logprob_std": 32.749613132925354, "trajectory_balance_rmse": 220.97234441496877} {"canonical_length": 203, "canonical_nll": 1.517657049771013, "generation_rouge_l_f1": 0.19206680584551145, "generation_token_f1": 0.44258872651356995, "prompt_id": "f380c029d22e1217", "reading_canonical_log_prob": -308.0843811035156, "reading_distances": [19, 12, 12, 11], "reading_log_probs": [-301.75689697265625, -303.41876220703125, -303.464111328125, -303.2075500488281], "reading_logprob_std": 0.8110313428356218, "tokenization_path_logprob_std": 146.9930077487924, "trajectory_balance_rmse": 3969.437665790531} {"canonical_length": 191, "canonical_nll": 0.7728597431282722, "generation_rouge_l_f1": 0.3954022988505747, "generation_token_f1": 0.6528735632183907, "prompt_id": "b8396f86fe7c1b16", "reading_canonical_log_prob": -147.6162109375, "reading_distances": [41, 36, 35, 34], "reading_log_probs": [-149.17735290527344, -148.2930908203125, -151.40367126464844, -148.1031494140625], "reading_logprob_std": 1.5137579875211238, "tokenization_path_logprob_std": 179.6376286161198, "trajectory_balance_rmse": 2789.7967781623774} {"canonical_length": 149, "canonical_nll": 1.8192708060245386, "generation_rouge_l_f1": 0.20388349514563106, "generation_token_f1": 0.41262135922330095, "prompt_id": "6cb0fdf8b6463b4b", "reading_canonical_log_prob": -271.07135009765625, "reading_distances": [45, 51, 38, 49], "reading_log_probs": [-265.4512939453125, -266.2560119628906, -266.4205017089844, -267.08050537109375], "reading_logprob_std": 0.6698081651743689, "tokenization_path_logprob_std": 91.55013930292932, "trajectory_balance_rmse": 3256.76509427645} {"canonical_length": 118, "canonical_nll": 1.50177454544326, "generation_rouge_l_f1": 0.28758169934640526, "generation_token_f1": 0.32679738562091504, "prompt_id": "7c85ca0add25adb8", "reading_canonical_log_prob": -177.2093963623047, "reading_distances": [33, 33, 38, 35], "reading_log_probs": [-170.6507568359375, -170.41290283203125, -170.70672607421875, -173.36166381835938], "reading_logprob_std": 1.3916100276165313, "tokenization_path_logprob_std": 175.62409700979458, "trajectory_balance_rmse": 2683.075159758913} {"canonical_length": 114, "canonical_nll": 1.4117926882024396, "generation_rouge_l_f1": 0.33027522935779813, "generation_token_f1": 0.4587155963302752, "prompt_id": "9a1872631342b4a7", "reading_canonical_log_prob": -160.94436645507812, "reading_distances": [30, 29, 23, 28], "reading_log_probs": [-160.3418426513672, -160.28326416015625, -157.8402557373047, -159.68902587890625], "reading_logprob_std": 1.1700035111112452, "tokenization_path_logprob_std": 157.68319290340386, "trajectory_balance_rmse": 2547.8120366515227} {"canonical_length": 13, "canonical_nll": 1.9756587101862981, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.6666666666666666, "prompt_id": "d49306af1b19f211", "reading_canonical_log_prob": -25.683563232421875, "reading_distances": [22, 22, 20, 22], "reading_log_probs": [-24.700483322143555, -24.381574630737305, -23.956689834594727, -25.474931716918945], "reading_logprob_std": 0.641334353963463, "tokenization_path_logprob_std": 83.30248081969219, "trajectory_balance_rmse": 240.3880226591189} {"canonical_length": 253, "canonical_nll": 1.624481201171875, "generation_rouge_l_f1": 0.13829787234042554, "generation_token_f1": 0.3191489361702128, "prompt_id": "807f35218626237b", "reading_canonical_log_prob": -410.9937438964844, "reading_distances": [17, 15, 21, 21], "reading_log_probs": [-409.21466064453125, -411.3133544921875, -411.3798828125, -411.4564208984375], "reading_logprob_std": 1.0858538894746679, "tokenization_path_logprob_std": 263.3764839548101, "trajectory_balance_rmse": 5274.350130264755} {"canonical_length": 4, "canonical_nll": 2.486392021179199, "generation_rouge_l_f1": 0.022641509433962263, "generation_token_f1": 0.022641509433962263, "prompt_id": "1a44d44c0ed30f8a", "reading_canonical_log_prob": -9.945568084716797, "reading_distances": [20, 20, 28, 22], "reading_log_probs": [-9.950069427490234, -10.334871292114258, -9.67657470703125, -8.600663185119629], "reading_logprob_std": 0.7439862677589019, "tokenization_path_logprob_std": 39.848872556802576, "trajectory_balance_rmse": 138.61531031468218} {"canonical_length": 119, "canonical_nll": 0.23275202262301406, "generation_rouge_l_f1": 0.3981264637002342, "generation_token_f1": 0.5292740046838408, "prompt_id": "c78f29d81d80f8fc", "reading_canonical_log_prob": -27.697490692138672, "reading_distances": [21, 27, 20, 22], "reading_log_probs": [-31.82027244567871, -29.223878860473633, -30.275598526000977, -29.05815315246582], "reading_logprob_std": 1.2705750882985798, "tokenization_path_logprob_std": 83.88887649431481, "trajectory_balance_rmse": 1603.3818234716948} {"canonical_length": 279, "canonical_nll": 1.0154269091971886, "generation_rouge_l_f1": 0.22182468694096602, "generation_token_f1": 0.4722719141323793, "prompt_id": "dbd0e34044b57777", "reading_canonical_log_prob": -283.3041076660156, "reading_distances": [25, 18, 19, 23], "reading_log_probs": [-275.12640380859375, -275.6009521484375, -286.85418701171875, -284.09320068359375], "reading_logprob_std": 5.948012040162272, "tokenization_path_logprob_std": 264.7715721528402, "trajectory_balance_rmse": 4493.3054818763885} {"canonical_length": 11, "canonical_nll": 0.12487856908278032, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "1c94de058e5cb066", "reading_canonical_log_prob": -1.3736642599105835, "reading_distances": [20, 22, 23, 12], "reading_log_probs": [-1.3759856224060059, -1.465625286102295, -1.4315369129180908, -1.2607733011245728], "reading_logprob_std": 0.0897597061643957, "tokenization_path_logprob_std": 41.12591194703127, "trajectory_balance_rmse": 208.14168382662461} {"canonical_length": 40, "canonical_nll": 0.6010672569274902, "generation_rouge_l_f1": 0.10830324909747294, "generation_token_f1": 0.18050541516245486, "prompt_id": "91069c7b26de61af", "reading_canonical_log_prob": -24.04269027709961, "reading_distances": [45, 43, 46, 36], "reading_log_probs": [-22.726085662841797, -22.942481994628906, -23.11967658996582, -22.056880950927734], "reading_logprob_std": 0.46500876697406107, "tokenization_path_logprob_std": 71.71929045008142, "trajectory_balance_rmse": 663.7878835071841} {"canonical_length": 10, "canonical_nll": 0.26801066398620604, "generation_rouge_l_f1": 0.875, "generation_token_f1": 1.0, "prompt_id": "c9ec0273bafaec5d", "reading_canonical_log_prob": -2.6801066398620605, "reading_distances": [41, 38, 42, 49], "reading_log_probs": [-5.288188457489014, -5.718630313873291, -7.465456962585449, -7.799215316772461], "reading_logprob_std": 1.2490885395954663, "tokenization_path_logprob_std": 62.986345174409394, "trajectory_balance_rmse": 181.05062851810214} {"canonical_length": 12, "canonical_nll": 0.09349671999613444, "generation_rouge_l_f1": 0.9565217391304348, "generation_token_f1": 0.9565217391304348, "prompt_id": "1cb768d05dd2b1b9", "reading_canonical_log_prob": -1.1219606399536133, "reading_distances": [41, 43, 37, 40], "reading_log_probs": [-1.1951185464859009, -1.2426759004592896, -1.1433043479919434, -1.1371415853500366], "reading_logprob_std": 0.04946203750841063, "tokenization_path_logprob_std": 31.04644720757625, "trajectory_balance_rmse": 169.4015473064146} {"canonical_length": 117, "canonical_nll": 0.8515400030674078, "generation_rouge_l_f1": 0.3581267217630854, "generation_token_f1": 0.38016528925619836, "prompt_id": "31716783a47d0f33", "reading_canonical_log_prob": -99.63018035888672, "reading_distances": [29, 30, 29, 26], "reading_log_probs": [-97.73245239257812, -97.49195098876953, -97.75056457519531, -97.09241485595703], "reading_logprob_std": 0.30652463007579855, "tokenization_path_logprob_std": 119.84569371158763, "trajectory_balance_rmse": 2251.0691512173617} {"canonical_length": 38, "canonical_nll": 0.4564416283055356, "generation_rouge_l_f1": 0.28415300546448086, "generation_token_f1": 0.36065573770491804, "prompt_id": "64444cea846a2353", "reading_canonical_log_prob": -17.34478187561035, "reading_distances": [621, 595, 616, 591], "reading_log_probs": [-29.552404403686523, -23.717748641967773, -26.362152099609375, -29.285991668701172], "reading_logprob_std": 2.7513471131448894, "tokenization_path_logprob_std": 34.98714706161509, "trajectory_balance_rmse": 891.59595251274} {"canonical_length": 17, "canonical_nll": 1.3758519116569967, "generation_rouge_l_f1": 0.45161290322580644, "generation_token_f1": 0.45161290322580644, "prompt_id": "bcb6a50b95a69177", "reading_canonical_log_prob": -23.389482498168945, "reading_distances": [31, 30, 30, 36], "reading_log_probs": [-24.388002395629883, -25.9091796875, -25.05304718017578, -25.09035301208496], "reading_logprob_std": 0.622788957482841, "tokenization_path_logprob_std": 52.05379412360195, "trajectory_balance_rmse": 413.6133086684212} {"canonical_length": 47, "canonical_nll": 0.6667694741107048, "generation_rouge_l_f1": 0.5378151260504201, "generation_token_f1": 0.6218487394957983, "prompt_id": "5a3dfb14391da9f8", "reading_canonical_log_prob": -31.338165283203125, "reading_distances": [25, 28, 24, 23], "reading_log_probs": [-30.708545684814453, -29.969039916992188, -30.159658432006836, -31.3502254486084], "reading_logprob_std": 0.6205748429367481, "tokenization_path_logprob_std": 86.21430498910617, "trajectory_balance_rmse": 715.2455631869376} {"canonical_length": 255, "canonical_nll": 0.16803424311619178, "generation_rouge_l_f1": 0.9053356282271945, "generation_token_f1": 0.9569707401032702, "prompt_id": "1b64247b2a4cbf46", "reading_canonical_log_prob": -42.848731994628906, "reading_distances": [40, 50, 41, 46], "reading_log_probs": [-46.12364959716797, -45.571533203125, -44.17359924316406, -44.8639030456543], "reading_logprob_std": 0.8478388902418762, "tokenization_path_logprob_std": 79.79382244889135, "trajectory_balance_rmse": 2982.413379853797} {"canonical_length": 8, "canonical_nll": 0.15970128774642944, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "ffe7a66de245047f", "reading_canonical_log_prob": -1.2776103019714355, "reading_distances": [23, 21, 28, 26], "reading_log_probs": [-1.5187122821807861, -1.160536527633667, -1.2081072330474854, -1.2976090908050537], "reading_logprob_std": 0.1588275876170526, "tokenization_path_logprob_std": 25.886408540162385, "trajectory_balance_rmse": 158.2760443528706} {"canonical_length": 195, "canonical_nll": 1.3138509114583334, "generation_rouge_l_f1": 0.2352941176470588, "generation_token_f1": 0.4611764705882353, "prompt_id": "9b0fbce06c332eb0", "reading_canonical_log_prob": -256.200927734375, "reading_distances": [22, 24, 21, 17], "reading_log_probs": [-255.46905517578125, -254.7978515625, -255.98167419433594, -256.2442321777344], "reading_logprob_std": 0.6374808190213648, "tokenization_path_logprob_std": 313.9356087763986, "trajectory_balance_rmse": 4479.677172931548} {"canonical_length": 103, "canonical_nll": 1.4317296592934619, "generation_rouge_l_f1": 0.30845771144278605, "generation_token_f1": 0.5174129353233831, "prompt_id": "f63c128779f1525d", "reading_canonical_log_prob": -147.46815490722656, "reading_distances": [33, 28, 31, 28], "reading_log_probs": [-143.71710205078125, -142.16062927246094, -141.22251892089844, -140.92691040039062], "reading_logprob_std": 1.2556611681987044, "tokenization_path_logprob_std": 105.05738128503336, "trajectory_balance_rmse": 2199.5859473327932} {"canonical_length": 45, "canonical_nll": 0.7941862318250869, "generation_rouge_l_f1": 0.3, "generation_token_f1": 0.43333333333333335, "prompt_id": "e0f5d30e662ee161", "reading_canonical_log_prob": -35.738380432128906, "reading_distances": [36, 27, 29, 30], "reading_log_probs": [-36.69671630859375, -40.162410736083984, -40.19831085205078, -40.50007247924805], "reading_logprob_std": 1.8014828318899498, "tokenization_path_logprob_std": 27.80202311113053, "trajectory_balance_rmse": 1008.6583657069592} {"canonical_length": 10, "canonical_nll": 1.4896364212036133, "generation_rouge_l_f1": 0.5714285714285714, "generation_token_f1": 0.5714285714285714, "prompt_id": "10787d0a825f897d", "reading_canonical_log_prob": -14.896364212036133, "reading_distances": [35, 37, 43, 32], "reading_log_probs": [-17.45283317565918, -19.705642700195312, -19.06102180480957, -16.424270629882812], "reading_logprob_std": 1.4959600555166288, "tokenization_path_logprob_std": 50.92589196237534, "trajectory_balance_rmse": 216.45898703764556} {"canonical_length": 138, "canonical_nll": 0.9226226806640625, "generation_rouge_l_f1": 0.2056555269922879, "generation_token_f1": 0.40102827763496146, "prompt_id": "55c070ce04e00179", "reading_canonical_log_prob": -127.32192993164062, "reading_distances": [20, 14, 18, 20], "reading_log_probs": [-126.48007202148438, -125.74244689941406, -133.1714630126953, -127.069091796875], "reading_logprob_std": 3.4138803687396697, "tokenization_path_logprob_std": 143.4104327510021, "trajectory_balance_rmse": 2820.952849375482} {"canonical_length": 57, "canonical_nll": 1.8799636573122258, "generation_rouge_l_f1": 0.3619047619047619, "generation_token_f1": 0.49523809523809526, "prompt_id": "9718a6939e1171c6", "reading_canonical_log_prob": -107.15792846679688, "reading_distances": [19, 17, 25, 27], "reading_log_probs": [-107.38554382324219, -104.74191284179688, -104.7408447265625, -107.378173828125], "reading_logprob_std": 1.5244848942544016, "tokenization_path_logprob_std": 92.06744439632942, "trajectory_balance_rmse": 1327.647343124576} {"canonical_length": 180, "canonical_nll": 1.4352810329861112, "generation_rouge_l_f1": 0.18426966292134833, "generation_token_f1": 0.40898876404494383, "prompt_id": "abf26ab3207cd57f", "reading_canonical_log_prob": -258.3505859375, "reading_distances": [23, 22, 24, 26], "reading_log_probs": [-259.6297607421875, -259.5810852050781, -262.0283508300781, -262.6881103515625], "reading_logprob_std": 1.612118248795314, "tokenization_path_logprob_std": 283.9589784932638, "trajectory_balance_rmse": 2938.6234470002964} {"canonical_length": 8, "canonical_nll": 0.07098463177680969, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "9706792c2f395e2c", "reading_canonical_log_prob": -0.5678770542144775, "reading_distances": [26, 27, 19, 24], "reading_log_probs": [-0.5698226094245911, -0.5909004211425781, -0.5662716627120972, -0.5415862798690796], "reading_logprob_std": 0.020211387206261637, "tokenization_path_logprob_std": 39.76812135190276, "trajectory_balance_rmse": 152.52679178186654} {"canonical_length": 198, "canonical_nll": 1.3169959482520517, "generation_rouge_l_f1": 0.24628450106157113, "generation_token_f1": 0.4670912951167729, "prompt_id": "4d13949d8c9ff709", "reading_canonical_log_prob": -260.76519775390625, "reading_distances": [31, 26, 23, 26], "reading_log_probs": [-263.34661865234375, -258.4197998046875, -258.36083984375, -257.13519287109375], "reading_logprob_std": 2.751806168595347, "tokenization_path_logprob_std": 280.3863699498667, "trajectory_balance_rmse": 4614.307512415376} {"canonical_length": 259, "canonical_nll": 1.428127215175555, "generation_rouge_l_f1": 0.3281519861830743, "generation_token_f1": 0.6183074265975821, "prompt_id": "cf2b1f1029bc9513", "reading_canonical_log_prob": -369.88494873046875, "reading_distances": [29, 30, 35, 24], "reading_log_probs": [-368.2051086425781, -367.477294921875, -365.84112548828125, -366.86273193359375], "reading_logprob_std": 1.000771460997405, "tokenization_path_logprob_std": 56.43214030706871, "trajectory_balance_rmse": 4167.0867175985495} {"canonical_length": 159, "canonical_nll": 1.6865554905537539, "generation_rouge_l_f1": 0.19338422391857504, "generation_token_f1": 0.38167938931297707, "prompt_id": "0c1e711645beb310", "reading_canonical_log_prob": -268.1623229980469, "reading_distances": [20, 22, 27, 25], "reading_log_probs": [-268.2364196777344, -270.8128356933594, -268.5130310058594, -269.79248046875], "reading_logprob_std": 1.1938375032808408, "tokenization_path_logprob_std": 68.79695605367749, "trajectory_balance_rmse": 2800.5927997106965} {"canonical_length": 36, "canonical_nll": 0.9833842383490669, "generation_rouge_l_f1": 0.2692307692307692, "generation_token_f1": 0.4230769230769231, "prompt_id": "602cd10603b86b5c", "reading_canonical_log_prob": -35.401832580566406, "reading_distances": [11, 13, 14, 14], "reading_log_probs": [-36.090187072753906, -35.553993225097656, -35.656517028808594, -34.36236572265625], "reading_logprob_std": 0.7397215132090337, "tokenization_path_logprob_std": 70.25365630995026, "trajectory_balance_rmse": 662.0964535034738} {"canonical_length": 8, "canonical_nll": 3.061927080154419, "generation_rouge_l_f1": 0.3333333333333333, "generation_token_f1": 0.3333333333333333, "prompt_id": "b36657ba534be993", "reading_canonical_log_prob": -24.49541664123535, "reading_distances": [31, 35, 29, 34], "reading_log_probs": [-21.23352813720703, -20.968637466430664, -22.57891845703125, -22.89031982421875], "reading_logprob_std": 0.9577769242148729, "tokenization_path_logprob_std": 9.102665591334862, "trajectory_balance_rmse": 136.1484430841556} {"canonical_length": 33, "canonical_nll": 1.784502665201823, "generation_rouge_l_f1": 0.32558139534883723, "generation_token_f1": 0.37209302325581395, "prompt_id": "f11461f699664b03", "reading_canonical_log_prob": -58.888587951660156, "reading_distances": [21, 20, 20, 18], "reading_log_probs": [-59.37810516357422, -61.264400482177734, -59.89767837524414, -59.117549896240234], "reading_logprob_std": 0.9566195951665843, "tokenization_path_logprob_std": 74.51331465912943, "trajectory_balance_rmse": 741.9471982532449} {"canonical_length": 323, "canonical_nll": 1.2925435825029026, "generation_rouge_l_f1": 0.2576271186440678, "generation_token_f1": 0.45762711864406774, "prompt_id": "167bb9079aab9f14", "reading_canonical_log_prob": -417.4915771484375, "reading_distances": [19, 26, 28, 23], "reading_log_probs": [-413.855224609375, -412.756591796875, -415.8049011230469, -416.8543701171875], "reading_logprob_std": 1.852666286324689, "tokenization_path_logprob_std": 188.84586171466594, "trajectory_balance_rmse": 7530.897490237462} {"canonical_length": 330, "canonical_nll": 1.130511474609375, "generation_rouge_l_f1": 0.19999999999999998, "generation_token_f1": 0.44561403508771924, "prompt_id": "e95dc10521ada79e", "reading_canonical_log_prob": -373.06878662109375, "reading_distances": [39, 43, 31, 31], "reading_log_probs": [-364.37762451171875, -371.0466003417969, -367.58587646484375, -366.818359375], "reading_logprob_std": 2.7563414731419527, "tokenization_path_logprob_std": 225.12812576280885, "trajectory_balance_rmse": 5799.024917928362} {"canonical_length": 126, "canonical_nll": 1.3139989943731398, "generation_rouge_l_f1": 0.17045454545454544, "generation_token_f1": 0.3068181818181818, "prompt_id": "78584ee870db48f1", "reading_canonical_log_prob": -165.56387329101562, "reading_distances": [32, 35, 34, 38], "reading_log_probs": [-166.06939697265625, -167.18624877929688, -166.5220947265625, -165.5773162841797], "reading_logprob_std": 0.6841544235501323, "tokenization_path_logprob_std": 132.78198336482677, "trajectory_balance_rmse": 2605.8694573969406} {"canonical_length": 318, "canonical_nll": 1.335528679613797, "generation_rouge_l_f1": 0.23287671232876714, "generation_token_f1": 0.4726027397260274, "prompt_id": "16eff2d1d4178803", "reading_canonical_log_prob": -424.6981201171875, "reading_distances": [17, 25, 25, 22], "reading_log_probs": [-424.3406982421875, -425.48114013671875, -424.985107421875, -426.12481689453125], "reading_logprob_std": 0.7559903094679897, "tokenization_path_logprob_std": 193.19916207022902, "trajectory_balance_rmse": 7052.222930627279} {"canonical_length": 104, "canonical_nll": 1.133683131291316, "generation_rouge_l_f1": 0.40677966101694923, "generation_token_f1": 0.5084745762711864, "prompt_id": "22619d053543f675", "reading_canonical_log_prob": -117.90304565429688, "reading_distances": [43, 39, 43, 38], "reading_log_probs": [-112.30435943603516, -113.26463317871094, -115.09410095214844, -112.30821228027344], "reading_logprob_std": 1.3142699872648922, "tokenization_path_logprob_std": 107.51370990558995, "trajectory_balance_rmse": 1991.211435623263} {"canonical_length": 180, "canonical_nll": 1.0437676323784721, "generation_rouge_l_f1": 0.255952380952381, "generation_token_f1": 0.5416666666666667, "prompt_id": "3c84043f06521168", "reading_canonical_log_prob": -187.878173828125, "reading_distances": [20, 28, 25, 24], "reading_log_probs": [-185.10916137695312, -184.8044891357422, -186.13803100585938, -185.27293395996094], "reading_logprob_std": 0.5718673154901126, "tokenization_path_logprob_std": 38.997327274863004, "trajectory_balance_rmse": 3531.7864796611098} {"canonical_length": 238, "canonical_nll": 1.5930173216747636, "generation_rouge_l_f1": 0.22352941176470587, "generation_token_f1": 0.48627450980392156, "prompt_id": "4c550a924c0f2e5b", "reading_canonical_log_prob": -379.13812255859375, "reading_distances": [31, 22, 30, 26], "reading_log_probs": [-370.3701171875, -370.38946533203125, -376.6054992675781, -369.66790771484375], "reading_logprob_std": 3.2488890667183887, "tokenization_path_logprob_std": 156.01165066596292, "trajectory_balance_rmse": 4373.869142867493} {"canonical_length": 72, "canonical_nll": 1.327966160244412, "generation_rouge_l_f1": 0.2772277227722772, "generation_token_f1": 0.396039603960396, "prompt_id": "9e6e21f46355104c", "reading_canonical_log_prob": -95.61356353759766, "reading_distances": [19, 13, 20, 21], "reading_log_probs": [-98.85467529296875, -96.7885513305664, -95.32999420166016, -96.7420654296875], "reading_logprob_std": 1.4514033567597355, "tokenization_path_logprob_std": 72.59613497222631, "trajectory_balance_rmse": 1254.2915846171943} {"canonical_length": 27, "canonical_nll": 1.4182621991192852, "generation_rouge_l_f1": 0.35294117647058826, "generation_token_f1": 0.5490196078431373, "prompt_id": "4a624dc7ffd3988d", "reading_canonical_log_prob": -38.2930793762207, "reading_distances": [30, 26, 25, 36], "reading_log_probs": [-41.34730529785156, -39.537864685058594, -39.67870330810547, -40.95723342895508], "reading_logprob_std": 0.9073563436859373, "tokenization_path_logprob_std": 84.51645705326635, "trajectory_balance_rmse": 682.0105186211131} {"canonical_length": 7, "canonical_nll": 0.09634110757282802, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "0cd953773dcd678b", "reading_canonical_log_prob": -0.6743877530097961, "reading_distances": [27, 29, 36, 31], "reading_log_probs": [-0.7625654935836792, -0.8093612194061279, -0.839173436164856, -0.7670555710792542], "reading_logprob_std": 0.036467222989302904, "tokenization_path_logprob_std": 19.12210142536464, "trajectory_balance_rmse": 165.33010395996965} {"canonical_length": 165, "canonical_nll": 0.6048128070253315, "generation_rouge_l_f1": 0.2601156069364162, "generation_token_f1": 0.5664739884393064, "prompt_id": "a86300073620913b", "reading_canonical_log_prob": -99.79411315917969, "reading_distances": [135, 135, 154, 137], "reading_log_probs": [-92.70115661621094, -95.18380737304688, -96.87760925292969, -90.74345397949219], "reading_logprob_std": 2.7026603407195053, "tokenization_path_logprob_std": 74.5733337424428, "trajectory_balance_rmse": 2700.9717740281712} {"canonical_length": 294, "canonical_nll": 1.4277520212186436, "generation_rouge_l_f1": 0.1843478260869565, "generation_token_f1": 0.3060869565217392, "prompt_id": "2837da3fee02c3ea", "reading_canonical_log_prob": -419.75909423828125, "reading_distances": [32, 24, 25, 27], "reading_log_probs": [-415.614990234375, -413.6763916015625, -414.9804992675781, -414.7646789550781], "reading_logprob_std": 0.8070221647854112, "tokenization_path_logprob_std": 197.6500260279087, "trajectory_balance_rmse": 5652.6990556066885} {"canonical_length": 23, "canonical_nll": 1.240005410235861, "generation_rouge_l_f1": 0.1240875912408759, "generation_token_f1": 0.1386861313868613, "prompt_id": "8041588a81157d9c", "reading_canonical_log_prob": -28.520124435424805, "reading_distances": [54, 53, 49, 37], "reading_log_probs": [-33.890716552734375, -29.089101791381836, -29.636884689331055, -28.917675018310547], "reading_logprob_std": 2.3581144963428793, "tokenization_path_logprob_std": 89.17770290449872, "trajectory_balance_rmse": 532.7397510049709} {"canonical_length": 308, "canonical_nll": 0.8999980579723011, "generation_rouge_l_f1": 0.23193916349809887, "generation_token_f1": 0.42965779467680604, "prompt_id": "bc95295408e3badf", "reading_canonical_log_prob": -277.19940185546875, "reading_distances": [30, 32, 28, 32], "reading_log_probs": [-272.1217041015625, -280.09716796875, -277.5162658691406, -277.3736877441406], "reading_logprob_std": 3.3465328008536455, "tokenization_path_logprob_std": 132.77596715118665, "trajectory_balance_rmse": 5066.278222646706} {"canonical_length": 126, "canonical_nll": 1.4132476080031622, "generation_rouge_l_f1": 0.3673469387755102, "generation_token_f1": 0.5247813411078718, "prompt_id": "2a5a99cace11afef", "reading_canonical_log_prob": -178.06919860839844, "reading_distances": [27, 28, 22, 22], "reading_log_probs": [-181.8336181640625, -176.75979614257812, -180.074951171875, -178.41421508789062], "reading_logprob_std": 2.179722768074959, "tokenization_path_logprob_std": 184.42573666943196, "trajectory_balance_rmse": 2966.9173934819783} {"canonical_length": 14, "canonical_nll": 1.8985582079206194, "generation_rouge_l_f1": 0.10232558139534884, "generation_token_f1": 0.11162790697674418, "prompt_id": "90db691d0627dd6d", "reading_canonical_log_prob": -26.579814910888672, "reading_distances": [16, 21, 16, 22], "reading_log_probs": [-23.60295295715332, -24.36617088317871, -25.658309936523438, -24.93228530883789], "reading_logprob_std": 0.8704088623845667, "tokenization_path_logprob_std": 23.920489278445924, "trajectory_balance_rmse": 313.299562022671} {"canonical_length": 14, "canonical_nll": 1.0218473161969865, "generation_rouge_l_f1": 0.4799999999999999, "generation_token_f1": 0.4799999999999999, "prompt_id": "07c0c6eb902aa919", "reading_canonical_log_prob": -14.305862426757812, "reading_distances": [26, 26, 23, 27], "reading_log_probs": [-14.627385139465332, -13.885225296020508, -15.338825225830078, -14.055880546569824], "reading_logprob_std": 0.6564798010305939, "tokenization_path_logprob_std": 44.03583615692099, "trajectory_balance_rmse": 283.41495775366366} {"canonical_length": 6, "canonical_nll": 0.6983706951141357, "generation_rouge_l_f1": 0.6, "generation_token_f1": 0.6, "prompt_id": "285daef948799d4d", "reading_canonical_log_prob": -4.1902241706848145, "reading_distances": [43, 43, 52, 50], "reading_log_probs": [-3.8869173526763916, -4.082125663757324, -4.60286283493042, -4.549809455871582], "reading_logprob_std": 0.35152289525136876, "tokenization_path_logprob_std": 29.550179146505055, "trajectory_balance_rmse": 131.8583867931698} {"canonical_length": 68, "canonical_nll": 0.6734592213350183, "generation_rouge_l_f1": 0.22966507177033496, "generation_token_f1": 0.3253588516746411, "prompt_id": "924060b32d38a1e4", "reading_canonical_log_prob": -45.79522705078125, "reading_distances": [34, 30, 33, 36], "reading_log_probs": [-46.99643325805664, -46.4252815246582, -46.66866683959961, -46.368350982666016], "reading_logprob_std": 0.2858903613215016, "tokenization_path_logprob_std": 73.78069665034123, "trajectory_balance_rmse": 1154.5089374728486} {"canonical_length": 147, "canonical_nll": 1.0240452565303466, "generation_rouge_l_f1": 0.2395543175487465, "generation_token_f1": 0.4623955431754875, "prompt_id": "ac4b8108300a4ae5", "reading_canonical_log_prob": -150.53465270996094, "reading_distances": [32, 32, 30, 29], "reading_log_probs": [-149.57797241210938, -147.24151611328125, -149.20606994628906, -148.29446411132812], "reading_logprob_std": 1.0425901034723832, "tokenization_path_logprob_std": 73.54889000683896, "trajectory_balance_rmse": 2223.7696171762313} {"canonical_length": 53, "canonical_nll": 1.6368603976267688, "generation_rouge_l_f1": 0.3489932885906041, "generation_token_f1": 0.4697986577181208, "prompt_id": "10e2361fe555104f", "reading_canonical_log_prob": -86.75360107421875, "reading_distances": [25, 24, 18, 28], "reading_log_probs": [-92.65791320800781, -93.01486206054688, -87.69377136230469, -88.07640075683594], "reading_logprob_std": 2.866606364800283, "tokenization_path_logprob_std": 75.74603671060504, "trajectory_balance_rmse": 989.3404923946562} {"canonical_length": 27, "canonical_nll": 1.4048425179940682, "generation_rouge_l_f1": 0.12182741116751268, "generation_token_f1": 0.18274111675126906, "prompt_id": "e53f2608a52ca6d9", "reading_canonical_log_prob": -37.930747985839844, "reading_distances": [24, 33, 32, 36], "reading_log_probs": [-40.18574523925781, -41.88410568237305, -39.48980712890625, -40.30674743652344], "reading_logprob_std": 1.011249872064846, "tokenization_path_logprob_std": 45.765639113861496, "trajectory_balance_rmse": 409.1229428040057} {"canonical_length": 18, "canonical_nll": 1.47910647922092, "generation_rouge_l_f1": 0.37037037037037035, "generation_token_f1": 0.44444444444444453, "prompt_id": "3f4c8b5b47d9a555", "reading_canonical_log_prob": -26.623916625976562, "reading_distances": [17, 19, 15, 16], "reading_log_probs": [-24.824121475219727, -26.18099594116211, -26.073657989501953, -24.398332595825195], "reading_logprob_std": 0.893489003851364, "tokenization_path_logprob_std": 52.50366259299851, "trajectory_balance_rmse": 387.27017465269853} {"canonical_length": 11, "canonical_nll": 0.6725495945323597, "generation_rouge_l_f1": 0.6153846153846153, "generation_token_f1": 0.7692307692307692, "prompt_id": "94a2d875be1c7a06", "reading_canonical_log_prob": -7.398045539855957, "reading_distances": [17, 16, 13, 17], "reading_log_probs": [-8.153398513793945, -7.798654556274414, -8.599254608154297, -7.979004383087158], "reading_logprob_std": 0.34317647184030986, "tokenization_path_logprob_std": 33.81362168406366, "trajectory_balance_rmse": 210.94949372778538} {"canonical_length": 18, "canonical_nll": 0.5394313600328233, "generation_rouge_l_f1": 0.22222222222222224, "generation_token_f1": 0.22222222222222224, "prompt_id": "a86c032862f0a2ee", "reading_canonical_log_prob": -9.70976448059082, "reading_distances": [17, 17, 15, 14], "reading_log_probs": [-9.04365062713623, -8.6689453125, -9.145967483520508, -8.669099807739258], "reading_logprob_std": 0.2493515062361515, "tokenization_path_logprob_std": 42.309176554808516, "trajectory_balance_rmse": 281.2794502043024} {"canonical_length": 80, "canonical_nll": 1.4587530136108398, "generation_rouge_l_f1": 0.16568047337278105, "generation_token_f1": 0.2781065088757396, "prompt_id": "26b8508f97bd840b", "reading_canonical_log_prob": -116.70024108886719, "reading_distances": [33, 29, 38, 28], "reading_log_probs": [-113.36448669433594, -116.07937622070312, -115.7657699584961, -115.05300903320312], "reading_logprob_std": 1.2127001818209933, "tokenization_path_logprob_std": 80.4253250570213, "trajectory_balance_rmse": 1859.2527689529386} {"canonical_length": 282, "canonical_nll": 0.8935213562444593, "generation_rouge_l_f1": 0.3053435114503817, "generation_token_f1": 0.4541984732824427, "prompt_id": "d2afb79833a3e8c4", "reading_canonical_log_prob": -251.9730224609375, "reading_distances": [21, 26, 22, 22], "reading_log_probs": [-254.20420837402344, -256.05596923828125, -253.6258087158203, -253.30149841308594], "reading_logprob_std": 1.230732051830046, "tokenization_path_logprob_std": 101.09645784172163, "trajectory_balance_rmse": 5737.315850484224} {"canonical_length": 206, "canonical_nll": 1.0425413557626668, "generation_rouge_l_f1": 0.2764578833693304, "generation_token_f1": 0.5745140388768899, "prompt_id": "8d6230df8e48bdfc", "reading_canonical_log_prob": -214.76351928710938, "reading_distances": [14, 14, 12, 17], "reading_log_probs": [-205.1332244873047, -205.85260009765625, -209.02896118164062, -205.98855590820312], "reading_logprob_std": 1.7266942270554335, "tokenization_path_logprob_std": 67.50469005091338, "trajectory_balance_rmse": 3726.4159884480036} {"canonical_length": 14, "canonical_nll": 2.7941256931849887, "generation_rouge_l_f1": 0.32, "generation_token_f1": 0.32, "prompt_id": "74ebf94d0f403056", "reading_canonical_log_prob": -39.117759704589844, "reading_distances": [22, 24, 30, 26], "reading_log_probs": [-42.5113639831543, -43.80811309814453, -44.60009002685547, -42.886714935302734], "reading_logprob_std": 0.9397286543019672, "tokenization_path_logprob_std": 13.467583258067862, "trajectory_balance_rmse": 265.15164234353387} {"canonical_length": 129, "canonical_nll": 1.2016913835392442, "generation_rouge_l_f1": 0.3005780346820809, "generation_token_f1": 0.5433526011560693, "prompt_id": "c307d4b735027f71", "reading_canonical_log_prob": -155.0181884765625, "reading_distances": [18, 17, 16, 15], "reading_log_probs": [-152.57810974121094, -152.63751220703125, -153.22047424316406, -152.2149658203125], "reading_logprob_std": 0.41608008909815397, "tokenization_path_logprob_std": 183.12841703791776, "trajectory_balance_rmse": 2433.8508336686423} {"canonical_length": 73, "canonical_nll": 1.066115287885274, "generation_rouge_l_f1": 0.31884057971014496, "generation_token_f1": 0.47342995169082125, "prompt_id": "1044800b3405ca7b", "reading_canonical_log_prob": -77.826416015625, "reading_distances": [12, 13, 13, 12], "reading_log_probs": [-83.81057739257812, -84.65620422363281, -81.874267578125, -84.86851501464844], "reading_logprob_std": 1.3642268118445637, "tokenization_path_logprob_std": 88.800172686042, "trajectory_balance_rmse": 1250.8034376591092} {"canonical_length": 205, "canonical_nll": 1.380770204125381, "generation_rouge_l_f1": 0.2026431718061674, "generation_token_f1": 0.4625550660792952, "prompt_id": "c87b3b9596ed8821", "reading_canonical_log_prob": -283.0578918457031, "reading_distances": [30, 29, 32, 27], "reading_log_probs": [-280.916015625, -283.33746337890625, -280.98773193359375, -282.8067626953125], "reading_logprob_std": 1.2434907495449916, "tokenization_path_logprob_std": 91.15113863957922, "trajectory_balance_rmse": 4461.598632145616} {"canonical_length": 11, "canonical_nll": 1.2742887843738904, "generation_rouge_l_f1": 0.18181818181818182, "generation_token_f1": 0.18181818181818182, "prompt_id": "f2f04e5d3550740c", "reading_canonical_log_prob": -14.017176628112793, "reading_distances": [19, 11, 13, 17], "reading_log_probs": [-17.537288665771484, -14.285562515258789, -15.799567222595215, -16.28167152404785], "reading_logprob_std": 1.3440938448411317, "tokenization_path_logprob_std": 55.67723356243766, "trajectory_balance_rmse": 209.571731551592} {"canonical_length": 149, "canonical_nll": 1.3991772952495805, "generation_rouge_l_f1": 0.24664879356568364, "generation_token_f1": 0.35388739946380704, "prompt_id": "e64143ec52b6e7c2", "reading_canonical_log_prob": -208.4774169921875, "reading_distances": [20, 18, 16, 15], "reading_log_probs": [-212.15704345703125, -209.33375549316406, -209.0986328125, -211.71368408203125], "reading_logprob_std": 1.5832254534716672, "tokenization_path_logprob_std": 53.61515173627115, "trajectory_balance_rmse": 2686.733615753578} {"canonical_length": 48, "canonical_nll": 1.121137301127116, "generation_rouge_l_f1": 0.22972972972972971, "generation_token_f1": 0.28378378378378377, "prompt_id": "c95d33d4bb11096d", "reading_canonical_log_prob": -53.81459045410156, "reading_distances": [18, 26, 25, 24], "reading_log_probs": [-57.28855514526367, -57.29355239868164, -56.657447814941406, -57.30449676513672], "reading_logprob_std": 0.3191129286440895, "tokenization_path_logprob_std": 139.4199704225181, "trajectory_balance_rmse": 1202.7487104399108} {"canonical_length": 211, "canonical_nll": 1.362888137311167, "generation_rouge_l_f1": 0.22767857142857142, "generation_token_f1": 0.44642857142857145, "prompt_id": "0accd9d43b5400d9", "reading_canonical_log_prob": -287.56939697265625, "reading_distances": [23, 20, 20, 30], "reading_log_probs": [-282.91033935546875, -284.5279846191406, -283.4444885253906, -284.5555419921875], "reading_logprob_std": 0.8174118836678317, "tokenization_path_logprob_std": 103.98621373852059, "trajectory_balance_rmse": 4266.97034055904} {"canonical_length": 28, "canonical_nll": 0.8813458851405552, "generation_rouge_l_f1": 0.5142857142857142, "generation_token_f1": 0.5142857142857142, "prompt_id": "c5287ae056861247", "reading_canonical_log_prob": -24.677684783935547, "reading_distances": [19, 16, 23, 19], "reading_log_probs": [-21.47394561767578, -21.073732376098633, -21.874210357666016, -21.990245819091797], "reading_logprob_std": 0.4164507291049993, "tokenization_path_logprob_std": 25.73273664675861, "trajectory_balance_rmse": 605.7314611311989} {"canonical_length": 9, "canonical_nll": 0.7886150148179796, "generation_rouge_l_f1": 0.08641975308641975, "generation_token_f1": 0.08641975308641975, "prompt_id": "9ba45d22569ee03c", "reading_canonical_log_prob": -7.097535133361816, "reading_distances": [41, 44, 43, 46], "reading_log_probs": [-10.62436294555664, -9.490984916687012, -9.196732521057129, -9.556042671203613], "reading_logprob_std": 0.6247601780970136, "tokenization_path_logprob_std": 35.94174209400178, "trajectory_balance_rmse": 219.94887475308482} {"canonical_length": 9, "canonical_nll": 1.4876430299546983, "generation_rouge_l_f1": 0.36363636363636365, "generation_token_f1": 0.36363636363636365, "prompt_id": "61752b6441afd01d", "reading_canonical_log_prob": -13.388787269592285, "reading_distances": [20, 20, 16, 23], "reading_log_probs": [-12.844592094421387, -12.734367370605469, -12.847146987915039, -12.288049697875977], "reading_logprob_std": 0.2655817424657543, "tokenization_path_logprob_std": 41.407568763264315, "trajectory_balance_rmse": 250.9960620790524} {"canonical_length": 147, "canonical_nll": 1.0715997397494155, "generation_rouge_l_f1": 0.32891246684350134, "generation_token_f1": 0.4774535809018568, "prompt_id": "1c1ecaad77fec061", "reading_canonical_log_prob": -157.52516174316406, "reading_distances": [14, 14, 17, 16], "reading_log_probs": [-154.98988342285156, -158.25485229492188, -158.28492736816406, -156.37017822265625], "reading_logprob_std": 1.5979600336684538, "tokenization_path_logprob_std": 125.36114900431622, "trajectory_balance_rmse": 2767.0864575215805} {"canonical_length": 58, "canonical_nll": 1.102270060572131, "generation_rouge_l_f1": 0.4375, "generation_token_f1": 0.5937499999999999, "prompt_id": "697abdbcde745481", "reading_canonical_log_prob": -63.931663513183594, "reading_distances": [21, 25, 17, 22], "reading_log_probs": [-60.145294189453125, -59.27898406982422, -62.519012451171875, -60.68840408325195], "reading_logprob_std": 1.369774179959909, "tokenization_path_logprob_std": 56.66390789370639, "trajectory_balance_rmse": 990.2697860306134} {"canonical_length": 338, "canonical_nll": 1.229379326634153, "generation_rouge_l_f1": 0.15656565656565655, "generation_token_f1": 0.24747474747474746, "prompt_id": "4034832cbc1502c1", "reading_canonical_log_prob": -415.53021240234375, "reading_distances": [29, 20, 22, 17], "reading_log_probs": [-407.9009704589844, -413.30908203125, -409.40191650390625, -413.77899169921875], "reading_logprob_std": 2.8967969139496743, "tokenization_path_logprob_std": 155.64850018461135, "trajectory_balance_rmse": 6956.75803614307} {"canonical_length": 13, "canonical_nll": 1.6523865919846754, "generation_rouge_l_f1": 0.6666666666666667, "generation_token_f1": 0.7500000000000001, "prompt_id": "228b1bd7f7bc8ed3", "reading_canonical_log_prob": -21.48102569580078, "reading_distances": [14, 16, 14, 15], "reading_log_probs": [-17.78649139404297, -18.35053253173828, -19.164358139038086, -18.38513946533203], "reading_logprob_std": 0.5661073356437554, "tokenization_path_logprob_std": 28.65154115421669, "trajectory_balance_rmse": 217.64218049260117} {"canonical_length": 115, "canonical_nll": 1.1822954260784646, "generation_rouge_l_f1": 0.3228346456692914, "generation_token_f1": 0.5433070866141733, "prompt_id": "28e1d60f144a6d02", "reading_canonical_log_prob": -135.96397399902344, "reading_distances": [20, 24, 27, 23], "reading_log_probs": [-130.93115234375, -132.66415405273438, -133.34442138671875, -132.3117218017578], "reading_logprob_std": 1.0159796173919684, "tokenization_path_logprob_std": 56.17558673144751, "trajectory_balance_rmse": 2160.301220037207} {"canonical_length": 131, "canonical_nll": 1.3838555721836237, "generation_rouge_l_f1": 0.2571428571428572, "generation_token_f1": 0.44, "prompt_id": "f38548affd15660d", "reading_canonical_log_prob": -181.2850799560547, "reading_distances": [27, 26, 30, 26], "reading_log_probs": [-177.60594177246094, -176.046142578125, -177.7438201904297, -176.0027618408203], "reading_logprob_std": 0.9547008789162712, "tokenization_path_logprob_std": 120.17469640645453, "trajectory_balance_rmse": 2571.8659200946186} {"canonical_length": 39, "canonical_nll": 1.246071546505659, "generation_rouge_l_f1": 0.27272727272727276, "generation_token_f1": 0.33333333333333337, "prompt_id": "b15afd8e011cefe5", "reading_canonical_log_prob": -48.5967903137207, "reading_distances": [20, 19, 30, 25], "reading_log_probs": [-50.19990539550781, -50.11723327636719, -51.235260009765625, -50.15813446044922], "reading_logprob_std": 0.5394746518610241, "tokenization_path_logprob_std": 88.33545663977897, "trajectory_balance_rmse": 752.2017860499291} {"canonical_length": 286, "canonical_nll": 0.9309881250341455, "generation_rouge_l_f1": 0.11063829787234045, "generation_token_f1": 0.25106382978723407, "prompt_id": "60d1476fd22fd625", "reading_canonical_log_prob": -266.2626037597656, "reading_distances": [17, 24, 20, 22], "reading_log_probs": [-259.6629333496094, -261.9156188964844, -256.53155517578125, -258.9363098144531], "reading_logprob_std": 2.2183960577699113, "tokenization_path_logprob_std": 130.5445589999653, "trajectory_balance_rmse": 4665.829032657508} {"canonical_length": 58, "canonical_nll": 1.2247460463951374, "generation_rouge_l_f1": 0.3787878787878788, "generation_token_f1": 0.40909090909090906, "prompt_id": "f1aff688c2307612", "reading_canonical_log_prob": -71.03527069091797, "reading_distances": [35, 51, 43, 44], "reading_log_probs": [-68.90673828125, -69.16342163085938, -67.61141204833984, -69.31268310546875], "reading_logprob_std": 0.7764171738291116, "tokenization_path_logprob_std": 92.08940756590326, "trajectory_balance_rmse": 1246.6934827285502} {"canonical_length": 123, "canonical_nll": 1.4533058724752286, "generation_rouge_l_f1": 0.19662921348314608, "generation_token_f1": 0.34831460674157305, "prompt_id": "a3f93481aca6425b", "reading_canonical_log_prob": -178.75662231445312, "reading_distances": [24, 22, 23, 28], "reading_log_probs": [-176.36070251464844, -176.08853149414062, -176.50189208984375, -178.30540466308594], "reading_logprob_std": 1.0088743412750445, "tokenization_path_logprob_std": 136.21766179754167, "trajectory_balance_rmse": 2283.7545028970053} {"canonical_length": 163, "canonical_nll": 1.2403787250167753, "generation_rouge_l_f1": 0.2119815668202765, "generation_token_f1": 0.3963133640552996, "prompt_id": "39b8bde022351616", "reading_canonical_log_prob": -202.18173217773438, "reading_distances": [17, 22, 26, 16], "reading_log_probs": [-194.98013305664062, -196.73182678222656, -197.39303588867188, -196.88638305664062], "reading_logprob_std": 1.050480138801373, "tokenization_path_logprob_std": 133.398620222943, "trajectory_balance_rmse": 3344.0769283309187} {"canonical_length": 56, "canonical_nll": 1.5738751547677177, "generation_rouge_l_f1": 0.15547703180212016, "generation_token_f1": 0.21201413427561838, "prompt_id": "969cc5c017329874", "reading_canonical_log_prob": -88.13700866699219, "reading_distances": [32, 31, 33, 33], "reading_log_probs": [-89.54141235351562, -90.90090942382812, -86.19064331054688, -88.44744110107422], "reading_logprob_std": 1.9910614526906207, "tokenization_path_logprob_std": 97.54304810303412, "trajectory_balance_rmse": 847.4665791691287} {"canonical_length": 262, "canonical_nll": 1.6860393495050094, "generation_rouge_l_f1": 0.2934362934362934, "generation_token_f1": 0.5444015444015443, "prompt_id": "a4cec9116b07e32b", "reading_canonical_log_prob": -441.7423095703125, "reading_distances": [12, 20, 13, 16], "reading_log_probs": [-430.9896545410156, -428.712646484375, -431.9361877441406, -433.4580078125], "reading_logprob_std": 1.9874411390824667, "tokenization_path_logprob_std": 130.70447955086127, "trajectory_balance_rmse": 4864.672911113288} {"canonical_length": 412, "canonical_nll": 0.9169941133665807, "generation_rouge_l_f1": 0.2660944206008583, "generation_token_f1": 0.5608011444921316, "prompt_id": "63a1199c25b071e9", "reading_canonical_log_prob": -377.80157470703125, "reading_distances": [209, 221, 217, 245], "reading_log_probs": [-376.35076904296875, -374.70343017578125, -372.7243957519531, -375.19873046875], "reading_logprob_std": 1.5131571131050243, "tokenization_path_logprob_std": 156.0469201560474, "trajectory_balance_rmse": 6929.615327367712} {"canonical_length": 34, "canonical_nll": 1.472651762120864, "generation_rouge_l_f1": 0.08227848101265821, "generation_token_f1": 0.13924050632911392, "prompt_id": "85280303ae668b99", "reading_canonical_log_prob": -50.070159912109375, "reading_distances": [10, 14, 10, 15], "reading_log_probs": [-46.761268615722656, -47.972381591796875, -47.3139762878418, -49.10013198852539], "reading_logprob_std": 1.0057432151132024, "tokenization_path_logprob_std": 47.43200247153164, "trajectory_balance_rmse": 642.9501458725679} {"canonical_length": 18, "canonical_nll": 1.3730822669135199, "generation_rouge_l_f1": 0.23529411764705882, "generation_token_f1": 0.35294117647058826, "prompt_id": "963484b861b43113", "reading_canonical_log_prob": -24.71548080444336, "reading_distances": [15, 15, 17, 12], "reading_log_probs": [-24.20232391357422, -24.366580963134766, -23.9224796295166, -24.818082809448242], "reading_logprob_std": 0.37501524270661646, "tokenization_path_logprob_std": 23.26048066882543, "trajectory_balance_rmse": 344.5230119026916} {"canonical_length": 242, "canonical_nll": 1.2033333266076962, "generation_rouge_l_f1": 0.20606060606060606, "generation_token_f1": 0.5212121212121211, "prompt_id": "edfd3e88db49ba9c", "reading_canonical_log_prob": -291.2066650390625, "reading_distances": [24, 28, 23, 31], "reading_log_probs": [-294.80712890625, -295.7913818359375, -294.1402893066406, -301.32684326171875], "reading_logprob_std": 3.2778817047494266, "tokenization_path_logprob_std": 117.03174506041562, "trajectory_balance_rmse": 5280.450510591034} {"canonical_length": 3, "canonical_nll": 4.055494626363118, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.6666666666666666, "prompt_id": "315bed08881fe880", "reading_canonical_log_prob": -12.166483879089355, "reading_distances": [31, 32, 31, 29], "reading_log_probs": [-13.735618591308594, -13.82612419128418, -12.51414966583252, -13.53546142578125], "reading_logprob_std": 0.6047782100445342, "tokenization_path_logprob_std": 24.419234196251082, "trajectory_balance_rmse": 87.9326231914647} {"canonical_length": 40, "canonical_nll": 1.3079820632934571, "generation_rouge_l_f1": 0.1884057971014493, "generation_token_f1": 0.30434782608695654, "prompt_id": "ec3686b4d6045b07", "reading_canonical_log_prob": -52.31928253173828, "reading_distances": [20, 30, 28, 25], "reading_log_probs": [-51.78125, -52.282440185546875, -52.580142974853516, -52.3427734375], "reading_logprob_std": 0.3358250421891637, "tokenization_path_logprob_std": 111.4684821894367, "trajectory_balance_rmse": 1019.0737803472474} {"canonical_length": 266, "canonical_nll": 1.5329867484874296, "generation_rouge_l_f1": 0.23921568627450981, "generation_token_f1": 0.4980392156862745, "prompt_id": "a1b8dba044d5aaa9", "reading_canonical_log_prob": -407.77447509765625, "reading_distances": [26, 27, 19, 23], "reading_log_probs": [-405.4901123046875, -407.166015625, -408.2994384765625, -408.0697021484375], "reading_logprob_std": 1.2750596297997672, "tokenization_path_logprob_std": 85.26522514289583, "trajectory_balance_rmse": 5933.509945182138} {"canonical_length": 154, "canonical_nll": 1.4463563398881392, "generation_rouge_l_f1": 0.3667481662591687, "generation_token_f1": 0.5183374083129584, "prompt_id": "e45ec8d7efaf0407", "reading_canonical_log_prob": -222.73887634277344, "reading_distances": [22, 15, 16, 15], "reading_log_probs": [-228.77145385742188, -231.014892578125, -230.3079071044922, -226.6528778076172], "reading_logprob_std": 1.9315000826615676, "tokenization_path_logprob_std": 105.64506158764382, "trajectory_balance_rmse": 3648.3447891024134} {"canonical_length": 3, "canonical_nll": 2.738016128540039, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.6666666666666666, "prompt_id": "3521244aca15fc91", "reading_canonical_log_prob": -8.214048385620117, "reading_distances": [49, 49, 46, 48], "reading_log_probs": [-6.41398286819458, -6.647892951965332, -6.437809467315674, -6.551921367645264], "reading_logprob_std": 0.10827272858477227, "tokenization_path_logprob_std": 16.885972535740297, "trajectory_balance_rmse": 67.75985377759513} {"canonical_length": 23, "canonical_nll": 0.6448122107464335, "generation_rouge_l_f1": 0.14569536423841062, "generation_token_f1": 0.18543046357615892, "prompt_id": "99de541f6c007900", "reading_canonical_log_prob": -14.830680847167969, "reading_distances": [26, 23, 26, 26], "reading_log_probs": [-15.22268295288086, -14.712902069091797, -14.800811767578125, -16.451749801635742], "reading_logprob_std": 0.8013204264178051, "tokenization_path_logprob_std": 71.14904344982119, "trajectory_balance_rmse": 276.2974881071805} {"canonical_length": 176, "canonical_nll": 1.336721333590421, "generation_rouge_l_f1": 0.27415730337078653, "generation_token_f1": 0.4898876404494383, "prompt_id": "124ab50eaad8157c", "reading_canonical_log_prob": -235.26295471191406, "reading_distances": [18, 19, 23, 25], "reading_log_probs": [-234.211181640625, -230.96609497070312, -233.67349243164062, -235.1411590576172], "reading_logprob_std": 1.7935012694729768, "tokenization_path_logprob_std": 220.28891920702492, "trajectory_balance_rmse": 3906.7448136207477} {"canonical_length": 15, "canonical_nll": 0.20555707613627117, "generation_rouge_l_f1": 0.8888888888888888, "generation_token_f1": 0.8888888888888888, "prompt_id": "7efa869db692612b", "reading_canonical_log_prob": -3.0833561420440674, "reading_distances": [48, 52, 54, 51], "reading_log_probs": [-8.897991180419922, -7.9401350021362305, -11.678871154785156, -10.206399917602539], "reading_logprob_std": 1.6239096738059189, "tokenization_path_logprob_std": 34.16731775072323, "trajectory_balance_rmse": 208.39605422641247} {"canonical_length": 248, "canonical_nll": 1.4020870577904485, "generation_rouge_l_f1": 0.29959514170040485, "generation_token_f1": 0.5829959514170041, "prompt_id": "b9153e23659ffc43", "reading_canonical_log_prob": -347.71759033203125, "reading_distances": [30, 27, 29, 36], "reading_log_probs": [-343.87640380859375, -344.95428466796875, -344.36822509765625, -343.8279724121094], "reading_logprob_std": 0.5251967099289268, "tokenization_path_logprob_std": 208.83442506811775, "trajectory_balance_rmse": 5556.567900929203} {"canonical_length": 48, "canonical_nll": 1.1113770008087158, "generation_rouge_l_f1": 0.2, "generation_token_f1": 0.2066666666666667, "prompt_id": "2f3de773aa866413", "reading_canonical_log_prob": -53.34609603881836, "reading_distances": [23, 33, 25, 29], "reading_log_probs": [-46.50278854370117, -52.1449089050293, -49.851341247558594, -51.722408294677734], "reading_logprob_std": 2.5695429066108093, "tokenization_path_logprob_std": 73.55675571777294, "trajectory_balance_rmse": 795.1555180544996} {"canonical_length": 301, "canonical_nll": 1.3598631798626972, "generation_rouge_l_f1": 0.17314487632508835, "generation_token_f1": 0.2968197879858657, "prompt_id": "54066b815b8806cf", "reading_canonical_log_prob": -409.3188171386719, "reading_distances": [32, 37, 45, 40], "reading_log_probs": [-409.563720703125, -409.1851806640625, -411.04833984375, -408.4515075683594], "reading_logprob_std": 1.0930683425473904, "tokenization_path_logprob_std": 285.39788707878375, "trajectory_balance_rmse": 5792.807530704702} {"canonical_length": 394, "canonical_nll": 1.322364458577887, "generation_rouge_l_f1": 0.20763358778625954, "generation_token_f1": 0.4122137404580153, "prompt_id": "38de2822dc803e85", "reading_canonical_log_prob": -521.0115966796875, "reading_distances": [26, 31, 38, 36], "reading_log_probs": [-512.82421875, -519.5489501953125, -514.5335693359375, -516.3168334960938], "reading_logprob_std": 2.874063685050355, "tokenization_path_logprob_std": 246.5615040574066, "trajectory_balance_rmse": 7439.571575248368} {"canonical_length": 329, "canonical_nll": 0.9220273096148366, "generation_rouge_l_f1": 0.21640488656195464, "generation_token_f1": 0.48865619546247824, "prompt_id": "93d5783f4b9e5595", "reading_canonical_log_prob": -303.34698486328125, "reading_distances": [32, 31, 30, 32], "reading_log_probs": [-298.94854736328125, -297.6834411621094, -297.90386962890625, -301.0001220703125], "reading_logprob_std": 1.5148306886744929, "tokenization_path_logprob_std": 211.49528101052195, "trajectory_balance_rmse": 7144.993488540171} {"canonical_length": 25, "canonical_nll": 0.6175448608398437, "generation_rouge_l_f1": 0.20408163265306123, "generation_token_f1": 0.23129251700680273, "prompt_id": "fffceba44eca0753", "reading_canonical_log_prob": -15.438621520996094, "reading_distances": [27, 31, 30, 28], "reading_log_probs": [-15.48336124420166, -15.536999702453613, -15.910709381103516, -15.954368591308594], "reading_logprob_std": 0.24547798191864895, "tokenization_path_logprob_std": 72.03769174144193, "trajectory_balance_rmse": 396.64629318113595} {"canonical_length": 49, "canonical_nll": 1.8894491390306123, "generation_rouge_l_f1": 0.19354838709677416, "generation_token_f1": 0.25806451612903225, "prompt_id": "4d1f490ef0029856", "reading_canonical_log_prob": -92.5830078125, "reading_distances": [21, 21, 25, 25], "reading_log_probs": [-90.63809204101562, -90.891845703125, -91.33415222167969, -89.89390563964844], "reading_logprob_std": 0.6033609850813554, "tokenization_path_logprob_std": 104.14228641338852, "trajectory_balance_rmse": 1179.1885800102816} {"canonical_length": 58, "canonical_nll": 1.4259839551202182, "generation_rouge_l_f1": 0.30344827586206896, "generation_token_f1": 0.46896551724137925, "prompt_id": "04c2c0f3f6e1a037", "reading_canonical_log_prob": -82.70706939697266, "reading_distances": [11, 14, 10, 10], "reading_log_probs": [-93.85697174072266, -99.53230285644531, -94.90422058105469, -103.58805847167969], "reading_logprob_std": 4.484022580609764, "tokenization_path_logprob_std": 69.43917494753651, "trajectory_balance_rmse": 1277.6267663067424} {"canonical_length": 68, "canonical_nll": 0.44102006800034466, "generation_rouge_l_f1": 0.27397260273972607, "generation_token_f1": 0.4246575342465754, "prompt_id": "42b3ed92d904a820", "reading_canonical_log_prob": -29.989364624023438, "reading_distances": [10, 11, 11, 14], "reading_log_probs": [-30.87799644470215, -33.8665771484375, -30.396984100341797, -30.0844669342041], "reading_logprob_std": 1.7376415326847232, "tokenization_path_logprob_std": 40.871498093316525, "trajectory_balance_rmse": 1240.8255699781707} {"canonical_length": 4, "canonical_nll": 0.1489853709936142, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "a0febc02441ecdc9", "reading_canonical_log_prob": -0.5959414839744568, "reading_distances": [48, 47, 48, 41], "reading_log_probs": [-2.812072277069092, -3.1004862785339355, -2.521332263946533, -2.302161693572998], "reading_logprob_std": 0.3474308623208861, "tokenization_path_logprob_std": 28.560269520583876, "trajectory_balance_rmse": 115.6575867211814} {"canonical_length": 84, "canonical_nll": 1.9655551002139138, "generation_rouge_l_f1": 0.18181818181818182, "generation_token_f1": 0.24848484848484848, "prompt_id": "95d007f697f6007d", "reading_canonical_log_prob": -165.10662841796875, "reading_distances": [28, 23, 21, 24], "reading_log_probs": [-168.3111572265625, -171.932861328125, -170.1363525390625, -169.7880401611328], "reading_logprob_std": 1.488239589544394, "tokenization_path_logprob_std": 120.52727739734294, "trajectory_balance_rmse": 1611.7767044602745} {"canonical_length": 178, "canonical_nll": 1.1161389297313904, "generation_rouge_l_f1": 0.27539503386004516, "generation_token_f1": 0.4920993227990971, "prompt_id": "42731119470564d3", "reading_canonical_log_prob": -198.6727294921875, "reading_distances": [57, 55, 51, 60], "reading_log_probs": [-204.76492309570312, -203.9625244140625, -203.13238525390625, -202.4883575439453], "reading_logprob_std": 0.9903217001737922, "tokenization_path_logprob_std": 178.28763156500125, "trajectory_balance_rmse": 2465.462971471184} {"canonical_length": 302, "canonical_nll": 1.2197046343064466, "generation_rouge_l_f1": 0.24957841483979767, "generation_token_f1": 0.4924114671163575, "prompt_id": "f73070b2f6b0ff25", "reading_canonical_log_prob": -368.3507995605469, "reading_distances": [25, 27, 22, 27], "reading_log_probs": [-363.9478759765625, -365.01983642578125, -367.0733642578125, -364.89202880859375], "reading_logprob_std": 1.3165844844745516, "tokenization_path_logprob_std": 121.49774875986847, "trajectory_balance_rmse": 6175.959590709587} {"canonical_length": 143, "canonical_nll": 1.1749006151319383, "generation_rouge_l_f1": 0.2205128205128205, "generation_token_f1": 0.39999999999999997, "prompt_id": "08c2f6d9de2bf434", "reading_canonical_log_prob": -168.0107879638672, "reading_distances": [21, 22, 20, 18], "reading_log_probs": [-171.0604248046875, -167.09982299804688, -167.5380401611328, -169.3083953857422], "reading_logprob_std": 1.811241545362823, "tokenization_path_logprob_std": 183.36781282696043, "trajectory_balance_rmse": 2749.963227916255} {"canonical_length": 118, "canonical_nll": 1.0488811428264035, "generation_rouge_l_f1": 0.26628895184135976, "generation_token_f1": 0.40226628895184136, "prompt_id": "5f825922a6155ea5", "reading_canonical_log_prob": -123.76797485351562, "reading_distances": [37, 29, 34, 36], "reading_log_probs": [-122.4361572265625, -123.21753692626953, -122.35406494140625, -123.52672576904297], "reading_logprob_std": 0.5790040407211727, "tokenization_path_logprob_std": 138.98916155960714, "trajectory_balance_rmse": 2287.237450620009} {"canonical_length": 55, "canonical_nll": 1.1806958285245028, "generation_rouge_l_f1": 0.15032679738562094, "generation_token_f1": 0.24836601307189546, "prompt_id": "3828b93ba3b0437a", "reading_canonical_log_prob": -64.93827056884766, "reading_distances": [22, 30, 26, 26], "reading_log_probs": [-64.97093963623047, -64.57481384277344, -65.36831665039062, -64.41346740722656], "reading_logprob_std": 0.42749423747250787, "tokenization_path_logprob_std": 137.63628164892495, "trajectory_balance_rmse": 1030.4434073404284} {"canonical_length": 109, "canonical_nll": 2.0055589588410263, "generation_rouge_l_f1": 0.22695035460992904, "generation_token_f1": 0.43971631205673756, "prompt_id": "bf57ac1eadcdbbef", "reading_canonical_log_prob": -218.60592651367188, "reading_distances": [36, 45, 38, 37], "reading_log_probs": [-211.57884216308594, -214.021728515625, -213.13388061523438, -213.42298889160156], "reading_logprob_std": 1.0415158065068941, "tokenization_path_logprob_std": 119.35861262040419, "trajectory_balance_rmse": 2407.27622464865} {"canonical_length": 15, "canonical_nll": 2.793748474121094, "generation_rouge_l_f1": 0.1904761904761905, "generation_token_f1": 0.1904761904761905, "prompt_id": "fd15860ce78ef664", "reading_canonical_log_prob": -41.906227111816406, "reading_distances": [34, 33, 22, 31], "reading_log_probs": [-40.614784240722656, -43.100624084472656, -40.27565002441406, -41.32783889770508], "reading_logprob_std": 1.2594048155376973, "tokenization_path_logprob_std": 51.29485038885643, "trajectory_balance_rmse": 283.5053505202796} {"canonical_length": 25, "canonical_nll": 1.9059603881835938, "generation_rouge_l_f1": 0.13043478260869565, "generation_token_f1": 0.13043478260869565, "prompt_id": "b2df929cec29e4f9", "reading_canonical_log_prob": -47.649009704589844, "reading_distances": [31, 39, 30, 29], "reading_log_probs": [-49.73674011230469, -49.53236770629883, -47.32014465332031, -53.78561782836914], "reading_logprob_std": 2.6935342839637992, "tokenization_path_logprob_std": 47.63956441126642, "trajectory_balance_rmse": 587.5597640904136} {"canonical_length": 280, "canonical_nll": 0.8100490025111607, "generation_rouge_l_f1": 0.4047619047619047, "generation_token_f1": 0.6230158730158731, "prompt_id": "190309fcf6be3c92", "reading_canonical_log_prob": -226.813720703125, "reading_distances": [24, 29, 28, 27], "reading_log_probs": [-230.91734313964844, -228.22108459472656, -230.91265869140625, -230.2527313232422], "reading_logprob_std": 1.2753824468985646, "tokenization_path_logprob_std": 201.2895995365053, "trajectory_balance_rmse": 3560.262697293013} {"canonical_length": 20, "canonical_nll": 0.7728681564331055, "generation_rouge_l_f1": 0.22448979591836735, "generation_token_f1": 0.26530612244897955, "prompt_id": "48be81c0aa95d468", "reading_canonical_log_prob": -15.45736312866211, "reading_distances": [14, 23, 18, 23], "reading_log_probs": [-15.230480194091797, -14.509721755981445, -13.757757186889648, -13.959622383117676], "reading_logprob_std": 0.6590583527785717, "tokenization_path_logprob_std": 47.21248603859581, "trajectory_balance_rmse": 499.2423716146166} {"canonical_length": 86, "canonical_nll": 2.141904964003452, "generation_rouge_l_f1": 0.23893805309734512, "generation_token_f1": 0.3805309734513274, "prompt_id": "87fbf3fe9c551bf6", "reading_canonical_log_prob": -184.20382690429688, "reading_distances": [32, 30, 29, 27], "reading_log_probs": [-179.69615173339844, -176.81134033203125, -176.530517578125, -178.64279174804688], "reading_logprob_std": 1.5096280326206188, "tokenization_path_logprob_std": 79.76829843604612, "trajectory_balance_rmse": 1715.7966605228212} {"canonical_length": 122, "canonical_nll": 1.320910469430392, "generation_rouge_l_f1": 0.15342465753424658, "generation_token_f1": 0.2684931506849315, "prompt_id": "31fa13f44ebbb510", "reading_canonical_log_prob": -161.1510772705078, "reading_distances": [22, 22, 21, 22], "reading_log_probs": [-158.14779663085938, -151.7327880859375, -154.9269561767578, -157.90557861328125], "reading_logprob_std": 3.0105794256321814, "tokenization_path_logprob_std": 113.36351992298276, "trajectory_balance_rmse": 2046.6861856836676} {"canonical_length": 113, "canonical_nll": 1.7177804592436394, "generation_rouge_l_f1": 0.24175824175824173, "generation_token_f1": 0.3681318681318681, "prompt_id": "1e48963a206d1f06", "reading_canonical_log_prob": -194.10919189453125, "reading_distances": [13, 13, 15, 13], "reading_log_probs": [-187.16879272460938, -188.59104919433594, -189.27011108398438, -186.93601989746094], "reading_logprob_std": 1.1232677104723683, "tokenization_path_logprob_std": 195.80874009102533, "trajectory_balance_rmse": 2135.6026674698132} {"canonical_length": 7, "canonical_nll": 2.1026954650878906, "generation_rouge_l_f1": 0.7499999999999999, "generation_token_f1": 0.7499999999999999, "prompt_id": "b92fcbf5058fa547", "reading_canonical_log_prob": -14.718868255615234, "reading_distances": [21, 17, 20, 17], "reading_log_probs": [-14.886558532714844, -14.152251243591309, -15.38680648803711, -14.938735008239746], "reading_logprob_std": 0.5111774814838671, "tokenization_path_logprob_std": 27.564820854752686, "trajectory_balance_rmse": 117.04012374173635} {"canonical_length": 14, "canonical_nll": 0.992785998753139, "generation_rouge_l_f1": 0.5454545454545455, "generation_token_f1": 0.6363636363636364, "prompt_id": "5f82c19500742c50", "reading_canonical_log_prob": -13.899003982543945, "reading_distances": [28, 27, 25, 23], "reading_log_probs": [-15.660356521606445, -15.662328720092773, -14.708478927612305, -15.494027137756348], "reading_logprob_std": 0.45542847473005976, "tokenization_path_logprob_std": 49.438392298738854, "trajectory_balance_rmse": 391.96330872472356} {"canonical_length": 107, "canonical_nll": 1.481352048499562, "generation_rouge_l_f1": 0.3113207547169811, "generation_token_f1": 0.5188679245283018, "prompt_id": "cb0faecefbcafa87", "reading_canonical_log_prob": -158.50466918945312, "reading_distances": [26, 27, 26, 28], "reading_log_probs": [-159.68939208984375, -161.1600341796875, -162.81307983398438, -161.072265625], "reading_logprob_std": 1.2781252500613771, "tokenization_path_logprob_std": 136.84694643934233, "trajectory_balance_rmse": 2247.9923647263563} {"canonical_length": 111, "canonical_nll": 1.6764289924690314, "generation_rouge_l_f1": 0.10084033613445378, "generation_token_f1": 0.1176470588235294, "prompt_id": "935eaf327e8c6a5b", "reading_canonical_log_prob": -186.0836181640625, "reading_distances": [36, 33, 34, 33], "reading_log_probs": [-176.91165161132812, -177.77691650390625, -178.42770385742188, -187.30291748046875], "reading_logprob_std": 4.838759652966032, "tokenization_path_logprob_std": 97.28318424558205, "trajectory_balance_rmse": 2041.6261637967068} {"canonical_length": 72, "canonical_nll": 1.2813384797837999, "generation_rouge_l_f1": 0.1577287066246057, "generation_token_f1": 0.24605678233438485, "prompt_id": "116cdf21f0043438", "reading_canonical_log_prob": -92.2563705444336, "reading_distances": [12, 18, 18, 16], "reading_log_probs": [-93.6827392578125, -91.99829864501953, -90.06169128417969, -90.970458984375], "reading_logprob_std": 1.552915378087992, "tokenization_path_logprob_std": 220.17332268118594, "trajectory_balance_rmse": 1567.983226250884} {"canonical_length": 31, "canonical_nll": 1.5927648236674647, "generation_rouge_l_f1": 0.2456140350877193, "generation_token_f1": 0.3157894736842105, "prompt_id": "da79267701ae23c7", "reading_canonical_log_prob": -49.375709533691406, "reading_distances": [33, 35, 40, 31], "reading_log_probs": [-48.84964370727539, -49.93863296508789, -47.06840133666992, -50.10314178466797], "reading_logprob_std": 1.3965708525916847, "tokenization_path_logprob_std": 50.465439238116105, "trajectory_balance_rmse": 864.1282181010926} {"canonical_length": 269, "canonical_nll": 0.7935770027699524, "generation_rouge_l_f1": 0.27516778523489926, "generation_token_f1": 0.40939597315436244, "prompt_id": "3348c8386d3ef99b", "reading_canonical_log_prob": -213.4722137451172, "reading_distances": [15, 17, 11, 12], "reading_log_probs": [-204.26364135742188, -206.95223999023438, -208.15904235839844, -206.9571990966797], "reading_logprob_std": 1.6471892381611772, "tokenization_path_logprob_std": 149.96117191572947, "trajectory_balance_rmse": 4249.737619090359} {"canonical_length": 10, "canonical_nll": 0.3871044397354126, "generation_rouge_l_f1": 0.9, "generation_token_f1": 0.9, "prompt_id": "d04f01367f709567", "reading_canonical_log_prob": -3.871044397354126, "reading_distances": [20, 27, 14, 25], "reading_log_probs": [-2.894686222076416, -3.8610644340515137, -3.2595701217651367, -3.515517234802246], "reading_logprob_std": 0.4081631061320462, "tokenization_path_logprob_std": 43.50262062261517, "trajectory_balance_rmse": 157.867209374551} {"canonical_length": 281, "canonical_nll": 0.8902563997859209, "generation_rouge_l_f1": 0.2447418738049713, "generation_token_f1": 0.3671128107074569, "prompt_id": "f04ba139f8e40a51", "reading_canonical_log_prob": -250.16204833984375, "reading_distances": [14, 12, 15, 9], "reading_log_probs": [-250.1905517578125, -258.30755615234375, -250.9016571044922, -251.19906616210938], "reading_logprob_std": 3.7955561214659754, "tokenization_path_logprob_std": 197.0843085813957, "trajectory_balance_rmse": 3867.1326304157606} {"canonical_length": 501, "canonical_nll": 0.766096164604385, "generation_rouge_l_f1": 0.3945578231292517, "generation_token_f1": 0.5333333333333333, "prompt_id": "16085c364de7d67f", "reading_canonical_log_prob": -383.8141784667969, "reading_distances": [14, 13, 17, 15], "reading_log_probs": [-379.27984619140625, -381.0476379394531, -380.0523681640625, -380.04461669921875], "reading_logprob_std": 0.7247657904278315, "tokenization_path_logprob_std": 114.57205847992095, "trajectory_balance_rmse": 6616.09395416882} {"canonical_length": 405, "canonical_nll": 1.1258619520399304, "generation_rouge_l_f1": 0.18547140649149924, "generation_token_f1": 0.3585780525502319, "prompt_id": "122921a791ee9365", "reading_canonical_log_prob": -455.9740905761719, "reading_distances": [24, 22, 23, 21], "reading_log_probs": [-451.9836120605469, -452.0415344238281, -452.2821960449219, -451.89300537109375], "reading_logprob_std": 0.16637438806184052, "tokenization_path_logprob_std": 207.28060281592025, "trajectory_balance_rmse": 6940.34121539844} {"canonical_length": 67, "canonical_nll": 1.630200172538188, "generation_rouge_l_f1": 0.2696629213483146, "generation_token_f1": 0.29213483146067415, "prompt_id": "5090b423053cf048", "reading_canonical_log_prob": -109.2234115600586, "reading_distances": [33, 38, 36, 33], "reading_log_probs": [-108.19210815429688, -105.61567687988281, -107.49545288085938, -106.41193389892578], "reading_logprob_std": 1.1414151851870196, "tokenization_path_logprob_std": 120.43288238814405, "trajectory_balance_rmse": 1435.8709828067933} {"canonical_length": 349, "canonical_nll": 1.069049037288458, "generation_rouge_l_f1": 0.26031746031746034, "generation_token_f1": 0.45079365079365086, "prompt_id": "4811a788bad2e4f1", "reading_canonical_log_prob": -373.0981140136719, "reading_distances": [31, 24, 29, 22], "reading_log_probs": [-369.60791015625, -371.3446960449219, -371.30401611328125, -371.43572998046875], "reading_logprob_std": 0.8785127834130927, "tokenization_path_logprob_std": 126.9998655643903, "trajectory_balance_rmse": 7310.105474596944} {"canonical_length": 14, "canonical_nll": 1.2583633150373186, "generation_rouge_l_f1": 0.05263157894736842, "generation_token_f1": 0.05263157894736842, "prompt_id": "b2fd6c1ae0104b61", "reading_canonical_log_prob": -17.61708641052246, "reading_distances": [24, 18, 27, 20], "reading_log_probs": [-19.228734970092773, -18.13253402709961, -18.663244247436523, -18.83831024169922], "reading_logprob_std": 0.4549990145394426, "tokenization_path_logprob_std": 51.227472459987794, "trajectory_balance_rmse": 306.4568308820318} {"canonical_length": 12, "canonical_nll": 0.34230470657348633, "generation_rouge_l_f1": 0.0782122905027933, "generation_token_f1": 0.08938547486033518, "prompt_id": "3b02f9d0f5b11977", "reading_canonical_log_prob": -4.107656478881836, "reading_distances": [26, 20, 18, 19], "reading_log_probs": [-4.396998405456543, -3.849215269088745, -4.328787803649902, -4.0758843421936035], "reading_logprob_std": 0.25052646861779, "tokenization_path_logprob_std": 18.575907945246897, "trajectory_balance_rmse": 206.23145721139218} {"canonical_length": 6, "canonical_nll": 0.7027630011240641, "generation_rouge_l_f1": 0.8000000000000002, "generation_token_f1": 0.8000000000000002, "prompt_id": "842d7e43c22f686e", "reading_canonical_log_prob": -4.216578006744385, "reading_distances": [23, 27, 22, 26], "reading_log_probs": [-7.602456092834473, -8.019532203674316, -8.753995895385742, -6.1053972244262695], "reading_logprob_std": 1.1165275861498878, "tokenization_path_logprob_std": 28.0984397544601, "trajectory_balance_rmse": 157.07194767315193} {"canonical_length": 74, "canonical_nll": 1.7383892987225507, "generation_rouge_l_f1": 0.43209876543209885, "generation_token_f1": 0.654320987654321, "prompt_id": "f726d0133e785ebc", "reading_canonical_log_prob": -128.64080810546875, "reading_distances": [22, 22, 20, 22], "reading_log_probs": [-130.27687072753906, -129.00558471679688, -129.1986083984375, -128.69590759277344], "reading_logprob_std": 0.687030346721131, "tokenization_path_logprob_std": 100.83039801812095, "trajectory_balance_rmse": 1251.3528225135228} {"canonical_length": 48, "canonical_nll": 1.2328759829203289, "generation_rouge_l_f1": 0.32608695652173914, "generation_token_f1": 0.40217391304347827, "prompt_id": "cb06f10a99b5f8fb", "reading_canonical_log_prob": -59.17804718017578, "reading_distances": [21, 19, 24, 17], "reading_log_probs": [-59.59224319458008, -59.64935302734375, -60.69331741333008, -61.16217041015625], "reading_logprob_std": 0.7788131267074131, "tokenization_path_logprob_std": 19.685907025800223, "trajectory_balance_rmse": 979.7972454143191} {"canonical_length": 171, "canonical_nll": 0.8608889217265168, "generation_rouge_l_f1": 0.2606060606060606, "generation_token_f1": 0.42424242424242425, "prompt_id": "443ce871037b41c0", "reading_canonical_log_prob": -147.21200561523438, "reading_distances": [30, 36, 30, 28], "reading_log_probs": [-150.18463134765625, -149.607421875, -148.81869506835938, -150.51138305664062], "reading_logprob_std": 0.7421897664370147, "tokenization_path_logprob_std": 26.484134311670115, "trajectory_balance_rmse": 2259.835413238323} {"canonical_length": 307, "canonical_nll": 1.3557285967401262, "generation_rouge_l_f1": 0.29441624365482233, "generation_token_f1": 0.5279187817258884, "prompt_id": "d1c3cdcce82ccde0", "reading_canonical_log_prob": -416.20867919921875, "reading_distances": [44, 41, 34, 36], "reading_log_probs": [-413.4886474609375, -413.496826171875, -414.86944580078125, -414.4812927246094], "reading_logprob_std": 0.7009480137506149, "tokenization_path_logprob_std": 224.1637242305485, "trajectory_balance_rmse": 6437.137199690176} {"canonical_length": 61, "canonical_nll": 0.5209815853931865, "generation_rouge_l_f1": 0.4654088050314465, "generation_token_f1": 0.5660377358490566, "prompt_id": "66bf9508c6f9f94b", "reading_canonical_log_prob": -31.779876708984375, "reading_distances": [41, 46, 45, 47], "reading_log_probs": [-36.147621154785156, -36.216758728027344, -36.999324798583984, -36.78030776977539], "reading_logprob_std": 0.4191695251557505, "tokenization_path_logprob_std": 142.30533300487323, "trajectory_balance_rmse": 1218.7486370428862} {"canonical_length": 265, "canonical_nll": 0.860369873046875, "generation_rouge_l_f1": 0.22488038277511962, "generation_token_f1": 0.5119617224880382, "prompt_id": "c9e81615c39cdb45", "reading_canonical_log_prob": -227.99801635742188, "reading_distances": [19, 27, 23, 24], "reading_log_probs": [-224.01548767089844, -224.23622131347656, -225.26914978027344, -222.074462890625], "reading_logprob_std": 1.3333583615025304, "tokenization_path_logprob_std": 290.16014024488607, "trajectory_balance_rmse": 5370.702615367763} {"canonical_length": 41, "canonical_nll": 1.5657047178687118, "generation_rouge_l_f1": 0.06535947712418301, "generation_token_f1": 0.09803921568627451, "prompt_id": "768e651293911174", "reading_canonical_log_prob": -64.19389343261719, "reading_distances": [55, 54, 52, 61], "reading_log_probs": [-66.08355712890625, -65.12137603759766, -66.26961517333984, -66.91387939453125], "reading_logprob_std": 0.7414199719310248, "tokenization_path_logprob_std": 88.89402384808821, "trajectory_balance_rmse": 797.6387419165312} {"canonical_length": 367, "canonical_nll": 1.1249415426228286, "generation_rouge_l_f1": 0.17257318952234205, "generation_token_f1": 0.3020030816640986, "prompt_id": "bd6b470fc22dea74", "reading_canonical_log_prob": -412.8535461425781, "reading_distances": [48, 35, 46, 46], "reading_log_probs": [-412.4340515136719, -410.7909851074219, -414.28558349609375, -414.560546875], "reading_logprob_std": 1.7594326523168087, "tokenization_path_logprob_std": 245.6139921730388, "trajectory_balance_rmse": 7271.004561991211} {"canonical_length": 303, "canonical_nll": 1.2070785874974217, "generation_rouge_l_f1": 0.1903114186851211, "generation_token_f1": 0.301038062283737, "prompt_id": "ca8e3b0a9ea7294b", "reading_canonical_log_prob": -365.74481201171875, "reading_distances": [24, 17, 25, 21], "reading_log_probs": [-366.7465515136719, -364.810546875, -363.87237548828125, -365.8936767578125], "reading_logprob_std": 1.2541727635902513, "tokenization_path_logprob_std": 110.67619091224657, "trajectory_balance_rmse": 6297.641870657221} {"canonical_length": 246, "canonical_nll": 1.510790817136687, "generation_rouge_l_f1": 0.28625235404896426, "generation_token_f1": 0.5310734463276836, "prompt_id": "5cd1f02fb9a1877d", "reading_canonical_log_prob": -371.654541015625, "reading_distances": [25, 17, 30, 29], "reading_log_probs": [-370.2637939453125, -367.2607421875, -371.2112121582031, -366.1097106933594], "reading_logprob_std": 2.4174492838545882, "tokenization_path_logprob_std": 191.93669573897864, "trajectory_balance_rmse": 4823.668959838601} {"canonical_length": 210, "canonical_nll": 1.0009264264787947, "generation_rouge_l_f1": 0.2832244008714597, "generation_token_f1": 0.48366013071895425, "prompt_id": "754987e5b5772585", "reading_canonical_log_prob": -210.19454956054688, "reading_distances": [27, 24, 27, 33], "reading_log_probs": [-206.093505859375, -206.83274841308594, -206.19754028320312, -206.3235626220703], "reading_logprob_std": 0.32772829149504695, "tokenization_path_logprob_std": 242.71854449687024, "trajectory_balance_rmse": 4641.820234690279} {"canonical_length": 221, "canonical_nll": 1.1850137149586397, "generation_rouge_l_f1": 0.26834381551362685, "generation_token_f1": 0.4863731656184486, "prompt_id": "5aac80ddf699bdd3", "reading_canonical_log_prob": -261.8880310058594, "reading_distances": [20, 20, 20, 20], "reading_log_probs": [-261.00048828125, -258.48272705078125, -262.3948974609375, -260.731689453125], "reading_logprob_std": 1.6197973501480354, "tokenization_path_logprob_std": 211.85273882349446, "trajectory_balance_rmse": 3876.0624684727236} {"canonical_length": 25, "canonical_nll": 1.1156793212890626, "generation_rouge_l_f1": 0.15, "generation_token_f1": 0.16999999999999998, "prompt_id": "f552b8f246a9b742", "reading_canonical_log_prob": -27.891983032226562, "reading_distances": [23, 22, 27, 26], "reading_log_probs": [-25.587955474853516, -25.24493408203125, -29.153148651123047, -27.759605407714844], "reading_logprob_std": 1.850315406817499, "tokenization_path_logprob_std": 54.39471270589736, "trajectory_balance_rmse": 455.26878449078396} {"canonical_length": 212, "canonical_nll": 1.1667395537754275, "generation_rouge_l_f1": 0.2591792656587473, "generation_token_f1": 0.548596112311015, "prompt_id": "1dcab7a046130541", "reading_canonical_log_prob": -247.34878540039062, "reading_distances": [20, 18, 19, 26], "reading_log_probs": [-244.45132446289062, -244.95120239257812, -245.13694763183594, -244.03175354003906], "reading_logprob_std": 0.4997781424934839, "tokenization_path_logprob_std": 58.041705474740134, "trajectory_balance_rmse": 4018.688913739216} {"canonical_length": 21, "canonical_nll": 1.5472564697265625, "generation_rouge_l_f1": 0.25, "generation_token_f1": 0.25, "prompt_id": "721929550e55e7e5", "reading_canonical_log_prob": -32.49238586425781, "reading_distances": [24, 25, 26, 27], "reading_log_probs": [-34.878501892089844, -35.92402648925781, -34.76097869873047, -33.86351013183594], "reading_logprob_std": 0.8436527380218575, "tokenization_path_logprob_std": 24.34818793095511, "trajectory_balance_rmse": 325.2444917696875} {"canonical_length": 106, "canonical_nll": 2.334591127791495, "generation_rouge_l_f1": 0.2716049382716049, "generation_token_f1": 0.411522633744856, "prompt_id": "a21c76c0694212ca", "reading_canonical_log_prob": -247.46665954589844, "reading_distances": [34, 36, 35, 33], "reading_log_probs": [-244.21044921875, -244.05467224121094, -246.22787475585938, -243.83505249023438], "reading_logprob_std": 1.107995012317916, "tokenization_path_logprob_std": 101.12378081741348, "trajectory_balance_rmse": 1787.7200308609445} {"canonical_length": 10, "canonical_nll": 1.023246955871582, "generation_rouge_l_f1": 0.1956521739130435, "generation_token_f1": 0.1956521739130435, "prompt_id": "7cb87ccdae4ac547", "reading_canonical_log_prob": -10.23246955871582, "reading_distances": [19, 12, 18, 15], "reading_log_probs": [-10.958892822265625, -12.490010261535645, -11.712911605834961, -10.422700881958008], "reading_logprob_std": 0.9010485021745307, "tokenization_path_logprob_std": 19.41900001159748, "trajectory_balance_rmse": 247.70773743117658} {"canonical_length": 263, "canonical_nll": 0.8282560051167419, "generation_rouge_l_f1": 0.3310344827586207, "generation_token_f1": 0.5471264367816092, "prompt_id": "02274883d489f1d7", "reading_canonical_log_prob": -217.83132934570312, "reading_distances": [52, 51, 65, 54], "reading_log_probs": [-214.53839111328125, -212.63265991210938, -210.69186401367188, -214.29713439941406], "reading_logprob_std": 1.7800029385210985, "tokenization_path_logprob_std": 213.68862857629364, "trajectory_balance_rmse": 4118.657648370411} {"canonical_length": 219, "canonical_nll": 1.1597853011736587, "generation_rouge_l_f1": 0.21810699588477367, "generation_token_f1": 0.5102880658436214, "prompt_id": "dd480c5be4f865e4", "reading_canonical_log_prob": -253.99298095703125, "reading_distances": [25, 18, 27, 22], "reading_log_probs": [-253.07090759277344, -254.17868041992188, -250.29299926757812, -252.2733154296875], "reading_logprob_std": 1.6388683703492855, "tokenization_path_logprob_std": 249.90788962738333, "trajectory_balance_rmse": 4865.495702528767} {"canonical_length": 17, "canonical_nll": 0.682769214405733, "generation_rouge_l_f1": 0.07317073170731708, "generation_token_f1": 0.08943089430894309, "prompt_id": "48b6b6146fd695d3", "reading_canonical_log_prob": -11.607076644897461, "reading_distances": [17, 19, 18, 17], "reading_log_probs": [-10.280009269714355, -10.833154678344727, -15.922334671020508, -10.61612319946289], "reading_logprob_std": 2.6826217616455494, "tokenization_path_logprob_std": 27.243354003498073, "trajectory_balance_rmse": 245.6087988421163} {"canonical_length": 230, "canonical_nll": 2.148482082201087, "generation_rouge_l_f1": 0.23076923076923078, "generation_token_f1": 0.35256410256410253, "prompt_id": "619b20aa15b49441", "reading_canonical_log_prob": -494.15087890625, "reading_distances": [26, 22, 19, 24], "reading_log_probs": [-487.8032531738281, -484.8244934082031, -488.9512939453125, -487.66729736328125], "reading_logprob_std": 1.7552339720457069, "tokenization_path_logprob_std": 96.04260515277178, "trajectory_balance_rmse": 5723.07109358039} {"canonical_length": 264, "canonical_nll": 1.2241148515181108, "generation_rouge_l_f1": 0.2824858757062147, "generation_token_f1": 0.5009416195856874, "prompt_id": "49784b3e9fd7ae19", "reading_canonical_log_prob": -323.16632080078125, "reading_distances": [38, 31, 38, 35], "reading_log_probs": [-323.03802490234375, -323.417724609375, -324.5867919921875, -324.9163818359375], "reading_logprob_std": 0.9033455240322698, "tokenization_path_logprob_std": 283.4080579418692, "trajectory_balance_rmse": 5860.554875087515} {"canonical_length": 128, "canonical_nll": 1.2200442552566528, "generation_rouge_l_f1": 0.2898550724637681, "generation_token_f1": 0.4782608695652174, "prompt_id": "211ba51e9f80e5d0", "reading_canonical_log_prob": -156.16566467285156, "reading_distances": [11, 9, 14, 12], "reading_log_probs": [-158.60813903808594, -155.98013305664062, -157.88211059570312, -161.47727966308594], "reading_logprob_std": 2.280841961838585, "tokenization_path_logprob_std": 31.252228744105526, "trajectory_balance_rmse": 2513.4367868293193} {"canonical_length": 19, "canonical_nll": 0.48341269242136103, "generation_rouge_l_f1": 0.75, "generation_token_f1": 0.75, "prompt_id": "c295bb5aa04e8e4c", "reading_canonical_log_prob": -9.18484115600586, "reading_distances": [23, 23, 25, 18], "reading_log_probs": [-11.813644409179688, -9.157255172729492, -9.515364646911621, -9.012048721313477], "reading_logprob_std": 1.3099009590245907, "tokenization_path_logprob_std": 81.86265134310999, "trajectory_balance_rmse": 568.8760577125495} {"canonical_length": 163, "canonical_nll": 0.5986044480025403, "generation_rouge_l_f1": 0.26737967914438504, "generation_token_f1": 0.4224598930481283, "prompt_id": "55b91ee2bfd58fff", "reading_canonical_log_prob": -97.57252502441406, "reading_distances": [39, 24, 36, 32], "reading_log_probs": [-100.38150024414062, -99.02623748779297, -101.77558898925781, -101.27363586425781], "reading_logprob_std": 1.2054672888389386, "tokenization_path_logprob_std": 150.3644629273716, "trajectory_balance_rmse": 2221.5537485564028} {"canonical_length": 79, "canonical_nll": 0.9170953291880933, "generation_rouge_l_f1": 0.4383561643835616, "generation_token_f1": 0.5753424657534246, "prompt_id": "aacdb99cf2030827", "reading_canonical_log_prob": -72.45053100585938, "reading_distances": [13, 29, 28, 24], "reading_log_probs": [-74.54776000976562, -76.41947174072266, -75.87995147705078, -74.37277221679688], "reading_logprob_std": 1.0025097678529837, "tokenization_path_logprob_std": 103.5251449935173, "trajectory_balance_rmse": 1689.8479060855543} {"canonical_length": 7, "canonical_nll": 0.09546126638139997, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "732d6c9c54c4fcf0", "reading_canonical_log_prob": -0.6682288646697998, "reading_distances": [19, 17, 23, 18], "reading_log_probs": [-1.1529288291931152, -0.7845209240913391, -1.0093051195144653, -0.8027669787406921], "reading_logprob_std": 0.17618274902597061, "tokenization_path_logprob_std": 50.113455891302365, "trajectory_balance_rmse": 168.5657039742736} {"canonical_length": 234, "canonical_nll": 0.6114989712706997, "generation_rouge_l_f1": 0.43495145631067955, "generation_token_f1": 0.6485436893203883, "prompt_id": "1e288eaee9c72d47", "reading_canonical_log_prob": -143.09075927734375, "reading_distances": [34, 34, 31, 32], "reading_log_probs": [-147.56301879882812, -148.54934692382812, -147.7952423095703, -150.3450164794922], "reading_logprob_std": 1.2603152140209524, "tokenization_path_logprob_std": 322.629214310578, "trajectory_balance_rmse": 3822.081884625644} {"canonical_length": 85, "canonical_nll": 1.1562137379365809, "generation_rouge_l_f1": 0.22564102564102562, "generation_token_f1": 0.32820512820512815, "prompt_id": "34251c7523ed14e6", "reading_canonical_log_prob": -98.27816772460938, "reading_distances": [26, 25, 27, 25], "reading_log_probs": [-98.56838989257812, -98.04003143310547, -96.74283599853516, -98.95932006835938], "reading_logprob_std": 0.9663179819228636, "tokenization_path_logprob_std": 128.4428407727539, "trajectory_balance_rmse": 1709.6539425394087} {"canonical_length": 62, "canonical_nll": 2.4111214914629535, "generation_rouge_l_f1": 0.25, "generation_token_f1": 0.5208333333333334, "prompt_id": "915708e38c0c1bff", "reading_canonical_log_prob": -149.48953247070312, "reading_distances": [13, 14, 11, 14], "reading_log_probs": [-147.30856323242188, -148.2090301513672, -147.37939453125, -147.0673828125], "reading_logprob_std": 0.49690969678515073, "tokenization_path_logprob_std": 96.53965716386453, "trajectory_balance_rmse": 1038.6835994287894} {"canonical_length": 145, "canonical_nll": 1.6474110570447198, "generation_rouge_l_f1": 0.30526315789473685, "generation_token_f1": 0.4578947368421052, "prompt_id": "fc76d6d236be83db", "reading_canonical_log_prob": -238.87460327148438, "reading_distances": [23, 27, 22, 25], "reading_log_probs": [-241.02053833007812, -237.1947021484375, -237.7364044189453, -236.6182098388672], "reading_logprob_std": 1.9722913499977996, "tokenization_path_logprob_std": 209.23423447285055, "trajectory_balance_rmse": 2377.674423659355} {"canonical_length": 391, "canonical_nll": 0.7308972448949009, "generation_rouge_l_f1": 0.34931506849315064, "generation_token_f1": 0.48287671232876705, "prompt_id": "cf72c2330315d4cd", "reading_canonical_log_prob": -285.78082275390625, "reading_distances": [23, 19, 28, 22], "reading_log_probs": [-281.7581481933594, -282.866455078125, -280.8724365234375, -282.71026611328125], "reading_logprob_std": 0.9263494857491814, "tokenization_path_logprob_std": 112.23024221907738, "trajectory_balance_rmse": 4024.8373639082747} {"canonical_length": 247, "canonical_nll": 1.2258678853270495, "generation_rouge_l_f1": 0.2729044834307992, "generation_token_f1": 0.5341130604288498, "prompt_id": "12725fe24fa84c62", "reading_canonical_log_prob": -302.78936767578125, "reading_distances": [34, 37, 32, 31], "reading_log_probs": [-295.6282958984375, -295.83477783203125, -295.29827880859375, -296.2425842285156], "reading_logprob_std": 0.39525795440198647, "tokenization_path_logprob_std": 330.7815390263194, "trajectory_balance_rmse": 5799.7870967277395} {"canonical_length": 164, "canonical_nll": 1.6488829822075077, "generation_rouge_l_f1": 0.2816229116945107, "generation_token_f1": 0.4916467780429594, "prompt_id": "767d0dc7cec8c49a", "reading_canonical_log_prob": -270.41680908203125, "reading_distances": [13, 12, 17, 13], "reading_log_probs": [-271.89056396484375, -273.27349853515625, -272.75299072265625, -275.83795166015625], "reading_logprob_std": 1.6980991763968913, "tokenization_path_logprob_std": 162.03368398975812, "trajectory_balance_rmse": 3411.2426128844786} {"canonical_length": 211, "canonical_nll": 1.3836736453087974, "generation_rouge_l_f1": 0.2639175257731959, "generation_token_f1": 0.39999999999999997, "prompt_id": "0305770999fedf01", "reading_canonical_log_prob": -291.95513916015625, "reading_distances": [33, 26, 28, 37], "reading_log_probs": [-289.14837646484375, -291.1999816894531, -289.0836181640625, -292.9288330078125], "reading_logprob_std": 1.842976034290551, "tokenization_path_logprob_std": 147.9572692135689, "trajectory_balance_rmse": 2631.762264813768} {"canonical_length": 28, "canonical_nll": 0.654026712690081, "generation_rouge_l_f1": 0.6415094339622641, "generation_token_f1": 0.7924528301886792, "prompt_id": "47deaa7fb6aebf76", "reading_canonical_log_prob": -18.312747955322266, "reading_distances": [17, 18, 14, 15], "reading_log_probs": [-11.951783180236816, -11.723502159118652, -13.542119979858398, -12.472478866577148], "reading_logprob_std": 0.8095717157884522, "tokenization_path_logprob_std": 33.67728043806057, "trajectory_balance_rmse": 756.244847648896} {"canonical_length": 55, "canonical_nll": 1.1158058860085227, "generation_rouge_l_f1": 0.33663366336633666, "generation_token_f1": 0.3960396039603961, "prompt_id": "83b035c8880eabe5", "reading_canonical_log_prob": -61.36932373046875, "reading_distances": [38, 38, 29, 28], "reading_log_probs": [-61.3311767578125, -61.30156707763672, -62.5941162109375, -62.211952209472656], "reading_logprob_std": 0.646606002262783, "tokenization_path_logprob_std": 94.62892757221975, "trajectory_balance_rmse": 1470.2940976540317} {"canonical_length": 235, "canonical_nll": 1.443663641747008, "generation_rouge_l_f1": 0.2608695652173913, "generation_token_f1": 0.4782608695652174, "prompt_id": "3f8f300bcd0a09f9", "reading_canonical_log_prob": -339.2609558105469, "reading_distances": [30, 24, 30, 33], "reading_log_probs": [-327.26177978515625, -331.86578369140625, -327.365966796875, -329.04510498046875], "reading_logprob_std": 2.1488764732734933, "tokenization_path_logprob_std": 339.2730261015702, "trajectory_balance_rmse": 5012.2545499534635} {"canonical_length": 32, "canonical_nll": 1.0390411615371704, "generation_rouge_l_f1": 0.19696969696969696, "generation_token_f1": 0.3333333333333333, "prompt_id": "de8521c6b6a04f6f", "reading_canonical_log_prob": -33.24931716918945, "reading_distances": [20, 21, 24, 20], "reading_log_probs": [-31.198246002197266, -29.126567840576172, -30.934749603271484, -31.10893440246582], "reading_logprob_std": 0.9831452985335587, "tokenization_path_logprob_std": 99.0688130444714, "trajectory_balance_rmse": 649.0964781832662} {"canonical_length": 12, "canonical_nll": 0.38265395164489746, "generation_rouge_l_f1": 0.8571428571428571, "generation_token_f1": 0.8571428571428571, "prompt_id": "89d6bb48c1695f7e", "reading_canonical_log_prob": -4.5918474197387695, "reading_distances": [23, 26, 34, 26], "reading_log_probs": [-5.529541015625, -6.4298882484436035, -6.368083477020264, -7.441681385040283], "reading_logprob_std": 0.7826353472240632, "tokenization_path_logprob_std": 64.194526037384, "trajectory_balance_rmse": 262.7769057759103} {"canonical_length": 325, "canonical_nll": 1.6416856971153846, "generation_rouge_l_f1": 0.22913256955810146, "generation_token_f1": 0.4713584288052374, "prompt_id": "b1d79661f636c046", "reading_canonical_log_prob": -533.5478515625, "reading_distances": [16, 19, 22, 14], "reading_log_probs": [-527.4952392578125, -529.9193115234375, -527.3577270507812, -531.2071533203125], "reading_logprob_std": 1.8866123977544442, "tokenization_path_logprob_std": 432.9488099896533, "trajectory_balance_rmse": 6902.9585983569405} {"canonical_length": 15, "canonical_nll": 0.9568319320678711, "generation_rouge_l_f1": 0.11956521739130434, "generation_token_f1": 0.13043478260869565, "prompt_id": "db698be75c6d5259", "reading_canonical_log_prob": -14.352478981018066, "reading_distances": [22, 19, 26, 22], "reading_log_probs": [-16.36578369140625, -16.273582458496094, -16.876018524169922, -14.9765625], "reading_logprob_std": 0.8089035451505174, "tokenization_path_logprob_std": 47.6780124298587, "trajectory_balance_rmse": 348.0255503278217} {"canonical_length": 33, "canonical_nll": 0.8748397827148438, "generation_rouge_l_f1": 0.16417910447761194, "generation_token_f1": 0.17910447761194032, "prompt_id": "4c3924ee604d1034", "reading_canonical_log_prob": -28.869712829589844, "reading_distances": [31, 23, 26, 19], "reading_log_probs": [-26.758682250976562, -26.647258758544922, -26.302364349365234, -26.66904067993164], "reading_logprob_std": 0.20053157902526197, "tokenization_path_logprob_std": 60.09963570489683, "trajectory_balance_rmse": 760.1295935253685} {"canonical_length": 63, "canonical_nll": 1.2098472231910342, "generation_rouge_l_f1": 0.28, "generation_token_f1": 0.4699999999999999, "prompt_id": "1fd45a1cb9cbeccc", "reading_canonical_log_prob": -76.22037506103516, "reading_distances": [16, 16, 20, 19], "reading_log_probs": [-76.49174499511719, -74.07271575927734, -78.2569580078125, -76.85850524902344], "reading_logprob_std": 1.7398843089425329, "tokenization_path_logprob_std": 104.61231421601671, "trajectory_balance_rmse": 1194.7979238061787} {"canonical_length": 205, "canonical_nll": 1.0519603450123858, "generation_rouge_l_f1": 0.3896713615023474, "generation_token_f1": 0.596244131455399, "prompt_id": "326b7dac8c34a0d5", "reading_canonical_log_prob": -215.65187072753906, "reading_distances": [21, 21, 21, 12], "reading_log_probs": [-214.1114501953125, -214.63232421875, -213.47750854492188, -215.4939727783203], "reading_logprob_std": 0.8527761499079737, "tokenization_path_logprob_std": 72.32470494456388, "trajectory_balance_rmse": 4033.424555359185} {"canonical_length": 322, "canonical_nll": 1.4643171796147127, "generation_rouge_l_f1": 0.21516754850088182, "generation_token_f1": 0.40564373897707234, "prompt_id": "1e2cec78b7ee8d07", "reading_canonical_log_prob": -471.5101318359375, "reading_distances": [15, 19, 18, 17], "reading_log_probs": [-464.75286865234375, -472.2133483886719, -464.14019775390625, -465.572021484375], "reading_logprob_std": 3.7420831799131524, "tokenization_path_logprob_std": 240.31682434866627, "trajectory_balance_rmse": 6031.072910167809} {"canonical_length": 360, "canonical_nll": 1.1447061326768664, "generation_rouge_l_f1": 0.28760330578512394, "generation_token_f1": 0.4661157024793388, "prompt_id": "e3633ae650b9ffe5", "reading_canonical_log_prob": -412.0942077636719, "reading_distances": [22, 15, 20, 21], "reading_log_probs": [-405.83404541015625, -409.2882080078125, -406.755859375, -406.652099609375], "reading_logprob_std": 1.4950701233573842, "tokenization_path_logprob_std": 242.92740193578072, "trajectory_balance_rmse": 5799.94600408997} {"canonical_length": 199, "canonical_nll": 0.9763040973912531, "generation_rouge_l_f1": 0.24282560706401768, "generation_token_f1": 0.48123620309050774, "prompt_id": "e7d8d0860929d3e3", "reading_canonical_log_prob": -194.28451538085938, "reading_distances": [24, 30, 28, 25], "reading_log_probs": [-193.69451904296875, -192.3585205078125, -190.52301025390625, -195.2330780029297], "reading_logprob_std": 2.0005715983945795, "tokenization_path_logprob_std": 205.2632720510583, "trajectory_balance_rmse": 3444.012452029607} {"canonical_length": 58, "canonical_nll": 1.9287261962890625, "generation_rouge_l_f1": 0.1492537313432836, "generation_token_f1": 0.1492537313432836, "prompt_id": "8c4c85aa12a6cb37", "reading_canonical_log_prob": -111.86611938476562, "reading_distances": [30, 24, 30, 29], "reading_log_probs": [-102.84518432617188, -104.00716400146484, -103.37004852294922, -104.36626434326172], "reading_logprob_std": 0.6749493041657803, "tokenization_path_logprob_std": 57.423494662107906, "trajectory_balance_rmse": 1250.9759448297918} {"canonical_length": 35, "canonical_nll": 2.147682843889509, "generation_rouge_l_f1": 0.3589743589743589, "generation_token_f1": 0.4871794871794872, "prompt_id": "5f2bc0710dc7dd88", "reading_canonical_log_prob": -75.16889953613281, "reading_distances": [13, 16, 14, 25], "reading_log_probs": [-73.63868713378906, -72.35991668701172, -72.10623168945312, -73.85588073730469], "reading_logprob_std": 0.8847967240293677, "tokenization_path_logprob_std": 49.22603765095408, "trajectory_balance_rmse": 728.529903577751} {"canonical_length": 24, "canonical_nll": 0.8581141630808512, "generation_rouge_l_f1": 0.42857142857142855, "generation_token_f1": 0.42857142857142855, "prompt_id": "e06153205819c799", "reading_canonical_log_prob": -20.59473991394043, "reading_distances": [70, 71, 64, 77], "reading_log_probs": [-22.605634689331055, -18.411563873291016, -22.05942153930664, -19.771854400634766], "reading_logprob_std": 1.9644572792513524, "tokenization_path_logprob_std": 72.12224943533477, "trajectory_balance_rmse": 467.46136754485275} {"canonical_length": 81, "canonical_nll": 0.8008508564513407, "generation_rouge_l_f1": 0.15709969788519634, "generation_token_f1": 0.19939577039274922, "prompt_id": "788acaefe94c61a6", "reading_canonical_log_prob": -64.8689193725586, "reading_distances": [19, 19, 26, 26], "reading_log_probs": [-65.51185607910156, -63.266685485839844, -65.39681243896484, -66.35739135742188], "reading_logprob_std": 1.3159778182816702, "tokenization_path_logprob_std": 35.825027994298345, "trajectory_balance_rmse": 937.0718730291384} {"canonical_length": 21, "canonical_nll": 2.1935551961263022, "generation_rouge_l_f1": 0.15517241379310348, "generation_token_f1": 0.18965517241379307, "prompt_id": "24304bef6f934812", "reading_canonical_log_prob": -46.064659118652344, "reading_distances": [17, 17, 14, 14], "reading_log_probs": [-45.229217529296875, -45.92027282714844, -44.944942474365234, -42.602928161621094], "reading_logprob_std": 1.4403946664708858, "tokenization_path_logprob_std": 62.09907806588569, "trajectory_balance_rmse": 365.131395806007} {"canonical_length": 315, "canonical_nll": 1.4507246713789683, "generation_rouge_l_f1": 0.2191304347826087, "generation_token_f1": 0.4765217391304348, "prompt_id": "969a799cc7257588", "reading_canonical_log_prob": -456.978271484375, "reading_distances": [20, 15, 16, 18], "reading_log_probs": [-446.78045654296875, -447.3963317871094, -446.7530517578125, -445.4803161621094], "reading_logprob_std": 0.8049430267852301, "tokenization_path_logprob_std": 284.4175274065448, "trajectory_balance_rmse": 6264.021049502877} {"canonical_length": 152, "canonical_nll": 1.2858244243421053, "generation_rouge_l_f1": 0.18316831683168316, "generation_token_f1": 0.31683168316831684, "prompt_id": "7f301b40ae5af8a5", "reading_canonical_log_prob": -195.4453125, "reading_distances": [23, 32, 31, 24], "reading_log_probs": [-188.69747924804688, -191.8609619140625, -192.09231567382812, -188.99575805664062], "reading_logprob_std": 1.8136770653140326, "tokenization_path_logprob_std": 277.0236943374086, "trajectory_balance_rmse": 2815.8232239214162} {"canonical_length": 354, "canonical_nll": 1.4348953160862465, "generation_rouge_l_f1": 0.21103896103896103, "generation_token_f1": 0.4383116883116883, "prompt_id": "d838a6ece3beb4b3", "reading_canonical_log_prob": -507.95294189453125, "reading_distances": [15, 19, 22, 17], "reading_log_probs": [-501.8462219238281, -502.21295166015625, -504.34942626953125, -503.5665283203125], "reading_logprob_std": 1.1679611695869179, "tokenization_path_logprob_std": 200.34035899917635, "trajectory_balance_rmse": 7704.102359660615} {"canonical_length": 15, "canonical_nll": 2.348475138346354, "generation_rouge_l_f1": 0.2962962962962963, "generation_token_f1": 0.3703703703703704, "prompt_id": "3ad7d2823c1e6bd8", "reading_canonical_log_prob": -35.22712707519531, "reading_distances": [40, 44, 31, 41], "reading_log_probs": [-37.35004425048828, -36.74269104003906, -37.76383972167969, -37.693023681640625], "reading_logprob_std": 0.46624664213729916, "tokenization_path_logprob_std": 58.577337697585115, "trajectory_balance_rmse": 529.0536155937168} {"canonical_length": 8, "canonical_nll": 0.2641701102256775, "generation_rouge_l_f1": 0.7692307692307692, "generation_token_f1": 0.7692307692307692, "prompt_id": "7a4144ac9a54046a", "reading_canonical_log_prob": -2.11336088180542, "reading_distances": [32, 28, 27, 30], "reading_log_probs": [-1.4163339138031006, -1.00885808467865, -1.4987013339996338, -1.2347412109375], "reading_logprob_std": 0.21726354870780654, "tokenization_path_logprob_std": 29.635653158319766, "trajectory_balance_rmse": 117.89252632439882} {"canonical_length": 34, "canonical_nll": 0.4355659484863281, "generation_rouge_l_f1": 0.23076923076923075, "generation_token_f1": 0.3333333333333333, "prompt_id": "c898f115e1329204", "reading_canonical_log_prob": -14.809242248535156, "reading_distances": [10, 7, 6, 7], "reading_log_probs": [-14.950019836425781, -14.600432395935059, -15.047080039978027, -14.927997589111328], "reading_logprob_std": 0.1943130517261167, "tokenization_path_logprob_std": 65.58274519507673, "trajectory_balance_rmse": 520.0912561330492} {"canonical_length": 104, "canonical_nll": 1.7020665682279146, "generation_rouge_l_f1": 0.19999999999999998, "generation_token_f1": 0.3277777777777778, "prompt_id": "5328de30202d5ac5", "reading_canonical_log_prob": -177.01492309570312, "reading_distances": [14, 12, 11, 14], "reading_log_probs": [-173.8485870361328, -174.85238647460938, -175.1392364501953, -174.89410400390625], "reading_logprob_std": 0.5708625147949262, "tokenization_path_logprob_std": 22.888212010724548, "trajectory_balance_rmse": 2424.258441456795} {"canonical_length": 35, "canonical_nll": 1.2694826398577008, "generation_rouge_l_f1": 0.4285714285714286, "generation_token_f1": 0.4285714285714286, "prompt_id": "f9e71d1877230558", "reading_canonical_log_prob": -44.43189239501953, "reading_distances": [29, 22, 28, 26], "reading_log_probs": [-41.49590301513672, -42.122840881347656, -40.09239196777344, -41.19683074951172], "reading_logprob_std": 0.8491339286655081, "tokenization_path_logprob_std": 126.72594276371944, "trajectory_balance_rmse": 761.2201281214952} {"canonical_length": 211, "canonical_nll": 1.5396207836566944, "generation_rouge_l_f1": 0.2217741935483871, "generation_token_f1": 0.435483870967742, "prompt_id": "4df4b024265f29c3", "reading_canonical_log_prob": -324.8599853515625, "reading_distances": [27, 28, 31, 30], "reading_log_probs": [-325.9013671875, -325.06201171875, -325.6783142089844, -327.1478576660156], "reading_logprob_std": 0.875506938601489, "tokenization_path_logprob_std": 112.44220309708176, "trajectory_balance_rmse": 4571.515443883466} {"canonical_length": 21, "canonical_nll": 0.2628281457083566, "generation_rouge_l_f1": 0.36666666666666664, "generation_token_f1": 0.36666666666666664, "prompt_id": "6e5cc5a69c7269eb", "reading_canonical_log_prob": -5.519391059875488, "reading_distances": [19, 26, 29, 24], "reading_log_probs": [-4.775003433227539, -8.198307991027832, -6.174673080444336, -10.419737815856934], "reading_logprob_std": 2.4595314389982708, "tokenization_path_logprob_std": 72.2359236598139, "trajectory_balance_rmse": 466.828049974417} {"canonical_length": 81, "canonical_nll": 1.4931556795850212, "generation_rouge_l_f1": 0.24778761061946902, "generation_token_f1": 0.33038348082595875, "prompt_id": "15d75a153e79b8a7", "reading_canonical_log_prob": -120.94561004638672, "reading_distances": [37, 41, 41, 42], "reading_log_probs": [-115.35467529296875, -116.31802368164062, -117.6943359375, -117.53741455078125], "reading_logprob_std": 1.101974612385381, "tokenization_path_logprob_std": 107.78047993521658, "trajectory_balance_rmse": 2069.395383155086} {"canonical_length": 4, "canonical_nll": 1.4427919387817383, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.6666666666666666, "prompt_id": "9acb0ff71abeda7d", "reading_canonical_log_prob": -5.771167755126953, "reading_distances": [21, 16, 14, 15], "reading_log_probs": [-14.486515045166016, -14.348896026611328, -7.046442985534668, -7.5589799880981445], "reading_logprob_std": 4.113553083202165, "tokenization_path_logprob_std": 16.98548105070163, "trajectory_balance_rmse": 60.50038724478442} {"canonical_length": 114, "canonical_nll": 0.5785397981342516, "generation_rouge_l_f1": 0.4385964912280702, "generation_token_f1": 0.6228070175438596, "prompt_id": "1b111d554cbb9115", "reading_canonical_log_prob": -65.95353698730469, "reading_distances": [9, 12, 11, 10], "reading_log_probs": [-66.87576293945312, -66.37046813964844, -66.49694061279297, -66.01715850830078], "reading_logprob_std": 0.3543826332437836, "tokenization_path_logprob_std": 38.69423675588581, "trajectory_balance_rmse": 1440.567439223723} {"canonical_length": 135, "canonical_nll": 1.395934154369213, "generation_rouge_l_f1": 0.2305764411027569, "generation_token_f1": 0.431077694235589, "prompt_id": "4c468581b163e3ed", "reading_canonical_log_prob": -188.45111083984375, "reading_distances": [19, 23, 22, 28], "reading_log_probs": [-186.2537384033203, -185.0509033203125, -185.80691528320312, -187.22677612304688], "reading_logprob_std": 0.908994265100676, "tokenization_path_logprob_std": 293.9536404793398, "trajectory_balance_rmse": 2809.484016482215} {"canonical_length": 38, "canonical_nll": 0.9495177018015009, "generation_rouge_l_f1": 0.417910447761194, "generation_token_f1": 0.47761194029850745, "prompt_id": "5a9339c4c26875b0", "reading_canonical_log_prob": -36.08167266845703, "reading_distances": [20, 17, 15, 19], "reading_log_probs": [-42.3175163269043, -46.21460723876953, -38.683834075927734, -44.83026123046875], "reading_logprob_std": 3.305454788347961, "tokenization_path_logprob_std": 47.70794362288216, "trajectory_balance_rmse": 452.6948648079666} {"canonical_length": 245, "canonical_nll": 1.31660118881537, "generation_rouge_l_f1": 0.28793774319066145, "generation_token_f1": 0.5175097276264591, "prompt_id": "dadb86ce514e9684", "reading_canonical_log_prob": -322.5672912597656, "reading_distances": [14, 16, 17, 17], "reading_log_probs": [-323.0438537597656, -323.3607177734375, -324.2247619628906, -324.20489501953125], "reading_logprob_std": 0.5987880867381837, "tokenization_path_logprob_std": 83.46965303578786, "trajectory_balance_rmse": 5186.485769680506} {"canonical_length": 153, "canonical_nll": 1.6008831348294525, "generation_rouge_l_f1": 0.21212121212121213, "generation_token_f1": 0.47878787878787876, "prompt_id": "8be6dabc44180a81", "reading_canonical_log_prob": -244.93511962890625, "reading_distances": [21, 22, 22, 24], "reading_log_probs": [-244.4972381591797, -246.53814697265625, -245.91697692871094, -244.4573974609375], "reading_logprob_std": 1.0419651792259133, "tokenization_path_logprob_std": 168.82099142158293, "trajectory_balance_rmse": 3391.857417327905} {"canonical_length": 8, "canonical_nll": 1.1783535480499268, "generation_rouge_l_f1": 0.5, "generation_token_f1": 0.5, "prompt_id": "5786bd96f280e551", "reading_canonical_log_prob": -9.426828384399414, "reading_distances": [14, 23, 19, 22], "reading_log_probs": [-15.548836708068848, -14.730828285217285, -16.872501373291016, -13.061188697814941], "reading_logprob_std": 1.5945266493795518, "tokenization_path_logprob_std": 29.553462002439037, "trajectory_balance_rmse": 101.6092166465614} {"canonical_length": 103, "canonical_nll": 1.2514928428872119, "generation_rouge_l_f1": 0.21183800623052962, "generation_token_f1": 0.3613707165109034, "prompt_id": "edcbad1e6d6e267f", "reading_canonical_log_prob": -128.9037628173828, "reading_distances": [20, 28, 20, 24], "reading_log_probs": [-127.90735626220703, -130.73500061035156, -128.14776611328125, -129.56094360351562], "reading_logprob_std": 1.3183633593860364, "tokenization_path_logprob_std": 199.850569945547, "trajectory_balance_rmse": 2165.8644024778173} {"canonical_length": 15, "canonical_nll": 1.9900747934977214, "generation_rouge_l_f1": 0.10752688172043011, "generation_token_f1": 0.11827956989247312, "prompt_id": "f125519b3e50ff0b", "reading_canonical_log_prob": -29.85112190246582, "reading_distances": [14, 14, 15, 14], "reading_log_probs": [-26.6387882232666, -24.350528717041016, -26.035491943359375, -24.67840576171875], "reading_logprob_std": 1.0890165555673685, "tokenization_path_logprob_std": 42.41901768579202, "trajectory_balance_rmse": 313.8402255966914} {"canonical_length": 335, "canonical_nll": 0.6616614042822995, "generation_rouge_l_f1": 0.18563922942206654, "generation_token_f1": 0.3432574430823117, "prompt_id": "805fd7652d88dc11", "reading_canonical_log_prob": -221.6565704345703, "reading_distances": [24, 26, 23, 25], "reading_log_probs": [-229.52566528320312, -228.62391662597656, -227.7135009765625, -228.80667114257812], "reading_logprob_std": 0.7456160522921195, "tokenization_path_logprob_std": 182.98565588102068, "trajectory_balance_rmse": 4589.964683013623} {"canonical_length": 58, "canonical_nll": 1.4497341287547145, "generation_rouge_l_f1": 0.20454545454545456, "generation_token_f1": 0.29545454545454547, "prompt_id": "7efa21552d311157", "reading_canonical_log_prob": -84.08457946777344, "reading_distances": [38, 36, 33, 36], "reading_log_probs": [-82.31584167480469, -82.62744140625, -84.73959350585938, -82.31600189208984], "reading_logprob_std": 1.169175091323149, "tokenization_path_logprob_std": 81.78582412978405, "trajectory_balance_rmse": 1388.1590549827813} {"canonical_length": 337, "canonical_nll": 1.0968675019125556, "generation_rouge_l_f1": 0.25370675453047775, "generation_token_f1": 0.5205930807248764, "prompt_id": "961872748d95ea69", "reading_canonical_log_prob": -369.64434814453125, "reading_distances": [27, 32, 27, 29], "reading_log_probs": [-372.0887451171875, -371.99200439453125, -371.4029541015625, -372.68023681640625], "reading_logprob_std": 0.5229424572058435, "tokenization_path_logprob_std": 342.81146112479917, "trajectory_balance_rmse": 7702.532145347496} {"canonical_length": 62, "canonical_nll": 0.7634740029611895, "generation_rouge_l_f1": 0.25384615384615383, "generation_token_f1": 0.3076923076923077, "prompt_id": "e754a1aad6ea85da", "reading_canonical_log_prob": -47.33538818359375, "reading_distances": [6, 6, 9, 7], "reading_log_probs": [-45.73603820800781, -45.60808563232422, -45.01537322998047, -45.25224304199219], "reading_logprob_std": 0.32962439276496713, "tokenization_path_logprob_std": 102.80503916533512, "trajectory_balance_rmse": 931.5152834379085} {"canonical_length": 200, "canonical_nll": 1.8839646911621093, "generation_rouge_l_f1": 0.20502092050209206, "generation_token_f1": 0.38075313807531375, "prompt_id": "4b3ba24fb950898f", "reading_canonical_log_prob": -376.7929382324219, "reading_distances": [42, 39, 42, 44], "reading_log_probs": [-373.51800537109375, -372.6773376464844, -373.41766357421875, -372.0159606933594], "reading_logprob_std": 0.7025668315138136, "tokenization_path_logprob_std": 170.0885420613516, "trajectory_balance_rmse": 4677.107561105439} {"canonical_length": 33, "canonical_nll": 1.1356270530007102, "generation_rouge_l_f1": 0.10317460317460317, "generation_token_f1": 0.15079365079365079, "prompt_id": "7b02e9bcfb31548c", "reading_canonical_log_prob": -37.47569274902344, "reading_distances": [11, 16, 15, 12], "reading_log_probs": [-38.410858154296875, -38.22425842285156, -40.572322845458984, -41.52070617675781], "reading_logprob_std": 1.624225688950326, "tokenization_path_logprob_std": 78.06008761494802, "trajectory_balance_rmse": 707.7608855147869} {"canonical_length": 64, "canonical_nll": 0.7954563498497009, "generation_rouge_l_f1": 0.28688524590163933, "generation_token_f1": 0.38524590163934425, "prompt_id": "2e9002691e2c9828", "reading_canonical_log_prob": -50.90920639038086, "reading_distances": [33, 45, 41, 34], "reading_log_probs": [-53.24272537231445, -55.391845703125, -53.66706466674805, -54.393985748291016], "reading_logprob_std": 0.9408857023259063, "tokenization_path_logprob_std": 107.47186981519985, "trajectory_balance_rmse": 1128.3235329819697} {"canonical_length": 110, "canonical_nll": 1.2580542824485086, "generation_rouge_l_f1": 0.22631578947368422, "generation_token_f1": 0.43157894736842106, "prompt_id": "660c03c6bae9ac45", "reading_canonical_log_prob": -138.38597106933594, "reading_distances": [31, 30, 29, 28], "reading_log_probs": [-133.70651245117188, -133.71885681152344, -133.5811767578125, -134.0032958984375], "reading_logprob_std": 0.17841613204098009, "tokenization_path_logprob_std": 137.77111380439683, "trajectory_balance_rmse": 2280.580347038125} {"canonical_length": 129, "canonical_nll": 1.153824207394622, "generation_rouge_l_f1": 0.21333333333333332, "generation_token_f1": 0.42666666666666664, "prompt_id": "dc5df7204109382e", "reading_canonical_log_prob": -148.84332275390625, "reading_distances": [34, 21, 31, 29], "reading_log_probs": [-149.4482879638672, -146.03311157226562, -148.58563232421875, -148.23843383789062], "reading_logprob_std": 1.4540376776936728, "tokenization_path_logprob_std": 187.56107441676753, "trajectory_balance_rmse": 2516.959692044802} {"canonical_length": 330, "canonical_nll": 1.3236167214133523, "generation_rouge_l_f1": 0.32450331125827814, "generation_token_f1": 0.5596026490066225, "prompt_id": "e7afe4dd1ceeee4d", "reading_canonical_log_prob": -436.79351806640625, "reading_distances": [24, 17, 18, 16], "reading_log_probs": [-442.3766174316406, -442.749755859375, -443.25665283203125, -445.0372009277344], "reading_logprob_std": 1.1779961847222078, "tokenization_path_logprob_std": 165.9186942471634, "trajectory_balance_rmse": 5681.398412312774} {"canonical_length": 142, "canonical_nll": 0.8004171881877201, "generation_rouge_l_f1": 0.2228915662650603, "generation_token_f1": 0.35542168674698793, "prompt_id": "64cb8edf4fa19a8d", "reading_canonical_log_prob": -113.65924072265625, "reading_distances": [14, 17, 19, 20], "reading_log_probs": [-114.48094177246094, -112.10084533691406, -111.60625457763672, -111.90431213378906], "reading_logprob_std": 1.320977839561293, "tokenization_path_logprob_std": 150.88195520833398, "trajectory_balance_rmse": 1869.6924159361838} {"canonical_length": 67, "canonical_nll": 1.5386486907503498, "generation_rouge_l_f1": 0.16296296296296295, "generation_token_f1": 0.1925925925925926, "prompt_id": "d242cdc5b38c5c82", "reading_canonical_log_prob": -103.08946228027344, "reading_distances": [24, 28, 31, 27], "reading_log_probs": [-100.24375915527344, -97.760498046875, -98.05107116699219, -98.41761016845703], "reading_logprob_std": 1.1165372954864823, "tokenization_path_logprob_std": 126.59658451796467, "trajectory_balance_rmse": 1396.5498311152803} {"canonical_length": 71, "canonical_nll": 3.396483515349912, "generation_rouge_l_f1": 0.22857142857142856, "generation_token_f1": 0.2857142857142857, "prompt_id": "406fdefc259c7cdd", "reading_canonical_log_prob": -241.15032958984375, "reading_distances": [32, 31, 32, 35], "reading_log_probs": [-238.681884765625, -239.6492156982422, -235.32723999023438, -234.13687133789062], "reading_logprob_std": 2.6351614380769193, "tokenization_path_logprob_std": 104.07636865080175, "trajectory_balance_rmse": 1227.5890073148555} {"canonical_length": 112, "canonical_nll": 1.1341735294886999, "generation_rouge_l_f1": 0.2652519893899204, "generation_token_f1": 0.41379310344827586, "prompt_id": "8b97cb84ee36bc33", "reading_canonical_log_prob": -127.02743530273438, "reading_distances": [27, 21, 23, 24], "reading_log_probs": [-129.57615661621094, -128.04257202148438, -127.7175064086914, -129.05282592773438], "reading_logprob_std": 0.8655295170321741, "tokenization_path_logprob_std": 118.32198322980894, "trajectory_balance_rmse": 2325.0902680929935} {"canonical_length": 3, "canonical_nll": 2.5141989390055337, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.6666666666666666, "prompt_id": "3c3cb4aabb9bd15f", "reading_canonical_log_prob": -7.542596817016602, "reading_distances": [24, 22, 21, 22], "reading_log_probs": [-5.336086273193359, -5.480493545532227, -5.661382675170898, -5.682884216308594], "reading_logprob_std": 0.16357597620649195, "tokenization_path_logprob_std": 14.596093611866415, "trajectory_balance_rmse": 41.5491635958695} {"canonical_length": 76, "canonical_nll": 2.1171232524671053, "generation_rouge_l_f1": 0.15942028985507245, "generation_token_f1": 0.36231884057971014, "prompt_id": "4d20c1324c854a9e", "reading_canonical_log_prob": -160.9013671875, "reading_distances": [11, 17, 12, 15], "reading_log_probs": [-157.46311950683594, -157.84365844726562, -157.5803680419922, -156.8498992919922], "reading_logprob_std": 0.4208186294789043, "tokenization_path_logprob_std": 182.73495114405063, "trajectory_balance_rmse": 1303.1558390969603} {"canonical_length": 194, "canonical_nll": 0.7933392082293009, "generation_rouge_l_f1": 0.3670588235294118, "generation_token_f1": 0.5223529411764706, "prompt_id": "45005e42ada68a08", "reading_canonical_log_prob": -153.90780639648438, "reading_distances": [25, 26, 20, 23], "reading_log_probs": [-151.35186767578125, -148.96575927734375, -150.23846435546875, -149.9142303466797], "reading_logprob_std": 0.9842291658327932, "tokenization_path_logprob_std": 83.81577085165942, "trajectory_balance_rmse": 2385.039291785329} {"canonical_length": 3, "canonical_nll": 0.6094028552373251, "generation_rouge_l_f1": 0.14285714285714288, "generation_token_f1": 0.14285714285714288, "prompt_id": "1b11c060139de682", "reading_canonical_log_prob": -1.828208565711975, "reading_distances": [17, 18, 19, 19], "reading_log_probs": [-4.0805134773254395, -4.541658401489258, -5.487237930297852, -4.458542823791504], "reading_logprob_std": 0.5981692194500765, "tokenization_path_logprob_std": 20.35731495068812, "trajectory_balance_rmse": 35.053049014475704} {"canonical_length": 180, "canonical_nll": 1.5306259155273438, "generation_rouge_l_f1": 0.19457013574660634, "generation_token_f1": 0.39366515837104077, "prompt_id": "fb27819d14f6e258", "reading_canonical_log_prob": -275.5126647949219, "reading_distances": [18, 20, 19, 19], "reading_log_probs": [-276.4112854003906, -278.053955078125, -276.09893798828125, -278.4400939941406], "reading_logprob_std": 1.1677683636566336, "tokenization_path_logprob_std": 214.58305059587192, "trajectory_balance_rmse": 3918.3335512370168} {"canonical_length": 92, "canonical_nll": 1.8548849354619565, "generation_rouge_l_f1": 0.17366946778711487, "generation_token_f1": 0.31932773109243695, "prompt_id": "b1a0ce36dec4c1a2", "reading_canonical_log_prob": -170.6494140625, "reading_distances": [31, 34, 28, 28], "reading_log_probs": [-174.46646118164062, -176.81832885742188, -173.54916381835938, -173.04922485351562], "reading_logprob_std": 1.6714464504405644, "tokenization_path_logprob_std": 180.79945862846301, "trajectory_balance_rmse": 1737.9305818518894} {"canonical_length": 205, "canonical_nll": 1.2613973477991616, "generation_rouge_l_f1": 0.21691973969631234, "generation_token_f1": 0.3774403470715836, "prompt_id": "fe99f791079b2601", "reading_canonical_log_prob": -258.5864562988281, "reading_distances": [21, 31, 27, 25], "reading_log_probs": [-253.42684936523438, -249.69879150390625, -248.90533447265625, -253.59645080566406], "reading_logprob_std": 2.452875379574007, "tokenization_path_logprob_std": 217.18708842775973, "trajectory_balance_rmse": 3589.748248037952} {"canonical_length": 265, "canonical_nll": 1.8238244398584906, "generation_rouge_l_f1": 0.21292775665399238, "generation_token_f1": 0.46768060836501896, "prompt_id": "7fb35b5d3ec740ba", "reading_canonical_log_prob": -483.3134765625, "reading_distances": [22, 26, 27, 27], "reading_log_probs": [-481.6382141113281, -482.35919189453125, -480.0182189941406, -481.0885009765625], "reading_logprob_std": 0.9868591053347882, "tokenization_path_logprob_std": 252.17231265870217, "trajectory_balance_rmse": 6301.138795632864} {"canonical_length": 11, "canonical_nll": 0.42148936878551135, "generation_rouge_l_f1": 0.88, "generation_token_f1": 0.88, "prompt_id": "cb7972227cbb9dc7", "reading_canonical_log_prob": -4.636383056640625, "reading_distances": [29, 19, 21, 21], "reading_log_probs": [-4.447629928588867, -4.3831586837768555, -4.391536712646484, -3.9890358448028564], "reading_logprob_std": 0.21115189559284853, "tokenization_path_logprob_std": 71.27874629691168, "trajectory_balance_rmse": 240.8945713706135} {"canonical_length": 109, "canonical_nll": 1.8321477207568808, "generation_rouge_l_f1": 0.011235955056179777, "generation_token_f1": 0.011235955056179777, "prompt_id": "ce54506448843c44", "reading_canonical_log_prob": -199.7041015625, "reading_distances": [22, 21, 20, 20], "reading_log_probs": [-194.87445068359375, -196.44595336914062, -196.17727661132812, -195.61004638671875], "reading_logprob_std": 0.6952670313312853, "tokenization_path_logprob_std": 56.95926185706619, "trajectory_balance_rmse": 1970.9740373961235} {"canonical_length": 90, "canonical_nll": 0.7178610059950087, "generation_rouge_l_f1": 0.27947598253275113, "generation_token_f1": 0.39301310043668114, "prompt_id": "110898cad92a62bf", "reading_canonical_log_prob": -64.60749053955078, "reading_distances": [12, 16, 16, 17], "reading_log_probs": [-63.628082275390625, -66.427490234375, -66.35323333740234, -66.41853332519531], "reading_logprob_std": 1.3862301604934564, "tokenization_path_logprob_std": 115.05855834492324, "trajectory_balance_rmse": 1077.4286149354502} {"canonical_length": 6, "canonical_nll": 1.0865565141042073, "generation_rouge_l_f1": 0.03745318352059925, "generation_token_f1": 0.03745318352059925, "prompt_id": "bb43b5a4398228cc", "reading_canonical_log_prob": -6.519339084625244, "reading_distances": [34, 27, 22, 30], "reading_log_probs": [-6.966850757598877, -6.475916862487793, -5.827216625213623, -5.663180828094482], "reading_logprob_std": 0.6019134027096329, "tokenization_path_logprob_std": 19.711281831586657, "trajectory_balance_rmse": 107.98172621679535} {"canonical_length": 344, "canonical_nll": 1.3379764113315316, "generation_rouge_l_f1": 0.19784172661870503, "generation_token_f1": 0.3669064748201439, "prompt_id": "6dc0804e4c40522b", "reading_canonical_log_prob": -460.2638854980469, "reading_distances": [16, 21, 19, 19], "reading_log_probs": [-458.1216735839844, -453.3402404785156, -459.8026123046875, -460.8653564453125], "reading_logprob_std": 3.325839470598004, "tokenization_path_logprob_std": 185.91569052279954, "trajectory_balance_rmse": 5446.45950212658} {"canonical_length": 325, "canonical_nll": 1.1576014122596154, "generation_rouge_l_f1": 0.2538071065989848, "generation_token_f1": 0.47715736040609136, "prompt_id": "8ed2c6a41c30f4da", "reading_canonical_log_prob": -376.220458984375, "reading_distances": [30, 35, 29, 23], "reading_log_probs": [-372.9217529296875, -372.85162353515625, -371.6064758300781, -374.1844482421875], "reading_logprob_std": 1.0528543622645905, "tokenization_path_logprob_std": 62.71698600431099, "trajectory_balance_rmse": 7298.449535731437} {"canonical_length": 116, "canonical_nll": 0.7967262925772831, "generation_rouge_l_f1": 0.3024691358024691, "generation_token_f1": 0.4444444444444445, "prompt_id": "b2cdbde556d30889", "reading_canonical_log_prob": -92.42024993896484, "reading_distances": [15, 15, 13, 12], "reading_log_probs": [-97.01390075683594, -95.0387954711914, -94.14301300048828, -92.68479919433594], "reading_logprob_std": 1.81094519400067, "tokenization_path_logprob_std": 125.65937674561856, "trajectory_balance_rmse": 1381.1339173010647} {"canonical_length": 15, "canonical_nll": 1.7258776346842448, "generation_rouge_l_f1": 0.5185185185185186, "generation_token_f1": 0.5185185185185186, "prompt_id": "eb54d1c2e52f2aa1", "reading_canonical_log_prob": -25.888164520263672, "reading_distances": [21, 23, 21, 19], "reading_log_probs": [-28.4465274810791, -25.942041397094727, -26.05672264099121, -28.972623825073242], "reading_logprob_std": 1.5800964220986562, "tokenization_path_logprob_std": 37.877588875891746, "trajectory_balance_rmse": 321.89303964768277} {"canonical_length": 61, "canonical_nll": 1.4788908411244877, "generation_rouge_l_f1": 0.12, "generation_token_f1": 0.22666666666666666, "prompt_id": "2d9f537706729121", "reading_canonical_log_prob": -90.21234130859375, "reading_distances": [11, 15, 10, 10], "reading_log_probs": [-88.78913879394531, -87.29988861083984, -88.4613037109375, -88.35623168945312], "reading_logprob_std": 0.6447568259182467, "tokenization_path_logprob_std": 40.414147444952306, "trajectory_balance_rmse": 1185.8313494725096} {"canonical_length": 2, "canonical_nll": 0.4105542302131653, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "e6a02b338700cb67", "reading_canonical_log_prob": -0.8211084604263306, "reading_distances": [22, 19, 25, 21], "reading_log_probs": [-2.0157835483551025, -1.8812719583511353, -2.794081211090088, -1.7741302251815796], "reading_logprob_std": 0.4625326154980049, "tokenization_path_logprob_std": 27.00298824199575, "trajectory_balance_rmse": 44.09846255591452} {"canonical_length": 169, "canonical_nll": 0.8471204768976516, "generation_rouge_l_f1": 0.2350119904076739, "generation_token_f1": 0.4124700239808154, "prompt_id": "56062aaca5480bb3", "reading_canonical_log_prob": -143.16336059570312, "reading_distances": [27, 24, 27, 23], "reading_log_probs": [-139.9671173095703, -140.44973754882812, -141.6337432861328, -138.65859985351562], "reading_logprob_std": 1.230999074196747, "tokenization_path_logprob_std": 134.4053687984213, "trajectory_balance_rmse": 2026.9627088258114} {"canonical_length": 3, "canonical_nll": 0.21751171350479126, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "a6378b47a0666677", "reading_canonical_log_prob": -0.6525351405143738, "reading_distances": [29, 31, 27, 30], "reading_log_probs": [-2.0428578853607178, -3.502448081970215, -2.952786922454834, -3.1314737796783447], "reading_logprob_std": 0.620157055116665, "tokenization_path_logprob_std": 24.300229740858235, "trajectory_balance_rmse": 60.3936881001855} {"canonical_length": 199, "canonical_nll": 1.0191609751639055, "generation_rouge_l_f1": 0.3295194508009154, "generation_token_f1": 0.5766590389016018, "prompt_id": "41d1448457af2a4f", "reading_canonical_log_prob": -202.8130340576172, "reading_distances": [13, 18, 21, 16], "reading_log_probs": [-203.27957153320312, -200.55459594726562, -203.07901000976562, -200.80128479003906], "reading_logprob_std": 1.4499763208025387, "tokenization_path_logprob_std": 122.83785300955076, "trajectory_balance_rmse": 4174.32549486255} {"canonical_length": 94, "canonical_nll": 1.8586914387155087, "generation_rouge_l_f1": 0.15211267605633802, "generation_token_f1": 0.2535211267605634, "prompt_id": "5749ffa6095e5a40", "reading_canonical_log_prob": -174.7169952392578, "reading_distances": [31, 23, 22, 19], "reading_log_probs": [-170.5155487060547, -169.78494262695312, -173.36293029785156, -176.30491638183594], "reading_logprob_std": 2.9738510514332015, "tokenization_path_logprob_std": 100.69051065192298, "trajectory_balance_rmse": 2103.461332172085} {"canonical_length": 58, "canonical_nll": 1.3337262909987877, "generation_rouge_l_f1": 0.28346456692913385, "generation_token_f1": 0.42519685039370075, "prompt_id": "e503fe8212044208", "reading_canonical_log_prob": -77.35612487792969, "reading_distances": [31, 31, 34, 34], "reading_log_probs": [-81.17835235595703, -78.55719757080078, -85.20008850097656, -78.59869384765625], "reading_logprob_std": 3.127933093746348, "tokenization_path_logprob_std": 94.79429868083456, "trajectory_balance_rmse": 1387.9979365842698} {"canonical_length": 161, "canonical_nll": 0.9715117460452252, "generation_rouge_l_f1": 0.31123919308357356, "generation_token_f1": 0.5590778097982709, "prompt_id": "cf2e46f2d41b486e", "reading_canonical_log_prob": -156.41339111328125, "reading_distances": [25, 21, 22, 22], "reading_log_probs": [-155.9801483154297, -158.02487182617188, -155.86114501953125, -157.59046936035156], "reading_logprob_std": 1.1048820359120712, "tokenization_path_logprob_std": 134.24489042560288, "trajectory_balance_rmse": 2790.8406239191318} {"canonical_length": 62, "canonical_nll": 0.7173467451526273, "generation_rouge_l_f1": 0.2320441988950276, "generation_token_f1": 0.2983425414364641, "prompt_id": "45e020ec1f1a052d", "reading_canonical_log_prob": -44.47549819946289, "reading_distances": [12, 12, 12, 18], "reading_log_probs": [-44.820281982421875, -44.48395919799805, -44.968204498291016, -45.80682373046875], "reading_logprob_std": 0.5624355522610397, "tokenization_path_logprob_std": 47.40378845107957, "trajectory_balance_rmse": 1012.9895496322526} {"canonical_length": 284, "canonical_nll": 1.642393300231074, "generation_rouge_l_f1": 0.20363636363636362, "generation_token_f1": 0.40727272727272723, "prompt_id": "238ad0e8d97d1bdf", "reading_canonical_log_prob": -466.439697265625, "reading_distances": [13, 19, 17, 26], "reading_log_probs": [-444.685546875, -443.84613037109375, -444.0928649902344, -443.11480712890625], "reading_logprob_std": 0.650347603880035, "tokenization_path_logprob_std": 345.78811988677097, "trajectory_balance_rmse": 5312.602317753727} {"canonical_length": 30, "canonical_nll": 0.6689069112141927, "generation_rouge_l_f1": 0.06250000000000001, "generation_token_f1": 0.0703125, "prompt_id": "aca807bc21ebbe28", "reading_canonical_log_prob": -20.06720733642578, "reading_distances": [9, 11, 9, 11], "reading_log_probs": [-19.689437866210938, -19.626689910888672, -17.658327102661133, -20.119449615478516], "reading_logprob_std": 1.0988117308631848, "tokenization_path_logprob_std": 20.11795048412405, "trajectory_balance_rmse": 392.8607071103301} {"canonical_length": 39, "canonical_nll": 0.9804421449318911, "generation_rouge_l_f1": 0.17985611510791366, "generation_token_f1": 0.2661870503597122, "prompt_id": "1c53f5181f02d61c", "reading_canonical_log_prob": -38.23724365234375, "reading_distances": [21, 21, 21, 22], "reading_log_probs": [-40.81343078613281, -39.06655502319336, -40.051185607910156, -40.56713104248047], "reading_logprob_std": 0.7735568108885859, "tokenization_path_logprob_std": 39.05264001417716, "trajectory_balance_rmse": 483.015511085845} {"canonical_length": 10, "canonical_nll": 3.0090072631835936, "generation_rouge_l_f1": 0.36363636363636365, "generation_token_f1": 0.36363636363636365, "prompt_id": "1a83cc65795284e0", "reading_canonical_log_prob": -30.090072631835938, "reading_distances": [21, 14, 22, 17], "reading_log_probs": [-20.93377685546875, -21.332340240478516, -20.994426727294922, -22.1655216217041], "reading_logprob_std": 0.5671247919767054, "tokenization_path_logprob_std": 62.141842461402355, "trajectory_balance_rmse": 134.9029156958154} {"canonical_length": 30, "canonical_nll": 1.6751057942708334, "generation_rouge_l_f1": 0.43636363636363634, "generation_token_f1": 0.4727272727272727, "prompt_id": "a5ddf34551bf57f2", "reading_canonical_log_prob": -50.253173828125, "reading_distances": [24, 20, 18, 17], "reading_log_probs": [-50.963134765625, -52.13035583496094, -50.78300857543945, -51.55662155151367], "reading_logprob_std": 0.611696393687953, "tokenization_path_logprob_std": 91.66454245659827, "trajectory_balance_rmse": 800.7211387474018} {"canonical_length": 33, "canonical_nll": 1.0540815411191997, "generation_rouge_l_f1": 0.125, "generation_token_f1": 0.19642857142857142, "prompt_id": "626308b96b881737", "reading_canonical_log_prob": -34.784690856933594, "reading_distances": [22, 18, 14, 14], "reading_log_probs": [-35.025970458984375, -35.237823486328125, -36.344173431396484, -35.474212646484375], "reading_logprob_std": 0.5788050653143545, "tokenization_path_logprob_std": 38.42053710634693, "trajectory_balance_rmse": 682.6947687595275} {"canonical_length": 79, "canonical_nll": 0.7891548253312896, "generation_rouge_l_f1": 0.191044776119403, "generation_token_f1": 0.26865671641791045, "prompt_id": "b884d1bbe7793313", "reading_canonical_log_prob": -62.343231201171875, "reading_distances": [19, 23, 24, 21], "reading_log_probs": [-62.19643020629883, -63.256980895996094, -62.525535583496094, -63.38025665283203], "reading_logprob_std": 0.5712034396778287, "tokenization_path_logprob_std": 50.07322841857989, "trajectory_balance_rmse": 927.3215337661776} {"canonical_length": 50, "canonical_nll": 1.3095436096191406, "generation_rouge_l_f1": 0.3619047619047619, "generation_token_f1": 0.5142857142857142, "prompt_id": "d8395ff4ce8fbfae", "reading_canonical_log_prob": -65.47718048095703, "reading_distances": [38, 31, 29, 29], "reading_log_probs": [-69.57872009277344, -70.40005493164062, -69.76695251464844, -68.07981872558594], "reading_logprob_std": 0.9826606236481673, "tokenization_path_logprob_std": 145.60758748033382, "trajectory_balance_rmse": 1060.8802081145575} {"canonical_length": 228, "canonical_nll": 1.4143560309159129, "generation_rouge_l_f1": 0.26229508196721313, "generation_token_f1": 0.5245901639344263, "prompt_id": "cabaeed82d3c2f7b", "reading_canonical_log_prob": -322.4731750488281, "reading_distances": [24, 21, 27, 23], "reading_log_probs": [-320.1527404785156, -317.585205078125, -318.1007385253906, -318.9789733886719], "reading_logprob_std": 1.1239936440695528, "tokenization_path_logprob_std": 206.48085551553828, "trajectory_balance_rmse": 5111.369341412478} {"canonical_length": 60, "canonical_nll": 0.031255032618840536, "generation_rouge_l_f1": 0.6486486486486486, "generation_token_f1": 0.954954954954955, "prompt_id": "ef7598410b16ef36", "reading_canonical_log_prob": -1.8753019571304321, "reading_distances": [32, 36, 31, 32], "reading_log_probs": [-2.003511428833008, -3.2306859493255615, -2.072591543197632, -2.6956045627593994], "reading_logprob_std": 0.5777371257957385, "tokenization_path_logprob_std": 126.1622064607502, "trajectory_balance_rmse": 1020.5714737685901} {"canonical_length": 239, "canonical_nll": 1.217131291473261, "generation_rouge_l_f1": 0.25619834710743805, "generation_token_f1": 0.5330578512396694, "prompt_id": "39157986b3a30c4b", "reading_canonical_log_prob": -290.8943786621094, "reading_distances": [42, 44, 40, 40], "reading_log_probs": [-286.35601806640625, -285.5018615722656, -292.699951171875, -292.74786376953125], "reading_logprob_std": 3.938592170216616, "tokenization_path_logprob_std": 95.48120563987807, "trajectory_balance_rmse": 4114.676092627049} {"canonical_length": 113, "canonical_nll": 1.8041745076137305, "generation_rouge_l_f1": 0.28070175438596495, "generation_token_f1": 0.42105263157894735, "prompt_id": "952e6be856268bb7", "reading_canonical_log_prob": -203.87171936035156, "reading_distances": [25, 27, 29, 33], "reading_log_probs": [-199.13438415527344, -200.0362091064453, -196.95545959472656, -199.8826141357422], "reading_logprob_std": 1.4202021940188603, "tokenization_path_logprob_std": 55.55559328052776, "trajectory_balance_rmse": 2237.1042541391766} {"canonical_length": 324, "canonical_nll": 0.9603349662121431, "generation_rouge_l_f1": 0.22262773722627738, "generation_token_f1": 0.4233576642335766, "prompt_id": "b30cec3b9127ceb8", "reading_canonical_log_prob": -311.1485290527344, "reading_distances": [24, 19, 21, 20], "reading_log_probs": [-308.2913513183594, -307.5998229980469, -305.849365234375, -305.9933166503906], "reading_logprob_std": 1.2037496218409127, "tokenization_path_logprob_std": 264.21725143809994, "trajectory_balance_rmse": 6367.25830804516} {"canonical_length": 211, "canonical_nll": 1.2609076477340047, "generation_rouge_l_f1": 0.2645739910313901, "generation_token_f1": 0.4215246636771301, "prompt_id": "ce18c132bb5bd7f2", "reading_canonical_log_prob": -266.051513671875, "reading_distances": [25, 28, 28, 24], "reading_log_probs": [-265.1300964355469, -264.51971435546875, -264.55426025390625, -264.36199951171875], "reading_logprob_std": 0.3362982381985781, "tokenization_path_logprob_std": 452.99302722121996, "trajectory_balance_rmse": 4277.417166074767} {"canonical_length": 53, "canonical_nll": 1.4360619311062794, "generation_rouge_l_f1": 0.08668730650154799, "generation_token_f1": 0.14860681114551086, "prompt_id": "f1a703b446eae2b2", "reading_canonical_log_prob": -76.11128234863281, "reading_distances": [54, 59, 45, 56], "reading_log_probs": [-74.43734741210938, -73.40818786621094, -75.79168701171875, -75.17906951904297], "reading_logprob_std": 1.0261557306765507, "tokenization_path_logprob_std": 95.5712841532283, "trajectory_balance_rmse": 934.7667123078172} {"canonical_length": 56, "canonical_nll": 1.7216089793613978, "generation_rouge_l_f1": 0.21359223300970875, "generation_token_f1": 0.4271844660194175, "prompt_id": "7188e613a859c6bb", "reading_canonical_log_prob": -96.41010284423828, "reading_distances": [26, 25, 28, 24], "reading_log_probs": [-97.38328552246094, -94.25, -94.7089614868164, -98.5055923461914], "reading_logprob_std": 2.060828667574921, "tokenization_path_logprob_std": 74.83296487883635, "trajectory_balance_rmse": 973.841848017175} {"canonical_length": 198, "canonical_nll": 1.584247011126894, "generation_rouge_l_f1": 0.26067415730337073, "generation_token_f1": 0.49438202247191015, "prompt_id": "85df6700a3ed9560", "reading_canonical_log_prob": -313.680908203125, "reading_distances": [40, 30, 32, 41], "reading_log_probs": [-312.081298828125, -311.2445983886719, -310.77001953125, -314.40399169921875], "reading_logprob_std": 1.6131534359093957, "tokenization_path_logprob_std": 98.46914359340617, "trajectory_balance_rmse": 3770.2226847871525} {"canonical_length": 94, "canonical_nll": 1.3051364979845412, "generation_rouge_l_f1": 0.1935483870967742, "generation_token_f1": 0.3225806451612903, "prompt_id": "0baa5b0629968fb8", "reading_canonical_log_prob": -122.68283081054688, "reading_distances": [26, 24, 29, 21], "reading_log_probs": [-125.85734558105469, -123.87383270263672, -124.28184509277344, -124.50628662109375], "reading_logprob_std": 0.8592017375557875, "tokenization_path_logprob_std": 67.6922300408481, "trajectory_balance_rmse": 2220.8623671046785} {"canonical_length": 331, "canonical_nll": 1.1761810210536254, "generation_rouge_l_f1": 0.2068965517241379, "generation_token_f1": 0.46823956442831216, "prompt_id": "3d704d3c66ac7a28", "reading_canonical_log_prob": -389.31591796875, "reading_distances": [25, 32, 31, 32], "reading_log_probs": [-385.85015869140625, -383.11273193359375, -384.3275146484375, -384.8234558105469], "reading_logprob_std": 1.1370392768529771, "tokenization_path_logprob_std": 159.05542276291416, "trajectory_balance_rmse": 5674.5965281347} {"canonical_length": 14, "canonical_nll": 0.5220974172864642, "generation_rouge_l_f1": 0.6, "generation_token_f1": 0.6, "prompt_id": "d990b55cc5500b31", "reading_canonical_log_prob": -7.309363842010498, "reading_distances": [59, 60, 60, 60], "reading_log_probs": [-7.843471527099609, -7.10836124420166, -7.489202499389648, -7.949077606201172], "reading_logprob_std": 0.380829740261515, "tokenization_path_logprob_std": 43.087279664390316, "trajectory_balance_rmse": 304.1573095686968} {"canonical_length": 125, "canonical_nll": 1.570748779296875, "generation_rouge_l_f1": 0.39175257731958757, "generation_token_f1": 0.5257731958762887, "prompt_id": "96948513433b0e29", "reading_canonical_log_prob": -196.34359741210938, "reading_distances": [14, 19, 14, 18], "reading_log_probs": [-196.55162048339844, -196.10183715820312, -194.84373474121094, -194.5948486328125], "reading_logprob_std": 0.9514863471384708, "tokenization_path_logprob_std": 160.3426358664189, "trajectory_balance_rmse": 2730.3056031539822} {"canonical_length": 88, "canonical_nll": 1.821729139848189, "generation_rouge_l_f1": 0.26595744680851063, "generation_token_f1": 0.4468085106382979, "prompt_id": "5b5a0a9c17c96287", "reading_canonical_log_prob": -160.31216430664062, "reading_distances": [43, 45, 35, 38], "reading_log_probs": [-159.40768432617188, -158.95071411132812, -158.66226196289062, -156.79293823242188], "reading_logprob_std": 1.1487292256567865, "tokenization_path_logprob_std": 137.65444937873153, "trajectory_balance_rmse": 2266.0408756193287} {"canonical_length": 310, "canonical_nll": 1.2529508529170867, "generation_rouge_l_f1": 0.273224043715847, "generation_token_f1": 0.5500910746812387, "prompt_id": "9ae64366f18d52ad", "reading_canonical_log_prob": -388.4147644042969, "reading_distances": [20, 27, 15, 21], "reading_log_probs": [-391.35076904296875, -386.94903564453125, -388.9758605957031, -389.6929016113281], "reading_logprob_std": 1.8237995848693647, "tokenization_path_logprob_std": 173.30909541619891, "trajectory_balance_rmse": 6416.134362600891} {"canonical_length": 192, "canonical_nll": 1.3868261973063152, "generation_rouge_l_f1": 0.23498694516971277, "generation_token_f1": 0.5065274151436031, "prompt_id": "eaabe2e9ef0c1d39", "reading_canonical_log_prob": -266.2706298828125, "reading_distances": [28, 29, 31, 35], "reading_log_probs": [-275.4154357910156, -269.0404052734375, -269.90386962890625, -272.50482177734375], "reading_logprob_std": 2.8723212749753073, "tokenization_path_logprob_std": 293.8534707210016, "trajectory_balance_rmse": 3972.8345155354427} {"canonical_length": 208, "canonical_nll": 1.4159892155573919, "generation_rouge_l_f1": 0.2716627634660421, "generation_token_f1": 0.46370023419203743, "prompt_id": "74426ee6c2f5003e", "reading_canonical_log_prob": -294.5257568359375, "reading_distances": [20, 21, 28, 22], "reading_log_probs": [-291.3486328125, -289.33892822265625, -292.13397216796875, -291.2860107421875], "reading_logprob_std": 1.1896088831727611, "tokenization_path_logprob_std": 136.96934757686887, "trajectory_balance_rmse": 4410.077174318724} {"canonical_length": 341, "canonical_nll": 1.1798487688439332, "generation_rouge_l_f1": 0.28955223880597014, "generation_token_f1": 0.5313432835820896, "prompt_id": "6909a2a1a6b2d043", "reading_canonical_log_prob": -402.32843017578125, "reading_distances": [36, 44, 34, 33], "reading_log_probs": [-408.5203857421875, -410.556396484375, -407.57208251953125, -411.59429931640625], "reading_logprob_std": 1.8406331224321681, "tokenization_path_logprob_std": 334.73106417932576, "trajectory_balance_rmse": 5282.47821572713} {"canonical_length": 19, "canonical_nll": 1.4625473022460938, "generation_rouge_l_f1": 0.3225806451612903, "generation_token_f1": 0.3870967741935484, "prompt_id": "fcaeb274bce92488", "reading_canonical_log_prob": -27.78839874267578, "reading_distances": [24, 30, 32, 34], "reading_log_probs": [-32.28585433959961, -34.168556213378906, -23.442581176757812, -32.951324462890625], "reading_logprob_std": 4.908625822343717, "tokenization_path_logprob_std": 25.43824538730262, "trajectory_balance_rmse": 324.5895391347049} {"canonical_length": 20, "canonical_nll": 0.6021696090698242, "generation_rouge_l_f1": 0.10091743119266056, "generation_token_f1": 0.11926605504587157, "prompt_id": "59d62ad713859ec4", "reading_canonical_log_prob": -12.043392181396484, "reading_distances": [13, 15, 14, 16], "reading_log_probs": [-15.737308502197266, -15.264395713806152, -13.019241333007812, -12.729918479919434], "reading_logprob_std": 1.533077817762168, "tokenization_path_logprob_std": 43.95990308056927, "trajectory_balance_rmse": 187.57385648204175} {"canonical_length": 227, "canonical_nll": 1.3314213017535104, "generation_rouge_l_f1": 0.19921875, "generation_token_f1": 0.37499999999999994, "prompt_id": "0205b5c76a68ce56", "reading_canonical_log_prob": -302.2326354980469, "reading_distances": [32, 25, 25, 26], "reading_log_probs": [-306.8705749511719, -304.95562744140625, -311.41143798828125, -307.2349853515625], "reading_logprob_std": 2.7192979325760103, "tokenization_path_logprob_std": 124.31782398998675, "trajectory_balance_rmse": 3788.044036002809} {"canonical_length": 251, "canonical_nll": 1.553826442277764, "generation_rouge_l_f1": 0.21497120921305185, "generation_token_f1": 0.4721689059500959, "prompt_id": "28e138f0fd885003", "reading_canonical_log_prob": -390.01043701171875, "reading_distances": [23, 25, 29, 28], "reading_log_probs": [-395.5797424316406, -394.4797058105469, -395.35198974609375, -391.9327087402344], "reading_logprob_std": 1.6708870532643674, "tokenization_path_logprob_std": 167.15508974529268, "trajectory_balance_rmse": 4448.611560942535} {"canonical_length": 25, "canonical_nll": 0.3572559356689453, "generation_rouge_l_f1": 0.8800000000000001, "generation_token_f1": 0.8800000000000001, "prompt_id": "3cdea1f3e20549fa", "reading_canonical_log_prob": -8.931398391723633, "reading_distances": [31, 36, 33, 37], "reading_log_probs": [-8.401426315307617, -6.950188636779785, -9.247598648071289, -6.988340854644775], "reading_logprob_std": 1.1255629356133745, "tokenization_path_logprob_std": 41.74084723569413, "trajectory_balance_rmse": 416.9816012536851} {"canonical_length": 143, "canonical_nll": 2.2010540728802446, "generation_rouge_l_f1": 0.18886198547215496, "generation_token_f1": 0.33898305084745756, "prompt_id": "b7ff5b60544bcb0d", "reading_canonical_log_prob": -314.750732421875, "reading_distances": [17, 16, 16, 18], "reading_log_probs": [-320.91033935546875, -324.4774475097656, -326.8658142089844, -324.5500793457031], "reading_logprob_std": 2.458181023683374, "tokenization_path_logprob_std": 142.48823514664434, "trajectory_balance_rmse": 2769.2989978245582} {"canonical_length": 297, "canonical_nll": 1.2874484592013888, "generation_rouge_l_f1": 0.24184261036468327, "generation_token_f1": 0.48368522072936654, "prompt_id": "2c96188d7d0e9844", "reading_canonical_log_prob": -382.3721923828125, "reading_distances": [20, 19, 18, 15], "reading_log_probs": [-380.2451171875, -379.89617919921875, -378.30328369140625, -381.96875], "reading_logprob_std": 1.5036592644013544, "tokenization_path_logprob_std": 268.7040357196264, "trajectory_balance_rmse": 5021.883980170412} {"canonical_length": 101, "canonical_nll": 1.102525012327893, "generation_rouge_l_f1": 0.39688715953307396, "generation_token_f1": 0.5680933852140079, "prompt_id": "f320a0730139386c", "reading_canonical_log_prob": -111.35502624511719, "reading_distances": [35, 41, 38, 37], "reading_log_probs": [-110.11767578125, -110.06453704833984, -108.94326782226562, -110.24638366699219], "reading_logprob_std": 0.6046381337728072, "tokenization_path_logprob_std": 171.392157288032, "trajectory_balance_rmse": 1568.2548154199683} {"canonical_length": 109, "canonical_nll": 2.000144468534977, "generation_rouge_l_f1": 0.16759776536312848, "generation_token_f1": 0.33519553072625696, "prompt_id": "2a4224ce9ad0c001", "reading_canonical_log_prob": -218.0157470703125, "reading_distances": [17, 13, 19, 14], "reading_log_probs": [-218.79054260253906, -227.97584533691406, -232.37277221679688, -222.85153198242188], "reading_logprob_std": 5.927224978174891, "tokenization_path_logprob_std": 127.00140722507247, "trajectory_balance_rmse": 2076.704304977328} {"canonical_length": 20, "canonical_nll": 0.6331688404083252, "generation_rouge_l_f1": 0.5, "generation_token_f1": 0.5833333333333334, "prompt_id": "af538126b4c7deeb", "reading_canonical_log_prob": -12.663376808166504, "reading_distances": [27, 37, 35, 32], "reading_log_probs": [-12.794758796691895, -13.272939682006836, -15.559258460998535, -12.28659725189209], "reading_logprob_std": 1.4445235617915182, "tokenization_path_logprob_std": 42.05047151872859, "trajectory_balance_rmse": 325.4366562678797} {"canonical_length": 48, "canonical_nll": 1.7409907976786296, "generation_rouge_l_f1": 0.36363636363636365, "generation_token_f1": 0.5151515151515151, "prompt_id": "6c3b3c80c0d299e0", "reading_canonical_log_prob": -83.56755828857422, "reading_distances": [29, 23, 19, 32], "reading_log_probs": [-82.14674377441406, -81.4279556274414, -80.60443115234375, -83.36994934082031], "reading_logprob_std": 1.172221519053878, "tokenization_path_logprob_std": 33.19091920973081, "trajectory_balance_rmse": 1083.5951135633734} {"canonical_length": 131, "canonical_nll": 1.315042044370229, "generation_rouge_l_f1": 0.24289405684754523, "generation_token_f1": 0.4444444444444444, "prompt_id": "783057aa1290e5ce", "reading_canonical_log_prob": -172.2705078125, "reading_distances": [35, 27, 24, 27], "reading_log_probs": [-174.11862182617188, -172.48683166503906, -175.82350158691406, -169.98684692382812], "reading_logprob_std": 2.4847995476186733, "tokenization_path_logprob_std": 232.22826279106778, "trajectory_balance_rmse": 2727.9291257903433} {"canonical_length": 358, "canonical_nll": 1.5732761148633905, "generation_rouge_l_f1": 0.20136518771331055, "generation_token_f1": 0.3720136518771331, "prompt_id": "609ea42a956a4ffa", "reading_canonical_log_prob": -563.2328491210938, "reading_distances": [13, 18, 13, 14], "reading_log_probs": [-556.793212890625, -557.9474487304688, -558.7901611328125, -558.1053466796875], "reading_logprob_std": 0.828946345619718, "tokenization_path_logprob_std": 224.83695132978485, "trajectory_balance_rmse": 6393.347048460694} {"canonical_length": 185, "canonical_nll": 0.9529633393158784, "generation_rouge_l_f1": 0.37727272727272726, "generation_token_f1": 0.6409090909090909, "prompt_id": "0fd242df3fc4b686", "reading_canonical_log_prob": -176.2982177734375, "reading_distances": [21, 23, 25, 23], "reading_log_probs": [-173.9859619140625, -173.1856689453125, -171.69509887695312, -175.91915893554688], "reading_logprob_std": 1.759787464641402, "tokenization_path_logprob_std": 127.03528390714567, "trajectory_balance_rmse": 3605.6027533702622} {"canonical_length": 78, "canonical_nll": 0.5936010311811398, "generation_rouge_l_f1": 0.2807017543859649, "generation_token_f1": 0.4350877192982456, "prompt_id": "c7ad7826ab1191ed", "reading_canonical_log_prob": -46.300880432128906, "reading_distances": [33, 30, 29, 24], "reading_log_probs": [-48.06100845336914, -48.33966064453125, -49.59864044189453, -49.32710266113281], "reading_logprob_std": 0.7460316320041195, "tokenization_path_logprob_std": 103.73729505327051, "trajectory_balance_rmse": 1457.979846409068} {"canonical_length": 9, "canonical_nll": 2.061953650580512, "generation_rouge_l_f1": 0.055299539170506916, "generation_token_f1": 0.055299539170506916, "prompt_id": "f4e51cd8f0bc2528", "reading_canonical_log_prob": -18.55758285522461, "reading_distances": [30, 27, 27, 24], "reading_log_probs": [-18.790742874145508, -19.397876739501953, -19.79863929748535, -18.47451400756836], "reading_logprob_std": 0.5951879655973957, "tokenization_path_logprob_std": 47.68150974993933, "trajectory_balance_rmse": 188.57156377670105} {"canonical_length": 110, "canonical_nll": 1.263899647105824, "generation_rouge_l_f1": 0.24169184290030213, "generation_token_f1": 0.3444108761329305, "prompt_id": "67c67cc47e6c2592", "reading_canonical_log_prob": -139.02896118164062, "reading_distances": [20, 31, 25, 28], "reading_log_probs": [-132.93399047851562, -129.69850158691406, -132.47625732421875, -133.546875], "reading_logprob_std": 1.7011188111373905, "tokenization_path_logprob_std": 112.3119112234163, "trajectory_balance_rmse": 1814.7915403706772} {"canonical_length": 18, "canonical_nll": 1.0174950493706598, "generation_rouge_l_f1": 0.5999999999999999, "generation_token_f1": 0.5999999999999999, "prompt_id": "82aeadf42ab79a24", "reading_canonical_log_prob": -18.314910888671875, "reading_distances": [41, 46, 37, 40], "reading_log_probs": [-19.755224227905273, -18.541820526123047, -17.960620880126953, -18.685775756835938], "reading_logprob_std": 0.7483891302768743, "tokenization_path_logprob_std": 29.85889819186783, "trajectory_balance_rmse": 396.8486004331931} {"canonical_length": 12, "canonical_nll": 0.12506441275278726, "generation_rouge_l_f1": 0.7142857142857143, "generation_token_f1": 0.7142857142857143, "prompt_id": "736b52db67801c89", "reading_canonical_log_prob": -1.5007729530334473, "reading_distances": [24, 21, 27, 28], "reading_log_probs": [-1.5107122659683228, -1.626281499862671, -1.6282851696014404, -1.7795929908752441], "reading_logprob_std": 0.1102568835589817, "tokenization_path_logprob_std": 46.80444439185496, "trajectory_balance_rmse": 203.39921043790113} {"canonical_length": 29, "canonical_nll": 1.0574001443797145, "generation_rouge_l_f1": 0.22222222222222224, "generation_token_f1": 0.3703703703703704, "prompt_id": "3c9993234c0b5914", "reading_canonical_log_prob": -30.66460418701172, "reading_distances": [21, 21, 19, 29], "reading_log_probs": [-29.71632957458496, -31.114267349243164, -29.742536544799805, -31.05717658996582], "reading_logprob_std": 0.7834736088284299, "tokenization_path_logprob_std": 40.18928277874893, "trajectory_balance_rmse": 727.9719314735753} {"canonical_length": 221, "canonical_nll": 1.0605372087987839, "generation_rouge_l_f1": 0.23983739837398374, "generation_token_f1": 0.5121951219512195, "prompt_id": "1560e8f9b6710dda", "reading_canonical_log_prob": -234.37872314453125, "reading_distances": [14, 14, 16, 10], "reading_log_probs": [-231.9498291015625, -233.06202697753906, -233.216552734375, -229.71060180664062], "reading_logprob_std": 1.6176972458022565, "tokenization_path_logprob_std": 191.12070541139147, "trajectory_balance_rmse": 4195.729987691512} {"canonical_length": 184, "canonical_nll": 1.8819925059442935, "generation_rouge_l_f1": 0.2151029748283753, "generation_token_f1": 0.3569794050343249, "prompt_id": "1073dfb1d10b700d", "reading_canonical_log_prob": -346.28662109375, "reading_distances": [15, 18, 21, 20], "reading_log_probs": [-342.7598571777344, -343.6996154785156, -346.10015869140625, -342.9791259765625], "reading_logprob_std": 1.530560810040181, "tokenization_path_logprob_std": 86.98587966875179, "trajectory_balance_rmse": 4041.7806908704915} {"canonical_length": 67, "canonical_nll": 1.4178859653757578, "generation_rouge_l_f1": 0.24113475177304966, "generation_token_f1": 0.4822695035460993, "prompt_id": "2045b4674c0f9cd0", "reading_canonical_log_prob": -94.99835968017578, "reading_distances": [29, 22, 24, 21], "reading_log_probs": [-91.78121185302734, -92.69369506835938, -92.22422790527344, -93.29908752441406], "reading_logprob_std": 0.6503241524117731, "tokenization_path_logprob_std": 177.05967476014789, "trajectory_balance_rmse": 1475.7374269958489} {"canonical_length": 18, "canonical_nll": 0.5475238164265951, "generation_rouge_l_f1": 0.5833333333333334, "generation_token_f1": 0.5833333333333334, "prompt_id": "e0eb2633a5eb27ba", "reading_canonical_log_prob": -9.855428695678711, "reading_distances": [17, 23, 24, 27], "reading_log_probs": [-8.035215377807617, -7.743188858032227, -7.88921594619751, -7.82850980758667], "reading_logprob_std": 0.12302144976212813, "tokenization_path_logprob_std": 46.53098222103543, "trajectory_balance_rmse": 480.2696104844319} {"canonical_length": 10, "canonical_nll": 1.4222087860107422, "generation_rouge_l_f1": 0.4, "generation_token_f1": 0.6666666666666667, "prompt_id": "d46e9e081fc93a1e", "reading_canonical_log_prob": -14.222087860107422, "reading_distances": [15, 16, 10, 18], "reading_log_probs": [-14.119085311889648, -14.607093811035156, -14.900962829589844, -14.7317476272583], "reading_logprob_std": 0.33607677933045077, "tokenization_path_logprob_std": 36.44000597523915, "trajectory_balance_rmse": 291.05099041768386} {"canonical_length": 104, "canonical_nll": 1.6589824969951923, "generation_rouge_l_f1": 0.1865284974093264, "generation_token_f1": 0.36269430051813467, "prompt_id": "d0a49199c6983c98", "reading_canonical_log_prob": -172.5341796875, "reading_distances": [26, 23, 34, 26], "reading_log_probs": [-167.1111602783203, -167.001708984375, -167.94097900390625, -165.95118713378906], "reading_logprob_std": 0.8160480180748292, "tokenization_path_logprob_std": 158.4406822516246, "trajectory_balance_rmse": 1588.673499380629} {"canonical_length": 2, "canonical_nll": 2.12251615524292, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "22a76db8343e67a9", "reading_canonical_log_prob": -4.24503231048584, "reading_distances": [28, 20, 26, 22], "reading_log_probs": [-4.606000900268555, -4.82442569732666, -4.035910606384277, -4.251782417297363], "reading_logprob_std": 0.3528999011030878, "tokenization_path_logprob_std": 17.177644029516642, "trajectory_balance_rmse": 35.64552852579182} {"canonical_length": 193, "canonical_nll": 1.2507718733555295, "generation_rouge_l_f1": 0.2332613390928726, "generation_token_f1": 0.5226781857451405, "prompt_id": "ab11d19cfada82a0", "reading_canonical_log_prob": -241.3989715576172, "reading_distances": [23, 24, 17, 23], "reading_log_probs": [-238.72946166992188, -239.8244171142578, -239.74880981445312, -240.78610229492188], "reading_logprob_std": 0.8403520723368589, "tokenization_path_logprob_std": 176.5410079655848, "trajectory_balance_rmse": 4253.956965529933} {"canonical_length": 246, "canonical_nll": 1.333812682609248, "generation_rouge_l_f1": 0.23970037453183518, "generation_token_f1": 0.3895131086142322, "prompt_id": "65ef693d02b228c6", "reading_canonical_log_prob": -328.117919921875, "reading_distances": [76, 81, 86, 85], "reading_log_probs": [-329.95281982421875, -332.0284423828125, -330.2689208984375, -328.53668212890625], "reading_logprob_std": 1.4347628011380573, "tokenization_path_logprob_std": 169.99284621338015, "trajectory_balance_rmse": 5100.384507081214} {"canonical_length": 24, "canonical_nll": 0.5677564938863119, "generation_rouge_l_f1": 0.5714285714285714, "generation_token_f1": 0.6428571428571429, "prompt_id": "4fb4dab7cec52a74", "reading_canonical_log_prob": -13.626155853271484, "reading_distances": [19, 27, 21, 27], "reading_log_probs": [-18.142227172851562, -17.94369888305664, -16.46897315979004, -17.362709045410156], "reading_logprob_std": 0.7504426454124153, "tokenization_path_logprob_std": 90.94740463616218, "trajectory_balance_rmse": 427.5282717329562} {"canonical_length": 36, "canonical_nll": 1.3310446209377713, "generation_rouge_l_f1": 0.15873015873015872, "generation_token_f1": 0.18412698412698417, "prompt_id": "d445e9b3b0bb59b8", "reading_canonical_log_prob": -47.917606353759766, "reading_distances": [49, 43, 38, 42], "reading_log_probs": [-52.61128234863281, -48.749916076660156, -52.33634948730469, -53.000877380371094], "reading_logprob_std": 1.9687622124196815, "tokenization_path_logprob_std": 116.04623952798251, "trajectory_balance_rmse": 830.3118674405607} {"canonical_length": 17, "canonical_nll": 0.49495938244987936, "generation_rouge_l_f1": 0.8148148148148148, "generation_token_f1": 0.8148148148148148, "prompt_id": "9384aa5dbac32997", "reading_canonical_log_prob": -8.41430950164795, "reading_distances": [27, 24, 26, 25], "reading_log_probs": [-8.350696563720703, -8.666142463684082, -8.329366683959961, -8.558084487915039], "reading_logprob_std": 0.1633955110710603, "tokenization_path_logprob_std": 62.903830494715294, "trajectory_balance_rmse": 284.515063307007} {"canonical_length": 227, "canonical_nll": 1.4677250395787445, "generation_rouge_l_f1": 0.17817371937639195, "generation_token_f1": 0.3741648106904231, "prompt_id": "7e391a560a4818b8", "reading_canonical_log_prob": -333.173583984375, "reading_distances": [12, 17, 13, 11], "reading_log_probs": [-333.18701171875, -332.1761474609375, -330.46209716796875, -334.00396728515625], "reading_logprob_std": 1.525837472175075, "tokenization_path_logprob_std": 170.5751021554339, "trajectory_balance_rmse": 4225.823314874846} {"canonical_length": 30, "canonical_nll": 4.548360188802083, "generation_rouge_l_f1": 0.007575757575757575, "generation_token_f1": 0.007575757575757575, "prompt_id": "62ca6c153160a733", "reading_canonical_log_prob": -136.4508056640625, "reading_distances": [47, 42, 48, 45], "reading_log_probs": [-135.3988494873047, -135.44058227539062, -136.3195037841797, -135.79791259765625], "reading_logprob_std": 0.4263060979390977, "tokenization_path_logprob_std": 0.0, "trajectory_balance_rmse": 136.45993041992188} {"canonical_length": 85, "canonical_nll": 0.36797171200022977, "generation_rouge_l_f1": 0.1744186046511628, "generation_token_f1": 0.2558139534883721, "prompt_id": "5fd334d7d6294d39", "reading_canonical_log_prob": -31.27759552001953, "reading_distances": [25, 29, 20, 25], "reading_log_probs": [-40.79457473754883, -36.394989013671875, -34.44814682006836, -39.57476806640625], "reading_logprob_std": 2.9055244065241137, "tokenization_path_logprob_std": 111.83282871099807, "trajectory_balance_rmse": 1515.3359301621533} {"canonical_length": 161, "canonical_nll": 1.7063473529697204, "generation_rouge_l_f1": 0.2685851318944844, "generation_token_f1": 0.4844124700239808, "prompt_id": "ec23e8fead4fc5cf", "reading_canonical_log_prob": -274.721923828125, "reading_distances": [25, 31, 28, 26], "reading_log_probs": [-271.62225341796875, -272.12603759765625, -273.61376953125, -274.18438720703125], "reading_logprob_std": 1.2096906779138774, "tokenization_path_logprob_std": 246.40153904251392, "trajectory_balance_rmse": 2735.0936311406153} {"canonical_length": 296, "canonical_nll": 1.1428918580751162, "generation_rouge_l_f1": 0.31046931407942235, "generation_token_f1": 0.516245487364621, "prompt_id": "6d37aa3b89a989f6", "reading_canonical_log_prob": -338.2959899902344, "reading_distances": [16, 20, 24, 16], "reading_log_probs": [-345.54071044921875, -341.3702392578125, -342.3275146484375, -342.2953186035156], "reading_logprob_std": 1.8262714748752316, "tokenization_path_logprob_std": 97.17723889372004, "trajectory_balance_rmse": 5781.517672713813} {"canonical_length": 19, "canonical_nll": 3.6686770790501644, "generation_rouge_l_f1": 0.07407407407407408, "generation_token_f1": 0.14814814814814817, "prompt_id": "b2ecf966e39675a0", "reading_canonical_log_prob": -69.70486450195312, "reading_distances": [23, 29, 29, 24], "reading_log_probs": [-68.8846664428711, -71.14924621582031, -71.62154388427734, -70.09552764892578], "reading_logprob_std": 1.216111220039178, "tokenization_path_logprob_std": 57.27646036171875, "trajectory_balance_rmse": 422.9386538326041} {"canonical_length": 17, "canonical_nll": 1.1498649821561926, "generation_rouge_l_f1": 0.15254237288135591, "generation_token_f1": 0.20338983050847456, "prompt_id": "efd422672564486e", "reading_canonical_log_prob": -19.547704696655273, "reading_distances": [14, 19, 18, 17], "reading_log_probs": [-22.669612884521484, -22.4221134185791, -21.547361373901367, -23.471139907836914], "reading_logprob_std": 0.7921343304799721, "tokenization_path_logprob_std": 36.739381885951055, "trajectory_balance_rmse": 219.66392168868222} {"canonical_length": 196, "canonical_nll": 1.6057194301060267, "generation_rouge_l_f1": 0.2468513853904282, "generation_token_f1": 0.5037783375314862, "prompt_id": "743badce8c66694d", "reading_canonical_log_prob": -314.72100830078125, "reading_distances": [18, 18, 22, 15], "reading_log_probs": [-311.048583984375, -314.99102783203125, -312.7606201171875, -314.8033447265625], "reading_logprob_std": 1.865360260005865, "tokenization_path_logprob_std": 326.5235110895894, "trajectory_balance_rmse": 4293.015087166765} {"canonical_length": 14, "canonical_nll": 0.19360988480704172, "generation_rouge_l_f1": 0.6923076923076924, "generation_token_f1": 0.6923076923076924, "prompt_id": "685c7dde1bc34a6f", "reading_canonical_log_prob": -2.710538387298584, "reading_distances": [33, 28, 26, 35], "reading_log_probs": [-3.1778831481933594, -2.986227512359619, -3.028813362121582, -2.919481039047241], "reading_logprob_std": 0.10952455472248508, "tokenization_path_logprob_std": 30.342039204239914, "trajectory_balance_rmse": 210.08767011489437} {"canonical_length": 214, "canonical_nll": 1.3353068984557535, "generation_rouge_l_f1": 0.24886877828054302, "generation_token_f1": 0.4796380090497738, "prompt_id": "d0bf7e97f3a575eb", "reading_canonical_log_prob": -285.75567626953125, "reading_distances": [28, 25, 35, 29], "reading_log_probs": [-288.34210205078125, -289.5163269042969, -289.384033203125, -290.8232421875], "reading_logprob_std": 1.0172402449937887, "tokenization_path_logprob_std": 283.6790389982723, "trajectory_balance_rmse": 4340.866187545391} {"canonical_length": 120, "canonical_nll": 0.6901986440022786, "generation_rouge_l_f1": 0.17600000000000002, "generation_token_f1": 0.33066666666666666, "prompt_id": "181490950025f621", "reading_canonical_log_prob": -82.82383728027344, "reading_distances": [31, 28, 22, 22], "reading_log_probs": [-80.08421325683594, -82.45574951171875, -81.98966979980469, -82.13543701171875], "reading_logprob_std": 1.0725183065082893, "tokenization_path_logprob_std": 175.595242369033, "trajectory_balance_rmse": 2106.9256629161137} {"canonical_length": 38, "canonical_nll": 1.7517368918971012, "generation_rouge_l_f1": 0.2884615384615385, "generation_token_f1": 0.46153846153846156, "prompt_id": "70f652d8f02698c0", "reading_canonical_log_prob": -66.56600189208984, "reading_distances": [28, 29, 27, 30], "reading_log_probs": [-66.09146118164062, -65.30194091796875, -65.74775695800781, -65.03376007080078], "reading_logprob_std": 0.46910117452142375, "tokenization_path_logprob_std": 38.587093519756635, "trajectory_balance_rmse": 929.4465675117825} {"canonical_length": 278, "canonical_nll": 1.1361816186699079, "generation_rouge_l_f1": 0.1895910780669145, "generation_token_f1": 0.44609665427509293, "prompt_id": "fe3deab131dcdb79", "reading_canonical_log_prob": -315.8584899902344, "reading_distances": [25, 32, 23, 30], "reading_log_probs": [-314.4703063964844, -314.5343322753906, -315.0379333496094, -313.06610107421875], "reading_logprob_std": 0.8463417199179386, "tokenization_path_logprob_std": 183.52966303425873, "trajectory_balance_rmse": 6165.554098490048} {"canonical_length": 16, "canonical_nll": 1.074468731880188, "generation_rouge_l_f1": 0.19999999999999998, "generation_token_f1": 0.19999999999999998, "prompt_id": "16c98792a8b466f0", "reading_canonical_log_prob": -17.191499710083008, "reading_distances": [33, 33, 34, 31], "reading_log_probs": [-25.847307205200195, -27.60173797607422, -28.4993839263916, -30.00136375427246], "reading_logprob_std": 1.73655884842795, "tokenization_path_logprob_std": 35.81675545955357, "trajectory_balance_rmse": 303.01182294267153} {"canonical_length": 53, "canonical_nll": 1.0548294355284493, "generation_rouge_l_f1": 0.2708333333333333, "generation_token_f1": 0.31249999999999994, "prompt_id": "52f4380d6ba7596b", "reading_canonical_log_prob": -55.90596008300781, "reading_distances": [13, 13, 14, 16], "reading_log_probs": [-57.783660888671875, -57.36845397949219, -57.68788146972656, -56.92588806152344], "reading_logprob_std": 0.38685453391512276, "tokenization_path_logprob_std": 120.0368256930548, "trajectory_balance_rmse": 965.7212165084997} {"canonical_length": 78, "canonical_nll": 0.7742763421474359, "generation_rouge_l_f1": 0.13468013468013468, "generation_token_f1": 0.16835016835016836, "prompt_id": "9b002f163aa12acd", "reading_canonical_log_prob": -60.3935546875, "reading_distances": [18, 19, 21, 20], "reading_log_probs": [-59.75643539428711, -59.86894989013672, -60.357505798339844, -59.31888961791992], "reading_logprob_std": 0.42674818858848507, "tokenization_path_logprob_std": 102.43683622968197, "trajectory_balance_rmse": 1331.1609887316451} {"canonical_length": 22, "canonical_nll": 2.7455955852161753, "generation_rouge_l_f1": 0.2777777777777778, "generation_token_f1": 0.4444444444444445, "prompt_id": "6e455bb516dd4dc1", "reading_canonical_log_prob": -60.40310287475586, "reading_distances": [12, 14, 16, 15], "reading_log_probs": [-60.746280670166016, -60.99793243408203, -58.62807846069336, -60.50572204589844], "reading_logprob_std": 1.0798148767531903, "tokenization_path_logprob_std": 51.091660642192544, "trajectory_balance_rmse": 494.1682947400323} {"canonical_length": 130, "canonical_nll": 1.4715256911057693, "generation_rouge_l_f1": 0.2644628099173554, "generation_token_f1": 0.5537190082644629, "prompt_id": "7c910d294599c2d9", "reading_canonical_log_prob": -191.29833984375, "reading_distances": [14, 14, 13, 6], "reading_log_probs": [-192.28427124023438, -193.21377563476562, -193.62396240234375, -192.90145874023438], "reading_logprob_std": 0.5647623252388577, "tokenization_path_logprob_std": 205.85856331937285, "trajectory_balance_rmse": 2445.581504010341} {"canonical_length": 22, "canonical_nll": 1.4232817563143643, "generation_rouge_l_f1": 0.0966542750929368, "generation_token_f1": 0.11152416356877323, "prompt_id": "e55841df337a493f", "reading_canonical_log_prob": -31.312198638916016, "reading_distances": [19, 21, 21, 24], "reading_log_probs": [-33.16242980957031, -31.589529037475586, -31.94015121459961, -35.074485778808594], "reading_logprob_std": 1.573633068002411, "tokenization_path_logprob_std": 31.08587972692601, "trajectory_balance_rmse": 476.03297655704273} {"canonical_length": 440, "canonical_nll": 2.101877940784801, "generation_rouge_l_f1": 0.23906705539358603, "generation_token_f1": 0.45772594752186585, "prompt_id": "58ceb01ab8ce3495", "reading_canonical_log_prob": -924.8262939453125, "reading_distances": [80, 81, 74, 88], "reading_log_probs": [-928.4237060546875, -934.1204833984375, -933.9678955078125, -947.5465698242188], "reading_logprob_std": 8.131926296560984, "tokenization_path_logprob_std": 99.04567049113214, "trajectory_balance_rmse": 7487.568338817783} {"canonical_length": 32, "canonical_nll": 1.3395122289657593, "generation_rouge_l_f1": 0.1504702194357367, "generation_token_f1": 0.17554858934169282, "prompt_id": "a8059315f9c843b5", "reading_canonical_log_prob": -42.8643913269043, "reading_distances": [23, 31, 21, 20], "reading_log_probs": [-42.742279052734375, -43.19984436035156, -41.56172180175781, -41.74530029296875], "reading_logprob_std": 0.7868657818819075, "tokenization_path_logprob_std": 97.69696626412993, "trajectory_balance_rmse": 804.0425775348331} {"canonical_length": 12, "canonical_nll": 1.2106920878092449, "generation_rouge_l_f1": 0.045112781954887216, "generation_token_f1": 0.05263157894736842, "prompt_id": "3715d1d4c18fe319", "reading_canonical_log_prob": -14.528305053710938, "reading_distances": [13, 9, 11, 19], "reading_log_probs": [-17.538782119750977, -14.914619445800781, -15.762765884399414, -16.007068634033203], "reading_logprob_std": 1.0938879838132163, "tokenization_path_logprob_std": 63.29844538827275, "trajectory_balance_rmse": 234.63884418242014} {"canonical_length": 264, "canonical_nll": 1.9156057184392756, "generation_rouge_l_f1": 0.2549019607843137, "generation_token_f1": 0.48235294117647065, "prompt_id": "996c63537529bc50", "reading_canonical_log_prob": -505.71990966796875, "reading_distances": [31, 19, 25, 22], "reading_log_probs": [-487.857666015625, -500.2733459472656, -500.98480224609375, -496.4087829589844], "reading_logprob_std": 6.027529789977327, "tokenization_path_logprob_std": 95.02395165862139, "trajectory_balance_rmse": 5698.412063922453} {"canonical_length": 15, "canonical_nll": 0.6263481140136719, "generation_rouge_l_f1": 0.8148148148148148, "generation_token_f1": 0.888888888888889, "prompt_id": "a9d8f6ac4429b707", "reading_canonical_log_prob": -9.395221710205078, "reading_distances": [23, 23, 20, 23], "reading_log_probs": [-10.169934272766113, -9.643800735473633, -10.396140098571777, -9.807717323303223], "reading_logprob_std": 0.3413589533831272, "tokenization_path_logprob_std": 64.17631698912909, "trajectory_balance_rmse": 203.98009277623038} {"canonical_length": 17, "canonical_nll": 0.9713230133056641, "generation_rouge_l_f1": 0.5925925925925927, "generation_token_f1": 0.6666666666666667, "prompt_id": "b1822d4ae9e3de29", "reading_canonical_log_prob": -16.51249122619629, "reading_distances": [21, 22, 25, 16], "reading_log_probs": [-16.367429733276367, -15.71788501739502, -17.634376525878906, -16.69027328491211], "reading_logprob_std": 0.7979713887178622, "tokenization_path_logprob_std": 43.61275251643171, "trajectory_balance_rmse": 410.1711117988188} {"canonical_length": 14, "canonical_nll": 0.8405520575387138, "generation_rouge_l_f1": 0.171875, "generation_token_f1": 0.18749999999999997, "prompt_id": "92e5aa90e47873e8", "reading_canonical_log_prob": -11.767728805541992, "reading_distances": [27, 17, 21, 21], "reading_log_probs": [-12.148247718811035, -11.976567268371582, -11.460240364074707, -12.459383964538574], "reading_logprob_std": 0.4180936327732207, "tokenization_path_logprob_std": 57.008756051386506, "trajectory_balance_rmse": 291.5560536364316} {"canonical_length": 176, "canonical_nll": 1.7769473682750354, "generation_rouge_l_f1": 0.19631901840490798, "generation_token_f1": 0.3619631901840491, "prompt_id": "d3105cc3dd2ecc61", "reading_canonical_log_prob": -312.74273681640625, "reading_distances": [31, 31, 33, 35], "reading_log_probs": [-313.65606689453125, -317.8315124511719, -311.2886962890625, -316.9817199707031], "reading_logprob_std": 3.0281989530893982, "tokenization_path_logprob_std": 124.00504003290759, "trajectory_balance_rmse": 3541.452331817741} {"canonical_length": 46, "canonical_nll": 1.0990807906441067, "generation_rouge_l_f1": 0.1716738197424893, "generation_token_f1": 0.20600858369098712, "prompt_id": "f2effbcb126018c5", "reading_canonical_log_prob": -50.557716369628906, "reading_distances": [23, 26, 29, 23], "reading_log_probs": [-42.146095275878906, -43.45587921142578, -40.98987579345703, -41.608463287353516], "reading_logprob_std": 1.0495314250491665, "tokenization_path_logprob_std": 44.20962441655698, "trajectory_balance_rmse": 451.3042093410097} {"canonical_length": 226, "canonical_nll": 1.3341815239560288, "generation_rouge_l_f1": 0.2550607287449393, "generation_token_f1": 0.4939271255060728, "prompt_id": "e73472036e75a1c6", "reading_canonical_log_prob": -301.5250244140625, "reading_distances": [41, 37, 54, 44], "reading_log_probs": [-298.5244140625, -298.402587890625, -296.1688537597656, -299.39208984375], "reading_logprob_std": 1.3745999525026296, "tokenization_path_logprob_std": 156.36771596704435, "trajectory_balance_rmse": 3969.041261110897} {"canonical_length": 8, "canonical_nll": 0.603020429611206, "generation_rouge_l_f1": 0.35, "generation_token_f1": 0.35, "prompt_id": "dea67577bfdb2da1", "reading_canonical_log_prob": -4.824163436889648, "reading_distances": [22, 26, 30, 27], "reading_log_probs": [-5.656219005584717, -6.036113262176514, -5.911689758300781, -5.709631443023682], "reading_logprob_std": 0.17685610664336573, "tokenization_path_logprob_std": 14.115144591333518, "trajectory_balance_rmse": 125.04400725627431} {"canonical_length": 107, "canonical_nll": 1.3798448794356017, "generation_rouge_l_f1": 0.1694915254237288, "generation_token_f1": 0.2711864406779661, "prompt_id": "90fd3570044a7e38", "reading_canonical_log_prob": -147.64340209960938, "reading_distances": [24, 25, 24, 18], "reading_log_probs": [-148.6017303466797, -150.05799865722656, -147.7899932861328, -147.57403564453125], "reading_logprob_std": 1.1253276990617325, "tokenization_path_logprob_std": 97.39739599426066, "trajectory_balance_rmse": 2084.141687839076} {"canonical_length": 222, "canonical_nll": 0.6516268618471988, "generation_rouge_l_f1": 0.29064039408867, "generation_token_f1": 0.5123152709359605, "prompt_id": "1c7402f9ee727b52", "reading_canonical_log_prob": -144.66116333007812, "reading_distances": [78, 90, 86, 86], "reading_log_probs": [-134.6470947265625, -134.2178955078125, -134.03306579589844, -134.38519287109375], "reading_logprob_std": 0.26076489829901306, "tokenization_path_logprob_std": 60.93080288867627, "trajectory_balance_rmse": 2343.971504764199} {"canonical_length": 19, "canonical_nll": 1.703636370207134, "generation_rouge_l_f1": 0.29411764705882354, "generation_token_f1": 0.4117647058823529, "prompt_id": "219202257fbe0a6c", "reading_canonical_log_prob": -32.36909103393555, "reading_distances": [21, 22, 20, 19], "reading_log_probs": [-29.674270629882812, -28.43879508972168, -27.275142669677734, -31.62438201904297], "reading_logprob_std": 1.8597285457585397, "tokenization_path_logprob_std": 43.49102023038074, "trajectory_balance_rmse": 415.44255624998976} {"canonical_length": 47, "canonical_nll": 1.5989553573283743, "generation_rouge_l_f1": 0.26, "generation_token_f1": 0.37999999999999995, "prompt_id": "1eb21676f35636bf", "reading_canonical_log_prob": -75.1509017944336, "reading_distances": [20, 29, 28, 23], "reading_log_probs": [-76.2965316772461, -81.93026733398438, -74.54409790039062, -77.09007263183594], "reading_logprob_std": 3.161021457120283, "tokenization_path_logprob_std": 26.151596786542605, "trajectory_balance_rmse": 988.4608001815254} {"canonical_length": 15, "canonical_nll": 0.13934728304545085, "generation_rouge_l_f1": 0.962962962962963, "generation_token_f1": 0.962962962962963, "prompt_id": "5f8f43706c38c575", "reading_canonical_log_prob": -2.0902092456817627, "reading_distances": [29, 31, 29, 33], "reading_log_probs": [-3.5490217208862305, -3.1343154907226562, -3.537341594696045, -3.2280642986297607], "reading_logprob_std": 0.2125250328227759, "tokenization_path_logprob_std": 38.13259719630578, "trajectory_balance_rmse": 279.7495433258845} {"canonical_length": 18, "canonical_nll": 0.9519704182942709, "generation_rouge_l_f1": 0.7741935483870968, "generation_token_f1": 0.8387096774193549, "prompt_id": "46d2c294deb9298b", "reading_canonical_log_prob": -17.135467529296875, "reading_distances": [73, 64, 59, 69], "reading_log_probs": [-16.916629791259766, -14.425045013427734, -16.894241333007812, -13.34366226196289], "reading_logprob_std": 1.7992480784375062, "tokenization_path_logprob_std": 43.60715388359487, "trajectory_balance_rmse": 389.3223803101039} {"canonical_length": 228, "canonical_nll": 1.2377402322334157, "generation_rouge_l_f1": 0.11607142857142856, "generation_token_f1": 0.11607142857142856, "prompt_id": "7cb804cc64427944", "reading_canonical_log_prob": -282.20477294921875, "reading_distances": [24, 19, 19, 19], "reading_log_probs": [-280.3944091796875, -279.9927978515625, -279.27081298828125, -281.34698486328125], "reading_logprob_std": 0.8658682902752914, "tokenization_path_logprob_std": 110.75501431214127, "trajectory_balance_rmse": 4494.651830518963} {"canonical_length": 171, "canonical_nll": 1.461685269896747, "generation_rouge_l_f1": 0.3206997084548105, "generation_token_f1": 0.5014577259475219, "prompt_id": "d5a4beae5409050e", "reading_canonical_log_prob": -249.94818115234375, "reading_distances": [29, 18, 28, 22], "reading_log_probs": [-249.14889526367188, -251.47557067871094, -250.67132568359375, -255.26724243164062], "reading_logprob_std": 2.6030658268561515, "tokenization_path_logprob_std": 324.8227721093103, "trajectory_balance_rmse": 3814.6856289725215} {"canonical_length": 168, "canonical_nll": 1.8011934189569383, "generation_rouge_l_f1": 0.21247113163972287, "generation_token_f1": 0.4295612009237876, "prompt_id": "ad2f9057c92483ba", "reading_canonical_log_prob": -302.6004943847656, "reading_distances": [13, 14, 14, 11], "reading_log_probs": [-293.43487548828125, -290.89373779296875, -293.27099609375, -291.2198791503906], "reading_logprob_std": 1.3340181142057623, "tokenization_path_logprob_std": 160.8795725035517, "trajectory_balance_rmse": 3383.619507355629} {"canonical_length": 159, "canonical_nll": 1.260937768708235, "generation_rouge_l_f1": 0.2959183673469387, "generation_token_f1": 0.520408163265306, "prompt_id": "240077b10fcaa781", "reading_canonical_log_prob": -200.48910522460938, "reading_distances": [27, 25, 26, 24], "reading_log_probs": [-199.82101440429688, -199.35671997070312, -200.21463012695312, -199.9383544921875], "reading_logprob_std": 0.35764369298215054, "tokenization_path_logprob_std": 127.70358090963227, "trajectory_balance_rmse": 3432.959979527545} {"canonical_length": 327, "canonical_nll": 0.9013037258696484, "generation_rouge_l_f1": 0.355475763016158, "generation_token_f1": 0.5421903052064632, "prompt_id": "dcea5e8dbdd12a1a", "reading_canonical_log_prob": -294.726318359375, "reading_distances": [68, 68, 70, 69], "reading_log_probs": [-296.9514465332031, -296.3368835449219, -293.3078308105469, -293.9591064453125], "reading_logprob_std": 1.7762522990377119, "tokenization_path_logprob_std": 183.24314102312928, "trajectory_balance_rmse": 5274.7991475391955} {"canonical_length": 4, "canonical_nll": 2.33766770362854, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.6666666666666666, "prompt_id": "09f4cd01fd363213", "reading_canonical_log_prob": -9.35067081451416, "reading_distances": [30, 33, 33, 28], "reading_log_probs": [-10.490139961242676, -10.445145606994629, -11.633661270141602, -10.879531860351562], "reading_logprob_std": 0.5500957183582669, "tokenization_path_logprob_std": 17.37553189845042, "trajectory_balance_rmse": 94.23901561551959} {"canonical_length": 139, "canonical_nll": 1.102315340110724, "generation_rouge_l_f1": 0.4314868804664723, "generation_token_f1": 0.565597667638484, "prompt_id": "0eb3ed1212b155f2", "reading_canonical_log_prob": -153.22183227539062, "reading_distances": [26, 19, 22, 22], "reading_log_probs": [-153.56991577148438, -153.9429931640625, -155.599365234375, -155.74954223632812], "reading_logprob_std": 1.1194628364852024, "tokenization_path_logprob_std": 163.41896395768734, "trajectory_balance_rmse": 2523.8575215734313} {"canonical_length": 3, "canonical_nll": 3.7504860560099282, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "423d33be2ee3658e", "reading_canonical_log_prob": -11.251458168029785, "reading_distances": [22, 21, 21, 23], "reading_log_probs": [-10.597761154174805, -9.31564712524414, -10.103910446166992, -9.141087532043457], "reading_logprob_std": 0.6824258827304951, "tokenization_path_logprob_std": 9.648579107732994, "trajectory_balance_rmse": 33.82889820112082} {"canonical_length": 33, "canonical_nll": 1.2057265079382695, "generation_rouge_l_f1": 0.34782608695652173, "generation_token_f1": 0.34782608695652173, "prompt_id": "82dd428e952ac035", "reading_canonical_log_prob": -39.78897476196289, "reading_distances": [36, 31, 30, 32], "reading_log_probs": [-39.87472915649414, -42.6895637512207, -39.287940979003906, -40.470359802246094], "reading_logprob_std": 1.4865061932889392, "tokenization_path_logprob_std": 42.69931462616551, "trajectory_balance_rmse": 683.4480119922988} {"canonical_length": 293, "canonical_nll": 1.1847973716136946, "generation_rouge_l_f1": 0.2254025044722719, "generation_token_f1": 0.47942754919499103, "prompt_id": "a3746b2776781f66", "reading_canonical_log_prob": -347.1456298828125, "reading_distances": [21, 25, 23, 27], "reading_log_probs": [-348.73651123046875, -349.10528564453125, -349.0722351074219, -352.1302795410156], "reading_logprob_std": 1.5882297431758319, "tokenization_path_logprob_std": 234.10068305673772, "trajectory_balance_rmse": 6165.895508688852} {"canonical_length": 64, "canonical_nll": 1.9897284507751465, "generation_rouge_l_f1": 0.22568093385214008, "generation_token_f1": 0.28793774319066145, "prompt_id": "d4e9c302f1022491", "reading_canonical_log_prob": -127.34262084960938, "reading_distances": [30, 25, 32, 28], "reading_log_probs": [-121.80170440673828, -124.32052612304688, -124.34288024902344, -127.57254791259766], "reading_logprob_std": 2.3648743961163308, "tokenization_path_logprob_std": 165.21893418012115, "trajectory_balance_rmse": 1305.8398021059077} {"canonical_length": 99, "canonical_nll": 1.025008615821299, "generation_rouge_l_f1": 0.2840909090909091, "generation_token_f1": 0.44886363636363635, "prompt_id": "ed00a510b9a4ea44", "reading_canonical_log_prob": -101.4758529663086, "reading_distances": [30, 33, 35, 28], "reading_log_probs": [-101.0608139038086, -101.32211303710938, -101.82136535644531, -101.92120361328125], "reading_logprob_std": 0.4087701346403119, "tokenization_path_logprob_std": 235.74392444899885, "trajectory_balance_rmse": 2198.830527956804} {"canonical_length": 275, "canonical_nll": 1.4020518909801136, "generation_rouge_l_f1": 0.22627737226277372, "generation_token_f1": 0.4416058394160584, "prompt_id": "00a5d180fc30eed4", "reading_canonical_log_prob": -385.56427001953125, "reading_distances": [30, 25, 32, 26], "reading_log_probs": [-388.92071533203125, -385.6417236328125, -388.2535095214844, -388.318359375], "reading_logprob_std": 1.4591605116260677, "tokenization_path_logprob_std": 170.50451903277622, "trajectory_balance_rmse": 5575.491636003307} {"canonical_length": 115, "canonical_nll": 1.1890938136888587, "generation_rouge_l_f1": 0.28108108108108104, "generation_token_f1": 0.4108108108108108, "prompt_id": "300b5dc6095f893e", "reading_canonical_log_prob": -136.74578857421875, "reading_distances": [6, 10, 5, 6], "reading_log_probs": [-133.34947204589844, -134.67721557617188, -133.87338256835938, -133.0704345703125], "reading_logprob_std": 0.7063913707072774, "tokenization_path_logprob_std": 95.46507250169452, "trajectory_balance_rmse": 2184.107539136552} {"canonical_length": 6, "canonical_nll": 3.2210779190063477, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "2c288dbc9a356333", "reading_canonical_log_prob": -19.326467514038086, "reading_distances": [19, 21, 19, 22], "reading_log_probs": [-20.031085968017578, -18.665170669555664, -20.027496337890625, -20.775135040283203], "reading_logprob_std": 0.8796870187560697, "tokenization_path_logprob_std": 32.04177891672284, "trajectory_balance_rmse": 141.14209137791934} {"canonical_length": 233, "canonical_nll": 1.0517852521249664, "generation_rouge_l_f1": 0.3073394495412844, "generation_token_f1": 0.518348623853211, "prompt_id": "810a6641955f3c0f", "reading_canonical_log_prob": -245.0659637451172, "reading_distances": [42, 39, 33, 35], "reading_log_probs": [-239.45437622070312, -234.46090698242188, -243.32826232910156, -235.98440551757812], "reading_logprob_std": 3.946155268043827, "tokenization_path_logprob_std": 78.21083126379077, "trajectory_balance_rmse": 2786.2857555813907} {"canonical_length": 167, "canonical_nll": 1.9312835510619386, "generation_rouge_l_f1": 0.2765273311897106, "generation_token_f1": 0.5530546623794212, "prompt_id": "eb867c21f2710dc4", "reading_canonical_log_prob": -322.52435302734375, "reading_distances": [23, 31, 25, 26], "reading_log_probs": [-324.156494140625, -320.4400939941406, -319.668212890625, -322.89520263671875], "reading_logprob_std": 2.093322251667933, "tokenization_path_logprob_std": 147.1964071552932, "trajectory_balance_rmse": 3651.683133430205} {"canonical_length": 48, "canonical_nll": 1.9963006973266602, "generation_rouge_l_f1": 0.14173228346456693, "generation_token_f1": 0.2047244094488189, "prompt_id": "ee8008c4016a6c6e", "reading_canonical_log_prob": -95.82243347167969, "reading_distances": [24, 23, 32, 28], "reading_log_probs": [-90.09809875488281, -90.15037536621094, -89.7467041015625, -90.64048767089844], "reading_logprob_std": 0.3676962528834669, "tokenization_path_logprob_std": 59.847507788382686, "trajectory_balance_rmse": 1124.3606330219138} {"canonical_length": 232, "canonical_nll": 1.2138892864358837, "generation_rouge_l_f1": 0.2515991471215352, "generation_token_f1": 0.4904051172707889, "prompt_id": "5adec8000ecf22fe", "reading_canonical_log_prob": -281.622314453125, "reading_distances": [22, 18, 20, 23], "reading_log_probs": [-285.4549560546875, -287.1239013671875, -285.6300964355469, -282.5120849609375], "reading_logprob_std": 1.9300009105805032, "tokenization_path_logprob_std": 160.38933478572127, "trajectory_balance_rmse": 4506.69740596374} {"canonical_length": 283, "canonical_nll": 1.512600740358602, "generation_rouge_l_f1": 0.23076923076923075, "generation_token_f1": 0.3794871794871795, "prompt_id": "893df9640ff6dc30", "reading_canonical_log_prob": -428.0660095214844, "reading_distances": [22, 23, 24, 19], "reading_log_probs": [-416.6351623535156, -414.66790771484375, -417.95343017578125, -417.760009765625], "reading_logprob_std": 1.5073809227144488, "tokenization_path_logprob_std": 30.09668846348504, "trajectory_balance_rmse": 5408.681803376661} {"canonical_length": 13, "canonical_nll": 0.789787219120906, "generation_rouge_l_f1": 0.1568627450980392, "generation_token_f1": 0.1568627450980392, "prompt_id": "5f407731bdd7e46c", "reading_canonical_log_prob": -10.267233848571777, "reading_distances": [25, 31, 31, 29], "reading_log_probs": [-10.823906898498535, -9.662908554077148, -10.852519989013672, -10.25579833984375], "reading_logprob_std": 0.5623057879467525, "tokenization_path_logprob_std": 40.93924671149354, "trajectory_balance_rmse": 139.89910067779152} {"canonical_length": 226, "canonical_nll": 1.2341268083690542, "generation_rouge_l_f1": 0.2809917355371901, "generation_token_f1": 0.5, "prompt_id": "c371f7349e2c9b37", "reading_canonical_log_prob": -278.91265869140625, "reading_distances": [24, 30, 21, 19], "reading_log_probs": [-277.35577392578125, -274.9281005859375, -275.4907531738281, -276.51739501953125], "reading_logprob_std": 1.079012189122906, "tokenization_path_logprob_std": 208.16237887981035, "trajectory_balance_rmse": 4734.9658099862} {"canonical_length": 9, "canonical_nll": 1.4322820239596896, "generation_rouge_l_f1": 0.35294117647058826, "generation_token_f1": 0.35294117647058826, "prompt_id": "59b66b10bf9d5bcd", "reading_canonical_log_prob": -12.890538215637207, "reading_distances": [64, 50, 67, 58], "reading_log_probs": [-14.82833194732666, -16.800552368164062, -20.012184143066406, -14.332496643066406], "reading_logprob_std": 2.576699397840848, "tokenization_path_logprob_std": 18.693647240833094, "trajectory_balance_rmse": 196.04754293491095} {"canonical_length": 387, "canonical_nll": 1.3791305187136629, "generation_rouge_l_f1": 0.1897590361445783, "generation_token_f1": 0.3825301204819277, "prompt_id": "d9970f787452110a", "reading_canonical_log_prob": -533.7235107421875, "reading_distances": [32, 28, 31, 31], "reading_log_probs": [-523.554931640625, -526.0621337890625, -527.7239990234375, -528.1522827148438], "reading_logprob_std": 2.0840019166792674, "tokenization_path_logprob_std": 339.9614254653972, "trajectory_balance_rmse": 7130.498000818196} {"canonical_length": 199, "canonical_nll": 1.0833875186479272, "generation_rouge_l_f1": 0.3542116630669547, "generation_token_f1": 0.5442764578833693, "prompt_id": "4f2d283e14485973", "reading_canonical_log_prob": -215.5941162109375, "reading_distances": [20, 21, 24, 20], "reading_log_probs": [-217.21588134765625, -216.310546875, -214.7726287841797, -214.24993896484375], "reading_logprob_std": 1.3684062955127274, "tokenization_path_logprob_std": 182.60620430764956, "trajectory_balance_rmse": 4011.422741187991} {"canonical_length": 366, "canonical_nll": 1.359709692783043, "generation_rouge_l_f1": 0.21283783783783786, "generation_token_f1": 0.37162162162162166, "prompt_id": "74c803f89db0a539", "reading_canonical_log_prob": -497.65374755859375, "reading_distances": [12, 15, 17, 12], "reading_log_probs": [-487.4759521484375, -486.34716796875, -485.5465087890625, -488.20556640625], "reading_logprob_std": 1.1794959959164593, "tokenization_path_logprob_std": 51.37862157857594, "trajectory_balance_rmse": 6868.181465488696} {"canonical_length": 132, "canonical_nll": 1.2056410818388967, "generation_rouge_l_f1": 0.25609756097560976, "generation_token_f1": 0.5, "prompt_id": "456727f2988746cd", "reading_canonical_log_prob": -159.14462280273438, "reading_distances": [26, 29, 30, 31], "reading_log_probs": [-161.39007568359375, -161.37179565429688, -162.34329223632812, -163.00601196289062], "reading_logprob_std": 0.7944532712696715, "tokenization_path_logprob_std": 155.34245403970388, "trajectory_balance_rmse": 3228.8224898328035} {"canonical_length": 12, "canonical_nll": 0.5420205593109131, "generation_rouge_l_f1": 0.5714285714285715, "generation_token_f1": 0.5714285714285715, "prompt_id": "4105efabcbebe823", "reading_canonical_log_prob": -6.504246711730957, "reading_distances": [29, 18, 24, 23], "reading_log_probs": [-6.459219455718994, -5.422332763671875, -6.31493616104126, -5.7819013595581055], "reading_logprob_std": 0.48000590138643423, "tokenization_path_logprob_std": 19.181457371411394, "trajectory_balance_rmse": 258.2959183756194} {"canonical_length": 57, "canonical_nll": 1.5382477609734786, "generation_rouge_l_f1": 0.06930693069306931, "generation_token_f1": 0.099009900990099, "prompt_id": "61331d091d23a9af", "reading_canonical_log_prob": -87.68012237548828, "reading_distances": [16, 16, 15, 16], "reading_log_probs": [-89.11447143554688, -90.93453979492188, -92.01475524902344, -88.56280517578125], "reading_logprob_std": 1.6004317264606724, "tokenization_path_logprob_std": 127.66333553132885, "trajectory_balance_rmse": 1187.242881210014} {"canonical_length": 32, "canonical_nll": 2.5753679275512695, "generation_rouge_l_f1": 0.3380281690140845, "generation_token_f1": 0.4225352112676056, "prompt_id": "1a1747f5e15bef6b", "reading_canonical_log_prob": -82.41177368164062, "reading_distances": [21, 38, 30, 33], "reading_log_probs": [-82.45565032958984, -81.47174835205078, -81.4774169921875, -79.51334381103516], "reading_logprob_std": 1.2340696615133981, "tokenization_path_logprob_std": 25.28120454504444, "trajectory_balance_rmse": 867.0869870211486} {"canonical_length": 9, "canonical_nll": 0.15399955378638375, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "8d4efdacc1d27322", "reading_canonical_log_prob": -1.3859959840774536, "reading_distances": [33, 36, 36, 33], "reading_log_probs": [-2.335984945297241, -2.0778024196624756, -1.932105302810669, -1.6542280912399292], "reading_logprob_std": 0.2846676419499619, "tokenization_path_logprob_std": 18.96181483114336, "trajectory_balance_rmse": 236.33087385228967} {"canonical_length": 263, "canonical_nll": 1.6415802117989544, "generation_rouge_l_f1": 0.196, "generation_token_f1": 0.392, "prompt_id": "123b7e61f56166d9", "reading_canonical_log_prob": -431.735595703125, "reading_distances": [34, 36, 32, 39], "reading_log_probs": [-443.7769775390625, -437.0277099609375, -440.51263427734375, -444.3382873535156], "reading_logprob_std": 3.375752746208819, "tokenization_path_logprob_std": 178.27600896555427, "trajectory_balance_rmse": 4658.044974427324} {"canonical_length": 63, "canonical_nll": 1.2289334251767112, "generation_rouge_l_f1": 0.3212435233160622, "generation_token_f1": 0.4145077720207254, "prompt_id": "0407e3355bf3ebea", "reading_canonical_log_prob": -77.42280578613281, "reading_distances": [18, 15, 20, 19], "reading_log_probs": [-79.99678039550781, -83.01678466796875, -83.48365783691406, -83.42697143554688], "reading_logprob_std": 1.66919111539371, "tokenization_path_logprob_std": 17.946496414430346, "trajectory_balance_rmse": 1080.236942826719} {"canonical_length": 29, "canonical_nll": 0.6731823230611866, "generation_rouge_l_f1": 0.07142857142857142, "generation_token_f1": 0.07142857142857142, "prompt_id": "f217efc1d8ea8fee", "reading_canonical_log_prob": -19.522287368774414, "reading_distances": [10, 17, 13, 15], "reading_log_probs": [-19.573558807373047, -19.700410842895508, -18.077590942382812, -18.558555603027344], "reading_logprob_std": 0.7880856351488186, "tokenization_path_logprob_std": 24.64726083556425, "trajectory_balance_rmse": 527.4065369188588} {"canonical_length": 125, "canonical_nll": 1.5328892822265625, "generation_rouge_l_f1": 0.28654970760233917, "generation_token_f1": 0.4444444444444445, "prompt_id": "65c5ecab43401060", "reading_canonical_log_prob": -191.6111602783203, "reading_distances": [31, 17, 21, 32], "reading_log_probs": [-197.31051635742188, -194.68508911132812, -197.3790283203125, -196.29580688476562], "reading_logprob_std": 1.2567247880001384, "tokenization_path_logprob_std": 187.29659460628125, "trajectory_balance_rmse": 2651.3036848408706} {"canonical_length": 41, "canonical_nll": 1.0476384046601086, "generation_rouge_l_f1": 0.4814814814814815, "generation_token_f1": 0.4814814814814815, "prompt_id": "5cd588db0ae30114", "reading_canonical_log_prob": -42.95317459106445, "reading_distances": [26, 32, 31, 25], "reading_log_probs": [-47.88648223876953, -47.35129928588867, -46.08941650390625, -47.93521499633789], "reading_logprob_std": 0.8591914181935938, "tokenization_path_logprob_std": 13.289181609239487, "trajectory_balance_rmse": 762.0616894198876} {"canonical_length": 145, "canonical_nll": 1.7528647587217134, "generation_rouge_l_f1": 0.27246376811594203, "generation_token_f1": 0.4579710144927537, "prompt_id": "f88ab53f7308d0a9", "reading_canonical_log_prob": -254.16539001464844, "reading_distances": [26, 29, 27, 18], "reading_log_probs": [-261.6346740722656, -261.70758056640625, -263.9364929199219, -259.70703125], "reading_logprob_std": 1.729115527249698, "tokenization_path_logprob_std": 214.09415103886116, "trajectory_balance_rmse": 3351.7854150378753} {"canonical_length": 55, "canonical_nll": 1.3344365900213069, "generation_rouge_l_f1": 0.26993865030674846, "generation_token_f1": 0.42944785276073616, "prompt_id": "8c737ded95e1b468", "reading_canonical_log_prob": -73.39401245117188, "reading_distances": [12, 20, 18, 16], "reading_log_probs": [-86.43901062011719, -84.25442504882812, -83.7051010131836, -84.46070861816406], "reading_logprob_std": 1.1928819209488348, "tokenization_path_logprob_std": 110.51958194367872, "trajectory_balance_rmse": 1048.9008230840186} {"canonical_length": 22, "canonical_nll": 0.29621856862848456, "generation_rouge_l_f1": 0.11371237458193979, "generation_token_f1": 0.12040133779264213, "prompt_id": "cd208c8b818d2286", "reading_canonical_log_prob": -6.51680850982666, "reading_distances": [24, 27, 28, 30], "reading_log_probs": [-4.559353828430176, -9.093342781066895, -8.099761962890625, -9.639179229736328], "reading_logprob_std": 2.283110405443769, "tokenization_path_logprob_std": 43.52583483620586, "trajectory_balance_rmse": 315.72514715399103} {"canonical_length": 165, "canonical_nll": 1.0159174138849432, "generation_rouge_l_f1": 0.32547169811320753, "generation_token_f1": 0.4575471698113207, "prompt_id": "a88030476f29d6bf", "reading_canonical_log_prob": -167.62637329101562, "reading_distances": [14, 12, 16, 17], "reading_log_probs": [-169.7622528076172, -168.77725219726562, -168.45736694335938, -168.80548095703125], "reading_logprob_std": 0.5636689705451953, "tokenization_path_logprob_std": 150.75406042833154, "trajectory_balance_rmse": 3774.363467293304} {"canonical_length": 377, "canonical_nll": 1.2772150419118866, "generation_rouge_l_f1": 0.21981424148606812, "generation_token_f1": 0.43962848297213625, "prompt_id": "842add8204336566", "reading_canonical_log_prob": -481.51007080078125, "reading_distances": [35, 31, 24, 24], "reading_log_probs": [-486.2920837402344, -487.36724853515625, -483.7928771972656, -486.9927673339844], "reading_logprob_std": 1.6085354026111949, "tokenization_path_logprob_std": 108.07679260816504, "trajectory_balance_rmse": 7239.619793563703} {"canonical_length": 112, "canonical_nll": 1.1496877670288086, "generation_rouge_l_f1": 0.27237354085603105, "generation_token_f1": 0.443579766536965, "prompt_id": "95ed700077e4b7ce", "reading_canonical_log_prob": -128.76502990722656, "reading_distances": [12, 13, 12, 14], "reading_log_probs": [-125.81501770019531, -124.52552032470703, -123.89372253417969, -126.51737976074219], "reading_logprob_std": 1.1936547785398766, "tokenization_path_logprob_std": 169.19922149401626, "trajectory_balance_rmse": 1518.0255857295058} {"canonical_length": 11, "canonical_nll": 0.7995121695778586, "generation_rouge_l_f1": 0.6, "generation_token_f1": 0.6, "prompt_id": "e5cef34a838082a8", "reading_canonical_log_prob": -8.794633865356445, "reading_distances": [35, 36, 30, 38], "reading_log_probs": [-10.220836639404297, -10.426891326904297, -10.137409210205078, -9.80631160736084], "reading_logprob_std": 0.2581657056976906, "tokenization_path_logprob_std": 34.33743768973001, "trajectory_balance_rmse": 342.43389886777453} {"canonical_length": 10, "canonical_nll": 0.27648110389709474, "generation_rouge_l_f1": 0.8888888888888888, "generation_token_f1": 0.8888888888888888, "prompt_id": "47116aef05367776", "reading_canonical_log_prob": -2.7648110389709473, "reading_distances": [32, 37, 32, 38], "reading_log_probs": [-6.059532642364502, -4.444855690002441, -6.64250373840332, -6.261201858520508], "reading_logprob_std": 0.9687564582229259, "tokenization_path_logprob_std": 30.282248998790486, "trajectory_balance_rmse": 134.08958936030115} {"canonical_length": 375, "canonical_nll": 1.4455887044270834, "generation_rouge_l_f1": 0.18238021638330756, "generation_token_f1": 0.37403400309119006, "prompt_id": "7dea199ced8c5f4e", "reading_canonical_log_prob": -542.0957641601562, "reading_distances": [32, 26, 28, 27], "reading_log_probs": [-530.01513671875, -533.1146240234375, -528.09130859375, -535.1190185546875], "reading_logprob_std": 3.135782048120457, "tokenization_path_logprob_std": 142.45873326893212, "trajectory_balance_rmse": 6728.484567791309} {"canonical_length": 377, "canonical_nll": 1.166606225132626, "generation_rouge_l_f1": 0.24528301886792453, "generation_token_f1": 0.5, "prompt_id": "c009e5e48dec12c7", "reading_canonical_log_prob": -439.810546875, "reading_distances": [27, 29, 32, 26], "reading_log_probs": [-435.10296630859375, -438.38165283203125, -434.08990478515625, -436.33709716796875], "reading_logprob_std": 1.8472578937887416, "tokenization_path_logprob_std": 270.77120270018656, "trajectory_balance_rmse": 6442.177688352532} {"canonical_length": 206, "canonical_nll": 1.0173136146323194, "generation_rouge_l_f1": 0.22589531680440772, "generation_token_f1": 0.4352617079889807, "prompt_id": "12ce96bbad2517ca", "reading_canonical_log_prob": -209.5666046142578, "reading_distances": [21, 23, 23, 20], "reading_log_probs": [-208.73568725585938, -209.91722106933594, -211.50819396972656, -211.38543701171875], "reading_logprob_std": 1.316748251485696, "tokenization_path_logprob_std": 96.40498694113356, "trajectory_balance_rmse": 4184.039964632197} {"canonical_length": 193, "canonical_nll": 1.3473618265260687, "generation_rouge_l_f1": 0.3547671840354767, "generation_token_f1": 0.6385809312638582, "prompt_id": "cda8b27fc8b243fa", "reading_canonical_log_prob": -260.04083251953125, "reading_distances": [22, 12, 18, 13], "reading_log_probs": [-258.87432861328125, -261.1020812988281, -260.2588195800781, -258.3568115234375], "reading_logprob_std": 1.2587279397898812, "tokenization_path_logprob_std": 208.38937749935232, "trajectory_balance_rmse": 3734.368361112731} {"canonical_length": 247, "canonical_nll": 0.9920426959450911, "generation_rouge_l_f1": 0.2751937984496124, "generation_token_f1": 0.5193798449612403, "prompt_id": "1ba77d2f5f7df1e4", "reading_canonical_log_prob": -245.0345458984375, "reading_distances": [17, 17, 17, 19], "reading_log_probs": [-239.5672607421875, -241.87728881835938, -241.47848510742188, -241.74339294433594], "reading_logprob_std": 1.0790315223690097, "tokenization_path_logprob_std": 206.37824596891252, "trajectory_balance_rmse": 5642.982694496449} {"canonical_length": 10, "canonical_nll": 0.9268220901489258, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.6666666666666666, "prompt_id": "78c9a1c087a98da8", "reading_canonical_log_prob": -9.268220901489258, "reading_distances": [16, 21, 21, 21], "reading_log_probs": [-9.099365234375, -8.277695655822754, -8.25010871887207, -8.383316993713379], "reading_logprob_std": 0.40194982678100205, "tokenization_path_logprob_std": 43.18607381428, "trajectory_balance_rmse": 175.5410075633675} {"canonical_length": 128, "canonical_nll": 1.1765756607055664, "generation_rouge_l_f1": 0.3023255813953488, "generation_token_f1": 0.5271317829457365, "prompt_id": "6de4ae8779beb5f7", "reading_canonical_log_prob": -150.6016845703125, "reading_distances": [38, 40, 33, 35], "reading_log_probs": [-150.1439208984375, -151.23411560058594, -149.90023803710938, -149.77462768554688], "reading_logprob_std": 0.6651688752674991, "tokenization_path_logprob_std": 186.69384853139283, "trajectory_balance_rmse": 2535.5774727912144} {"canonical_length": 292, "canonical_nll": 0.8534383904444028, "generation_rouge_l_f1": 0.39784946236559143, "generation_token_f1": 0.6021505376344085, "prompt_id": "bcb2851b6b713edf", "reading_canonical_log_prob": -249.20401000976562, "reading_distances": [46, 47, 43, 50], "reading_log_probs": [-248.2137908935547, -249.9696044921875, -251.51638793945312, -250.49716186523438], "reading_logprob_std": 1.3818311896413893, "tokenization_path_logprob_std": 233.69475896687283, "trajectory_balance_rmse": 4934.130620356133} {"canonical_length": 207, "canonical_nll": 1.5402554866772342, "generation_rouge_l_f1": 0.25053995680345575, "generation_token_f1": 0.5313174946004319, "prompt_id": "790529293961a686", "reading_canonical_log_prob": -318.8328857421875, "reading_distances": [30, 26, 26, 35], "reading_log_probs": [-312.40478515625, -314.906982421875, -312.7676086425781, -313.029052734375], "reading_logprob_std": 1.1163271301946287, "tokenization_path_logprob_std": 140.57059932191785, "trajectory_balance_rmse": 4494.945375814339} {"canonical_length": 130, "canonical_nll": 1.7808701735276442, "generation_rouge_l_f1": 0.2, "generation_token_f1": 0.4, "prompt_id": "ecb22237f8778fd5", "reading_canonical_log_prob": -231.51312255859375, "reading_distances": [19, 21, 20, 14], "reading_log_probs": [-230.31790161132812, -229.90972900390625, -229.65762329101562, -230.79736328125], "reading_logprob_std": 0.49857421520234796, "tokenization_path_logprob_std": 143.16624222014195, "trajectory_balance_rmse": 2422.3898361460233} {"canonical_length": 37, "canonical_nll": 1.3529475444072, "generation_rouge_l_f1": 0.09523809523809522, "generation_token_f1": 0.09523809523809522, "prompt_id": "db3b67bebed71ebb", "reading_canonical_log_prob": -50.059059143066406, "reading_distances": [29, 33, 32, 29], "reading_log_probs": [-49.21493148803711, -49.8493537902832, -49.060447692871094, -50.04836654663086], "reading_logprob_std": 0.47949048189948484, "tokenization_path_logprob_std": 15.785784450720238, "trajectory_balance_rmse": 773.6417773204023} {"canonical_length": 10, "canonical_nll": 0.11959205865859986, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "a75ceffa04e21449", "reading_canonical_log_prob": -1.1959205865859985, "reading_distances": [38, 37, 35, 32], "reading_log_probs": [-1.5386695861816406, -1.2783305644989014, -1.5548796653747559, -1.5941548347473145], "reading_logprob_std": 0.1440152478874479, "tokenization_path_logprob_std": 31.752615626814485, "trajectory_balance_rmse": 142.7758548588324} {"canonical_length": 7, "canonical_nll": 0.013738919581685747, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "7af908712adcd78e", "reading_canonical_log_prob": -0.09617243707180023, "reading_distances": [31, 36, 38, 38], "reading_log_probs": [-0.2845968008041382, -1.352447271347046, -1.3649468421936035, -0.20526699721813202], "reading_logprob_std": 0.6438678837418237, "tokenization_path_logprob_std": 64.12201418545241, "trajectory_balance_rmse": 180.00332823606794} {"canonical_length": 45, "canonical_nll": 0.7000284830729167, "generation_rouge_l_f1": 0.05128205128205128, "generation_token_f1": 0.05128205128205128, "prompt_id": "3c24680236bfa43e", "reading_canonical_log_prob": -31.50128173828125, "reading_distances": [26, 20, 23, 28], "reading_log_probs": [-31.56985092163086, -31.106422424316406, -31.723011016845703, -31.188161849975586], "reading_logprob_std": 0.29676547864102315, "tokenization_path_logprob_std": 134.49350583896603, "trajectory_balance_rmse": 782.8628449719355} {"canonical_length": 22, "canonical_nll": 0.2832221334630793, "generation_rouge_l_f1": 0.7222222222222222, "generation_token_f1": 0.8333333333333333, "prompt_id": "8be7f076246178b6", "reading_canonical_log_prob": -6.230886936187744, "reading_distances": [28, 32, 38, 35], "reading_log_probs": [-6.061198711395264, -6.472847938537598, -5.276413917541504, -5.809472560882568], "reading_logprob_std": 0.5003649575759144, "tokenization_path_logprob_std": 66.73382310973564, "trajectory_balance_rmse": 455.49697607104764} {"canonical_length": 20, "canonical_nll": 1.579991340637207, "generation_rouge_l_f1": 0.4444444444444445, "generation_token_f1": 0.5185185185185186, "prompt_id": "ba65e8996b196d2c", "reading_canonical_log_prob": -31.59982681274414, "reading_distances": [39, 40, 26, 40], "reading_log_probs": [-31.21441078186035, -29.839096069335938, -27.212963104248047, -29.05604362487793], "reading_logprob_std": 1.670037777669358, "tokenization_path_logprob_std": 62.015937145121775, "trajectory_balance_rmse": 331.2036996945467} {"canonical_length": 8, "canonical_nll": 1.267653226852417, "generation_rouge_l_f1": 0.39999999999999997, "generation_token_f1": 0.39999999999999997, "prompt_id": "e026b81849f6617b", "reading_canonical_log_prob": -10.141225814819336, "reading_distances": [29, 25, 43, 28], "reading_log_probs": [-12.117707252502441, -12.602433204650879, -13.718365669250488, -14.277261734008789], "reading_logprob_std": 0.9926170314664654, "tokenization_path_logprob_std": 55.26510881067931, "trajectory_balance_rmse": 196.59767619005186} {"canonical_length": 84, "canonical_nll": 2.635734194800967, "generation_rouge_l_f1": 0.462962962962963, "generation_token_f1": 0.462962962962963, "prompt_id": "1c6873b5ff06ab16", "reading_canonical_log_prob": -221.40167236328125, "reading_distances": [22, 21, 23, 21], "reading_log_probs": [-217.66165161132812, -227.5791015625, -218.08216857910156, -216.3620147705078], "reading_logprob_std": 5.157481663302903, "tokenization_path_logprob_std": 76.26199927735402, "trajectory_balance_rmse": 1890.8860619383092} {"canonical_length": 212, "canonical_nll": 0.9509128354630381, "generation_rouge_l_f1": 0.2488479262672811, "generation_token_f1": 0.4976958525345622, "prompt_id": "9cabb6227a1b5412", "reading_canonical_log_prob": -201.59352111816406, "reading_distances": [16, 14, 17, 20], "reading_log_probs": [-196.84722900390625, -196.48690795898438, -199.886962890625, -196.22036743164062], "reading_logprob_std": 1.7038708000461396, "tokenization_path_logprob_std": 92.26974157901527, "trajectory_balance_rmse": 3402.4578459327613} {"canonical_length": 123, "canonical_nll": 1.4313657186864837, "generation_rouge_l_f1": 0.2365038560411311, "generation_token_f1": 0.4061696658097686, "prompt_id": "3aaeb505baa59e0c", "reading_canonical_log_prob": -176.0579833984375, "reading_distances": [16, 18, 18, 19], "reading_log_probs": [-175.03060913085938, -173.3607177734375, -173.56651306152344, -175.6773681640625], "reading_logprob_std": 1.126030701541318, "tokenization_path_logprob_std": 170.54067990782065, "trajectory_balance_rmse": 2572.9535396378355} {"canonical_length": 305, "canonical_nll": 1.0936067174692623, "generation_rouge_l_f1": 0.21238938053097345, "generation_token_f1": 0.48141592920353987, "prompt_id": "5964386ba6fbd7d7", "reading_canonical_log_prob": -333.550048828125, "reading_distances": [30, 32, 37, 40], "reading_log_probs": [-329.0951843261719, -322.9464111328125, -323.9886474609375, -324.2069091796875], "reading_logprob_std": 2.7462415910176263, "tokenization_path_logprob_std": 110.35776838342697, "trajectory_balance_rmse": 5722.105263058299} {"canonical_length": 125, "canonical_nll": 1.033601806640625, "generation_rouge_l_f1": 0.24456521739130432, "generation_token_f1": 0.3967391304347826, "prompt_id": "a562f2a21ae863a6", "reading_canonical_log_prob": -129.20022583007812, "reading_distances": [10, 8, 8, 7], "reading_log_probs": [-131.03573608398438, -130.7133026123047, -129.39682006835938, -129.3898162841797], "reading_logprob_std": 0.8652481551343135, "tokenization_path_logprob_std": 101.02542234813302, "trajectory_balance_rmse": 2355.588517915196} {"canonical_length": 166, "canonical_nll": 1.1801951764577843, "generation_rouge_l_f1": 0.17307692307692307, "generation_token_f1": 0.31730769230769235, "prompt_id": "f84f62888219bcaf", "reading_canonical_log_prob": -195.9123992919922, "reading_distances": [87, 91, 90, 94], "reading_log_probs": [-188.28598022460938, -188.45166015625, -188.3433380126953, -188.04566955566406], "reading_logprob_std": 0.1716725415025349, "tokenization_path_logprob_std": 85.41277512153238, "trajectory_balance_rmse": 2438.071358205917} {"canonical_length": 86, "canonical_nll": 0.6230600933695949, "generation_rouge_l_f1": 0.38009049773755654, "generation_token_f1": 0.5610859728506787, "prompt_id": "6b3af0ffe0ba6c9d", "reading_canonical_log_prob": -53.583168029785156, "reading_distances": [32, 26, 25, 21], "reading_log_probs": [-54.42681884765625, -54.01393508911133, -55.15946960449219, -54.700199127197266], "reading_logprob_std": 0.4809819759285823, "tokenization_path_logprob_std": 132.27228535777925, "trajectory_balance_rmse": 1824.6890175546537} {"canonical_length": 61, "canonical_nll": 0.973929045630283, "generation_rouge_l_f1": 0.4523809523809524, "generation_token_f1": 0.6190476190476191, "prompt_id": "53e2d36ebe126c1f", "reading_canonical_log_prob": -59.409671783447266, "reading_distances": [22, 21, 23, 24], "reading_log_probs": [-64.99688720703125, -66.455078125, -65.41256713867188, -67.39103698730469], "reading_logprob_std": 1.07657848528294, "tokenization_path_logprob_std": 70.97762529989856, "trajectory_balance_rmse": 1154.3008514149833} {"canonical_length": 133, "canonical_nll": 0.9838276339652843, "generation_rouge_l_f1": 0.32022471910112354, "generation_token_f1": 0.5112359550561798, "prompt_id": "bce0ef5bdff3466e", "reading_canonical_log_prob": -130.8490753173828, "reading_distances": [23, 21, 18, 19], "reading_log_probs": [-133.24069213867188, -135.42074584960938, -132.19717407226562, -129.56753540039062], "reading_logprob_std": 2.43070860163281, "tokenization_path_logprob_std": 107.786051359704, "trajectory_balance_rmse": 2399.682410590515} {"canonical_length": 74, "canonical_nll": 2.317640974714949, "generation_rouge_l_f1": 0.23357664233576642, "generation_token_f1": 0.3065693430656934, "prompt_id": "2c030e502077ba6b", "reading_canonical_log_prob": -171.50543212890625, "reading_distances": [32, 25, 29, 23], "reading_log_probs": [-172.9685516357422, -168.84848022460938, -169.8938446044922, -171.617919921875], "reading_logprob_std": 1.8254688076145231, "tokenization_path_logprob_std": 43.34197807609345, "trajectory_balance_rmse": 1232.2788103232863} {"canonical_length": 152, "canonical_nll": 1.5235599718595807, "generation_rouge_l_f1": 0.19138755980861247, "generation_token_f1": 0.3014354066985646, "prompt_id": "b73bca8ae43941f1", "reading_canonical_log_prob": -231.58111572265625, "reading_distances": [18, 23, 21, 12], "reading_log_probs": [-230.67764282226562, -230.23277282714844, -231.04769897460938, -232.4166259765625], "reading_logprob_std": 0.9427881354467076, "tokenization_path_logprob_std": 99.34666072224326, "trajectory_balance_rmse": 3063.181144917282} {"canonical_length": 139, "canonical_nll": 1.4757971454867356, "generation_rouge_l_f1": 0.205, "generation_token_f1": 0.38499999999999995, "prompt_id": "98eb458a641b4b4a", "reading_canonical_log_prob": -205.13580322265625, "reading_distances": [37, 38, 30, 23], "reading_log_probs": [-205.3494415283203, -203.39707946777344, -203.37734985351562, -203.39907836914062], "reading_logprob_std": 0.9791852490573848, "tokenization_path_logprob_std": 109.42183427448124, "trajectory_balance_rmse": 3255.066606972958} {"canonical_length": 79, "canonical_nll": 1.8962566520594344, "generation_rouge_l_f1": 0.21333333333333335, "generation_token_f1": 0.3, "prompt_id": "3cfac8c1932346e3", "reading_canonical_log_prob": -149.8042755126953, "reading_distances": [22, 28, 26, 21], "reading_log_probs": [-147.9852752685547, -150.42111206054688, -152.33676147460938, -147.8748779296875], "reading_logprob_std": 2.1397497195219617, "tokenization_path_logprob_std": 107.0455015651185, "trajectory_balance_rmse": 1563.6259979433771} {"canonical_length": 22, "canonical_nll": 1.1681486476551404, "generation_rouge_l_f1": 0.37499999999999994, "generation_token_f1": 0.37499999999999994, "prompt_id": "7f70df4099ae17f5", "reading_canonical_log_prob": -25.699270248413086, "reading_distances": [28, 29, 27, 31], "reading_log_probs": [-28.338430404663086, -30.445674896240234, -30.4467830657959, -29.39510726928711], "reading_logprob_std": 1.008791018787001, "tokenization_path_logprob_std": 49.45370879989707, "trajectory_balance_rmse": 463.1235027788224} {"canonical_length": 87, "canonical_nll": 0.5814951754164421, "generation_rouge_l_f1": 0.023255813953488372, "generation_token_f1": 0.023255813953488372, "prompt_id": "0f424e2f0c95df2d", "reading_canonical_log_prob": -50.59008026123047, "reading_distances": [21, 22, 21, 18], "reading_log_probs": [-49.12583923339844, -47.93488693237305, -50.13682556152344, -49.77320098876953], "reading_logprob_std": 0.9669407560139757, "tokenization_path_logprob_std": 102.94788528946131, "trajectory_balance_rmse": 1448.251529466121} {"canonical_length": 222, "canonical_nll": 1.1698894328899212, "generation_rouge_l_f1": 0.26175869120654394, "generation_token_f1": 0.46625766871165647, "prompt_id": "7d4605b87d9b1016", "reading_canonical_log_prob": -259.7154541015625, "reading_distances": [26, 19, 24, 24], "reading_log_probs": [-261.026123046875, -261.0337219238281, -262.24798583984375, -262.8072814941406], "reading_logprob_std": 0.8943477082087101, "tokenization_path_logprob_std": 275.79225109416103, "trajectory_balance_rmse": 5068.305901070996} {"canonical_length": 68, "canonical_nll": 2.172198127297794, "generation_rouge_l_f1": 0.16289592760180996, "generation_token_f1": 0.24434389140271492, "prompt_id": "7bba6cec3efa18f3", "reading_canonical_log_prob": -147.70947265625, "reading_distances": [16, 18, 20, 17], "reading_log_probs": [-147.16580200195312, -146.51199340820312, -146.87716674804688, -145.51754760742188], "reading_logprob_std": 0.7187000279746075, "tokenization_path_logprob_std": 141.98148013064832, "trajectory_balance_rmse": 1458.128160632574} {"canonical_length": 10, "canonical_nll": 0.6369214057922363, "generation_rouge_l_f1": 0.7777777777777778, "generation_token_f1": 0.7777777777777778, "prompt_id": "23ac4432bd5b4331", "reading_canonical_log_prob": -6.369214057922363, "reading_distances": [23, 18, 24, 23], "reading_log_probs": [-8.134020805358887, -6.994666576385498, -8.201574325561523, -8.086098670959473], "reading_logprob_std": 0.5749038344968193, "tokenization_path_logprob_std": 49.62940580363892, "trajectory_balance_rmse": 214.3487222818178} {"canonical_length": 35, "canonical_nll": 0.7086807250976562, "generation_rouge_l_f1": 0.1951219512195122, "generation_token_f1": 0.25365853658536586, "prompt_id": "f25ec758946024eb", "reading_canonical_log_prob": -24.80382537841797, "reading_distances": [37, 39, 42, 43], "reading_log_probs": [-26.781896591186523, -27.79637908935547, -28.288373947143555, -27.176315307617188], "reading_logprob_std": 0.6656717996525475, "tokenization_path_logprob_std": 42.56139714530875, "trajectory_balance_rmse": 380.7661010861058} {"canonical_length": 225, "canonical_nll": 1.5860245768229166, "generation_rouge_l_f1": 0.28690228690228686, "generation_token_f1": 0.5654885654885654, "prompt_id": "e70426e74522ced9", "reading_canonical_log_prob": -356.85552978515625, "reading_distances": [29, 30, 30, 31], "reading_log_probs": [-362.79119873046875, -367.2619934082031, -361.9299011230469, -362.5577392578125], "reading_logprob_std": 2.44505858683621, "tokenization_path_logprob_std": 295.8806578401487, "trajectory_balance_rmse": 4261.808084194045} {"canonical_length": 17, "canonical_nll": 0.23275208473205566, "generation_rouge_l_f1": 0.7878787878787877, "generation_token_f1": 0.7878787878787877, "prompt_id": "fa1698c1de8ba5af", "reading_canonical_log_prob": -3.9567854404449463, "reading_distances": [22, 23, 21, 22], "reading_log_probs": [-6.867685317993164, -5.670562744140625, -6.30073881149292, -6.560423851013184], "reading_logprob_std": 0.5087035312348399, "tokenization_path_logprob_std": 37.48091158788631, "trajectory_balance_rmse": 302.08508751463046} {"canonical_length": 41, "canonical_nll": 1.418546257949457, "generation_rouge_l_f1": 0.20238095238095236, "generation_token_f1": 0.2857142857142857, "prompt_id": "2ce41f448f87f273", "reading_canonical_log_prob": -58.160396575927734, "reading_distances": [38, 37, 28, 41], "reading_log_probs": [-53.18607711791992, -53.121665954589844, -53.405517578125, -53.09891891479492], "reading_logprob_std": 0.13993789642412083, "tokenization_path_logprob_std": 68.8731413763441, "trajectory_balance_rmse": 763.3354113461472} {"canonical_length": 17, "canonical_nll": 0.34619976492489085, "generation_rouge_l_f1": 0.9696969696969697, "generation_token_f1": 0.9696969696969697, "prompt_id": "f7f7943a60f6d582", "reading_canonical_log_prob": -5.8853960037231445, "reading_distances": [33, 25, 33, 33], "reading_log_probs": [-6.7661027908325195, -8.18160343170166, -8.807665824890137, -7.589276313781738], "reading_logprob_std": 0.869698935972639, "tokenization_path_logprob_std": 16.28363213559841, "trajectory_balance_rmse": 313.96049263801916} {"canonical_length": 45, "canonical_nll": 2.373726399739583, "generation_rouge_l_f1": 0.24460431654676257, "generation_token_f1": 0.28776978417266186, "prompt_id": "d6cc1411da921bc9", "reading_canonical_log_prob": -106.81768798828125, "reading_distances": [17, 21, 19, 19], "reading_log_probs": [-106.26222229003906, -106.89016723632812, -107.69693756103516, -106.67269897460938], "reading_logprob_std": 0.6033551577485651, "tokenization_path_logprob_std": 30.604710058575723, "trajectory_balance_rmse": 1038.481070834443} {"canonical_length": 19, "canonical_nll": 0.2755141258239746, "generation_rouge_l_f1": 0.5454545454545454, "generation_token_f1": 0.5454545454545454, "prompt_id": "9f836d8d503579f9", "reading_canonical_log_prob": -5.234768390655518, "reading_distances": [34, 42, 46, 38], "reading_log_probs": [-6.960628509521484, -8.248427391052246, -6.922781467437744, -6.066443920135498], "reading_logprob_std": 0.8995891383862888, "tokenization_path_logprob_std": 60.47420709923481, "trajectory_balance_rmse": 302.87328748292686} {"canonical_length": 193, "canonical_nll": 1.4675740454481057, "generation_rouge_l_f1": 0.23162583518930957, "generation_token_f1": 0.4053452115812918, "prompt_id": "d3c6858742c763af", "reading_canonical_log_prob": -283.2417907714844, "reading_distances": [35, 34, 29, 42], "reading_log_probs": [-291.044189453125, -291.267578125, -288.95941162109375, -292.4625549316406], "reading_logprob_std": 1.4558953994789883, "tokenization_path_logprob_std": 47.93834632149668, "trajectory_balance_rmse": 4401.3590785007445} {"canonical_length": 19, "canonical_nll": 0.8881833929764597, "generation_rouge_l_f1": 0.6046511627906976, "generation_token_f1": 0.6046511627906976, "prompt_id": "923c3d05ff40a1cb", "reading_canonical_log_prob": -16.875484466552734, "reading_distances": [31, 34, 27, 31], "reading_log_probs": [-18.11456298828125, -18.65579605102539, -18.399982452392578, -19.288673400878906], "reading_logprob_std": 0.500722702600605, "tokenization_path_logprob_std": 23.810781414851157, "trajectory_balance_rmse": 549.1468132697179} {"canonical_length": 200, "canonical_nll": 1.4191412353515624, "generation_rouge_l_f1": 0.4066985645933014, "generation_token_f1": 0.6172248803827751, "prompt_id": "180d045694c4d4f7", "reading_canonical_log_prob": -283.8282470703125, "reading_distances": [12, 11, 15, 8], "reading_log_probs": [-282.86590576171875, -282.21807861328125, -281.7013854980469, -283.870361328125], "reading_logprob_std": 0.9347987546438674, "tokenization_path_logprob_std": 245.9105628047027, "trajectory_balance_rmse": 3998.1151265545936} {"canonical_length": 314, "canonical_nll": 1.003275780161475, "generation_rouge_l_f1": 0.2072538860103627, "generation_token_f1": 0.5043177892918826, "prompt_id": "1350704a8f43bb33", "reading_canonical_log_prob": -315.0285949707031, "reading_distances": [33, 45, 36, 44], "reading_log_probs": [-315.8636474609375, -317.046875, -317.11663818359375, -318.3266296386719], "reading_logprob_std": 1.0059412063023818, "tokenization_path_logprob_std": 367.3586944126338, "trajectory_balance_rmse": 6406.548413027281} {"canonical_length": 79, "canonical_nll": 0.6652024184601216, "generation_rouge_l_f1": 0.24207492795389052, "generation_token_f1": 0.39769452449567727, "prompt_id": "3269989441292641", "reading_canonical_log_prob": -52.55099105834961, "reading_distances": [29, 31, 32, 34], "reading_log_probs": [-53.03243637084961, -54.946189880371094, -55.27705383300781, -53.944549560546875], "reading_logprob_std": 1.0173917812651156, "tokenization_path_logprob_std": 50.78291259727659, "trajectory_balance_rmse": 1760.346835922714} {"canonical_length": 204, "canonical_nll": 1.1504398420745252, "generation_rouge_l_f1": 0.2413793103448276, "generation_token_f1": 0.5, "prompt_id": "0f5981f2ed369cdc", "reading_canonical_log_prob": -234.68972778320312, "reading_distances": [28, 42, 38, 33], "reading_log_probs": [-237.19235229492188, -237.66741943359375, -238.9500274658203, -237.55001831054688], "reading_logprob_std": 0.7671336147333355, "tokenization_path_logprob_std": 301.3525424455073, "trajectory_balance_rmse": 5053.145634274235} {"canonical_length": 53, "canonical_nll": 1.0322573679798055, "generation_rouge_l_f1": 0.15325670498084293, "generation_token_f1": 0.26819923371647514, "prompt_id": "801510c180dde0a4", "reading_canonical_log_prob": -54.70964050292969, "reading_distances": [34, 36, 39, 36], "reading_log_probs": [-51.15039825439453, -50.64629364013672, -51.20689392089844, -50.712921142578125], "reading_logprob_std": 0.2903188549402979, "tokenization_path_logprob_std": 71.04210373956343, "trajectory_balance_rmse": 1170.8644902471817} {"canonical_length": 18, "canonical_nll": 2.087320751614041, "generation_rouge_l_f1": 0.1176470588235294, "generation_token_f1": 0.13148788927335642, "prompt_id": "a62ba8086bae83ba", "reading_canonical_log_prob": -37.571773529052734, "reading_distances": [21, 20, 17, 18], "reading_log_probs": [-41.027828216552734, -42.62221908569336, -42.05200958251953, -42.33699035644531], "reading_logprob_std": 0.6947807549583694, "tokenization_path_logprob_std": 35.86148107372534, "trajectory_balance_rmse": 342.3503243373997} {"canonical_length": 278, "canonical_nll": 1.0826338075047774, "generation_rouge_l_f1": 0.2567567567567568, "generation_token_f1": 0.5045045045045045, "prompt_id": "6197952fa284f196", "reading_canonical_log_prob": -300.9721984863281, "reading_distances": [20, 24, 23, 23], "reading_log_probs": [-296.920654296875, -297.7932434082031, -298.7583312988281, -297.98760986328125], "reading_logprob_std": 0.7549860477796482, "tokenization_path_logprob_std": 120.84898140389505, "trajectory_balance_rmse": 5549.545722954268} {"canonical_length": 13, "canonical_nll": 0.861271491417518, "generation_rouge_l_f1": 0.3636363636363636, "generation_token_f1": 0.5454545454545454, "prompt_id": "d2ae9ca225f37e77", "reading_canonical_log_prob": -11.196529388427734, "reading_distances": [24, 28, 33, 31], "reading_log_probs": [-11.0472412109375, -9.692922592163086, -11.882989883422852, -12.911685943603516], "reading_logprob_std": 1.3608776813432832, "tokenization_path_logprob_std": 34.75176154383968, "trajectory_balance_rmse": 262.20634558691756} {"canonical_length": 14, "canonical_nll": 2.4849720001220703, "generation_rouge_l_f1": 0.3225806451612903, "generation_token_f1": 0.3870967741935484, "prompt_id": "3d0956e900302c45", "reading_canonical_log_prob": -34.789608001708984, "reading_distances": [19, 18, 20, 17], "reading_log_probs": [-34.707271575927734, -36.801795959472656, -35.28776931762695, -33.75691223144531], "reading_logprob_std": 1.2758758401180736, "tokenization_path_logprob_std": 43.0038405430713, "trajectory_balance_rmse": 287.17199047505244} {"canonical_length": 13, "canonical_nll": 0.4413195756765512, "generation_rouge_l_f1": 0.059259259259259255, "generation_token_f1": 0.06666666666666667, "prompt_id": "360b4e609ec3ff2b", "reading_canonical_log_prob": -5.737154483795166, "reading_distances": [22, 16, 20, 23], "reading_log_probs": [-4.601842403411865, -5.130538463592529, -4.74680233001709, -4.966667175292969], "reading_logprob_std": 0.23382290094546254, "tokenization_path_logprob_std": 20.564915467232858, "trajectory_balance_rmse": 115.42663710233543} {"canonical_length": 86, "canonical_nll": 1.739495920580487, "generation_rouge_l_f1": 0.35714285714285715, "generation_token_f1": 0.4999999999999999, "prompt_id": "1ad837078d8b4fb2", "reading_canonical_log_prob": -149.59664916992188, "reading_distances": [31, 36, 32, 30], "reading_log_probs": [-141.63877868652344, -143.03253173828125, -145.07725524902344, -142.90977478027344], "reading_logprob_std": 1.4222937694386397, "tokenization_path_logprob_std": 197.12306062281917, "trajectory_balance_rmse": 1791.1689248969967} {"canonical_length": 20, "canonical_nll": 0.6045084476470948, "generation_rouge_l_f1": 0.11486486486486486, "generation_token_f1": 0.11486486486486486, "prompt_id": "24f992c3a95d7b8a", "reading_canonical_log_prob": -12.090168952941895, "reading_distances": [38, 41, 32, 31], "reading_log_probs": [-11.82815170288086, -11.5346040725708, -11.539813041687012, -11.241999626159668], "reading_logprob_std": 0.2393082004673677, "tokenization_path_logprob_std": 26.711562796461163, "trajectory_balance_rmse": 509.73349639160966} {"canonical_length": 29, "canonical_nll": 0.7761619830953663, "generation_rouge_l_f1": 0.15555555555555556, "generation_token_f1": 0.24444444444444446, "prompt_id": "429af0c846ac90d6", "reading_canonical_log_prob": -22.508697509765625, "reading_distances": [20, 20, 14, 22], "reading_log_probs": [-21.459291458129883, -21.706172943115234, -22.009002685546875, -22.226425170898438], "reading_logprob_std": 0.33680704230983477, "tokenization_path_logprob_std": 61.48749522537201, "trajectory_balance_rmse": 636.7779721671557} {"canonical_length": 132, "canonical_nll": 1.1873774672999526, "generation_rouge_l_f1": 0.24786324786324784, "generation_token_f1": 0.4358974358974359, "prompt_id": "a4d2b99ec03649f1", "reading_canonical_log_prob": -156.73382568359375, "reading_distances": [19, 21, 19, 21], "reading_log_probs": [-154.47454833984375, -156.82577514648438, -154.3963623046875, -155.14016723632812], "reading_logprob_std": 1.1281989934817132, "tokenization_path_logprob_std": 66.93240694330163, "trajectory_balance_rmse": 2654.1894017487566} {"canonical_length": 27, "canonical_nll": 0.33229238015634044, "generation_rouge_l_f1": 0.49438202247191015, "generation_token_f1": 0.49438202247191015, "prompt_id": "363a6f01446a3b83", "reading_canonical_log_prob": -8.971894264221191, "reading_distances": [45, 40, 44, 33], "reading_log_probs": [-11.108551979064941, -10.498077392578125, -11.82990837097168, -9.923587799072266], "reading_logprob_std": 0.818282986078107, "tokenization_path_logprob_std": 50.790840939827284, "trajectory_balance_rmse": 633.8026596724465} {"canonical_length": 56, "canonical_nll": 1.9996910095214844, "generation_rouge_l_f1": 0.1081081081081081, "generation_token_f1": 0.222972972972973, "prompt_id": "15d4cbd9fbd2e327", "reading_canonical_log_prob": -111.98269653320312, "reading_distances": [28, 29, 30, 28], "reading_log_probs": [-111.59133911132812, -110.88117980957031, -112.0465087890625, -111.5016098022461], "reading_logprob_std": 0.47953075417535496, "tokenization_path_logprob_std": 66.84438871630917, "trajectory_balance_rmse": 1148.983049239462} {"canonical_length": 56, "canonical_nll": 1.0458759580339705, "generation_rouge_l_f1": 0.2748091603053435, "generation_token_f1": 0.4580152671755725, "prompt_id": "c741581c2c318b9d", "reading_canonical_log_prob": -58.569053649902344, "reading_distances": [19, 28, 30, 29], "reading_log_probs": [-57.47691345214844, -58.51870346069336, -58.80830383300781, -58.81092071533203], "reading_logprob_std": 0.6329012754767928, "tokenization_path_logprob_std": 97.23119575882761, "trajectory_balance_rmse": 1035.073028356188} {"canonical_length": 10, "canonical_nll": 0.9132015228271484, "generation_rouge_l_f1": 0.72, "generation_token_f1": 0.72, "prompt_id": "1551065a718dd716", "reading_canonical_log_prob": -9.132015228271484, "reading_distances": [19, 24, 17, 29], "reading_log_probs": [-10.948968887329102, -11.326212882995605, -9.824115753173828, -11.451601028442383], "reading_logprob_std": 0.74055007542923, "tokenization_path_logprob_std": 53.36199515114404, "trajectory_balance_rmse": 175.69642242490198} {"canonical_length": 22, "canonical_nll": 1.5294896906072444, "generation_rouge_l_f1": 0.5625000000000001, "generation_token_f1": 0.625, "prompt_id": "0281e958164e52e0", "reading_canonical_log_prob": -33.648773193359375, "reading_distances": [20, 23, 25, 22], "reading_log_probs": [-32.63550567626953, -32.733062744140625, -32.52509689331055, -33.368988037109375], "reading_logprob_std": 0.37853961598493263, "tokenization_path_logprob_std": 34.7131045715783, "trajectory_balance_rmse": 365.833316340384} {"canonical_length": 338, "canonical_nll": 1.2788672814002404, "generation_rouge_l_f1": 0.22331691297208536, "generation_token_f1": 0.3973727422003284, "prompt_id": "059afc2f6690f050", "reading_canonical_log_prob": -432.25714111328125, "reading_distances": [28, 20, 24, 34], "reading_log_probs": [-426.37152099609375, -426.24517822265625, -428.95208740234375, -427.6160888671875], "reading_logprob_std": 1.2654340509462203, "tokenization_path_logprob_std": 317.7590692883091, "trajectory_balance_rmse": 6127.823880969553} {"canonical_length": 11, "canonical_nll": 0.971763784235174, "generation_rouge_l_f1": 0.19354838709677416, "generation_token_f1": 0.25806451612903225, "prompt_id": "3463b16713cbef14", "reading_canonical_log_prob": -10.689401626586914, "reading_distances": [37, 40, 33, 39], "reading_log_probs": [-10.269775390625, -9.114330291748047, -10.351277351379395, -10.556295394897461], "reading_logprob_std": 0.6503283429784519, "tokenization_path_logprob_std": 40.40808702424947, "trajectory_balance_rmse": 269.6216737447756} {"canonical_length": 184, "canonical_nll": 1.9533724577530571, "generation_rouge_l_f1": 0.2135922330097087, "generation_token_f1": 0.3834951456310679, "prompt_id": "75052e7415fb5476", "reading_canonical_log_prob": -359.4205322265625, "reading_distances": [25, 20, 30, 28], "reading_log_probs": [-356.44036865234375, -358.7503356933594, -356.12451171875, -358.04266357421875], "reading_logprob_std": 1.2608899061187138, "tokenization_path_logprob_std": 263.08296822405117, "trajectory_balance_rmse": 3830.626664201024} {"canonical_length": 155, "canonical_nll": 2.352900154359879, "generation_rouge_l_f1": 0.22895622895622897, "generation_token_f1": 0.41750841750841755, "prompt_id": "3e8a52f189b6e093", "reading_canonical_log_prob": -364.69952392578125, "reading_distances": [27, 25, 28, 31], "reading_log_probs": [-359.7615966796875, -364.3701171875, -360.44854736328125, -363.2142333984375], "reading_logprob_std": 2.198386895153238, "tokenization_path_logprob_std": 110.21877193378604, "trajectory_balance_rmse": 3266.572295345522} {"canonical_length": 278, "canonical_nll": 1.4765212244267085, "generation_rouge_l_f1": 0.23529411764705882, "generation_token_f1": 0.46691176470588236, "prompt_id": "39b34b632ecf238e", "reading_canonical_log_prob": -410.472900390625, "reading_distances": [27, 21, 17, 23], "reading_log_probs": [-411.3916931152344, -410.5257568359375, -411.34716796875, -412.4618835449219], "reading_logprob_std": 0.7938844050184438, "tokenization_path_logprob_std": 108.01824731603686, "trajectory_balance_rmse": 6040.978180221498} {"canonical_length": 144, "canonical_nll": 2.011315027872721, "generation_rouge_l_f1": 0.15201900237529692, "generation_token_f1": 0.32779097387173395, "prompt_id": "76cab7f49b66aca5", "reading_canonical_log_prob": -289.6293640136719, "reading_distances": [17, 14, 15, 11], "reading_log_probs": [-279.7672119140625, -286.2930908203125, -283.2376708984375, -283.211669921875], "reading_logprob_std": 2.6665660459967313, "tokenization_path_logprob_std": 163.6989047276649, "trajectory_balance_rmse": 2561.7364803726605} {"canonical_length": 40, "canonical_nll": 1.8350421905517578, "generation_rouge_l_f1": 0.3695652173913044, "generation_token_f1": 0.3695652173913044, "prompt_id": "b6ca714d2fc1e411", "reading_canonical_log_prob": -73.40168762207031, "reading_distances": [14, 16, 13, 18], "reading_log_probs": [-72.77488708496094, -72.87682342529297, -73.35065460205078, -73.2384033203125], "reading_logprob_std": 0.2775792539156598, "tokenization_path_logprob_std": 73.0192769530103, "trajectory_balance_rmse": 782.9124455936953} {"canonical_length": 39, "canonical_nll": 1.3836392133663862, "generation_rouge_l_f1": 0.31147540983606564, "generation_token_f1": 0.42622950819672134, "prompt_id": "56f7c951f93e4273", "reading_canonical_log_prob": -53.96192932128906, "reading_distances": [34, 27, 29, 26], "reading_log_probs": [-51.17605972290039, -50.81965637207031, -52.810604095458984, -55.376251220703125], "reading_logprob_std": 2.0766427612793565, "tokenization_path_logprob_std": 92.68291489690962, "trajectory_balance_rmse": 762.1377593118667} {"canonical_length": 17, "canonical_nll": 0.5350250356337604, "generation_rouge_l_f1": 0.34285714285714286, "generation_token_f1": 0.34285714285714286, "prompt_id": "3f9be20b25197a77", "reading_canonical_log_prob": -9.095425605773926, "reading_distances": [19, 14, 11, 10], "reading_log_probs": [-15.657196044921875, -9.361700057983398, -10.087038040161133, -9.715221405029297], "reading_logprob_std": 2.982676827421591, "tokenization_path_logprob_std": 60.766947655387646, "trajectory_balance_rmse": 255.03672254946085} {"canonical_length": 231, "canonical_nll": 1.3562080416328464, "generation_rouge_l_f1": 0.27364185110663986, "generation_token_f1": 0.46680080482897385, "prompt_id": "053c343603667759", "reading_canonical_log_prob": -313.2840576171875, "reading_distances": [22, 20, 18, 20], "reading_log_probs": [-305.30267333984375, -302.6485595703125, -303.62615966796875, -303.84521484375], "reading_logprob_std": 1.096010729336734, "tokenization_path_logprob_std": 192.70075922939998, "trajectory_balance_rmse": 4610.175104112669} {"canonical_length": 53, "canonical_nll": 1.24275049173607, "generation_rouge_l_f1": 0.5822784810126582, "generation_token_f1": 0.6329113924050633, "prompt_id": "e23c57e239d810b9", "reading_canonical_log_prob": -65.86577606201172, "reading_distances": [29, 21, 23, 29], "reading_log_probs": [-65.37730407714844, -64.34894561767578, -64.15355682373047, -65.58662414550781], "reading_logprob_std": 0.720104171875901, "tokenization_path_logprob_std": 87.67983859001251, "trajectory_balance_rmse": 964.6579735529834} {"canonical_length": 66, "canonical_nll": 1.1411925807143704, "generation_rouge_l_f1": 0.19469026548672566, "generation_token_f1": 0.247787610619469, "prompt_id": "e9b590e25ca62203", "reading_canonical_log_prob": -75.31871032714844, "reading_distances": [21, 22, 30, 20], "reading_log_probs": [-72.90281677246094, -72.95044708251953, -74.3886489868164, -73.49261474609375], "reading_logprob_std": 0.6905960902757751, "tokenization_path_logprob_std": 149.20080162028, "trajectory_balance_rmse": 1322.4121079234021} {"canonical_length": 16, "canonical_nll": 1.8232314586639404, "generation_rouge_l_f1": 0.125, "generation_token_f1": 0.125, "prompt_id": "d046e49475ca84ec", "reading_canonical_log_prob": -29.171703338623047, "reading_distances": [37, 44, 42, 37], "reading_log_probs": [-29.227977752685547, -29.41577911376953, -29.404991149902344, -29.584739685058594], "reading_logprob_std": 0.1457325635983378, "tokenization_path_logprob_std": 107.41300750692426, "trajectory_balance_rmse": 489.81796379136523} {"canonical_length": 56, "canonical_nll": 1.032827513558524, "generation_rouge_l_f1": 0.3870967741935484, "generation_token_f1": 0.5376344086021505, "prompt_id": "750881d0bf75959b", "reading_canonical_log_prob": -57.838340759277344, "reading_distances": [48, 44, 52, 48], "reading_log_probs": [-60.32270050048828, -58.53151321411133, -58.2793083190918, -58.68149185180664], "reading_logprob_std": 0.9275964381670279, "tokenization_path_logprob_std": 75.41017590937226, "trajectory_balance_rmse": 1289.2046049451665} {"canonical_length": 2, "canonical_nll": 0.14813746511936188, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "73f8da9f02677521", "reading_canonical_log_prob": -0.29627493023872375, "reading_distances": [18, 14, 15, 17], "reading_log_probs": [-0.46907833218574524, -0.41107961535453796, -0.4852566719055176, -0.43986988067626953], "reading_logprob_std": 0.03274879054215412, "tokenization_path_logprob_std": 10.687951305827877, "trajectory_balance_rmse": 51.17704158663531} {"canonical_length": 156, "canonical_nll": 2.3992888621794872, "generation_rouge_l_f1": 0.19011406844106463, "generation_token_f1": 0.2813688212927757, "prompt_id": "9c25897ff633e0cd", "reading_canonical_log_prob": -374.2890625, "reading_distances": [29, 34, 28, 35], "reading_log_probs": [-377.676513671875, -381.35040283203125, -378.03863525390625, -381.8555603027344], "reading_logprob_std": 2.177248449578087, "tokenization_path_logprob_std": 219.53644594746396, "trajectory_balance_rmse": 2619.0058914002407} {"canonical_length": 263, "canonical_nll": 1.1062466581511408, "generation_rouge_l_f1": 0.37404580152671757, "generation_token_f1": 0.6030534351145038, "prompt_id": "5d0b3ece32e43007", "reading_canonical_log_prob": -290.94287109375, "reading_distances": [13, 21, 23, 25], "reading_log_probs": [-288.54925537109375, -286.4359130859375, -284.8235168457031, -287.96026611328125], "reading_logprob_std": 1.669753506628393, "tokenization_path_logprob_std": 98.46367381609214, "trajectory_balance_rmse": 6009.628673593051} {"canonical_length": 21, "canonical_nll": 0.5346130189441499, "generation_rouge_l_f1": 0.5, "generation_token_f1": 0.5, "prompt_id": "16cea9158da9fd5a", "reading_canonical_log_prob": -11.226873397827148, "reading_distances": [31, 29, 34, 38], "reading_log_probs": [-12.36432933807373, -12.214235305786133, -12.348370552062988, -12.252613067626953], "reading_logprob_std": 0.07297147798112942, "tokenization_path_logprob_std": 26.667916581260833, "trajectory_balance_rmse": 423.68487923324193} {"canonical_length": 18, "canonical_nll": 1.2953724331325955, "generation_rouge_l_f1": 0.39999999999999997, "generation_token_f1": 0.5142857142857143, "prompt_id": "d643de98c824ec5b", "reading_canonical_log_prob": -23.31670379638672, "reading_distances": [27, 23, 18, 22], "reading_log_probs": [-25.621864318847656, -24.072547912597656, -23.620922088623047, -24.252212524414062], "reading_logprob_std": 0.8619198553862141, "tokenization_path_logprob_std": 57.74285446197999, "trajectory_balance_rmse": 309.2949462325638} {"canonical_length": 97, "canonical_nll": 1.7987003916317654, "generation_rouge_l_f1": 0.31055900621118016, "generation_token_f1": 0.49689440993788814, "prompt_id": "cdeaf78291feca5e", "reading_canonical_log_prob": -174.47393798828125, "reading_distances": [37, 29, 32, 39], "reading_log_probs": [-184.71128845214844, -186.9932098388672, -190.30419921875, -192.23101806640625], "reading_logprob_std": 3.35589116460062, "tokenization_path_logprob_std": 56.339768014380105, "trajectory_balance_rmse": 1744.6274665273793} {"canonical_length": 308, "canonical_nll": 1.756129079050832, "generation_rouge_l_f1": 0.2845849802371541, "generation_token_f1": 0.45454545454545453, "prompt_id": "26a8450ba4613a47", "reading_canonical_log_prob": -540.8877563476562, "reading_distances": [22, 17, 20, 17], "reading_log_probs": [-539.4769287109375, -537.3326416015625, -534.4698486328125, -540.49169921875], "reading_logprob_std": 2.663589219920255, "tokenization_path_logprob_std": 145.3704491603012, "trajectory_balance_rmse": 5165.551797904214} {"canonical_length": 104, "canonical_nll": 0.6480702620286208, "generation_rouge_l_f1": 0.45714285714285713, "generation_token_f1": 0.6628571428571428, "prompt_id": "c15330337653f483", "reading_canonical_log_prob": -67.39930725097656, "reading_distances": [281, 304, 302, 287], "reading_log_probs": [-66.62759399414062, -69.13043975830078, -75.02974700927734, -69.45238494873047], "reading_logprob_std": 3.5455641117642953, "tokenization_path_logprob_std": 156.00700326283882, "trajectory_balance_rmse": 1803.8810110029826} {"canonical_length": 11, "canonical_nll": 1.3045597076416016, "generation_rouge_l_f1": 0.7000000000000001, "generation_token_f1": 0.7000000000000001, "prompt_id": "8609ed9bb93a15ed", "reading_canonical_log_prob": -14.350156784057617, "reading_distances": [17, 14, 15, 19], "reading_log_probs": [-17.775270462036133, -19.363285064697266, -17.679405212402344, -19.77145767211914], "reading_logprob_std": 1.076045368880605, "tokenization_path_logprob_std": 26.622748616926472, "trajectory_balance_rmse": 143.99855712302096} {"canonical_length": 367, "canonical_nll": 1.388716944556795, "generation_rouge_l_f1": 0.2624, "generation_token_f1": 0.5055999999999999, "prompt_id": "4cb8d591dfd52931", "reading_canonical_log_prob": -509.65911865234375, "reading_distances": [22, 25, 30, 31], "reading_log_probs": [-507.76959228515625, -505.9894714355469, -510.139404296875, -508.8536376953125], "reading_logprob_std": 1.7568575291360835, "tokenization_path_logprob_std": 170.8938227322901, "trajectory_balance_rmse": 7637.331939388849} {"canonical_length": 277, "canonical_nll": 0.45260801900595105, "generation_rouge_l_f1": 0.4084507042253521, "generation_token_f1": 0.6384976525821596, "prompt_id": "7e37b68aba5b5c9c", "reading_canonical_log_prob": -125.37242126464844, "reading_distances": [35, 28, 36, 36], "reading_log_probs": [-126.92733764648438, -127.20276641845703, -125.77515411376953, -128.7101593017578], "reading_logprob_std": 1.2078357575515206, "tokenization_path_logprob_std": 112.34487835985502, "trajectory_balance_rmse": 4135.872203048549} {"canonical_length": 87, "canonical_nll": 1.3463648653578484, "generation_rouge_l_f1": 0.16574585635359115, "generation_token_f1": 0.29834254143646405, "prompt_id": "fef695db887a3a23", "reading_canonical_log_prob": -117.13374328613281, "reading_distances": [17, 19, 19, 15], "reading_log_probs": [-123.41607666015625, -116.13331604003906, -120.6855239868164, -116.4142074584961], "reading_logprob_std": 3.5185974304844105, "tokenization_path_logprob_std": 29.222492720572976, "trajectory_balance_rmse": 1682.9148859050683} {"canonical_length": 174, "canonical_nll": 0.5673230927566002, "generation_rouge_l_f1": 0.4495412844036697, "generation_token_f1": 0.6422018348623854, "prompt_id": "471b9b9cabf2decb", "reading_canonical_log_prob": -98.71421813964844, "reading_distances": [22, 24, 22, 22], "reading_log_probs": [-92.35244750976562, -89.04962921142578, -90.29881286621094, -90.54695129394531], "reading_logprob_std": 1.361673187904605, "tokenization_path_logprob_std": 101.53034755441824, "trajectory_balance_rmse": 2603.0158013804844} {"canonical_length": 10, "canonical_nll": 0.9672113418579101, "generation_rouge_l_f1": 0.35294117647058826, "generation_token_f1": 0.35294117647058826, "prompt_id": "71f95c366e84f3b3", "reading_canonical_log_prob": -9.672113418579102, "reading_distances": [29, 41, 33, 31], "reading_log_probs": [-7.44532585144043, -9.02548885345459, -8.436574935913086, -7.752277374267578], "reading_logprob_std": 0.707687508936079, "tokenization_path_logprob_std": 30.288556301758813, "trajectory_balance_rmse": 265.98548725641893} {"canonical_length": 7, "canonical_nll": 0.4768147809164865, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "78998e3018151a20", "reading_canonical_log_prob": -3.3377034664154053, "reading_distances": [15, 19, 16, 24], "reading_log_probs": [-4.718006610870361, -7.641833305358887, -4.795309066772461, -6.632103443145752], "reading_logprob_std": 1.4351124552045973, "tokenization_path_logprob_std": 24.96658794280203, "trajectory_balance_rmse": 107.17923364290388} {"canonical_length": 179, "canonical_nll": 1.5042624020709672, "generation_rouge_l_f1": 0.17738359201773835, "generation_token_f1": 0.4124168514412417, "prompt_id": "3025be82a097d5b7", "reading_canonical_log_prob": -269.2629699707031, "reading_distances": [24, 21, 30, 26], "reading_log_probs": [-264.53265380859375, -259.71148681640625, -264.4884948730469, -263.55902099609375], "reading_logprob_std": 2.285476043714869, "tokenization_path_logprob_std": 141.07174465300824, "trajectory_balance_rmse": 3490.6281794739166} {"canonical_length": 259, "canonical_nll": 1.838718370120958, "generation_rouge_l_f1": 0.125, "generation_token_f1": 0.2056451612903226, "prompt_id": "317ccf6fb834b830", "reading_canonical_log_prob": -476.2280578613281, "reading_distances": [20, 24, 29, 20], "reading_log_probs": [-471.60943603515625, -468.50469970703125, -468.1386413574219, -470.1153869628906], "reading_logprob_std": 1.5956665622019508, "tokenization_path_logprob_std": 225.72619539589977, "trajectory_balance_rmse": 5269.769518476737} {"canonical_length": 35, "canonical_nll": 1.1890626089913505, "generation_rouge_l_f1": 0.5416666666666667, "generation_token_f1": 0.5416666666666667, "prompt_id": "43805108665c2c05", "reading_canonical_log_prob": -41.617191314697266, "reading_distances": [31, 17, 23, 25], "reading_log_probs": [-37.35930633544922, -37.46039581298828, -37.910823822021484, -37.3945426940918], "reading_logprob_std": 0.2564826909495808, "tokenization_path_logprob_std": 63.936012186684245, "trajectory_balance_rmse": 724.5861036838112} {"canonical_length": 118, "canonical_nll": 1.4557647705078125, "generation_rouge_l_f1": 0.2695417789757413, "generation_token_f1": 0.463611859838275, "prompt_id": "5056b26f2553e0fc", "reading_canonical_log_prob": -171.78024291992188, "reading_distances": [26, 26, 33, 14], "reading_log_probs": [-169.99497985839844, -172.6917266845703, -170.1476287841797, -171.86630249023438], "reading_logprob_std": 1.3198861083026099, "tokenization_path_logprob_std": 96.47184017501078, "trajectory_balance_rmse": 2395.9748157512363} {"canonical_length": 372, "canonical_nll": 0.9104220482610887, "generation_rouge_l_f1": 0.3291139240506329, "generation_token_f1": 0.5411392405063292, "prompt_id": "1ea4408820560cde", "reading_canonical_log_prob": -338.677001953125, "reading_distances": [12, 13, 13, 14], "reading_log_probs": [-343.64044189453125, -338.35675048828125, -336.87786865234375, -337.80419921875], "reading_logprob_std": 3.0422307195464557, "tokenization_path_logprob_std": 391.3507405327144, "trajectory_balance_rmse": 6175.791085349605} {"canonical_length": 9, "canonical_nll": 0.7872042655944824, "generation_rouge_l_f1": 0.8571428571428571, "generation_token_f1": 0.8571428571428571, "prompt_id": "2b79f5bffeeb5daf", "reading_canonical_log_prob": -7.084838390350342, "reading_distances": [35, 37, 32, 32], "reading_log_probs": [-10.361313819885254, -9.917959213256836, -10.08831787109375, -9.889395713806152], "reading_logprob_std": 0.21664194049577798, "tokenization_path_logprob_std": 49.282138486671386, "trajectory_balance_rmse": 175.72063400131861} {"canonical_length": 36, "canonical_nll": 1.0949365827772353, "generation_rouge_l_f1": 0.11387900355871884, "generation_token_f1": 0.1708185053380783, "prompt_id": "91748c9b4543a8c6", "reading_canonical_log_prob": -39.41771697998047, "reading_distances": [10, 13, 13, 11], "reading_log_probs": [-37.69305419921875, -36.800453186035156, -36.28948974609375, -37.11109161376953], "reading_logprob_std": 0.5872265102713468, "tokenization_path_logprob_std": 77.28942063023466, "trajectory_balance_rmse": 489.62187386751316} {"canonical_length": 187, "canonical_nll": 1.209268172157002, "generation_rouge_l_f1": 0.2624113475177305, "generation_token_f1": 0.4113475177304965, "prompt_id": "9a3c9e43bea33f6c", "reading_canonical_log_prob": -226.13314819335938, "reading_distances": [21, 17, 25, 16], "reading_log_probs": [-220.9715576171875, -220.46043395996094, -220.51031494140625, -220.03179931640625], "reading_logprob_std": 0.38431007838290643, "tokenization_path_logprob_std": 66.95006681644405, "trajectory_balance_rmse": 4071.883519468347} {"canonical_length": 376, "canonical_nll": 1.260187676612367, "generation_rouge_l_f1": 0.20097244732576985, "generation_token_f1": 0.45705024311183146, "prompt_id": "7473e8ab94e776ba", "reading_canonical_log_prob": -473.83056640625, "reading_distances": [22, 19, 22, 14], "reading_log_probs": [-478.9835510253906, -477.81732177734375, -480.452392578125, -475.7659912109375], "reading_logprob_std": 1.978724132654721, "tokenization_path_logprob_std": 106.73995232328298, "trajectory_balance_rmse": 7392.66066285783} {"canonical_length": 87, "canonical_nll": 0.43014201898684445, "generation_rouge_l_f1": 0.24242424242424243, "generation_token_f1": 0.3973063973063973, "prompt_id": "81563152acabd240", "reading_canonical_log_prob": -37.42235565185547, "reading_distances": [40, 35, 40, 34], "reading_log_probs": [-36.59346008300781, -34.0255241394043, -35.293312072753906, -36.104576110839844], "reading_logprob_std": 1.1221846283265666, "tokenization_path_logprob_std": 95.97146126678851, "trajectory_balance_rmse": 985.6147595397207} {"canonical_length": 19, "canonical_nll": 2.3571379812140214, "generation_rouge_l_f1": 0.13333333333333333, "generation_token_f1": 0.13333333333333333, "prompt_id": "49b74500a1071228", "reading_canonical_log_prob": -44.785621643066406, "reading_distances": [22, 20, 18, 28], "reading_log_probs": [-43.159698486328125, -43.825416564941406, -44.651153564453125, -44.186622619628906], "reading_logprob_std": 0.6291722333628644, "tokenization_path_logprob_std": 67.30418432084791, "trajectory_balance_rmse": 483.4080080019651} {"canonical_length": 83, "canonical_nll": 1.102916855409921, "generation_rouge_l_f1": 0.3092783505154639, "generation_token_f1": 0.4742268041237113, "prompt_id": "c1a613d92b944071", "reading_canonical_log_prob": -91.54209899902344, "reading_distances": [36, 36, 31, 34], "reading_log_probs": [-92.87660217285156, -92.59947204589844, -92.42900848388672, -93.86524963378906], "reading_logprob_std": 0.6421710823667577, "tokenization_path_logprob_std": 73.55804292585601, "trajectory_balance_rmse": 1587.4561963073857} {"canonical_length": 190, "canonical_nll": 1.3246424624794408, "generation_rouge_l_f1": 0.25431034482758624, "generation_token_f1": 0.4913793103448276, "prompt_id": "00050e95a0ce36d0", "reading_canonical_log_prob": -251.68206787109375, "reading_distances": [34, 29, 22, 38], "reading_log_probs": [-252.32260131835938, -249.99252319335938, -252.07379150390625, -254.54017639160156], "reading_logprob_std": 1.8597645774859441, "tokenization_path_logprob_std": 117.76566607981664, "trajectory_balance_rmse": 3630.95440312321} {"canonical_length": 309, "canonical_nll": 1.3170582792905543, "generation_rouge_l_f1": 0.23404255319148937, "generation_token_f1": 0.5283687943262411, "prompt_id": "3e9a72eac16389eb", "reading_canonical_log_prob": -406.97100830078125, "reading_distances": [18, 19, 20, 22], "reading_log_probs": [-403.27899169921875, -405.6185607910156, -403.6673583984375, -405.45977783203125], "reading_logprob_std": 1.205038402438265, "tokenization_path_logprob_std": 327.73410596588695, "trajectory_balance_rmse": 6563.291146391428} {"canonical_length": 335, "canonical_nll": 1.181411333226446, "generation_rouge_l_f1": 0.22752293577981653, "generation_token_f1": 0.4513761467889908, "prompt_id": "bef02b456531b6b6", "reading_canonical_log_prob": -395.7727966308594, "reading_distances": [18, 20, 20, 21], "reading_log_probs": [-394.52288818359375, -393.382568359375, -394.0020751953125, -392.86456298828125], "reading_logprob_std": 0.7227073075755576, "tokenization_path_logprob_std": 152.92078954145703, "trajectory_balance_rmse": 5958.135405345483} {"canonical_length": 110, "canonical_nll": 1.4484956221147016, "generation_rouge_l_f1": 0.20728291316526606, "generation_token_f1": 0.3025210084033613, "prompt_id": "9d2a25bad82737d8", "reading_canonical_log_prob": -159.3345184326172, "reading_distances": [16, 19, 17, 18], "reading_log_probs": [-159.01730346679688, -161.081298828125, -161.65390014648438, -159.0344696044922], "reading_logprob_std": 1.3720669348356704, "tokenization_path_logprob_std": 103.71246251203324, "trajectory_balance_rmse": 2625.1685975982855} {"canonical_length": 10, "canonical_nll": 1.4565784454345703, "generation_rouge_l_f1": 0.0625, "generation_token_f1": 0.0625, "prompt_id": "77bd9a6bbbf6a3af", "reading_canonical_log_prob": -14.565784454345703, "reading_distances": [21, 22, 17, 19], "reading_log_probs": [-15.259183883666992, -13.958781242370605, -14.380802154541016, -14.590434074401855], "reading_logprob_std": 0.5424374227573706, "tokenization_path_logprob_std": 38.37180412927734, "trajectory_balance_rmse": 162.42770100021914} {"canonical_length": 18, "canonical_nll": 0.7811597188313802, "generation_rouge_l_f1": 0.1, "generation_token_f1": 0.1, "prompt_id": "45f03bee685e0c93", "reading_canonical_log_prob": -14.060874938964844, "reading_distances": [23, 23, 27, 23], "reading_log_probs": [-12.354080200195312, -11.631677627563477, -12.187971115112305, -12.279600143432617], "reading_logprob_std": 0.32821066578722763, "tokenization_path_logprob_std": 82.23488509086069, "trajectory_balance_rmse": 419.34502490617314} {"canonical_length": 300, "canonical_nll": 1.0055543009440104, "generation_rouge_l_f1": 0.2640845070422535, "generation_token_f1": 0.4823943661971831, "prompt_id": "1c2335f122ec1aa9", "reading_canonical_log_prob": -301.6662902832031, "reading_distances": [29, 22, 25, 22], "reading_log_probs": [-306.8091125488281, -305.83062744140625, -306.3684997558594, -303.91326904296875], "reading_logprob_std": 1.2757746031802737, "tokenization_path_logprob_std": 145.95381309575225, "trajectory_balance_rmse": 4976.726473856904} {"canonical_length": 24, "canonical_nll": 0.5785550673802694, "generation_rouge_l_f1": 0.5185185185185185, "generation_token_f1": 0.5185185185185185, "prompt_id": "0b6723f81e5854b3", "reading_canonical_log_prob": -13.885321617126465, "reading_distances": [20, 28, 25, 33], "reading_log_probs": [-14.429092407226562, -14.489065170288086, -14.001641273498535, -14.53407096862793], "reading_logprob_std": 0.2450205194961243, "tokenization_path_logprob_std": 77.5497383328208, "trajectory_balance_rmse": 565.8451923903785} {"canonical_length": 43, "canonical_nll": 0.28790811050769893, "generation_rouge_l_f1": 0.49350649350649345, "generation_token_f1": 0.935064935064935, "prompt_id": "6d26cacb241c8ba0", "reading_canonical_log_prob": -12.380048751831055, "reading_distances": [32, 37, 42, 38], "reading_log_probs": [-10.624897003173828, -11.767257690429688, -10.710294723510742, -10.421581268310547], "reading_logprob_std": 0.6031161582542595, "tokenization_path_logprob_std": 53.7363198021057, "trajectory_balance_rmse": 804.9359474400983} {"canonical_length": 10, "canonical_nll": 1.6422578811645507, "generation_rouge_l_f1": 0.4444444444444445, "generation_token_f1": 0.888888888888889, "prompt_id": "d7683fe01587262b", "reading_canonical_log_prob": -16.422578811645508, "reading_distances": [40, 33, 45, 37], "reading_log_probs": [-17.11345100402832, -17.39969825744629, -17.427759170532227, -17.94487190246582], "reading_logprob_std": 0.34609670127171116, "tokenization_path_logprob_std": 26.20249758498611, "trajectory_balance_rmse": 224.1980219385066} {"canonical_length": 71, "canonical_nll": 0.7172979435450594, "generation_rouge_l_f1": 0.6438356164383562, "generation_token_f1": 0.6712328767123289, "prompt_id": "e455ade71af9293b", "reading_canonical_log_prob": -50.92815399169922, "reading_distances": [52, 34, 39, 53], "reading_log_probs": [-48.21617889404297, -50.52249526977539, -48.01166534423828, -49.658477783203125], "reading_logprob_std": 1.1973558174822327, "tokenization_path_logprob_std": 125.53820185958112, "trajectory_balance_rmse": 1490.435489316846} {"canonical_length": 334, "canonical_nll": 1.7695975846159244, "generation_rouge_l_f1": 0.1849912739965096, "generation_token_f1": 0.33158813263525305, "prompt_id": "83deb2f6c2a5fe75", "reading_canonical_log_prob": -591.0455932617188, "reading_distances": [43, 35, 31, 37], "reading_log_probs": [-576.6292724609375, -577.2310180664062, -582.0105590820312, -578.5148315429688], "reading_logprob_std": 2.408110893658817, "tokenization_path_logprob_std": 145.2460490286189, "trajectory_balance_rmse": 5769.938085148599} {"canonical_length": 251, "canonical_nll": 1.3133433064616533, "generation_rouge_l_f1": 0.19953596287703015, "generation_token_f1": 0.35266821345707655, "prompt_id": "cc738e3b9180c6b5", "reading_canonical_log_prob": -329.649169921875, "reading_distances": [11, 11, 12, 15], "reading_log_probs": [-322.14715576171875, -330.13739013671875, -325.00347900390625, -325.2291259765625], "reading_logprob_std": 3.3166246066625678, "tokenization_path_logprob_std": 150.34251751470853, "trajectory_balance_rmse": 4554.95908897549} {"canonical_length": 15, "canonical_nll": 0.07571935653686523, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "d829b51f56a6d0c5", "reading_canonical_log_prob": -1.1357903480529785, "reading_distances": [18, 16, 15, 19], "reading_log_probs": [-1.1384222507476807, -0.991034746170044, -1.1748957633972168, -0.9887286424636841], "reading_logprob_std": 0.09743799681442604, "tokenization_path_logprob_std": 46.69262576182369, "trajectory_balance_rmse": 386.1312945624656} {"canonical_length": 158, "canonical_nll": 1.3921147600004944, "generation_rouge_l_f1": 0.19935691318327975, "generation_token_f1": 0.4823151125401929, "prompt_id": "f53fbba7ee710f29", "reading_canonical_log_prob": -219.95413208007812, "reading_distances": [22, 23, 22, 29], "reading_log_probs": [-222.0466766357422, -218.5423583984375, -220.16635131835938, -221.29592895507812], "reading_logprob_std": 1.5241101540256512, "tokenization_path_logprob_std": 176.0489584787314, "trajectory_balance_rmse": 3643.7796143737264} {"canonical_length": 268, "canonical_nll": 1.2747948489972014, "generation_rouge_l_f1": 0.20040080160320642, "generation_token_f1": 0.4208416833667335, "prompt_id": "2ce5516156798d8a", "reading_canonical_log_prob": -341.64501953125, "reading_distances": [31, 29, 26, 30], "reading_log_probs": [-339.69610595703125, -335.5217590332031, -339.1085205078125, -336.05047607421875], "reading_logprob_std": 2.112602689537609, "tokenization_path_logprob_std": 316.48284495531664, "trajectory_balance_rmse": 5064.0929181749225} {"canonical_length": 28, "canonical_nll": 0.24428420407431467, "generation_rouge_l_f1": 0.8771929824561403, "generation_token_f1": 0.8771929824561403, "prompt_id": "83deea31656f880e", "reading_canonical_log_prob": -6.8399577140808105, "reading_distances": [33, 26, 28, 23], "reading_log_probs": [-10.256535530090332, -8.80905532836914, -9.221556663513184, -6.702960968017578], "reading_logprob_std": 1.4928548556489558, "tokenization_path_logprob_std": 56.09284094875642, "trajectory_balance_rmse": 504.4074248247155} {"canonical_length": 16, "canonical_nll": 0.2635642886161804, "generation_rouge_l_f1": 0.9444444444444444, "generation_token_f1": 0.9444444444444444, "prompt_id": "2a011ac439c0c89f", "reading_canonical_log_prob": -4.217028617858887, "reading_distances": [50, 50, 40, 47], "reading_log_probs": [-4.211326599121094, -4.015399932861328, -4.2902960777282715, -4.099938869476318], "reading_logprob_std": 0.12109956386473589, "tokenization_path_logprob_std": 44.48163331653873, "trajectory_balance_rmse": 361.7983168249415} {"canonical_length": 27, "canonical_nll": 1.2003970675998263, "generation_rouge_l_f1": 0.07999999999999999, "generation_token_f1": 0.07999999999999999, "prompt_id": "a62ae1af2ae45a72", "reading_canonical_log_prob": -32.41072082519531, "reading_distances": [27, 23, 29, 22], "reading_log_probs": [-31.052688598632812, -35.076576232910156, -35.987911224365234, -34.98240280151367], "reading_logprob_std": 2.1954725206169883, "tokenization_path_logprob_std": 45.40012874184822, "trajectory_balance_rmse": 411.8299114487697} {"canonical_length": 99, "canonical_nll": 1.6173453282828283, "generation_rouge_l_f1": 0.24260355029585798, "generation_token_f1": 0.3254437869822485, "prompt_id": "53aea8a2764798f8", "reading_canonical_log_prob": -160.1171875, "reading_distances": [34, 31, 28, 33], "reading_log_probs": [-162.23284912109375, -159.74246215820312, -161.21401977539062, -162.78561401367188], "reading_logprob_std": 1.3367171532343902, "tokenization_path_logprob_std": 132.48815133831414, "trajectory_balance_rmse": 2053.167326924525} {"canonical_length": 159, "canonical_nll": 0.5223499993858097, "generation_rouge_l_f1": 0.38153846153846155, "generation_token_f1": 0.5784615384615384, "prompt_id": "7dff62aaf925a1c6", "reading_canonical_log_prob": -83.05364990234375, "reading_distances": [30, 32, 34, 32], "reading_log_probs": [-86.65611267089844, -84.7322998046875, -85.78079223632812, -85.35044860839844], "reading_logprob_std": 0.8082202882628249, "tokenization_path_logprob_std": 51.78969386336405, "trajectory_balance_rmse": 2159.4513553301804} {"canonical_length": 217, "canonical_nll": 1.3784266880580358, "generation_rouge_l_f1": 0.26552462526766596, "generation_token_f1": 0.4967880085653106, "prompt_id": "5470eb105173aff8", "reading_canonical_log_prob": -299.11859130859375, "reading_distances": [25, 38, 33, 35], "reading_log_probs": [-301.0945129394531, -338.8706970214844, -297.66961669921875, -302.7510070800781], "reading_logprob_std": 19.299169930944366, "tokenization_path_logprob_std": 187.7995622785241, "trajectory_balance_rmse": 4619.768298230665} {"canonical_length": 15, "canonical_nll": 0.5914922078450521, "generation_rouge_l_f1": 0.5882352941176471, "generation_token_f1": 0.5882352941176471, "prompt_id": "861fa387c936a923", "reading_canonical_log_prob": -8.872383117675781, "reading_distances": [9, 9, 11, 10], "reading_log_probs": [-6.300222396850586, -8.39136028289795, -6.817190647125244, -6.736891269683838], "reading_logprob_std": 0.9152649562944448, "tokenization_path_logprob_std": 65.94790140300823, "trajectory_balance_rmse": 183.20866231638325} {"canonical_length": 11, "canonical_nll": 1.00372652574019, "generation_rouge_l_f1": 0.29090909090909095, "generation_token_f1": 0.29090909090909095, "prompt_id": "6c2751afcc2de989", "reading_canonical_log_prob": -11.04099178314209, "reading_distances": [11, 9, 9, 14], "reading_log_probs": [-14.53222942352295, -11.103523254394531, -16.681625366210938, -11.200464248657227], "reading_logprob_std": 2.717908755150928, "tokenization_path_logprob_std": 37.41273226518032, "trajectory_balance_rmse": 188.53867807069207} {"canonical_length": 69, "canonical_nll": 0.7159290037293365, "generation_rouge_l_f1": 0.2880658436213992, "generation_token_f1": 0.38683127572016457, "prompt_id": "2dc95060fe386398", "reading_canonical_log_prob": -49.39910125732422, "reading_distances": [21, 18, 24, 20], "reading_log_probs": [-48.66535186767578, -48.090049743652344, -48.997013092041016, -48.333465576171875], "reading_logprob_std": 0.39510012452533433, "tokenization_path_logprob_std": 131.1944462229792, "trajectory_balance_rmse": 1541.4946242379306} {"canonical_length": 367, "canonical_nll": 1.290272642546194, "generation_rouge_l_f1": 0.2360655737704918, "generation_token_f1": 0.43934426229508194, "prompt_id": "346df7037bb677e7", "reading_canonical_log_prob": -473.5300598144531, "reading_distances": [27, 25, 27, 28], "reading_log_probs": [-474.7356872558594, -475.0154113769531, -471.8922119140625, -474.32000732421875], "reading_logprob_std": 1.4279513193159141, "tokenization_path_logprob_std": 223.15523777812157, "trajectory_balance_rmse": 6495.549836204063} {"canonical_length": 168, "canonical_nll": 1.758023943219866, "generation_rouge_l_f1": 0.18625277161862527, "generation_token_f1": 0.3680709534368071, "prompt_id": "cccaf997f15db0c8", "reading_canonical_log_prob": -295.3480224609375, "reading_distances": [19, 12, 16, 12], "reading_log_probs": [-294.2404479980469, -293.0748596191406, -295.2312316894531, -293.345458984375], "reading_logprob_std": 0.9755578380298621, "tokenization_path_logprob_std": 104.92935381802606, "trajectory_balance_rmse": 3751.3416005580257} {"canonical_length": 15, "canonical_nll": 0.3540458679199219, "generation_rouge_l_f1": 0.2105263157894737, "generation_token_f1": 0.2105263157894737, "prompt_id": "c7d4acf2c03f4f98", "reading_canonical_log_prob": -5.310688018798828, "reading_distances": [15, 14, 19, 20], "reading_log_probs": [-7.551694393157959, -4.427370548248291, -5.161056041717529, -5.457561492919922], "reading_logprob_std": 1.3400698703025302, "tokenization_path_logprob_std": 28.40926011805485, "trajectory_balance_rmse": 216.1293545911208} {"canonical_length": 31, "canonical_nll": 2.010653588079637, "generation_rouge_l_f1": 0.2413793103448276, "generation_token_f1": 0.3103448275862069, "prompt_id": "f4a99dd0e66e3bbe", "reading_canonical_log_prob": -62.33026123046875, "reading_distances": [22, 18, 23, 19], "reading_log_probs": [-60.80146026611328, -61.45758056640625, -60.718379974365234, -61.074588775634766], "reading_logprob_std": 0.33316416164068885, "tokenization_path_logprob_std": 32.19873818365801, "trajectory_balance_rmse": 742.0532476788828} {"canonical_length": 51, "canonical_nll": 1.2905871821384804, "generation_rouge_l_f1": 0.20242914979757082, "generation_token_f1": 0.26720647773279355, "prompt_id": "b701439d534ac688", "reading_canonical_log_prob": -65.8199462890625, "reading_distances": [21, 24, 23, 21], "reading_log_probs": [-63.97413635253906, -63.18135452270508, -62.58613204956055, -63.144039154052734], "reading_logprob_std": 0.5708957070730648, "tokenization_path_logprob_std": 44.27296332430517, "trajectory_balance_rmse": 1110.2857392000667} {"canonical_length": 448, "canonical_nll": 1.3350230625697546, "generation_rouge_l_f1": 0.17872340425531913, "generation_token_f1": 0.3801418439716312, "prompt_id": "7fdb8e03f79c4b21", "reading_canonical_log_prob": -598.09033203125, "reading_distances": [15, 14, 13, 13], "reading_log_probs": [-597.7196044921875, -593.2154541015625, -595.8768310546875, -591.957275390625], "reading_logprob_std": 2.5967363958920093, "tokenization_path_logprob_std": 261.3414902205266, "trajectory_balance_rmse": 7348.868847033676} {"canonical_length": 392, "canonical_nll": 1.2566396752182318, "generation_rouge_l_f1": 0.19536423841059605, "generation_token_f1": 0.39403973509933765, "prompt_id": "79079ac51debbea1", "reading_canonical_log_prob": -492.6027526855469, "reading_distances": [26, 23, 23, 25], "reading_log_probs": [-485.13446044921875, -484.509521484375, -489.0760498046875, -487.3139953613281], "reading_logprob_std": 2.091652820899445, "tokenization_path_logprob_std": 356.19586189348746, "trajectory_balance_rmse": 6848.582460254167} {"canonical_length": 5, "canonical_nll": 2.3745468139648436, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "e0c4de6979cf5d26", "reading_canonical_log_prob": -11.872734069824219, "reading_distances": [19, 19, 26, 24], "reading_log_probs": [-10.74741268157959, -11.03858470916748, -11.196876525878906, -11.144475936889648], "reading_logprob_std": 0.20072227886064306, "tokenization_path_logprob_std": 13.32529589861882, "trajectory_balance_rmse": 107.60245484788582} {"canonical_length": 199, "canonical_nll": 1.5850351611573492, "generation_rouge_l_f1": 0.24821002386634847, "generation_token_f1": 0.3436754176610978, "prompt_id": "a5453cc0769ffb94", "reading_canonical_log_prob": -315.4219970703125, "reading_distances": [22, 13, 22, 20], "reading_log_probs": [-331.9806823730469, -318.31146240234375, -329.6263427734375, -328.226806640625], "reading_logprob_std": 6.019283893835513, "tokenization_path_logprob_std": 108.03765996545394, "trajectory_balance_rmse": 3261.0331825584735} {"canonical_length": 16, "canonical_nll": 1.16176438331604, "generation_rouge_l_f1": 0.23529411764705882, "generation_token_f1": 0.4117647058823529, "prompt_id": "c1820a57ac4bcbc9", "reading_canonical_log_prob": -18.58823013305664, "reading_distances": [17, 13, 13, 14], "reading_log_probs": [-18.54509925842285, -18.144763946533203, -18.605932235717773, -18.971012115478516], "reading_logprob_std": 0.33838049602957915, "tokenization_path_logprob_std": 53.939255397408694, "trajectory_balance_rmse": 406.30545036709566} {"canonical_length": 87, "canonical_nll": 0.38779217073287087, "generation_rouge_l_f1": 0.3674911660777386, "generation_token_f1": 0.4240282685512367, "prompt_id": "326b224fa8e4dc3a", "reading_canonical_log_prob": -33.737918853759766, "reading_distances": [25, 28, 21, 30], "reading_log_probs": [-34.330039978027344, -35.608192443847656, -31.548694610595703, -32.208335876464844], "reading_logprob_std": 1.8784935899543262, "tokenization_path_logprob_std": 133.6487348106153, "trajectory_balance_rmse": 1073.6045140022486} {"canonical_length": 210, "canonical_nll": 0.8757173810686384, "generation_rouge_l_f1": 0.22670025188916879, "generation_token_f1": 0.41309823677581864, "prompt_id": "a8f7e780841cbd23", "reading_canonical_log_prob": -183.90065002441406, "reading_distances": [15, 21, 12, 16], "reading_log_probs": [-181.67579650878906, -181.43667602539062, -183.40310668945312, -182.85336303710938], "reading_logprob_std": 0.9400138174275642, "tokenization_path_logprob_std": 160.10881666257862, "trajectory_balance_rmse": 2905.486471889692} {"canonical_length": 221, "canonical_nll": 1.4576708763433257, "generation_rouge_l_f1": 0.25263157894736843, "generation_token_f1": 0.5263157894736842, "prompt_id": "675093cc227e317b", "reading_canonical_log_prob": -322.145263671875, "reading_distances": [15, 17, 14, 19], "reading_log_probs": [-317.1756286621094, -318.6690979003906, -320.8636474609375, -320.4544677734375], "reading_logprob_std": 1.7018064010074592, "tokenization_path_logprob_std": 131.34271489238537, "trajectory_balance_rmse": 4526.807713285737} {"canonical_length": 327, "canonical_nll": 1.183395245753297, "generation_rouge_l_f1": 0.24790619765494137, "generation_token_f1": 0.47571189279731996, "prompt_id": "df931a7b15ed70cf", "reading_canonical_log_prob": -386.9702453613281, "reading_distances": [34, 35, 42, 34], "reading_log_probs": [-386.30413818359375, -387.70147705078125, -387.10211181640625, -386.3941955566406], "reading_logprob_std": 0.6561756103105723, "tokenization_path_logprob_std": 235.7659546210633, "trajectory_balance_rmse": 6886.407672509672} {"canonical_length": 267, "canonical_nll": 1.3367910778031367, "generation_rouge_l_f1": 0.23985239852398524, "generation_token_f1": 0.4723247232472325, "prompt_id": "64f98636b7dde269", "reading_canonical_log_prob": -356.9232177734375, "reading_distances": [27, 26, 32, 24], "reading_log_probs": [-354.49871826171875, -356.06243896484375, -355.45465087890625, -357.273193359375], "reading_logprob_std": 1.1618652743201727, "tokenization_path_logprob_std": 13.703621220638771, "trajectory_balance_rmse": 4572.50076332525} {"canonical_length": 110, "canonical_nll": 1.3888105912642046, "generation_rouge_l_f1": 0.19883040935672516, "generation_token_f1": 0.32163742690058483, "prompt_id": "3485bbb40ac08228", "reading_canonical_log_prob": -152.7691650390625, "reading_distances": [16, 15, 10, 14], "reading_log_probs": [-152.37416076660156, -155.99952697753906, -158.99766540527344, -156.16839599609375], "reading_logprob_std": 2.714658219826208, "tokenization_path_logprob_std": 37.633963118487046, "trajectory_balance_rmse": 2288.8093901215843} {"canonical_length": 20, "canonical_nll": 0.9610662460327148, "generation_rouge_l_f1": 0.0625, "generation_token_f1": 0.0703125, "prompt_id": "5de8b78dfeffe43c", "reading_canonical_log_prob": -19.221324920654297, "reading_distances": [13, 19, 12, 11], "reading_log_probs": [-17.493635177612305, -19.24123764038086, -18.783824920654297, -19.563983917236328], "reading_logprob_std": 0.9095368537488306, "tokenization_path_logprob_std": 31.343019032629158, "trajectory_balance_rmse": 266.25316606464384} {"canonical_length": 14, "canonical_nll": 2.229457582746233, "generation_rouge_l_f1": 0.2647058823529411, "generation_token_f1": 0.2647058823529411, "prompt_id": "75d59497e8843334", "reading_canonical_log_prob": -31.212406158447266, "reading_distances": [7, 7, 11, 10], "reading_log_probs": [-29.63637924194336, -28.930158615112305, -31.228538513183594, -28.098957061767578], "reading_logprob_std": 1.328066009844214, "tokenization_path_logprob_std": 45.49916584251871, "trajectory_balance_rmse": 234.69454977491026} {"canonical_length": 53, "canonical_nll": 2.38724100364829, "generation_rouge_l_f1": 0.2105263157894737, "generation_token_f1": 0.2894736842105263, "prompt_id": "9fe6000f50208baa", "reading_canonical_log_prob": -126.52377319335938, "reading_distances": [24, 22, 20, 23], "reading_log_probs": [-124.38655090332031, -127.17491912841797, -128.0873260498047, -125.03430938720703], "reading_logprob_std": 1.747043509606024, "tokenization_path_logprob_std": 57.54124845671711, "trajectory_balance_rmse": 1281.3061670142329} {"canonical_length": 13, "canonical_nll": 0.034613263148527876, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "2321fcc6e5fd7b96", "reading_canonical_log_prob": -0.4499724209308624, "reading_distances": [31, 35, 36, 45], "reading_log_probs": [-0.6351961493492126, -0.7115888595581055, -0.778737485408783, -0.4924358129501343], "reading_logprob_std": 0.12292481388482068, "tokenization_path_logprob_std": 62.417199658743314, "trajectory_balance_rmse": 178.43372837523142} {"canonical_length": 6, "canonical_nll": 1.6422452926635742, "generation_rouge_l_f1": 0.4444444444444444, "generation_token_f1": 0.4444444444444444, "prompt_id": "9ecc7ea82de4a5d6", "reading_canonical_log_prob": -9.853471755981445, "reading_distances": [47, 46, 47, 40], "reading_log_probs": [-10.76755142211914, -10.485591888427734, -9.172586441040039, -10.192938804626465], "reading_logprob_std": 0.6954819796308216, "tokenization_path_logprob_std": 13.328684851952959, "trajectory_balance_rmse": 66.4983259681438} {"canonical_length": 268, "canonical_nll": 1.122039225564074, "generation_rouge_l_f1": 0.22834645669291337, "generation_token_f1": 0.4173228346456693, "prompt_id": "aceb2f4e30eb8855", "reading_canonical_log_prob": -300.7065124511719, "reading_distances": [25, 27, 32, 27], "reading_log_probs": [-302.10015869140625, -303.2509460449219, -309.216552734375, -301.5475158691406], "reading_logprob_std": 3.5305459182935177, "tokenization_path_logprob_std": 190.94384703937382, "trajectory_balance_rmse": 4456.896663914957} {"canonical_length": 118, "canonical_nll": 1.3886050208140228, "generation_rouge_l_f1": 0.3333333333333333, "generation_token_f1": 0.5042735042735043, "prompt_id": "c575c5fadea559fc", "reading_canonical_log_prob": -163.8553924560547, "reading_distances": [53, 46, 38, 46], "reading_log_probs": [-157.74172973632812, -159.31997680664062, -155.44712829589844, -158.2427978515625], "reading_logprob_std": 1.6325380851699547, "tokenization_path_logprob_std": 43.40530935813775, "trajectory_balance_rmse": 2481.592547575121} {"canonical_length": 61, "canonical_nll": 2.383665991611168, "generation_rouge_l_f1": 0.23255813953488372, "generation_token_f1": 0.3488372093023256, "prompt_id": "3adbad55ce76ffec", "reading_canonical_log_prob": -145.40362548828125, "reading_distances": [25, 21, 24, 19], "reading_log_probs": [-141.56187438964844, -139.88107299804688, -144.1549072265625, -140.04364013671875], "reading_logprob_std": 1.9800789552762224, "tokenization_path_logprob_std": 90.3298176121744, "trajectory_balance_rmse": 1093.1524157513531} {"canonical_length": 12, "canonical_nll": 2.1006007194519043, "generation_rouge_l_f1": 0.6, "generation_token_f1": 0.6, "prompt_id": "70c54675265e6b63", "reading_canonical_log_prob": -25.20720863342285, "reading_distances": [27, 34, 22, 28], "reading_log_probs": [-24.684627532958984, -25.70970916748047, -25.740093231201172, -25.562070846557617], "reading_logprob_std": 0.4990924430691953, "tokenization_path_logprob_std": 42.73756427391558, "trajectory_balance_rmse": 323.7801696624095} {"canonical_length": 11, "canonical_nll": 1.6462149186567827, "generation_rouge_l_f1": 0.3636363636363636, "generation_token_f1": 0.3636363636363636, "prompt_id": "5e3e21467d43fd88", "reading_canonical_log_prob": -18.10836410522461, "reading_distances": [16, 20, 12, 17], "reading_log_probs": [-15.347213745117188, -15.917953491210938, -16.431211471557617, -16.430763244628906], "reading_logprob_std": 0.5165012298379038, "tokenization_path_logprob_std": 16.398870958933422, "trajectory_balance_rmse": 237.38210662164408} {"canonical_length": 17, "canonical_nll": 0.23558911155251896, "generation_rouge_l_f1": 0.4210526315789474, "generation_token_f1": 0.4210526315789474, "prompt_id": "46d8b242e85d5c4e", "reading_canonical_log_prob": -4.005014896392822, "reading_distances": [12, 8, 11, 10], "reading_log_probs": [-4.624279022216797, -4.624998569488525, -4.629193305969238, -4.409730434417725], "reading_logprob_std": 0.10823496132654001, "tokenization_path_logprob_std": 14.318198488996648, "trajectory_balance_rmse": 118.41846489550795} {"canonical_length": 48, "canonical_nll": 1.3021903038024902, "generation_rouge_l_f1": 0.17737003058103973, "generation_token_f1": 0.20183486238532108, "prompt_id": "6509666af0922805", "reading_canonical_log_prob": -62.50513458251953, "reading_distances": [30, 28, 32, 23], "reading_log_probs": [-61.46528625488281, -61.156150817871094, -59.2548942565918, -58.639034271240234], "reading_logprob_std": 1.393408097742625, "tokenization_path_logprob_std": 80.19281451511038, "trajectory_balance_rmse": 852.6432987401387} {"canonical_length": 175, "canonical_nll": 1.054526628766741, "generation_rouge_l_f1": 0.23232323232323232, "generation_token_f1": 0.505050505050505, "prompt_id": "42f6dd40e9d218d5", "reading_canonical_log_prob": -184.5421600341797, "reading_distances": [8, 15, 12, 13], "reading_log_probs": [-184.74147033691406, -182.3465576171875, -181.77244567871094, -181.06854248046875], "reading_logprob_std": 1.594248002321088, "tokenization_path_logprob_std": 140.94138744946153, "trajectory_balance_rmse": 2971.3642222665003} {"canonical_length": 85, "canonical_nll": 1.0817124310661765, "generation_rouge_l_f1": 0.25, "generation_token_f1": 0.3928571428571429, "prompt_id": "68880c3a241cb55b", "reading_canonical_log_prob": -91.945556640625, "reading_distances": [35, 41, 42, 37], "reading_log_probs": [-94.30908966064453, -93.69947814941406, -94.0931396484375, -93.33010864257812], "reading_logprob_std": 0.43304002513505446, "tokenization_path_logprob_std": 161.74617795901486, "trajectory_balance_rmse": 1638.7103162697717} {"canonical_length": 14, "canonical_nll": 1.3706406184605189, "generation_rouge_l_f1": 0.3333333333333333, "generation_token_f1": 0.4666666666666667, "prompt_id": "a0d2e6c5b9eebb8b", "reading_canonical_log_prob": -19.188968658447266, "reading_distances": [26, 28, 25, 32], "reading_log_probs": [-19.440223693847656, -20.40995216369629, -20.288124084472656, -19.083858489990234], "reading_logprob_std": 0.6461376704801373, "tokenization_path_logprob_std": 36.02315620830845, "trajectory_balance_rmse": 342.3265279625378} {"canonical_length": 256, "canonical_nll": 0.6092175245285034, "generation_rouge_l_f1": 0.2851782363977486, "generation_token_f1": 0.42026266416510316, "prompt_id": "c51ea1a2a74eb3ff", "reading_canonical_log_prob": -155.95968627929688, "reading_distances": [28, 24, 27, 20], "reading_log_probs": [-158.71142578125, -159.28387451171875, -160.76504516601562, -161.1189727783203], "reading_logprob_std": 1.1557122471874468, "tokenization_path_logprob_std": 196.59947923733478, "trajectory_balance_rmse": 3154.682591595317} {"canonical_length": 61, "canonical_nll": 0.767165512335105, "generation_rouge_l_f1": 0.33149171270718225, "generation_token_f1": 0.430939226519337, "prompt_id": "9ce480763a8f4a02", "reading_canonical_log_prob": -46.797096252441406, "reading_distances": [30, 32, 30, 29], "reading_log_probs": [-51.38280487060547, -50.510658264160156, -50.185791015625, -51.62793731689453], "reading_logprob_std": 0.6884287032491789, "tokenization_path_logprob_std": 82.09690730134194, "trajectory_balance_rmse": 909.7848378622364} {"canonical_length": 17, "canonical_nll": 0.5998995724846335, "generation_rouge_l_f1": 0.8125000000000001, "generation_token_f1": 0.8125000000000001, "prompt_id": "690ac1ef381d5161", "reading_canonical_log_prob": -10.19829273223877, "reading_distances": [32, 37, 39, 41], "reading_log_probs": [-9.27123737335205, -10.294473648071289, -8.842964172363281, -11.244124412536621], "reading_logprob_std": 1.0761428231372585, "tokenization_path_logprob_std": 15.007278105823852, "trajectory_balance_rmse": 279.5412974907455} {"canonical_length": 49, "canonical_nll": 1.079357692173549, "generation_rouge_l_f1": 0.37593984962406013, "generation_token_f1": 0.46616541353383456, "prompt_id": "f544199aa0f474b9", "reading_canonical_log_prob": -52.888526916503906, "reading_distances": [20, 18, 19, 19], "reading_log_probs": [-52.380157470703125, -52.15701675415039, -51.61628341674805, -51.82936096191406], "reading_logprob_std": 0.3393406651633332, "tokenization_path_logprob_std": 86.71426834430946, "trajectory_balance_rmse": 958.9360831623835} {"canonical_length": 236, "canonical_nll": 0.5226457240217823, "generation_rouge_l_f1": 0.14334470989761092, "generation_token_f1": 0.30716723549488056, "prompt_id": "672d37d3ef4c74d7", "reading_canonical_log_prob": -123.34439086914062, "reading_distances": [20, 24, 25, 26], "reading_log_probs": [-122.7887191772461, -125.0074691772461, -124.8199691772461, -125.24481964111328], "reading_logprob_std": 1.1311221206894044, "tokenization_path_logprob_std": 152.57090595178192, "trajectory_balance_rmse": 3852.965623084283} {"canonical_length": 76, "canonical_nll": 1.0826329683002673, "generation_rouge_l_f1": 0.2518518518518519, "generation_token_f1": 0.4222222222222222, "prompt_id": "e905bcd075fa835d", "reading_canonical_log_prob": -82.28010559082031, "reading_distances": [19, 17, 18, 19], "reading_log_probs": [-81.47286987304688, -82.36693572998047, -81.34915161132812, -81.11781311035156], "reading_logprob_std": 0.5469944090692623, "tokenization_path_logprob_std": 75.98163827563918, "trajectory_balance_rmse": 1617.5682313758273} {"canonical_length": 10, "canonical_nll": 0.6779260635375977, "generation_rouge_l_f1": 0.5555555555555556, "generation_token_f1": 0.7777777777777778, "prompt_id": "1d805df4e7b3cce0", "reading_canonical_log_prob": -6.779260635375977, "reading_distances": [34, 41, 36, 34], "reading_log_probs": [-7.384090900421143, -6.711420059204102, -7.795100688934326, -7.685545921325684], "reading_logprob_std": 0.48713254408821205, "tokenization_path_logprob_std": 29.3677250358441, "trajectory_balance_rmse": 212.52761296596134} {"canonical_length": 140, "canonical_nll": 1.5278885977608816, "generation_rouge_l_f1": 0.1553398058252427, "generation_token_f1": 0.30097087378640774, "prompt_id": "6a438344f95ac9cf", "reading_canonical_log_prob": -213.90440368652344, "reading_distances": [25, 26, 18, 25], "reading_log_probs": [-217.967529296875, -213.2364044189453, -213.42413330078125, -211.90206909179688], "reading_logprob_std": 2.644934933697097, "tokenization_path_logprob_std": 166.17019965547277, "trajectory_balance_rmse": 3076.4887287530223} {"canonical_length": 103, "canonical_nll": 1.8371436850538532, "generation_rouge_l_f1": 0.2916666666666667, "generation_token_f1": 0.4166666666666667, "prompt_id": "208ec2e52d9d1ce6", "reading_canonical_log_prob": -189.22579956054688, "reading_distances": [32, 29, 36, 30], "reading_log_probs": [-186.7635498046875, -185.04010009765625, -188.78433227539062, -185.79994201660156], "reading_logprob_std": 1.6198148097635021, "tokenization_path_logprob_std": 135.12489587733558, "trajectory_balance_rmse": 2007.580306734313} {"canonical_length": 14, "canonical_nll": 0.4061450617653983, "generation_rouge_l_f1": 0.761904761904762, "generation_token_f1": 0.8571428571428571, "prompt_id": "6a81e01870d01f92", "reading_canonical_log_prob": -5.686030864715576, "reading_distances": [25, 28, 31, 24], "reading_log_probs": [-5.656795501708984, -6.478857040405273, -6.375177383422852, -6.274765968322754], "reading_logprob_std": 0.3692594264474386, "tokenization_path_logprob_std": 39.38012806986793, "trajectory_balance_rmse": 377.0392188775353} {"canonical_length": 274, "canonical_nll": 1.1899004191377738, "generation_rouge_l_f1": 0.19713261648745517, "generation_token_f1": 0.3799283154121864, "prompt_id": "7857d9054d4c3033", "reading_canonical_log_prob": -326.03271484375, "reading_distances": [29, 31, 23, 30], "reading_log_probs": [-320.625732421875, -321.18206787109375, -320.2533264160156, -321.3183288574219], "reading_logprob_std": 0.4952476766533772, "tokenization_path_logprob_std": 171.39284138906, "trajectory_balance_rmse": 4240.304753575118} {"canonical_length": 158, "canonical_nll": 1.2903350636928896, "generation_rouge_l_f1": 0.2557544757033248, "generation_token_f1": 0.35805626598465473, "prompt_id": "77515334ad84a428", "reading_canonical_log_prob": -203.87294006347656, "reading_distances": [19, 18, 20, 16], "reading_log_probs": [-205.4605712890625, -206.83901977539062, -204.16622924804688, -203.0662078857422], "reading_logprob_std": 1.6303471872328854, "tokenization_path_logprob_std": 103.37603287743488, "trajectory_balance_rmse": 2538.4479396067495} {"canonical_length": 43, "canonical_nll": 0.7534680033839026, "generation_rouge_l_f1": 0.40336134453781514, "generation_token_f1": 0.5042016806722689, "prompt_id": "78695f86d2d683db", "reading_canonical_log_prob": -32.39912414550781, "reading_distances": [40, 37, 35, 31], "reading_log_probs": [-34.70299530029297, -34.97455978393555, -33.64483642578125, -34.37504577636719], "reading_logprob_std": 0.574585743014193, "tokenization_path_logprob_std": 140.56213058084538, "trajectory_balance_rmse": 1123.8882934070068} {"canonical_length": 8, "canonical_nll": 0.17922836542129517, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "73a081a89d8a40bd", "reading_canonical_log_prob": -1.4338269233703613, "reading_distances": [28, 28, 32, 25], "reading_log_probs": [-1.1988227367401123, -1.1974457502365112, -1.4848518371582031, -1.1289262771606445], "reading_logprob_std": 0.15829304014215922, "tokenization_path_logprob_std": 29.276354899276143, "trajectory_balance_rmse": 101.75887268032653} {"canonical_length": 84, "canonical_nll": 1.1016110011509486, "generation_rouge_l_f1": 0.1474358974358974, "generation_token_f1": 0.2692307692307692, "prompt_id": "7879ca5705298aac", "reading_canonical_log_prob": -92.53532409667969, "reading_distances": [46, 46, 41, 35], "reading_log_probs": [-92.8335952758789, -95.03125762939453, -96.73382568359375, -94.17691040039062], "reading_logprob_std": 1.6333114499238797, "tokenization_path_logprob_std": 112.46663946280907, "trajectory_balance_rmse": 1223.90648142819} {"canonical_length": 355, "canonical_nll": 1.3441913443551936, "generation_rouge_l_f1": 0.18357487922705315, "generation_token_f1": 0.3639291465378422, "prompt_id": "2e88e8dc9b895643", "reading_canonical_log_prob": -477.18792724609375, "reading_distances": [23, 20, 20, 21], "reading_log_probs": [-473.37945556640625, -473.95892333984375, -471.2742919921875, -471.3862609863281], "reading_logprob_std": 1.37169892597911, "tokenization_path_logprob_std": 197.4225837738071, "trajectory_balance_rmse": 7538.413223104325} {"canonical_length": 343, "canonical_nll": 1.3675983751480503, "generation_rouge_l_f1": 0.20066889632107024, "generation_token_f1": 0.3913043478260869, "prompt_id": "a81b52bde10fe226", "reading_canonical_log_prob": -469.08624267578125, "reading_distances": [19, 20, 20, 19], "reading_log_probs": [-471.6629638671875, -471.5849914550781, -469.6875305175781, -471.5572509765625], "reading_logprob_std": 0.9581481256950287, "tokenization_path_logprob_std": 76.39745267839301, "trajectory_balance_rmse": 6728.620315039443} {"canonical_length": 269, "canonical_nll": 1.1432229648292287, "generation_rouge_l_f1": 0.27087198515769945, "generation_token_f1": 0.5194805194805194, "prompt_id": "ff000a10ec04c16c", "reading_canonical_log_prob": -307.5269775390625, "reading_distances": [33, 25, 32, 26], "reading_log_probs": [-306.2564392089844, -307.27178955078125, -306.1556396484375, -306.5751953125], "reading_logprob_std": 0.5041298951392604, "tokenization_path_logprob_std": 318.6487385004565, "trajectory_balance_rmse": 5513.192083853545} {"canonical_length": 405, "canonical_nll": 1.2220398702739197, "generation_rouge_l_f1": 0.209375, "generation_token_f1": 0.425, "prompt_id": "e7f10884569482cf", "reading_canonical_log_prob": -494.9261474609375, "reading_distances": [31, 38, 41, 31], "reading_log_probs": [-487.61627197265625, -485.587890625, -485.51678466796875, -488.5953369140625], "reading_logprob_std": 1.5277439566292021, "tokenization_path_logprob_std": 264.1732842886547, "trajectory_balance_rmse": 6609.105027564463} {"canonical_length": 10, "canonical_nll": 1.1115167617797852, "generation_rouge_l_f1": 0.8750000000000001, "generation_token_f1": 0.8750000000000001, "prompt_id": "7d1f1626ed9a7e46", "reading_canonical_log_prob": -11.115167617797852, "reading_distances": [34, 38, 35, 40], "reading_log_probs": [-12.585197448730469, -10.961586952209473, -11.57926082611084, -11.021236419677734], "reading_logprob_std": 0.7522446380212588, "tokenization_path_logprob_std": 23.73282879485396, "trajectory_balance_rmse": 245.51205687209907} {"canonical_length": 41, "canonical_nll": 1.8238756133288871, "generation_rouge_l_f1": 0.16574585635359118, "generation_token_f1": 0.25414364640883974, "prompt_id": "8de4f5d47e6be004", "reading_canonical_log_prob": -74.77890014648438, "reading_distances": [13, 16, 15, 15], "reading_log_probs": [-75.7681884765625, -76.13119506835938, -76.37394714355469, -74.02117919921875], "reading_logprob_std": 1.0644784051896385, "tokenization_path_logprob_std": 93.24831172505247, "trajectory_balance_rmse": 928.546141213633} {"canonical_length": 187, "canonical_nll": 1.2441889308990641, "generation_rouge_l_f1": 0.34500000000000003, "generation_token_f1": 0.575, "prompt_id": "37dc889754565fdd", "reading_canonical_log_prob": -232.663330078125, "reading_distances": [32, 40, 40, 36], "reading_log_probs": [-231.3343505859375, -229.6781463623047, -228.226806640625, -229.9311065673828], "reading_logprob_std": 1.2729214976928134, "tokenization_path_logprob_std": 126.55732957388392, "trajectory_balance_rmse": 3768.219079857312} {"canonical_length": 81, "canonical_nll": 1.3399364330150463, "generation_rouge_l_f1": 0.2857142857142857, "generation_token_f1": 0.4242424242424242, "prompt_id": "fccb0a4a57193f97", "reading_canonical_log_prob": -108.53485107421875, "reading_distances": [24, 26, 24, 21], "reading_log_probs": [-106.22552490234375, -104.75182342529297, -104.07381439208984, -104.41322326660156], "reading_logprob_std": 0.9476126178886243, "tokenization_path_logprob_std": 105.9908324214743, "trajectory_balance_rmse": 1771.9307900135316} {"canonical_length": 8, "canonical_nll": 0.4374772310256958, "generation_rouge_l_f1": 0.7272727272727272, "generation_token_f1": 0.7272727272727272, "prompt_id": "a29eafa220e503a1", "reading_canonical_log_prob": -3.4998178482055664, "reading_distances": [35, 27, 28, 37], "reading_log_probs": [-3.5018203258514404, -4.019262313842773, -3.645768404006958, -3.8422186374664307], "reading_logprob_std": 0.22615870698600915, "tokenization_path_logprob_std": 46.925297755644095, "trajectory_balance_rmse": 191.9399938608073} {"canonical_length": 20, "canonical_nll": 1.4957195281982423, "generation_rouge_l_f1": 0.10416666666666667, "generation_token_f1": 0.125, "prompt_id": "276e3276fb9b91f2", "reading_canonical_log_prob": -29.914390563964844, "reading_distances": [40, 49, 46, 48], "reading_log_probs": [-30.838966369628906, -29.60580825805664, -30.196277618408203, -32.73514175415039], "reading_logprob_std": 1.357584317301432, "tokenization_path_logprob_std": 31.305273202257595, "trajectory_balance_rmse": 456.7987274537406} {"canonical_length": 229, "canonical_nll": 1.272408031480281, "generation_rouge_l_f1": 0.19277108433734938, "generation_token_f1": 0.3614457831325302, "prompt_id": "14d153fd0e1b36b1", "reading_canonical_log_prob": -291.3814392089844, "reading_distances": [15, 21, 16, 22], "reading_log_probs": [-287.4231872558594, -286.9220886230469, -285.7887878417969, -286.08746337890625], "reading_logprob_std": 0.7514816806019137, "tokenization_path_logprob_std": 131.62126358745076, "trajectory_balance_rmse": 4211.61740980579} {"canonical_length": 315, "canonical_nll": 0.45615040612599206, "generation_rouge_l_f1": 0.3266055045871559, "generation_token_f1": 0.47339449541284406, "prompt_id": "0ee97cbd046603bc", "reading_canonical_log_prob": -143.6873779296875, "reading_distances": [25, 28, 24, 29], "reading_log_probs": [-147.02281188964844, -147.515869140625, -147.597412109375, -147.62071228027344], "reading_logprob_std": 0.2812086977092085, "tokenization_path_logprob_std": 140.06675171686499, "trajectory_balance_rmse": 4669.689491994661} {"canonical_length": 282, "canonical_nll": 0.8560579989818816, "generation_rouge_l_f1": 0.25362318840579706, "generation_token_f1": 0.4927536231884058, "prompt_id": "740c83c01085ccfc", "reading_canonical_log_prob": -241.40835571289062, "reading_distances": [16, 17, 28, 24], "reading_log_probs": [-246.61264038085938, -243.04110717773438, -243.57730102539062, -242.7896728515625], "reading_logprob_std": 1.7690698204798139, "tokenization_path_logprob_std": 246.62377004269183, "trajectory_balance_rmse": 6231.744527830849} {"canonical_length": 128, "canonical_nll": 1.202700138092041, "generation_rouge_l_f1": 0.19130434782608693, "generation_token_f1": 0.4115942028985507, "prompt_id": "055891d31fa97386", "reading_canonical_log_prob": -153.94561767578125, "reading_distances": [26, 24, 21, 22], "reading_log_probs": [-153.81692504882812, -153.41297912597656, -156.27835083007812, -156.4629669189453], "reading_logprob_std": 1.6013065083392772, "tokenization_path_logprob_std": 341.8912268937635, "trajectory_balance_rmse": 2425.130276906349} {"canonical_length": 197, "canonical_nll": 1.195664459073604, "generation_rouge_l_f1": 0.16204690831556504, "generation_token_f1": 0.42217484008528783, "prompt_id": "54415ac83a3ddec9", "reading_canonical_log_prob": -235.5458984375, "reading_distances": [20, 19, 19, 19], "reading_log_probs": [-230.57371520996094, -230.68096923828125, -230.02554321289062, -231.28326416015625], "reading_logprob_std": 0.5155628078899768, "tokenization_path_logprob_std": 136.67199532716268, "trajectory_balance_rmse": 3358.981674509489} {"canonical_length": 71, "canonical_nll": 1.4735277202767385, "generation_rouge_l_f1": 0.3146067415730337, "generation_token_f1": 0.4606741573033708, "prompt_id": "6bb4c6466fd9ddad", "reading_canonical_log_prob": -104.62046813964844, "reading_distances": [11, 17, 13, 18], "reading_log_probs": [-101.9673843383789, -101.76094818115234, -101.30044555664062, -105.76661682128906], "reading_logprob_std": 2.0640922212061534, "tokenization_path_logprob_std": 111.82629292739473, "trajectory_balance_rmse": 1340.1261879666886} {"canonical_length": 96, "canonical_nll": 0.9092378616333008, "generation_rouge_l_f1": 0.28181818181818186, "generation_token_f1": 0.43636363636363634, "prompt_id": "5ca3d0261fdb6645", "reading_canonical_log_prob": -87.28683471679688, "reading_distances": [20, 27, 25, 23], "reading_log_probs": [-90.97488403320312, -91.51575469970703, -91.95162963867188, -92.69664001464844], "reading_logprob_std": 0.7274690294154691, "tokenization_path_logprob_std": 171.27797200996454, "trajectory_balance_rmse": 1911.5109473398247} {"canonical_length": 120, "canonical_nll": 1.3683268229166667, "generation_rouge_l_f1": 0.2105263157894737, "generation_token_f1": 0.38421052631578945, "prompt_id": "2749992f70103af2", "reading_canonical_log_prob": -164.19921875, "reading_distances": [38, 44, 37, 44], "reading_log_probs": [-162.39950561523438, -169.91104125976562, -164.0196533203125, -167.00079345703125], "reading_logprob_std": 3.3202043959372274, "tokenization_path_logprob_std": 158.26016994706936, "trajectory_balance_rmse": 2703.548083665986} {"canonical_length": 18, "canonical_nll": 0.026426596773995295, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "67a8cb2f3ba290b7", "reading_canonical_log_prob": -0.4756787419319153, "reading_distances": [39, 50, 47, 39], "reading_log_probs": [-0.7337608337402344, -0.7780328989028931, -0.7354730367660522, -0.7581502199172974], "reading_logprob_std": 0.020973608367597936, "tokenization_path_logprob_std": 51.443634063179914, "trajectory_balance_rmse": 345.3723202976844} {"canonical_length": 144, "canonical_nll": 0.973701900906033, "generation_rouge_l_f1": 0.23844282238442824, "generation_token_f1": 0.3746958637469586, "prompt_id": "14aa2b70fa0eb52a", "reading_canonical_log_prob": -140.21307373046875, "reading_distances": [32, 26, 28, 30], "reading_log_probs": [-141.92515563964844, -140.87442016601562, -139.2043914794922, -140.27667236328125], "reading_logprob_std": 1.1372548932417568, "tokenization_path_logprob_std": 219.11607092840597, "trajectory_balance_rmse": 2398.9566003624786} {"canonical_length": 46, "canonical_nll": 1.4505006541376528, "generation_rouge_l_f1": 0.12063492063492062, "generation_token_f1": 0.15238095238095237, "prompt_id": "632380e80642e576", "reading_canonical_log_prob": -66.72303009033203, "reading_distances": [26, 32, 28, 25], "reading_log_probs": [-67.05091094970703, -68.81867980957031, -66.62120056152344, -67.50083923339844], "reading_logprob_std": 0.9509414739676152, "tokenization_path_logprob_std": 70.31270202077877, "trajectory_balance_rmse": 901.696985635842} {"canonical_length": 186, "canonical_nll": 1.3404755130890877, "generation_rouge_l_f1": 0.3175355450236967, "generation_token_f1": 0.4928909952606635, "prompt_id": "e0e28697fbbba739", "reading_canonical_log_prob": -249.3284454345703, "reading_distances": [36, 28, 34, 28], "reading_log_probs": [-248.8291015625, -247.95578002929688, -250.40985107421875, -246.1273651123047], "reading_logprob_std": 1.7857326036686783, "tokenization_path_logprob_std": 341.35204769241886, "trajectory_balance_rmse": 3903.084633812913} {"canonical_length": 41, "canonical_nll": 1.066709751036109, "generation_rouge_l_f1": 0.3259259259259259, "generation_token_f1": 0.42962962962962964, "prompt_id": "dd2bdffacb538a3a", "reading_canonical_log_prob": -43.73509979248047, "reading_distances": [145, 138, 159, 126], "reading_log_probs": [-54.07097625732422, -47.29780578613281, -53.69395065307617, -48.61852264404297], "reading_logprob_std": 3.46605165132261, "tokenization_path_logprob_std": 91.79301076503452, "trajectory_balance_rmse": 975.9459034164787} {"canonical_length": 11, "canonical_nll": 1.2662467089566318, "generation_rouge_l_f1": 0.05511811023622047, "generation_token_f1": 0.06299212598425198, "prompt_id": "fbdd1075d5b5f58e", "reading_canonical_log_prob": -13.92871379852295, "reading_distances": [25, 19, 24, 15], "reading_log_probs": [-12.801197052001953, -12.071893692016602, -12.292097091674805, -12.848426818847656], "reading_logprob_std": 0.38234940510291016, "tokenization_path_logprob_std": 35.34768986727675, "trajectory_balance_rmse": 180.09781665670963} {"canonical_length": 25, "canonical_nll": 2.0523846435546873, "generation_rouge_l_f1": 0.2571428571428572, "generation_token_f1": 0.2571428571428572, "prompt_id": "71ff4b9aa3b9c919", "reading_canonical_log_prob": -51.30961608886719, "reading_distances": [30, 22, 22, 23], "reading_log_probs": [-48.463478088378906, -50.9462890625, -49.075748443603516, -47.73174285888672], "reading_logprob_std": 1.375779847972267, "tokenization_path_logprob_std": 104.06262869359173, "trajectory_balance_rmse": 517.5398115895033} {"canonical_length": 227, "canonical_nll": 0.5589798578606828, "generation_rouge_l_f1": 0.17463617463617467, "generation_token_f1": 0.4033264033264033, "prompt_id": "84c9e255e3abf202", "reading_canonical_log_prob": -126.888427734375, "reading_distances": [29, 23, 23, 24], "reading_log_probs": [-136.7382049560547, -136.34217834472656, -138.6600341796875, -134.78302001953125], "reading_logprob_std": 1.5944612969703822, "tokenization_path_logprob_std": 119.2284331545223, "trajectory_balance_rmse": 3951.769691660965} {"canonical_length": 20, "canonical_nll": 0.32970609664916994, "generation_rouge_l_f1": 0.7096774193548387, "generation_token_f1": 0.7096774193548387, "prompt_id": "7aac721bb785059a", "reading_canonical_log_prob": -6.594121932983398, "reading_distances": [19, 22, 28, 23], "reading_log_probs": [-6.070326805114746, -6.644958019256592, -6.129371643066406, -6.205312728881836], "reading_logprob_std": 0.26089475303695764, "tokenization_path_logprob_std": 54.75325685105369, "trajectory_balance_rmse": 374.22521722972743} {"canonical_length": 16, "canonical_nll": 0.5788739919662476, "generation_rouge_l_f1": 0.8666666666666667, "generation_token_f1": 0.8666666666666667, "prompt_id": "e8e5a8f77a9d33c4", "reading_canonical_log_prob": -9.261983871459961, "reading_distances": [43, 41, 42, 43], "reading_log_probs": [-8.502872467041016, -9.775146484375, -8.339458465576172, -9.455971717834473], "reading_logprob_std": 0.7049502319495132, "tokenization_path_logprob_std": 44.404903421082885, "trajectory_balance_rmse": 262.1364001634378} {"canonical_length": 139, "canonical_nll": 1.515576040144447, "generation_rouge_l_f1": 0.13725490196078433, "generation_token_f1": 0.2303921568627451, "prompt_id": "b0bb5fcadef34afc", "reading_canonical_log_prob": -210.66506958007812, "reading_distances": [16, 10, 16, 17], "reading_log_probs": [-208.31288146972656, -211.01675415039062, -208.9970245361328, -208.919921875], "reading_logprob_std": 1.177194241560958, "tokenization_path_logprob_std": 98.584398408738, "trajectory_balance_rmse": 2413.971379308298} {"canonical_length": 214, "canonical_nll": 1.5327693190530083, "generation_rouge_l_f1": 0.19867549668874174, "generation_token_f1": 0.31125827814569534, "prompt_id": "6db457998b4e4d8a", "reading_canonical_log_prob": -328.01263427734375, "reading_distances": [39, 38, 39, 35], "reading_log_probs": [-316.4171142578125, -315.6484375, -313.4912109375, -317.6844482421875], "reading_logprob_std": 1.7592639299682982, "tokenization_path_logprob_std": 236.83377245906948, "trajectory_balance_rmse": 3633.531329187535} {"canonical_length": 124, "canonical_nll": 1.5307822688933341, "generation_rouge_l_f1": 0.1547277936962751, "generation_token_f1": 0.24641833810888253, "prompt_id": "b7fe2d1d82e7517d", "reading_canonical_log_prob": -189.81700134277344, "reading_distances": [21, 28, 32, 26], "reading_log_probs": [-186.7884979248047, -188.5567169189453, -188.69593811035156, -187.69583129882812], "reading_logprob_std": 0.8826522791097788, "tokenization_path_logprob_std": 85.11295275678981, "trajectory_balance_rmse": 2144.618816327513} {"canonical_length": 3, "canonical_nll": 0.03722161054611206, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "fe8dff1c5ccd0457", "reading_canonical_log_prob": -0.11166483163833618, "reading_distances": [57, 77, 54, 55], "reading_log_probs": [-0.18968729674816132, -0.38235414028167725, -0.2428065687417984, -0.5206857919692993], "reading_logprob_std": 0.1486966169688533, "tokenization_path_logprob_std": 14.024203072054759, "trajectory_balance_rmse": 66.55741568304607} {"canonical_length": 36, "canonical_nll": 1.4913809034559462, "generation_rouge_l_f1": 0.26666666666666666, "generation_token_f1": 0.3259259259259259, "prompt_id": "007cda46c19b41d2", "reading_canonical_log_prob": -53.68971252441406, "reading_distances": [22, 16, 22, 16], "reading_log_probs": [-57.14442825317383, -61.241294860839844, -58.778839111328125, -57.87916946411133], "reading_logprob_std": 1.783546104372429, "tokenization_path_logprob_std": 134.81584420833218, "trajectory_balance_rmse": 809.5685650811487} {"canonical_length": 121, "canonical_nll": 1.62306099883781, "generation_rouge_l_f1": 0.144, "generation_token_f1": 0.25599999999999995, "prompt_id": "0f24b7ddfdbc9dca", "reading_canonical_log_prob": -196.390380859375, "reading_distances": [20, 21, 20, 22], "reading_log_probs": [-198.42320251464844, -199.86019897460938, -199.1976776123047, -200.77407836914062], "reading_logprob_std": 0.9979369233355081, "tokenization_path_logprob_std": 88.42942476403465, "trajectory_balance_rmse": 1913.5660871108707} {"canonical_length": 4, "canonical_nll": 0.9966735243797302, "generation_rouge_l_f1": 0.25, "generation_token_f1": 0.25, "prompt_id": "0131c4e6a39cb194", "reading_canonical_log_prob": -3.986694097518921, "reading_distances": [17, 27, 22, 21], "reading_log_probs": [-3.8642518520355225, -4.28325891494751, -3.7799232006073, -4.522546768188477], "reading_logprob_std": 0.35096601426404667, "tokenization_path_logprob_std": 4.798844941617778, "trajectory_balance_rmse": 53.69040657990535} {"canonical_length": 182, "canonical_nll": 1.1186848734761332, "generation_rouge_l_f1": 0.13524590163934427, "generation_token_f1": 0.25, "prompt_id": "21b0719467934ebb", "reading_canonical_log_prob": -203.60064697265625, "reading_distances": [46, 42, 48, 46], "reading_log_probs": [-199.99386596679688, -195.3975372314453, -194.21913146972656, -188.69699096679688], "reading_logprob_std": 4.644649047010441, "tokenization_path_logprob_std": 177.58184667423726, "trajectory_balance_rmse": 2863.6293611930423} {"canonical_length": 246, "canonical_nll": 1.3444913538490855, "generation_rouge_l_f1": 0.22134387351778656, "generation_token_f1": 0.4822134387351779, "prompt_id": "5d16c846337cb10b", "reading_canonical_log_prob": -330.744873046875, "reading_distances": [17, 19, 21, 15], "reading_log_probs": [-330.2099609375, -329.8199157714844, -328.9136962890625, -330.8695983886719], "reading_logprob_std": 0.8173218056408591, "tokenization_path_logprob_std": 142.76529798943056, "trajectory_balance_rmse": 5010.655066705819} {"canonical_length": 116, "canonical_nll": 0.5391659572206694, "generation_rouge_l_f1": 0.3722397476340694, "generation_token_f1": 0.47318611987381703, "prompt_id": "d66394bc1b63c0cb", "reading_canonical_log_prob": -62.543251037597656, "reading_distances": [22, 14, 17, 19], "reading_log_probs": [-62.9832649230957, -63.99497985839844, -63.65300369262695, -63.664825439453125], "reading_logprob_std": 0.4245322189567545, "tokenization_path_logprob_std": 57.9087891884343, "trajectory_balance_rmse": 1098.8666255177993} {"canonical_length": 75, "canonical_nll": 0.9754289754231771, "generation_rouge_l_f1": 0.2912621359223301, "generation_token_f1": 0.4077669902912622, "prompt_id": "4b9d4e7ec5d5b8b5", "reading_canonical_log_prob": -73.15717315673828, "reading_distances": [39, 32, 36, 40], "reading_log_probs": [-68.80231475830078, -69.76715087890625, -69.81651306152344, -68.79991912841797], "reading_logprob_std": 0.5723453055906216, "tokenization_path_logprob_std": 139.42823315440094, "trajectory_balance_rmse": 1445.830423156337} {"canonical_length": 363, "canonical_nll": 1.2543361022781079, "generation_rouge_l_f1": 0.2677165354330709, "generation_token_f1": 0.43779527559055115, "prompt_id": "cddcde885956c90e", "reading_canonical_log_prob": -455.3240051269531, "reading_distances": [21, 21, 25, 24], "reading_log_probs": [-452.9868469238281, -453.2498474121094, -453.802978515625, -453.4234313964844], "reading_logprob_std": 0.3422946424117525, "tokenization_path_logprob_std": 123.56841735872248, "trajectory_balance_rmse": 7285.344813345987} {"canonical_length": 204, "canonical_nll": 1.6374894684436274, "generation_rouge_l_f1": 0.2173913043478261, "generation_token_f1": 0.36086956521739133, "prompt_id": "e708ffc93eb420d5", "reading_canonical_log_prob": -334.0478515625, "reading_distances": [40, 42, 33, 39], "reading_log_probs": [-327.34539794921875, -327.5752868652344, -327.0009765625, -328.8227844238281], "reading_logprob_std": 0.7936840348870994, "tokenization_path_logprob_std": 64.87358683956175, "trajectory_balance_rmse": 3961.1463383294667} {"canonical_length": 73, "canonical_nll": 1.8636819499812713, "generation_rouge_l_f1": 0.17073170731707318, "generation_token_f1": 0.2378048780487805, "prompt_id": "276028c9e0375416", "reading_canonical_log_prob": -136.0487823486328, "reading_distances": [30, 25, 29, 25], "reading_log_probs": [-133.09693908691406, -135.46884155273438, -134.21470642089844, -134.04490661621094], "reading_logprob_std": 0.9748181523972091, "tokenization_path_logprob_std": 115.7242462435486, "trajectory_balance_rmse": 1532.6090012254815} {"canonical_length": 100, "canonical_nll": 1.0620518493652344, "generation_rouge_l_f1": 0.25671641791044775, "generation_token_f1": 0.3701492537313433, "prompt_id": "328ee04612e4bcf9", "reading_canonical_log_prob": -106.20518493652344, "reading_distances": [28, 27, 25, 29], "reading_log_probs": [-106.90956115722656, -106.71490478515625, -105.41984558105469, -105.29370880126953], "reading_logprob_std": 0.8456264586726541, "tokenization_path_logprob_std": 101.80455441342167, "trajectory_balance_rmse": 1635.5612896779817} {"canonical_length": 90, "canonical_nll": 1.1373841179741753, "generation_rouge_l_f1": 0.20307692307692307, "generation_token_f1": 0.2953846153846154, "prompt_id": "bdf95ab3a5ffa3ff", "reading_canonical_log_prob": -102.36457061767578, "reading_distances": [25, 34, 30, 30], "reading_log_probs": [-91.12435913085938, -94.3551025390625, -90.25965881347656, -91.52288818359375], "reading_logprob_std": 1.7732757099196954, "tokenization_path_logprob_std": 116.66790672320863, "trajectory_balance_rmse": 1139.8632981675862} {"canonical_length": 132, "canonical_nll": 1.703051133589311, "generation_rouge_l_f1": 0.2198391420911528, "generation_token_f1": 0.3914209115281501, "prompt_id": "80019145cee29a1a", "reading_canonical_log_prob": -224.80274963378906, "reading_distances": [18, 18, 19, 24], "reading_log_probs": [-223.5462188720703, -219.09173583984375, -222.46743774414062, -218.59536743164062], "reading_logprob_std": 2.452070404777967, "tokenization_path_logprob_std": 104.20489454983425, "trajectory_balance_rmse": 2446.4330368175497} {"canonical_length": 250, "canonical_nll": 1.394816162109375, "generation_rouge_l_f1": 0.22709163346613545, "generation_token_f1": 0.5179282868525896, "prompt_id": "77663077eb493ddd", "reading_canonical_log_prob": -348.70404052734375, "reading_distances": [31, 30, 29, 30], "reading_log_probs": [-344.6419677734375, -345.7681884765625, -345.7678527832031, -345.8920593261719], "reading_logprob_std": 0.5866210008010126, "tokenization_path_logprob_std": 150.9097233863174, "trajectory_balance_rmse": 4782.282076131043} {"canonical_length": 369, "canonical_nll": 1.2200462940908705, "generation_rouge_l_f1": 0.3094098883572568, "generation_token_f1": 0.5390749601275917, "prompt_id": "4825c22b71594c0c", "reading_canonical_log_prob": -450.19708251953125, "reading_distances": [32, 32, 36, 32], "reading_log_probs": [-453.419189453125, -456.0968933105469, -452.6853332519531, -450.2100830078125], "reading_logprob_std": 2.422587260112052, "tokenization_path_logprob_std": 158.24034972393764, "trajectory_balance_rmse": 7032.096097980576} {"canonical_length": 198, "canonical_nll": 1.133654970111269, "generation_rouge_l_f1": 0.31549295774647884, "generation_token_f1": 0.5746478873239436, "prompt_id": "c520430df28e5bdb", "reading_canonical_log_prob": -224.46368408203125, "reading_distances": [17, 21, 22, 21], "reading_log_probs": [-220.9581298828125, -223.10797119140625, -223.75677490234375, -222.10452270507812], "reading_logprob_std": 1.2222329477856932, "tokenization_path_logprob_std": 150.60511459495615, "trajectory_balance_rmse": 3804.809557441675} {"canonical_length": 290, "canonical_nll": 1.369360246329472, "generation_rouge_l_f1": 0.21739130434782608, "generation_token_f1": 0.4746376811594203, "prompt_id": "c3e62d8cbe21b9a4", "reading_canonical_log_prob": -397.1144714355469, "reading_distances": [25, 23, 28, 26], "reading_log_probs": [-395.7477722167969, -398.3518371582031, -396.42431640625, -395.884765625], "reading_logprob_std": 1.2024463500829492, "tokenization_path_logprob_std": 139.55365567093804, "trajectory_balance_rmse": 5904.404060760956} {"canonical_length": 38, "canonical_nll": 0.6760834643715307, "generation_rouge_l_f1": 0.20422535211267606, "generation_token_f1": 0.2464788732394366, "prompt_id": "ef7fdf70b61d5a5a", "reading_canonical_log_prob": -25.691171646118164, "reading_distances": [34, 27, 27, 31], "reading_log_probs": [-22.862239837646484, -24.98856544494629, -25.697816848754883, -24.05699348449707], "reading_logprob_std": 1.2265248684407124, "tokenization_path_logprob_std": 98.4820269316607, "trajectory_balance_rmse": 611.3224900240218} {"canonical_length": 262, "canonical_nll": 0.9594067872025585, "generation_rouge_l_f1": 0.31211498973305957, "generation_token_f1": 0.5954825462012321, "prompt_id": "f0bfbb195bb03bbd", "reading_canonical_log_prob": -251.3645782470703, "reading_distances": [38, 26, 27, 32], "reading_log_probs": [-254.38671875, -252.12698364257812, -252.43817138671875, -252.59576416015625], "reading_logprob_std": 1.0186668736454658, "tokenization_path_logprob_std": 92.22595545067948, "trajectory_balance_rmse": 5450.512450208146} {"canonical_length": 12, "canonical_nll": 0.6288140614827474, "generation_rouge_l_f1": 0.7000000000000001, "generation_token_f1": 0.7000000000000001, "prompt_id": "064f3342c7039feb", "reading_canonical_log_prob": -7.545768737792969, "reading_distances": [32, 31, 33, 30], "reading_log_probs": [-7.194114685058594, -7.77641487121582, -7.781383991241455, -8.280250549316406], "reading_logprob_std": 0.4440713893299102, "tokenization_path_logprob_std": 49.066697143264555, "trajectory_balance_rmse": 271.7825552481304} {"canonical_length": 174, "canonical_nll": 1.493584819223689, "generation_rouge_l_f1": 0.22857142857142856, "generation_token_f1": 0.41428571428571426, "prompt_id": "5319439232a42bef", "reading_canonical_log_prob": -259.8837585449219, "reading_distances": [27, 27, 25, 31], "reading_log_probs": [-252.41612243652344, -250.1717529296875, -252.04653930664062, -250.64865112304688], "reading_logprob_std": 1.0798953744437862, "tokenization_path_logprob_std": 171.59086639549744, "trajectory_balance_rmse": 3888.7545480633244} {"canonical_length": 274, "canonical_nll": 1.116197349381273, "generation_rouge_l_f1": 0.19852941176470587, "generation_token_f1": 0.4595588235294117, "prompt_id": "7f189f92ecaccbe7", "reading_canonical_log_prob": -305.83807373046875, "reading_distances": [21, 13, 20, 19], "reading_log_probs": [-308.27484130859375, -307.6048278808594, -308.7074890136719, -308.8925476074219], "reading_logprob_std": 0.5719843713468015, "tokenization_path_logprob_std": 87.83984936339122, "trajectory_balance_rmse": 5262.102782322503} {"canonical_length": 225, "canonical_nll": 1.0507347276475694, "generation_rouge_l_f1": 0.2789699570815451, "generation_token_f1": 0.47210300429184554, "prompt_id": "89ab6e8e7235950e", "reading_canonical_log_prob": -236.41531372070312, "reading_distances": [30, 18, 26, 22], "reading_log_probs": [-235.46737670898438, -235.30935668945312, -233.75350952148438, -235.20445251464844], "reading_logprob_std": 0.7941631371192329, "tokenization_path_logprob_std": 193.21589979693414, "trajectory_balance_rmse": 4527.026422050638} {"canonical_length": 188, "canonical_nll": 1.411251960916722, "generation_rouge_l_f1": 0.2706422018348624, "generation_token_f1": 0.5321100917431192, "prompt_id": "3d38b4ece7a08884", "reading_canonical_log_prob": -265.31536865234375, "reading_distances": [26, 21, 21, 29], "reading_log_probs": [-270.9110107421875, -267.80108642578125, -269.45941162109375, -269.794677734375], "reading_logprob_std": 1.286527003612511, "tokenization_path_logprob_std": 71.8371468459324, "trajectory_balance_rmse": 3772.426109669692} {"canonical_length": 169, "canonical_nll": 1.815282697508321, "generation_rouge_l_f1": 0.18309859154929578, "generation_token_f1": 0.323943661971831, "prompt_id": "49135037e7b65f9f", "reading_canonical_log_prob": -306.78277587890625, "reading_distances": [10, 15, 17, 16], "reading_log_probs": [-305.89483642578125, -312.31622314453125, -314.4647521972656, -307.5101318359375], "reading_logprob_std": 4.014227593939371, "tokenization_path_logprob_std": 190.45438745478583, "trajectory_balance_rmse": 4078.2289343708076} {"canonical_length": 250, "canonical_nll": 1.500998291015625, "generation_rouge_l_f1": 0.2684824902723735, "generation_token_f1": 0.5019455252918288, "prompt_id": "daaf9a26e9fd85ac", "reading_canonical_log_prob": -375.24957275390625, "reading_distances": [37, 30, 33, 34], "reading_log_probs": [-368.39300537109375, -369.03985595703125, -369.4910583496094, -367.0028076171875], "reading_logprob_std": 1.0840251789990514, "tokenization_path_logprob_std": 40.6428135308649, "trajectory_balance_rmse": 5987.960903024249} {"canonical_length": 32, "canonical_nll": 1.4940141439437866, "generation_rouge_l_f1": 0.18478260869565216, "generation_token_f1": 0.2282608695652174, "prompt_id": "06a9af54aedbd3a8", "reading_canonical_log_prob": -47.80845260620117, "reading_distances": [43, 44, 45, 39], "reading_log_probs": [-48.86762619018555, -50.432865142822266, -49.315277099609375, -47.11956024169922], "reading_logprob_std": 1.3770216591756388, "tokenization_path_logprob_std": 65.80440550982608, "trajectory_balance_rmse": 696.1926945114094} {"canonical_length": 48, "canonical_nll": 0.5817713340123495, "generation_rouge_l_f1": 0.20224719101123595, "generation_token_f1": 0.23220973782771534, "prompt_id": "4148986bb9e9364b", "reading_canonical_log_prob": -27.925024032592773, "reading_distances": [11, 9, 6, 10], "reading_log_probs": [-28.036888122558594, -27.60076904296875, -25.42232894897461, -27.650310516357422], "reading_logprob_std": 1.1862939220767543, "tokenization_path_logprob_std": 87.11448683702633, "trajectory_balance_rmse": 909.761535855942} {"canonical_length": 19, "canonical_nll": 0.8492911489386308, "generation_rouge_l_f1": 0.12734082397003746, "generation_token_f1": 0.14981273408239698, "prompt_id": "1c5e2235e187365f", "reading_canonical_log_prob": -16.136531829833984, "reading_distances": [23, 24, 19, 15], "reading_log_probs": [-18.16604232788086, -17.3823184967041, -17.841503143310547, -17.874601364135742], "reading_logprob_std": 0.3238794805031931, "tokenization_path_logprob_std": 61.41750845069316, "trajectory_balance_rmse": 309.05873736717047} {"canonical_length": 369, "canonical_nll": 1.1202930150639396, "generation_rouge_l_f1": 0.2564935064935065, "generation_token_f1": 0.5357142857142857, "prompt_id": "6125f7b585f4ca3f", "reading_canonical_log_prob": -413.38812255859375, "reading_distances": [32, 19, 29, 26], "reading_log_probs": [-411.4645690917969, -414.196044921875, -415.85491943359375, -414.85601806640625], "reading_logprob_std": 1.880232601300026, "tokenization_path_logprob_std": 376.71046248928906, "trajectory_balance_rmse": 7192.816356749697} {"canonical_length": 21, "canonical_nll": 0.9085188366117931, "generation_rouge_l_f1": 0.2727272727272727, "generation_token_f1": 0.2727272727272727, "prompt_id": "bac6bf1b8815878e", "reading_canonical_log_prob": -19.078895568847656, "reading_distances": [19, 20, 17, 16], "reading_log_probs": [-17.829086303710938, -18.222402572631836, -18.721454620361328, -17.60596466064453], "reading_logprob_std": 0.4894017320047687, "tokenization_path_logprob_std": 70.84396186711156, "trajectory_balance_rmse": 404.0639246057426} {"canonical_length": 81, "canonical_nll": 2.269473605685764, "generation_rouge_l_f1": 0.2090909090909091, "generation_token_f1": 0.24545454545454545, "prompt_id": "3e2b40881e16b7da", "reading_canonical_log_prob": -183.82736206054688, "reading_distances": [22, 14, 19, 18], "reading_log_probs": [-185.19949340820312, -185.82882690429688, -184.7104034423828, -184.352294921875], "reading_logprob_std": 0.6398095007463267, "tokenization_path_logprob_std": 73.90523210561145, "trajectory_balance_rmse": 1476.8747009396552} {"canonical_length": 32, "canonical_nll": 1.1381046772003174, "generation_rouge_l_f1": 0.14983713355048858, "generation_token_f1": 0.20846905537459284, "prompt_id": "e01d08fb030556a3", "reading_canonical_log_prob": -36.419349670410156, "reading_distances": [32, 27, 24, 30], "reading_log_probs": [-37.129539489746094, -37.769073486328125, -35.75832748413086, -37.24897003173828], "reading_logprob_std": 0.8582497715823663, "tokenization_path_logprob_std": 134.47831997273397, "trajectory_balance_rmse": 723.9655741154164} {"canonical_length": 64, "canonical_nll": 1.4768487215042114, "generation_rouge_l_f1": 0.1212121212121212, "generation_token_f1": 0.1212121212121212, "prompt_id": "68a9acf034dc69e5", "reading_canonical_log_prob": -94.51831817626953, "reading_distances": [28, 30, 23, 21], "reading_log_probs": [-91.54637908935547, -92.61825561523438, -91.93130493164062, -91.24620056152344], "reading_logprob_std": 0.5923821014498644, "tokenization_path_logprob_std": 76.18791731979952, "trajectory_balance_rmse": 1412.5908834238924} {"canonical_length": 196, "canonical_nll": 1.0047437706772162, "generation_rouge_l_f1": 0.1870824053452116, "generation_token_f1": 0.34743875278396436, "prompt_id": "12fd41da356aa8b2", "reading_canonical_log_prob": -196.92977905273438, "reading_distances": [32, 31, 33, 31], "reading_log_probs": [-196.65396118164062, -199.25616455078125, -191.41041564941406, -201.17138671875], "reading_logprob_std": 4.234518603086467, "tokenization_path_logprob_std": 140.72501784788, "trajectory_balance_rmse": 3545.030569324572} {"canonical_length": 404, "canonical_nll": 0.9240369891176129, "generation_rouge_l_f1": 0.2787136294027565, "generation_token_f1": 0.48392036753445633, "prompt_id": "7a350261a78f28df", "reading_canonical_log_prob": -373.3109436035156, "reading_distances": [20, 20, 20, 25], "reading_log_probs": [-368.04949951171875, -368.9938659667969, -368.695068359375, -371.3409118652344], "reading_logprob_std": 1.4358613772601176, "tokenization_path_logprob_std": 238.96661400354802, "trajectory_balance_rmse": 6299.778290431446} {"canonical_length": 316, "canonical_nll": 1.3110663498504251, "generation_rouge_l_f1": 0.210896309314587, "generation_token_f1": 0.421792618629174, "prompt_id": "732a9ef0a365d23a", "reading_canonical_log_prob": -414.2969665527344, "reading_distances": [20, 24, 26, 24], "reading_log_probs": [-411.0557861328125, -409.948486328125, -408.8059997558594, -410.80279541015625], "reading_logprob_std": 1.0154624712252294, "tokenization_path_logprob_std": 136.89687913508638, "trajectory_balance_rmse": 7486.2063050270135} {"canonical_length": 244, "canonical_nll": 1.488156928390753, "generation_rouge_l_f1": 0.23799582463465555, "generation_token_f1": 0.488517745302714, "prompt_id": "26e0cf383cca3d0a", "reading_canonical_log_prob": -363.11029052734375, "reading_distances": [27, 33, 27, 30], "reading_log_probs": [-360.163330078125, -360.62701416015625, -360.84295654296875, -360.8396911621094], "reading_logprob_std": 0.3196655359043732, "tokenization_path_logprob_std": 62.11125091007016, "trajectory_balance_rmse": 5394.227722325153} {"canonical_length": 123, "canonical_nll": 1.778093291491997, "generation_rouge_l_f1": 0.2202898550724638, "generation_token_f1": 0.4057971014492754, "prompt_id": "92f09a8894ee3f5f", "reading_canonical_log_prob": -218.70547485351562, "reading_distances": [14, 20, 14, 18], "reading_log_probs": [-227.54739379882812, -228.1238250732422, -230.70094299316406, -227.8970947265625], "reading_logprob_std": 1.4420455609853395, "tokenization_path_logprob_std": 74.98446692717256, "trajectory_balance_rmse": 2708.025828275795} {"canonical_length": 114, "canonical_nll": 1.2883128116005345, "generation_rouge_l_f1": 0.28254847645429365, "generation_token_f1": 0.40443213296398894, "prompt_id": "eb8e055ee1d3ca60", "reading_canonical_log_prob": -146.86766052246094, "reading_distances": [21, 21, 17, 21], "reading_log_probs": [-149.04505920410156, -148.18658447265625, -145.99737548828125, -148.35671997070312], "reading_logprob_std": 1.319319718899996, "tokenization_path_logprob_std": 181.02273751093153, "trajectory_balance_rmse": 2418.0872230929667} {"canonical_length": 221, "canonical_nll": 1.6434138371394231, "generation_rouge_l_f1": 0.19230769230769232, "generation_token_f1": 0.4423076923076923, "prompt_id": "5a05098dc8854389", "reading_canonical_log_prob": -363.1944580078125, "reading_distances": [22, 20, 19, 25], "reading_log_probs": [-366.85882568359375, -360.61041259765625, -354.4818420410156, -362.86981201171875], "reading_logprob_std": 5.173384798377275, "tokenization_path_logprob_std": 118.04300992248322, "trajectory_balance_rmse": 4399.679183665011} {"canonical_length": 6, "canonical_nll": 0.7351329326629639, "generation_rouge_l_f1": 0.7272727272727272, "generation_token_f1": 0.7272727272727272, "prompt_id": "0744be99f0e84455", "reading_canonical_log_prob": -4.410797595977783, "reading_distances": [20, 24, 17, 15], "reading_log_probs": [-4.2660088539123535, -3.5947587490081787, -4.239288806915283, -4.663015365600586], "reading_logprob_std": 0.44206336513584527, "tokenization_path_logprob_std": 45.19339946681578, "trajectory_balance_rmse": 149.10314586478594} {"canonical_length": 3, "canonical_nll": 2.412130037943522, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.6666666666666666, "prompt_id": "1e2019bd6514f4f1", "reading_canonical_log_prob": -7.236390113830566, "reading_distances": [45, 55, 50, 54], "reading_log_probs": [-7.436141490936279, -7.675426006317139, -6.9944047927856445, -7.5102434158325195], "reading_logprob_std": 0.29083846484108383, "tokenization_path_logprob_std": 14.780708553802445, "trajectory_balance_rmse": 38.05851204108853} {"canonical_length": 17, "canonical_nll": 0.8434412900139304, "generation_rouge_l_f1": 0.5, "generation_token_f1": 0.5, "prompt_id": "982c9409d78c0796", "reading_canonical_log_prob": -14.338501930236816, "reading_distances": [16, 16, 19, 17], "reading_log_probs": [-22.68870735168457, -21.380691528320312, -18.08443832397461, -18.475378036499023], "reading_logprob_std": 2.2383272795335585, "tokenization_path_logprob_std": 62.563069154157105, "trajectory_balance_rmse": 418.7902519151937} {"canonical_length": 185, "canonical_nll": 1.438520606788429, "generation_rouge_l_f1": 0.2694063926940639, "generation_token_f1": 0.4885844748858448, "prompt_id": "f2e5fb29aaf06273", "reading_canonical_log_prob": -266.1263122558594, "reading_distances": [35, 39, 36, 37], "reading_log_probs": [-261.05902099609375, -263.56890869140625, -268.6222839355469, -265.46563720703125], "reading_logprob_std": 3.188773901695474, "tokenization_path_logprob_std": 206.06550848116203, "trajectory_balance_rmse": 3618.4632203496335} {"canonical_length": 309, "canonical_nll": 1.0309459106050263, "generation_rouge_l_f1": 0.23172905525846701, "generation_token_f1": 0.4848484848484849, "prompt_id": "3aa4d1f8b5499bdf", "reading_canonical_log_prob": -318.5622863769531, "reading_distances": [65, 56, 54, 65], "reading_log_probs": [-331.391845703125, -330.1853332519531, -326.14764404296875, -334.62762451171875], "reading_logprob_std": 3.5044560013013326, "tokenization_path_logprob_std": 263.242780579903, "trajectory_balance_rmse": 5021.727978920758} {"canonical_length": 253, "canonical_nll": 1.5471567749505928, "generation_rouge_l_f1": 0.22700587084148727, "generation_token_f1": 0.49706457925636005, "prompt_id": "aa727754a09ad816", "reading_canonical_log_prob": -391.4306640625, "reading_distances": [24, 22, 21, 24], "reading_log_probs": [-387.94268798828125, -390.1656799316406, -391.6126708984375, -396.46435546875], "reading_logprob_std": 3.6094231295589148, "tokenization_path_logprob_std": 72.84074573753144, "trajectory_balance_rmse": 5128.807231164816} {"canonical_length": 218, "canonical_nll": 1.3216808914044582, "generation_rouge_l_f1": 0.2127659574468085, "generation_token_f1": 0.3343465045592705, "prompt_id": "4d9af47029652710", "reading_canonical_log_prob": -288.1264343261719, "reading_distances": [22, 24, 22, 19], "reading_log_probs": [-293.427978515625, -286.98529052734375, -289.1731262207031, -291.2190856933594], "reading_logprob_std": 2.7596618264875086, "tokenization_path_logprob_std": 175.51882025451914, "trajectory_balance_rmse": 4675.003992265314} {"canonical_length": 286, "canonical_nll": 1.1684670615029502, "generation_rouge_l_f1": 0.2034548944337812, "generation_token_f1": 0.45681381957773515, "prompt_id": "4d68d14019e3f5eb", "reading_canonical_log_prob": -334.18157958984375, "reading_distances": [25, 24, 24, 22], "reading_log_probs": [-321.96868896484375, -324.6272277832031, -320.99658203125, -320.7041015625], "reading_logprob_std": 1.7858250796568118, "tokenization_path_logprob_std": 381.15208288586666, "trajectory_balance_rmse": 5033.898993985633} {"canonical_length": 104, "canonical_nll": 1.268396230844351, "generation_rouge_l_f1": 0.2151394422310757, "generation_token_f1": 0.39043824701195223, "prompt_id": "9752f3aa2c81e0d1", "reading_canonical_log_prob": -131.9132080078125, "reading_distances": [23, 32, 17, 28], "reading_log_probs": [-129.34698486328125, -130.43478393554688, -131.46405029296875, -131.62503051757812], "reading_logprob_std": 1.0550179955626984, "tokenization_path_logprob_std": 122.14519487439593, "trajectory_balance_rmse": 2151.889474348112} {"canonical_length": 129, "canonical_nll": 0.3381823458412821, "generation_rouge_l_f1": 0.5416666666666666, "generation_token_f1": 0.5468750000000001, "prompt_id": "46cccd71028358c4", "reading_canonical_log_prob": -43.62552261352539, "reading_distances": [20, 25, 24, 25], "reading_log_probs": [-44.08087921142578, -43.434364318847656, -43.57740020751953, -45.051551818847656], "reading_logprob_std": 0.731582506926174, "tokenization_path_logprob_std": 72.2164936224894, "trajectory_balance_rmse": 1254.0736854286454} {"canonical_length": 3, "canonical_nll": 1.061321496963501, "generation_rouge_l_f1": 0.028985507246376812, "generation_token_f1": 0.028985507246376812, "prompt_id": "f385579e0d438649", "reading_canonical_log_prob": -3.183964490890503, "reading_distances": [15, 11, 14, 12], "reading_log_probs": [-2.679447650909424, -2.8020424842834473, -2.8784472942352295, -2.5976696014404297], "reading_logprob_std": 0.12508666448548797, "tokenization_path_logprob_std": 16.58190378119232, "trajectory_balance_rmse": 35.997448534902496} {"canonical_length": 3, "canonical_nll": 2.493165651957194, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.6666666666666666, "prompt_id": "f015c3c34e66dddf", "reading_canonical_log_prob": -7.479496955871582, "reading_distances": [26, 19, 21, 22], "reading_log_probs": [-7.295690059661865, -7.347428321838379, -7.118887901306152, -6.9061431884765625], "reading_logprob_std": 0.19956331940069605, "tokenization_path_logprob_std": 17.771355174466336, "trajectory_balance_rmse": 32.0567695140766} {"canonical_length": 47, "canonical_nll": 0.7341905959109043, "generation_rouge_l_f1": 0.3356643356643357, "generation_token_f1": 0.39160839160839156, "prompt_id": "f91ba1a01be59af5", "reading_canonical_log_prob": -34.5069580078125, "reading_distances": [25, 28, 29, 28], "reading_log_probs": [-33.77035903930664, -35.067134857177734, -33.34775161743164, -34.48340606689453], "reading_logprob_std": 0.76132503312214, "tokenization_path_logprob_std": 107.77255642652223, "trajectory_balance_rmse": 746.2349545951141} {"canonical_length": 179, "canonical_nll": 1.4475087426894204, "generation_rouge_l_f1": 0.2334096109839817, "generation_token_f1": 0.47139588100686497, "prompt_id": "0764d55b71d1b589", "reading_canonical_log_prob": -259.10406494140625, "reading_distances": [14, 15, 18, 17], "reading_log_probs": [-258.76654052734375, -253.943115234375, -252.8415069580078, -255.79421997070312], "reading_logprob_std": 2.5910870871893743, "tokenization_path_logprob_std": 237.50436711473802, "trajectory_balance_rmse": 3684.3108595359763} {"canonical_length": 75, "canonical_nll": 1.5182753499348958, "generation_rouge_l_f1": 0.11730205278592375, "generation_token_f1": 0.24046920821114368, "prompt_id": "a40186552cfde2fc", "reading_canonical_log_prob": -113.87065124511719, "reading_distances": [12, 12, 6, 5], "reading_log_probs": [-121.78653717041016, -118.28317260742188, -114.11000061035156, -113.86065673828125], "reading_logprob_std": 3.775563323752648, "tokenization_path_logprob_std": 70.78409168610602, "trajectory_balance_rmse": 1421.3928208956224} {"canonical_length": 263, "canonical_nll": 1.1101563892437025, "generation_rouge_l_f1": 0.25515947467166983, "generation_token_f1": 0.5590994371482175, "prompt_id": "a3ec61d54b8539e4", "reading_canonical_log_prob": -291.97113037109375, "reading_distances": [22, 25, 26, 23], "reading_log_probs": [-291.85546875, -293.9755859375, -292.76129150390625, -292.7001037597656], "reading_logprob_std": 0.8724453061728664, "tokenization_path_logprob_std": 166.06877953842536, "trajectory_balance_rmse": 5579.538216947247} {"canonical_length": 210, "canonical_nll": 1.5841859363374255, "generation_rouge_l_f1": 0.20985010706638116, "generation_token_f1": 0.44967880085653106, "prompt_id": "ca23abf600fc132e", "reading_canonical_log_prob": -332.6790466308594, "reading_distances": [43, 45, 50, 45], "reading_log_probs": [-334.9521789550781, -332.7335205078125, -334.74468994140625, -334.927734375], "reading_logprob_std": 1.074669592020505, "tokenization_path_logprob_std": 58.56095949936928, "trajectory_balance_rmse": 4726.625211804462} {"canonical_length": 132, "canonical_nll": 1.5175452954841382, "generation_rouge_l_f1": 0.24468085106382978, "generation_token_f1": 0.43617021276595747, "prompt_id": "a83cfaa0169d4436", "reading_canonical_log_prob": -200.31597900390625, "reading_distances": [40, 56, 39, 47], "reading_log_probs": [-206.71939086914062, -205.62533569335938, -207.6134033203125, -208.1279754638672], "reading_logprob_std": 1.0977525811172104, "tokenization_path_logprob_std": 78.49852015087097, "trajectory_balance_rmse": 975.2863617738984} {"canonical_length": 115, "canonical_nll": 1.6405679453974185, "generation_rouge_l_f1": 0.2090395480225989, "generation_token_f1": 0.3163841807909604, "prompt_id": "bd9008593d2eaa20", "reading_canonical_log_prob": -188.66531372070312, "reading_distances": [41, 42, 41, 41], "reading_log_probs": [-186.75604248046875, -188.88668823242188, -187.19302368164062, -189.28204345703125], "reading_logprob_std": 1.241641850208166, "tokenization_path_logprob_std": 24.144673139123032, "trajectory_balance_rmse": 2713.9361871455394} {"canonical_length": 7, "canonical_nll": 0.8833706719534737, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "48bde41f90ff68c1", "reading_canonical_log_prob": -6.183594703674316, "reading_distances": [31, 34, 27, 35], "reading_log_probs": [-6.016020774841309, -6.059188365936279, -5.57352352142334, -6.188913345336914], "reading_logprob_std": 0.2675448012309682, "tokenization_path_logprob_std": 39.913963408304944, "trajectory_balance_rmse": 154.9730230790537} {"canonical_length": 130, "canonical_nll": 0.6900716928335336, "generation_rouge_l_f1": 0.3089430894308943, "generation_token_f1": 0.5203252032520326, "prompt_id": "8dbb7f1da28c2195", "reading_canonical_log_prob": -89.70932006835938, "reading_distances": [11, 9, 18, 15], "reading_log_probs": [-87.87445068359375, -87.80990600585938, -87.6129150390625, -88.8826904296875], "reading_logprob_std": 0.5694380049686475, "tokenization_path_logprob_std": 148.8966359192197, "trajectory_balance_rmse": 1703.9763554877698} {"canonical_length": 260, "canonical_nll": 2.119525850736178, "generation_rouge_l_f1": 0.26335877862595425, "generation_token_f1": 0.5038167938931297, "prompt_id": "d859a6a8a580026f", "reading_canonical_log_prob": -551.0767211914062, "reading_distances": [38, 35, 33, 31], "reading_log_probs": [-557.69970703125, -562.3976440429688, -565.4285888671875, -562.4957275390625], "reading_logprob_std": 3.1964298024123092, "tokenization_path_logprob_std": 218.57136659142697, "trajectory_balance_rmse": 5714.490227718962} {"canonical_length": 242, "canonical_nll": 1.2937562800635976, "generation_rouge_l_f1": 0.22355289421157687, "generation_token_f1": 0.4351297405189621, "prompt_id": "b4d85be43e32fd7f", "reading_canonical_log_prob": -313.0890197753906, "reading_distances": [37, 43, 34, 38], "reading_log_probs": [-307.1357116699219, -310.6778564453125, -311.84100341796875, -314.33380126953125], "reading_logprob_std": 2.9920998548198505, "tokenization_path_logprob_std": 117.39951629415484, "trajectory_balance_rmse": 5382.74000357637} {"canonical_length": 19, "canonical_nll": 1.200868104633532, "generation_rouge_l_f1": 0.2916666666666667, "generation_token_f1": 0.4166666666666667, "prompt_id": "64c4e6c526576026", "reading_canonical_log_prob": -22.81649398803711, "reading_distances": [27, 28, 30, 29], "reading_log_probs": [-23.53719711303711, -22.275676727294922, -21.954795837402344, -21.559127807617188], "reading_logprob_std": 0.855430970690688, "tokenization_path_logprob_std": 15.175946106336443, "trajectory_balance_rmse": 440.78146349452544} {"canonical_length": 56, "canonical_nll": 1.4429969787597656, "generation_rouge_l_f1": 0.0718562874251497, "generation_token_f1": 0.1377245508982036, "prompt_id": "505a139a155b8b27", "reading_canonical_log_prob": -80.80783081054688, "reading_distances": [26, 29, 18, 32], "reading_log_probs": [-79.04776000976562, -80.2714614868164, -80.37358093261719, -80.34857177734375], "reading_logprob_std": 0.6431923970218346, "tokenization_path_logprob_std": 76.38214903296682, "trajectory_balance_rmse": 1155.4282668605454} {"canonical_length": 171, "canonical_nll": 1.135448790433114, "generation_rouge_l_f1": 0.30434782608695654, "generation_token_f1": 0.5314009661835749, "prompt_id": "ca5c2935eb3033fb", "reading_canonical_log_prob": -194.1617431640625, "reading_distances": [18, 24, 21, 20], "reading_log_probs": [-190.3785858154297, -191.03294372558594, -189.70672607421875, -189.8922119140625], "reading_logprob_std": 0.5923587510803177, "tokenization_path_logprob_std": 216.96582378586453, "trajectory_balance_rmse": 3498.7129030846504} {"canonical_length": 78, "canonical_nll": 0.837485093336839, "generation_rouge_l_f1": 0.3072289156626506, "generation_token_f1": 0.3855421686746988, "prompt_id": "f8aa35c9d5781c5a", "reading_canonical_log_prob": -65.32383728027344, "reading_distances": [22, 28, 23, 24], "reading_log_probs": [-70.19774627685547, -70.09751892089844, -68.79971313476562, -69.87580871582031], "reading_logprob_std": 0.6428856142159622, "tokenization_path_logprob_std": 105.03194540146481, "trajectory_balance_rmse": 1405.8066881528093} {"canonical_length": 339, "canonical_nll": 1.1968351381014934, "generation_rouge_l_f1": 0.24666666666666667, "generation_token_f1": 0.5200000000000001, "prompt_id": "de613d5bcfdf1d4f", "reading_canonical_log_prob": -405.72711181640625, "reading_distances": [30, 37, 37, 36], "reading_log_probs": [-397.87103271484375, -397.72747802734375, -396.6076965332031, -396.67913818359375], "reading_logprob_std": 0.6705265195969331, "tokenization_path_logprob_std": 414.139999593164, "trajectory_balance_rmse": 7304.223136930876} {"canonical_length": 19, "canonical_nll": 1.6745522147730778, "generation_rouge_l_f1": 0.24000000000000005, "generation_token_f1": 0.24000000000000005, "prompt_id": "cb6d19eef2166daa", "reading_canonical_log_prob": -31.816492080688477, "reading_distances": [23, 30, 27, 26], "reading_log_probs": [-31.494070053100586, -32.65349578857422, -33.29409408569336, -31.81100845336914], "reading_logprob_std": 0.816726116098819, "tokenization_path_logprob_std": 89.64966579227905, "trajectory_balance_rmse": 494.6146932503168} {"canonical_length": 389, "canonical_nll": 1.195074164775466, "generation_rouge_l_f1": 0.18241042345276876, "generation_token_f1": 0.35179153094462545, "prompt_id": "980fa61309c5ee0c", "reading_canonical_log_prob": -464.88385009765625, "reading_distances": [20, 23, 21, 26], "reading_log_probs": [-465.35809326171875, -466.019287109375, -465.0528869628906, -465.2781066894531], "reading_logprob_std": 0.41540418309552435, "tokenization_path_logprob_std": 220.03062235543996, "trajectory_balance_rmse": 7144.2406030199645} {"canonical_length": 14, "canonical_nll": 2.358839307512556, "generation_rouge_l_f1": 0.09523809523809525, "generation_token_f1": 0.09523809523809525, "prompt_id": "0f4edb0d2ff1676b", "reading_canonical_log_prob": -33.02375030517578, "reading_distances": [14, 13, 14, 17], "reading_log_probs": [-35.683963775634766, -36.36985397338867, -33.91979217529297, -35.78984069824219], "reading_logprob_std": 1.0579168372162224, "tokenization_path_logprob_std": 111.23688486225177, "trajectory_balance_rmse": 337.48250663141334} {"canonical_length": 80, "canonical_nll": 0.9438306808471679, "generation_rouge_l_f1": 0.3055555555555555, "generation_token_f1": 0.3944444444444445, "prompt_id": "7bc8145a69124b53", "reading_canonical_log_prob": -75.50645446777344, "reading_distances": [25, 28, 26, 23], "reading_log_probs": [-76.86080932617188, -76.60484313964844, -76.36248779296875, -77.2406234741211], "reading_logprob_std": 0.37551929026781705, "tokenization_path_logprob_std": 63.045426040868364, "trajectory_balance_rmse": 1304.3580556019078} {"canonical_length": 12, "canonical_nll": 0.6137715975443522, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "5167b2d9f21714f8", "reading_canonical_log_prob": -7.365259170532227, "reading_distances": [27, 20, 22, 22], "reading_log_probs": [-5.610215663909912, -5.701549053192139, -6.198378562927246, -6.1097493171691895], "reading_logprob_std": 0.29228031849792996, "tokenization_path_logprob_std": 20.6990311200133, "trajectory_balance_rmse": 227.21227746323584} {"canonical_length": 130, "canonical_nll": 1.0168621356670673, "generation_rouge_l_f1": 0.3152454780361757, "generation_token_f1": 0.5064599483204134, "prompt_id": "786583b4f8ed0931", "reading_canonical_log_prob": -132.19207763671875, "reading_distances": [23, 23, 21, 20], "reading_log_probs": [-131.02304077148438, -132.91641235351562, -129.52996826171875, -130.56179809570312], "reading_logprob_std": 1.4172670581463962, "tokenization_path_logprob_std": 159.54130526001876, "trajectory_balance_rmse": 3362.615043561208} {"canonical_length": 351, "canonical_nll": 1.1606759182747952, "generation_rouge_l_f1": 0.19269102990033224, "generation_token_f1": 0.3787375415282392, "prompt_id": "99a05025a306c903", "reading_canonical_log_prob": -407.3972473144531, "reading_distances": [18, 15, 13, 19], "reading_log_probs": [-405.09356689453125, -403.2464904785156, -404.28204345703125, -403.669189453125], "reading_logprob_std": 0.8023791236250832, "tokenization_path_logprob_std": 342.3073170668703, "trajectory_balance_rmse": 6025.0895129919045} {"canonical_length": 32, "canonical_nll": 1.2998607158660889, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "4d4e1a944b4c608a", "reading_canonical_log_prob": -41.595542907714844, "reading_distances": [12, 11, 14, 14], "reading_log_probs": [-42.83415985107422, -42.08043670654297, -42.809539794921875, -43.23811340332031], "reading_logprob_std": 0.48195480283423847, "tokenization_path_logprob_std": 17.21733632888828, "trajectory_balance_rmse": 620.2541880579646} {"canonical_length": 375, "canonical_nll": 1.2934869791666668, "generation_rouge_l_f1": 0.30906389301634474, "generation_token_f1": 0.5705794947994056, "prompt_id": "8de18ae965c9405f", "reading_canonical_log_prob": -485.0576171875, "reading_distances": [29, 25, 28, 29], "reading_log_probs": [-482.39178466796875, -482.43194580078125, -480.4967041015625, -480.2342529296875], "reading_logprob_std": 1.186443617115181, "tokenization_path_logprob_std": 376.23774874552436, "trajectory_balance_rmse": 6812.753612215037} {"canonical_length": 154, "canonical_nll": 1.0747160478071733, "generation_rouge_l_f1": 0.38725490196078427, "generation_token_f1": 0.5245098039215687, "prompt_id": "577845af12167b4c", "reading_canonical_log_prob": -165.5062713623047, "reading_distances": [9, 11, 10, 8], "reading_log_probs": [-167.14144897460938, -166.55177307128906, -165.55958557128906, -166.6749725341797], "reading_logprob_std": 0.6652889088581404, "tokenization_path_logprob_std": 41.6424151409755, "trajectory_balance_rmse": 2646.870453499229} {"canonical_length": 94, "canonical_nll": 1.1479234086706283, "generation_rouge_l_f1": 0.19628647214854114, "generation_token_f1": 0.35013262599469497, "prompt_id": "aea07a40db5e5411", "reading_canonical_log_prob": -107.90480041503906, "reading_distances": [18, 24, 30, 29], "reading_log_probs": [-109.87823486328125, -111.7093505859375, -109.60560607910156, -110.68292236328125], "reading_logprob_std": 0.9449311896837094, "tokenization_path_logprob_std": 118.22504864113363, "trajectory_balance_rmse": 1957.7827506646022} {"canonical_length": 111, "canonical_nll": 1.4852859909470018, "generation_rouge_l_f1": 0.15706806282722513, "generation_token_f1": 0.324607329842932, "prompt_id": "fdb4e9b8596975db", "reading_canonical_log_prob": -164.8667449951172, "reading_distances": [24, 23, 23, 20], "reading_log_probs": [-161.916748046875, -161.18923950195312, -162.46194458007812, -160.584228515625], "reading_logprob_std": 0.8222808913869799, "tokenization_path_logprob_std": 142.87656136109058, "trajectory_balance_rmse": 2552.215807809981} {"canonical_length": 364, "canonical_nll": 1.645067655123197, "generation_rouge_l_f1": 0.2360876897133221, "generation_token_f1": 0.5092748735244519, "prompt_id": "219a3ab27e90cb77", "reading_canonical_log_prob": -598.8046264648438, "reading_distances": [13, 15, 18, 13], "reading_log_probs": [-600.6026000976562, -600.9107055664062, -599.8384399414062, -597.6856689453125], "reading_logprob_std": 1.4540860477565174, "tokenization_path_logprob_std": 399.7423474065624, "trajectory_balance_rmse": 7457.662327274127} {"canonical_length": 42, "canonical_nll": 1.2402064005533855, "generation_rouge_l_f1": 0.09523809523809523, "generation_token_f1": 0.17582417582417584, "prompt_id": "c1bdd5a5ad1d792e", "reading_canonical_log_prob": -52.08866882324219, "reading_distances": [14, 21, 19, 18], "reading_log_probs": [-54.54993438720703, -56.045066833496094, -53.71782684326172, -55.76691436767578], "reading_logprob_std": 1.0840140957953492, "tokenization_path_logprob_std": 27.459653587740558, "trajectory_balance_rmse": 534.462579722049} {"canonical_length": 262, "canonical_nll": 1.7406362286050812, "generation_rouge_l_f1": 0.27819548872180455, "generation_token_f1": 0.5037593984962406, "prompt_id": "52d1f167e2d1bb05", "reading_canonical_log_prob": -456.04669189453125, "reading_distances": [30, 24, 24, 28], "reading_log_probs": [-449.399658203125, -450.23138427734375, -449.13482666015625, -450.51141357421875], "reading_logprob_std": 0.6566169964757929, "tokenization_path_logprob_std": 224.67789702782713, "trajectory_balance_rmse": 5482.419864339731} {"canonical_length": 17, "canonical_nll": 2.15593427770278, "generation_rouge_l_f1": 0.1111111111111111, "generation_token_f1": 0.1111111111111111, "prompt_id": "c86b63feb26be380", "reading_canonical_log_prob": -36.650882720947266, "reading_distances": [37, 37, 36, 36], "reading_log_probs": [-33.47163772583008, -31.683069229125977, -31.34952163696289, -32.04188537597656], "reading_logprob_std": 0.9338949046208962, "tokenization_path_logprob_std": 39.22048576983473, "trajectory_balance_rmse": 397.41442236054695} {"canonical_length": 10, "canonical_nll": 2.0558258056640626, "generation_rouge_l_f1": 0.2, "generation_token_f1": 0.3, "prompt_id": "197710f492b62fe4", "reading_canonical_log_prob": -20.558258056640625, "reading_distances": [14, 7, 9, 16], "reading_log_probs": [-18.942474365234375, -19.143301010131836, -19.163217544555664, -20.34369659423828], "reading_logprob_std": 0.6381849027383314, "tokenization_path_logprob_std": 40.79547444866202, "trajectory_balance_rmse": 221.01646538762174} {"canonical_length": 25, "canonical_nll": 1.227169189453125, "generation_rouge_l_f1": 0.18867924528301885, "generation_token_f1": 0.22641509433962265, "prompt_id": "23c24264db31cb91", "reading_canonical_log_prob": -30.679229736328125, "reading_distances": [31, 23, 27, 30], "reading_log_probs": [-29.800487518310547, -31.24103355407715, -31.354358673095703, -32.55235290527344], "reading_logprob_std": 1.1265745558459572, "tokenization_path_logprob_std": 90.87971060432298, "trajectory_balance_rmse": 444.10009504724513} {"canonical_length": 169, "canonical_nll": 0.8433055990546413, "generation_rouge_l_f1": 0.33115468409586063, "generation_token_f1": 0.5577342047930283, "prompt_id": "742a95dae13136e2", "reading_canonical_log_prob": -142.51864624023438, "reading_distances": [19, 18, 18, 16], "reading_log_probs": [-140.25473022460938, -143.9477081298828, -140.0282745361328, -143.34347534179688], "reading_logprob_std": 2.040166000732142, "tokenization_path_logprob_std": 172.76729131757654, "trajectory_balance_rmse": 2560.550356589912} {"canonical_length": 19, "canonical_nll": 2.1425969977127877, "generation_rouge_l_f1": 0.4615384615384615, "generation_token_f1": 0.4615384615384615, "prompt_id": "87a7422c0b830ad1", "reading_canonical_log_prob": -40.70934295654297, "reading_distances": [48, 51, 43, 43], "reading_log_probs": [-42.71561813354492, -42.01451873779297, -40.45277404785156, -40.38336944580078], "reading_logprob_std": 1.1603124273400842, "tokenization_path_logprob_std": 51.92365376530294, "trajectory_balance_rmse": 389.1749381700629} {"canonical_length": 44, "canonical_nll": 0.8328087546608665, "generation_rouge_l_f1": 0.5945945945945946, "generation_token_f1": 0.6216216216216217, "prompt_id": "73b66ff67ed60622", "reading_canonical_log_prob": -36.643585205078125, "reading_distances": [29, 26, 25, 22], "reading_log_probs": [-37.80821228027344, -37.95408630371094, -38.6278076171875, -37.91679763793945], "reading_logprob_std": 0.372562362824723, "tokenization_path_logprob_std": 66.38789375018591, "trajectory_balance_rmse": 1150.7841757790316} {"canonical_length": 126, "canonical_nll": 1.0559745667472717, "generation_rouge_l_f1": 0.36000000000000004, "generation_token_f1": 0.5800000000000001, "prompt_id": "f610e308eb24edb1", "reading_canonical_log_prob": -133.05279541015625, "reading_distances": [23, 21, 24, 25], "reading_log_probs": [-130.95741271972656, -132.58294677734375, -131.58409118652344, -130.5785369873047], "reading_logprob_std": 0.8758397893413683, "tokenization_path_logprob_std": 115.43390210129324, "trajectory_balance_rmse": 2009.523257411523} {"canonical_length": 251, "canonical_nll": 1.2289727807519921, "generation_rouge_l_f1": 0.21176470588235294, "generation_token_f1": 0.4509803921568627, "prompt_id": "da58850d3c7fdcfd", "reading_canonical_log_prob": -308.47216796875, "reading_distances": [18, 23, 25, 22], "reading_log_probs": [-303.6155090332031, -305.67919921875, -306.485595703125, -305.49798583984375], "reading_logprob_std": 1.2144415871462286, "tokenization_path_logprob_std": 81.90321018572567, "trajectory_balance_rmse": 5130.797427847756} {"canonical_length": 296, "canonical_nll": 1.2881446013579498, "generation_rouge_l_f1": 0.23529411764705882, "generation_token_f1": 0.47438330170777987, "prompt_id": "0a2b0e3667ca975a", "reading_canonical_log_prob": -381.2908020019531, "reading_distances": [12, 16, 15, 19], "reading_log_probs": [-380.93035888671875, -379.8670349121094, -378.62921142578125, -378.97930908203125], "reading_logprob_std": 1.0277550516349727, "tokenization_path_logprob_std": 238.19448552894207, "trajectory_balance_rmse": 5846.258446791843} {"canonical_length": 111, "canonical_nll": 1.1043646185247749, "generation_rouge_l_f1": 0.23728813559322037, "generation_token_f1": 0.37853107344632764, "prompt_id": "b3bdd02a596bc2d5", "reading_canonical_log_prob": -122.58447265625, "reading_distances": [39, 36, 44, 46], "reading_log_probs": [-126.14192199707031, -124.04051208496094, -125.17347717285156, -125.33033752441406], "reading_logprob_std": 0.8652717818242673, "tokenization_path_logprob_std": 84.53028200491971, "trajectory_balance_rmse": 1457.3085382339882} {"canonical_length": 14, "canonical_nll": 0.07202623571668353, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "3a6cc5e39c6939e7", "reading_canonical_log_prob": -1.0083673000335693, "reading_distances": [22, 27, 23, 27], "reading_log_probs": [-1.1475694179534912, -0.766061007976532, -1.1036145687103271, -0.9519121646881104], "reading_logprob_std": 0.17254447122691438, "tokenization_path_logprob_std": 57.483966714997884, "trajectory_balance_rmse": 284.72364506861777} {"canonical_length": 79, "canonical_nll": 0.7063537790805479, "generation_rouge_l_f1": 0.3179190751445087, "generation_token_f1": 0.42774566473988446, "prompt_id": "9431e7948140e38d", "reading_canonical_log_prob": -55.80194854736328, "reading_distances": [36, 34, 37, 37], "reading_log_probs": [-54.37127685546875, -53.869300842285156, -53.980018615722656, -54.031341552734375], "reading_logprob_std": 0.2163655469892277, "tokenization_path_logprob_std": 110.01349456767471, "trajectory_balance_rmse": 1306.1450459305242} {"canonical_length": 56, "canonical_nll": 0.9630532264709473, "generation_rouge_l_f1": 0.41142857142857137, "generation_token_f1": 0.4685714285714286, "prompt_id": "8230c7c1426f816d", "reading_canonical_log_prob": -53.93098068237305, "reading_distances": [24, 23, 25, 21], "reading_log_probs": [-54.28096008300781, -53.08531188964844, -53.136451721191406, -52.88922882080078], "reading_logprob_std": 0.6310423198626816, "tokenization_path_logprob_std": 93.21355472171598, "trajectory_balance_rmse": 1225.855225078026} {"canonical_length": 101, "canonical_nll": 1.1748796217512376, "generation_rouge_l_f1": 0.3115577889447236, "generation_token_f1": 0.6030150753768845, "prompt_id": "211cd27653ee878e", "reading_canonical_log_prob": -118.662841796875, "reading_distances": [27, 24, 31, 27], "reading_log_probs": [-120.73692321777344, -118.74771881103516, -120.2535629272461, -119.64654541015625], "reading_logprob_std": 0.8574876641123547, "tokenization_path_logprob_std": 79.64759236224596, "trajectory_balance_rmse": 1994.7744830671952} {"canonical_length": 19, "canonical_nll": 2.0848308362458883, "generation_rouge_l_f1": 0.3636363636363636, "generation_token_f1": 0.6363636363636365, "prompt_id": "bd8cff3238489dd7", "reading_canonical_log_prob": -39.611785888671875, "reading_distances": [31, 29, 37, 36], "reading_log_probs": [-42.60755157470703, -43.63191223144531, -42.65155792236328, -36.91162872314453], "reading_logprob_std": 3.0627449651684784, "tokenization_path_logprob_std": 23.670748291626065, "trajectory_balance_rmse": 451.63815998129854} {"canonical_length": 21, "canonical_nll": 1.2164986928304036, "generation_rouge_l_f1": 0.22222222222222224, "generation_token_f1": 0.24691358024691357, "prompt_id": "9444b92ac8847c28", "reading_canonical_log_prob": -25.546472549438477, "reading_distances": [28, 21, 24, 23], "reading_log_probs": [-27.6888427734375, -26.921268463134766, -26.896320343017578, -27.044719696044922], "reading_logprob_std": 0.37305513968764914, "tokenization_path_logprob_std": 59.517440360611204, "trajectory_balance_rmse": 534.8438689003851} {"canonical_length": 13, "canonical_nll": 2.099725283109225, "generation_rouge_l_f1": 0.32, "generation_token_f1": 0.4, "prompt_id": "d790284919634e0e", "reading_canonical_log_prob": -27.296428680419922, "reading_distances": [32, 30, 32, 24], "reading_log_probs": [-23.963376998901367, -25.059890747070312, -24.974641799926758, -24.344118118286133], "reading_logprob_std": 0.5233800921193367, "tokenization_path_logprob_std": 22.6190977932192, "trajectory_balance_rmse": 425.53089530117336} {"canonical_length": 86, "canonical_nll": 1.0799863948378452, "generation_rouge_l_f1": 0.3417085427135678, "generation_token_f1": 0.4623115577889447, "prompt_id": "58457960c61ffd91", "reading_canonical_log_prob": -92.87882995605469, "reading_distances": [32, 30, 29, 29], "reading_log_probs": [-94.15685272216797, -92.93258666992188, -93.7931900024414, -93.37158966064453], "reading_logprob_std": 0.52905769898036, "tokenization_path_logprob_std": 114.49944248440165, "trajectory_balance_rmse": 1853.8197735136775} {"canonical_length": 6, "canonical_nll": 2.26253350575765, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "02cbb58a2f84ec5b", "reading_canonical_log_prob": -13.575201034545898, "reading_distances": [19, 16, 17, 20], "reading_log_probs": [-12.150174140930176, -12.195215225219727, -12.339442253112793, -11.879308700561523], "reading_logprob_std": 0.192254271393641, "tokenization_path_logprob_std": 18.85977934646447, "trajectory_balance_rmse": 170.32159437784674} {"canonical_length": 50, "canonical_nll": 1.909886474609375, "generation_rouge_l_f1": 0.2315112540192926, "generation_token_f1": 0.3215434083601286, "prompt_id": "d0f9c73543784d9e", "reading_canonical_log_prob": -95.49432373046875, "reading_distances": [40, 32, 29, 40], "reading_log_probs": [-95.56359100341797, -98.31854248046875, -97.94447326660156, -98.57286071777344], "reading_logprob_std": 1.3818326761093647, "tokenization_path_logprob_std": 34.455091948431935, "trajectory_balance_rmse": 496.3930931024857} {"canonical_length": 76, "canonical_nll": 1.5469426606830798, "generation_rouge_l_f1": 0.18018018018018017, "generation_token_f1": 0.3123123123123123, "prompt_id": "e67d8b619feb7d26", "reading_canonical_log_prob": -117.56764221191406, "reading_distances": [6, 10, 9, 8], "reading_log_probs": [-117.35481262207031, -117.38297271728516, -119.23639678955078, -117.2276611328125], "reading_logprob_std": 0.9596718575496275, "tokenization_path_logprob_std": 90.87864923941615, "trajectory_balance_rmse": 1848.6344443842152} {"canonical_length": 260, "canonical_nll": 1.240191180889423, "generation_rouge_l_f1": 0.27949183303085295, "generation_token_f1": 0.44646098003629764, "prompt_id": "aa422d282fda03cd", "reading_canonical_log_prob": -322.44970703125, "reading_distances": [28, 30, 37, 23], "reading_log_probs": [-320.1761474609375, -321.1136169433594, -321.46551513671875, -323.2353515625], "reading_logprob_std": 1.279907512134186, "tokenization_path_logprob_std": 220.3583746626729, "trajectory_balance_rmse": 3679.100364300558} {"canonical_length": 281, "canonical_nll": 1.407435950011121, "generation_rouge_l_f1": 0.18111753371868977, "generation_token_f1": 0.3737957610789981, "prompt_id": "050b55490a11894a", "reading_canonical_log_prob": -395.489501953125, "reading_distances": [19, 17, 26, 18], "reading_log_probs": [-391.28106689453125, -393.5398864746094, -388.5443115234375, -396.35113525390625], "reading_logprob_std": 3.317919460743614, "tokenization_path_logprob_std": 133.00629124066944, "trajectory_balance_rmse": 5864.718388031553} {"canonical_length": 251, "canonical_nll": 0.9244280811325013, "generation_rouge_l_f1": 0.25203252032520324, "generation_token_f1": 0.41463414634146345, "prompt_id": "2d149cddaf627a2b", "reading_canonical_log_prob": -232.0314483642578, "reading_distances": [26, 29, 33, 35], "reading_log_probs": [-225.76724243164062, -227.29608154296875, -228.96136474609375, -229.57687377929688], "reading_logprob_std": 1.7177283966263577, "tokenization_path_logprob_std": 118.96831920029071, "trajectory_balance_rmse": 5266.349283728874} {"canonical_length": 157, "canonical_nll": 1.6310673610419983, "generation_rouge_l_f1": 0.23978201634877383, "generation_token_f1": 0.43596730245231613, "prompt_id": "f7af4006bdd897fe", "reading_canonical_log_prob": -256.07757568359375, "reading_distances": [29, 22, 22, 32], "reading_log_probs": [-253.1239776611328, -249.48614501953125, -253.4927215576172, -253.47314453125], "reading_logprob_std": 1.9459556310167878, "tokenization_path_logprob_std": 73.34586863281152, "trajectory_balance_rmse": 3476.7599302128388} {"canonical_length": 21, "canonical_nll": 0.35681234087262836, "generation_rouge_l_f1": 0.45333333333333325, "generation_token_f1": 0.4799999999999999, "prompt_id": "2d0d3b6d72c67323", "reading_canonical_log_prob": -7.493059158325195, "reading_distances": [16, 31, 24, 23], "reading_log_probs": [-10.017794609069824, -9.349709510803223, -10.17068099975586, -8.809438705444336], "reading_logprob_std": 0.6290669843392552, "tokenization_path_logprob_std": 27.31320737818417, "trajectory_balance_rmse": 376.6295491920562} {"canonical_length": 7, "canonical_nll": 0.7143885067531041, "generation_rouge_l_f1": 0.13888888888888887, "generation_token_f1": 0.13888888888888887, "prompt_id": "974b052b2f839275", "reading_canonical_log_prob": -5.0007195472717285, "reading_distances": [15, 25, 19, 23], "reading_log_probs": [-4.879350662231445, -5.1589131355285645, -5.022687911987305, -5.4354376792907715], "reading_logprob_std": 0.2368753222383638, "tokenization_path_logprob_std": 31.14466472623429, "trajectory_balance_rmse": 101.73365498544442} {"canonical_length": 71, "canonical_nll": 1.3750038684253962, "generation_rouge_l_f1": 0.16296296296296295, "generation_token_f1": 0.29629629629629634, "prompt_id": "7f8aa3c30a8d7e14", "reading_canonical_log_prob": -97.62527465820312, "reading_distances": [32, 30, 32, 37], "reading_log_probs": [-98.1722412109375, -97.94032287597656, -97.99513244628906, -99.05392456054688], "reading_logprob_std": 0.5185454576404853, "tokenization_path_logprob_std": 114.23389681987906, "trajectory_balance_rmse": 1521.360404302696} {"canonical_length": 12, "canonical_nll": 0.69346817334493, "generation_rouge_l_f1": 0.6956521739130435, "generation_token_f1": 0.6956521739130435, "prompt_id": "537e5de8a8d8abfa", "reading_canonical_log_prob": -8.32161808013916, "reading_distances": [57, 49, 64, 54], "reading_log_probs": [-10.668313980102539, -10.62191104888916, -9.259182929992676, -13.784584999084473], "reading_logprob_std": 1.9156754017919961, "tokenization_path_logprob_std": 27.7195138629133, "trajectory_balance_rmse": 297.070847128722} {"canonical_length": 235, "canonical_nll": 1.520505994431516, "generation_rouge_l_f1": 0.27572016460905346, "generation_token_f1": 0.5761316872427984, "prompt_id": "219d66b292bdf703", "reading_canonical_log_prob": -357.31890869140625, "reading_distances": [38, 33, 45, 36], "reading_log_probs": [-357.4572448730469, -357.28900146484375, -357.4549865722656, -359.0263671875], "reading_logprob_std": 0.8167865658645695, "tokenization_path_logprob_std": 267.71040130222224, "trajectory_balance_rmse": 5178.20891613708} {"canonical_length": 233, "canonical_nll": 1.6583794196787822, "generation_rouge_l_f1": 0.16521739130434784, "generation_token_f1": 0.31304347826086953, "prompt_id": "b3b45f7b904e5cc3", "reading_canonical_log_prob": -386.40240478515625, "reading_distances": [26, 14, 21, 19], "reading_log_probs": [-371.4143981933594, -377.5194091796875, -365.09332275390625, -367.7789001464844], "reading_logprob_std": 5.376968367659265, "tokenization_path_logprob_std": 147.97258857298593, "trajectory_balance_rmse": 5116.881229180563} {"canonical_length": 121, "canonical_nll": 1.4309853798101757, "generation_rouge_l_f1": 0.2772861356932153, "generation_token_f1": 0.4188790560471976, "prompt_id": "dd6de628b051978a", "reading_canonical_log_prob": -173.14923095703125, "reading_distances": [38, 39, 33, 35], "reading_log_probs": [-170.93252563476562, -169.4112548828125, -170.43553161621094, -170.44183349609375], "reading_logprob_std": 0.6398776802773635, "tokenization_path_logprob_std": 99.58767498999288, "trajectory_balance_rmse": 2422.30709314784} {"canonical_length": 105, "canonical_nll": 0.8979322887602307, "generation_rouge_l_f1": 0.23564954682779454, "generation_token_f1": 0.2900302114803625, "prompt_id": "d2e124ed83b7b0d6", "reading_canonical_log_prob": -94.28289031982422, "reading_distances": [29, 23, 31, 31], "reading_log_probs": [-92.52989196777344, -94.47471618652344, -92.56793212890625, -92.94479370117188], "reading_logprob_std": 0.9162625461756069, "tokenization_path_logprob_std": 117.65242694422564, "trajectory_balance_rmse": 1701.7281165723516} {"canonical_length": 102, "canonical_nll": 0.9217768650428921, "generation_rouge_l_f1": 0.22754491017964074, "generation_token_f1": 0.3473053892215569, "prompt_id": "40e867680ebfa198", "reading_canonical_log_prob": -94.021240234375, "reading_distances": [36, 38, 40, 29], "reading_log_probs": [-94.16881561279297, -99.77159881591797, -101.12335205078125, -101.07450103759766], "reading_logprob_std": 3.3036901754401415, "tokenization_path_logprob_std": 87.80373577637546, "trajectory_balance_rmse": 1435.158219656891} {"canonical_length": 217, "canonical_nll": 1.4600987588205645, "generation_rouge_l_f1": 0.20374220374220373, "generation_token_f1": 0.42411642411642414, "prompt_id": "7763d96632870ad4", "reading_canonical_log_prob": -316.8414306640625, "reading_distances": [13, 12, 12, 12], "reading_log_probs": [-306.1675720214844, -306.6239013671875, -307.84454345703125, -306.1866760253906], "reading_logprob_std": 0.787955802602396, "tokenization_path_logprob_std": 148.16645024664732, "trajectory_balance_rmse": 4748.704225445092} {"canonical_length": 11, "canonical_nll": 0.7090448899702593, "generation_rouge_l_f1": 0.07258064516129031, "generation_token_f1": 0.07258064516129031, "prompt_id": "151a13661f1584a8", "reading_canonical_log_prob": -7.799493789672852, "reading_distances": [15, 12, 13, 15], "reading_log_probs": [-7.463176727294922, -7.849888801574707, -7.371212482452393, -7.694645404815674], "reading_logprob_std": 0.21783437618801493, "tokenization_path_logprob_std": 35.68883137989002, "trajectory_balance_rmse": 205.8336347385688} {"canonical_length": 166, "canonical_nll": 0.9184454492775791, "generation_rouge_l_f1": 0.31400966183574874, "generation_token_f1": 0.48792270531400966, "prompt_id": "f8789af8051fda52", "reading_canonical_log_prob": -152.46194458007812, "reading_distances": [19, 21, 25, 28], "reading_log_probs": [-152.987548828125, -150.65386962890625, -151.99258422851562, -152.1159210205078], "reading_logprob_std": 0.9635313180130651, "tokenization_path_logprob_std": 163.05243012756492, "trajectory_balance_rmse": 3470.7659090975367} {"canonical_length": 199, "canonical_nll": 0.6378936000804806, "generation_rouge_l_f1": 0.26991150442477874, "generation_token_f1": 0.4646017699115044, "prompt_id": "b3a3308a3d93f328", "reading_canonical_log_prob": -126.94082641601562, "reading_distances": [7, 10, 10, 6], "reading_log_probs": [-126.67021179199219, -124.48085021972656, -125.2349853515625, -126.60538482666016], "reading_logprob_std": 1.073069715789382, "tokenization_path_logprob_std": 89.6289423110323, "trajectory_balance_rmse": 3530.1156794313833} {"canonical_length": 33, "canonical_nll": 0.441596811467951, "generation_rouge_l_f1": 0.8461538461538461, "generation_token_f1": 0.9615384615384615, "prompt_id": "06bad79ff1fc809f", "reading_canonical_log_prob": -14.572694778442383, "reading_distances": [21, 25, 19, 26], "reading_log_probs": [-14.91126823425293, -15.356618881225586, -16.240570068359375, -13.762166976928711], "reading_logprob_std": 1.0308549333953372, "tokenization_path_logprob_std": 44.26606616035617, "trajectory_balance_rmse": 559.2513573480124} {"canonical_length": 153, "canonical_nll": 1.7310352200776145, "generation_rouge_l_f1": 0.2675585284280937, "generation_token_f1": 0.4280936454849499, "prompt_id": "df98426df5120860", "reading_canonical_log_prob": -264.848388671875, "reading_distances": [21, 18, 21, 20], "reading_log_probs": [-260.8895568847656, -258.5898742675781, -257.692626953125, -261.027099609375], "reading_logprob_std": 1.6681238746462101, "tokenization_path_logprob_std": 220.71029941147302, "trajectory_balance_rmse": 3218.9678945744567} {"canonical_length": 25, "canonical_nll": 0.7929039001464844, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.8000000000000002, "prompt_id": "e01eb25b0bbbab91", "reading_canonical_log_prob": -19.82259750366211, "reading_distances": [28, 30, 28, 22], "reading_log_probs": [-18.49089241027832, -19.740400314331055, -20.885534286499023, -20.563356399536133], "reading_logprob_std": 1.0678277727723544, "tokenization_path_logprob_std": 34.02886705812893, "trajectory_balance_rmse": 218.93767695553362} {"canonical_length": 42, "canonical_nll": 2.100397382463728, "generation_rouge_l_f1": 0.1935483870967742, "generation_token_f1": 0.29493087557603687, "prompt_id": "050fda8d8b928503", "reading_canonical_log_prob": -88.21669006347656, "reading_distances": [29, 30, 28, 33], "reading_log_probs": [-82.43810272216797, -82.92462158203125, -82.85253143310547, -82.23709106445312], "reading_logprob_std": 0.32983828026392636, "tokenization_path_logprob_std": 79.67776770566297, "trajectory_balance_rmse": 1012.2987109608114} {"canonical_length": 191, "canonical_nll": 1.1202459684841295, "generation_rouge_l_f1": 0.2959427207637232, "generation_token_f1": 0.5823389021479713, "prompt_id": "c86e9e8e522d0430", "reading_canonical_log_prob": -213.96697998046875, "reading_distances": [30, 21, 19, 24], "reading_log_probs": [-216.80848693847656, -215.2880859375, -218.07733154296875, -220.74041748046875], "reading_logprob_std": 2.3090672834458403, "tokenization_path_logprob_std": 215.56067830743984, "trajectory_balance_rmse": 3270.0208068899346} {"canonical_length": 71, "canonical_nll": 1.133115741568552, "generation_rouge_l_f1": 0.3378378378378379, "generation_token_f1": 0.4189189189189189, "prompt_id": "0d01c13272391fbb", "reading_canonical_log_prob": -80.45121765136719, "reading_distances": [27, 26, 27, 32], "reading_log_probs": [-80.33091735839844, -80.73896789550781, -81.48297119140625, -78.80279541015625], "reading_logprob_std": 1.1296952752631264, "tokenization_path_logprob_std": 36.51861890185422, "trajectory_balance_rmse": 1401.9499381028052} {"canonical_length": 96, "canonical_nll": 1.454250971476237, "generation_rouge_l_f1": 0.19875776397515527, "generation_token_f1": 0.3726708074534161, "prompt_id": "f4314cbcdd2435c8", "reading_canonical_log_prob": -139.60809326171875, "reading_distances": [23, 30, 30, 30], "reading_log_probs": [-135.62701416015625, -134.3909912109375, -135.35800170898438, -137.87435913085938], "reading_logprob_std": 1.4734227694590538, "tokenization_path_logprob_std": 98.02153718660524, "trajectory_balance_rmse": 2103.881795629171} {"canonical_length": 28, "canonical_nll": 1.0274744033813477, "generation_rouge_l_f1": 0.7647058823529412, "generation_token_f1": 0.8235294117647058, "prompt_id": "6039b41734030698", "reading_canonical_log_prob": -28.769283294677734, "reading_distances": [13, 18, 17, 15], "reading_log_probs": [-27.190914154052734, -28.81241226196289, -29.077539443969727, -28.436769485473633], "reading_logprob_std": 0.834803003868727, "tokenization_path_logprob_std": 52.58908513075887, "trajectory_balance_rmse": 318.5284935338302} {"canonical_length": 180, "canonical_nll": 1.0418399386935764, "generation_rouge_l_f1": 0.26198083067092653, "generation_token_f1": 0.536741214057508, "prompt_id": "bc991d771f2ffcd2", "reading_canonical_log_prob": -187.53118896484375, "reading_distances": [28, 34, 27, 26], "reading_log_probs": [-187.41644287109375, -188.0231475830078, -188.3741455078125, -189.79800415039062], "reading_logprob_std": 1.0106849116204342, "tokenization_path_logprob_std": 71.71379373620721, "trajectory_balance_rmse": 3165.7553071506454} {"canonical_length": 322, "canonical_nll": 0.977053529727533, "generation_rouge_l_f1": 0.3443708609271523, "generation_token_f1": 0.6125827814569536, "prompt_id": "ee15b1eaf2aa94fb", "reading_canonical_log_prob": -314.6112365722656, "reading_distances": [31, 29, 29, 32], "reading_log_probs": [-320.2530517578125, -318.70831298828125, -320.035888671875, -319.2342529296875], "reading_logprob_std": 0.7160667220403893, "tokenization_path_logprob_std": 145.11817415082567, "trajectory_balance_rmse": 5818.86839914709} {"canonical_length": 42, "canonical_nll": 1.5174714951288133, "generation_rouge_l_f1": 0.1390728476821192, "generation_token_f1": 0.2052980132450331, "prompt_id": "e36603aac9ffa742", "reading_canonical_log_prob": -63.733802795410156, "reading_distances": [24, 27, 27, 29], "reading_log_probs": [-66.84574127197266, -66.55935668945312, -66.58086395263672, -68.6495361328125], "reading_logprob_std": 1.0022711409693772, "tokenization_path_logprob_std": 118.38394393424589, "trajectory_balance_rmse": 978.3339248399343} {"canonical_length": 248, "canonical_nll": 1.4047118156186995, "generation_rouge_l_f1": 0.21960784313725493, "generation_token_f1": 0.42745098039215684, "prompt_id": "46c0bc7a74b4f8f8", "reading_canonical_log_prob": -348.3685302734375, "reading_distances": [19, 11, 10, 18], "reading_log_probs": [-341.60601806640625, -342.6474304199219, -337.8866271972656, -337.16845703125], "reading_logprob_std": 2.705095826617827, "tokenization_path_logprob_std": 178.04982002744075, "trajectory_balance_rmse": 4703.147072637387} {"canonical_length": 193, "canonical_nll": 1.5253080852291128, "generation_rouge_l_f1": 0.25221238938053103, "generation_token_f1": 0.42035398230088494, "prompt_id": "17af0bbd6fc62e57", "reading_canonical_log_prob": -294.38446044921875, "reading_distances": [22, 30, 30, 29], "reading_log_probs": [-288.73516845703125, -290.481689453125, -290.4703369140625, -289.70660400390625], "reading_logprob_std": 0.8260854716469707, "tokenization_path_logprob_std": 75.0829630595575, "trajectory_balance_rmse": 3388.749458304667} {"canonical_length": 27, "canonical_nll": 1.563160366482205, "generation_rouge_l_f1": 0.07086614173228348, "generation_token_f1": 0.11023622047244094, "prompt_id": "917c56964f6f4a40", "reading_canonical_log_prob": -42.20532989501953, "reading_distances": [36, 35, 28, 33], "reading_log_probs": [-42.55859375, -41.37364196777344, -42.83369445800781, -41.54570770263672], "reading_logprob_std": 0.7260623850639211, "tokenization_path_logprob_std": 49.068920283399066, "trajectory_balance_rmse": 504.39957234480727} {"canonical_length": 350, "canonical_nll": 1.3519955008370537, "generation_rouge_l_f1": 0.16262975778546712, "generation_token_f1": 0.29411764705882354, "prompt_id": "74f8c2e7c18cb610", "reading_canonical_log_prob": -473.19842529296875, "reading_distances": [43, 41, 32, 39], "reading_log_probs": [-476.8258361816406, -470.71368408203125, -474.117919921875, -472.6937561035156], "reading_logprob_std": 2.5707172957459057, "tokenization_path_logprob_std": 193.2241463195297, "trajectory_balance_rmse": 5980.157525389471} {"canonical_length": 26, "canonical_nll": 0.8375932253324069, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.6666666666666666, "prompt_id": "c533f2b5441efb8c", "reading_canonical_log_prob": -21.777423858642578, "reading_distances": [25, 25, 23, 27], "reading_log_probs": [-22.86315155029297, -23.175453186035156, -23.120975494384766, -21.985698699951172], "reading_logprob_std": 0.55085282051494, "tokenization_path_logprob_std": 42.09819730272292, "trajectory_balance_rmse": 245.42693164093586} {"canonical_length": 91, "canonical_nll": 1.1290244636954843, "generation_rouge_l_f1": 0.2458100558659218, "generation_token_f1": 0.46927374301675984, "prompt_id": "ce28dbb2751f3b46", "reading_canonical_log_prob": -102.74122619628906, "reading_distances": [21, 16, 15, 17], "reading_log_probs": [-102.50321960449219, -104.58920288085938, -102.25392150878906, -104.10749816894531], "reading_logprob_std": 1.1586094551035226, "tokenization_path_logprob_std": 17.893737371627672, "trajectory_balance_rmse": 1997.838152517036} {"canonical_length": 13, "canonical_nll": 0.9766280834491436, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.7777777777777778, "prompt_id": "a24d7fb2090fb702", "reading_canonical_log_prob": -12.696165084838867, "reading_distances": [25, 25, 28, 31], "reading_log_probs": [-12.67221736907959, -13.021867752075195, -13.431234359741211, -13.512370109558105], "reading_logprob_std": 0.3893344108691761, "tokenization_path_logprob_std": 40.53647100104021, "trajectory_balance_rmse": 214.31152694319994} {"canonical_length": 133, "canonical_nll": 1.1228267124720983, "generation_rouge_l_f1": 0.5062240663900415, "generation_token_f1": 0.6390041493775933, "prompt_id": "4973bf9719e2751b", "reading_canonical_log_prob": -149.33595275878906, "reading_distances": [18, 18, 23, 24], "reading_log_probs": [-156.33668518066406, -155.8686065673828, -154.9134521484375, -156.15927124023438], "reading_logprob_std": 0.634102823158513, "tokenization_path_logprob_std": 83.61460670145964, "trajectory_balance_rmse": 2656.647316748991} {"canonical_length": 15, "canonical_nll": 3.110218302408854, "generation_rouge_l_f1": 0.2580645161290323, "generation_token_f1": 0.2580645161290323, "prompt_id": "c4461f4a56e11ca5", "reading_canonical_log_prob": -46.65327453613281, "reading_distances": [18, 23, 21, 22], "reading_log_probs": [-49.247398376464844, -46.84668731689453, -45.9578857421875, -46.521446228027344], "reading_logprob_std": 1.4499561057860153, "tokenization_path_logprob_std": 4.558235014259806, "trajectory_balance_rmse": 302.95105105219096} {"canonical_length": 51, "canonical_nll": 0.8788037767597273, "generation_rouge_l_f1": 0.5436893203883496, "generation_token_f1": 0.6213592233009708, "prompt_id": "cdeadd1b86cf8871", "reading_canonical_log_prob": -44.818992614746094, "reading_distances": [13, 10, 11, 10], "reading_log_probs": [-45.622432708740234, -45.21968078613281, -45.852088928222656, -45.176841735839844], "reading_logprob_std": 0.3254786755269911, "tokenization_path_logprob_std": 93.92675842149804, "trajectory_balance_rmse": 692.2534647551893} {"canonical_length": 11, "canonical_nll": 1.1011008349331943, "generation_rouge_l_f1": 0.19999999999999998, "generation_token_f1": 0.3, "prompt_id": "55ee0fbefdef0ed3", "reading_canonical_log_prob": -12.112109184265137, "reading_distances": [15, 14, 16, 12], "reading_log_probs": [-13.931760787963867, -10.82613754272461, -14.713760375976562, -14.081215858459473], "reading_logprob_std": 1.7413599718072241, "tokenization_path_logprob_std": 53.498981364437945, "trajectory_balance_rmse": 254.21148817641313} {"canonical_length": 282, "canonical_nll": 1.5508220483225288, "generation_rouge_l_f1": 0.22761194029850748, "generation_token_f1": 0.40671641791044777, "prompt_id": "0b1cafe0e71072e8", "reading_canonical_log_prob": -437.3318176269531, "reading_distances": [30, 22, 21, 21], "reading_log_probs": [-431.830078125, -427.3461608886719, -435.6373291015625, -431.6255798339844], "reading_logprob_std": 3.388627086319595, "tokenization_path_logprob_std": 278.66104061795386, "trajectory_balance_rmse": 5665.435372162867} {"canonical_length": 98, "canonical_nll": 1.257784162248884, "generation_rouge_l_f1": 0.32608695652173914, "generation_token_f1": 0.5760869565217391, "prompt_id": "3a37b4e0a87fe5c3", "reading_canonical_log_prob": -123.26284790039062, "reading_distances": [15, 15, 20, 16], "reading_log_probs": [-128.64041137695312, -125.51666259765625, -125.57107543945312, -125.64033508300781], "reading_logprob_std": 1.5330291322001224, "tokenization_path_logprob_std": 130.9281584754749, "trajectory_balance_rmse": 2104.8994644424506} {"canonical_length": 261, "canonical_nll": 0.9024569924307052, "generation_rouge_l_f1": 0.22535211267605634, "generation_token_f1": 0.4668008048289738, "prompt_id": "7236a1de81b123d6", "reading_canonical_log_prob": -235.54127502441406, "reading_distances": [24, 22, 23, 24], "reading_log_probs": [-232.9237060546875, -232.42184448242188, -232.32876586914062, -233.11715698242188], "reading_logprob_std": 0.3826362254091372, "tokenization_path_logprob_std": 213.2899537603778, "trajectory_balance_rmse": 5067.5821529302975} {"canonical_length": 181, "canonical_nll": 1.5349505514071133, "generation_rouge_l_f1": 0.26291079812206575, "generation_token_f1": 0.5070422535211268, "prompt_id": "564e5eeb9b5a33fe", "reading_canonical_log_prob": -277.8260498046875, "reading_distances": [42, 46, 36, 36], "reading_log_probs": [-270.7541198730469, -268.5836181640625, -272.6142578125, -270.42633056640625], "reading_logprob_std": 1.6509418337106898, "tokenization_path_logprob_std": 177.56515896887643, "trajectory_balance_rmse": 3380.1634333565303} {"canonical_length": 141, "canonical_nll": 1.0671866126094305, "generation_rouge_l_f1": 0.3002680965147453, "generation_token_f1": 0.45576407506702415, "prompt_id": "8686f6b687c7ab8c", "reading_canonical_log_prob": -150.4733123779297, "reading_distances": [22, 20, 25, 24], "reading_log_probs": [-151.13751220703125, -148.4554901123047, -149.92578125, -152.358154296875], "reading_logprob_std": 1.6698424064015904, "tokenization_path_logprob_std": 153.74015132631473, "trajectory_balance_rmse": 2777.093909851633} {"canonical_length": 101, "canonical_nll": 1.2988377939356435, "generation_rouge_l_f1": 0.31390134529147984, "generation_token_f1": 0.5022421524663677, "prompt_id": "79b5d532fcbe3323", "reading_canonical_log_prob": -131.1826171875, "reading_distances": [28, 26, 24, 28], "reading_log_probs": [-131.4662322998047, -132.42039489746094, -130.84376525878906, -130.30087280273438], "reading_logprob_std": 0.9096162828240001, "tokenization_path_logprob_std": 153.66543426260506, "trajectory_balance_rmse": 2562.2068018574832} {"canonical_length": 333, "canonical_nll": 1.0192690554323856, "generation_rouge_l_f1": 0.23934426229508196, "generation_token_f1": 0.4295081967213115, "prompt_id": "f5411f1c63da4002", "reading_canonical_log_prob": -339.4165954589844, "reading_distances": [45, 44, 44, 54], "reading_log_probs": [-339.1833190917969, -339.5838623046875, -337.9820556640625, -339.9779052734375], "reading_logprob_std": 0.8631003189410213, "tokenization_path_logprob_std": 216.27319315933616, "trajectory_balance_rmse": 7205.353955426432} {"canonical_length": 105, "canonical_nll": 1.3697012765066965, "generation_rouge_l_f1": 0.1945945945945946, "generation_token_f1": 0.3621621621621622, "prompt_id": "0709a316fd9c3cd1", "reading_canonical_log_prob": -143.81863403320312, "reading_distances": [30, 26, 31, 25], "reading_log_probs": [-138.66754150390625, -141.24752807617188, -139.78131103515625, -137.91665649414062], "reading_logprob_std": 1.448622220260245, "tokenization_path_logprob_std": 60.24188399544527, "trajectory_balance_rmse": 1769.5521372759947} {"canonical_length": 68, "canonical_nll": 0.7493219375610352, "generation_rouge_l_f1": 0.25316455696202533, "generation_token_f1": 0.25316455696202533, "prompt_id": "28c4e87a8705b13a", "reading_canonical_log_prob": -50.95389175415039, "reading_distances": [29, 29, 25, 27], "reading_log_probs": [-51.48341369628906, -50.702091217041016, -52.062408447265625, -52.31605911254883], "reading_logprob_std": 0.7163967779816344, "tokenization_path_logprob_std": 120.5593517729214, "trajectory_balance_rmse": 1228.497802613345} {"canonical_length": 10, "canonical_nll": 0.6880414009094238, "generation_rouge_l_f1": 0.8888888888888888, "generation_token_f1": 0.8888888888888888, "prompt_id": "d86571d08d8aa4d7", "reading_canonical_log_prob": -6.880414009094238, "reading_distances": [36, 42, 39, 47], "reading_log_probs": [-5.832896709442139, -5.722782135009766, -6.247067928314209, -5.509298801422119], "reading_logprob_std": 0.30999094549881784, "tokenization_path_logprob_std": 27.18397151650318, "trajectory_balance_rmse": 239.28008325450799} {"canonical_length": 30, "canonical_nll": 0.6510667165120443, "generation_rouge_l_f1": 0.34782608695652173, "generation_token_f1": 0.38260869565217387, "prompt_id": "f8bb815af5d18371", "reading_canonical_log_prob": -19.532001495361328, "reading_distances": [26, 20, 27, 18], "reading_log_probs": [-17.73379135131836, -18.33287239074707, -16.97064971923828, -17.520355224609375], "reading_logprob_std": 0.5630903655481379, "tokenization_path_logprob_std": 107.16857783732118, "trajectory_balance_rmse": 534.3865847095908} {"canonical_length": 39, "canonical_nll": 1.6734824547400842, "generation_rouge_l_f1": 0.4285714285714286, "generation_token_f1": 0.4857142857142857, "prompt_id": "e8ce7622bc908c26", "reading_canonical_log_prob": -65.26581573486328, "reading_distances": [19, 20, 22, 15], "reading_log_probs": [-68.6649169921875, -69.58474731445312, -66.67562866210938, -64.48040771484375], "reading_logprob_std": 2.2666002967548247, "tokenization_path_logprob_std": 103.32278786216537, "trajectory_balance_rmse": 847.8309446716036} {"canonical_length": 35, "canonical_nll": 0.8718447004045758, "generation_rouge_l_f1": 0.10631229235880399, "generation_token_f1": 0.16611295681063123, "prompt_id": "300a5e4b8236fdd5", "reading_canonical_log_prob": -30.514564514160156, "reading_distances": [33, 33, 27, 25], "reading_log_probs": [-30.362953186035156, -33.06111145019531, -33.59503173828125, -30.899173736572266], "reading_logprob_std": 1.5874670123123045, "tokenization_path_logprob_std": 129.64687481494582, "trajectory_balance_rmse": 741.5826437824142} {"canonical_length": 64, "canonical_nll": 0.8888909816741943, "generation_rouge_l_f1": 0.5433526011560693, "generation_token_f1": 0.6358381502890174, "prompt_id": "20d20afcda52df5b", "reading_canonical_log_prob": -56.88902282714844, "reading_distances": [28, 29, 26, 27], "reading_log_probs": [-57.955936431884766, -59.50926208496094, -57.433650970458984, -57.499267578125], "reading_logprob_std": 0.9681035621715872, "tokenization_path_logprob_std": 174.87038039328255, "trajectory_balance_rmse": 1284.3384067048032} {"canonical_length": 92, "canonical_nll": 1.9330877221148948, "generation_rouge_l_f1": 0.32335329341317365, "generation_token_f1": 0.4790419161676646, "prompt_id": "26bff204aae37f53", "reading_canonical_log_prob": -177.8440704345703, "reading_distances": [15, 14, 19, 14], "reading_log_probs": [-172.89535522460938, -171.731201171875, -170.96835327148438, -171.20904541015625], "reading_logprob_std": 0.8575521480941787, "tokenization_path_logprob_std": 100.44854629381015, "trajectory_balance_rmse": 1960.996762599987} {"canonical_length": 259, "canonical_nll": 1.546646059249819, "generation_rouge_l_f1": 0.23880597014925375, "generation_token_f1": 0.4477611940298507, "prompt_id": "d7bb721e18ff9d29", "reading_canonical_log_prob": -400.5813293457031, "reading_distances": [21, 22, 26, 25], "reading_log_probs": [-413.95111083984375, -416.152099609375, -405.5218505859375, -416.08282470703125], "reading_logprob_std": 5.0413469736275704, "tokenization_path_logprob_std": 149.35079586746673, "trajectory_balance_rmse": 5686.22985350661} {"canonical_length": 110, "canonical_nll": 0.7610369595614347, "generation_rouge_l_f1": 0.3252032520325204, "generation_token_f1": 0.4010840108401084, "prompt_id": "6df1e56bb497e6a2", "reading_canonical_log_prob": -83.71406555175781, "reading_distances": [41, 43, 35, 42], "reading_log_probs": [-89.0523681640625, -87.9012222290039, -86.03083801269531, -86.74030303955078], "reading_logprob_std": 1.3275867098107685, "tokenization_path_logprob_std": 105.87464444680691, "trajectory_balance_rmse": 1789.862513499677} {"canonical_length": 9, "canonical_nll": 0.3176446755727132, "generation_rouge_l_f1": 0.823529411764706, "generation_token_f1": 0.823529411764706, "prompt_id": "196c78145c93b0b5", "reading_canonical_log_prob": -2.858802080154419, "reading_distances": [36, 41, 38, 35], "reading_log_probs": [-3.493543863296509, -3.4802613258361816, -3.852203369140625, -3.8766403198242188], "reading_logprob_std": 0.21825640825815085, "tokenization_path_logprob_std": 46.47288636192153, "trajectory_balance_rmse": 181.14899517505745} {"canonical_length": 61, "canonical_nll": 0.4634282159023598, "generation_rouge_l_f1": 0.2, "generation_token_f1": 0.3103448275862069, "prompt_id": "31d1718162ea0ff4", "reading_canonical_log_prob": -28.269121170043945, "reading_distances": [25, 22, 22, 19], "reading_log_probs": [-30.463146209716797, -29.530757904052734, -30.47743797302246, -30.44931411743164], "reading_logprob_std": 0.4664121170514768, "tokenization_path_logprob_std": 67.70221026163634, "trajectory_balance_rmse": 921.3512112699127} {"canonical_length": 117, "canonical_nll": 0.9513233543461205, "generation_rouge_l_f1": 0.2254335260115607, "generation_token_f1": 0.369942196531792, "prompt_id": "823548660cad1f9e", "reading_canonical_log_prob": -111.3048324584961, "reading_distances": [23, 23, 25, 25], "reading_log_probs": [-107.89691162109375, -107.54134368896484, -109.75303649902344, -107.19085693359375], "reading_logprob_std": 1.1419767771949052, "tokenization_path_logprob_std": 132.20856323697237, "trajectory_balance_rmse": 1595.8701774774256} {"canonical_length": 14, "canonical_nll": 0.7205044882638114, "generation_rouge_l_f1": 0.14285714285714288, "generation_token_f1": 0.14285714285714288, "prompt_id": "912aaeeaa0cce8d9", "reading_canonical_log_prob": -10.08706283569336, "reading_distances": [32, 32, 28, 26], "reading_log_probs": [-7.60540246963501, -8.191961288452148, -8.145828247070312, -7.645939826965332], "reading_logprob_std": 0.31463082502281214, "tokenization_path_logprob_std": 46.178290182354985, "trajectory_balance_rmse": 344.86225607585095} {"canonical_length": 368, "canonical_nll": 1.5070747707201086, "generation_rouge_l_f1": 0.25039619651347067, "generation_token_f1": 0.4152139461172742, "prompt_id": "4547f9ca1d45d608", "reading_canonical_log_prob": -554.603515625, "reading_distances": [25, 26, 27, 25], "reading_log_probs": [-555.3868408203125, -553.6582641601562, -554.6441650390625, -555.471923828125], "reading_logprob_std": 0.8412951449712077, "tokenization_path_logprob_std": 237.66924544515246, "trajectory_balance_rmse": 7254.401521647021} {"canonical_length": 228, "canonical_nll": 1.2840932210286458, "generation_rouge_l_f1": 0.308, "generation_token_f1": 0.504, "prompt_id": "99e3f7e86895d3aa", "reading_canonical_log_prob": -292.77325439453125, "reading_distances": [20, 23, 20, 19], "reading_log_probs": [-293.47607421875, -294.888427734375, -294.9603576660156, -293.32147216796875], "reading_logprob_std": 0.8835633790178129, "tokenization_path_logprob_std": 31.237800264028472, "trajectory_balance_rmse": 4827.381718393167} {"canonical_length": 318, "canonical_nll": 0.7255755730395047, "generation_rouge_l_f1": 0.2711864406779661, "generation_token_f1": 0.4896421845574388, "prompt_id": "d3e72bee48c9454b", "reading_canonical_log_prob": -230.7330322265625, "reading_distances": [17, 25, 23, 29], "reading_log_probs": [-232.47183227539062, -230.2112579345703, -232.069580078125, -232.2169189453125], "reading_logprob_std": 1.0341957425721726, "tokenization_path_logprob_std": 58.184930481056895, "trajectory_balance_rmse": 4936.1918342844765} {"canonical_length": 281, "canonical_nll": 1.4492261350366993, "generation_rouge_l_f1": 0.21092278719397364, "generation_token_f1": 0.45574387947269307, "prompt_id": "3126c4e8c381cc6a", "reading_canonical_log_prob": -407.2325439453125, "reading_distances": [25, 25, 24, 26], "reading_log_probs": [-405.0392761230469, -407.06976318359375, -406.73614501953125, -406.3789367675781], "reading_logprob_std": 0.8903684793916836, "tokenization_path_logprob_std": 214.59268823441258, "trajectory_balance_rmse": 6038.860064236259} {"canonical_length": 23, "canonical_nll": 0.23701427293860394, "generation_rouge_l_f1": 0.20408163265306123, "generation_token_f1": 0.20408163265306123, "prompt_id": "963018756c1975a3", "reading_canonical_log_prob": -5.451328277587891, "reading_distances": [44, 40, 38, 46], "reading_log_probs": [-4.542697906494141, -5.152806282043457, -4.958066463470459, -4.769632339477539], "reading_logprob_std": 0.26085049856157133, "tokenization_path_logprob_std": 49.50317704149836, "trajectory_balance_rmse": 336.22716873034136} {"canonical_length": 50, "canonical_nll": 2.0315943908691407, "generation_rouge_l_f1": 0.296875, "generation_token_f1": 0.42187499999999994, "prompt_id": "b9207d2d35a02140", "reading_canonical_log_prob": -101.57971954345703, "reading_distances": [35, 37, 39, 39], "reading_log_probs": [-103.40495300292969, -104.01337432861328, -103.02572631835938, -104.869140625], "reading_logprob_std": 0.804352515618306, "tokenization_path_logprob_std": 19.00428112531955, "trajectory_balance_rmse": 1157.3600030412952} {"canonical_length": 59, "canonical_nll": 0.6397355936341367, "generation_rouge_l_f1": 0.31788079470198677, "generation_token_f1": 0.5298013245033112, "prompt_id": "cc824ceca446b0f2", "reading_canonical_log_prob": -37.74440002441406, "reading_distances": [12, 11, 16, 14], "reading_log_probs": [-42.27400207519531, -41.859161376953125, -40.42253112792969, -44.095001220703125], "reading_logprob_std": 1.5128890194770686, "tokenization_path_logprob_std": 106.96239016737485, "trajectory_balance_rmse": 1125.2180690108867} {"canonical_length": 27, "canonical_nll": 0.9485952589246962, "generation_rouge_l_f1": 0.68, "generation_token_f1": 0.8000000000000002, "prompt_id": "9cebf3474b92652d", "reading_canonical_log_prob": -25.612071990966797, "reading_distances": [21, 26, 27, 25], "reading_log_probs": [-24.9827880859375, -25.272972106933594, -25.201095581054688, -25.202932357788086], "reading_logprob_std": 0.1259641496868527, "tokenization_path_logprob_std": 50.07911552696789, "trajectory_balance_rmse": 610.7735250264956} {"canonical_length": 5, "canonical_nll": 3.8448657989501953, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "c004648702d6143d", "reading_canonical_log_prob": -19.224328994750977, "reading_distances": [22, 19, 24, 21], "reading_log_probs": [-20.49344825744629, -20.099279403686523, -24.296337127685547, -18.99453353881836], "reading_logprob_std": 2.3059516552794337, "tokenization_path_logprob_std": 2.409517766137863, "trajectory_balance_rmse": 58.72391163475374} {"canonical_length": 120, "canonical_nll": 1.2470474243164062, "generation_rouge_l_f1": 0.22916666666666663, "generation_token_f1": 0.45138888888888884, "prompt_id": "ebe4917bfacc9af5", "reading_canonical_log_prob": -149.64569091796875, "reading_distances": [28, 27, 18, 32], "reading_log_probs": [-151.9681396484375, -153.31797790527344, -152.59088134765625, -153.67762756347656], "reading_logprob_std": 0.7621927847539895, "tokenization_path_logprob_std": 144.7329044320698, "trajectory_balance_rmse": 2704.9783054670374} {"canonical_length": 127, "canonical_nll": 1.4844581423781988, "generation_rouge_l_f1": 0.24810126582278486, "generation_token_f1": 0.45063291139240513, "prompt_id": "bd51aa5053398c97", "reading_canonical_log_prob": -188.52618408203125, "reading_distances": [31, 21, 23, 18], "reading_log_probs": [-186.50408935546875, -186.1471710205078, -186.90097045898438, -186.3021240234375], "reading_logprob_std": 0.326156453182782, "tokenization_path_logprob_std": 83.15407497490638, "trajectory_balance_rmse": 2881.060997408104} {"canonical_length": 168, "canonical_nll": 1.4118394397553944, "generation_rouge_l_f1": 0.16075650118203308, "generation_token_f1": 0.3309692671394799, "prompt_id": "0d7273ba3d245c40", "reading_canonical_log_prob": -237.18902587890625, "reading_distances": [28, 22, 20, 15], "reading_log_probs": [-232.6678466796875, -229.52285766601562, -234.05433654785156, -234.39410400390625], "reading_logprob_std": 2.220586597898347, "tokenization_path_logprob_std": 220.69725146531133, "trajectory_balance_rmse": 3341.340278610717} {"canonical_length": 211, "canonical_nll": 1.6421669331772069, "generation_rouge_l_f1": 0.210752688172043, "generation_token_f1": 0.3655913978494624, "prompt_id": "16d4d8a6fd8e0475", "reading_canonical_log_prob": -346.4972229003906, "reading_distances": [32, 31, 29, 33], "reading_log_probs": [-341.72308349609375, -339.94488525390625, -343.17840576171875, -343.54901123046875], "reading_logprob_std": 1.6380041544347608, "tokenization_path_logprob_std": 269.2432851120829, "trajectory_balance_rmse": 4367.818653113781} {"canonical_length": 131, "canonical_nll": 1.7351124304851502, "generation_rouge_l_f1": 0.3225806451612903, "generation_token_f1": 0.4700460829493088, "prompt_id": "5669cff2cb390e91", "reading_canonical_log_prob": -227.2997283935547, "reading_distances": [26, 24, 21, 20], "reading_log_probs": [-233.6702880859375, -231.73312377929688, -230.37667846679688, -232.77438354492188], "reading_logprob_std": 1.4164589136237353, "tokenization_path_logprob_std": 176.41277029202138, "trajectory_balance_rmse": 2761.39197669405} {"canonical_length": 314, "canonical_nll": 1.0727788840129877, "generation_rouge_l_f1": 0.231433506044905, "generation_token_f1": 0.4559585492227979, "prompt_id": "ea8af27e7f1b048e", "reading_canonical_log_prob": -336.8525695800781, "reading_distances": [27, 27, 28, 30], "reading_log_probs": [-332.6726379394531, -331.5997009277344, -332.9975280761719, -334.1137390136719], "reading_logprob_std": 1.0349619489629942, "tokenization_path_logprob_std": 208.64258757790233, "trajectory_balance_rmse": 6384.316654997128} {"canonical_length": 8, "canonical_nll": 0.8191691637039185, "generation_rouge_l_f1": 0.049180327868852465, "generation_token_f1": 0.049180327868852465, "prompt_id": "b75f2d55eed057da", "reading_canonical_log_prob": -6.553353309631348, "reading_distances": [14, 17, 18, 17], "reading_log_probs": [-6.663610458374023, -6.3567328453063965, -5.196533203125, -6.277020454406738], "reading_logprob_std": 0.6400398001037165, "tokenization_path_logprob_std": 36.40392662021228, "trajectory_balance_rmse": 148.7479617579309} {"canonical_length": 22, "canonical_nll": 0.6706272472034801, "generation_rouge_l_f1": 0.2439024390243902, "generation_token_f1": 0.2601626016260163, "prompt_id": "55725f89c2c9fced", "reading_canonical_log_prob": -14.753799438476562, "reading_distances": [35, 29, 30, 32], "reading_log_probs": [-15.00441837310791, -15.53650188446045, -14.464569091796875, -15.793094635009766], "reading_logprob_std": 0.5899449286631107, "tokenization_path_logprob_std": 37.65713364841164, "trajectory_balance_rmse": 431.6176166005752} {"canonical_length": 6, "canonical_nll": 0.6928532123565674, "generation_rouge_l_f1": 0.8571428571428571, "generation_token_f1": 0.8571428571428571, "prompt_id": "e8e15c8f276bd3e5", "reading_canonical_log_prob": -4.157119274139404, "reading_distances": [21, 24, 21, 25], "reading_log_probs": [-4.484629154205322, -4.551519393920898, -4.438352108001709, -5.0613908767700195], "reading_logprob_std": 0.2887073236577926, "tokenization_path_logprob_std": 17.42411080014611, "trajectory_balance_rmse": 75.80114723132543} {"canonical_length": 37, "canonical_nll": 1.3771106616870776, "generation_rouge_l_f1": 0.12244897959183672, "generation_token_f1": 0.17006802721088438, "prompt_id": "6e9b2ca73b16e2a8", "reading_canonical_log_prob": -50.953094482421875, "reading_distances": [36, 33, 30, 32], "reading_log_probs": [-55.5059814453125, -56.96484375, -57.437782287597656, -60.41755676269531], "reading_logprob_std": 2.0617076204657776, "tokenization_path_logprob_std": 84.61123336975773, "trajectory_balance_rmse": 878.5632293357323} {"canonical_length": 266, "canonical_nll": 1.437238191303454, "generation_rouge_l_f1": 0.22672064777327935, "generation_token_f1": 0.4736842105263158, "prompt_id": "fb1406f3dcf9e9ba", "reading_canonical_log_prob": -382.30535888671875, "reading_distances": [24, 25, 22, 33], "reading_log_probs": [-375.7265930175781, -372.3326416015625, -373.4534606933594, -372.1551513671875], "reading_logprob_std": 1.6435372819543637, "tokenization_path_logprob_std": 55.35498169321822, "trajectory_balance_rmse": 5263.520580340451} {"canonical_length": 3, "canonical_nll": 0.1692753036816915, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "f6b05fa9b7a6a33d", "reading_canonical_log_prob": -0.5078259110450745, "reading_distances": [34, 24, 24, 32], "reading_log_probs": [-3.6331663131713867, -2.1915104389190674, -1.8267751932144165, -2.5918664932250977], "reading_logprob_std": 0.7801926749248413, "tokenization_path_logprob_std": 30.403329754851164, "trajectory_balance_rmse": 110.43255833945891} {"canonical_length": 274, "canonical_nll": 1.2806008888857208, "generation_rouge_l_f1": 0.19136960600375236, "generation_token_f1": 0.36397748592870544, "prompt_id": "b4b2d4b733fbf2fd", "reading_canonical_log_prob": -350.8846435546875, "reading_distances": [27, 40, 31, 33], "reading_log_probs": [-356.21502685546875, -352.46600341796875, -353.06121826171875, -360.7021484375], "reading_logprob_std": 3.7716941917945457, "tokenization_path_logprob_std": 270.3899071348181, "trajectory_balance_rmse": 5067.581077744096} {"canonical_length": 309, "canonical_nll": 0.7175047050402003, "generation_rouge_l_f1": 0.26692456479690524, "generation_token_f1": 0.4564796905222437, "prompt_id": "8c9f3f4649945cbf", "reading_canonical_log_prob": -221.70895385742188, "reading_distances": [52, 46, 43, 42], "reading_log_probs": [-219.10745239257812, -212.76473999023438, -218.46481323242188, -215.10281372070312], "reading_logprob_std": 2.971259287651946, "tokenization_path_logprob_std": 118.49317979507168, "trajectory_balance_rmse": 2715.742300403683} {"canonical_length": 8, "canonical_nll": 0.13581109046936035, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "eccb7866ba7ed3da", "reading_canonical_log_prob": -1.0864887237548828, "reading_distances": [20, 26, 26, 26], "reading_log_probs": [-1.5598716735839844, -1.1289697885513306, -1.8619954586029053, -1.2788276672363281], "reading_logprob_std": 0.3234977639217792, "tokenization_path_logprob_std": 39.18017823338099, "trajectory_balance_rmse": 188.26910353949367} {"canonical_length": 71, "canonical_nll": 2.034702784578565, "generation_rouge_l_f1": 0.18902439024390244, "generation_token_f1": 0.2682926829268293, "prompt_id": "f78297865f8b46fc", "reading_canonical_log_prob": -144.46389770507812, "reading_distances": [26, 17, 21, 20], "reading_log_probs": [-141.394775390625, -140.38616943359375, -141.4348907470703, -142.92926025390625], "reading_logprob_std": 1.047767728342806, "tokenization_path_logprob_std": 67.07974783374165, "trajectory_balance_rmse": 1714.0041236326708} {"canonical_length": 147, "canonical_nll": 1.021930201523969, "generation_rouge_l_f1": 0.19559902200489, "generation_token_f1": 0.35207823960880197, "prompt_id": "65091372e7851e26", "reading_canonical_log_prob": -150.22373962402344, "reading_distances": [27, 26, 19, 24], "reading_log_probs": [-148.99986267089844, -149.4447784423828, -149.716796875, -151.12088012695312], "reading_logprob_std": 0.9158520808848442, "tokenization_path_logprob_std": 164.79144038688406, "trajectory_balance_rmse": 2875.086898289848} {"canonical_length": 16, "canonical_nll": 1.674704670906067, "generation_rouge_l_f1": 0.16666666666666669, "generation_token_f1": 0.16666666666666669, "prompt_id": "aa225c057d9ef9cd", "reading_canonical_log_prob": -26.79527473449707, "reading_distances": [28, 29, 24, 25], "reading_log_probs": [-24.109376907348633, -24.577898025512695, -26.74039077758789, -24.863557815551758], "reading_logprob_std": 1.1543762114833342, "tokenization_path_logprob_std": 20.620885759254453, "trajectory_balance_rmse": 245.0220647166394} {"canonical_length": 352, "canonical_nll": 0.9854642694646661, "generation_rouge_l_f1": 0.30158730158730157, "generation_token_f1": 0.4841269841269842, "prompt_id": "1b2704eec1f82afa", "reading_canonical_log_prob": -346.8834228515625, "reading_distances": [31, 31, 32, 34], "reading_log_probs": [-347.94647216796875, -346.4415283203125, -351.29766845703125, -351.4684143066406], "reading_logprob_std": 2.49633536196443, "tokenization_path_logprob_std": 351.49795099909716, "trajectory_balance_rmse": 6181.437078214564} {"canonical_length": 284, "canonical_nll": 1.3830701801138865, "generation_rouge_l_f1": 0.2554744525547445, "generation_token_f1": 0.5218978102189782, "prompt_id": "aa3d50bf234f7b48", "reading_canonical_log_prob": -392.79193115234375, "reading_distances": [36, 33, 34, 30], "reading_log_probs": [-386.927978515625, -385.6453857421875, -388.2804870605469, -387.28765869140625], "reading_logprob_std": 1.088968203688558, "tokenization_path_logprob_std": 256.9542513958653, "trajectory_balance_rmse": 6834.138809127833} {"canonical_length": 11, "canonical_nll": 0.7235522703690962, "generation_rouge_l_f1": 0.048, "generation_token_f1": 0.056, "prompt_id": "38ebe3c79fd5a7f3", "reading_canonical_log_prob": -7.959074974060059, "reading_distances": [23, 13, 19, 23], "reading_log_probs": [-8.594779968261719, -8.165348052978516, -7.050817012786865, -8.099137306213379], "reading_logprob_std": 0.6557078723494518, "tokenization_path_logprob_std": 37.17639814160661, "trajectory_balance_rmse": 115.10682471177542} {"canonical_length": 266, "canonical_nll": 1.3063111269384398, "generation_rouge_l_f1": 0.20727272727272725, "generation_token_f1": 0.39272727272727276, "prompt_id": "df8d0cc99eb20244", "reading_canonical_log_prob": -347.478759765625, "reading_distances": [24, 18, 23, 28], "reading_log_probs": [-347.8343505859375, -348.7777099609375, -346.82965087890625, -348.75579833984375], "reading_logprob_std": 0.9243859548545057, "tokenization_path_logprob_std": 285.3726900627965, "trajectory_balance_rmse": 5407.935055258936} {"canonical_length": 131, "canonical_nll": 1.6610186627802959, "generation_rouge_l_f1": 0.24309392265193375, "generation_token_f1": 0.430939226519337, "prompt_id": "0f4d389ccaad372e", "reading_canonical_log_prob": -217.59344482421875, "reading_distances": [32, 29, 29, 26], "reading_log_probs": [-213.45236206054688, -210.04913330078125, -212.37301635742188, -212.0467071533203], "reading_logprob_std": 1.4206814770363703, "tokenization_path_logprob_std": 119.04347570330664, "trajectory_balance_rmse": 2300.4414979887033} {"canonical_length": 30, "canonical_nll": 1.4981575012207031, "generation_rouge_l_f1": 0.2727272727272727, "generation_token_f1": 0.36363636363636365, "prompt_id": "936a9a79e3069ca4", "reading_canonical_log_prob": -44.944725036621094, "reading_distances": [27, 27, 24, 26], "reading_log_probs": [-45.12260818481445, -45.182125091552734, -43.7993049621582, -47.11725616455078], "reading_logprob_std": 1.3662301601409734, "tokenization_path_logprob_std": 112.10788094570078, "trajectory_balance_rmse": 703.2115605054423} {"canonical_length": 427, "canonical_nll": 1.5282597061621779, "generation_rouge_l_f1": 0.20364741641337386, "generation_token_f1": 0.34346504559270513, "prompt_id": "82391a0f7d399b89", "reading_canonical_log_prob": -652.56689453125, "reading_distances": [24, 29, 30, 30], "reading_log_probs": [-648.979248046875, -644.9671630859375, -646.718505859375, -649.4476318359375], "reading_logprob_std": 2.0820081386273492, "tokenization_path_logprob_std": 215.93961415909752, "trajectory_balance_rmse": 7086.565951869513} {"canonical_length": 189, "canonical_nll": 0.9358807922040344, "generation_rouge_l_f1": 0.3257328990228013, "generation_token_f1": 0.4820846905537459, "prompt_id": "0bf403e155ca208a", "reading_canonical_log_prob": -176.8814697265625, "reading_distances": [23, 21, 22, 24], "reading_log_probs": [-173.4970703125, -174.9725341796875, -177.364990234375, -175.93890380859375], "reading_logprob_std": 1.6276718469217428, "tokenization_path_logprob_std": 87.18380698195584, "trajectory_balance_rmse": 3160.775097608903} {"canonical_length": 175, "canonical_nll": 0.6589038521902901, "generation_rouge_l_f1": 0.20535714285714285, "generation_token_f1": 0.3214285714285714, "prompt_id": "3b3aada8838b1247", "reading_canonical_log_prob": -115.30817413330078, "reading_distances": [23, 18, 21, 16], "reading_log_probs": [-111.1463851928711, -109.9287109375, -110.82870483398438, -111.68618774414062], "reading_logprob_std": 0.7364910543568473, "tokenization_path_logprob_std": 69.67341346789499, "trajectory_balance_rmse": 2684.9373127129575} {"canonical_length": 62, "canonical_nll": 1.7849477952526462, "generation_rouge_l_f1": 0.27368421052631575, "generation_token_f1": 0.33684210526315794, "prompt_id": "0221b725811d035d", "reading_canonical_log_prob": -110.66676330566406, "reading_distances": [14, 20, 17, 12], "reading_log_probs": [-106.84304809570312, -109.74130249023438, -106.61796569824219, -111.8822021484375], "reading_logprob_std": 2.5148638739589697, "tokenization_path_logprob_std": 112.6018980100328, "trajectory_balance_rmse": 1322.648543272384} {"canonical_length": 109, "canonical_nll": 0.8112819566639191, "generation_rouge_l_f1": 0.2604166666666667, "generation_token_f1": 0.3854166666666667, "prompt_id": "7ac2b23c73ad8de9", "reading_canonical_log_prob": -88.42973327636719, "reading_distances": [39, 38, 40, 33], "reading_log_probs": [-92.89500427246094, -92.34259796142578, -90.78864288330078, -92.59008026123047], "reading_logprob_std": 0.9379094744426001, "tokenization_path_logprob_std": 125.15618153632518, "trajectory_balance_rmse": 2026.5769686276712} {"canonical_length": 98, "canonical_nll": 1.3950772966657365, "generation_rouge_l_f1": 0.2472527472527473, "generation_token_f1": 0.3846153846153846, "prompt_id": "a2a34e9410c7790e", "reading_canonical_log_prob": -136.7175750732422, "reading_distances": [21, 22, 25, 22], "reading_log_probs": [-132.93984985351562, -133.22988891601562, -130.51223754882812, -134.07424926757812], "reading_logprob_std": 1.5289099338976375, "tokenization_path_logprob_std": 121.0571709551797, "trajectory_balance_rmse": 2101.5650062544746} {"canonical_length": 139, "canonical_nll": 1.1581360521934014, "generation_rouge_l_f1": 0.24535315985130113, "generation_token_f1": 0.37918215613382905, "prompt_id": "1c2356a54b0c56f0", "reading_canonical_log_prob": -160.9809112548828, "reading_distances": [31, 27, 25, 29], "reading_log_probs": [-154.7257843017578, -154.8245086669922, -154.89974975585938, -153.48439025878906], "reading_logprob_std": 0.6699434506506201, "tokenization_path_logprob_std": 174.00685253765408, "trajectory_balance_rmse": 3361.4246330657866} {"canonical_length": 150, "canonical_nll": 1.5090629069010417, "generation_rouge_l_f1": 0.24509803921568632, "generation_token_f1": 0.49019607843137264, "prompt_id": "55d3c9fe027b6cf4", "reading_canonical_log_prob": -226.35943603515625, "reading_distances": [20, 18, 19, 24], "reading_log_probs": [-226.62193298339844, -224.9960174560547, -226.9239959716797, -226.6459197998047], "reading_logprob_std": 0.8780671944815398, "tokenization_path_logprob_std": 57.175104897203326, "trajectory_balance_rmse": 3119.185932547275} {"canonical_length": 22, "canonical_nll": 2.8044492548162285, "generation_rouge_l_f1": 0.25, "generation_token_f1": 0.25, "prompt_id": "99817bb6bf526c93", "reading_canonical_log_prob": -61.69788360595703, "reading_distances": [17, 20, 20, 23], "reading_log_probs": [-61.91029739379883, -60.73130798339844, -61.02944564819336, -60.77959060668945], "reading_logprob_std": 0.5475748416604663, "tokenization_path_logprob_std": 43.22008914786923, "trajectory_balance_rmse": 434.2860602936078} {"canonical_length": 177, "canonical_nll": 1.5013872566869704, "generation_rouge_l_f1": 0.26605504587155965, "generation_token_f1": 0.5137614678899083, "prompt_id": "a99bf80957c76958", "reading_canonical_log_prob": -265.74554443359375, "reading_distances": [22, 28, 30, 26], "reading_log_probs": [-272.054931640625, -270.9697265625, -271.9958190917969, -269.69482421875], "reading_logprob_std": 1.1077037310119229, "tokenization_path_logprob_std": 257.73282557234995, "trajectory_balance_rmse": 3795.864530539381} {"canonical_length": 333, "canonical_nll": 1.65960693359375, "generation_rouge_l_f1": 0.2536585365853659, "generation_token_f1": 0.4943089430894309, "prompt_id": "6b033b3db0a5c2fe", "reading_canonical_log_prob": -552.6491088867188, "reading_distances": [23, 27, 26, 29], "reading_log_probs": [-549.6344604492188, -551.5863037109375, -553.2952270507812, -554.2562866210938], "reading_logprob_std": 2.0319334924348014, "tokenization_path_logprob_std": 96.19646302209041, "trajectory_balance_rmse": 7339.031646932113} {"canonical_length": 190, "canonical_nll": 1.1519980982730262, "generation_rouge_l_f1": 0.33035714285714285, "generation_token_f1": 0.48660714285714285, "prompt_id": "f5608ead26273538", "reading_canonical_log_prob": -218.879638671875, "reading_distances": [34, 36, 38, 46], "reading_log_probs": [-228.6458740234375, -225.0758056640625, -227.14437866210938, -228.8513946533203], "reading_logprob_std": 1.7438005715082774, "tokenization_path_logprob_std": 113.39387791192526, "trajectory_balance_rmse": 3436.0875990305003} {"canonical_length": 305, "canonical_nll": 1.2609765224769467, "generation_rouge_l_f1": 0.2281639928698752, "generation_token_f1": 0.42424242424242425, "prompt_id": "eb21c6c6320f2679", "reading_canonical_log_prob": -384.59783935546875, "reading_distances": [32, 25, 34, 31], "reading_log_probs": [-382.0513000488281, -375.6722717285156, -384.67156982421875, -380.5930480957031], "reading_logprob_std": 3.7806531680731483, "tokenization_path_logprob_std": 293.56988572040416, "trajectory_balance_rmse": 6402.103757004033} {"canonical_length": 222, "canonical_nll": 0.875575709987331, "generation_rouge_l_f1": 0.3178807947019867, "generation_token_f1": 0.609271523178808, "prompt_id": "bd9aa33662e77c04", "reading_canonical_log_prob": -194.3778076171875, "reading_distances": [18, 17, 19, 19], "reading_log_probs": [-198.94326782226562, -194.76126098632812, -190.58251953125, -190.4950714111328], "reading_logprob_std": 4.025266208341916, "tokenization_path_logprob_std": 113.51361489972084, "trajectory_balance_rmse": 2093.970031059511} {"canonical_length": 85, "canonical_nll": 0.7934184354894301, "generation_rouge_l_f1": 0.20245398773006135, "generation_token_f1": 0.3190184049079755, "prompt_id": "4fcfacdaa4123cbd", "reading_canonical_log_prob": -67.44056701660156, "reading_distances": [28, 26, 23, 26], "reading_log_probs": [-65.3846664428711, -62.79735565185547, -61.99938201904297, -63.888999938964844], "reading_logprob_std": 1.4660164591492697, "tokenization_path_logprob_std": 149.6502344370103, "trajectory_balance_rmse": 1165.1796418587337} {"canonical_length": 214, "canonical_nll": 1.6930787273656542, "generation_rouge_l_f1": 0.2360801781737194, "generation_token_f1": 0.4721603563474388, "prompt_id": "42e23f667d7b63a7", "reading_canonical_log_prob": -362.31884765625, "reading_distances": [17, 18, 22, 22], "reading_log_probs": [-352.96514892578125, -349.80645751953125, -358.98138427734375, -351.92083740234375], "reading_logprob_std": 3.934507712084518, "tokenization_path_logprob_std": 208.48110773064712, "trajectory_balance_rmse": 4317.144830800286} {"canonical_length": 13, "canonical_nll": 1.3860169924222505, "generation_rouge_l_f1": 0.06271777003484322, "generation_token_f1": 0.07665505226480836, "prompt_id": "f4ba682f79ee92c5", "reading_canonical_log_prob": -18.018220901489258, "reading_distances": [12, 12, 19, 18], "reading_log_probs": [-17.954132080078125, -17.348731994628906, -18.88076400756836, -17.17964744567871], "reading_logprob_std": 0.7689002791827236, "tokenization_path_logprob_std": 55.74333520679351, "trajectory_balance_rmse": 297.9407438281405} {"canonical_length": 112, "canonical_nll": 1.0542110715593611, "generation_rouge_l_f1": 0.35135135135135137, "generation_token_f1": 0.5405405405405405, "prompt_id": "8c5355410cf1d176", "reading_canonical_log_prob": -118.07164001464844, "reading_distances": [22, 20, 30, 24], "reading_log_probs": [-117.37846374511719, -117.7855453491211, -117.69512176513672, -119.54061126708984], "reading_logprob_std": 0.9761802234350668, "tokenization_path_logprob_std": 76.7660941436965, "trajectory_balance_rmse": 2141.2803748960446} {"canonical_length": 255, "canonical_nll": 1.2458259133731617, "generation_rouge_l_f1": 0.2633587786259542, "generation_token_f1": 0.47709923664122134, "prompt_id": "131fb5156c48d9ba", "reading_canonical_log_prob": -317.68560791015625, "reading_distances": [22, 25, 26, 26], "reading_log_probs": [-319.98284912109375, -316.4265441894531, -318.7041015625, -317.1789245605469], "reading_logprob_std": 1.5870308158180422, "tokenization_path_logprob_std": 221.00394204919058, "trajectory_balance_rmse": 5563.2085714590075} {"canonical_length": 17, "canonical_nll": 1.197336196899414, "generation_rouge_l_f1": 0.29629629629629634, "generation_token_f1": 0.29629629629629634, "prompt_id": "20a85de31d07820a", "reading_canonical_log_prob": -20.35471534729004, "reading_distances": [17, 18, 17, 23], "reading_log_probs": [-19.704872131347656, -19.5040225982666, -19.503276824951172, -19.127347946166992], "reading_logprob_std": 0.24112967786799738, "tokenization_path_logprob_std": 70.1639517798013, "trajectory_balance_rmse": 265.5766816767906} {"canonical_length": 236, "canonical_nll": 1.175004215563758, "generation_rouge_l_f1": 0.242798353909465, "generation_token_f1": 0.5185185185185185, "prompt_id": "8bdd0bdfad8b9b1b", "reading_canonical_log_prob": -277.3009948730469, "reading_distances": [20, 18, 17, 16], "reading_log_probs": [-276.5860595703125, -276.074951171875, -276.19476318359375, -276.7342529296875], "reading_logprob_std": 0.313101267047685, "tokenization_path_logprob_std": 157.7816298908932, "trajectory_balance_rmse": 5115.315014249864} {"canonical_length": 314, "canonical_nll": 1.4190409472034236, "generation_rouge_l_f1": 0.2482758620689655, "generation_token_f1": 0.5275862068965517, "prompt_id": "f543bc422c7c86fd", "reading_canonical_log_prob": -445.578857421875, "reading_distances": [23, 23, 27, 26], "reading_log_probs": [-437.085693359375, -438.34503173828125, -439.3045654296875, -439.364990234375], "reading_logprob_std": 1.0672859682821234, "tokenization_path_logprob_std": 236.9686291227452, "trajectory_balance_rmse": 7510.613786703741} {"canonical_length": 17, "canonical_nll": 0.8983244615442613, "generation_rouge_l_f1": 0.5294117647058824, "generation_token_f1": 0.6470588235294117, "prompt_id": "6b9813758e34b6e5", "reading_canonical_log_prob": -15.271515846252441, "reading_distances": [28, 34, 32, 32], "reading_log_probs": [-12.93232536315918, -13.704547882080078, -13.202130317687988, -13.344274520874023], "reading_logprob_std": 0.3216169718602368, "tokenization_path_logprob_std": 57.85598261123578, "trajectory_balance_rmse": 466.4206223674405} {"canonical_length": 133, "canonical_nll": 1.4048279496960174, "generation_rouge_l_f1": 0.18090452261306533, "generation_token_f1": 0.30150753768844224, "prompt_id": "89680ea889169bf7", "reading_canonical_log_prob": -186.8421173095703, "reading_distances": [24, 22, 25, 25], "reading_log_probs": [-176.70391845703125, -176.93911743164062, -177.01483154296875, -177.33334350585938], "reading_logprob_std": 0.2599291716842209, "tokenization_path_logprob_std": 88.7862813054994, "trajectory_balance_rmse": 2887.4153820307265} {"canonical_length": 15, "canonical_nll": 0.871837043762207, "generation_rouge_l_f1": 0.5714285714285715, "generation_token_f1": 0.5714285714285715, "prompt_id": "5d244886eb8450ed", "reading_canonical_log_prob": -13.077555656433105, "reading_distances": [33, 28, 32, 31], "reading_log_probs": [-13.459224700927734, -12.464946746826172, -12.843618392944336, -13.424696922302246], "reading_logprob_std": 0.480530680265674, "tokenization_path_logprob_std": 77.66795758824131, "trajectory_balance_rmse": 454.3795375422549} {"canonical_length": 245, "canonical_nll": 0.9561735192123725, "generation_rouge_l_f1": 0.2674418604651163, "generation_token_f1": 0.5232558139534883, "prompt_id": "8dfd747741dad44f", "reading_canonical_log_prob": -234.26251220703125, "reading_distances": [21, 20, 17, 20], "reading_log_probs": [-231.2464599609375, -230.86270141601562, -231.07870483398438, -232.44125366210938], "reading_logprob_std": 0.7069870561654705, "tokenization_path_logprob_std": 153.10158582168341, "trajectory_balance_rmse": 4600.56007163068} {"canonical_length": 18, "canonical_nll": 0.5880859692891439, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "73b880653e3d865c", "reading_canonical_log_prob": -10.58554744720459, "reading_distances": [18, 22, 27, 17], "reading_log_probs": [-12.200920104980469, -12.86375904083252, -11.617486953735352, -12.270580291748047], "reading_logprob_std": 0.5095903820777116, "tokenization_path_logprob_std": 26.175741978618806, "trajectory_balance_rmse": 311.44181672665553} {"canonical_length": 18, "canonical_nll": 1.936169942220052, "generation_rouge_l_f1": 0.3448275862068966, "generation_token_f1": 0.3448275862068966, "prompt_id": "50d05075f2282d29", "reading_canonical_log_prob": -34.85105895996094, "reading_distances": [22, 24, 29, 30], "reading_log_probs": [-33.09531784057617, -34.574153900146484, -33.96306610107422, -38.6429443359375], "reading_logprob_std": 2.458757258393359, "tokenization_path_logprob_std": 35.32843334296667, "trajectory_balance_rmse": 421.8225105572587} {"canonical_length": 12, "canonical_nll": 1.025974988937378, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "4907b79891256514", "reading_canonical_log_prob": -12.311699867248535, "reading_distances": [36, 32, 29, 40], "reading_log_probs": [-12.367198944091797, -12.242463111877441, -12.378866195678711, -11.68189811706543], "reading_logprob_std": 0.329638159376535, "tokenization_path_logprob_std": 39.05295646647871, "trajectory_balance_rmse": 245.711766875542} {"canonical_length": 58, "canonical_nll": 2.5647972370016165, "generation_rouge_l_f1": 0.1683168316831683, "generation_token_f1": 0.20792079207920794, "prompt_id": "d9bc5b8ac130964c", "reading_canonical_log_prob": -148.75823974609375, "reading_distances": [13, 12, 10, 14], "reading_log_probs": [-145.24368286132812, -148.59793090820312, -148.3797607421875, -146.78265380859375], "reading_logprob_std": 1.5638620926610551, "tokenization_path_logprob_std": 47.772117404199946, "trajectory_balance_rmse": 1287.008187424657} {"canonical_length": 14, "canonical_nll": 1.0357497760227747, "generation_rouge_l_f1": 0.15942028985507245, "generation_token_f1": 0.15942028985507245, "prompt_id": "0946ba5ceddc203c", "reading_canonical_log_prob": -14.500496864318848, "reading_distances": [17, 14, 12, 15], "reading_log_probs": [-11.53205680847168, -13.164165496826172, -11.667993545532227, -11.871116638183594], "reading_logprob_std": 0.7499435525787381, "tokenization_path_logprob_std": 13.854002340517214, "trajectory_balance_rmse": 198.96867389052716} {"canonical_length": 17, "canonical_nll": 1.161053264842314, "generation_rouge_l_f1": 0.13559322033898305, "generation_token_f1": 0.1440677966101695, "prompt_id": "d08a5a0e92cb2564", "reading_canonical_log_prob": -19.737905502319336, "reading_distances": [27, 16, 21, 20], "reading_log_probs": [-22.5662841796875, -23.604970932006836, -22.136241912841797, -23.18901824951172], "reading_logprob_std": 0.6512887111528597, "tokenization_path_logprob_std": 58.57152065047575, "trajectory_balance_rmse": 293.07332262992804} {"canonical_length": 65, "canonical_nll": 0.5878412686861478, "generation_rouge_l_f1": 0.5590062111801243, "generation_token_f1": 0.5838509316770186, "prompt_id": "02aff4ae4d6fb42e", "reading_canonical_log_prob": -38.20968246459961, "reading_distances": [18, 26, 25, 20], "reading_log_probs": [-36.92510986328125, -36.88886260986328, -37.38609313964844, -36.84008026123047], "reading_logprob_std": 0.2531135216956882, "tokenization_path_logprob_std": 59.15499693269193, "trajectory_balance_rmse": 1042.5496940365463} {"canonical_length": 281, "canonical_nll": 1.1536523133410252, "generation_rouge_l_f1": 0.22684310018903592, "generation_token_f1": 0.4158790170132325, "prompt_id": "a79f040945146e56", "reading_canonical_log_prob": -324.1763000488281, "reading_distances": [35, 35, 41, 35], "reading_log_probs": [-320.96063232421875, -322.1802978515625, -322.2328796386719, -322.7418518066406], "reading_logprob_std": 0.755871485375736, "tokenization_path_logprob_std": 276.4440117123031, "trajectory_balance_rmse": 5767.738985292434} {"canonical_length": 341, "canonical_nll": 1.3864380061801227, "generation_rouge_l_f1": 0.20244328097731237, "generation_token_f1": 0.40488656195462475, "prompt_id": "79e68a828bcf94f4", "reading_canonical_log_prob": -472.7753601074219, "reading_distances": [30, 25, 29, 23], "reading_log_probs": [-465.7816162109375, -466.0842590332031, -465.7205810546875, -465.94219970703125], "reading_logprob_std": 0.16397722877006798, "tokenization_path_logprob_std": 217.97638293411427, "trajectory_balance_rmse": 6909.386045680084} {"canonical_length": 50, "canonical_nll": 0.5877598571777344, "generation_rouge_l_f1": 0.5263157894736842, "generation_token_f1": 0.631578947368421, "prompt_id": "5debb498a4c42600", "reading_canonical_log_prob": -29.38799285888672, "reading_distances": [33, 32, 30, 30], "reading_log_probs": [-28.666526794433594, -29.20290184020996, -29.174230575561523, -28.981178283691406], "reading_logprob_std": 0.246934445268955, "tokenization_path_logprob_std": 68.16980786785528, "trajectory_balance_rmse": 916.0009632151397} {"canonical_length": 34, "canonical_nll": 1.3747989430147058, "generation_rouge_l_f1": 0.3571428571428571, "generation_token_f1": 0.3571428571428571, "prompt_id": "fa77d10032456ab0", "reading_canonical_log_prob": -46.7431640625, "reading_distances": [27, 28, 32, 27], "reading_log_probs": [-48.21878433227539, -48.382659912109375, -50.030818939208984, -47.57792282104492], "reading_logprob_std": 1.0448986166855552, "tokenization_path_logprob_std": 103.23953917561832, "trajectory_balance_rmse": 955.5715777943959} {"canonical_length": 8, "canonical_nll": 0.04337511956691742, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "253388f751e77ae9", "reading_canonical_log_prob": -0.34700095653533936, "reading_distances": [48, 51, 45, 55], "reading_log_probs": [-2.924142837524414, -0.6096610426902771, -1.3818445205688477, -2.9327168464660645], "reading_logprob_std": 1.1595129779685813, "tokenization_path_logprob_std": 32.2254365624299, "trajectory_balance_rmse": 141.30696018322308} {"canonical_length": 51, "canonical_nll": 1.1987318899117263, "generation_rouge_l_f1": 0.3164556962025316, "generation_token_f1": 0.46835443037974683, "prompt_id": "0182330e9a378205", "reading_canonical_log_prob": -61.13532638549805, "reading_distances": [25, 22, 24, 26], "reading_log_probs": [-62.01488494873047, -63.647857666015625, -61.304222106933594, -64.89468383789062], "reading_logprob_std": 1.6177016049076571, "tokenization_path_logprob_std": 89.79932029592372, "trajectory_balance_rmse": 776.9291483597112} {"canonical_length": 111, "canonical_nll": 1.7295335649370074, "generation_rouge_l_f1": 0.23343848580441642, "generation_token_f1": 0.41009463722397477, "prompt_id": "8709e23cb34a4482", "reading_canonical_log_prob": -191.9782257080078, "reading_distances": [22, 24, 26, 28], "reading_log_probs": [-184.40313720703125, -182.72341918945312, -183.4020233154297, -183.1587371826172], "reading_logprob_std": 0.711887544998014, "tokenization_path_logprob_std": 115.88276667030252, "trajectory_balance_rmse": 2741.546139027168} {"canonical_length": 6, "canonical_nll": 0.7521193027496338, "generation_rouge_l_f1": 0.888888888888889, "generation_token_f1": 0.888888888888889, "prompt_id": "0852cc501290063e", "reading_canonical_log_prob": -4.512715816497803, "reading_distances": [25, 34, 28, 28], "reading_log_probs": [-4.544562339782715, -4.120385646820068, -3.900848150253296, -4.683294773101807], "reading_logprob_std": 0.3641001038137263, "tokenization_path_logprob_std": 17.842458239765477, "trajectory_balance_rmse": 73.98202431915159} {"canonical_length": 12, "canonical_nll": 1.2700586318969727, "generation_rouge_l_f1": 0.7058823529411765, "generation_token_f1": 0.7058823529411765, "prompt_id": "0108b707d4dec974", "reading_canonical_log_prob": -15.240703582763672, "reading_distances": [35, 40, 31, 43], "reading_log_probs": [-18.147451400756836, -21.60265350341797, -16.56951332092285, -26.678974151611328], "reading_logprob_std": 4.476958792058861, "tokenization_path_logprob_std": 78.16548289611232, "trajectory_balance_rmse": 314.23594316808754} {"canonical_length": 104, "canonical_nll": 1.6448402404785156, "generation_rouge_l_f1": 0.20052770448548812, "generation_token_f1": 0.3641160949868074, "prompt_id": "5ffba76c8fed8b0d", "reading_canonical_log_prob": -171.06338500976562, "reading_distances": [18, 19, 19, 19], "reading_log_probs": [-167.92422485351562, -166.11007690429688, -168.0238494873047, -166.8813018798828], "reading_logprob_std": 0.9106533678220949, "tokenization_path_logprob_std": 23.800120420958535, "trajectory_balance_rmse": 2354.750686391915} {"canonical_length": 245, "canonical_nll": 1.1060019979671556, "generation_rouge_l_f1": 0.31451612903225806, "generation_token_f1": 0.5685483870967741, "prompt_id": "66d201ed6635605c", "reading_canonical_log_prob": -270.9704895019531, "reading_distances": [43, 40, 50, 36], "reading_log_probs": [-269.941650390625, -272.211181640625, -270.20172119140625, -271.34124755859375], "reading_logprob_std": 1.0516066317109432, "tokenization_path_logprob_std": 233.496977941056, "trajectory_balance_rmse": 4807.048326233718} {"canonical_length": 62, "canonical_nll": 1.0541427366195186, "generation_rouge_l_f1": 0.29906542056074764, "generation_token_f1": 0.3738317757009346, "prompt_id": "d9f443de104a8944", "reading_canonical_log_prob": -65.35684967041016, "reading_distances": [22, 17, 18, 17], "reading_log_probs": [-67.5877685546875, -68.01434326171875, -67.49201965332031, -66.07107543945312], "reading_logprob_std": 0.8445751436105395, "tokenization_path_logprob_std": 20.8221594245079, "trajectory_balance_rmse": 587.9558733799751} {"canonical_length": 19, "canonical_nll": 1.6612739562988281, "generation_rouge_l_f1": 0.3478260869565218, "generation_token_f1": 0.3478260869565218, "prompt_id": "d416f7f8cc539e7b", "reading_canonical_log_prob": -31.564205169677734, "reading_distances": [23, 30, 33, 28], "reading_log_probs": [-28.40756607055664, -29.481672286987305, -29.096637725830078, -29.32038116455078], "reading_logprob_std": 0.47311757452849745, "tokenization_path_logprob_std": 103.81697676752148, "trajectory_balance_rmse": 524.9893811796144} {"canonical_length": 12, "canonical_nll": 1.5938334465026855, "generation_rouge_l_f1": 0.3846153846153846, "generation_token_f1": 0.4615384615384615, "prompt_id": "111f88e5ddbba7a9", "reading_canonical_log_prob": -19.126001358032227, "reading_distances": [24, 22, 18, 19], "reading_log_probs": [-32.804664611816406, -27.662092208862305, -31.9047908782959, -27.077184677124023], "reading_logprob_std": 2.911303266467042, "tokenization_path_logprob_std": 33.225560151419835, "trajectory_balance_rmse": 271.1799160258753} {"canonical_length": 11, "canonical_nll": 1.1014806574041194, "generation_rouge_l_f1": 0.09523809523809525, "generation_token_f1": 0.09523809523809525, "prompt_id": "47566df03369ba07", "reading_canonical_log_prob": -12.116287231445312, "reading_distances": [50, 40, 54, 48], "reading_log_probs": [-10.831623077392578, -15.38536262512207, -10.170693397521973, -14.637323379516602], "reading_logprob_std": 2.6356505832237938, "tokenization_path_logprob_std": 32.85553488260934, "trajectory_balance_rmse": 200.72989511707837} {"canonical_length": 26, "canonical_nll": 0.7902049284714919, "generation_rouge_l_f1": 0.21212121212121213, "generation_token_f1": 0.21212121212121213, "prompt_id": "0ad3db1386b28ee8", "reading_canonical_log_prob": -20.54532814025879, "reading_distances": [30, 31, 33, 37], "reading_log_probs": [-22.101036071777344, -22.17828369140625, -22.213956832885742, -21.682147979736328], "reading_logprob_std": 0.2457012753302143, "tokenization_path_logprob_std": 42.64525848152781, "trajectory_balance_rmse": 599.752822982912} {"canonical_length": 104, "canonical_nll": 1.7408420856182392, "generation_rouge_l_f1": 0.18028169014084508, "generation_token_f1": 0.37183098591549296, "prompt_id": "136b8765ce15fa6a", "reading_canonical_log_prob": -181.04757690429688, "reading_distances": [18, 21, 18, 21], "reading_log_probs": [-181.8020782470703, -180.3323211669922, -187.486572265625, -182.45285034179688], "reading_logprob_std": 3.1079933452660033, "tokenization_path_logprob_std": 83.42337173697173, "trajectory_balance_rmse": 2070.582800458443} {"canonical_length": 35, "canonical_nll": 0.7973073141915458, "generation_rouge_l_f1": 0.4507042253521127, "generation_token_f1": 0.5915492957746479, "prompt_id": "92e8091e1c9b04af", "reading_canonical_log_prob": -27.9057559967041, "reading_distances": [28, 31, 24, 38], "reading_log_probs": [-28.282855987548828, -26.593303680419922, -26.1561279296875, -25.671682357788086], "reading_logprob_std": 1.1354508270164534, "tokenization_path_logprob_std": 113.32920810332502, "trajectory_balance_rmse": 703.4876739751453} {"canonical_length": 20, "canonical_nll": 2.581010437011719, "generation_rouge_l_f1": 0.16, "generation_token_f1": 0.16, "prompt_id": "3bfad973aaa202e7", "reading_canonical_log_prob": -51.620208740234375, "reading_distances": [21, 15, 21, 17], "reading_log_probs": [-44.59266662597656, -44.36795425415039, -44.844642639160156, -48.728607177734375], "reading_logprob_std": 2.0725929428739978, "tokenization_path_logprob_std": 115.80617149782712, "trajectory_balance_rmse": 531.1526960021604} {"canonical_length": 292, "canonical_nll": 0.9943368990127355, "generation_rouge_l_f1": 0.22823984526112184, "generation_token_f1": 0.4951644100580271, "prompt_id": "be5dbfe122219a1b", "reading_canonical_log_prob": -290.34637451171875, "reading_distances": [18, 28, 24, 21], "reading_log_probs": [-289.7808837890625, -292.3546142578125, -292.98040771484375, -289.947265625], "reading_logprob_std": 1.6400106864666497, "tokenization_path_logprob_std": 168.1578004521335, "trajectory_balance_rmse": 5808.87026027587} {"canonical_length": 39, "canonical_nll": 1.2631190373347356, "generation_rouge_l_f1": 0.1413427561837456, "generation_token_f1": 0.18374558303886923, "prompt_id": "6acd370678d1b813", "reading_canonical_log_prob": -49.26164245605469, "reading_distances": [16, 16, 14, 17], "reading_log_probs": [-44.266082763671875, -46.57601547241211, -46.755043029785156, -47.599151611328125], "reading_logprob_std": 1.4268683071358486, "tokenization_path_logprob_std": 35.30358142080507, "trajectory_balance_rmse": 372.4660255537445} {"canonical_length": 197, "canonical_nll": 1.2336994306690197, "generation_rouge_l_f1": 0.2744063324538259, "generation_token_f1": 0.5118733509234829, "prompt_id": "2bb939d56ab15ee3", "reading_canonical_log_prob": -243.03878784179688, "reading_distances": [22, 20, 21, 20], "reading_log_probs": [-239.34320068359375, -240.00357055664062, -239.2406463623047, -239.72677612304688], "reading_logprob_std": 0.3522222907063976, "tokenization_path_logprob_std": 390.11590511024286, "trajectory_balance_rmse": 4312.419454693384} {"canonical_length": 14, "canonical_nll": 1.4846816744123186, "generation_rouge_l_f1": 0.4666666666666667, "generation_token_f1": 0.4666666666666667, "prompt_id": "9d34cf3c82e91821", "reading_canonical_log_prob": -20.78554344177246, "reading_distances": [48, 30, 40, 42], "reading_log_probs": [-20.923173904418945, -19.49335479736328, -19.120502471923828, -19.77765464782715], "reading_logprob_std": 0.7777057918171847, "tokenization_path_logprob_std": 54.3632026452054, "trajectory_balance_rmse": 210.6594330869183} {"canonical_length": 13, "canonical_nll": 0.07192703393789437, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "b8a06c7365e05ce4", "reading_canonical_log_prob": -0.935051441192627, "reading_distances": [30, 24, 29, 22], "reading_log_probs": [-2.251075506210327, -3.037006378173828, -1.7107049226760864, -2.8946146965026855], "reading_logprob_std": 0.6127005581116074, "tokenization_path_logprob_std": 33.70263288380636, "trajectory_balance_rmse": 285.0020837075593} {"canonical_length": 44, "canonical_nll": 1.5686288313432173, "generation_rouge_l_f1": 0.2302158273381295, "generation_token_f1": 0.28776978417266186, "prompt_id": "3c94fb62a787deaa", "reading_canonical_log_prob": -69.01966857910156, "reading_distances": [26, 19, 21, 24], "reading_log_probs": [-73.23451232910156, -73.10386657714844, -72.87300109863281, -72.43772888183594], "reading_logprob_std": 0.3498955493068233, "tokenization_path_logprob_std": 71.56278732735922, "trajectory_balance_rmse": 775.9896970047693} {"canonical_length": 230, "canonical_nll": 1.0723983764648437, "generation_rouge_l_f1": 0.31752577319587627, "generation_token_f1": 0.5938144329896906, "prompt_id": "84f970de38e4a1fd", "reading_canonical_log_prob": -246.65162658691406, "reading_distances": [25, 25, 21, 29], "reading_log_probs": [-241.63037109375, -239.64991760253906, -240.93130493164062, -240.2327423095703], "reading_logprob_std": 0.8579959018688726, "tokenization_path_logprob_std": 189.32617070993697, "trajectory_balance_rmse": 3769.028972526568} {"canonical_length": 43, "canonical_nll": 1.5695630450581395, "generation_rouge_l_f1": 0.09271523178807947, "generation_token_f1": 0.16556291390728478, "prompt_id": "b9d9b86bb406a853", "reading_canonical_log_prob": -67.4912109375, "reading_distances": [16, 18, 17, 12], "reading_log_probs": [-71.96843719482422, -72.86617279052734, -72.77987670898438, -71.67040252685547], "reading_logprob_std": 0.5931156749338151, "tokenization_path_logprob_std": 53.57349392467189, "trajectory_balance_rmse": 906.1809811934902} {"canonical_length": 217, "canonical_nll": 1.031647431685628, "generation_rouge_l_f1": 0.3493449781659389, "generation_token_f1": 0.6026200873362445, "prompt_id": "2dfd04cd2fcb5555", "reading_canonical_log_prob": -223.86749267578125, "reading_distances": [42, 36, 40, 40], "reading_log_probs": [-219.4297637939453, -215.65106201171875, -211.1903076171875, -212.43177795410156], "reading_logprob_std": 3.6849061416333235, "tokenization_path_logprob_std": 177.46397332437044, "trajectory_balance_rmse": 4858.847092635302} {"canonical_length": 193, "canonical_nll": 0.9778190968567844, "generation_rouge_l_f1": 0.3090909090909091, "generation_token_f1": 0.49090909090909096, "prompt_id": "daff264becf42549", "reading_canonical_log_prob": -188.71908569335938, "reading_distances": [16, 11, 14, 14], "reading_log_probs": [-186.19866943359375, -187.0741729736328, -196.74696350097656, -186.18914794921875], "reading_logprob_std": 5.14657426469231, "tokenization_path_logprob_std": 254.031542258013, "trajectory_balance_rmse": 3962.6377002365202} {"canonical_length": 254, "canonical_nll": 1.2200004998154528, "generation_rouge_l_f1": 0.22102425876010784, "generation_token_f1": 0.3827493261455525, "prompt_id": "f80fc3edc248f324", "reading_canonical_log_prob": -309.880126953125, "reading_distances": [30, 28, 30, 27], "reading_log_probs": [-310.5760192871094, -312.0292053222656, -312.7411804199219, -310.4408874511719], "reading_logprob_std": 1.1232001505481999, "tokenization_path_logprob_std": 182.42049798782006, "trajectory_balance_rmse": 5547.1245595756045} {"canonical_length": 72, "canonical_nll": 1.8862090640597873, "generation_rouge_l_f1": 0.22360248447204967, "generation_token_f1": 0.2732919254658385, "prompt_id": "7da5b774aaa5c557", "reading_canonical_log_prob": -135.8070526123047, "reading_distances": [25, 18, 24, 29], "reading_log_probs": [-133.30862426757812, -132.267822265625, -132.7225341796875, -133.8330841064453], "reading_logprob_std": 0.6826393152022907, "tokenization_path_logprob_std": 63.77924168247079, "trajectory_balance_rmse": 1614.4439259291323} {"canonical_length": 138, "canonical_nll": 1.560023874476336, "generation_rouge_l_f1": 0.23979591836734698, "generation_token_f1": 0.413265306122449, "prompt_id": "0e6878e78767cf2b", "reading_canonical_log_prob": -215.28329467773438, "reading_distances": [16, 20, 24, 22], "reading_log_probs": [-216.40509033203125, -214.01039123535156, -216.83123779296875, -218.5550537109375], "reading_logprob_std": 1.873526122725142, "tokenization_path_logprob_std": 80.81322581526912, "trajectory_balance_rmse": 2592.7846294084256} {"canonical_length": 152, "canonical_nll": 1.2184392025596218, "generation_rouge_l_f1": 0.21377672209026127, "generation_token_f1": 0.4228028503562945, "prompt_id": "0ac01e4320c0e1b4", "reading_canonical_log_prob": -185.2027587890625, "reading_distances": [29, 28, 26, 29], "reading_log_probs": [-174.96328735351562, -178.4933624267578, -174.68667602539062, -173.89341735839844], "reading_logprob_std": 2.0404693283462345, "tokenization_path_logprob_std": 82.86752822506588, "trajectory_balance_rmse": 3092.6455094440657} {"canonical_length": 259, "canonical_nll": 1.2362568387653836, "generation_rouge_l_f1": 0.20784313725490194, "generation_token_f1": 0.48627450980392156, "prompt_id": "cc034e8024f676e6", "reading_canonical_log_prob": -320.1905212402344, "reading_distances": [23, 15, 30, 24], "reading_log_probs": [-316.6720275878906, -316.3326416015625, -317.3768310546875, -317.8157043457031], "reading_logprob_std": 0.6709659568815454, "tokenization_path_logprob_std": 253.19157028175272, "trajectory_balance_rmse": 5212.972836644682} {"canonical_length": 253, "canonical_nll": 1.2870811492558054, "generation_rouge_l_f1": 0.26294820717131473, "generation_token_f1": 0.4701195219123506, "prompt_id": "e0ae35419a54967e", "reading_canonical_log_prob": -325.63153076171875, "reading_distances": [23, 23, 27, 30], "reading_log_probs": [-313.57177734375, -315.3541259765625, -322.30657958984375, -319.07135009765625], "reading_logprob_std": 3.898075526046469, "tokenization_path_logprob_std": 177.85992998329647, "trajectory_balance_rmse": 4805.591469168311} {"canonical_length": 6, "canonical_nll": 0.6407334009806315, "generation_rouge_l_f1": 0.8333333333333333, "generation_token_f1": 0.8333333333333333, "prompt_id": "9ad445292250739f", "reading_canonical_log_prob": -3.844400405883789, "reading_distances": [27, 19, 18, 27], "reading_log_probs": [-5.041286945343018, -4.806690216064453, -3.7009239196777344, -5.146244525909424], "reading_logprob_std": 0.6639272453887662, "tokenization_path_logprob_std": 28.784730972191944, "trajectory_balance_rmse": 140.73187285969624} {"canonical_length": 21, "canonical_nll": 1.236229487827846, "generation_rouge_l_f1": 0.26666666666666666, "generation_token_f1": 0.3111111111111111, "prompt_id": "1a95335c2e86bbe5", "reading_canonical_log_prob": -25.960819244384766, "reading_distances": [17, 23, 23, 26], "reading_log_probs": [-27.485706329345703, -28.368183135986328, -28.59939193725586, -26.685998916625977], "reading_logprob_std": 0.8757314562946967, "tokenization_path_logprob_std": 11.59687846205494, "trajectory_balance_rmse": 435.74579985452266} {"canonical_length": 25, "canonical_nll": 0.24293712615966798, "generation_rouge_l_f1": 0.17674418604651163, "generation_token_f1": 0.17674418604651163, "prompt_id": "8fe26f83cb600778", "reading_canonical_log_prob": -6.073428153991699, "reading_distances": [22, 23, 25, 23], "reading_log_probs": [-5.581175804138184, -5.262277603149414, -5.955746650695801, -5.463171005249023], "reading_logprob_std": 0.2915208339674588, "tokenization_path_logprob_std": 36.80058992537419, "trajectory_balance_rmse": 189.03657486698282} {"canonical_length": 17, "canonical_nll": 0.9377504236557904, "generation_rouge_l_f1": 0.05309734513274336, "generation_token_f1": 0.05309734513274336, "prompt_id": "8bd44913279a5599", "reading_canonical_log_prob": -15.941757202148438, "reading_distances": [14, 12, 11, 13], "reading_log_probs": [-17.866931915283203, -19.307052612304688, -19.98945426940918, -19.08960723876953], "reading_logprob_std": 0.8849040862836639, "tokenization_path_logprob_std": 9.497197596666227, "trajectory_balance_rmse": 251.87649036556036} {"canonical_length": 355, "canonical_nll": 1.3114281882702465, "generation_rouge_l_f1": 0.27597955706984667, "generation_token_f1": 0.5383304940374786, "prompt_id": "e131aed6216cf197", "reading_canonical_log_prob": -465.5570068359375, "reading_distances": [45, 39, 43, 45], "reading_log_probs": [-463.0271301269531, -462.8785705566406, -460.5729064941406, -462.21307373046875], "reading_logprob_std": 1.1238732416779715, "tokenization_path_logprob_std": 104.84310778683155, "trajectory_balance_rmse": 7258.994379009327} {"canonical_length": 46, "canonical_nll": 1.3744606349779211, "generation_rouge_l_f1": 0.10526315789473685, "generation_token_f1": 0.15789473684210525, "prompt_id": "87d0cec84789988b", "reading_canonical_log_prob": -63.225189208984375, "reading_distances": [15, 21, 25, 22], "reading_log_probs": [-62.628692626953125, -65.35324096679688, -63.385868072509766, -64.61663818359375], "reading_logprob_std": 1.2205306455050042, "tokenization_path_logprob_std": 125.0575456526788, "trajectory_balance_rmse": 1000.2219348292989} {"canonical_length": 138, "canonical_nll": 0.9415363919907722, "generation_rouge_l_f1": 0.26993865030674846, "generation_token_f1": 0.2822085889570552, "prompt_id": "d60af44d5acef8d0", "reading_canonical_log_prob": -129.93202209472656, "reading_distances": [28, 34, 31, 35], "reading_log_probs": [-126.11372375488281, -127.32362365722656, -124.73145294189453, -124.85189056396484], "reading_logprob_std": 1.2182662577885357, "tokenization_path_logprob_std": 138.5135072991797, "trajectory_balance_rmse": 2998.8453650912797} {"canonical_length": 35, "canonical_nll": 1.677165331159319, "generation_rouge_l_f1": 0.2191780821917808, "generation_token_f1": 0.2191780821917808, "prompt_id": "a9f740ec35ab9678", "reading_canonical_log_prob": -58.70078659057617, "reading_distances": [14, 15, 17, 11], "reading_log_probs": [-64.60757446289062, -62.403961181640625, -64.58269500732422, -63.183040618896484], "reading_logprob_std": 1.0877616701692665, "tokenization_path_logprob_std": 112.19625785782306, "trajectory_balance_rmse": 768.3597569552607} {"canonical_length": 257, "canonical_nll": 1.3276333938776752, "generation_rouge_l_f1": 0.10051993067590988, "generation_token_f1": 0.194107452339688, "prompt_id": "a25427203b062343", "reading_canonical_log_prob": -341.2017822265625, "reading_distances": [32, 33, 30, 31], "reading_log_probs": [-337.60394287109375, -337.8993835449219, -337.48492431640625, -338.11212158203125], "reading_logprob_std": 0.2843276491573895, "tokenization_path_logprob_std": 235.7554965919028, "trajectory_balance_rmse": 5152.479113620472} {"canonical_length": 28, "canonical_nll": 1.206782613481794, "generation_rouge_l_f1": 0.3103448275862069, "generation_token_f1": 0.44827586206896547, "prompt_id": "b0bf00b191d5ab96", "reading_canonical_log_prob": -33.789913177490234, "reading_distances": [26, 29, 27, 31], "reading_log_probs": [-34.407684326171875, -35.00149917602539, -34.52272033691406, -34.550682067871094], "reading_logprob_std": 0.26133388828416, "tokenization_path_logprob_std": 57.899226530310834, "trajectory_balance_rmse": 774.878974717663} {"canonical_length": 31, "canonical_nll": 0.9390150500882056, "generation_rouge_l_f1": 0.16417910447761194, "generation_token_f1": 0.23880597014925373, "prompt_id": "77df38b1cb264cc8", "reading_canonical_log_prob": -29.109466552734375, "reading_distances": [8, 10, 12, 10], "reading_log_probs": [-31.64356231689453, -31.391672134399414, -28.8214168548584, -30.412235260009766], "reading_logprob_std": 1.2793335499729286, "tokenization_path_logprob_std": 69.47637447107704, "trajectory_balance_rmse": 669.0481536497106} {"canonical_length": 54, "canonical_nll": 1.6923068011248554, "generation_rouge_l_f1": 0.1464968152866242, "generation_token_f1": 0.2229299363057325, "prompt_id": "93f9306a97ca7347", "reading_canonical_log_prob": -91.38456726074219, "reading_distances": [13, 14, 13, 13], "reading_log_probs": [-90.19798278808594, -91.52911376953125, -89.95681762695312, -90.96955871582031], "reading_logprob_std": 0.7208941940480549, "tokenization_path_logprob_std": 108.79350944406505, "trajectory_balance_rmse": 1071.0921086720286} {"canonical_length": 31, "canonical_nll": 1.5513506243305821, "generation_rouge_l_f1": 0.2564102564102564, "generation_token_f1": 0.2564102564102564, "prompt_id": "1cc77327b9384870", "reading_canonical_log_prob": -48.09186935424805, "reading_distances": [26, 23, 26, 19], "reading_log_probs": [-42.142127990722656, -42.27997589111328, -43.42909622192383, -43.42163848876953], "reading_logprob_std": 0.7033469405710869, "tokenization_path_logprob_std": 76.45362678405094, "trajectory_balance_rmse": 644.7149942556514} {"canonical_length": 182, "canonical_nll": 1.4093458573896807, "generation_rouge_l_f1": 0.24519230769230768, "generation_token_f1": 0.49038461538461536, "prompt_id": "0d40240cbf3216d7", "reading_canonical_log_prob": -256.5009460449219, "reading_distances": [27, 22, 27, 25], "reading_log_probs": [-261.4355163574219, -256.02081298828125, -258.2375793457031, -260.60772705078125], "reading_logprob_std": 2.446127597634965, "tokenization_path_logprob_std": 194.48103170820198, "trajectory_balance_rmse": 3521.921825715926} {"canonical_length": 136, "canonical_nll": 1.824822706334731, "generation_rouge_l_f1": 0.24309392265193372, "generation_token_f1": 0.38121546961325964, "prompt_id": "c887cead4589c177", "reading_canonical_log_prob": -248.17588806152344, "reading_distances": [12, 12, 19, 17], "reading_log_probs": [-254.37767028808594, -255.09417724609375, -252.10540771484375, -255.57986450195312], "reading_logprob_std": 1.5373764377831451, "tokenization_path_logprob_std": 238.65880194334994, "trajectory_balance_rmse": 3085.199375965859} {"canonical_length": 169, "canonical_nll": 2.127655029296875, "generation_rouge_l_f1": 0.2456140350877193, "generation_token_f1": 0.38596491228070173, "prompt_id": "5686c4dfdf197ade", "reading_canonical_log_prob": -359.5736999511719, "reading_distances": [13, 11, 11, 10], "reading_log_probs": [-354.0474548339844, -353.9437255859375, -354.8175048828125, -355.49456787109375], "reading_logprob_std": 0.7259986983430051, "tokenization_path_logprob_std": 150.560177090137, "trajectory_balance_rmse": 3100.346373006806} {"canonical_length": 348, "canonical_nll": 1.1587717341280532, "generation_rouge_l_f1": 0.2926829268292683, "generation_token_f1": 0.5300813008130081, "prompt_id": "497e17449fccf12b", "reading_canonical_log_prob": -403.2525634765625, "reading_distances": [35, 24, 33, 27], "reading_log_probs": [-398.06878662109375, -403.3156433105469, -398.666259765625, -397.837158203125], "reading_logprob_std": 2.5861507242598876, "tokenization_path_logprob_std": 331.22696740297505, "trajectory_balance_rmse": 7456.553726436488} {"canonical_length": 38, "canonical_nll": 0.9675000843248869, "generation_rouge_l_f1": 0.4444444444444444, "generation_token_f1": 0.4444444444444444, "prompt_id": "faaabffbce1c5a62", "reading_canonical_log_prob": -36.7650032043457, "reading_distances": [35, 38, 39, 31], "reading_log_probs": [-38.23138427734375, -37.90204620361328, -36.05097579956055, -38.89759063720703], "reading_logprob_std": 1.2188540384533408, "tokenization_path_logprob_std": 56.04502622266794, "trajectory_balance_rmse": 784.3298406915087} {"canonical_length": 76, "canonical_nll": 0.35764367956864207, "generation_rouge_l_f1": 0.14693877551020407, "generation_token_f1": 0.20408163265306126, "prompt_id": "bf9932d84bc4ffec", "reading_canonical_log_prob": -27.180919647216797, "reading_distances": [21, 18, 17, 16], "reading_log_probs": [-24.536434173583984, -25.776878356933594, -26.521739959716797, -25.827716827392578], "reading_logprob_std": 0.8259652532071068, "tokenization_path_logprob_std": 17.942561052520837, "trajectory_balance_rmse": 355.58463119550214} {"canonical_length": 165, "canonical_nll": 1.8241984049479167, "generation_rouge_l_f1": 0.2535211267605634, "generation_token_f1": 0.460093896713615, "prompt_id": "d69cbb797502e917", "reading_canonical_log_prob": -300.99273681640625, "reading_distances": [18, 20, 13, 19], "reading_log_probs": [-300.58392333984375, -297.63446044921875, -301.1617431640625, -301.3316345214844], "reading_logprob_std": 1.7255907599036897, "tokenization_path_logprob_std": 146.625950329447, "trajectory_balance_rmse": 4041.0807787661925} {"canonical_length": 119, "canonical_nll": 1.429702374113708, "generation_rouge_l_f1": 0.21917808219178084, "generation_token_f1": 0.4, "prompt_id": "98ec80f0ace1a8dc", "reading_canonical_log_prob": -170.13458251953125, "reading_distances": [34, 25, 35, 36], "reading_log_probs": [-165.23892211914062, -164.28782653808594, -166.86412048339844, -166.93016052246094], "reading_logprob_std": 1.2919533185822978, "tokenization_path_logprob_std": 243.28052738227746, "trajectory_balance_rmse": 2778.9805030862735} {"canonical_length": 14, "canonical_nll": 1.3566357748849052, "generation_rouge_l_f1": 0.05185185185185185, "generation_token_f1": 0.06666666666666667, "prompt_id": "e2aae622b59df148", "reading_canonical_log_prob": -18.992900848388672, "reading_distances": [33, 23, 26, 32], "reading_log_probs": [-18.935924530029297, -18.971139907836914, -20.23370933532715, -20.91465950012207], "reading_logprob_std": 0.9762138041658563, "tokenization_path_logprob_std": 54.20487990609289, "trajectory_balance_rmse": 257.92943250341585} {"canonical_length": 10, "canonical_nll": 0.673405933380127, "generation_rouge_l_f1": 0.56, "generation_token_f1": 0.56, "prompt_id": "aefa9cb1b8d05c05", "reading_canonical_log_prob": -6.7340593338012695, "reading_distances": [33, 28, 33, 26], "reading_log_probs": [-11.952970504760742, -12.571687698364258, -10.574649810791016, -14.283055305480957], "reading_logprob_std": 1.5378850413194913, "tokenization_path_logprob_std": 23.392387732016875, "trajectory_balance_rmse": 284.82444821391834} {"canonical_length": 121, "canonical_nll": 1.1976409155475207, "generation_rouge_l_f1": 0.25, "generation_token_f1": 0.45312499999999994, "prompt_id": "7467b21b2ac04b77", "reading_canonical_log_prob": -144.91455078125, "reading_distances": [11, 5, 12, 9], "reading_log_probs": [-143.94940185546875, -143.40187072753906, -141.31103515625, -141.70065307617188], "reading_logprob_std": 1.2824188491220663, "tokenization_path_logprob_std": 88.00677663577554, "trajectory_balance_rmse": 2586.1712852893934} {"canonical_length": 92, "canonical_nll": 1.2761964383332625, "generation_rouge_l_f1": 0.14432989690721648, "generation_token_f1": 0.14432989690721648, "prompt_id": "80fb331e54c070fd", "reading_canonical_log_prob": -117.41007232666016, "reading_distances": [18, 21, 19, 23], "reading_log_probs": [-116.885986328125, -115.9222412109375, -115.4107666015625, -113.83919525146484], "reading_logprob_std": 1.2733988970178813, "tokenization_path_logprob_std": 63.2969665960633, "trajectory_balance_rmse": 1566.5861834853458} {"canonical_length": 288, "canonical_nll": 1.3086633682250977, "generation_rouge_l_f1": 0.3132969034608379, "generation_token_f1": 0.5027322404371585, "prompt_id": "1baa92d9506c1060", "reading_canonical_log_prob": -376.8950500488281, "reading_distances": [17, 27, 17, 16], "reading_log_probs": [-374.67449951171875, -373.6689453125, -377.7953186035156, -376.96990966796875], "reading_logprob_std": 1.9283883806706952, "tokenization_path_logprob_std": 96.7893515687816, "trajectory_balance_rmse": 6362.699656128394} {"canonical_length": 232, "canonical_nll": 1.501263322501347, "generation_rouge_l_f1": 0.24066390041493776, "generation_token_f1": 0.4605809128630705, "prompt_id": "23fd1d0b807b2a3d", "reading_canonical_log_prob": -348.2930908203125, "reading_distances": [27, 26, 23, 21], "reading_log_probs": [-348.3208312988281, -342.5465393066406, -344.5553894042969, -344.84027099609375], "reading_logprob_std": 2.398144108689431, "tokenization_path_logprob_std": 72.76712072603272, "trajectory_balance_rmse": 4864.469020126604} {"canonical_length": 23, "canonical_nll": 1.1358176521632983, "generation_rouge_l_f1": 0.4888888888888889, "generation_token_f1": 0.5333333333333332, "prompt_id": "e7e3fd3a1ae85dcc", "reading_canonical_log_prob": -26.12380599975586, "reading_distances": [20, 26, 14, 28], "reading_log_probs": [-24.889171600341797, -24.527175903320312, -24.103784561157227, -24.278108596801758], "reading_logprob_std": 0.3407045765120437, "tokenization_path_logprob_std": 67.09898508030433, "trajectory_balance_rmse": 489.08707763560136} {"canonical_length": 11, "canonical_nll": 0.6140107675032183, "generation_rouge_l_f1": 0.7368421052631577, "generation_token_f1": 0.7368421052631577, "prompt_id": "7a61f927f0ede558", "reading_canonical_log_prob": -6.7541184425354, "reading_distances": [24, 18, 17, 20], "reading_log_probs": [-7.249121189117432, -8.459272384643555, -8.684333801269531, -6.180747032165527], "reading_logprob_std": 1.1610340804260042, "tokenization_path_logprob_std": 34.201446183984125, "trajectory_balance_rmse": 247.52663672746138} {"canonical_length": 244, "canonical_nll": 1.645537579645876, "generation_rouge_l_f1": 0.2357563850687623, "generation_token_f1": 0.5304518664047152, "prompt_id": "d3428656e283f9ef", "reading_canonical_log_prob": -401.51116943359375, "reading_distances": [27, 34, 40, 40], "reading_log_probs": [-402.55877685546875, -405.0367431640625, -401.8836364746094, -403.87017822265625], "reading_logprob_std": 1.401344853482399, "tokenization_path_logprob_std": 127.13234294763309, "trajectory_balance_rmse": 5909.402462169837} {"canonical_length": 18, "canonical_nll": 0.46171463860405815, "generation_rouge_l_f1": 0.8333333333333334, "generation_token_f1": 0.8333333333333334, "prompt_id": "f6a9e161c915b813", "reading_canonical_log_prob": -8.310863494873047, "reading_distances": [12, 15, 12, 19], "reading_log_probs": [-7.479347229003906, -9.192076683044434, -9.243673324584961, -8.08387565612793], "reading_logprob_std": 0.865430804798142, "tokenization_path_logprob_std": 20.42089723068002, "trajectory_balance_rmse": 354.1020011899029} {"canonical_length": 4, "canonical_nll": 2.1312038898468018, "generation_rouge_l_f1": 0.01702127659574468, "generation_token_f1": 0.01702127659574468, "prompt_id": "60d17bcb4782ac31", "reading_canonical_log_prob": -8.524815559387207, "reading_distances": [14, 12, 12, 19], "reading_log_probs": [-7.07981538772583, -6.740745544433594, -6.483692169189453, -6.905826091766357], "reading_logprob_std": 0.25366128972166757, "tokenization_path_logprob_std": 27.24503180119928, "trajectory_balance_rmse": 63.64857209572206} {"canonical_length": 12, "canonical_nll": 1.5059386889139812, "generation_rouge_l_f1": 0.15384615384615385, "generation_token_f1": 0.15384615384615385, "prompt_id": "469967c1d14dd489", "reading_canonical_log_prob": -18.071264266967773, "reading_distances": [26, 23, 30, 26], "reading_log_probs": [-20.542949676513672, -19.450225830078125, -20.85126304626465, -20.349132537841797], "reading_logprob_std": 0.602061819892709, "tokenization_path_logprob_std": 37.47955805083707, "trajectory_balance_rmse": 201.89895972732373} {"canonical_length": 269, "canonical_nll": 1.3279289614312269, "generation_rouge_l_f1": 0.20350877192982456, "generation_token_f1": 0.4245614035087719, "prompt_id": "812c384ebd7a8b2e", "reading_canonical_log_prob": -357.212890625, "reading_distances": [30, 30, 26, 37], "reading_log_probs": [-351.7396240234375, -349.3194274902344, -350.4129943847656, -349.0936279296875], "reading_logprob_std": 1.2112732496156904, "tokenization_path_logprob_std": 102.96161426397317, "trajectory_balance_rmse": 4275.976443969447} {"canonical_length": 99, "canonical_nll": 1.0533239961874605, "generation_rouge_l_f1": 0.2, "generation_token_f1": 0.3263157894736842, "prompt_id": "40633ecaff73b957", "reading_canonical_log_prob": -104.2790756225586, "reading_distances": [19, 26, 23, 15], "reading_log_probs": [-109.70753479003906, -107.28002166748047, -110.10945892333984, -114.20581817626953], "reading_logprob_std": 2.8728825243853264, "tokenization_path_logprob_std": 203.67626605847983, "trajectory_balance_rmse": 2425.6757560067226} {"canonical_length": 210, "canonical_nll": 1.2602716355096726, "generation_rouge_l_f1": 0.21524663677130046, "generation_token_f1": 0.4260089686098655, "prompt_id": "cd8cb6a38db69f38", "reading_canonical_log_prob": -264.65704345703125, "reading_distances": [16, 23, 20, 20], "reading_log_probs": [-266.08148193359375, -265.8788146972656, -264.91424560546875, -267.90533447265625], "reading_logprob_std": 1.2487907912234026, "tokenization_path_logprob_std": 147.32293403345503, "trajectory_balance_rmse": 4221.983524089785} {"canonical_length": 105, "canonical_nll": 1.2063274565197173, "generation_rouge_l_f1": 0.12698412698412698, "generation_token_f1": 0.20634920634920634, "prompt_id": "2de9ac9d5457eaa3", "reading_canonical_log_prob": -126.66438293457031, "reading_distances": [30, 28, 31, 30], "reading_log_probs": [-124.908203125, -123.30239868164062, -126.13569641113281, -125.05095672607422], "reading_logprob_std": 1.1678833314831272, "tokenization_path_logprob_std": 42.24445023210775, "trajectory_balance_rmse": 1146.0219707750127} {"canonical_length": 99, "canonical_nll": 1.21844104805378, "generation_rouge_l_f1": 0.2696629213483146, "generation_token_f1": 0.3932584269662921, "prompt_id": "0b44770e70f57065", "reading_canonical_log_prob": -120.62566375732422, "reading_distances": [28, 23, 29, 32], "reading_log_probs": [-116.34033966064453, -119.056396484375, -114.72561645507812, -115.09249877929688], "reading_logprob_std": 1.9609552494425262, "tokenization_path_logprob_std": 49.52504570748834, "trajectory_balance_rmse": 2094.141650183265} {"canonical_length": 111, "canonical_nll": 1.6650143185177364, "generation_rouge_l_f1": 0.19618528610354224, "generation_token_f1": 0.3651226158038148, "prompt_id": "f5df13177d4b7c6c", "reading_canonical_log_prob": -184.81658935546875, "reading_distances": [16, 19, 15, 16], "reading_log_probs": [-187.05386352539062, -185.54803466796875, -188.22715759277344, -187.229248046875], "reading_logprob_std": 1.1058518880708776, "tokenization_path_logprob_std": 105.84419155075676, "trajectory_balance_rmse": 2196.0993345237466} {"canonical_length": 129, "canonical_nll": 1.9053672376529192, "generation_rouge_l_f1": 0.2506393861892583, "generation_token_f1": 0.36317135549872126, "prompt_id": "b85464f813712359", "reading_canonical_log_prob": -245.79237365722656, "reading_distances": [41, 46, 44, 43], "reading_log_probs": [-238.66302490234375, -241.15846252441406, -238.20082092285156, -239.84129333496094], "reading_logprob_std": 1.322966559020366, "tokenization_path_logprob_std": 142.2382278869013, "trajectory_balance_rmse": 3081.010360642829} {"canonical_length": 52, "canonical_nll": 1.2327323326697717, "generation_rouge_l_f1": 0.1208053691275168, "generation_token_f1": 0.16778523489932884, "prompt_id": "fa52ad4bc1072fe5", "reading_canonical_log_prob": -64.10208129882812, "reading_distances": [47, 48, 45, 59], "reading_log_probs": [-63.768035888671875, -67.62300109863281, -65.66827392578125, -64.06843566894531], "reading_logprob_std": 1.7695874650589327, "tokenization_path_logprob_std": 13.737649685467419, "trajectory_balance_rmse": 797.0572304508804} {"canonical_length": 192, "canonical_nll": 1.2141224543253581, "generation_rouge_l_f1": 0.36206896551724144, "generation_token_f1": 0.5344827586206896, "prompt_id": "63e9efbaa63103ce", "reading_canonical_log_prob": -233.11151123046875, "reading_distances": [16, 24, 25, 23], "reading_log_probs": [-231.90664672851562, -231.65875244140625, -230.23220825195312, -230.65542602539062], "reading_logprob_std": 0.7985170913738404, "tokenization_path_logprob_std": 132.22542820018063, "trajectory_balance_rmse": 4201.492189354855} {"canonical_length": 265, "canonical_nll": 1.3406692216981133, "generation_rouge_l_f1": 0.21013133208255164, "generation_token_f1": 0.4090056285178237, "prompt_id": "867f08ace4ea14ea", "reading_canonical_log_prob": -355.27734375, "reading_distances": [22, 25, 33, 21], "reading_log_probs": [-347.561767578125, -343.3123779296875, -348.20477294921875, -343.7215881347656], "reading_logprob_std": 2.5400077845733806, "tokenization_path_logprob_std": 17.607978764883388, "trajectory_balance_rmse": 5500.1238987684155} {"canonical_length": 6, "canonical_nll": 0.08784961700439453, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "126ead873a35c47e", "reading_canonical_log_prob": -0.5270977020263672, "reading_distances": [31, 37, 32, 27], "reading_log_probs": [-2.1159067153930664, -3.139136552810669, -2.2583324909210205, -2.4692485332489014], "reading_logprob_std": 0.45287840283632846, "tokenization_path_logprob_std": 31.516110642360246, "trajectory_balance_rmse": 100.69126844829296} {"canonical_length": 51, "canonical_nll": 2.380688835592831, "generation_rouge_l_f1": 0.13636363636363638, "generation_token_f1": 0.25757575757575757, "prompt_id": "9800df60020cf919", "reading_canonical_log_prob": -121.41513061523438, "reading_distances": [51, 49, 47, 41], "reading_log_probs": [-122.70555114746094, -121.16656494140625, -121.95686340332031, -122.00731658935547], "reading_logprob_std": 0.6291875644564635, "tokenization_path_logprob_std": 106.60294835799763, "trajectory_balance_rmse": 1024.6674882657378} {"canonical_length": 11, "canonical_nll": 0.9344233599576083, "generation_rouge_l_f1": 0.06802721088435373, "generation_token_f1": 0.06802721088435373, "prompt_id": "0aba661f60975419", "reading_canonical_log_prob": -10.278656959533691, "reading_distances": [12, 13, 20, 19], "reading_log_probs": [-8.083600997924805, -8.174630165100098, -8.530616760253906, -9.509860038757324], "reading_logprob_std": 0.6526105859376, "tokenization_path_logprob_std": 32.48604075466185, "trajectory_balance_rmse": 210.91792246436205} {"canonical_length": 245, "canonical_nll": 1.1388361716757016, "generation_rouge_l_f1": 0.3652173913043478, "generation_token_f1": 0.6086956521739131, "prompt_id": "a78682b2f45a9311", "reading_canonical_log_prob": -279.0148620605469, "reading_distances": [15, 17, 17, 15], "reading_log_probs": [-276.706787109375, -277.89788818359375, -278.75592041015625, -277.36993408203125], "reading_logprob_std": 0.8657051989392212, "tokenization_path_logprob_std": 168.2399521742307, "trajectory_balance_rmse": 4627.558445768914} {"canonical_length": 41, "canonical_nll": 0.6988930585907727, "generation_rouge_l_f1": 0.12075471698113206, "generation_token_f1": 0.13584905660377358, "prompt_id": "e970a5e256be139f", "reading_canonical_log_prob": -28.65461540222168, "reading_distances": [14, 16, 12, 15], "reading_log_probs": [-23.940099716186523, -23.027313232421875, -23.188648223876953, -24.486148834228516], "reading_logprob_std": 0.6790781911275743, "tokenization_path_logprob_std": 44.520602229074534, "trajectory_balance_rmse": 326.6405506313711} {"canonical_length": 112, "canonical_nll": 1.689689772469657, "generation_rouge_l_f1": 0.2909090909090909, "generation_token_f1": 0.4727272727272727, "prompt_id": "0b4d03b967102211", "reading_canonical_log_prob": -189.24525451660156, "reading_distances": [23, 28, 26, 24], "reading_log_probs": [-183.5105743408203, -183.9015350341797, -183.9467315673828, -186.03582763671875], "reading_logprob_std": 1.1416931003726463, "tokenization_path_logprob_std": 211.2420803084704, "trajectory_balance_rmse": 2195.898774816165} {"canonical_length": 354, "canonical_nll": 1.154104286667991, "generation_rouge_l_f1": 0.2558139534883721, "generation_token_f1": 0.5083056478405316, "prompt_id": "e03c0c9ce1e8af04", "reading_canonical_log_prob": -408.55291748046875, "reading_distances": [11, 15, 15, 18], "reading_log_probs": [-407.98223876953125, -406.3995361328125, -407.4217529296875, -407.93280029296875], "reading_logprob_std": 0.734763666542765, "tokenization_path_logprob_std": 218.29398960464403, "trajectory_balance_rmse": 7270.787089719079} {"canonical_length": 114, "canonical_nll": 2.0062602528354576, "generation_rouge_l_f1": 0.11789473684210527, "generation_token_f1": 0.16842105263157894, "prompt_id": "6aab22322acba4ab", "reading_canonical_log_prob": -228.7136688232422, "reading_distances": [49, 44, 49, 43], "reading_log_probs": [-235.7593231201172, -233.703369140625, -236.26419067382812, -242.38653564453125], "reading_logprob_std": 3.73987610374217, "tokenization_path_logprob_std": 93.91701449095453, "trajectory_balance_rmse": 2128.1823592984774} {"canonical_length": 72, "canonical_nll": 0.7679963641696506, "generation_rouge_l_f1": 0.3252032520325204, "generation_token_f1": 0.5365853658536585, "prompt_id": "1498d74fffe04fb0", "reading_canonical_log_prob": -55.295738220214844, "reading_distances": [27, 29, 20, 31], "reading_log_probs": [-60.97087478637695, -60.218055725097656, -56.534271240234375, -62.69941329956055], "reading_logprob_std": 2.5976658290210852, "tokenization_path_logprob_std": 59.800351001026414, "trajectory_balance_rmse": 1329.3912340511117} {"canonical_length": 203, "canonical_nll": 1.6496466274919181, "generation_rouge_l_f1": 0.20985010706638116, "generation_token_f1": 0.4668094218415418, "prompt_id": "363570662fa8a4e6", "reading_canonical_log_prob": -334.8782653808594, "reading_distances": [31, 30, 29, 32], "reading_log_probs": [-335.7384948730469, -339.03326416015625, -335.2137451171875, -337.9500427246094], "reading_logprob_std": 1.8090335784546268, "tokenization_path_logprob_std": 173.12897602133629, "trajectory_balance_rmse": 4580.627537335686} {"canonical_length": 267, "canonical_nll": 1.6830030577013109, "generation_rouge_l_f1": 0.21621621621621623, "generation_token_f1": 0.4517374517374517, "prompt_id": "2b4383e7d7d3624e", "reading_canonical_log_prob": -449.36181640625, "reading_distances": [25, 34, 34, 33], "reading_log_probs": [-439.39923095703125, -439.02947998046875, -437.1824951171875, -438.4598388671875], "reading_logprob_std": 0.9704183893890016, "tokenization_path_logprob_std": 262.67699007290395, "trajectory_balance_rmse": 5520.990534813914} {"canonical_length": 72, "canonical_nll": 1.3947522905137804, "generation_rouge_l_f1": 0.3529411764705882, "generation_token_f1": 0.392156862745098, "prompt_id": "fadbe59576c37aff", "reading_canonical_log_prob": -100.42216491699219, "reading_distances": [19, 22, 16, 20], "reading_log_probs": [-105.86363983154297, -104.7685775756836, -101.44979095458984, -104.33106231689453], "reading_logprob_std": 1.8827618452788846, "tokenization_path_logprob_std": 89.74972671778737, "trajectory_balance_rmse": 1142.4601767205156} {"canonical_length": 119, "canonical_nll": 1.5338711778656775, "generation_rouge_l_f1": 0.24126984126984127, "generation_token_f1": 0.43809523809523804, "prompt_id": "6caa7e0c0b6b5ae9", "reading_canonical_log_prob": -182.53067016601562, "reading_distances": [26, 30, 24, 30], "reading_log_probs": [-185.26995849609375, -184.73809814453125, -184.82620239257812, -186.29986572265625], "reading_logprob_std": 0.7164186244392136, "tokenization_path_logprob_std": 35.029123805717205, "trajectory_balance_rmse": 2873.5501121515954} {"canonical_length": 246, "canonical_nll": 1.2167548202886813, "generation_rouge_l_f1": 0.23838383838383836, "generation_token_f1": 0.501010101010101, "prompt_id": "09ec255dc57dfab9", "reading_canonical_log_prob": -299.3216857910156, "reading_distances": [26, 27, 21, 29], "reading_log_probs": [-307.385498046875, -302.1292724609375, -309.41094970703125, -308.4903564453125], "reading_logprob_std": 3.2568489224576815, "tokenization_path_logprob_std": 141.5301207728788, "trajectory_balance_rmse": 5229.976165532665} {"canonical_length": 83, "canonical_nll": 1.4056804611022213, "generation_rouge_l_f1": 0.2548476454293629, "generation_token_f1": 0.3213296398891966, "prompt_id": "a8f654f305639b3f", "reading_canonical_log_prob": -116.67147827148438, "reading_distances": [17, 20, 13, 15], "reading_log_probs": [-128.88307189941406, -123.77351379394531, -126.5439453125, -123.9349365234375], "reading_logprob_std": 2.42506728222407, "tokenization_path_logprob_std": 49.93238073310682, "trajectory_balance_rmse": 708.2524270079839} {"canonical_length": 286, "canonical_nll": 1.257517141062063, "generation_rouge_l_f1": 0.2303030303030303, "generation_token_f1": 0.4565656565656566, "prompt_id": "0e8185255effe518", "reading_canonical_log_prob": -359.64990234375, "reading_distances": [30, 29, 26, 25], "reading_log_probs": [-360.33465576171875, -360.87139892578125, -362.7657775878906, -356.2711486816406], "reading_logprob_std": 2.733156200588302, "tokenization_path_logprob_std": 155.71310182825238, "trajectory_balance_rmse": 4874.887608789416} {"canonical_length": 51, "canonical_nll": 1.0461971806544883, "generation_rouge_l_f1": 0.449438202247191, "generation_token_f1": 0.5842696629213482, "prompt_id": "d03a9ea989bf62b9", "reading_canonical_log_prob": -53.356056213378906, "reading_distances": [30, 36, 34, 33], "reading_log_probs": [-52.495513916015625, -51.591941833496094, -52.266845703125, -52.03287124633789], "reading_logprob_std": 0.3859434248781077, "tokenization_path_logprob_std": 117.49234857741996, "trajectory_balance_rmse": 870.6934802244217} {"canonical_length": 69, "canonical_nll": 0.6971376391424648, "generation_rouge_l_f1": 0.38961038961038963, "generation_token_f1": 0.5844155844155844, "prompt_id": "fd11f4c3417e5ef0", "reading_canonical_log_prob": -48.10249710083008, "reading_distances": [44, 42, 39, 41], "reading_log_probs": [-47.662841796875, -46.92658233642578, -46.34774398803711, -46.01704025268555], "reading_logprob_std": 0.7217980187815075, "tokenization_path_logprob_std": 90.31457741471982, "trajectory_balance_rmse": 1331.5570143022717} {"canonical_length": 40, "canonical_nll": 1.6261564254760743, "generation_rouge_l_f1": 0.23115577889447236, "generation_token_f1": 0.35175879396984927, "prompt_id": "43fe74e6e2773223", "reading_canonical_log_prob": -65.04625701904297, "reading_distances": [16, 14, 13, 20], "reading_log_probs": [-62.27006530761719, -62.425777435302734, -65.48084259033203, -65.43753814697266], "reading_logprob_std": 1.7975034115546678, "tokenization_path_logprob_std": 105.88172965844115, "trajectory_balance_rmse": 647.5438780862678} {"canonical_length": 82, "canonical_nll": 1.1683655715570218, "generation_rouge_l_f1": 0.10919540229885057, "generation_token_f1": 0.2126436781609195, "prompt_id": "7eeeafb10f4366ae", "reading_canonical_log_prob": -95.80597686767578, "reading_distances": [30, 28, 34, 28], "reading_log_probs": [-90.49090576171875, -91.96751403808594, -91.91426086425781, -91.691650390625], "reading_logprob_std": 0.6938174285004375, "tokenization_path_logprob_std": 18.52941879543171, "trajectory_balance_rmse": 1083.169639032718} {"canonical_length": 3, "canonical_nll": 1.254802942276001, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.6666666666666666, "prompt_id": "6b50f3d234f6e0b4", "reading_canonical_log_prob": -3.764408826828003, "reading_distances": [20, 25, 26, 21], "reading_log_probs": [-5.627631187438965, -5.986844539642334, -6.366832733154297, -5.895333290100098], "reading_logprob_std": 0.3058042176186146, "tokenization_path_logprob_std": 19.510807423932913, "trajectory_balance_rmse": 25.64536669740563} {"canonical_length": 348, "canonical_nll": 1.1726104034774605, "generation_rouge_l_f1": 0.2436548223350254, "generation_token_f1": 0.46700507614213194, "prompt_id": "b62b5836a5079a74", "reading_canonical_log_prob": -408.06842041015625, "reading_distances": [31, 37, 29, 30], "reading_log_probs": [-406.0713195800781, -409.50860595703125, -405.9452209472656, -407.49237060546875], "reading_logprob_std": 1.6584836021885383, "tokenization_path_logprob_std": 434.04792432835455, "trajectory_balance_rmse": 7251.348862257849} {"canonical_length": 39, "canonical_nll": 1.2855553260216346, "generation_rouge_l_f1": 0.4727272727272728, "generation_token_f1": 0.5454545454545454, "prompt_id": "495dc4d7314e6352", "reading_canonical_log_prob": -50.13665771484375, "reading_distances": [27, 21, 27, 27], "reading_log_probs": [-50.46769714355469, -50.15919494628906, -49.8502082824707, -49.25117492675781], "reading_logprob_std": 0.519230719332671, "tokenization_path_logprob_std": 92.80167357688556, "trajectory_balance_rmse": 864.0224508345376} {"canonical_length": 54, "canonical_nll": 1.1395821747956452, "generation_rouge_l_f1": 0.31884057971014496, "generation_token_f1": 0.3478260869565218, "prompt_id": "0c8a53a7cd7794b9", "reading_canonical_log_prob": -61.537437438964844, "reading_distances": [24, 25, 26, 30], "reading_log_probs": [-64.40132141113281, -61.1769905090332, -63.11534881591797, -62.689369201660156], "reading_logprob_std": 1.3293731767560233, "tokenization_path_logprob_std": 56.444866336348355, "trajectory_balance_rmse": 1056.8186462704516} {"canonical_length": 185, "canonical_nll": 1.8777360245988175, "generation_rouge_l_f1": 0.21411192214111924, "generation_token_f1": 0.44768856447688565, "prompt_id": "bfd0b674c803f580", "reading_canonical_log_prob": -347.38116455078125, "reading_distances": [31, 33, 30, 37], "reading_log_probs": [-347.82684326171875, -346.63958740234375, -347.0437927246094, -347.3893127441406], "reading_logprob_std": 0.5048953429240647, "tokenization_path_logprob_std": 137.75946541068964, "trajectory_balance_rmse": 4254.479392464341} {"canonical_length": 215, "canonical_nll": 1.515332882903343, "generation_rouge_l_f1": 0.17748917748917747, "generation_token_f1": 0.3809523809523809, "prompt_id": "f65fb83f66989dab", "reading_canonical_log_prob": -325.79656982421875, "reading_distances": [23, 31, 22, 26], "reading_log_probs": [-323.85638427734375, -324.7803039550781, -324.408447265625, -325.30609130859375], "reading_logprob_std": 0.6110473356511062, "tokenization_path_logprob_std": 414.8093914454394, "trajectory_balance_rmse": 3837.539087069883} {"canonical_length": 20, "canonical_nll": 0.6940269947052002, "generation_rouge_l_f1": 0.7027027027027027, "generation_token_f1": 0.7027027027027027, "prompt_id": "ed802f2ffb8ef795", "reading_canonical_log_prob": -13.880539894104004, "reading_distances": [22, 18, 19, 22], "reading_log_probs": [-14.687093734741211, -14.184446334838867, -13.800649642944336, -14.538202285766602], "reading_logprob_std": 0.3954989754040681, "tokenization_path_logprob_std": 61.384371814339765, "trajectory_balance_rmse": 465.1589765499244} {"canonical_length": 17, "canonical_nll": 0.8108756121467141, "generation_rouge_l_f1": 0.27272727272727276, "generation_token_f1": 0.36363636363636365, "prompt_id": "578c6b312be28f5a", "reading_canonical_log_prob": -13.78488540649414, "reading_distances": [29, 21, 29, 32], "reading_log_probs": [-11.971048355102539, -12.429841995239258, -12.067338943481445, -12.517707824707031], "reading_logprob_std": 0.2677936165916293, "tokenization_path_logprob_std": 39.809160490167336, "trajectory_balance_rmse": 381.58375767762186} {"canonical_length": 20, "canonical_nll": 1.2491803169250488, "generation_rouge_l_f1": 0.4615384615384615, "generation_token_f1": 0.4615384615384615, "prompt_id": "016449fd389e919a", "reading_canonical_log_prob": -24.983606338500977, "reading_distances": [13, 19, 14, 16], "reading_log_probs": [-26.480743408203125, -25.540796279907227, -25.326004028320312, -26.70612144470215], "reading_logprob_std": 0.6816989458225289, "tokenization_path_logprob_std": 34.46698488756942, "trajectory_balance_rmse": 450.8428893634817} {"canonical_length": 172, "canonical_nll": 1.3861217055209847, "generation_rouge_l_f1": 0.29235880398671094, "generation_token_f1": 0.5249169435215948, "prompt_id": "977004fae436eabd", "reading_canonical_log_prob": -238.41293334960938, "reading_distances": [26, 27, 25, 25], "reading_log_probs": [-238.82070922851562, -240.1552276611328, -239.33224487304688, -240.4073486328125], "reading_logprob_std": 0.7335272381898515, "tokenization_path_logprob_std": 197.99932281823064, "trajectory_balance_rmse": 3472.254880842141} {"canonical_length": 223, "canonical_nll": 1.5981959868974216, "generation_rouge_l_f1": 0.2459677419354839, "generation_token_f1": 0.4838709677419354, "prompt_id": "3d21a3aabdc36c10", "reading_canonical_log_prob": -356.397705078125, "reading_distances": [32, 23, 34, 31], "reading_log_probs": [-353.9928894042969, -351.47064208984375, -354.24798583984375, -351.84393310546875], "reading_logprob_std": 1.4340290826500612, "tokenization_path_logprob_std": 268.17094146181364, "trajectory_balance_rmse": 4203.856938359319} {"canonical_length": 248, "canonical_nll": 0.3892511244743101, "generation_rouge_l_f1": 0.2653465346534653, "generation_token_f1": 0.5782178217821783, "prompt_id": "99f1a0976d255260", "reading_canonical_log_prob": -96.5342788696289, "reading_distances": [34, 32, 34, 32], "reading_log_probs": [-99.52870178222656, -100.52734375, -100.5597152709961, -99.47049713134766], "reading_logprob_std": 0.6033262813819978, "tokenization_path_logprob_std": 137.3672267787423, "trajectory_balance_rmse": 3730.1483329656658} {"canonical_length": 183, "canonical_nll": 1.0001451669494963, "generation_rouge_l_f1": 0.3905882352941177, "generation_token_f1": 0.5929411764705882, "prompt_id": "24640c4b7c2fb283", "reading_canonical_log_prob": -183.0265655517578, "reading_distances": [72, 61, 66, 65], "reading_log_probs": [-168.8457489013672, -171.27352905273438, -172.29014587402344, -170.74545288085938], "reading_logprob_std": 1.445259862045423, "tokenization_path_logprob_std": 179.95996748976998, "trajectory_balance_rmse": 2972.7663398075506} {"canonical_length": 6, "canonical_nll": 0.10522122184435527, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "94e6b38d533edcd5", "reading_canonical_log_prob": -0.6313273310661316, "reading_distances": [35, 35, 38, 30], "reading_log_probs": [-1.0695514678955078, -1.3186771869659424, -1.9313230514526367, -1.0495954751968384], "reading_logprob_std": 0.41132869757497237, "tokenization_path_logprob_std": 32.11290950512889, "trajectory_balance_rmse": 100.32647501444804} {"canonical_length": 252, "canonical_nll": 1.4645327613467263, "generation_rouge_l_f1": 0.24609375, "generation_token_f1": 0.5078125, "prompt_id": "22a8bdc8dd7cf586", "reading_canonical_log_prob": -369.062255859375, "reading_distances": [27, 22, 23, 22], "reading_log_probs": [-368.9110412597656, -368.95556640625, -368.5185546875, -368.7092590332031], "reading_logprob_std": 0.20098900816689233, "tokenization_path_logprob_std": 116.27971984174653, "trajectory_balance_rmse": 5426.037639942187} {"canonical_length": 27, "canonical_nll": 1.5279049343532987, "generation_rouge_l_f1": 0.09395973154362416, "generation_token_f1": 0.1342281879194631, "prompt_id": "6585d8646962bd31", "reading_canonical_log_prob": -41.25343322753906, "reading_distances": [16, 14, 18, 12], "reading_log_probs": [-39.882816314697266, -38.645503997802734, -39.02006912231445, -38.96562194824219], "reading_logprob_std": 0.5293281136512243, "tokenization_path_logprob_std": 65.22755318473435, "trajectory_balance_rmse": 301.4894893431944} {"canonical_length": 81, "canonical_nll": 1.373867647147473, "generation_rouge_l_f1": 0.3976608187134503, "generation_token_f1": 0.5964912280701755, "prompt_id": "04a5aa53ab7903f1", "reading_canonical_log_prob": -111.28327941894531, "reading_distances": [22, 17, 27, 27], "reading_log_probs": [-111.18795013427734, -111.80867767333984, -109.79986572265625, -110.1707534790039], "reading_logprob_std": 0.9220651140440312, "tokenization_path_logprob_std": 95.84228294318075, "trajectory_balance_rmse": 1887.6116214624133} {"canonical_length": 198, "canonical_nll": 1.021892046687579, "generation_rouge_l_f1": 0.23605150214592274, "generation_token_f1": 0.41201716738197425, "prompt_id": "f2143d23c11c6f8b", "reading_canonical_log_prob": -202.33462524414062, "reading_distances": [20, 24, 23, 20], "reading_log_probs": [-201.64491271972656, -202.42855834960938, -202.62796020507812, -202.341552734375], "reading_logprob_std": 0.42770223852008393, "tokenization_path_logprob_std": 193.79149586260203, "trajectory_balance_rmse": 3797.4305196372243} {"canonical_length": 12, "canonical_nll": 0.7801425457000732, "generation_rouge_l_f1": 0.6, "generation_token_f1": 0.6, "prompt_id": "5633164a0de34e8b", "reading_canonical_log_prob": -9.361710548400879, "reading_distances": [36, 42, 40, 33], "reading_log_probs": [-9.052642822265625, -9.796172142028809, -9.97542667388916, -8.389897346496582], "reading_logprob_std": 0.7284253386346837, "tokenization_path_logprob_std": 45.335834399510254, "trajectory_balance_rmse": 245.47292599964064} {"canonical_length": 134, "canonical_nll": 1.1999865859302121, "generation_rouge_l_f1": 0.158974358974359, "generation_token_f1": 0.2923076923076924, "prompt_id": "fe842e52ca168c74", "reading_canonical_log_prob": -160.79820251464844, "reading_distances": [26, 25, 19, 27], "reading_log_probs": [-150.20721435546875, -150.99554443359375, -152.57574462890625, -146.7352294921875], "reading_logprob_std": 2.4672017208873913, "tokenization_path_logprob_std": 198.78039594389685, "trajectory_balance_rmse": 2252.4437069272476} {"canonical_length": 46, "canonical_nll": 0.3483080656632133, "generation_rouge_l_f1": 0.3020408163265306, "generation_token_f1": 0.3020408163265306, "prompt_id": "6327942806f18cd4", "reading_canonical_log_prob": -16.022171020507812, "reading_distances": [7, 13, 10, 12], "reading_log_probs": [-16.533721923828125, -18.051700592041016, -16.267759323120117, -17.296289443969727], "reading_logprob_std": 0.8045428944429698, "tokenization_path_logprob_std": 35.87304804770387, "trajectory_balance_rmse": 501.90337266750964} {"canonical_length": 54, "canonical_nll": 0.8918911969220197, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "95494e563c3fc530", "reading_canonical_log_prob": -48.16212463378906, "reading_distances": [29, 29, 30, 30], "reading_log_probs": [-50.332767486572266, -50.436073303222656, -50.39949035644531, -50.49524688720703], "reading_logprob_std": 0.06802740071013186, "tokenization_path_logprob_std": 76.01040893600565, "trajectory_balance_rmse": 1048.281476729799} {"canonical_length": 44, "canonical_nll": 1.8175898465243252, "generation_rouge_l_f1": 0.1988950276243094, "generation_token_f1": 0.34254143646408836, "prompt_id": "15a442e093e950e5", "reading_canonical_log_prob": -79.97395324707031, "reading_distances": [23, 29, 30, 26], "reading_log_probs": [-75.37625885009766, -74.67139434814453, -74.60590362548828, -73.29977416992188], "reading_logprob_std": 0.8657210322404234, "tokenization_path_logprob_std": 81.84449318439341, "trajectory_balance_rmse": 786.2120866517039} {"canonical_length": 303, "canonical_nll": 0.7605324420991904, "generation_rouge_l_f1": 0.2435129740518962, "generation_token_f1": 0.4191616766467066, "prompt_id": "452a884e801c54f0", "reading_canonical_log_prob": -230.4413299560547, "reading_distances": [15, 24, 29, 28], "reading_log_probs": [-226.5922393798828, -225.96685791015625, -227.31350708007812, -225.13571166992188], "reading_logprob_std": 0.9255567223708786, "tokenization_path_logprob_std": 153.58175824581545, "trajectory_balance_rmse": 5004.797259602423} {"canonical_length": 340, "canonical_nll": 1.1846174352309282, "generation_rouge_l_f1": 0.21739130434782605, "generation_token_f1": 0.3946488294314382, "prompt_id": "6f1aa7f491e54018", "reading_canonical_log_prob": -402.7699279785156, "reading_distances": [25, 23, 16, 18], "reading_log_probs": [-398.1920471191406, -397.6441650390625, -398.5601501464844, -402.9765625], "reading_logprob_std": 2.4512828532016844, "tokenization_path_logprob_std": 208.53752097699055, "trajectory_balance_rmse": 6678.10473249446} {"canonical_length": 197, "canonical_nll": 0.8423111813927665, "generation_rouge_l_f1": 0.27374301675977647, "generation_token_f1": 0.45810055865921784, "prompt_id": "b6dbd4688fd838f3", "reading_canonical_log_prob": -165.935302734375, "reading_distances": [20, 19, 18, 22], "reading_log_probs": [-164.76214599609375, -165.08541870117188, -164.92733764648438, -165.27186584472656], "reading_logprob_std": 0.2179563572278456, "tokenization_path_logprob_std": 290.00877150459945, "trajectory_balance_rmse": 4323.388456410127} {"canonical_length": 138, "canonical_nll": 1.727140011994735, "generation_rouge_l_f1": 0.21176470588235294, "generation_token_f1": 0.4235294117647059, "prompt_id": "4e91817c9d8f7b21", "reading_canonical_log_prob": -238.34532165527344, "reading_distances": [17, 20, 19, 20], "reading_log_probs": [-240.9678955078125, -243.92822265625, -240.48223876953125, -242.89349365234375], "reading_logprob_std": 1.6193349117352027, "tokenization_path_logprob_std": 63.88757063929295, "trajectory_balance_rmse": 1894.4091385218337} {"canonical_length": 10, "canonical_nll": 1.1835392951965331, "generation_rouge_l_f1": 0.43478260869565216, "generation_token_f1": 0.43478260869565216, "prompt_id": "8f6ce2573cd1b9d1", "reading_canonical_log_prob": -11.835392951965332, "reading_distances": [37, 33, 38, 35], "reading_log_probs": [-12.497818946838379, -12.270181655883789, -11.17691421508789, -12.830906867980957], "reading_logprob_std": 0.7160594578556966, "tokenization_path_logprob_std": 25.279450533119395, "trajectory_balance_rmse": 198.61944110641815} {"canonical_length": 9, "canonical_nll": 0.7764274809095595, "generation_rouge_l_f1": 0.33333333333333337, "generation_token_f1": 0.33333333333333337, "prompt_id": "c246d0eff1fb5aac", "reading_canonical_log_prob": -6.987847328186035, "reading_distances": [12, 18, 17, 17], "reading_log_probs": [-7.816011905670166, -7.058235168457031, -7.750655174255371, -7.649885654449463], "reading_logprob_std": 0.3471005228165203, "tokenization_path_logprob_std": 21.96750168298759, "trajectory_balance_rmse": 161.76867461675096} {"canonical_length": 222, "canonical_nll": 0.623102686426661, "generation_rouge_l_f1": 0.3858093126385809, "generation_token_f1": 0.5809312638580931, "prompt_id": "877651642570fc39", "reading_canonical_log_prob": -138.32879638671875, "reading_distances": [31, 33, 32, 30], "reading_log_probs": [-145.97216796875, -144.94873046875, -146.8865203857422, -142.60708618164062], "reading_logprob_std": 1.8429850092034539, "tokenization_path_logprob_std": 151.1880985259727, "trajectory_balance_rmse": 2997.2348005168537} {"canonical_length": 53, "canonical_nll": 0.8712373409631118, "generation_rouge_l_f1": 0.28415300546448086, "generation_token_f1": 0.3825136612021858, "prompt_id": "77a973a8d7f91484", "reading_canonical_log_prob": -46.17557907104492, "reading_distances": [31, 29, 31, 33], "reading_log_probs": [-43.90362548828125, -43.42157745361328, -44.97697830200195, -45.64775085449219], "reading_logprob_std": 1.0104242298259785, "tokenization_path_logprob_std": 46.300019342232616, "trajectory_balance_rmse": 1060.6724394267217} {"canonical_length": 21, "canonical_nll": 0.8166329520089286, "generation_rouge_l_f1": 0.4736842105263158, "generation_token_f1": 0.5263157894736842, "prompt_id": "f9bfdd6eb133d424", "reading_canonical_log_prob": -17.1492919921875, "reading_distances": [24, 24, 22, 26], "reading_log_probs": [-16.005271911621094, -16.817827224731445, -15.951248168945312, -16.337749481201172], "reading_logprob_std": 0.39838457132589883, "tokenization_path_logprob_std": 38.74025449250947, "trajectory_balance_rmse": 399.80925937350224} {"canonical_length": 247, "canonical_nll": 1.4317835757606907, "generation_rouge_l_f1": 0.28817204301075267, "generation_token_f1": 0.49032258064516127, "prompt_id": "3a822df2306dda39", "reading_canonical_log_prob": -353.6505432128906, "reading_distances": [39, 36, 48, 42], "reading_log_probs": [-347.99957275390625, -349.01202392578125, -352.9641418457031, -351.6783447265625], "reading_logprob_std": 2.3019401919303815, "tokenization_path_logprob_std": 246.80176153023672, "trajectory_balance_rmse": 5429.220557404328} {"canonical_length": 88, "canonical_nll": 1.2620533162897283, "generation_rouge_l_f1": 0.1956521739130435, "generation_token_f1": 0.40217391304347827, "prompt_id": "c28f9f6ed8615ffa", "reading_canonical_log_prob": -111.0606918334961, "reading_distances": [26, 27, 29, 29], "reading_log_probs": [-105.57855224609375, -100.92573547363281, -102.8575439453125, -100.38517761230469], "reading_logprob_std": 2.3480561505776345, "tokenization_path_logprob_std": 106.06921234238388, "trajectory_balance_rmse": 1753.4978861701459}