CoolFace
Modelpublic

MARS-Retokenization/olmo2-7b-instruct-inv-reference-mixed-self-sft

sourceHugging Faceapache-2.0updated 16d agoView on Hugging Face
0likes77downloads
test_metrics.per_example.jsonl2478 linesDownload Raw Back to root
1{"canonical_length": 14, "canonical_nll": 1.9497860499790736, "generation_rouge_l_f1": 0.3157894736842105, "generation_token_f1": 0.3157894736842105, "prompt_id": "54ef1334aadf1dfd", "reading_canonical_log_prob": -27.29700469970703, "reading_distances": [37, 36, 33, 38], "reading_log_probs": [-30.131423950195312, -31.792617797851562, -30.130577087402344, -31.724775314331055], "reading_logprob_std": 0.9401588606349203, "tokenization_path_logprob_std": 35.42418200284078, "trajectory_balance_rmse": 252.34524507633725}2{"canonical_length": 23, "canonical_nll": 0.585747594418733, "generation_rouge_l_f1": 0.15384615384615385, "generation_token_f1": 0.2051282051282051, "prompt_id": "d109cb9b45bb2b84", "reading_canonical_log_prob": -13.47219467163086, "reading_distances": [12, 12, 9, 10], "reading_log_probs": [-20.349119186401367, -16.827327728271484, -17.873336791992188, -16.198217391967773], "reading_logprob_std": 1.827075053809213, "tokenization_path_logprob_std": 49.206924834885726, "trajectory_balance_rmse": 216.76323617175117}3{"canonical_length": 79, "canonical_nll": 1.7142867076246044, "generation_rouge_l_f1": 0.35374149659863946, "generation_token_f1": 0.5306122448979591, "prompt_id": "f461f81672c6f974", "reading_canonical_log_prob": -135.42864990234375, "reading_distances": [24, 27, 34, 27], "reading_log_probs": [-131.45204162597656, -132.30120849609375, -135.8720245361328, -130.16302490234375], "reading_logprob_std": 2.4466666184860673, "tokenization_path_logprob_std": 169.09541892221063, "trajectory_balance_rmse": 1891.391975956774}4{"canonical_length": 22, "canonical_nll": 0.7279287685047496, "generation_rouge_l_f1": 0.14423076923076922, "generation_token_f1": 0.14423076923076922, "prompt_id": "db2b92b024f2b9fd", "reading_canonical_log_prob": -16.014432907104492, "reading_distances": [20, 13, 17, 14], "reading_log_probs": [-23.62776756286621, -18.14665985107422, -23.78668785095215, -19.875160217285156], "reading_logprob_std": 2.802492002377508, "tokenization_path_logprob_std": 58.104934064858675, "trajectory_balance_rmse": 181.65991212671057}5{"canonical_length": 22, "canonical_nll": 2.2437083504416724, "generation_rouge_l_f1": 0.44444444444444436, "generation_token_f1": 0.44444444444444436, "prompt_id": "a3ca3e528a4c44fb", "reading_canonical_log_prob": -49.3615837097168, "reading_distances": [27, 30, 28, 27], "reading_log_probs": [-50.08561325073242, -51.33319854736328, -52.262977600097656, -51.677608489990234], "reading_logprob_std": 0.920036127670522, "tokenization_path_logprob_std": 75.48698727608443, "trajectory_balance_rmse": 655.3405011600305}6{"canonical_length": 46, "canonical_nll": 1.3379868217136548, "generation_rouge_l_f1": 0.24324324324324326, "generation_token_f1": 0.35135135135135126, "prompt_id": "bea2699cff758312", "reading_canonical_log_prob": -61.547393798828125, "reading_distances": [21, 15, 20, 25], "reading_log_probs": [-59.93962860107422, -61.71258544921875, -61.139259338378906, -58.488853454589844], "reading_logprob_std": 1.4269151040009944, "tokenization_path_logprob_std": 104.08720425569213, "trajectory_balance_rmse": 1208.0705007922395}7{"canonical_length": 41, "canonical_nll": 1.4509633692299448, "generation_rouge_l_f1": 0.2018348623853211, "generation_token_f1": 0.2110091743119266, "prompt_id": "cf9306b8e22601b5", "reading_canonical_log_prob": -59.489498138427734, "reading_distances": [16, 21, 21, 18], "reading_log_probs": [-60.593387603759766, -61.17976760864258, -62.26262283325195, -62.15650939941406], "reading_logprob_std": 0.8016336202664475, "tokenization_path_logprob_std": 27.656336351048985, "trajectory_balance_rmse": 355.02941918607024}8{"canonical_length": 231, "canonical_nll": 1.22139446043865, "generation_rouge_l_f1": 0.2336065573770492, "generation_token_f1": 0.5081967213114754, "prompt_id": "ff6c469935e35f5e", "reading_canonical_log_prob": -282.1421203613281, "reading_distances": [17, 21, 23, 17], "reading_log_probs": [-285.5497741699219, -283.59710693359375, -285.81842041015625, -279.8642578125], "reading_logprob_std": 2.7466741214007393, "tokenization_path_logprob_std": 327.7292653238587, "trajectory_balance_rmse": 5198.63038742627}9{"canonical_length": 203, "canonical_nll": 1.4411894699622845, "generation_rouge_l_f1": 0.2990033222591362, "generation_token_f1": 0.5448504983388703, "prompt_id": "41c1bf9aacb45501", "reading_canonical_log_prob": -292.56146240234375, "reading_distances": [23, 22, 21, 22], "reading_log_probs": [-308.26318359375, -299.7464599609375, -303.59967041015625, -305.3548583984375], "reading_logprob_std": 3.560469781349985, "tokenization_path_logprob_std": 130.68999970835327, "trajectory_balance_rmse": 5103.771975252342}10{"canonical_length": 33, "canonical_nll": 1.3438383160215435, "generation_rouge_l_f1": 0.1176470588235294, "generation_token_f1": 0.1176470588235294, "prompt_id": "aff22e4e1c81f916", "reading_canonical_log_prob": -44.34666442871094, "reading_distances": [25, 29, 34, 25], "reading_log_probs": [-43.83816146850586, -42.36849594116211, -44.00465393066406, -44.3736572265625], "reading_logprob_std": 0.880730775576844, "tokenization_path_logprob_std": 157.69646457136005, "trajectory_balance_rmse": 842.1142957208738}11{"canonical_length": 10, "canonical_nll": 1.3745299339294434, "generation_rouge_l_f1": 0.07179487179487179, "generation_token_f1": 0.08205128205128205, "prompt_id": "f599bb889055c49d", "reading_canonical_log_prob": -13.745299339294434, "reading_distances": [63, 53, 60, 54], "reading_log_probs": [-16.73047637939453, -23.28733253479004, -23.449031829833984, -18.390525817871094], "reading_logprob_std": 3.4215062457845096, "tokenization_path_logprob_std": 52.21980017097139, "trajectory_balance_rmse": 190.30158021330953}12{"canonical_length": 90, "canonical_nll": 1.4008102416992188, "generation_rouge_l_f1": 0.14594594594594595, "generation_token_f1": 0.2756756756756757, "prompt_id": "7fbc599a6c4c6b54", "reading_canonical_log_prob": -126.07292175292969, "reading_distances": [21, 34, 28, 33], "reading_log_probs": [-137.07778930664062, -141.08282470703125, -136.9124755859375, -135.23550415039062], "reading_logprob_std": 2.480874416765412, "tokenization_path_logprob_std": 152.63952075619554, "trajectory_balance_rmse": 1949.403003493134}13{"canonical_length": 294, "canonical_nll": 1.6962220068691538, "generation_rouge_l_f1": 0.2162162162162162, "generation_token_f1": 0.3675675675675676, "prompt_id": "9e624d36638d2bbd", "reading_canonical_log_prob": -498.68927001953125, "reading_distances": [20, 13, 22, 16], "reading_log_probs": [-491.0934753417969, -493.2012023925781, -489.84619140625, -491.46453857421875], "reading_logprob_std": 1.3852516603905454, "tokenization_path_logprob_std": 226.15448614864366, "trajectory_balance_rmse": 6454.177480840409}14{"canonical_length": 27, "canonical_nll": 0.7758024710196035, "generation_rouge_l_f1": 0.07352941176470588, "generation_token_f1": 0.08088235294117646, "prompt_id": "3982276a331d29f5", "reading_canonical_log_prob": -20.946666717529297, "reading_distances": [26, 23, 24, 25], "reading_log_probs": [-16.079570770263672, -17.354854583740234, -17.343957901000977, -17.31911849975586], "reading_logprob_std": 0.6300472863355938, "tokenization_path_logprob_std": 34.677652439108854, "trajectory_balance_rmse": 335.39319908421663}15{"canonical_length": 10, "canonical_nll": 1.2277457237243652, "generation_rouge_l_f1": 0.03053435114503817, "generation_token_f1": 0.03816793893129771, "prompt_id": "3b79dbc85baf9416", "reading_canonical_log_prob": -12.277457237243652, "reading_distances": [21, 25, 21, 19], "reading_log_probs": [-10.123486518859863, -11.248552322387695, -9.640289306640625, -11.270515441894531], "reading_logprob_std": 0.8195302270684031, "tokenization_path_logprob_std": 17.176371363568755, "trajectory_balance_rmse": 189.51031111859012}16{"canonical_length": 174, "canonical_nll": 1.5014560743309986, "generation_rouge_l_f1": 0.24713958810068645, "generation_token_f1": 0.448512585812357, "prompt_id": "bc017f1a688cb1fb", "reading_canonical_log_prob": -261.25335693359375, "reading_distances": [23, 20, 25, 28], "reading_log_probs": [-264.40093994140625, -269.5585021972656, -266.40911865234375, -265.134033203125], "reading_logprob_std": 2.2783590839397823, "tokenization_path_logprob_std": 268.8466015076654, "trajectory_balance_rmse": 4505.631366394592}17{"canonical_length": 64, "canonical_nll": 0.845356285572052, "generation_rouge_l_f1": 0.14634146341463417, "generation_token_f1": 0.20209059233449478, "prompt_id": "2060bf62bca6703a", "reading_canonical_log_prob": -54.10280227661133, "reading_distances": [17, 14, 23, 18], "reading_log_probs": [-58.909828186035156, -54.94184494018555, -56.85252380371094, -56.3802490234375], "reading_logprob_std": 1.6411100561064045, "tokenization_path_logprob_std": 70.67329766601227, "trajectory_balance_rmse": 1307.717189149067}18{"canonical_length": 12, "canonical_nll": 0.7714955806732178, "generation_rouge_l_f1": 0.42857142857142855, "generation_token_f1": 0.42857142857142855, "prompt_id": "8300e5ea4c06c29e", "reading_canonical_log_prob": -9.257946968078613, "reading_distances": [27, 29, 30, 27], "reading_log_probs": [-16.552738189697266, -16.943790435791016, -14.881085395812988, -9.917574882507324], "reading_logprob_std": 3.23047507773506, "tokenization_path_logprob_std": 16.01189973498564, "trajectory_balance_rmse": 134.04585344628944}19{"canonical_length": 142, "canonical_nll": 1.3487664880886885, "generation_rouge_l_f1": 0.14423076923076922, "generation_token_f1": 0.33653846153846156, "prompt_id": "5c25240178b5a183", "reading_canonical_log_prob": -191.52484130859375, "reading_distances": [20, 19, 20, 21], "reading_log_probs": [-211.66073608398438, -214.79019165039062, -219.22576904296875, -217.4642791748047], "reading_logprob_std": 3.299397238752619, "tokenization_path_logprob_std": 43.141548145940206, "trajectory_balance_rmse": 2830.348005627447}20{"canonical_length": 14, "canonical_nll": 1.2294213431222099, "generation_rouge_l_f1": 0.6363636363636365, "generation_token_f1": 0.6363636363636365, "prompt_id": "571e075c79645e81", "reading_canonical_log_prob": -17.211898803710938, "reading_distances": [28, 15, 24, 19], "reading_log_probs": [-20.391035079956055, -17.317710876464844, -19.563589096069336, -18.96645164489746], "reading_logprob_std": 1.299945983295593, "tokenization_path_logprob_std": 18.593561204100716, "trajectory_balance_rmse": 318.4728569548606}21{"canonical_length": 153, "canonical_nll": 1.5208706325954862, "generation_rouge_l_f1": 0.21052631578947367, "generation_token_f1": 0.4976076555023922, "prompt_id": "fc69046f95700385", "reading_canonical_log_prob": -232.69320678710938, "reading_distances": [39, 41, 39, 38], "reading_log_probs": [-244.11441040039062, -240.1039276123047, -239.96826171875, -240.55917358398438], "reading_logprob_std": 1.9682704096577965, "tokenization_path_logprob_std": 298.0368377478985, "trajectory_balance_rmse": 3658.718313421244}22{"canonical_length": 56, "canonical_nll": 1.7218502589634486, "generation_rouge_l_f1": 0.03703703703703703, "generation_token_f1": 0.07407407407407406, "prompt_id": "fcbb32fd544b170a", "reading_canonical_log_prob": -96.42361450195312, "reading_distances": [24, 22, 24, 25], "reading_log_probs": [-101.10171508789062, -99.61028289794922, -100.95167541503906, -98.7996826171875], "reading_logprob_std": 1.10429898875844, "tokenization_path_logprob_std": 65.67758114209133, "trajectory_balance_rmse": 1129.05928123398}23{"canonical_length": 83, "canonical_nll": 0.8529988484210279, "generation_rouge_l_f1": 0.3225806451612903, "generation_token_f1": 0.4331797235023041, "prompt_id": "f6d789446e38536d", "reading_canonical_log_prob": -70.79890441894531, "reading_distances": [8, 7, 8, 9], "reading_log_probs": [-72.28863525390625, -70.06280517578125, -70.4365463256836, -69.06256866455078], "reading_logprob_std": 1.3484567942808838, "tokenization_path_logprob_std": 157.62826792129223, "trajectory_balance_rmse": 1771.7513881017837}24{"canonical_length": 270, "canonical_nll": 1.332774070457176, "generation_rouge_l_f1": 0.14382022471910114, "generation_token_f1": 0.2831460674157303, "prompt_id": "86891b80d6ccb140", "reading_canonical_log_prob": -359.8489990234375, "reading_distances": [20, 24, 27, 23], "reading_log_probs": [-361.638427734375, -367.2666015625, -360.9367980957031, -370.4036865234375], "reading_logprob_std": 4.550898051596304, "tokenization_path_logprob_std": 105.7928460822157, "trajectory_balance_rmse": 3543.5057797313793}25{"canonical_length": 27, "canonical_nll": 1.6250678168402777, "generation_rouge_l_f1": 0.35789473684210527, "generation_token_f1": 0.4210526315789474, "prompt_id": "01a2b8a844748c41", "reading_canonical_log_prob": -43.8768310546875, "reading_distances": [33, 37, 33, 36], "reading_log_probs": [-41.667518615722656, -42.79216003417969, -42.820396423339844, -44.52349090576172], "reading_logprob_std": 1.177899188438816, "tokenization_path_logprob_std": 23.91034541576919, "trajectory_balance_rmse": 660.0470592876661}26{"canonical_length": 66, "canonical_nll": 1.1378740252870503, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "fc8d39645d6796a9", "reading_canonical_log_prob": -75.09968566894531, "reading_distances": [28, 32, 30, 26], "reading_log_probs": [-84.13359069824219, -82.9926528930664, -84.29202270507812, -78.40328979492188], "reading_logprob_std": 2.762712722741883, "tokenization_path_logprob_std": 61.33859127518875, "trajectory_balance_rmse": 1574.0843312028092}27{"canonical_length": 340, "canonical_nll": 1.227778715245864, "generation_rouge_l_f1": 0.22185970636215335, "generation_token_f1": 0.4469820554649266, "prompt_id": "7a3c4e3465a24fae", "reading_canonical_log_prob": -417.44476318359375, "reading_distances": [28, 28, 21, 20], "reading_log_probs": [-434.563720703125, -433.4266662597656, -432.45672607421875, -431.85205078125], "reading_logprob_std": 1.1857242167149433, "tokenization_path_logprob_std": 270.48772320895137, "trajectory_balance_rmse": 7164.407659105661}28{"canonical_length": 22, "canonical_nll": 0.6750792590054598, "generation_rouge_l_f1": 0.8333333333333334, "generation_token_f1": 0.8333333333333334, "prompt_id": "9652191d0c5b1125", "reading_canonical_log_prob": -14.851743698120117, "reading_distances": [20, 17, 15, 23], "reading_log_probs": [-17.177465438842773, -24.283308029174805, -22.751739501953125, -23.871257781982422], "reading_logprob_std": 3.2931898786412677, "tokenization_path_logprob_std": 46.424743745904486, "trajectory_balance_rmse": 418.57524946722225}29{"canonical_length": 229, "canonical_nll": 1.0042640652719024, "generation_rouge_l_f1": 0.29661016949152547, "generation_token_f1": 0.5127118644067797, "prompt_id": "1fb3dd895fb36ca1", "reading_canonical_log_prob": -229.97647094726562, "reading_distances": [46, 47, 47, 53], "reading_log_probs": [-228.78228759765625, -227.33395385742188, -230.16900634765625, -230.35150146484375], "reading_logprob_std": 1.4041460135646893, "tokenization_path_logprob_std": 205.74387005847728, "trajectory_balance_rmse": 4936.589344213509}30{"canonical_length": 84, "canonical_nll": 0.24569852011544363, "generation_rouge_l_f1": 0.5217391304347826, "generation_token_f1": 0.6521739130434783, "prompt_id": "fba6cc8156910498", "reading_canonical_log_prob": -20.638675689697266, "reading_distances": [22, 14, 24, 27], "reading_log_probs": [-24.706499099731445, -21.638282775878906, -22.54901885986328, -24.702205657958984], "reading_logprob_std": 1.5524702857735513, "tokenization_path_logprob_std": 195.84893710736944, "trajectory_balance_rmse": 1476.6136564014707}31{"canonical_length": 24, "canonical_nll": 0.5704621076583862, "generation_rouge_l_f1": 0.13793103448275862, "generation_token_f1": 0.13793103448275862, "prompt_id": "02b01e693f3f83a0", "reading_canonical_log_prob": -13.69109058380127, "reading_distances": [34, 32, 25, 24], "reading_log_probs": [-11.166200637817383, -11.215022087097168, -10.601247787475586, -11.135313034057617], "reading_logprob_std": 0.28734523554780333, "tokenization_path_logprob_std": 56.136266458109425, "trajectory_balance_rmse": 372.015656283551}32{"canonical_length": 331, "canonical_nll": 0.8866763388642371, "generation_rouge_l_f1": 0.2596491228070175, "generation_token_f1": 0.5473684210526316, "prompt_id": "40bc9c44485ded59", "reading_canonical_log_prob": -293.4898681640625, "reading_distances": [27, 34, 28, 33], "reading_log_probs": [-304.820556640625, -305.47393798828125, -305.040771484375, -305.97601318359375], "reading_logprob_std": 0.5102998582022023, "tokenization_path_logprob_std": 197.3292615416616, "trajectory_balance_rmse": 4693.584315825606}33{"canonical_length": 11, "canonical_nll": 1.8179811997847124, "generation_rouge_l_f1": 0.3, "generation_token_f1": 0.3, "prompt_id": "f0c9ccf7c1a3a5c6", "reading_canonical_log_prob": -19.997793197631836, "reading_distances": [22, 26, 27, 28], "reading_log_probs": [-18.434232711791992, -20.299884796142578, -20.90508460998535, -20.831409454345703], "reading_logprob_std": 1.1542110019113936, "tokenization_path_logprob_std": 39.75180634155779, "trajectory_balance_rmse": 263.1295894079302}34{"canonical_length": 149, "canonical_nll": 1.9166034468068371, "generation_rouge_l_f1": 0.23940149625935164, "generation_token_f1": 0.4139650872817956, "prompt_id": "7b4b39937f27934d", "reading_canonical_log_prob": -285.57391357421875, "reading_distances": [27, 27, 23, 30], "reading_log_probs": [-282.19049072265625, -284.7328186035156, -282.2171630859375, -283.1026306152344], "reading_logprob_std": 1.1925543299324357, "tokenization_path_logprob_std": 126.5330486718361, "trajectory_balance_rmse": 3403.248526333784}35{"canonical_length": 10, "canonical_nll": 0.9396727561950684, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.7777777777777778, "prompt_id": "177155c261ac8893", "reading_canonical_log_prob": -9.396727561950684, "reading_distances": [18, 15, 11, 15], "reading_log_probs": [-12.289041519165039, -10.631338119506836, -8.5715913772583, -10.48171329498291], "reading_logprob_std": 1.520618333650693, "tokenization_path_logprob_std": 45.334042071787565, "trajectory_balance_rmse": 205.0905158262058}36{"canonical_length": 9, "canonical_nll": 0.4697124163309733, "generation_rouge_l_f1": 0.6666666666666665, "generation_token_f1": 0.7777777777777777, "prompt_id": "316f7da66f20e71c", "reading_canonical_log_prob": -4.22741174697876, "reading_distances": [31, 31, 30, 26], "reading_log_probs": [-4.425214767456055, -5.053736209869385, -4.718759536743164, -4.704432964324951], "reading_logprob_std": 0.2571636826639979, "tokenization_path_logprob_std": 39.97814127796114, "trajectory_balance_rmse": 204.53128934722557}37{"canonical_length": 10, "canonical_nll": 1.5975896835327148, "generation_rouge_l_f1": 0.17977528089887643, "generation_token_f1": 0.20224719101123595, "prompt_id": "4c771fd22dee11bd", "reading_canonical_log_prob": -15.975896835327148, "reading_distances": [25, 31, 29, 30], "reading_log_probs": [-15.922562599182129, -18.94124984741211, -17.85403060913086, -19.243030548095703], "reading_logprob_std": 1.5019705631181715, "tokenization_path_logprob_std": 42.64066578607948, "trajectory_balance_rmse": 287.3071747815683}38{"canonical_length": 150, "canonical_nll": 1.1588053385416666, "generation_rouge_l_f1": 0.1848341232227488, "generation_token_f1": 0.3744075829383886, "prompt_id": "fb7e151d45f5c282", "reading_canonical_log_prob": -173.82080078125, "reading_distances": [37, 33, 38, 39], "reading_log_probs": [-182.91925048828125, -184.49581909179688, -184.72935485839844, -182.03643798828125], "reading_logprob_std": 1.2876431939282993, "tokenization_path_logprob_std": 123.00500859381785, "trajectory_balance_rmse": 2947.655309016205}39{"canonical_length": 186, "canonical_nll": 1.1706890803511425, "generation_rouge_l_f1": 0.2583732057416268, "generation_token_f1": 0.4736842105263158, "prompt_id": "c6ce7ec89de605a5", "reading_canonical_log_prob": -217.7481689453125, "reading_distances": [17, 19, 20, 19], "reading_log_probs": [-221.7207794189453, -215.6871795654297, -213.72265625, -218.21986389160156], "reading_logprob_std": 3.4536144022379314, "tokenization_path_logprob_std": 272.7170111869327, "trajectory_balance_rmse": 4391.604296496903}40{"canonical_length": 220, "canonical_nll": 1.005317549272017, "generation_rouge_l_f1": 0.33472803347280333, "generation_token_f1": 0.5564853556485355, "prompt_id": "320fa6e957218c02", "reading_canonical_log_prob": -221.16986083984375, "reading_distances": [14, 13, 14, 13], "reading_log_probs": [-229.19802856445312, -220.642822265625, -227.8655548095703, -231.6480255126953], "reading_logprob_std": 4.730741438541889, "tokenization_path_logprob_std": 192.23185377336532, "trajectory_balance_rmse": 4264.228768831172}41{"canonical_length": 58, "canonical_nll": 1.1937336099558864, "generation_rouge_l_f1": 0.04545454545454545, "generation_token_f1": 0.04545454545454545, "prompt_id": "d3de64d71666f0d5", "reading_canonical_log_prob": -69.2365493774414, "reading_distances": [22, 20, 23, 23], "reading_log_probs": [-75.10354614257812, -77.43204498291016, -86.99222564697266, -75.92497253417969], "reading_logprob_std": 5.504466078273172, "tokenization_path_logprob_std": 63.0700815330257, "trajectory_balance_rmse": 930.0715923227201}42{"canonical_length": 11, "canonical_nll": 0.11106746846979315, "generation_rouge_l_f1": 0.6, "generation_token_f1": 0.9, "prompt_id": "bd4efb1cb33c7494", "reading_canonical_log_prob": -1.2217421531677246, "reading_distances": [44, 47, 47, 43], "reading_log_probs": [-2.4241766929626465, -2.865818738937378, -3.492772340774536, -3.059509515762329], "reading_logprob_std": 0.4433673949111005, "tokenization_path_logprob_std": 36.40830884216114, "trajectory_balance_rmse": 188.49520761358417}43{"canonical_length": 38, "canonical_nll": 0.6213798522949219, "generation_rouge_l_f1": 0.18357487922705315, "generation_token_f1": 0.27053140096618356, "prompt_id": "d19ca6fde647262d", "reading_canonical_log_prob": -23.61243438720703, "reading_distances": [17, 15, 15, 17], "reading_log_probs": [-25.39507293701172, -24.19778060913086, -25.550437927246094, -24.12137222290039], "reading_logprob_std": 0.7614519878054649, "tokenization_path_logprob_std": 68.95746257396517, "trajectory_balance_rmse": 966.0253862748135}44{"canonical_length": 88, "canonical_nll": 1.3089749596335671, "generation_rouge_l_f1": 0.2751322751322752, "generation_token_f1": 0.5291005291005292, "prompt_id": "b0e33c086e3feda9", "reading_canonical_log_prob": -115.1897964477539, "reading_distances": [14, 17, 19, 20], "reading_log_probs": [-119.43267059326172, -122.50727081298828, -120.7280044555664, -122.02059936523438], "reading_logprob_std": 1.3814808215580192, "tokenization_path_logprob_std": 88.85382649781239, "trajectory_balance_rmse": 1814.0977992990863}45{"canonical_length": 53, "canonical_nll": 0.9078232387326798, "generation_rouge_l_f1": 0.46511627906976744, "generation_token_f1": 0.7674418604651162, "prompt_id": "85224f0bc07faa4c", "reading_canonical_log_prob": -48.11463165283203, "reading_distances": [18, 20, 17, 19], "reading_log_probs": [-49.24110412597656, -54.89569091796875, -49.67790985107422, -51.797706604003906], "reading_logprob_std": 2.582273733321932, "tokenization_path_logprob_std": 100.03850354511165, "trajectory_balance_rmse": 1062.7700177608617}46{"canonical_length": 26, "canonical_nll": 1.357811267559345, "generation_rouge_l_f1": 0.4210526315789474, "generation_token_f1": 0.4912280701754386, "prompt_id": "7327a63351d4e2c9", "reading_canonical_log_prob": -35.30309295654297, "reading_distances": [40, 29, 31, 36], "reading_log_probs": [-43.78435516357422, -38.98894500732422, -37.598968505859375, -41.54953384399414], "reading_logprob_std": 2.7438569168011027, "tokenization_path_logprob_std": 93.98915077608285, "trajectory_balance_rmse": 614.3653482881413}47{"canonical_length": 84, "canonical_nll": 1.395102092197963, "generation_rouge_l_f1": 0.31724137931034485, "generation_token_f1": 0.5379310344827586, "prompt_id": "7fabfcdc4a3bff58", "reading_canonical_log_prob": -117.1885757446289, "reading_distances": [29, 25, 32, 33], "reading_log_probs": [-118.19733428955078, -118.39522552490234, -118.77702331542969, -119.51531219482422], "reading_logprob_std": 0.581499515403082, "tokenization_path_logprob_std": 135.86680451922382, "trajectory_balance_rmse": 1846.088944593068}48{"canonical_length": 13, "canonical_nll": 0.6232903553889348, "generation_rouge_l_f1": 0.14285714285714285, "generation_token_f1": 0.14285714285714285, "prompt_id": "173c8d85c343f33f", "reading_canonical_log_prob": -8.102774620056152, "reading_distances": [24, 32, 28, 32], "reading_log_probs": [-7.360267639160156, -6.12272834777832, -5.301791191101074, -8.886127471923828], "reading_logprob_std": 1.5613790945489274, "tokenization_path_logprob_std": 54.275459024437104, "trajectory_balance_rmse": 211.45364951429073}49{"canonical_length": 118, "canonical_nll": 1.4928612143306408, "generation_rouge_l_f1": 0.26063829787234044, "generation_token_f1": 0.48936170212765956, "prompt_id": "e1f635b187b10d1b", "reading_canonical_log_prob": -176.15762329101562, "reading_distances": [29, 26, 23, 20], "reading_log_probs": [-177.0896453857422, -174.4681396484375, -176.765869140625, -175.94068908691406], "reading_logprob_std": 1.1699753826936081, "tokenization_path_logprob_std": 78.79068290000521, "trajectory_balance_rmse": 2348.703663968562}50{"canonical_length": 242, "canonical_nll": 0.8136358339924458, "generation_rouge_l_f1": 0.39917695473251025, "generation_token_f1": 0.5679012345679012, "prompt_id": "5ff715a6459080e1", "reading_canonical_log_prob": -196.89987182617188, "reading_distances": [20, 30, 28, 26], "reading_log_probs": [-187.7091064453125, -193.48721313476562, -189.03094482421875, -191.59136962890625], "reading_logprob_std": 2.5854410661125184, "tokenization_path_logprob_std": 146.54608657862562, "trajectory_balance_rmse": 2983.6055446860587}51{"canonical_length": 147, "canonical_nll": 0.7271813632679635, "generation_rouge_l_f1": 0.22023809523809526, "generation_token_f1": 0.3214285714285714, "prompt_id": "3e82cf9b74d0c14b", "reading_canonical_log_prob": -106.89566040039062, "reading_distances": [24, 26, 25, 23], "reading_log_probs": [-110.21884155273438, -108.0694351196289, -109.56341552734375, -109.22599029541016], "reading_logprob_std": 0.8999415735207692, "tokenization_path_logprob_std": 136.83071151463213, "trajectory_balance_rmse": 1657.0238988452584}52{"canonical_length": 42, "canonical_nll": 1.1982054937453497, "generation_rouge_l_f1": 0.34782608695652173, "generation_token_f1": 0.5, "prompt_id": "b1354ebe20eb5cb8", "reading_canonical_log_prob": -50.32463073730469, "reading_distances": [46, 54, 47, 50], "reading_log_probs": [-66.8611831665039, -64.5502700805664, -63.392032623291016, -66.43626403808594], "reading_logprob_std": 1.6258770241008185, "tokenization_path_logprob_std": 42.173977833990804, "trajectory_balance_rmse": 864.7905741268034}53{"canonical_length": 124, "canonical_nll": 1.1890359693957913, "generation_rouge_l_f1": 0.18592964824120603, "generation_token_f1": 0.37185929648241206, "prompt_id": "751bafc67843f928", "reading_canonical_log_prob": -147.44046020507812, "reading_distances": [31, 31, 34, 36], "reading_log_probs": [-151.4251708984375, -155.15557861328125, -157.4653778076172, -155.28384399414062], "reading_logprob_std": 2.5066596862091606, "tokenization_path_logprob_std": 67.12314533327023, "trajectory_balance_rmse": 2733.134758166058}54{"canonical_length": 283, "canonical_nll": 1.1704819750027606, "generation_rouge_l_f1": 0.17482517482517484, "generation_token_f1": 0.3286713286713287, "prompt_id": "2c0f185491268c7a", "reading_canonical_log_prob": -331.24639892578125, "reading_distances": [23, 30, 38, 21], "reading_log_probs": [-327.4717712402344, -328.46441650390625, -325.13885498046875, -329.3976745605469], "reading_logprob_std": 1.8304102626612873, "tokenization_path_logprob_std": 164.0094983574784, "trajectory_balance_rmse": 4476.261145969822}55{"canonical_length": 384, "canonical_nll": 1.4805256525675456, "generation_rouge_l_f1": 0.22259136212624586, "generation_token_f1": 0.3853820598006645, "prompt_id": "d93394062daf791b", "reading_canonical_log_prob": -568.5218505859375, "reading_distances": [25, 28, 29, 25], "reading_log_probs": [-583.126220703125, -579.2742919921875, -572.9810791015625, -572.6715698242188], "reading_logprob_std": 5.085577994163089, "tokenization_path_logprob_std": 161.19802467356953, "trajectory_balance_rmse": 7176.613365851401}56{"canonical_length": 13, "canonical_nll": 1.5513904278094952, "generation_rouge_l_f1": 0.24999999999999994, "generation_token_f1": 0.41666666666666663, "prompt_id": "1bc766d97da83b42", "reading_canonical_log_prob": -20.168075561523438, "reading_distances": [20, 20, 19, 19], "reading_log_probs": [-23.757640838623047, -22.195987701416016, -22.63919448852539, -20.44833755493164], "reading_logprob_std": 1.3751284296557096, "tokenization_path_logprob_std": 38.24004572358623, "trajectory_balance_rmse": 295.01103371759683}57{"canonical_length": 5, "canonical_nll": 0.49666194915771483, "generation_rouge_l_f1": 0.9090909090909091, "generation_token_f1": 0.9090909090909091, "prompt_id": "03cbf10fd71003f1", "reading_canonical_log_prob": -2.483309745788574, "reading_distances": [27, 20, 23, 23], "reading_log_probs": [-3.065438747406006, -3.670248508453369, -3.6780107021331787, -3.263700485229492], "reading_logprob_std": 0.3051422436244518, "tokenization_path_logprob_std": 24.325359091211247, "trajectory_balance_rmse": 103.29419923624981}58{"canonical_length": 218, "canonical_nll": 1.336617006074398, "generation_rouge_l_f1": 0.2292134831460674, "generation_token_f1": 0.4629213483146068, "prompt_id": "3b33c4e6381082ec", "reading_canonical_log_prob": -291.38250732421875, "reading_distances": [23, 20, 19, 25], "reading_log_probs": [-293.674072265625, -296.6055908203125, -294.0398864746094, -294.10089111328125], "reading_logprob_std": 1.346906219780117, "tokenization_path_logprob_std": 158.0049547447675, "trajectory_balance_rmse": 3739.1784889533415}59{"canonical_length": 14, "canonical_nll": 0.8792092459542411, "generation_rouge_l_f1": 0.5185185185185186, "generation_token_f1": 0.6666666666666666, "prompt_id": "275fffbea930b52f", "reading_canonical_log_prob": -12.308929443359375, "reading_distances": [12, 17, 19, 17], "reading_log_probs": [-25.22675895690918, -28.32659912109375, -30.61043930053711, -30.13938331604004], "reading_logprob_std": 2.4401316122803856, "tokenization_path_logprob_std": 13.955170831601494, "trajectory_balance_rmse": 287.5199774571916}60{"canonical_length": 181, "canonical_nll": 0.9791215085192938, "generation_rouge_l_f1": 0.2127659574468085, "generation_token_f1": 0.5295508274231678, "prompt_id": "a7f0f685867c31e2", "reading_canonical_log_prob": -177.2209930419922, "reading_distances": [81, 74, 72, 75], "reading_log_probs": [-197.41781616210938, -205.66775512695312, -202.90750122070312, -204.4287872314453], "reading_logprob_std": 3.6379962949038243, "tokenization_path_logprob_std": 116.50228580197404, "trajectory_balance_rmse": 2829.579150823272}61{"canonical_length": 128, "canonical_nll": 1.1345092058181763, "generation_rouge_l_f1": 0.22572178477690288, "generation_token_f1": 0.42519685039370075, "prompt_id": "858a5416234d0d2a", "reading_canonical_log_prob": -145.21717834472656, "reading_distances": [19, 21, 21, 19], "reading_log_probs": [-163.4263458251953, -165.78042602539062, -159.54144287109375, -158.85894775390625], "reading_logprob_std": 3.2760858446802157, "tokenization_path_logprob_std": 39.087014621253424, "trajectory_balance_rmse": 2796.5945285144326}62{"canonical_length": 205, "canonical_nll": 0.8387918611852134, "generation_rouge_l_f1": 0.3378016085790884, "generation_token_f1": 0.5308310991957105, "prompt_id": "17197512e608a44f", "reading_canonical_log_prob": -171.95233154296875, "reading_distances": [16, 14, 12, 13], "reading_log_probs": [-196.851318359375, -183.16036987304688, -177.20553588867188, -195.59117126464844], "reading_logprob_std": 9.587332220344615, "tokenization_path_logprob_std": 200.1304381327198, "trajectory_balance_rmse": 3790.2487714978506}63{"canonical_length": 160, "canonical_nll": 1.540947151184082, "generation_rouge_l_f1": 0.2099737532808399, "generation_token_f1": 0.39370078740157477, "prompt_id": "7c90d4dc6f1c3493", "reading_canonical_log_prob": -246.55154418945312, "reading_distances": [22, 24, 15, 17], "reading_log_probs": [-251.52842712402344, -253.31747436523438, -246.74337768554688, -248.58580017089844], "reading_logprob_std": 2.9404995323209815, "tokenization_path_logprob_std": 65.21103653714212, "trajectory_balance_rmse": 3315.540905336944}64{"canonical_length": 23, "canonical_nll": 4.225556083347486, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "d1e684b980513357", "reading_canonical_log_prob": -97.18778991699219, "reading_distances": [9, 19, 23, 13], "reading_log_probs": [-100.49637603759766, -95.93827819824219, -93.23497009277344, -95.8656997680664], "reading_logprob_std": 3.0163625429750125, "tokenization_path_logprob_std": 46.65136401557521, "trajectory_balance_rmse": 379.5198730685012}65{"canonical_length": 392, "canonical_nll": 1.0270040862414302, "generation_rouge_l_f1": 0.2385321100917431, "generation_token_f1": 0.47094801223241595, "prompt_id": "5b2aa008c5db43fd", "reading_canonical_log_prob": -402.5856018066406, "reading_distances": [7, 12, 10, 9], "reading_log_probs": [-399.9477233886719, -397.9456787109375, -397.9828796386719, -394.0704345703125], "reading_logprob_std": 2.462001983132376, "tokenization_path_logprob_std": 264.42522258033716, "trajectory_balance_rmse": 7205.265904800515}66{"canonical_length": 82, "canonical_nll": 0.9839157476657774, "generation_rouge_l_f1": 0.20886075949367086, "generation_token_f1": 0.310126582278481, "prompt_id": "180245122ca6704c", "reading_canonical_log_prob": -80.68109130859375, "reading_distances": [29, 24, 35, 31], "reading_log_probs": [-97.88654327392578, -95.59281921386719, -94.1202163696289, -88.53141784667969], "reading_logprob_std": 3.981534225947791, "tokenization_path_logprob_std": 70.20459837039209, "trajectory_balance_rmse": 1178.0327010758144}67{"canonical_length": 99, "canonical_nll": 2.084275833284012, "generation_rouge_l_f1": 0.16201117318435754, "generation_token_f1": 0.2681564245810056, "prompt_id": "0280ead696ae0c9e", "reading_canonical_log_prob": -206.3433074951172, "reading_distances": [34, 24, 24, 32], "reading_log_probs": [-206.44024658203125, -208.23870849609375, -206.15835571289062, -209.02517700195312], "reading_logprob_std": 1.3892701234517393, "tokenization_path_logprob_std": 173.79026768904498, "trajectory_balance_rmse": 2685.395655797215}68{"canonical_length": 148, "canonical_nll": 1.5668788600612331, "generation_rouge_l_f1": 0.2032967032967033, "generation_token_f1": 0.3131868131868132, "prompt_id": "1d5af645718b1db0", "reading_canonical_log_prob": -231.8980712890625, "reading_distances": [58, 46, 51, 52], "reading_log_probs": [-236.31564331054688, -229.95330810546875, -239.28570556640625, -240.66769409179688], "reading_logprob_std": 4.761324454964223, "tokenization_path_logprob_std": 110.68328778424484, "trajectory_balance_rmse": 2543.1963900569044}69{"canonical_length": 120, "canonical_nll": 1.4862744649251303, "generation_rouge_l_f1": 0.321285140562249, "generation_token_f1": 0.4417670682730923, "prompt_id": "0211ac7bc9c6651a", "reading_canonical_log_prob": -178.35293579101562, "reading_distances": [17, 21, 20, 24], "reading_log_probs": [-208.4502716064453, -203.91168212890625, -207.02938842773438, -213.82931518554688], "reading_logprob_std": 4.141963656715089, "tokenization_path_logprob_std": 129.25618098781686, "trajectory_balance_rmse": 2482.5786580563017}70{"canonical_length": 115, "canonical_nll": 0.6067611694335937, "generation_rouge_l_f1": 0.2857142857142857, "generation_token_f1": 0.4651162790697674, "prompt_id": "0da3d3229c8dfac7", "reading_canonical_log_prob": -69.77753448486328, "reading_distances": [14, 15, 19, 15], "reading_log_probs": [-73.39289855957031, -75.1070785522461, -74.85179901123047, -71.59677124023438], "reading_logprob_std": 1.6144015346558234, "tokenization_path_logprob_std": 148.0888114054629, "trajectory_balance_rmse": 1495.7746400872397}71{"canonical_length": 196, "canonical_nll": 1.6112988530373087, "generation_rouge_l_f1": 0.22675736961451246, "generation_token_f1": 0.46258503401360546, "prompt_id": "dd26708537c260a2", "reading_canonical_log_prob": -315.8145751953125, "reading_distances": [35, 38, 37, 36], "reading_log_probs": [-318.1626892089844, -318.0522155761719, -315.6744689941406, -315.1344299316406], "reading_logprob_std": 1.5767210854793505, "tokenization_path_logprob_std": 229.45270775720354, "trajectory_balance_rmse": 4927.868856595305}72{"canonical_length": 95, "canonical_nll": 1.6347175999691612, "generation_rouge_l_f1": 0.2613065326633166, "generation_token_f1": 0.4824120603015075, "prompt_id": "9fc2e86bc1fbb199", "reading_canonical_log_prob": -155.2981719970703, "reading_distances": [13, 14, 14, 17], "reading_log_probs": [-158.4520721435547, -158.71730041503906, -157.61068725585938, -155.92788696289062], "reading_logprob_std": 1.2578759315022796, "tokenization_path_logprob_std": 114.69322209042274, "trajectory_balance_rmse": 2488.778435723449}73{"canonical_length": 178, "canonical_nll": 1.185450693194786, "generation_rouge_l_f1": 0.26146788990825687, "generation_token_f1": 0.4862385321100918, "prompt_id": "0d0deb69605ca531", "reading_canonical_log_prob": -211.01022338867188, "reading_distances": [17, 19, 18, 14], "reading_log_probs": [-224.35423278808594, -219.05361938476562, -219.63449096679688, -227.93212890625], "reading_logprob_std": 4.195135520612033, "tokenization_path_logprob_std": 104.69314772217625, "trajectory_balance_rmse": 4206.438360112152}74{"canonical_length": 20, "canonical_nll": 0.6773427963256836, "generation_rouge_l_f1": 0.42857142857142855, "generation_token_f1": 0.46428571428571425, "prompt_id": "5b98bbd785876818", "reading_canonical_log_prob": -13.546855926513672, "reading_distances": [25, 35, 29, 29], "reading_log_probs": [-14.239872932434082, -15.337894439697266, -14.023865699768066, -14.055354118347168], "reading_logprob_std": 0.623092440538984, "tokenization_path_logprob_std": 6.014026538279708, "trajectory_balance_rmse": 246.93956302206226}75{"canonical_length": 322, "canonical_nll": 1.0143533078780085, "generation_rouge_l_f1": 0.1657848324514991, "generation_token_f1": 0.3492063492063492, "prompt_id": "9d8ebc1fd91528b6", "reading_canonical_log_prob": -326.62176513671875, "reading_distances": [21, 16, 23, 24], "reading_log_probs": [-335.62872314453125, -331.0585632324219, -334.64306640625, -336.54925537109375], "reading_logprob_std": 2.4037328599260643, "tokenization_path_logprob_std": 193.67718099009556, "trajectory_balance_rmse": 5829.973256754019}76{"canonical_length": 202, "canonical_nll": 1.1862169020246751, "generation_rouge_l_f1": 0.26914153132250584, "generation_token_f1": 0.4965197215777263, "prompt_id": "5f17b0be639454cf", "reading_canonical_log_prob": -239.61581420898438, "reading_distances": [20, 15, 22, 24], "reading_log_probs": [-233.71585083007812, -237.51718139648438, -238.7593994140625, -234.63101196289062], "reading_logprob_std": 2.3741959839546367, "tokenization_path_logprob_std": 255.10166249544721, "trajectory_balance_rmse": 4274.687750337663}77{"canonical_length": 11, "canonical_nll": 0.38014936447143555, "generation_rouge_l_f1": 0.9, "generation_token_f1": 0.9, "prompt_id": "1932b722766373ae", "reading_canonical_log_prob": -4.181643009185791, "reading_distances": [39, 43, 40, 40], "reading_log_probs": [-5.512370586395264, -7.392113208770752, -6.305584907531738, -7.202442646026611], "reading_logprob_std": 0.8679399617888697, "tokenization_path_logprob_std": 9.482893794193625, "trajectory_balance_rmse": 180.35993970348247}78{"canonical_length": 31, "canonical_nll": 1.7752243780320691, "generation_rouge_l_f1": 0.08843537414965986, "generation_token_f1": 0.16326530612244897, "prompt_id": "3f8935161866e686", "reading_canonical_log_prob": -55.03195571899414, "reading_distances": [29, 21, 32, 27], "reading_log_probs": [-61.71436309814453, -62.52757263183594, -55.15277099609375, -71.20614624023438], "reading_logprob_std": 6.590561076569817, "tokenization_path_logprob_std": 71.73365829409585, "trajectory_balance_rmse": 556.1202112054387}79{"canonical_length": 45, "canonical_nll": 1.067165035671658, "generation_rouge_l_f1": 0.4473684210526316, "generation_token_f1": 0.6315789473684211, "prompt_id": "88f54467af0c5001", "reading_canonical_log_prob": -48.02242660522461, "reading_distances": [23, 28, 28, 26], "reading_log_probs": [-50.729270935058594, -55.408531188964844, -50.14244079589844, -49.703857421875], "reading_logprob_std": 2.6419482268752628, "tokenization_path_logprob_std": 40.85013061901521, "trajectory_balance_rmse": 811.062279352168}80{"canonical_length": 272, "canonical_nll": 1.3916435241699219, "generation_rouge_l_f1": 0.2433460076045627, "generation_token_f1": 0.5171102661596959, "prompt_id": "05940c01ebeedfd9", "reading_canonical_log_prob": -378.52703857421875, "reading_distances": [21, 29, 30, 24], "reading_log_probs": [-381.76226806640625, -385.24176025390625, -388.5406494140625, -385.7702331542969], "reading_logprob_std": 2.7831975506704403, "tokenization_path_logprob_std": 240.64523000992148, "trajectory_balance_rmse": 6764.68129008645}81{"canonical_length": 26, "canonical_nll": 1.1543677403376653, "generation_rouge_l_f1": 0.16216216216216214, "generation_token_f1": 0.16216216216216214, "prompt_id": "cc3b11b7124977c4", "reading_canonical_log_prob": -30.013561248779297, "reading_distances": [22, 25, 18, 24], "reading_log_probs": [-36.12590789794922, -34.66185760498047, -34.12157440185547, -30.603363037109375], "reading_logprob_std": 2.3416588327497774, "tokenization_path_logprob_std": 75.30868284857581, "trajectory_balance_rmse": 540.7789191322325}82{"canonical_length": 25, "canonical_nll": 0.9942060852050781, "generation_rouge_l_f1": 0.7741935483870969, "generation_token_f1": 0.7741935483870969, "prompt_id": "9335f10e0202a2ca", "reading_canonical_log_prob": -24.855152130126953, "reading_distances": [21, 24, 28, 25], "reading_log_probs": [-28.384748458862305, -27.661962509155273, -27.434612274169922, -28.467510223388672], "reading_logprob_std": 0.5163575336736211, "tokenization_path_logprob_std": 15.48587627223928, "trajectory_balance_rmse": 259.6699108439691}83{"canonical_length": 160, "canonical_nll": 1.262875461578369, "generation_rouge_l_f1": 0.27338129496402874, "generation_token_f1": 0.43165467625899284, "prompt_id": "ee0b55cb338c65fa", "reading_canonical_log_prob": -202.06007385253906, "reading_distances": [31, 27, 24, 29], "reading_log_probs": [-220.13034057617188, -226.61351013183594, -231.57156372070312, -217.68624877929688], "reading_logprob_std": 6.298061347951014, "tokenization_path_logprob_std": 208.66881313788807, "trajectory_balance_rmse": 3212.0805910124695}84{"canonical_length": 205, "canonical_nll": 0.47595977783203125, "generation_rouge_l_f1": 0.37280701754385964, "generation_token_f1": 0.5833333333333333, "prompt_id": "9ed2cebe3982e4a6", "reading_canonical_log_prob": -97.5717544555664, "reading_distances": [11, 9, 12, 13], "reading_log_probs": [-94.90589904785156, -95.7464828491211, -102.68312072753906, -99.24523162841797], "reading_logprob_std": 3.5613531482840557, "tokenization_path_logprob_std": 236.12776391427755, "trajectory_balance_rmse": 2020.159938430683}85{"canonical_length": 67, "canonical_nll": 1.5109899720149254, "generation_rouge_l_f1": 0.2938388625592417, "generation_token_f1": 0.4360189573459715, "prompt_id": "57c5e7bb82e63751", "reading_canonical_log_prob": -101.236328125, "reading_distances": [25, 31, 31, 31], "reading_log_probs": [-100.09224700927734, -99.27998352050781, -103.1185302734375, -102.28221893310547], "reading_logprob_std": 1.8041945271674047, "tokenization_path_logprob_std": 139.02338005817077, "trajectory_balance_rmse": 1371.0814768683476}86{"canonical_length": 233, "canonical_nll": 1.2846656111688572, "generation_rouge_l_f1": 0.27642276422764234, "generation_token_f1": 0.5447154471544715, "prompt_id": "2049d32643760e75", "reading_canonical_log_prob": -299.32708740234375, "reading_distances": [33, 23, 35, 30], "reading_log_probs": [-304.79461669921875, -309.26898193359375, -306.9161376953125, -308.32171630859375], "reading_logprob_std": 1.9444379362242683, "tokenization_path_logprob_std": 263.26287052025134, "trajectory_balance_rmse": 4608.354092720017}87{"canonical_length": 65, "canonical_nll": 1.2572606013371395, "generation_rouge_l_f1": 0.2649006622516556, "generation_token_f1": 0.43708609271523186, "prompt_id": "9327dc9ed2ef5314", "reading_canonical_log_prob": -81.72193908691406, "reading_distances": [14, 17, 18, 19], "reading_log_probs": [-73.07252502441406, -76.3521957397461, -77.89594268798828, -73.01271057128906], "reading_logprob_std": 2.4393723144694968, "tokenization_path_logprob_std": 84.79276333804067, "trajectory_balance_rmse": 1340.3009508306375}88{"canonical_length": 27, "canonical_nll": 1.22370359632704, "generation_rouge_l_f1": 0.4680851063829788, "generation_token_f1": 0.5106382978723405, "prompt_id": "fb2aa3ee59bdbf8a", "reading_canonical_log_prob": -33.03999710083008, "reading_distances": [103, 96, 95, 98], "reading_log_probs": [-35.25102233886719, -37.890724182128906, -36.5616340637207, -40.47060775756836], "reading_logprob_std": 2.229203071402915, "tokenization_path_logprob_std": 101.82202979967529, "trajectory_balance_rmse": 682.1172670145515}89{"canonical_length": 176, "canonical_nll": 0.7155953320589933, "generation_rouge_l_f1": 0.4586666666666667, "generation_token_f1": 0.6346666666666666, "prompt_id": "24055c0763c8403a", "reading_canonical_log_prob": -125.94477844238281, "reading_distances": [140, 136, 147, 136], "reading_log_probs": [-127.96427917480469, -129.44105529785156, -131.93641662597656, -136.19970703125], "reading_logprob_std": 3.6039661268095435, "tokenization_path_logprob_std": 51.89816498415684, "trajectory_balance_rmse": 3193.0339489762528}90{"canonical_length": 73, "canonical_nll": 0.49964622602070846, "generation_rouge_l_f1": 0.3687150837988827, "generation_token_f1": 0.4245810055865921, "prompt_id": "c59968b8c3d9c69d", "reading_canonical_log_prob": -36.47417449951172, "reading_distances": [15, 15, 12, 15], "reading_log_probs": [-38.8802375793457, -38.228973388671875, -37.31223678588867, -38.70320129394531], "reading_logprob_std": 0.70203239172539, "tokenization_path_logprob_std": 123.38105581026515, "trajectory_balance_rmse": 1099.9111878302851}91{"canonical_length": 283, "canonical_nll": 1.5044065010421268, "generation_rouge_l_f1": 0.28516377649325625, "generation_token_f1": 0.4778420038535646, "prompt_id": "c3213521900f0885", "reading_canonical_log_prob": -425.7470397949219, "reading_distances": [39, 42, 35, 42], "reading_log_probs": [-429.5526123046875, -429.36279296875, -429.2222900390625, -426.76611328125], "reading_logprob_std": 1.3135515383057783, "tokenization_path_logprob_std": 208.1398127062026, "trajectory_balance_rmse": 5297.20853895582}92{"canonical_length": 70, "canonical_nll": 1.4394677298409597, "generation_rouge_l_f1": 0.27312775330396477, "generation_token_f1": 0.36123348017621143, "prompt_id": "48bb22e3d5050293", "reading_canonical_log_prob": -100.76274108886719, "reading_distances": [16, 17, 20, 23], "reading_log_probs": [-106.39407348632812, -108.12251281738281, -110.49842834472656, -115.08838653564453], "reading_logprob_std": 3.7711669079402843, "tokenization_path_logprob_std": 49.92046277435872, "trajectory_balance_rmse": 1125.445920318238}93{"canonical_length": 183, "canonical_nll": 1.2718916095671107, "generation_rouge_l_f1": 0.3066666666666667, "generation_token_f1": 0.5777777777777777, "prompt_id": "5a342c18ce379992", "reading_canonical_log_prob": -232.75616455078125, "reading_distances": [30, 32, 26, 26], "reading_log_probs": [-235.36378479003906, -233.38327026367188, -233.47572326660156, -231.94911193847656], "reading_logprob_std": 1.400687467973967, "tokenization_path_logprob_std": 232.13446719272363, "trajectory_balance_rmse": 3928.026044803697}94{"canonical_length": 280, "canonical_nll": 1.5723279680524553, "generation_rouge_l_f1": 0.29545454545454547, "generation_token_f1": 0.5318181818181817, "prompt_id": "66f2043754e3c9e5", "reading_canonical_log_prob": -440.2518310546875, "reading_distances": [21, 15, 28, 21], "reading_log_probs": [-450.9353942871094, -443.9150695800781, -448.598876953125, -449.15313720703125], "reading_logprob_std": 2.9944855003969977, "tokenization_path_logprob_std": 251.9063796980962, "trajectory_balance_rmse": 5871.7494975464115}95{"canonical_length": 61, "canonical_nll": 1.8064682757268187, "generation_rouge_l_f1": 0.21951219512195122, "generation_token_f1": 0.3536585365853659, "prompt_id": "b6ca714dbce654ce", "reading_canonical_log_prob": -110.19456481933594, "reading_distances": [44, 40, 44, 45], "reading_log_probs": [-127.99491119384766, -123.2599105834961, -122.73528289794922, -131.44737243652344], "reading_logprob_std": 4.135352455299086, "tokenization_path_logprob_std": 132.2593993447681, "trajectory_balance_rmse": 1728.3014619201228}96{"canonical_length": 38, "canonical_nll": 1.2158171001233553, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "4112128a24dc114d", "reading_canonical_log_prob": -46.2010498046875, "reading_distances": [28, 33, 29, 29], "reading_log_probs": [-50.84478759765625, -50.14234161376953, -50.59428405761719, -51.07964324951172], "reading_logprob_std": 0.4010090850105128, "tokenization_path_logprob_std": 158.64868775995424, "trajectory_balance_rmse": 804.4822938745144}97{"canonical_length": 24, "canonical_nll": 0.6962072054545084, "generation_rouge_l_f1": 0.5384615384615385, "generation_token_f1": 0.576923076923077, "prompt_id": "07948a71ef97c7bb", "reading_canonical_log_prob": -16.708972930908203, "reading_distances": [22, 17, 20, 21], "reading_log_probs": [-18.61897087097168, -17.05960464477539, -17.40824317932129, -17.635265350341797], "reading_logprob_std": 0.6689316272912873, "tokenization_path_logprob_std": 86.58042410225825, "trajectory_balance_rmse": 635.0072124646224}98{"canonical_length": 130, "canonical_nll": 1.6172778789813702, "generation_rouge_l_f1": 0.18556701030927833, "generation_token_f1": 0.3608247422680413, "prompt_id": "d894cb33b4a83ecb", "reading_canonical_log_prob": -210.24612426757812, "reading_distances": [22, 32, 25, 33], "reading_log_probs": [-218.9940185546875, -222.00265502929688, -228.24462890625, -225.48568725585938], "reading_logprob_std": 4.036014517654948, "tokenization_path_logprob_std": 146.48999911336108, "trajectory_balance_rmse": 3087.2993540741695}99{"canonical_length": 49, "canonical_nll": 0.4983270606216119, "generation_rouge_l_f1": 0.13924050632911392, "generation_token_f1": 0.1518987341772152, "prompt_id": "98f5a9670dc63974", "reading_canonical_log_prob": -24.418025970458984, "reading_distances": [14, 19, 21, 21], "reading_log_probs": [-25.350229263305664, -24.578529357910156, -26.350418090820312, -26.87905502319336], "reading_logprob_std": 1.0265099310884358, "tokenization_path_logprob_std": 60.606018860799814, "trajectory_balance_rmse": 753.8084287448397}100{"canonical_length": 155, "canonical_nll": 1.3793449155745967, "generation_rouge_l_f1": 0.21649484536082472, "generation_token_f1": 0.37628865979381443, "prompt_id": "9866aa5bc569efc5", "reading_canonical_log_prob": -213.7984619140625, "reading_distances": [11, 8, 10, 13], "reading_log_probs": [-223.9603271484375, -223.5155792236328, -223.23211669921875, -223.97015380859375], "reading_logprob_std": 0.36054074770561456, "tokenization_path_logprob_std": 187.43450563794076, "trajectory_balance_rmse": 3351.6824376371433}101{"canonical_length": 173, "canonical_nll": 1.873979689757948, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "63dbcd016e93c526", "reading_canonical_log_prob": -324.198486328125, "reading_distances": [23, 18, 27, 28], "reading_log_probs": [-323.632080078125, -328.01251220703125, -323.42681884765625, -321.11138916015625], "reading_logprob_std": 2.880964798533925, "tokenization_path_logprob_std": 126.07984412154264, "trajectory_balance_rmse": 3694.856188266734}102{"canonical_length": 45, "canonical_nll": 0.972894287109375, "generation_rouge_l_f1": 0.326530612244898, "generation_token_f1": 0.38775510204081637, "prompt_id": "bb113220673ce1f4", "reading_canonical_log_prob": -43.780242919921875, "reading_distances": [24, 30, 27, 22], "reading_log_probs": [-45.83067321777344, -44.659263610839844, -44.892494201660156, -47.59858322143555], "reading_logprob_std": 1.3352527906485212, "tokenization_path_logprob_std": 25.6255619928763, "trajectory_balance_rmse": 584.2122565517343}103{"canonical_length": 16, "canonical_nll": 2.2273945808410645, "generation_rouge_l_f1": 0.4444444444444445, "generation_token_f1": 0.5925925925925927, "prompt_id": "5d38e8c442eebce3", "reading_canonical_log_prob": -35.63831329345703, "reading_distances": [37, 40, 42, 45], "reading_log_probs": [-39.40292739868164, -45.146095275878906, -41.80894088745117, -44.37932586669922], "reading_logprob_std": 2.611972970235645, "tokenization_path_logprob_std": 46.98050513363756, "trajectory_balance_rmse": 348.0554752056059}104{"canonical_length": 7, "canonical_nll": 1.6335875647408622, "generation_rouge_l_f1": 0.6, "generation_token_f1": 1.0, "prompt_id": "8ee1f7d9e89653dc", "reading_canonical_log_prob": -11.435112953186035, "reading_distances": [21, 23, 24, 19], "reading_log_probs": [-10.19539737701416, -9.536157608032227, -9.22426986694336, -11.38166618347168], "reading_logprob_std": 0.9549194611165669, "tokenization_path_logprob_std": 40.91803236802802, "trajectory_balance_rmse": 128.33844469545107}105{"canonical_length": 220, "canonical_nll": 1.7617155595259233, "generation_rouge_l_f1": 0.2181818181818182, "generation_token_f1": 0.3717171717171717, "prompt_id": "8768fc8d898485e9", "reading_canonical_log_prob": -387.5774230957031, "reading_distances": [16, 15, 12, 16], "reading_log_probs": [-395.45355224609375, -401.8929443359375, -389.03717041015625, -399.13641357421875], "reading_logprob_std": 5.560750035377808, "tokenization_path_logprob_std": 119.3483570660833, "trajectory_balance_rmse": 4794.528838310525}106{"canonical_length": 238, "canonical_nll": 1.1914749786633403, "generation_rouge_l_f1": 0.32057416267942584, "generation_token_f1": 0.5454545454545455, "prompt_id": "f7a53c2bee8bfe30", "reading_canonical_log_prob": -283.571044921875, "reading_distances": [14, 15, 19, 24], "reading_log_probs": [-284.2467346191406, -283.57391357421875, -289.5729064941406, -287.49273681640625], "reading_logprob_std": 2.814094330480817, "tokenization_path_logprob_std": 84.28695117834087, "trajectory_balance_rmse": 4383.77561936986}107{"canonical_length": 218, "canonical_nll": 1.2288188409367833, "generation_rouge_l_f1": 0.25052192066805845, "generation_token_f1": 0.49686847599164935, "prompt_id": "90b0a660ae2019c3", "reading_canonical_log_prob": -267.88250732421875, "reading_distances": [31, 29, 40, 26], "reading_log_probs": [-271.6253662109375, -270.27557373046875, -264.40704345703125, -272.7942199707031], "reading_logprob_std": 3.724024464617067, "tokenization_path_logprob_std": 205.2797532108577, "trajectory_balance_rmse": 3817.2696306040802}108{"canonical_length": 153, "canonical_nll": 1.2038778666577308, "generation_rouge_l_f1": 0.23493975903614456, "generation_token_f1": 0.3433734939759036, "prompt_id": "e11b424a8c471248", "reading_canonical_log_prob": -184.1933135986328, "reading_distances": [38, 51, 52, 57], "reading_log_probs": [-190.853759765625, -186.42410278320312, -187.40338134765625, -195.02967834472656], "reading_logprob_std": 3.8959443798198605, "tokenization_path_logprob_std": 271.5117057294532, "trajectory_balance_rmse": 2685.5905757300666}109{"canonical_length": 69, "canonical_nll": 1.7637957144474639, "generation_rouge_l_f1": 0.25757575757575757, "generation_token_f1": 0.3333333333333333, "prompt_id": "cc633ec6fcbbe3de", "reading_canonical_log_prob": -121.701904296875, "reading_distances": [18, 20, 20, 22], "reading_log_probs": [-123.7466049194336, -122.79342651367188, -122.46684265136719, -121.23414611816406], "reading_logprob_std": 1.0374786294167504, "tokenization_path_logprob_std": 139.26353544153886, "trajectory_balance_rmse": 1273.3918973911505}110{"canonical_length": 51, "canonical_nll": 1.7523778279622395, "generation_rouge_l_f1": 0.1554054054054054, "generation_token_f1": 0.23648648648648649, "prompt_id": "66e12e5aa4182ef3", "reading_canonical_log_prob": -89.37126922607422, "reading_distances": [33, 39, 39, 49], "reading_log_probs": [-90.10105895996094, -87.1590347290039, -95.96902465820312, -91.01496124267578], "reading_logprob_std": 3.6623146318293083, "tokenization_path_logprob_std": 92.17348869003027, "trajectory_balance_rmse": 891.2909668245221}111{"canonical_length": 147, "canonical_nll": 1.5330451394424958, "generation_rouge_l_f1": 0.16786570743405274, "generation_token_f1": 0.3021582733812949, "prompt_id": "831cd880037811e2", "reading_canonical_log_prob": -225.35763549804688, "reading_distances": [19, 19, 28, 19], "reading_log_probs": [-218.6829376220703, -221.54971313476562, -219.07614135742188, -220.3008575439453], "reading_logprob_std": 1.2964311028489932, "tokenization_path_logprob_std": 229.35461201522256, "trajectory_balance_rmse": 3547.515564525693}112{"canonical_length": 239, "canonical_nll": 1.012785588348261, "generation_rouge_l_f1": 0.22088353413654618, "generation_token_f1": 0.4899598393574297, "prompt_id": "61ee3d334ec189e4", "reading_canonical_log_prob": -242.05575561523438, "reading_distances": [22, 26, 31, 29], "reading_log_probs": [-262.58026123046875, -261.9789123535156, -257.24822998046875, -260.752685546875], "reading_logprob_std": 2.3856586671705005, "tokenization_path_logprob_std": 247.60003831124837, "trajectory_balance_rmse": 5680.192074705777}113{"canonical_length": 239, "canonical_nll": 1.6740068890559623, "generation_rouge_l_f1": 0.2857142857142857, "generation_token_f1": 0.5373134328358209, "prompt_id": "978f2098f6623381", "reading_canonical_log_prob": -400.087646484375, "reading_distances": [16, 19, 23, 20], "reading_log_probs": [-398.16925048828125, -398.34515380859375, -397.9270935058594, -397.7976989746094], "reading_logprob_std": 0.24475441198561942, "tokenization_path_logprob_std": 314.23121614350976, "trajectory_balance_rmse": 4976.78139402694}114{"canonical_length": 366, "canonical_nll": 1.946721478238132, "generation_rouge_l_f1": 0.22585924713584288, "generation_token_f1": 0.45171849427168576, "prompt_id": "f4a2dbf6cf78f4f4", "reading_canonical_log_prob": -712.5000610351562, "reading_distances": [41, 37, 37, 36], "reading_log_probs": [-725.3699340820312, -722.5311279296875, -722.2054443359375, -722.6027221679688], "reading_logprob_std": 1.4719404062626282, "tokenization_path_logprob_std": 157.68403907427162, "trajectory_balance_rmse": 8145.741434863139}115{"canonical_length": 323, "canonical_nll": 1.623653447295859, "generation_rouge_l_f1": 0.21071428571428572, "generation_token_f1": 0.4357142857142857, "prompt_id": "607d603874554f5a", "reading_canonical_log_prob": -524.4400634765625, "reading_distances": [21, 20, 22, 23], "reading_log_probs": [-522.29150390625, -524.783935546875, -528.4216918945312, -522.4332885742188], "reading_logprob_std": 2.86401975535302, "tokenization_path_logprob_std": 184.9807989805545, "trajectory_balance_rmse": 5994.893689497195}116{"canonical_length": 16, "canonical_nll": 1.776570439338684, "generation_rouge_l_f1": 0.13333333333333333, "generation_token_f1": 0.13333333333333333, "prompt_id": "65ee5494d4aae49e", "reading_canonical_log_prob": -28.425127029418945, "reading_distances": [15, 17, 22, 18], "reading_log_probs": [-31.84292221069336, -36.61409378051758, -33.1880989074707, -34.323421478271484], "reading_logprob_std": 2.0207266208985715, "tokenization_path_logprob_std": 50.84548029673221, "trajectory_balance_rmse": 320.9502805502754}117{"canonical_length": 120, "canonical_nll": 1.9149976094563803, "generation_rouge_l_f1": 0.24719101123595508, "generation_token_f1": 0.4419475655430712, "prompt_id": "bc654a75738c95db", "reading_canonical_log_prob": -229.79971313476562, "reading_distances": [27, 29, 26, 27], "reading_log_probs": [-226.0516357421875, -228.89772033691406, -227.16624450683594, -229.0208740234375], "reading_logprob_std": 1.4321240697506572, "tokenization_path_logprob_std": 146.1317987431496, "trajectory_balance_rmse": 2401.5364186874926}118{"canonical_length": 166, "canonical_nll": 1.059880543904132, "generation_rouge_l_f1": 0.2512315270935961, "generation_token_f1": 0.5172413793103449, "prompt_id": "1277e399b9abf4ff", "reading_canonical_log_prob": -175.94017028808594, "reading_distances": [28, 32, 24, 24], "reading_log_probs": [-187.19786071777344, -189.51959228515625, -184.64535522460938, -184.40394592285156], "reading_logprob_std": 2.4100135670856524, "tokenization_path_logprob_std": 70.08444306106972, "trajectory_balance_rmse": 4031.9261764103053}119{"canonical_length": 13, "canonical_nll": 1.82622190622183, "generation_rouge_l_f1": 0.17391304347826086, "generation_token_f1": 0.17391304347826086, "prompt_id": "888761fd7cf1d9af", "reading_canonical_log_prob": -23.74088478088379, "reading_distances": [29, 23, 26, 29], "reading_log_probs": [-24.676197052001953, -24.733196258544922, -23.751136779785156, -25.035037994384766], "reading_logprob_std": 0.5546548341744915, "tokenization_path_logprob_std": 58.836374213134995, "trajectory_balance_rmse": 361.1851272454957}120{"canonical_length": 196, "canonical_nll": 1.3558875881895727, "generation_rouge_l_f1": 0.1894736842105263, "generation_token_f1": 0.4126315789473684, "prompt_id": "4049144029e8d577", "reading_canonical_log_prob": -265.75396728515625, "reading_distances": [20, 24, 26, 23], "reading_log_probs": [-265.62652587890625, -270.2095947265625, -266.3697509765625, -267.04510498046875], "reading_logprob_std": 2.0162623978472305, "tokenization_path_logprob_std": 187.99369343386954, "trajectory_balance_rmse": 4803.887532426119}121{"canonical_length": 105, "canonical_nll": 1.1616864885602678, "generation_rouge_l_f1": 0.3173076923076923, "generation_token_f1": 0.38461538461538464, "prompt_id": "35fc3a98ce123749", "reading_canonical_log_prob": -121.97708129882812, "reading_distances": [37, 34, 41, 33], "reading_log_probs": [-124.44586944580078, -119.11756134033203, -119.99779510498047, -119.5045394897461], "reading_logprob_std": 2.479261298138735, "tokenization_path_logprob_std": 237.16037300936708, "trajectory_balance_rmse": 2095.254695115631}122{"canonical_length": 17, "canonical_nll": 3.879818187040441, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "30594b15ee7b3362", "reading_canonical_log_prob": -65.9569091796875, "reading_distances": [20, 22, 21, 20], "reading_log_probs": [-64.02786254882812, -64.41773223876953, -67.56487274169922, -64.11534881591797], "reading_logprob_std": 1.697184868163345, "tokenization_path_logprob_std": 81.09657464925445, "trajectory_balance_rmse": 388.7505151934599}123{"canonical_length": 89, "canonical_nll": 1.1611084670163272, "generation_rouge_l_f1": 0.21369863013698628, "generation_token_f1": 0.3013698630136986, "prompt_id": "55622d0a348ebb7e", "reading_canonical_log_prob": -103.33865356445312, "reading_distances": [18, 15, 16, 18], "reading_log_probs": [-106.99649047851562, -108.00550079345703, -107.1644287109375, -108.37901306152344], "reading_logprob_std": 0.6633125561133711, "tokenization_path_logprob_std": 169.1223337041733, "trajectory_balance_rmse": 1675.0943496073892}124{"canonical_length": 181, "canonical_nll": 1.235990071164969, "generation_rouge_l_f1": 0.2571428571428571, "generation_token_f1": 0.49523809523809526, "prompt_id": "96bafc4191db5d29", "reading_canonical_log_prob": -223.71420288085938, "reading_distances": [35, 26, 23, 33], "reading_log_probs": [-250.69244384765625, -245.19931030273438, -242.95294189453125, -249.5555419921875], "reading_logprob_std": 3.6398741514517474, "tokenization_path_logprob_std": 112.58967492257536, "trajectory_balance_rmse": 3767.452140441261}125{"canonical_length": 186, "canonical_nll": 1.3218177877446657, "generation_rouge_l_f1": 0.31279620853080564, "generation_token_f1": 0.4644549763033175, "prompt_id": "a69ab9afafd56708", "reading_canonical_log_prob": -245.8581085205078, "reading_distances": [26, 30, 32, 31], "reading_log_probs": [-251.13795471191406, -249.38177490234375, -251.13731384277344, -252.5281219482422], "reading_logprob_std": 1.288813949918433, "tokenization_path_logprob_std": 145.27162483459267, "trajectory_balance_rmse": 3660.561342376146}126{"canonical_length": 7, "canonical_nll": 1.5508989606584822, "generation_rouge_l_f1": 0.03755868544600939, "generation_token_f1": 0.03755868544600939, "prompt_id": "7d277a000cf626ea", "reading_canonical_log_prob": -10.856292724609375, "reading_distances": [35, 27, 28, 28], "reading_log_probs": [-11.044538497924805, -8.265727043151855, -8.148492813110352, -9.0921630859375], "reading_logprob_std": 1.3387773476935543, "tokenization_path_logprob_std": 55.169745633065624, "trajectory_balance_rmse": 179.09354695345192}127{"canonical_length": 8, "canonical_nll": 0.4528108239173889, "generation_rouge_l_f1": 0.2, "generation_token_f1": 0.2, "prompt_id": "4b1173c764ce0a81", "reading_canonical_log_prob": -3.6224865913391113, "reading_distances": [45, 57, 50, 50], "reading_log_probs": [-6.628847599029541, -8.869915008544922, -6.209334850311279, -8.802679061889648], "reading_logprob_std": 1.406311930781061, "tokenization_path_logprob_std": 26.76160031604973, "trajectory_balance_rmse": 127.46120548170839}128{"canonical_length": 48, "canonical_nll": 1.0377055009206135, "generation_rouge_l_f1": 0.2694300518134715, "generation_token_f1": 0.3212435233160622, "prompt_id": "c86574795b337627", "reading_canonical_log_prob": -49.80986404418945, "reading_distances": [8, 10, 9, 8], "reading_log_probs": [-55.872802734375, -52.570030212402344, -51.54332733154297, -53.03675842285156], "reading_logprob_std": 1.8528829200516803, "tokenization_path_logprob_std": 13.604061830211856, "trajectory_balance_rmse": 509.12836041516215}129{"canonical_length": 251, "canonical_nll": 1.2328564161323456, "generation_rouge_l_f1": 0.2256637168141593, "generation_token_f1": 0.4070796460176991, "prompt_id": "f0455cfad098aced", "reading_canonical_log_prob": -309.44696044921875, "reading_distances": [40, 40, 42, 42], "reading_log_probs": [-330.107177734375, -336.12054443359375, -333.59771728515625, -334.482421875], "reading_logprob_std": 2.5383395368835147, "tokenization_path_logprob_std": 217.36270396052856, "trajectory_balance_rmse": 5235.77945683031}130{"canonical_length": 92, "canonical_nll": 1.342961021091627, "generation_rouge_l_f1": 0.2657342657342658, "generation_token_f1": 0.34965034965034963, "prompt_id": "765c339341bf24b5", "reading_canonical_log_prob": -123.55241394042969, "reading_distances": [73, 64, 58, 62], "reading_log_probs": [-134.4619598388672, -132.66012573242188, -136.42800903320312, -132.74488830566406], "reading_logprob_std": 1.7755233450918027, "tokenization_path_logprob_std": 146.26096286652563, "trajectory_balance_rmse": 1358.7366383867197}131{"canonical_length": 25, "canonical_nll": 2.181556396484375, "generation_rouge_l_f1": 0.27118644067796616, "generation_token_f1": 0.4406779661016949, "prompt_id": "ef02d7967a69b355", "reading_canonical_log_prob": -54.538909912109375, "reading_distances": [34, 29, 29, 42], "reading_log_probs": [-52.324806213378906, -52.544273376464844, -52.187294006347656, -50.88588333129883], "reading_logprob_std": 0.7477151898202552, "tokenization_path_logprob_std": 18.457304243082195, "trajectory_balance_rmse": 486.7297063327723}132{"canonical_length": 26, "canonical_nll": 1.141248996441181, "generation_rouge_l_f1": 0.48148148148148145, "generation_token_f1": 0.5555555555555556, "prompt_id": "27392a7bee3d725d", "reading_canonical_log_prob": -29.672473907470703, "reading_distances": [30, 32, 35, 30], "reading_log_probs": [-33.82945251464844, -35.629825592041016, -35.96915817260742, -40.25110626220703], "reading_logprob_std": 2.7212708650056734, "tokenization_path_logprob_std": 43.83541179036776, "trajectory_balance_rmse": 594.0170403013653}133{"canonical_length": 53, "canonical_nll": 1.6254633777546432, "generation_rouge_l_f1": 0.04166666666666667, "generation_token_f1": 0.04166666666666667, "prompt_id": "deecb5d2b0490e50", "reading_canonical_log_prob": -86.1495590209961, "reading_distances": [15, 23, 21, 17], "reading_log_probs": [-89.67301940917969, -91.0243911743164, -91.66471862792969, -90.902587890625], "reading_logprob_std": 0.8321963222228719, "tokenization_path_logprob_std": 80.6054514872991, "trajectory_balance_rmse": 1107.9522759792146}134{"canonical_length": 124, "canonical_nll": 1.2365506079889113, "generation_rouge_l_f1": 0.20526315789473687, "generation_token_f1": 0.4368421052631579, "prompt_id": "8c1c19dea7da1533", "reading_canonical_log_prob": -153.332275390625, "reading_distances": [26, 29, 26, 20], "reading_log_probs": [-156.48751831054688, -156.17160034179688, -152.72061157226562, -158.26715087890625], "reading_logprob_std": 2.3187862564528032, "tokenization_path_logprob_std": 211.07024667640027, "trajectory_balance_rmse": 2151.8891947028196}135{"canonical_length": 81, "canonical_nll": 1.4417977980625483, "generation_rouge_l_f1": 0.21621621621621626, "generation_token_f1": 0.3783783783783784, "prompt_id": "8ddcb00e61289cea", "reading_canonical_log_prob": -116.7856216430664, "reading_distances": [17, 20, 20, 22], "reading_log_probs": [-130.05490112304688, -127.02366638183594, -123.48745727539062, -126.89889526367188], "reading_logprob_std": 2.6838766668939678, "tokenization_path_logprob_std": 56.39809067020346, "trajectory_balance_rmse": 1678.7267026548225}136{"canonical_length": 78, "canonical_nll": 0.8153157845521585, "generation_rouge_l_f1": 0.35514018691588783, "generation_token_f1": 0.41121495327102814, "prompt_id": "e509fb17129966c3", "reading_canonical_log_prob": -63.59463119506836, "reading_distances": [48, 40, 43, 52], "reading_log_probs": [-69.7365493774414, -71.84226989746094, -69.90609741210938, -71.17782592773438], "reading_logprob_std": 1.0143810361622707, "tokenization_path_logprob_std": 43.015150609826, "trajectory_balance_rmse": 1177.4758057986662}137{"canonical_length": 170, "canonical_nll": 1.1468864889705883, "generation_rouge_l_f1": 0.23017902813299235, "generation_token_f1": 0.4245524296675192, "prompt_id": "07a3f091145dd3c4", "reading_canonical_log_prob": -194.970703125, "reading_distances": [27, 25, 30, 29], "reading_log_probs": [-186.9697723388672, -189.69662475585938, -191.95677185058594, -190.73892211914062], "reading_logprob_std": 2.1250526454608107, "tokenization_path_logprob_std": 119.25785539220011, "trajectory_balance_rmse": 3861.377383766603}138{"canonical_length": 159, "canonical_nll": 1.4578095442094143, "generation_rouge_l_f1": 0.24697336561743344, "generation_token_f1": 0.5230024213075061, "prompt_id": "1cfb5bc2275fa797", "reading_canonical_log_prob": -231.79171752929688, "reading_distances": [24, 27, 18, 22], "reading_log_probs": [-242.98391723632812, -243.468994140625, -248.98985290527344, -234.00389099121094], "reading_logprob_std": 6.202113536553222, "tokenization_path_logprob_std": 86.7244433398197, "trajectory_balance_rmse": 3176.9902182021947}139{"canonical_length": 11, "canonical_nll": 0.35730177705938165, "generation_rouge_l_f1": 0.4, "generation_token_f1": 0.4, "prompt_id": "c9f832251ee1421d", "reading_canonical_log_prob": -3.9303195476531982, "reading_distances": [20, 22, 20, 18], "reading_log_probs": [-2.950313091278076, -3.2896859645843506, -3.6884512901306152, -3.824432373046875], "reading_logprob_std": 0.39660694518566847, "tokenization_path_logprob_std": 28.545258395019893, "trajectory_balance_rmse": 125.34856319881996}140{"canonical_length": 174, "canonical_nll": 1.0525912580818966, "generation_rouge_l_f1": 0.2871046228710462, "generation_token_f1": 0.4914841849148418, "prompt_id": "13d2c1dd5e0685fa", "reading_canonical_log_prob": -183.15087890625, "reading_distances": [54, 61, 58, 50], "reading_log_probs": [-177.26177978515625, -176.7695770263672, -172.48985290527344, -170.92323303222656], "reading_logprob_std": 3.137684909808828, "tokenization_path_logprob_std": 104.0161707009056, "trajectory_balance_rmse": 1871.2520256338767}141{"canonical_length": 266, "canonical_nll": 1.2487221624618186, "generation_rouge_l_f1": 0.22348484848484848, "generation_token_f1": 0.48484848484848486, "prompt_id": "e2b16f584507a85d", "reading_canonical_log_prob": -332.16009521484375, "reading_distances": [32, 28, 27, 23], "reading_log_probs": [-333.614990234375, -336.7932434082031, -338.5971984863281, -336.1915588378906], "reading_logprob_std": 2.060861034365257, "tokenization_path_logprob_std": 472.41879628772574, "trajectory_balance_rmse": 6418.357109641469}142{"canonical_length": 279, "canonical_nll": 1.2234976163474462, "generation_rouge_l_f1": 0.3236363636363636, "generation_token_f1": 0.5745454545454546, "prompt_id": "76e96744ab3b125b", "reading_canonical_log_prob": -341.3558349609375, "reading_distances": [18, 25, 16, 24], "reading_log_probs": [-347.51629638671875, -349.68157958984375, -349.41058349609375, -349.9233093261719], "reading_logprob_std": 1.0979240341912189, "tokenization_path_logprob_std": 161.39062854761795, "trajectory_balance_rmse": 6796.737961264032}143{"canonical_length": 157, "canonical_nll": 1.5452710777331309, "generation_rouge_l_f1": 0.2992874109263658, "generation_token_f1": 0.5035629453681709, "prompt_id": "2fad87f3c29a8aa9", "reading_canonical_log_prob": -242.60755920410156, "reading_distances": [27, 26, 28, 22], "reading_log_probs": [-239.92330932617188, -241.6875762939453, -241.18466186523438, -243.00486755371094], "reading_logprob_std": 1.2747868189463607, "tokenization_path_logprob_std": 69.70379670201507, "trajectory_balance_rmse": 3624.9523797177253}144{"canonical_length": 135, "canonical_nll": 2.1193827311197917, "generation_rouge_l_f1": 0.16296296296296295, "generation_token_f1": 0.2765432098765432, "prompt_id": "057edeee1c342070", "reading_canonical_log_prob": -286.1166687011719, "reading_distances": [19, 19, 23, 21], "reading_log_probs": [-282.47698974609375, -282.7742004394531, -284.8822937011719, -284.1727294921875], "reading_logprob_std": 1.1421024719481703, "tokenization_path_logprob_std": 125.69842170837163, "trajectory_balance_rmse": 3301.719638751714}145{"canonical_length": 273, "canonical_nll": 1.0304464822287087, "generation_rouge_l_f1": 0.24390243902439027, "generation_token_f1": 0.4840525328330207, "prompt_id": "b42b291887a21049", "reading_canonical_log_prob": -281.3118896484375, "reading_distances": [31, 28, 27, 29], "reading_log_probs": [-305.4306640625, -303.91680908203125, -300.83062744140625, -305.8578796386719], "reading_logprob_std": 2.2766961377823627, "tokenization_path_logprob_std": 124.15450251869994, "trajectory_balance_rmse": 5911.963957346555}146{"canonical_length": 8, "canonical_nll": 0.5802167654037476, "generation_rouge_l_f1": 0.7058823529411764, "generation_token_f1": 0.7058823529411764, "prompt_id": "751d7c04030e09ea", "reading_canonical_log_prob": -4.6417341232299805, "reading_distances": [22, 26, 26, 24], "reading_log_probs": [-4.885215759277344, -3.0522265434265137, -2.3966727256774902, -4.017339706420898], "reading_logprob_std": 1.091392857256565, "tokenization_path_logprob_std": 76.66809048662431, "trajectory_balance_rmse": 212.1323492799328}147{"canonical_length": 45, "canonical_nll": 1.190960693359375, "generation_rouge_l_f1": 0.6373626373626373, "generation_token_f1": 0.7692307692307692, "prompt_id": "bdda8d0e92726300", "reading_canonical_log_prob": -53.593231201171875, "reading_distances": [18, 23, 23, 24], "reading_log_probs": [-55.772769927978516, -55.88298797607422, -52.56964874267578, -55.562198638916016], "reading_logprob_std": 1.5904127129338117, "tokenization_path_logprob_std": 45.1985240584316, "trajectory_balance_rmse": 1035.387681791469}148{"canonical_length": 225, "canonical_nll": 0.758565673828125, "generation_rouge_l_f1": 0.3613963039014374, "generation_token_f1": 0.5010266940451745, "prompt_id": "ce2f11428dfe632d", "reading_canonical_log_prob": -170.67727661132812, "reading_distances": [35, 33, 26, 26], "reading_log_probs": [-169.89820861816406, -171.84518432617188, -169.2128448486328, -172.79515075683594], "reading_logprob_std": 1.6662651505517172, "tokenization_path_logprob_std": 50.84236004415784, "trajectory_balance_rmse": 3423.0792510610027}149{"canonical_length": 371, "canonical_nll": 1.1597685698229026, "generation_rouge_l_f1": 0.2464285714285714, "generation_token_f1": 0.4285714285714285, "prompt_id": "3ce75f08c1522ce9", "reading_canonical_log_prob": -430.2741394042969, "reading_distances": [20, 17, 24, 20], "reading_log_probs": [-438.50860595703125, -446.51556396484375, -435.2967834472656, -446.9334716796875], "reading_logprob_std": 5.822755211210871, "tokenization_path_logprob_std": 267.2731317844713, "trajectory_balance_rmse": 6302.060546083426}150{"canonical_length": 3, "canonical_nll": 1.6111644109090169, "generation_rouge_l_f1": 0.03418803418803419, "generation_token_f1": 0.03418803418803419, "prompt_id": "cc1031d1c53d6dff", "reading_canonical_log_prob": -4.833493232727051, "reading_distances": [40, 29, 40, 40], "reading_log_probs": [-4.176474571228027, -4.282686710357666, -4.590517044067383, -4.032347202301025], "reading_logprob_std": 0.236725551745511, "tokenization_path_logprob_std": 12.56495697359529, "trajectory_balance_rmse": 32.003501919197966}151{"canonical_length": 14, "canonical_nll": 3.7210426330566406, "generation_rouge_l_f1": 0.07142857142857142, "generation_token_f1": 0.07142857142857142, "prompt_id": "6b9afce682e6088d", "reading_canonical_log_prob": -52.09459686279297, "reading_distances": [55, 50, 45, 53], "reading_log_probs": [-55.46272277832031, -55.27147674560547, -51.44103240966797, -58.65870666503906], "reading_logprob_std": 2.953322048487298, "tokenization_path_logprob_std": 84.182577366367, "trajectory_balance_rmse": 370.21508336564125}152{"canonical_length": 15, "canonical_nll": 0.4247511545817057, "generation_rouge_l_f1": 0.7142857142857143, "generation_token_f1": 0.7142857142857143, "prompt_id": "16fc84421f512267", "reading_canonical_log_prob": -6.371267318725586, "reading_distances": [24, 30, 36, 24], "reading_log_probs": [-8.49226188659668, -4.659757137298584, -8.232742309570312, -5.986100673675537], "reading_logprob_std": 1.839589989356806, "tokenization_path_logprob_std": 22.076144410398676, "trajectory_balance_rmse": 268.81774191477115}153{"canonical_length": 12, "canonical_nll": 0.7075645128885905, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "7550a792cc616b59", "reading_canonical_log_prob": -8.490774154663086, "reading_distances": [24, 37, 26, 31], "reading_log_probs": [-10.864445686340332, -14.308194160461426, -11.336352348327637, -14.462608337402344], "reading_logprob_std": 1.9073987067414928, "tokenization_path_logprob_std": 23.439265992972924, "trajectory_balance_rmse": 181.5608871082561}154{"canonical_length": 242, "canonical_nll": 1.5986977569327867, "generation_rouge_l_f1": 0.22519083969465647, "generation_token_f1": 0.4427480916030534, "prompt_id": "d8ba10df8798bb3c", "reading_canonical_log_prob": -386.8848571777344, "reading_distances": [33, 38, 35, 38], "reading_log_probs": [-393.70819091796875, -391.3500061035156, -394.95367431640625, -395.95989990234375], "reading_logprob_std": 1.988141787785842, "tokenization_path_logprob_std": 467.63379615208504, "trajectory_balance_rmse": 5739.0312287176475}155{"canonical_length": 15, "canonical_nll": 1.0250547409057618, "generation_rouge_l_f1": 0.3636363636363637, "generation_token_f1": 0.3636363636363637, "prompt_id": "c62abdde12be0560", "reading_canonical_log_prob": -15.375821113586426, "reading_distances": [25, 28, 29, 29], "reading_log_probs": [-20.233850479125977, -17.99492645263672, -20.963199615478516, -19.23859214782715], "reading_logprob_std": 1.286691200755773, "tokenization_path_logprob_std": 60.12849634986213, "trajectory_balance_rmse": 280.07093925320567}156{"canonical_length": 318, "canonical_nll": 1.537877784585053, "generation_rouge_l_f1": 0.22145328719723184, "generation_token_f1": 0.4013840830449827, "prompt_id": "e65b179e41f3017d", "reading_canonical_log_prob": -489.0451354980469, "reading_distances": [21, 27, 27, 30], "reading_log_probs": [-495.76165771484375, -497.1871643066406, -496.0802001953125, -503.5509033203125], "reading_logprob_std": 3.6553594718032008, "tokenization_path_logprob_std": 191.14256871137235, "trajectory_balance_rmse": 6947.6591146129085}157{"canonical_length": 83, "canonical_nll": 1.1428651924592903, "generation_rouge_l_f1": 0.18435754189944137, "generation_token_f1": 0.3575418994413408, "prompt_id": "688f5dc10c3a0a47", "reading_canonical_log_prob": -94.8578109741211, "reading_distances": [33, 34, 24, 33], "reading_log_probs": [-125.51163482666016, -121.26952362060547, -107.80577087402344, -120.62548828125], "reading_logprob_std": 7.6452606659159565, "tokenization_path_logprob_std": 35.65651524306081, "trajectory_balance_rmse": 1679.0852888320987}158{"canonical_length": 324, "canonical_nll": 1.6087047435619213, "generation_rouge_l_f1": 0.34615384615384615, "generation_token_f1": 0.5314685314685315, "prompt_id": "d7836a3fff052985", "reading_canonical_log_prob": -521.2203369140625, "reading_distances": [22, 21, 22, 22], "reading_log_probs": [-528.8226928710938, -524.5599365234375, -526.5816650390625, -524.575927734375], "reading_logprob_std": 2.027702632599386, "tokenization_path_logprob_std": 250.22447988613007, "trajectory_balance_rmse": 6042.894639750076}159{"canonical_length": 18, "canonical_nll": 1.4257148106892903, "generation_rouge_l_f1": 0.4210526315789473, "generation_token_f1": 0.4210526315789473, "prompt_id": "97d1bc52cae659b0", "reading_canonical_log_prob": -25.662866592407227, "reading_distances": [9, 12, 13, 10], "reading_log_probs": [-27.69719696044922, -27.293071746826172, -27.4661922454834, -27.04833221435547], "reading_logprob_std": 0.27419290301555327, "tokenization_path_logprob_std": 23.94798566769602, "trajectory_balance_rmse": 220.34596891846743}160{"canonical_length": 311, "canonical_nll": 0.9545270423030546, "generation_rouge_l_f1": 0.29360967184801384, "generation_token_f1": 0.5492227979274611, "prompt_id": "d655055a06f3b252", "reading_canonical_log_prob": -296.85791015625, "reading_distances": [30, 34, 35, 31], "reading_log_probs": [-305.40399169921875, -302.9452209472656, -305.96722412109375, -307.3623352050781], "reading_logprob_std": 1.8436293496006433, "tokenization_path_logprob_std": 359.79244647273805, "trajectory_balance_rmse": 6153.058026144095}161{"canonical_length": 270, "canonical_nll": 1.432325122974537, "generation_rouge_l_f1": 0.29328621908127206, "generation_token_f1": 0.46289752650176674, "prompt_id": "62d551bf467199ab", "reading_canonical_log_prob": -386.727783203125, "reading_distances": [18, 19, 24, 20], "reading_log_probs": [-387.2098388671875, -386.6791687011719, -388.84295654296875, -384.66851806640625], "reading_logprob_std": 1.7213760477819406, "tokenization_path_logprob_std": 208.84911650420628, "trajectory_balance_rmse": 4947.756050502163}162{"canonical_length": 11, "canonical_nll": 0.8358870419588956, "generation_rouge_l_f1": 0.7407407407407407, "generation_token_f1": 0.7407407407407407, "prompt_id": "1237b2c3cf7851d2", "reading_canonical_log_prob": -9.194757461547852, "reading_distances": [20, 20, 16, 18], "reading_log_probs": [-12.49440860748291, -11.721536636352539, -11.733133316040039, -14.959097862243652], "reading_logprob_std": 1.5313487710563485, "tokenization_path_logprob_std": 67.15753231217808, "trajectory_balance_rmse": 294.8373374506349}163{"canonical_length": 3, "canonical_nll": 1.94081449508667, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.6666666666666666, "prompt_id": "9af597ad9c807de8", "reading_canonical_log_prob": -5.82244348526001, "reading_distances": [17, 19, 21, 18], "reading_log_probs": [-5.596325874328613, -6.900830268859863, -7.209153175354004, -10.542126655578613], "reading_logprob_std": 2.106076597146494, "tokenization_path_logprob_std": 14.380339314602542, "trajectory_balance_rmse": 36.170628253476266}164{"canonical_length": 158, "canonical_nll": 1.2686240280730814, "generation_rouge_l_f1": 0.32634032634032634, "generation_token_f1": 0.5174825174825175, "prompt_id": "0366eeefae975865", "reading_canonical_log_prob": -200.44259643554688, "reading_distances": [31, 34, 31, 32], "reading_log_probs": [-199.49205017089844, -201.46273803710938, -204.74722290039062, -198.43865966796875], "reading_logprob_std": 2.7740004730200396, "tokenization_path_logprob_std": 157.0944020028253, "trajectory_balance_rmse": 2760.52610564238}165{"canonical_length": 94, "canonical_nll": 1.7816454298952793, "generation_rouge_l_f1": 0.1756373937677054, "generation_token_f1": 0.31728045325779036, "prompt_id": "e99bd769bc8f61be", "reading_canonical_log_prob": -167.47467041015625, "reading_distances": [18, 16, 21, 22], "reading_log_probs": [-168.34442138671875, -173.59457397460938, -173.77706909179688, -171.2227783203125], "reading_logprob_std": 2.5420762898272335, "tokenization_path_logprob_std": 317.0219145822422, "trajectory_balance_rmse": 2267.9430313083867}166{"canonical_length": 287, "canonical_nll": 0.3326626555014155, "generation_rouge_l_f1": 0.596401028277635, "generation_token_f1": 0.6940874035989717, "prompt_id": "e1ea57924db21e38", "reading_canonical_log_prob": -95.47418212890625, "reading_distances": [20, 19, 16, 21], "reading_log_probs": [-97.17064666748047, -95.32450866699219, -96.02246856689453, -96.05879211425781], "reading_logprob_std": 0.7632386377540004, "tokenization_path_logprob_std": 117.35618705084428, "trajectory_balance_rmse": 2507.0717229389043}167{"canonical_length": 248, "canonical_nll": 0.8499796467442666, "generation_rouge_l_f1": 0.4007285974499089, "generation_token_f1": 0.6302367941712205, "prompt_id": "d13dad8f3ef63531", "reading_canonical_log_prob": -210.79495239257812, "reading_distances": [29, 34, 32, 32], "reading_log_probs": [-215.68955993652344, -217.67654418945312, -223.22760009765625, -222.66372680664062], "reading_logprob_std": 3.7127407291870953, "tokenization_path_logprob_std": 218.36919498588787, "trajectory_balance_rmse": 3411.235134620288}168{"canonical_length": 21, "canonical_nll": 1.4473465510777064, "generation_rouge_l_f1": 0.5294117647058824, "generation_token_f1": 0.5294117647058824, "prompt_id": "146de30da91fe96b", "reading_canonical_log_prob": -30.394277572631836, "reading_distances": [19, 19, 18, 14], "reading_log_probs": [-29.251296997070312, -29.343189239501953, -27.3831787109375, -31.427648544311523], "reading_logprob_std": 1.652754374545346, "tokenization_path_logprob_std": 55.66468320308638, "trajectory_balance_rmse": 452.1812740496736}169{"canonical_length": 141, "canonical_nll": 1.077667236328125, "generation_rouge_l_f1": 0.31976744186046513, "generation_token_f1": 0.48837209302325585, "prompt_id": "e8d40d5d31440017", "reading_canonical_log_prob": -151.95108032226562, "reading_distances": [76, 71, 76, 68], "reading_log_probs": [-168.0336151123047, -167.21139526367188, -167.434326171875, -164.91763305664062], "reading_logprob_std": 1.365934490051066, "tokenization_path_logprob_std": 106.12514241421594, "trajectory_balance_rmse": 2673.582021215148}170{"canonical_length": 7, "canonical_nll": 0.18645589692252024, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "e77edca13c4c71bd", "reading_canonical_log_prob": -1.3051912784576416, "reading_distances": [13, 20, 17, 13], "reading_log_probs": [-4.4195380210876465, -2.9704198837280273, -4.866613388061523, -3.5542070865631104], "reading_logprob_std": 0.851830626274371, "tokenization_path_logprob_std": 29.57360995870941, "trajectory_balance_rmse": 212.851735631348}171{"canonical_length": 255, "canonical_nll": 1.383279718137255, "generation_rouge_l_f1": 0.16015625, "generation_token_f1": 0.34765625, "prompt_id": "5dc83250d862853a", "reading_canonical_log_prob": -352.736328125, "reading_distances": [31, 29, 31, 27], "reading_log_probs": [-370.9789123535156, -371.2861022949219, -366.09765625, -365.0389404296875], "reading_logprob_std": 3.2438682693116068, "tokenization_path_logprob_std": 323.5801175891341, "trajectory_balance_rmse": 5640.452886915203}172{"canonical_length": 25, "canonical_nll": 1.3657568359375, "generation_rouge_l_f1": 0.18181818181818182, "generation_token_f1": 0.3272727272727273, "prompt_id": "a9f2944f6ac15039", "reading_canonical_log_prob": -34.1439208984375, "reading_distances": [26, 27, 34, 28], "reading_log_probs": [-36.79048156738281, -37.9655647277832, -37.28938293457031, -36.23051452636719], "reading_logprob_std": 0.7377956706669592, "tokenization_path_logprob_std": 63.56909509459799, "trajectory_balance_rmse": 424.8990933263347}173{"canonical_length": 51, "canonical_nll": 1.460252649643842, "generation_rouge_l_f1": 0.16597510373443983, "generation_token_f1": 0.2987551867219917, "prompt_id": "0e9e30fa66e7e5e1", "reading_canonical_log_prob": -74.47288513183594, "reading_distances": [14, 18, 21, 19], "reading_log_probs": [-91.57363891601562, -86.79145812988281, -89.86898803710938, -88.72236633300781], "reading_logprob_std": 2.0087139580652065, "tokenization_path_logprob_std": 73.81440902177852, "trajectory_balance_rmse": 921.6389758368086}174{"canonical_length": 102, "canonical_nll": 2.3627813002642464, "generation_rouge_l_f1": 0.15160349854227406, "generation_token_f1": 0.2099125364431487, "prompt_id": "f6e26ca6d4af6e82", "reading_canonical_log_prob": -241.00369262695312, "reading_distances": [27, 34, 25, 22], "reading_log_probs": [-254.53152465820312, -272.5957946777344, -272.01470947265625, -254.6605224609375], "reading_logprob_std": 10.227315138767052, "tokenization_path_logprob_std": 51.659319972172284, "trajectory_balance_rmse": 2063.2446700679366}175{"canonical_length": 92, "canonical_nll": 1.8865313322647759, "generation_rouge_l_f1": 0.1949860724233983, "generation_token_f1": 0.350974930362117, "prompt_id": "e5d1110bf220108e", "reading_canonical_log_prob": -173.56088256835938, "reading_distances": [30, 29, 24, 27], "reading_log_probs": [-184.5202178955078, -187.25653076171875, -183.93243408203125, -185.49667358398438], "reading_logprob_std": 1.454301804593818, "tokenization_path_logprob_std": 126.54297989163568, "trajectory_balance_rmse": 1864.9711907848018}176{"canonical_length": 10, "canonical_nll": 1.117439079284668, "generation_rouge_l_f1": 0.4444444444444444, "generation_token_f1": 0.5555555555555556, "prompt_id": "dd04d981a73d4900", "reading_canonical_log_prob": -11.17439079284668, "reading_distances": [20, 14, 16, 15], "reading_log_probs": [-15.922994613647461, -14.990853309631348, -16.202701568603516, -13.836228370666504], "reading_logprob_std": 1.0686302405866812, "tokenization_path_logprob_std": 23.823092775543504, "trajectory_balance_rmse": 132.3181764591026}177{"canonical_length": 299, "canonical_nll": 1.6451598712831836, "generation_rouge_l_f1": 0.26642984014209586, "generation_token_f1": 0.5186500888099467, "prompt_id": "d9940a473d55b366", "reading_canonical_log_prob": -491.9028015136719, "reading_distances": [26, 27, 16, 26], "reading_log_probs": [-495.6156311035156, -491.7284240722656, -493.9034729003906, -494.64569091796875], "reading_logprob_std": 1.6526448252031896, "tokenization_path_logprob_std": 264.72468586568414, "trajectory_balance_rmse": 6018.04596896303}178{"canonical_length": 416, "canonical_nll": 1.1316114572378306, "generation_rouge_l_f1": 0.23076923076923075, "generation_token_f1": 0.38141025641025644, "prompt_id": "7c0396dbe5594b7c", "reading_canonical_log_prob": -470.7503662109375, "reading_distances": [25, 21, 23, 25], "reading_log_probs": [-476.1591796875, -476.1550598144531, -477.8043518066406, -475.21722412109375], "reading_logprob_std": 1.075748502539476, "tokenization_path_logprob_std": 91.40725257215028, "trajectory_balance_rmse": 6269.093614758333}179{"canonical_length": 138, "canonical_nll": 1.4942968617314878, "generation_rouge_l_f1": 0.20197044334975367, "generation_token_f1": 0.40394088669950734, "prompt_id": "5ee117416302abcd", "reading_canonical_log_prob": -206.2129669189453, "reading_distances": [20, 17, 19, 22], "reading_log_probs": [-228.3773651123047, -224.24879455566406, -229.89906311035156, -227.791748046875], "reading_logprob_std": 2.3913461680169394, "tokenization_path_logprob_std": 81.78823773672138, "trajectory_balance_rmse": 2696.782866383991}180{"canonical_length": 326, "canonical_nll": 1.5321578511431173, "generation_rouge_l_f1": 0.22984562607204115, "generation_token_f1": 0.48370497427101206, "prompt_id": "7d6d0adc840aad0a", "reading_canonical_log_prob": -499.48345947265625, "reading_distances": [26, 24, 27, 19], "reading_log_probs": [-506.56317138671875, -508.1313171386719, -508.0253601074219, -508.8442687988281], "reading_logprob_std": 0.9570188418949634, "tokenization_path_logprob_std": 398.9852874823108, "trajectory_balance_rmse": 7240.31189957735}181{"canonical_length": 17, "canonical_nll": 1.576725679285386, "generation_rouge_l_f1": 0.2580645161290323, "generation_token_f1": 0.38709677419354843, "prompt_id": "305ea9936db72d06", "reading_canonical_log_prob": -26.804336547851562, "reading_distances": [62, 54, 57, 53], "reading_log_probs": [-27.788713455200195, -22.536026000976562, -22.529050827026367, -23.14691162109375], "reading_logprob_std": 2.542244457427398, "tokenization_path_logprob_std": 14.757533584867845, "trajectory_balance_rmse": 331.3343045240164}182{"canonical_length": 8, "canonical_nll": 1.3999344110488892, "generation_rouge_l_f1": 0.5454545454545454, "generation_token_f1": 0.5454545454545454, "prompt_id": "f9a917228bf6b5fe", "reading_canonical_log_prob": -11.199475288391113, "reading_distances": [32, 38, 38, 37], "reading_log_probs": [-12.235350608825684, -12.521687507629395, -12.625722885131836, -11.77822494506836], "reading_logprob_std": 0.3791609054109131, "tokenization_path_logprob_std": 24.772457416756435, "trajectory_balance_rmse": 118.69863798161961}183{"canonical_length": 206, "canonical_nll": 1.2632992938884253, "generation_rouge_l_f1": 0.3661417322834646, "generation_token_f1": 0.5196850393700788, "prompt_id": "6fb0440b40107b60", "reading_canonical_log_prob": -260.2396545410156, "reading_distances": [13, 11, 13, 12], "reading_log_probs": [-265.54217529296875, -266.25799560546875, -264.78277587890625, -267.29205322265625], "reading_logprob_std": 1.0682220486956024, "tokenization_path_logprob_std": 166.70171928574365, "trajectory_balance_rmse": 2909.2914329258588}184{"canonical_length": 94, "canonical_nll": 1.8769083225980718, "generation_rouge_l_f1": 0.22478386167146977, "generation_token_f1": 0.3688760806916426, "prompt_id": "16cf88f25ee512b8", "reading_canonical_log_prob": -176.42938232421875, "reading_distances": [17, 17, 16, 22], "reading_log_probs": [-180.2920684814453, -180.52491760253906, -180.25802612304688, -175.504638671875], "reading_logprob_std": 2.429745968226843, "tokenization_path_logprob_std": 51.073205549500166, "trajectory_balance_rmse": 2364.2631487435524}185{"canonical_length": 12, "canonical_nll": 1.0896989504496257, "generation_rouge_l_f1": 0.28571428571428575, "generation_token_f1": 0.28571428571428575, "prompt_id": "024090bc27e03da4", "reading_canonical_log_prob": -13.076387405395508, "reading_distances": [26, 19, 17, 29], "reading_log_probs": [-13.06888198852539, -11.524077415466309, -10.431750297546387, -14.176523208618164], "reading_logprob_std": 1.6537766165811676, "tokenization_path_logprob_std": 62.63412387500427, "trajectory_balance_rmse": 308.57779438179847}186{"canonical_length": 58, "canonical_nll": 1.9525925208782327, "generation_rouge_l_f1": 0.12322274881516586, "generation_token_f1": 0.1990521327014218, "prompt_id": "d97dcaaab25ed5d3", "reading_canonical_log_prob": -113.2503662109375, "reading_distances": [11, 17, 10, 15], "reading_log_probs": [-137.5779571533203, -133.84109497070312, -135.99058532714844, -137.17236328125], "reading_logprob_std": 1.6773652742114495, "tokenization_path_logprob_std": 79.54705404113824, "trajectory_balance_rmse": 1314.9675624520044}187{"canonical_length": 386, "canonical_nll": 0.8933710286036675, "generation_rouge_l_f1": 0.2901023890784983, "generation_token_f1": 0.4709897610921502, "prompt_id": "5a59813799d04978", "reading_canonical_log_prob": -344.8412170410156, "reading_distances": [9, 11, 12, 14], "reading_log_probs": [-345.63946533203125, -349.9737854003906, -345.4590759277344, -347.6976013183594], "reading_logprob_std": 2.1140286589952995, "tokenization_path_logprob_std": 319.06589787072215, "trajectory_balance_rmse": 5892.416463927329}188{"canonical_length": 120, "canonical_nll": 1.1458994547526042, "generation_rouge_l_f1": 0.26666666666666666, "generation_token_f1": 0.4266666666666667, "prompt_id": "84eecbd6c7c38a00", "reading_canonical_log_prob": -137.5079345703125, "reading_distances": [20, 24, 27, 23], "reading_log_probs": [-135.450439453125, -135.99533081054688, -136.2484130859375, -137.37628173828125], "reading_logprob_std": 0.8106427911036562, "tokenization_path_logprob_std": 98.2077873910674, "trajectory_balance_rmse": 2265.0855331993166}189{"canonical_length": 12, "canonical_nll": 0.3581600983937581, "generation_rouge_l_f1": 0.8000000000000002, "generation_token_f1": 0.8000000000000002, "prompt_id": "1914d368729c6a5a", "reading_canonical_log_prob": -4.297921180725098, "reading_distances": [23, 22, 23, 27], "reading_log_probs": [-5.356057167053223, -4.17108678817749, -6.462409019470215, -4.639513969421387], "reading_logprob_std": 0.9972516954723096, "tokenization_path_logprob_std": 35.17624319733041, "trajectory_balance_rmse": 250.0974209810788}190{"canonical_length": 313, "canonical_nll": 1.1933443599615614, "generation_rouge_l_f1": 0.20350877192982456, "generation_token_f1": 0.4035087719298246, "prompt_id": "dc582c1cebaf17e1", "reading_canonical_log_prob": -373.51678466796875, "reading_distances": [43, 44, 34, 50], "reading_log_probs": [-396.2080078125, -397.46148681640625, -387.88812255859375, -398.82745361328125], "reading_logprob_std": 4.923052313534834, "tokenization_path_logprob_std": 257.889722532804, "trajectory_balance_rmse": 7307.137536042447}191{"canonical_length": 34, "canonical_nll": 1.350354138542624, "generation_rouge_l_f1": 0.14583333333333334, "generation_token_f1": 0.16666666666666666, "prompt_id": "c32c06c0fe76b404", "reading_canonical_log_prob": -45.91204071044922, "reading_distances": [19, 19, 20, 22], "reading_log_probs": [-46.362701416015625, -55.712860107421875, -46.229164123535156, -45.474342346191406], "reading_logprob_std": 4.861155323731138, "tokenization_path_logprob_std": 42.35751494696103, "trajectory_balance_rmse": 522.6316942763283}192{"canonical_length": 124, "canonical_nll": 1.8177786796323714, "generation_rouge_l_f1": 0.2198952879581152, "generation_token_f1": 0.39267015706806285, "prompt_id": "3415a33aecea8511", "reading_canonical_log_prob": -225.40455627441406, "reading_distances": [22, 20, 19, 23], "reading_log_probs": [-229.77435302734375, -230.0027618408203, -227.15028381347656, -227.71697998046875], "reading_logprob_std": 1.4391336466349534, "tokenization_path_logprob_std": 44.34482054418409, "trajectory_balance_rmse": 3201.1107787608216}193{"canonical_length": 333, "canonical_nll": 1.449696950368337, "generation_rouge_l_f1": 0.22628951747088186, "generation_token_f1": 0.48252911813643923, "prompt_id": "f7e2c3707f2870f0", "reading_canonical_log_prob": -482.74908447265625, "reading_distances": [58, 53, 59, 53], "reading_log_probs": [-487.7235107421875, -483.6065368652344, -482.7237548828125, -487.34796142578125], "reading_logprob_std": 2.5535745823046656, "tokenization_path_logprob_std": 190.2006374626729, "trajectory_balance_rmse": 6859.339873477022}194{"canonical_length": 39, "canonical_nll": 1.7264196933844151, "generation_rouge_l_f1": 0.11501597444089456, "generation_token_f1": 0.16613418530351437, "prompt_id": "ec9f4061e1384314", "reading_canonical_log_prob": -67.33036804199219, "reading_distances": [19, 18, 19, 16], "reading_log_probs": [-72.27804565429688, -73.33887481689453, -73.52194213867188, -71.72428131103516], "reading_logprob_std": 0.858841380805741, "tokenization_path_logprob_std": 67.20041398439304, "trajectory_balance_rmse": 832.0600743557645}195{"canonical_length": 14, "canonical_nll": 1.3416384288242884, "generation_rouge_l_f1": 0.3870967741935484, "generation_token_f1": 0.3870967741935484, "prompt_id": "a05868d3246d849c", "reading_canonical_log_prob": -18.78293800354004, "reading_distances": [10, 9, 8, 9], "reading_log_probs": [-18.62823486328125, -18.60723304748535, -19.01409149169922, -18.860361099243164], "reading_logprob_std": 0.19503197198625757, "tokenization_path_logprob_std": 78.20637031700414, "trajectory_balance_rmse": 391.041063626299}196{"canonical_length": 127, "canonical_nll": 1.4371631051611713, "generation_rouge_l_f1": 0.23728813559322035, "generation_token_f1": 0.41162227602905566, "prompt_id": "420ee5406fba3002", "reading_canonical_log_prob": -182.51971435546875, "reading_distances": [21, 24, 21, 25], "reading_log_probs": [-188.307861328125, -189.379638671875, -186.75987243652344, -186.51812744140625], "reading_logprob_std": 1.3496291453895193, "tokenization_path_logprob_std": 134.82762509135554, "trajectory_balance_rmse": 2937.2783251592246}197{"canonical_length": 9, "canonical_nll": 3.0420989990234375, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "0a4e72c4105acd5f", "reading_canonical_log_prob": -27.378890991210938, "reading_distances": [16, 8, 15, 16], "reading_log_probs": [-26.754789352416992, -27.448244094848633, -24.874956130981445, -27.171524047851562], "reading_logprob_std": 1.1604930472570545, "tokenization_path_logprob_std": 26.091260591488798, "trajectory_balance_rmse": 151.61543158127367}198{"canonical_length": 16, "canonical_nll": 0.7182347774505615, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.6666666666666666, "prompt_id": "0ce850d0751da7a2", "reading_canonical_log_prob": -11.491756439208984, "reading_distances": [61, 60, 73, 64], "reading_log_probs": [-23.667892456054688, -18.773284912109375, -20.84900665283203, -25.730281829833984], "reading_logprob_std": 3.0644741843923815, "tokenization_path_logprob_std": 26.175614047298023, "trajectory_balance_rmse": 338.5418633722003}199{"canonical_length": 9, "canonical_nll": 3.8234613206651478, "generation_rouge_l_f1": 0.30769230769230765, "generation_token_f1": 0.30769230769230765, "prompt_id": "eeee24292d24cae3", "reading_canonical_log_prob": -34.41115188598633, "reading_distances": [19, 19, 26, 23], "reading_log_probs": [-42.794551849365234, -39.9150390625, -43.84022521972656, -42.279876708984375], "reading_logprob_std": 1.660425350259621, "tokenization_path_logprob_std": 33.554867365362924, "trajectory_balance_rmse": 182.474901403933}200{"canonical_length": 20, "canonical_nll": 2.2813703536987306, "generation_rouge_l_f1": 0.325, "generation_token_f1": 0.375, "prompt_id": "aa417b16bf366ab6", "reading_canonical_log_prob": -45.62740707397461, "reading_distances": [31, 33, 27, 38], "reading_log_probs": [-35.899513244628906, -37.109962463378906, -38.703128814697266, -37.45338821411133], "reading_logprob_std": 1.1531820770762342, "tokenization_path_logprob_std": 56.810095335457085, "trajectory_balance_rmse": 420.72870125422224}201{"canonical_length": 277, "canonical_nll": 1.4702505394009477, "generation_rouge_l_f1": 0.23300970873786409, "generation_token_f1": 0.46601941747572817, "prompt_id": "31ffef5f718912de", "reading_canonical_log_prob": -407.2593994140625, "reading_distances": [15, 18, 20, 15], "reading_log_probs": [-415.7868347167969, -420.38818359375, -420.3815612792969, -414.98193359375], "reading_logprob_std": 2.905674786080048, "tokenization_path_logprob_std": 266.0358857446212, "trajectory_balance_rmse": 6012.366388210549}202{"canonical_length": 170, "canonical_nll": 1.0882064819335937, "generation_rouge_l_f1": 0.18952618453865339, "generation_token_f1": 0.2793017456359102, "prompt_id": "fe5c1f86ece4d243", "reading_canonical_log_prob": -184.99510192871094, "reading_distances": [18, 14, 18, 19], "reading_log_probs": [-185.30398559570312, -176.71597290039062, -181.05148315429688, -171.91445922851562], "reading_logprob_std": 5.747851369061398, "tokenization_path_logprob_std": 56.178271510666356, "trajectory_balance_rmse": 3143.3020110921857}203{"canonical_length": 166, "canonical_nll": 1.4711618078760353, "generation_rouge_l_f1": 0.24875621890547267, "generation_token_f1": 0.4626865671641791, "prompt_id": "8db7426fa0df6c62", "reading_canonical_log_prob": -244.21286010742188, "reading_distances": [18, 17, 24, 22], "reading_log_probs": [-256.47064208984375, -254.68003845214844, -254.86720275878906, -258.5567321777344], "reading_logprob_std": 1.7982737783644274, "tokenization_path_logprob_std": 94.09638648972678, "trajectory_balance_rmse": 2158.560247484685}204{"canonical_length": 39, "canonical_nll": 0.7905854934301132, "generation_rouge_l_f1": 0.34710743801652894, "generation_token_f1": 0.42975206611570255, "prompt_id": "6c3aa6c41d3147aa", "reading_canonical_log_prob": -30.832834243774414, "reading_distances": [49, 49, 41, 43], "reading_log_probs": [-37.907203674316406, -38.28547668457031, -37.822139739990234, -35.902034759521484], "reading_logprob_std": 1.0705648193976707, "tokenization_path_logprob_std": 92.91355468565575, "trajectory_balance_rmse": 737.7146532811097}205{"canonical_length": 8, "canonical_nll": 2.2708425521850586, "generation_rouge_l_f1": 0.0390625, "generation_token_f1": 0.0390625, "prompt_id": "e07ed69174530f07", "reading_canonical_log_prob": -18.16674041748047, "reading_distances": [35, 38, 33, 38], "reading_log_probs": [-16.574121475219727, -17.342222213745117, -18.13024139404297, -17.464126586914062], "reading_logprob_std": 0.6379094035803689, "tokenization_path_logprob_std": 48.587411739039865, "trajectory_balance_rmse": 118.65712406651407}206{"canonical_length": 57, "canonical_nll": 1.9452858640436541, "generation_rouge_l_f1": 0.12751677852348994, "generation_token_f1": 0.18120805369127518, "prompt_id": "93c96f6a3b9968fc", "reading_canonical_log_prob": -110.88129425048828, "reading_distances": [25, 18, 29, 22], "reading_log_probs": [-115.6377944946289, -115.27538299560547, -117.4475326538086, -118.35078430175781], "reading_logprob_std": 1.4651222606560663, "tokenization_path_logprob_std": 165.14142074139315, "trajectory_balance_rmse": 1256.214799977173}207{"canonical_length": 264, "canonical_nll": 1.3240487069794626, "generation_rouge_l_f1": 0.2598425196850394, "generation_token_f1": 0.515748031496063, "prompt_id": "0a9cdb872ed6cdad", "reading_canonical_log_prob": -349.5488586425781, "reading_distances": [23, 20, 21, 19], "reading_log_probs": [-347.91748046875, -345.5214538574219, -345.69354248046875, -345.3865661621094], "reading_logprob_std": 1.198416127298637, "tokenization_path_logprob_std": 245.04807841764307, "trajectory_balance_rmse": 5000.443423470675}208{"canonical_length": 141, "canonical_nll": 1.5484667838887964, "generation_rouge_l_f1": 0.2123287671232877, "generation_token_f1": 0.4109589041095891, "prompt_id": "c6291fd5f11c978c", "reading_canonical_log_prob": -218.3338165283203, "reading_distances": [23, 17, 21, 19], "reading_log_probs": [-222.14283752441406, -225.71539306640625, -228.91488647460938, -228.85260009765625], "reading_logprob_std": 3.21101514961333, "tokenization_path_logprob_std": 186.26609207453743, "trajectory_balance_rmse": 3204.918977142485}209{"canonical_length": 9, "canonical_nll": 0.3405393759409587, "generation_rouge_l_f1": 0.6956521739130436, "generation_token_f1": 0.6956521739130436, "prompt_id": "c5b68b2a56c5748c", "reading_canonical_log_prob": -3.064854383468628, "reading_distances": [48, 43, 38, 41], "reading_log_probs": [-2.071141481399536, -2.494964599609375, -2.3323965072631836, -2.723292589187622], "reading_logprob_std": 0.27455161837001846, "tokenization_path_logprob_std": 49.08680685042078, "trajectory_balance_rmse": 120.54389516239215}210{"canonical_length": 101, "canonical_nll": 1.924111545676052, "generation_rouge_l_f1": 0.234375, "generation_token_f1": 0.359375, "prompt_id": "4478e3a9a8e3ebd2", "reading_canonical_log_prob": -194.33526611328125, "reading_distances": [28, 21, 23, 19], "reading_log_probs": [-196.10623168945312, -195.9689483642578, -192.75547790527344, -190.67845153808594], "reading_logprob_std": 2.635289082713679, "tokenization_path_logprob_std": 95.85275723265573, "trajectory_balance_rmse": 1187.108095931738}211{"canonical_length": 153, "canonical_nll": 1.6524331086601307, "generation_rouge_l_f1": 0.24056603773584903, "generation_token_f1": 0.4245283018867924, "prompt_id": "1679ec014dbdb2b7", "reading_canonical_log_prob": -252.822265625, "reading_distances": [25, 25, 20, 28], "reading_log_probs": [-266.4521789550781, -263.2154541015625, -259.601318359375, -268.8554992675781], "reading_logprob_std": 4.017654561671142, "tokenization_path_logprob_std": 130.65483709059652, "trajectory_balance_rmse": 3223.83124074605}212{"canonical_length": 6, "canonical_nll": 0.8120741844177246, "generation_rouge_l_f1": 0.20833333333333334, "generation_token_f1": 0.20833333333333334, "prompt_id": "d608b8ecb5eb0aa8", "reading_canonical_log_prob": -4.872445106506348, "reading_distances": [6, 12, 17, 13], "reading_log_probs": [-5.190810203552246, -4.534075736999512, -5.322995662689209, -8.725790977478027], "reading_logprob_std": 1.8867244430482621, "tokenization_path_logprob_std": 34.88985982472092, "trajectory_balance_rmse": 133.62545949285678}213{"canonical_length": 8, "canonical_nll": 0.3435807526111603, "generation_rouge_l_f1": 0.923076923076923, "generation_token_f1": 0.923076923076923, "prompt_id": "c7c5da8841670598", "reading_canonical_log_prob": -2.7486460208892822, "reading_distances": [23, 23, 28, 23], "reading_log_probs": [-2.3532485961914062, -2.513376235961914, -2.68441104888916, -1.927518367767334], "reading_logprob_std": 0.32428440940317865, "tokenization_path_logprob_std": 42.692184463621416, "trajectory_balance_rmse": 214.59238870010438}214{"canonical_length": 72, "canonical_nll": 1.6951235665215387, "generation_rouge_l_f1": 0.2763157894736842, "generation_token_f1": 0.4342105263157895, "prompt_id": "8b986ced31f57ea9", "reading_canonical_log_prob": -122.04889678955078, "reading_distances": [31, 31, 35, 33], "reading_log_probs": [-138.27511596679688, -127.1943588256836, -155.0845947265625, -136.41799926757812], "reading_logprob_std": 11.619571223684291, "tokenization_path_logprob_std": 151.7731281328934, "trajectory_balance_rmse": 1488.4813853184826}215{"canonical_length": 288, "canonical_nll": 1.2767437828911676, "generation_rouge_l_f1": 0.20493358633776093, "generation_token_f1": 0.4478178368121442, "prompt_id": "3d6847202e1bf0e3", "reading_canonical_log_prob": -367.70220947265625, "reading_distances": [24, 20, 25, 23], "reading_log_probs": [-375.49517822265625, -371.0196228027344, -369.73931884765625, -371.701416015625], "reading_logprob_std": 2.4749914249708267, "tokenization_path_logprob_std": 315.1708910952582, "trajectory_balance_rmse": 5922.777364284756}216{"canonical_length": 6, "canonical_nll": 0.6992433071136475, "generation_rouge_l_f1": 0.8000000000000002, "generation_token_f1": 0.8000000000000002, "prompt_id": "81930576a5952ad4", "reading_canonical_log_prob": -4.195459842681885, "reading_distances": [48, 43, 45, 39], "reading_log_probs": [-1.9793037176132202, -3.2920403480529785, -2.2023940086364746, -2.0455081462860107], "reading_logprob_std": 0.6153057921387748, "tokenization_path_logprob_std": 18.890995787452606, "trajectory_balance_rmse": 80.4340645319547}217{"canonical_length": 271, "canonical_nll": 1.1150502560323454, "generation_rouge_l_f1": 0.260707635009311, "generation_token_f1": 0.5363128491620113, "prompt_id": "52dc7c85e884efbf", "reading_canonical_log_prob": -302.1786193847656, "reading_distances": [16, 11, 11, 19], "reading_log_probs": [-308.7040710449219, -306.25421142578125, -307.54595947265625, -310.36517333984375], "reading_logprob_std": 1.7468753796881213, "tokenization_path_logprob_std": 129.8533334493668, "trajectory_balance_rmse": 5853.490305134392}218{"canonical_length": 220, "canonical_nll": 1.128641718084162, "generation_rouge_l_f1": 0.2979683972911964, "generation_token_f1": 0.4966139954853273, "prompt_id": "02a68089e544097d", "reading_canonical_log_prob": -248.30117797851562, "reading_distances": [27, 21, 23, 25], "reading_log_probs": [-246.69076538085938, -244.91583251953125, -246.73655700683594, -244.37283325195312], "reading_logprob_std": 1.215262959956489, "tokenization_path_logprob_std": 369.57774585333544, "trajectory_balance_rmse": 3792.047278500309}219{"canonical_length": 116, "canonical_nll": 1.7027141965668777, "generation_rouge_l_f1": 0.27705627705627706, "generation_token_f1": 0.45021645021645024, "prompt_id": "4fcd7cfb63238642", "reading_canonical_log_prob": -197.5148468017578, "reading_distances": [28, 26, 36, 35], "reading_log_probs": [-199.23382568359375, -199.5247802734375, -201.82937622070312, -203.30520629882812], "reading_logprob_std": 1.9403293344010895, "tokenization_path_logprob_std": 60.91163101629825, "trajectory_balance_rmse": 3047.1566467320295}220{"canonical_length": 260, "canonical_nll": 1.5273456280048077, "generation_rouge_l_f1": 0.16296296296296295, "generation_token_f1": 0.3851851851851852, "prompt_id": "0b49b7e13d1bd9b0", "reading_canonical_log_prob": -397.10986328125, "reading_distances": [35, 29, 27, 24], "reading_log_probs": [-407.38720703125, -403.7064208984375, -403.5939636230469, -402.99212646484375], "reading_logprob_std": 2.0028872167496194, "tokenization_path_logprob_std": 184.41498947745538, "trajectory_balance_rmse": 4974.254579075872}221{"canonical_length": 169, "canonical_nll": 1.2326342340051775, "generation_rouge_l_f1": 0.20737327188940094, "generation_token_f1": 0.4193548387096774, "prompt_id": "112cdc2afc089b69", "reading_canonical_log_prob": -208.315185546875, "reading_distances": [23, 27, 25, 31], "reading_log_probs": [-221.88006591796875, -218.67877197265625, -221.676513671875, -221.99281311035156], "reading_logprob_std": 1.5909063890050967, "tokenization_path_logprob_std": 86.9836626856821, "trajectory_balance_rmse": 3040.0027705040548}222{"canonical_length": 250, "canonical_nll": 1.14650048828125, "generation_rouge_l_f1": 0.30452674897119336, "generation_token_f1": 0.5020576131687243, "prompt_id": "845212ac338f2658", "reading_canonical_log_prob": -286.6251220703125, "reading_distances": [41, 42, 28, 40], "reading_log_probs": [-291.2968444824219, -299.10772705078125, -293.2535400390625, -295.329345703125], "reading_logprob_std": 3.3411107238025006, "tokenization_path_logprob_std": 118.98718920105834, "trajectory_balance_rmse": 5292.754779800335}223{"canonical_length": 30, "canonical_nll": 0.5365698496500652, "generation_rouge_l_f1": 0.4193548387096774, "generation_token_f1": 0.5806451612903226, "prompt_id": "9ca4c748ffa8cd48", "reading_canonical_log_prob": -16.097095489501953, "reading_distances": [46, 53, 45, 52], "reading_log_probs": [-15.197861671447754, -14.448274612426758, -15.166574478149414, -15.021445274353027], "reading_logprob_std": 0.3487506491349843, "tokenization_path_logprob_std": 114.2359498286305, "trajectory_balance_rmse": 594.4337987383811}224{"canonical_length": 501, "canonical_nll": 0.5833335769866517, "generation_rouge_l_f1": 0.29009433962264153, "generation_token_f1": 0.4528301886792453, "prompt_id": "585bee96b6b0ece7", "reading_canonical_log_prob": -292.2501220703125, "reading_distances": [19, 22, 22, 17], "reading_log_probs": [-302.2265625, -301.1133117675781, -303.6290283203125, -302.18707275390625], "reading_logprob_std": 1.0315370931842947, "tokenization_path_logprob_std": 154.1228823105096, "trajectory_balance_rmse": 8101.397981956908}225{"canonical_length": 168, "canonical_nll": 1.2066795712425595, "generation_rouge_l_f1": 0.13106796116504857, "generation_token_f1": 0.24271844660194175, "prompt_id": "b3cb4c0e99b9a63e", "reading_canonical_log_prob": -202.72216796875, "reading_distances": [21, 25, 25, 26], "reading_log_probs": [-208.54617309570312, -209.91122436523438, -209.0460205078125, -210.22113037109375], "reading_logprob_std": 0.7715891024573671, "tokenization_path_logprob_std": 214.98294537526118, "trajectory_balance_rmse": 3513.0000089831715}226{"canonical_length": 430, "canonical_nll": 0.9874715405841207, "generation_rouge_l_f1": 0.21693907875185736, "generation_token_f1": 0.424962852897474, "prompt_id": "2fa618fc412d053e", "reading_canonical_log_prob": -424.6127624511719, "reading_distances": [27, 29, 26, 25], "reading_log_probs": [-459.38702392578125, -459.14459228515625, -456.9351806640625, -453.6020812988281], "reading_logprob_std": 2.6808945770810184, "tokenization_path_logprob_std": 168.77711010781957, "trajectory_balance_rmse": 7002.357056989572}227{"canonical_length": 160, "canonical_nll": 1.085921859741211, "generation_rouge_l_f1": 0.26107226107226106, "generation_token_f1": 0.5314685314685315, "prompt_id": "72f894b5c3eed051", "reading_canonical_log_prob": -173.74749755859375, "reading_distances": [27, 26, 28, 25], "reading_log_probs": [-183.71107482910156, -181.74148559570312, -185.94473266601562, -184.1089324951172], "reading_logprob_std": 1.724071074188471, "tokenization_path_logprob_std": 271.5862328335781, "trajectory_balance_rmse": 3344.7274039694257}228{"canonical_length": 98, "canonical_nll": 1.3949510224011479, "generation_rouge_l_f1": 0.22284122562674094, "generation_token_f1": 0.3231197771587743, "prompt_id": "9c00b56f3aa02b5b", "reading_canonical_log_prob": -136.7052001953125, "reading_distances": [37, 31, 42, 34], "reading_log_probs": [-142.61355590820312, -135.6832733154297, -137.62815856933594, -138.5054931640625], "reading_logprob_std": 2.9194234942858976, "tokenization_path_logprob_std": 172.97366727329612, "trajectory_balance_rmse": 2165.939666217033}229{"canonical_length": 10, "canonical_nll": 0.6755030632019043, "generation_rouge_l_f1": 0.4444444444444444, "generation_token_f1": 0.4444444444444444, "prompt_id": "363eae7c1280c1a6", "reading_canonical_log_prob": -6.755030632019043, "reading_distances": [65, 53, 54, 43], "reading_log_probs": [-2.553571939468384, -4.5313849449157715, -3.0991857051849365, -4.176835060119629], "reading_logprob_std": 0.9211700096439104, "tokenization_path_logprob_std": 51.68933061938301, "trajectory_balance_rmse": 210.48886449414695}230{"canonical_length": 7, "canonical_nll": 1.803654670715332, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.6666666666666666, "prompt_id": "86c615edc3ae2a5f", "reading_canonical_log_prob": -12.625582695007324, "reading_distances": [42, 39, 44, 43], "reading_log_probs": [-20.823503494262695, -25.72819709777832, -24.23175811767578, -22.777664184570312], "reading_logprob_std": 2.0926523478056005, "tokenization_path_logprob_std": 22.774349478257463, "trajectory_balance_rmse": 123.34294935252794}231{"canonical_length": 28, "canonical_nll": 0.9224012919834682, "generation_rouge_l_f1": 0.27118644067796616, "generation_token_f1": 0.27118644067796616, "prompt_id": "b424089c85e11874", "reading_canonical_log_prob": -25.82723617553711, "reading_distances": [23, 24, 20, 16], "reading_log_probs": [-27.790895462036133, -27.551193237304688, -26.780292510986328, -28.637008666992188], "reading_logprob_std": 0.7646001677331964, "tokenization_path_logprob_std": 66.8185098157577, "trajectory_balance_rmse": 523.7601534462649}232{"canonical_length": 7, "canonical_nll": 0.18448638916015625, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "69e0fa630dc3c290", "reading_canonical_log_prob": -1.2914047241210938, "reading_distances": [41, 32, 25, 23], "reading_log_probs": [-3.6350131034851074, -4.451851844787598, -3.775564432144165, -3.5716054439544678], "reading_logprob_std": 0.4046395923236925, "tokenization_path_logprob_std": 13.774925354394831, "trajectory_balance_rmse": 101.33553880089637}233{"canonical_length": 158, "canonical_nll": 1.3048841259147548, "generation_rouge_l_f1": 0.20652173913043478, "generation_token_f1": 0.3804347826086957, "prompt_id": "edfb4a427bd0504a", "reading_canonical_log_prob": -206.17169189453125, "reading_distances": [31, 25, 27, 23], "reading_log_probs": [-213.89691162109375, -211.21102905273438, -208.45193481445312, -208.1778106689453], "reading_logprob_std": 2.684182248583967, "tokenization_path_logprob_std": 127.15538163627677, "trajectory_balance_rmse": 3263.6341915528924}234{"canonical_length": 9, "canonical_nll": 0.38617833455403644, "generation_rouge_l_f1": 0.8421052631578948, "generation_token_f1": 0.8421052631578948, "prompt_id": "9a5bfd989c2d5e52", "reading_canonical_log_prob": -3.475605010986328, "reading_distances": [37, 40, 36, 44], "reading_log_probs": [-3.891145706176758, -5.418067932128906, -4.371074199676514, -3.4802913665771484], "reading_logprob_std": 0.8354286365337721, "tokenization_path_logprob_std": 35.95963969650777, "trajectory_balance_rmse": 126.29549619939462}235{"canonical_length": 272, "canonical_nll": 0.8588823430678424, "generation_rouge_l_f1": 0.41637010676156583, "generation_token_f1": 0.5836298932384343, "prompt_id": "a2cdcfce41950902", "reading_canonical_log_prob": -233.61599731445312, "reading_distances": [29, 18, 27, 24], "reading_log_probs": [-243.7224884033203, -241.18978881835938, -244.618408203125, -244.13763427734375], "reading_logprob_std": 1.5293231856886504, "tokenization_path_logprob_std": 233.67469214366602, "trajectory_balance_rmse": 3519.0503116203795}236{"canonical_length": 72, "canonical_nll": 1.3114681243896484, "generation_rouge_l_f1": 0.20261437908496732, "generation_token_f1": 0.30065359477124187, "prompt_id": "41dfa00e84da9369", "reading_canonical_log_prob": -94.42570495605469, "reading_distances": [44, 37, 39, 41], "reading_log_probs": [-112.98055267333984, -108.263916015625, -118.80311584472656, -116.8988037109375], "reading_logprob_std": 4.661580719836471, "tokenization_path_logprob_std": 90.6006739321975, "trajectory_balance_rmse": 831.2270000902117}237{"canonical_length": 13, "canonical_nll": 0.979535176203801, "generation_rouge_l_f1": 0.06324110671936758, "generation_token_f1": 0.07905138339920949, "prompt_id": "398d73f7e39625a8", "reading_canonical_log_prob": -12.733957290649414, "reading_distances": [27, 27, 24, 31], "reading_log_probs": [-18.91150665283203, -19.644580841064453, -22.35198974609375, -17.570968627929688], "reading_logprob_std": 2.013691829255116, "tokenization_path_logprob_std": 34.08411525331541, "trajectory_balance_rmse": 143.22407829753988}238{"canonical_length": 89, "canonical_nll": 1.393503767720769, "generation_rouge_l_f1": 0.28820960698689957, "generation_token_f1": 0.445414847161572, "prompt_id": "d769bfcaa2a347d7", "reading_canonical_log_prob": -124.02183532714844, "reading_distances": [67, 74, 63, 79], "reading_log_probs": [-125.32975006103516, -127.72976684570312, -124.21145629882812, -125.16958618164062], "reading_logprob_std": 1.4968678683273813, "tokenization_path_logprob_std": 98.33681434617124, "trajectory_balance_rmse": 1966.4915044072688}239{"canonical_length": 351, "canonical_nll": 1.3342682493378295, "generation_rouge_l_f1": 0.2288261515601783, "generation_token_f1": 0.3833580980683507, "prompt_id": "026d0e5d489e80b2", "reading_canonical_log_prob": -468.3281555175781, "reading_distances": [7, 12, 10, 9], "reading_log_probs": [-474.92578125, -481.90008544921875, -481.54949951171875, -475.6282958984375], "reading_logprob_std": 3.7363863603301315, "tokenization_path_logprob_std": 56.61418313853281, "trajectory_balance_rmse": 7210.613331329534}240{"canonical_length": 55, "canonical_nll": 1.1180191733620384, "generation_rouge_l_f1": 0.3220338983050847, "generation_token_f1": 0.38983050847457623, "prompt_id": "4c12c24de5859788", "reading_canonical_log_prob": -61.49105453491211, "reading_distances": [20, 25, 24, 27], "reading_log_probs": [-62.79865264892578, -62.19993591308594, -62.35779571533203, -66.1134262084961], "reading_logprob_std": 1.8480987240836917, "tokenization_path_logprob_std": 86.44583270540849, "trajectory_balance_rmse": 1269.3256956448315}241{"canonical_length": 298, "canonical_nll": 1.7559603492685611, "generation_rouge_l_f1": 0.18114602587800369, "generation_token_f1": 0.36598890942698714, "prompt_id": "1259351bdd54ab2c", "reading_canonical_log_prob": -523.2761840820312, "reading_distances": [18, 17, 17, 15], "reading_log_probs": [-528.4362182617188, -528.07275390625, -528.690185546875, -530.44677734375], "reading_logprob_std": 1.0544262814226286, "tokenization_path_logprob_std": 121.81680599292014, "trajectory_balance_rmse": 5409.13328838983}242{"canonical_length": 17, "canonical_nll": 1.4025582706227022, "generation_rouge_l_f1": 0.07168458781362008, "generation_token_f1": 0.07168458781362008, "prompt_id": "1e2afb46510ad456", "reading_canonical_log_prob": -23.843490600585938, "reading_distances": [26, 30, 36, 28], "reading_log_probs": [-36.6720085144043, -35.32721710205078, -43.84822463989258, -47.387611389160156], "reading_logprob_std": 5.764247747779135, "tokenization_path_logprob_std": 58.31003183864337, "trajectory_balance_rmse": 290.0399950392798}243{"canonical_length": 23, "canonical_nll": 1.305503098861031, "generation_rouge_l_f1": 0.10144927536231883, "generation_token_f1": 0.10144927536231883, "prompt_id": "dfa52b61eae824fc", "reading_canonical_log_prob": -30.02657127380371, "reading_distances": [14, 16, 20, 16], "reading_log_probs": [-26.327220916748047, -27.305757522583008, -27.023014068603516, -25.083667755126953], "reading_logprob_std": 0.9902347105569669, "tokenization_path_logprob_std": 27.873599019120068, "trajectory_balance_rmse": 257.5940968403463}244{"canonical_length": 5, "canonical_nll": 1.1991461753845214, "generation_rouge_l_f1": 0.4444444444444445, "generation_token_f1": 0.4444444444444445, "prompt_id": "30fd5bfb1635c7a5", "reading_canonical_log_prob": -5.995730876922607, "reading_distances": [19, 19, 28, 23], "reading_log_probs": [-11.692852973937988, -9.757084846496582, -11.098726272583008, -9.798137664794922], "reading_logprob_std": 0.9653733586878737, "tokenization_path_logprob_std": 9.680748675399009, "trajectory_balance_rmse": 14.851865305718434}245{"canonical_length": 11, "canonical_nll": 0.26299166679382324, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "bd172f5e24a2f430", "reading_canonical_log_prob": -2.8929083347320557, "reading_distances": [36, 40, 44, 39], "reading_log_probs": [-2.8635149002075195, -3.204739570617676, -2.233527898788452, -2.6895804405212402], "reading_logprob_std": 0.4041654009148393, "tokenization_path_logprob_std": 25.947922200985857, "trajectory_balance_rmse": 171.71778256386503}246{"canonical_length": 8, "canonical_nll": 0.13502609729766846, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "74eb19d7e1933dab", "reading_canonical_log_prob": -1.0802087783813477, "reading_distances": [27, 23, 22, 23], "reading_log_probs": [-1.953102946281433, -2.5129873752593994, -1.8272545337677002, -1.4709663391113281], "reading_logprob_std": 0.4325065271226731, "tokenization_path_logprob_std": 37.23015743900947, "trajectory_balance_rmse": 161.6600158889415}247{"canonical_length": 2, "canonical_nll": 1.7751214504241943, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "69770982bd8468eb", "reading_canonical_log_prob": -3.5502429008483887, "reading_distances": [27, 24, 27, 30], "reading_log_probs": [-3.0115928649902344, -3.2603137493133545, -3.262455940246582, -2.5286781787872314], "reading_logprob_std": 0.345413514890663, "tokenization_path_logprob_std": 17.6030092427689, "trajectory_balance_rmse": 26.671744666215773}248{"canonical_length": 33, "canonical_nll": 0.7175994641853102, "generation_rouge_l_f1": 0.16546762589928057, "generation_token_f1": 0.2014388489208633, "prompt_id": "decae069198bb200", "reading_canonical_log_prob": -23.680782318115234, "reading_distances": [16, 19, 18, 15], "reading_log_probs": [-22.633872985839844, -24.65536880493164, -22.88011360168457, -22.118995666503906], "reading_logprob_std": 1.1021266823249416, "tokenization_path_logprob_std": 68.107369671343, "trajectory_balance_rmse": 483.31132922083214}249{"canonical_length": 296, "canonical_nll": 1.506135064202386, "generation_rouge_l_f1": 0.20614828209764918, "generation_token_f1": 0.4954792043399638, "prompt_id": "78378b5209c8e413", "reading_canonical_log_prob": -445.81597900390625, "reading_distances": [15, 20, 20, 20], "reading_log_probs": [-453.4945068359375, -451.93817138671875, -455.6890869140625, -456.0087890625], "reading_logprob_std": 1.9214031739845077, "tokenization_path_logprob_std": 295.6867393095692, "trajectory_balance_rmse": 5944.754121910845}250{"canonical_length": 46, "canonical_nll": 0.945492288340693, "generation_rouge_l_f1": 0.19178082191780824, "generation_token_f1": 0.29223744292237447, "prompt_id": "6c87b1ea5d43b53e", "reading_canonical_log_prob": -43.492645263671875, "reading_distances": [22, 18, 15, 22], "reading_log_probs": [-49.58728790283203, -48.269866943359375, -46.78033447265625, -47.20465850830078], "reading_logprob_std": 1.25249535108726, "tokenization_path_logprob_std": 59.83210802268152, "trajectory_balance_rmse": 595.152863348236}251{"canonical_length": 120, "canonical_nll": 2.1858037312825522, "generation_rouge_l_f1": 0.15625000000000003, "generation_token_f1": 0.2760416666666667, "prompt_id": "c823f34b5de71522", "reading_canonical_log_prob": -262.29644775390625, "reading_distances": [21, 26, 20, 22], "reading_log_probs": [-258.04443359375, -259.1336669921875, -261.27752685546875, -261.5220947265625], "reading_logprob_std": 1.685575667964532, "tokenization_path_logprob_std": 118.46398473614698, "trajectory_balance_rmse": 2776.953261850539}252{"canonical_length": 23, "canonical_nll": 1.1952715334684954, "generation_rouge_l_f1": 0.15384615384615383, "generation_token_f1": 0.23076923076923075, "prompt_id": "95ee759ef85d853e", "reading_canonical_log_prob": -27.49124526977539, "reading_distances": [29, 27, 29, 26], "reading_log_probs": [-29.060749053955078, -29.964523315429688, -30.240962982177734, -31.703062057495117], "reading_logprob_std": 1.0965172286915847, "tokenization_path_logprob_std": 45.03446405544156, "trajectory_balance_rmse": 226.26337364970686}253{"canonical_length": 11, "canonical_nll": 0.752770337191495, "generation_rouge_l_f1": 0.7058823529411765, "generation_token_f1": 0.7058823529411765, "prompt_id": "036acca12e6542a8", "reading_canonical_log_prob": -8.280473709106445, "reading_distances": [18, 20, 25, 21], "reading_log_probs": [-8.756172180175781, -8.552827835083008, -9.01192569732666, -9.193500518798828], "reading_logprob_std": 0.28169377737595674, "tokenization_path_logprob_std": 37.43312699843526, "trajectory_balance_rmse": 186.86105121555875}254{"canonical_length": 13, "canonical_nll": 1.225680791414701, "generation_rouge_l_f1": 0.4615384615384615, "generation_token_f1": 0.4615384615384615, "prompt_id": "5e5a1ac1eccda1ac", "reading_canonical_log_prob": -15.933850288391113, "reading_distances": [20, 36, 27, 27], "reading_log_probs": [-16.56709861755371, -17.693471908569336, -19.61092758178711, -17.05092430114746], "reading_logprob_std": 1.3357525759463855, "tokenization_path_logprob_std": 106.78686011842738, "trajectory_balance_rmse": 366.9988078275745}255{"canonical_length": 8, "canonical_nll": 0.11748841404914856, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "05b4960bf40b9d90", "reading_canonical_log_prob": -0.9399073123931885, "reading_distances": [24, 30, 24, 25], "reading_log_probs": [-2.447437047958374, -1.1648701429367065, -2.2321739196777344, -1.4836642742156982], "reading_logprob_std": 0.6069876288549166, "tokenization_path_logprob_std": 35.31979262481461, "trajectory_balance_rmse": 140.3246404964965}256{"canonical_length": 199, "canonical_nll": 1.2776384976641018, "generation_rouge_l_f1": 0.31947483588621445, "generation_token_f1": 0.5689277899343546, "prompt_id": "55f01821168c15c2", "reading_canonical_log_prob": -254.25006103515625, "reading_distances": [29, 33, 37, 34], "reading_log_probs": [-255.8157958984375, -255.5681915283203, -255.8306427001953, -258.1424560546875], "reading_logprob_std": 1.208134763786871, "tokenization_path_logprob_std": 72.69302472397857, "trajectory_balance_rmse": 3780.5686860466744}257{"canonical_length": 181, "canonical_nll": 0.9364586087221599, "generation_rouge_l_f1": 0.31890660592255127, "generation_token_f1": 0.469248291571754, "prompt_id": "28aed3c6f636dfb4", "reading_canonical_log_prob": -169.49900817871094, "reading_distances": [13, 12, 11, 15], "reading_log_probs": [-170.25790405273438, -167.82896423339844, -170.38621520996094, -168.51734924316406], "reading_logprob_std": 1.2731787621538186, "tokenization_path_logprob_std": 146.58316592176118, "trajectory_balance_rmse": 3271.5961736347244}258{"canonical_length": 102, "canonical_nll": 2.0386003980449603, "generation_rouge_l_f1": 0.1890243902439024, "generation_token_f1": 0.426829268292683, "prompt_id": "33e9d73a4126f71a", "reading_canonical_log_prob": -207.93724060058594, "reading_distances": [18, 20, 12, 20], "reading_log_probs": [-205.3733673095703, -208.05445861816406, -207.9438934326172, -208.33541870117188], "reading_logprob_std": 1.3788297835842445, "tokenization_path_logprob_std": 187.46724296606754, "trajectory_balance_rmse": 2418.098209139256}259{"canonical_length": 169, "canonical_nll": 1.9017584986940643, "generation_rouge_l_f1": 0.23908045977011494, "generation_token_f1": 0.45977011494252873, "prompt_id": "ba9c16c03160cc50", "reading_canonical_log_prob": -321.3971862792969, "reading_distances": [23, 21, 23, 16], "reading_log_probs": [-327.2271423339844, -326.92169189453125, -324.7744140625, -322.59716796875], "reading_logprob_std": 2.152489806262452, "tokenization_path_logprob_std": 264.1484916486938, "trajectory_balance_rmse": 3169.5354866536863}260{"canonical_length": 278, "canonical_nll": 0.7183388360112691, "generation_rouge_l_f1": 0.25447316103379725, "generation_token_f1": 0.3618290258449304, "prompt_id": "d04a82b0b7e2a434", "reading_canonical_log_prob": -199.6981964111328, "reading_distances": [23, 27, 24, 28], "reading_log_probs": [-187.03488159179688, -191.01564025878906, -193.22171020507812, -190.49441528320312], "reading_logprob_std": 2.560407176129471, "tokenization_path_logprob_std": 103.55258071511382, "trajectory_balance_rmse": 3258.8658063627668}261{"canonical_length": 38, "canonical_nll": 0.6808747743305407, "generation_rouge_l_f1": 0.3409090909090909, "generation_token_f1": 0.38636363636363635, "prompt_id": "e1a492712f46f15d", "reading_canonical_log_prob": -25.873241424560547, "reading_distances": [19, 21, 19, 17], "reading_log_probs": [-27.049175262451172, -30.38935089111328, -27.34511947631836, -29.46578598022461], "reading_logprob_std": 1.6253693693415836, "tokenization_path_logprob_std": 67.18427235303281, "trajectory_balance_rmse": 554.1182321453048}262{"canonical_length": 25, "canonical_nll": 0.5811145782470704, "generation_rouge_l_f1": 0.14046822742474915, "generation_token_f1": 0.14046822742474915, "prompt_id": "fcc6cd8702ae37ce", "reading_canonical_log_prob": -14.527864456176758, "reading_distances": [23, 26, 26, 22], "reading_log_probs": [-14.715049743652344, -17.653926849365234, -16.133237838745117, -17.203289031982422], "reading_logprob_std": 1.3070376791674285, "tokenization_path_logprob_std": 22.86668554804116, "trajectory_balance_rmse": 285.4447794271348}263{"canonical_length": 342, "canonical_nll": 1.2200527971948099, "generation_rouge_l_f1": 0.24999999999999994, "generation_token_f1": 0.46381578947368424, "prompt_id": "38014139c1d9865d", "reading_canonical_log_prob": -417.258056640625, "reading_distances": [32, 30, 25, 38], "reading_log_probs": [-430.8861083984375, -431.77484130859375, -434.0836181640625, -437.4408264160156], "reading_logprob_std": 2.9252197670956015, "tokenization_path_logprob_std": 156.58493781583124, "trajectory_balance_rmse": 5714.899093487363}264{"canonical_length": 6, "canonical_nll": 0.5518355369567871, "generation_rouge_l_f1": 0.8000000000000002, "generation_token_f1": 0.8000000000000002, "prompt_id": "bd70ea68bde60a24", "reading_canonical_log_prob": -3.3110132217407227, "reading_distances": [25, 25, 19, 25], "reading_log_probs": [-2.716811418533325, -3.321553945541382, -2.431436777114868, -3.027616262435913], "reading_logprob_std": 0.38491230329396964, "tokenization_path_logprob_std": 15.9407547310517, "trajectory_balance_rmse": 106.73781328984641}265{"canonical_length": 11, "canonical_nll": 2.3421573638916016, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "57765c15fcf7839d", "reading_canonical_log_prob": -25.763731002807617, "reading_distances": [10, 12, 12, 14], "reading_log_probs": [-24.38497543334961, -25.47216796875, -24.682708740234375, -24.99078369140625], "reading_logprob_std": 0.46435647657899387, "tokenization_path_logprob_std": 40.39360181994491, "trajectory_balance_rmse": 200.74718532529275}266{"canonical_length": 50, "canonical_nll": 1.4618154907226562, "generation_rouge_l_f1": 0.25, "generation_token_f1": 0.29545454545454547, "prompt_id": "253987951dace9e0", "reading_canonical_log_prob": -73.09077453613281, "reading_distances": [32, 23, 29, 26], "reading_log_probs": [-72.21365356445312, -69.9797134399414, -71.19098663330078, -71.0291748046875], "reading_logprob_std": 0.9144242741217387, "tokenization_path_logprob_std": 135.25178804575125, "trajectory_balance_rmse": 1127.8391897811427}267{"canonical_length": 38, "canonical_nll": 1.2579662925318669, "generation_rouge_l_f1": 0.30656934306569344, "generation_token_f1": 0.3503649635036496, "prompt_id": "e17273b705c2c377", "reading_canonical_log_prob": -47.80271911621094, "reading_distances": [29, 22, 28, 29], "reading_log_probs": [-56.80362319946289, -51.66608428955078, -51.92521667480469, -53.94892120361328], "reading_logprob_std": 2.375506152118087, "tokenization_path_logprob_std": 48.693883969690354, "trajectory_balance_rmse": 737.7124351686963}268{"canonical_length": 344, "canonical_nll": 1.5900037898573764, "generation_rouge_l_f1": 0.23240589198036007, "generation_token_f1": 0.46808510638297873, "prompt_id": "3e31d044adc4af42", "reading_canonical_log_prob": -546.9613037109375, "reading_distances": [23, 27, 31, 26], "reading_log_probs": [-542.5718994140625, -545.3155517578125, -545.4965209960938, -546.1376953125], "reading_logprob_std": 1.5789161462983305, "tokenization_path_logprob_std": 182.32294445963535, "trajectory_balance_rmse": 8062.456273713472}269{"canonical_length": 23, "canonical_nll": 0.8669013977050781, "generation_rouge_l_f1": 0.41379310344827586, "generation_token_f1": 0.4482758620689656, "prompt_id": "31f5a769dd1b152b", "reading_canonical_log_prob": -19.938732147216797, "reading_distances": [24, 19, 24, 19], "reading_log_probs": [-21.260570526123047, -21.46466064453125, -21.091442108154297, -21.136289596557617], "reading_logprob_std": 0.1670413785834907, "tokenization_path_logprob_std": 31.66998600754786, "trajectory_balance_rmse": 495.03028563704333}270{"canonical_length": 22, "canonical_nll": 1.12218562039462, "generation_rouge_l_f1": 0.7333333333333333, "generation_token_f1": 0.7333333333333333, "prompt_id": "74fb912570f58841", "reading_canonical_log_prob": -24.68808364868164, "reading_distances": [26, 24, 17, 29], "reading_log_probs": [-26.032453536987305, -23.573192596435547, -24.28256607055664, -25.512304306030273], "reading_logprob_std": 1.1238418786028743, "tokenization_path_logprob_std": 45.59534642102105, "trajectory_balance_rmse": 221.18042950467503}271{"canonical_length": 10, "canonical_nll": 0.4536423683166504, "generation_rouge_l_f1": 0.17391304347826086, "generation_token_f1": 0.17391304347826086, "prompt_id": "434a6990c55f121a", "reading_canonical_log_prob": -4.536423683166504, "reading_distances": [14, 20, 19, 15], "reading_log_probs": [-6.1137166023254395, -6.093165397644043, -7.930542945861816, -4.673203945159912], "reading_logprob_std": 1.3347553201731077, "tokenization_path_logprob_std": 36.106860696973115, "trajectory_balance_rmse": 228.70209170896197}272{"canonical_length": 222, "canonical_nll": 1.1682493192655545, "generation_rouge_l_f1": 0.225130890052356, "generation_token_f1": 0.4450261780104712, "prompt_id": "2e39c9a4bb908fa5", "reading_canonical_log_prob": -259.3513488769531, "reading_distances": [16, 23, 18, 23], "reading_log_probs": [-260.7913513183594, -262.21728515625, -259.97308349609375, -260.2344665527344], "reading_logprob_std": 1.0020521439100423, "tokenization_path_logprob_std": 104.00969393295522, "trajectory_balance_rmse": 5483.1025147710225}273{"canonical_length": 226, "canonical_nll": 1.5365533069171737, "generation_rouge_l_f1": 0.1714285714285714, "generation_token_f1": 0.29890109890109895, "prompt_id": "1a576b185bace9d2", "reading_canonical_log_prob": -347.26104736328125, "reading_distances": [10, 16, 12, 16], "reading_log_probs": [-349.5121765136719, -347.06170654296875, -345.53472900390625, -357.88031005859375], "reading_logprob_std": 5.504829431257455, "tokenization_path_logprob_std": 185.89483272982253, "trajectory_balance_rmse": 4099.275086487588}274{"canonical_length": 270, "canonical_nll": 1.1776294849537037, "generation_rouge_l_f1": 0.2681564245810056, "generation_token_f1": 0.5400372439478586, "prompt_id": "c2df8f1e4778ecd3", "reading_canonical_log_prob": -317.9599609375, "reading_distances": [35, 28, 35, 35], "reading_log_probs": [-319.71759033203125, -321.6524658203125, -320.58245849609375, -322.0745544433594], "reading_logprob_std": 1.0644421353414466, "tokenization_path_logprob_std": 150.74038751434804, "trajectory_balance_rmse": 5685.180332086187}275{"canonical_length": 2, "canonical_nll": 1.3475195169448853, "generation_rouge_l_f1": 0.022988505747126436, "generation_token_f1": 0.022988505747126436, "prompt_id": "42c831675952234b", "reading_canonical_log_prob": -2.6950390338897705, "reading_distances": [68, 67, 67, 49], "reading_log_probs": [-0.7835102677345276, -2.2077293395996094, -1.4160490036010742, -3.2224955558776855], "reading_logprob_std": 1.052651882436005, "tokenization_path_logprob_std": 7.321127908001274, "trajectory_balance_rmse": 24.547909330179653}276{"canonical_length": 16, "canonical_nll": 2.6162567138671875, "generation_rouge_l_f1": 0.21621621621621623, "generation_token_f1": 0.27027027027027023, "prompt_id": "a9f4c15e6cb406fd", "reading_canonical_log_prob": -41.860107421875, "reading_distances": [21, 18, 20, 19], "reading_log_probs": [-45.03369140625, -51.295555114746094, -54.982791900634766, -55.88299560546875], "reading_logprob_std": 4.927420145130567, "tokenization_path_logprob_std": 50.476305133618126, "trajectory_balance_rmse": 430.8240966832067}277{"canonical_length": 161, "canonical_nll": 1.7674983243764557, "generation_rouge_l_f1": 0.165, "generation_token_f1": 0.33999999999999997, "prompt_id": "25e2da0cacffb5d9", "reading_canonical_log_prob": -284.5672302246094, "reading_distances": [23, 21, 19, 23], "reading_log_probs": [-282.2342224121094, -284.07159423828125, -285.7535705566406, -284.8739318847656], "reading_logprob_std": 1.4993433817731325, "tokenization_path_logprob_std": 195.85199209564036, "trajectory_balance_rmse": 3337.515791067373}278{"canonical_length": 116, "canonical_nll": 1.1804333390860722, "generation_rouge_l_f1": 0.19117647058823525, "generation_token_f1": 0.29411764705882354, "prompt_id": "d71e3178e4fc08e6", "reading_canonical_log_prob": -136.93026733398438, "reading_distances": [36, 35, 42, 35], "reading_log_probs": [-141.82850646972656, -140.83641052246094, -142.2989501953125, -143.77178955078125], "reading_logprob_std": 1.2215648741887908, "tokenization_path_logprob_std": 48.36860856240383, "trajectory_balance_rmse": 3050.7894906810347}279{"canonical_length": 9, "canonical_nll": 1.0147751702202692, "generation_rouge_l_f1": 0.2105263157894737, "generation_token_f1": 0.2105263157894737, "prompt_id": "c69244953386b7ea", "reading_canonical_log_prob": -9.132976531982422, "reading_distances": [16, 19, 15, 5], "reading_log_probs": [-10.384641647338867, -9.780542373657227, -10.41486644744873, -8.031624794006348], "reading_logprob_std": 1.1196524123129077, "tokenization_path_logprob_std": 21.554142599635796, "trajectory_balance_rmse": 212.0970907370001}280{"canonical_length": 16, "canonical_nll": 1.9754126071929932, "generation_rouge_l_f1": 0.23728813559322035, "generation_token_f1": 0.2711864406779661, "prompt_id": "44b99fb81cd83d68", "reading_canonical_log_prob": -31.60660171508789, "reading_distances": [25, 21, 24, 30], "reading_log_probs": [-31.442838668823242, -30.381996154785156, -32.71379852294922, -29.52735137939453], "reading_logprob_std": 1.3763170274716072, "tokenization_path_logprob_std": 76.91240700107433, "trajectory_balance_rmse": 396.9519197826757}281{"canonical_length": 23, "canonical_nll": 0.27188929267551587, "generation_rouge_l_f1": 0.3333333333333333, "generation_token_f1": 0.3333333333333333, "prompt_id": "f3f998d8d70ee428", "reading_canonical_log_prob": -6.253453731536865, "reading_distances": [51, 41, 40, 42], "reading_log_probs": [-8.737499237060547, -10.437334060668945, -8.469767570495605, -9.296992301940918], "reading_logprob_std": 0.8722647313530739, "tokenization_path_logprob_std": 20.977186730495195, "trajectory_balance_rmse": 298.6243653038239}282{"canonical_length": 45, "canonical_nll": 0.817822265625, "generation_rouge_l_f1": 0.2463768115942029, "generation_token_f1": 0.34782608695652173, "prompt_id": "8565123aabdca8e0", "reading_canonical_log_prob": -36.802001953125, "reading_distances": [21, 22, 24, 22], "reading_log_probs": [-45.79608154296875, -42.84437942504883, -44.22321701049805, -43.58203125], "reading_logprob_std": 1.256480183276512, "tokenization_path_logprob_std": 39.68382901274896, "trajectory_balance_rmse": 967.6732217136275}283{"canonical_length": 293, "canonical_nll": 1.6397434273677474, "generation_rouge_l_f1": 0.21532846715328469, "generation_token_f1": 0.4416058394160584, "prompt_id": "c754674f366c507b", "reading_canonical_log_prob": -480.44482421875, "reading_distances": [27, 30, 36, 32], "reading_log_probs": [-476.1412048339844, -481.3006591796875, -478.39190673828125, -481.1693420410156], "reading_logprob_std": 2.4691433028026717, "tokenization_path_logprob_std": 116.92867290795527, "trajectory_balance_rmse": 6515.567303399202}284{"canonical_length": 15, "canonical_nll": 1.2183638254801432, "generation_rouge_l_f1": 0.10596026490066225, "generation_token_f1": 0.10596026490066225, "prompt_id": "f7c03b54285f64b4", "reading_canonical_log_prob": -18.27545738220215, "reading_distances": [26, 26, 29, 25], "reading_log_probs": [-16.43271255493164, -19.660913467407227, -15.056312561035156, -16.05518341064453], "reading_logprob_std": 1.992897138437246, "tokenization_path_logprob_std": 82.5968371967089, "trajectory_balance_rmse": 270.1925179558276}285{"canonical_length": 33, "canonical_nll": 0.36456657178474194, "generation_rouge_l_f1": 0.1282051282051282, "generation_token_f1": 0.15384615384615385, "prompt_id": "278cabb1c777cfb5", "reading_canonical_log_prob": -12.030696868896484, "reading_distances": [58, 44, 50, 56], "reading_log_probs": [-20.82574462890625, -28.653667449951172, -22.63113021850586, -24.304866790771484], "reading_logprob_std": 3.3494340296448817, "tokenization_path_logprob_std": 78.00203052693165, "trajectory_balance_rmse": 438.92489135841384}286{"canonical_length": 82, "canonical_nll": 0.494830945642983, "generation_rouge_l_f1": 0.4088888888888889, "generation_token_f1": 0.4533333333333333, "prompt_id": "26496139eab8348f", "reading_canonical_log_prob": -40.57613754272461, "reading_distances": [17, 20, 18, 17], "reading_log_probs": [-44.98204803466797, -45.753318786621094, -45.05589294433594, -44.83495330810547], "reading_logprob_std": 0.4083055210826355, "tokenization_path_logprob_std": 44.24902957658725, "trajectory_balance_rmse": 801.1782963620309}287{"canonical_length": 130, "canonical_nll": 1.9943026029146635, "generation_rouge_l_f1": 0.16666666666666669, "generation_token_f1": 0.3585858585858586, "prompt_id": "5e25b104005e6f72", "reading_canonical_log_prob": -259.25933837890625, "reading_distances": [25, 23, 23, 19], "reading_log_probs": [-268.2918701171875, -262.57342529296875, -261.0420227050781, -250.60891723632812], "reading_logprob_std": 7.372765525561365, "tokenization_path_logprob_std": 229.4139162245407, "trajectory_balance_rmse": 1725.1017147150278}288{"canonical_length": 107, "canonical_nll": 2.410878689489632, "generation_rouge_l_f1": 0.24669603524229075, "generation_token_f1": 0.4581497797356828, "prompt_id": "14f8273006e150ca", "reading_canonical_log_prob": -257.9640197753906, "reading_distances": [25, 25, 34, 30], "reading_log_probs": [-262.88739013671875, -258.6769104003906, -262.11309814453125, -260.0794372558594], "reading_logprob_std": 1.9175181764399019, "tokenization_path_logprob_std": 118.99371959655102, "trajectory_balance_rmse": 2041.8238048699177}289{"canonical_length": 185, "canonical_nll": 1.4674786541913007, "generation_rouge_l_f1": 0.3112391930835735, "generation_token_f1": 0.478386167146974, "prompt_id": "c0e6a5cad32d62b6", "reading_canonical_log_prob": -271.4835510253906, "reading_distances": [13, 17, 17, 12], "reading_log_probs": [-279.75531005859375, -279.0694274902344, -275.00360107421875, -277.07574462890625], "reading_logprob_std": 2.141431350944632, "tokenization_path_logprob_std": 243.75389080955674, "trajectory_balance_rmse": 3117.825572816133}290{"canonical_length": 16, "canonical_nll": 1.0834076404571533, "generation_rouge_l_f1": 0.4, "generation_token_f1": 0.4571428571428572, "prompt_id": "937bfa51bfbd0542", "reading_canonical_log_prob": -17.334522247314453, "reading_distances": [31, 34, 37, 39], "reading_log_probs": [-22.574331283569336, -22.833839416503906, -21.635696411132812, -18.855396270751953], "reading_logprob_std": 1.8205421106608681, "tokenization_path_logprob_std": 18.302943668380138, "trajectory_balance_rmse": 429.3900528285655}291{"canonical_length": 75, "canonical_nll": 0.4320518493652344, "generation_rouge_l_f1": 0.28664495114006516, "generation_token_f1": 0.3127035830618892, "prompt_id": "19b52394b9bc5537", "reading_canonical_log_prob": -32.40388870239258, "reading_distances": [13, 13, 17, 12], "reading_log_probs": [-34.341590881347656, -34.520263671875, -34.51499938964844, -32.93827438354492], "reading_logprob_std": 0.7648568077186007, "tokenization_path_logprob_std": 58.642113332163746, "trajectory_balance_rmse": 538.2035967015161}292{"canonical_length": 339, "canonical_nll": 1.4618659005404222, "generation_rouge_l_f1": 0.22463768115942032, "generation_token_f1": 0.36231884057971014, "prompt_id": "c03437022c1ce69c", "reading_canonical_log_prob": -495.5725402832031, "reading_distances": [29, 29, 24, 28], "reading_log_probs": [-504.65228271484375, -504.0814208984375, -502.28564453125, -507.03680419921875], "reading_logprob_std": 1.9609823988020914, "tokenization_path_logprob_std": 213.72586811957459, "trajectory_balance_rmse": 6036.010432100742}293{"canonical_length": 22, "canonical_nll": 0.7633348811756481, "generation_rouge_l_f1": 0.0842911877394636, "generation_token_f1": 0.10727969348659006, "prompt_id": "01ef5d8df6df3761", "reading_canonical_log_prob": -16.793367385864258, "reading_distances": [21, 20, 21, 26], "reading_log_probs": [-21.09416389465332, -21.3870849609375, -20.009525299072266, -24.920150756835938], "reading_logprob_std": 2.1290625094603155, "tokenization_path_logprob_std": 49.10548299112536, "trajectory_balance_rmse": 359.4974837433676}294{"canonical_length": 110, "canonical_nll": 1.6866174871271307, "generation_rouge_l_f1": 0.23218997361477572, "generation_token_f1": 0.33245382585751976, "prompt_id": "1275798cfa3be44f", "reading_canonical_log_prob": -185.52792358398438, "reading_distances": [36, 34, 37, 38], "reading_log_probs": [-190.3336181640625, -193.10720825195312, -190.46395874023438, -195.1759033203125], "reading_logprob_std": 2.3206716909956704, "tokenization_path_logprob_std": 121.32189684967484, "trajectory_balance_rmse": 2678.2944789039666}295{"canonical_length": 4, "canonical_nll": 0.30223894119262695, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "1f2fdcd7eec76cc5", "reading_canonical_log_prob": -1.2089557647705078, "reading_distances": [29, 29, 25, 33], "reading_log_probs": [-2.6569883823394775, -1.3799457550048828, -0.7706301212310791, -1.2795040607452393], "reading_logprob_std": 0.8024418555957323, "tokenization_path_logprob_std": 29.811001605970493, "trajectory_balance_rmse": 52.012149516197596}296{"canonical_length": 128, "canonical_nll": 1.3304717540740967, "generation_rouge_l_f1": 0.11285266457680251, "generation_token_f1": 0.13793103448275862, "prompt_id": "b8b97c6b0ef87a98", "reading_canonical_log_prob": -170.30038452148438, "reading_distances": [32, 28, 24, 26], "reading_log_probs": [-164.25997924804688, -166.36660766601562, -166.74581909179688, -170.6531219482422], "reading_logprob_std": 2.665748459851226, "tokenization_path_logprob_std": 29.83867002328512, "trajectory_balance_rmse": 1602.8842795567468}297{"canonical_length": 168, "canonical_nll": 1.1785697937011719, "generation_rouge_l_f1": 0.22686567164179103, "generation_token_f1": 0.47164179104477616, "prompt_id": "7ec56cfda55797c0", "reading_canonical_log_prob": -197.99972534179688, "reading_distances": [27, 35, 24, 27], "reading_log_probs": [-209.979736328125, -220.7691192626953, -213.97386169433594, -207.11093139648438], "reading_logprob_std": 5.919003975669093, "tokenization_path_logprob_std": 76.8120497491383, "trajectory_balance_rmse": 3323.5961826520224}298{"canonical_length": 32, "canonical_nll": 1.4721683263778687, "generation_rouge_l_f1": 0.5789473684210527, "generation_token_f1": 0.6315789473684211, "prompt_id": "c6a2889afcb5222e", "reading_canonical_log_prob": -47.1093864440918, "reading_distances": [15, 16, 13, 9], "reading_log_probs": [-46.629676818847656, -46.56907653808594, -49.274410247802734, -48.3275032043457], "reading_logprob_std": 1.3287973732554268, "tokenization_path_logprob_std": 41.59882303225125, "trajectory_balance_rmse": 474.2134943012681}299{"canonical_length": 32, "canonical_nll": 1.512040138244629, "generation_rouge_l_f1": 0.35714285714285715, "generation_token_f1": 0.3928571428571428, "prompt_id": "5371fa87b29b339c", "reading_canonical_log_prob": -48.385284423828125, "reading_distances": [12, 17, 19, 16], "reading_log_probs": [-47.124671936035156, -44.95265197753906, -47.2088623046875, -46.731204986572266], "reading_logprob_std": 1.0552040607713509, "tokenization_path_logprob_std": 100.75479419510545, "trajectory_balance_rmse": 695.6127195504781}300{"canonical_length": 63, "canonical_nll": 1.2324752807617188, "generation_rouge_l_f1": 0.3270440251572327, "generation_token_f1": 0.4779874213836478, "prompt_id": "c89f55288eb64d0c", "reading_canonical_log_prob": -77.64594268798828, "reading_distances": [22, 29, 25, 31], "reading_log_probs": [-80.14511108398438, -78.89424896240234, -79.91160583496094, -82.55094909667969], "reading_logprob_std": 1.5486510534226834, "tokenization_path_logprob_std": 66.5556769733046, "trajectory_balance_rmse": 1546.5235235431978}301{"canonical_length": 221, "canonical_nll": 1.490878696355345, "generation_rouge_l_f1": 0.21118012422360252, "generation_token_f1": 0.3312629399585922, "prompt_id": "0e1e524655779cac", "reading_canonical_log_prob": -329.48419189453125, "reading_distances": [18, 24, 18, 17], "reading_log_probs": [-333.6396789550781, -331.9216613769531, -328.8470153808594, -328.8038330078125], "reading_logprob_std": 2.3889115551182596, "tokenization_path_logprob_std": 135.82672925355064, "trajectory_balance_rmse": 4598.898596437147}302{"canonical_length": 9, "canonical_nll": 0.4927203920152452, "generation_rouge_l_f1": 0.8235294117647058, "generation_token_f1": 0.8235294117647058, "prompt_id": "5f85bd9588d13639", "reading_canonical_log_prob": -4.434483528137207, "reading_distances": [21, 13, 20, 14], "reading_log_probs": [-5.789300918579102, -4.959333419799805, -4.898360729217529, -3.674121379852295], "reading_logprob_std": 0.8713425748165508, "tokenization_path_logprob_std": 45.036118055547135, "trajectory_balance_rmse": 205.90148114682367}303{"canonical_length": 26, "canonical_nll": 0.8778312389667218, "generation_rouge_l_f1": 0.1299638989169675, "generation_token_f1": 0.15884476534296027, "prompt_id": "570dbf2f8a32567a", "reading_canonical_log_prob": -22.823612213134766, "reading_distances": [20, 14, 21, 23], "reading_log_probs": [-23.171558380126953, -21.39267349243164, -21.726123809814453, -23.272350311279297], "reading_logprob_std": 0.9703548322400256, "tokenization_path_logprob_std": 114.41691365921052, "trajectory_balance_rmse": 498.65394849515747}304{"canonical_length": 26, "canonical_nll": 1.037784796494704, "generation_rouge_l_f1": 0.14046822742474915, "generation_token_f1": 0.15384615384615385, "prompt_id": "b4dad9a8bbd784f6", "reading_canonical_log_prob": -26.982404708862305, "reading_distances": [40, 29, 38, 37], "reading_log_probs": [-29.2044677734375, -32.65895462036133, -30.215106964111328, -28.31130027770996], "reading_logprob_std": 1.8764236522802327, "tokenization_path_logprob_std": 39.63234267247943, "trajectory_balance_rmse": 599.6202952301321}305{"canonical_length": 10, "canonical_nll": 2.2278654098510744, "generation_rouge_l_f1": 0.26666666666666666, "generation_token_f1": 0.39999999999999997, "prompt_id": "9e76d45f4e1b2541", "reading_canonical_log_prob": -22.278654098510742, "reading_distances": [28, 41, 31, 40], "reading_log_probs": [-25.577659606933594, -26.279645919799805, -24.411712646484375, -29.04827308654785], "reading_logprob_std": 1.9695502625237462, "tokenization_path_logprob_std": 44.71142568906065, "trajectory_balance_rmse": 316.5312776571182}306{"canonical_length": 256, "canonical_nll": 1.3891205787658691, "generation_rouge_l_f1": 0.2149712092130518, "generation_token_f1": 0.44145873320537427, "prompt_id": "6a9973c4375e58ef", "reading_canonical_log_prob": -355.6148681640625, "reading_distances": [32, 35, 29, 33], "reading_log_probs": [-357.7562561035156, -360.03045654296875, -358.82244873046875, -357.7474365234375], "reading_logprob_std": 1.0853550368227614, "tokenization_path_logprob_std": 338.7979226051527, "trajectory_balance_rmse": 5990.153497214603}307{"canonical_length": 277, "canonical_nll": 1.136051549808213, "generation_rouge_l_f1": 0.2169117647058824, "generation_token_f1": 0.4448529411764706, "prompt_id": "4a7ad2175af88b68", "reading_canonical_log_prob": -314.686279296875, "reading_distances": [30, 31, 28, 36], "reading_log_probs": [-320.26885986328125, -318.6861877441406, -316.8348388671875, -324.2000732421875], "reading_logprob_std": 3.1335402398519654, "tokenization_path_logprob_std": 235.05590496715126, "trajectory_balance_rmse": 5381.306092653906}308{"canonical_length": 24, "canonical_nll": 0.34895992279052734, "generation_rouge_l_f1": 0.37037037037037035, "generation_token_f1": 0.37037037037037035, "prompt_id": "7271cafab8ca2bf4", "reading_canonical_log_prob": -8.375038146972656, "reading_distances": [28, 34, 34, 26], "reading_log_probs": [-9.141197204589844, -11.706948280334473, -17.505094528198242, -13.343385696411133], "reading_logprob_std": 3.5096592055861935, "tokenization_path_logprob_std": 85.06863701988586, "trajectory_balance_rmse": 506.4814929098489}309{"canonical_length": 11, "canonical_nll": 0.5947467197071422, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "7798c57a62239065", "reading_canonical_log_prob": -6.5422139167785645, "reading_distances": [14, 14, 17, 15], "reading_log_probs": [-9.172545433044434, -8.258142471313477, -8.61686897277832, -10.000656127929688], "reading_logprob_std": 0.7588687779083131, "tokenization_path_logprob_std": 56.090315863813586, "trajectory_balance_rmse": 233.85448746885913}310{"canonical_length": 97, "canonical_nll": 1.027299468050298, "generation_rouge_l_f1": 0.2904109589041096, "generation_token_f1": 0.4054794520547945, "prompt_id": "fd06dfbeb5280c45", "reading_canonical_log_prob": -99.6480484008789, "reading_distances": [39, 44, 39, 45], "reading_log_probs": [-100.14124298095703, -106.44834899902344, -98.27680969238281, -101.19160461425781], "reading_logprob_std": 3.503122452121695, "tokenization_path_logprob_std": 236.77931078081943, "trajectory_balance_rmse": 2008.5610062755675}311{"canonical_length": 210, "canonical_nll": 0.9984141031901042, "generation_rouge_l_f1": 0.3333333333333333, "generation_token_f1": 0.5024875621890547, "prompt_id": "3efd48165ac7ab7c", "reading_canonical_log_prob": -209.66696166992188, "reading_distances": [16, 19, 17, 19], "reading_log_probs": [-209.6393280029297, -212.21499633789062, -214.31398010253906, -211.91400146484375], "reading_logprob_std": 1.9130431081066228, "tokenization_path_logprob_std": 131.7204752823774, "trajectory_balance_rmse": 2139.8839390588105}312{"canonical_length": 69, "canonical_nll": 1.479774364526721, "generation_rouge_l_f1": 0.26190476190476186, "generation_token_f1": 0.4642857142857143, "prompt_id": "18e11e5d4d3c48a7", "reading_canonical_log_prob": -102.10443115234375, "reading_distances": [32, 40, 35, 39], "reading_log_probs": [-106.18336486816406, -107.93204498291016, -102.38762664794922, -102.85296630859375], "reading_logprob_std": 2.66632114217342, "tokenization_path_logprob_std": 119.56553475699398, "trajectory_balance_rmse": 1767.6052559789257}313{"canonical_length": 17, "canonical_nll": 0.3987610760857077, "generation_rouge_l_f1": 0.08450704225352113, "generation_token_f1": 0.09154929577464789, "prompt_id": "b05a564949c24a61", "reading_canonical_log_prob": -6.778938293457031, "reading_distances": [70, 64, 51, 67], "reading_log_probs": [-3.975156307220459, -6.821446895599365, -5.433920860290527, -5.988356590270996], "reading_logprob_std": 1.1975324090476553, "tokenization_path_logprob_std": 53.45421828677862, "trajectory_balance_rmse": 142.6457826126644}314{"canonical_length": 129, "canonical_nll": 0.7668585962103319, "generation_rouge_l_f1": 0.4086021505376344, "generation_token_f1": 0.5232974910394265, "prompt_id": "977aa631a6e3fb74", "reading_canonical_log_prob": -98.92475891113281, "reading_distances": [9, 15, 14, 11], "reading_log_probs": [-101.84632873535156, -103.1265869140625, -102.79967498779297, -100.29816436767578], "reading_logprob_std": 1.2685012114547602, "tokenization_path_logprob_std": 80.12704747802925, "trajectory_balance_rmse": 1267.878236139785}315{"canonical_length": 85, "canonical_nll": 1.2229063146254595, "generation_rouge_l_f1": 0.28444444444444444, "generation_token_f1": 0.4711111111111111, "prompt_id": "308d6723913890bf", "reading_canonical_log_prob": -103.94703674316406, "reading_distances": [24, 23, 23, 21], "reading_log_probs": [-106.0806884765625, -107.71539306640625, -105.48213195800781, -105.67994689941406], "reading_logprob_std": 1.014919941081878, "tokenization_path_logprob_std": 55.93266230363385, "trajectory_balance_rmse": 1730.2873004677335}316{"canonical_length": 189, "canonical_nll": 1.159871257802166, "generation_rouge_l_f1": 0.27790432801822323, "generation_token_f1": 0.47835990888382685, "prompt_id": "ff140f7a02786106", "reading_canonical_log_prob": -219.21566772460938, "reading_distances": [15, 15, 20, 21], "reading_log_probs": [-224.50262451171875, -222.20449829101562, -224.8125, -225.10260009765625], "reading_logprob_std": 1.3235747904387904, "tokenization_path_logprob_std": 228.66783995426982, "trajectory_balance_rmse": 3507.3343408924075}317{"canonical_length": 27, "canonical_nll": 1.8854350337275751, "generation_rouge_l_f1": 0.17142857142857143, "generation_token_f1": 0.17142857142857143, "prompt_id": "536f5c5bf3146493", "reading_canonical_log_prob": -50.90674591064453, "reading_distances": [31, 42, 32, 33], "reading_log_probs": [-59.732452392578125, -60.327476501464844, -60.7141227722168, -53.60608673095703], "reading_logprob_std": 3.350383545132404, "tokenization_path_logprob_std": 53.23990376782177, "trajectory_balance_rmse": 808.1039818456921}318{"canonical_length": 210, "canonical_nll": 1.5986900692894346, "generation_rouge_l_f1": 0.22633744855967078, "generation_token_f1": 0.4897119341563786, "prompt_id": "18e930a2b5a84e85", "reading_canonical_log_prob": -335.72491455078125, "reading_distances": [14, 19, 17, 14], "reading_log_probs": [-341.4771423339844, -344.8926086425781, -344.5156555175781, -344.2242126464844], "reading_logprob_std": 1.5577268049678372, "tokenization_path_logprob_std": 166.73188510432578, "trajectory_balance_rmse": 4328.8602297194075}319{"canonical_length": 21, "canonical_nll": 1.5565685090564547, "generation_rouge_l_f1": 0.19178082191780824, "generation_token_f1": 0.273972602739726, "prompt_id": "2631f3bcf5083818", "reading_canonical_log_prob": -32.68793869018555, "reading_distances": [32, 21, 35, 28], "reading_log_probs": [-30.618881225585938, -29.04625129699707, -27.086524963378906, -30.468921661376953], "reading_logprob_std": 1.640079246493234, "tokenization_path_logprob_std": 62.45314382962191, "trajectory_balance_rmse": 297.9262082048582}320{"canonical_length": 169, "canonical_nll": 0.8697605471639238, "generation_rouge_l_f1": 0.3535353535353536, "generation_token_f1": 0.5202020202020202, "prompt_id": "62ad42126bc04505", "reading_canonical_log_prob": -146.98953247070312, "reading_distances": [43, 45, 43, 38], "reading_log_probs": [-148.33535766601562, -148.28378295898438, -151.71499633789062, -147.7393035888672], "reading_logprob_std": 1.8178676573674086, "tokenization_path_logprob_std": 176.6248944373444, "trajectory_balance_rmse": 2859.9266363886945}321{"canonical_length": 221, "canonical_nll": 1.3023277041059813, "generation_rouge_l_f1": 0.2809224318658281, "generation_token_f1": 0.5157232704402515, "prompt_id": "44a0d1ef8b9a47d0", "reading_canonical_log_prob": -287.8144226074219, "reading_distances": [15, 7, 13, 19], "reading_log_probs": [-297.5082702636719, -289.8904113769531, -291.29730224609375, -294.6443786621094], "reading_logprob_std": 3.4228672220753458, "tokenization_path_logprob_std": 191.5373372913855, "trajectory_balance_rmse": 5322.683650773419}322{"canonical_length": 84, "canonical_nll": 1.1725309462774367, "generation_rouge_l_f1": 0.024390243902439022, "generation_token_f1": 0.024390243902439022, "prompt_id": "1bc849b3b651f6ed", "reading_canonical_log_prob": -98.49259948730469, "reading_distances": [23, 20, 21, 26], "reading_log_probs": [-108.37522888183594, -105.63133239746094, -104.4886703491211, -107.10520935058594], "reading_logprob_std": 1.6973391168432919, "tokenization_path_logprob_std": 184.4537623910664, "trajectory_balance_rmse": 1959.5234922342058}323{"canonical_length": 225, "canonical_nll": 1.2557892523871528, "generation_rouge_l_f1": 0.1963927855711423, "generation_token_f1": 0.42885771543086165, "prompt_id": "a8febb0397bb60c3", "reading_canonical_log_prob": -282.5525817871094, "reading_distances": [16, 15, 18, 12], "reading_log_probs": [-285.2103271484375, -287.3564758300781, -286.19342041015625, -285.41357421875], "reading_logprob_std": 0.9725165191354989, "tokenization_path_logprob_std": 62.67492194921525, "trajectory_balance_rmse": 4421.871691365713}324{"canonical_length": 7, "canonical_nll": 2.85873658316476, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "37af528e072ecf7b", "reading_canonical_log_prob": -20.01115608215332, "reading_distances": [9, 15, 22, 20], "reading_log_probs": [-19.961559295654297, -20.693849563598633, -19.579174041748047, -20.820335388183594], "reading_logprob_std": 0.5929408041526528, "tokenization_path_logprob_std": 58.80688673932306, "trajectory_balance_rmse": 144.52790100232815}325{"canonical_length": 252, "canonical_nll": 1.5981951516772073, "generation_rouge_l_f1": 0.21285140562248997, "generation_token_f1": 0.42570281124497994, "prompt_id": "192d126d4e011ab4", "reading_canonical_log_prob": -402.74517822265625, "reading_distances": [21, 25, 16, 20], "reading_log_probs": [-412.55352783203125, -411.09881591796875, -405.50408935546875, -411.302001953125], "reading_logprob_std": 3.140264412665143, "tokenization_path_logprob_std": 231.25977416818628, "trajectory_balance_rmse": 6140.611869166205}326{"canonical_length": 182, "canonical_nll": 1.3383456219683636, "generation_rouge_l_f1": 0.29032258064516125, "generation_token_f1": 0.52073732718894, "prompt_id": "55712d2518889f49", "reading_canonical_log_prob": -243.5789031982422, "reading_distances": [23, 25, 23, 19], "reading_log_probs": [-282.9347839355469, -278.9729919433594, -279.1080017089844, -277.49365234375], "reading_logprob_std": 2.323050583445627, "tokenization_path_logprob_std": 106.82253310062463, "trajectory_balance_rmse": 3195.4862835698664}327{"canonical_length": 12, "canonical_nll": 1.0116907755533855, "generation_rouge_l_f1": 0.3225806451612903, "generation_token_f1": 0.3225806451612903, "prompt_id": "587eb61a8e257f2b", "reading_canonical_log_prob": -12.140289306640625, "reading_distances": [21, 25, 23, 29], "reading_log_probs": [-13.135499954223633, -14.05309009552002, -10.65376091003418, -12.117817878723145], "reading_logprob_std": 1.4571907170311795, "tokenization_path_logprob_std": 53.73126413432721, "trajectory_balance_rmse": 368.05754402698494}328{"canonical_length": 109, "canonical_nll": 0.8415711429140983, "generation_rouge_l_f1": 0.13293051359516614, "generation_token_f1": 0.21752265861027192, "prompt_id": "fff5e81d18ab40f8", "reading_canonical_log_prob": -91.73125457763672, "reading_distances": [15, 15, 14, 16], "reading_log_probs": [-93.00189208984375, -95.04917907714844, -92.47992706298828, -95.09943389892578], "reading_logprob_std": 1.364090376447361, "tokenization_path_logprob_std": 143.9065969626025, "trajectory_balance_rmse": 1260.6799061780673}329{"canonical_length": 15, "canonical_nll": 0.40414098103841145, "generation_rouge_l_f1": 0.8571428571428571, "generation_token_f1": 0.9285714285714286, "prompt_id": "e02d75efb09e53b0", "reading_canonical_log_prob": -6.062114715576172, "reading_distances": [25, 28, 20, 18], "reading_log_probs": [-7.365386009216309, -10.72225570678711, -7.704286098480225, -7.577710151672363], "reading_logprob_std": 1.5927135585118064, "tokenization_path_logprob_std": 39.840976880442696, "trajectory_balance_rmse": 240.14351812735694}330{"canonical_length": 27, "canonical_nll": 1.2876653318051938, "generation_rouge_l_f1": 0.10600706713780919, "generation_token_f1": 0.127208480565371, "prompt_id": "93316a155b82ba8e", "reading_canonical_log_prob": -34.766963958740234, "reading_distances": [27, 27, 28, 24], "reading_log_probs": [-42.23338317871094, -38.206016540527344, -37.449703216552734, -41.30898666381836], "reading_logprob_std": 2.3283092415572417, "tokenization_path_logprob_std": 128.68795483154892, "trajectory_balance_rmse": 571.129078356283}331{"canonical_length": 110, "canonical_nll": 1.3480331420898437, "generation_rouge_l_f1": 0.38211382113821135, "generation_token_f1": 0.5203252032520325, "prompt_id": "37f1255c888facef", "reading_canonical_log_prob": -148.2836456298828, "reading_distances": [14, 19, 18, 12], "reading_log_probs": [-147.327880859375, -147.83050537109375, -147.57620239257812, -147.35369873046875], "reading_logprob_std": 0.23389527627743564, "tokenization_path_logprob_std": 181.75078626695844, "trajectory_balance_rmse": 2478.758821589076}332{"canonical_length": 150, "canonical_nll": 0.9391184488932292, "generation_rouge_l_f1": 0.4021739130434783, "generation_token_f1": 0.5706521739130435, "prompt_id": "cf1ad5289e6ad557", "reading_canonical_log_prob": -140.86776733398438, "reading_distances": [19, 21, 21, 19], "reading_log_probs": [-140.75424194335938, -141.13211059570312, -142.54254150390625, -141.29571533203125], "reading_logprob_std": 0.7748468035011924, "tokenization_path_logprob_std": 109.85599007354116, "trajectory_balance_rmse": 2853.539151977751}333{"canonical_length": 314, "canonical_nll": 0.9491549935310509, "generation_rouge_l_f1": 0.32051282051282054, "generation_token_f1": 0.4903846153846154, "prompt_id": "ff9bcfc5d56ada93", "reading_canonical_log_prob": -298.03466796875, "reading_distances": [41, 34, 36, 40], "reading_log_probs": [-302.83367919921875, -304.4349365234375, -306.34515380859375, -305.73675537109375], "reading_logprob_std": 1.5555315612820029, "tokenization_path_logprob_std": 156.597102103461, "trajectory_balance_rmse": 5225.022740785851}334{"canonical_length": 12, "canonical_nll": 1.6536814371744792, "generation_rouge_l_f1": 0.10526315789473685, "generation_token_f1": 0.10526315789473685, "prompt_id": "6f26ffa7e9fb10c0", "reading_canonical_log_prob": -19.84417724609375, "reading_distances": [22, 21, 20, 22], "reading_log_probs": [-18.985368728637695, -19.658889770507812, -19.96495246887207, -20.226289749145508], "reading_logprob_std": 0.5351802003578103, "tokenization_path_logprob_std": 45.35923963404357, "trajectory_balance_rmse": 253.29845048196927}335{"canonical_length": 41, "canonical_nll": 0.9469495633753334, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "07e67bead15cc6c8", "reading_canonical_log_prob": -38.82493209838867, "reading_distances": [31, 38, 35, 24], "reading_log_probs": [-37.57271957397461, -36.77773666381836, -39.497474670410156, -39.942623138427734], "reading_logprob_std": 1.5156066248026017, "tokenization_path_logprob_std": 93.97283483570095, "trajectory_balance_rmse": 767.4534105527542}336{"canonical_length": 413, "canonical_nll": 1.2226831468485169, "generation_rouge_l_f1": 0.23283582089552238, "generation_token_f1": 0.4447761194029851, "prompt_id": "52cb3482c2bf1893", "reading_canonical_log_prob": -504.9681396484375, "reading_distances": [26, 24, 22, 33], "reading_log_probs": [-514.2741088867188, -507.5346374511719, -510.4327392578125, -511.82867431640625], "reading_logprob_std": 2.8128148336907954, "tokenization_path_logprob_std": 153.30790799652445, "trajectory_balance_rmse": 7586.5326105539625}337{"canonical_length": 265, "canonical_nll": 1.695814370209316, "generation_rouge_l_f1": 0.2248062015503876, "generation_token_f1": 0.4728682170542636, "prompt_id": "3b670d936478ce43", "reading_canonical_log_prob": -449.39080810546875, "reading_distances": [31, 26, 30, 29], "reading_log_probs": [-456.80462646484375, -455.5249328613281, -455.0368957519531, -454.286865234375], "reading_logprob_std": 1.0581094245619589, "tokenization_path_logprob_std": 379.3074772515009, "trajectory_balance_rmse": 6433.778983419708}338{"canonical_length": 78, "canonical_nll": 2.305433615660056, "generation_rouge_l_f1": 0.1533923303834808, "generation_token_f1": 0.2831858407079646, "prompt_id": "d940482038c5842a", "reading_canonical_log_prob": -179.82382202148438, "reading_distances": [20, 30, 28, 26], "reading_log_probs": [-177.736572265625, -179.31625366210938, -175.06204223632812, -178.68724060058594], "reading_logprob_std": 1.8750179986278201, "tokenization_path_logprob_std": 195.9400349641723, "trajectory_balance_rmse": 2001.9478098065713}339{"canonical_length": 291, "canonical_nll": 0.9788302392074743, "generation_rouge_l_f1": 0.29338103756708406, "generation_token_f1": 0.5152057245080501, "prompt_id": "07d10dd03fee840c", "reading_canonical_log_prob": -284.839599609375, "reading_distances": [26, 27, 19, 24], "reading_log_probs": [-294.2298889160156, -297.55950927734375, -294.85943603515625, -294.16351318359375], "reading_logprob_std": 1.6019415475424579, "tokenization_path_logprob_std": 347.9867695456292, "trajectory_balance_rmse": 5725.291022954588}340{"canonical_length": 4, "canonical_nll": 1.6795105934143066, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "7ef3a945dcbdd086", "reading_canonical_log_prob": -6.718042373657227, "reading_distances": [21, 23, 17, 24], "reading_log_probs": [-7.368317127227783, -9.610788345336914, -7.87083101272583, -12.711002349853516], "reading_logprob_std": 2.413357269501864, "tokenization_path_logprob_std": 17.85651646662951, "trajectory_balance_rmse": 45.64551465728045}341{"canonical_length": 156, "canonical_nll": 1.0265819842998798, "generation_rouge_l_f1": 0.23409669211195927, "generation_token_f1": 0.3969465648854961, "prompt_id": "ef63cfbdbb1633a2", "reading_canonical_log_prob": -160.14678955078125, "reading_distances": [25, 27, 34, 30], "reading_log_probs": [-167.60198974609375, -161.93499755859375, -170.39810180664062, -169.855712890625], "reading_logprob_std": 3.8694138417112365, "tokenization_path_logprob_std": 154.55373477891217, "trajectory_balance_rmse": 2959.695970310861}342{"canonical_length": 152, "canonical_nll": 1.5223972922877262, "generation_rouge_l_f1": 0.22950819672131145, "generation_token_f1": 0.43559718969555034, "prompt_id": "1d1550150b11ce18", "reading_canonical_log_prob": -231.40438842773438, "reading_distances": [24, 30, 29, 27], "reading_log_probs": [-238.59449768066406, -241.22018432617188, -237.68771362304688, -239.25900268554688], "reading_logprob_std": 1.4986545800304294, "tokenization_path_logprob_std": 244.52214981300975, "trajectory_balance_rmse": 3787.43784703109}343{"canonical_length": 436, "canonical_nll": 0.8003895610844324, "generation_rouge_l_f1": 0.3304347826086957, "generation_token_f1": 0.47652173913043483, "prompt_id": "becce7dcacc67b08", "reading_canonical_log_prob": -348.9698486328125, "reading_distances": [22, 25, 20, 29], "reading_log_probs": [-371.56829833984375, -370.08905029296875, -359.79693603515625, -366.30059814453125], "reading_logprob_std": 5.25261377038363, "tokenization_path_logprob_std": 112.52856819737691, "trajectory_balance_rmse": 5534.69942319988}344{"canonical_length": 289, "canonical_nll": 1.067862857171821, "generation_rouge_l_f1": 0.2003454231433506, "generation_token_f1": 0.3799654576856649, "prompt_id": "11042f17f722f10a", "reading_canonical_log_prob": -308.61236572265625, "reading_distances": [23, 29, 28, 26], "reading_log_probs": [-321.3117370605469, -322.9449157714844, -318.2641906738281, -323.1906433105469], "reading_logprob_std": 2.26797643694526, "tokenization_path_logprob_std": 150.96181681394887, "trajectory_balance_rmse": 6346.871439179859}345{"canonical_length": 31, "canonical_nll": 1.1003267841954385, "generation_rouge_l_f1": 0.11643835616438354, "generation_token_f1": 0.15753424657534248, "prompt_id": "b29cd52aed559d3c", "reading_canonical_log_prob": -34.110130310058594, "reading_distances": [21, 19, 19, 18], "reading_log_probs": [-41.08506393432617, -41.30901336669922, -43.20124816894531, -41.49608612060547], "reading_logprob_std": 0.9669735722822924, "tokenization_path_logprob_std": 54.8831448195251, "trajectory_balance_rmse": 542.3235353722555}346{"canonical_length": 192, "canonical_nll": 0.8761210441589355, "generation_rouge_l_f1": 0.2577777777777778, "generation_token_f1": 0.4355555555555556, "prompt_id": "3d6b176b83ba1bb0", "reading_canonical_log_prob": -168.21524047851562, "reading_distances": [20, 22, 20, 16], "reading_log_probs": [-180.96591186523438, -182.2447509765625, -178.89913940429688, -176.8134765625], "reading_logprob_std": 2.3838264908952356, "tokenization_path_logprob_std": 42.32870699033084, "trajectory_balance_rmse": 2113.8690843464774}347{"canonical_length": 146, "canonical_nll": 1.9193328439372859, "generation_rouge_l_f1": 0.18735362997658078, "generation_token_f1": 0.3372365339578454, "prompt_id": "ed74cc90058602c2", "reading_canonical_log_prob": -280.22259521484375, "reading_distances": [28, 29, 22, 30], "reading_log_probs": [-277.1527404785156, -280.6053466796875, -276.8936462402344, -277.4167175292969], "reading_logprob_std": 1.7386532832535229, "tokenization_path_logprob_std": 254.71560175622304, "trajectory_balance_rmse": 3550.5583920215595}348{"canonical_length": 217, "canonical_nll": 1.686430900327621, "generation_rouge_l_f1": 0.22997946611909653, "generation_token_f1": 0.4681724845995893, "prompt_id": "b2e99fb7bff1b5b7", "reading_canonical_log_prob": -365.95550537109375, "reading_distances": [26, 37, 39, 36], "reading_log_probs": [-378.62725830078125, -381.75042724609375, -383.38848876953125, -383.47039794921875], "reading_logprob_std": 2.2643570717328134, "tokenization_path_logprob_std": 256.67732288361077, "trajectory_balance_rmse": 5135.1027417398}349{"canonical_length": 318, "canonical_nll": 0.7288065976316824, "generation_rouge_l_f1": 0.21428571428571425, "generation_token_f1": 0.21978021978021975, "prompt_id": "5a32b0efbf18f769", "reading_canonical_log_prob": -231.760498046875, "reading_distances": [26, 29, 29, 25], "reading_log_probs": [-231.97653198242188, -231.9650421142578, -231.00270080566406, -231.28375244140625], "reading_logprob_std": 0.4913984385454802, "tokenization_path_logprob_std": 175.87992066441896, "trajectory_balance_rmse": 5280.332489791671}350{"canonical_length": 25, "canonical_nll": 0.564500961303711, "generation_rouge_l_f1": 0.09999999999999999, "generation_token_f1": 0.1090909090909091, "prompt_id": "3697c04353d3a1f1", "reading_canonical_log_prob": -14.112524032592773, "reading_distances": [29, 29, 32, 26], "reading_log_probs": [-16.507225036621094, -15.49989128112793, -21.0084228515625, -18.669160842895508], "reading_logprob_std": 2.446252335332801, "tokenization_path_logprob_std": 61.11180201093262, "trajectory_balance_rmse": 301.7575947444175}351{"canonical_length": 83, "canonical_nll": 1.7381922710372741, "generation_rouge_l_f1": 0.4203821656050955, "generation_token_f1": 0.5987261146496815, "prompt_id": "4983a8fb02a849d4", "reading_canonical_log_prob": -144.26995849609375, "reading_distances": [18, 19, 13, 10], "reading_log_probs": [-140.94664001464844, -139.79623413085938, -140.2790985107422, -143.26173400878906], "reading_logprob_std": 1.5348090947003672, "tokenization_path_logprob_std": 105.50437928987323, "trajectory_balance_rmse": 1423.9843655714476}352{"canonical_length": 244, "canonical_nll": 1.8038266291383838, "generation_rouge_l_f1": 0.2304147465437788, "generation_token_f1": 0.4792626728110599, "prompt_id": "fe89f27185bbc3b2", "reading_canonical_log_prob": -440.1336975097656, "reading_distances": [43, 39, 29, 41], "reading_log_probs": [-448.313232421875, -449.80279541015625, -450.066162109375, -447.923583984375], "reading_logprob_std": 1.0659434363118234, "tokenization_path_logprob_std": 157.08922670477097, "trajectory_balance_rmse": 6479.227465452336}353{"canonical_length": 87, "canonical_nll": 1.5456069420123923, "generation_rouge_l_f1": 0.1875, "generation_token_f1": 0.35714285714285715, "prompt_id": "6b767f6a353fd065", "reading_canonical_log_prob": -134.46780395507812, "reading_distances": [20, 17, 23, 18], "reading_log_probs": [-144.44004821777344, -144.54684448242188, -140.52879333496094, -151.2041015625], "reading_logprob_std": 4.429896071557433, "tokenization_path_logprob_std": 139.32588564213953, "trajectory_balance_rmse": 1679.3564009397996}354{"canonical_length": 136, "canonical_nll": 1.359443888944738, "generation_rouge_l_f1": 0.3488372093023256, "generation_token_f1": 0.5058139534883721, "prompt_id": "e77b6aefa4b6596d", "reading_canonical_log_prob": -184.88436889648438, "reading_distances": [31, 25, 29, 28], "reading_log_probs": [-182.26541137695312, -186.98129272460938, -179.50758361816406, -186.48968505859375], "reading_logprob_std": 3.5653118774194743, "tokenization_path_logprob_std": 145.19045800654737, "trajectory_balance_rmse": 2556.693052755231}355{"canonical_length": 9, "canonical_nll": 1.474744372897678, "generation_rouge_l_f1": 0.5714285714285715, "generation_token_f1": 0.8571428571428571, "prompt_id": "383239282305a169", "reading_canonical_log_prob": -13.272699356079102, "reading_distances": [31, 34, 35, 32], "reading_log_probs": [-14.230514526367188, -14.805618286132812, -14.5974702835083, -13.69361686706543], "reading_logprob_std": 0.4873810713205811, "tokenization_path_logprob_std": 29.611701429503743, "trajectory_balance_rmse": 120.36285870797941}356{"canonical_length": 10, "canonical_nll": 1.6265445709228517, "generation_rouge_l_f1": 0.5714285714285714, "generation_token_f1": 0.5714285714285714, "prompt_id": "29653e40ee0ea781", "reading_canonical_log_prob": -16.265445709228516, "reading_distances": [30, 24, 20, 25], "reading_log_probs": [-14.081696510314941, -16.366357803344727, -14.364483833312988, -14.88750171661377], "reading_logprob_std": 1.0172278648291924, "tokenization_path_logprob_std": 30.5501388481815, "trajectory_balance_rmse": 180.33602602106197}357{"canonical_length": 251, "canonical_nll": 1.6846349951755477, "generation_rouge_l_f1": 0.2012072434607646, "generation_token_f1": 0.4225352112676056, "prompt_id": "b8e0545bd0cc139e", "reading_canonical_log_prob": -422.8433837890625, "reading_distances": [48, 52, 34, 45], "reading_log_probs": [-427.4844665527344, -430.415283203125, -432.2012023925781, -438.7467346191406], "reading_logprob_std": 4.770766425134796, "tokenization_path_logprob_std": 88.12269996579236, "trajectory_balance_rmse": 6034.354207167745}358{"canonical_length": 430, "canonical_nll": 1.2203850324763807, "generation_rouge_l_f1": 0.19844961240310077, "generation_token_f1": 0.3503875968992247, "prompt_id": "a7d20d598bd1e5dd", "reading_canonical_log_prob": -524.7655639648438, "reading_distances": [36, 35, 38, 29], "reading_log_probs": [-607.803955078125, -607.8934936523438, -611.75390625, -609.142822265625], "reading_logprob_std": 1.841287225100055, "tokenization_path_logprob_std": 369.7758678257468, "trajectory_balance_rmse": 8244.97587093843}359{"canonical_length": 47, "canonical_nll": 2.942827265313331, "generation_rouge_l_f1": 0.20437956204379562, "generation_token_f1": 0.3211678832116789, "prompt_id": "ecbe9c8ff4f546fd", "reading_canonical_log_prob": -138.31288146972656, "reading_distances": [12, 14, 10, 13], "reading_log_probs": [-138.3053436279297, -163.94699096679688, -156.11862182617188, -155.57345581054688], "reading_logprob_std": 10.819316510951197, "tokenization_path_logprob_std": 112.93102974796486, "trajectory_balance_rmse": 857.5447691600737}360{"canonical_length": 12, "canonical_nll": 1.2401835123697917, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "641a36ce0ce7d08c", "reading_canonical_log_prob": -14.8822021484375, "reading_distances": [25, 23, 26, 29], "reading_log_probs": [-15.46908187866211, -17.720569610595703, -18.83487892150879, -16.169309616088867], "reading_logprob_std": 1.5177136905525177, "tokenization_path_logprob_std": 43.63386761364857, "trajectory_balance_rmse": 293.2221669000599}361{"canonical_length": 105, "canonical_nll": 1.6729096912202381, "generation_rouge_l_f1": 0.19607843137254904, "generation_token_f1": 0.3081232492997199, "prompt_id": "32d37daf5bb72f67", "reading_canonical_log_prob": -175.655517578125, "reading_distances": [19, 20, 24, 20], "reading_log_probs": [-177.51751708984375, -177.96310424804688, -187.89816284179688, -180.52239990234375], "reading_logprob_std": 4.801414178016676, "tokenization_path_logprob_std": 147.51094082346208, "trajectory_balance_rmse": 2565.744270294446}362{"canonical_length": 19, "canonical_nll": 2.9871842233758223, "generation_rouge_l_f1": 0.26666666666666666, "generation_token_f1": 0.33333333333333337, "prompt_id": "b21a0da1ea5ee3b7", "reading_canonical_log_prob": -56.756500244140625, "reading_distances": [20, 26, 23, 24], "reading_log_probs": [-55.38643264770508, -58.42890167236328, -55.69841766357422, -58.45707702636719], "reading_logprob_std": 1.6795174795431589, "tokenization_path_logprob_std": 66.4783604368206, "trajectory_balance_rmse": 418.4309211081814}363{"canonical_length": 281, "canonical_nll": 1.4870874805382561, "generation_rouge_l_f1": 0.21923076923076923, "generation_token_f1": 0.47692307692307695, "prompt_id": "11a80e6e30ed7f8f", "reading_canonical_log_prob": -417.87158203125, "reading_distances": [19, 17, 19, 21], "reading_log_probs": [-419.7655334472656, -425.878662109375, -421.292236328125, -421.06744384765625], "reading_logprob_std": 2.67129476541275, "tokenization_path_logprob_std": 114.44540217872606, "trajectory_balance_rmse": 6361.991396816223}364{"canonical_length": 119, "canonical_nll": 1.9088421989889706, "generation_rouge_l_f1": 0.21390374331550804, "generation_token_f1": 0.4117647058823529, "prompt_id": "e2dff42d7f8d6988", "reading_canonical_log_prob": -227.1522216796875, "reading_distances": [23, 21, 26, 21], "reading_log_probs": [-226.82186889648438, -224.8249969482422, -225.82513427734375, -225.75823974609375], "reading_logprob_std": 0.8158827593138007, "tokenization_path_logprob_std": 172.27005586219678, "trajectory_balance_rmse": 2499.7227090979295}365{"canonical_length": 338, "canonical_nll": 1.736863175792807, "generation_rouge_l_f1": 0.19931271477663232, "generation_token_f1": 0.43986254295532645, "prompt_id": "bd63a203fd3cb2b2", "reading_canonical_log_prob": -587.0597534179688, "reading_distances": [29, 29, 31, 30], "reading_log_probs": [-589.7119140625, -595.198974609375, -592.9389038085938, -596.9820556640625], "reading_logprob_std": 3.135955864003643, "tokenization_path_logprob_std": 127.79690746340076, "trajectory_balance_rmse": 5963.615900223177}366{"canonical_length": 17, "canonical_nll": 0.28997886882108803, "generation_rouge_l_f1": 0.14444444444444446, "generation_token_f1": 0.14444444444444446, "prompt_id": "f5560d8544cd62d8", "reading_canonical_log_prob": -4.929640769958496, "reading_distances": [25, 21, 22, 18], "reading_log_probs": [-6.9208502769470215, -5.945829391479492, -5.1912641525268555, -6.675077438354492], "reading_logprob_std": 0.7802462996804141, "tokenization_path_logprob_std": 59.19343783672161, "trajectory_balance_rmse": 410.9457339969162}367{"canonical_length": 143, "canonical_nll": 1.5466895470252404, "generation_rouge_l_f1": 0.2702702702702703, "generation_token_f1": 0.4668304668304668, "prompt_id": "d5ad06fe63008db3", "reading_canonical_log_prob": -221.17660522460938, "reading_distances": [21, 24, 21, 26], "reading_log_probs": [-223.61993408203125, -231.34793090820312, -227.6561737060547, -228.0998077392578], "reading_logprob_std": 3.168314743259185, "tokenization_path_logprob_std": 266.8935505849011, "trajectory_balance_rmse": 3466.077629030523}368{"canonical_length": 68, "canonical_nll": 1.3337964450611788, "generation_rouge_l_f1": 0.21021021021021022, "generation_token_f1": 0.31231231231231227, "prompt_id": "32cf7dbccdbc97bc", "reading_canonical_log_prob": -90.69815826416016, "reading_distances": [16, 20, 16, 23], "reading_log_probs": [-94.20989990234375, -95.7437515258789, -91.53021240234375, -100.02537536621094], "reading_logprob_std": 3.5544243463820404, "tokenization_path_logprob_std": 33.13016291521256, "trajectory_balance_rmse": 1845.936334831786}369{"canonical_length": 47, "canonical_nll": 1.8509002036236701, "generation_rouge_l_f1": 0.1527272727272727, "generation_token_f1": 0.17454545454545456, "prompt_id": "c180433ec97360b0", "reading_canonical_log_prob": -86.9923095703125, "reading_distances": [32, 24, 28, 36], "reading_log_probs": [-90.35628509521484, -87.33595275878906, -93.52543640136719, -102.04197692871094], "reading_logprob_std": 6.343169806000918, "tokenization_path_logprob_std": 12.7049721119778, "trajectory_balance_rmse": 856.9329826579326}370{"canonical_length": 122, "canonical_nll": 0.9525388498775295, "generation_rouge_l_f1": 0.311614730878187, "generation_token_f1": 0.48725212464589235, "prompt_id": "aaf9ee6d37011f66", "reading_canonical_log_prob": -116.2097396850586, "reading_distances": [35, 34, 46, 32], "reading_log_probs": [-142.8516845703125, -135.0140380859375, -122.35284423828125, -134.9126434326172], "reading_logprob_std": 8.479014279796214, "tokenization_path_logprob_std": 105.57423459001862, "trajectory_balance_rmse": 2115.028636355806}371{"canonical_length": 19, "canonical_nll": 0.2161637356406764, "generation_rouge_l_f1": 0.08695652173913043, "generation_token_f1": 0.08695652173913043, "prompt_id": "f459f20b3852c4a3", "reading_canonical_log_prob": -4.107110977172852, "reading_distances": [15, 21, 20, 22], "reading_log_probs": [-0.7202613949775696, -0.8531240820884705, -0.9828231334686279, -0.7317484021186829], "reading_logprob_std": 0.12292122539060134, "tokenization_path_logprob_std": 71.15029453306873, "trajectory_balance_rmse": 242.64030378324256}372{"canonical_length": 19, "canonical_nll": 1.4983951167056435, "generation_rouge_l_f1": 0.34285714285714286, "generation_token_f1": 0.34285714285714286, "prompt_id": "a05d7928d7de0d6c", "reading_canonical_log_prob": -28.469507217407227, "reading_distances": [26, 28, 28, 20], "reading_log_probs": [-32.34461212158203, -32.583900451660156, -34.8013916015625, -28.088897705078125], "reading_logprob_std": 2.804520465769194, "tokenization_path_logprob_std": 20.032814200580507, "trajectory_balance_rmse": 362.4853184665956}373{"canonical_length": 15, "canonical_nll": 1.5440205891927083, "generation_rouge_l_f1": 0.14285714285714288, "generation_token_f1": 0.14285714285714288, "prompt_id": "2c22dccbca7a08c7", "reading_canonical_log_prob": -23.160308837890625, "reading_distances": [26, 28, 26, 28], "reading_log_probs": [-24.23041534423828, -19.511266708374023, -22.36921501159668, -21.26633644104004], "reading_logprob_std": 1.9787343109884719, "tokenization_path_logprob_std": 45.3142451792725, "trajectory_balance_rmse": 359.6157687703864}374{"canonical_length": 166, "canonical_nll": 1.892740272613893, "generation_rouge_l_f1": 0.3359375, "generation_token_f1": 0.5078125, "prompt_id": "6e818f0ad7674362", "reading_canonical_log_prob": -314.19488525390625, "reading_distances": [25, 24, 28, 23], "reading_log_probs": [-313.7330322265625, -311.269775390625, -311.4764404296875, -309.2030029296875], "reading_logprob_std": 1.8521110279473163, "tokenization_path_logprob_std": 246.2579777808618, "trajectory_balance_rmse": 3181.400477588779}375{"canonical_length": 70, "canonical_nll": 1.129691423688616, "generation_rouge_l_f1": 0.3366336633663366, "generation_token_f1": 0.4752475247524752, "prompt_id": "8ebd94ecdb15c9f4", "reading_canonical_log_prob": -79.07839965820312, "reading_distances": [22, 16, 20, 23], "reading_log_probs": [-80.0999526977539, -88.26864624023438, -83.24171447753906, -84.8144760131836], "reading_logprob_std": 3.3973011974685265, "tokenization_path_logprob_std": 113.68633303714871, "trajectory_balance_rmse": 1789.9552427154922}376{"canonical_length": 3, "canonical_nll": 0.5935000578562418, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "2148a4bb4107090b", "reading_canonical_log_prob": -1.7805001735687256, "reading_distances": [15, 22, 21, 16], "reading_log_probs": [-1.5220491886138916, -2.4057726860046387, -1.5290918350219727, -1.607964277267456], "reading_logprob_std": 0.4281439258439198, "tokenization_path_logprob_std": 20.94996750604537, "trajectory_balance_rmse": 49.564128738314885}377{"canonical_length": 368, "canonical_nll": 0.952703393023947, "generation_rouge_l_f1": 0.20408163265306123, "generation_token_f1": 0.43956043956043955, "prompt_id": "5d2176dd2ade1ce9", "reading_canonical_log_prob": -350.5948486328125, "reading_distances": [25, 27, 20, 22], "reading_log_probs": [-348.9720153808594, -347.06878662109375, -353.24237060546875, -347.0079345703125], "reading_logprob_std": 2.9254751999657627, "tokenization_path_logprob_std": 176.68672976389485, "trajectory_balance_rmse": 7526.662715840852}378{"canonical_length": 12, "canonical_nll": 0.6787544091542562, "generation_rouge_l_f1": 0.06896551724137931, "generation_token_f1": 0.08429118773946359, "prompt_id": "9d6ee475e283101d", "reading_canonical_log_prob": -8.145052909851074, "reading_distances": [30, 28, 33, 33], "reading_log_probs": [-7.03036642074585, -7.017210960388184, -6.705418109893799, -7.614633560180664], "reading_logprob_std": 0.37946605474982115, "tokenization_path_logprob_std": 15.477248419472591, "trajectory_balance_rmse": 264.1725672524799}379{"canonical_length": 12, "canonical_nll": 0.5585266749064127, "generation_rouge_l_f1": 0.2, "generation_token_f1": 0.2, "prompt_id": "a41c0de97fa957ce", "reading_canonical_log_prob": -6.702320098876953, "reading_distances": [20, 17, 24, 23], "reading_log_probs": [-7.950334072113037, -7.437896728515625, -8.030607223510742, -7.496982574462891], "reading_logprob_std": 0.3047014282471484, "tokenization_path_logprob_std": 18.291129587356867, "trajectory_balance_rmse": 199.99519568229294}380{"canonical_length": 9, "canonical_nll": 0.16862399048275417, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "916ba9265d9ad9c6", "reading_canonical_log_prob": -1.5176159143447876, "reading_distances": [28, 36, 39, 32], "reading_log_probs": [-1.380875825881958, -0.9586178660392761, -1.0371030569076538, -0.768051266670227], "reading_logprob_std": 0.25607613801408596, "tokenization_path_logprob_std": 28.473144834889595, "trajectory_balance_rmse": 166.30111276801767}381{"canonical_length": 99, "canonical_nll": 0.8416353476167929, "generation_rouge_l_f1": 0.5504587155963303, "generation_token_f1": 0.6880733944954128, "prompt_id": "5baddab4b656da75", "reading_canonical_log_prob": -83.3218994140625, "reading_distances": [20, 25, 26, 30], "reading_log_probs": [-83.85621643066406, -83.71109008789062, -84.72248840332031, -85.05862426757812], "reading_logprob_std": 0.6563177348770257, "tokenization_path_logprob_std": 159.32912011192897, "trajectory_balance_rmse": 2072.109039194174}382{"canonical_length": 54, "canonical_nll": 0.6192998532895688, "generation_rouge_l_f1": 0.39688715953307385, "generation_token_f1": 0.4046692607003891, "prompt_id": "22a09297862230a9", "reading_canonical_log_prob": -33.44219207763672, "reading_distances": [31, 27, 27, 34], "reading_log_probs": [-35.593650817871094, -37.37334060668945, -36.27288818359375, -35.58229446411133], "reading_logprob_std": 0.842839922322079, "tokenization_path_logprob_std": 73.65669120079616, "trajectory_balance_rmse": 870.8005951039622}383{"canonical_length": 40, "canonical_nll": 0.2922965526580811, "generation_rouge_l_f1": 0.1865671641791045, "generation_token_f1": 0.20149253731343286, "prompt_id": "c6508f3625213115", "reading_canonical_log_prob": -11.691862106323242, "reading_distances": [10, 15, 15, 12], "reading_log_probs": [-9.455276489257812, -8.658224105834961, -9.350266456604004, -8.909402847290039], "reading_logprob_std": 0.37424053531868595, "tokenization_path_logprob_std": 27.671022932500488, "trajectory_balance_rmse": 212.1250751125649}384{"canonical_length": 272, "canonical_nll": 1.533188314998851, "generation_rouge_l_f1": 0.27547169811320754, "generation_token_f1": 0.5056603773584906, "prompt_id": "a4b74ba3880275a6", "reading_canonical_log_prob": -417.0272216796875, "reading_distances": [20, 15, 19, 14], "reading_log_probs": [-427.6695556640625, -426.5563049316406, -425.325927734375, -419.8209533691406], "reading_logprob_std": 3.482288621603759, "tokenization_path_logprob_std": 162.54563295872035, "trajectory_balance_rmse": 6187.77659584335}385{"canonical_length": 17, "canonical_nll": 0.4011688793406767, "generation_rouge_l_f1": 0.2222222222222222, "generation_token_f1": 0.2222222222222222, "prompt_id": "5f38035dc4f5627c", "reading_canonical_log_prob": -6.819870948791504, "reading_distances": [20, 17, 20, 22], "reading_log_probs": [-7.839777946472168, -6.652517318725586, -7.83181619644165, -8.376341819763184], "reading_logprob_std": 0.7278035956215231, "tokenization_path_logprob_std": 31.333860814778554, "trajectory_balance_rmse": 379.02037679427764}386{"canonical_length": 34, "canonical_nll": 0.954249662511489, "generation_rouge_l_f1": 0.16438356164383564, "generation_token_f1": 0.19863013698630136, "prompt_id": "0e027c7a776ba5aa", "reading_canonical_log_prob": -32.444488525390625, "reading_distances": [18, 25, 24, 27], "reading_log_probs": [-35.660400390625, -41.939125061035156, -44.73474884033203, -39.50645446777344], "reading_logprob_std": 3.8473668031382835, "tokenization_path_logprob_std": 76.06572514575123, "trajectory_balance_rmse": 544.6742515622417}387{"canonical_length": 23, "canonical_nll": 2.5470362124235733, "generation_rouge_l_f1": 0.27027027027027023, "generation_token_f1": 0.37837837837837834, "prompt_id": "9d315888010aca18", "reading_canonical_log_prob": -58.58183288574219, "reading_distances": [19, 24, 26, 19], "reading_log_probs": [-57.77538299560547, -66.3636245727539, -55.46567916870117, -55.00309753417969], "reading_logprob_std": 5.282194637234893, "tokenization_path_logprob_std": 69.40950463828646, "trajectory_balance_rmse": 411.1335280299716}388{"canonical_length": 390, "canonical_nll": 0.9680700057592148, "generation_rouge_l_f1": 0.2523076923076923, "generation_token_f1": 0.47384615384615386, "prompt_id": "b3017655183fdef3", "reading_canonical_log_prob": -377.54730224609375, "reading_distances": [23, 22, 20, 23], "reading_log_probs": [-394.9715576171875, -390.18328857421875, -387.0314025878906, -387.90338134765625], "reading_logprob_std": 3.556981483462159, "tokenization_path_logprob_std": 241.64474086173036, "trajectory_balance_rmse": 7559.872926262616}389{"canonical_length": 24, "canonical_nll": 0.9711410999298096, "generation_rouge_l_f1": 0.47058823529411764, "generation_token_f1": 0.47058823529411764, "prompt_id": "70f7bc293fb9b6e2", "reading_canonical_log_prob": -23.30738639831543, "reading_distances": [36, 38, 32, 32], "reading_log_probs": [-25.18489646911621, -27.005340576171875, -26.258729934692383, -25.321163177490234], "reading_logprob_std": 0.8543328810114136, "tokenization_path_logprob_std": 52.31694072404289, "trajectory_balance_rmse": 407.16011701677775}390{"canonical_length": 85, "canonical_nll": 1.3632431927849264, "generation_rouge_l_f1": 0.19847328244274812, "generation_token_f1": 0.33587786259541985, "prompt_id": "6a1786b00d1d1907", "reading_canonical_log_prob": -115.87567138671875, "reading_distances": [30, 26, 28, 25], "reading_log_probs": [-117.73941040039062, -116.08768463134766, -114.87942504882812, -122.25216674804688], "reading_logprob_std": 3.2286572961247915, "tokenization_path_logprob_std": 152.465252649941, "trajectory_balance_rmse": 1241.6959854291283}391{"canonical_length": 108, "canonical_nll": 1.5481089839228876, "generation_rouge_l_f1": 0.20915032679738563, "generation_token_f1": 0.35947712418300654, "prompt_id": "7988aedbe67fdb66", "reading_canonical_log_prob": -167.19577026367188, "reading_distances": [15, 15, 17, 16], "reading_log_probs": [-169.99710083007812, -169.87252807617188, -173.1154022216797, -170.81527709960938], "reading_logprob_std": 1.5028949389394692, "tokenization_path_logprob_std": 59.81771657564745, "trajectory_balance_rmse": 2230.8354202274863}392{"canonical_length": 98, "canonical_nll": 2.2681377177335778, "generation_rouge_l_f1": 0.21917808219178084, "generation_token_f1": 0.2904109589041096, "prompt_id": "0b1fe757c0710416", "reading_canonical_log_prob": -222.27749633789062, "reading_distances": [12, 14, 16, 18], "reading_log_probs": [-221.0269775390625, -216.8408203125, -220.53570556640625, -216.68991088867188], "reading_logprob_std": 2.328098204607744, "tokenization_path_logprob_std": 139.5816820436514, "trajectory_balance_rmse": 2060.9704117798806}393{"canonical_length": 13, "canonical_nll": 1.0355417545025165, "generation_rouge_l_f1": 0.0885608856088561, "generation_token_f1": 0.09594095940959409, "prompt_id": "78fa2b118c48dd30", "reading_canonical_log_prob": -13.462042808532715, "reading_distances": [22, 33, 29, 27], "reading_log_probs": [-19.906930923461914, -12.719850540161133, -11.658503532409668, -13.309852600097656], "reading_logprob_std": 3.7351219849673445, "tokenization_path_logprob_std": 34.19438800502191, "trajectory_balance_rmse": 264.8969001914141}394{"canonical_length": 424, "canonical_nll": 1.154949260207842, "generation_rouge_l_f1": 0.1758957654723127, "generation_token_f1": 0.42996742671009774, "prompt_id": "995817411c15ddee", "reading_canonical_log_prob": -489.698486328125, "reading_distances": [18, 16, 10, 10], "reading_log_probs": [-488.78009033203125, -491.04443359375, -493.0199890136719, -494.1326599121094], "reading_logprob_std": 2.3528704080666136, "tokenization_path_logprob_std": 153.8209581554189, "trajectory_balance_rmse": 7960.183838519671}395{"canonical_length": 64, "canonical_nll": 1.226598858833313, "generation_rouge_l_f1": 0.22, "generation_token_f1": 0.42, "prompt_id": "fd32715308ec2a93", "reading_canonical_log_prob": -78.50232696533203, "reading_distances": [37, 47, 42, 46], "reading_log_probs": [-81.48258972167969, -83.5191650390625, -84.32918548583984, -80.91765594482422], "reading_logprob_std": 1.6235869020208455, "tokenization_path_logprob_std": 179.87358687340281, "trajectory_balance_rmse": 1450.2037148684446}396{"canonical_length": 333, "canonical_nll": 1.2576271034217812, "generation_rouge_l_f1": 0.1992619926199262, "generation_token_f1": 0.3837638376383764, "prompt_id": "6c1da32736122296", "reading_canonical_log_prob": -418.7898254394531, "reading_distances": [35, 35, 36, 41], "reading_log_probs": [-440.78387451171875, -442.00177001953125, -438.28302001953125, -440.83831787109375], "reading_logprob_std": 1.5666518522214836, "tokenization_path_logprob_std": 211.8144884373544, "trajectory_balance_rmse": 5512.367196924553}397{"canonical_length": 53, "canonical_nll": 0.8698786609577682, "generation_rouge_l_f1": 0.29885057471264365, "generation_token_f1": 0.33333333333333337, "prompt_id": "ffb0593302e5d193", "reading_canonical_log_prob": -46.10356903076172, "reading_distances": [40, 37, 35, 40], "reading_log_probs": [-54.68674850463867, -56.74581527709961, -55.343658447265625, -56.04718780517578], "reading_logprob_std": 0.8884049102201369, "tokenization_path_logprob_std": 92.20662307846209, "trajectory_balance_rmse": 840.1049776669003}398{"canonical_length": 142, "canonical_nll": 1.3711628443758253, "generation_rouge_l_f1": 0.27, "generation_token_f1": 0.47500000000000003, "prompt_id": "780311bcf262bb9d", "reading_canonical_log_prob": -194.7051239013672, "reading_distances": [35, 31, 31, 37], "reading_log_probs": [-206.6183624267578, -205.41683959960938, -215.19679260253906, -212.17576599121094], "reading_logprob_std": 4.622184045954031, "tokenization_path_logprob_std": 217.93780734951304, "trajectory_balance_rmse": 3315.0743035139635}399{"canonical_length": 32, "canonical_nll": 1.6802161931991577, "generation_rouge_l_f1": 0.16513761467889906, "generation_token_f1": 0.1743119266055046, "prompt_id": "ddc773449a2312dd", "reading_canonical_log_prob": -53.76691818237305, "reading_distances": [34, 28, 33, 31], "reading_log_probs": [-56.39167785644531, -56.31227493286133, -61.79896926879883, -57.543914794921875], "reading_logprob_std": 2.5868094129361534, "tokenization_path_logprob_std": 123.67756796279244, "trajectory_balance_rmse": 833.8891196922827}400{"canonical_length": 259, "canonical_nll": 0.6937935140602377, "generation_rouge_l_f1": 0.42125984251968507, "generation_token_f1": 0.5984251968503937, "prompt_id": "fa8dcc0f054c9984", "reading_canonical_log_prob": -179.69252014160156, "reading_distances": [21, 28, 27, 21], "reading_log_probs": [-180.43031311035156, -181.9407958984375, -183.1328582763672, -180.508544921875], "reading_logprob_std": 1.2894057093320868, "tokenization_path_logprob_std": 175.37713971416483, "trajectory_balance_rmse": 3062.8807527910285}401{"canonical_length": 79, "canonical_nll": 1.1975532242014437, "generation_rouge_l_f1": 0.24203821656050956, "generation_token_f1": 0.356687898089172, "prompt_id": "2d9359e9dd847731", "reading_canonical_log_prob": -94.60670471191406, "reading_distances": [21, 13, 17, 16], "reading_log_probs": [-98.6854248046875, -101.30604553222656, -103.56187438964844, -104.89466857910156], "reading_logprob_std": 2.7225209114647244, "tokenization_path_logprob_std": 137.93797568468372, "trajectory_balance_rmse": 1544.7977341847504}402{"canonical_length": 91, "canonical_nll": 1.8248128366994334, "generation_rouge_l_f1": 0.15254237288135594, "generation_token_f1": 0.2711864406779661, "prompt_id": "5cb4ccbc96ab0eb7", "reading_canonical_log_prob": -166.05796813964844, "reading_distances": [27, 35, 32, 31], "reading_log_probs": [-165.23590087890625, -168.80691528320312, -165.7862091064453, -168.65428161621094], "reading_logprob_std": 1.8733684601550462, "tokenization_path_logprob_std": 139.86929792645398, "trajectory_balance_rmse": 2046.183896731873}403{"canonical_length": 76, "canonical_nll": 0.5185963982029965, "generation_rouge_l_f1": 0.5, "generation_token_f1": 0.671875, "prompt_id": "23716ea232c0bbd4", "reading_canonical_log_prob": -39.413326263427734, "reading_distances": [4, 4, 2, 5], "reading_log_probs": [-42.489219665527344, -41.457515716552734, -41.719642639160156, -43.0225715637207], "reading_logprob_std": 0.7162902226004028, "tokenization_path_logprob_std": 22.97786989767884, "trajectory_balance_rmse": 898.9405766856223}404{"canonical_length": 7, "canonical_nll": 1.828620365687779, "generation_rouge_l_f1": 0.5, "generation_token_f1": 0.5, "prompt_id": "d26a8398a22a7919", "reading_canonical_log_prob": -12.800342559814453, "reading_distances": [39, 35, 40, 40], "reading_log_probs": [-16.118366241455078, -14.92773723602295, -14.693953514099121, -18.111234664916992], "reading_logprob_std": 1.5621884961971395, "tokenization_path_logprob_std": 26.416623095467454, "trajectory_balance_rmse": 187.6549864074895}405{"canonical_length": 177, "canonical_nll": 2.0149432683395125, "generation_rouge_l_f1": 0.18181818181818182, "generation_token_f1": 0.3545454545454546, "prompt_id": "cf319e0486bfdfdc", "reading_canonical_log_prob": -356.64495849609375, "reading_distances": [31, 27, 24, 18], "reading_log_probs": [-361.73944091796875, -366.71307373046875, -362.7591552734375, -359.95428466796875], "reading_logprob_std": 2.859860477757677, "tokenization_path_logprob_std": 208.20541018798733, "trajectory_balance_rmse": 4827.670492102212}406{"canonical_length": 8, "canonical_nll": 1.4560474157333374, "generation_rouge_l_f1": 0.23529411764705882, "generation_token_f1": 0.23529411764705882, "prompt_id": "2145da07c0ca4001", "reading_canonical_log_prob": -11.6483793258667, "reading_distances": [22, 22, 21, 25], "reading_log_probs": [-14.199138641357422, -18.612945556640625, -16.501550674438477, -14.797028541564941], "reading_logprob_std": 1.9804223393053837, "tokenization_path_logprob_std": 18.065679313569234, "trajectory_balance_rmse": 206.1316210974182}407{"canonical_length": 272, "canonical_nll": 1.2727304346421187, "generation_rouge_l_f1": 0.21172022684310018, "generation_token_f1": 0.4423440453686201, "prompt_id": "924cae405c1c753d", "reading_canonical_log_prob": -346.18267822265625, "reading_distances": [27, 22, 24, 21], "reading_log_probs": [-356.0998229980469, -354.5130615234375, -354.97796630859375, -354.1609191894531], "reading_logprob_std": 0.8437727313232184, "tokenization_path_logprob_std": 116.31918742122373, "trajectory_balance_rmse": 6195.003651690809}408{"canonical_length": 147, "canonical_nll": 1.3542856229405824, "generation_rouge_l_f1": 0.23821339950372208, "generation_token_f1": 0.4665012406947891, "prompt_id": "eb989912613f5027", "reading_canonical_log_prob": -199.07998657226562, "reading_distances": [30, 30, 34, 36], "reading_log_probs": [-201.70069885253906, -206.35906982421875, -201.6322479248047, -211.59413146972656], "reading_logprob_std": 4.730864562052036, "tokenization_path_logprob_std": 92.28016283523274, "trajectory_balance_rmse": 3173.448167389912}409{"canonical_length": 7, "canonical_nll": 1.5643997192382812, "generation_rouge_l_f1": 0.22222222222222224, "generation_token_f1": 0.888888888888889, "prompt_id": "1a61687e59394e25", "reading_canonical_log_prob": -10.950798034667969, "reading_distances": [19, 19, 20, 19], "reading_log_probs": [-13.478920936584473, -13.861639022827148, -14.018777847290039, -12.006739616394043], "reading_logprob_std": 0.9182809445651698, "tokenization_path_logprob_std": 16.238318627129956, "trajectory_balance_rmse": 69.42804630532325}410{"canonical_length": 12, "canonical_nll": 1.2177095413208008, "generation_rouge_l_f1": 0.3529411764705882, "generation_token_f1": 0.3529411764705882, "prompt_id": "687a259eb6971a8a", "reading_canonical_log_prob": -14.61251449584961, "reading_distances": [22, 21, 20, 22], "reading_log_probs": [-14.865496635437012, -15.758440017700195, -13.935758590698242, -15.60689926147461], "reading_logprob_std": 0.8341303803121306, "tokenization_path_logprob_std": 30.339476519780312, "trajectory_balance_rmse": 238.15359403469716}411{"canonical_length": 177, "canonical_nll": 1.2648030943789725, "generation_rouge_l_f1": 0.22277227722772278, "generation_token_f1": 0.43069306930693074, "prompt_id": "ae453e9c5b7b5add", "reading_canonical_log_prob": -223.87014770507812, "reading_distances": [27, 28, 24, 25], "reading_log_probs": [-232.10850524902344, -223.7992401123047, -225.55328369140625, -226.0338897705078], "reading_logprob_std": 3.6195888367916935, "tokenization_path_logprob_std": 187.07856693766283, "trajectory_balance_rmse": 3293.33240922311}412{"canonical_length": 5, "canonical_nll": 2.082530975341797, "generation_rouge_l_f1": 0.06451612903225806, "generation_token_f1": 0.06451612903225806, "prompt_id": "ffdd99b122d4c9bc", "reading_canonical_log_prob": -10.412654876708984, "reading_distances": [42, 43, 43, 40], "reading_log_probs": [-10.129598617553711, -12.31428337097168, -11.96825122833252, -15.256508827209473], "reading_logprob_std": 2.1218656513482483, "tokenization_path_logprob_std": 35.28547692848821, "trajectory_balance_rmse": 75.22854632950946}413{"canonical_length": 350, "canonical_nll": 0.95015869140625, "generation_rouge_l_f1": 0.2916666666666667, "generation_token_f1": 0.5381944444444444, "prompt_id": "56f7657e17072dd3", "reading_canonical_log_prob": -332.5555419921875, "reading_distances": [16, 12, 11, 15], "reading_log_probs": [-334.8297424316406, -333.821533203125, -333.98760986328125, -337.7784729003906], "reading_logprob_std": 1.8365792541084265, "tokenization_path_logprob_std": 223.39990318472547, "trajectory_balance_rmse": 7193.7686816275855}414{"canonical_length": 172, "canonical_nll": 1.2124674597451852, "generation_rouge_l_f1": 0.26327944572748263, "generation_token_f1": 0.48036951501154734, "prompt_id": "f0c44899ac8f3454", "reading_canonical_log_prob": -208.54440307617188, "reading_distances": [26, 32, 27, 32], "reading_log_probs": [-212.52716064453125, -214.69375610351562, -211.64013671875, -214.432861328125], "reading_logprob_std": 1.4805576257472792, "tokenization_path_logprob_std": 264.5044263630924, "trajectory_balance_rmse": 3953.7609178948437}415{"canonical_length": 316, "canonical_nll": 1.199454681782783, "generation_rouge_l_f1": 0.22654867256637168, "generation_token_f1": 0.37168141592920345, "prompt_id": "fb86a83720d2950d", "reading_canonical_log_prob": -379.0276794433594, "reading_distances": [25, 22, 27, 26], "reading_log_probs": [-395.44000244140625, -386.4945983886719, -395.403564453125, -395.2755126953125], "reading_logprob_std": 4.4397744680056555, "tokenization_path_logprob_std": 143.46794958403046, "trajectory_balance_rmse": 5952.261934435302}416{"canonical_length": 60, "canonical_nll": 1.522460683186849, "generation_rouge_l_f1": 0.13496932515337423, "generation_token_f1": 0.21472392638036808, "prompt_id": "4f164749bc77ca14", "reading_canonical_log_prob": -91.34764099121094, "reading_distances": [26, 25, 19, 23], "reading_log_probs": [-93.162109375, -92.49610900878906, -92.46174621582031, -97.38688659667969], "reading_logprob_std": 2.362214737668132, "tokenization_path_logprob_std": 96.55433776859495, "trajectory_balance_rmse": 1533.9472871883106}417{"canonical_length": 161, "canonical_nll": 1.4557332104037266, "generation_rouge_l_f1": 0.2946859903381642, "generation_token_f1": 0.5072463768115942, "prompt_id": "8d15f97380395bd9", "reading_canonical_log_prob": -234.373046875, "reading_distances": [26, 31, 27, 30], "reading_log_probs": [-246.29531860351562, -244.67605590820312, -249.3922576904297, -244.37294006347656], "reading_logprob_std": 2.2992130998440383, "tokenization_path_logprob_std": 181.13056332166138, "trajectory_balance_rmse": 2947.6903861139035}418{"canonical_length": 9, "canonical_nll": 1.0154796176486545, "generation_rouge_l_f1": 0.8571428571428571, "generation_token_f1": 0.8571428571428571, "prompt_id": "a75869a5f1810362", "reading_canonical_log_prob": -9.13931655883789, "reading_distances": [44, 44, 46, 48], "reading_log_probs": [-8.813225746154785, -8.861950874328613, -9.703242301940918, -8.9580659866333], "reading_logprob_std": 0.4171110388567184, "tokenization_path_logprob_std": 15.247612289827648, "trajectory_balance_rmse": 146.69811531686068}419{"canonical_length": 228, "canonical_nll": 1.287009791324013, "generation_rouge_l_f1": 0.24680851063829787, "generation_token_f1": 0.5702127659574467, "prompt_id": "f3ee3068b8bfca4b", "reading_canonical_log_prob": -293.438232421875, "reading_distances": [22, 24, 28, 26], "reading_log_probs": [-305.23065185546875, -309.18695068359375, -307.4166259765625, -304.7054443359375], "reading_logprob_std": 2.0671050774565063, "tokenization_path_logprob_std": 109.67893529449665, "trajectory_balance_rmse": 4779.225244046551}420{"canonical_length": 126, "canonical_nll": 1.421876211015005, "generation_rouge_l_f1": 0.0851063829787234, "generation_token_f1": 0.1276595744680851, "prompt_id": "6a31e531aca13031", "reading_canonical_log_prob": -179.15640258789062, "reading_distances": [22, 30, 22, 21], "reading_log_probs": [-174.74270629882812, -176.03208923339844, -175.02252197265625, -174.36953735351562], "reading_logprob_std": 0.7123759509045106, "tokenization_path_logprob_std": 135.66957609220265, "trajectory_balance_rmse": 2430.009143953337}421{"canonical_length": 11, "canonical_nll": 0.8036344701593573, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "01a0ba7b191885ca", "reading_canonical_log_prob": -8.83997917175293, "reading_distances": [12, 6, 7, 9], "reading_log_probs": [-8.675495147705078, -9.21739387512207, -12.99514102935791, -14.539042472839355], "reading_logprob_std": 2.862241706452205, "tokenization_path_logprob_std": 28.279263182242808, "trajectory_balance_rmse": 153.92515977326707}422{"canonical_length": 9, "canonical_nll": 2.0081808302137585, "generation_rouge_l_f1": 0.06106870229007635, "generation_token_f1": 0.06106870229007635, "prompt_id": "8b612a049e9d2b42", "reading_canonical_log_prob": -18.073627471923828, "reading_distances": [16, 11, 14, 13], "reading_log_probs": [-16.783618927001953, -15.88621711730957, -17.141433715820312, -16.65938949584961], "reading_logprob_std": 0.5287189903296164, "tokenization_path_logprob_std": 41.70310591800446, "trajectory_balance_rmse": 159.89029211091807}423{"canonical_length": 168, "canonical_nll": 0.635583105541411, "generation_rouge_l_f1": 0.30071599045346065, "generation_token_f1": 0.5011933174224344, "prompt_id": "1792c7fe245165d5", "reading_canonical_log_prob": -106.77796173095703, "reading_distances": [53, 51, 52, 52], "reading_log_probs": [-112.46949005126953, -110.92605590820312, -111.9970474243164, -112.57698059082031], "reading_logprob_std": 0.7542017202927738, "tokenization_path_logprob_std": 65.63094349222024, "trajectory_balance_rmse": 2298.183799666539}424{"canonical_length": 240, "canonical_nll": 1.526435089111328, "generation_rouge_l_f1": 0.31643002028397565, "generation_token_f1": 0.5760649087221095, "prompt_id": "9148af04e8535f5f", "reading_canonical_log_prob": -366.34442138671875, "reading_distances": [24, 22, 23, 22], "reading_log_probs": [-362.0837707519531, -367.4881591796875, -367.52044677734375, -366.0819091796875], "reading_logprob_std": 2.5625173713483673, "tokenization_path_logprob_std": 120.9593193453343, "trajectory_balance_rmse": 4388.835788951643}425{"canonical_length": 11, "canonical_nll": 1.072273687882857, "generation_rouge_l_f1": 0.07111111111111111, "generation_token_f1": 0.08, "prompt_id": "4f964e48306df1f7", "reading_canonical_log_prob": -11.795010566711426, "reading_distances": [19, 16, 11, 12], "reading_log_probs": [-12.742400169372559, -12.66817569732666, -13.03221607208252, -10.866519927978516], "reading_logprob_std": 0.9864565332555817, "tokenization_path_logprob_std": 24.154980118107826, "trajectory_balance_rmse": 156.06749560119758}426{"canonical_length": 13, "canonical_nll": 1.1911512521597056, "generation_rouge_l_f1": 0.4444444444444444, "generation_token_f1": 0.4444444444444444, "prompt_id": "65511a559eb975a4", "reading_canonical_log_prob": -15.484966278076172, "reading_distances": [22, 21, 18, 18], "reading_log_probs": [-16.463462829589844, -20.108642578125, -16.79416275024414, -20.23331069946289], "reading_logprob_std": 2.0501525976773967, "tokenization_path_logprob_std": 7.71784970956862, "trajectory_balance_rmse": 320.44410551060855}427{"canonical_length": 273, "canonical_nll": 1.149850586831788, "generation_rouge_l_f1": 0.2380106571936057, "generation_token_f1": 0.5079928952042629, "prompt_id": "dfd43bd8f8a6af69", "reading_canonical_log_prob": -313.9092102050781, "reading_distances": [16, 9, 16, 19], "reading_log_probs": [-312.97772216796875, -314.853515625, -313.8534240722656, -317.45709228515625], "reading_logprob_std": 1.9389737554769817, "tokenization_path_logprob_std": 227.40179365533257, "trajectory_balance_rmse": 4677.631504618023}428{"canonical_length": 137, "canonical_nll": 1.8882914076756387, "generation_rouge_l_f1": 0.19746835443037974, "generation_token_f1": 0.3848101265822785, "prompt_id": "9ebcabae9c8c462a", "reading_canonical_log_prob": -258.6959228515625, "reading_distances": [13, 16, 16, 21], "reading_log_probs": [-266.5311279296875, -257.57611083984375, -264.031005859375, -260.822509765625], "reading_logprob_std": 3.889413898337957, "tokenization_path_logprob_std": 178.15416759017953, "trajectory_balance_rmse": 3300.636360815767}429{"canonical_length": 224, "canonical_nll": 1.0538294655936105, "generation_rouge_l_f1": 0.25274725274725274, "generation_token_f1": 0.4468864468864469, "prompt_id": "4e90bb5af9e629db", "reading_canonical_log_prob": -236.05780029296875, "reading_distances": [57, 54, 44, 47], "reading_log_probs": [-237.9176788330078, -240.0740966796875, -234.8072052001953, -239.2624053955078], "reading_logprob_std": 2.316269829123015, "tokenization_path_logprob_std": 102.31237911224214, "trajectory_balance_rmse": 4178.649492603931}430{"canonical_length": 17, "canonical_nll": 0.7255789812873391, "generation_rouge_l_f1": 0.1016949152542373, "generation_token_f1": 0.1016949152542373, "prompt_id": "3779413f4fba856c", "reading_canonical_log_prob": -12.334842681884766, "reading_distances": [12, 16, 21, 18], "reading_log_probs": [-11.902875900268555, -14.050712585449219, -13.825162887573242, -18.813152313232422], "reading_logprob_std": 2.9392750399942877, "tokenization_path_logprob_std": 75.17128812571416, "trajectory_balance_rmse": 416.27789071911}431{"canonical_length": 125, "canonical_nll": 1.220606689453125, "generation_rouge_l_f1": 0.1907216494845361, "generation_token_f1": 0.34536082474226804, "prompt_id": "f749a091b4c13176", "reading_canonical_log_prob": -152.57583618164062, "reading_distances": [18, 22, 15, 19], "reading_log_probs": [-158.82644653320312, -158.65147399902344, -160.07366943359375, -158.32888793945312], "reading_logprob_std": 0.7640194689615724, "tokenization_path_logprob_std": 101.68036024320739, "trajectory_balance_rmse": 2785.403267080521}432{"canonical_length": 45, "canonical_nll": 1.65579104953342, "generation_rouge_l_f1": 0.25287356321839083, "generation_token_f1": 0.4367816091954023, "prompt_id": "6378420685c9b069", "reading_canonical_log_prob": -74.5105972290039, "reading_distances": [23, 19, 22, 21], "reading_log_probs": [-76.66085815429688, -75.52236938476562, -76.15478515625, -74.16487121582031], "reading_logprob_std": 1.0795340751058222, "tokenization_path_logprob_std": 156.07630395914921, "trajectory_balance_rmse": 1224.194868811541}433{"canonical_length": 119, "canonical_nll": 0.9417804108948267, "generation_rouge_l_f1": 0.30769230769230765, "generation_token_f1": 0.41687344913151364, "prompt_id": "634524f69adfb3f0", "reading_canonical_log_prob": -112.07186889648438, "reading_distances": [28, 25, 28, 25], "reading_log_probs": [-115.97245788574219, -117.4068603515625, -117.49293518066406, -117.1717529296875], "reading_logprob_std": 0.7055442730689117, "tokenization_path_logprob_std": 86.09569934558145, "trajectory_balance_rmse": 1939.6824200754545}434{"canonical_length": 162, "canonical_nll": 1.601744098427855, "generation_rouge_l_f1": 0.18097447795823665, "generation_token_f1": 0.3990719257540603, "prompt_id": "c5c8a284983e4726", "reading_canonical_log_prob": -259.4825439453125, "reading_distances": [20, 22, 23, 19], "reading_log_probs": [-270.94610595703125, -272.44775390625, -271.0970764160156, -275.5394592285156], "reading_logprob_std": 2.131008553374862, "tokenization_path_logprob_std": 46.664765615215245, "trajectory_balance_rmse": 2886.0998836296317}435{"canonical_length": 88, "canonical_nll": 1.406823071566495, "generation_rouge_l_f1": 0.23602484472049687, "generation_token_f1": 0.31677018633540377, "prompt_id": "626ba72af7ac67f2", "reading_canonical_log_prob": -123.80043029785156, "reading_distances": [31, 26, 29, 27], "reading_log_probs": [-118.77208709716797, -120.87725830078125, -118.7196044921875, -120.1730728149414], "reading_logprob_std": 1.066973075386978, "tokenization_path_logprob_std": 27.100401143441314, "trajectory_balance_rmse": 1340.5673122143746}436{"canonical_length": 297, "canonical_nll": 0.9811882250236742, "generation_rouge_l_f1": 0.30601092896174864, "generation_token_f1": 0.5063752276867031, "prompt_id": "c0257d516383b045", "reading_canonical_log_prob": -291.41290283203125, "reading_distances": [26, 29, 34, 23], "reading_log_probs": [-292.7803649902344, -294.20220947265625, -295.5119323730469, -293.3603515625], "reading_logprob_std": 1.1857798180965455, "tokenization_path_logprob_std": 233.60802189515869, "trajectory_balance_rmse": 6470.208585792425}437{"canonical_length": 38, "canonical_nll": 0.5403120141280325, "generation_rouge_l_f1": 0.19178082191780818, "generation_token_f1": 0.27397260273972607, "prompt_id": "dae9d23ce12cb989", "reading_canonical_log_prob": -20.531856536865234, "reading_distances": [21, 14, 18, 13], "reading_log_probs": [-22.993539810180664, -19.683027267456055, -18.75877571105957, -19.90225601196289], "reading_logprob_std": 1.8407110304702652, "tokenization_path_logprob_std": 29.741541319485773, "trajectory_balance_rmse": 272.6075256811076}438{"canonical_length": 10, "canonical_nll": 0.4533041477203369, "generation_rouge_l_f1": 0.10191082802547771, "generation_token_f1": 0.10191082802547771, "prompt_id": "cde09f27bb3893ac", "reading_canonical_log_prob": -4.533041477203369, "reading_distances": [29, 28, 22, 26], "reading_log_probs": [-6.487037181854248, -7.982783794403076, -5.77309513092041, -7.327946662902832], "reading_logprob_std": 0.9653668940349256, "tokenization_path_logprob_std": 29.930037522571848, "trajectory_balance_rmse": 204.4629089725426}439{"canonical_length": 174, "canonical_nll": 1.6286675464147808, "generation_rouge_l_f1": 0.18181818181818182, "generation_token_f1": 0.4449760765550239, "prompt_id": "64bb51344f196fcb", "reading_canonical_log_prob": -283.3881530761719, "reading_distances": [27, 18, 20, 18], "reading_log_probs": [-302.3888244628906, -297.1694030761719, -293.0627746582031, -295.41363525390625], "reading_logprob_std": 3.9617388040920285, "tokenization_path_logprob_std": 111.79118158727897, "trajectory_balance_rmse": 3447.344602607872}440{"canonical_length": 312, "canonical_nll": 1.3740428044245794, "generation_rouge_l_f1": 0.2146690518783542, "generation_token_f1": 0.42218246869409654, "prompt_id": "bc3436553628cc88", "reading_canonical_log_prob": -428.70135498046875, "reading_distances": [37, 30, 30, 31], "reading_log_probs": [-434.0008239746094, -437.2213439941406, -430.4400329589844, -444.0113525390625], "reading_logprob_std": 5.770129797306304, "tokenization_path_logprob_std": 429.394966901521, "trajectory_balance_rmse": 7080.0468603161125}441{"canonical_length": 39, "canonical_nll": 1.3790797698192108, "generation_rouge_l_f1": 0.16058394160583941, "generation_token_f1": 0.2116788321167883, "prompt_id": "b3b9c2d52d21969e", "reading_canonical_log_prob": -53.78411102294922, "reading_distances": [27, 30, 29, 26], "reading_log_probs": [-47.688926696777344, -49.15877914428711, -45.810691833496094, -50.142417907714844], "reading_logprob_std": 1.8852262503677373, "tokenization_path_logprob_std": 62.96241863028577, "trajectory_balance_rmse": 853.3207139254423}442{"canonical_length": 21, "canonical_nll": 0.44665027800060453, "generation_rouge_l_f1": 0.3125, "generation_token_f1": 0.375, "prompt_id": "cc11564e8f134b4b", "reading_canonical_log_prob": -9.379655838012695, "reading_distances": [15, 14, 14, 14], "reading_log_probs": [-10.397066116333008, -10.067957878112793, -10.780561447143555, -12.74848747253418], "reading_logprob_std": 1.2024398422301934, "tokenization_path_logprob_std": 15.021306672920716, "trajectory_balance_rmse": 167.78748072232764}443{"canonical_length": 20, "canonical_nll": 1.1066983222961426, "generation_rouge_l_f1": 0.17687074829931973, "generation_token_f1": 0.21768707482993196, "prompt_id": "e07efb6f1797853d", "reading_canonical_log_prob": -22.13396644592285, "reading_distances": [27, 28, 25, 24], "reading_log_probs": [-25.45095443725586, -25.115036010742188, -26.20442008972168, -26.49037742614746], "reading_logprob_std": 0.640380427771403, "tokenization_path_logprob_std": 60.89318772598331, "trajectory_balance_rmse": 342.80729298852174}444{"canonical_length": 10, "canonical_nll": 0.4856422901153564, "generation_rouge_l_f1": 0.72, "generation_token_f1": 0.72, "prompt_id": "302fe191c6bb7b5b", "reading_canonical_log_prob": -4.8564229011535645, "reading_distances": [24, 30, 26, 22], "reading_log_probs": [-3.394963502883911, -4.072648525238037, -3.2775416374206543, -2.847193956375122], "reading_logprob_std": 0.5076275989080096, "tokenization_path_logprob_std": 29.91239020872504, "trajectory_balance_rmse": 175.5198659283155}445{"canonical_length": 22, "canonical_nll": 1.7806927074085583, "generation_rouge_l_f1": 0.6451612903225806, "generation_token_f1": 0.6451612903225806, "prompt_id": "de0e448f061c6724", "reading_canonical_log_prob": -39.17523956298828, "reading_distances": [32, 25, 34, 28], "reading_log_probs": [-42.751556396484375, -38.50579071044922, -42.388206481933594, -40.63836669921875], "reading_logprob_std": 1.9430880052437518, "tokenization_path_logprob_std": 60.53024149999271, "trajectory_balance_rmse": 671.6319781392443}446{"canonical_length": 79, "canonical_nll": 1.0847809224189082, "generation_rouge_l_f1": 0.4204545454545454, "generation_token_f1": 0.5454545454545455, "prompt_id": "f979fa68d4ceeb01", "reading_canonical_log_prob": -85.69769287109375, "reading_distances": [26, 27, 25, 24], "reading_log_probs": [-91.84272766113281, -94.74813079833984, -91.2036361694336, -93.29007720947266], "reading_logprob_std": 1.580802135400658, "tokenization_path_logprob_std": 107.30373161099268, "trajectory_balance_rmse": 1600.0810973715124}447{"canonical_length": 193, "canonical_nll": 1.4736173165276878, "generation_rouge_l_f1": 0.20734341252699787, "generation_token_f1": 0.4319654427645788, "prompt_id": "e2f73d0c042e6f14", "reading_canonical_log_prob": -284.40814208984375, "reading_distances": [12, 13, 12, 16], "reading_log_probs": [-283.0965576171875, -284.97198486328125, -284.9179992675781, -275.4324035644531], "reading_logprob_std": 4.532817244182064, "tokenization_path_logprob_std": 273.6799200809332, "trajectory_balance_rmse": 4521.17724502577}448{"canonical_length": 216, "canonical_nll": 1.9129702250162761, "generation_rouge_l_f1": 0.21757322175732216, "generation_token_f1": 0.44769874476987453, "prompt_id": "409538e18660e490", "reading_canonical_log_prob": -413.2015686035156, "reading_distances": [28, 38, 30, 31], "reading_log_probs": [-417.87274169921875, -417.4302978515625, -417.89532470703125, -415.48785400390625], "reading_logprob_std": 1.142701810543534, "tokenization_path_logprob_std": 105.80963258299863, "trajectory_balance_rmse": 4974.494620913845}449{"canonical_length": 64, "canonical_nll": 2.094329833984375, "generation_rouge_l_f1": 0.28235294117647064, "generation_token_f1": 0.4, "prompt_id": "cc38da16fec412a8", "reading_canonical_log_prob": -134.037109375, "reading_distances": [32, 31, 26, 30], "reading_log_probs": [-137.23764038085938, -139.24740600585938, -135.27137756347656, -139.82675170898438], "reading_logprob_std": 2.0717287875066757, "tokenization_path_logprob_std": 81.48937713387564, "trajectory_balance_rmse": 1133.465868351297}450{"canonical_length": 53, "canonical_nll": 0.7230063384433962, "generation_rouge_l_f1": 0.3132530120481928, "generation_token_f1": 0.3373493975903614, "prompt_id": "0e93241c78cc22bd", "reading_canonical_log_prob": -38.3193359375, "reading_distances": [32, 28, 26, 35], "reading_log_probs": [-41.96589660644531, -42.27337646484375, -40.932411193847656, -41.085609436035156], "reading_logprob_std": 0.6563782914545387, "tokenization_path_logprob_std": 141.08975866925144, "trajectory_balance_rmse": 886.6182907605396}451{"canonical_length": 19, "canonical_nll": 0.262618667200992, "generation_rouge_l_f1": 0.21052631578947367, "generation_token_f1": 0.21052631578947367, "prompt_id": "32e3b4bcc84e208d", "reading_canonical_log_prob": -4.989754676818848, "reading_distances": [31, 28, 36, 29], "reading_log_probs": [-5.224601745605469, -4.248859405517578, -6.037315368652344, -4.792425632476807], "reading_logprob_std": 0.7551568363858465, "tokenization_path_logprob_std": 24.037412173864414, "trajectory_balance_rmse": 375.7036482893923}452{"canonical_length": 28, "canonical_nll": 1.6701654706682478, "generation_rouge_l_f1": 0.24390243902439027, "generation_token_f1": 0.34146341463414637, "prompt_id": "08748791c0916706", "reading_canonical_log_prob": -46.76463317871094, "reading_distances": [22, 19, 23, 21], "reading_log_probs": [-61.09075927734375, -66.70602416992188, -58.751949310302734, -59.68964767456055], "reading_logprob_std": 3.5630098444365412, "tokenization_path_logprob_std": 113.03800626249817, "trajectory_balance_rmse": 883.3063945550324}453{"canonical_length": 268, "canonical_nll": 1.5313757142024254, "generation_rouge_l_f1": 0.2085661080074488, "generation_token_f1": 0.4022346368715084, "prompt_id": "340800ab7642e87c", "reading_canonical_log_prob": -410.40869140625, "reading_distances": [20, 20, 25, 20], "reading_log_probs": [-421.4994812011719, -417.80828857421875, -425.68670654296875, -414.42388916015625], "reading_logprob_std": 4.844209999972094, "tokenization_path_logprob_std": 291.3080075962173, "trajectory_balance_rmse": 4808.825084112349}454{"canonical_length": 335, "canonical_nll": 1.3183905302588619, "generation_rouge_l_f1": 0.26513911620294595, "generation_token_f1": 0.5040916530278233, "prompt_id": "ba9377f6ac227fec", "reading_canonical_log_prob": -441.66082763671875, "reading_distances": [21, 25, 30, 26], "reading_log_probs": [-442.04742431640625, -443.58856201171875, -442.7471008300781, -443.2300109863281], "reading_logprob_std": 0.6666446972588527, "tokenization_path_logprob_std": 140.32440054821126, "trajectory_balance_rmse": 7146.905193428687}455{"canonical_length": 12, "canonical_nll": 0.13283525904019675, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "39897df6fcbea057", "reading_canonical_log_prob": -1.5940231084823608, "reading_distances": [22, 20, 26, 18], "reading_log_probs": [-2.6171364784240723, -3.0727105140686035, -4.429746150970459, -3.7065258026123047], "reading_logprob_std": 0.7877280536829799, "tokenization_path_logprob_std": 67.97425683918271, "trajectory_balance_rmse": 216.5630521213075}456{"canonical_length": 55, "canonical_nll": 0.6142262545498934, "generation_rouge_l_f1": 0.07407407407407407, "generation_token_f1": 0.07407407407407407, "prompt_id": "d9f80153d939b659", "reading_canonical_log_prob": -33.78244400024414, "reading_distances": [21, 17, 22, 21], "reading_log_probs": [-38.625099182128906, -38.01087951660156, -41.87369918823242, -39.161468505859375], "reading_logprob_std": 1.703422680975981, "tokenization_path_logprob_std": 84.74238732340827, "trajectory_balance_rmse": 1135.8671199106561}457{"canonical_length": 105, "canonical_nll": 1.5314878917875745, "generation_rouge_l_f1": 0.2569832402234637, "generation_token_f1": 0.36312849162011174, "prompt_id": "ea105e2e33c793e5", "reading_canonical_log_prob": -160.8062286376953, "reading_distances": [34, 44, 32, 30], "reading_log_probs": [-167.46926879882812, -170.61509704589844, -169.10720825195312, -165.695556640625], "reading_logprob_std": 2.1181774587962026, "tokenization_path_logprob_std": 111.7531280155133, "trajectory_balance_rmse": 2290.047860399186}458{"canonical_length": 311, "canonical_nll": 1.4810010903901225, "generation_rouge_l_f1": 0.22141560798548093, "generation_token_f1": 0.4355716878402904, "prompt_id": "2fb44fa5d32cf514", "reading_canonical_log_prob": -460.5913391113281, "reading_distances": [33, 30, 30, 31], "reading_log_probs": [-470.1924743652344, -475.5543518066406, -472.46099853515625, -474.8463134765625], "reading_logprob_std": 2.437793233457039, "tokenization_path_logprob_std": 83.40479140318233, "trajectory_balance_rmse": 6510.200236838438}459{"canonical_length": 134, "canonical_nll": 1.0879507491837686, "generation_rouge_l_f1": 0.23404255319148934, "generation_token_f1": 0.36170212765957444, "prompt_id": "f67931e5043d993b", "reading_canonical_log_prob": -145.785400390625, "reading_distances": [27, 26, 23, 29], "reading_log_probs": [-149.60708618164062, -151.29493713378906, -149.0601806640625, -151.93682861328125], "reading_logprob_std": 1.3618893388043194, "tokenization_path_logprob_std": 222.5264791984602, "trajectory_balance_rmse": 2626.40018546209}460{"canonical_length": 42, "canonical_nll": 1.5037651062011719, "generation_rouge_l_f1": 0.184, "generation_token_f1": 0.26399999999999996, "prompt_id": "f4dd2f85181ef8ef", "reading_canonical_log_prob": -63.15813446044922, "reading_distances": [23, 21, 17, 16], "reading_log_probs": [-70.77220153808594, -64.49278259277344, -67.48384094238281, -66.9707260131836], "reading_logprob_std": 2.5827239244746227, "tokenization_path_logprob_std": 73.0951918670028, "trajectory_balance_rmse": 967.0356042399493}461{"canonical_length": 386, "canonical_nll": 0.9010006603181671, "generation_rouge_l_f1": 0.22087745839636916, "generation_token_f1": 0.36006051437216335, "prompt_id": "044b1de06d0cdd9b", "reading_canonical_log_prob": -347.7862548828125, "reading_distances": [44, 47, 39, 43], "reading_log_probs": [-358.032470703125, -355.3140869140625, -356.61968994140625, -358.4769287109375], "reading_logprob_std": 1.4358680931219407, "tokenization_path_logprob_std": 349.00137800087634, "trajectory_balance_rmse": 6566.449618428324}462{"canonical_length": 120, "canonical_nll": 0.8130342483520507, "generation_rouge_l_f1": 0.254335260115607, "generation_token_f1": 0.3815028901734104, "prompt_id": "c7dd76b068a7ae00", "reading_canonical_log_prob": -97.5641098022461, "reading_distances": [41, 43, 39, 47], "reading_log_probs": [-106.33161926269531, -106.6903305053711, -104.8701400756836, -106.54872131347656], "reading_logprob_std": 0.839767319240347, "tokenization_path_logprob_std": 227.14267505410456, "trajectory_balance_rmse": 2189.365969965909}463{"canonical_length": 35, "canonical_nll": 2.624527849469866, "generation_rouge_l_f1": 0.47457627118644063, "generation_token_f1": 0.47457627118644063, "prompt_id": "2fa90c11cfdc9fd0", "reading_canonical_log_prob": -91.85847473144531, "reading_distances": [20, 19, 24, 27], "reading_log_probs": [-96.81599426269531, -97.5782241821289, -99.71719360351562, -101.49478149414062], "reading_logprob_std": 2.120601960434587, "tokenization_path_logprob_std": 17.014336412633238, "trajectory_balance_rmse": 704.7686791812469}464{"canonical_length": 33, "canonical_nll": 1.2507958845658735, "generation_rouge_l_f1": 0.2282608695652174, "generation_token_f1": 0.2934782608695652, "prompt_id": "acac7a68437d400e", "reading_canonical_log_prob": -41.27626419067383, "reading_distances": [24, 27, 34, 27], "reading_log_probs": [-40.82102584838867, -40.25585174560547, -44.45432662963867, -43.378623962402344], "reading_logprob_std": 2.0124122826525546, "tokenization_path_logprob_std": 67.68673660859952, "trajectory_balance_rmse": 522.7966391946442}465{"canonical_length": 41, "canonical_nll": 0.7476232575207222, "generation_rouge_l_f1": 0.2531645569620253, "generation_token_f1": 0.2531645569620253, "prompt_id": "98a669e9058b6ddc", "reading_canonical_log_prob": -30.65255355834961, "reading_distances": [12, 12, 9, 12], "reading_log_probs": [-28.665794372558594, -33.35615539550781, -29.542098999023438, -33.36289596557617], "reading_logprob_std": 2.482869992393883, "tokenization_path_logprob_std": 66.84393364532727, "trajectory_balance_rmse": 537.2777433795804}466{"canonical_length": 133, "canonical_nll": 1.3770439893679511, "generation_rouge_l_f1": 0.20565552699228792, "generation_token_f1": 0.38046272493573263, "prompt_id": "610adc3864696d51", "reading_canonical_log_prob": -183.1468505859375, "reading_distances": [21, 19, 21, 21], "reading_log_probs": [-188.90208435058594, -190.26113891601562, -191.7925567626953, -189.04769897460938], "reading_logprob_std": 1.3408636857468408, "tokenization_path_logprob_std": 161.44894524543784, "trajectory_balance_rmse": 3383.7186961309267}467{"canonical_length": 34, "canonical_nll": 0.76397424585679, "generation_rouge_l_f1": 0.31304347826086953, "generation_token_f1": 0.4695652173913043, "prompt_id": "74b66fda46123ad6", "reading_canonical_log_prob": -25.97512435913086, "reading_distances": [38, 42, 44, 39], "reading_log_probs": [-30.18025016784668, -26.557661056518555, -28.61836814880371, -29.233095169067383], "reading_logprob_std": 1.5341143107903763, "tokenization_path_logprob_std": 72.94176510352287, "trajectory_balance_rmse": 612.0763801056185}468{"canonical_length": 23, "canonical_nll": 0.6720493980076002, "generation_rouge_l_f1": 0.47727272727272724, "generation_token_f1": 0.5, "prompt_id": "9ce2e67b60ca246c", "reading_canonical_log_prob": -15.457136154174805, "reading_distances": [19, 18, 17, 19], "reading_log_probs": [-16.455381393432617, -16.244495391845703, -16.34274673461914, -15.566496849060059], "reading_logprob_std": 0.39991399496876034, "tokenization_path_logprob_std": 74.60361738996957, "trajectory_balance_rmse": 433.0897708178684}469{"canonical_length": 9, "canonical_nll": 0.8121350606282552, "generation_rouge_l_f1": 0.4615384615384615, "generation_token_f1": 0.4615384615384615, "prompt_id": "f37797ee8ed6a36e", "reading_canonical_log_prob": -7.309215545654297, "reading_distances": [42, 32, 42, 48], "reading_log_probs": [-8.101794242858887, -7.4295172691345215, -7.933881759643555, -7.355274677276611], "reading_logprob_std": 0.36879579384736333, "tokenization_path_logprob_std": 28.78139134012331, "trajectory_balance_rmse": 172.2497367389381}470{"canonical_length": 25, "canonical_nll": 1.2963720703125, "generation_rouge_l_f1": 0.2342342342342342, "generation_token_f1": 0.32432432432432434, "prompt_id": "b9be7345d7f0f09d", "reading_canonical_log_prob": -32.4093017578125, "reading_distances": [29, 23, 24, 28], "reading_log_probs": [-33.5026741027832, -32.49449920654297, -33.61565399169922, -33.75472640991211], "reading_logprob_std": 0.5742540985295208, "tokenization_path_logprob_std": 22.264623979596347, "trajectory_balance_rmse": 598.8446254882932}471{"canonical_length": 102, "canonical_nll": 0.843317892037186, "generation_rouge_l_f1": 0.44881889763779526, "generation_token_f1": 0.6141732283464567, "prompt_id": "101280c93f2c70da", "reading_canonical_log_prob": -86.01842498779297, "reading_distances": [18, 16, 10, 18], "reading_log_probs": [-90.13233947753906, -88.6663589477539, -90.51823425292969, -91.38253021240234], "reading_logprob_std": 1.1333955824315227, "tokenization_path_logprob_std": 83.2417633996584, "trajectory_balance_rmse": 844.0723613928112}472{"canonical_length": 309, "canonical_nll": 1.1472667706437096, "generation_rouge_l_f1": 0.3079584775086505, "generation_token_f1": 0.5570934256055362, "prompt_id": "bc039a3de3397314", "reading_canonical_log_prob": -354.50543212890625, "reading_distances": [35, 38, 27, 35], "reading_log_probs": [-357.4788818359375, -361.71466064453125, -358.12646484375, -357.35577392578125], "reading_logprob_std": 2.0584247303776024, "tokenization_path_logprob_std": 274.16761708139535, "trajectory_balance_rmse": 7036.429305840145}473{"canonical_length": 144, "canonical_nll": 1.0009341769748263, "generation_rouge_l_f1": 0.30319148936170215, "generation_token_f1": 0.4042553191489362, "prompt_id": "3bf14cf65dcd8fce", "reading_canonical_log_prob": -144.134521484375, "reading_distances": [35, 36, 35, 39], "reading_log_probs": [-148.44955444335938, -148.1936492919922, -147.87136840820312, -148.66323852539062], "reading_logprob_std": 0.3411847890481103, "tokenization_path_logprob_std": 159.13775204665342, "trajectory_balance_rmse": 2228.8527436535633}474{"canonical_length": 38, "canonical_nll": 1.768587212813528, "generation_rouge_l_f1": 0.07860262008733623, "generation_token_f1": 0.11353711790393012, "prompt_id": "ab60c5be348f9273", "reading_canonical_log_prob": -67.20631408691406, "reading_distances": [15, 20, 25, 22], "reading_log_probs": [-71.50709533691406, -74.06808471679688, -75.06143188476562, -73.79474639892578], "reading_logprob_std": 1.5025330510218047, "tokenization_path_logprob_std": 80.77179384749951, "trajectory_balance_rmse": 965.0802446656214}475{"canonical_length": 12, "canonical_nll": 2.6842727661132812, "generation_rouge_l_f1": 0.4705882352941177, "generation_token_f1": 0.4705882352941177, "prompt_id": "8392d28ac48cc769", "reading_canonical_log_prob": -32.211273193359375, "reading_distances": [38, 35, 33, 32], "reading_log_probs": [-42.30015182495117, -38.82861328125, -37.77113342285156, -39.93864440917969], "reading_logprob_std": 1.9405506403563186, "tokenization_path_logprob_std": 66.39971939720753, "trajectory_balance_rmse": 354.415232654713}476{"canonical_length": 90, "canonical_nll": 1.8579216851128473, "generation_rouge_l_f1": 0.1756373937677054, "generation_token_f1": 0.3002832861189802, "prompt_id": "2a39144255d482e3", "reading_canonical_log_prob": -167.21295166015625, "reading_distances": [29, 29, 25, 23], "reading_log_probs": [-185.22898864746094, -186.06015014648438, -187.71826171875, -200.93710327148438], "reading_logprob_std": 7.373618600389884, "tokenization_path_logprob_std": 157.76798533648912, "trajectory_balance_rmse": 1759.4135486332027}477{"canonical_length": 292, "canonical_nll": 1.4357329068118578, "generation_rouge_l_f1": 0.21428571428571427, "generation_token_f1": 0.4642857142857143, "prompt_id": "7938846720b7f660", "reading_canonical_log_prob": -419.2340087890625, "reading_distances": [20, 13, 22, 17], "reading_log_probs": [-435.8888854980469, -432.77587890625, -437.81787109375, -435.06195068359375], "reading_logprob_std": 2.0884308972976577, "tokenization_path_logprob_std": 315.0597538608927, "trajectory_balance_rmse": 6018.966703989607}478{"canonical_length": 196, "canonical_nll": 1.4962759212571748, "generation_rouge_l_f1": 0.29310344827586204, "generation_token_f1": 0.5431034482758621, "prompt_id": "2a7c37e2ef87098f", "reading_canonical_log_prob": -293.27008056640625, "reading_distances": [27, 30, 28, 31], "reading_log_probs": [-304.1692810058594, -302.710205078125, -301.85552978515625, -305.9185485839844], "reading_logprob_std": 1.7812529298621615, "tokenization_path_logprob_std": 211.1605143974339, "trajectory_balance_rmse": 4663.353800867616}479{"canonical_length": 284, "canonical_nll": 1.2908374625192562, "generation_rouge_l_f1": 0.2340036563071298, "generation_token_f1": 0.45338208409506403, "prompt_id": "f9f1a4c90d1c6936", "reading_canonical_log_prob": -366.59783935546875, "reading_distances": [27, 25, 28, 28], "reading_log_probs": [-377.5123291015625, -377.4571533203125, -382.6075134277344, -377.55767822265625], "reading_logprob_std": 2.5495613073273407, "tokenization_path_logprob_std": 277.819351910311, "trajectory_balance_rmse": 6488.997189180762}480{"canonical_length": 149, "canonical_nll": 1.3334452993917785, "generation_rouge_l_f1": 0.22748815165876776, "generation_token_f1": 0.4170616113744076, "prompt_id": "4c7fd497c60d26bd", "reading_canonical_log_prob": -198.683349609375, "reading_distances": [20, 22, 27, 23], "reading_log_probs": [-217.7579803466797, -214.26889038085938, -219.86676025390625, -222.1246795654297], "reading_logprob_std": 3.3396174181966627, "tokenization_path_logprob_std": 155.23869082699355, "trajectory_balance_rmse": 3105.7145056818563}481{"canonical_length": 25, "canonical_nll": 1.1624423217773439, "generation_rouge_l_f1": 0.36363636363636365, "generation_token_f1": 0.5194805194805194, "prompt_id": "c2e4f5be5a2ad0b2", "reading_canonical_log_prob": -29.061058044433594, "reading_distances": [13, 17, 20, 16], "reading_log_probs": [-30.69232177734375, -34.34397506713867, -35.568565368652344, -33.54328155517578], "reading_logprob_std": 2.0712878642134553, "tokenization_path_logprob_std": 105.67301773532088, "trajectory_balance_rmse": 496.4865662528749}482{"canonical_length": 14, "canonical_nll": 1.6785718372889928, "generation_rouge_l_f1": 0.22222222222222224, "generation_token_f1": 0.22222222222222224, "prompt_id": "619e2fb6adbffc57", "reading_canonical_log_prob": -23.5000057220459, "reading_distances": [36, 30, 30, 30], "reading_log_probs": [-32.11142349243164, -27.97895050048828, -35.15129089355469, -29.007062911987305], "reading_logprob_std": 3.2430207213997067, "tokenization_path_logprob_std": 63.737878489233296, "trajectory_balance_rmse": 367.213831137916}483{"canonical_length": 136, "canonical_nll": 1.444884692921358, "generation_rouge_l_f1": 0.19597989949748745, "generation_token_f1": 0.42211055276381915, "prompt_id": "237ba288f6603bf1", "reading_canonical_log_prob": -196.5043182373047, "reading_distances": [18, 17, 18, 21], "reading_log_probs": [-198.91807556152344, -194.83901977539062, -195.85227966308594, -196.4698486328125], "reading_logprob_std": 1.7344380811064466, "tokenization_path_logprob_std": 215.60616746496612, "trajectory_balance_rmse": 2853.0923973247072}484{"canonical_length": 9, "canonical_nll": 0.7298073238796658, "generation_rouge_l_f1": 0.7142857142857143, "generation_token_f1": 0.7142857142857143, "prompt_id": "52afc3f6a37f4218", "reading_canonical_log_prob": -6.568265914916992, "reading_distances": [27, 26, 24, 22], "reading_log_probs": [-6.9712066650390625, -7.048493385314941, -6.648263454437256, -6.411247730255127], "reading_logprob_std": 0.29527699411198527, "tokenization_path_logprob_std": 26.30645772873089, "trajectory_balance_rmse": 208.13145530139087}485{"canonical_length": 115, "canonical_nll": 1.6382422405740489, "generation_rouge_l_f1": 0.2124352331606217, "generation_token_f1": 0.33678756476683935, "prompt_id": "0096b48ae77be6ff", "reading_canonical_log_prob": -188.39785766601562, "reading_distances": [35, 31, 30, 25], "reading_log_probs": [-206.78240966796875, -202.76815795898438, -205.95448303222656, -213.82177734375], "reading_logprob_std": 4.659952300648729, "tokenization_path_logprob_std": 133.70009388089585, "trajectory_balance_rmse": 2927.8016659837945}486{"canonical_length": 252, "canonical_nll": 1.0745876251705109, "generation_rouge_l_f1": 0.32193158953722334, "generation_token_f1": 0.5392354124748491, "prompt_id": "2061c66b7116786e", "reading_canonical_log_prob": -270.79608154296875, "reading_distances": [31, 38, 33, 29], "reading_log_probs": [-274.17425537109375, -277.5560607910156, -282.865234375, -275.3291931152344], "reading_logprob_std": 3.854021566240901, "tokenization_path_logprob_std": 183.67930930731222, "trajectory_balance_rmse": 4845.009358418934}487{"canonical_length": 94, "canonical_nll": 1.463074379778923, "generation_rouge_l_f1": 0.2376237623762376, "generation_token_f1": 0.4158415841584158, "prompt_id": "5622ae2789dbdaf4", "reading_canonical_log_prob": -137.52899169921875, "reading_distances": [18, 18, 16, 17], "reading_log_probs": [-139.47918701171875, -137.99563598632812, -140.32147216796875, -136.25228881835938], "reading_logprob_std": 1.7872293805476802, "tokenization_path_logprob_std": 125.32891302529224, "trajectory_balance_rmse": 1983.5029013430592}488{"canonical_length": 279, "canonical_nll": 1.2267677006328406, "generation_rouge_l_f1": 0.2734082397003746, "generation_token_f1": 0.4981273408239701, "prompt_id": "531c2e72ca7ad121", "reading_canonical_log_prob": -342.2681884765625, "reading_distances": [27, 16, 22, 25], "reading_log_probs": [-347.0062255859375, -345.9197692871094, -344.7030029296875, -346.58355712890625], "reading_logprob_std": 1.005049305663082, "tokenization_path_logprob_std": 207.78444983338775, "trajectory_balance_rmse": 5823.343242079168}489{"canonical_length": 33, "canonical_nll": 0.7458625562263258, "generation_rouge_l_f1": 0.6557377049180327, "generation_token_f1": 0.7868852459016394, "prompt_id": "50acd3d956ee080e", "reading_canonical_log_prob": -24.61346435546875, "reading_distances": [26, 23, 37, 36], "reading_log_probs": [-22.774341583251953, -24.90546417236328, -21.449359893798828, -22.785728454589844], "reading_logprob_std": 1.4294872873351052, "tokenization_path_logprob_std": 59.704915743231936, "trajectory_balance_rmse": 675.304732076572}490{"canonical_length": 26, "canonical_nll": 1.3701597360464244, "generation_rouge_l_f1": 0.27586206896551724, "generation_token_f1": 0.31034482758620685, "prompt_id": "3fcc8e674911ba73", "reading_canonical_log_prob": -35.62415313720703, "reading_distances": [16, 15, 15, 15], "reading_log_probs": [-36.22119903564453, -36.13423156738281, -37.36806106567383, -34.083892822265625], "reading_logprob_std": 1.3663491183258996, "tokenization_path_logprob_std": 41.01854271142664, "trajectory_balance_rmse": 496.36864545146307}491{"canonical_length": 347, "canonical_nll": 0.9631698564424982, "generation_rouge_l_f1": 0.19354838709677422, "generation_token_f1": 0.4161290322580645, "prompt_id": "915d77693c098234", "reading_canonical_log_prob": -334.2199401855469, "reading_distances": [17, 13, 11, 17], "reading_log_probs": [-340.56732177734375, -340.8631591796875, -340.18865966796875, -337.990478515625], "reading_logprob_std": 1.304168640829406, "tokenization_path_logprob_std": 225.7371750316421, "trajectory_balance_rmse": 5973.995820715071}492{"canonical_length": 16, "canonical_nll": 0.18422217667102814, "generation_rouge_l_f1": 0.9333333333333333, "generation_token_f1": 0.9333333333333333, "prompt_id": "d4ba094ba07a6bba", "reading_canonical_log_prob": -2.94755482673645, "reading_distances": [21, 21, 28, 21], "reading_log_probs": [-2.192138671875, -1.6134285926818848, -1.9873449802398682, -1.6639785766601562], "reading_logprob_std": 0.27427685143532327, "tokenization_path_logprob_std": 41.99690457532884, "trajectory_balance_rmse": 251.11902424578767}493{"canonical_length": 35, "canonical_nll": 0.8547585623604911, "generation_rouge_l_f1": 0.14765100671140938, "generation_token_f1": 0.20805369127516776, "prompt_id": "40f13ac2314179d4", "reading_canonical_log_prob": -29.916549682617188, "reading_distances": [20, 25, 28, 30], "reading_log_probs": [-30.164011001586914, -31.576152801513672, -31.11124610900879, -28.59696388244629], "reading_logprob_std": 1.3153059038514463, "tokenization_path_logprob_std": 93.99020490617562, "trajectory_balance_rmse": 767.2853649859229}494{"canonical_length": 228, "canonical_nll": 1.1480026245117188, "generation_rouge_l_f1": 0.2063157894736842, "generation_token_f1": 0.37894736842105264, "prompt_id": "c60f4d0722caa17f", "reading_canonical_log_prob": -261.7445983886719, "reading_distances": [62, 53, 67, 59], "reading_log_probs": [-257.4425964355469, -259.5035095214844, -253.00753784179688, -256.6296691894531], "reading_logprob_std": 2.7103874262149827, "tokenization_path_logprob_std": 222.96401596777807, "trajectory_balance_rmse": 4871.850454854969}495{"canonical_length": 20, "canonical_nll": 1.856924819946289, "generation_rouge_l_f1": 0.45614035087719296, "generation_token_f1": 0.5263157894736843, "prompt_id": "2ca8b957d3b2b5bc", "reading_canonical_log_prob": -37.13849639892578, "reading_distances": [24, 23, 30, 30], "reading_log_probs": [-45.36573791503906, -42.21501159667969, -42.221397399902344, -43.55662536621094], "reading_logprob_std": 1.4907951969153714, "tokenization_path_logprob_std": 68.97314752561189, "trajectory_balance_rmse": 474.0506511026531}496{"canonical_length": 159, "canonical_nll": 1.2718592229879127, "generation_rouge_l_f1": 0.2795180722891566, "generation_token_f1": 0.5397590361445783, "prompt_id": "762bfb8c99f50dd6", "reading_canonical_log_prob": -202.22561645507812, "reading_distances": [22, 20, 31, 20], "reading_log_probs": [-224.27301025390625, -224.7239990234375, -221.96066284179688, -226.97366333007812], "reading_logprob_std": 2.054893726027657, "tokenization_path_logprob_std": 88.18625027462478, "trajectory_balance_rmse": 2904.9905430122676}497{"canonical_length": 231, "canonical_nll": 1.1060879529812635, "generation_rouge_l_f1": 0.26640159045725653, "generation_token_f1": 0.5208747514910537, "prompt_id": "9f2ef447478c22f4", "reading_canonical_log_prob": -255.50631713867188, "reading_distances": [29, 27, 19, 24], "reading_log_probs": [-258.28033447265625, -256.23236083984375, -256.6250305175781, -258.201904296875], "reading_logprob_std": 1.0590955377760711, "tokenization_path_logprob_std": 242.32761282447237, "trajectory_balance_rmse": 4947.2207670783}498{"canonical_length": 361, "canonical_nll": 1.2217618567131232, "generation_rouge_l_f1": 0.26870748299319724, "generation_token_f1": 0.46938775510204084, "prompt_id": "dc0376ea25a8c1bc", "reading_canonical_log_prob": -441.0560302734375, "reading_distances": [27, 26, 30, 30], "reading_log_probs": [-441.4711608886719, -449.47027587890625, -443.85906982421875, -449.50640869140625], "reading_logprob_std": 4.058248105319502, "tokenization_path_logprob_std": 177.26023119872497, "trajectory_balance_rmse": 6356.144663261003}499{"canonical_length": 18, "canonical_nll": 3.1914776696099176, "generation_rouge_l_f1": 0.16666666666666666, "generation_token_f1": 0.25, "prompt_id": "9cf805580809e94d", "reading_canonical_log_prob": -57.446598052978516, "reading_distances": [31, 22, 23, 20], "reading_log_probs": [-66.4383316040039, -69.1578369140625, -65.52198028564453, -63.044036865234375], "reading_logprob_std": 2.52479130652334, "tokenization_path_logprob_std": 42.112332046540466, "trajectory_balance_rmse": 406.7745726026968}500{"canonical_length": 280, "canonical_nll": 1.6038649422781808, "generation_rouge_l_f1": 0.25136612021857924, "generation_token_f1": 0.49544626593806923, "prompt_id": "d3ad116b6d046506", "reading_canonical_log_prob": -449.0821838378906, "reading_distances": [24, 22, 22, 24], "reading_log_probs": [-451.6078796386719, -450.74249267578125, -452.28106689453125, -451.01953125], "reading_logprob_std": 0.6821328489912806, "tokenization_path_logprob_std": 144.60130208022719, "trajectory_balance_rmse": 6988.715636500745}501{"canonical_length": 141, "canonical_nll": 1.3806570066627881, "generation_rouge_l_f1": 0.31521739130434784, "generation_token_f1": 0.5054347826086957, "prompt_id": "d4c38137aba16167", "reading_canonical_log_prob": -194.67263793945312, "reading_distances": [12, 14, 9, 16], "reading_log_probs": [-212.72943115234375, -230.49798583984375, -208.43763732910156, -214.19049072265625], "reading_logprob_std": 9.669386487570504, "tokenization_path_logprob_std": 22.928518015998208, "trajectory_balance_rmse": 1344.6668965534716}502{"canonical_length": 41, "canonical_nll": 1.492904104837557, "generation_rouge_l_f1": 0.23333333333333334, "generation_token_f1": 0.26666666666666666, "prompt_id": "84547b634d24f63a", "reading_canonical_log_prob": -61.209068298339844, "reading_distances": [24, 28, 29, 24], "reading_log_probs": [-66.34956359863281, -61.8855094909668, -65.79124450683594, -64.57820129394531], "reading_logprob_std": 1.9864971829675584, "tokenization_path_logprob_std": 146.6855796618001, "trajectory_balance_rmse": 977.4235555886478}503{"canonical_length": 6, "canonical_nll": 0.28261741002400714, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "58c9afa198f60472", "reading_canonical_log_prob": -1.695704460144043, "reading_distances": [43, 40, 52, 40], "reading_log_probs": [-2.9544925689697266, -4.636956691741943, -4.03112268447876, -4.326112270355225], "reading_logprob_std": 0.731541550978987, "tokenization_path_logprob_std": 13.700173307291184, "trajectory_balance_rmse": 61.574946722902844}504{"canonical_length": 238, "canonical_nll": 1.566974030823267, "generation_rouge_l_f1": 0.20491803278688522, "generation_token_f1": 0.38524590163934425, "prompt_id": "74608c6aad72d41d", "reading_canonical_log_prob": -372.9398193359375, "reading_distances": [37, 39, 36, 32], "reading_log_probs": [-387.5786437988281, -390.0888671875, -384.40875244140625, -385.98321533203125], "reading_logprob_std": 2.4237340123698243, "tokenization_path_logprob_std": 197.2544197274615, "trajectory_balance_rmse": 5502.125293812229}505{"canonical_length": 296, "canonical_nll": 1.5081440693623311, "generation_rouge_l_f1": 0.23290203327171904, "generation_token_f1": 0.4658040665434381, "prompt_id": "5c242c11a7bd0e44", "reading_canonical_log_prob": -446.41064453125, "reading_distances": [26, 35, 28, 36], "reading_log_probs": [-447.58624267578125, -446.2974853515625, -448.89532470703125, -447.14385986328125], "reading_logprob_std": 1.0840910934920769, "tokenization_path_logprob_std": 282.42208825275253, "trajectory_balance_rmse": 7235.993046339855}506{"canonical_length": 42, "canonical_nll": 0.6803854533604213, "generation_rouge_l_f1": 0.7391304347826086, "generation_token_f1": 0.7391304347826086, "prompt_id": "c3513f19426dd701", "reading_canonical_log_prob": -28.576189041137695, "reading_distances": [23, 27, 26, 33], "reading_log_probs": [-27.61651611328125, -27.425813674926758, -25.63974380493164, -30.289443969726562], "reading_logprob_std": 1.9170001559099885, "tokenization_path_logprob_std": 28.589321865529545, "trajectory_balance_rmse": 330.5497928656245}507{"canonical_length": 356, "canonical_nll": 1.0775454231862271, "generation_rouge_l_f1": 0.23723228995057663, "generation_token_f1": 0.4316309719934102, "prompt_id": "80a15e1bafbe7e14", "reading_canonical_log_prob": -383.6061706542969, "reading_distances": [20, 20, 17, 16], "reading_log_probs": [-386.8230895996094, -383.41168212890625, -382.0411376953125, -385.4070739746094], "reading_logprob_std": 2.115407228231757, "tokenization_path_logprob_std": 375.0872726835657, "trajectory_balance_rmse": 7345.277647668734}508{"canonical_length": 51, "canonical_nll": 1.455877565870098, "generation_rouge_l_f1": 0.42372881355932207, "generation_token_f1": 0.5932203389830508, "prompt_id": "74f44ff0ad2c7451", "reading_canonical_log_prob": -74.249755859375, "reading_distances": [23, 25, 25, 22], "reading_log_probs": [-79.43572235107422, -75.97425842285156, -75.63771057128906, -79.24729919433594], "reading_logprob_std": 2.0473014691618747, "tokenization_path_logprob_std": 57.27556666445529, "trajectory_balance_rmse": 1262.5322291314774}509{"canonical_length": 21, "canonical_nll": 1.522920426868257, "generation_rouge_l_f1": 0.42857142857142855, "generation_token_f1": 0.42857142857142855, "prompt_id": "0b30260636e7066b", "reading_canonical_log_prob": -31.9813289642334, "reading_distances": [32, 34, 23, 31], "reading_log_probs": [-35.43981170654297, -34.33555603027344, -33.90818786621094, -30.824800491333008], "reading_logprob_std": 1.9765058902174715, "tokenization_path_logprob_std": 94.29886198289628, "trajectory_balance_rmse": 545.8459541378468}510{"canonical_length": 142, "canonical_nll": 1.260041465221996, "generation_rouge_l_f1": 0.3206521739130435, "generation_token_f1": 0.4782608695652174, "prompt_id": "270a3b8ee410411a", "reading_canonical_log_prob": -178.92588806152344, "reading_distances": [24, 31, 31, 29], "reading_log_probs": [-182.241455078125, -181.06454467773438, -181.67462158203125, -183.90481567382812], "reading_logprob_std": 1.2208701002998117, "tokenization_path_logprob_std": 139.13368040157542, "trajectory_balance_rmse": 2452.5985852785866}511{"canonical_length": 2, "canonical_nll": 2.0206985473632812, "generation_rouge_l_f1": 0.09523809523809523, "generation_token_f1": 0.09523809523809523, "prompt_id": "aa51f5831004ecdf", "reading_canonical_log_prob": -4.0413970947265625, "reading_distances": [35, 33, 39, 28], "reading_log_probs": [-2.7577199935913086, -2.635730504989624, -2.442366600036621, -3.028972625732422], "reading_logprob_std": 0.24563570303175083, "tokenization_path_logprob_std": 25.507922400862387, "trajectory_balance_rmse": 44.448525168120014}512{"canonical_length": 53, "canonical_nll": 0.8317480986973025, "generation_rouge_l_f1": 0.22754491017964076, "generation_token_f1": 0.2994011976047904, "prompt_id": "d23e5c49d3b87636", "reading_canonical_log_prob": -44.08264923095703, "reading_distances": [20, 16, 18, 13], "reading_log_probs": [-39.881675720214844, -39.68144607543945, -40.696563720703125, -41.36808776855469], "reading_logprob_std": 0.7767305261457451, "tokenization_path_logprob_std": 135.5072456426632, "trajectory_balance_rmse": 883.2246899117914}513{"canonical_length": 78, "canonical_nll": 1.648043705866887, "generation_rouge_l_f1": 0.18627450980392157, "generation_token_f1": 0.29411764705882354, "prompt_id": "1cf9ae29e6e4ee3d", "reading_canonical_log_prob": -128.5474090576172, "reading_distances": [19, 22, 27, 21], "reading_log_probs": [-135.13223266601562, -140.9285430908203, -134.74647521972656, -141.25997924804688], "reading_logprob_std": 3.5595979330271543, "tokenization_path_logprob_std": 74.60312336435912, "trajectory_balance_rmse": 1716.7110534983306}514{"canonical_length": 113, "canonical_nll": 1.3522650794645326, "generation_rouge_l_f1": 0.25263157894736843, "generation_token_f1": 0.39473684210526316, "prompt_id": "4b26a788bce312c3", "reading_canonical_log_prob": -152.8059539794922, "reading_distances": [27, 37, 31, 34], "reading_log_probs": [-167.42271423339844, -173.77603149414062, -171.5370635986328, -167.88430786132812], "reading_logprob_std": 3.0355338447457463, "tokenization_path_logprob_std": 256.9828005432629, "trajectory_balance_rmse": 2383.4541577335463}515{"canonical_length": 320, "canonical_nll": 1.5254377365112304, "generation_rouge_l_f1": 0.2747826086956522, "generation_token_f1": 0.5217391304347826, "prompt_id": "c59752dc85c30f5b", "reading_canonical_log_prob": -488.14007568359375, "reading_distances": [35, 38, 41, 35], "reading_log_probs": [-493.74945068359375, -491.06951904296875, -494.5927734375, -494.52801513671875], "reading_logprob_std": 1.6552467054273385, "tokenization_path_logprob_std": 337.2181815196017, "trajectory_balance_rmse": 8135.746287011251}516{"canonical_length": 9, "canonical_nll": 1.7420507007175021, "generation_rouge_l_f1": 0.5454545454545454, "generation_token_f1": 0.5454545454545454, "prompt_id": "b4e868a234198084", "reading_canonical_log_prob": -15.67845630645752, "reading_distances": [25, 32, 30, 31], "reading_log_probs": [-14.703886032104492, -15.436767578125, -15.079291343688965, -17.495201110839844], "reading_logprob_std": 1.2473652982401573, "tokenization_path_logprob_std": 33.58975684704447, "trajectory_balance_rmse": 247.81142597901436}517{"canonical_length": 68, "canonical_nll": 1.7076667336856617, "generation_rouge_l_f1": 0.23214285714285715, "generation_token_f1": 0.3660714285714286, "prompt_id": "466e7627e6013e86", "reading_canonical_log_prob": -116.121337890625, "reading_distances": [33, 36, 37, 37], "reading_log_probs": [-112.8511962890625, -114.12803649902344, -114.45323181152344, -111.87359619140625], "reading_logprob_std": 1.1900740072363751, "tokenization_path_logprob_std": 96.52013366405907, "trajectory_balance_rmse": 1618.2871377394356}518{"canonical_length": 14, "canonical_nll": 0.9596988133021763, "generation_rouge_l_f1": 0.4827586206896552, "generation_token_f1": 0.4827586206896552, "prompt_id": "ca45cac53e43d02f", "reading_canonical_log_prob": -13.435783386230469, "reading_distances": [24, 24, 23, 29], "reading_log_probs": [-15.832087516784668, -14.18946647644043, -15.689343452453613, -15.773165702819824], "reading_logprob_std": 0.789874034404107, "tokenization_path_logprob_std": 36.2137540491147, "trajectory_balance_rmse": 178.52545880930202}519{"canonical_length": 20, "canonical_nll": 2.200551223754883, "generation_rouge_l_f1": 0.10309278350515463, "generation_token_f1": 0.11340206185567009, "prompt_id": "685a9c22a80a27a9", "reading_canonical_log_prob": -44.011024475097656, "reading_distances": [24, 21, 20, 16], "reading_log_probs": [-46.7094612121582, -44.565757751464844, -47.7338752746582, -51.06361770629883], "reading_logprob_std": 2.707241904730892, "tokenization_path_logprob_std": 42.28608524073632, "trajectory_balance_rmse": 403.4796587144569}520{"canonical_length": 12, "canonical_nll": 1.092390775680542, "generation_rouge_l_f1": 0.18181818181818182, "generation_token_f1": 0.24242424242424246, "prompt_id": "426d94bf331d4f8d", "reading_canonical_log_prob": -13.108689308166504, "reading_distances": [23, 26, 23, 31], "reading_log_probs": [-13.418874740600586, -11.714008331298828, -12.915202140808105, -9.194031715393066], "reading_logprob_std": 1.8852456116885365, "tokenization_path_logprob_std": 48.405934926332975, "trajectory_balance_rmse": 154.4719693866629}521{"canonical_length": 123, "canonical_nll": 1.2456552148834477, "generation_rouge_l_f1": 0.1763085399449036, "generation_token_f1": 0.28099173553719003, "prompt_id": "efba508a45047679", "reading_canonical_log_prob": -153.21559143066406, "reading_distances": [29, 34, 36, 30], "reading_log_probs": [-159.383056640625, -156.40768432617188, -159.44580078125, -160.56741333007812], "reading_logprob_std": 1.7807065473872388, "tokenization_path_logprob_std": 108.6294457274989, "trajectory_balance_rmse": 2529.978118733691}522{"canonical_length": 138, "canonical_nll": 1.228504346764606, "generation_rouge_l_f1": 0.233502538071066, "generation_token_f1": 0.4416243654822335, "prompt_id": "79eb49a82dc5be50", "reading_canonical_log_prob": -169.53359985351562, "reading_distances": [20, 28, 20, 15], "reading_log_probs": [-174.39503479003906, -170.68133544921875, -177.81802368164062, -172.98953247070312], "reading_logprob_std": 2.9868918112671143, "tokenization_path_logprob_std": 214.59913199023953, "trajectory_balance_rmse": 2908.452790464147}523{"canonical_length": 355, "canonical_nll": 1.1600440656635123, "generation_rouge_l_f1": 0.2469135802469136, "generation_token_f1": 0.4722222222222222, "prompt_id": "144d91c0bdbf405c", "reading_canonical_log_prob": -411.8156433105469, "reading_distances": [30, 28, 30, 35], "reading_log_probs": [-422.4185485839844, -419.675048828125, -419.50390625, -422.6962585449219], "reading_logprob_std": 1.718700206010013, "tokenization_path_logprob_std": 263.352467444773, "trajectory_balance_rmse": 6025.720822154606}524{"canonical_length": 194, "canonical_nll": 1.1434537749929525, "generation_rouge_l_f1": 0.24255319148936172, "generation_token_f1": 0.4127659574468085, "prompt_id": "6bebc48cdc1b16ab", "reading_canonical_log_prob": -221.8300323486328, "reading_distances": [58, 61, 53, 54], "reading_log_probs": [-221.71873474121094, -221.82916259765625, -224.1741485595703, -223.68038940429688], "reading_logprob_std": 1.2602624826694926, "tokenization_path_logprob_std": 157.7770332606063, "trajectory_balance_rmse": 3472.313153906613}525{"canonical_length": 136, "canonical_nll": 2.7720673504997704, "generation_rouge_l_f1": 0.234375, "generation_token_f1": 0.4453125, "prompt_id": "f19548b23e650d6d", "reading_canonical_log_prob": -377.00115966796875, "reading_distances": [20, 12, 22, 19], "reading_log_probs": [-406.369873046875, -382.933837890625, -401.7843017578125, -395.2781982421875], "reading_logprob_std": 10.179035077543768, "tokenization_path_logprob_std": 160.11788306884372, "trajectory_balance_rmse": 2871.5303302353204}526{"canonical_length": 283, "canonical_nll": 1.1575533054742713, "generation_rouge_l_f1": 0.19607843137254902, "generation_token_f1": 0.46345811051693403, "prompt_id": "b85ad6ca4cfd648f", "reading_canonical_log_prob": -327.58758544921875, "reading_distances": [25, 23, 24, 23], "reading_log_probs": [-319.50225830078125, -319.41796875, -315.3514099121094, -320.834228515625], "reading_logprob_std": 2.373723939161019, "tokenization_path_logprob_std": 335.63843466022416, "trajectory_balance_rmse": 5314.86373469354}527{"canonical_length": 146, "canonical_nll": 1.450117241846372, "generation_rouge_l_f1": 0.2430379746835443, "generation_token_f1": 0.4658227848101266, "prompt_id": "6b7b364a0fc8e714", "reading_canonical_log_prob": -211.7171173095703, "reading_distances": [33, 22, 34, 27], "reading_log_probs": [-214.25772094726562, -231.58840942382812, -211.44107055664062, -216.50238037109375], "reading_logprob_std": 9.002060743105593, "tokenization_path_logprob_std": 82.41095442887972, "trajectory_balance_rmse": 3180.067444810603}528{"canonical_length": 112, "canonical_nll": 2.6390348161969865, "generation_rouge_l_f1": 0.14395886889460155, "generation_token_f1": 0.2570694087403599, "prompt_id": "1d91ffc1f9553c07", "reading_canonical_log_prob": -295.5718994140625, "reading_distances": [12, 16, 19, 16], "reading_log_probs": [-299.0094909667969, -294.6256103515625, -297.2107849121094, -297.4510803222656], "reading_logprob_std": 1.8167403594710387, "tokenization_path_logprob_std": 116.58067101997142, "trajectory_balance_rmse": 2271.764962934459}529{"canonical_length": 37, "canonical_nll": 0.9509100011877112, "generation_rouge_l_f1": 0.19889502762430938, "generation_token_f1": 0.2541436464088398, "prompt_id": "1ac7a2e92a59b9ce", "reading_canonical_log_prob": -35.18367004394531, "reading_distances": [13, 20, 20, 13], "reading_log_probs": [-36.98466110229492, -39.26052474975586, -37.71028137207031, -38.35227584838867], "reading_logprob_std": 0.9668151200372656, "tokenization_path_logprob_std": 144.54899743817475, "trajectory_balance_rmse": 940.5744604709579}530{"canonical_length": 43, "canonical_nll": 1.9926750715388808, "generation_rouge_l_f1": 0.3209876543209877, "generation_token_f1": 0.4444444444444445, "prompt_id": "66c94013abca2b3a", "reading_canonical_log_prob": -85.68502807617188, "reading_distances": [36, 24, 28, 26], "reading_log_probs": [-89.74918365478516, -92.42298889160156, -89.93660736083984, -89.62535095214844], "reading_logprob_std": 1.332461642255679, "tokenization_path_logprob_std": 143.2954113502702, "trajectory_balance_rmse": 863.129100685976}531{"canonical_length": 15, "canonical_nll": 1.0494525909423829, "generation_rouge_l_f1": 0.4, "generation_token_f1": 0.4333333333333333, "prompt_id": "785090c8f7a0ea91", "reading_canonical_log_prob": -15.741788864135742, "reading_distances": [17, 16, 18, 13], "reading_log_probs": [-15.916172981262207, -14.743136405944824, -16.0114688873291, -15.577088356018066], "reading_logprob_std": 0.5768379382583568, "tokenization_path_logprob_std": 64.68056673636944, "trajectory_balance_rmse": 359.2104886526354}532{"canonical_length": 217, "canonical_nll": 1.1622438211045507, "generation_rouge_l_f1": 0.1735159817351598, "generation_token_f1": 0.3013698630136986, "prompt_id": "30ed32c8494397b3", "reading_canonical_log_prob": -252.2069091796875, "reading_distances": [23, 23, 22, 15], "reading_log_probs": [-261.4447937011719, -260.16778564453125, -261.8519592285156, -258.585205078125], "reading_logprob_std": 1.471576960998336, "tokenization_path_logprob_std": 90.78978642233416, "trajectory_balance_rmse": 3464.085287980777}533{"canonical_length": 16, "canonical_nll": 1.4244794845581055, "generation_rouge_l_f1": 0.26666666666666666, "generation_token_f1": 0.3555555555555555, "prompt_id": "e8d7fb5246616f5e", "reading_canonical_log_prob": -22.791671752929688, "reading_distances": [79, 78, 81, 78], "reading_log_probs": [-14.032383918762207, -16.28727149963379, -15.554600715637207, -15.671107292175293], "reading_logprob_std": 0.9581716471262423, "tokenization_path_logprob_std": 33.14207026133653, "trajectory_balance_rmse": 299.6448790115227}534{"canonical_length": 6, "canonical_nll": 1.647186279296875, "generation_rouge_l_f1": 0.3636363636363636, "generation_token_f1": 0.3636363636363636, "prompt_id": "96f83a49dbeabc97", "reading_canonical_log_prob": -9.88311767578125, "reading_distances": [31, 28, 32, 30], "reading_log_probs": [-10.677366256713867, -9.460308074951172, -8.702747344970703, -9.328128814697266], "reading_logprob_std": 0.8257957646252464, "tokenization_path_logprob_std": 7.810740335019535, "trajectory_balance_rmse": 108.24962651059298}535{"canonical_length": 80, "canonical_nll": 1.2439531326293944, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "dea31d77a681f384", "reading_canonical_log_prob": -99.51625061035156, "reading_distances": [13, 18, 15, 14], "reading_log_probs": [-101.32166290283203, -100.25662231445312, -100.60781860351562, -100.5847396850586], "reading_logprob_std": 0.4489314578600352, "tokenization_path_logprob_std": 88.34531390888607, "trajectory_balance_rmse": 1459.62154762257}536{"canonical_length": 9, "canonical_nll": 1.4242497550116644, "generation_rouge_l_f1": 0.14285714285714285, "generation_token_f1": 0.14285714285714285, "prompt_id": "428bb97daf254741", "reading_canonical_log_prob": -12.81824779510498, "reading_distances": [21, 22, 25, 22], "reading_log_probs": [-13.787065505981445, -14.75721549987793, -14.205289840698242, -14.871932983398438], "reading_logprob_std": 0.5045819360585526, "tokenization_path_logprob_std": 76.06536309222064, "trajectory_balance_rmse": 273.37556075495456}537{"canonical_length": 249, "canonical_nll": 1.3681227595930598, "generation_rouge_l_f1": 0.2345679012345679, "generation_token_f1": 0.46502057613168724, "prompt_id": "ed06fac728bea8f5", "reading_canonical_log_prob": -340.6625671386719, "reading_distances": [21, 18, 19, 17], "reading_log_probs": [-343.103271484375, -346.4327392578125, -346.82415771484375, -344.2516784667969], "reading_logprob_std": 1.7742855993321267, "tokenization_path_logprob_std": 142.68288539285555, "trajectory_balance_rmse": 5546.593877110266}538{"canonical_length": 42, "canonical_nll": 2.054227556501116, "generation_rouge_l_f1": 0.2222222222222222, "generation_token_f1": 0.3333333333333333, "prompt_id": "2455341860ece852", "reading_canonical_log_prob": -86.27755737304688, "reading_distances": [18, 16, 24, 16], "reading_log_probs": [-90.54824829101562, -87.37312316894531, -88.19829559326172, -86.79998779296875], "reading_logprob_std": 1.6486816040175267, "tokenization_path_logprob_std": 104.3432964259588, "trajectory_balance_rmse": 893.7503231275508}539{"canonical_length": 56, "canonical_nll": 1.6726012911115373, "generation_rouge_l_f1": 0.391304347826087, "generation_token_f1": 0.4565217391304348, "prompt_id": "2b713b1aba655ea9", "reading_canonical_log_prob": -93.6656723022461, "reading_distances": [19, 18, 10, 18], "reading_log_probs": [-98.0113525390625, -102.074951171875, -97.43927001953125, -99.44462585449219], "reading_logprob_std": 2.068090659294695, "tokenization_path_logprob_std": 149.40580933201332, "trajectory_balance_rmse": 1176.212356774365}540{"canonical_length": 21, "canonical_nll": 2.4101504371279763, "generation_rouge_l_f1": 0.2222222222222222, "generation_token_f1": 0.2222222222222222, "prompt_id": "12ac8fa81a314c1b", "reading_canonical_log_prob": -50.6131591796875, "reading_distances": [27, 26, 32, 32], "reading_log_probs": [-52.8082275390625, -57.928428649902344, -57.751068115234375, -59.812618255615234], "reading_logprob_std": 2.993621523309537, "tokenization_path_logprob_std": 28.634942187741743, "trajectory_balance_rmse": 437.5005760898367}541{"canonical_length": 174, "canonical_nll": 1.2590439018161816, "generation_rouge_l_f1": 0.23580786026200876, "generation_token_f1": 0.4148471615720524, "prompt_id": "46239be323c13eb4", "reading_canonical_log_prob": -219.07363891601562, "reading_distances": [21, 15, 22, 21], "reading_log_probs": [-261.98236083984375, -257.9957275390625, -262.96820068359375, -258.721435546875], "reading_logprob_std": 2.4287499262460766, "tokenization_path_logprob_std": 202.0347138924754, "trajectory_balance_rmse": 3151.5597901484775}542{"canonical_length": 298, "canonical_nll": 1.2630390960898175, "generation_rouge_l_f1": 0.16727272727272727, "generation_token_f1": 0.35272727272727267, "prompt_id": "3fde994857b8ea58", "reading_canonical_log_prob": -376.3856506347656, "reading_distances": [39, 36, 35, 35], "reading_log_probs": [-386.9288330078125, -380.5010070800781, -387.72991943359375, -384.16986083984375], "reading_logprob_std": 3.265505218169947, "tokenization_path_logprob_std": 166.1723928262031, "trajectory_balance_rmse": 6799.766251392165}543{"canonical_length": 111, "canonical_nll": 1.1386746243313626, "generation_rouge_l_f1": 0.2936288088642659, "generation_token_f1": 0.45983379501385047, "prompt_id": "366758412bdbb007", "reading_canonical_log_prob": -126.39288330078125, "reading_distances": [39, 36, 33, 39], "reading_log_probs": [-134.5443878173828, -133.3060302734375, -135.44989013671875, -134.8837890625], "reading_logprob_std": 0.9071285005554622, "tokenization_path_logprob_std": 101.32189503637686, "trajectory_balance_rmse": 2446.665006589736}544{"canonical_length": 17, "canonical_nll": 1.0540203767664291, "generation_rouge_l_f1": 0.17307692307692307, "generation_token_f1": 0.1923076923076923, "prompt_id": "f77e57c7031a2b6c", "reading_canonical_log_prob": -17.918346405029297, "reading_distances": [10, 15, 16, 13], "reading_log_probs": [-18.29625129699707, -17.276473999023438, -17.038604736328125, -16.6370849609375], "reading_logprob_std": 0.7071699346649042, "tokenization_path_logprob_std": 24.705974131618614, "trajectory_balance_rmse": 518.0319093656318}545{"canonical_length": 8, "canonical_nll": 1.7676820755004883, "generation_rouge_l_f1": 0.16, "generation_token_f1": 0.16, "prompt_id": "1f37960a14b0dc02", "reading_canonical_log_prob": -14.141456604003906, "reading_distances": [24, 23, 21, 30], "reading_log_probs": [-31.69003677368164, -33.10376739501953, -26.05426025390625, -30.542505264282227], "reading_logprob_std": 3.0479145564870187, "tokenization_path_logprob_std": 52.81061930983032, "trajectory_balance_rmse": 179.36681623052633}546{"canonical_length": 28, "canonical_nll": 2.6310653686523438, "generation_rouge_l_f1": 0.13953488372093023, "generation_token_f1": 0.1627906976744186, "prompt_id": "8d8a8351f2dceabb", "reading_canonical_log_prob": -73.66983032226562, "reading_distances": [20, 20, 21, 19], "reading_log_probs": [-75.75115966796875, -73.89763641357422, -77.36091613769531, -71.73916625976562], "reading_logprob_std": 2.421781833627642, "tokenization_path_logprob_std": 93.7080162529253, "trajectory_balance_rmse": 673.2989406866047}547{"canonical_length": 2, "canonical_nll": 0.685929000377655, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "7a9b64f9112cd0c5", "reading_canonical_log_prob": -1.37185800075531, "reading_distances": [45, 42, 52, 41], "reading_log_probs": [-1.63774573802948, -1.3755491971969604, -1.9379358291625977, -1.7785377502441406], "reading_logprob_std": 0.23853192532763642, "tokenization_path_logprob_std": 0.0, "trajectory_balance_rmse": 1.3932312726974487}548{"canonical_length": 5, "canonical_nll": 1.1178220748901366, "generation_rouge_l_f1": 0.8, "generation_token_f1": 0.8, "prompt_id": "2797e2b7991dcf3e", "reading_canonical_log_prob": -5.589110374450684, "reading_distances": [28, 20, 20, 23], "reading_log_probs": [-8.475947380065918, -7.8756303787231445, -8.436339378356934, -6.221312046051025], "reading_logprob_std": 1.056836458344613, "tokenization_path_logprob_std": 48.50873629575173, "trajectory_balance_rmse": 75.7304900967065}549{"canonical_length": 313, "canonical_nll": 1.695204859724441, "generation_rouge_l_f1": 0.2041522491349481, "generation_token_f1": 0.5224913494809689, "prompt_id": "d353ae8508e0bb05", "reading_canonical_log_prob": -530.59912109375, "reading_distances": [42, 41, 40, 45], "reading_log_probs": [-549.2760009765625, -543.599853515625, -545.4051513671875, -546.0079345703125], "reading_logprob_std": 2.3682568663670525, "tokenization_path_logprob_std": 308.0601745096098, "trajectory_balance_rmse": 6806.708940973058}550{"canonical_length": 7, "canonical_nll": 0.02801424264907837, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "1099a8369d87bee2", "reading_canonical_log_prob": -0.19609969854354858, "reading_distances": [20, 19, 25, 19], "reading_log_probs": [-1.8269903659820557, -2.8332037925720215, -2.2390167713165283, -3.173032522201538], "reading_logprob_std": 0.6010399258420177, "tokenization_path_logprob_std": 26.622939037749777, "trajectory_balance_rmse": 163.7939222958913}551{"canonical_length": 264, "canonical_nll": 1.34537529222893, "generation_rouge_l_f1": 0.27099236641221375, "generation_token_f1": 0.5267175572519084, "prompt_id": "d8d494771b6c2d81", "reading_canonical_log_prob": -355.1790771484375, "reading_distances": [25, 25, 31, 25], "reading_log_probs": [-365.9305114746094, -365.30963134765625, -362.9358825683594, -361.5097351074219], "reading_logprob_std": 2.0616407177895986, "tokenization_path_logprob_std": 351.93133452139284, "trajectory_balance_rmse": 6311.361500140354}552{"canonical_length": 372, "canonical_nll": 1.217383764123404, "generation_rouge_l_f1": 0.24960998439937598, "generation_token_f1": 0.4149765990639626, "prompt_id": "27da9cdd8cf02e41", "reading_canonical_log_prob": -452.86676025390625, "reading_distances": [33, 34, 34, 31], "reading_log_probs": [-482.0646667480469, -475.8299255371094, -478.80438232421875, -477.40411376953125], "reading_logprob_std": 2.653746099885542, "tokenization_path_logprob_std": 296.1545633012609, "trajectory_balance_rmse": 7110.610670987103}553{"canonical_length": 51, "canonical_nll": 1.296866024241728, "generation_rouge_l_f1": 0.23076923076923075, "generation_token_f1": 0.2769230769230769, "prompt_id": "47574188e8957081", "reading_canonical_log_prob": -66.14016723632812, "reading_distances": [33, 28, 28, 33], "reading_log_probs": [-64.57081604003906, -61.798431396484375, -58.40520477294922, -60.54924774169922], "reading_logprob_std": 2.574640063206111, "tokenization_path_logprob_std": 110.18215869493612, "trajectory_balance_rmse": 993.1117875606583}554{"canonical_length": 238, "canonical_nll": 2.0841933819426206, "generation_rouge_l_f1": 0.2015810276679842, "generation_token_f1": 0.32411067193675885, "prompt_id": "b7c16abbc7061a65", "reading_canonical_log_prob": -496.03802490234375, "reading_distances": [27, 30, 34, 27], "reading_log_probs": [-508.4760437011719, -519.716064453125, -510.53436279296875, -517.367919921875], "reading_logprob_std": 5.37087157624408, "tokenization_path_logprob_std": 130.1524289407179, "trajectory_balance_rmse": 5467.853713529507}555{"canonical_length": 334, "canonical_nll": 1.3116197414740831, "generation_rouge_l_f1": 0.22857142857142856, "generation_token_f1": 0.48739495798319327, "prompt_id": "01fef5f51abd900f", "reading_canonical_log_prob": -438.08099365234375, "reading_distances": [20, 31, 34, 33], "reading_log_probs": [-443.8563232421875, -448.1868896484375, -443.95501708984375, -442.7982177734375], "reading_logprob_std": 2.3834122535512274, "tokenization_path_logprob_std": 379.124560612646, "trajectory_balance_rmse": 7244.182104375077}556{"canonical_length": 60, "canonical_nll": 0.9624937057495118, "generation_rouge_l_f1": 0.2830188679245283, "generation_token_f1": 0.3113207547169811, "prompt_id": "1d0839cb7b1b8564", "reading_canonical_log_prob": -57.7496223449707, "reading_distances": [29, 34, 29, 30], "reading_log_probs": [-65.09310913085938, -60.24584197998047, -63.540924072265625, -58.986812591552734], "reading_logprob_std": 2.833942881814482, "tokenization_path_logprob_std": 112.33907819472684, "trajectory_balance_rmse": 1027.6469429086692}557{"canonical_length": 198, "canonical_nll": 1.4735579057173296, "generation_rouge_l_f1": 0.35159817351598166, "generation_token_f1": 0.5251141552511416, "prompt_id": "b900859423a57fd5", "reading_canonical_log_prob": -291.76446533203125, "reading_distances": [19, 22, 22, 25], "reading_log_probs": [-295.53936767578125, -303.78790283203125, -297.9642333984375, -297.41094970703125], "reading_logprob_std": 3.562629003160223, "tokenization_path_logprob_std": 274.76037252288666, "trajectory_balance_rmse": 4339.966596763634}558{"canonical_length": 48, "canonical_nll": 1.4566109975179036, "generation_rouge_l_f1": 0.1647058823529412, "generation_token_f1": 0.24313725490196078, "prompt_id": "5fa2bbf4741331df", "reading_canonical_log_prob": -69.91732788085938, "reading_distances": [18, 16, 18, 16], "reading_log_probs": [-68.92947387695312, -70.10079193115234, -69.83271789550781, -71.97080993652344], "reading_logprob_std": 1.277296277730286, "tokenization_path_logprob_std": 44.33176425898593, "trajectory_balance_rmse": 1153.1647827241695}559{"canonical_length": 501, "canonical_nll": 0.5966182868637725, "generation_rouge_l_f1": 0.2777777777777778, "generation_token_f1": 0.48055555555555557, "prompt_id": "97c9524265f7d241", "reading_canonical_log_prob": -298.90576171875, "reading_distances": [22, 18, 21, 19], "reading_log_probs": [-311.7008056640625, -311.4873046875, -309.6994323730469, -305.7588806152344], "reading_logprob_std": 2.7522237815291355, "tokenization_path_logprob_std": 181.0017807841185, "trajectory_balance_rmse": 6690.257687402177}560{"canonical_length": 11, "canonical_nll": 1.3115141608498313, "generation_rouge_l_f1": 0.588235294117647, "generation_token_f1": 0.588235294117647, "prompt_id": "990e2b768a18fc67", "reading_canonical_log_prob": -14.426655769348145, "reading_distances": [25, 26, 23, 31], "reading_log_probs": [-16.812057495117188, -16.554012298583984, -15.344621658325195, -18.30451774597168], "reading_logprob_std": 1.2157053079961597, "tokenization_path_logprob_std": 51.423165908770244, "trajectory_balance_rmse": 212.18783430654264}561{"canonical_length": 30, "canonical_nll": 1.3624001820882161, "generation_rouge_l_f1": 0.5116279069767442, "generation_token_f1": 0.6976744186046512, "prompt_id": "4a59564b57b1ad24", "reading_canonical_log_prob": -40.872005462646484, "reading_distances": [31, 30, 32, 32], "reading_log_probs": [-38.857147216796875, -36.89848327636719, -39.348876953125, -37.12933349609375], "reading_logprob_std": 1.226363084193515, "tokenization_path_logprob_std": 27.85655004146212, "trajectory_balance_rmse": 578.712546049616}562{"canonical_length": 118, "canonical_nll": 1.6313598115565413, "generation_rouge_l_f1": 0.18384401114206125, "generation_token_f1": 0.30083565459610023, "prompt_id": "fc01869b83f5b227", "reading_canonical_log_prob": -192.50045776367188, "reading_distances": [26, 26, 28, 33], "reading_log_probs": [-214.60606384277344, -209.979248046875, -212.58847045898438, -211.70669555664062], "reading_logprob_std": 1.9247096621605484, "tokenization_path_logprob_std": 218.87370445892705, "trajectory_balance_rmse": 2767.87233902422}563{"canonical_length": 10, "canonical_nll": 0.37330658435821534, "generation_rouge_l_f1": 0.3636363636363636, "generation_token_f1": 0.3636363636363636, "prompt_id": "e35706a135002934", "reading_canonical_log_prob": -3.7330658435821533, "reading_distances": [32, 30, 33, 36], "reading_log_probs": [-2.4643397331237793, -2.428572416305542, -2.494415760040283, -2.9529876708984375], "reading_logprob_std": 0.24674473618682768, "tokenization_path_logprob_std": 9.566440570794063, "trajectory_balance_rmse": 196.52276370455368}564{"canonical_length": 123, "canonical_nll": 1.0658604071392277, "generation_rouge_l_f1": 0.212987012987013, "generation_token_f1": 0.37922077922077924, "prompt_id": "5df9638e97e9c3b6", "reading_canonical_log_prob": -131.100830078125, "reading_distances": [11, 11, 10, 14], "reading_log_probs": [-132.65533447265625, -136.5913543701172, -136.27333068847656, -133.28192138671875], "reading_logprob_std": 2.0202470095697413, "tokenization_path_logprob_std": 125.15126770290618, "trajectory_balance_rmse": 2301.1743062632654}565{"canonical_length": 25, "canonical_nll": 0.5485408782958985, "generation_rouge_l_f1": 0.13076923076923075, "generation_token_f1": 0.14615384615384616, "prompt_id": "6f6b5ba97837e4b7", "reading_canonical_log_prob": -13.713521957397461, "reading_distances": [27, 27, 33, 26], "reading_log_probs": [-12.58556842803955, -10.641364097595215, -11.515645980834961, -12.507814407348633], "reading_logprob_std": 0.9202859141771672, "tokenization_path_logprob_std": 52.21502359791924, "trajectory_balance_rmse": 454.75213666714336}566{"canonical_length": 211, "canonical_nll": 0.753381955115151, "generation_rouge_l_f1": 0.47357293868921774, "generation_token_f1": 0.6046511627906976, "prompt_id": "111777e66227bbb0", "reading_canonical_log_prob": -158.96359252929688, "reading_distances": [40, 43, 44, 36], "reading_log_probs": [-172.14968872070312, -174.19058227539062, -176.2589111328125, -168.58041381835938], "reading_logprob_std": 3.2723841213125775, "tokenization_path_logprob_std": 73.39375560446132, "trajectory_balance_rmse": 3649.6161749835383}567{"canonical_length": 264, "canonical_nll": 1.1135794899680398, "generation_rouge_l_f1": 0.23300970873786409, "generation_token_f1": 0.4893203883495146, "prompt_id": "7eae8a9735d3d72c", "reading_canonical_log_prob": -293.9849853515625, "reading_distances": [31, 30, 35, 31], "reading_log_probs": [-314.0224609375, -309.6321105957031, -316.0758361816406, -311.64599609375], "reading_logprob_std": 2.8038653171550094, "tokenization_path_logprob_std": 82.18702686630522, "trajectory_balance_rmse": 5989.641989013794}568{"canonical_length": 6, "canonical_nll": 0.05202531814575195, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "88814d5eacafd109", "reading_canonical_log_prob": -0.3121519088745117, "reading_distances": [38, 42, 36, 41], "reading_log_probs": [-2.171250581741333, -1.8860113620758057, -0.7582056522369385, -1.1791558265686035], "reading_logprob_std": 0.646213258463418, "tokenization_path_logprob_std": 12.97758685685304, "trajectory_balance_rmse": 101.56385543831252}569{"canonical_length": 150, "canonical_nll": 1.3457236735026041, "generation_rouge_l_f1": 0.17682926829268295, "generation_token_f1": 0.3170731707317073, "prompt_id": "c6b7c5355ef042f7", "reading_canonical_log_prob": -201.85855102539062, "reading_distances": [40, 40, 38, 41], "reading_log_probs": [-210.417724609375, -204.32296752929688, -205.7998046875, -205.2477569580078], "reading_logprob_std": 2.7163218691387634, "tokenization_path_logprob_std": 112.58292779331342, "trajectory_balance_rmse": 3110.5696792921017}570{"canonical_length": 21, "canonical_nll": 1.590982891264416, "generation_rouge_l_f1": 0.1111111111111111, "generation_token_f1": 0.16666666666666669, "prompt_id": "62bbc871380d57bb", "reading_canonical_log_prob": -33.410640716552734, "reading_distances": [13, 17, 17, 13], "reading_log_probs": [-33.513607025146484, -29.341264724731445, -32.95329666137695, -28.72622299194336], "reading_logprob_std": 2.4483776381545885, "tokenization_path_logprob_std": 31.27454412136362, "trajectory_balance_rmse": 379.0862711063914}571{"canonical_length": 28, "canonical_nll": 1.7369469233921595, "generation_rouge_l_f1": 0.25, "generation_token_f1": 0.2857142857142857, "prompt_id": "085c601e3edc073f", "reading_canonical_log_prob": -48.63451385498047, "reading_distances": [15, 11, 13, 11], "reading_log_probs": [-53.88550567626953, -55.394561767578125, -57.45441436767578, -54.85753631591797], "reading_logprob_std": 1.5064961495770939, "tokenization_path_logprob_std": 50.066245180111345, "trajectory_balance_rmse": 555.798420997865}572{"canonical_length": 314, "canonical_nll": 1.420926622524383, "generation_rouge_l_f1": 0.27303754266211605, "generation_token_f1": 0.5085324232081911, "prompt_id": "575d828d632e226f", "reading_canonical_log_prob": -446.17095947265625, "reading_distances": [37, 32, 28, 34], "reading_log_probs": [-458.3624572753906, -457.24505615234375, -456.80133056640625, -460.24127197265625], "reading_logprob_std": 1.53359721607652, "tokenization_path_logprob_std": 309.67538097187037, "trajectory_balance_rmse": 7230.3333542171695}573{"canonical_length": 382, "canonical_nll": 1.0185204950302684, "generation_rouge_l_f1": 0.011799410029498525, "generation_token_f1": 0.011799410029498525, "prompt_id": "acf1157605893744", "reading_canonical_log_prob": -389.0748291015625, "reading_distances": [20, 17, 14, 14], "reading_log_probs": [-397.32879638671875, -400.36541748046875, -396.615966796875, -397.6417236328125], "reading_logprob_std": 1.6420562019270923, "tokenization_path_logprob_std": 237.80600833245077, "trajectory_balance_rmse": 7814.719599765904}574{"canonical_length": 19, "canonical_nll": 1.7900053325452303, "generation_rouge_l_f1": 0.3225806451612903, "generation_token_f1": 0.3225806451612903, "prompt_id": "c0621d0076f956e9", "reading_canonical_log_prob": -34.010101318359375, "reading_distances": [34, 34, 32, 34], "reading_log_probs": [-40.129859924316406, -42.783653259277344, -44.28599548339844, -41.884193420410156], "reading_logprob_std": 1.7375385169724198, "tokenization_path_logprob_std": 34.719084046350986, "trajectory_balance_rmse": 316.7469105996817}575{"canonical_length": 252, "canonical_nll": 1.1630188472687253, "generation_rouge_l_f1": 0.34285714285714286, "generation_token_f1": 0.5676190476190476, "prompt_id": "71ec6348f5f044b8", "reading_canonical_log_prob": -293.08074951171875, "reading_distances": [19, 20, 17, 19], "reading_log_probs": [-306.96282958984375, -298.4658203125, -297.72674560546875, -310.9787292480469], "reading_logprob_std": 6.4959349853419255, "tokenization_path_logprob_std": 139.2292556403041, "trajectory_balance_rmse": 5645.0941926251035}576{"canonical_length": 13, "canonical_nll": 0.700949082007775, "generation_rouge_l_f1": 0.1138211382113821, "generation_token_f1": 0.13008130081300812, "prompt_id": "73a9260fbda0ad11", "reading_canonical_log_prob": -9.112338066101074, "reading_distances": [40, 30, 36, 38], "reading_log_probs": [-7.8369035720825195, -8.169059753417969, -9.171812057495117, -9.691399574279785], "reading_logprob_std": 0.86238360163868, "tokenization_path_logprob_std": 39.136513054337, "trajectory_balance_rmse": 156.47381679025034}577{"canonical_length": 142, "canonical_nll": 1.005886453977773, "generation_rouge_l_f1": 0.3163538873994638, "generation_token_f1": 0.4772117962466488, "prompt_id": "0e246b28f0fca81c", "reading_canonical_log_prob": -142.83587646484375, "reading_distances": [10, 7, 10, 8], "reading_log_probs": [-144.2518310546875, -145.09072875976562, -149.16770935058594, -146.40652465820312], "reading_logprob_std": 2.1503787223498847, "tokenization_path_logprob_std": 103.69451177555607, "trajectory_balance_rmse": 2905.5198729359354}578{"canonical_length": 38, "canonical_nll": 1.348841014661287, "generation_rouge_l_f1": 0.05957446808510638, "generation_token_f1": 0.07659574468085106, "prompt_id": "7dd4910bb043bf96", "reading_canonical_log_prob": -51.255958557128906, "reading_distances": [14, 9, 11, 12], "reading_log_probs": [-53.98908996582031, -52.56848907470703, -56.69135665893555, -51.6143798828125], "reading_logprob_std": 2.210649504921639, "tokenization_path_logprob_std": 83.9876247318649, "trajectory_balance_rmse": 584.839155725075}579{"canonical_length": 187, "canonical_nll": 1.29362495952749, "generation_rouge_l_f1": 0.20785219399538107, "generation_token_f1": 0.4387990762124711, "prompt_id": "b7fe1fa68d367af3", "reading_canonical_log_prob": -241.90786743164062, "reading_distances": [17, 13, 15, 15], "reading_log_probs": [-247.52731323242188, -246.736083984375, -249.19284057617188, -246.10716247558594], "reading_logprob_std": 1.334461744920525, "tokenization_path_logprob_std": 247.07228000103748, "trajectory_balance_rmse": 4004.9401310528497}580{"canonical_length": 96, "canonical_nll": 1.4730661710103352, "generation_rouge_l_f1": 0.2184873949579832, "generation_token_f1": 0.3865546218487395, "prompt_id": "b435d26419c9a382", "reading_canonical_log_prob": -141.4143524169922, "reading_distances": [41, 46, 48, 52], "reading_log_probs": [-137.82916259765625, -141.91171264648438, -139.07879638671875, -141.48863220214844], "reading_logprob_std": 1.9500478102837002, "tokenization_path_logprob_std": 176.38536135234614, "trajectory_balance_rmse": 2248.397473180315}581{"canonical_length": 113, "canonical_nll": 1.3907860950031112, "generation_rouge_l_f1": 0.17692307692307696, "generation_token_f1": 0.36923076923076925, "prompt_id": "fc2765baa4652639", "reading_canonical_log_prob": -157.15882873535156, "reading_distances": [19, 26, 24, 21], "reading_log_probs": [-170.63412475585938, -167.59814453125, -168.56105041503906, -168.54672241210938], "reading_logprob_std": 1.2812511697633349, "tokenization_path_logprob_std": 216.89556344052474, "trajectory_balance_rmse": 2208.2971696952554}582{"canonical_length": 216, "canonical_nll": 1.4324326691804108, "generation_rouge_l_f1": 0.25053078556263264, "generation_token_f1": 0.47983014861995754, "prompt_id": "6389a5bbf9eaba60", "reading_canonical_log_prob": -309.40545654296875, "reading_distances": [34, 29, 31, 26], "reading_log_probs": [-310.48297119140625, -308.1460876464844, -308.6387939453125, -311.1037902832031], "reading_logprob_std": 1.4234465828944958, "tokenization_path_logprob_std": 140.42611594463867, "trajectory_balance_rmse": 4552.051280140208}583{"canonical_length": 13, "canonical_nll": 0.1549972937657283, "generation_rouge_l_f1": 0.888888888888889, "generation_token_f1": 0.888888888888889, "prompt_id": "286cf1906a688fcb", "reading_canonical_log_prob": -2.0149648189544678, "reading_distances": [25, 25, 24, 24], "reading_log_probs": [-3.929429531097412, -4.444279670715332, -3.98948073387146, -3.957287073135376], "reading_logprob_std": 0.24401044217942425, "tokenization_path_logprob_std": 18.96035343211193, "trajectory_balance_rmse": 244.87809908248383}584{"canonical_length": 93, "canonical_nll": 1.352346686906712, "generation_rouge_l_f1": 0.2515723270440251, "generation_token_f1": 0.4150943396226416, "prompt_id": "33333486aa163f12", "reading_canonical_log_prob": -125.76824188232422, "reading_distances": [34, 26, 32, 28], "reading_log_probs": [-129.20046997070312, -134.7128143310547, -131.842041015625, -137.3461151123047], "reading_logprob_std": 3.525925608973973, "tokenization_path_logprob_std": 168.21785608483205, "trajectory_balance_rmse": 2325.376836748184}585{"canonical_length": 60, "canonical_nll": 2.1996388753255207, "generation_rouge_l_f1": 0.21359223300970875, "generation_token_f1": 0.2459546925566343, "prompt_id": "c5556f6cb2c9f8dd", "reading_canonical_log_prob": -131.97833251953125, "reading_distances": [25, 23, 27, 23], "reading_log_probs": [-130.8897247314453, -136.62765502929688, -126.45147705078125, -133.4180908203125], "reading_logprob_std": 4.295386378944843, "tokenization_path_logprob_std": 35.63973890072076, "trajectory_balance_rmse": 375.7148643914324}586{"canonical_length": 11, "canonical_nll": 3.220491582697088, "generation_rouge_l_f1": 0.18181818181818185, "generation_token_f1": 0.18181818181818185, "prompt_id": "b1950a5b21116b51", "reading_canonical_log_prob": -35.42540740966797, "reading_distances": [17, 18, 12, 19], "reading_log_probs": [-38.40334701538086, -35.724571228027344, -37.80607604980469, -37.64240264892578], "reading_logprob_std": 1.1600695854952832, "tokenization_path_logprob_std": 37.7846230566521, "trajectory_balance_rmse": 239.3837807975457}587{"canonical_length": 263, "canonical_nll": 1.476326481924311, "generation_rouge_l_f1": 0.1954022988505747, "generation_token_f1": 0.4750957854406131, "prompt_id": "355011c0ec753a64", "reading_canonical_log_prob": -388.27386474609375, "reading_distances": [18, 21, 20, 18], "reading_log_probs": [-399.738525390625, -404.5162048339844, -403.5853271484375, -404.704833984375], "reading_logprob_std": 2.3173932095729923, "tokenization_path_logprob_std": 303.9758019271018, "trajectory_balance_rmse": 6253.37004662441}588{"canonical_length": 29, "canonical_nll": 0.984411963101091, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "c2e4ff98064c8812", "reading_canonical_log_prob": -28.54794692993164, "reading_distances": [25, 22, 22, 29], "reading_log_probs": [-29.183555603027344, -26.857086181640625, -26.221006393432617, -27.179162979125977], "reading_logprob_std": 1.279099376573895, "tokenization_path_logprob_std": 100.97230815858657, "trajectory_balance_rmse": 606.7444976982719}589{"canonical_length": 57, "canonical_nll": 1.9742938928436815, "generation_rouge_l_f1": 0.13953488372093026, "generation_token_f1": 0.2059800664451827, "prompt_id": "e95bb93274990ecf", "reading_canonical_log_prob": -112.53475189208984, "reading_distances": [28, 30, 27, 28], "reading_log_probs": [-114.28002166748047, -113.38314056396484, -112.66490173339844, -113.70406341552734], "reading_logprob_std": 0.6735138899395984, "tokenization_path_logprob_std": 140.52426078231653, "trajectory_balance_rmse": 1032.6566477270933}590{"canonical_length": 40, "canonical_nll": 1.9163951873779297, "generation_rouge_l_f1": 0.3376623376623376, "generation_token_f1": 0.3896103896103896, "prompt_id": "0eb7b4a395752869", "reading_canonical_log_prob": -76.65580749511719, "reading_distances": [30, 29, 36, 29], "reading_log_probs": [-94.14556884765625, -80.55047607421875, -88.55122375488281, -93.2432632446289], "reading_logprob_std": 6.218731409439429, "tokenization_path_logprob_std": 95.14638808941602, "trajectory_balance_rmse": 704.2284490158112}591{"canonical_length": 124, "canonical_nll": 1.8465953949959046, "generation_rouge_l_f1": 0.3549488054607508, "generation_token_f1": 0.5324232081911262, "prompt_id": "5d63d37b660c138f", "reading_canonical_log_prob": -228.9778289794922, "reading_distances": [38, 28, 37, 39], "reading_log_probs": [-239.10226440429688, -238.79351806640625, -244.29986572265625, -241.45980834960938], "reading_logprob_std": 2.552177036176546, "tokenization_path_logprob_std": 74.06515497617646, "trajectory_balance_rmse": 2951.6488355218885}592{"canonical_length": 241, "canonical_nll": 1.213645460200013, "generation_rouge_l_f1": 0.23300970873786409, "generation_token_f1": 0.45825242718446596, "prompt_id": "17f90ddfcc762d15", "reading_canonical_log_prob": -292.4885559082031, "reading_distances": [23, 25, 31, 31], "reading_log_probs": [-292.45849609375, -289.7715148925781, -291.939697265625, -282.84442138671875], "reading_logprob_std": 4.428403918898782, "tokenization_path_logprob_std": 95.17158892754553, "trajectory_balance_rmse": 5266.270045326833}593{"canonical_length": 61, "canonical_nll": 1.4294213466956966, "generation_rouge_l_f1": 0.30303030303030304, "generation_token_f1": 0.40404040404040403, "prompt_id": "9e46c646f8e1ae11", "reading_canonical_log_prob": -87.1947021484375, "reading_distances": [21, 19, 23, 26], "reading_log_probs": [-88.69444274902344, -90.74790954589844, -91.7992172241211, -90.54729461669922], "reading_logprob_std": 1.2910659313695387, "tokenization_path_logprob_std": 43.97894735223644, "trajectory_balance_rmse": 1247.91563120905}594{"canonical_length": 317, "canonical_nll": 1.0879706253388703, "generation_rouge_l_f1": 0.2517241379310345, "generation_token_f1": 0.5206896551724137, "prompt_id": "c405755d9e3e8c9b", "reading_canonical_log_prob": -344.8866882324219, "reading_distances": [23, 26, 23, 21], "reading_log_probs": [-352.1580505371094, -352.23455810546875, -354.3860168457031, -351.616943359375], "reading_logprob_std": 1.2227181099313635, "tokenization_path_logprob_std": 214.86219375525167, "trajectory_balance_rmse": 6209.270551507829}595{"canonical_length": 225, "canonical_nll": 0.5766236029730902, "generation_rouge_l_f1": 0.24778761061946902, "generation_token_f1": 0.4557522123893806, "prompt_id": "d5c04061128f24af", "reading_canonical_log_prob": -129.7403106689453, "reading_distances": [27, 26, 23, 22], "reading_log_probs": [-129.3516387939453, -127.94256591796875, -130.3804931640625, -130.218994140625], "reading_logprob_std": 1.116095486762496, "tokenization_path_logprob_std": 209.7281939578012, "trajectory_balance_rmse": 3203.3437197978924}596{"canonical_length": 61, "canonical_nll": 0.9991213063724705, "generation_rouge_l_f1": 0.5137614678899083, "generation_token_f1": 0.5871559633027522, "prompt_id": "1419b413999c869f", "reading_canonical_log_prob": -60.9463996887207, "reading_distances": [17, 13, 12, 12], "reading_log_probs": [-60.0377197265625, -68.23643493652344, -61.2508544921875, -60.07994842529297], "reading_logprob_std": 3.930543194927314, "tokenization_path_logprob_std": 35.61081817747015, "trajectory_balance_rmse": 698.7397829683506}597{"canonical_length": 116, "canonical_nll": 1.4660482735469424, "generation_rouge_l_f1": 0.2065217391304348, "generation_token_f1": 0.40217391304347827, "prompt_id": "3501affcfd13e96c", "reading_canonical_log_prob": -170.0615997314453, "reading_distances": [24, 21, 25, 18], "reading_log_probs": [-189.83108520507812, -184.51272583007812, -187.94622802734375, -188.91720581054688], "reading_logprob_std": 2.323861507317578, "tokenization_path_logprob_std": 90.44839257785763, "trajectory_balance_rmse": 2415.159706519565}598{"canonical_length": 26, "canonical_nll": 1.9241975637582631, "generation_rouge_l_f1": 0.2033898305084746, "generation_token_f1": 0.3050847457627119, "prompt_id": "6fdd3868e470d9e0", "reading_canonical_log_prob": -50.029136657714844, "reading_distances": [57, 60, 47, 39], "reading_log_probs": [-63.83183288574219, -63.87525177001953, -66.25203704833984, -70.72050476074219], "reading_logprob_std": 3.2376286853781684, "tokenization_path_logprob_std": 61.91984692599404, "trajectory_balance_rmse": 605.1082469269325}599{"canonical_length": 304, "canonical_nll": 0.8293057993838662, "generation_rouge_l_f1": 0.368888888888889, "generation_token_f1": 0.5822222222222222, "prompt_id": "1d8b10c0b2c4c1d8", "reading_canonical_log_prob": -252.1089630126953, "reading_distances": [27, 19, 19, 23], "reading_log_probs": [-262.26495361328125, -259.15618896484375, -255.79949951171875, -259.1845397949219], "reading_logprob_std": 2.6407405723319375, "tokenization_path_logprob_std": 257.92269524342, "trajectory_balance_rmse": 4747.887515135564}600{"canonical_length": 31, "canonical_nll": 0.8372014568698022, "generation_rouge_l_f1": 0.07692307692307693, "generation_token_f1": 0.07692307692307693, "prompt_id": "6f83d657c17ec17d", "reading_canonical_log_prob": -25.953245162963867, "reading_distances": [16, 19, 23, 24], "reading_log_probs": [-22.601646423339844, -22.261470794677734, -22.379518508911133, -24.757064819335938], "reading_logprob_std": 1.1798848809475209, "tokenization_path_logprob_std": 79.89698117425739, "trajectory_balance_rmse": 502.9907149448188}601{"canonical_length": 39, "canonical_nll": 1.5331037472455928, "generation_rouge_l_f1": 0.09756097560975609, "generation_token_f1": 0.09756097560975609, "prompt_id": "6587efe9dc00b7ac", "reading_canonical_log_prob": -59.791046142578125, "reading_distances": [24, 22, 21, 23], "reading_log_probs": [-64.33503723144531, -71.9206314086914, -66.12963104248047, -64.37442779541016], "reading_logprob_std": 3.586141964076034, "tokenization_path_logprob_std": 121.4252212867948, "trajectory_balance_rmse": 946.7784232602986}602{"canonical_length": 15, "canonical_nll": 0.816888173421224, "generation_rouge_l_f1": 0.6153846153846153, "generation_token_f1": 0.6153846153846153, "prompt_id": "ea95653fa930225e", "reading_canonical_log_prob": -12.25332260131836, "reading_distances": [15, 23, 15, 16], "reading_log_probs": [-13.351081848144531, -14.162079811096191, -13.852507591247559, -14.507847785949707], "reading_logprob_std": 0.49092719497166915, "tokenization_path_logprob_std": 61.92765766420515, "trajectory_balance_rmse": 444.8580380063919}603{"canonical_length": 218, "canonical_nll": 1.4986852243405964, "generation_rouge_l_f1": 0.2911392405063291, "generation_token_f1": 0.48523206751054854, "prompt_id": "dbfaa5f0bb7fb12c", "reading_canonical_log_prob": -326.71337890625, "reading_distances": [38, 42, 40, 45], "reading_log_probs": [-330.7364807128906, -338.5401611328125, -333.2481689453125, -337.128173828125], "reading_logprob_std": 3.5720334323778573, "tokenization_path_logprob_std": 100.25803942513357, "trajectory_balance_rmse": 5151.713164478147}604{"canonical_length": 236, "canonical_nll": 1.2549481149447166, "generation_rouge_l_f1": 0.24497991967871488, "generation_token_f1": 0.5220883534136546, "prompt_id": "6ef66ee37f69a6e9", "reading_canonical_log_prob": -296.1677551269531, "reading_distances": [37, 45, 46, 48], "reading_log_probs": [-296.8299560546875, -298.6761474609375, -295.90960693359375, -295.3232421875], "reading_logprob_std": 1.4652984840504488, "tokenization_path_logprob_std": 240.08758838476533, "trajectory_balance_rmse": 5670.627341022409}605{"canonical_length": 254, "canonical_nll": 0.3875891317532757, "generation_rouge_l_f1": 0.2426614481409002, "generation_token_f1": 0.4540117416829746, "prompt_id": "e2bfc09d77b30452", "reading_canonical_log_prob": -98.44763946533203, "reading_distances": [21, 26, 24, 23], "reading_log_probs": [-98.7232666015625, -98.63441467285156, -101.86810302734375, -99.3814697265625], "reading_logprob_std": 1.5146313898031347, "tokenization_path_logprob_std": 73.9131636441769, "trajectory_balance_rmse": 3355.031948015234}606{"canonical_length": 315, "canonical_nll": 0.8550653366815476, "generation_rouge_l_f1": 0.17317487266553483, "generation_token_f1": 0.3870967741935484, "prompt_id": "3024b6f11a619ed8", "reading_canonical_log_prob": -269.3455810546875, "reading_distances": [16, 13, 15, 12], "reading_log_probs": [-272.31439208984375, -270.87042236328125, -271.85137939453125, -271.9912109375], "reading_logprob_std": 0.6219506927232318, "tokenization_path_logprob_std": 125.0317836600873, "trajectory_balance_rmse": 5850.5273200502625}607{"canonical_length": 147, "canonical_nll": 1.550362308009141, "generation_rouge_l_f1": 0.16826923076923075, "generation_token_f1": 0.3365384615384615, "prompt_id": "2027b774c4dd5ef1", "reading_canonical_log_prob": -227.90325927734375, "reading_distances": [79, 81, 68, 82], "reading_log_probs": [-236.03250122070312, -235.77403259277344, -232.49037170410156, -237.36517333984375], "reading_logprob_std": 2.070977048354918, "tokenization_path_logprob_std": 66.82888295636688, "trajectory_balance_rmse": 3203.248528881607}608{"canonical_length": 320, "canonical_nll": 1.0204129219055176, "generation_rouge_l_f1": 0.24542124542124544, "generation_token_f1": 0.45054945054945056, "prompt_id": "ea4a6ac8af4f585f", "reading_canonical_log_prob": -326.5321350097656, "reading_distances": [29, 33, 37, 34], "reading_log_probs": [-326.2271728515625, -332.7806091308594, -334.17315673828125, -334.599609375], "reading_logprob_std": 3.8903051820334213, "tokenization_path_logprob_std": 163.9351035262409, "trajectory_balance_rmse": 6745.5742439954065}609{"canonical_length": 142, "canonical_nll": 1.207552305409606, "generation_rouge_l_f1": 0.26815642458100564, "generation_token_f1": 0.4357541899441341, "prompt_id": "e9a730206e7c88d2", "reading_canonical_log_prob": -171.47242736816406, "reading_distances": [33, 22, 21, 22], "reading_log_probs": [-179.41937255859375, -178.35572814941406, -176.79258728027344, -179.626220703125], "reading_logprob_std": 1.296185607959937, "tokenization_path_logprob_std": 115.32962309394476, "trajectory_balance_rmse": 2728.72290633459}610{"canonical_length": 324, "canonical_nll": 1.049944559733073, "generation_rouge_l_f1": 0.1966604823747681, "generation_token_f1": 0.40074211502782936, "prompt_id": "21922a55d95c3a47", "reading_canonical_log_prob": -340.1820373535156, "reading_distances": [18, 19, 18, 19], "reading_log_probs": [-347.521728515625, -346.51019287109375, -345.6522216796875, -343.88031005859375], "reading_logprob_std": 1.543001877151461, "tokenization_path_logprob_std": 124.48435826096171, "trajectory_balance_rmse": 6813.137569826408}611{"canonical_length": 17, "canonical_nll": 0.5540488187004539, "generation_rouge_l_f1": 0.8666666666666666, "generation_token_f1": 0.8666666666666666, "prompt_id": "732a8e545760e4c1", "reading_canonical_log_prob": -9.418829917907715, "reading_distances": [24, 22, 16, 17], "reading_log_probs": [-15.945259094238281, -12.19761848449707, -11.151209831237793, -11.032313346862793], "reading_logprob_std": 2.30274831994151, "tokenization_path_logprob_std": 40.92405376646405, "trajectory_balance_rmse": 291.6607620270277}612{"canonical_length": 8, "canonical_nll": 0.588773250579834, "generation_rouge_l_f1": 0.875, "generation_token_f1": 0.875, "prompt_id": "530dfa24f61035c2", "reading_canonical_log_prob": -4.710186004638672, "reading_distances": [34, 32, 26, 35], "reading_log_probs": [-4.233458995819092, -5.0284271240234375, -3.920172929763794, -4.20265531539917], "reading_logprob_std": 0.47618222072188193, "tokenization_path_logprob_std": 29.963275428463163, "trajectory_balance_rmse": 111.21129389203045}613{"canonical_length": 8, "canonical_nll": 0.8330545425415039, "generation_rouge_l_f1": 0.7142857142857143, "generation_token_f1": 0.7142857142857143, "prompt_id": "ab024353313a6e65", "reading_canonical_log_prob": -6.664436340332031, "reading_distances": [27, 29, 36, 33], "reading_log_probs": [-6.113619327545166, -8.244169235229492, -7.174607276916504, -7.940580368041992], "reading_logprob_std": 0.9498056266335485, "tokenization_path_logprob_std": 10.559186594987409, "trajectory_balance_rmse": 134.9270029871085}614{"canonical_length": 12, "canonical_nll": 0.9634536107381185, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "2321d8df80c63fef", "reading_canonical_log_prob": -11.561443328857422, "reading_distances": [20, 20, 20, 16], "reading_log_probs": [-12.844955444335938, -13.509904861450195, -13.502254486083984, -13.360779762268066], "reading_logprob_std": 0.31392493867984744, "tokenization_path_logprob_std": 61.2602651048439, "trajectory_balance_rmse": 234.91778856198843}615{"canonical_length": 16, "canonical_nll": 0.5935214757919312, "generation_rouge_l_f1": 0.2777777777777778, "generation_token_f1": 0.3055555555555556, "prompt_id": "f3f8000f2b6cf3b2", "reading_canonical_log_prob": -9.496343612670898, "reading_distances": [5, 7, 4, 12], "reading_log_probs": [-9.275550842285156, -8.619678497314453, -9.478434562683105, -9.879039764404297], "reading_logprob_std": 0.5259488666939708, "tokenization_path_logprob_std": 11.64437552606575, "trajectory_balance_rmse": 167.83774411166536}616{"canonical_length": 8, "canonical_nll": 0.1274949461221695, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "30a8ec8b58b46182", "reading_canonical_log_prob": -1.019959568977356, "reading_distances": [28, 24, 26, 26], "reading_log_probs": [-1.208789348602295, -1.6718144416809082, -1.3481662273406982, -1.4263540506362915], "reading_logprob_std": 0.1941358752368096, "tokenization_path_logprob_std": 22.93064434570944, "trajectory_balance_rmse": 48.49226468778034}617{"canonical_length": 267, "canonical_nll": 1.455431305960323, "generation_rouge_l_f1": 0.22180451127819548, "generation_token_f1": 0.4699248120300752, "prompt_id": "f1f8d8a792fc708b", "reading_canonical_log_prob": -388.60015869140625, "reading_distances": [24, 28, 29, 30], "reading_log_probs": [-401.4442138671875, -423.9420166015625, -403.8633728027344, -399.5810852050781], "reading_logprob_std": 11.293138066304316, "tokenization_path_logprob_std": 246.562410582803, "trajectory_balance_rmse": 5739.609238240769}618{"canonical_length": 194, "canonical_nll": 0.9341663478576031, "generation_rouge_l_f1": 0.23194748358862144, "generation_token_f1": 0.4814004376367615, "prompt_id": "1d19c88736fcd9d4", "reading_canonical_log_prob": -181.228271484375, "reading_distances": [52, 48, 29, 47], "reading_log_probs": [-188.73194885253906, -186.9727783203125, -185.87252807617188, -186.3923797607422], "reading_logprob_std": 1.2437245988930206, "tokenization_path_logprob_std": 166.5930989043474, "trajectory_balance_rmse": 3585.704298714392}619{"canonical_length": 15, "canonical_nll": 2.3578478495279946, "generation_rouge_l_f1": 0.0606060606060606, "generation_token_f1": 0.0606060606060606, "prompt_id": "1dc2cb66ddb80668", "reading_canonical_log_prob": -35.36771774291992, "reading_distances": [15, 15, 15, 13], "reading_log_probs": [-39.56227111816406, -38.8720588684082, -39.8592529296875, -41.031497955322266], "reading_logprob_std": 0.9006990956084006, "tokenization_path_logprob_std": 36.91021326773666, "trajectory_balance_rmse": 312.7893400060905}620{"canonical_length": 16, "canonical_nll": 1.753328800201416, "generation_rouge_l_f1": 0.09137055837563451, "generation_token_f1": 0.11167512690355329, "prompt_id": "bf66528e05d6425b", "reading_canonical_log_prob": -28.053260803222656, "reading_distances": [29, 37, 31, 36], "reading_log_probs": [-29.022377014160156, -32.65779495239258, -29.83681869506836, -32.20791244506836], "reading_logprob_std": 1.775049335109368, "tokenization_path_logprob_std": 76.08449005655066, "trajectory_balance_rmse": 515.7878544059292}621{"canonical_length": 118, "canonical_nll": 1.2646114543332891, "generation_rouge_l_f1": 0.26332288401253917, "generation_token_f1": 0.42006269592476486, "prompt_id": "471e2be2592f5add", "reading_canonical_log_prob": -149.22415161132812, "reading_distances": [20, 22, 21, 20], "reading_log_probs": [-153.96902465820312, -145.57247924804688, -152.0489501953125, -154.20599365234375], "reading_logprob_std": 4.035053680431272, "tokenization_path_logprob_std": 168.35979004097015, "trajectory_balance_rmse": 2782.027276930456}622{"canonical_length": 19, "canonical_nll": 4.583344308953536, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "57103ee2e4cc89e4", "reading_canonical_log_prob": -87.08354187011719, "reading_distances": [36, 32, 40, 39], "reading_log_probs": [-91.74303436279297, -88.22164154052734, -83.63468170166016, -89.71975708007812], "reading_logprob_std": 3.4466400830253754, "tokenization_path_logprob_std": 16.620540264174732, "trajectory_balance_rmse": 251.52855139336566}623{"canonical_length": 23, "canonical_nll": 2.2500918844471807, "generation_rouge_l_f1": 0.12403100775193798, "generation_token_f1": 0.15503875968992248, "prompt_id": "aa0821e4d154fd43", "reading_canonical_log_prob": -51.752113342285156, "reading_distances": [20, 22, 25, 21], "reading_log_probs": [-54.74601745605469, -52.161460876464844, -51.954429626464844, -53.125213623046875], "reading_logprob_std": 1.2728680492684512, "tokenization_path_logprob_std": 24.29409776121416, "trajectory_balance_rmse": 511.23269122664954}624{"canonical_length": 24, "canonical_nll": 1.3427181243896484, "generation_rouge_l_f1": 0.3, "generation_token_f1": 0.39999999999999997, "prompt_id": "2aaa42d342162266", "reading_canonical_log_prob": -32.22523498535156, "reading_distances": [31, 32, 27, 28], "reading_log_probs": [-36.267093658447266, -38.113861083984375, -39.383792877197266, -36.83338165283203], "reading_logprob_std": 1.3905034395004643, "tokenization_path_logprob_std": 106.58289010868288, "trajectory_balance_rmse": 702.4337815463634}625{"canonical_length": 10, "canonical_nll": 0.37697687149047854, "generation_rouge_l_f1": 0.07207207207207207, "generation_token_f1": 0.07207207207207207, "prompt_id": "b5604eaaebf6e52c", "reading_canonical_log_prob": -3.769768714904785, "reading_distances": [19, 20, 12, 15], "reading_log_probs": [-6.37447452545166, -4.34709358215332, -5.025496482849121, -4.778307914733887], "reading_logprob_std": 0.8748829246602822, "tokenization_path_logprob_std": 54.763919768907314, "trajectory_balance_rmse": 233.92631239349947}626{"canonical_length": 223, "canonical_nll": 1.3949092300483463, "generation_rouge_l_f1": 0.20140515222482436, "generation_token_f1": 0.3700234192037471, "prompt_id": "5dbcfc1c200ae9f7", "reading_canonical_log_prob": -311.06475830078125, "reading_distances": [23, 24, 25, 24], "reading_log_probs": [-308.6935729980469, -318.706787109375, -317.89434814453125, -307.7208251953125], "reading_logprob_std": 5.850334348495951, "tokenization_path_logprob_std": 203.08911535172757, "trajectory_balance_rmse": 5101.997937733006}627{"canonical_length": 163, "canonical_nll": 1.376762530555023, "generation_rouge_l_f1": 0.1783132530120482, "generation_token_f1": 0.3710843373493976, "prompt_id": "d526cee0379fafdd", "reading_canonical_log_prob": -224.41229248046875, "reading_distances": [34, 29, 22, 30], "reading_log_probs": [-225.97593688964844, -223.3191375732422, -220.60452270507812, -222.7720947265625], "reading_logprob_std": 2.208733343319758, "tokenization_path_logprob_std": 214.7919720047841, "trajectory_balance_rmse": 4081.180530606745}628{"canonical_length": 85, "canonical_nll": 1.4661962172564338, "generation_rouge_l_f1": 0.25974025974025977, "generation_token_f1": 0.44155844155844154, "prompt_id": "b641694c0de50f89", "reading_canonical_log_prob": -124.62667846679688, "reading_distances": [35, 25, 31, 37], "reading_log_probs": [-148.50222778320312, -143.75704956054688, -140.98219299316406, -144.9931640625], "reading_logprob_std": 3.1184513351269416, "tokenization_path_logprob_std": 63.1447038282181, "trajectory_balance_rmse": 1845.1891884812223}629{"canonical_length": 13, "canonical_nll": 0.39441735927875227, "generation_rouge_l_f1": 0.2857142857142857, "generation_token_f1": 0.2857142857142857, "prompt_id": "2e5bfbd760499e1d", "reading_canonical_log_prob": -5.127425670623779, "reading_distances": [22, 24, 25, 19], "reading_log_probs": [-4.6389265060424805, -3.687152624130249, -3.424534797668457, -4.257866859436035], "reading_logprob_std": 0.5488588452610618, "tokenization_path_logprob_std": 17.13065941680172, "trajectory_balance_rmse": 359.8209991178775}630{"canonical_length": 363, "canonical_nll": 1.380949330395575, "generation_rouge_l_f1": 0.2360655737704918, "generation_token_f1": 0.4590163934426229, "prompt_id": "1d0a6d1714584f14", "reading_canonical_log_prob": -501.28460693359375, "reading_distances": [36, 39, 33, 28], "reading_log_probs": [-500.3889465332031, -498.9797058105469, -501.5426025390625, -506.86279296875], "reading_logprob_std": 3.4429093590824786, "tokenization_path_logprob_std": 101.25014796422813, "trajectory_balance_rmse": 7299.147006420668}631{"canonical_length": 37, "canonical_nll": 1.978715432656778, "generation_rouge_l_f1": 0.05405405405405406, "generation_token_f1": 0.05405405405405406, "prompt_id": "c7b4cbd5453e19a7", "reading_canonical_log_prob": -73.21247100830078, "reading_distances": [24, 21, 16, 21], "reading_log_probs": [-69.87948608398438, -74.912841796875, -72.61984252929688, -71.74185180664062], "reading_logprob_std": 2.089587681280892, "tokenization_path_logprob_std": 54.294605723842245, "trajectory_balance_rmse": 733.1152100697511}632{"canonical_length": 10, "canonical_nll": 0.6394563674926758, "generation_rouge_l_f1": 0.5833333333333334, "generation_token_f1": 0.6666666666666666, "prompt_id": "88beef2f88591814", "reading_canonical_log_prob": -6.394563674926758, "reading_distances": [62, 63, 57, 70], "reading_log_probs": [-9.429936408996582, -9.157651901245117, -9.25576400756836, -8.878621101379395], "reading_logprob_std": 0.23060507210440104, "tokenization_path_logprob_std": 33.14028549468664, "trajectory_balance_rmse": 268.5822891812413}633{"canonical_length": 65, "canonical_nll": 0.8661203824556791, "generation_rouge_l_f1": 0.7746478873239436, "generation_token_f1": 0.7887323943661972, "prompt_id": "cfc50376a0f13c13", "reading_canonical_log_prob": -56.29782485961914, "reading_distances": [28, 24, 30, 17], "reading_log_probs": [-59.560096740722656, -60.401336669921875, -59.4659538269043, -59.17596435546875], "reading_logprob_std": 0.5263615251331142, "tokenization_path_logprob_std": 66.64533473377651, "trajectory_balance_rmse": 708.5663138794527}634{"canonical_length": 90, "canonical_nll": 1.5453548855251735, "generation_rouge_l_f1": 0.26022304832713755, "generation_token_f1": 0.39405204460966536, "prompt_id": "6a7a68d7cf446159", "reading_canonical_log_prob": -139.08193969726562, "reading_distances": [21, 18, 14, 18], "reading_log_probs": [-156.81521606445312, -153.08326721191406, -149.3349151611328, -153.218994140625], "reading_logprob_std": 3.054638418077099, "tokenization_path_logprob_std": 79.33829345169502, "trajectory_balance_rmse": 2072.1735761512427}635{"canonical_length": 216, "canonical_nll": 1.6467501322428386, "generation_rouge_l_f1": 0.2208067940552017, "generation_token_f1": 0.42462845010615713, "prompt_id": "a929bd728f13cca9", "reading_canonical_log_prob": -355.6980285644531, "reading_distances": [35, 36, 29, 38], "reading_log_probs": [-361.11627197265625, -366.558837890625, -364.03704833984375, -365.48583984375], "reading_logprob_std": 2.3603553035413403, "tokenization_path_logprob_std": 256.6633658023985, "trajectory_balance_rmse": 5856.588834630319}636{"canonical_length": 321, "canonical_nll": 0.9285406665267232, "generation_rouge_l_f1": 0.2273476112026359, "generation_token_f1": 0.46787479406919275, "prompt_id": "167bbbea26df8184", "reading_canonical_log_prob": -298.0615539550781, "reading_distances": [31, 34, 34, 37], "reading_log_probs": [-310.5487976074219, -314.5894775390625, -307.4354248046875, -309.5858459472656], "reading_logprob_std": 2.9970620494864106, "tokenization_path_logprob_std": 278.302581439145, "trajectory_balance_rmse": 6203.89353097579}637{"canonical_length": 195, "canonical_nll": 1.6795835837339743, "generation_rouge_l_f1": 0.2403846153846154, "generation_token_f1": 0.4278846153846154, "prompt_id": "00b0a2b88466f41c", "reading_canonical_log_prob": -327.518798828125, "reading_distances": [21, 26, 16, 25], "reading_log_probs": [-339.9766845703125, -340.30078125, -338.25677490234375, -342.37335205078125], "reading_logprob_std": 1.688857158658905, "tokenization_path_logprob_std": 104.8102261582482, "trajectory_balance_rmse": 4291.961230250935}638{"canonical_length": 89, "canonical_nll": 1.3383590183900984, "generation_rouge_l_f1": 0.17901234567901234, "generation_token_f1": 0.2901234567901234, "prompt_id": "a26fac121253e753", "reading_canonical_log_prob": -119.11395263671875, "reading_distances": [29, 27, 29, 31], "reading_log_probs": [-121.15413665771484, -118.25807189941406, -119.18081665039062, -121.43389892578125], "reading_logprob_std": 1.5376704452609995, "tokenization_path_logprob_std": 69.79278549768378, "trajectory_balance_rmse": 2140.8481252760143}639{"canonical_length": 223, "canonical_nll": 0.9883807400416901, "generation_rouge_l_f1": 0.20140515222482436, "generation_token_f1": 0.28103044496487123, "prompt_id": "109439f51e6651d1", "reading_canonical_log_prob": -220.40890502929688, "reading_distances": [38, 33, 38, 41], "reading_log_probs": [-228.2572021484375, -230.22994995117188, -228.42636108398438, -226.64187622070312], "reading_logprob_std": 1.4674586483635037, "tokenization_path_logprob_std": 260.96046175448606, "trajectory_balance_rmse": 3629.516839245051}640{"canonical_length": 174, "canonical_nll": 1.0217946370442708, "generation_rouge_l_f1": 0.2818181818181818, "generation_token_f1": 0.5409090909090909, "prompt_id": "25f04ac4bbf2d3c3", "reading_canonical_log_prob": -177.79226684570312, "reading_distances": [22, 20, 16, 20], "reading_log_probs": [-180.73764038085938, -180.0430908203125, -178.9196319580078, -181.40650939941406], "reading_logprob_std": 1.0622526742455927, "tokenization_path_logprob_std": 97.162472562456, "trajectory_balance_rmse": 3740.5976551442445}641{"canonical_length": 13, "canonical_nll": 1.484188960148738, "generation_rouge_l_f1": 0.06796116504854369, "generation_token_f1": 0.06796116504854369, "prompt_id": "68cd67de83285a28", "reading_canonical_log_prob": -19.294456481933594, "reading_distances": [10, 9, 11, 10], "reading_log_probs": [-18.332916259765625, -16.368274688720703, -14.996641159057617, -15.36241626739502], "reading_logprob_std": 1.4955862671062838, "tokenization_path_logprob_std": 15.689847059043576, "trajectory_balance_rmse": 178.7387951166794}642{"canonical_length": 12, "canonical_nll": 0.33476706345876056, "generation_rouge_l_f1": 0.145985401459854, "generation_token_f1": 0.145985401459854, "prompt_id": "41752c3a8cc3cb46", "reading_canonical_log_prob": -4.017204761505127, "reading_distances": [20, 21, 21, 19], "reading_log_probs": [-3.0427355766296387, -2.8204784393310547, -2.8336753845214844, -3.191840171813965], "reading_logprob_std": 0.17834944394850177, "tokenization_path_logprob_std": 26.593502376423636, "trajectory_balance_rmse": 201.13358925173375}643{"canonical_length": 33, "canonical_nll": 0.41971374280524976, "generation_rouge_l_f1": 0.22950819672131148, "generation_token_f1": 0.3770491803278689, "prompt_id": "6595c3b5c5d331ea", "reading_canonical_log_prob": -13.850553512573242, "reading_distances": [24, 21, 22, 23], "reading_log_probs": [-17.41798210144043, -17.379091262817383, -19.609493255615234, -21.488372802734375], "reading_logprob_std": 1.9740691957169634, "tokenization_path_logprob_std": 53.78306755159017, "trajectory_balance_rmse": 332.9118755485722}644{"canonical_length": 7, "canonical_nll": 0.3842164788927351, "generation_rouge_l_f1": 0.4615384615384615, "generation_token_f1": 0.7692307692307692, "prompt_id": "77f51b75ebf88292", "reading_canonical_log_prob": -2.6895153522491455, "reading_distances": [19, 22, 19, 26], "reading_log_probs": [-4.327537536621094, -3.7768797874450684, -2.9673187732696533, -3.5260848999023438], "reading_logprob_std": 0.5646718859484358, "tokenization_path_logprob_std": 38.50850163029403, "trajectory_balance_rmse": 131.28609530522792}645{"canonical_length": 250, "canonical_nll": 1.0323345947265625, "generation_rouge_l_f1": 0.3076923076923077, "generation_token_f1": 0.5256410256410255, "prompt_id": "3d42fc04af88620e", "reading_canonical_log_prob": -258.0836486816406, "reading_distances": [29, 25, 33, 29], "reading_log_probs": [-270.6935729980469, -269.32012939453125, -272.2176513671875, -270.89678955078125], "reading_logprob_std": 1.1859112510005978, "tokenization_path_logprob_std": 59.15667949479624, "trajectory_balance_rmse": 5067.980331117482}646{"canonical_length": 311, "canonical_nll": 0.6012236665682777, "generation_rouge_l_f1": 0.3961165048543689, "generation_token_f1": 0.5631067961165048, "prompt_id": "391687d0204ea44e", "reading_canonical_log_prob": -186.98056030273438, "reading_distances": [21, 24, 23, 26], "reading_log_probs": [-199.38998413085938, -194.42575073242188, -190.9603729248047, -198.5046844482422], "reading_logprob_std": 3.8949635622782193, "tokenization_path_logprob_std": 175.3854250291541, "trajectory_balance_rmse": 4060.944355821429}647{"canonical_length": 45, "canonical_nll": 1.5734771728515624, "generation_rouge_l_f1": 0.13215859030837004, "generation_token_f1": 0.1762114537444934, "prompt_id": "9043379228e84740", "reading_canonical_log_prob": -70.80647277832031, "reading_distances": [32, 31, 36, 33], "reading_log_probs": [-76.9529800415039, -79.878662109375, -75.92924499511719, -76.80668640136719], "reading_logprob_std": 1.7183782216330203, "tokenization_path_logprob_std": 61.79537592775453, "trajectory_balance_rmse": 985.2187613619909}648{"canonical_length": 14, "canonical_nll": 1.4326340811593192, "generation_rouge_l_f1": 0.3157894736842105, "generation_token_f1": 0.5263157894736842, "prompt_id": "82a1a337beb39341", "reading_canonical_log_prob": -20.05687713623047, "reading_distances": [26, 23, 25, 34], "reading_log_probs": [-19.631364822387695, -27.00426483154297, -20.11553192138672, -25.77166748046875], "reading_logprob_std": 3.799815739585818, "tokenization_path_logprob_std": 69.9270266203691, "trajectory_balance_rmse": 353.08635556849936}649{"canonical_length": 15, "canonical_nll": 1.1341517130533854, "generation_rouge_l_f1": 0.056818181818181816, "generation_token_f1": 0.0909090909090909, "prompt_id": "32a6cf563f35ecba", "reading_canonical_log_prob": -17.01227569580078, "reading_distances": [33, 25, 30, 33], "reading_log_probs": [-26.709611892700195, -31.003684997558594, -23.417856216430664, -23.438697814941406], "reading_logprob_std": 3.5910520963939105, "tokenization_path_logprob_std": 27.225887918055165, "trajectory_balance_rmse": 182.7872488325012}650{"canonical_length": 147, "canonical_nll": 1.533760745509141, "generation_rouge_l_f1": 0.16986301369863016, "generation_token_f1": 0.27945205479452057, "prompt_id": "2655c987f350c6c7", "reading_canonical_log_prob": -225.46282958984375, "reading_distances": [14, 17, 14, 15], "reading_log_probs": [-228.8407440185547, -228.51467895507812, -231.005615234375, -229.2171630859375], "reading_logprob_std": 1.1117360939042578, "tokenization_path_logprob_std": 89.95321131115624, "trajectory_balance_rmse": 3304.3767276327185}651{"canonical_length": 392, "canonical_nll": 1.0269138569734535, "generation_rouge_l_f1": 0.24731182795698922, "generation_token_f1": 0.4336917562724014, "prompt_id": "dd04ff95cd33338f", "reading_canonical_log_prob": -402.55023193359375, "reading_distances": [25, 30, 24, 23], "reading_log_probs": [-415.50006103515625, -438.5897216796875, -418.50732421875, -408.85821533203125], "reading_logprob_std": 12.801889947278232, "tokenization_path_logprob_std": 86.33557273107203, "trajectory_balance_rmse": 4333.319171325524}652{"canonical_length": 197, "canonical_nll": 0.6224484516279346, "generation_rouge_l_f1": 0.17177914110429446, "generation_token_f1": 0.3803680981595092, "prompt_id": "f51cb0e119af1f71", "reading_canonical_log_prob": -122.62234497070312, "reading_distances": [38, 44, 38, 39], "reading_log_probs": [-145.75982666015625, -141.9145050048828, -145.21717834472656, -141.67864990234375], "reading_logprob_std": 2.145177380232285, "tokenization_path_logprob_std": 234.0767112770556, "trajectory_balance_rmse": 3605.632799779413}653{"canonical_length": 11, "canonical_nll": 1.6177347356622869, "generation_rouge_l_f1": 0.15384615384615383, "generation_token_f1": 0.15384615384615383, "prompt_id": "40f09143e7b23121", "reading_canonical_log_prob": -17.795082092285156, "reading_distances": [42, 47, 49, 42], "reading_log_probs": [-16.18756103515625, -17.463369369506836, -16.261856079101562, -16.91912078857422], "reading_logprob_std": 0.6014028000960242, "tokenization_path_logprob_std": 86.8732402605505, "trajectory_balance_rmse": 291.82022235277714}654{"canonical_length": 98, "canonical_nll": 1.3198847868004624, "generation_rouge_l_f1": 0.25958702064896755, "generation_token_f1": 0.40117994100294985, "prompt_id": "09e3575b08c40649", "reading_canonical_log_prob": -129.3487091064453, "reading_distances": [16, 17, 13, 13], "reading_log_probs": [-134.04446411132812, -136.84637451171875, -133.30511474609375, -132.23883056640625], "reading_logprob_std": 1.9698687723151533, "tokenization_path_logprob_std": 52.11255204130168, "trajectory_balance_rmse": 2144.162716057061}655{"canonical_length": 11, "canonical_nll": 0.6279218847101385, "generation_rouge_l_f1": 0.06926406926406928, "generation_token_f1": 0.06926406926406928, "prompt_id": "fc90d5db5297a0af", "reading_canonical_log_prob": -6.907140731811523, "reading_distances": [18, 22, 20, 20], "reading_log_probs": [-6.753106117248535, -7.216258525848389, -7.339608192443848, -7.324772834777832], "reading_logprob_std": 0.27575778276410423, "tokenization_path_logprob_std": 43.75186116377477, "trajectory_balance_rmse": 248.97618077729632}656{"canonical_length": 244, "canonical_nll": 1.0681304931640625, "generation_rouge_l_f1": 0.33946830265848676, "generation_token_f1": 0.6298568507157464, "prompt_id": "e15ab7087e38fd8b", "reading_canonical_log_prob": -260.62384033203125, "reading_distances": [21, 16, 16, 15], "reading_log_probs": [-270.8970642089844, -268.90167236328125, -268.69256591796875, -265.1257019042969], "reading_logprob_std": 2.4009411607669344, "tokenization_path_logprob_std": 133.97977833113188, "trajectory_balance_rmse": 4760.541334374687}657{"canonical_length": 83, "canonical_nll": 0.9577029124799982, "generation_rouge_l_f1": 0.43283582089552236, "generation_token_f1": 0.5671641791044777, "prompt_id": "514750b55e58033f", "reading_canonical_log_prob": -79.48934173583984, "reading_distances": [17, 18, 16, 13], "reading_log_probs": [-88.26336669921875, -79.38911437988281, -85.41803741455078, -79.921630859375], "reading_logprob_std": 4.313494151746569, "tokenization_path_logprob_std": 171.27798746015102, "trajectory_balance_rmse": 1711.4711704282324}658{"canonical_length": 145, "canonical_nll": 1.3381583378232758, "generation_rouge_l_f1": 0.16346153846153846, "generation_token_f1": 0.2932692307692308, "prompt_id": "612827ad6442cf24", "reading_canonical_log_prob": -194.032958984375, "reading_distances": [27, 26, 28, 29], "reading_log_probs": [-205.2562713623047, -201.22589111328125, -199.14210510253906, -202.605224609375], "reading_logprob_std": 2.564061957140016, "tokenization_path_logprob_std": 245.25927737299426, "trajectory_balance_rmse": 3588.093636828092}659{"canonical_length": 184, "canonical_nll": 0.7947608284328295, "generation_rouge_l_f1": 0.31529411764705884, "generation_token_f1": 0.4423529411764706, "prompt_id": "637d5437d0c071a7", "reading_canonical_log_prob": -146.23599243164062, "reading_distances": [249, 215, 215, 227], "reading_log_probs": [-164.65899658203125, -165.4076385498047, -158.52976989746094, -157.632080078125], "reading_logprob_std": 4.042231993350255, "tokenization_path_logprob_std": 94.76820730885204, "trajectory_balance_rmse": 2962.910266770765}660{"canonical_length": 65, "canonical_nll": 1.0580731905423677, "generation_rouge_l_f1": 0.14285714285714288, "generation_token_f1": 0.14285714285714288, "prompt_id": "c97d2c584d8df586", "reading_canonical_log_prob": -68.7747573852539, "reading_distances": [15, 19, 16, 17], "reading_log_probs": [-71.36539459228516, -72.28498840332031, -72.34886169433594, -72.2350082397461], "reading_logprob_std": 0.4644556524523048, "tokenization_path_logprob_std": 38.7343906462762, "trajectory_balance_rmse": 874.8849622107581}661{"canonical_length": 17, "canonical_nll": 1.0733529259176815, "generation_rouge_l_f1": 0.6923076923076923, "generation_token_f1": 0.6923076923076923, "prompt_id": "a17a07dc87573da6", "reading_canonical_log_prob": -18.246999740600586, "reading_distances": [29, 25, 21, 24], "reading_log_probs": [-19.824642181396484, -19.980958938598633, -21.85655975341797, -19.79990577697754], "reading_logprob_std": 0.9972554036148262, "tokenization_path_logprob_std": 116.91472483226923, "trajectory_balance_rmse": 488.0745428397508}662{"canonical_length": 13, "canonical_nll": 0.4029342211209811, "generation_rouge_l_f1": 0.7, "generation_token_f1": 0.7, "prompt_id": "4b95b246a8b6c8d6", "reading_canonical_log_prob": -5.238144874572754, "reading_distances": [10, 12, 14, 12], "reading_log_probs": [-4.565258502960205, -3.772770404815674, -8.36904525756836, -4.9940104484558105], "reading_logprob_std": 2.0266698674582257, "tokenization_path_logprob_std": 20.326674021944537, "trajectory_balance_rmse": 326.0634881555477}663{"canonical_length": 99, "canonical_nll": 1.229902325254498, "generation_rouge_l_f1": 0.32773109243697485, "generation_token_f1": 0.5546218487394958, "prompt_id": "0c93e5e51c22b30d", "reading_canonical_log_prob": -121.76033020019531, "reading_distances": [17, 18, 18, 15], "reading_log_probs": [-127.68283081054688, -126.42900848388672, -127.92005157470703, -125.73341369628906], "reading_logprob_std": 1.0375057438270614, "tokenization_path_logprob_std": 123.23210447162555, "trajectory_balance_rmse": 2066.1069515310883}664{"canonical_length": 306, "canonical_nll": 1.1934272915709252, "generation_rouge_l_f1": 0.2097902097902098, "generation_token_f1": 0.4615384615384615, "prompt_id": "6c10022ee752e034", "reading_canonical_log_prob": -365.1887512207031, "reading_distances": [29, 34, 30, 36], "reading_log_probs": [-382.15301513671875, -383.4436950683594, -382.0418701171875, -388.07781982421875], "reading_logprob_std": 2.8380514559396763, "tokenization_path_logprob_std": 253.94382232192018, "trajectory_balance_rmse": 6571.119360450844}665{"canonical_length": 69, "canonical_nll": 1.1289042597231658, "generation_rouge_l_f1": 0.23134328358208955, "generation_token_f1": 0.3731343283582089, "prompt_id": "7884c89879ae1c04", "reading_canonical_log_prob": -77.89439392089844, "reading_distances": [35, 28, 26, 37], "reading_log_probs": [-83.47035217285156, -82.6593017578125, -80.75751495361328, -84.0531234741211], "reading_logprob_std": 1.436937672312382, "tokenization_path_logprob_std": 129.90854478722719, "trajectory_balance_rmse": 1263.810665021518}666{"canonical_length": 46, "canonical_nll": 0.953792489093283, "generation_rouge_l_f1": 0.4444444444444444, "generation_token_f1": 0.5185185185185185, "prompt_id": "803eb754dd6450de", "reading_canonical_log_prob": -43.874454498291016, "reading_distances": [14, 15, 17, 13], "reading_log_probs": [-49.46244812011719, -45.5025634765625, -52.2197265625, -47.36164855957031], "reading_logprob_std": 2.8849325345112207, "tokenization_path_logprob_std": 50.938606345510976, "trajectory_balance_rmse": 1141.9344122950733}667{"canonical_length": 12, "canonical_nll": 0.9686092535654703, "generation_rouge_l_f1": 0.37499999999999994, "generation_token_f1": 0.37499999999999994, "prompt_id": "7f58078c29e97e0d", "reading_canonical_log_prob": -11.623311042785645, "reading_distances": [39, 46, 39, 39], "reading_log_probs": [-26.68197250366211, -17.64406967163086, -17.775184631347656, -24.919658660888672], "reading_logprob_std": 4.7268312440447735, "tokenization_path_logprob_std": 16.040308989321986, "trajectory_balance_rmse": 186.7209838555478}668{"canonical_length": 135, "canonical_nll": 1.4648118760850695, "generation_rouge_l_f1": 0.20069204152249134, "generation_token_f1": 0.46366782006920415, "prompt_id": "979bfe699ee39dca", "reading_canonical_log_prob": -197.74960327148438, "reading_distances": [22, 21, 20, 29], "reading_log_probs": [-196.91375732421875, -198.2593994140625, -196.0160675048828, -198.55966186523438], "reading_logprob_std": 1.1873699672025178, "tokenization_path_logprob_std": 200.76876698810815, "trajectory_balance_rmse": 2555.35216278159}669{"canonical_length": 13, "canonical_nll": 1.514238797701322, "generation_rouge_l_f1": 0.07881773399014778, "generation_token_f1": 0.09852216748768473, "prompt_id": "f30d075da952702b", "reading_canonical_log_prob": -19.685104370117188, "reading_distances": [34, 26, 31, 29], "reading_log_probs": [-34.707210540771484, -31.26356315612793, -27.914390563964844, -32.13501739501953], "reading_logprob_std": 2.8048673247843703, "tokenization_path_logprob_std": 78.6027615760875, "trajectory_balance_rmse": 203.22335552421063}670{"canonical_length": 117, "canonical_nll": 1.126392592731704, "generation_rouge_l_f1": 0.22857142857142856, "generation_token_f1": 0.37402597402597404, "prompt_id": "ad627413d7298d7b", "reading_canonical_log_prob": -131.78793334960938, "reading_distances": [24, 17, 22, 29], "reading_log_probs": [-136.36361694335938, -137.7444610595703, -134.2756805419922, -138.78802490234375], "reading_logprob_std": 1.9499290209407942, "tokenization_path_logprob_std": 145.53330737516328, "trajectory_balance_rmse": 2250.873348763687}671{"canonical_length": 314, "canonical_nll": 1.561489129522044, "generation_rouge_l_f1": 0.28835489833641403, "generation_token_f1": 0.5360443622920519, "prompt_id": "802f028e684a8238", "reading_canonical_log_prob": -490.3075866699219, "reading_distances": [24, 22, 23, 19], "reading_log_probs": [-493.48626708984375, -492.62969970703125, -495.2522277832031, -493.72357177734375], "reading_logprob_std": 1.0923841344839698, "tokenization_path_logprob_std": 307.4095015578933, "trajectory_balance_rmse": 6133.560237653461}672{"canonical_length": 190, "canonical_nll": 1.4865819027549343, "generation_rouge_l_f1": 0.24516129032258063, "generation_token_f1": 0.421505376344086, "prompt_id": "2e2423f516c40a85", "reading_canonical_log_prob": -282.4505615234375, "reading_distances": [8, 16, 21, 16], "reading_log_probs": [-285.5587463378906, -285.17327880859375, -282.6826477050781, -280.02984619140625], "reading_logprob_std": 2.560676402240409, "tokenization_path_logprob_std": 199.4472137266515, "trajectory_balance_rmse": 4308.7868867016}673{"canonical_length": 297, "canonical_nll": 1.574747824106955, "generation_rouge_l_f1": 0.23344947735191635, "generation_token_f1": 0.421602787456446, "prompt_id": "8d1b09dca5de1a98", "reading_canonical_log_prob": -467.7001037597656, "reading_distances": [26, 21, 28, 23], "reading_log_probs": [-463.3004150390625, -462.8702392578125, -461.2901611328125, -461.13824462890625], "reading_logprob_std": 1.0962315658938653, "tokenization_path_logprob_std": 185.83721667396196, "trajectory_balance_rmse": 7385.541287498474}674{"canonical_length": 82, "canonical_nll": 0.6367000021585604, "generation_rouge_l_f1": 0.23052959501557632, "generation_token_f1": 0.3613707165109034, "prompt_id": "0609bdeeadeb9baf", "reading_canonical_log_prob": -52.20940017700195, "reading_distances": [35, 36, 39, 34], "reading_log_probs": [-55.81857681274414, -55.13807678222656, -54.7835693359375, -55.69906997680664], "reading_logprob_std": 0.4853806899131879, "tokenization_path_logprob_std": 85.5790353186114, "trajectory_balance_rmse": 987.8274712437241}675{"canonical_length": 222, "canonical_nll": 1.089935302734375, "generation_rouge_l_f1": 0.3976143141153082, "generation_token_f1": 0.6242544731610338, "prompt_id": "e9f992cd65ad5461", "reading_canonical_log_prob": -241.96563720703125, "reading_distances": [25, 32, 29, 24], "reading_log_probs": [-241.8640594482422, -240.9583282470703, -246.92735290527344, -241.48611450195312], "reading_logprob_std": 2.770604944262551, "tokenization_path_logprob_std": 117.92715025846178, "trajectory_balance_rmse": 3236.6159801725353}676{"canonical_length": 6, "canonical_nll": 1.322067419687907, "generation_rouge_l_f1": 0.06153846153846154, "generation_token_f1": 0.06153846153846154, "prompt_id": "15eeee68ceb20913", "reading_canonical_log_prob": -7.932404518127441, "reading_distances": [32, 38, 36, 31], "reading_log_probs": [-6.417309761047363, -8.176240921020508, -8.986620903015137, -8.636940002441406], "reading_logprob_std": 1.1406574518430488, "tokenization_path_logprob_std": 49.060299399907834, "trajectory_balance_rmse": 152.7357999684061}677{"canonical_length": 105, "canonical_nll": 1.3476004464285714, "generation_rouge_l_f1": 0.2076502732240437, "generation_token_f1": 0.32786885245901637, "prompt_id": "728a7ca9e481ba81", "reading_canonical_log_prob": -141.498046875, "reading_distances": [29, 30, 30, 24], "reading_log_probs": [-141.82452392578125, -140.52830505371094, -136.43753051757812, -141.91940307617188], "reading_logprob_std": 2.5727644042894235, "tokenization_path_logprob_std": 60.64120179927275, "trajectory_balance_rmse": 1969.716512065201}678{"canonical_length": 123, "canonical_nll": 1.4166231232930004, "generation_rouge_l_f1": 0.12994350282485875, "generation_token_f1": 0.22033898305084745, "prompt_id": "92cf4663b97aed32", "reading_canonical_log_prob": -174.24464416503906, "reading_distances": [26, 24, 30, 22], "reading_log_probs": [-176.00527954101562, -175.9342041015625, -177.5452423095703, -171.59385681152344], "reading_logprob_std": 2.560765438904962, "tokenization_path_logprob_std": 139.44829184330266, "trajectory_balance_rmse": 1817.6260211496228}679{"canonical_length": 66, "canonical_nll": 1.0237234866980351, "generation_rouge_l_f1": 0.28571428571428575, "generation_token_f1": 0.4505494505494505, "prompt_id": "3226fde653ee5231", "reading_canonical_log_prob": -67.56575012207031, "reading_distances": [8, 13, 12, 12], "reading_log_probs": [-62.58056640625, -67.21707153320312, -66.59766387939453, -64.49691772460938], "reading_logprob_std": 2.1115296838081012, "tokenization_path_logprob_std": 69.33278027129403, "trajectory_balance_rmse": 1559.3597158098687}680{"canonical_length": 259, "canonical_nll": 1.2596417872601955, "generation_rouge_l_f1": 0.3202099737532808, "generation_token_f1": 0.4881889763779528, "prompt_id": "d40a7f2ec7ebb044", "reading_canonical_log_prob": -326.2472229003906, "reading_distances": [22, 24, 26, 20], "reading_log_probs": [-341.02490234375, -342.7364807128906, -339.12274169921875, -344.58831787109375], "reading_logprob_std": 2.3382076337412796, "tokenization_path_logprob_std": 71.33971298741346, "trajectory_balance_rmse": 4738.8717069773365}681{"canonical_length": 13, "canonical_nll": 1.3694592989408052, "generation_rouge_l_f1": 0.21428571428571427, "generation_token_f1": 0.21428571428571427, "prompt_id": "6844823da97b3fe4", "reading_canonical_log_prob": -17.80297088623047, "reading_distances": [20, 24, 22, 21], "reading_log_probs": [-20.155590057373047, -21.232736587524414, -23.452102661132812, -25.99329376220703], "reading_logprob_std": 2.5844433174586725, "tokenization_path_logprob_std": 72.45786509582554, "trajectory_balance_rmse": 280.39895259864204}682{"canonical_length": 118, "canonical_nll": 0.9292996293407375, "generation_rouge_l_f1": 0.185929648241206, "generation_token_f1": 0.3165829145728643, "prompt_id": "49597fd7a90f1c14", "reading_canonical_log_prob": -109.65735626220703, "reading_distances": [24, 25, 31, 24], "reading_log_probs": [-117.4195556640625, -119.83979797363281, -119.97705078125, -122.56255340576172], "reading_logprob_std": 2.100909310678328, "tokenization_path_logprob_std": 129.39176615442156, "trajectory_balance_rmse": 1815.601138314494}683{"canonical_length": 258, "canonical_nll": 1.6452310251635174, "generation_rouge_l_f1": 0.17509727626459143, "generation_token_f1": 0.32684824902723736, "prompt_id": "06dc86d1f57bb10a", "reading_canonical_log_prob": -424.4696044921875, "reading_distances": [29, 34, 39, 30], "reading_log_probs": [-432.6266174316406, -431.4878845214844, -426.61468505859375, -434.80322265625], "reading_logprob_std": 3.46373283399737, "tokenization_path_logprob_std": 158.9500252378858, "trajectory_balance_rmse": 5303.248256707006}684{"canonical_length": 250, "canonical_nll": 1.3201622314453125, "generation_rouge_l_f1": 0.3089108910891089, "generation_token_f1": 0.499009900990099, "prompt_id": "104414445caa5ddc", "reading_canonical_log_prob": -330.0405578613281, "reading_distances": [37, 32, 36, 35], "reading_log_probs": [-330.5101623535156, -330.85113525390625, -331.070068359375, -331.20684814453125], "reading_logprob_std": 0.3039051963545587, "tokenization_path_logprob_std": 322.0614173043568, "trajectory_balance_rmse": 5736.279590553715}685{"canonical_length": 274, "canonical_nll": 0.8694618949054802, "generation_rouge_l_f1": 0.3578154425612053, "generation_token_f1": 0.5725047080979284, "prompt_id": "96cee7006ba2cc41", "reading_canonical_log_prob": -238.23255920410156, "reading_distances": [20, 20, 21, 18], "reading_log_probs": [-247.53038024902344, -249.85494995117188, -249.9408416748047, -245.6497802734375], "reading_logprob_std": 2.058612846282019, "tokenization_path_logprob_std": 115.85481651317455, "trajectory_balance_rmse": 4647.693594006642}686{"canonical_length": 150, "canonical_nll": 2.0009311930338542, "generation_rouge_l_f1": 0.21052631578947367, "generation_token_f1": 0.39234449760765544, "prompt_id": "c61662a19e060e22", "reading_canonical_log_prob": -300.1396789550781, "reading_distances": [28, 29, 33, 33], "reading_log_probs": [-304.54296875, -309.95758056640625, -307.7597351074219, -311.710205078125], "reading_logprob_std": 3.0895430330105036, "tokenization_path_logprob_std": 230.86274779327576, "trajectory_balance_rmse": 3169.9905829489517}687{"canonical_length": 80, "canonical_nll": 1.324944019317627, "generation_rouge_l_f1": 0.32926829268292684, "generation_token_f1": 0.4878048780487805, "prompt_id": "448e404116b3a283", "reading_canonical_log_prob": -105.99552154541016, "reading_distances": [18, 27, 26, 24], "reading_log_probs": [-112.64912414550781, -121.94136047363281, -116.9033203125, -118.99832153320312], "reading_logprob_std": 3.907135624487951, "tokenization_path_logprob_std": 87.12052106931338, "trajectory_balance_rmse": 1415.5717074222314}688{"canonical_length": 40, "canonical_nll": 2.188016891479492, "generation_rouge_l_f1": 0.3703703703703704, "generation_token_f1": 0.5432098765432098, "prompt_id": "62e94a0db3b86cbd", "reading_canonical_log_prob": -87.52067565917969, "reading_distances": [16, 17, 15, 16], "reading_log_probs": [-91.61722564697266, -91.06156921386719, -91.70140075683594, -91.28230285644531], "reading_logprob_std": 0.2974563820744732, "tokenization_path_logprob_std": 53.44485130872228, "trajectory_balance_rmse": 1086.871359115219}689{"canonical_length": 203, "canonical_nll": 1.0125751965151633, "generation_rouge_l_f1": 0.2745098039215686, "generation_token_f1": 0.5343137254901961, "prompt_id": "6996ebbe576a8626", "reading_canonical_log_prob": -205.55276489257812, "reading_distances": [33, 32, 37, 29], "reading_log_probs": [-209.2108154296875, -206.32144165039062, -209.07843017578125, -204.86492919921875], "reading_logprob_std": 2.135586151956987, "tokenization_path_logprob_std": 352.3376391907335, "trajectory_balance_rmse": 4427.282146004138}690{"canonical_length": 97, "canonical_nll": 1.2851588455672116, "generation_rouge_l_f1": 0.2722513089005235, "generation_token_f1": 0.4712041884816754, "prompt_id": "539a3dd889cf333b", "reading_canonical_log_prob": -124.66040802001953, "reading_distances": [39, 25, 33, 33], "reading_log_probs": [-123.9139404296875, -125.84386444091797, -129.65554809570312, -127.7822494506836], "reading_logprob_std": 2.4740317382962416, "tokenization_path_logprob_std": 162.1420914416631, "trajectory_balance_rmse": 2555.0493547034725}691{"canonical_length": 83, "canonical_nll": 1.8684756727103728, "generation_rouge_l_f1": 0.06382978723404255, "generation_token_f1": 0.08510638297872342, "prompt_id": "22dd62a830e20aab", "reading_canonical_log_prob": -155.08348083496094, "reading_distances": [24, 37, 29, 27], "reading_log_probs": [-159.15206909179688, -158.4122772216797, -155.12071228027344, -163.68914794921875], "reading_logprob_std": 3.5294247473075098, "tokenization_path_logprob_std": 160.25390226108036, "trajectory_balance_rmse": 1845.5215898191643}692{"canonical_length": 26, "canonical_nll": 1.1219711303710938, "generation_rouge_l_f1": 0.23255813953488372, "generation_token_f1": 0.2790697674418604, "prompt_id": "23f859fe909ddbe9", "reading_canonical_log_prob": -29.171249389648438, "reading_distances": [17, 14, 14, 16], "reading_log_probs": [-43.714752197265625, -33.1866455078125, -39.82737350463867, -40.434940338134766], "reading_logprob_std": 4.413200776533635, "tokenization_path_logprob_std": 101.03265957695243, "trajectory_balance_rmse": 526.8589693553826}693{"canonical_length": 97, "canonical_nll": 0.68323469653572, "generation_rouge_l_f1": 0.25555555555555554, "generation_token_f1": 0.38333333333333336, "prompt_id": "9658518adf176971", "reading_canonical_log_prob": -66.27376556396484, "reading_distances": [16, 13, 18, 14], "reading_log_probs": [-58.021060943603516, -60.81249237060547, -59.88477325439453, -62.43540954589844], "reading_logprob_std": 1.842829685104138, "tokenization_path_logprob_std": 84.91718883739132, "trajectory_balance_rmse": 1092.8963815276777}694{"canonical_length": 243, "canonical_nll": 1.3437910668643904, "generation_rouge_l_f1": 0.23762376237623764, "generation_token_f1": 0.40792079207920795, "prompt_id": "c496b50e51f7e82e", "reading_canonical_log_prob": -326.5412292480469, "reading_distances": [30, 37, 26, 33], "reading_log_probs": [-326.28778076171875, -323.24273681640625, -326.781982421875, -325.6956787109375], "reading_logprob_std": 1.570306549013543, "tokenization_path_logprob_std": 128.72726258730384, "trajectory_balance_rmse": 5554.152267937754}695{"canonical_length": 157, "canonical_nll": 1.0639274257003881, "generation_rouge_l_f1": 0.23448275862068965, "generation_token_f1": 0.4689655172413793, "prompt_id": "9433be6b41e84522", "reading_canonical_log_prob": -167.03660583496094, "reading_distances": [41, 34, 38, 35], "reading_log_probs": [-177.86663818359375, -178.55953979492188, -181.55958557128906, -173.50804138183594], "reading_logprob_std": 3.3224064301344503, "tokenization_path_logprob_std": 182.70167617762507, "trajectory_balance_rmse": 3655.4002227814694}696{"canonical_length": 14, "canonical_nll": 0.8172859464372907, "generation_rouge_l_f1": 0.6666666666666667, "generation_token_f1": 0.6666666666666667, "prompt_id": "22dd55d64faca0d8", "reading_canonical_log_prob": -11.44200325012207, "reading_distances": [40, 42, 49, 49], "reading_log_probs": [-16.7855224609375, -17.663820266723633, -18.218780517578125, -18.270343780517578], "reading_logprob_std": 0.689736692909688, "tokenization_path_logprob_std": 41.37648617546838, "trajectory_balance_rmse": 414.1323922151088}697{"canonical_length": 52, "canonical_nll": 1.2744770050048828, "generation_rouge_l_f1": 0.13636363636363638, "generation_token_f1": 0.2077922077922078, "prompt_id": "9ce142b4fe68303a", "reading_canonical_log_prob": -66.2728042602539, "reading_distances": [31, 27, 33, 38], "reading_log_probs": [-72.13319396972656, -75.74485778808594, -74.6055908203125, -74.69722747802734], "reading_logprob_std": 1.5312028311856931, "tokenization_path_logprob_std": 78.00203871497628, "trajectory_balance_rmse": 1265.9608324457945}698{"canonical_length": 28, "canonical_nll": 1.1275969914027624, "generation_rouge_l_f1": 0.4800000000000001, "generation_token_f1": 0.5333333333333333, "prompt_id": "e62a412929959b77", "reading_canonical_log_prob": -31.572715759277344, "reading_distances": [27, 24, 28, 24], "reading_log_probs": [-38.38939666748047, -31.69305419921875, -30.209686279296875, -35.52544403076172], "reading_logprob_std": 3.7091751314723846, "tokenization_path_logprob_std": 48.613500078652486, "trajectory_balance_rmse": 776.6919943041125}699{"canonical_length": 21, "canonical_nll": 1.2547046116420202, "generation_rouge_l_f1": 0.4210526315789474, "generation_token_f1": 0.4210526315789474, "prompt_id": "273ee8648fccd69e", "reading_canonical_log_prob": -26.348796844482422, "reading_distances": [31, 32, 30, 28], "reading_log_probs": [-43.21636962890625, -37.39649200439453, -37.43067932128906, -32.82877731323242], "reading_logprob_std": 4.255290338516319, "tokenization_path_logprob_std": 43.861952306636965, "trajectory_balance_rmse": 396.9683644522459}700{"canonical_length": 109, "canonical_nll": 1.3102277003297018, "generation_rouge_l_f1": 0.3146067415730337, "generation_token_f1": 0.46067415730337075, "prompt_id": "0ffd3304b685d298", "reading_canonical_log_prob": -142.8148193359375, "reading_distances": [22, 24, 24, 19], "reading_log_probs": [-146.7559051513672, -163.693603515625, -163.04476928710938, -156.71856689453125], "reading_logprob_std": 7.855785066611252, "tokenization_path_logprob_std": 146.27135784760728, "trajectory_balance_rmse": 2403.5065630768936}701{"canonical_length": 288, "canonical_nll": 1.1458083258734808, "generation_rouge_l_f1": 0.39080459770114945, "generation_token_f1": 0.5900383141762452, "prompt_id": "6b90af69594f1bfc", "reading_canonical_log_prob": -329.9927978515625, "reading_distances": [30, 33, 33, 33], "reading_log_probs": [-335.2255859375, -337.51800537109375, -336.8992614746094, -337.56427001953125], "reading_logprob_std": 1.093657498584978, "tokenization_path_logprob_std": 217.90446825136496, "trajectory_balance_rmse": 6983.049962584422}702{"canonical_length": 256, "canonical_nll": 1.2957913875579834, "generation_rouge_l_f1": 0.2857142857142857, "generation_token_f1": 0.6240601503759399, "prompt_id": "7021b5d065291a9c", "reading_canonical_log_prob": -331.72259521484375, "reading_distances": [18, 24, 19, 23], "reading_log_probs": [-327.726318359375, -331.0401916503906, -329.1961364746094, -332.20489501953125], "reading_logprob_std": 1.9792564216713582, "tokenization_path_logprob_std": 140.71564869312064, "trajectory_balance_rmse": 5728.646753469314}703{"canonical_length": 68, "canonical_nll": 0.734378365909352, "generation_rouge_l_f1": 0.3756906077348066, "generation_token_f1": 0.4861878453038674, "prompt_id": "56589cd90846fa45", "reading_canonical_log_prob": -49.93772888183594, "reading_distances": [39, 45, 32, 27], "reading_log_probs": [-66.63157653808594, -64.85526275634766, -58.20077896118164, -56.17434310913086], "reading_logprob_std": 5.060750374085043, "tokenization_path_logprob_std": 22.544211105025788, "trajectory_balance_rmse": 953.691725532039}704{"canonical_length": 187, "canonical_nll": 1.5871023902281083, "generation_rouge_l_f1": 0.2329113924050633, "generation_token_f1": 0.49620253164556966, "prompt_id": "b2517684dcd5d425", "reading_canonical_log_prob": -296.78814697265625, "reading_distances": [47, 36, 38, 36], "reading_log_probs": [-303.79254150390625, -300.53253173828125, -298.6169128417969, -301.1263122558594], "reading_logprob_std": 2.137811002756473, "tokenization_path_logprob_std": 73.33342951385374, "trajectory_balance_rmse": 4505.187465536014}705{"canonical_length": 2, "canonical_nll": 7.353148460388184, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "44fca97732ef92a8", "reading_canonical_log_prob": -14.706296920776367, "reading_distances": [22, 24, 22, 18], "reading_log_probs": [-10.105454444885254, -10.180644989013672, -13.394423484802246, -8.648943901062012], "reading_logprob_std": 2.002882315777437, "tokenization_path_logprob_std": 21.57331284182077, "trajectory_balance_rmse": 39.24637006684672}706{"canonical_length": 14, "canonical_nll": 0.6937011310032436, "generation_rouge_l_f1": 0.37499999999999994, "generation_token_f1": 0.37499999999999994, "prompt_id": "907693271b754ffe", "reading_canonical_log_prob": -9.71181583404541, "reading_distances": [5, 6, 7, 8], "reading_log_probs": [-10.130441665649414, -10.966141700744629, -10.511618614196777, -10.792024612426758], "reading_logprob_std": 0.3647965107107942, "tokenization_path_logprob_std": 19.600296745140657, "trajectory_balance_rmse": 175.9603101287057}707{"canonical_length": 220, "canonical_nll": 1.1987394853071733, "generation_rouge_l_f1": 0.21900826446280994, "generation_token_f1": 0.41735537190082644, "prompt_id": "3d885c9990d3f7a8", "reading_canonical_log_prob": -263.7226867675781, "reading_distances": [43, 52, 46, 50], "reading_log_probs": [-273.36328125, -279.24005126953125, -273.0422058105469, -291.0233154296875], "reading_logprob_std": 8.401861045154616, "tokenization_path_logprob_std": 100.78872508239999, "trajectory_balance_rmse": 3981.3107313633673}708{"canonical_length": 12, "canonical_nll": 1.1556708017985027, "generation_rouge_l_f1": 0.22222222222222224, "generation_token_f1": 0.22222222222222224, "prompt_id": "946f59f054cd3b7a", "reading_canonical_log_prob": -13.868049621582031, "reading_distances": [27, 25, 33, 27], "reading_log_probs": [-14.323083877563477, -14.32302188873291, -13.614836692810059, -14.522661209106445], "reading_logprob_std": 0.3986406769857445, "tokenization_path_logprob_std": 44.71473085388957, "trajectory_balance_rmse": 260.465843595447}709{"canonical_length": 114, "canonical_nll": 1.0690850374991434, "generation_rouge_l_f1": 0.19767441860465118, "generation_token_f1": 0.36046511627906974, "prompt_id": "1e094a0368e082c2", "reading_canonical_log_prob": -121.87569427490234, "reading_distances": [30, 31, 29, 21], "reading_log_probs": [-132.74830627441406, -128.73190307617188, -129.36898803710938, -131.64013671875], "reading_logprob_std": 1.8885872732752016, "tokenization_path_logprob_std": 171.40071134788903, "trajectory_balance_rmse": 2681.449234884013}710{"canonical_length": 8, "canonical_nll": 0.13641682267189026, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "f01addec53ef2695", "reading_canonical_log_prob": -1.091334581375122, "reading_distances": [24, 25, 23, 24], "reading_log_probs": [-1.5159099102020264, -0.8188730478286743, -0.7126538753509521, -1.7799417972564697], "reading_logprob_std": 0.5224009191852752, "tokenization_path_logprob_std": 46.18178616133577, "trajectory_balance_rmse": 141.49946143085344}711{"canonical_length": 43, "canonical_nll": 1.0626777826353562, "generation_rouge_l_f1": 0.37209302325581395, "generation_token_f1": 0.5116279069767442, "prompt_id": "bd70996cafa5e523", "reading_canonical_log_prob": -45.69514465332031, "reading_distances": [32, 29, 25, 18], "reading_log_probs": [-52.37285614013672, -56.23794174194336, -58.68858337402344, -53.22639465332031], "reading_logprob_std": 2.893475290662504, "tokenization_path_logprob_std": 81.53335692475189, "trajectory_balance_rmse": 821.094424984841}712{"canonical_length": 280, "canonical_nll": 1.0499823434012276, "generation_rouge_l_f1": 0.2513863216266174, "generation_token_f1": 0.5175600739371534, "prompt_id": "40b3b0fca502cfb8", "reading_canonical_log_prob": -293.99505615234375, "reading_distances": [29, 29, 26, 34], "reading_log_probs": [-291.36572265625, -295.16534423828125, -293.2851867675781, -291.1662292480469], "reading_logprob_std": 1.8748146141811575, "tokenization_path_logprob_std": 147.0337212963293, "trajectory_balance_rmse": 5463.534585431631}713{"canonical_length": 74, "canonical_nll": 1.2139390481484902, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "88f331fe7934b488", "reading_canonical_log_prob": -89.83148956298828, "reading_distances": [15, 10, 13, 16], "reading_log_probs": [-91.48065185546875, -92.88520812988281, -91.77847290039062, -92.64724731445312], "reading_logprob_std": 0.6744552041046388, "tokenization_path_logprob_std": 97.62833897553746, "trajectory_balance_rmse": 1489.4067390953203}714{"canonical_length": 33, "canonical_nll": 0.6797325249874231, "generation_rouge_l_f1": 0.24590163934426232, "generation_token_f1": 0.3770491803278688, "prompt_id": "8b4c34e2f1db7af4", "reading_canonical_log_prob": -22.43117332458496, "reading_distances": [18, 17, 20, 21], "reading_log_probs": [-21.307405471801758, -21.233732223510742, -23.45065689086914, -23.135263442993164], "reading_logprob_std": 1.175091065338754, "tokenization_path_logprob_std": 61.11427536838884, "trajectory_balance_rmse": 844.0638022007186}715{"canonical_length": 155, "canonical_nll": 2.0442621046496976, "generation_rouge_l_f1": 0.10722610722610723, "generation_token_f1": 0.1678321678321678, "prompt_id": "7a99289ed340c72d", "reading_canonical_log_prob": -316.8606262207031, "reading_distances": [13, 21, 17, 17], "reading_log_probs": [-312.9859619140625, -313.42742919921875, -319.08489990234375, -317.0125732421875], "reading_logprob_std": 2.9263217396658883, "tokenization_path_logprob_std": 79.8977308697131, "trajectory_balance_rmse": 2459.0820007832695}716{"canonical_length": 21, "canonical_nll": 1.224790300641741, "generation_rouge_l_f1": 0.380952380952381, "generation_token_f1": 0.4285714285714286, "prompt_id": "321ea8e8622b4791", "reading_canonical_log_prob": -25.720596313476562, "reading_distances": [26, 33, 35, 27], "reading_log_probs": [-21.886180877685547, -35.68284606933594, -36.34856033325195, -32.08246994018555], "reading_logprob_std": 6.677562626485896, "tokenization_path_logprob_std": 63.62395987585002, "trajectory_balance_rmse": 464.79776768067956}717{"canonical_length": 176, "canonical_nll": 1.8915415677157315, "generation_rouge_l_f1": 0.19277108433734938, "generation_token_f1": 0.4289156626506025, "prompt_id": "e9ba47f9f768bdc0", "reading_canonical_log_prob": -332.91131591796875, "reading_distances": [25, 29, 29, 20], "reading_log_probs": [-359.93670654296875, -360.34320068359375, -355.9232177734375, -352.00982666015625], "reading_logprob_std": 3.909441091896836, "tokenization_path_logprob_std": 150.45561594691543, "trajectory_balance_rmse": 4508.161637351647}718{"canonical_length": 7, "canonical_nll": 0.8350035122462681, "generation_rouge_l_f1": 0.05853658536585366, "generation_token_f1": 0.05853658536585366, "prompt_id": "cd4e165565532ef7", "reading_canonical_log_prob": -5.845024585723877, "reading_distances": [31, 32, 36, 37], "reading_log_probs": [-9.56353759765625, -13.218168258666992, -10.975590705871582, -7.7035040855407715], "reading_logprob_std": 2.326606430487415, "tokenization_path_logprob_std": 14.596286877444081, "trajectory_balance_rmse": 156.4007261592743}719{"canonical_length": 270, "canonical_nll": 0.9826551649305556, "generation_rouge_l_f1": 0.29225352112676056, "generation_token_f1": 0.5422535211267606, "prompt_id": "df9cf174d34cff3e", "reading_canonical_log_prob": -265.31689453125, "reading_distances": [32, 36, 42, 30], "reading_log_probs": [-269.70440673828125, -271.25860595703125, -272.1864013671875, -268.67486572265625], "reading_logprob_std": 1.5679920215832686, "tokenization_path_logprob_std": 176.5947241473597, "trajectory_balance_rmse": 4134.816752281963}720{"canonical_length": 143, "canonical_nll": 1.5959461985768137, "generation_rouge_l_f1": 0.43729903536977494, "generation_token_f1": 0.5466237942122187, "prompt_id": "1fba2562017bea7b", "reading_canonical_log_prob": -228.22030639648438, "reading_distances": [29, 34, 33, 29], "reading_log_probs": [-222.89883422851562, -232.5050506591797, -225.75985717773438, -226.12008666992188], "reading_logprob_std": 4.054179728515533, "tokenization_path_logprob_std": 57.142286394379, "trajectory_balance_rmse": 3081.2933058099516}721{"canonical_length": 28, "canonical_nll": 0.9793647357395717, "generation_rouge_l_f1": 0.23529411764705885, "generation_token_f1": 0.2823529411764706, "prompt_id": "56997da805bc232c", "reading_canonical_log_prob": -27.422212600708008, "reading_distances": [25, 26, 29, 36], "reading_log_probs": [-26.239730834960938, -23.971525192260742, -24.42751693725586, -26.43392562866211], "reading_logprob_std": 1.2504535138686335, "tokenization_path_logprob_std": 17.212008568650823, "trajectory_balance_rmse": 597.6815444932103}722{"canonical_length": 110, "canonical_nll": 1.5965897993607954, "generation_rouge_l_f1": 0.4385026737967914, "generation_token_f1": 0.5454545454545455, "prompt_id": "02d66047fd43a021", "reading_canonical_log_prob": -175.6248779296875, "reading_distances": [22, 19, 16, 18], "reading_log_probs": [-178.89089965820312, -173.7450714111328, -180.3872528076172, -177.5194549560547], "reading_logprob_std": 2.845878467664038, "tokenization_path_logprob_std": 100.51537468829783, "trajectory_balance_rmse": 2188.8477508501132}723{"canonical_length": 79, "canonical_nll": 1.5108573044402689, "generation_rouge_l_f1": 0.1329479768786127, "generation_token_f1": 0.20231213872832368, "prompt_id": "83a9bce5dad053f6", "reading_canonical_log_prob": -119.35772705078125, "reading_distances": [20, 21, 22, 24], "reading_log_probs": [-147.2223358154297, -150.84432983398438, -151.84555053710938, -148.53048706054688], "reading_logprob_std": 2.112466406902468, "tokenization_path_logprob_std": 114.6678821136939, "trajectory_balance_rmse": 1432.989997729893}724{"canonical_length": 21, "canonical_nll": 6.2388494582403276, "generation_rouge_l_f1": 0.02390438247011952, "generation_token_f1": 0.02390438247011952, "prompt_id": "35f4df3b6373154b", "reading_canonical_log_prob": -131.01583862304688, "reading_distances": [20, 17, 23, 20], "reading_log_probs": [-133.0187225341797, -125.10787200927734, -129.4127197265625, -124.69119262695312], "reading_logprob_std": 3.936270658468692, "tokenization_path_logprob_std": 22.000066876279366, "trajectory_balance_rmse": 239.45502060464096}725{"canonical_length": 228, "canonical_nll": 1.2164507414165295, "generation_rouge_l_f1": 0.19999999999999998, "generation_token_f1": 0.43529411764705883, "prompt_id": "7c0757fa3b0cff4b", "reading_canonical_log_prob": -277.35076904296875, "reading_distances": [21, 29, 32, 26], "reading_log_probs": [-273.2645568847656, -273.50927734375, -275.2358093261719, -273.6033935546875], "reading_logprob_std": 0.8997724887509592, "tokenization_path_logprob_std": 217.66484661469258, "trajectory_balance_rmse": 3088.07779520301}726{"canonical_length": 211, "canonical_nll": 1.7017113563573756, "generation_rouge_l_f1": 0.25165562913907286, "generation_token_f1": 0.49448123620309054, "prompt_id": "5fa3ea0df830456f", "reading_canonical_log_prob": -359.06109619140625, "reading_distances": [20, 27, 23, 19], "reading_log_probs": [-364.9305419921875, -362.61273193359375, -362.3762512207031, -369.55499267578125], "reading_logprob_std": 3.3300078222748906, "tokenization_path_logprob_std": 139.0129022621403, "trajectory_balance_rmse": 4635.472946961422}727{"canonical_length": 145, "canonical_nll": 1.1993761786099137, "generation_rouge_l_f1": 0.3404255319148936, "generation_token_f1": 0.4302600472813239, "prompt_id": "70669892361b488c", "reading_canonical_log_prob": -173.9095458984375, "reading_distances": [35, 36, 29, 32], "reading_log_probs": [-189.3448486328125, -194.01028442382812, -191.15176391601562, -198.9019775390625], "reading_logprob_std": 4.168688434043327, "tokenization_path_logprob_std": 135.3600696130654, "trajectory_balance_rmse": 2495.748908007007}728{"canonical_length": 12, "canonical_nll": 1.041111707687378, "generation_rouge_l_f1": 0.7272727272727273, "generation_token_f1": 0.9090909090909091, "prompt_id": "e90b400c44335d9a", "reading_canonical_log_prob": -12.493340492248535, "reading_distances": [19, 18, 21, 25], "reading_log_probs": [-15.076462745666504, -14.00628662109375, -16.59127426147461, -18.17074966430664], "reading_logprob_std": 1.815080312462258, "tokenization_path_logprob_std": 35.60589548782557, "trajectory_balance_rmse": 191.93108568883085}729{"canonical_length": 232, "canonical_nll": 1.409936049888874, "generation_rouge_l_f1": 0.3130081300813008, "generation_token_f1": 0.540650406504065, "prompt_id": "9f0a3bc3aaf621a1", "reading_canonical_log_prob": -327.10516357421875, "reading_distances": [20, 31, 25, 26], "reading_log_probs": [-327.915283203125, -327.863037109375, -325.67779541015625, -328.9070739746094], "reading_logprob_std": 1.3627871110096146, "tokenization_path_logprob_std": 318.353590042071, "trajectory_balance_rmse": 5328.337434312005}730{"canonical_length": 93, "canonical_nll": 0.5373414562594506, "generation_rouge_l_f1": 0.3647798742138365, "generation_token_f1": 0.5408805031446541, "prompt_id": "f2be334fa1e04283", "reading_canonical_log_prob": -49.972755432128906, "reading_distances": [25, 15, 23, 24], "reading_log_probs": [-50.26274490356445, -53.052330017089844, -52.58436584472656, -54.002838134765625], "reading_logprob_std": 1.5888813167370723, "tokenization_path_logprob_std": 28.125578806426983, "trajectory_balance_rmse": 1675.6904354412782}731{"canonical_length": 54, "canonical_nll": 1.5391529224537037, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "1a42eb844c9543a9", "reading_canonical_log_prob": -83.1142578125, "reading_distances": [45, 31, 40, 36], "reading_log_probs": [-84.1326675415039, -83.49754333496094, -82.7173080444336, -84.99696350097656], "reading_logprob_std": 0.9664147536396863, "tokenization_path_logprob_std": 37.13205299887194, "trajectory_balance_rmse": 1061.0616868769746}732{"canonical_length": 40, "canonical_nll": 0.538700008392334, "generation_rouge_l_f1": 0.11726384364820848, "generation_token_f1": 0.1758957654723127, "prompt_id": "3afed29700596cb5", "reading_canonical_log_prob": -21.54800033569336, "reading_distances": [11, 9, 16, 14], "reading_log_probs": [-24.444419860839844, -27.12662696838379, -25.80370330810547, -25.341896057128906], "reading_logprob_std": 1.1178851817249416, "tokenization_path_logprob_std": 43.76350621175208, "trajectory_balance_rmse": 438.8358979554244}733{"canonical_length": 28, "canonical_nll": 0.7047033309936523, "generation_rouge_l_f1": 0.3928571428571428, "generation_token_f1": 0.5357142857142857, "prompt_id": "e24f8c298fcc6b27", "reading_canonical_log_prob": -19.731693267822266, "reading_distances": [13, 12, 13, 12], "reading_log_probs": [-19.310823440551758, -21.716495513916016, -21.13595199584961, -20.92420768737793], "reading_logprob_std": 1.0300059678004572, "tokenization_path_logprob_std": 16.997954893193878, "trajectory_balance_rmse": 561.1670877545465}734{"canonical_length": 232, "canonical_nll": 1.5596142472891972, "generation_rouge_l_f1": 0.20907297830374755, "generation_token_f1": 0.44181459566074954, "prompt_id": "d2492eb6d00f3204", "reading_canonical_log_prob": -361.83050537109375, "reading_distances": [31, 27, 20, 34], "reading_log_probs": [-377.79132080078125, -372.9911193847656, -375.6042175292969, -378.71710205078125], "reading_logprob_std": 2.5493076629722546, "tokenization_path_logprob_std": 177.31155788845675, "trajectory_balance_rmse": 5972.357219657684}735{"canonical_length": 83, "canonical_nll": 1.5096077976456608, "generation_rouge_l_f1": 0.22857142857142856, "generation_token_f1": 0.35555555555555557, "prompt_id": "d1d32462252d2ba1", "reading_canonical_log_prob": -125.29744720458984, "reading_distances": [18, 20, 14, 14], "reading_log_probs": [-121.448486328125, -121.58560180664062, -121.43827056884766, -121.62442016601562], "reading_logprob_std": 0.09474648748154121, "tokenization_path_logprob_std": 155.85458537322776, "trajectory_balance_rmse": 1661.5240690366709}736{"canonical_length": 171, "canonical_nll": 1.3224282069512976, "generation_rouge_l_f1": 0.24311926605504589, "generation_token_f1": 0.41743119266055045, "prompt_id": "981d7ddcc229e43f", "reading_canonical_log_prob": -226.13522338867188, "reading_distances": [22, 19, 20, 21], "reading_log_probs": [-214.64508056640625, -225.3516387939453, -221.05062866210938, -215.33181762695312], "reading_logprob_std": 5.064034518912628, "tokenization_path_logprob_std": 101.7174232332657, "trajectory_balance_rmse": 3152.635293248071}737{"canonical_length": 14, "canonical_nll": 1.760364259992327, "generation_rouge_l_f1": 0.18867924528301888, "generation_token_f1": 0.30188679245283023, "prompt_id": "a353a5c04a14d88b", "reading_canonical_log_prob": -24.645099639892578, "reading_distances": [35, 31, 31, 35], "reading_log_probs": [-20.231250762939453, -21.718008041381836, -22.62590789794922, -21.708370208740234], "reading_logprob_std": 0.9913358960223595, "tokenization_path_logprob_std": 49.73032897808985, "trajectory_balance_rmse": 329.89818510107665}738{"canonical_length": 37, "canonical_nll": 0.6388453921756229, "generation_rouge_l_f1": 0.26126126126126126, "generation_token_f1": 0.2702702702702703, "prompt_id": "104ca723f309949c", "reading_canonical_log_prob": -23.637279510498047, "reading_distances": [17, 15, 21, 14], "reading_log_probs": [-26.581438064575195, -26.666414260864258, -26.775440216064453, -26.239646911621094], "reading_logprob_std": 0.23143942247911553, "tokenization_path_logprob_std": 72.47036576374376, "trajectory_balance_rmse": 576.2663249266169}739{"canonical_length": 270, "canonical_nll": 0.8968722873263889, "generation_rouge_l_f1": 0.3070362473347548, "generation_token_f1": 0.4690831556503198, "prompt_id": "5768cefb5b534d32", "reading_canonical_log_prob": -242.155517578125, "reading_distances": [22, 20, 25, 20], "reading_log_probs": [-241.911376953125, -240.42886352539062, -241.5691680908203, -243.49188232421875], "reading_logprob_std": 1.2646527482585883, "tokenization_path_logprob_std": 219.7711640591679, "trajectory_balance_rmse": 4373.196223881381}740{"canonical_length": 460, "canonical_nll": 0.7538102523140285, "generation_rouge_l_f1": 0.34210526315789475, "generation_token_f1": 0.5263157894736843, "prompt_id": "2c20880d9a9f5257", "reading_canonical_log_prob": -346.7527160644531, "reading_distances": [31, 27, 35, 21], "reading_log_probs": [-345.388427734375, -351.0662536621094, -351.2027282714844, -346.12847900390625], "reading_logprob_std": 3.119023931279518, "tokenization_path_logprob_std": 187.61124132518538, "trajectory_balance_rmse": 8048.920883066553}741{"canonical_length": 43, "canonical_nll": 1.295146764710892, "generation_rouge_l_f1": 0.2809917355371901, "generation_token_f1": 0.4132231404958678, "prompt_id": "1a2898fe9876d2bc", "reading_canonical_log_prob": -55.69131088256836, "reading_distances": [19, 17, 12, 16], "reading_log_probs": [-60.98249053955078, -65.38202667236328, -71.73664855957031, -70.10774230957031], "reading_logprob_std": 4.861798329443202, "tokenization_path_logprob_std": 66.75801182723268, "trajectory_balance_rmse": 822.9029710945059}742{"canonical_length": 85, "canonical_nll": 1.011567777745864, "generation_rouge_l_f1": 0.4878048780487805, "generation_token_f1": 0.646341463414634, "prompt_id": "52568183284a12f5", "reading_canonical_log_prob": -85.98326110839844, "reading_distances": [17, 20, 15, 17], "reading_log_probs": [-88.63740539550781, -88.30722045898438, -92.47531127929688, -93.59648132324219], "reading_logprob_std": 2.677642955507499, "tokenization_path_logprob_std": 117.25132798317617, "trajectory_balance_rmse": 1824.9256279488263}743{"canonical_length": 202, "canonical_nll": 1.667326067933942, "generation_rouge_l_f1": 0.2554347826086957, "generation_token_f1": 0.5652173913043478, "prompt_id": "ea255680cda4b618", "reading_canonical_log_prob": -336.79986572265625, "reading_distances": [29, 17, 25, 26], "reading_log_probs": [-339.33953857421875, -336.3436584472656, -336.60699462890625, -345.46832275390625], "reading_logprob_std": 4.241226930005858, "tokenization_path_logprob_std": 267.15356317158086, "trajectory_balance_rmse": 4451.943212649789}744{"canonical_length": 19, "canonical_nll": 1.1167055431165194, "generation_rouge_l_f1": 0.18750000000000003, "generation_token_f1": 0.3125, "prompt_id": "559aadd3a6bc5deb", "reading_canonical_log_prob": -21.217405319213867, "reading_distances": [32, 35, 35, 39], "reading_log_probs": [-25.804523468017578, -26.735431671142578, -26.912317276000977, -25.3803653717041], "reading_logprob_std": 0.7353047624669469, "tokenization_path_logprob_std": 74.05510616300475, "trajectory_balance_rmse": 434.4750369282803}745{"canonical_length": 24, "canonical_nll": 0.9988199869791666, "generation_rouge_l_f1": 0.3283582089552239, "generation_token_f1": 0.44776119402985076, "prompt_id": "4196e10bb50daa8f", "reading_canonical_log_prob": -23.9716796875, "reading_distances": [25, 23, 20, 18], "reading_log_probs": [-25.234403610229492, -25.793750762939453, -30.671287536621094, -27.45877456665039], "reading_logprob_std": 2.444437751403508, "tokenization_path_logprob_std": 20.13190978787891, "trajectory_balance_rmse": 486.0583302816336}746{"canonical_length": 293, "canonical_nll": 1.339147053480962, "generation_rouge_l_f1": 0.2401500938086304, "generation_token_f1": 0.5140712945590995, "prompt_id": "0d2dbbf4ae30ce19", "reading_canonical_log_prob": -392.3700866699219, "reading_distances": [18, 13, 20, 21], "reading_log_probs": [-404.3046875, -397.9330749511719, -392.713623046875, -396.34881591796875], "reading_logprob_std": 4.840913277594246, "tokenization_path_logprob_std": 125.84206738194517, "trajectory_balance_rmse": 6887.638732611314}747{"canonical_length": 227, "canonical_nll": 1.0829617000361371, "generation_rouge_l_f1": 0.28832951945080093, "generation_token_f1": 0.5034324942791761, "prompt_id": "cd0a37baf57f0d68", "reading_canonical_log_prob": -245.83230590820312, "reading_distances": [23, 29, 26, 27], "reading_log_probs": [-253.9263916015625, -256.24578857421875, -253.2523193359375, -252.69418334960938], "reading_logprob_std": 1.5609448645899375, "tokenization_path_logprob_std": 284.6846727261882, "trajectory_balance_rmse": 4466.052096644015}748{"canonical_length": 27, "canonical_nll": 1.6618982950846355, "generation_rouge_l_f1": 0.25, "generation_token_f1": 0.2857142857142857, "prompt_id": "f0dd5c7b12d6ff90", "reading_canonical_log_prob": -44.871253967285156, "reading_distances": [23, 23, 19, 24], "reading_log_probs": [-54.55548095703125, -53.36397933959961, -53.91365051269531, -52.41978454589844], "reading_logprob_std": 0.9045302334123049, "tokenization_path_logprob_std": 83.59059670250213, "trajectory_balance_rmse": 527.1817214510031}749{"canonical_length": 29, "canonical_nll": 0.9300013575060614, "generation_rouge_l_f1": 0.5263157894736842, "generation_token_f1": 0.5263157894736842, "prompt_id": "c737e978a37038bd", "reading_canonical_log_prob": -26.97003936767578, "reading_distances": [23, 22, 27, 26], "reading_log_probs": [-29.022411346435547, -28.453235626220703, -32.638343811035156, -28.288650512695312], "reading_logprob_std": 2.049377631501336, "tokenization_path_logprob_std": 66.75626084054723, "trajectory_balance_rmse": 457.4732527639439}750{"canonical_length": 90, "canonical_nll": 1.414278835720486, "generation_rouge_l_f1": 0.3439490445859873, "generation_token_f1": 0.535031847133758, "prompt_id": "6fe0c1612b081574", "reading_canonical_log_prob": -127.28509521484375, "reading_distances": [20, 25, 24, 27], "reading_log_probs": [-126.31346130371094, -135.4749755859375, -129.96485900878906, -126.58177185058594], "reading_logprob_std": 4.264522010982897, "tokenization_path_logprob_std": 201.80197385523664, "trajectory_balance_rmse": 2091.658805113953}751{"canonical_length": 276, "canonical_nll": 1.3486012997834578, "generation_rouge_l_f1": 0.2631578947368421, "generation_token_f1": 0.5020242914979757, "prompt_id": "a45d6c8378dedb7b", "reading_canonical_log_prob": -372.2139587402344, "reading_distances": [13, 14, 15, 14], "reading_log_probs": [-375.36199951171875, -371.62786865234375, -376.64190673828125, -380.4620666503906], "reading_logprob_std": 3.6442858760041386, "tokenization_path_logprob_std": 197.02482000109666, "trajectory_balance_rmse": 5926.438771577069}752{"canonical_length": 103, "canonical_nll": 1.1661217698773134, "generation_rouge_l_f1": 0.23978201634877386, "generation_token_f1": 0.34877384196185285, "prompt_id": "2e7570c8d7724f9e", "reading_canonical_log_prob": -120.11054229736328, "reading_distances": [25, 22, 17, 25], "reading_log_probs": [-127.96102142333984, -130.5218048095703, -128.525390625, -130.81607055664062], "reading_logprob_std": 1.4243977141237492, "tokenization_path_logprob_std": 77.17918934357871, "trajectory_balance_rmse": 1490.2533493257024}753{"canonical_length": 30, "canonical_nll": 2.06608034769694, "generation_rouge_l_f1": 0.12631578947368421, "generation_token_f1": 0.1473684210526316, "prompt_id": "7aa76ccd6b778b98", "reading_canonical_log_prob": -61.9824104309082, "reading_distances": [40, 36, 27, 35], "reading_log_probs": [-59.96065139770508, -59.91017532348633, -61.048946380615234, -61.200172424316406], "reading_logprob_std": 0.6896319994693763, "tokenization_path_logprob_std": 92.58919089106774, "trajectory_balance_rmse": 719.8592795629248}754{"canonical_length": 252, "canonical_nll": 1.543016585092696, "generation_rouge_l_f1": 0.24379232505643342, "generation_token_f1": 0.38374717832957106, "prompt_id": "0667a89ba9993054", "reading_canonical_log_prob": -388.8401794433594, "reading_distances": [28, 27, 25, 26], "reading_log_probs": [-394.9930114746094, -390.34759521484375, -395.2584228515625, -392.7796936035156], "reading_logprob_std": 2.2862731821109725, "tokenization_path_logprob_std": 123.71518144393531, "trajectory_balance_rmse": 5008.1092074075}755{"canonical_length": 248, "canonical_nll": 1.1653332864084551, "generation_rouge_l_f1": 0.2664015904572564, "generation_token_f1": 0.5646123260437377, "prompt_id": "45c2d1c36166f6b6", "reading_canonical_log_prob": -289.0026550292969, "reading_distances": [24, 24, 22, 32], "reading_log_probs": [-294.74468994140625, -297.040283203125, -298.0632629394531, -295.52862548828125], "reading_logprob_std": 1.490335915851865, "tokenization_path_logprob_std": 60.796018019425254, "trajectory_balance_rmse": 4380.721883569176}756{"canonical_length": 20, "canonical_nll": 1.3397778511047362, "generation_rouge_l_f1": 0.5454545454545454, "generation_token_f1": 0.7272727272727272, "prompt_id": "e5f7552dff198384", "reading_canonical_log_prob": -26.795557022094727, "reading_distances": [29, 20, 21, 27], "reading_log_probs": [-26.719961166381836, -26.155683517456055, -27.10870361328125, -27.734046936035156], "reading_logprob_std": 0.6638544367793908, "tokenization_path_logprob_std": 85.1524935972831, "trajectory_balance_rmse": 463.02130294977246}757{"canonical_length": 117, "canonical_nll": 1.6637748000968215, "generation_rouge_l_f1": 0.2058823529411765, "generation_token_f1": 0.3725490196078432, "prompt_id": "4a930bccec9843cd", "reading_canonical_log_prob": -194.66165161132812, "reading_distances": [34, 32, 26, 27], "reading_log_probs": [-204.55682373046875, -197.06298828125, -197.48837280273438, -208.15093994140625], "reading_logprob_std": 5.4455807892826105, "tokenization_path_logprob_std": 205.7625378017723, "trajectory_balance_rmse": 2787.5818162970772}758{"canonical_length": 49, "canonical_nll": 0.9345281951281489, "generation_rouge_l_f1": 0.2626262626262626, "generation_token_f1": 0.3838383838383838, "prompt_id": "35925631dfc295f8", "reading_canonical_log_prob": -45.7918815612793, "reading_distances": [33, 29, 36, 28], "reading_log_probs": [-56.176231384277344, -54.443138122558594, -56.781089782714844, -52.78596878051758], "reading_logprob_std": 1.803620582653425, "tokenization_path_logprob_std": 156.99955132882405, "trajectory_balance_rmse": 982.9062865835592}759{"canonical_length": 295, "canonical_nll": 1.169253798662606, "generation_rouge_l_f1": 0.26807760141093484, "generation_token_f1": 0.5185185185185185, "prompt_id": "f6597a51ed7cc783", "reading_canonical_log_prob": -344.92987060546875, "reading_distances": [21, 21, 19, 23], "reading_log_probs": [-359.41400146484375, -353.1893005371094, -356.4183349609375, -353.5068054199219], "reading_logprob_std": 2.910045267182099, "tokenization_path_logprob_std": 258.7543389211125, "trajectory_balance_rmse": 5814.5812162171715}760{"canonical_length": 213, "canonical_nll": 1.225994844391872, "generation_rouge_l_f1": 0.22881355932203393, "generation_token_f1": 0.4406779661016949, "prompt_id": "87270c0f1888627a", "reading_canonical_log_prob": -261.13690185546875, "reading_distances": [13, 15, 9, 10], "reading_log_probs": [-256.70465087890625, -261.8726501464844, -257.7978210449219, -260.1383056640625], "reading_logprob_std": 2.323490080824225, "tokenization_path_logprob_std": 367.10046460647385, "trajectory_balance_rmse": 4904.329335146753}761{"canonical_length": 14, "canonical_nll": 1.7994296210152763, "generation_rouge_l_f1": 0.15384615384615385, "generation_token_f1": 0.15384615384615385, "prompt_id": "a53e27e32808bf25", "reading_canonical_log_prob": -25.192014694213867, "reading_distances": [25, 20, 27, 27], "reading_log_probs": [-25.036354064941406, -25.5306339263916, -24.41357421875, -25.109817504882812], "reading_logprob_std": 0.46072657833381875, "tokenization_path_logprob_std": 23.575951026300192, "trajectory_balance_rmse": 270.35633683515226}762{"canonical_length": 128, "canonical_nll": 1.1767538785934448, "generation_rouge_l_f1": 0.23896103896103896, "generation_token_f1": 0.43636363636363634, "prompt_id": "1e1844cf1025dc97", "reading_canonical_log_prob": -150.62449645996094, "reading_distances": [15, 16, 18, 16], "reading_log_probs": [-156.2438201904297, -156.83033752441406, -157.33119201660156, -161.61849975585938], "reading_logprob_std": 2.449012061535352, "tokenization_path_logprob_std": 177.42147223710896, "trajectory_balance_rmse": 2935.6096008128775}763{"canonical_length": 155, "canonical_nll": 1.5611719931325605, "generation_rouge_l_f1": 0.21333333333333332, "generation_token_f1": 0.37866666666666665, "prompt_id": "87981277aeac45ee", "reading_canonical_log_prob": -241.98165893554688, "reading_distances": [21, 21, 23, 18], "reading_log_probs": [-247.3105010986328, -243.0272979736328, -246.19677734375, -247.03750610351562], "reading_logprob_std": 1.9683946388201894, "tokenization_path_logprob_std": 122.07433203961644, "trajectory_balance_rmse": 3098.895632281101}764{"canonical_length": 26, "canonical_nll": 0.18903614924504206, "generation_rouge_l_f1": 0.17674418604651163, "generation_token_f1": 0.17674418604651163, "prompt_id": "40b80040395896ae", "reading_canonical_log_prob": -4.914939880371094, "reading_distances": [39, 46, 45, 47], "reading_log_probs": [-4.572225570678711, -5.929042816162109, -5.324554443359375, -4.900821208953857], "reading_logprob_std": 0.5857418788828792, "tokenization_path_logprob_std": 103.44981284740432, "trajectory_balance_rmse": 306.3083021548979}765{"canonical_length": 356, "canonical_nll": 1.5343576495567064, "generation_rouge_l_f1": 0.22553897180762852, "generation_token_f1": 0.4709784411276949, "prompt_id": "63d17b19976fa874", "reading_canonical_log_prob": -546.2313232421875, "reading_distances": [29, 29, 27, 20], "reading_log_probs": [-559.8518676757812, -555.6793212890625, -557.5709228515625, -553.4818115234375], "reading_logprob_std": 2.7129090339758344, "tokenization_path_logprob_std": 262.0540719199768, "trajectory_balance_rmse": 8277.194369171419}766{"canonical_length": 19, "canonical_nll": 0.5544486798738179, "generation_rouge_l_f1": 0.1217391304347826, "generation_token_f1": 0.1217391304347826, "prompt_id": "53bd5d756df134c2", "reading_canonical_log_prob": -10.534524917602539, "reading_distances": [27, 21, 25, 25], "reading_log_probs": [-5.132636070251465, -5.616946220397949, -5.887775897979736, -5.919641494750977], "reading_logprob_std": 0.3640236460703648, "tokenization_path_logprob_std": 57.14330304872518, "trajectory_balance_rmse": 212.49561299974215}767{"canonical_length": 114, "canonical_nll": 1.0456813343784266, "generation_rouge_l_f1": 0.18857142857142858, "generation_token_f1": 0.3485714285714286, "prompt_id": "1cf0a058de57af94", "reading_canonical_log_prob": -119.20767211914062, "reading_distances": [38, 28, 33, 36], "reading_log_probs": [-131.50161743164062, -118.37771606445312, -126.7669448852539, -132.30288696289062], "reading_logprob_std": 6.3916100183774605, "tokenization_path_logprob_std": 83.29909543843026, "trajectory_balance_rmse": 1519.5016601646366}768{"canonical_length": 257, "canonical_nll": 1.4660542410171449, "generation_rouge_l_f1": 0.1904761904761905, "generation_token_f1": 0.43478260869565216, "prompt_id": "a5e93d67499f0604", "reading_canonical_log_prob": -376.77593994140625, "reading_distances": [15, 19, 20, 19], "reading_log_probs": [-378.2296142578125, -378.52557373046875, -382.840087890625, -382.57464599609375], "reading_logprob_std": 2.505058848024672, "tokenization_path_logprob_std": 243.13951857560463, "trajectory_balance_rmse": 4649.095418544854}769{"canonical_length": 118, "canonical_nll": 1.3174068644895391, "generation_rouge_l_f1": 0.3571428571428571, "generation_token_f1": 0.48412698412698413, "prompt_id": "9493ef83593c3f2d", "reading_canonical_log_prob": -155.45401000976562, "reading_distances": [32, 36, 37, 32], "reading_log_probs": [-165.2293701171875, -160.3260955810547, -165.48623657226562, -161.20510864257812], "reading_logprob_std": 2.677538317969086, "tokenization_path_logprob_std": 102.49411173677726, "trajectory_balance_rmse": 2638.6941833338947}770{"canonical_length": 22, "canonical_nll": 2.1381603587757456, "generation_rouge_l_f1": 0.21428571428571427, "generation_token_f1": 0.21428571428571427, "prompt_id": "be453082a5e24aed", "reading_canonical_log_prob": -47.039527893066406, "reading_distances": [35, 24, 25, 37], "reading_log_probs": [-41.93138122558594, -42.73185729980469, -41.762657165527344, -47.7342529296875], "reading_logprob_std": 2.8279235120018025, "tokenization_path_logprob_std": 35.939104899863736, "trajectory_balance_rmse": 413.49330891119615}771{"canonical_length": 280, "canonical_nll": 1.119319806780134, "generation_rouge_l_f1": 0.34892086330935246, "generation_token_f1": 0.5683453237410071, "prompt_id": "6b6295e1afd98591", "reading_canonical_log_prob": -313.4095458984375, "reading_distances": [14, 17, 17, 20], "reading_log_probs": [-316.63494873046875, -317.6562805175781, -320.68487548828125, -316.8618469238281], "reading_logprob_std": 1.8689472925782482, "tokenization_path_logprob_std": 309.7680650303401, "trajectory_balance_rmse": 6066.7825597721085}772{"canonical_length": 200, "canonical_nll": 1.253603973388672, "generation_rouge_l_f1": 0.2844827586206896, "generation_token_f1": 0.5043103448275862, "prompt_id": "900ff807cdf33dd9", "reading_canonical_log_prob": -250.72079467773438, "reading_distances": [23, 20, 27, 21], "reading_log_probs": [-257.53875732421875, -257.5404968261719, -261.5161437988281, -250.94073486328125], "reading_logprob_std": 4.383257992648104, "tokenization_path_logprob_std": 221.01772970588794, "trajectory_balance_rmse": 3688.00473665084}773{"canonical_length": 120, "canonical_nll": 1.5241629282633464, "generation_rouge_l_f1": 0.23140495867768596, "generation_token_f1": 0.39118457300275483, "prompt_id": "6b5f709d21a7a023", "reading_canonical_log_prob": -182.89955139160156, "reading_distances": [13, 19, 16, 18], "reading_log_probs": [-185.30609130859375, -182.760986328125, -182.45773315429688, -184.0924072265625], "reading_logprob_std": 1.310233312763507, "tokenization_path_logprob_std": 56.677328212332654, "trajectory_balance_rmse": 2381.514689322541}774{"canonical_length": 52, "canonical_nll": 0.8888505055354192, "generation_rouge_l_f1": 0.15894039735099336, "generation_token_f1": 0.23841059602649006, "prompt_id": "e072285f618dcbd9", "reading_canonical_log_prob": -46.2202262878418, "reading_distances": [28, 34, 21, 27], "reading_log_probs": [-48.35102081298828, -47.38457489013672, -47.87305450439453, -47.4222297668457], "reading_logprob_std": 0.45354120435865297, "tokenization_path_logprob_std": 11.431911222797606, "trajectory_balance_rmse": 1133.7326446246188}775{"canonical_length": 147, "canonical_nll": 1.463953446368782, "generation_rouge_l_f1": 0.16625916870415647, "generation_token_f1": 0.30806845965770174, "prompt_id": "21874077bcbf7b19", "reading_canonical_log_prob": -215.20115661621094, "reading_distances": [27, 20, 22, 23], "reading_log_probs": [-213.16421508789062, -215.96185302734375, -215.26986694335938, -216.14309692382812], "reading_logprob_std": 1.3665195893953617, "tokenization_path_logprob_std": 130.8130524065936, "trajectory_balance_rmse": 4082.767871039697}776{"canonical_length": 108, "canonical_nll": 1.1453503502739801, "generation_rouge_l_f1": 0.21546961325966849, "generation_token_f1": 0.39779005524861877, "prompt_id": "8ba54ff004576b32", "reading_canonical_log_prob": -123.69783782958984, "reading_distances": [20, 23, 22, 21], "reading_log_probs": [-124.01066589355469, -123.40670776367188, -124.16252899169922, -123.64823913574219], "reading_logprob_std": 0.34318097344055876, "tokenization_path_logprob_std": 28.482576897216788, "trajectory_balance_rmse": 2700.2890507299635}777{"canonical_length": 8, "canonical_nll": 0.3828320801258087, "generation_rouge_l_f1": 0.030418250950570342, "generation_token_f1": 0.030418250950570342, "prompt_id": "bccef35e52f5dc4b", "reading_canonical_log_prob": -3.0626566410064697, "reading_distances": [34, 34, 34, 34], "reading_log_probs": [-5.279749393463135, -6.3738813400268555, -7.567453384399414, -3.6549670696258545], "reading_logprob_std": 1.6632120045432035, "tokenization_path_logprob_std": 36.006084482474144, "trajectory_balance_rmse": 149.61188392632963}778{"canonical_length": 53, "canonical_nll": 0.9870195928609596, "generation_rouge_l_f1": 0.1377049180327869, "generation_token_f1": 0.22295081967213118, "prompt_id": "492f8c42a0de57b2", "reading_canonical_log_prob": -52.31203842163086, "reading_distances": [47, 39, 37, 45], "reading_log_probs": [-53.17278289794922, -58.74184799194336, -55.579246520996094, -57.5674934387207], "reading_logprob_std": 2.440172746854784, "tokenization_path_logprob_std": 115.89161164654622, "trajectory_balance_rmse": 958.3073928491417}779{"canonical_length": 211, "canonical_nll": 1.2776301234819314, "generation_rouge_l_f1": 0.21123595505617979, "generation_token_f1": 0.39101123595505616, "prompt_id": "6aa6f6caad65a47f", "reading_canonical_log_prob": -269.5799560546875, "reading_distances": [21, 21, 29, 27], "reading_log_probs": [-303.40252685546875, -298.858154296875, -292.449951171875, -294.21185302734375], "reading_logprob_std": 4.923042778590142, "tokenization_path_logprob_std": 179.78612727018862, "trajectory_balance_rmse": 4033.941258965207}780{"canonical_length": 107, "canonical_nll": 1.5175022588712033, "generation_rouge_l_f1": 0.24031007751937983, "generation_token_f1": 0.4108527131782946, "prompt_id": "83dae6df424d2709", "reading_canonical_log_prob": -162.37274169921875, "reading_distances": [31, 32, 29, 25], "reading_log_probs": [-166.46499633789062, -163.51368713378906, -177.21905517578125, -165.79025268554688], "reading_logprob_std": 6.1131870898227385, "tokenization_path_logprob_std": 171.65762045918737, "trajectory_balance_rmse": 2204.1288609381877}781{"canonical_length": 104, "canonical_nll": 1.2185645470252404, "generation_rouge_l_f1": 0.19672131147540983, "generation_token_f1": 0.34972677595628415, "prompt_id": "d7aad76280327bd4", "reading_canonical_log_prob": -126.730712890625, "reading_distances": [28, 32, 36, 32], "reading_log_probs": [-130.57501220703125, -131.61883544921875, -130.7520294189453, -130.54995727539062], "reading_logprob_std": 0.5046629402919585, "tokenization_path_logprob_std": 156.81032937598712, "trajectory_balance_rmse": 2457.456690696018}782{"canonical_length": 362, "canonical_nll": 1.5589989087858254, "generation_rouge_l_f1": 0.14446227929373995, "generation_token_f1": 0.32423756019261635, "prompt_id": "117db47d424e536b", "reading_canonical_log_prob": -564.3576049804688, "reading_distances": [27, 15, 15, 23], "reading_log_probs": [-584.3198852539062, -580.96728515625, -586.99951171875, -583.326171875], "reading_logprob_std": 2.4975541560064314, "tokenization_path_logprob_std": 294.19344983899606, "trajectory_balance_rmse": 7142.947192104117}783{"canonical_length": 207, "canonical_nll": 1.4463226760643115, "generation_rouge_l_f1": 0.21748400852878466, "generation_token_f1": 0.47334754797441364, "prompt_id": "f593668e2032bc26", "reading_canonical_log_prob": -299.3887939453125, "reading_distances": [32, 31, 31, 33], "reading_log_probs": [-307.1547546386719, -304.1673583984375, -311.2171630859375, -308.5782470703125], "reading_logprob_std": 2.937878942034741, "tokenization_path_logprob_std": 221.97939894359888, "trajectory_balance_rmse": 4632.911131133171}784{"canonical_length": 243, "canonical_nll": 1.0464398419415508, "generation_rouge_l_f1": 0.28884026258205686, "generation_token_f1": 0.5514223194748359, "prompt_id": "62a2987fac290e0c", "reading_canonical_log_prob": -254.28488159179688, "reading_distances": [13, 16, 18, 14], "reading_log_probs": [-252.73092651367188, -261.0446472167969, -255.24081420898438, -254.50534057617188], "reading_logprob_std": 3.6003930783678166, "tokenization_path_logprob_std": 234.47141208838127, "trajectory_balance_rmse": 5269.455362239289}785{"canonical_length": 28, "canonical_nll": 0.8642869676862445, "generation_rouge_l_f1": 0.1411764705882353, "generation_token_f1": 0.16470588235294117, "prompt_id": "0597c4917add4bdd", "reading_canonical_log_prob": -24.200035095214844, "reading_distances": [17, 16, 17, 20], "reading_log_probs": [-23.688817977905273, -27.285865783691406, -25.069807052612305, -29.096113204956055], "reading_logprob_std": 2.3889289558336517, "tokenization_path_logprob_std": 17.857694307525595, "trajectory_balance_rmse": 321.28517996282443}786{"canonical_length": 51, "canonical_nll": 1.6526961981081496, "generation_rouge_l_f1": 0.09433962264150944, "generation_token_f1": 0.16352201257861637, "prompt_id": "ea67625f5161437c", "reading_canonical_log_prob": -84.28750610351562, "reading_distances": [19, 21, 20, 20], "reading_log_probs": [-95.88897705078125, -97.1307373046875, -98.53396606445312, -96.45120239257812], "reading_logprob_std": 1.1410067158535901, "tokenization_path_logprob_std": 105.31671319797871, "trajectory_balance_rmse": 996.9193987154131}787{"canonical_length": 98, "canonical_nll": 0.9500726972307477, "generation_rouge_l_f1": 0.30625, "generation_token_f1": 0.41250000000000003, "prompt_id": "f9bf7abf6deb13c3", "reading_canonical_log_prob": -93.10712432861328, "reading_distances": [17, 20, 14, 24], "reading_log_probs": [-97.69297790527344, -94.46056365966797, -94.78704833984375, -96.56810760498047], "reading_logprob_std": 1.5242147574148617, "tokenization_path_logprob_std": 82.91315789398547, "trajectory_balance_rmse": 2136.049079608463}788{"canonical_length": 144, "canonical_nll": 1.5034221013387044, "generation_rouge_l_f1": 0.28211586901763225, "generation_token_f1": 0.47355163727959704, "prompt_id": "dd61adf3859cf501", "reading_canonical_log_prob": -216.49278259277344, "reading_distances": [40, 41, 39, 36], "reading_log_probs": [-249.11587524414062, -246.12814331054688, -256.4361877441406, -246.37376403808594], "reading_logprob_std": 4.809722544821358, "tokenization_path_logprob_std": 173.68944926989468, "trajectory_balance_rmse": 2464.7820392387503}789{"canonical_length": 38, "canonical_nll": 2.5052446064196134, "generation_rouge_l_f1": 0.19696969696969696, "generation_token_f1": 0.33333333333333337, "prompt_id": "932b65e0c79cc6cd", "reading_canonical_log_prob": -95.19929504394531, "reading_distances": [31, 23, 29, 24], "reading_log_probs": [-100.07051849365234, -98.19245910644531, -101.51792907714844, -101.53324890136719], "reading_logprob_std": 1.5806524597367766, "tokenization_path_logprob_std": 50.55594088806873, "trajectory_balance_rmse": 622.9872910407049}790{"canonical_length": 246, "canonical_nll": 1.4364518576521215, "generation_rouge_l_f1": 0.21765913757700206, "generation_token_f1": 0.3819301848049281, "prompt_id": "5a38778847ffa7e8", "reading_canonical_log_prob": -353.3671569824219, "reading_distances": [28, 29, 34, 30], "reading_log_probs": [-364.7378234863281, -363.1058654785156, -367.708740234375, -362.9343566894531], "reading_logprob_std": 2.212704515357343, "tokenization_path_logprob_std": 242.49648364850594, "trajectory_balance_rmse": 5702.484692352706}791{"canonical_length": 163, "canonical_nll": 0.8726598821534701, "generation_rouge_l_f1": 0.31666666666666665, "generation_token_f1": 0.5111111111111112, "prompt_id": "bea3ed585e39be34", "reading_canonical_log_prob": -142.24356079101562, "reading_distances": [19, 15, 16, 14], "reading_log_probs": [-153.48355102539062, -152.2627716064453, -150.8661346435547, -152.1920166015625], "reading_logprob_std": 1.0693766142382353, "tokenization_path_logprob_std": 300.1475337161013, "trajectory_balance_rmse": 3088.407125829386}792{"canonical_length": 256, "canonical_nll": 0.6479513645172119, "generation_rouge_l_f1": 0.32727272727272727, "generation_token_f1": 0.602020202020202, "prompt_id": "e77970903b119833", "reading_canonical_log_prob": -165.87554931640625, "reading_distances": [21, 26, 26, 33], "reading_log_probs": [-171.10386657714844, -173.70594787597656, -177.16213989257812, -183.85379028320312], "reading_logprob_std": 5.520689954527511, "tokenization_path_logprob_std": 153.47284612267092, "trajectory_balance_rmse": 2354.803442612328}793{"canonical_length": 28, "canonical_nll": 1.0143534796578544, "generation_rouge_l_f1": 0.09090909090909091, "generation_token_f1": 0.12121212121212122, "prompt_id": "4ea6cdd31bfaf581", "reading_canonical_log_prob": -28.401897430419922, "reading_distances": [24, 17, 19, 19], "reading_log_probs": [-43.98585510253906, -41.12461853027344, -36.5130615234375, -32.150543212890625], "reading_logprob_std": 5.203652386888138, "tokenization_path_logprob_std": 9.511040933727918, "trajectory_balance_rmse": 255.7398613458556}794{"canonical_length": 92, "canonical_nll": 0.7255196778670602, "generation_rouge_l_f1": 0.43601895734597146, "generation_token_f1": 0.5876777251184835, "prompt_id": "c7f44052e9719f7b", "reading_canonical_log_prob": -66.74781036376953, "reading_distances": [30, 36, 40, 40], "reading_log_probs": [-69.33727264404297, -69.7576904296875, -68.1507339477539, -68.42899322509766], "reading_logprob_std": 0.754694768151246, "tokenization_path_logprob_std": 132.00527990896387, "trajectory_balance_rmse": 1797.5351724042782}795{"canonical_length": 22, "canonical_nll": 1.8729967637495561, "generation_rouge_l_f1": 0.3673469387755102, "generation_token_f1": 0.4897959183673469, "prompt_id": "4d4bec1ba05bf153", "reading_canonical_log_prob": -41.205928802490234, "reading_distances": [41, 31, 45, 38], "reading_log_probs": [-38.942386627197266, -40.52156066894531, -45.5158805847168, -41.56782150268555], "reading_logprob_std": 2.8021210034782396, "tokenization_path_logprob_std": 54.45089043535473, "trajectory_balance_rmse": 666.991813586226}796{"canonical_length": 56, "canonical_nll": 1.9184979030064173, "generation_rouge_l_f1": 0.3838383838383838, "generation_token_f1": 0.3838383838383838, "prompt_id": "2dc7b73ea6e1f7f9", "reading_canonical_log_prob": -107.43588256835938, "reading_distances": [45, 47, 42, 49], "reading_log_probs": [-125.8658676147461, -128.89930725097656, -120.68162536621094, -128.46939086914062], "reading_logprob_std": 3.7773543677809043, "tokenization_path_logprob_std": 139.99449540033723, "trajectory_balance_rmse": 1210.7018888404764}797{"canonical_length": 14, "canonical_nll": 2.8491532461983815, "generation_rouge_l_f1": 0.03731343283582089, "generation_token_f1": 0.04477611940298507, "prompt_id": "cc99639a6604cf81", "reading_canonical_log_prob": -39.888145446777344, "reading_distances": [37, 35, 37, 33], "reading_log_probs": [-46.01288986206055, -35.85075378417969, -45.78478240966797, -42.27179718017578], "reading_logprob_std": 4.7396607242991955, "tokenization_path_logprob_std": 25.56910685013247, "trajectory_balance_rmse": 296.70065310119463}798{"canonical_length": 41, "canonical_nll": 1.0633038776676829, "generation_rouge_l_f1": 0.1889763779527559, "generation_token_f1": 0.2125984251968504, "prompt_id": "34af9dd76e319507", "reading_canonical_log_prob": -43.595458984375, "reading_distances": [29, 35, 37, 35], "reading_log_probs": [-47.606361389160156, -44.51984405517578, -48.87495422363281, -45.649810791015625], "reading_logprob_std": 1.94955907066652, "tokenization_path_logprob_std": 87.45784792054562, "trajectory_balance_rmse": 821.0966904565186}799{"canonical_length": 319, "canonical_nll": 1.3765379328712775, "generation_rouge_l_f1": 0.3103448275862069, "generation_token_f1": 0.5379310344827586, "prompt_id": "761d4b8b4ce8e2b9", "reading_canonical_log_prob": -439.1156005859375, "reading_distances": [29, 23, 27, 31], "reading_log_probs": [-444.59765625, -445.3570556640625, -447.2808837890625, -450.4403991699219], "reading_logprob_std": 2.6050923787163116, "tokenization_path_logprob_std": 223.14184863071083, "trajectory_balance_rmse": 7185.807155801198}800{"canonical_length": 30, "canonical_nll": 0.6991831461588541, "generation_rouge_l_f1": 0.5714285714285714, "generation_token_f1": 0.6984126984126985, "prompt_id": "667ddd428394372e", "reading_canonical_log_prob": -20.975494384765625, "reading_distances": [32, 33, 31, 34], "reading_log_probs": [-27.84893798828125, -24.398534774780273, -23.577117919921875, -26.29897117614746], "reading_logprob_std": 1.920308036583469, "tokenization_path_logprob_std": 58.06922619968707, "trajectory_balance_rmse": 612.3337081080202}801{"canonical_length": 15, "canonical_nll": 0.5307352066040039, "generation_rouge_l_f1": 0.11353711790393015, "generation_token_f1": 0.12227074235807858, "prompt_id": "eaab92bbea496d12", "reading_canonical_log_prob": -7.961028099060059, "reading_distances": [28, 24, 20, 23], "reading_log_probs": [-11.907941818237305, -10.501361846923828, -11.664210319519043, -10.117084503173828], "reading_logprob_std": 0.8726647944095556, "tokenization_path_logprob_std": 50.9583046009483, "trajectory_balance_rmse": 317.68618391921484}802{"canonical_length": 77, "canonical_nll": 0.8912206872717127, "generation_rouge_l_f1": 0.6888888888888889, "generation_token_f1": 0.7555555555555555, "prompt_id": "7cb40862b1b15473", "reading_canonical_log_prob": -68.62399291992188, "reading_distances": [33, 25, 25, 25], "reading_log_probs": [-71.24822998046875, -88.74046325683594, -86.09971618652344, -78.36802673339844], "reading_logprob_std": 7.913998039592973, "tokenization_path_logprob_std": 121.33451016621227, "trajectory_balance_rmse": 769.2190984758879}803{"canonical_length": 37, "canonical_nll": 1.4717768076303843, "generation_rouge_l_f1": 0.2545454545454545, "generation_token_f1": 0.32727272727272727, "prompt_id": "b218ec9e88427665", "reading_canonical_log_prob": -54.45574188232422, "reading_distances": [28, 35, 29, 27], "reading_log_probs": [-65.38687133789062, -65.28277587890625, -60.805145263671875, -66.92472839355469], "reading_logprob_std": 2.6388529022696536, "tokenization_path_logprob_std": 114.30864027893553, "trajectory_balance_rmse": 1007.6312109297228}804{"canonical_length": 68, "canonical_nll": 1.3993262122659123, "generation_rouge_l_f1": 0.23529411764705882, "generation_token_f1": 0.2823529411764706, "prompt_id": "605c5bbf05c7cb07", "reading_canonical_log_prob": -95.15418243408203, "reading_distances": [34, 31, 35, 28], "reading_log_probs": [-93.50746154785156, -92.1400146484375, -91.337158203125, -90.40834045410156], "reading_logprob_std": 1.3130960736171224, "tokenization_path_logprob_std": 109.44603480070002, "trajectory_balance_rmse": 1925.599492650854}805{"canonical_length": 163, "canonical_nll": 2.194194208624904, "generation_rouge_l_f1": 0.20095693779904306, "generation_token_f1": 0.3827751196172249, "prompt_id": "371fd9b264246fc1", "reading_canonical_log_prob": -357.6536560058594, "reading_distances": [18, 15, 18, 17], "reading_log_probs": [-359.6326904296875, -359.5281677246094, -359.150146484375, -358.156005859375], "reading_logprob_std": 0.6732022232076592, "tokenization_path_logprob_std": 154.69244363191842, "trajectory_balance_rmse": 3515.60134633138}806{"canonical_length": 9, "canonical_nll": 1.1824671427408855, "generation_rouge_l_f1": 0.2105263157894737, "generation_token_f1": 0.2105263157894737, "prompt_id": "ed3c13f24a96d3b3", "reading_canonical_log_prob": -10.642204284667969, "reading_distances": [28, 24, 28, 29], "reading_log_probs": [-13.244285583496094, -10.934606552124023, -11.665451049804688, -11.155872344970703], "reading_logprob_std": 1.0420959750617562, "tokenization_path_logprob_std": 28.424667566607848, "trajectory_balance_rmse": 168.94620340675417}807{"canonical_length": 27, "canonical_nll": 0.8400069342719184, "generation_rouge_l_f1": 0.5652173913043479, "generation_token_f1": 0.5652173913043479, "prompt_id": "9e311e757597dd6a", "reading_canonical_log_prob": -22.680187225341797, "reading_distances": [32, 34, 33, 34], "reading_log_probs": [-27.878740310668945, -24.88071060180664, -24.55006217956543, -24.15873146057129], "reading_logprob_std": 1.7002565510252339, "tokenization_path_logprob_std": 55.04390060808531, "trajectory_balance_rmse": 646.4507854460965}808{"canonical_length": 23, "canonical_nll": 0.938609911047894, "generation_rouge_l_f1": 0.24561403508771928, "generation_token_f1": 0.2807017543859649, "prompt_id": "6d2468c671bba59a", "reading_canonical_log_prob": -21.588027954101562, "reading_distances": [38, 41, 45, 45], "reading_log_probs": [-20.69329833984375, -24.31011962890625, -24.72235870361328, -21.846805572509766], "reading_logprob_std": 1.9397586961759412, "tokenization_path_logprob_std": 63.675521483159585, "trajectory_balance_rmse": 427.9747043881981}809{"canonical_length": 34, "canonical_nll": 1.7512783723718979, "generation_rouge_l_f1": 0.1111111111111111, "generation_token_f1": 0.1111111111111111, "prompt_id": "d8ebdfdb02245d42", "reading_canonical_log_prob": -59.54346466064453, "reading_distances": [20, 20, 21, 20], "reading_log_probs": [-61.16478729248047, -63.05513000488281, -58.41001892089844, -64.67520141601562], "reading_logprob_std": 2.6916437314858315, "tokenization_path_logprob_std": 103.57525144386715, "trajectory_balance_rmse": 928.0390548839933}810{"canonical_length": 282, "canonical_nll": 1.1836179909131206, "generation_rouge_l_f1": 0.25585585585585585, "generation_token_f1": 0.5153153153153153, "prompt_id": "0659b1d0d1c9b272", "reading_canonical_log_prob": -333.7802734375, "reading_distances": [22, 21, 19, 21], "reading_log_probs": [-341.5354309082031, -338.25927734375, -341.46826171875, -337.87408447265625], "reading_logprob_std": 1.9897071799274082, "tokenization_path_logprob_std": 88.35978422185394, "trajectory_balance_rmse": 5247.707290225362}811{"canonical_length": 126, "canonical_nll": 1.6179733276367188, "generation_rouge_l_f1": 0.24404761904761907, "generation_token_f1": 0.3988095238095238, "prompt_id": "11e44461a25c9604", "reading_canonical_log_prob": -203.86463928222656, "reading_distances": [13, 16, 12, 13], "reading_log_probs": [-204.22689819335938, -207.91937255859375, -203.95950317382812, -204.47190856933594], "reading_logprob_std": 1.8617650655067959, "tokenization_path_logprob_std": 283.5852139562946, "trajectory_balance_rmse": 3514.1439014324033}812{"canonical_length": 4, "canonical_nll": 0.6438968181610107, "generation_rouge_l_f1": 0.017857142857142856, "generation_token_f1": 0.017857142857142856, "prompt_id": "72cbdeacc12e1d9a", "reading_canonical_log_prob": -2.575587272644043, "reading_distances": [20, 24, 19, 21], "reading_log_probs": [-3.2365236282348633, -3.054765462875366, -3.2853362560272217, -2.6569125652313232], "reading_logprob_std": 0.2854431994920232, "tokenization_path_logprob_std": 31.820232585476795, "trajectory_balance_rmse": 77.93822437509519}813{"canonical_length": 9, "canonical_nll": 1.6783118777804904, "generation_rouge_l_f1": 0.05303030303030303, "generation_token_f1": 0.05303030303030303, "prompt_id": "e46df451982e05b1", "reading_canonical_log_prob": -15.104806900024414, "reading_distances": [16, 22, 15, 22], "reading_log_probs": [-14.601043701171875, -11.04335880279541, -12.398832321166992, -13.771626472473145], "reading_logprob_std": 1.5641852633212665, "tokenization_path_logprob_std": 27.720119944387633, "trajectory_balance_rmse": 118.58562112290296}814{"canonical_length": 112, "canonical_nll": 0.9613617488316127, "generation_rouge_l_f1": 0.35449735449735453, "generation_token_f1": 0.5185185185185185, "prompt_id": "ad4b8c7e2797f458", "reading_canonical_log_prob": -107.67251586914062, "reading_distances": [36, 35, 37, 30], "reading_log_probs": [-106.49899291992188, -112.38934326171875, -122.7408676147461, -115.6265640258789], "reading_logprob_std": 6.770365852421221, "tokenization_path_logprob_std": 119.77867297210148, "trajectory_balance_rmse": 1698.1408142243376}815{"canonical_length": 229, "canonical_nll": 1.314555472161572, "generation_rouge_l_f1": 0.2597402597402597, "generation_token_f1": 0.4935064935064935, "prompt_id": "9d2ca51ff7ce1141", "reading_canonical_log_prob": -301.033203125, "reading_distances": [29, 33, 32, 26], "reading_log_probs": [-299.9564208984375, -308.6119689941406, -295.4560241699219, -304.5466613769531], "reading_logprob_std": 5.68981316529652, "tokenization_path_logprob_std": 244.677592494875, "trajectory_balance_rmse": 5349.295790876275}816{"canonical_length": 3, "canonical_nll": 1.8078675270080566, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.6666666666666666, "prompt_id": "f22d1f7967fa63cf", "reading_canonical_log_prob": -5.42360258102417, "reading_distances": [24, 31, 24, 32], "reading_log_probs": [-6.477930068969727, -7.323730945587158, -6.760849952697754, -7.519737243652344], "reading_logprob_std": 0.48407543955164756, "tokenization_path_logprob_std": 37.44813491143834, "trajectory_balance_rmse": 59.90185708100575}817{"canonical_length": 341, "canonical_nll": 1.4314043598790323, "generation_rouge_l_f1": 0.2006472491909385, "generation_token_f1": 0.3883495145631068, "prompt_id": "a546d753c3fe436f", "reading_canonical_log_prob": -488.10888671875, "reading_distances": [20, 22, 25, 22], "reading_log_probs": [-500.3086853027344, -497.2439270019531, -494.585205078125, -493.4822998046875], "reading_logprob_std": 3.043922219950162, "tokenization_path_logprob_std": 74.503639878332, "trajectory_balance_rmse": 7046.965693857133}818{"canonical_length": 175, "canonical_nll": 1.3617583356584821, "generation_rouge_l_f1": 0.2512820512820513, "generation_token_f1": 0.46153846153846156, "prompt_id": "55985253cc61ade7", "reading_canonical_log_prob": -238.30770874023438, "reading_distances": [25, 30, 26, 24], "reading_log_probs": [-239.7417755126953, -242.4608154296875, -240.9292449951172, -240.08706665039062], "reading_logprob_std": 1.2114778660531647, "tokenization_path_logprob_std": 190.18376610865292, "trajectory_balance_rmse": 3337.8791472996695}819{"canonical_length": 134, "canonical_nll": 1.5136424961374766, "generation_rouge_l_f1": 0.33187772925764186, "generation_token_f1": 0.5327510917030569, "prompt_id": "e4db8e939e6495ed", "reading_canonical_log_prob": -202.82809448242188, "reading_distances": [17, 15, 20, 15], "reading_log_probs": [-201.1246795654297, -201.78672790527344, -201.01446533203125, -202.10977172851562], "reading_logprob_std": 0.5260944661094593, "tokenization_path_logprob_std": 134.37649045374528, "trajectory_balance_rmse": 3167.4392093216643}820{"canonical_length": 36, "canonical_nll": 0.6898475223117404, "generation_rouge_l_f1": 0.223463687150838, "generation_token_f1": 0.24581005586592178, "prompt_id": "8ba984bcebc3168b", "reading_canonical_log_prob": -24.834510803222656, "reading_distances": [16, 17, 15, 15], "reading_log_probs": [-29.666818618774414, -33.732818603515625, -28.090457916259766, -29.337844848632812], "reading_logprob_std": 2.446655800476155, "tokenization_path_logprob_std": 32.905988424396575, "trajectory_balance_rmse": 691.2876366580431}821{"canonical_length": 4, "canonical_nll": 2.193131446838379, "generation_rouge_l_f1": 0.015267175572519083, "generation_token_f1": 0.015267175572519083, "prompt_id": "fe48fe173961fdd5", "reading_canonical_log_prob": -8.772525787353516, "reading_distances": [18, 27, 24, 23], "reading_log_probs": [-9.643614768981934, -10.573003768920898, -10.512044906616211, -10.274267196655273], "reading_logprob_std": 0.4247700420566875, "tokenization_path_logprob_std": 18.138146582266973, "trajectory_balance_rmse": 41.2030502299409}822{"canonical_length": 267, "canonical_nll": 1.8187271861101357, "generation_rouge_l_f1": 0.14122137404580154, "generation_token_f1": 0.25190839694656486, "prompt_id": "51d00b55119644b0", "reading_canonical_log_prob": -485.60015869140625, "reading_distances": [25, 28, 26, 25], "reading_log_probs": [-505.8533935546875, -519.8877563476562, -509.5389404296875, -507.77337646484375], "reading_logprob_std": 6.266354177342877, "tokenization_path_logprob_std": 98.56474023745216, "trajectory_balance_rmse": 6028.586184642286}823{"canonical_length": 7, "canonical_nll": 0.13014593294688634, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "e708a1adc69b9f0e", "reading_canonical_log_prob": -0.9110215306282043, "reading_distances": [22, 28, 24, 25], "reading_log_probs": [-1.2398120164871216, -1.3884541988372803, -1.3293638229370117, -1.3971251249313354], "reading_logprob_std": 0.07246832068334351, "tokenization_path_logprob_std": 9.80763571879683, "trajectory_balance_rmse": 106.71557753660173}824{"canonical_length": 22, "canonical_nll": 1.0067856528542258, "generation_rouge_l_f1": 0.11764705882352941, "generation_token_f1": 0.11764705882352941, "prompt_id": "d7a89d7235636e3d", "reading_canonical_log_prob": -22.14928436279297, "reading_distances": [14, 16, 15, 14], "reading_log_probs": [-23.56981658935547, -23.482501983642578, -27.598146438598633, -29.120601654052734], "reading_logprob_std": 2.8590623041085443, "tokenization_path_logprob_std": 59.32606599897191, "trajectory_balance_rmse": 504.16060746937205}825{"canonical_length": 12, "canonical_nll": 0.6681664784749349, "generation_rouge_l_f1": 0.8571428571428572, "generation_token_f1": 0.8571428571428572, "prompt_id": "17e2c9ee9612ad40", "reading_canonical_log_prob": -8.017997741699219, "reading_distances": [21, 22, 23, 29], "reading_log_probs": [-8.901680946350098, -9.333252906799316, -8.853804588317871, -8.989473342895508], "reading_logprob_std": 0.21654774246818795, "tokenization_path_logprob_std": 43.48478238973027, "trajectory_balance_rmse": 264.85347372146725}826{"canonical_length": 324, "canonical_nll": 1.141529224537037, "generation_rouge_l_f1": 0.26436781609195403, "generation_token_f1": 0.5287356321839081, "prompt_id": "36541d1f10617d25", "reading_canonical_log_prob": -369.85546875, "reading_distances": [11, 11, 15, 17], "reading_log_probs": [-383.68853759765625, -388.9610595703125, -377.5042724609375, -381.48187255859375], "reading_logprob_std": 4.777826437397007, "tokenization_path_logprob_std": 223.92631034401495, "trajectory_balance_rmse": 6689.168488037935}827{"canonical_length": 315, "canonical_nll": 1.0976046123201886, "generation_rouge_l_f1": 0.253968253968254, "generation_token_f1": 0.5255731922398589, "prompt_id": "e15e866d282a743f", "reading_canonical_log_prob": -345.7454528808594, "reading_distances": [24, 27, 30, 33], "reading_log_probs": [-350.5931091308594, -351.87164306640625, -351.21319580078125, -352.3629150390625], "reading_logprob_std": 0.7718011798638672, "tokenization_path_logprob_std": 305.11248601956424, "trajectory_balance_rmse": 6006.29193595124}828{"canonical_length": 10, "canonical_nll": 1.5683156967163085, "generation_rouge_l_f1": 0.3636363636363636, "generation_token_f1": 0.5454545454545454, "prompt_id": "718643fa581e9e44", "reading_canonical_log_prob": -15.683156967163086, "reading_distances": [18, 21, 22, 21], "reading_log_probs": [-21.220699310302734, -21.597217559814453, -20.156267166137695, -20.145727157592773], "reading_logprob_std": 0.7423846951664179, "tokenization_path_logprob_std": 28.457333628382386, "trajectory_balance_rmse": 135.94451714300354}829{"canonical_length": 159, "canonical_nll": 0.9840579242826258, "generation_rouge_l_f1": 0.3333333333333333, "generation_token_f1": 0.5211267605633801, "prompt_id": "482bbea97f7ae0fa", "reading_canonical_log_prob": -156.4652099609375, "reading_distances": [12, 10, 15, 12], "reading_log_probs": [-160.15792846679688, -161.8282012939453, -161.21493530273438, -160.20840454101562], "reading_logprob_std": 0.812535305564277, "tokenization_path_logprob_std": 175.74305161722054, "trajectory_balance_rmse": 3346.6067391789584}830{"canonical_length": 11, "canonical_nll": 1.0646774985573508, "generation_rouge_l_f1": 0.20224719101123595, "generation_token_f1": 0.20224719101123595, "prompt_id": "f9bf285092a25c3b", "reading_canonical_log_prob": -11.71145248413086, "reading_distances": [47, 46, 43, 42], "reading_log_probs": [-13.385721206665039, -14.370804786682129, -13.436212539672852, -14.136367797851562], "reading_logprob_std": 0.4962400372337657, "tokenization_path_logprob_std": 9.2156407325003, "trajectory_balance_rmse": 223.89370101593315}831{"canonical_length": 150, "canonical_nll": 0.9897731526692708, "generation_rouge_l_f1": 0.34803921568627455, "generation_token_f1": 0.5245098039215687, "prompt_id": "9d4eb4020bd4213f", "reading_canonical_log_prob": -148.46597290039062, "reading_distances": [31, 18, 28, 24], "reading_log_probs": [-150.57952880859375, -147.91574096679688, -151.71363830566406, -151.60458374023438], "reading_logprob_std": 1.7672055561509223, "tokenization_path_logprob_std": 105.29011439636068, "trajectory_balance_rmse": 2819.1132320430484}832{"canonical_length": 134, "canonical_nll": 1.0873967640435518, "generation_rouge_l_f1": 0.2464985994397759, "generation_token_f1": 0.39775910364145656, "prompt_id": "13778e3d12464b89", "reading_canonical_log_prob": -145.71116638183594, "reading_distances": [18, 19, 21, 19], "reading_log_probs": [-149.9272003173828, -149.7200164794922, -150.44085693359375, -148.33950805664062], "reading_logprob_std": 0.8976179841385111, "tokenization_path_logprob_std": 99.06321453128608, "trajectory_balance_rmse": 1564.4964999692559}833{"canonical_length": 395, "canonical_nll": 1.4018304366099683, "generation_rouge_l_f1": 0.2954545454545454, "generation_token_f1": 0.45454545454545453, "prompt_id": "513a1cb695ee9eb0", "reading_canonical_log_prob": -553.7230224609375, "reading_distances": [25, 24, 16, 27], "reading_log_probs": [-555.7860717773438, -557.2012939453125, -566.1708984375, -557.5628051757812], "reading_logprob_std": 4.72306419634717, "tokenization_path_logprob_std": 376.5329088725188, "trajectory_balance_rmse": 5712.643407105133}834{"canonical_length": 193, "canonical_nll": 1.0578632255909974, "generation_rouge_l_f1": 0.3162583518930957, "generation_token_f1": 0.49888641425389757, "prompt_id": "444facd57b3cdbe1", "reading_canonical_log_prob": -204.1676025390625, "reading_distances": [24, 22, 22, 14], "reading_log_probs": [-220.2838897705078, -218.2888641357422, -221.05712890625, -224.63580322265625], "reading_logprob_std": 2.6500159152815126, "tokenization_path_logprob_std": 260.36435293767516, "trajectory_balance_rmse": 3779.3477763296964}835{"canonical_length": 226, "canonical_nll": 1.177513055041828, "generation_rouge_l_f1": 0.29508196721311475, "generation_token_f1": 0.6434426229508196, "prompt_id": "5870502ae5fa3ae7", "reading_canonical_log_prob": -266.1179504394531, "reading_distances": [27, 20, 23, 23], "reading_log_probs": [-272.1998291015625, -271.75042724609375, -274.89825439453125, -271.8232421875], "reading_logprob_std": 1.4998637893377382, "tokenization_path_logprob_std": 187.62217589225108, "trajectory_balance_rmse": 2704.823204789533}836{"canonical_length": 17, "canonical_nll": 0.9378174052518957, "generation_rouge_l_f1": 0.8387096774193549, "generation_token_f1": 0.9032258064516129, "prompt_id": "8c61a7a56532124f", "reading_canonical_log_prob": -15.942895889282227, "reading_distances": [29, 40, 31, 29], "reading_log_probs": [-14.337903022766113, -16.28921890258789, -12.905845642089844, -12.507465362548828], "reading_logprob_std": 1.7106652053347084, "tokenization_path_logprob_std": 69.15641887970315, "trajectory_balance_rmse": 365.3807598084602}837{"canonical_length": 12, "canonical_nll": 0.7215415636698405, "generation_rouge_l_f1": 0.3333333333333333, "generation_token_f1": 0.3333333333333333, "prompt_id": "fb0ad86c22711d8d", "reading_canonical_log_prob": -8.658498764038086, "reading_distances": [9, 12, 14, 16], "reading_log_probs": [-7.894575595855713, -5.668924331665039, -6.937466621398926, -10.403553009033203], "reading_logprob_std": 2.004249899142611, "tokenization_path_logprob_std": 40.887298649797735, "trajectory_balance_rmse": 185.2872496507928}838{"canonical_length": 219, "canonical_nll": 1.4475585380101312, "generation_rouge_l_f1": 0.21848739495798317, "generation_token_f1": 0.4327731092436975, "prompt_id": "a18ee8ea214b4b09", "reading_canonical_log_prob": -317.01531982421875, "reading_distances": [36, 42, 52, 45], "reading_log_probs": [-335.6184997558594, -336.5424499511719, -339.2633056640625, -339.70294189453125], "reading_logprob_std": 2.0084406639148096, "tokenization_path_logprob_std": 127.72705820100865, "trajectory_balance_rmse": 5166.572853765273}839{"canonical_length": 283, "canonical_nll": 1.3194897115862412, "generation_rouge_l_f1": 0.20357142857142857, "generation_token_f1": 0.43214285714285716, "prompt_id": "6933d659ab69515c", "reading_canonical_log_prob": -373.41558837890625, "reading_distances": [32, 27, 25, 23], "reading_log_probs": [-383.59991455078125, -391.23809814453125, -388.2349853515625, -377.9000244140625], "reading_logprob_std": 5.816992295401424, "tokenization_path_logprob_std": 348.7772447965443, "trajectory_balance_rmse": 5318.692795730042}840{"canonical_length": 49, "canonical_nll": 1.3043105456293846, "generation_rouge_l_f1": 0.49438202247191015, "generation_token_f1": 0.7191011235955055, "prompt_id": "e748f45af7778fcd", "reading_canonical_log_prob": -63.911216735839844, "reading_distances": [21, 31, 38, 33], "reading_log_probs": [-68.34943389892578, -64.98450469970703, -67.01899719238281, -66.55636596679688], "reading_logprob_std": 1.388399502776471, "tokenization_path_logprob_std": 57.350781649057275, "trajectory_balance_rmse": 856.8262411885696}841{"canonical_length": 8, "canonical_nll": 1.3088672161102295, "generation_rouge_l_f1": 0.761904761904762, "generation_token_f1": 0.761904761904762, "prompt_id": "3f21812c287e05df", "reading_canonical_log_prob": -10.470937728881836, "reading_distances": [41, 27, 42, 32], "reading_log_probs": [-16.283123016357422, -12.988566398620605, -14.994033813476562, -12.153190612792969], "reading_logprob_std": 1.8788812780689361, "tokenization_path_logprob_std": 19.539500642146347, "trajectory_balance_rmse": 179.83941059731075}842{"canonical_length": 38, "canonical_nll": 1.278693952058491, "generation_rouge_l_f1": 0.256, "generation_token_f1": 0.304, "prompt_id": "cc2b221229d6b7ba", "reading_canonical_log_prob": -48.590370178222656, "reading_distances": [27, 19, 21, 26], "reading_log_probs": [-59.72818374633789, -60.835655212402344, -63.493473052978516, -65.73988342285156], "reading_logprob_std": 2.7034890908966736, "tokenization_path_logprob_std": 192.52850852646887, "trajectory_balance_rmse": 913.0121549248048}843{"canonical_length": 57, "canonical_nll": 1.1338691042180646, "generation_rouge_l_f1": 0.2644628099173554, "generation_token_f1": 0.3305785123966943, "prompt_id": "6c3587030700c60e", "reading_canonical_log_prob": -64.63053894042969, "reading_distances": [33, 25, 30, 23], "reading_log_probs": [-69.17877197265625, -69.37555694580078, -69.42037963867188, -68.00503540039062], "reading_logprob_std": 0.668224876324858, "tokenization_path_logprob_std": 36.98872899063368, "trajectory_balance_rmse": 1033.829572632409}844{"canonical_length": 108, "canonical_nll": 1.7251418784812644, "generation_rouge_l_f1": 0.24786324786324784, "generation_token_f1": 0.5042735042735043, "prompt_id": "3f751ca73a73cf7d", "reading_canonical_log_prob": -186.31532287597656, "reading_distances": [26, 21, 13, 23], "reading_log_probs": [-189.55416870117188, -190.0286865234375, -190.04196166992188, -186.99237060546875], "reading_logprob_std": 1.4590326381651744, "tokenization_path_logprob_std": 81.54541134154982, "trajectory_balance_rmse": 1707.3288171136992}845{"canonical_length": 77, "canonical_nll": 0.7191560918634589, "generation_rouge_l_f1": 0.2333333333333333, "generation_token_f1": 0.32, "prompt_id": "7c7baba7466fe6f6", "reading_canonical_log_prob": -55.37501907348633, "reading_distances": [8, 7, 7, 11], "reading_log_probs": [-58.08014678955078, -57.040191650390625, -60.563838958740234, -61.86246109008789], "reading_logprob_std": 2.2157189821348005, "tokenization_path_logprob_std": 160.63669289127293, "trajectory_balance_rmse": 1562.6886517909204}846{"canonical_length": 294, "canonical_nll": 0.7391870200228529, "generation_rouge_l_f1": 0.2780748663101604, "generation_token_f1": 0.5525846702317291, "prompt_id": "fd3d2e71f4e4243f", "reading_canonical_log_prob": -217.32098388671875, "reading_distances": [15, 19, 17, 20], "reading_log_probs": [-223.82717895507812, -225.5134735107422, -222.4161834716797, -224.03643798828125], "reading_logprob_std": 1.2674893538298837, "tokenization_path_logprob_std": 225.45011464130178, "trajectory_balance_rmse": 4965.719413720546}847{"canonical_length": 399, "canonical_nll": 1.033715652045152, "generation_rouge_l_f1": 0.19783616692426587, "generation_token_f1": 0.4482225656877898, "prompt_id": "5bc59064e72854cb", "reading_canonical_log_prob": -412.4525451660156, "reading_distances": [22, 20, 21, 24], "reading_log_probs": [-418.8797302246094, -417.90740966796875, -418.41729736328125, -414.3267822265625], "reading_logprob_std": 2.075688225399166, "tokenization_path_logprob_std": 119.52332950022566, "trajectory_balance_rmse": 7577.2530251044645}848{"canonical_length": 280, "canonical_nll": 1.2095714024135045, "generation_rouge_l_f1": 0.21268656716417908, "generation_token_f1": 0.38059701492537307, "prompt_id": "cd25f5705dfa5cbb", "reading_canonical_log_prob": -338.67999267578125, "reading_distances": [22, 18, 25, 28], "reading_log_probs": [-361.80755615234375, -355.31890869140625, -366.91241455078125, -361.13702392578125], "reading_logprob_std": 4.74540759112125, "tokenization_path_logprob_std": 151.12950835189824, "trajectory_balance_rmse": 6346.010432409051}849{"canonical_length": 365, "canonical_nll": 0.8490466810252568, "generation_rouge_l_f1": 0.20615384615384616, "generation_token_f1": 0.44, "prompt_id": "ee3200ffbd24e2bc", "reading_canonical_log_prob": -309.90203857421875, "reading_distances": [26, 28, 28, 31], "reading_log_probs": [-328.183349609375, -322.7317199707031, -328.5569152832031, -328.1929626464844], "reading_logprob_std": 2.7950915750459115, "tokenization_path_logprob_std": 194.4921203904816, "trajectory_balance_rmse": 4975.676421598184}850{"canonical_length": 115, "canonical_nll": 2.162905485733696, "generation_rouge_l_f1": 0.2588235294117647, "generation_token_f1": 0.38823529411764707, "prompt_id": "4180e393ac001d5f", "reading_canonical_log_prob": -248.734130859375, "reading_distances": [23, 25, 20, 17], "reading_log_probs": [-259.123779296875, -272.8232421875, -301.66107177734375, -263.327392578125], "reading_logprob_std": 19.16165753939406, "tokenization_path_logprob_std": 120.12492312670933, "trajectory_balance_rmse": 2688.6751832963478}851{"canonical_length": 9, "canonical_nll": 0.9737368689643012, "generation_rouge_l_f1": 0.8571428571428571, "generation_token_f1": 0.8571428571428571, "prompt_id": "79896777ff1e0459", "reading_canonical_log_prob": -8.763631820678711, "reading_distances": [36, 51, 52, 45], "reading_log_probs": [-7.619384288787842, -10.602462768554688, -9.392345428466797, -6.151922225952148], "reading_logprob_std": 1.9572010404879923, "tokenization_path_logprob_std": 40.39470100646328, "trajectory_balance_rmse": 192.87166615662952}852{"canonical_length": 147, "canonical_nll": 1.5321315843231824, "generation_rouge_l_f1": 0.2287104622871046, "generation_token_f1": 0.43795620437956206, "prompt_id": "4be1dda9c6193970", "reading_canonical_log_prob": -225.2233428955078, "reading_distances": [43, 40, 45, 43], "reading_log_probs": [-236.80697631835938, -233.52162170410156, -233.09732055664062, -234.94740295410156], "reading_logprob_std": 1.6745348215479223, "tokenization_path_logprob_std": 276.8610828288928, "trajectory_balance_rmse": 4056.380423102885}853{"canonical_length": 9, "canonical_nll": 1.25760989718967, "generation_rouge_l_f1": 0.09090909090909091, "generation_token_f1": 0.09090909090909091, "prompt_id": "e275be5afb765439", "reading_canonical_log_prob": -11.318489074707031, "reading_distances": [32, 31, 30, 24], "reading_log_probs": [-32.66938400268555, -27.178266525268555, -30.576570510864258, -39.58979797363281], "reading_logprob_std": 5.238139150813973, "tokenization_path_logprob_std": 26.244132725553214, "trajectory_balance_rmse": 121.18735492312516}854{"canonical_length": 8, "canonical_nll": 0.9828038215637207, "generation_rouge_l_f1": 0.05, "generation_token_f1": 0.05, "prompt_id": "b4100d452af762d4", "reading_canonical_log_prob": -7.862430572509766, "reading_distances": [17, 18, 23, 23], "reading_log_probs": [-9.705924987792969, -13.875401496887207, -10.960836410522461, -18.471500396728516], "reading_logprob_std": 3.8925520499837534, "tokenization_path_logprob_std": 12.985389819976483, "trajectory_balance_rmse": 113.64728320799263}855{"canonical_length": 65, "canonical_nll": 0.5099753159743089, "generation_rouge_l_f1": 0.25641025641025644, "generation_token_f1": 0.336996336996337, "prompt_id": "660009a4d953661a", "reading_canonical_log_prob": -33.14839553833008, "reading_distances": [24, 31, 30, 34], "reading_log_probs": [-45.66142654418945, -45.652374267578125, -39.883853912353516, -48.037567138671875], "reading_logprob_std": 3.4698040660652443, "tokenization_path_logprob_std": 73.82041531142234, "trajectory_balance_rmse": 781.123413683086}856{"canonical_length": 245, "canonical_nll": 0.9587403589365434, "generation_rouge_l_f1": 0.4444444444444444, "generation_token_f1": 0.6711111111111111, "prompt_id": "9c1c5be112145501", "reading_canonical_log_prob": -234.89138793945312, "reading_distances": [26, 30, 16, 16], "reading_log_probs": [-235.75201416015625, -236.8090057373047, -233.31527709960938, -234.32931518554688], "reading_logprob_std": 1.5400830425343148, "tokenization_path_logprob_std": 72.5930666666251, "trajectory_balance_rmse": 4338.400885264315}857{"canonical_length": 41, "canonical_nll": 0.7183907671672541, "generation_rouge_l_f1": 0.1384083044982699, "generation_token_f1": 0.17301038062283738, "prompt_id": "8a1926c471e546ee", "reading_canonical_log_prob": -29.454021453857422, "reading_distances": [21, 15, 27, 24], "reading_log_probs": [-27.864105224609375, -27.490419387817383, -29.284351348876953, -24.256643295288086], "reading_logprob_std": 2.1237519805614844, "tokenization_path_logprob_std": 36.63323141257339, "trajectory_balance_rmse": 458.4265028267761}858{"canonical_length": 311, "canonical_nll": 1.1911226622161375, "generation_rouge_l_f1": 0.23488773747841105, "generation_token_f1": 0.5077720207253885, "prompt_id": "a1e93e4a3a2191cd", "reading_canonical_log_prob": -370.43914794921875, "reading_distances": [37, 35, 36, 42], "reading_log_probs": [-381.6045227050781, -383.43463134765625, -378.4156494140625, -381.2262878417969], "reading_logprob_std": 2.07420509237198, "tokenization_path_logprob_std": 232.26652747525404, "trajectory_balance_rmse": 6360.184871421859}859{"canonical_length": 130, "canonical_nll": 1.5403663048377403, "generation_rouge_l_f1": 0.21025641025641023, "generation_token_f1": 0.37948717948717947, "prompt_id": "c4f0abd89b25e9b2", "reading_canonical_log_prob": -200.24761962890625, "reading_distances": [22, 21, 18, 21], "reading_log_probs": [-203.114013671875, -201.79367065429688, -204.56982421875, -204.11135864257812], "reading_logprob_std": 1.2297114814740828, "tokenization_path_logprob_std": 203.427994476658, "trajectory_balance_rmse": 2993.5454818325097}860{"canonical_length": 226, "canonical_nll": 1.1280028756740874, "generation_rouge_l_f1": 0.26552462526766596, "generation_token_f1": 0.4411134903640257, "prompt_id": "8d709bdbfc4890e5", "reading_canonical_log_prob": -254.92864990234375, "reading_distances": [23, 27, 30, 20], "reading_log_probs": [-253.0482177734375, -253.92379760742188, -253.40103149414062, -249.25308227539062], "reading_logprob_std": 2.1328486401041156, "tokenization_path_logprob_std": 125.50199346523257, "trajectory_balance_rmse": 3872.12348552073}861{"canonical_length": 410, "canonical_nll": 1.225685975609756, "generation_rouge_l_f1": 0.2524271844660194, "generation_token_f1": 0.4304207119741101, "prompt_id": "fbcfdeca84611d04", "reading_canonical_log_prob": -502.53125, "reading_distances": [30, 19, 29, 28], "reading_log_probs": [-532.4943237304688, -525.622802734375, -542.24365234375, -537.176513671875], "reading_logprob_std": 7.0687509760782525, "tokenization_path_logprob_std": 222.6698966263762, "trajectory_balance_rmse": 7760.729760695796}862{"canonical_length": 45, "canonical_nll": 1.3341722276475694, "generation_rouge_l_f1": 0.23703703703703702, "generation_token_f1": 0.4148148148148148, "prompt_id": "8c6898d598b2732f", "reading_canonical_log_prob": -60.037750244140625, "reading_distances": [31, 31, 33, 26], "reading_log_probs": [-68.78485870361328, -69.61959838867188, -69.13404083251953, -62.320655822753906], "reading_logprob_std": 3.4464617290548216, "tokenization_path_logprob_std": 15.305556877195945, "trajectory_balance_rmse": 964.618390431262}863{"canonical_length": 93, "canonical_nll": 0.5513881150112357, "generation_rouge_l_f1": 0.526829268292683, "generation_token_f1": 0.673170731707317, "prompt_id": "f5259298cfe12698", "reading_canonical_log_prob": -51.27909469604492, "reading_distances": [19, 16, 17, 21], "reading_log_probs": [-55.47312927246094, -55.06672668457031, -56.57714080810547, -54.97869110107422], "reading_logprob_std": 0.7344281332012058, "tokenization_path_logprob_std": 30.73671536511653, "trajectory_balance_rmse": 1070.1745475951643}864{"canonical_length": 183, "canonical_nll": 0.39290894054975667, "generation_rouge_l_f1": 0.32628398791540786, "generation_token_f1": 0.42900302114803623, "prompt_id": "af8c4385ce2b0480", "reading_canonical_log_prob": -71.90233612060547, "reading_distances": [31, 31, 37, 35], "reading_log_probs": [-81.27145385742188, -77.3173828125, -76.3974609375, -79.36418914794922], "reading_logprob_std": 2.176872607460677, "tokenization_path_logprob_std": 57.109753945689825, "trajectory_balance_rmse": 2653.349577543547}865{"canonical_length": 164, "canonical_nll": 1.6390306426257621, "generation_rouge_l_f1": 0.18401937046004843, "generation_token_f1": 0.3922518159806295, "prompt_id": "8db3810d387bcbf4", "reading_canonical_log_prob": -268.801025390625, "reading_distances": [24, 22, 19, 25], "reading_log_probs": [-278.8796081542969, -277.28472900390625, -273.12408447265625, -276.2156982421875], "reading_logprob_std": 2.4286136950770234, "tokenization_path_logprob_std": 353.3839967807177, "trajectory_balance_rmse": 3756.2683147037674}866{"canonical_length": 170, "canonical_nll": 0.8208959242876839, "generation_rouge_l_f1": 0.33971291866028713, "generation_token_f1": 0.5454545454545455, "prompt_id": "d92a9462a284aee7", "reading_canonical_log_prob": -139.55230712890625, "reading_distances": [12, 16, 13, 14], "reading_log_probs": [-143.1276092529297, -141.50845336914062, -144.9711456298828, -142.15895080566406], "reading_logprob_std": 1.5077692040867134, "tokenization_path_logprob_std": 95.50525094235712, "trajectory_balance_rmse": 3319.8080594576672}867{"canonical_length": 76, "canonical_nll": 0.9126317877518503, "generation_rouge_l_f1": 0.3253968253968254, "generation_token_f1": 0.46031746031746035, "prompt_id": "32c20ba8cc10f798", "reading_canonical_log_prob": -69.36001586914062, "reading_distances": [33, 33, 28, 28], "reading_log_probs": [-72.36796569824219, -73.3801498413086, -71.91822814941406, -69.75914001464844], "reading_logprob_std": 1.5259789432078863, "tokenization_path_logprob_std": 89.0187604756268, "trajectory_balance_rmse": 1783.5392338438783}868{"canonical_length": 21, "canonical_nll": 0.7110481262207031, "generation_rouge_l_f1": 0.26666666666666666, "generation_token_f1": 0.26666666666666666, "prompt_id": "2e575375b1e838f0", "reading_canonical_log_prob": -14.932010650634766, "reading_distances": [25, 23, 27, 29], "reading_log_probs": [-16.13966178894043, -15.106914520263672, -15.004173278808594, -15.3798189163208], "reading_logprob_std": 0.5131117881560637, "tokenization_path_logprob_std": 82.89785741174292, "trajectory_balance_rmse": 246.7809156177762}869{"canonical_length": 9, "canonical_nll": 0.24837761455112034, "generation_rouge_l_f1": 0.5714285714285715, "generation_token_f1": 0.7142857142857143, "prompt_id": "f19b5cf2d17aa856", "reading_canonical_log_prob": -2.235398530960083, "reading_distances": [37, 36, 30, 42], "reading_log_probs": [-1.9100587368011475, -5.107854843139648, -6.079684257507324, -3.87357234954834], "reading_logprob_std": 1.7981911759491374, "tokenization_path_logprob_std": 21.05950808326975, "trajectory_balance_rmse": 178.9866521355175}870{"canonical_length": 3, "canonical_nll": 1.4549781481424968, "generation_rouge_l_f1": 0.6666666666666666, "generation_token_f1": 0.6666666666666666, "prompt_id": "745badf44c39e597", "reading_canonical_log_prob": -4.36493444442749, "reading_distances": [24, 26, 19, 20], "reading_log_probs": [-5.5652570724487305, -5.451928615570068, -4.473116397857666, -5.7043609619140625], "reading_logprob_std": 0.5599645158282606, "tokenization_path_logprob_std": 8.015248978065562, "trajectory_balance_rmse": 39.35558906022294}871{"canonical_length": 272, "canonical_nll": 1.300362979664522, "generation_rouge_l_f1": 0.25918762088974856, "generation_token_f1": 0.5145067698259187, "prompt_id": "8f529e5554e9c902", "reading_canonical_log_prob": -353.69873046875, "reading_distances": [21, 22, 18, 19], "reading_log_probs": [-352.05499267578125, -354.66387939453125, -355.78033447265625, -354.33740234375], "reading_logprob_std": 1.5633555695479477, "tokenization_path_logprob_std": 100.743335025795, "trajectory_balance_rmse": 5571.6465905641835}872{"canonical_length": 324, "canonical_nll": 0.9853743564935378, "generation_rouge_l_f1": 0.22468354430379747, "generation_token_f1": 0.5316455696202531, "prompt_id": "642f20171ea337c8", "reading_canonical_log_prob": -319.26129150390625, "reading_distances": [26, 21, 25, 27], "reading_log_probs": [-314.447021484375, -318.1177978515625, -318.2789001464844, -318.70574951171875], "reading_logprob_std": 1.9758647709004353, "tokenization_path_logprob_std": 201.92180918098873, "trajectory_balance_rmse": 4805.040453948712}873{"canonical_length": 14, "canonical_nll": 1.4311210087367467, "generation_rouge_l_f1": 0.4210526315789474, "generation_token_f1": 0.4210526315789474, "prompt_id": "c076cc444fb03f81", "reading_canonical_log_prob": -20.035694122314453, "reading_distances": [196, 181, 192, 186], "reading_log_probs": [-23.67742347717285, -23.087646484375, -23.064655303955078, -22.015056610107422], "reading_logprob_std": 0.691581544176366, "tokenization_path_logprob_std": 57.357119301658834, "trajectory_balance_rmse": 384.6351526668144}874{"canonical_length": 183, "canonical_nll": 1.2081358862704918, "generation_rouge_l_f1": 0.2393162393162393, "generation_token_f1": 0.2564102564102564, "prompt_id": "8082ddc02c38399e", "reading_canonical_log_prob": -221.0888671875, "reading_distances": [18, 26, 25, 25], "reading_log_probs": [-223.43797302246094, -220.47535705566406, -220.34934997558594, -221.83758544921875], "reading_logprob_std": 1.442345164492267, "tokenization_path_logprob_std": 153.28731705811623, "trajectory_balance_rmse": 1651.133480452621}875{"canonical_length": 7, "canonical_nll": 0.41781394822256906, "generation_rouge_l_f1": 0.08695652173913045, "generation_token_f1": 0.08695652173913045, "prompt_id": "3af881f54016c0d1", "reading_canonical_log_prob": -2.9246976375579834, "reading_distances": [35, 38, 36, 40], "reading_log_probs": [-6.290343284606934, -4.4900360107421875, -6.103565216064453, -6.717667102813721], "reading_logprob_std": 0.9747450897083437, "tokenization_path_logprob_std": 9.16936764259556, "trajectory_balance_rmse": 78.42960915175814}876{"canonical_length": 241, "canonical_nll": 0.6713363204259596, "generation_rouge_l_f1": 0.3140495867768595, "generation_token_f1": 0.5041322314049587, "prompt_id": "edd1c37dcb8ffb27", "reading_canonical_log_prob": -161.79205322265625, "reading_distances": [14, 14, 15, 16], "reading_log_probs": [-172.63739013671875, -173.55775451660156, -167.75338745117188, -164.91912841796875], "reading_logprob_std": 4.08882296585079, "tokenization_path_logprob_std": 72.37564053204666, "trajectory_balance_rmse": 3616.1753159035966}877{"canonical_length": 277, "canonical_nll": 1.624097693267712, "generation_rouge_l_f1": 0.18691588785046728, "generation_token_f1": 0.3588785046728972, "prompt_id": "b7a747a09567587c", "reading_canonical_log_prob": -449.87506103515625, "reading_distances": [36, 29, 37, 27], "reading_log_probs": [-449.76812744140625, -449.5173034667969, -450.6380920410156, -450.38568115234375], "reading_logprob_std": 0.522420787159817, "tokenization_path_logprob_std": 351.1199418229044, "trajectory_balance_rmse": 5752.612918156292}878{"canonical_length": 203, "canonical_nll": 1.5134220217249077, "generation_rouge_l_f1": 0.20085470085470086, "generation_token_f1": 0.4444444444444445, "prompt_id": "85ba85bd1384cec6", "reading_canonical_log_prob": -307.22467041015625, "reading_distances": [19, 22, 16, 16], "reading_log_probs": [-306.62982177734375, -305.5265197753906, -305.51495361328125, -310.0132141113281], "reading_logprob_std": 2.1266651782573147, "tokenization_path_logprob_std": 213.3961695417484, "trajectory_balance_rmse": 4731.57739869219}879{"canonical_length": 25, "canonical_nll": 0.54549072265625, "generation_rouge_l_f1": 0.7272727272727272, "generation_token_f1": 0.9090909090909091, "prompt_id": "ff4cb1405a9ba9bb", "reading_canonical_log_prob": -13.63726806640625, "reading_distances": [16, 15, 16, 22], "reading_log_probs": [-15.197519302368164, -13.963701248168945, -14.921001434326172, -15.760229110717773], "reading_logprob_std": 0.7507579809088422, "tokenization_path_logprob_std": 82.63690746949143, "trajectory_balance_rmse": 343.14950617364354}880{"canonical_length": 103, "canonical_nll": 1.501685281401699, "generation_rouge_l_f1": 0.3270440251572327, "generation_token_f1": 0.4150943396226416, "prompt_id": "ea60f11b47ae285e", "reading_canonical_log_prob": -154.673583984375, "reading_distances": [22, 29, 23, 29], "reading_log_probs": [-156.52053833007812, -154.03375244140625, -155.1604461669922, -155.42994689941406], "reading_logprob_std": 1.0212236441905373, "tokenization_path_logprob_std": 159.41199403731218, "trajectory_balance_rmse": 2609.8860297938586}881{"canonical_length": 22, "canonical_nll": 0.41535429521040484, "generation_rouge_l_f1": 0.22047244094488186, "generation_token_f1": 0.31496062992125984, "prompt_id": "6245bcac7beebc0c", "reading_canonical_log_prob": -9.137794494628906, "reading_distances": [25, 30, 21, 25], "reading_log_probs": [-9.762015342712402, -8.728254318237305, -9.372873306274414, -8.392000198364258], "reading_logprob_std": 0.6183140654689633, "tokenization_path_logprob_std": 76.38195270615446, "trajectory_balance_rmse": 489.48505083169624}882{"canonical_length": 156, "canonical_nll": 1.4283115680401142, "generation_rouge_l_f1": 0.1898148148148148, "generation_token_f1": 0.3796296296296296, "prompt_id": "5016da9f4e182301", "reading_canonical_log_prob": -222.8166046142578, "reading_distances": [14, 16, 20, 18], "reading_log_probs": [-223.14976501464844, -225.9716033935547, -225.560302734375, -228.70565795898438], "reading_logprob_std": 2.2763071781539193, "tokenization_path_logprob_std": 161.414449432669, "trajectory_balance_rmse": 3973.2753869909793}883{"canonical_length": 103, "canonical_nll": 1.6219171320350425, "generation_rouge_l_f1": 0.23676012461059193, "generation_token_f1": 0.41744548286604366, "prompt_id": "d6b153dd541c60bc", "reading_canonical_log_prob": -167.05746459960938, "reading_distances": [30, 33, 31, 34], "reading_log_probs": [-166.7840576171875, -167.48472595214844, -166.2203369140625, -168.12850952148438], "reading_logprob_std": 0.8301871121023546, "tokenization_path_logprob_std": 164.72750045963738, "trajectory_balance_rmse": 1776.9748340919612}884{"canonical_length": 50, "canonical_nll": 1.1491517639160156, "generation_rouge_l_f1": 0.15172413793103445, "generation_token_f1": 0.21379310344827585, "prompt_id": "f50b16ccb1ed30ed", "reading_canonical_log_prob": -57.45758819580078, "reading_distances": [14, 17, 13, 16], "reading_log_probs": [-77.52726745605469, -72.22250366210938, -67.5192642211914, -71.00254821777344], "reading_logprob_std": 4.149444026238135, "tokenization_path_logprob_std": 59.92999354466092, "trajectory_balance_rmse": 926.1529634973518}885{"canonical_length": 26, "canonical_nll": 0.5048969342158391, "generation_rouge_l_f1": 0.21428571428571425, "generation_token_f1": 0.21428571428571425, "prompt_id": "ac38f5f5447a0903", "reading_canonical_log_prob": -13.127320289611816, "reading_distances": [35, 43, 36, 36], "reading_log_probs": [-14.847053527832031, -15.637397766113281, -15.490274429321289, -16.08060073852539], "reading_logprob_std": 0.5104389586774526, "tokenization_path_logprob_std": 105.10724148371243, "trajectory_balance_rmse": 548.5776714841809}886{"canonical_length": 10, "canonical_nll": 1.4388532638549805, "generation_rouge_l_f1": 0.07633587786259541, "generation_token_f1": 0.07633587786259541, "prompt_id": "0027f65734731d27", "reading_canonical_log_prob": -14.388532638549805, "reading_distances": [25, 23, 18, 21], "reading_log_probs": [-15.82016372680664, -18.276782989501953, -14.555577278137207, -18.472475051879883], "reading_logprob_std": 1.9126044187755578, "tokenization_path_logprob_std": 32.14800692829297, "trajectory_balance_rmse": 264.3766412540554}887{"canonical_length": 146, "canonical_nll": 1.2791259974649507, "generation_rouge_l_f1": 0.23295454545454544, "generation_token_f1": 0.38636363636363635, "prompt_id": "488c8f3398b00ac7", "reading_canonical_log_prob": -186.7523956298828, "reading_distances": [15, 14, 19, 12], "reading_log_probs": [-193.038330078125, -193.47415161132812, -192.17913818359375, -192.39712524414062], "reading_logprob_std": 0.5932859549261748, "tokenization_path_logprob_std": 150.89444112340098, "trajectory_balance_rmse": 2525.50325874749}888{"canonical_length": 4, "canonical_nll": 4.388460159301758, "generation_rouge_l_f1": 0.05, "generation_token_f1": 0.05, "prompt_id": "91adc07e055d938f", "reading_canonical_log_prob": -17.55384063720703, "reading_distances": [31, 31, 34, 30], "reading_log_probs": [-23.03875732421875, -17.892745971679688, -19.597488403320312, -21.032489776611328], "reading_logprob_std": 2.182739886762637, "tokenization_path_logprob_std": 5.431723994769332, "trajectory_balance_rmse": 30.485671253642067}889{"canonical_length": 263, "canonical_nll": 1.2739140615717086, "generation_rouge_l_f1": 0.2702702702702703, "generation_token_f1": 0.4787644787644788, "prompt_id": "5f3070e7c5d93f6b", "reading_canonical_log_prob": -335.0393981933594, "reading_distances": [19, 21, 16, 20], "reading_log_probs": [-340.2715759277344, -340.6298522949219, -339.6258850097656, -338.363037109375], "reading_logprob_std": 0.9970366656843641, "tokenization_path_logprob_std": 152.72402496749697, "trajectory_balance_rmse": 6032.149225207823}890{"canonical_length": 17, "canonical_nll": 0.13953241179971135, "generation_rouge_l_f1": 0.2857142857142857, "generation_token_f1": 0.2857142857142857, "prompt_id": "f969cf0286bd838d", "reading_canonical_log_prob": -2.3720510005950928, "reading_distances": [18, 22, 21, 15], "reading_log_probs": [-3.0198426246643066, -3.937525510787964, -2.154365062713623, -2.0867061614990234], "reading_logprob_std": 0.8694686175477884, "tokenization_path_logprob_std": 36.02856530785446, "trajectory_balance_rmse": 239.0466986053857}891{"canonical_length": 291, "canonical_nll": 0.6174261348763692, "generation_rouge_l_f1": 0.21729490022172948, "generation_token_f1": 0.4257206208425721, "prompt_id": "60292acfc5a172b6", "reading_canonical_log_prob": -179.67100524902344, "reading_distances": [25, 36, 33, 23], "reading_log_probs": [-185.04116821289062, -186.13525390625, -192.7250213623047, -186.67791748046875], "reading_logprob_std": 3.4545206508823116, "tokenization_path_logprob_std": 235.18244699933365, "trajectory_balance_rmse": 4867.299604325085}892{"canonical_length": 174, "canonical_nll": 0.853436261758037, "generation_rouge_l_f1": 0.2071197411003236, "generation_token_f1": 0.34951456310679613, "prompt_id": "2b8f8cb7903f03af", "reading_canonical_log_prob": -148.49790954589844, "reading_distances": [37, 33, 41, 42], "reading_log_probs": [-147.57537841796875, -149.0791015625, -147.2239532470703, -149.6976318359375], "reading_logprob_std": 1.184338185395829, "tokenization_path_logprob_std": 245.07498525237386, "trajectory_balance_rmse": 3428.768364284533}893{"canonical_length": 71, "canonical_nll": 1.1708915602992958, "generation_rouge_l_f1": 0.21262458471760795, "generation_token_f1": 0.292358803986711, "prompt_id": "2a8ed34f73d4ba31", "reading_canonical_log_prob": -83.13330078125, "reading_distances": [19, 14, 21, 19], "reading_log_probs": [-86.24940490722656, -85.79452514648438, -85.00385284423828, -84.38418579101562], "reading_logprob_std": 0.8284304906366989, "tokenization_path_logprob_std": 38.04931037027333, "trajectory_balance_rmse": 1173.3949540555855}894{"canonical_length": 370, "canonical_nll": 0.753729248046875, "generation_rouge_l_f1": 0.36018957345971564, "generation_token_f1": 0.5624012638230648, "prompt_id": "a2f2bd1ddbd7ac6c", "reading_canonical_log_prob": -278.87982177734375, "reading_distances": [22, 19, 21, 22], "reading_log_probs": [-279.8619384765625, -274.426025390625, -277.95086669921875, -278.7859802246094], "reading_logprob_std": 2.3539003020061435, "tokenization_path_logprob_std": 173.6581382751773, "trajectory_balance_rmse": 7053.046500625551}895{"canonical_length": 292, "canonical_nll": 1.4641619120558647, "generation_rouge_l_f1": 0.2018348623853211, "generation_token_f1": 0.3889908256880734, "prompt_id": "439bf3be7d027444", "reading_canonical_log_prob": -427.5352783203125, "reading_distances": [22, 23, 22, 22], "reading_log_probs": [-428.3792724609375, -427.9268798828125, -429.3828430175781, -427.10113525390625], "reading_logprob_std": 0.9510222915219871, "tokenization_path_logprob_std": 127.86448029122276, "trajectory_balance_rmse": 5756.77038703452}896{"canonical_length": 152, "canonical_nll": 1.0294713472065173, "generation_rouge_l_f1": 0.38421052631578945, "generation_token_f1": 0.5315789473684212, "prompt_id": "53583cdc2f7841b7", "reading_canonical_log_prob": -156.47964477539062, "reading_distances": [23, 21, 25, 25], "reading_log_probs": [-173.1112060546875, -173.56280517578125, -168.26242065429688, -169.8485107421875], "reading_logprob_std": 2.561991141497581, "tokenization_path_logprob_std": 120.57752856899675, "trajectory_balance_rmse": 2481.3827771125975}897{"canonical_length": 73, "canonical_nll": 1.114371939881207, "generation_rouge_l_f1": 0.1453488372093023, "generation_token_f1": 0.2674418604651163, "prompt_id": "11b8b10c9b5e1733", "reading_canonical_log_prob": -81.34915161132812, "reading_distances": [17, 16, 12, 20], "reading_log_probs": [-81.53487396240234, -81.43714904785156, -82.87368774414062, -81.87683868408203], "reading_logprob_std": 0.6563529806599142, "tokenization_path_logprob_std": 118.96740193002374, "trajectory_balance_rmse": 1682.6611562151766}898{"canonical_length": 17, "canonical_nll": 0.9638510311351103, "generation_rouge_l_f1": 0.10778443113772454, "generation_token_f1": 0.11976047904191615, "prompt_id": "d62e67048e61fa64", "reading_canonical_log_prob": -16.385467529296875, "reading_distances": [15, 9, 17, 13], "reading_log_probs": [-27.647624969482422, -25.519073486328125, -25.027652740478516, -23.051523208618164], "reading_logprob_std": 1.8875584746478296, "tokenization_path_logprob_std": 78.8250275364158, "trajectory_balance_rmse": 279.1737691063804}899{"canonical_length": 17, "canonical_nll": 1.196514466229607, "generation_rouge_l_f1": 0.18823529411764706, "generation_token_f1": 0.21176470588235294, "prompt_id": "15d2843edae95663", "reading_canonical_log_prob": -20.34074592590332, "reading_distances": [21, 22, 23, 26], "reading_log_probs": [-26.349227905273438, -25.975204467773438, -24.84611701965332, -21.900737762451172], "reading_logprob_std": 2.0153510195516526, "tokenization_path_logprob_std": 33.574750925827765, "trajectory_balance_rmse": 326.10425184379176}900{"canonical_length": 243, "canonical_nll": 1.3947431148324974, "generation_rouge_l_f1": 0.2105263157894737, "generation_token_f1": 0.48582995951417, "prompt_id": "960951e022fd8e15", "reading_canonical_log_prob": -338.9225769042969, "reading_distances": [33, 34, 33, 30], "reading_log_probs": [-354.37408447265625, -354.1444091796875, -351.4924621582031, -353.59454345703125], "reading_logprob_std": 1.3139667424441401, "tokenization_path_logprob_std": 215.858040670274, "trajectory_balance_rmse": 5974.089831165828}901{"canonical_length": 192, "canonical_nll": 1.2980165481567383, "generation_rouge_l_f1": 0.29977628635346754, "generation_token_f1": 0.5145413870246085, "prompt_id": "e17ce1bfa59fa95a", "reading_canonical_log_prob": -249.21917724609375, "reading_distances": [18, 25, 15, 25], "reading_log_probs": [-250.79266357421875, -245.22613525390625, -245.63333129882812, -248.2996826171875], "reading_logprob_std": 2.590717926352637, "tokenization_path_logprob_std": 132.94559320481207, "trajectory_balance_rmse": 4194.790407632609}902{"canonical_length": 37, "canonical_nll": 1.2992667120856207, "generation_rouge_l_f1": 0.19999999999999998, "generation_token_f1": 0.28, "prompt_id": "305cbe322d895404", "reading_canonical_log_prob": -48.07286834716797, "reading_distances": [13, 24, 14, 13], "reading_log_probs": [-46.98246765136719, -48.75138854980469, -49.746551513671875, -49.612266540527344], "reading_logprob_std": 1.272614619388831, "tokenization_path_logprob_std": 24.381609237075498, "trajectory_balance_rmse": 533.2752770523638}903{"canonical_length": 25, "canonical_nll": 1.3035293579101563, "generation_rouge_l_f1": 0.11718749999999999, "generation_token_f1": 0.1484375, "prompt_id": "fee5a1b64fc319ac", "reading_canonical_log_prob": -32.588233947753906, "reading_distances": [27, 26, 24, 24], "reading_log_probs": [-38.52960968017578, -36.52599334716797, -37.841922760009766, -33.17755889892578], "reading_logprob_std": 2.377530458576105, "tokenization_path_logprob_std": 46.37585626576449, "trajectory_balance_rmse": 562.5622655220646}904{"canonical_length": 12, "canonical_nll": 1.2830328146616619, "generation_rouge_l_f1": 0.2857142857142857, "generation_token_f1": 0.2857142857142857, "prompt_id": "dce2143d59fcc65f", "reading_canonical_log_prob": -15.396393775939941, "reading_distances": [11, 11, 13, 20], "reading_log_probs": [-16.97539520263672, -21.43335723876953, -19.700368881225586, -18.23191261291504], "reading_logprob_std": 1.921080494987827, "tokenization_path_logprob_std": 33.217631117669924, "trajectory_balance_rmse": 232.57777919781012}905{"canonical_length": 30, "canonical_nll": 0.4184571901957194, "generation_rouge_l_f1": 0.14915254237288136, "generation_token_f1": 0.14915254237288136, "prompt_id": "bc27bf2b83eeb772", "reading_canonical_log_prob": -12.553715705871582, "reading_distances": [35, 42, 39, 37], "reading_log_probs": [-10.187223434448242, -14.688032150268555, -17.43046760559082, -12.3626070022583], "reading_logprob_std": 3.1100097042548227, "tokenization_path_logprob_std": 56.7134820863859, "trajectory_balance_rmse": 364.33797139826294}906{"canonical_length": 171, "canonical_nll": 2.024245234260782, "generation_rouge_l_f1": 0.28423772609819126, "generation_token_f1": 0.47545219638242897, "prompt_id": "8e8e4001993842cc", "reading_canonical_log_prob": -346.14593505859375, "reading_distances": [23, 25, 30, 23], "reading_log_probs": [-340.181396484375, -342.03515625, -340.652587890625, -341.8384094238281], "reading_logprob_std": 0.9018737800662014, "tokenization_path_logprob_std": 130.8821269597097, "trajectory_balance_rmse": 2666.7035703011247}907{"canonical_length": 93, "canonical_nll": 1.8173299809937835, "generation_rouge_l_f1": 0.26143790849673204, "generation_token_f1": 0.36601307189542487, "prompt_id": "ae4b947d6478d1ef", "reading_canonical_log_prob": -169.01168823242188, "reading_distances": [21, 24, 18, 27], "reading_log_probs": [-175.263916015625, -179.90097045898438, -178.17706298828125, -177.28761291503906], "reading_logprob_std": 1.9295220562897983, "tokenization_path_logprob_std": 87.15434661271787, "trajectory_balance_rmse": 2050.1022405456797}908{"canonical_length": 153, "canonical_nll": 2.184887455958946, "generation_rouge_l_f1": 0.28571428571428575, "generation_token_f1": 0.4857142857142857, "prompt_id": "e8d78f3e32f3fc0a", "reading_canonical_log_prob": -334.28778076171875, "reading_distances": [18, 20, 23, 24], "reading_log_probs": [-345.0582275390625, -344.33380126953125, -348.309326171875, -345.1626281738281], "reading_logprob_std": 1.7677385697116803, "tokenization_path_logprob_std": 77.0875751266899, "trajectory_balance_rmse": 4092.819427762184}909{"canonical_length": 14, "canonical_nll": 1.3732305254255022, "generation_rouge_l_f1": 0.5333333333333333, "generation_token_f1": 0.5333333333333333, "prompt_id": "0652d72e6ba6a111", "reading_canonical_log_prob": -19.22522735595703, "reading_distances": [35, 27, 25, 28], "reading_log_probs": [-21.8012752532959, -22.496850967407227, -21.74398422241211, -23.012977600097656], "reading_logprob_std": 0.6054522856804225, "tokenization_path_logprob_std": 29.34012187754935, "trajectory_balance_rmse": 195.47321726284878}910{"canonical_length": 35, "canonical_nll": 1.6850494384765624, "generation_rouge_l_f1": 0.198019801980198, "generation_token_f1": 0.37623762376237624, "prompt_id": "abb413900ce0e656", "reading_canonical_log_prob": -58.97673034667969, "reading_distances": [28, 24, 23, 27], "reading_log_probs": [-57.63823699951172, -61.651161193847656, -60.343196868896484, -56.45169448852539], "reading_logprob_std": 2.392997894663724, "tokenization_path_logprob_std": 89.97196667116154, "trajectory_balance_rmse": 925.0012476646384}911{"canonical_length": 294, "canonical_nll": 0.8864903871704932, "generation_rouge_l_f1": 0.18439716312056736, "generation_token_f1": 0.42907801418439717, "prompt_id": "7607920df7068b2c", "reading_canonical_log_prob": -260.628173828125, "reading_distances": [14, 8, 13, 9], "reading_log_probs": [-269.752685546875, -271.93145751953125, -268.45721435546875, -265.5114440917969], "reading_logprob_std": 2.6829387248955094, "tokenization_path_logprob_std": 214.08735120634358, "trajectory_balance_rmse": 5451.364164533425}912{"canonical_length": 50, "canonical_nll": 2.6402703857421876, "generation_rouge_l_f1": 0.30769230769230765, "generation_token_f1": 0.4358974358974359, "prompt_id": "6e3d23fca9ec801f", "reading_canonical_log_prob": -132.01351928710938, "reading_distances": [28, 32, 27, 24], "reading_log_probs": [-141.11859130859375, -143.4976043701172, -141.27816772460938, -139.677490234375], "reading_logprob_std": 1.577003475657373, "tokenization_path_logprob_std": 113.48477070062017, "trajectory_balance_rmse": 1248.4387201862187}913{"canonical_length": 18, "canonical_nll": 1.9276771545410156, "generation_rouge_l_f1": 0.4666666666666667, "generation_token_f1": 0.4666666666666667, "prompt_id": "4afc9f7951fa42e2", "reading_canonical_log_prob": -34.69818878173828, "reading_distances": [18, 20, 24, 21], "reading_log_probs": [-32.23305892944336, -32.58955383300781, -31.52487564086914, -32.49953842163086], "reading_logprob_std": 0.48228911363266075, "tokenization_path_logprob_std": 41.404577877357404, "trajectory_balance_rmse": 376.7111534276547}914{"canonical_length": 501, "canonical_nll": 0.6360296984156687, "generation_rouge_l_f1": 0.2421052631578947, "generation_token_f1": 0.3763157894736842, "prompt_id": "031d7ff67f65087b", "reading_canonical_log_prob": -318.65087890625, "reading_distances": [34, 27, 22, 24], "reading_log_probs": [-322.5064392089844, -323.2408142089844, -319.81524658203125, -321.135498046875], "reading_logprob_std": 1.5157846621922682, "tokenization_path_logprob_std": 163.93652242013226, "trajectory_balance_rmse": 6064.579233641631}915{"canonical_length": 128, "canonical_nll": 1.152758240699768, "generation_rouge_l_f1": 0.2136986301369863, "generation_token_f1": 0.3561643835616438, "prompt_id": "81ff4f6d7865cbd6", "reading_canonical_log_prob": -147.5530548095703, "reading_distances": [20, 19, 18, 19], "reading_log_probs": [-163.55479431152344, -163.99484252929688, -163.07345581054688, -164.94183349609375], "reading_logprob_std": 0.7950779541736086, "tokenization_path_logprob_std": 155.16164011048966, "trajectory_balance_rmse": 2323.009173820879}916{"canonical_length": 112, "canonical_nll": 1.394688333783831, "generation_rouge_l_f1": 0.20105820105820105, "generation_token_f1": 0.3280423280423281, "prompt_id": "fa3e1b2e7d03f031", "reading_canonical_log_prob": -156.20509338378906, "reading_distances": [26, 29, 30, 28], "reading_log_probs": [-158.9398193359375, -161.24758911132812, -162.71893310546875, -157.3484649658203], "reading_logprob_std": 2.386592082878725, "tokenization_path_logprob_std": 166.72829332432383, "trajectory_balance_rmse": 2722.255134328017}917{"canonical_length": 136, "canonical_nll": 1.175948647891774, "generation_rouge_l_f1": 0.3, "generation_token_f1": 0.46111111111111114, "prompt_id": "c4484ce7220f53d7", "reading_canonical_log_prob": -159.92901611328125, "reading_distances": [19, 15, 13, 15], "reading_log_probs": [-162.4453125, -161.00106811523438, -163.09214782714844, -158.80967712402344], "reading_logprob_std": 1.8981668027224208, "tokenization_path_logprob_std": 107.77347938974158, "trajectory_balance_rmse": 2749.7797056857357}918{"canonical_length": 5, "canonical_nll": 0.15211386680603028, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "e265516b7f250f1e", "reading_canonical_log_prob": -0.7605693340301514, "reading_distances": [36, 36, 40, 41], "reading_log_probs": [-2.5922112464904785, -1.795020341873169, -2.3801469802856445, -2.4480109214782715], "reading_logprob_std": 0.35055357646707314, "tokenization_path_logprob_std": 41.18227207734934, "trajectory_balance_rmse": 117.73926614394152}919{"canonical_length": 20, "canonical_nll": 0.46318774223327636, "generation_rouge_l_f1": 0.8823529411764706, "generation_token_f1": 0.8823529411764706, "prompt_id": "d3f0dad3a37f19c3", "reading_canonical_log_prob": -9.263754844665527, "reading_distances": [16, 14, 16, 17], "reading_log_probs": [-8.247053146362305, -8.021060943603516, -7.640684604644775, -8.077155113220215], "reading_logprob_std": 0.2559231404345693, "tokenization_path_logprob_std": 56.20376006922601, "trajectory_balance_rmse": 463.49924812851674}920{"canonical_length": 23, "canonical_nll": 1.45605817048446, "generation_rouge_l_f1": 0.1904761904761905, "generation_token_f1": 0.1904761904761905, "prompt_id": "448ac2a15956c368", "reading_canonical_log_prob": -33.48933792114258, "reading_distances": [29, 35, 41, 31], "reading_log_probs": [-32.39417266845703, -33.56029510498047, -34.98176956176758, -33.074195861816406], "reading_logprob_std": 1.0959641983579542, "tokenization_path_logprob_std": 31.695803097909547, "trajectory_balance_rmse": 463.08613594323333}921{"canonical_length": 165, "canonical_nll": 1.446948612097538, "generation_rouge_l_f1": 0.3130841121495327, "generation_token_f1": 0.46728971962616817, "prompt_id": "e00cf91341eb5691", "reading_canonical_log_prob": -238.74652099609375, "reading_distances": [14, 14, 9, 17], "reading_log_probs": [-241.31076049804688, -242.74908447265625, -242.48773193359375, -242.763916015625], "reading_logprob_std": 0.6898370185416881, "tokenization_path_logprob_std": 103.64450303583402, "trajectory_balance_rmse": 3918.1270701294025}922{"canonical_length": 224, "canonical_nll": 1.0358702114650182, "generation_rouge_l_f1": 0.29959514170040485, "generation_token_f1": 0.6356275303643725, "prompt_id": "bf8645b4ff7492c2", "reading_canonical_log_prob": -232.03492736816406, "reading_distances": [52, 55, 60, 48], "reading_log_probs": [-236.21670532226562, -237.80862426757812, -238.10308837890625, -233.01312255859375], "reading_logprob_std": 2.3335753557971053, "tokenization_path_logprob_std": 244.13892452711835, "trajectory_balance_rmse": 4817.803443385328}923{"canonical_length": 261, "canonical_nll": 1.3828300388380028, "generation_rouge_l_f1": 0.22304832713754646, "generation_token_f1": 0.5055762081784387, "prompt_id": "93af13c34520a089", "reading_canonical_log_prob": -360.91864013671875, "reading_distances": [25, 28, 26, 22], "reading_log_probs": [-368.52960205078125, -358.216064453125, -357.51495361328125, -362.21282958984375], "reading_logprob_std": 5.050815767133138, "tokenization_path_logprob_std": 255.09824519317118, "trajectory_balance_rmse": 5764.785676562648}924{"canonical_length": 107, "canonical_nll": 1.0739164263288552, "generation_rouge_l_f1": 0.2958904109589041, "generation_token_f1": 0.410958904109589, "prompt_id": "8f7942ab9e010ef0", "reading_canonical_log_prob": -114.9090576171875, "reading_distances": [34, 31, 34, 35], "reading_log_probs": [-123.83770751953125, -121.0377197265625, -120.87286376953125, -120.16697692871094], "reading_logprob_std": 1.6173058850985913, "tokenization_path_logprob_std": 131.74841967055798, "trajectory_balance_rmse": 1803.8296032362866}925{"canonical_length": 39, "canonical_nll": 2.748377873347356, "generation_rouge_l_f1": 0.2692307692307692, "generation_token_f1": 0.4230769230769231, "prompt_id": "ba9caab6c5bf7b20", "reading_canonical_log_prob": -107.18673706054688, "reading_distances": [38, 37, 40, 36], "reading_log_probs": [-111.89613342285156, -116.83683013916016, -117.68527221679688, -120.79004669189453], "reading_logprob_std": 3.6857123335534836, "tokenization_path_logprob_std": 30.505169000417954, "trajectory_balance_rmse": 755.79325907922}926{"canonical_length": 292, "canonical_nll": 1.4304444822546554, "generation_rouge_l_f1": 0.20071684587813618, "generation_token_f1": 0.37992831541218636, "prompt_id": "90e4d685d3873c08", "reading_canonical_log_prob": -417.6897888183594, "reading_distances": [30, 30, 22, 20], "reading_log_probs": [-423.179443359375, -419.99420166015625, -418.98272705078125, -420.78662109375], "reading_logprob_std": 1.7885942807722621, "tokenization_path_logprob_std": 271.8061542316994, "trajectory_balance_rmse": 6464.720206868021}927{"canonical_length": 15, "canonical_nll": 2.004952239990234, "generation_rouge_l_f1": 0.0970873786407767, "generation_token_f1": 0.0970873786407767, "prompt_id": "ec455005996bdab4", "reading_canonical_log_prob": -30.074283599853516, "reading_distances": [42, 34, 48, 44], "reading_log_probs": [-33.727718353271484, -32.8824462890625, -34.02927017211914, -34.54839324951172], "reading_logprob_std": 0.6975551776599235, "tokenization_path_logprob_std": 59.933861507741604, "trajectory_balance_rmse": 258.45064980437746}928{"canonical_length": 243, "canonical_nll": 1.631054787969393, "generation_rouge_l_f1": 0.19413092550790068, "generation_token_f1": 0.41986455981941306, "prompt_id": "ab90737e4575d4fd", "reading_canonical_log_prob": -396.3463134765625, "reading_distances": [11, 17, 14, 17], "reading_log_probs": [-399.6031799316406, -403.7411804199219, -407.8140869140625, -406.402099609375], "reading_logprob_std": 3.610518391804415, "tokenization_path_logprob_std": 391.792573086596, "trajectory_balance_rmse": 5155.736807990678}929{"canonical_length": 240, "canonical_nll": 1.4541946411132813, "generation_rouge_l_f1": 0.19253438113948917, "generation_token_f1": 0.4086444007858547, "prompt_id": "ec5a98de132967a0", "reading_canonical_log_prob": -349.0067138671875, "reading_distances": [33, 32, 29, 37], "reading_log_probs": [-362.9737548828125, -357.5810546875, -365.186279296875, -362.3010559082031], "reading_logprob_std": 3.199887634417912, "tokenization_path_logprob_std": 178.8213868782278, "trajectory_balance_rmse": 5364.0319617590185}930{"canonical_length": 37, "canonical_nll": 0.8292725021774704, "generation_rouge_l_f1": 0.06060606060606061, "generation_token_f1": 0.06060606060606061, "prompt_id": "71251a192c44eeb3", "reading_canonical_log_prob": -30.683082580566406, "reading_distances": [22, 24, 25, 22], "reading_log_probs": [-36.53017807006836, -39.6115608215332, -38.9028205871582, -37.51078796386719], "reading_logprob_std": 1.3826081055496247, "tokenization_path_logprob_std": 69.14778397795415, "trajectory_balance_rmse": 652.9821174925908}931{"canonical_length": 221, "canonical_nll": 1.8076597187853507, "generation_rouge_l_f1": 0.2707423580786026, "generation_token_f1": 0.5021834061135371, "prompt_id": "646766dc50567beb", "reading_canonical_log_prob": -399.4927978515625, "reading_distances": [44, 41, 39, 54], "reading_log_probs": [-395.6224060058594, -388.610595703125, -393.56365966796875, -395.18621826171875], "reading_logprob_std": 3.2145312127965764, "tokenization_path_logprob_std": 243.23832692809933, "trajectory_balance_rmse": 4155.003107850369}932{"canonical_length": 87, "canonical_nll": 1.1432204849418552, "generation_rouge_l_f1": 0.27472527472527475, "generation_token_f1": 0.5494505494505495, "prompt_id": "3175f6135d7796a6", "reading_canonical_log_prob": -99.4601821899414, "reading_distances": [40, 40, 42, 47], "reading_log_probs": [-98.74130249023438, -95.73139953613281, -95.13105010986328, -95.95489501953125], "reading_logprob_std": 1.605890063064253, "tokenization_path_logprob_std": 32.44683611069706, "trajectory_balance_rmse": 1470.9095191239646}933{"canonical_length": 315, "canonical_nll": 0.9634448823474703, "generation_rouge_l_f1": 0.24772313296903462, "generation_token_f1": 0.42622950819672134, "prompt_id": "364270b26598f5f6", "reading_canonical_log_prob": -303.4851379394531, "reading_distances": [25, 17, 25, 26], "reading_log_probs": [-309.9404296875, -308.0914306640625, -309.80047607421875, -311.01348876953125], "reading_logprob_std": 1.2083218314662048, "tokenization_path_logprob_std": 252.3841502459774, "trajectory_balance_rmse": 5546.23697610117}934{"canonical_length": 132, "canonical_nll": 1.9197305621522847, "generation_rouge_l_f1": 0.1945137157107232, "generation_token_f1": 0.3391521197007481, "prompt_id": "3b49b131c410b62c", "reading_canonical_log_prob": -253.40443420410156, "reading_distances": [18, 13, 10, 23], "reading_log_probs": [-250.92926025390625, -251.32357788085938, -256.4004211425781, -252.74691772460938], "reading_logprob_std": 2.4923430673002653, "tokenization_path_logprob_std": 212.13965225796295, "trajectory_balance_rmse": 2662.7179088218973}935{"canonical_length": 9, "canonical_nll": 0.4803614086574978, "generation_rouge_l_f1": 0.8571428571428571, "generation_token_f1": 0.8571428571428571, "prompt_id": "300deaab60201ff0", "reading_canonical_log_prob": -4.3232526779174805, "reading_distances": [31, 19, 23, 20], "reading_log_probs": [-4.658048629760742, -3.1460824012756348, -3.7048749923706055, -4.385225772857666], "reading_logprob_std": 0.681885670405448, "tokenization_path_logprob_std": 55.637550906077585, "trajectory_balance_rmse": 165.14596868178594}936{"canonical_length": 239, "canonical_nll": 1.4592088515788442, "generation_rouge_l_f1": 0.24796747967479674, "generation_token_f1": 0.540650406504065, "prompt_id": "166e7bed707f11c1", "reading_canonical_log_prob": -348.75091552734375, "reading_distances": [21, 26, 23, 26], "reading_log_probs": [-357.8338623046875, -355.40753173828125, -360.614501953125, -355.92962646484375], "reading_logprob_std": 2.3554605188115225, "tokenization_path_logprob_std": 321.97982009234977, "trajectory_balance_rmse": 5620.235781947884}937{"canonical_length": 278, "canonical_nll": 1.3704316941954249, "generation_rouge_l_f1": 0.2209737827715356, "generation_token_f1": 0.47191011235955066, "prompt_id": "af085ada31a636ce", "reading_canonical_log_prob": -380.9800109863281, "reading_distances": [29, 31, 32, 32], "reading_log_probs": [-381.679931640625, -381.10455322265625, -381.5078125, -376.78619384765625], "reading_logprob_std": 2.33477128329056, "tokenization_path_logprob_std": 66.80964729655896, "trajectory_balance_rmse": 5679.4046284743645}938{"canonical_length": 157, "canonical_nll": 1.5365124356215167, "generation_rouge_l_f1": 0.195, "generation_token_f1": 0.405, "prompt_id": "f0f7445e4deb0e6e", "reading_canonical_log_prob": -241.23245239257812, "reading_distances": [28, 29, 13, 35], "reading_log_probs": [-241.73605346679688, -243.29873657226562, -243.3192138671875, -246.86488342285156], "reading_logprob_std": 2.17069242672165, "tokenization_path_logprob_std": 152.55275934533447, "trajectory_balance_rmse": 4097.211018779074}939{"canonical_length": 20, "canonical_nll": 2.1304805755615233, "generation_rouge_l_f1": 0.1739130434782609, "generation_token_f1": 0.1739130434782609, "prompt_id": "b5d6efcaf5d67694", "reading_canonical_log_prob": -42.60961151123047, "reading_distances": [41, 35, 40, 28], "reading_log_probs": [-35.65675735473633, -37.68815612792969, -42.91698455810547, -40.49740982055664], "reading_logprob_std": 3.1800981547674256, "tokenization_path_logprob_std": 66.1586085969764, "trajectory_balance_rmse": 484.61380634878975}940{"canonical_length": 91, "canonical_nll": 2.0406903277386674, "generation_rouge_l_f1": 0.2389078498293515, "generation_token_f1": 0.3959044368600683, "prompt_id": "b7652a39606177d7", "reading_canonical_log_prob": -185.70281982421875, "reading_distances": [24, 13, 22, 24], "reading_log_probs": [-189.59625244140625, -189.15524291992188, -189.8567657470703, -185.81809997558594], "reading_logprob_std": 1.8814061930099197, "tokenization_path_logprob_std": 133.91459060658042, "trajectory_balance_rmse": 1853.8861987177133}941{"canonical_length": 18, "canonical_nll": 0.8484606213039823, "generation_rouge_l_f1": 0.09022556390977444, "generation_token_f1": 0.09022556390977444, "prompt_id": "f3723ab195d4cd31", "reading_canonical_log_prob": -15.27229118347168, "reading_distances": [25, 26, 22, 20], "reading_log_probs": [-13.790154457092285, -12.326101303100586, -13.389537811279297, -12.31911849975586], "reading_logprob_std": 0.7497019180616818, "tokenization_path_logprob_std": 83.3170105754033, "trajectory_balance_rmse": 388.2742079434421}942{"canonical_length": 7, "canonical_nll": 1.057474204472133, "generation_rouge_l_f1": 0.15, "generation_token_f1": 0.15, "prompt_id": "ff3fea8e176d412d", "reading_canonical_log_prob": -7.402319431304932, "reading_distances": [15, 23, 26, 19], "reading_log_probs": [-8.499775886535645, -6.714561462402344, -6.979274749755859, -8.505514144897461], "reading_logprob_std": 0.9620264420054246, "tokenization_path_logprob_std": 30.60139358713389, "trajectory_balance_rmse": 132.59057771509129}943{"canonical_length": 69, "canonical_nll": 1.1432956197987432, "generation_rouge_l_f1": 0.16296296296296298, "generation_token_f1": 0.2296296296296296, "prompt_id": "f5b5083636173713", "reading_canonical_log_prob": -78.88739776611328, "reading_distances": [16, 25, 19, 24], "reading_log_probs": [-82.37828063964844, -84.92060852050781, -87.44746398925781, -90.73457336425781], "reading_logprob_std": 3.5704802757313225, "tokenization_path_logprob_std": 105.96048596505854, "trajectory_balance_rmse": 1034.6225417409241}944{"canonical_length": 291, "canonical_nll": 0.9978612526175902, "generation_rouge_l_f1": 0.26591760299625467, "generation_token_f1": 0.5093632958801498, "prompt_id": "640dfdad1fff517f", "reading_canonical_log_prob": -290.37762451171875, "reading_distances": [22, 25, 24, 22], "reading_log_probs": [-294.94281005859375, -298.0516357421875, -296.8277282714844, -293.87530517578125], "reading_logprob_std": 1.8711353306631502, "tokenization_path_logprob_std": 414.69645604672513, "trajectory_balance_rmse": 6646.265541350625}945{"canonical_length": 17, "canonical_nll": 1.803090263815487, "generation_rouge_l_f1": 0.1764705882352941, "generation_token_f1": 0.1764705882352941, "prompt_id": "f39da6ed7768e21d", "reading_canonical_log_prob": -30.65253448486328, "reading_distances": [34, 32, 27, 33], "reading_log_probs": [-32.69873046875, -31.154001235961914, -33.94247055053711, -29.83245849609375], "reading_logprob_std": 1.7926430262372386, "tokenization_path_logprob_std": 43.3681886958574, "trajectory_balance_rmse": 413.66465874230306}946{"canonical_length": 16, "canonical_nll": 2.1582088470458984, "generation_rouge_l_f1": 0.22222222222222224, "generation_token_f1": 0.3703703703703704, "prompt_id": "968e82be4a52a789", "reading_canonical_log_prob": -34.531341552734375, "reading_distances": [34, 29, 30, 27], "reading_log_probs": [-35.846458435058594, -36.708213806152344, -36.954833984375, -38.0878791809082], "reading_logprob_std": 0.9239036346687785, "tokenization_path_logprob_std": 33.561155933396805, "trajectory_balance_rmse": 405.9092636204615}947{"canonical_length": 15, "canonical_nll": 2.462449900309245, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "e36a9df5c4b484e7", "reading_canonical_log_prob": -36.93674850463867, "reading_distances": [23, 25, 23, 26], "reading_log_probs": [-38.37334442138672, -36.59328079223633, -40.402549743652344, -40.25420379638672], "reading_logprob_std": 1.7971928680322153, "tokenization_path_logprob_std": 67.24916343698939, "trajectory_balance_rmse": 313.55427436914954}948{"canonical_length": 243, "canonical_nll": 1.1957823199990354, "generation_rouge_l_f1": 0.2412451361867704, "generation_token_f1": 0.5097276264591439, "prompt_id": "87e83988cc685281", "reading_canonical_log_prob": -290.5751037597656, "reading_distances": [28, 33, 37, 39], "reading_log_probs": [-295.1458740234375, -296.08929443359375, -298.4263610839844, -295.3436584472656], "reading_logprob_std": 1.5058681107014924, "tokenization_path_logprob_std": 268.60716732895264, "trajectory_balance_rmse": 5391.333222587818}949{"canonical_length": 106, "canonical_nll": 0.921961802356648, "generation_rouge_l_f1": 0.24332344213649854, "generation_token_f1": 0.4035608308605341, "prompt_id": "753de9bc3063e720", "reading_canonical_log_prob": -97.72795104980469, "reading_distances": [21, 16, 16, 21], "reading_log_probs": [-97.37825012207031, -97.46588897705078, -97.63923645019531, -99.2432632446289], "reading_logprob_std": 0.8811015345597045, "tokenization_path_logprob_std": 115.17022864876277, "trajectory_balance_rmse": 2301.3954684706555}950{"canonical_length": 18, "canonical_nll": 1.4631724887424045, "generation_rouge_l_f1": 0.5384615384615384, "generation_token_f1": 0.5384615384615384, "prompt_id": "52785eaaf4af8786", "reading_canonical_log_prob": -26.33710479736328, "reading_distances": [20, 24, 24, 27], "reading_log_probs": [-20.081600189208984, -20.681682586669922, -24.43603515625, -22.47723960876465], "reading_logprob_std": 1.9624890493880387, "tokenization_path_logprob_std": 57.155300409917125, "trajectory_balance_rmse": 370.0186465902879}951{"canonical_length": 293, "canonical_nll": 1.6848225772584258, "generation_rouge_l_f1": 0.21428571428571427, "generation_token_f1": 0.38928571428571423, "prompt_id": "9296c91d93887868", "reading_canonical_log_prob": -493.65301513671875, "reading_distances": [19, 19, 23, 17], "reading_log_probs": [-500.71563720703125, -506.68438720703125, -500.2035827636719, -498.4983825683594], "reading_logprob_std": 3.567487448374251, "tokenization_path_logprob_std": 197.67828676718761, "trajectory_balance_rmse": 6055.606580957803}952{"canonical_length": 130, "canonical_nll": 1.6522533710186298, "generation_rouge_l_f1": 0.2914572864321608, "generation_token_f1": 0.4321608040201005, "prompt_id": "dd7d0916ee5ef6b2", "reading_canonical_log_prob": -214.79293823242188, "reading_distances": [54, 60, 62, 57], "reading_log_probs": [-217.78402709960938, -215.89305114746094, -216.8459930419922, -216.9727783203125], "reading_logprob_std": 0.7748013753918763, "tokenization_path_logprob_std": 160.30870312182944, "trajectory_balance_rmse": 2595.0138284233626}953{"canonical_length": 172, "canonical_nll": 1.8339873912722566, "generation_rouge_l_f1": 0.21226415094339623, "generation_token_f1": 0.4150943396226416, "prompt_id": "0d595f7cde9d737e", "reading_canonical_log_prob": -315.4458312988281, "reading_distances": [17, 18, 19, 20], "reading_log_probs": [-318.77703857421875, -321.1348876953125, -320.83837890625, -322.323486328125], "reading_logprob_std": 1.4745648041572361, "tokenization_path_logprob_std": 307.0773639401986, "trajectory_balance_rmse": 3685.2017555467096}954{"canonical_length": 154, "canonical_nll": 0.8693789197252942, "generation_rouge_l_f1": 0.22377622377622378, "generation_token_f1": 0.41958041958041953, "prompt_id": "8c2391cc4e076841", "reading_canonical_log_prob": -133.8843536376953, "reading_distances": [233, 246, 247, 230], "reading_log_probs": [-134.03306579589844, -137.25759887695312, -133.96566772460938, -134.45306396484375], "reading_logprob_std": 1.568394519690787, "tokenization_path_logprob_std": 95.4047004579287, "trajectory_balance_rmse": 2607.3066928636786}955{"canonical_length": 395, "canonical_nll": 0.7286645816851266, "generation_rouge_l_f1": 0.27816901408450706, "generation_token_f1": 0.39788732394366194, "prompt_id": "a3d283a16147427b", "reading_canonical_log_prob": -287.822509765625, "reading_distances": [14, 17, 17, 14], "reading_log_probs": [-294.71197509765625, -295.1353759765625, -294.95733642578125, -292.45367431640625], "reading_logprob_std": 1.2526950526045655, "tokenization_path_logprob_std": 261.6943263684734, "trajectory_balance_rmse": 5952.779481731995}956{"canonical_length": 11, "canonical_nll": 1.4684443040327593, "generation_rouge_l_f1": 0.25316455696202533, "generation_token_f1": 0.25316455696202533, "prompt_id": "977c41664864ae25", "reading_canonical_log_prob": -16.15288734436035, "reading_distances": [30, 32, 35, 27], "reading_log_probs": [-16.618789672851562, -18.281152725219727, -17.471546173095703, -15.575787544250488], "reading_logprob_std": 1.1599878705306133, "tokenization_path_logprob_std": 35.219432206238025, "trajectory_balance_rmse": 277.2772033203385}957{"canonical_length": 20, "canonical_nll": 1.1473117828369142, "generation_rouge_l_f1": 0.11827956989247312, "generation_token_f1": 0.16129032258064516, "prompt_id": "4c8a56abb260b4fb", "reading_canonical_log_prob": -22.94623565673828, "reading_distances": [28, 29, 25, 28], "reading_log_probs": [-25.945817947387695, -28.109580993652344, -25.621469497680664, -29.01453399658203], "reading_logprob_std": 1.6514290800039773, "tokenization_path_logprob_std": 52.43966174245402, "trajectory_balance_rmse": 241.58426894104016}958{"canonical_length": 244, "canonical_nll": 0.7427403809594326, "generation_rouge_l_f1": 0.31434184675834975, "generation_token_f1": 0.5893909626719057, "prompt_id": "7a98c8dec132eec9", "reading_canonical_log_prob": -181.22865295410156, "reading_distances": [32, 35, 31, 33], "reading_log_probs": [-185.92568969726562, -183.1800537109375, -186.59603881835938, -182.80612182617188], "reading_logprob_std": 1.9124992098511555, "tokenization_path_logprob_std": 213.90119801145386, "trajectory_balance_rmse": 4647.428781247349}959{"canonical_length": 419, "canonical_nll": 1.1579372069147151, "generation_rouge_l_f1": 0.23661971830985915, "generation_token_f1": 0.3887323943661972, "prompt_id": "7f79088a335cbb01", "reading_canonical_log_prob": -485.1756896972656, "reading_distances": [23, 13, 20, 16], "reading_log_probs": [-511.531494140625, -513.6880493164062, -515.63916015625, -510.48492431640625], "reading_logprob_std": 2.295864925737599, "tokenization_path_logprob_std": 172.320184329906, "trajectory_balance_rmse": 7664.206821541106}960{"canonical_length": 12, "canonical_nll": 2.24161958694458, "generation_rouge_l_f1": 0.4285714285714285, "generation_token_f1": 0.4285714285714285, "prompt_id": "48b64f1a7aabf6e4", "reading_canonical_log_prob": -26.89943504333496, "reading_distances": [30, 23, 21, 27], "reading_log_probs": [-26.46677017211914, -26.968368530273438, -26.386463165283203, -26.435714721679688], "reading_logprob_std": 0.2713813765830527, "tokenization_path_logprob_std": 45.29882442355132, "trajectory_balance_rmse": 303.35834184001766}961{"canonical_length": 319, "canonical_nll": 1.4750834976244123, "generation_rouge_l_f1": 0.21342512908777972, "generation_token_f1": 0.42340791738382094, "prompt_id": "735c6c6b04d8c3fe", "reading_canonical_log_prob": -470.5516357421875, "reading_distances": [35, 38, 37, 39], "reading_log_probs": [-491.60791015625, -492.5673828125, -491.998779296875, -490.170654296875], "reading_logprob_std": 1.0226081613415543, "tokenization_path_logprob_std": 329.8693705837692, "trajectory_balance_rmse": 6642.84170082723}962{"canonical_length": 57, "canonical_nll": 1.2228738885176809, "generation_rouge_l_f1": 0.33082706766917297, "generation_token_f1": 0.42105263157894735, "prompt_id": "eab393eed1fb7651", "reading_canonical_log_prob": -69.70381164550781, "reading_distances": [35, 35, 37, 35], "reading_log_probs": [-73.27839660644531, -80.99259948730469, -73.61702728271484, -78.61376953125], "reading_logprob_std": 3.7981918008435978, "tokenization_path_logprob_std": 90.44909989781576, "trajectory_balance_rmse": 1237.399308867034}963{"canonical_length": 20, "canonical_nll": 0.02195672392845154, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "691dfdc26548dcb9", "reading_canonical_log_prob": -0.43913447856903076, "reading_distances": [30, 31, 33, 31], "reading_log_probs": [-1.0295456647872925, -0.6387479305267334, -0.608250617980957, -0.7907854318618774], "reading_logprob_std": 0.1924801462828111, "tokenization_path_logprob_std": 42.554801584074596, "trajectory_balance_rmse": 315.98360747370293}964{"canonical_length": 230, "canonical_nll": 1.9428148352581522, "generation_rouge_l_f1": 0.2619047619047619, "generation_token_f1": 0.46190476190476193, "prompt_id": "2b073d802d00719d", "reading_canonical_log_prob": -446.847412109375, "reading_distances": [24, 28, 25, 21], "reading_log_probs": [-460.0274963378906, -448.8948669433594, -457.33966064453125, -452.0867614746094], "reading_logprob_std": 5.027514703797078, "tokenization_path_logprob_std": 50.15612113120061, "trajectory_balance_rmse": 4277.492977159043}965{"canonical_length": 11, "canonical_nll": 4.296413074840199, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "11a956113e5f3fb8", "reading_canonical_log_prob": -47.26054382324219, "reading_distances": [29, 30, 27, 26], "reading_log_probs": [-38.402549743652344, -42.59178924560547, -43.77184295654297, -38.98084259033203], "reading_logprob_std": 2.647303767423877, "tokenization_path_logprob_std": 0.0, "trajectory_balance_rmse": 47.12293243408203}966{"canonical_length": 162, "canonical_nll": 1.227442611882716, "generation_rouge_l_f1": 0.3, "generation_token_f1": 0.4263157894736842, "prompt_id": "b48c6b78ff2012e6", "reading_canonical_log_prob": -198.845703125, "reading_distances": [30, 20, 32, 22], "reading_log_probs": [-202.38882446289062, -200.84457397460938, -198.75750732421875, -197.65634155273438], "reading_logprob_std": 2.1154354843712913, "tokenization_path_logprob_std": 190.30816862917177, "trajectory_balance_rmse": 3006.238092689396}967{"canonical_length": 142, "canonical_nll": 0.8904308265363666, "generation_rouge_l_f1": 0.2934782608695652, "generation_token_f1": 0.4673913043478261, "prompt_id": "c057418a3f03722b", "reading_canonical_log_prob": -126.44117736816406, "reading_distances": [34, 36, 34, 30], "reading_log_probs": [-125.81670379638672, -122.88922882080078, -123.20645141601562, -126.41250610351562], "reading_logprob_std": 1.79191255497548, "tokenization_path_logprob_std": 207.66151845513002, "trajectory_balance_rmse": 2794.5671068544475}968{"canonical_length": 239, "canonical_nll": 1.2809803216527196, "generation_rouge_l_f1": 0.20338983050847456, "generation_token_f1": 0.461864406779661, "prompt_id": "14f504daadaa6ae8", "reading_canonical_log_prob": -306.154296875, "reading_distances": [23, 17, 20, 23], "reading_log_probs": [-313.483154296875, -314.3140563964844, -314.2359619140625, -316.6964111328125], "reading_logprob_std": 1.3939649773597595, "tokenization_path_logprob_std": 328.91427473159837, "trajectory_balance_rmse": 5803.9988769953725}969{"canonical_length": 97, "canonical_nll": 1.509129475072487, "generation_rouge_l_f1": 0.23574144486692014, "generation_token_f1": 0.37262357414448666, "prompt_id": "20abb31273f6c931", "reading_canonical_log_prob": -146.38555908203125, "reading_distances": [20, 23, 30, 29], "reading_log_probs": [-146.87161254882812, -145.15060424804688, -148.7051239013672, -144.3602294921875], "reading_logprob_std": 1.931495274735143, "tokenization_path_logprob_std": 209.53236493405004, "trajectory_balance_rmse": 2316.277401119364}970{"canonical_length": 5, "canonical_nll": 0.46824045181274415, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "609097a0b2afbf5d", "reading_canonical_log_prob": -2.3412022590637207, "reading_distances": [26, 26, 23, 33], "reading_log_probs": [-1.9891033172607422, -1.8504395484924316, -2.1703388690948486, -1.864703893661499], "reading_logprob_std": 0.14818371696495622, "tokenization_path_logprob_std": 32.59278132254577, "trajectory_balance_rmse": 74.59762449835642}971{"canonical_length": 70, "canonical_nll": 2.1439442225864958, "generation_rouge_l_f1": 0.10591900311526481, "generation_token_f1": 0.18691588785046728, "prompt_id": "b13cf3764c885d75", "reading_canonical_log_prob": -150.0760955810547, "reading_distances": [30, 22, 25, 35], "reading_log_probs": [-158.11087036132812, -156.6298828125, -157.46780395507812, -158.4902801513672], "reading_logprob_std": 0.814425051955366, "tokenization_path_logprob_std": 74.57845929856326, "trajectory_balance_rmse": 1497.0168374948844}972{"canonical_length": 42, "canonical_nll": 0.8199640001569476, "generation_rouge_l_f1": 0.07142857142857142, "generation_token_f1": 0.07142857142857142, "prompt_id": "d289e388118f2de0", "reading_canonical_log_prob": -34.4384880065918, "reading_distances": [21, 22, 26, 20], "reading_log_probs": [-36.02216339111328, -37.258487701416016, -39.096458435058594, -38.71445083618164], "reading_logprob_std": 1.4104400725030195, "tokenization_path_logprob_std": 90.33121244850567, "trajectory_balance_rmse": 755.7071334307919}973{"canonical_length": 10, "canonical_nll": 2.7958442687988283, "generation_rouge_l_f1": 0.09090909090909093, "generation_token_f1": 0.09090909090909093, "prompt_id": "58e5e1069ca06eb2", "reading_canonical_log_prob": -27.95844268798828, "reading_distances": [23, 12, 20, 22], "reading_log_probs": [-36.45415115356445, -33.88186264038086, -29.00971221923828, -32.718204498291016], "reading_logprob_std": 3.093521478222701, "tokenization_path_logprob_std": 61.607546791625104, "trajectory_balance_rmse": 280.25421467578235}974{"canonical_length": 113, "canonical_nll": 1.5215959126970409, "generation_rouge_l_f1": 0.3223684210526316, "generation_token_f1": 0.4539473684210526, "prompt_id": "ec3868510227235c", "reading_canonical_log_prob": -171.94033813476562, "reading_distances": [34, 28, 26, 26], "reading_log_probs": [-174.6104736328125, -174.25167846679688, -172.39987182617188, -172.22528076171875], "reading_logprob_std": 1.2339164235673552, "tokenization_path_logprob_std": 137.9640761313327, "trajectory_balance_rmse": 2518.7581536235743}975{"canonical_length": 16, "canonical_nll": 2.354593276977539, "generation_rouge_l_f1": 0.1818181818181818, "generation_token_f1": 0.1818181818181818, "prompt_id": "d625c7ada7581821", "reading_canonical_log_prob": -37.673492431640625, "reading_distances": [11, 10, 11, 10], "reading_log_probs": [-41.25695037841797, -41.26506042480469, -42.10004425048828, -54.54655456542969], "reading_logprob_std": 6.51495303202542, "tokenization_path_logprob_std": 21.284500517769548, "trajectory_balance_rmse": 396.0066764246753}976{"canonical_length": 7, "canonical_nll": 2.592205047607422, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "34ca0227975643ca", "reading_canonical_log_prob": -18.145435333251953, "reading_distances": [23, 19, 21, 16], "reading_log_probs": [-17.204221725463867, -14.047368049621582, -16.645023345947266, -13.48835563659668], "reading_logprob_std": 1.850922015564755, "tokenization_path_logprob_std": 20.459531200974208, "trajectory_balance_rmse": 86.9465797734239}977{"canonical_length": 92, "canonical_nll": 0.4850777335788893, "generation_rouge_l_f1": 0.37820512820512825, "generation_token_f1": 0.467948717948718, "prompt_id": "03d8c4a193a46caa", "reading_canonical_log_prob": -44.62715148925781, "reading_distances": [29, 29, 32, 23], "reading_log_probs": [-43.81451416015625, -44.30254364013672, -50.09685516357422, -49.354331970214844], "reading_logprob_std": 3.291928445461082, "tokenization_path_logprob_std": 87.03006782037026, "trajectory_balance_rmse": 1294.5560708093253}978{"canonical_length": 22, "canonical_nll": 1.1012698086825283, "generation_rouge_l_f1": 0.2608695652173913, "generation_token_f1": 0.4782608695652174, "prompt_id": "cfb9b15522adbd99", "reading_canonical_log_prob": -24.227935791015625, "reading_distances": [22, 14, 16, 16], "reading_log_probs": [-27.264705657958984, -24.631160736083984, -26.983150482177734, -29.233890533447266], "reading_logprob_std": 1.885809494135285, "tokenization_path_logprob_std": 18.40661026768406, "trajectory_balance_rmse": 538.2047564931102}979{"canonical_length": 115, "canonical_nll": 1.440857862389606, "generation_rouge_l_f1": 0.3322259136212624, "generation_token_f1": 0.4053156146179402, "prompt_id": "7e152e3b35975991", "reading_canonical_log_prob": -165.6986541748047, "reading_distances": [25, 29, 33, 23], "reading_log_probs": [-176.27651977539062, -174.37417602539062, -182.71737670898438, -172.54104614257812], "reading_logprob_std": 4.430804852153253, "tokenization_path_logprob_std": 73.77905002460972, "trajectory_balance_rmse": 3088.0125046204107}980{"canonical_length": 10, "canonical_nll": 2.679336929321289, "generation_rouge_l_f1": 0.4, "generation_token_f1": 0.6, "prompt_id": "6df3b4a03782fc55", "reading_canonical_log_prob": -26.79336929321289, "reading_distances": [26, 26, 24, 27], "reading_log_probs": [-22.77791976928711, -24.503883361816406, -23.00358009338379, -23.55160140991211], "reading_logprob_std": 0.7684689519975163, "tokenization_path_logprob_std": 53.0951651246041, "trajectory_balance_rmse": 285.796289394488}981{"canonical_length": 150, "canonical_nll": 1.5840472412109374, "generation_rouge_l_f1": 0.1766109785202864, "generation_token_f1": 0.39618138424821003, "prompt_id": "68f61446f4cfbca0", "reading_canonical_log_prob": -237.60708618164062, "reading_distances": [20, 26, 25, 32], "reading_log_probs": [-242.44003295898438, -238.77468872070312, -241.42832946777344, -239.21719360351562], "reading_logprob_std": 1.7552712859275277, "tokenization_path_logprob_std": 216.88814585687604, "trajectory_balance_rmse": 2964.8097451947447}982{"canonical_length": 206, "canonical_nll": 1.380712416565534, "generation_rouge_l_f1": 0.24539877300613494, "generation_token_f1": 0.38036809815950917, "prompt_id": "6a128630a7f46740", "reading_canonical_log_prob": -284.4267578125, "reading_distances": [26, 28, 25, 28], "reading_log_probs": [-283.3546447753906, -283.4618835449219, -284.12158203125, -285.49713134765625], "reading_logprob_std": 0.98571096396284, "tokenization_path_logprob_std": 131.41942195751992, "trajectory_balance_rmse": 4234.201505402516}983{"canonical_length": 324, "canonical_nll": 1.2371469191562983, "generation_rouge_l_f1": 0.19130434782608693, "generation_token_f1": 0.3652173913043479, "prompt_id": "fc8ef010d8778b0f", "reading_canonical_log_prob": -400.8356018066406, "reading_distances": [41, 46, 39, 39], "reading_log_probs": [-405.92584228515625, -405.77935791015625, -408.5776062011719, -404.79638671875], "reading_logprob_std": 1.6182044074230593, "tokenization_path_logprob_std": 283.8285523481133, "trajectory_balance_rmse": 7262.1711108888285}984{"canonical_length": 54, "canonical_nll": 0.4472892196090133, "generation_rouge_l_f1": 0.5217391304347827, "generation_token_f1": 0.5797101449275361, "prompt_id": "68eb421daa21c33e", "reading_canonical_log_prob": -24.15361785888672, "reading_distances": [20, 21, 24, 17], "reading_log_probs": [-25.4620361328125, -21.392616271972656, -24.48172378540039, -27.03857421875], "reading_logprob_std": 2.3798384599154327, "tokenization_path_logprob_std": 48.20344526135814, "trajectory_balance_rmse": 686.2764466593791}985{"canonical_length": 133, "canonical_nll": 1.035209254214638, "generation_rouge_l_f1": 0.3557692307692307, "generation_token_f1": 0.5480769230769231, "prompt_id": "a12c87820d75347d", "reading_canonical_log_prob": -137.68283081054688, "reading_distances": [22, 28, 23, 20], "reading_log_probs": [-139.33572387695312, -141.76797485351562, -143.54095458984375, -141.2254180908203], "reading_logprob_std": 1.731335915991782, "tokenization_path_logprob_std": 81.65878301580624, "trajectory_balance_rmse": 2528.6236798923046}986{"canonical_length": 220, "canonical_nll": 0.0498377496545965, "generation_rouge_l_f1": 0.8924731182795699, "generation_token_f1": 0.9032258064516129, "prompt_id": "c386599fdaa650e0", "reading_canonical_log_prob": -10.96430492401123, "reading_distances": [14, 10, 11, 12], "reading_log_probs": [-14.03879451751709, -14.13764762878418, -14.15438461303711, -15.91136360168457], "reading_logprob_std": 0.9019872333021989, "tokenization_path_logprob_std": 154.48082532721043, "trajectory_balance_rmse": 2489.65633977571}987{"canonical_length": 13, "canonical_nll": 1.0687343157254732, "generation_rouge_l_f1": 0.3157894736842105, "generation_token_f1": 0.3157894736842105, "prompt_id": "28aa44aad6506016", "reading_canonical_log_prob": -13.893546104431152, "reading_distances": [24, 28, 30, 33], "reading_log_probs": [-19.765056610107422, -17.500770568847656, -14.13548469543457, -15.170601844787598], "reading_logprob_std": 2.5125416848109756, "tokenization_path_logprob_std": 38.83506644582159, "trajectory_balance_rmse": 273.90001456878394}988{"canonical_length": 150, "canonical_nll": 1.3763034057617187, "generation_rouge_l_f1": 0.26107226107226106, "generation_token_f1": 0.4289044289044289, "prompt_id": "764869f6f55c420d", "reading_canonical_log_prob": -206.4455108642578, "reading_distances": [11, 13, 15, 13], "reading_log_probs": [-210.1085662841797, -206.5556182861328, -206.43124389648438, -203.40347290039062], "reading_logprob_std": 2.7420081362577595, "tokenization_path_logprob_std": 136.65815166591327, "trajectory_balance_rmse": 2435.3614826664248}989{"canonical_length": 266, "canonical_nll": 0.7573311597781074, "generation_rouge_l_f1": 0.3431952662721893, "generation_token_f1": 0.5364891518737673, "prompt_id": "8fb6a1df0bc7d753", "reading_canonical_log_prob": -201.45008850097656, "reading_distances": [44, 32, 42, 44], "reading_log_probs": [-213.2329864501953, -213.01451110839844, -215.59774780273438, -215.38772583007812], "reading_logprob_std": 1.373320146209321, "tokenization_path_logprob_std": 173.8889912358208, "trajectory_balance_rmse": 4283.193716321541}990{"canonical_length": 81, "canonical_nll": 1.7153218587239583, "generation_rouge_l_f1": 0.1941176470588235, "generation_token_f1": 0.2823529411764706, "prompt_id": "f5aeb1160ca2b326", "reading_canonical_log_prob": -138.94107055664062, "reading_distances": [52, 48, 48, 53], "reading_log_probs": [-147.32296752929688, -143.7086181640625, -145.85177612304688, -145.63137817382812], "reading_logprob_std": 1.4840290704604635, "tokenization_path_logprob_std": 123.51526653493941, "trajectory_balance_rmse": 1721.4780119965392}991{"canonical_length": 29, "canonical_nll": 0.8835405152419518, "generation_rouge_l_f1": 0.46938775510204084, "generation_token_f1": 0.48979591836734687, "prompt_id": "7d27494a4e6427f6", "reading_canonical_log_prob": -25.6226749420166, "reading_distances": [38, 36, 37, 27], "reading_log_probs": [-25.16912078857422, -27.638965606689453, -26.95913314819336, -27.541013717651367], "reading_logprob_std": 1.1452985593820024, "tokenization_path_logprob_std": 28.505739069134062, "trajectory_balance_rmse": 528.6257909500721}992{"canonical_length": 13, "canonical_nll": 0.3535900115966797, "generation_rouge_l_f1": 0.8421052631578948, "generation_token_f1": 0.8421052631578948, "prompt_id": "b015ce761d144f7f", "reading_canonical_log_prob": -4.596670150756836, "reading_distances": [29, 23, 31, 25], "reading_log_probs": [-6.134666919708252, -5.377636432647705, -6.664711952209473, -6.686061382293701], "reading_logprob_std": 0.6142118112036276, "tokenization_path_logprob_std": 54.34840060421439, "trajectory_balance_rmse": 184.26172852469116}993{"canonical_length": 246, "canonical_nll": 1.409233031234121, "generation_rouge_l_f1": 0.26459143968871596, "generation_token_f1": 0.49416342412451364, "prompt_id": "e7d47f51d6085f91", "reading_canonical_log_prob": -346.67132568359375, "reading_distances": [24, 16, 22, 18], "reading_log_probs": [-357.73480224609375, -358.8616638183594, -356.4908752441406, -351.8775329589844], "reading_logprob_std": 3.0660292488206258, "tokenization_path_logprob_std": 194.38023866925928, "trajectory_balance_rmse": 5716.768526970191}994{"canonical_length": 205, "canonical_nll": 1.5418461497237044, "generation_rouge_l_f1": 0.28837209302325584, "generation_token_f1": 0.5627906976744186, "prompt_id": "321f893720a9d001", "reading_canonical_log_prob": -316.0784606933594, "reading_distances": [29, 23, 37, 23], "reading_log_probs": [-319.5018615722656, -316.3453063964844, -317.2861022949219, -318.25299072265625], "reading_logprob_std": 1.3506892621428406, "tokenization_path_logprob_std": 387.38530700796963, "trajectory_balance_rmse": 4715.169961245625}995{"canonical_length": 159, "canonical_nll": 1.2211730765096795, "generation_rouge_l_f1": 0.30267062314540055, "generation_token_f1": 0.5163204747774481, "prompt_id": "46617b87eb38bf31", "reading_canonical_log_prob": -194.16651916503906, "reading_distances": [34, 38, 37, 34], "reading_log_probs": [-201.6573486328125, -192.08018493652344, -195.66741943359375, -198.44711303710938], "reading_logprob_std": 4.072669580687809, "tokenization_path_logprob_std": 107.06863634051093, "trajectory_balance_rmse": 1831.1870212250813}996{"canonical_length": 4, "canonical_nll": 0.7052232623100281, "generation_rouge_l_f1": 0.04081632653061225, "generation_token_f1": 0.04081632653061225, "prompt_id": "03d95aafdc590416", "reading_canonical_log_prob": -2.8208930492401123, "reading_distances": [38, 34, 23, 35], "reading_log_probs": [-4.603021144866943, -4.259190082550049, -4.627546310424805, -3.575974941253662], "reading_logprob_std": 0.49006094620017165, "tokenization_path_logprob_std": 19.482936776841495, "trajectory_balance_rmse": 69.48325201821048}997{"canonical_length": 153, "canonical_nll": 0.6095874761444291, "generation_rouge_l_f1": 0.34408602150537637, "generation_token_f1": 0.5483870967741936, "prompt_id": "3e79c37d02563d09", "reading_canonical_log_prob": -93.26688385009766, "reading_distances": [39, 36, 39, 30], "reading_log_probs": [-112.94430541992188, -117.19778442382812, -111.33922576904297, -118.29144287109375], "reading_logprob_std": 3.330570073687869, "tokenization_path_logprob_std": 119.97001609000017, "trajectory_balance_rmse": 2161.8138925013577}998{"canonical_length": 385, "canonical_nll": 0.971715040330763, "generation_rouge_l_f1": 0.19047619047619047, "generation_token_f1": 0.4217687074829932, "prompt_id": "8a6a50a6d59a29d8", "reading_canonical_log_prob": -374.11029052734375, "reading_distances": [35, 25, 33, 31], "reading_log_probs": [-402.3468017578125, -401.1480712890625, -396.1358642578125, -390.9732666015625], "reading_logprob_std": 5.201532511587911, "tokenization_path_logprob_std": 201.8928895804661, "trajectory_balance_rmse": 7198.508853301693}999{"canonical_length": 28, "canonical_nll": 0.7999698775155204, "generation_rouge_l_f1": 0.2676056338028169, "generation_token_f1": 0.29577464788732394, "prompt_id": "280d0259825e43a6", "reading_canonical_log_prob": -22.39915657043457, "reading_distances": [32, 33, 36, 36], "reading_log_probs": [-23.85266876220703, -25.100730895996094, -24.005046844482422, -24.424297332763672], "reading_logprob_std": 0.5583826548289156, "tokenization_path_logprob_std": 97.97095973912788, "trajectory_balance_rmse": 409.76869004337755}1000{"canonical_length": 245, "canonical_nll": 1.6818837691326531, "generation_rouge_l_f1": 0.15324165029469547, "generation_token_f1": 0.33398821218074654, "prompt_id": "fcecb04380951f2a", "reading_canonical_log_prob": -412.0615234375, "reading_distances": [26, 23, 25, 26], "reading_log_probs": [-415.69952392578125, -414.77667236328125, -416.2493896484375, -414.6155700683594], "reading_logprob_std": 0.7742297169905112, "tokenization_path_logprob_std": 178.45003493391204, "trajectory_balance_rmse": 5442.7653811761165}1001{"canonical_length": 126, "canonical_nll": 1.6398107135106648, "generation_rouge_l_f1": 0.21022727272727273, "generation_token_f1": 0.31818181818181823, "prompt_id": "1ae061f91192abf8", "reading_canonical_log_prob": -206.61614990234375, "reading_distances": [29, 32, 30, 21], "reading_log_probs": [-226.31370544433594, -213.54978942871094, -226.6016845703125, -234.0084228515625], "reading_logprob_std": 8.49498175539005, "tokenization_path_logprob_std": 131.13979837684008, "trajectory_balance_rmse": 2458.335920018982}1002{"canonical_length": 205, "canonical_nll": 1.269740555926067, "generation_rouge_l_f1": 0.184, "generation_token_f1": 0.32, "prompt_id": "eedaf90695c8e20b", "reading_canonical_log_prob": -260.29681396484375, "reading_distances": [16, 17, 16, 21], "reading_log_probs": [-262.2005310058594, -263.95159912109375, -259.31951904296875, -258.81134033203125], "reading_logprob_std": 2.4322444156413057, "tokenization_path_logprob_std": 90.30073591740707, "trajectory_balance_rmse": 4507.938391821997}1003{"canonical_length": 119, "canonical_nll": 1.1565241453026522, "generation_rouge_l_f1": 0.19169329073482427, "generation_token_f1": 0.35143769968051114, "prompt_id": "a0b38c7913178787", "reading_canonical_log_prob": -137.62637329101562, "reading_distances": [48, 44, 45, 41], "reading_log_probs": [-121.50823974609375, -119.39783477783203, -121.41709899902344, -125.32916259765625], "reading_logprob_std": 2.4769585937930874, "tokenization_path_logprob_std": 98.04276930686918, "trajectory_balance_rmse": 1501.2545014637024}1004{"canonical_length": 68, "canonical_nll": 1.6770899155560661, "generation_rouge_l_f1": 0.24161073825503357, "generation_token_f1": 0.33557046979865773, "prompt_id": "4101000b96f3bbde", "reading_canonical_log_prob": -114.0421142578125, "reading_distances": [18, 19, 23, 15], "reading_log_probs": [-125.38677215576172, -121.03524780273438, -119.80836486816406, -119.40452575683594], "reading_logprob_std": 2.7411865769031865, "tokenization_path_logprob_std": 77.51909518447216, "trajectory_balance_rmse": 1171.1222738706108}1005{"canonical_length": 138, "canonical_nll": 1.3326126319774683, "generation_rouge_l_f1": 0.24999999999999994, "generation_token_f1": 0.45918367346938777, "prompt_id": "5539256ae94ad40f", "reading_canonical_log_prob": -183.90054321289062, "reading_distances": [63, 57, 65, 62], "reading_log_probs": [-193.08404541015625, -191.53814697265625, -195.27017211914062, -186.64682006835938], "reading_logprob_std": 3.6608527380729554, "tokenization_path_logprob_std": 281.91015470086893, "trajectory_balance_rmse": 2709.43923608802}1006{"canonical_length": 86, "canonical_nll": 1.2001984174861464, "generation_rouge_l_f1": 0.23283582089552238, "generation_token_f1": 0.35820895522388063, "prompt_id": "3b0f78eeffb87a9e", "reading_canonical_log_prob": -103.2170639038086, "reading_distances": [32, 33, 33, 35], "reading_log_probs": [-100.01875305175781, -103.98897552490234, -104.69995880126953, -102.9684066772461], "reading_logprob_std": 2.059978580312789, "tokenization_path_logprob_std": 27.378071950097574, "trajectory_balance_rmse": 1656.402122441084}1007{"canonical_length": 302, "canonical_nll": 1.4727215293227442, "generation_rouge_l_f1": 0.24456521739130435, "generation_token_f1": 0.3315217391304348, "prompt_id": "3a8d8459792a6678", "reading_canonical_log_prob": -444.76190185546875, "reading_distances": [14, 16, 18, 17], "reading_log_probs": [-439.8047180175781, -442.18621826171875, -439.7972717285156, -436.8725280761719], "reading_logprob_std": 2.174968027944168, "tokenization_path_logprob_std": 189.9044360752718, "trajectory_balance_rmse": 6508.597126604883}1008{"canonical_length": 289, "canonical_nll": 1.1199453809269573, "generation_rouge_l_f1": 0.23394495412844038, "generation_token_f1": 0.44954128440366975, "prompt_id": "a788a16b80a319d9", "reading_canonical_log_prob": -323.6642150878906, "reading_distances": [12, 20, 16, 20], "reading_log_probs": [-327.4328308105469, -329.73480224609375, -330.92010498046875, -329.9862976074219], "reading_logprob_std": 1.4810057114413906, "tokenization_path_logprob_std": 224.99250859593155, "trajectory_balance_rmse": 5714.962962085638}1009{"canonical_length": 252, "canonical_nll": 1.305753919813368, "generation_rouge_l_f1": 0.1845574387947269, "generation_token_f1": 0.42561205273069674, "prompt_id": "79833cd8d7978d26", "reading_canonical_log_prob": -329.04998779296875, "reading_distances": [30, 30, 26, 31], "reading_log_probs": [-325.6971435546875, -322.97845458984375, -334.2645568847656, -330.82208251953125], "reading_logprob_std": 5.064632876752994, "tokenization_path_logprob_std": 190.1455167816199, "trajectory_balance_rmse": 4237.6892487645755}1010{"canonical_length": 26, "canonical_nll": 1.674670439500075, "generation_rouge_l_f1": 0.08270676691729323, "generation_token_f1": 0.12030075187969924, "prompt_id": "7c29635f759f46ec", "reading_canonical_log_prob": -43.54143142700195, "reading_distances": [18, 22, 22, 16], "reading_log_probs": [-40.06965637207031, -39.08053207397461, -38.75297546386719, -39.904815673828125], "reading_logprob_std": 0.6359153902054822, "tokenization_path_logprob_std": 23.381595735958896, "trajectory_balance_rmse": 535.6443795121451}1011{"canonical_length": 346, "canonical_nll": 1.2971622709594022, "generation_rouge_l_f1": 0.34087481146304677, "generation_token_f1": 0.539969834087481, "prompt_id": "88945c2caccedba1", "reading_canonical_log_prob": -448.8181457519531, "reading_distances": [39, 36, 53, 44], "reading_log_probs": [-452.2125244140625, -459.90972900390625, -459.7850646972656, -460.12713623046875], "reading_logprob_std": 3.8666439482276154, "tokenization_path_logprob_std": 100.40065974647902, "trajectory_balance_rmse": 7365.8113973358495}1012{"canonical_length": 17, "canonical_nll": 1.0379810333251953, "generation_rouge_l_f1": 0.6153846153846153, "generation_token_f1": 0.7692307692307692, "prompt_id": "f3be16dc1ce038fa", "reading_canonical_log_prob": -17.64567756652832, "reading_distances": [34, 49, 48, 47], "reading_log_probs": [-17.130901336669922, -16.792709350585938, -19.9704532623291, -19.357433319091797], "reading_logprob_std": 1.5860467752867382, "tokenization_path_logprob_std": 63.018706506373626, "trajectory_balance_rmse": 448.9218140742737}1013{"canonical_length": 56, "canonical_nll": 1.297546659197126, "generation_rouge_l_f1": 0.33999999999999997, "generation_token_f1": 0.4, "prompt_id": "08f613a2192f5ff2", "reading_canonical_log_prob": -72.66261291503906, "reading_distances": [23, 24, 20, 20], "reading_log_probs": [-72.68090057373047, -71.22840881347656, -76.32035827636719, -71.14568328857422], "reading_logprob_std": 2.4225388003415347, "tokenization_path_logprob_std": 23.45040886051946, "trajectory_balance_rmse": 1157.4836826699884}1014{"canonical_length": 180, "canonical_nll": 1.784307861328125, "generation_rouge_l_f1": 0.2534562211981567, "generation_token_f1": 0.4055299539170507, "prompt_id": "ddaf5127b3985530", "reading_canonical_log_prob": -321.1754150390625, "reading_distances": [26, 27, 31, 27], "reading_log_probs": [-327.86102294921875, -338.4233703613281, -329.6527404785156, -330.294921875], "reading_logprob_std": 4.691351251877994, "tokenization_path_logprob_std": 86.28096417309833, "trajectory_balance_rmse": 3650.851444948136}1015{"canonical_length": 184, "canonical_nll": 1.0852619668711787, "generation_rouge_l_f1": 0.23797468354430382, "generation_token_f1": 0.40506329113924044, "prompt_id": "6fc1da282516cfc1", "reading_canonical_log_prob": -199.68820190429688, "reading_distances": [18, 12, 9, 14], "reading_log_probs": [-215.5947723388672, -211.42083740234375, -206.1622772216797, -210.49908447265625], "reading_logprob_std": 3.869428387900922, "tokenization_path_logprob_std": 204.04886131976096, "trajectory_balance_rmse": 3509.8176858501565}1016{"canonical_length": 80, "canonical_nll": 1.2141057968139648, "generation_rouge_l_f1": 0.13836477987421383, "generation_token_f1": 0.25786163522012584, "prompt_id": "9d2634370eff9e7e", "reading_canonical_log_prob": -97.12846374511719, "reading_distances": [39, 26, 33, 34], "reading_log_probs": [-127.24267578125, -121.67189025878906, -116.47618865966797, -126.21986389160156], "reading_logprob_std": 4.921175560230517, "tokenization_path_logprob_std": 151.75678764669405, "trajectory_balance_rmse": 1798.977635398718}1017{"canonical_length": 132, "canonical_nll": 0.782440185546875, "generation_rouge_l_f1": 0.4280442804428044, "generation_token_f1": 0.6420664206642067, "prompt_id": "8a81f9867fa3fdd7", "reading_canonical_log_prob": -103.2821044921875, "reading_distances": [33, 32, 31, 34], "reading_log_probs": [-108.18905639648438, -111.11659240722656, -112.76344299316406, -108.84872436523438], "reading_logprob_std": 2.103786386162294, "tokenization_path_logprob_std": 72.64999368595076, "trajectory_balance_rmse": 1771.692799165525}1018{"canonical_length": 33, "canonical_nll": 1.0860795685739228, "generation_rouge_l_f1": 0.71875, "generation_token_f1": 0.75, "prompt_id": "32094cae1c9e7173", "reading_canonical_log_prob": -35.84062576293945, "reading_distances": [44, 46, 39, 44], "reading_log_probs": [-39.136314392089844, -39.072017669677734, -39.998878479003906, -36.17654800415039], "reading_logprob_std": 1.6673680294246849, "tokenization_path_logprob_std": 57.01990875758484, "trajectory_balance_rmse": 639.638560921968}1019{"canonical_length": 23, "canonical_nll": 1.7080831113068953, "generation_rouge_l_f1": 0.10309278350515463, "generation_token_f1": 0.11340206185567009, "prompt_id": "896dee39b103c9ac", "reading_canonical_log_prob": -39.285911560058594, "reading_distances": [27, 29, 30, 26], "reading_log_probs": [-36.11298370361328, -51.51982879638672, -51.14982986450195, -48.467132568359375], "reading_logprob_std": 7.261515742851692, "tokenization_path_logprob_std": 87.92491619977568, "trajectory_balance_rmse": 616.1459011950775}1020{"canonical_length": 105, "canonical_nll": 1.7438334147135417, "generation_rouge_l_f1": 0.18617021276595744, "generation_token_f1": 0.33510638297872336, "prompt_id": "1057782952469d89", "reading_canonical_log_prob": -183.10250854492188, "reading_distances": [11, 10, 10, 9], "reading_log_probs": [-190.3383331298828, -188.279052734375, -186.85043334960938, -183.5902557373047], "reading_logprob_std": 2.83721045234597, "tokenization_path_logprob_std": 110.13940399704174, "trajectory_balance_rmse": 2573.428813566016}1021{"canonical_length": 25, "canonical_nll": 0.723299789428711, "generation_rouge_l_f1": 0.10561056105610563, "generation_token_f1": 0.1254125412541254, "prompt_id": "03db11358469fd8f", "reading_canonical_log_prob": -18.082494735717773, "reading_distances": [19, 20, 25, 24], "reading_log_probs": [-21.748138427734375, -21.08578872680664, -22.298431396484375, -21.687313079833984], "reading_logprob_std": 0.4959022576136405, "tokenization_path_logprob_std": 31.450999792561305, "trajectory_balance_rmse": 232.07486534854456}1022{"canonical_length": 240, "canonical_nll": 1.2088943481445313, "generation_rouge_l_f1": 0.21074380165289258, "generation_token_f1": 0.5247933884297521, "prompt_id": "85339fcc838b71e5", "reading_canonical_log_prob": -290.1346435546875, "reading_distances": [21, 19, 22, 25], "reading_log_probs": [-289.04901123046875, -281.95428466796875, -285.04595947265625, -287.93792724609375], "reading_logprob_std": 3.179622425121081, "tokenization_path_logprob_std": 175.66262689006234, "trajectory_balance_rmse": 5883.977894227989}1023{"canonical_length": 28, "canonical_nll": 0.47996207645961214, "generation_rouge_l_f1": 0.25, "generation_token_f1": 0.275, "prompt_id": "5af4615607f58b8b", "reading_canonical_log_prob": -13.43893814086914, "reading_distances": [47, 45, 38, 39], "reading_log_probs": [-18.093000411987305, -17.239721298217773, -17.182279586791992, -16.713199615478516], "reading_logprob_std": 0.5745946801572095, "tokenization_path_logprob_std": 89.47365611118524, "trajectory_balance_rmse": 618.0415886082958}1024{"canonical_length": 65, "canonical_nll": 2.0002483661358172, "generation_rouge_l_f1": 0.2649006622516556, "generation_token_f1": 0.4105960264900662, "prompt_id": "13f18463e2c1c5f7", "reading_canonical_log_prob": -130.01614379882812, "reading_distances": [25, 18, 20, 18], "reading_log_probs": [-128.73880004882812, -127.22338104248047, -127.25580596923828, -128.72296142578125], "reading_logprob_std": 0.8611211152908864, "tokenization_path_logprob_std": 57.71413590506524, "trajectory_balance_rmse": 1435.3170291929332}1025{"canonical_length": 64, "canonical_nll": 1.8651926517486572, "generation_rouge_l_f1": 0.16413373860182373, "generation_token_f1": 0.2613981762917933, "prompt_id": "d02d88b60737f21f", "reading_canonical_log_prob": -119.37232971191406, "reading_distances": [21, 23, 16, 19], "reading_log_probs": [-125.92314910888672, -133.63673400878906, -133.38494873046875, -130.1039581298828], "reading_logprob_std": 3.605156148129826, "tokenization_path_logprob_std": 29.31657781252133, "trajectory_balance_rmse": 1452.411739944483}1026{"canonical_length": 236, "canonical_nll": 1.5218880863512976, "generation_rouge_l_f1": 0.21157684630738524, "generation_token_f1": 0.4311377245508982, "prompt_id": "559f28b026670bdd", "reading_canonical_log_prob": -359.16558837890625, "reading_distances": [19, 13, 16, 14], "reading_log_probs": [-364.61907958984375, -357.3251953125, -363.4163818359375, -361.30596923828125], "reading_logprob_std": 3.2019120014973965, "tokenization_path_logprob_std": 135.12413707953718, "trajectory_balance_rmse": 5317.631968498965}1027{"canonical_length": 50, "canonical_nll": 0.9023011016845703, "generation_rouge_l_f1": 0.44827586206896547, "generation_token_f1": 0.5689655172413793, "prompt_id": "44cafb7b8f27398a", "reading_canonical_log_prob": -45.115055084228516, "reading_distances": [30, 23, 28, 26], "reading_log_probs": [-42.48210906982422, -41.97079849243164, -43.03776931762695, -43.08708572387695], "reading_logprob_std": 0.5262405662117051, "tokenization_path_logprob_std": 62.313456259651446, "trajectory_balance_rmse": 963.0484626370076}1028{"canonical_length": 274, "canonical_nll": 1.0878891770857093, "generation_rouge_l_f1": 0.2448979591836735, "generation_token_f1": 0.48237476808905383, "prompt_id": "55f3c7490c4e7162", "reading_canonical_log_prob": -298.0816345214844, "reading_distances": [14, 17, 22, 21], "reading_log_probs": [-302.7383117675781, -305.1336975097656, -305.3734130859375, -306.3791809082031], "reading_logprob_std": 1.5426660169059514, "tokenization_path_logprob_std": 301.3509789001351, "trajectory_balance_rmse": 4883.297482160699}1029{"canonical_length": 308, "canonical_nll": 1.1798935927353895, "generation_rouge_l_f1": 0.24179620034542315, "generation_token_f1": 0.40759930915371334, "prompt_id": "41cd8f9fa13781b8", "reading_canonical_log_prob": -363.4072265625, "reading_distances": [28, 27, 31, 17], "reading_log_probs": [-374.8011169433594, -371.0362854003906, -379.9033203125, -374.30963134765625], "reading_logprob_std": 3.663746396748457, "tokenization_path_logprob_std": 219.86850195154062, "trajectory_balance_rmse": 6423.44009363113}1030{"canonical_length": 373, "canonical_nll": 1.2715230742344588, "generation_rouge_l_f1": 0.20421393841166938, "generation_token_f1": 0.40842787682333875, "prompt_id": "28de6d5b6a032c7b", "reading_canonical_log_prob": -474.2781066894531, "reading_distances": [26, 21, 23, 25], "reading_log_probs": [-473.7197570800781, -472.92474365234375, -472.9901428222656, -470.7062683105469], "reading_logprob_std": 1.3034406570828225, "tokenization_path_logprob_std": 226.128323477807, "trajectory_balance_rmse": 6103.413239303773}1031{"canonical_length": 132, "canonical_nll": 1.7377508914831914, "generation_rouge_l_f1": 0.19774011299435026, "generation_token_f1": 0.35593220338983045, "prompt_id": "dc17c375b90b9957", "reading_canonical_log_prob": -229.38311767578125, "reading_distances": [23, 24, 22, 18], "reading_log_probs": [-226.06581115722656, -225.94590759277344, -228.12107849121094, -226.9673309326172], "reading_logprob_std": 1.0065223167703141, "tokenization_path_logprob_std": 253.456940937009, "trajectory_balance_rmse": 2925.0580938536827}1032{"canonical_length": 24, "canonical_nll": 0.5045696496963501, "generation_rouge_l_f1": 0.8444444444444443, "generation_token_f1": 0.8444444444444443, "prompt_id": "c1a810543bd499fe", "reading_canonical_log_prob": -12.109671592712402, "reading_distances": [28, 25, 24, 27], "reading_log_probs": [-12.936262130737305, -10.894858360290527, -11.37703800201416, -11.927830696105957], "reading_logprob_std": 0.8764675712415635, "tokenization_path_logprob_std": 41.0108104097279, "trajectory_balance_rmse": 388.95591850079364}1033{"canonical_length": 15, "canonical_nll": 0.8746392567952473, "generation_rouge_l_f1": 0.34782608695652173, "generation_token_f1": 0.34782608695652173, "prompt_id": "f21bbbf6325c37df", "reading_canonical_log_prob": -13.119588851928711, "reading_distances": [42, 39, 38, 36], "reading_log_probs": [-13.606658935546875, -12.837121963500977, -12.893911361694336, -14.07944107055664], "reading_logprob_std": 0.5969213859923634, "tokenization_path_logprob_std": 56.562067975785865, "trajectory_balance_rmse": 321.0940913959869}1034{"canonical_length": 141, "canonical_nll": 1.700297254197141, "generation_rouge_l_f1": 0.22164948453608244, "generation_token_f1": 0.422680412371134, "prompt_id": "ff640b48674add94", "reading_canonical_log_prob": -239.74191284179688, "reading_distances": [48, 54, 50, 41], "reading_log_probs": [-249.4596405029297, -251.47970581054688, -254.4873809814453, -244.80442810058594], "reading_logprob_std": 4.066067506499793, "tokenization_path_logprob_std": 159.92025870106, "trajectory_balance_rmse": 3363.625122878226}1035{"canonical_length": 133, "canonical_nll": 2.0493416463522087, "generation_rouge_l_f1": 0.18639798488664988, "generation_token_f1": 0.3324937027707808, "prompt_id": "294ad70d174d121a", "reading_canonical_log_prob": -272.56243896484375, "reading_distances": [29, 22, 27, 28], "reading_log_probs": [-285.99700927734375, -285.01025390625, -285.8017272949219, -284.4919128417969], "reading_logprob_std": 0.7004675808002687, "tokenization_path_logprob_std": 360.450696684168, "trajectory_balance_rmse": 3537.954080497838}1036{"canonical_length": 49, "canonical_nll": 2.463639551279496, "generation_rouge_l_f1": 0.2857142857142857, "generation_token_f1": 0.4857142857142857, "prompt_id": "fdd72207f77ed671", "reading_canonical_log_prob": -120.71833801269531, "reading_distances": [19, 23, 20, 23], "reading_log_probs": [-122.53704833984375, -121.38117218017578, -124.81944274902344, -123.13701629638672], "reading_logprob_std": 1.4329627617280145, "tokenization_path_logprob_std": 84.66951807170733, "trajectory_balance_rmse": 993.8854402705509}1037{"canonical_length": 9, "canonical_nll": 1.5689780977037218, "generation_rouge_l_f1": 0.4761904761904762, "generation_token_f1": 0.5714285714285715, "prompt_id": "bead008713b25f77", "reading_canonical_log_prob": -14.120802879333496, "reading_distances": [36, 42, 33, 40], "reading_log_probs": [-20.865982055664062, -20.8564453125, -17.368921279907227, -19.65953826904297], "reading_logprob_std": 1.646394246239094, "tokenization_path_logprob_std": 62.76286308909611, "trajectory_balance_rmse": 224.671932106486}1038{"canonical_length": 353, "canonical_nll": 1.0246397023835871, "generation_rouge_l_f1": 0.21664050235478807, "generation_token_f1": 0.4175824175824176, "prompt_id": "5e76d06be1ed709a", "reading_canonical_log_prob": -361.69781494140625, "reading_distances": [31, 31, 23, 25], "reading_log_probs": [-384.16607666015625, -384.03631591796875, -376.90264892578125, -380.009521484375], "reading_logprob_std": 3.4977148074755515, "tokenization_path_logprob_std": 408.754259665412, "trajectory_balance_rmse": 6282.110961228426}1039{"canonical_length": 238, "canonical_nll": 1.4829837574678308, "generation_rouge_l_f1": 0.21763602251407127, "generation_token_f1": 0.4577861163227017, "prompt_id": "f6f871799dc3c062", "reading_canonical_log_prob": -352.95013427734375, "reading_distances": [30, 29, 31, 29], "reading_log_probs": [-364.44976806640625, -360.6031188964844, -359.6120910644531, -351.7735290527344], "reading_logprob_std": 5.317218191307821, "tokenization_path_logprob_std": 186.74246540917872, "trajectory_balance_rmse": 3851.1794933419515}1040{"canonical_length": 108, "canonical_nll": 0.8857612609863281, "generation_rouge_l_f1": 0.33333333333333326, "generation_token_f1": 0.40123456790123463, "prompt_id": "e5243c35b3e882de", "reading_canonical_log_prob": -95.66221618652344, "reading_distances": [26, 27, 23, 32], "reading_log_probs": [-95.7811508178711, -96.77645111083984, -99.61434936523438, -94.71524810791016], "reading_logprob_std": 2.1040358008458284, "tokenization_path_logprob_std": 55.24429964996057, "trajectory_balance_rmse": 1478.3873388872873}1041{"canonical_length": 147, "canonical_nll": 1.6091709266714498, "generation_rouge_l_f1": 0.187192118226601, "generation_token_f1": 0.33004926108374383, "prompt_id": "26ddb945cf9f0d7d", "reading_canonical_log_prob": -236.54812622070312, "reading_distances": [33, 39, 29, 28], "reading_log_probs": [-247.50604248046875, -244.41098022460938, -250.27545166015625, -245.61370849609375], "reading_logprob_std": 2.556045045926417, "tokenization_path_logprob_std": 361.7806274239504, "trajectory_balance_rmse": 3842.609110549213}1042{"canonical_length": 48, "canonical_nll": 1.7629261016845703, "generation_rouge_l_f1": 0.12258064516129033, "generation_token_f1": 0.16774193548387095, "prompt_id": "e4581a40938db0d9", "reading_canonical_log_prob": -84.62045288085938, "reading_distances": [19, 20, 15, 22], "reading_log_probs": [-83.92082214355469, -82.8893051147461, -81.8038558959961, -76.68573760986328], "reading_logprob_std": 3.2113030360049466, "tokenization_path_logprob_std": 84.56147525520173, "trajectory_balance_rmse": 1052.2406811059157}1043{"canonical_length": 10, "canonical_nll": 0.08796284794807434, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "267c9641a2f8cfb9", "reading_canonical_log_prob": -0.8796284794807434, "reading_distances": [22, 25, 25, 23], "reading_log_probs": [-4.323722839355469, -2.546391487121582, -6.408597469329834, -5.4110517501831055], "reading_logprob_std": 1.653428358800354, "tokenization_path_logprob_std": 61.12025165440978, "trajectory_balance_rmse": 194.91685984567965}1044{"canonical_length": 63, "canonical_nll": 0.7804948110429067, "generation_rouge_l_f1": 0.36042402826855124, "generation_token_f1": 0.4381625441696113, "prompt_id": "df9c5775b476f764", "reading_canonical_log_prob": -49.171173095703125, "reading_distances": [29, 29, 30, 25], "reading_log_probs": [-51.3177490234375, -52.518310546875, -50.25736999511719, -50.262168884277344], "reading_logprob_std": 1.075563863557099, "tokenization_path_logprob_std": 86.81339790629498, "trajectory_balance_rmse": 905.6394139124088}1045{"canonical_length": 106, "canonical_nll": 1.8954994633512676, "generation_rouge_l_f1": 0.16666666666666669, "generation_token_f1": 0.3148148148148148, "prompt_id": "518ad8fe3c294839", "reading_canonical_log_prob": -200.92294311523438, "reading_distances": [18, 25, 21, 17], "reading_log_probs": [-204.64645385742188, -213.2547607421875, -207.02967834472656, -206.05191040039062], "reading_logprob_std": 3.8007318212838728, "tokenization_path_logprob_std": 53.01350913119842, "trajectory_balance_rmse": 2401.0346325131595}1046{"canonical_length": 203, "canonical_nll": 0.6925949322179034, "generation_rouge_l_f1": 0.38048780487804873, "generation_token_f1": 0.5756097560975609, "prompt_id": "6369662548b008b0", "reading_canonical_log_prob": -140.59677124023438, "reading_distances": [23, 23, 26, 23], "reading_log_probs": [-147.28628540039062, -148.78050231933594, -148.00576782226562, -144.4176483154297], "reading_logprob_std": 1.9036965811536004, "tokenization_path_logprob_std": 130.30674271257197, "trajectory_balance_rmse": 1950.6594980882878}1047{"canonical_length": 68, "canonical_nll": 0.34849901760325713, "generation_rouge_l_f1": 0.2459546925566343, "generation_token_f1": 0.3042071197411003, "prompt_id": "3f0899fb36d7151b", "reading_canonical_log_prob": -23.697933197021484, "reading_distances": [32, 35, 29, 37], "reading_log_probs": [-31.352413177490234, -39.033748626708984, -35.723670959472656, -38.69707489013672], "reading_logprob_std": 3.558627531039933, "tokenization_path_logprob_std": 83.04620064403751, "trajectory_balance_rmse": 1279.6742816040392}1048{"canonical_length": 59, "canonical_nll": 0.8161825406349311, "generation_rouge_l_f1": 0.33628318584070793, "generation_token_f1": 0.47787610619469023, "prompt_id": "ca44a0ae267d5905", "reading_canonical_log_prob": -48.15476989746094, "reading_distances": [16, 19, 18, 20], "reading_log_probs": [-61.16929626464844, -61.049190521240234, -59.11278533935547, -56.873985290527344], "reading_logprob_std": 2.018406722010568, "tokenization_path_logprob_std": 72.49328823409648, "trajectory_balance_rmse": 1091.1559308535602}1049{"canonical_length": 144, "canonical_nll": 2.5788182152642145, "generation_rouge_l_f1": 0.26771653543307083, "generation_token_f1": 0.39370078740157477, "prompt_id": "b3b2b51b9f629425", "reading_canonical_log_prob": -371.3498229980469, "reading_distances": [9, 12, 15, 11], "reading_log_probs": [-397.43182373046875, -408.5779113769531, -403.2015380859375, -394.29693603515625], "reading_logprob_std": 6.321233915602566, "tokenization_path_logprob_std": 160.94454294478751, "trajectory_balance_rmse": 2964.2983033704472}1050{"canonical_length": 120, "canonical_nll": 1.1903419494628906, "generation_rouge_l_f1": 0.15463917525773196, "generation_token_f1": 0.365979381443299, "prompt_id": "4412ea6c558d478d", "reading_canonical_log_prob": -142.84103393554688, "reading_distances": [24, 26, 31, 24], "reading_log_probs": [-145.104248046875, -144.35198974609375, -146.4036865234375, -146.0590362548828], "reading_logprob_std": 0.9313209563375833, "tokenization_path_logprob_std": 58.19290251480254, "trajectory_balance_rmse": 2414.3590223559313}1051{"canonical_length": 30, "canonical_nll": 0.9533123016357422, "generation_rouge_l_f1": 0.07905138339920949, "generation_token_f1": 0.13438735177865613, "prompt_id": "7938d8548e06b8bf", "reading_canonical_log_prob": -28.599369049072266, "reading_distances": [8, 9, 6, 8], "reading_log_probs": [-29.513214111328125, -29.269187927246094, -29.789987564086914, -31.502227783203125], "reading_logprob_std": 1.0116732648423183, "tokenization_path_logprob_std": 75.61635114291626, "trajectory_balance_rmse": 317.0088348667839}1052{"canonical_length": 332, "canonical_nll": 1.1430117136024567, "generation_rouge_l_f1": 0.4021543985637343, "generation_token_f1": 0.5888689407540395, "prompt_id": "8829bdef994324cc", "reading_canonical_log_prob": -379.4798889160156, "reading_distances": [34, 33, 33, 35], "reading_log_probs": [-386.4954833984375, -389.44580078125, -386.59515380859375, -385.35833740234375], "reading_logprob_std": 1.7409448518405521, "tokenization_path_logprob_std": 406.5793607738287, "trajectory_balance_rmse": 7180.028305415162}1053{"canonical_length": 14, "canonical_nll": 0.798762457711356, "generation_rouge_l_f1": 0.055999999999999994, "generation_token_f1": 0.06399999999999999, "prompt_id": "75996f4d9b7d5f4d", "reading_canonical_log_prob": -11.182674407958984, "reading_distances": [10, 18, 19, 23], "reading_log_probs": [-11.772798538208008, -12.119029998779297, -10.228736877441406, -10.448341369628906], "reading_logprob_std": 0.9429927598289528, "tokenization_path_logprob_std": 54.64570950121636, "trajectory_balance_rmse": 230.90349502686246}1054{"canonical_length": 164, "canonical_nll": 1.423182231623952, "generation_rouge_l_f1": 0.1788235294117647, "generation_token_f1": 0.32941176470588235, "prompt_id": "f5d8be75fb333542", "reading_canonical_log_prob": -233.40188598632812, "reading_distances": [18, 26, 19, 19], "reading_log_probs": [-222.75009155273438, -223.9676055908203, -229.87783813476562, -223.59469604492188], "reading_logprob_std": 3.2602179540243976, "tokenization_path_logprob_std": 186.88291058223413, "trajectory_balance_rmse": 3904.9556167615056}1055{"canonical_length": 45, "canonical_nll": 1.4235470241970487, "generation_rouge_l_f1": 0.23655913978494625, "generation_token_f1": 0.3655913978494624, "prompt_id": "61f6f6b3601f9e9d", "reading_canonical_log_prob": -64.05961608886719, "reading_distances": [19, 15, 27, 22], "reading_log_probs": [-67.61820983886719, -66.45625305175781, -81.04208374023438, -76.11770629882812], "reading_logprob_std": 6.976935830956799, "tokenization_path_logprob_std": 92.77302407549378, "trajectory_balance_rmse": 883.4008265285205}1056{"canonical_length": 26, "canonical_nll": 1.4990114065317006, "generation_rouge_l_f1": 0.5217391304347826, "generation_token_f1": 0.5217391304347826, "prompt_id": "4478590e9673bdfe", "reading_canonical_log_prob": -38.97429656982422, "reading_distances": [35, 29, 29, 31], "reading_log_probs": [-43.06544494628906, -42.73284149169922, -37.319793701171875, -42.926307678222656], "reading_logprob_std": 2.797528775404783, "tokenization_path_logprob_std": 124.49521390368533, "trajectory_balance_rmse": 886.833666008349}1057{"canonical_length": 90, "canonical_nll": 0.8542109171549479, "generation_rouge_l_f1": 0.17522658610271902, "generation_token_f1": 0.2537764350453172, "prompt_id": "b3f4ac8bd91cf595", "reading_canonical_log_prob": -76.87898254394531, "reading_distances": [30, 32, 23, 32], "reading_log_probs": [-85.47344207763672, -88.66556549072266, -92.8803939819336, -81.30076599121094], "reading_logprob_std": 4.903711599399541, "tokenization_path_logprob_std": 48.348554877969406, "trajectory_balance_rmse": 1662.7385662642944}1058{"canonical_length": 116, "canonical_nll": 1.0842450898269127, "generation_rouge_l_f1": 0.21024258760107817, "generation_token_f1": 0.42048517520215634, "prompt_id": "47d49f308f57d626", "reading_canonical_log_prob": -125.77243041992188, "reading_distances": [32, 35, 39, 36], "reading_log_probs": [-137.0469207763672, -136.58961486816406, -136.68971252441406, -134.80392456054688], "reading_logprob_std": 1.005098009868366, "tokenization_path_logprob_std": 155.5500402498612, "trajectory_balance_rmse": 2859.4683295136565}1059{"canonical_length": 264, "canonical_nll": 1.2018337827740293, "generation_rouge_l_f1": 0.23874755381604695, "generation_token_f1": 0.5401174168297455, "prompt_id": "d8021c0fb98991e7", "reading_canonical_log_prob": -317.28411865234375, "reading_distances": [15, 24, 20, 24], "reading_log_probs": [-325.6406555175781, -323.43804931640625, -325.937255859375, -326.40863037109375], "reading_logprob_std": 1.3172512508063705, "tokenization_path_logprob_std": 307.1617168443645, "trajectory_balance_rmse": 5217.195354346857}1060{"canonical_length": 165, "canonical_nll": 1.5906653201941288, "generation_rouge_l_f1": 0.1747126436781609, "generation_token_f1": 0.34022988505747126, "prompt_id": "fb0e2e105f1d060e", "reading_canonical_log_prob": -262.45977783203125, "reading_distances": [35, 27, 23, 27], "reading_log_probs": [-261.4080505371094, -264.178955078125, -257.9647521972656, -265.0614929199219], "reading_logprob_std": 3.196883587292662, "tokenization_path_logprob_std": 181.76391660128255, "trajectory_balance_rmse": 3124.423474154165}1061{"canonical_length": 12, "canonical_nll": 1.0195817947387695, "generation_rouge_l_f1": 0.5, "generation_token_f1": 0.6, "prompt_id": "d8467aeeb0933718", "reading_canonical_log_prob": -12.234981536865234, "reading_distances": [41, 42, 45, 36], "reading_log_probs": [-16.200132369995117, -16.857746124267578, -16.574880599975586, -14.592338562011719], "reading_logprob_std": 1.0124413676408013, "tokenization_path_logprob_std": 25.63186706978468, "trajectory_balance_rmse": 222.80624395772864}1062{"canonical_length": 62, "canonical_nll": 0.4315523947438886, "generation_rouge_l_f1": 0.7703703703703704, "generation_token_f1": 0.7851851851851852, "prompt_id": "b80e624242a44cf6", "reading_canonical_log_prob": -26.756248474121094, "reading_distances": [23, 29, 22, 21], "reading_log_probs": [-29.713289260864258, -27.936420440673828, -28.28411102294922, -28.471139907836914], "reading_logprob_std": 0.7737646987548187, "tokenization_path_logprob_std": 162.839674204213, "trajectory_balance_rmse": 1340.4201278020294}1063{"canonical_length": 225, "canonical_nll": 1.2395791286892361, "generation_rouge_l_f1": 0.19542619542619544, "generation_token_f1": 0.41164241164241167, "prompt_id": "aad9f0c226359f45", "reading_canonical_log_prob": -278.9053039550781, "reading_distances": [13, 20, 17, 19], "reading_log_probs": [-284.4671630859375, -288.9100036621094, -287.9930419921875, -285.00628662109375], "reading_logprob_std": 2.1882611384414408, "tokenization_path_logprob_std": 240.49735252604495, "trajectory_balance_rmse": 4404.513424054222}1064{"canonical_length": 221, "canonical_nll": 0.9436954826251414, "generation_rouge_l_f1": 0.26694045174537984, "generation_token_f1": 0.5503080082135523, "prompt_id": "2fd9ba0f00f7f678", "reading_canonical_log_prob": -208.55670166015625, "reading_distances": [36, 26, 21, 28], "reading_log_probs": [-217.77528381347656, -217.0169677734375, -213.36984252929688, -217.6605224609375], "reading_logprob_std": 2.08410144008968, "tokenization_path_logprob_std": 365.120526629954, "trajectory_balance_rmse": 4227.455449269165}1065{"canonical_length": 8, "canonical_nll": 1.6237720251083374, "generation_rouge_l_f1": 0.0975609756097561, "generation_token_f1": 0.14634146341463417, "prompt_id": "b1a346c0437945a9", "reading_canonical_log_prob": -12.9901762008667, "reading_distances": [28, 31, 24, 31], "reading_log_probs": [-14.21794605255127, -14.854523658752441, -15.379568099975586, -13.31866455078125], "reading_logprob_std": 0.8871847962837288, "tokenization_path_logprob_std": 49.81312661315018, "trajectory_balance_rmse": 164.9344868800165}1066{"canonical_length": 21, "canonical_nll": 2.3531851995558966, "generation_rouge_l_f1": 0.14906832298136646, "generation_token_f1": 0.16149068322981366, "prompt_id": "c2ce9d8e95420539", "reading_canonical_log_prob": -49.41688919067383, "reading_distances": [19, 16, 19, 22], "reading_log_probs": [-62.65310287475586, -57.15496826171875, -57.24235534667969, -57.92546081542969], "reading_logprob_std": 2.628754471166701, "tokenization_path_logprob_std": 56.888699915668454, "trajectory_balance_rmse": 329.25009071399217}1067{"canonical_length": 86, "canonical_nll": 1.2479394424793333, "generation_rouge_l_f1": 0.245014245014245, "generation_token_f1": 0.3247863247863248, "prompt_id": "ecdc1eebf63862d2", "reading_canonical_log_prob": -107.32279205322266, "reading_distances": [32, 35, 28, 36], "reading_log_probs": [-116.16070556640625, -117.29222869873047, -117.26922607421875, -116.60374450683594], "reading_logprob_std": 0.5494581526158394, "tokenization_path_logprob_std": 131.20954712886967, "trajectory_balance_rmse": 2136.3701327869553}1068{"canonical_length": 149, "canonical_nll": 0.9108795575647546, "generation_rouge_l_f1": 0.32020997375328086, "generation_token_f1": 0.4881889763779528, "prompt_id": "a473b04dd30c4a84", "reading_canonical_log_prob": -135.72105407714844, "reading_distances": [50, 50, 49, 47], "reading_log_probs": [-141.81570434570312, -144.667724609375, -142.46395874023438, -148.6134490966797], "reading_logprob_std": 3.06872862414242, "tokenization_path_logprob_std": 100.23850109818082, "trajectory_balance_rmse": 3612.9152279593272}1069{"canonical_length": 147, "canonical_nll": 1.4726192967421343, "generation_rouge_l_f1": 0.2105263157894737, "generation_token_f1": 0.31578947368421056, "prompt_id": "30cb3266dd228382", "reading_canonical_log_prob": -216.47503662109375, "reading_distances": [17, 20, 16, 12], "reading_log_probs": [-223.25173950195312, -226.3710479736328, -222.34976196289062, -220.36788940429688], "reading_logprob_std": 2.499945216658209, "tokenization_path_logprob_std": 259.2514064373747, "trajectory_balance_rmse": 2653.3872533002104}1070{"canonical_length": 151, "canonical_nll": 1.1027471277097993, "generation_rouge_l_f1": 0.22222222222222224, "generation_token_f1": 0.456140350877193, "prompt_id": "d49ee101a0892105", "reading_canonical_log_prob": -166.5148162841797, "reading_distances": [21, 17, 24, 23], "reading_log_probs": [-172.5528564453125, -176.8837890625, -173.1692352294922, -169.7387237548828], "reading_logprob_std": 2.939310014632186, "tokenization_path_logprob_std": 144.13073783623156, "trajectory_balance_rmse": 2544.1596323370754}1071{"canonical_length": 18, "canonical_nll": 0.7790531582302518, "generation_rouge_l_f1": 0.08888888888888889, "generation_token_f1": 0.12222222222222223, "prompt_id": "c17572e82dc23485", "reading_canonical_log_prob": -14.022956848144531, "reading_distances": [26, 29, 24, 21], "reading_log_probs": [-16.567026138305664, -17.181446075439453, -16.644481658935547, -15.223287582397461], "reading_logprob_std": 0.8332492921843482, "tokenization_path_logprob_std": 45.12111039427707, "trajectory_balance_rmse": 488.7114619462669}1072{"canonical_length": 267, "canonical_nll": 1.4420511195722143, "generation_rouge_l_f1": 0.24045801526717558, "generation_token_f1": 0.5458015267175572, "prompt_id": "50c4d271f8ba07ce", "reading_canonical_log_prob": -385.02764892578125, "reading_distances": [13, 16, 12, 13], "reading_log_probs": [-387.04302978515625, -390.40484619140625, -390.2977600097656, -386.9766845703125], "reading_logprob_std": 1.9298700668056081, "tokenization_path_logprob_std": 187.22806791462648, "trajectory_balance_rmse": 5372.605988613084}1073{"canonical_length": 27, "canonical_nll": 2.0359827677408853, "generation_rouge_l_f1": 0.2641509433962264, "generation_token_f1": 0.3773584905660378, "prompt_id": "71a578aa99112e47", "reading_canonical_log_prob": -54.971534729003906, "reading_distances": [10, 13, 11, 16], "reading_log_probs": [-55.637908935546875, -59.21019744873047, -58.055267333984375, -59.245418548583984], "reading_logprob_std": 1.6923986051732947, "tokenization_path_logprob_std": 58.53710788235749, "trajectory_balance_rmse": 676.8606012121944}1074{"canonical_length": 260, "canonical_nll": 1.089328355055589, "generation_rouge_l_f1": 0.22519083969465647, "generation_token_f1": 0.4732824427480916, "prompt_id": "c22b91c04ec53dee", "reading_canonical_log_prob": -283.2253723144531, "reading_distances": [38, 39, 38, 43], "reading_log_probs": [-281.03131103515625, -277.1014404296875, -280.88140869140625, -278.283447265625], "reading_logprob_std": 1.946188856327989, "tokenization_path_logprob_std": 175.53803920282056, "trajectory_balance_rmse": 5610.996607022098}1075{"canonical_length": 42, "canonical_nll": 1.8268623352050781, "generation_rouge_l_f1": 0.1896551724137931, "generation_token_f1": 0.24137931034482757, "prompt_id": "ce56aa7776fbfa4a", "reading_canonical_log_prob": -76.72821807861328, "reading_distances": [18, 19, 16, 16], "reading_log_probs": [-84.53604888916016, -79.50222778320312, -85.42313385009766, -76.31365203857422], "reading_logprob_std": 4.300589954005594, "tokenization_path_logprob_std": 36.48518928962086, "trajectory_balance_rmse": 913.4330294597593}1076{"canonical_length": 92, "canonical_nll": 1.2375999118970789, "generation_rouge_l_f1": 0.3492063492063492, "generation_token_f1": 0.36507936507936506, "prompt_id": "61cafca95080c624", "reading_canonical_log_prob": -113.85919189453125, "reading_distances": [28, 27, 23, 27], "reading_log_probs": [-140.20712280273438, -112.12403106689453, -109.78811645507812, -109.14283752441406], "reading_logprob_std": 14.982562657372178, "tokenization_path_logprob_std": 36.676335215646745, "trajectory_balance_rmse": 963.3130210351528}1077{"canonical_length": 9, "canonical_nll": 1.9871732923719618, "generation_rouge_l_f1": 0.6250000000000001, "generation_token_f1": 0.6250000000000001, "prompt_id": "33cbcf32b9a54c98", "reading_canonical_log_prob": -17.884559631347656, "reading_distances": [24, 22, 17, 21], "reading_log_probs": [-22.426525115966797, -17.967784881591797, -18.871700286865234, -17.564512252807617], "reading_logprob_std": 2.214436451149541, "tokenization_path_logprob_std": 27.072437029135745, "trajectory_balance_rmse": 170.08601012350962}1078{"canonical_length": 13, "canonical_nll": 0.5550167010380671, "generation_rouge_l_f1": 0.787878787878788, "generation_token_f1": 0.787878787878788, "prompt_id": "5e0db48c41c73148", "reading_canonical_log_prob": -7.215217113494873, "reading_distances": [39, 37, 34, 28], "reading_log_probs": [-7.487349987030029, -6.641359329223633, -8.035120010375977, -9.525455474853516], "reading_logprob_std": 1.2128240247905988, "tokenization_path_logprob_std": 33.808226372272564, "trajectory_balance_rmse": 247.45765392378672}1079{"canonical_length": 45, "canonical_nll": 1.4527511596679688, "generation_rouge_l_f1": 0.2247191011235955, "generation_token_f1": 0.31460674157303364, "prompt_id": "c70b50c5f56eb753", "reading_canonical_log_prob": -65.3738021850586, "reading_distances": [26, 33, 32, 25], "reading_log_probs": [-78.54763793945312, -103.07257843017578, -107.08184814453125, -90.2860107421875], "reading_logprob_std": 12.958697095917762, "tokenization_path_logprob_std": 64.98685376117874, "trajectory_balance_rmse": 794.7335065087146}1080{"canonical_length": 28, "canonical_nll": 1.0001674379621233, "generation_rouge_l_f1": 0.10810810810810811, "generation_token_f1": 0.10810810810810811, "prompt_id": "a944937cb2c4cc95", "reading_canonical_log_prob": -28.004688262939453, "reading_distances": [16, 13, 12, 14], "reading_log_probs": [-35.650760650634766, -27.305509567260742, -29.277021408081055, -31.31901741027832], "reading_logprob_std": 3.573007823650896, "tokenization_path_logprob_std": 71.0851062604962, "trajectory_balance_rmse": 611.6458515162212}1081{"canonical_length": 158, "canonical_nll": 1.30721833434286, "generation_rouge_l_f1": 0.2167487684729064, "generation_token_f1": 0.37438423645320196, "prompt_id": "b725cba739a4a837", "reading_canonical_log_prob": -206.54049682617188, "reading_distances": [35, 37, 30, 31], "reading_log_probs": [-202.40869140625, -205.04556274414062, -200.35488891601562, -203.54937744140625], "reading_logprob_std": 1.9773321275797782, "tokenization_path_logprob_std": 83.27675775921541, "trajectory_balance_rmse": 2999.186078874606}1082{"canonical_length": 501, "canonical_nll": 0.6830049008428456, "generation_rouge_l_f1": 0.21515151515151512, "generation_token_f1": 0.4090909090909091, "prompt_id": "c3741a95aac9caea", "reading_canonical_log_prob": -342.1854553222656, "reading_distances": [33, 35, 25, 30], "reading_log_probs": [-354.6100769042969, -354.6698913574219, -354.107666015625, -361.3112487792969], "reading_logprob_std": 3.433620894202807, "tokenization_path_logprob_std": 144.9569845770976, "trajectory_balance_rmse": 5156.221213965406}1083{"canonical_length": 14, "canonical_nll": 0.628596578325544, "generation_rouge_l_f1": 0.4651162790697674, "generation_token_f1": 0.4651162790697674, "prompt_id": "77d6fb3ebd2588d3", "reading_canonical_log_prob": -8.800352096557617, "reading_distances": [23, 18, 19, 26], "reading_log_probs": [-12.410881042480469, -12.803668975830078, -12.013792037963867, -12.607954025268555], "reading_logprob_std": 0.3373966736451933, "tokenization_path_logprob_std": 43.057134371879975, "trajectory_balance_rmse": 211.27139301128935}1084{"canonical_length": 242, "canonical_nll": 1.41348039611312, "generation_rouge_l_f1": 0.1936842105263158, "generation_token_f1": 0.4042105263157894, "prompt_id": "8d10c80bdd9627ab", "reading_canonical_log_prob": -342.062255859375, "reading_distances": [16, 20, 18, 18], "reading_log_probs": [-347.6668701171875, -348.120361328125, -346.68829345703125, -346.54974365234375], "reading_logprob_std": 0.7609255557066036, "tokenization_path_logprob_std": 341.8783770377282, "trajectory_balance_rmse": 5308.780470238561}1085{"canonical_length": 233, "canonical_nll": 1.3515797889283798, "generation_rouge_l_f1": 0.21473684210526314, "generation_token_f1": 0.47578947368421054, "prompt_id": "a741547771fa6f03", "reading_canonical_log_prob": -314.9180908203125, "reading_distances": [22, 28, 26, 20], "reading_log_probs": [-315.2017822265625, -315.9852600097656, -314.298828125, -313.593994140625], "reading_logprob_std": 1.0437569153457515, "tokenization_path_logprob_std": 135.71625572370203, "trajectory_balance_rmse": 4903.182094957655}1086{"canonical_length": 177, "canonical_nll": 1.0412483861890889, "generation_rouge_l_f1": 0.2529274004683841, "generation_token_f1": 0.4355971896955504, "prompt_id": "51a6d546f1967143", "reading_canonical_log_prob": -184.30096435546875, "reading_distances": [27, 33, 28, 28], "reading_log_probs": [-181.79087829589844, -184.2711181640625, -183.82037353515625, -183.67626953125], "reading_logprob_std": 1.0955562674021528, "tokenization_path_logprob_std": 220.47103505485794, "trajectory_balance_rmse": 3039.605144513512}1087{"canonical_length": 163, "canonical_nll": 1.2253228872100268, "generation_rouge_l_f1": 0.27433628318584075, "generation_token_f1": 0.5, "prompt_id": "715a0325621f71c4", "reading_canonical_log_prob": -199.72763061523438, "reading_distances": [46, 42, 32, 36], "reading_log_probs": [-192.2281036376953, -190.43280029296875, -189.763427734375, -190.51510620117188], "reading_logprob_std": 1.050870909757697, "tokenization_path_logprob_std": 116.9160270517153, "trajectory_balance_rmse": 2687.8006317425593}1088{"canonical_length": 264, "canonical_nll": 1.3142970692027698, "generation_rouge_l_f1": 0.23828125, "generation_token_f1": 0.44921875, "prompt_id": "0857b1f32ec14a13", "reading_canonical_log_prob": -346.97442626953125, "reading_distances": [17, 23, 21, 17], "reading_log_probs": [-356.3417663574219, -361.4893493652344, -354.5234375, -355.726806640625], "reading_logprob_std": 3.0735569807242613, "tokenization_path_logprob_std": 131.85018245523824, "trajectory_balance_rmse": 4519.900484437096}1089{"canonical_length": 2, "canonical_nll": 5.505666732788086, "generation_rouge_l_f1": 0.01785714285714286, "generation_token_f1": 0.01785714285714286, "prompt_id": "7c6d4e136a035811", "reading_canonical_log_prob": -11.011333465576172, "reading_distances": [65, 71, 83, 67], "reading_log_probs": [-12.485105514526367, -12.077385902404785, -10.538238525390625, -11.349777221679688], "reading_logprob_std": 0.8564695355568618, "tokenization_path_logprob_std": 0.0, "trajectory_balance_rmse": 10.891670227050781}1090{"canonical_length": 13, "canonical_nll": 0.9798189309927133, "generation_rouge_l_f1": 0.5599999999999999, "generation_token_f1": 0.7199999999999999, "prompt_id": "d3b27b0b1b980505", "reading_canonical_log_prob": -12.737646102905273, "reading_distances": [31, 27, 34, 37], "reading_log_probs": [-13.24582290649414, -13.234413146972656, -13.382360458374023, -13.284899711608887], "reading_logprob_std": 0.06722820570226033, "tokenization_path_logprob_std": 23.79733848693559, "trajectory_balance_rmse": 242.0164682162875}1091{"canonical_length": 131, "canonical_nll": 0.47141935625149095, "generation_rouge_l_f1": 0.459214501510574, "generation_token_f1": 0.5438066465256798, "prompt_id": "37ec341c729bc100", "reading_canonical_log_prob": -61.75593566894531, "reading_distances": [9, 8, 14, 7], "reading_log_probs": [-53.9974365234375, -56.35185623168945, -55.86846923828125, -54.9313850402832], "reading_logprob_std": 1.0426665973953528, "tokenization_path_logprob_std": 96.5966515649517, "trajectory_balance_rmse": 1222.7594368693658}1092{"canonical_length": 12, "canonical_nll": 3.73262087504069, "generation_rouge_l_f1": 0.23076923076923078, "generation_token_f1": 0.38461538461538464, "prompt_id": "b0f40cc1f3136930", "reading_canonical_log_prob": -44.79145050048828, "reading_distances": [10, 12, 9, 13], "reading_log_probs": [-63.84538269042969, -50.933265686035156, -45.28761672973633, -58.184940338134766], "reading_logprob_std": 8.134059791334785, "tokenization_path_logprob_std": 52.0630271735715, "trajectory_balance_rmse": 298.1475519629203}1093{"canonical_length": 264, "canonical_nll": 1.2554679639411694, "generation_rouge_l_f1": 0.29629629629629634, "generation_token_f1": 0.5614035087719299, "prompt_id": "dc37089a4982f612", "reading_canonical_log_prob": -331.44354248046875, "reading_distances": [16, 16, 14, 14], "reading_log_probs": [-355.3581848144531, -354.35577392578125, -335.81005859375, -334.024658203125], "reading_logprob_std": 11.542453892284406, "tokenization_path_logprob_std": 239.07799119228622, "trajectory_balance_rmse": 5477.509014471944}1094{"canonical_length": 3, "canonical_nll": 0.8658781051635742, "generation_rouge_l_f1": 0.05882352941176471, "generation_token_f1": 0.05882352941176471, "prompt_id": "9ba6dc4bdecfc908", "reading_canonical_log_prob": -2.5976343154907227, "reading_distances": [24, 25, 31, 15], "reading_log_probs": [-2.2138845920562744, -2.5517778396606445, -1.956922173500061, -2.232792854309082], "reading_logprob_std": 0.24363024762287977, "tokenization_path_logprob_std": 13.53078043460846, "trajectory_balance_rmse": 14.59539341221242}1095{"canonical_length": 261, "canonical_nll": 0.9591437445746528, "generation_rouge_l_f1": 0.6344294003868473, "generation_token_f1": 0.793036750483559, "prompt_id": "5cf08716fd03e235", "reading_canonical_log_prob": -250.33651733398438, "reading_distances": [30, 30, 30, 31], "reading_log_probs": [-269.9743347167969, -270.131591796875, -265.5967102050781, -271.3695983886719], "reading_logprob_std": 2.5258519391503773, "tokenization_path_logprob_std": 120.11992486877074, "trajectory_balance_rmse": 4265.2643112659225}1096{"canonical_length": 102, "canonical_nll": 0.9521930170994178, "generation_rouge_l_f1": 0.2142857142857143, "generation_token_f1": 0.34615384615384615, "prompt_id": "8cfc1971edf74b1e", "reading_canonical_log_prob": -97.12368774414062, "reading_distances": [38, 31, 32, 39], "reading_log_probs": [-96.15725708007812, -95.50528717041016, -95.34181213378906, -94.076904296875], "reading_logprob_std": 0.8700994427370189, "tokenization_path_logprob_std": 119.18882489648969, "trajectory_balance_rmse": 1879.1354502084305}1097{"canonical_length": 42, "canonical_nll": 2.7036485217866444, "generation_rouge_l_f1": 0.07843137254901962, "generation_token_f1": 0.1176470588235294, "prompt_id": "73cb687de2d9ad9d", "reading_canonical_log_prob": -113.55323791503906, "reading_distances": [24, 27, 23, 23], "reading_log_probs": [-114.4521713256836, -115.80624389648438, -117.13917541503906, -112.0700912475586], "reading_logprob_std": 2.1633108566123216, "tokenization_path_logprob_std": 58.58203351855345, "trajectory_balance_rmse": 793.445661492345}1098{"canonical_length": 50, "canonical_nll": 1.3209599304199218, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "5b7c88eabb38a156", "reading_canonical_log_prob": -66.0479965209961, "reading_distances": [29, 26, 30, 26], "reading_log_probs": [-58.88529968261719, -60.92375946044922, -59.38827133178711, -58.55931091308594], "reading_logprob_std": 1.0468239077375925, "tokenization_path_logprob_std": 89.75669697099656, "trajectory_balance_rmse": 987.8994442156314}1099{"canonical_length": 297, "canonical_nll": 1.8541937934027777, "generation_rouge_l_f1": 0.22263450834879409, "generation_token_f1": 0.43042671614100186, "prompt_id": "7564700da05daae9", "reading_canonical_log_prob": -550.695556640625, "reading_distances": [25, 22, 21, 30], "reading_log_probs": [-568.5263671875, -567.09130859375, -569.1441650390625, -572.203857421875], "reading_logprob_std": 2.1540484047346498, "tokenization_path_logprob_std": 326.2365351977119, "trajectory_balance_rmse": 6065.322303294232}1100{"canonical_length": 240, "canonical_nll": 1.5221951802571614, "generation_rouge_l_f1": 0.2300194931773879, "generation_token_f1": 0.40545808966861596, "prompt_id": "ca2d7d9e3b473a6f", "reading_canonical_log_prob": -365.32684326171875, "reading_distances": [17, 14, 21, 13], "reading_log_probs": [-374.93780517578125, -374.54730224609375, -370.80731201171875, -377.3034973144531], "reading_logprob_std": 2.686300543296277, "tokenization_path_logprob_std": 182.193239666402, "trajectory_balance_rmse": 5504.062880200658}1101{"canonical_length": 282, "canonical_nll": 1.237128399788065, "generation_rouge_l_f1": 0.26953125, "generation_token_f1": 0.39843749999999994, "prompt_id": "9e92f8d560ad3563", "reading_canonical_log_prob": -348.8702087402344, "reading_distances": [22, 25, 24, 15], "reading_log_probs": [-366.949462890625, -357.28179931640625, -366.15966796875, -362.6773681640625], "reading_logprob_std": 4.400695165571493, "tokenization_path_logprob_std": 192.889729691358, "trajectory_balance_rmse": 4314.8725583427595}1102{"canonical_length": 16, "canonical_nll": 2.014431953430176, "generation_rouge_l_f1": 0.45161290322580644, "generation_token_f1": 0.5161290322580646, "prompt_id": "8efbf41d41230763", "reading_canonical_log_prob": -32.23091125488281, "reading_distances": [37, 36, 42, 28], "reading_log_probs": [-35.93354797363281, -36.14146041870117, -33.13258361816406, -34.851776123046875], "reading_logprob_std": 1.3763207987921362, "tokenization_path_logprob_std": 36.4084196251416, "trajectory_balance_rmse": 363.68542458246304}1103{"canonical_length": 39, "canonical_nll": 1.9306034185947516, "generation_rouge_l_f1": 0.33962264150943394, "generation_token_f1": 0.3773584905660377, "prompt_id": "3f9b43f9811260d6", "reading_canonical_log_prob": -75.29353332519531, "reading_distances": [32, 31, 28, 35], "reading_log_probs": [-78.29763793945312, -81.93568420410156, -76.91874694824219, -78.64204406738281], "reading_logprob_std": 2.126084317679521, "tokenization_path_logprob_std": 64.41526139796262, "trajectory_balance_rmse": 676.4932832595478}1104{"canonical_length": 281, "canonical_nll": 1.0085476369620219, "generation_rouge_l_f1": 0.24518388791593695, "generation_token_f1": 0.4238178633975481, "prompt_id": "9e7206bd2d909c4b", "reading_canonical_log_prob": -283.4018859863281, "reading_distances": [28, 37, 33, 32], "reading_log_probs": [-291.921875, -299.17510986328125, -297.5544738769531, -289.0128479003906], "reading_logprob_std": 4.757935724335682, "tokenization_path_logprob_std": 63.42907552356942, "trajectory_balance_rmse": 4251.525783259128}1105{"canonical_length": 193, "canonical_nll": 1.3441003987208548, "generation_rouge_l_f1": 0.3830845771144279, "generation_token_f1": 0.5373134328358209, "prompt_id": "59cc1c6a894af7a8", "reading_canonical_log_prob": -259.411376953125, "reading_distances": [20, 23, 27, 23], "reading_log_probs": [-261.229736328125, -259.0027770996094, -261.2709655761719, -260.1835021972656], "reading_logprob_std": 1.0714873010154689, "tokenization_path_logprob_std": 190.0612040791962, "trajectory_balance_rmse": 3655.102388963147}1106{"canonical_length": 85, "canonical_nll": 1.1209692562327664, "generation_rouge_l_f1": 0.23906705539358597, "generation_token_f1": 0.36734693877551017, "prompt_id": "bb9e06240071dbbd", "reading_canonical_log_prob": -95.28238677978516, "reading_distances": [25, 25, 20, 24], "reading_log_probs": [-97.14262390136719, -97.91383361816406, -98.91827392578125, -99.14409637451172], "reading_logprob_std": 0.9276777483233662, "tokenization_path_logprob_std": 85.2374377918263, "trajectory_balance_rmse": 1636.4367926702641}1107{"canonical_length": 59, "canonical_nll": 0.9168294486352953, "generation_rouge_l_f1": 0.20689655172413793, "generation_token_f1": 0.3054187192118227, "prompt_id": "fbd2f42b1f0643a2", "reading_canonical_log_prob": -54.09293746948242, "reading_distances": [17, 17, 13, 17], "reading_log_probs": [-48.96332550048828, -48.72308349609375, -51.11525344848633, -50.02308654785156], "reading_logprob_std": 1.0960883915888806, "tokenization_path_logprob_std": 65.30650061908857, "trajectory_balance_rmse": 899.0855662225521}1108{"canonical_length": 73, "canonical_nll": 0.8639211262742134, "generation_rouge_l_f1": 0.2572347266881029, "generation_token_f1": 0.34083601286173637, "prompt_id": "1a0ff319dcddb777", "reading_canonical_log_prob": -63.06624221801758, "reading_distances": [48, 48, 55, 41], "reading_log_probs": [-64.57321166992188, -65.57461547851562, -66.39527130126953, -64.12638854980469], "reading_logprob_std": 1.0182103558483617, "tokenization_path_logprob_std": 104.48747894773147, "trajectory_balance_rmse": 643.2517247263177}1109{"canonical_length": 15, "canonical_nll": 0.8491415659586589, "generation_rouge_l_f1": 0.46153846153846156, "generation_token_f1": 0.5384615384615384, "prompt_id": "9f8d22c249bf7888", "reading_canonical_log_prob": -12.737123489379883, "reading_distances": [20, 21, 24, 21], "reading_log_probs": [-13.555301666259766, -16.250402450561523, -13.183633804321289, -14.228435516357422], "reading_logprob_std": 1.3674747178448947, "tokenization_path_logprob_std": 35.49901534957844, "trajectory_balance_rmse": 337.56010657012473}1110{"canonical_length": 168, "canonical_nll": 1.1611677805582683, "generation_rouge_l_f1": 0.17955112219451372, "generation_token_f1": 0.3441396508728179, "prompt_id": "9edad123b1e31f1d", "reading_canonical_log_prob": -195.07618713378906, "reading_distances": [28, 29, 28, 31], "reading_log_probs": [-218.79470825195312, -223.6187744140625, -221.59625244140625, -215.17620849609375], "reading_logprob_std": 3.660581503598963, "tokenization_path_logprob_std": 284.5733519455572, "trajectory_balance_rmse": 3323.030092123595}1111{"canonical_length": 82, "canonical_nll": 1.337618292831793, "generation_rouge_l_f1": 0.17791411042944782, "generation_token_f1": 0.26993865030674846, "prompt_id": "e027ef71856c6f38", "reading_canonical_log_prob": -109.68470001220703, "reading_distances": [29, 27, 25, 26], "reading_log_probs": [-119.43263244628906, -118.58676147460938, -122.30903625488281, -111.6940689086914], "reading_logprob_std": 4.499228029202058, "tokenization_path_logprob_std": 109.56155984742738, "trajectory_balance_rmse": 1954.2659163537635}1112{"canonical_length": 27, "canonical_nll": 1.9492051866319444, "generation_rouge_l_f1": 0.1891891891891892, "generation_token_f1": 0.1891891891891892, "prompt_id": "59a16e60dba35dc9", "reading_canonical_log_prob": -52.6285400390625, "reading_distances": [15, 14, 18, 15], "reading_log_probs": [-73.9598159790039, -65.74514770507812, -66.27659606933594, -63.47101593017578], "reading_logprob_std": 4.563012182265138, "tokenization_path_logprob_std": 64.82049597665394, "trajectory_balance_rmse": 634.9847296936279}1113{"canonical_length": 16, "canonical_nll": 0.5487178564071655, "generation_rouge_l_f1": 0.5, "generation_token_f1": 0.75, "prompt_id": "ca87800b07a107f6", "reading_canonical_log_prob": -8.779485702514648, "reading_distances": [27, 32, 30, 26], "reading_log_probs": [-9.109512329101562, -9.240385055541992, -8.984173774719238, -9.206440925598145], "reading_logprob_std": 0.11490274412798959, "tokenization_path_logprob_std": 14.07736083664683, "trajectory_balance_rmse": 235.5430348084208}1114{"canonical_length": 100, "canonical_nll": 1.5049606323242188, "generation_rouge_l_f1": 0.19662921348314608, "generation_token_f1": 0.34269662921348315, "prompt_id": "911e3e27940b6a5a", "reading_canonical_log_prob": -150.49606323242188, "reading_distances": [29, 27, 23, 30], "reading_log_probs": [-152.2185516357422, -149.85939025878906, -151.69180297851562, -151.75868225097656], "reading_logprob_std": 1.0417975525316099, "tokenization_path_logprob_std": 289.3734969377017, "trajectory_balance_rmse": 2061.1087840471214}1115{"canonical_length": 190, "canonical_nll": 2.196766421669408, "generation_rouge_l_f1": 0.2091503267973856, "generation_token_f1": 0.4139433551198257, "prompt_id": "1e23f4d5f326fadb", "reading_canonical_log_prob": -417.3856201171875, "reading_distances": [23, 25, 31, 29], "reading_log_probs": [-417.8870849609375, -428.8445739746094, -421.91424560546875, -430.2857666015625], "reading_logprob_std": 5.8466567800703455, "tokenization_path_logprob_std": 191.95310855891802, "trajectory_balance_rmse": 4235.409798633718}1116{"canonical_length": 151, "canonical_nll": 1.514494839093543, "generation_rouge_l_f1": 0.26344086021505375, "generation_token_f1": 0.521505376344086, "prompt_id": "75c88d19daf68f00", "reading_canonical_log_prob": -228.688720703125, "reading_distances": [18, 19, 24, 20], "reading_log_probs": [-235.15806579589844, -236.005126953125, -229.10397338867188, -234.42123413085938], "reading_logprob_std": 3.1134167615729607, "tokenization_path_logprob_std": 216.14201087366484, "trajectory_balance_rmse": 3761.822852757185}1117{"canonical_length": 165, "canonical_nll": 1.3388463800603694, "generation_rouge_l_f1": 0.25, "generation_token_f1": 0.48125, "prompt_id": "0efac862acbbc3cf", "reading_canonical_log_prob": -220.90965270996094, "reading_distances": [28, 33, 24, 31], "reading_log_probs": [-220.49880981445312, -223.6935577392578, -222.09115600585938, -220.0110626220703], "reading_logprob_std": 1.6692095961012579, "tokenization_path_logprob_std": 215.35484661179348, "trajectory_balance_rmse": 4066.0557506693103}1118{"canonical_length": 225, "canonical_nll": 0.9715812174479167, "generation_rouge_l_f1": 0.20833333333333334, "generation_token_f1": 0.425, "prompt_id": "867d14d7a7b70c9c", "reading_canonical_log_prob": -218.60577392578125, "reading_distances": [14, 16, 20, 20], "reading_log_probs": [-218.48703002929688, -214.6271209716797, -198.77996826171875, -204.30393981933594], "reading_logprob_std": 9.095061889263478, "tokenization_path_logprob_std": 62.74366136071494, "trajectory_balance_rmse": 3605.92655033307}1119{"canonical_length": 263, "canonical_nll": 1.5089322514407082, "generation_rouge_l_f1": 0.18687872763419483, "generation_token_f1": 0.34194831013916505, "prompt_id": "67b31fbc0903ce67", "reading_canonical_log_prob": -396.84918212890625, "reading_distances": [27, 20, 20, 14], "reading_log_probs": [-407.583251953125, -404.28472900390625, -404.48388671875, -405.86993408203125], "reading_logprob_std": 1.5246705343105647, "tokenization_path_logprob_std": 287.26711050518685, "trajectory_balance_rmse": 5148.953720465755}1120{"canonical_length": 38, "canonical_nll": 0.7189707003141704, "generation_rouge_l_f1": 0.36000000000000004, "generation_token_f1": 0.38, "prompt_id": "c50601d0ada030f4", "reading_canonical_log_prob": -27.320886611938477, "reading_distances": [28, 26, 24, 26], "reading_log_probs": [-28.655040740966797, -30.466136932373047, -28.91344451904297, -30.490493774414062], "reading_logprob_std": 0.9837962655395402, "tokenization_path_logprob_std": 32.69333495776394, "trajectory_balance_rmse": 229.7848585210045}1121{"canonical_length": 35, "canonical_nll": 0.5135156358991351, "generation_rouge_l_f1": 0.5542168674698795, "generation_token_f1": 0.5542168674698795, "prompt_id": "d0d4fbc6a94b1aa0", "reading_canonical_log_prob": -17.973047256469727, "reading_distances": [21, 19, 20, 21], "reading_log_probs": [-16.97222328186035, -16.88335418701172, -18.557748794555664, -15.858419418334961], "reading_logprob_std": 1.1144062536082389, "tokenization_path_logprob_std": 83.96355091485988, "trajectory_balance_rmse": 579.7565471717012}1122{"canonical_length": 74, "canonical_nll": 0.5694807928961676, "generation_rouge_l_f1": 0.425531914893617, "generation_token_f1": 0.5744680851063829, "prompt_id": "a83eef42db047836", "reading_canonical_log_prob": -42.141578674316406, "reading_distances": [20, 25, 26, 29], "reading_log_probs": [-41.206302642822266, -41.67047119140625, -40.727298736572266, -38.78705596923828], "reading_logprob_std": 1.2670783740073017, "tokenization_path_logprob_std": 89.9827206909935, "trajectory_balance_rmse": 1187.4937717912123}1123{"canonical_length": 107, "canonical_nll": 1.1646197310117918, "generation_rouge_l_f1": 0.26523297491039427, "generation_token_f1": 0.47311827956989244, "prompt_id": "8e875ab98664b402", "reading_canonical_log_prob": -124.61431121826172, "reading_distances": [18, 17, 18, 18], "reading_log_probs": [-135.71511840820312, -134.9852294921875, -134.4835205078125, -133.58274841308594], "reading_logprob_std": 0.8956668122560835, "tokenization_path_logprob_std": 73.29327450128892, "trajectory_balance_rmse": 2001.024640606028}1124{"canonical_length": 31, "canonical_nll": 0.7431251771988407, "generation_rouge_l_f1": 0.6222222222222222, "generation_token_f1": 0.6222222222222222, "prompt_id": "54f5b26b0d406b0a", "reading_canonical_log_prob": -23.036880493164062, "reading_distances": [29, 31, 27, 34], "reading_log_probs": [-27.33916664123535, -27.016515731811523, -29.405506134033203, -31.899791717529297], "reading_logprob_std": 2.2536689283255593, "tokenization_path_logprob_std": 70.73753278440807, "trajectory_balance_rmse": 612.513569677371}1125{"canonical_length": 262, "canonical_nll": 1.3443203991606036, "generation_rouge_l_f1": 0.22846441947565543, "generation_token_f1": 0.49063670411985016, "prompt_id": "652a0c6a273ec7b9", "reading_canonical_log_prob": -352.2119445800781, "reading_distances": [16, 17, 13, 13], "reading_log_probs": [-356.75, -352.06219482421875, -355.1026916503906, -353.09552001953125], "reading_logprob_std": 2.0893676467091, "tokenization_path_logprob_std": 238.73048494614562, "trajectory_balance_rmse": 4305.190520898793}1126{"canonical_length": 304, "canonical_nll": 1.6778772253739207, "generation_rouge_l_f1": 0.12457912457912458, "generation_token_f1": 0.2525252525252525, "prompt_id": "13a435b70dd47ea2", "reading_canonical_log_prob": -510.0746765136719, "reading_distances": [23, 14, 17, 11], "reading_log_probs": [-517.7884521484375, -521.5836181640625, -520.4683227539062, -515.899658203125], "reading_logprob_std": 2.5751458629088857, "tokenization_path_logprob_std": 370.55049314789403, "trajectory_balance_rmse": 6760.89394932125}1127{"canonical_length": 15, "canonical_nll": 0.793018086751302, "generation_rouge_l_f1": 0.3529411764705882, "generation_token_f1": 0.4705882352941177, "prompt_id": "1f8b04daa461e185", "reading_canonical_log_prob": -11.895271301269531, "reading_distances": [24, 18, 25, 23], "reading_log_probs": [-12.424539566040039, -10.279709815979004, -11.065324783325195, -12.421923637390137], "reading_logprob_std": 1.0604400701985985, "tokenization_path_logprob_std": 19.81389266669746, "trajectory_balance_rmse": 183.681845033232}1128{"canonical_length": 18, "canonical_nll": 0.3054545720418294, "generation_rouge_l_f1": 0.47619047619047616, "generation_token_f1": 0.47619047619047616, "prompt_id": "713c6185b310a4ca", "reading_canonical_log_prob": -5.49818229675293, "reading_distances": [14, 27, 24, 30], "reading_log_probs": [-5.4050750732421875, -10.939093589782715, -8.123337745666504, -6.678660869598389], "reading_logprob_std": 2.377028833290396, "tokenization_path_logprob_std": 58.779798715763036, "trajectory_balance_rmse": 391.2406802366449}1129{"canonical_length": 227, "canonical_nll": 0.8063396840368598, "generation_rouge_l_f1": 0.3123732251521298, "generation_token_f1": 0.4949290060851927, "prompt_id": "a7de6528410fa4b7", "reading_canonical_log_prob": -183.0391082763672, "reading_distances": [21, 28, 25, 25], "reading_log_probs": [-192.43504333496094, -198.64566040039062, -190.81935119628906, -189.8446807861328], "reading_logprob_std": 3.953379818711711, "tokenization_path_logprob_std": 86.73878609704282, "trajectory_balance_rmse": 3343.74942262196}1130{"canonical_length": 154, "canonical_nll": 1.729839176326603, "generation_rouge_l_f1": 0.2707838479809976, "generation_token_f1": 0.4370546318289786, "prompt_id": "b7245c913ba925d6", "reading_canonical_log_prob": -266.3952331542969, "reading_distances": [22, 28, 21, 28], "reading_log_probs": [-267.54693603515625, -268.79815673828125, -268.0393371582031, -266.97271728515625], "reading_logprob_std": 0.773706102280434, "tokenization_path_logprob_std": 123.61555373757567, "trajectory_balance_rmse": 3166.4388561824094}1131{"canonical_length": 9, "canonical_nll": 0.7106206681993272, "generation_rouge_l_f1": 0.8750000000000001, "generation_token_f1": 0.8750000000000001, "prompt_id": "402312ec99016f32", "reading_canonical_log_prob": -6.395586013793945, "reading_distances": [23, 21, 22, 20], "reading_log_probs": [-7.132452487945557, -7.360826015472412, -7.475787162780762, -6.691463947296143], "reading_logprob_std": 0.3465218379428095, "tokenization_path_logprob_std": 40.78940981776797, "trajectory_balance_rmse": 170.04873396746723}1132{"canonical_length": 290, "canonical_nll": 1.4355969659213361, "generation_rouge_l_f1": 0.2380106571936057, "generation_token_f1": 0.49378330373001783, "prompt_id": "083ca3a9f1344c38", "reading_canonical_log_prob": -416.3231201171875, "reading_distances": [30, 35, 32, 31], "reading_log_probs": [-424.122314453125, -425.85205078125, -426.0133361816406, -428.669921875], "reading_logprob_std": 1.8768888759855333, "tokenization_path_logprob_std": 223.76793754054282, "trajectory_balance_rmse": 6043.879281819117}1133{"canonical_length": 58, "canonical_nll": 1.6705818176269531, "generation_rouge_l_f1": 0.19591836734693877, "generation_token_f1": 0.2775510204081633, "prompt_id": "c06dd3bfbe8f1823", "reading_canonical_log_prob": -96.89374542236328, "reading_distances": [35, 29, 32, 34], "reading_log_probs": [-101.22756958007812, -101.20771026611328, -103.38756561279297, -101.84834289550781], "reading_logprob_std": 1.023992302808972, "tokenization_path_logprob_std": 121.88293775128167, "trajectory_balance_rmse": 1159.945515222223}1134{"canonical_length": 117, "canonical_nll": 0.9430011194995326, "generation_rouge_l_f1": 0.2011173184357542, "generation_token_f1": 0.32402234636871513, "prompt_id": "22135eb61f4e8a41", "reading_canonical_log_prob": -110.33113098144531, "reading_distances": [23, 16, 22, 21], "reading_log_probs": [-123.54287719726562, -125.29365539550781, -119.62701416015625, -117.16004943847656], "reading_logprob_std": 3.6911152505674676, "tokenization_path_logprob_std": 122.95212818946268, "trajectory_balance_rmse": 2225.321180638249}1135{"canonical_length": 21, "canonical_nll": 2.183457510811942, "generation_rouge_l_f1": 0.3529411764705882, "generation_token_f1": 0.4117647058823529, "prompt_id": "b36f1ed6b55c0058", "reading_canonical_log_prob": -45.85260772705078, "reading_distances": [27, 26, 24, 28], "reading_log_probs": [-59.48552322387695, -51.255043029785156, -56.30403518676758, -49.405670166015625], "reading_logprob_std": 4.618495875045387, "tokenization_path_logprob_std": 87.5854508063459, "trajectory_balance_rmse": 435.8118251054314}1136{"canonical_length": 25, "canonical_nll": 1.7651446533203126, "generation_rouge_l_f1": 0.2105263157894737, "generation_token_f1": 0.2631578947368421, "prompt_id": "017529d62a606652", "reading_canonical_log_prob": -44.12861633300781, "reading_distances": [25, 19, 26, 27], "reading_log_probs": [-43.71147918701172, -42.53459548950195, -42.91149139404297, -41.66218185424805], "reading_logprob_std": 0.8509105742524149, "tokenization_path_logprob_std": 92.50105275424599, "trajectory_balance_rmse": 488.9851702159808}1137{"canonical_length": 231, "canonical_nll": 1.649572331151921, "generation_rouge_l_f1": 0.17751479289940827, "generation_token_f1": 0.37475345167652857, "prompt_id": "47cd1ce3b4c7133a", "reading_canonical_log_prob": -381.05120849609375, "reading_distances": [15, 12, 12, 9], "reading_log_probs": [-389.7129211425781, -385.671142578125, -385.835693359375, -386.2374267578125], "reading_logprob_std": 1.913919950447398, "tokenization_path_logprob_std": 158.21984493481514, "trajectory_balance_rmse": 5121.077724353289}1138{"canonical_length": 34, "canonical_nll": 0.6977959801169002, "generation_rouge_l_f1": 0.4571428571428571, "generation_token_f1": 0.4857142857142857, "prompt_id": "549cddbffb2c458b", "reading_canonical_log_prob": -23.72506332397461, "reading_distances": [11, 13, 17, 10], "reading_log_probs": [-29.168651580810547, -29.181406021118164, -30.692413330078125, -27.295413970947266], "reading_logprob_std": 1.3907655396177365, "tokenization_path_logprob_std": 100.96137296613398, "trajectory_balance_rmse": 444.5291828794505}1139{"canonical_length": 356, "canonical_nll": 1.4196941075700054, "generation_rouge_l_f1": 0.24957841483979762, "generation_token_f1": 0.47892074198988194, "prompt_id": "6b4ba85e93ef4f52", "reading_canonical_log_prob": -505.4111022949219, "reading_distances": [29, 24, 24, 28], "reading_log_probs": [-503.0926513671875, -496.92559814453125, -501.9715576171875, -499.1402893066406], "reading_logprob_std": 2.7882671843299605, "tokenization_path_logprob_std": 343.71519934960804, "trajectory_balance_rmse": 7921.432920885689}1140{"canonical_length": 420, "canonical_nll": 1.2938700358072917, "generation_rouge_l_f1": 0.23547880690737835, "generation_token_f1": 0.43014128728414447, "prompt_id": "71f47b6572a134fd", "reading_canonical_log_prob": -543.4254150390625, "reading_distances": [28, 37, 36, 36], "reading_log_probs": [-541.7032470703125, -542.8193359375, -540.565185546875, -539.6295166015625], "reading_logprob_std": 1.383618872206404, "tokenization_path_logprob_std": 297.03958027486357, "trajectory_balance_rmse": 7828.369661348962}1141{"canonical_length": 144, "canonical_nll": 1.3292505476209853, "generation_rouge_l_f1": 0.274932614555256, "generation_token_f1": 0.4582210242587601, "prompt_id": "c1009daeb59a17f6", "reading_canonical_log_prob": -191.41207885742188, "reading_distances": [18, 17, 20, 19], "reading_log_probs": [-195.83627319335938, -196.5247344970703, -200.96148681640625, -197.65213012695312], "reading_logprob_std": 2.272054206353305, "tokenization_path_logprob_std": 100.9035006929341, "trajectory_balance_rmse": 2677.094080917692}1142{"canonical_length": 19, "canonical_nll": 2.151770541542455, "generation_rouge_l_f1": 0.411764705882353, "generation_token_f1": 0.411764705882353, "prompt_id": "231e10798735d9e9", "reading_canonical_log_prob": -40.88364028930664, "reading_distances": [17, 20, 16, 21], "reading_log_probs": [-41.385643005371094, -42.02881622314453, -41.84560775756836, -44.6905403137207], "reading_logprob_std": 1.4933063791837131, "tokenization_path_logprob_std": 17.93105294129818, "trajectory_balance_rmse": 475.5733597041861}1143{"canonical_length": 133, "canonical_nll": 1.8580839687720276, "generation_rouge_l_f1": 0.14790996784565916, "generation_token_f1": 0.26366559485530544, "prompt_id": "948eaf298aee4503", "reading_canonical_log_prob": -247.1251678466797, "reading_distances": [24, 19, 19, 19], "reading_log_probs": [-229.3079376220703, -229.83560180664062, -231.535400390625, -231.25889587402344], "reading_logprob_std": 1.081580400536757, "tokenization_path_logprob_std": 210.27235177188712, "trajectory_balance_rmse": 2827.8509424403896}1144{"canonical_length": 43, "canonical_nll": 0.31613764651986054, "generation_rouge_l_f1": 0.3829787234042553, "generation_token_f1": 0.5957446808510638, "prompt_id": "cc672b4406958c5b", "reading_canonical_log_prob": -13.593918800354004, "reading_distances": [44, 54, 40, 46], "reading_log_probs": [-23.33747100830078, -17.400869369506836, -18.68670654296875, -20.35967254638672], "reading_logprob_std": 2.5649373211033692, "tokenization_path_logprob_std": 35.8848828623337, "trajectory_balance_rmse": 787.8818387217954}1145{"canonical_length": 189, "canonical_nll": 1.1571185399615576, "generation_rouge_l_f1": 0.18981481481481483, "generation_token_f1": 0.34259259259259256, "prompt_id": "f0c1c49e728adae3", "reading_canonical_log_prob": -218.69540405273438, "reading_distances": [17, 19, 17, 21], "reading_log_probs": [-214.26712036132812, -219.0343780517578, -218.64736938476562, -217.7457275390625], "reading_logprob_std": 2.1725060253347817, "tokenization_path_logprob_std": 84.05780690362613, "trajectory_balance_rmse": 3683.3206127591825}1146{"canonical_length": 202, "canonical_nll": 1.5799119401686261, "generation_rouge_l_f1": 0.24242424242424243, "generation_token_f1": 0.4761904761904762, "prompt_id": "91ba22c36f60a796", "reading_canonical_log_prob": -319.1422119140625, "reading_distances": [28, 31, 29, 23], "reading_log_probs": [-325.3409729003906, -325.93231201171875, -320.43646240234375, -319.4364929199219], "reading_logprob_std": 3.3249906839191516, "tokenization_path_logprob_std": 282.7266676695962, "trajectory_balance_rmse": 4722.331721033755}1147{"canonical_length": 91, "canonical_nll": 1.21202640743046, "generation_rouge_l_f1": 0.28455284552845533, "generation_token_f1": 0.42276422764227634, "prompt_id": "f1222deb29a5be25", "reading_canonical_log_prob": -110.29440307617188, "reading_distances": [13, 13, 9, 15], "reading_log_probs": [-116.31722259521484, -115.73531341552734, -115.8082504272461, -115.06210327148438], "reading_logprob_std": 0.5154498651457398, "tokenization_path_logprob_std": 136.42449866181423, "trajectory_balance_rmse": 2193.4586960915044}1148{"canonical_length": 9, "canonical_nll": 0.2488141589694553, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "6b02223ad9dd90ef", "reading_canonical_log_prob": -2.2393274307250977, "reading_distances": [25, 29, 27, 25], "reading_log_probs": [-2.6756515502929688, -1.9005117416381836, -1.842444658279419, -1.8165372610092163], "reading_logprob_std": 0.4127395771661793, "tokenization_path_logprob_std": 48.46614675180617, "trajectory_balance_rmse": 180.84951497506364}1149{"canonical_length": 170, "canonical_nll": 1.3309161017922795, "generation_rouge_l_f1": 0.2068965517241379, "generation_token_f1": 0.37356321839080464, "prompt_id": "a027618ad6b82f3e", "reading_canonical_log_prob": -226.2557373046875, "reading_distances": [23, 19, 18, 19], "reading_log_probs": [-237.26693725585938, -238.31997680664062, -236.73348999023438, -237.36827087402344], "reading_logprob_std": 0.660136233043112, "tokenization_path_logprob_std": 201.4095488525846, "trajectory_balance_rmse": 3769.332601290178}1150{"canonical_length": 38, "canonical_nll": 1.5119420101768093, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "252eb77faa4d024a", "reading_canonical_log_prob": -57.45379638671875, "reading_distances": [20, 24, 22, 13], "reading_log_probs": [-59.76047897338867, -57.557315826416016, -57.21916580200195, -57.888816833496094], "reading_logprob_std": 1.1360747669731384, "tokenization_path_logprob_std": 44.35787262735552, "trajectory_balance_rmse": 663.6701139169618}1151{"canonical_length": 273, "canonical_nll": 0.7597573528359661, "generation_rouge_l_f1": 0.28270042194092826, "generation_token_f1": 0.5274261603375529, "prompt_id": "75be9357ba7c860a", "reading_canonical_log_prob": -207.41375732421875, "reading_distances": [24, 18, 17, 16], "reading_log_probs": [-220.16049194335938, -221.4410400390625, -214.2175750732422, -207.9789276123047], "reading_logprob_std": 6.175733513431951, "tokenization_path_logprob_std": 183.87501654677808, "trajectory_balance_rmse": 4853.956831392736}1152{"canonical_length": 347, "canonical_nll": 1.0733235383926962, "generation_rouge_l_f1": 0.25407166123778496, "generation_token_f1": 0.5635179153094463, "prompt_id": "8d16eaea1c875841", "reading_canonical_log_prob": -372.4432678222656, "reading_distances": [41, 42, 47, 53], "reading_log_probs": [-383.57781982421875, -386.8433532714844, -384.6949462890625, -387.4228515625], "reading_logprob_std": 1.8048288772603014, "tokenization_path_logprob_std": 161.48437726088943, "trajectory_balance_rmse": 7611.635499679803}1153{"canonical_length": 355, "canonical_nll": 0.9550095249229753, "generation_rouge_l_f1": 0.2601626016260163, "generation_token_f1": 0.5008130081300812, "prompt_id": "f4746ccf1632be92", "reading_canonical_log_prob": -339.02838134765625, "reading_distances": [23, 26, 23, 20], "reading_log_probs": [-342.7545471191406, -340.18951416015625, -344.10443115234375, -340.0690612792969], "reading_logprob_std": 1.9840763834484436, "tokenization_path_logprob_std": 241.6895517261005, "trajectory_balance_rmse": 7460.901488391807}1154{"canonical_length": 450, "canonical_nll": 1.0521318901909722, "generation_rouge_l_f1": 0.19774011299435026, "generation_token_f1": 0.3418079096045198, "prompt_id": "06190df6c1e1573b", "reading_canonical_log_prob": -473.4593505859375, "reading_distances": [16, 15, 17, 12], "reading_log_probs": [-477.75689697265625, -477.7242431640625, -474.7752380371094, -476.562255859375], "reading_logprob_std": 1.4011543485292695, "tokenization_path_logprob_std": 312.53421934876314, "trajectory_balance_rmse": 7592.401814239468}1155{"canonical_length": 102, "canonical_nll": 0.9073232014973959, "generation_rouge_l_f1": 0.27042253521126763, "generation_token_f1": 0.37183098591549296, "prompt_id": "56fa8783ee42ce19", "reading_canonical_log_prob": -92.54696655273438, "reading_distances": [13, 19, 16, 19], "reading_log_probs": [-93.839111328125, -93.09306335449219, -92.89100646972656, -91.70962524414062], "reading_logprob_std": 0.8822599283178342, "tokenization_path_logprob_std": 112.95397918313178, "trajectory_balance_rmse": 1950.5346608196216}1156{"canonical_length": 202, "canonical_nll": 2.2453555871944615, "generation_rouge_l_f1": 0.28710462287104627, "generation_token_f1": 0.4914841849148418, "prompt_id": "11f8f65e51ee2cf5", "reading_canonical_log_prob": -453.56182861328125, "reading_distances": [22, 23, 23, 20], "reading_log_probs": [-458.53228759765625, -460.79425048828125, -459.1257019042969, -460.6204833984375], "reading_logprob_std": 1.1134705300001257, "tokenization_path_logprob_std": 161.9668982969156, "trajectory_balance_rmse": 3549.9619809746587}1157{"canonical_length": 11, "canonical_nll": 2.659000049937855, "generation_rouge_l_f1": 0.2, "generation_token_f1": 0.2, "prompt_id": "8980941d8fb125ec", "reading_canonical_log_prob": -29.249000549316406, "reading_distances": [15, 18, 16, 21], "reading_log_probs": [-32.01899337768555, -40.81427001953125, -39.345523834228516, -36.76975631713867], "reading_logprob_std": 3.8595614279335275, "tokenization_path_logprob_std": 42.330532442358894, "trajectory_balance_rmse": 239.31123729450175}1158{"canonical_length": 69, "canonical_nll": 0.5367729076440784, "generation_rouge_l_f1": 0.20588235294117646, "generation_token_f1": 0.3431372549019608, "prompt_id": "2d0e3ff402d06b28", "reading_canonical_log_prob": -37.037330627441406, "reading_distances": [16, 14, 13, 14], "reading_log_probs": [-50.25616455078125, -44.902076721191406, -48.460121154785156, -41.56532287597656], "reading_logprob_std": 3.8595617961999116, "tokenization_path_logprob_std": 131.61262538515697, "trajectory_balance_rmse": 947.9396063660722}1159{"canonical_length": 53, "canonical_nll": 0.5856976059247863, "generation_rouge_l_f1": 0.18930041152263377, "generation_token_f1": 0.2716049382716049, "prompt_id": "54b6e30181a1bec1", "reading_canonical_log_prob": -31.041973114013672, "reading_distances": [19, 24, 19, 16], "reading_log_probs": [-27.688323974609375, -29.05846405029297, -31.64278793334961, -28.595294952392578], "reading_logprob_std": 1.6960280657386917, "tokenization_path_logprob_std": 21.47594736342194, "trajectory_balance_rmse": 429.0170959423824}1160{"canonical_length": 9, "canonical_nll": 0.8935994042290581, "generation_rouge_l_f1": 0.41025641025641024, "generation_token_f1": 0.41025641025641024, "prompt_id": "2996ca79ae58108b", "reading_canonical_log_prob": -8.042394638061523, "reading_distances": [30, 28, 26, 36], "reading_log_probs": [-8.060506820678711, -7.914007186889648, -8.034048080444336, -7.543618202209473], "reading_logprob_std": 0.23830221237982654, "tokenization_path_logprob_std": 42.22922306575307, "trajectory_balance_rmse": 140.69519750988718}1161{"canonical_length": 11, "canonical_nll": 1.4016671614213423, "generation_rouge_l_f1": 0.16, "generation_token_f1": 0.16, "prompt_id": "e6cc80b8ba91ddd7", "reading_canonical_log_prob": -15.418338775634766, "reading_distances": [21, 22, 27, 25], "reading_log_probs": [-16.904268264770508, -19.186016082763672, -17.836898803710938, -18.968942642211914], "reading_logprob_std": 1.060181378957692, "tokenization_path_logprob_std": 44.263482912838526, "trajectory_balance_rmse": 226.5533809594063}1162{"canonical_length": 218, "canonical_nll": 1.8638242669061784, "generation_rouge_l_f1": 0.22641509433962265, "generation_token_f1": 0.4528301886792453, "prompt_id": "1826061f4f79f3ac", "reading_canonical_log_prob": -406.3136901855469, "reading_distances": [14, 16, 13, 20], "reading_log_probs": [-410.0091552734375, -409.3782958984375, -407.9631652832031, -408.52117919921875], "reading_logprob_std": 0.9058492614361238, "tokenization_path_logprob_std": 158.25805299686058, "trajectory_balance_rmse": 4676.83206725419}1163{"canonical_length": 182, "canonical_nll": 1.8046948778760301, "generation_rouge_l_f1": 0.21479713603818615, "generation_token_f1": 0.4534606205250597, "prompt_id": "d48fbaa3dac1814d", "reading_canonical_log_prob": -328.4544677734375, "reading_distances": [29, 25, 30, 24], "reading_log_probs": [-342.60308837890625, -341.4290771484375, -351.630126953125, -347.2237548828125], "reading_logprob_std": 4.6661199862682805, "tokenization_path_logprob_std": 32.66820395278296, "trajectory_balance_rmse": 3692.5845544855433}1164{"canonical_length": 105, "canonical_nll": 0.577705819266183, "generation_rouge_l_f1": 0.3496932515337423, "generation_token_f1": 0.43558282208588966, "prompt_id": "7fbf1f51feeeeb73", "reading_canonical_log_prob": -60.65911102294922, "reading_distances": [17, 20, 17, 20], "reading_log_probs": [-65.66716003417969, -62.974365234375, -61.19239807128906, -62.404029846191406], "reading_logprob_std": 1.8905774705415723, "tokenization_path_logprob_std": 72.98940356337387, "trajectory_balance_rmse": 857.4781274108417}1165{"canonical_length": 18, "canonical_nll": 0.3793629010518392, "generation_rouge_l_f1": 0.1037037037037037, "generation_token_f1": 0.11851851851851852, "prompt_id": "8cc0717e876412c6", "reading_canonical_log_prob": -6.8285322189331055, "reading_distances": [20, 21, 29, 26], "reading_log_probs": [-8.787773132324219, -11.084543228149414, -7.9134955406188965, -10.639596939086914], "reading_logprob_std": 1.5042693426599847, "tokenization_path_logprob_std": 85.6119751077066, "trajectory_balance_rmse": 364.14480051889876}1166{"canonical_length": 152, "canonical_nll": 1.9959291156969572, "generation_rouge_l_f1": 0.23469387755102042, "generation_token_f1": 0.3673469387755102, "prompt_id": "3f19d2fd29921f16", "reading_canonical_log_prob": -303.3812255859375, "reading_distances": [19, 17, 16, 12], "reading_log_probs": [-312.19073486328125, -314.116943359375, -320.2994689941406, -313.4577331542969], "reading_logprob_std": 3.611724699572672, "tokenization_path_logprob_std": 303.410375344361, "trajectory_balance_rmse": 3031.61475780782}1167{"canonical_length": 257, "canonical_nll": 1.9743070491093142, "generation_rouge_l_f1": 0.19915254237288135, "generation_token_f1": 0.38135593220338987, "prompt_id": "774504b35575be36", "reading_canonical_log_prob": -507.39691162109375, "reading_distances": [31, 24, 28, 29], "reading_log_probs": [-514.38427734375, -510.4630126953125, -515.7324829101562, -514.1043090820312], "reading_logprob_std": 2.253694423526105, "tokenization_path_logprob_std": 314.72284965798167, "trajectory_balance_rmse": 5495.415951844061}1168{"canonical_length": 9, "canonical_nll": 0.431236637963189, "generation_rouge_l_f1": 0.0, "generation_token_f1": 0.0, "prompt_id": "18b528d0286abbc1", "reading_canonical_log_prob": -3.881129741668701, "reading_distances": [34, 30, 33, 31], "reading_log_probs": [-1.8169488906860352, -2.7718164920806885, -2.1719167232513428, -2.456108331680298], "reading_logprob_std": 0.40687993057409116, "tokenization_path_logprob_std": 17.301323130973074, "trajectory_balance_rmse": 191.7662528949343}1169{"canonical_length": 4, "canonical_nll": 1.7426186800003052, "generation_rouge_l_f1": 0.5714285714285715, "generation_token_f1": 0.5714285714285715, "prompt_id": "ad4fd52fcf4632b3", "reading_canonical_log_prob": -6.970474720001221, "reading_distances": [22, 29, 20, 25], "reading_log_probs": [-5.677864074707031, -5.1765336990356445, -5.813916206359863, -5.216721057891846], "reading_logprob_std": 0.32236221085708155, "tokenization_path_logprob_std": 30.30021798919276, "trajectory_balance_rmse": 92.86898957198696}1170{"canonical_length": 349, "canonical_nll": 1.5130268960422637, "generation_rouge_l_f1": 0.18493150684931509, "generation_token_f1": 0.410958904109589, "prompt_id": "ba9050b431bd2ba0", "reading_canonical_log_prob": -528.04638671875, "reading_distances": [18, 15, 21, 19], "reading_log_probs": [-522.6118774414062, -527.091064453125, -523.2296752929688, -521.246826171875], "reading_logprob_std": 2.5050830560252297, "tokenization_path_logprob_std": 274.0551410833943, "trajectory_balance_rmse": 6534.811135174135}1171{"canonical_length": 35, "canonical_nll": 0.6896263122558594, "generation_rouge_l_f1": 0.12781954887218044, "generation_token_f1": 0.19548872180451127, "prompt_id": "da7f7e74060fc4e6", "reading_canonical_log_prob": -24.136920928955078, "reading_distances": [27, 33, 26, 26], "reading_log_probs": [-24.015052795410156, -22.75440216064453, -23.454326629638672, -23.487041473388672], "reading_logprob_std": 0.5172180953568783, "tokenization_path_logprob_std": 76.21050499491386, "trajectory_balance_rmse": 682.642086344806}1172{"canonical_length": 366, "canonical_nll": 1.114656792312372, "generation_rouge_l_f1": 0.21664050235478807, "generation_token_f1": 0.43956043956043955, "prompt_id": "378c4d815b07ba59", "reading_canonical_log_prob": -407.9643859863281, "reading_distances": [29, 25, 29, 34], "reading_log_probs": [-415.68292236328125, -412.7955322265625, -418.077392578125, -420.3497009277344], "reading_logprob_std": 3.240064690145231, "tokenization_path_logprob_std": 365.54804097040125, "trajectory_balance_rmse": 7057.026579122546}1173{"canonical_length": 158, "canonical_nll": 0.5975696225709553, "generation_rouge_l_f1": 0.46913580246913583, "generation_token_f1": 0.6255144032921811, "prompt_id": "e2790ed28464c480", "reading_canonical_log_prob": -94.41600036621094, "reading_distances": [14, 24, 22, 19], "reading_log_probs": [-95.3733901977539, -95.28754425048828, -94.98316192626953, -95.69978332519531], "reading_logprob_std": 0.29471963903273946, "tokenization_path_logprob_std": 173.0602780608511, "trajectory_balance_rmse": 2080.2666355868305}1174{"canonical_length": 253, "canonical_nll": 1.2553328563102149, "generation_rouge_l_f1": 0.2671905697445972, "generation_token_f1": 0.5304518664047152, "prompt_id": "2481a184df55ff40", "reading_canonical_log_prob": -317.5992126464844, "reading_distances": [47, 39, 37, 41], "reading_log_probs": [-343.9942626953125, -332.8840637207031, -336.337890625, -343.7132873535156], "reading_logprob_std": 5.52066559509569, "tokenization_path_logprob_std": 240.5584767624945, "trajectory_balance_rmse": 4653.16967534523}1175{"canonical_length": 342, "canonical_nll": 1.5492291366844846, "generation_rouge_l_f1": 0.19865319865319866, "generation_token_f1": 0.393939393939394, "prompt_id": "130ff8d17ea30994", "reading_canonical_log_prob": -529.8363647460938, "reading_distances": [18, 20, 17, 21], "reading_log_probs": [-545.3275756835938, -550.9266357421875, -542.5791625976562, -541.7706909179688], "reading_logprob_std": 4.140450229607613, "tokenization_path_logprob_std": 139.5766657248354, "trajectory_balance_rmse": 7635.520147842602}1176{"canonical_length": 16, "canonical_nll": 0.754535436630249, "generation_rouge_l_f1": 0.7777777777777778, "generation_token_f1": 0.8888888888888888, "prompt_id": "16bf564574acdadf", "reading_canonical_log_prob": -12.072566986083984, "reading_distances": [27, 28, 22, 31], "reading_log_probs": [-16.984865188598633, -16.169652938842773, -14.849013328552246, -18.878997802734375], "reading_logprob_std": 1.6867025516784084, "tokenization_path_logprob_std": 27.07947410316818, "trajectory_balance_rmse": 220.2790189195804}1177{"canonical_length": 50, "canonical_nll": 0.6298949813842774, "generation_rouge_l_f1": 0.1890547263681592, "generation_token_f1": 0.31840796019900497, "prompt_id": "effedc7b02b1c679", "reading_canonical_log_prob": -31.494749069213867, "reading_distances": [29, 29, 27, 26], "reading_log_probs": [-37.19982147216797, -35.650917053222656, -36.36476135253906, -31.25717544555664], "reading_logprob_std": 2.6506830078275723, "tokenization_path_logprob_std": 62.4391268807624, "trajectory_balance_rmse": 781.8526077849417}1178{"canonical_length": 176, "canonical_nll": 1.4085471413352273, "generation_rouge_l_f1": 0.17674418604651163, "generation_token_f1": 0.3581395348837209, "prompt_id": "33cc463f90c8e3a1", "reading_canonical_log_prob": -247.904296875, "reading_distances": [25, 17, 25, 28], "reading_log_probs": [-246.01425170898438, -246.36546325683594, -246.04249572753906, -247.07485961914062], "reading_logprob_std": 0.49348782858414975, "tokenization_path_logprob_std": 100.68100019506808, "trajectory_balance_rmse": 3624.385102810858}1179{"canonical_length": 283, "canonical_nll": 1.2341067041188163, "generation_rouge_l_f1": 0.24686940966010731, "generation_token_f1": 0.5330948121645797, "prompt_id": "e32f1e4a8a806d4a", "reading_canonical_log_prob": -349.252197265625, "reading_distances": [15, 14, 15, 15], "reading_log_probs": [-358.95623779296875, -352.8197937011719, -354.6418151855469, -356.13177490234375], "reading_logprob_std": 2.5941720077868275, "tokenization_path_logprob_std": 358.71525461705687, "trajectory_balance_rmse": 6123.158634048985}1180{"canonical_length": 193, "canonical_nll": 1.2627100178614798, "generation_rouge_l_f1": 0.19438444924406048, "generation_token_f1": 0.38876889848812096, "prompt_id": "f16b4d6e1a4f11c6", "reading_canonical_log_prob": -243.70303344726562, "reading_distances": [11, 8, 9, 13], "reading_log_probs": [-250.18663024902344, -256.26788330078125, -253.5135498046875, -255.08648681640625], "reading_logprob_std": 2.6380991940873275, "tokenization_path_logprob_std": 189.83516905346087, "trajectory_balance_rmse": 4229.920723039991}1181{"canonical_length": 11, "canonical_nll": 0.900890437039462, "generation_rouge_l_f1": 0.7777777777777778, "generation_token_f1": 0.8888888888888888, "prompt_id": "7cf86e055cecbc26", "reading_canonical_log_prob": -9.909794807434082, "reading_distances": [9, 16, 13, 16], "reading_log_probs": [-10.2212495803833, -11.681831359863281, -10.98697280883789, -12.288783073425293], "reading_logprob_std": 0.8916396271059821, "tokenization_path_logprob_std": 35.457802930054974, "trajectory_balance_rmse": 197.45015838339444}1182{"canonical_length": 251, "canonical_nll": 1.5619380388601842, "generation_rouge_l_f1": 0.21696252465483237, "generation_token_f1": 0.42998027613412226, "prompt_id": "279ec9f263f853e8", "reading_canonical_log_prob": -392.04644775390625, "reading_distances": [29, 20, 27, 24], "reading_log_probs": [-390.5419921875, -393.2213439941406, -387.767822265625, -386.6112060546875], "reading_logprob_std": 2.9594285535455493, "tokenization_path_logprob_std": 201.9871168865536, "trajectory_balance_rmse": 5417.87170216087}1183{"canonical_length": 102, "canonical_nll": 1.2455711364746094, "generation_rouge_l_f1": 0.32098765432098764, "generation_token_f1": 0.4567901234567901, "prompt_id": "cf98e6698c25bbf0", "reading_canonical_log_prob": -127.04825592041016, "reading_distances": [110, 109, 100, 109], "reading_log_probs": [-119.18705749511719, -119.64510345458984, -125.99453735351562, -116.6880874633789], "reading_logprob_std": 3.9630134084825155, "tokenization_path_logprob_std": 159.3516753156645, "trajectory_balance_rmse": 1830.6312760839987}1184{"canonical_length": 125, "canonical_nll": 1.50201171875, "generation_rouge_l_f1": 0.1964735516372796, "generation_token_f1": 0.397984886649874, "prompt_id": "588a316f7d854f31", "reading_canonical_log_prob": -187.75146484375, "reading_distances": [34, 25, 33, 35], "reading_log_probs": [-194.733154296875, -191.7010498046875, -194.99783325195312, -190.99172973632812], "reading_logprob_std": 2.0551308104861437, "tokenization_path_logprob_std": 133.79307105630062, "trajectory_balance_rmse": 3213.279395413211}1185{"canonical_length": 116, "canonical_nll": 1.843872201853785, "generation_rouge_l_f1": 0.23268698060941828, "generation_token_f1": 0.371191135734072, "prompt_id": "6bdc8fa7262e4a79", "reading_canonical_log_prob": -213.88917541503906, "reading_distances": [18, 15, 18, 11], "reading_log_probs": [-213.30035400390625, -227.48992919921875, -216.11029052734375, -221.99244689941406], "reading_logprob_std": 6.318692625310894, "tokenization_path_logprob_std": 174.97913173619344, "trajectory_balance_rmse": 2485.320216215748}1186{"canonical_length": 7, "canonical_nll": 2.7416093008858815, "generation_rouge_l_f1": 0.2, "generation_token_f1": 0.2, "prompt_id": "73284d596fde1232", "reading_canonical_log_prob": -19.191265106201172, "reading_distances": [26, 25, 31, 27], "reading_log_probs": [-26.080543518066406, -22.304473876953125, -20.091121673583984, -21.73085594177246], "reading_logprob_std": 2.5325934000390795, "tokenization_path_logprob_std": 29.290445542640065, "trajectory_balance_rmse": 176.77938479364502}1187{"canonical_length": 47, "canonical_nll": 1.0833739422737283, "generation_rouge_l_f1": 0.15602836879432622, "generation_token_f1": 0.22695035460992907, "prompt_id": "08b3c472b8ed1840", "reading_canonical_log_prob": -50.918575286865234, "reading_distances": [37, 51, 48, 51], "reading_log_probs": [-50.27709197998047, -52.31874084472656, -54.178592681884766, -50.220787048339844], "reading_logprob_std": 1.8911612940219187, "tokenization_path_logprob_std": 63.6894411076199, "trajectory_balance_rmse": 638.2438920272166}1188{"canonical_length": 149, "canonical_nll": 0.7746859556876573, "generation_rouge_l_f1": 0.29649595687331537, "generation_token_f1": 0.45822102425876005, "prompt_id": "51fa8b33eba8ebe6", "reading_canonical_log_prob": -115.42820739746094, "reading_distances": [40, 36, 42, 45], "reading_log_probs": [-119.04322814941406, -122.62535858154297, -121.24957275390625, -121.67088317871094], "reading_logprob_std": 1.516176614374996, "tokenization_path_logprob_std": 71.72535434876143, "trajectory_balance_rmse": 2350.191411777714}1189{"canonical_length": 25, "canonical_nll": 0.8124961853027344, "generation_rouge_l_f1": 0.2686567164179105, "generation_token_f1": 0.3134328358208956, "prompt_id": "6e11957de8d37c71", "reading_canonical_log_prob": -20.31240463256836, "reading_distances": [26, 22, 27, 19], "reading_log_probs": [-21.340269088745117, -22.62666130065918, -21.595279693603516, -23.857208251953125], "reading_logprob_std": 1.1456123800738887, "tokenization_path_logprob_std": 30.037048314165784, "trajectory_balance_rmse": 477.7331151610306}1190{"canonical_length": 124, "canonical_nll": 1.1737483855216735, "generation_rouge_l_f1": 0.2210796915167095, "generation_token_f1": 0.4061696658097686, "prompt_id": "613e08997d787f9d", "reading_canonical_log_prob": -145.5447998046875, "reading_distances": [21, 18, 24, 20], "reading_log_probs": [-143.04635620117188, -145.76669311523438, -143.3068084716797, -147.36878967285156], "reading_logprob_std": 2.066981223639106, "tokenization_path_logprob_std": 200.7140928755919, "trajectory_balance_rmse": 2785.6575848175653}1191{"canonical_length": 7, "canonical_nll": 0.16052302292415074, "generation_rouge_l_f1": 1.0, "generation_token_f1": 1.0, "prompt_id": "82a7bb73a8e6bdca", "reading_canonical_log_prob": -1.1236611604690552, "reading_distances": [33, 34, 31, 40], "reading_log_probs": [-1.8113667964935303, -1.445715308189392, -2.253610610961914, -1.5671390295028687], "reading_logprob_std": 0.3567923739135296, "tokenization_path_logprob_std": 16.383844449223908, "trajectory_balance_rmse": 104.97244631531049}1192{"canonical_length": 500, "canonical_nll": 0.81297900390625, "generation_rouge_l_f1": 0.15172413793103448, "generation_token_f1": 0.2620689655172414, "prompt_id": "04c1fbf027769e8a", "reading_canonical_log_prob": -406.489501953125, "reading_distances": [24, 24, 28, 22], "reading_log_probs": [-417.69384765625, -417.3563537597656, -428.31036376953125, -417.96038818359375], "reading_logprob_std": 5.325821907423919, "tokenization_path_logprob_std": 200.21467146696898, "trajectory_balance_rmse": 8043.086034663614}1193{"canonical_length": 47, "canonical_nll": 1.896366038220994, "generation_rouge_l_f1": 0.17322834645669294, "generation_token_f1": 0.2677165354330709, "prompt_id": "5950f3bf48416566", "reading_canonical_log_prob": -89.12920379638672, "reading_distances": [32, 35, 36, 48], "reading_log_probs": [-96.4745101928711, -89.72310638427734, -92.9197006225586, -91.32283020019531], "reading_logprob_std": 2.8879816679367125, "tokenization_path_logprob_std": 73.11880625675266, "trajectory_balance_rmse": 1057.195269695691}1194{"canonical_length": 310, "canonical_nll": 1.1810854019657258, "generation_rouge_l_f1": 0.2756183745583039, "generation_token_f1": 0.4770318021201413, "prompt_id": "f93b23d1ff15b3d1", "reading_canonical_log_prob": -366.136474609375, "reading_distances": [22, 25, 23, 24], "reading_log_probs": [-361.2095947265625, -361.9642333984375, -361.508544921875, -359.565673828125], "reading_logprob_std": 1.0447015024886719, "tokenization_path_logprob_std": 249.32821451384623, "trajectory_balance_rmse": 5735.917032198753}1195{"canonical_length": 228, "canonical_nll": 1.3983487580951892, "generation_rouge_l_f1": 0.2414486921529175, "generation_token_f1": 0.45070422535211263, "prompt_id": "cb117bc3f0e8e883", "reading_canonical_log_prob": -318.8235168457031, "reading_distances": [25, 19, 19, 23], "reading_log_probs": [-328.0265808105469, -324.98382568359375, -320.8035583496094, -326.6986083984375], "reading_logprob_std": 3.1406030840963086, "tokenization_path_logprob_std": 300.3520167844739, "trajectory_balance_rmse": 4702.608865122326}1196{"canonical_length": 318, "canonical_nll": 1.8733478282232705, "generation_rouge_l_f1": 0.21832358674463936, "generation_token_f1": 0.49902534113060426, "prompt_id": "0ff4ea021b8e685b", "reading_canonical_log_prob": -595.724609375, "reading_distances": [16, 20, 14, 14], "reading_log_probs": [-599.964111328125, -606.0427856445312, -603.8404541015625, -600.5645141601562], "reading_logprob_std": 2.8567216566693108, "tokenization_path_logprob_std": 220.25505881231632, "trajectory_balance_rmse": 6493.275473559726}1197{"canonical_length": 70, "canonical_nll": 1.517994907924107, "generation_rouge_l_f1": 0.22352941176470587, "generation_token_f1": 0.3529411764705882, "prompt_id": "d8931b93e350f362", "reading_canonical_log_prob": -106.2596435546875, "reading_distances": [14, 16, 14, 11], "reading_log_probs": [-110.76921081542969, -112.06926727294922, -106.95133972167969, -112.42951965332031], "reading_logprob_std": 2.505931820691401, "tokenization_path_logprob_std": 115.00975243022906, "trajectory_balance_rmse": 1643.7055263595494}1198{"canonical_length": 43, "canonical_nll": 1.9972502242687136, "generation_rouge_l_f1": 0.3488372093023256, "generation_token_f1": 0.46511627906976744, "prompt_id": "cfd710328af17995", "reading_canonical_log_prob": -85.88175964355469, "reading_distances": [32, 31, 27, 30], "reading_log_probs": [-87.07557678222656, -88.95108032226562, -85.93759155273438, -83.38697814941406], "reading_logprob_std": 2.3267334053068076, "tokenization_path_logprob_std": 80.18251608949525, "trajectory_balance_rmse": 840.5582333979407}1199{"canonical_length": 237, "canonical_nll": 0.7680912581174183, "generation_rouge_l_f1": 0.2993492407809111, "generation_token_f1": 0.3774403470715835, "prompt_id": "3f921f925896c713", "reading_canonical_log_prob": -182.03762817382812, "reading_distances": [37, 35, 42, 39], "reading_log_probs": [-205.52755737304688, -196.45669555664062, -205.33743286132812, -201.58236694335938], "reading_logprob_std": 4.2536428626587925, "tokenization_path_logprob_std": 109.53918205693012, "trajectory_balance_rmse": 2244.669832983569}1200{"canonical_length": 242, "canonical_nll": 1.0406487835340263, "generation_rouge_l_f1": 0.3220675944333996, "generation_token_f1": 0.5049701789264414, "prompt_id": "74c9b8835b365ad8", "reading_canonical_log_prob": -251.83700561523438, "reading_distances": [21, 15, 17, 19], "reading_log_probs": [-262.7458190917969, -254.2945556640625, -252.0796661376953, -256.7156677246094], "reading_logprob_std": 4.599034570918568, "tokenization_path_logprob_std": 137.88174392557318, "trajectory_balance_rmse": 4514.873475395476}

Showing the first 1,200 of 2478 lines. Download the file for the rest.