Характеристики

Библиотека: kraken
Опубликована: 30 авг 2026

Ссылки

kraken-medieval-german-v2

Описание изображенийИзображение → текст3 профессии

Модель Kraken для распознавания немецких рукописей XV–XVI веков и других исторических документов по изображениям страниц или строк.

24скачиваний за 30 дней2лайкавеса обновлены1 окт 2026

Бенчмарки

Как модель справилась с задачами пользователей ailist и что о ней публикуют открытые лидерборды

Публичные оценки

Из открытых лидербордов. Сравнивайте результаты одного теста при одинаковых условиях — отсутствие оценки не значит, что модель хуже.

В подключённых источниках результатов для этой модели пока нет.

Что умеет модель

Описание составлено по карточке модели и метаданным Hugging Face

kraken-medieval-german-v2 — модель распознавания рукописного текста (HTR) на базе Kraken, разработанная dh-unibe. Она дообучена на CATMuS Medieval, модели для средневековых рукописей, и предназначена для чтения немецких исторических документов. Для обработки страницы Kraken сначала находит строки с помощью сегментации blla, а распознаватель считывает их по отдельности; на вход также можно подавать готовые изображения строк или их контуры. Модель обучена на наборах данных с судебными книгами XV–XVI веков, материалами авторов Bullinger и другими историческими документами.

На отложенной выборке собственного обучения указаны CER 21,31% и WER 59,70%; это результаты внутренней проверки, а не общего бенчмарка, и они могут не переноситься на другие корпуса. Базовая модель CATMuS Medieval обучалась на старофранцузском, среднефранцузском, латинском, испанском и итальянском, поэтому пригодность именно для немецких рукописей следует оценивать на целевом корпусе. В репозитории представлены веса Core ML; лицензия в предоставленных сведениях не указана.

Сценарии применения

распознавание рукописей в средневековых судебных книгах
оцифровка немецких исторических документов XV–XVI веков
извлечение текста из сканов архивных рукописей
транскрибирование отдельных строк исторических документов
подготовка транскрипций рукописей для цифровых гуманитарных исследований

Ограничения

CER и WER измерены на собственной отложенной выборке и не гарантируют такого же качества на других коллекциях; качество для целевого корпуса следует проверять отдельно. Базовая CATMuS Medieval обучалась на других языках, а лицензия в предоставленных сведениях не указана.

Для каких профессий подходит3

Разделы каталога профессий, в задачах которых используется эта модель

Все профессии

IV.9Отраслевые задачиМедиа, журналистика, культура, архивы и игрыAI помогает создавать, искать, сохранять и персонализировать медиа, культурный и игровой контент.44 задачи14 938моделей I.3Универсальные способностиИзвлечение и структурирование данныхAI помогает превращать разрозненные данные в структуру, пригодную для поиска, аналитики и автоматизации.18 задач25 791модель I.4Универсальные способностиОбработка документов и формAI помогает оцифровывать документы, проверять их комплектность и сокращать ручной ввод данных.16 задач7 180моделей

Карточка модели

README от автора на Hugging Face, как он его опубликовал

Открыть на Hugging Face

kraken-medieval-german-v2

Handwritten-text-recognition model trained on the serving-atr-inference training service. These are the weights of the best validation checkpoint of the run below — not its last epoch.

Base model

Fine-tuned from CATMuS Medieval (10.5281/zenodo.15030337, Pinche & Clérice 2025): a kraken HTR model for medieval manuscripts in Old and Middle French, Latin, Spanish and Italian, trained on graphematic transcriptions.

The training request named that base kraken-early_modern_german, the id it carried in this project's gateway registry at the time. That id was wrong. It described "early modern German, Kurrent, 16th–17th c." while the DOI behind it is CATMuS Medieval, which is neither German nor Kurrent nor early modern. The registry was corrected on 2026-10-01 (serving-atr-inference#101, #198) and the same DOI is now served as kraken-catmus_medieval. The weights this model was fine-tuned from are unchanged; only their name was.

Evaluation

metric	value
CER	21.31 %
WER	59.70 %
samples scored	—
characters scored	882255
character errors	188022

Measured on this run's own held-out validation split (page-level and seeded, so no page contributes lines to both sides). It is not a score on a shared benchmark and does not transfer to a different corpus.

Input: a page, and kraken segments it

Pass the page: kraken's own baseline segmentation (blla) finds the lines and this recogniser reads them one at a time — which is also the shape the CER below was measured in. The model never sees a whole page as a single input, so a line-level score is what it has. If you already have line polygons or crops, give kraken those instead of a scan.

This is the opposite arrangement from the VLM fine-tunes beside it (qwen3vl-german-xix-v2 and siblings): those are asked for one line per call because a page comes back as a fragment (thodel/serving-atr-inference#165).

Training data

dh-unibe/image-text_rats-und-richtebuecher_xv-xvi
- Training projects: Rats-undRichtebücher_MF_1_3543, Rats-undRichtebücher_MF_1_3544, Rats-undRichtebücher_MF_1_3545, Rats-undRichtebücher_MF_1_3546, Rats-undRichtebücher_MF_1_3547, Rats-undRichtebücher_MF_1_3548, Rats-undRichtebücher_MF_1_3549, Rats-undRichtebücher_MF_1_3550, Rats-undRichtebücher_MF_1_3551, Rats-undRichtebücher_MF_1_3552, Rats-undRichtebücher_MF_1_3553, Rats-undRichtebücher_MF_1_3554, Rats-undRichtebücher_MF_1_3555, Rats-undRichtebücher_MF_1_3556, Rats-undRichtebücher_MF_1_3557, Rats-undRichtebücher_MF_1_3558, Rats-undRichtebücher_MF_1_3559, Rats-undRichtebücher_MF_1_3560, Rats-undRichtebücher_MF_1_3561, Rats-undRichtebücher_MF_1_3562, Rats-undRichtebücher_MF_1_3563, Rats-undRichtebücher_MF_1_3564, Rats-undRichtebücher_MF_1_3565, Rats-undRichtebücher_MF_1_3566, Rats-undRichtebücher_MF_1_3567, Rats-undRichtebücher_MF_1_3568, Rats-undRichtebücher_MF_1_3569, Rats-undRichtebücher_MF_1_3570, Rats-undRichtebücher_MF_1_3571, Rats-undRichtebücher_MF_1_3572, Rats-undRichtebücher_MF_1_3573, Rats-undRichtebücher_MF_1_3574, Rats-undRichtebücher_MF_1_3575, TRAINING_VALIDATION_SET_Richtebuecher_M1, Test_MF_1_3556_p_204, escript_test, escript_test_2
- Evaluation: a seeded page-level split of the training projects (partition=0.9, seed=42)
dh-unibe/image-text_bullinger-autoren
- Training projects: 0008_Aberlin_Joachim, 0049_Adlischwyler_Hans_Jakob, 0066_Adlischwyler_Johannes, 0152_Altieri_Baldassare, 0164_Amerbach_Bonifacius, 0179_Aquilomontanus_Hermann, 0224_a_Lasco_Johannes, 0257_Beyel_Werner, 0260_Bucher_Melchior, 0261_Buchser_Johannes, 0278_Bartholomäus_Adam, 0315_Bedrot_Jakob, 0319_Belényesi_Gergely, 0328_Bersius_Marcus, 0330_Bertlin_Bartholomäus, 0343_Biberli_Lienhart, 0350_Bing_Simon, 0351_Birck_Sixt, 0391_Bitter_Dietrich, 0392_Blarer_Ambrosius, 0396_Blasius_Johannes, 0403_Blankenheim_Lorenz, 0405_Blarer_Thomas, 0431_Boltz_Valentin, 0437_Borrhaus_Martin, 0459_Brennwald_Heinrich, 0461_Brunner_Fridolin, 0465_Brunner_Leonhard, 0467_Bucer_Martin, 0495_Bullinger_Heinrich, 0505_Burcher_John, 0506_Butler_John, 0536_Caesarius_Johannes, 0540_Calvin_Johannes, 0578_Capito_Wolfgang, 0581_Cassander_Georg, 0599_Claudius_Matthias, 0646_Coletto_Andrea, 0654_Comander_Johannes, 0659_Comte_Béat, 0667_Cousin_Gilbert, 0671_Coverdale_Miles, 0703_Crodel_Markus, 0714_Curione_Celio_Secondo, 0756_Distel_Georg, 0761_Danmatter_Christian, 0840_Drome_Michael, 0985_de_Hotot_Jean, 0987_del_Prato_Bartolomeo, 0991_Edelmann_Martin, 1001_Eliott_Nicholas, 1010_Erb_Matthias, 1039_Euander_Benedikt, 1041_Edlibach_Hans, 1045_Fischer_Ulrich, 1047_Frosch_Johannes, 1049_Farel_Guillaume, 1055_Frecht_Martin, 1056_Frick_Konrad, 1057_Fuchs_Peter, 1083_Schmid_aus_Bergheim_Johannes, 1184_Frey_Johannes_Leopold, 1185_Frey_Johann_Leopold, 1188_Frey_Johannes, 1214_Fries_Johannes, 1218_Frölich_Georg, 1250_Furtmüller_Johann_Valentin, 1262_Gallicius_Philipp, 1278_Gast_Johannes, 1286_Gessner_Konrad, 1309_Grübel_Sebastian, 1365_Grynäus_Simon, 1381_Gwalther_Rudolf, 1384_Geldenhauer_Gerhard, 1385_Gering_Beat, 1387_Gassner_Thomas, 1388_Guldi_Niklaus, 1391_Haller_Berchtold, 1393_Hegner_Gebhart, 1395_Hewer_Jakob, 1396_Hirt_Balthasar, 1399_Huber_Peter, 1402_Im_Haag_Peter, 1406_Haller_Sulpitius, 1407_Heim_Luzi, 1409_Hentius_Martin, 1410_Herbrot_Jakob, 1414_Hindermann_Heinrich, 1417_Hospinian_Leonhard, 1419_Haab_Johannes, 1445_Haller_Johannes, 1453_Haller_Wolfgang, 1456_Happel_Wigand, 1458_Hardenberg_Albert, 1548_Hilles_Richard_____, 1553_Hochholzer_Christian, 1560_Hooper_John, 1569_Hospinian_Christian, 1571_Hospinian_Johannes, 1753_Jud_Leo, 1758_Kilchrat_Hans, 1760_Kromer_Benedikt, 1761_Kuster_Heinrich, 1764_Kambli_Johannes, 1767_Kappeler_Veit, 1780_Keller_Diethelm, 1804_Kilchmeyer_Jodocus, 1810_Klauser_Konrad, 1837_König_Nikolaus, 1857_Kappeler_d_J__Veit, 1858_Karlstadt_Andreas, 1860_Keller_Michael, 1863_Knight_Thomas, 1864_Kolin_Peter, 1866_Kunz_Peter, 1868_Lüthi_Heinrich, 1884_Lavater_Ludwig, 1888_Lavater_Hans_Rudolf, 1903_Lening_Johannes, 1915_Lindauer_Bernhard, 1999_Lüthard_Konrad, 2000_Maler_Hans, 2003_Muntprat_Heinrich, 2004_Mutschli_Hans, 2005_Mösel_Wolfgang, 2013_Müller_gen__Maier_Nikolaus, 2095_Melanchthon_Philipp, 2115_Meyer_Lorenz, 2229_Musculus_Wolfgang, 2234_Myconius_Oswald, 2237_Mötteli_Martin, 2258_Macarius_Joseph, 2261_Maurer_Georg, 2262_Medmann_Peter, 2263_Megander_Kaspar, 2264_Melander_Dionysius, 2268_Müller_Bartholomäus, 2272_Meyer_Jakob, 2293_Frey_Hans_Rudolf, 2300_von_Ulm_Heinrich, 2302_Negri_Francesco, 2319_Naogeorg_Thomas, 2380_Oporin_Johannes, 2441_Oechsli_Ludwig, 2442_Oekolampad_Johannes, 2444_Pfiffer_Heinrich, 2446_Partridge_Nicholas, 2448_Pergener_Oswald, 2449_Peutinger_Claudius_Pius, 2452_Piscatorius_Johannes, 2455_Pistorius_Johannes, 2458_Platter_Thomas, 2508_Pellikan_Konrad, 2594_Pfister_Nikolaus, 2608_Pincier_Johannes, 2629_Pontisella_Johannes, 2700_Ratgeb_Hans, 2804_Ryhiner_Heinrich, 2806_Regel_Johannes, 2807_Renato_Camillo, 2808_Rhellikan_Johannes, 2810_Ritter_Erasmus, 2812_Ruman_Thoman, 2817_Reublin_Wilhelm, 2820_Sam_Konrad, 2821_Schappeler_Christoph, 2822_Schlittler_Heinrich, 2823_Schmid_Andreas, 2825_Schütz_Benedikt, 2826_Sellarius_Heinrich, 2828_Spörli_Konrad, 2829_Stadler_Georg, 2834_Stoll_Heinrich, 2836_Strübi_Heinrich, 2838_Schmid_Erasmus, 2839_Schmid_Felix, 2841_Schnyder_Peter, 2845_Schwenckfeld_Kaspar, 2846_Seger_Martin, 2847_Spaldinus_Johannes_Arlius, 2851_Stoll_Balthasar, 2916_Schertlin_Sebastian, 2943_Schuler_Gervasius, 2958_Serin_Leonhard, 2977_Simler_Josias, 2982_Simler_Peter, 3025_Stancaro_Francesco, 3067_Stumpf_Johannes, 3088_Sturm_Johannes, 3099_Sulzer_Simon, 3130_Stier_Sigismund, 3138_Thamer_Theobald, 3147_Thomann_Heinrich, 3175_Toxites_Michael, 3181_Traheron_Bartholomew, 3199_Tschudi_Ägidius, 3206_Tschudi_Valentin, 3239_Tillmann_Bernhard, 3240_Trebellio_Teodosio, 3306_Vadian_Joachim, 3410_Vermigli_Peter_Martyr, 3429_Viret_Pierre, 3446_Vogler_Hans, 3462_Volmar_Melchior, 3481_Vogt_Simprecht, 3493_von_Cham_Bernhard, 3700_von_Rümlang_Eberhard, 3858_von_Landenberg_Hans, 3860_von_Meggen_Jost, 3872_Warner_Francis, 3873_Weingartner_Rudolf, 3875_Werdmüller_Otto, 3876_Werdmüller_Ulrich, 3877_Westerburg_Gerhard, 3878_Widenhuber_Hans, 3889_Wyttenbach_Niklaus, 3891_Wäber_Johannes, 3927_Welser_Hans, 3938_Wick_Johann_Jakob, 3963_Wimpfer_Georg, 3964_Winzürn_Johannes, 3988_Wolf_Johannes, 4021_Wagner_Johannes, 4023_Wetter_Wolfgang, 4025_Wirz_Melchior, 4027_Wirz_Ulrich, 4030_Wähinger_Hans, 4033_Zehnder_Johannes, 4034_Zipperli_Hans, 4079_Zwick_Johannes, 4107_Zili_Dominik, 4112_thom_Camph_Gerhard, 4710_Enzinas_Francisco_de, TRAINING_VALIDATION_SET_2022-05-19__Bullinger_comb_M1, TRAINING_VALIDATION_SET_Blarer_Ambriosius__Proj_Bullinger, TRAINING_VALIDATION_SET_Blasius_Johannes__Proj_Bullinger, TRAINING_VALIDATION_SET_Borrhaus_Martin__Proj_Bullinger, TRAINING_VALIDATION_SET_Brunner_Fridolin__Proj_Bullinger, TRAINING_VALIDATION_SET_Bucer_Martin__Proj_Bullinger, TRAINING_VALIDATION_SET_Bullinger_Heinrich__Proj_Bullinger, TRAINING_VALIDATION_SET_Bullinger_Heinrich__Proj_Bullinger--mitBasemodel, TRAINING_VALIDATION_SET_Curione_Celio_Secondo__Proj_Bullinger, TRAINING_VALIDATION_SET_Grynäus_Simon__Proj_Bullinger, TRAINING_VALIDATION_SET_Gwalther_Rudolf__Proj_Bullinger, TRAINING_VALIDATION_SET_Haller_Berchtold__Project-Bullinger, TRAINING_VALIDATION_SET_Haller_Johannes__Proj_Bullinger, TRAINING_VALIDATION_SET_Hilles_Richard__Proj_Bullinger, TRAINING_VALIDATION_SET_Hochholzer_Christian__Proj_Bullinger, TRAINING_VALIDATION_SET_Karlstadt_Andreas__Proj_Bullinger, TRAINING_VALIDATION_SET_Megander_Kaspar__Proj_Bullinger, TRAINING_VALIDATION_SET_Myconius_Oswald__Proj_Bullinger, TRAINING_VALIDATION_SET_Schuler_Gervasius__Proj_Bullinger, TRAINING_VALIDATION_SET_Vadian__Proj_Bullinger, TRAINING_VALIDATION_SET_Vogt_Simprecht__Project-Bullinger, TRAINING_VALIDATION_SET_Zwick_Johann__Proj_Bullinger, TRAINING_VALIDATION_SET_von_Rümlang__Proj_Bullinger
- Evaluation: a seeded page-level split of the training projects (partition=0.9, seed=42)
dh-unibe/image-text_koenigsfelden-charters-post-1500
- Training projects: FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_01, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_02, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_03, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_04, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_05, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_06, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_07, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_08, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_09, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_10, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_11, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_12, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_13, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_14, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_15, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_16, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_17, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_18, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_19, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_20, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_21, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_22, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_23, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_24, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_25, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_26, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_27, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_28, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_29, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_30, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_31, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_32, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_33, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_34, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_35, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_36, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_37, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_38, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_39, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_40, FRAD068_03G_SAINT_PIERRE_SAINT_GILLES_032_41, Lenzburg_0077, Lenzburg_0104a, Lenzburg_0104b, TRAINING_VALIDATION_SET_KF_Complete-charter-cartul, TRAINING_VALIDATION_SET_KF_Complete-charter-cartul__ABBREV, TRAINING_VALIDATION_SET_Königsfelden_M1, TRAINING_VALIDATION_SET_Königsfelden_M2, U-17_0388, u-17_0001, u-17_0002, u-17_0003, u-17_0004, u-17_0005, u-17_0006, u-17_0007-01, u-17_0007-02, u-17_0007a, u-17_0009, u-17_0010, u-17_0011, u-17_0011a, u-17_0012, u-17_0013, u-17_0014, u-17_0015, u-17_0016, u-17_0017, u-17_0018, u-17_0019, u-17_0020, u-17_0020a, u-17_0022, u-17_0023_01, u-17_0023_02, u-17_0024, u-17_0025, u-17_0026, u-17_0027, u-17_0028, u-17_0029, u-17_0030, u-17_0031, u-17_0032, u-17_0033, u-17_0034, u-17_0035, u-17_0036, u-17_0037, u-17_0038, u-17_0039, u-17_0040, u-17_0041, u-17_0042, u-17_0043, u-17_0044, u-17_0045, u-17_0046, u-17_0047, u-17_0048, u-17_0049, u-17_0050, u-17_0051, u-17_0052, u-17_0053, u-17_0054, u-17_0055_01, u-17_0055_02, u-17_0056, u-17_0057, u-17_0058, u-17_0059, u-17_0059a, u-17_0059b, u-17_0060, u-17_0061_01, u-17_0061_02, u-17_0062, u-17_0063, u-17_0063a, u-17_0064, u-17_0065, u-17_0066, u-17_0069, u-17_0070, u-17_0071, u-17_0072, u-17_0073, u-17_0074, u-17_0074c, u-17_0075, u-17_0076, u-17_0077, u-17_0078, u-17_0079, u-17_0080, u-17_0081, u-17_0082, u-17_0083, u-17_0084, u-17_0085, u-17_0086, u-17_0087, u-17_0088, u-17_0089, u-17_0090, u-17_0091, u-17_0092, u-17_0093, u-17_0094, u-17_0094b, u-17_0095, u-17_0096, u-17_0097, u-17_0098, u-17_0099, u-17_0100, u-17_0101, u-17_0102, u-17_0103, u-17_0104, u-17_0105, u-17_0106, u-17_0107, u-17_0108, u-17_0109, u-17_0110, u-17_0111, u-17_0112, u-17_0113, u-17_0114, u-17_0115, u-17_0116_01, u-17_0116_02, u-17_0117, u-17_0118, u-17_0119, u-17_0120, u-17_0121, u-17_0122, u-17_0123, u-17_0123a, u-17_0123b, u-17_0124, u-17_0125, u-17_0126, u-17_0127, u-17_0128, u-17_0129, u-17_0130, u-17_0130a, u-17_0130d, u-17_0131, u-17_0132, u-17_0133, u-17_0134, u-17_0136, u-17_0137, u-17_0138, u-17_0139, u-17_0140, u-17_0141, u-17_0142, u-17_0143, u-17_0144, u-17_0145, u-17_0146, u-17_0149, u-17_0150, u-17_0151, u-17_0152, u-17_0153, u-17_0154, u-17_0155, u-17_0156, u-17_0157, u-17_0158, u-17_0159, u-17_0160, u-17_0161, u-17_0162, u-17_0163, u-17_0164, u-17_0165, u-17_0166, u-17_0167, u-17_0168, u-17_0169, u-17_0170, u-17_0171, u-17_0172, u-17_0173, u-17_0173a, u-17_0174, u-17_0175, u-17_0176, u-17_0177, u-17_0179, u-17_0180, u-17_0181_01, u-17_0181_02, u-17_0182, u-17_0183, u-17_0184, u-17_0185, u-17_0185a, u-17_0186, u-17_0187, u-17_0187a, u-17_0188, u-17_0189, u-17_0190, u-17_0191, u-17_0192, u-17_0193, u-17_0194, u-17_0195, u-17_0196, u-17_0197, u-17_0198, u-17_0199, u-17_0200, u-17_0201_01, u-17_0201_02, u-17_0202, u-17_0203, u-17_0204, u-17_0205, u-17_0205a, u-17_0206, u-17_0207, u-17_0208, u-17_0209, u-17_0210, u-17_0211, u-17_0212, u-17_0213, u-17_0214, u-17_0215, u-17_0216, u-17_0217, u-17_0218, u-17_0219, u-17_0220, u-17_0221, u-17_0222, u-17_0223, u-17_0224, u-17_0225, u-17_0226, u-17_0227, u-17_0228, u-17_0229, u-17_0230, u-17_0231, u-17_0232, u-17_0233, u-17_0234, u-17_0235, u-17_0236, u-17_0237, u-17_0238, u-17_0239, u-17_0240, u-17_0241, u-17_0242, u-17_0244, u-17_0246, u-17_0247, u-17_0248, u-17_0249, u-17_0250, u-17_0251, u-17_0252, u-17_0253, u-17_0254, u-17_0254a, u-17_0255, u-17_0256, u-17_0257, u-17_0258, u-17_0259, u-17_0260, u-17_0261, u-17_0262, u-17_0263, u-17_0264, u-17_0265, u-17_0266_01, u-17_0266_02, u-17_0266a, u-17_0267, u-17_0268, u-17_0269, u-17_0270, u-17_0271, u-17_0272, u-17_0273, u-17_0274, u-17_0275, u-17_0276, u-17_0276a, u-17_0277_01, u-17_0277_02, u-17_0278, u-17_0279, u-17_0280, u-17_0281, u-17_0282, u-17_0283, u-17_0284, u-17_0285, u-17_0286, u-17_0287, u-17_0288, u-17_0289, u-17_0290, u-17_0291, u-17_0292, u-17_0293, u-17_0294, u-17_0295, u-17_0296, u-17_0297, u-17_0298, u-17_0299, u-17_0300, u-17_0301, u-17_0302, u-17_0303, u-17_0304, u-17_0305, u-17_0306, u-17_0306a, u-17_0307, u-17_0307a, u-17_0308_01, u-17_0308_02, u-17_0309, u-17_0310, u-17_0311, u-17_0312, u-17_0312a, u-17_0313, u-17_0314, u-17_0315, u-17_0316, u-17_0316a, u-17_0317, u-17_0318, u-17_0319, u-17_0320, u-17_0321, u-17_0322, u-17_0323, u-17_0324, u-17_0325, u-17_0326, u-17_0327, u-17_0328, u-17_0329, u-17_0330, u-17_0331, u-17_0332, u-17_0332a, u-17_0333, u-17_0334, u-17_0335, u-17_0336, u-17_0337, u-17_0338, u-17_0339, u-17_0340, u-17_0341, u-17_0342, u-17_0343, u-17_0344, u-17_0345, u-17_0346, u-17_0347, u-17_0348, u-17_0349, u-17_0350, u-17_0351, u-17_0352, u-17_0353, u-17_0354a, u-17_0355, u-17_0356, u-17_0357, u-17_0358, u-17_0359, u-17_0360, u-17_0361, u-17_0362, u-17_0363, u-17_0364, u-17_0365, u-17_0366, u-17_0367, u-17_0368, u-17_0369, u-17_0370, u-17_0371, u-17_0372, u-17_0373, u-17_0374, u-17_0375, u-17_0376, u-17_0377, u-17_0378, u-17_0379, u-17_0380, u-17_0382, u-17_0383, u-17_0385, u-17_0386, u-17_0387, u-17_0389, u-17_0390, u-17_0391, u-17_0392, u-17_0393, u-17_0394, u-17_0395, u-17_0396, u-17_0397, u-17_0398, u-17_0399, u-17_0400, u-17_0401, u-17_0402, u-17_0403, u-17_0404, u-17_0405, u-17_0406, u-17_0407, u-17_0408, u-17_0409, u-17_0410, u-17_0411, u-17_0412, u-17_0413, u-17_0414, u-17_0415, u-17_0416, u-17_0417, u-17_0418, u-17_0419, u-17_0420, u-17_0421, u-17_0422, u-17_0423, u-17_0424, u-17_0425, u-17_0426, u-17_0427, u-17_0428, u-17_0429, u-17_0430, u-17_0431, u-17_0432, u-17_0433, u-17_0434, u-17_0435, u-17_0436_01, u-17_0436_02, u-17_0436a, u-17_0437, u-17_0437a, u-17_0438, u-17_0439, u-17_0440, u-17_0441, u-17_0442, u-17_0443, u-17_0444, u-17_0445, u-17_0446, u-17_0447, u-17_0448, u-17_0449, u-17_0450, u-17_0451, u-17_0452, u-17_0453, u-17_0454, u-17_0456, u-17_0457, u-17_0458, u-17_0459, u-17_0460, u-17_0461, u-17_0462, u-17_0463, u-17_0464, u-17_0465, u-17_0466, u-17_0467, u-17_0468, u-17_0469, u-17_0470, u-17_0471, u-17_0472, u-17_0473, u-17_0474, u-17_0475, u-17_0476, u-17_0477, u-17_0478, u-17_0479, u-17_0480, u-17_0481, u-17_0482, u-17_0483, u-17_0484, u-17_0485, u-17_0486, u-17_0487, u-17_0488, u-17_0489, u-17_0490, u-17_0491, u-17_0492, u-17_0493, u-17_0494, u-17_0495, u-17_0496, u-17_0497, u-17_0498, u-17_0499, u-17_0500, u-17_0501, u-17_0502, u-17_0503, u-17_0504, u-17_0505, u-17_0506, u-17_0507, u-17_0508, u-17_0509, u-17_0510, u-17_0511, u-17_0512, u-17_0513, u-17_0514, u-17_0516, u-17_0517, u-17_0518, u-17_0519, u-17_0520, u-17_0521, u-17_0522, u-17_0523, u-17_0524, u-17_0525, u-17_0526, u-17_0527, u-17_0528, u-17_0529, u-17_0530, u-17_0531, u-17_0532, u-17_0533, u-17_0534, u-17_0535, u-17_0536, u-17_0537, u-17_0538, u-17_0539, u-17_0540, u-17_0541, u-17_0542, u-17_0543, u-17_0544, u-17_0545, u-17_0546, u-17_0546a, u-17_0547, u-17_0548, u-17_0549, u-17_0550, u-17_0551, u-17_0552, u-17_0553_01, u-17_0553_02, u-17_0554, u-17_0555, u-17_0556, u-17_0557, u-17_0558, u-17_0559, u-17_0560, u-17_0561, u-17_0562, u-17_0563, u-17_0564, u-17_0565, u-17_0566, u-17_0567, u-17_0568, u-17_0569, u-17_0570, u-17_0571, u-17_0572, u-17_0573, u-17_0574, u-17_0575, u-17_0576, u-17_0577, u-17_0578, u-17_0579, u-17_0581_01, u-17_0581_02, u-17_0581_03, u-17_0582, u-17_0583, u-17_0584, u-17_0585, u-17_0586, u-17_0587, u-17_0588, u-17_0589, u-17_0590, u-17_0591, u-17_0592, u-17_0593, u-17_0594, u-17_0595, u-17_0596, u-17_0597, u-17_0598, u-17_0599, u-17_0600, u-17_0601, u-17_0602, u-17_0604, u-17_0605, u-17_0606, u-17_0607, u-17_0608, u-17_0609, u-17_0610, u-17_0611, u-17_0612, u-17_0613, u-17_0614, u-17_0615, u-17_0616, u-17_0617, u-17_0618_01, u-17_0618_02, u-17_0619, u-17_0620, u-17_0621, u-17_0622, u-17_0623, u-17_0625, u-17_0626, u-17_0627, u-17_0628, u-17_0630, u-17_0631, u-17_0632, u-17_0633, u-17_0634, u-17_0635, u-17_0636, u-17_0637, u-17_0638, u-17_0639, u-17_0640, u-17_0641, u-17_0642, u-17_0643, u-17_0644, u-17_0645, u-17_0646, u-17_0647_02, u-17_0648, u-17_0649_01, u-17_0649_02, u-17_0650, u-17_0651, u-17_0653, u-17_0654, u-17_0655, u-17_0656, u-17_0657, u-17_0658, u-17_0659, u-17_0660, u-17_0661, u-17_0662, u-17_0663, u-17_0664, u-17_0665, u-17_0666, u-17_0667, u-17_0668_01, u-17_0668_02, u-17_0669_01, u-17_0669_02, u-17_0670, u-17_0671, u-17_0672, u-17_0673, u-17_0674, u-17_0675, u-17_0676, u-17_0677, u-17_0678_01, u-17_0678_02, u-17_0679, u-17_0680, u-17_0681, u-17_0682, u-17_0683, u-17_0684, u-17_0685, u-17_0686, u-17_0687, u-17_0688, u-17_0689, u-17_0690, u-17_0691, u-17_0692, u-17_0693, u-17_0694, u-17_0695, u-17_0696, u-17_0697_01, u-17_0697_02, u-17_0698, u-17_0699, u-17_0700, u-17_0701_01, u-17_0701_02, u-17_0701_03, u-17_0701_04, u-17_0702_01, u-17_0702_02, u-17_0703, u-17_0704, u-17_0705, u-17_0706, u-17_0707, u-17_0708, u-17_0709_01, u-17_0709_02, u-17_0710, u-17_0711, u-17_0712, u-17_0713, u-17_0714, u-17_0715, u-17_0716, u-17_0717, u-17_0718_01, u-17_0718_02, u-17_0719, u-17_0720, u-17_0722, u-17_0723, u-17_0724, u-17_0725, u-17_0726, u-17_0727_02, u-17_0728, u-17_0729, u-17_0730, u-17_0731, u-17_0732, u-17_0733, u-17_0734, u-17_0735, u-17_0736, u-17_0737, u-17_0738, u-17_0739, u-17_0740, u-17_0741, u-17_0742, u-17_0743, u-17_0744, u-17_0745, u-17_0746_01, u-17_0746_02, u-17_0747, u-17_0748_02, u-17_0748_04, u-17_0749, u-17_0750, u-17_0751, u-17_0752, u-17_0753, u-17_0754, u-17_0755, u-17_0756, u-17_0757, u-17_0758, u-17_0759, u-17_0760, u-17_0761, u-17_0762, u-17_0763, u-17_0764, u-17_0765, u-17_0766, u-17_0767, u-17_0768, u-17_0769, u-17_0770, u-17_0771, u-17_0772, u-17_0773, u-17_0774, u-17_0775, u-17_0776, u-17_0777, u-17_0778, u-17_0779, u-17_0780, u-17_0781, u-17_0782, u-17_0783, u-17_0784, u-17_0785, u-17_0786, u-17_0787_01, u-17_0787_02, u-17_0788, u-17_0789, u-17_0790, u-17_0791_01, u-17_0791_02, u-17_0792, u-17_0793, u-17_0794, u-17_0795, u-17_0796, u-17_0797_01, u-17_0797_02, u-17_0798, u-17_0799, u-17_0800, u-17_0801, u-17_0802, u-17_0803, u-17_0804, u-17_0805, u-17_0806, u-17_0807, u-17_0808, u-17_0809, u-17_0810, u-17_0811, u-17_0812, u-17_0813, u-17_0814, u-17_0815, u-17_0816, u-17_0817, u-17_0818, u-17_0819, u-17_0820, u-17_0821, u-17_0822, u-17_0824, u-17_0825_01, u-17_0825_02, u-17_0826, u-17_0827, u-17_0828_01, u-17_0828_02, u-17_0829, u-17_0830, u-17_0831, u-17_0832, u-17_0833, u-17_0834, u-17_0835, u-17_0836, u-17_0837, u-17_0838, u-17_0839, u-17_0840, u-17_0841, u-17_0842, u-17_0843, u-17_0844_01, u-17_0844_02, u-17_0846, u-17_0847, u-17_0848, u-17_0849, u-17_0850, u-17_0851_02, u-17_0852, u-17_0853a, u-17_0853b, u-17_0854, u-17_0855, u-17_0856, u-17_0857, u-17_0858, u-17_0859, u-17_0860, u-17_0861, u-17_0862, u-17_0863, u-17_0864, u-17_0865, u-17_0866, u-17_0867, u-17_0869, u-17_0870, u-17_0872, u-17_0873, u-17_0874, u-17_0875, u-17_0876, u-17_0876a, u-17_0877, u-17_0878, u-17_0879, u-17_0880, u-17_0880a, u-17_0881, u-17_0882, u-17_0883, u-17_0884, u-17_0885, u-17_0886, u-17_0887a, u-17_0887b, u-17_0887c, u-17_0888, u-17_0889, u-17_0890_01, u-17_0890_02, u-17_0891, u-17_0892a, u-17_0894, u-17_0895, u-17_0896, u-17_0897, u-17_0898, u-17_0898a, u-17_0899, u-17_0900, u-17_0901_01, u-17_0901_02, u-17_0901_03, u-17_0902, u-17_0903, u-17_0904, u-17_0905, u-17_0906, u-17_0907, u-17_0908, u-17_0909, u-17_0909a, u-17_0910, u-17_0911, u-17_0912, u-17_0913, u-17_0914, u-17_0915, u-17_0916, u-17_0917, u-17_0918, u-17_0919, u-17_0920, u-17_0921, u-17_0922, u-17_0923, u-17_0924, u-17_0924a, u-17_0924b, u-17_0924c, u-17_0924d, u-17_0924e, u-17_0924f, u-17_0924g, u-17_0924h, u-17_0924i, u-17_0924k, u-17_0925, u-17_0926, u-17_0927, u-17_0927a, u-17_0927b, u-17_0927c, u-17_0927d, u-17_0927e, u-17_0927f, u-17_0927g, u-17_0927h, u-17_0927i, u-17_0927k, u-17_0927l, u-17_0928, u-17_0928a, u-17_0928b, u-17_0929, u-17_0929a, u-17_0929b, u-17_0930, u-17_0930a, u-17_0930b, u-17_0931, u-17_0932, u-17_0933, u-17_0934, u-17_0934a, u-17_0934b, u-17_0935, u-17_0935a, u-17_0935b, u-17_0935c, u-17_0935d, u-17_0935e, u-17_0936, u-17_0936a, u-17_0936b, u-17_0936c, u-17_0936d, u-17_0936e, u-17_0937, u-17_0937a, u-17_0937b, u-17_0937c, u-17_0937d, u-17_0937e, u-17_0937f, u-17_0938, u-17_0938a, u-17_0938b, u-17_0938c, u-17_0938d, u-17_0939, u-17_0939a, u-17_0940, u-17_0941, u-17_0941a, u-17_0942, u-17_0943, u-17_0944, u-17_0945, u-17_0946, u-17_0947, u-17_0948, u-17_0949, u-17_0950, u-17_0951, u-17_0952, u-17_0953, u-17_0954, u-17_0955, u-17_0956, u-17_0957, u-17_0957a, u-17_0958, u-17_0959, u-17_0960, u-17_0961, u-17_0962, u-17_0963_01, u-17_0963_02, u-17_0964, u-17_0965, u-17_0966, u-17_0966a, u-17_0967, u-17_0968, u-17_0968a, u-17_0968b, u-17_0969, u-17_0970, u-17_0970a, u-17_0970b, u-17_0970c, u-17_0970d, u-17_0971, u-17_0972, u-17_0973, u-17_0973a, u-17_0974, u-17_0975, u-17_0976, u-17_0977, u-17_0978, u-17_0979, u-17_0980a, u-17_0980b, u-17_0980c, u-17_0980d, u-17_0981, u-17_0982, u-17_0984, u-17_0985, u-17_0986, u-17_0986a, u-17_0987, u-17_0988, u-17_0989, u-17_0990, u-17_0991, u-17_0992, u-17_0993, u-17_0994, u-17_0995, u-17_0996, u-17_0997, u-17_0998a, u-17_0998b, u-17_0998c, u-17_0998d, u-17_0998e, u-17_0998f, u-17_0999a, u-17_1000, u-17_1000a, u-17_1000b, u-17_1001, u-17_1002, u-17_1003, u-17_1004, u-17_1005, u-17_1006, u-17_1007, u-17_1008_01, u-17_1008_02, u-17_1008_03, u-17_1009, u-17_1010_01, u-17_1010_02, u-17_1011, u-17_1012, u-17_1013a, u-17_1013b, u-17_1014, u-17_1015, u-17_1016
- Evaluation: a seeded page-level split of the training projects (partition=0.9, seed=42)
dh-unibe/image-text_aaeb-xiv-xvii
- Training projects: B_168_14-11_1, B_168_14-11_2, B_168_14-11_3, B_168_14-12, B_168_14-14_1, B_168_14-15_1, B_168_14-15_2, B_168_14-16, B_168_14-17_1, B_168_14-17_2, B_168_14-17_3, B_168_14-17_4, B_168_14-17_5, B_168_14-17_6, B_168_14-18, B_168_14-19, B_168_14-20_1, B_168_14-20_2, B_168_14-20_3, B_168_14-21_1, B_168_14-21_2, B_168_14-21_3, B_168_14-21_4, B_168_14-21_5, B_168_14-22_1, B_168_14-22_2, B_168_14-22_3, B_168_14-22_4, B_168_14-22_5, B_168_14-22_6, B_168_14-23_1, B_168_14-23_2, B_168_14-23_3, B_168_14-23_4, B_168_14-24_1, B_168_14-24_2, B_168_14-24_3, B_168_14-24_4, B_168_14-24_5, B_168_14-24_6, B_168_14-25_1, B_168_14-25_2, B_168_14-26_1, B_168_14-26_2, B_168_14-26_3, B_168_14-26_4, B_168_14-26_5, B_168_14-26_6, B_168_14-26_7, B_168_14-26_8, B_168_14-27_1, B_168_14-27_2, B_168_14-27_3, B_168_14-27_4, B_168_14-28_1, B_168_14-28_2, B_168_14-28_3, B_168_14-29_1, B_168_14-29_2, B_168_14-29_3, B_168_14-29_5, B_168_14-29_6, B_168_14-29_8, B_168_14-30_1, B_168_14-30_2, B_168_14-30_3, B_168_14-31_1, B_168_14-31_2, B_168_14-31_3, B_168_14-31_4, B_168_14-32_1, B_168_14-32_2, B_168_14-32_3, B_168_14-32_4, B_168_14-33_1, B_168_14-33_2, B_168_14-33_3, B_168_14-34, B_168_14-35_1, B_168_14-35_2, B_168_14-35_3, B_168_14-35_4, B_168_14-36_1, B_168_14-36_2, B_168_14-36_3, B_168_14-36_4, B_168_14-36_5, B_168_14-37_1, B_168_14-37_2, B_168_14-39_1, B_168_14-40_1, B_168_14-40_2, B_168_14-42, B_168_14-5, B_168_14-6, B_168_14-7_1, B_168_14-7_2, B_168_14-8, B_168_14-9, B_168_15-10_1, B_168_15-10_2, B_168_15-10_3, B_168_15-11_1, B_168_15-11_2, B_168_15-11_3, B_168_15-11_4, B_168_15-12_1, B_168_15-12_2, B_168_15-12_3, B_168_15-12_5, B_168_15-12_6, B_168_15-12_8, B_168_15-17_1, B_168_15-20_2, B_168_15-20_3, B_168_15-20_4, B_168_15-20_5, B_168_15-20_6, B_168_15-23_1, B_168_15-23_2, B_168_15-23_3, B_168_15-23_4, B_168_15-23_5, B_168_15-24_2, B_168_15-27_3, B_168_15-27_4, B_168_15-27_6, B_168_15-27_7, B_168_15-2_1, B_168_15-2_2, B_168_15-2_3, B_168_15-2_4, B_168_15-30_3, B_168_15-30_4, B_168_15-30_6, B_168_15-30_7, B_168_15-30_8, B_168_15-33, B_168_15-5_1, B_168_15-5_2, B_168_15-5_3, B_168_15-5_4, B_168_15-5_5, B_168_15-5_6, B_168_15-6_1, B_168_15-6_3, B_168_15-7, B_168_15-8_16, B_168_15-8_17, B_168_15-8_2, B_168_15-8_3, B_168_15-9_2, B_168_15-9_4, B_168_16-1_1, B_168_16-1_2, B_168_16-33_2, B_168_16-33_3, B_168_16-35_3, B_168_16-35_4, B_168_16-35_5, B_168_16-36_5, B_168_16-36_6, B_168_16-37_3, B_168_16-38_2, B_168_16-38_5, B_168_16-3_2, B_168_16-3_4, B_168_16-3_6, B_168_16-3_7, B_168_16-40_2, B_168_16-40_3, B_168_16-40_5, B_168_16-40_6, B_168_16-40_7, B_168_16-40_8, B_168_16-4_1, B_168_16-4_2, B_168_16-5_1, B_168_16-5_2, B_168_16-5_3, B_168_16-6_1, B_168_16-6_2, B_168_16-9_12, B_168_16-9_3, B_168_16-9_6, B_168_16-9_7, B_168_16-9_8, B_168_17-14_4, B_168_17-14_5, B_168_17-15_5, B_168_17-15_6, B_168_17-15_7, B_168_17-16_3, B_168_17-16_4, B_168_17-1_2, B_168_17-1_3, B_168_17-1_4, B_168_17-1_5, B_168_17-1_6, B_168_17-21_2, B_168_17-21_4, B_168_17-22_2, B_168_17-22_3, B_168_17-2_1, B_168_17-2_2, B_168_17-2_4, B_168_17-30_1, B_168_17-3_3, B_168_17-45_2, B_168_17-48_2, B_168_17-48_3, B_168_17-48_4, B_168_17-4_2, B_168_17-6_5, B_168_17-6_6, B_168_17-8_2, B_168_17-8_3, B_168_17-8_4, B_168_17-8_5, B_168_17-8_6, B_168_17-8_7, B_168_18-12_6, B_168_18-13_2, B_168_18-15_1, B_168_18-15_4, B_168_18-1_4, B_168_18-22_2, B_168_18-24_10, B_168_18-5_1, B_168_18-8_4, B_168_18-9_1, B_168_18-9_2, B_168_18-9_3, B_168_19-1, B_168_19-20_2, B_168_19-20_4, B_168_19-20_5, B_168_19-21_1, B_168_19-24_5, B_168_19-28_1, B_168_19-28_10, B_168_19-28_15, B_168_19-28_16, B_168_19-28_2, B_168_19-28_3, B_168_19-28_33, B_168_19-29_1, B_168_19-29_2, B_168_19-29_3, B_168_19-29_4, B_168_19-29_5, B_168_19-29_6, B_168_19-2_1, B_168_19-2_2, B_168_19-30_1, B_168_19-30_15, B_168_19-30_16, B_168_19-30_6, B_168_19-35_1, B_168_19-35_10, B_168_19-35_11, B_168_19-35_12, B_168_19-35_13, B_168_19-35_14, B_168_19-35_15, B_168_19-35_16, B_168_19-35_17, B_168_19-35_18, B_168_19-35_2, B_168_19-35_3, B_168_19-35_4, B_168_19-35_5, B_168_19-35_6, B_168_19-35_7, B_168_19-35_8, B_168_19-35_9, B_168_19-3_1, B_168_19-3_2, PCrim_Dt_(ville)_6-1, PCrim_Dt_(ville)_6-10, PCrim_Dt_(ville)_6-7, PCrim_Dt_(ville)_6-9, PCrim_E_10-1, PCrim_E_10-2, PCrim_E_12-1, PCrim_E_12-2, PCrim_E_128-5, PCrim_E_14-1, PCrim_E_14-2, PCrim_E_142-3, PCrim_E_155-1, PCrim_E_156-1, PCrim_E_156-3, PCrim_E_156-7, PCrim_E_163-3, PCrim_E_30-1, PCrim_E_30-2, PCrim_E_300-17, PCrim_E_300-18, PCrim_E_332-10, PCrim_E_332-11, PCrim_E_332-12, PCrim_E_332-13, PCrim_E_332-14, PCrim_E_332-16, PCrim_E_332-18, PCrim_E_332-19, PCrim_E_332-2, PCrim_E_332-20, PCrim_E_332-22, PCrim_E_332-24, PCrim_E_332-28, PCrim_E_332-29, PCrim_E_332-3, PCrim_E_332-4, PCrim_E_332-5, PCrim_E_332-6, PCrim_E_332-7, PCrim_E_332-9, PCrim_E_399, PCrim_E_428-36, PCrim_E_55-1, PCrim_E_55-2, PCrim_E_6, PCrim_E_69-2, PCrim_E_7, PCrim_E_70-1, PCrim_E_70-2, PCrim_E_76-1, PCrim_E_76-2, PCrim_E_78-1, PCrim_E_78-11, PCrim_E_78-2, PCrim_E_85-1, PCrim_E_85-2, PCrim_LN_4, PCrim_LN_6-2, PCrim_LZ_24-2, PCrim_LZ_25-19, PCrim_SU_21, TRAINING_VALIDATION_SET_AAEB_FFC_V3, TRAINING_VALIDATION_SET_AAEB_FFC_v0, TRAINING_VALIDATION_SET_AAEB_FFC_v0_1, TRAINING_VALIDATION_SET_AAEB_FFC_v0_3, TRAINING_VALIDATION_SET_AAEB_FFC_v1, TRAINING_VALIDATION_SET_AAEB_FFC_v2, TRAINING_VALIDATION_SET_AAEB_FFC_v3_1, TRAINING_VALIDATION_SET_AAEB_mixte_v4, TRAINING_VALIDATION_SET_FFC_v0_2
- Evaluation: a seeded page-level split of the training projects (partition=0.9, seed=42)

Materialized from that selection: 12,286 pages, 325,454 transcribed lines.

Hyperparameters

spec: '[256,64,0,1 Cr4,2,8,4,2 Cr4,2,32,1,1 Mp4,2,4,2 Cr3,3,64,1,1 Mp1,2,1,2 S1(1x0)1,3
  Lbx256 Do0.5 Lbx256 Do0.5 Lbx256 Do0.5 Cr255,1,85,1,1]'
batch_size: 8
schedule: 1cycle
lrate: 0.0001
quit: early
epochs: 30
min_epochs: 10
lag: 8
augment: false
normalization: NFD
normalize_whitespace: true
weights_format: coreml
resize: union
freeze_backbone: null
accumulate_grad_batches: 32
pad: null
warmup: null
seed: 42
workers: 8
device: cuda:0

Provenance


engine	`kraken`
base model	CATMuS Medieval (`10.5281/zenodo.15030337`, requested as `kraken-early_modern_german`)
training job	`20260825T145122Z-kraken-medieval-german-v2`
trained	2026-08-30T19:34:23.949417+00:00
weights	`kraken-medieval-german-v2.mlmodel`

metadata.json in this repo is the record the trainer wrote, verbatim: the full request, the parsed metrics and the job id.

Using it

pip install kraken
python -c "from huggingface_hub import hf_hub_download; print(hf_hub_download('dh-unibe/kraken-medieval-german-v2', 'kraken-medieval-german-v2.mlmodel'))"
kraken -i page.jpg page.txt segment -bl ocr -m <the printed path>

Как запустить

На своей GPU в ailist. Цена — по текущим предложениям провайдеров

GPU в ailist

Универсальный рантайм · kraken

Нельзя запустить

В репозитории нет весов в формате PyTorch/safetensors.

Ревизия весов: b904f78