[ { "name": "Audio and Speech", "tables": [ { "subsection": "Hebrew-Specific Datasets", "columns": [ "Name", "Description", "Details", "Links" ], "rows": [ [ "[hebrew-speech-dataset](https://huggingface.co/datasets/shunyalabs/hebrew-speech-dataset)", "Hebrew speech with transcripts, 16 kHz.", "Author: shunyalabs · Size: 1K – 10K rows", "[Dataset](https://huggingface.co/datasets/shunyalabs/hebrew-speech-dataset)" ], [ "[Hebrew-talk](https://huggingface.co/datasets/sleeping-ai/Hebrew-talk)", "Hebrew speech with transcripts (audio and text columns).", "Author: sleeping-ai · Size: 10K – 100K rows · License: mit", "[Dataset](https://huggingface.co/datasets/sleeping-ai/Hebrew-talk)" ], [ "[hebrew_medical_audio](https://huggingface.co/datasets/verbit/hebrew_medical_audio)", "Hebrew medical-domain audio (gated).", "Author: verbit · Size: 1K – 10K rows · License: cc-by-nc-4.0", "[Dataset](https://huggingface.co/datasets/verbit/hebrew_medical_audio)" ], [ "[hebrew_speech_campus](https://huggingface.co/datasets/imvladikon/hebrew_speech_campus)", "Hebrew speech from Campus IL video lectures, with subtitles aligned to the audio.", "Author: imvladikon · Size: 10K – 100K rows", "[Dataset](https://huggingface.co/datasets/imvladikon/hebrew_speech_campus)" ], [ "[hebrew_speech_coursera](https://huggingface.co/datasets/imvladikon/hebrew_speech_coursera)", "Hebrew speech recognition dataset from Coursera course audio.", "Author: imvladikon · Size: 10K – 100K rows", "[Dataset](https://huggingface.co/datasets/imvladikon/hebrew_speech_coursera)" ], [ "[hebrew_speech_kan](https://huggingface.co/datasets/imvladikon/hebrew_speech_kan)", "Hebrew speech for ASR (Kan broadcast audio with transcripts).", "Author: imvladikon · Size: 10K – 100K rows", "[Dataset](https://huggingface.co/datasets/imvladikon/hebrew_speech_kan)" ], [ "[hebrew_speech_kan_nikud](https://huggingface.co/datasets/micsell/hebrew_speech_kan_nikud)", "Kan Hebrew speech with vocalised (nikud) transcripts.", "Author: micsell · Size: 10K – 100K rows", "[Dataset](https://huggingface.co/datasets/micsell/hebrew_speech_kan_nikud)" ], [ "[saspeech](https://huggingface.co/datasets/upai-inc/saspeech)", "Single-speaker studio Hebrew speech (Shaul Amsterdamski), gold-standard subset of the OpenSLR 134 corpus; for TTS.", "Author: upai-inc · Size: ~4 hours · License: other", "[Dataset](https://huggingface.co/datasets/upai-inc/saspeech)" ] ] }, { "subsection": "Hebrish / danielrosehill", "columns": [ "Name", "Description", "Details", "Links" ], "rows": [ [ "[English-Hebrew-Codeswitching-ASR-Benchmark](https://huggingface.co/datasets/Hebrish/English-Hebrew-Codeswitching-ASR-Benchmark)", "Benchmark test set for English-Hebrew code-switching ASR.", "Author: Hebrish · Size: <1K rows · License: cc-by-4.0", "[Dataset](https://huggingface.co/datasets/Hebrish/English-Hebrew-Codeswitching-ASR-Benchmark)" ], [ "[English-Hebrew-Codeswitching-Sentences](https://huggingface.co/datasets/danielrosehill/English-Hebrew-Codeswitching-Sentences)", "English-Hebrew code-switching sentences read aloud, for ASR training.", "Author: danielrosehill · Size: 10K – 100K rows · License: cc-by-4.0", "[Dataset](https://huggingface.co/datasets/danielrosehill/English-Hebrew-Codeswitching-Sentences)" ], [ "[English-Hebrew-Mixed-Sentences](https://huggingface.co/datasets/Hebrish/English-Hebrew-Mixed-Sentences)", "\"Hebrish\": English sentences mixed with Hebrew, audio dataset.", "Author: Hebrish · Size: <1K rows · License: mit", "[Dataset](https://huggingface.co/datasets/Hebrish/English-Hebrew-Mixed-Sentences)" ], [ "[Hebrish-TTS-Voice-Recordings](https://huggingface.co/datasets/Hebrish/Hebrish-TTS-Voice-Recordings)", "English-Hebrew voice recordings for TTS.", "Author: Hebrish · Size: 10K – 100K rows · License: cc-by-nc-4.0", "[Dataset](https://huggingface.co/datasets/Hebrish/Hebrish-TTS-Voice-Recordings)" ], [ "[Hebrish-Validation](https://huggingface.co/datasets/Hebrish/Hebrish-Validation)", "Validation benchmark for English-Hebrew code-switching ASR: roughly one-minute reads of a fixed script.", "Author: Hebrish · Size: <1K rows · License: cc-by-4.0", "[Dataset](https://huggingface.co/datasets/Hebrish/Hebrish-Validation)" ], [ "[Israeli-Name-Sentences](https://huggingface.co/datasets/Hebrish/Israeli-Name-Sentences)", "Sentences with Israeli names, English-Hebrew code-switching, for ASR.", "Author: Hebrish · Size: <1K rows · License: cc-by-4.0", "[Dataset](https://huggingface.co/datasets/Hebrish/Israeli-Name-Sentences)" ], [ "[Jewish-Religious-Terminology-Sentences](https://huggingface.co/datasets/Hebrish/Jewish-Religious-Terminology-Sentences)", "Sentences with Jewish religious terminology, English-Hebrew code-switching, for ASR.", "Author: Hebrish · Size: 1K – 10K rows · License: cc-by-4.0", "[Dataset](https://huggingface.co/datasets/Hebrish/Jewish-Religious-Terminology-Sentences)" ], [ "[Judaica-Sentences](https://huggingface.co/datasets/Hebrish/Judaica-Sentences)", "English sentences from traditional Jewish works with Hebrew terms read in; code-switching ASR.", "Author: Hebrish · Size: 1K – 10K rows · License: cc-by-4.0", "[Dataset](https://huggingface.co/datasets/Hebrish/Judaica-Sentences)" ], [ "[Noisy-Condition-Sentences](https://huggingface.co/datasets/Hebrish/Noisy-Condition-Sentences)", "Code-switching sentences recorded under noisy conditions, for ASR robustness.", "Author: Hebrish · Size: 10K – 100K rows · License: cc-by-4.0", "[Dataset](https://huggingface.co/datasets/Hebrish/Noisy-Condition-Sentences)" ], [ "[Pronunciation-Variants](https://huggingface.co/datasets/Hebrish/Pronunciation-Variants)", "Takes where a word is deliberately said one particular way (pronunciation variants), for ASR robustness.", "Author: Hebrish · Size: <1K rows · License: cc-by-4.0", "[Dataset](https://huggingface.co/datasets/Hebrish/Pronunciation-Variants)" ], [ "[Read-Aloud-Voice-Sentences](https://huggingface.co/datasets/danielrosehill/Read-Aloud-Voice-Sentences)", "Master set of read-aloud takes from the Hebrish capture app (English with Hebrew code-switching); the split-out datasets are copies of its rows.", "Author: danielrosehill · Size: 10K – 100K rows · License: cc-by-4.0", "[Dataset](https://huggingface.co/datasets/danielrosehill/Read-Aloud-Voice-Sentences)" ], [ "[Small-STT-Eval-Audio-Dataset](https://huggingface.co/datasets/danielrosehill/Small-STT-Eval-Audio-Dataset)", "Small STT evaluation set: 92 audio samples with ground truth, covering technical vocabulary and English/Hebrew code-switching.", "Author: danielrosehill · Size: 92 samples · License: cc-by-4.0", "[Dataset](https://huggingface.co/datasets/danielrosehill/Small-STT-Eval-Audio-Dataset)" ] ] }, { "subsection": "Multilingual datasets with a Hebrew portion", "columns": [ "Name", "Description", "Details", "Links" ], "rows": [ [ "[fleurs](https://huggingface.co/datasets/google/fleurs)", "FLEURS speech benchmark; Hebrew is the `he_il` config.", "Author: google · Size: 100K – 1M rows · License: cc-by-4.0", "[Dataset](https://huggingface.co/datasets/google/fleurs)" ], [ "[yodas](https://huggingface.co/datasets/espnet/yodas)", "YODAS multilingual YouTube speech; Hebrew is the `iw000` subset.", "Author: espnet · License: cc-by-3.0", "[Dataset](https://huggingface.co/datasets/espnet/yodas)" ] ] } ] }, { "name": "Instruction and Chat", "tables": [ { "subsection": null, "columns": [ "Name", "Description", "Details", "Links" ], "rows": [ [ "[hebrew_alpaca](https://huggingface.co/datasets/ashercn97/hebrew_alpaca)", "Hebrew Alpaca-style instruction data (instruction, input, output).", "Author: ashercn97 · Size: 1K – 10K rows", "[Dataset](https://huggingface.co/datasets/ashercn97/hebrew_alpaca)" ], [ "[Roleplay-Hebrew](https://huggingface.co/datasets/jojo-ai-mst/Roleplay-Hebrew)", "Hebrew roleplay instruction data, translated from the GPTeacher roleplay set.", "Author: jojo-ai-mst · Size: 1K – 10K rows · License: mit", "[Dataset](https://huggingface.co/datasets/jojo-ai-mst/Roleplay-Hebrew)" ] ] } ] }, { "name": "NLP Task Datasets", "tables": [ { "subsection": null, "columns": [ "Name", "Description", "Details", "Links" ], "rows": [ [ "[Hebrew-Paraphrase-Dataset](https://huggingface.co/datasets/HebArabNlpProject/Hebrew-Paraphrase-Dataset)", "9,785 LLM-generated paraphrase pairs (paragraph and sentence level); 300 manually validated gold.", "Author: HebArabNlpProject · Size: 9,785 instances · License: cc-by-sa-4.0", "[Dataset](https://huggingface.co/datasets/HebArabNlpProject/Hebrew-Paraphrase-Dataset)" ], [ "[hebrew_diacritized_text](https://huggingface.co/datasets/TigreGotico/hebrew_diacritized_text)", "~5.3M Hebrew sentences with full nikud, from the phonikud knesset_nikud_v6 corpus; for diacritization.", "Author: TigreGotico · Size: ~5.3M sentences · License: cc-by-4.0", "[Dataset](https://huggingface.co/datasets/TigreGotico/hebrew_diacritized_text)" ], [ "[hebrew_sentiment](https://huggingface.co/datasets/omilab/hebrew_sentiment)", "Hebrew sentiment analysis (HebrewSentiment, positive/negative).", "Author: omilab · Size: 10K – 100K rows · License: mit", "[Dataset](https://huggingface.co/datasets/omilab/hebrew_sentiment)" ], [ "[Hebrew_sentiment](https://huggingface.co/datasets/sepidmnorozy/Hebrew_sentiment)", "Hebrew sentiment classification with binary labels.", "Author: sepidmnorozy · Size: 10K – 100K rows", "[Dataset](https://huggingface.co/datasets/sepidmnorozy/Hebrew_sentiment)" ], [ "[HebrewMetaphors](https://huggingface.co/datasets/tdklab/HebrewMetaphors)", "Hebrew metaphor classification data.", "Author: tdklab · Size: 1K – 10K rows", "[Dataset](https://huggingface.co/datasets/tdklab/HebrewMetaphors)" ], [ "[HebrewSentiment](https://huggingface.co/datasets/HebArabNlpProject/HebrewSentiment)", "Hebrew sentiment (positive/negative/neutral) from the Hebrew NLP benchmark collection.", "Author: HebArabNlpProject · Size: 10K – 100K rows · License: other", "[Dataset](https://huggingface.co/datasets/HebArabNlpProject/HebrewSentiment)" ], [ "[wikianswers_hebrew](https://huggingface.co/datasets/imvladikon/wikianswers_hebrew)", "Hebrew version of WikiAnswers paraphrase/question data.", "Author: imvladikon · Size: 1M – 10M rows", "[Dataset](https://huggingface.co/datasets/imvladikon/wikianswers_hebrew)" ] ] } ] }, { "name": "OCR and Handwriting", "tables": [ { "subsection": null, "columns": [ "Name", "Description", "Details", "Links" ], "rows": [ [ "[hebrew-handwritten-dataset](https://huggingface.co/datasets/sivan22/hebrew-handwritten-dataset)", "HDD_v0: images of isolated handwritten Hebrew characters with train/test split.", "Author: sivan22 · Size: 1K – 10K rows · License: cc-by-3.0", "[Dataset](https://huggingface.co/datasets/sivan22/hebrew-handwritten-dataset)" ], [ "[Hebrew-Language-Signage](https://huggingface.co/datasets/danielrosehill/Hebrew-Language-Signage)", "68 photographs of Hebrew signage in Israel (street, commercial and public signs).", "Author: danielrosehill · Size: 68 images", "[Dataset](https://huggingface.co/datasets/danielrosehill/Hebrew-Language-Signage)" ], [ "[hebrew-ocr-dataset](https://huggingface.co/datasets/mbole/hebrew-ocr-dataset)", "Hebrew text with label and source columns for OCR work.", "Author: mbole · Size: <1K rows", "[Dataset](https://huggingface.co/datasets/mbole/hebrew-ocr-dataset)" ] ] } ] }, { "name": "Related", "tables": [ { "subsection": null, "columns": [ "Name", "Description", "Details", "Links" ], "rows": [ [ "[danielrosehill/Hebrew-AI-Models](https://github.com/danielrosehill/Hebrew-AI-Models)", "Companion list: Hebrew LLMs, ASR, TTS, OCR, NLP models.", "", "[View Repo](https://github.com/danielrosehill/Hebrew-AI-Models)" ], [ "[NNLP-IL/Hebrew-Resources](https://github.com/NNLP-IL/Hebrew-Resources)", "Comprehensive list of Hebrew NLP resources, including corpora.", "", "[View Repo](https://github.com/NNLP-IL/Hebrew-Resources)" ] ] } ] }, { "name": "Sentence Datasets", "tables": [ { "subsection": null, "columns": [ "Name", "Description", "Details", "Links" ], "rows": [ [ "[hebrew-wikipedia-sentences-corpus](https://huggingface.co/datasets/tomron87/hebrew-wikipedia-sentences-corpus)", "Cleaned, deduplicated Hebrew sentences from 366,610 Hebrew Wikipedia articles (crawled 2026-02).", "Author: tomron87 · Size: 10,999,257 sentences · License: cc-by-sa-3.0", "[Dataset](https://huggingface.co/datasets/tomron87/hebrew-wikipedia-sentences-corpus)" ] ] } ] }, { "name": "Text Corpora", "tables": [ { "subsection": "Hebrew-Specific Datasets", "columns": [ "Name", "Description", "Details", "Links" ], "rows": [ [ "[hebrew-BenIshHai](https://huggingface.co/datasets/AvishayDev/hebrew-BenIshHai)", "Halachot from the book Ben Ish Hai, from Hebrew Wikisource.", "Author: AvishayDev · Size: 1K – 10K rows", "[Dataset](https://huggingface.co/datasets/AvishayDev/hebrew-BenIshHai)" ], [ "[Hebrew-Mil](https://huggingface.co/datasets/MetroCat/Hebrew-Mil)", "Hebrew-only PDFs on military and defence topics.", "Author: MetroCat · Size: 1K – 10K rows · License: gpl", "[Dataset](https://huggingface.co/datasets/MetroCat/Hebrew-Mil)" ], [ "[hebrew_news](https://huggingface.co/datasets/imvladikon/hebrew_news)", "Hebrew news articles.", "Author: imvladikon · Size: 100K – 1M rows · License: other", "[Dataset](https://huggingface.co/datasets/imvladikon/hebrew_news)" ], [ "[Hebrew_Noam30B_Tokenized](https://huggingface.co/datasets/simsim314/Hebrew_Noam30B_Tokenized)", "Tokenized Hebrew text (input_ids and attention masks) for pretraining.", "Author: simsim314 · License: mit", "[Dataset](https://huggingface.co/datasets/simsim314/Hebrew_Noam30B_Tokenized)" ], [ "[Hebrew_wikipedia](https://huggingface.co/datasets/YanFren/Hebrew_wikipedia)", "Most of Hebrew Wikipedia as JSON, crawled breadth-first.", "Author: YanFren · Size: 100K – 1M rows · License: mit", "[Dataset](https://huggingface.co/datasets/YanFren/Hebrew_wikipedia)" ], [ "[HebrewLyricsDataet](https://huggingface.co/datasets/Norod78/HebrewLyricsDataet)", "Hebrew song lyrics with title and singer.", "Author: Norod78 · Size: 10K – 100K rows", "[Dataset](https://huggingface.co/datasets/Norod78/HebrewLyricsDataet)" ], [ "[Judaism-Hebrew](https://huggingface.co/datasets/guyhadad01/Judaism-Hebrew)", "Hebrew Judaism texts (responsa and related).", "Author: guyhadad01 · Size: 1K – 10K rows", "[Dataset](https://huggingface.co/datasets/guyhadad01/Judaism-Hebrew)" ], [ "[sefaria-hebrew](https://huggingface.co/datasets/sivan22/sefaria-hebrew)", "Jewish texts in Hebrew from the Sefaria project.", "Author: sivan22 · Size: 1M – 10M rows", "[Dataset](https://huggingface.co/datasets/sivan22/sefaria-hebrew)" ], [ "[synthetic_hebrew_medical_text](https://huggingface.co/datasets/cp500/synthetic_hebrew_medical_text)", "Synthetic Hebrew medical text.", "Author: cp500 · Size: 1K – 10K rows", "[Dataset](https://huggingface.co/datasets/cp500/synthetic_hebrew_medical_text)" ], [ "[Talmud-Hebrew](https://huggingface.co/datasets/guyhadad01/Talmud-Hebrew)", "Talmud Bavli text in Hebrew from Wikisource.", "Author: guyhadad01 · Size: <1K rows", "[Dataset](https://huggingface.co/datasets/guyhadad01/Talmud-Hebrew)" ] ] }, { "subsection": "Multilingual datasets with a Hebrew portion", "columns": [ "Name", "Description", "Details", "Links" ], "rows": [ [ "[fine-news](https://huggingface.co/datasets/ksolovev/fine-news)", "Multilingual news corpus from Common Crawl News; Hebrew is the `heb_Hebr` partition.", "Author: ksolovev · Size: 100M – 1B rows", "[Dataset](https://huggingface.co/datasets/ksolovev/fine-news)" ], [ "[wikipedia](https://huggingface.co/datasets/wikimedia/wikipedia)", "Wikipedia dumps; Hebrew is the `20231101.he` config.", "Author: wikimedia · Size: 10M – 100M rows · License: cc-by-sa-3.0", "[Dataset](https://huggingface.co/datasets/wikimedia/wikipedia)" ] ] } ] }, { "name": "Translation and Parallel Text", "tables": [ { "subsection": "Multilingual datasets with a Hebrew portion", "columns": [ "Name", "Description", "Details", "Links" ], "rows": [ [ "[flores_plus](https://huggingface.co/datasets/openlanguagedata/flores_plus)", "FLORES+ translation benchmark; Hebrew is `heb_Hebr` (also Yiddish `ydd_Hebr`).", "Author: openlanguagedata · Size: 100K – 1M rows · License: cc-by-sa-4.0", "[Dataset](https://huggingface.co/datasets/openlanguagedata/flores_plus)" ], [ "[parallel-sentences-global-voices](https://huggingface.co/datasets/sentence-transformers/parallel-sentences-global-voices)", "Global Voices news pairs; Hebrew is the `en-he` subset.", "Author: sentence-transformers · Size: 1M – 10M rows", "[Dataset](https://huggingface.co/datasets/sentence-transformers/parallel-sentences-global-voices)" ], [ "[parallel-sentences-talks](https://huggingface.co/datasets/sentence-transformers/parallel-sentences-talks)", "TED-style talk pairs; Hebrew is the `en-he` subset.", "Author: sentence-transformers · Size: 10M – 100M rows", "[Dataset](https://huggingface.co/datasets/sentence-transformers/parallel-sentences-talks)" ], [ "[parallel-sentences-tatoeba](https://huggingface.co/datasets/sentence-transformers/parallel-sentences-tatoeba)", "Tatoeba sentence pairs; Hebrew is the `en-he` subset.", "Author: sentence-transformers · Size: 1M – 10M rows", "[Dataset](https://huggingface.co/datasets/sentence-transformers/parallel-sentences-tatoeba)" ] ] } ] }, { "name": "Curated HF Collections", "tables": [ { "subsection": null, "columns": [ "Name", "Description", "Details", "Links" ], "rows": [ [ "[Audio Datasets - Mine](https://huggingface.co/collections/danielrosehill/audio-datasets-mine-6aee9e80cb04e786c3fdbeaa)", "The maintainer's own audio datasets.", "", "[Collection](https://huggingface.co/collections/danielrosehill/audio-datasets-mine-6aee9e80cb04e786c3fdbeaa)" ], [ "[Audio Datasets - Third Party Classics](https://huggingface.co/collections/danielrosehill/audio-datasets-third-party-classics-6aeea2c2b9ec48969f48e054)", "Well-known third-party audio datasets.", "", "[Collection](https://huggingface.co/collections/danielrosehill/audio-datasets-third-party-classics-6aeea2c2b9ec48969f48e054)" ], [ "[Code Switching (English And Hebrew) - All Resources](https://huggingface.co/collections/danielrosehill/code-switching-english-and-hebrew-all-resources-6acfdcb1e11911797af4c3b3)", "English-Hebrew code-switching models and datasets.", "", "[Collection](https://huggingface.co/collections/danielrosehill/code-switching-english-and-hebrew-all-resources-6acfdcb1e11911797af4c3b3)" ], [ "[Hebrew](https://huggingface.co/collections/danielrosehill/hebrew-6ad68f42575d0b5073d19e86)", "General Hebrew collection: models and datasets across tasks.", "", "[Collection](https://huggingface.co/collections/danielrosehill/hebrew-6ad68f42575d0b5073d19e86)" ], [ "[Hebrew audio datasets](https://huggingface.co/collections/danielrosehill/hebrew-audio-datasets-6ad6923f2ea298b0df13adf5)", "Hebrew speech and audio datasets.", "", "[Collection](https://huggingface.co/collections/danielrosehill/hebrew-audio-datasets-6ad6923f2ea298b0df13adf5)" ], [ "[Hebrew datasets](https://huggingface.co/collections/danielrosehill/hebrew-datasets-680428e2813deb0a8b8d1271)", "Hebrew text and NLP datasets.", "", "[Collection](https://huggingface.co/collections/danielrosehill/hebrew-datasets-680428e2813deb0a8b8d1271)" ], [ "[Israel Context Datasets](https://huggingface.co/collections/danielrosehill/israel-context-datasets-6ad676a3773a64192165a96c)", "Datasets about Israel, often Hebrew-language.", "", "[Collection](https://huggingface.co/collections/danielrosehill/israel-context-datasets-6ad676a3773a64192165a96c)" ], [ "[Israel Open Data](https://huggingface.co/collections/danielrosehill/israel-open-data-69fba702e180798fe3761073)", "Israeli government open data on the Hub.", "", "[Collection](https://huggingface.co/collections/danielrosehill/israel-open-data-69fba702e180798fe3761073)" ], [ "[Sentence Datasets - All](https://huggingface.co/collections/danielrosehill/sentence-datasets-all-6aeeb0facb33a09378459623)", "Sentence datasets (mixed languages; Hebrew ones are listed here).", "", "[Collection](https://huggingface.co/collections/danielrosehill/sentence-datasets-all-6aeeb0facb33a09378459623)" ], [ "[Sentence Datasets - Mine](https://huggingface.co/collections/danielrosehill/sentence-datasets-mine-6aeeb2ae61663e960fade58c)", "The maintainer's own sentence datasets.", "", "[Collection](https://huggingface.co/collections/danielrosehill/sentence-datasets-mine-6aeeb2ae61663e960fade58c)" ], [ "[Sentence Datasets - Third Party](https://huggingface.co/collections/danielrosehill/sentence-datasets-third-party-6aeea2c7b5a404455057151a)", "Third-party sentence datasets. All collections: [https://huggingface.co/danielrosehill/collections](https://huggingface.co/danielrosehill/collections)", "", "[Collection](https://huggingface.co/collections/danielrosehill/sentence-datasets-third-party-6aeea2c7b5a404455057151a)" ] ] } ] } ]