Datasets:

Hebrish
/

Dataset-Index

{
 "built": "2026-10-27",
 "master": "danielrosehill/Read-Aloud-Voice-Sentences",
 "blends": {
  "cs": {
   "line": "Jachnun (and the first Shakshuka runs)",
   "rule": "code-switching rows, Daniel's own English (about 2/3 of the English share), LibriSpeech-PC replay, plus the task tags ['currency']"
  },
  "all": {
   "line": "Sabich, Shakshuka v1.0+ and the Sabich variants",
   "rule": "every row of the master dataset at the run's revision, including every task tag present then, except ['name-cipher', 'paragraphs', 'weird-voices']; replay about 10%. From Sabich v2.1, priority-word takes (Hebrish/ASR-Error-List) appear 3 times."
  }
 },
 "models": [
  {
   "id": "danielrosehill/Parakeet-TDT-0.6B-v2-English-Hebrew-Codeswitching",
   "line": "Jachnun",
   "version": "PoC-1",
   "run": "pilot-en30",
   "released": "2026-10-15",
   "task": "speech-recognition",
   "base": null,
   "trained_on": [
    "danielrosehill/Read-Aloud-Voice-Sentences"
   ],
   "blend": "cs",
   "dataset_revision": "b31aeabb5edc",
   "minutes": null,
   "rows": null,
   "evidence": "ASR-Training data/checkpoints.json (pilot-en30)",
   "confidence": "confirmed"
  },
  {
   "id": "danielrosehill/Parakeet-TDT-0.6B-v2-English-Hebrew-Codeswitching",
   "line": "Jachnun",
   "version": "PoC-2",
   "run": "r2-en30",
   "released": "2026-10-15",
   "task": "speech-recognition",
   "base": null,
   "trained_on": [
    "danielrosehill/Read-Aloud-Voice-Sentences"
   ],
   "blend": "cs",
   "dataset_revision": "3bdaa2a50d52",
   "minutes": null,
   "rows": null,
   "evidence": "ASR-Training data/checkpoints.json (r2-en30)",
   "confidence": "confirmed"
  },
  {
   "id": "danielrosehill/Parakeet-TDT-0.6B-v2-English-Hebrew-Codeswitching",
   "line": "Jachnun",
   "version": "PoC-3",
   "run": "r3-en30",
   "released": "2026-10-18",
   "task": "speech-recognition",
   "base": null,
   "trained_on": [
    "danielrosehill/Read-Aloud-Voice-Sentences"
   ],
   "blend": "cs",
   "dataset_revision": "b47310621f79",
   "minutes": null,
   "rows": null,
   "evidence": "ASR-Training data/checkpoints.json (r3-en30)",
   "confidence": "confirmed"
  },
  {
   "id": "danielrosehill/Parakeet-TDT-0.6B-v2-English-Hebrew-Codeswitching",
   "line": "Jachnun",
   "version": "PoC-4",
   "run": "r4-en30",
   "released": "2026-10-18",
   "task": "speech-recognition",
   "base": null,
   "trained_on": [
    "danielrosehill/Read-Aloud-Voice-Sentences"
   ],
   "blend": "cs",
   "dataset_revision": "4fcca820e0c2",
   "minutes": null,
   "rows": null,
   "evidence": "ASR-Training data/checkpoints.json (r4-en30)",
   "confidence": "confirmed"
  },
  {
   "id": "danielrosehill/Parakeet-TDT-0.6B-v2-English-Hebrew-Codeswitching",
   "line": "Jachnun",
   "version": "PoC-5",
   "run": "r5-en30",
   "released": "2026-10-18",
   "task": "speech-recognition",
   "base": "nvidia/parakeet-tdt-0.6b-v2",
   "trained_on": [
    "danielrosehill/Read-Aloud-Voice-Sentences"
   ],
   "blend": "cs",
   "dataset_revision": "fcf5aec16449",
   "minutes": null,
   "rows": null,
   "evidence": "ASR-Training data/checkpoints.json (r5-en30)",
   "confidence": "confirmed"
  },
  {
   "id": "Hebrish/Nemotron-Speech-Streaming-EN-0.6B-English-Hebrew-Codeswitching",
   "line": "Shakshuka",
   "version": "PoC-1",
   "run": "r5-nemotron-en30",
   "released": "2026-10-18",
   "task": "speech-recognition",
   "base": "nvidia/nemotron-speech-streaming-en-0.6b",
   "trained_on": [
    "danielrosehill/Read-Aloud-Voice-Sentences"
   ],
   "blend": "cs",
   "dataset_revision": "fcf5aec16449",
   "minutes": null,
   "rows": null,
   "evidence": "ASR-Training data/checkpoints.json (r5-nemotron-en30)",
   "confidence": "confirmed"
  },
  {
   "id": "danielrosehill/Parakeet-TDT-0.6B-v2-English-Hebrew-Codeswitching",
   "line": "Jachnun",
   "version": "v1.0",
   "run": "r6-en30",
   "released": "2026-10-19",
   "task": "speech-recognition",
   "base": "nvidia/parakeet-tdt-0.6b-v2",
   "trained_on": [
    "danielrosehill/Read-Aloud-Voice-Sentences"
   ],
   "blend": "cs",
   "dataset_revision": "550bfcaa0f84",
   "minutes": null,
   "rows": null,
   "evidence": "ASR-Training data/checkpoints.json (r6-en30)",
   "confidence": "confirmed"
  },
  {
   "id": "Hebrish/Nemotron-Speech-Streaming-EN-0.6B-English-Hebrew-Codeswitching",
   "line": "Shakshuka",
   "version": "PoC-2",
   "run": "r6-nemotron-en30",
   "released": "2026-10-19",
   "task": "speech-recognition",
   "base": "nvidia/nemotron-speech-streaming-en-0.6b",
   "trained_on": [
    "danielrosehill/Read-Aloud-Voice-Sentences"
   ],
   "blend": "cs",
   "dataset_revision": "550bfcaa0f84",
   "minutes": null,
   "rows": null,
   "evidence": "ASR-Training data/checkpoints.json (r6-nemotron-en30)",
   "confidence": "confirmed"
  },
  {
   "id": "Hebrish/Sabich",
   "line": "Sabich",
   "version": "v1.0",
   "run": "r7-all",
   "released": "2026-10-20",
   "task": "speech-recognition",
   "base": "nvidia/parakeet-tdt-0.6b-v2",
   "trained_on": [
    "danielrosehill/Read-Aloud-Voice-Sentences"
   ],
   "blend": "all",
   "dataset_revision": "f956dbb184c2",
   "minutes": 623.0,
   "rows": 5334,
   "evidence": "ASR-Training data/checkpoints.json (r7-all)",
   "confidence": "confirmed"
  },
  {
   "id": "Hebrish/Shakshuka",
   "line": "Shakshuka",
   "version": "v1.0",
   "run": "r8-nemotron-all",
   "released": "2026-10-21",
   "task": "speech-recognition",
   "base": "nvidia/nemotron-speech-streaming-en-0.6b",
   "trained_on": [
    "danielrosehill/Read-Aloud-Voice-Sentences"
   ],
   "blend": "all",
   "dataset_revision": "e918e381e831",
   "minutes": 658.4,
   "rows": 5634,
   "evidence": "ASR-Training data/checkpoints.json (r8-nemotron-all)",
   "confidence": "confirmed"
  },
  {
   "id": "Hebrish/Sabich-Parakeet-110M",
   "line": "Sabich-Parakeet-110M",
   "version": "v1.0",
   "run": "r8-ctc110m-all",
   "released": "2026-10-21",
   "task": "speech-recognition",
   "base": "nvidia/parakeet-tdt_ctc-110m",
   "trained_on": [
    "danielrosehill/Read-Aloud-Voice-Sentences"
   ],
   "blend": "all",
   "dataset_revision": "e918e381e831",
   "minutes": 658.4,
   "rows": 5634,
   "evidence": "ASR-Training data/checkpoints.json (r8-ctc110m-all)",
   "confidence": "confirmed"
  },
  {
   "id": "Hebrish/Sabich-Qwen3-ASR-1.7B",
   "line": "Sabich-Qwen3-ASR-1.7B",
   "version": "PoC",
   "run": "q1-all",
   "released": null,
   "task": "speech-recognition",
   "base": "Qwen/Qwen3-ASR-1.7B",
   "trained_on": [
    "danielrosehill/Read-Aloud-Voice-Sentences"
   ],
   "blend": "all",
   "dataset_revision": "e918e381e831",
   "minutes": 658.4,
   "rows": 5634,
   "evidence": "ASR-Training data/checkpoints.json (q1-all)",
   "confidence": "confirmed"
  },
  {
   "id": "Hebrish/Sabich-Canary-Qwen-2.5B",
   "line": "Sabich-Canary-Qwen-2.5B",
   "version": "v1.0",
   "run": "cq1-all",
   "released": "2026-10-21",
   "task": "speech-recognition",
   "base": "nvidia/canary-qwen-2.5b",
   "trained_on": [
    "danielrosehill/Read-Aloud-Voice-Sentences"
   ],
   "blend": "all",
   "dataset_revision": "cf84f7e45d78",
   "minutes": 669.3,
   "rows": 5752,
   "evidence": "ASR-Training data/checkpoints.json (cq1-all)",
   "confidence": "confirmed"
  },
  {
   "id": "Hebrish/Sabich-asr4all-M",
   "line": "Sabich-asr4all-M",
   "version": "v1.0",
   "run": "a1-all",
   "released": "2026-10-21",
   "task": "speech-recognition",
   "base": "futo-org/asr4all-m",
   "trained_on": [
    "danielrosehill/Read-Aloud-Voice-Sentences"
   ],
   "blend": "all",
   "dataset_revision": "cf84f7e45d78",
   "minutes": 669.3,
   "rows": 5752,
   "evidence": "ASR-Training data/checkpoints.json (a1-all)",
   "confidence": "confirmed"
  },
  {
   "id": "Hebrish/Sabich-asr4all-S",
   "line": "Sabich-asr4all-S",
   "version": "v1.0",
   "run": "a1s-all",
   "released": "2026-10-21",
   "task": "speech-recognition",
   "base": "futo-org/asr4all-s",
   "trained_on": [
    "danielrosehill/Read-Aloud-Voice-Sentences"
   ],
   "blend": "all",
   "dataset_revision": "cf84f7e45d78",
   "minutes": 669.3,
   "rows": 5752,
   "evidence": "ASR-Training data/checkpoints.json (a1s-all)",
   "confidence": "confirmed"
  },
  {
   "id": "Hebrish/Sabich",
   "line": "Sabich",
   "version": "v2.0",
   "run": "r9-all",
   "released": "2026-10-22",
   "task": "speech-recognition",
   "base": "nvidia/parakeet-tdt-0.6b-v2",
   "trained_on": [
    "danielrosehill/Read-Aloud-Voice-Sentences"
   ],
   "blend": "all",
   "dataset_revision": "110d144a0502",
   "minutes": 713.9,
   "rows": 6225,
   "evidence": "ASR-Training data/checkpoints.json (r9-all)",
   "confidence": "confirmed"
  },
  {
   "id": "Hebrish/Shakshuka",
   "line": "Shakshuka",
   "version": "v2.0",
   "run": "r9-nemotron-all",
   "released": "2026-10-22",
   "task": "speech-recognition",
   "base": "nvidia/nemotron-speech-streaming-en-0.6b",
   "trained_on": [
    "danielrosehill/Read-Aloud-Voice-Sentences"
   ],
   "blend": "all",
   "dataset_revision": "110d144a0502",
   "minutes": 713.9,
   "rows": 6225,
   "evidence": "ASR-Training data/checkpoints.json (r9-nemotron-all)",
   "confidence": "confirmed"
  },
  {
   "id": "Hebrish/Sabich",
   "line": "Sabich",
   "version": "v2.1",
   "run": "r10-all",
   "released": "2026-10-25",
   "task": "speech-recognition",
   "base": "nvidia/parakeet-tdt-0.6b-v2",
   "trained_on": [
    "danielrosehill/Read-Aloud-Voice-Sentences"
   ],
   "blend": "all",
   "dataset_revision": "2cec2ebc85e9",
   "minutes": 848.4,
   "rows": 7765,
   "evidence": "ASR-Training data/checkpoints.json (r10-all)",
   "confidence": "confirmed"
  },
  {
   "id": "Hebrish/Meorav",
   "line": "Meorav",
   "version": "v2.1",
   "released": "2026-10-22",
   "task": "text-to-speech",
   "base": "Qwen3-TTS",
   "trained_on": [
    "Hebrish/Hebrish-TTS-Voice-Recordings"
   ],
   "data_note": "train split, 1,103 takes (v2.0 and v2.1)",
   "evidence": "Hebrish/Meorav model card",
   "confidence": "confirmed"
  },
  {
   "id": "Hebrish/Meorav",
   "line": "Meorav",
   "version": "v2.0",
   "released": "2026-10-22",
   "task": "text-to-speech",
   "base": "Qwen3-TTS",
   "trained_on": [
    "Hebrish/Hebrish-TTS-Voice-Recordings"
   ],
   "data_note": "train split, 1,103 takes; replaced by v2.1 the same day",
   "evidence": "Hebrish/Meorav model card",
   "confidence": "confirmed"
  },
  {
   "id": "Hebrish/TTS-V1",
   "line": "Meorav",
   "version": "PoC-1",
   "released": "2026-10-22",
   "task": "text-to-speech",
   "base": "Qwen3-TTS",
   "trained_on": [
    "Hebrish/Hebrish-TTS-Voice-Recordings"
   ],
   "data_note": "657 read takes, 50 min after trimming (run first-test); the repo is no longer on the Hub (404 on 2026-10-27)",
   "evidence": "Meorav-TTS README, model table",
   "confidence": "confirmed"
  },
  {
   "id": "danielrosehill/Whisper-Hebrish",
   "line": "Whisper-Hebrish",
   "version": "v2",
   "task": "speech-recognition",
   "base": "openai/whisper-large-v3-turbo",
   "trained_on": [
    "danielrosehill/Read-Aloud-Voice-Sentences"
   ],
   "data_note": "1,929 training rows / 215 min, mostly phone takes, 29% monolingual English; term-disjoint test split",
   "evidence": "model card, 'What changed from v1'",
   "confidence": "confirmed"
  },
  {
   "id": "Hebrish/Whisper-Hebrish-Small-ACFT",
   "line": "Whisper-Hebrish",
   "version": "small, ACFT",
   "task": "speech-recognition",
   "base": "openai/whisper-small",
   "trained_on": [
    "danielrosehill/Read-Aloud-Voice-Sentences"
   ],
   "data_note": "same 1,929 training clips and mix as Whisper-Hebrish v2",
   "evidence": "model card, 'Training'",
   "confidence": "confirmed"
  },
  {
   "id": "Hebrish/Hebrew-Latin-Token-Classifier",
   "line": "Token classifier",
   "version": "v1",
   "task": "token-classification",
   "trained_on": [
    "Hebrish/hebrew-latin-code-switching"
   ],
   "data_note": "4,737 sentences; marks the romanized Hebrew words in English text (TTS chain)",
   "evidence": "model card",
   "confidence": "confirmed"
  }
 ],
 "datasets": [
  {
   "id": "danielrosehill/Read-Aloud-Voice-Sentences",
   "title": "Read-Aloud Voice Sentences (master)",
   "kind": "audio-takes",
   "tag": null,
   "description": "Every take from the capture app; all speech models train on a pinned revision of it, selected by blend. Gated.",
   "in_blends": [
    "cs",
    "all"
   ],
   "status": "published",
   "url": "https://huggingface.co/datasets/danielrosehill/Read-Aloud-Voice-Sentences",
   "last_modified": "2026-10-27"
  },
  {
   "id": "danielrosehill/English-Hebrew-Codeswitching-Sentences",
   "title": "English-Hebrew Code-Switching Sentences",
   "kind": "audio-takes",
   "tag": "code_switch",
   "description": "English sentences with everyday Hebrew words and Israeli names written in Latin letters, read aloud by one English speaker living in Israel",
   "in_blends": [
    "cs",
    "all"
   ],
   "status": "published",
   "url": "https://huggingface.co/datasets/danielrosehill/English-Hebrew-Codeswitching-Sentences",
   "last_modified": "2026-10-27"
  },
  {
   "id": "danielrosehill/Tech-Sentences",
   "title": "Tech Sentences",
   "kind": "audio-takes",
   "tag": "tech",
   "description": "Dictation-style English sentences dense in software, DevOps, cloud, self-hosting and AI vocabulary (product names, CLI tools, acronyms, model names), read aloud by one speaker",
   "in_blends": [
    "all"
   ],
   "status": "published",
   "url": "https://huggingface.co/datasets/danielrosehill/Tech-Sentences",
   "last_modified": "2026-10-27"
  },
  {
   "id": "danielrosehill/AI-Interactions-Sentences",
   "title": "AI Interactions Sentences",
   "kind": "audio-takes",
   "tag": "ai_interactions",
   "description": "Instructions dictated to AI coding agents (Claude Code, Codex and others) and the git and GitHub workflow around them, read aloud by one speaker: subagents, worktrees, compaction, plan mode, hooks, MCP servers, pull requests, rebases",
   "in_blends": [
    "all"
   ],
   "status": "published",
   "url": "https://huggingface.co/datasets/danielrosehill/AI-Interactions-Sentences",
   "last_modified": "2026-10-27"
  },
  {
   "id": "danielrosehill/Linux-Admin-Sentences",
   "title": "Linux Admin Sentences",
   "kind": "audio-takes",
   "tag": "linux",
   "description": "Dictation-style English sentences about Linux desktop and server administration (KDE Plasma, GNOME, systemd, storage, networking, homelab), read aloud by one speaker",
   "in_blends": [
    "all"
   ],
   "status": "published",
   "url": "https://huggingface.co/datasets/danielrosehill/Linux-Admin-Sentences",
   "last_modified": "2026-10-27"
  },
  {
   "id": "danielrosehill/Smart-Home-Sentences",
   "title": "Smart Home Sentences",
   "kind": "audio-takes",
   "tag": "smart_home",
   "description": "Dictation-style English sentences about home automation (Home Assistant, Zigbee, Z-Wave, Matter, Thread, ESPHome, MQTT and device brands), read aloud by one speaker",
   "in_blends": [
    "all"
   ],
   "status": "published",
   "url": "https://huggingface.co/datasets/danielrosehill/Smart-Home-Sentences",
   "last_modified": "2026-10-27"
  },
  {
   "id": "danielrosehill/Personal-Vocabulary-Sentences",
   "title": "Personal Vocabulary Sentences",
   "kind": "audio-takes",
   "tag": "personal",
   "description": "English sentences built around names a speech model gets wrong for one speaker: his own name, his company and his project names (Daniel Rosehill, DSR Holdings, Carrot Cake AI, the podcast My Weird Prompts and its hosts Corn and Herman Poppl",
   "in_blends": [
    "all"
   ],
   "status": "published",
   "url": "https://huggingface.co/datasets/danielrosehill/Personal-Vocabulary-Sentences",
   "last_modified": "2026-10-27"
  },
  {
   "id": "danielrosehill/Homophone-Context-Sentences",
   "title": "Homophone Context Sentences",
   "kind": "audio-takes",
   "tag": "homophone",
   "description": "English sentences whose context settles which spelling of a homophone is meant (their/there, principal/principle, and 58 more sets), read aloud by one speaker",
   "in_blends": [
    "all"
   ],
   "status": "published",
   "url": "https://huggingface.co/datasets/danielrosehill/Homophone-Context-Sentences",
   "last_modified": "2026-10-27"
  },
  {
   "id": "Hebrish/Yiddish-Sentences",
   "title": "Yiddish Sentences",
   "kind": "audio-takes",
   "tag": "yiddish",
   "description": "English sentences with Yiddish words written in Latin letters (shul, daven, cholent, mensch, freier, tachles), the way English speakers in Israel use them, read aloud by one speaker",
   "in_blends": [
    "all"
   ],
   "status": "published",
   "url": "https://huggingface.co/datasets/Hebrish/Yiddish-Sentences",
   "last_modified": "2026-10-27"
  },
  {
   "id": "Hebrish/Jewish-Religious-Terminology-Sentences",
   "title": "Jewish Religious Terminology Sentences",
   "kind": "audio-takes",
   "tag": "religious",
   "description": "English sentences with Jewish religious vocabulary written in Latin letters: Shabbat and the festivals, prayer and the synagogue, Torah portions, Talmud and halacha, kashrut, and life-cycle events, read aloud by one speaker",
   "in_blends": [
    "all"
   ],
   "status": "published",
   "url": "https://huggingface.co/datasets/Hebrish/Jewish-Religious-Terminology-Sentences",
   "last_modified": "2026-10-27"
  },
  {
   "id": "danielrosehill/General-English-Sentences",
   "title": "General English Sentences",
   "kind": "audio-takes",
   "tag": "general",
   "description": "Plain, everyday English sentences (emails, messages, general prose and homophone pairs) read aloud by one speaker, with no Hebrew or Yiddish words, Israeli names or tech vocabulary",
   "in_blends": [
    "all"
   ],
   "status": "published",
   "url": "https://huggingface.co/datasets/danielrosehill/General-English-Sentences",
   "last_modified": "2026-10-27"
  },
  {
   "id": "Hebrish/Noisy-Condition-Sentences",
   "title": "Noisy Condition Sentences",
   "kind": "audio-takes",
   "tag": "noisy",
   "description": "Every take from the main dataset recorded in noise, whatever its category: outdoors, in a noisy room, with traffic, other people talking, background noise or a crying baby",
   "in_blends": [
    "all"
   ],
   "status": "published",
   "url": "https://huggingface.co/datasets/Hebrish/Noisy-Condition-Sentences",
   "last_modified": "2026-10-27"
  },
  {
   "id": "Hebrish/Disfluent-Read-Sentences",
   "title": "Disfluent Read Sentences",
   "kind": "audio-takes",
   "tag": "disfluency",
   "description": "Experimental",
   "in_blends": [
    "all"
   ],
   "status": "published",
   "url": "https://huggingface.co/datasets/Hebrish/Disfluent-Read-Sentences",
   "last_modified": "2026-10-27"
  },
  {
   "id": "Hebrish/Self-Correction-Sentences",
   "title": "Self-Correction Sentences",
   "kind": "audio-takes",
   "tag": "self-corrections",
   "description": "One speaker correcting himself mid-sentence (\"send it Tuesday, no, Wednesday\")",
   "in_blends": [
    "all"
   ],
   "status": "published",
   "url": "https://huggingface.co/datasets/Hebrish/Self-Correction-Sentences",
   "last_modified": "2026-10-27"
  },
  {
   "id": "Hebrish/Spoken-Aside-Sentences",
   "title": "Spoken Aside Sentences",
   "kind": "audio-takes",
   "tag": "asides",
   "description": "One speaker dictating a sentence and, partway through, saying something to someone in the room (\"come in, Hannah\")",
   "in_blends": [
    "all"
   ],
   "status": "published",
   "url": "https://huggingface.co/datasets/Hebrish/Spoken-Aside-Sentences",
   "last_modified": "2026-10-27"
  },
  {
   "id": "Hebrish/Hard-ASR-Sentences",
   "title": "Hard ASR Sentences",
   "kind": "audio-takes",
   "tag": "hard-asr",
   "description": "English sentences speech recognition finds hard, read by one speaker: Harvard sentences (IEEE 1969, public domain), tongue twisters, homographs, garden-path sentences, rare words, dense numbers and units, unusual proper nouns",
   "in_blends": [
    "all"
   ],
   "status": "published",
   "url": "https://huggingface.co/datasets/Hebrish/Hard-ASR-Sentences",
   "last_modified": "2026-10-27"
  },
  {
   "id": "Hebrish/Brand-Name-Sentences",
   "title": "Brand Name Sentences",
   "kind": "audio-takes",
   "tag": "brand-names",
   "description": "Natural sentences with company, product, AI model and open-source project names (Anthropic, OpenRouter, Amazon Bedrock, Qwen, Kubernetes, Nginx) read by one speaker, written in each name's official form",
   "in_blends": [
    "all"
   ],
   "status": "published",
   "url": "https://huggingface.co/datasets/Hebrish/Brand-Name-Sentences",
   "last_modified": "2026-10-27"
  },
  {
   "id": "Hebrish/Acronym-Sentences",
   "title": "Acronym Sentences",
   "kind": "audio-takes",
   "tag": "acronyms",
   "description": "Natural sentences with acronyms read by one speaker, both spelled out (API, SSH, HTTPS) and said as words (NASA, JSON, YAML), each in its standard written form",
   "in_blends": [
    "all"
   ],
   "status": "published",
   "url": "https://huggingface.co/datasets/Hebrish/Acronym-Sentences",
   "last_modified": "2026-10-27"
  },
  {
   "id": "Hebrish/Irish-English-Sentences",
   "title": "Irish English Sentences",
   "kind": "audio-takes",
   "tag": "irish-english",
   "description": "Everyday sentences in Irish English read by one speaker born in Ireland: vocabulary (craic, grand, giving out, press, messages), grammar (I'm after doing it, so it is, ye), Irish-language words used in English (Taoiseach, Dáil, Gardaí, slái",
   "in_blends": [
    "all"
   ],
   "status": "published",
   "url": "https://huggingface.co/datasets/Hebrish/Irish-English-Sentences",
   "last_modified": "2026-10-27"
  },
  {
   "id": "Hebrish/Repeated-Word-Sentences",
   "title": "Repeated Word Sentences",
   "kind": "audio-takes",
   "tag": "repetitions",
   "description": "Experimental",
   "in_blends": [
    "all"
   ],
   "status": "published",
   "url": "https://huggingface.co/datasets/Hebrish/Repeated-Word-Sentences",
   "last_modified": "2026-10-27"
  },
  {
   "id": "Hebrish/Singing-Sentences",
   "title": "Singing Sentences",
   "kind": "audio-takes",
   "tag": "singing",
   "description": "Experimental",
   "in_blends": [
    "all"
   ],
   "status": "published",
   "url": "https://huggingface.co/datasets/Hebrish/Singing-Sentences",
   "last_modified": "2026-10-27"
  },
  {
   "id": "Hebrish/Weird-Voice-Sentences",
   "title": "Weird Voice Sentences",
   "kind": "audio-takes",
   "tag": "weird-voices",
   "description": "Experimental",
   "in_blends": [],
   "status": "published",
   "url": "https://huggingface.co/datasets/Hebrish/Weird-Voice-Sentences",
   "last_modified": "2026-10-27"
  },
  {
   "id": "Hebrish/Currency-Sentences",
   "title": "Currency Sentences",
   "kind": "audio-takes",
   "tag": "currency",
   "description": "One speaker saying amounts of money the natural way (\"50 shekels\", \"a dollar fifty\", \"6 pounds 50\")",
   "in_blends": [
    "cs",
    "all"
   ],
   "status": "published",
   "url": "https://huggingface.co/datasets/Hebrish/Currency-Sentences",
   "last_modified": "2026-10-27"
  },
  {
   "id": "Hebrish/Paragraph-Dictation-Passages",
   "title": "Paragraph Dictation Passages",
   "kind": "audio-takes",
   "tag": "paragraphs",
   "description": "Experimental",
   "in_blends": [],
   "status": "planned",
   "url": "https://huggingface.co/datasets/Hebrish/Paragraph-Dictation-Passages"
  },
  {
   "id": "Hebrish/Expanded-Acronym-Sentences",
   "title": "Expanded Acronym Sentences",
   "kind": "audio-takes",
   "tag": "expanded-acronyms",
   "description": "Chat and business shorthand said as letters by one speaker (\"by E O D\", \"F Y I\", \"I M O\") with the transcript writing it out in full (\"by end of day\", \"For your information\", \"In my opinion\")",
   "in_blends": [
    "all"
   ],
   "status": "published",
   "url": "https://huggingface.co/datasets/Hebrish/Expanded-Acronym-Sentences",
   "last_modified": "2026-10-27"
  },
  {
   "id": "Hebrish/Procurement-Sentences",
   "title": "Procurement Sentences",
   "kind": "audio-takes",
   "tag": "procurement",
   "description": "Business sentences with purchasing and trade vocabulary read by one speaker: Incoterms 2020 (EXW, FCA, FOB, CIF, DDP, …) with their named place, purchase orders, RFQs, RFPs, invoices, bills of lading, MOQs, lead times, payment terms, freigh",
   "in_blends": [
    "all"
   ],
   "status": "planned",
   "url": "https://huggingface.co/datasets/Hebrish/Procurement-Sentences"
  },
  {
   "id": "Hebrish/Hardware-Sentences",
   "title": "Hardware Sentences",
   "kind": "audio-takes",
   "tag": "hardware",
   "description": "Short workshop and DIY sentences with hardware vocabulary read by one speaker: electronic components (capacitors, resistors, transistors, MOSFETs, regulators), fasteners and fixings (M3 and M4 screws, standoffs, wall anchors, Torx), connect",
   "in_blends": [
    "all"
   ],
   "status": "planned",
   "url": "https://huggingface.co/datasets/Hebrish/Hardware-Sentences"
  },
  {
   "id": "Hebrish/Israeli-Business-Sentences",
   "title": "Israeli Business Sentences",
   "kind": "audio-takes",
   "tag": "israeli-businesses",
   "description": "Everyday English sentences naming Israeli businesses read by one speaker: banks and insurers, supermarket and electronics chains, telecoms, food brands, cafes and public services, in the Latin spelling each uses itself (Hazi Hinam, Bank Yah",
   "in_blends": [
    "all"
   ],
   "status": "planned",
   "url": "https://huggingface.co/datasets/Hebrish/Israeli-Business-Sentences"
  },
  {
   "id": "Hebrish/Idiom-Sentences",
   "title": "Idiom Sentences",
   "kind": "audio-takes",
   "tag": "idioms",
   "description": "Natural sentences with English idioms that speech recognition tends to spell as a sound-alike (\"free rein\", not \"free reign\"; \"toe the line\", not \"tow\"; \"piqued my interest\"), read by one speaker, written in the standard form",
   "in_blends": [
    "all"
   ],
   "status": "planned",
   "url": "https://huggingface.co/datasets/Hebrish/Idiom-Sentences"
  },
  {
   "id": "Hebrish/Name-Cipher-Sentences",
   "title": "Name Cipher Sentences",
   "kind": "audio-takes",
   "tag": "name-cipher",
   "description": "Experimental, a privacy (PII obfuscation) test",
   "in_blends": [],
   "status": "planned",
   "url": "https://huggingface.co/datasets/Hebrish/Name-Cipher-Sentences"
  },
  {
   "id": "Hebrish/Numerics-Sentences",
   "title": "Numerics Sentences",
   "kind": "audio-takes",
   "tag": "numerics",
   "description": "Every kind of number that is not money, said by one speaker in a fixed way and written the way it is typed: the same number said several ways (\"thirteen hundred\", \"one thousand three hundred\" → 1,300), decimals, temperatures (written \"27 de",
   "in_blends": [
    "all"
   ],
   "status": "planned",
   "url": "https://huggingface.co/datasets/Hebrish/Numerics-Sentences"
  },
  {
   "id": "Hebrish/Fast-Speech-Sentences",
   "title": "Fast Speech Sentences",
   "kind": "audio-takes",
   "tag": "fast-speech",
   "description": "Robustness data",
   "in_blends": [
    "all"
   ],
   "status": "planned",
   "url": "https://huggingface.co/datasets/Hebrish/Fast-Speech-Sentences"
  },
  {
   "id": "Hebrish/Whispered-Sentences",
   "title": "Whispered Sentences",
   "kind": "audio-takes",
   "tag": "whispered",
   "description": "Experimental robustness data",
   "in_blends": [
    "all"
   ],
   "status": "planned",
   "url": "https://huggingface.co/datasets/Hebrish/Whispered-Sentences"
  },
  {
   "id": "Hebrish/Web-Address-Sentences",
   "title": "Web Address Sentences",
   "kind": "audio-takes",
   "tag": "web-addresses",
   "description": "One speaker saying web addresses aloud (\"hebrish dot app slash benchmarks\", \"h t t p s colon slash slash\", \"localhost colon three thousand\"), with the transcript writing them as typed: `hebrish.app/benchmarks`, `https://`, `localhost:3000`",
   "in_blends": [
    "all"
   ],
   "status": "planned",
   "url": "https://huggingface.co/datasets/Hebrish/Web-Address-Sentences"
  },
  {
   "id": "Hebrish/Spelled-Out-Sentences",
   "title": "Spelled-Out Sentences",
   "kind": "audio-takes",
   "tag": "spelled-out",
   "description": "One speaker spelling names, words and codes letter by letter in dictation (\"my surname is B R E N N A N\"), with the transcript writing the word the letters spell (`My surname is Brennan.`); codes as typed, uppercase letters and digits (`KQ7",
   "in_blends": [
    "all"
   ],
   "status": "planned",
   "url": "https://huggingface.co/datasets/Hebrish/Spelled-Out-Sentences"
  },
  {
   "id": "Hebrish/False-Start-Sentences",
   "title": "False-Start Sentences",
   "kind": "audio-takes",
   "tag": "false-starts",
   "description": "Experimental",
   "in_blends": [
    "all"
   ],
   "status": "planned",
   "url": "https://huggingface.co/datasets/Hebrish/False-Start-Sentences"
  },
  {
   "id": "Hebrish/NATO-Alphabet-Sentences",
   "title": "NATO Alphabet Sentences",
   "kind": "audio-takes",
   "tag": "nato-alphabet",
   "description": "Experimental",
   "in_blends": [
    "all"
   ],
   "status": "planned",
   "url": "https://huggingface.co/datasets/Hebrish/NATO-Alphabet-Sentences"
  },
  {
   "id": "Hebrish/Latin-Phrase-Sentences",
   "title": "Latin Phrase Sentences",
   "kind": "audio-takes",
   "tag": "latin-phrases",
   "description": "One speaker reading English sentences that use a Latin phrase (de facto, ad hominem, de minimis, bona fide, ad nauseam, per se, quid pro quo), with the phrase in its standard English spelling: lowercase, no italics, no hyphen (`ad hoc`, `po",
   "in_blends": [
    "all"
   ],
   "status": "planned",
   "url": "https://huggingface.co/datasets/Hebrish/Latin-Phrase-Sentences"
  },
  {
   "id": "Hebrish/Fragment-Sentences",
   "title": "Fragment Sentences",
   "kind": "audio-takes",
   "tag": "fragments",
   "description": "One speaker reading short and incomplete utterances, one to six words: message replies (\"Not yet.\"), single words (\"Apparently.\"), notes to self (\"Dentist, Thursday morning.\"), list items, headings, trailing clauses (\"Unless it rains.\") and",
   "in_blends": [
    "all"
   ],
   "status": "planned",
   "url": "https://huggingface.co/datasets/Hebrish/Fragment-Sentences"
  },
  {
   "id": "Hebrish/Entity-Dense-Sentences",
   "title": "Entity-Dense Sentences",
   "kind": "audio-takes",
   "tag": "entity-dense",
   "description": "One speaker reading sentences that pack three or more entities into one line: company and product names, people, places, amounts, percentages, dates, times and quantities, in the style of results calls, meeting recaps, itineraries, invoices",
   "in_blends": [
    "all"
   ],
   "status": "planned",
   "url": "https://huggingface.co/datasets/Hebrish/Entity-Dense-Sentences"
  },
  {
   "id": "Hebrish/Spoken-Punctuation-Sentences",
   "title": "Spoken Punctuation Sentences",
   "kind": "audio-takes",
   "tag": "spoken-punctuation",
   "description": "Experimental",
   "in_blends": [
    "all"
   ],
   "status": "planned",
   "url": "https://huggingface.co/datasets/Hebrish/Spoken-Punctuation-Sentences"
  },
  {
   "id": "Hebrish/English-Hebrew-Codeswitching-ASR-Benchmark",
   "kind": "benchmark",
   "description": "Held-out test sets (v1, v2) the models are scored on; never trained on.",
   "training": "never (evaluation only)",
   "title": "English Hebrew Codeswitching ASR Benchmark",
   "tag": null,
   "in_blends": [],
   "status": "published",
   "url": "https://huggingface.co/datasets/Hebrish/English-Hebrew-Codeswitching-ASR-Benchmark",
   "last_modified": "2026-10-25"
  },
  {
   "id": "Hebrish/ASR-Error-List",
   "kind": "error-list",
   "description": "Words the models mishear in Daniel's own dictation, with the wrong forms; drives the capture app's priority words. Sabich v2.1 repeated its takes 3 times.",
   "training": "via the master dataset's priority-word takes",
   "title": "ASR Error List",
   "tag": null,
   "in_blends": [],
   "status": "published",
   "url": "https://huggingface.co/datasets/Hebrish/ASR-Error-List",
   "last_modified": "2026-10-27"
  },
  {
   "id": "Hebrish/ASR-Error-Prediction",
   "kind": "error-list",
   "description": "Predicted mistranscription candidates for planning new sentences.",
   "training": "never",
   "title": "ASR Error Prediction",
   "tag": null,
   "in_blends": [],
   "status": "published",
   "url": "https://huggingface.co/datasets/Hebrish/ASR-Error-Prediction",
   "last_modified": "2026-10-25"
  },
  {
   "id": "Hebrish/Hebrish-TTS-Voice-Recordings",
   "kind": "audio-takes",
   "description": "Clean read takes for the Meorav voice (text-to-speech), a filtered view of the capture app's voice-clone takes.",
   "training": "Meorav",
   "title": "Hebrish TTS Voice Recordings",
   "tag": null,
   "in_blends": [],
   "status": "published",
   "url": "https://huggingface.co/datasets/Hebrish/Hebrish-TTS-Voice-Recordings",
   "last_modified": "2026-10-22"
  },
  {
   "id": "Hebrish/English-Hebrew-Mixed-Sentences",
   "kind": "sentence-text",
   "description": "The code-switching sentence bank the capture app serves (text only).",
   "training": "prompts for recordings, not audio",
   "title": "English Hebrew Mixed Sentences",
   "tag": null,
   "in_blends": [],
   "status": "published",
   "url": "https://huggingface.co/datasets/Hebrish/English-Hebrew-Mixed-Sentences",
   "last_modified": "2026-10-22"
  },
  {
   "id": "Hebrish/hebrew-latin-code-switching",
   "kind": "sentence-text",
   "description": "Sentences with romanized Hebrew words labelled, for the token classifier.",
   "training": "Hebrew-Latin token classifier",
   "title": "hebrew latin code switching",
   "tag": null,
   "in_blends": [],
   "status": "published",
   "url": "https://huggingface.co/datasets/Hebrish/hebrew-latin-code-switching",
   "last_modified": "2026-10-22"
  },
  {
   "id": "Hebrish/Linux-Admin-Sentence-Bank",
   "kind": "sentence-text",
   "description": "Linux administration sentence bank (text only).",
   "training": "prompts for recordings, not audio",
   "title": "Linux Admin Sentence Bank",
   "tag": null,
   "in_blends": [],
   "status": "published",
   "url": "https://huggingface.co/datasets/Hebrish/Linux-Admin-Sentence-Bank",
   "last_modified": "2026-10-22"
  },
  {
   "id": "Hebrish/Smart-Home-Sentence-Bank",
   "kind": "sentence-text",
   "description": "Smart-home sentence bank (text only).",
   "training": "prompts for recordings, not audio",
   "title": "Smart Home Sentence Bank",
   "tag": null,
   "in_blends": [],
   "status": "published",
   "url": "https://huggingface.co/datasets/Hebrish/Smart-Home-Sentence-Bank",
   "last_modified": "2026-10-22"
  },
  {
   "id": "Hebrish/Personal-Vocabulary-Sentence-Bank",
   "kind": "sentence-text",
   "description": "Personal vocabulary sentence bank (text only).",
   "training": "prompts for recordings, not audio",
   "title": "Personal Vocabulary Sentence Bank",
   "tag": null,
   "in_blends": [],
   "status": "published",
   "url": "https://huggingface.co/datasets/Hebrish/Personal-Vocabulary-Sentence-Bank",
   "last_modified": "2026-10-26"
  },
  {
   "id": "Hebrish/AI-Tech-Vocabulary-Sentence-Bank",
   "kind": "sentence-text",
   "description": "AI and tech vocabulary sentence bank (text only).",
   "training": "prompts for recordings, not audio",
   "title": "AI Tech Vocabulary Sentence Bank",
   "tag": null,
   "in_blends": [],
   "status": "published",
   "url": "https://huggingface.co/datasets/Hebrish/AI-Tech-Vocabulary-Sentence-Bank",
   "last_modified": "2026-10-22"
  },
  {
   "id": "Hebrish/Homophone-Context-Sentence-Bank",
   "kind": "sentence-text",
   "description": "Homophone-in-context sentence bank (text only).",
   "training": "prompts for recordings, not audio",
   "title": "Homophone Context Sentence Bank",
   "tag": null,
   "in_blends": [],
   "status": "published",
   "url": "https://huggingface.co/datasets/Hebrish/Homophone-Context-Sentence-Bank",
   "last_modified": "2026-10-22"
  },
  {
   "id": "Hebrish/AI-Interactions-Sentence-Bank",
   "kind": "sentence-text",
   "description": "Sentences dictated to AI agents (text only).",
   "training": "prompts for recordings, not audio",
   "title": "AI Interactions Sentence Bank",
   "tag": null,
   "in_blends": [],
   "status": "published",
   "url": "https://huggingface.co/datasets/Hebrish/AI-Interactions-Sentence-Bank",
   "last_modified": "2026-10-22"
  },
  {
   "id": "Hebrish/Parakeet-V2-Blend-All",
   "kind": "training-mix export",
   "description": "Export of the `all` training mix, for reproducing Sabich runs.",
   "training": "export of the all mix",
   "title": "Parakeet V2 Blend All",
   "tag": null,
   "in_blends": [],
   "status": "published",
   "url": "https://huggingface.co/datasets/Hebrish/Parakeet-V2-Blend-All",
   "last_modified": "2026-10-25"
  },
  {
   "id": "Hebrish/Room-Tone-Annotations",
   "kind": "annotations",
   "description": "Room-tone recordings and labels for noise conditions.",
   "training": "not yet used",
   "title": "Room Tone Annotations",
   "tag": null,
   "in_blends": [],
   "status": "published",
   "url": "https://huggingface.co/datasets/Hebrish/Room-Tone-Annotations",
   "last_modified": "2026-10-22"
  },
  {
   "id": "Hebrish/Hebrish-Ideas",
   "kind": "other",
   "description": "Ideas recorded in the capture app.",
   "training": "never",
   "title": "Hebrish Ideas",
   "tag": null,
   "in_blends": [],
   "status": "published",
   "url": "https://huggingface.co/datasets/Hebrish/Hebrish-Ideas",
   "last_modified": "2026-10-26"
  }
 ]
}