Parakeet V2 Fine-Tune - Training Data
Collection
Audio the Parakeet v2 fine-tunes blend from: master dataset, category and task subsets, imports, one blend dataset per model. • 28 items • Updated
October 7 Memorial Sentence Bank
Sentence dataset, split-out. Text to read for one topic, with no audio. Recordings: Hebrish/October-7-Memorial-Sentences. How every dataset fits together: the dataset index.
This dataset is dedicated to those who were murdered and who fell on October 7, 2023. Reading their names aloud is an act of remembrance, and it helps speech models write those names correctly.
It holds sentences to read aloud: three about the life of each person, and sentences about places and terms connected with that day. The sentences about people describe who they were, their work, families and passions, and not how they died. Where the sources say nothing about a person's life, their sentences simply remember their name.
Website: hebrish.app
Size
3,806 sentences: 3,687 about people and 119 about places and terms.
Columns
| Column | Meaning |
|---|---|
id |
Prompt id, shared with the recordings |
text |
The sentence |
term |
The person's name, or the place or term |
kind |
person, place or term |
source |
Where the fact came from: a record on oct7map.com, or a URL |
Attribution and licence
Sentences are licensed CC-BY-4.0. They are original wording, written by Claude from public records and checked line by line against those records by a separate pass. Only facts were used, never the sources' text.
Sources: oct7map.com for the list of people, October 7 Memorial for biographies, and Wikipedia and news reports for places. Names are spelled as those sources give them, which may differ from a family's own spelling. Survivors are not included.
If a name or a detail is wrong, please write to info@hebrish.app and it will be corrected.
Updating
Sentences are built in the project's private training repository and republished here after any correction.
- Downloads last month
- 77