Jerma-Subtitle-Search

KO | EN

gitlite — search

#hacktoberfest #llm #go #golang #android #rust #docker #macos #cli #kubernetes #ios #mcp #openai #open-source

Jerma-Subtitle-Search

(★ 31)

Webpage for searching through everything Jerma has verbally said online

Jerma-Subtitle-Search 최신버젼 다운로드

최종 버전 다운로드 (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

### Subtitles

```
Video Count    : 2543
Word Count     : 29,079,203
Duration       : 6128:43:45
Oldest Video   : 2005-07-16
Latest Video   : 2026-03-04
```

Subtitles are obtained using [a Python script](/src/scripts/get_subtitles/get_subtitles.py).
Audio gets downloaded with [yt-dlp](https://github.com/yt-dlp/yt-dlp), which gets transcribed using [WhisperX](https://github.com/m-bain/whisperX) (large-v3 model) and converted to [LRC format](https://en.wikipedia.org/wiki/LRC_(file_format)) with [ffmpeg](https://github.com/FFmpeg/FFmpeg).

Relevant information gets written to a [JSON file](https://subtitlefiles.jerma.io/file/jerma-subtitles/Subtitles.json), which gets indexed and compressed using [a JS script](/src/scripts/index-subtitles.js).

The Python script also supports downloading manually added/YouTube's auto-generated subtitles, and optionally only transcribing videos which don't have subtitles available.

Initially used YouTube's auto-generated subtitles, but far too many videos either didn't have them available or had [censored](https://support.google.com/youtube/thread/70343381/new-default-setting-for-automatic-captions-uses-to-better-avoid-mistakes) swears.

Tried using [OpenAI's Whisper](https://github.com/openai/whisper) next, but after transcribing a bunch of videos with it I realized it kinda sucks in some aspects.
It hallucinated a lot, especially during sections with no speech.
Timestamps were incorrect on some transcriptions, and the first timestamp would always start at zero seconds, which was normally wrong.
It's also pretty slow, especially if you use some of the bigger models.

Switching to WhisperX _mostly_ solved the aforementioned problems.
However, it's still far from perfect and does have some [limitations](https://github.com/m-bain/whisperX/tree/49161922461871e6732fbe1aeb20fc1d4cccc9df#limitations-%EF%B8%8F).

</details>

### Webpage

Uses [Mithril](https://github.com/MithrilJS/mithril.js), [MiniSearch](https://github.com/lucaong/minisearch), [lite-youtube-embed](https://github.com/paulirish/lite-youtube-embed) and [fflate](https://github.com/101arrowz/fflate).

Primary mouse (tap on touchscreen) timestamps to open them in the embeds.  
Middle mouse (hold on touchscreen) timestamps to open them in a new tab.  
Secondary mouse timestamps to copy their link to the clipboard.

Wildcard characters (*) can be used in searches to match zero or more of any character.

#### Running Locally

```bash
# feel free to substitute bun with node & npm/yarn/whatever
git clone https://github.com/JermaSites/Jerma-Subtitle-Search.git
cd Jerma-Subtitle-Search
curl -o src/assets/Subtitles.json https://subtitlefiles.jerma.io/file/jerma-subtitles/Subtitles.json
bun install
bun src/scripts/index-subtitles.js
bun run dev
```

#### Contributing

If you'd like to contribute have a look at [the contributing guide](/CONTRIBUTING.md).

<!--
Notes to self:

AVIF conversion:
- ffmpeg -i "x" -map 0 -map 0 -filter:0 "format=yuv420p" -filter:1 "format=yuva444p,alphaextract" -crf 21 "x.avif"
-->

Jerma-Subtitle-Search

Jerma-Subtitle-Search 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

Jerma-Subtitle-Search

Jerma-Subtitle-Search 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

Download README