<feed xmlns='http://www.w3.org/2005/Atom'>
<title>TaSTT.git/Scripts/string_matcher.py, branch v0.8.1</title>
<subtitle>Free self-hosted STT for VRChat.</subtitle>
<id>https://git.yummers.dev/TaSTT.git/atom?h=v0.8.1</id>
<link rel='self' href='https://git.yummers.dev/TaSTT.git/atom?h=v0.8.1'/>
<link rel='alternate' type='text/html' href='https://git.yummers.dev/TaSTT.git/'/>
<updated>2023-02-23T05:49:29+00:00</updated>
<entry>
<title>Begin work on C++ implementation</title>
<updated>2023-02-23T05:49:29+00:00</updated>
<author>
<name>yum</name>
<email>yum.food.vr@gmail.com</email>
</author>
<published>2023-02-21T21:19:43+00:00</published>
<link rel='alternate' type='text/html' href='https://git.yummers.dev/TaSTT.git/commit/?id=9a97fbc3c583ccd518d838faaaa36ed9aa5558e1'/>
<id>urn:sha1:9a97fbc3c583ccd518d838faaaa36ed9aa5558e1</id>
<content type='text'>
Use Const-me/Whisper to perform transcription. This implementation is
vastly more efficient: CPU usage, memory usage, and VRAM usage are all
dramatically reduced. It's slightly less accurate when comparing the
same model (due to the lack of beam search decoding), but since you can
use larger models, the impact is largely a wash.
</content>
</entry>
<entry>
<title>Bugfix: user-provided paths may now contain spaces</title>
<updated>2023-01-04T18:03:39+00:00</updated>
<author>
<name>yum</name>
<email>yum.food.vr@gmail.com</email>
</author>
<published>2023-01-04T17:52:02+00:00</published>
<link rel='alternate' type='text/html' href='https://git.yummers.dev/TaSTT.git/commit/?id=66d311b3267620995e5c35b16f3fba18ed0c48f3'/>
<id>urn:sha1:66d311b3267620995e5c35b16f3fba18ed0c48f3</id>
<content type='text'>
Previously, paths containing spaces would be interpreted by python's argument
parser as multiple separate arguments, causing it to fail. Now we escape paths
inside PythonWrapper using std::quoted().

* Improve PII filtering. Python output would contain multiple path separators
  (like C:\\Users\\foo\\), defeating the PII regex.
* Silence compiler warning in PII filter.
* Document usability improvements.
* Transcription layer exponential backoff goes to ~infinity when paused.
  This is a hack, since we really don't need to transcribe at all when paused,
  but it lets us keep the code simple. Good enough until the next rewrite.
* Shader only samples background when necessary.
* Limit matchStrings() print()s to DEBUG mode
</content>
</entry>
<entry>
<title>Fine-tune transcription</title>
<updated>2022-12-30T08:01:28+00:00</updated>
<author>
<name>yum</name>
<email>yum.food.vr@gmail.com</email>
</author>
<published>2022-12-30T08:01:28+00:00</published>
<link rel='alternate' type='text/html' href='https://git.yummers.dev/TaSTT.git/commit/?id=abdaa7ce215086bf1070d6093731cd35df866cbb'/>
<id>urn:sha1:abdaa7ce215086bf1070d6093731cd35df866cbb</id>
<content type='text'>
Bump up recording window to 28 seconds. This helps a lot with long-form
transcription tasks, s.a. transcribing an audiobook.

We should expose this as a parameter, since at 10s the transcription delay is
typically 300ms, while at 28s it's typically 1.1-1.2s.
</content>
</entry>
<entry>
<title>Finish python virtual env</title>
<updated>2022-12-18T01:51:12+00:00</updated>
<author>
<name>yum</name>
<email>yum.food.vr@gmail.com</email>
</author>
<published>2022-12-18T01:51:12+00:00</published>
<link rel='alternate' type='text/html' href='https://git.yummers.dev/TaSTT.git/commit/?id=ee8213d1d2c2008d2d996929500c9e87dac325a3'/>
<id>urn:sha1:ee8213d1d2c2008d2d996929500c9e87dac325a3</id>
<content type='text'>
GUI can now download all TaSTT dependencies and install them into a
virtual environment.

* Add buttons to check embedded python version &amp; install dependencies
* Add class to wrap interacting with embedded Python
* Put all TaSTT python scripts into a folder
</content>
</entry>
</feed>
