Nextcloud-Archive/text2speech_kokoro
1
0
Fork 0
mirror of https://github.com/nextcloud/text2speech_kokoro.git synced 2026-02-07 03:14:50 +01:00
Code Issues Projects Releases Packages Wiki Activity
No description
72 commits 24 branches 5 tags 471 KiB
Find a file
Exact
Lukas Schaefer c3ccc071eb
 Merge pull request #50 from nextcloud/1.4.0 
Prepare 1.4.0
2026-01-09 15:47:43 +01:00
.github Remove GitHub Actions from dependabot.yml 2026-01-07 15:45:11 +01:00
.run Update dependencies and add nextcloud 33 support 2025-10-17 11:33:02 -04:00
appinfo Prepare 1.4.0 2026-01-07 10:23:08 -05:00
ex_app/lib fix: Only add metadata if we are asked to includeWatermark 2025-12-03 15:05:17 +01:00
.gitattributes Initial commit 2025-04-04 15:24:01 -04:00
.gitignore Initial commit 2025-04-04 15:24:01 -04:00
.nextcloudignore Initial commit 2025-04-04 15:24:01 -04:00
.pre-commit-config.yaml Initial commit 2025-04-04 15:24:01 -04:00
CHANGELOG.md Prepare 1.4.0 2026-01-07 10:23:08 -05:00
Dockerfile fix: add missing packages to Dockerfile 2026-01-07 09:29:29 -05:00
healthcheck.sh Initial commit 2025-04-04 15:24:01 -04:00
krankerl.toml Initial commit 2025-04-04 15:24:01 -04:00
LICENSE Update license to agpl 2025-05-07 15:51:58 -04:00
Makefile Prepare 1.4.0 2026-01-07 10:23:08 -05:00
pyproject.toml chore: pylint style less strict 2025-10-27 10:00:04 -04:00
README.md fix: typo in readme 2025-06-20 08:23:11 -04:00
requirements.txt Bump torch from 2.9.0 to 2.9.1 2026-01-07 15:06:44 +00:00
start.sh First commit 2025-04-05 17:20:54 -04:00

README.md

Nextcloud Kokoro Text2Speech App

An exapp that converts text to speech using Kokoro. This is a very lightweight model that supports languages such as English, Japanese, Mandarin, Spanish, Hindi, Italian, and Brazilian Portuguese.

Kokoro Text2Speech is a Nextcloud app that provides an AI capable of converting text to speech using various voices and languages.

This app is built on top of the Kokoro AI model.

Ethical AI Rating

Rating: 🟡

Positive:

  • The software for training and inferencing of this model is open source
  • The trained model is freely available, and thus can be ran on-premises

Negative:

  • The training data isn't freely available, making it not possible to check or correct for bias or optimise the performance and CO2 usage.

Running outside of Docker

Make sure to use python 3.12.