0.3.0

This release fixes major issues with speaker diarization on different platforms due to issues with dockerization. This release smooths out all of that and also releases a NVIDIA image that provides support for NVIDIA GPUs.

Also new in this release is the addition of a configuration wizard allowing you to choose models to download. More importantly, it slims down the docker image, as models are now downloaded externally to your volume mounts. This allows using larger models and eventually supporting other languages (in the next minor release).

Pulls out pocketbase into a separate container as well making the overall image much lighter and easier to build. It's also better practice to have one process per container.

FIguring out docker was the major hurdle. Now that it is working development will be a lot smoother.

What's Changed

• Wizard by @rishikanthc in #26

Full Changelog: 0.2.1...0.3.0

0.2.1

Fix docker images for diarization

v0.2.0

What's Changed

• Speaker Diarization by @rishikanthc in #16
• Enhance diarization by @rishikanthc in #17

Full Changelog: v0.1.0...v0.2.0

v0.1.0

What's Changed

• Update Dockerfile due to not being able to find curl on docker machine by @RJGY in #14
• Fix UI by @rishikanthc in #15

New Contributors

• @RJGY made their first contribution in #14
• @rishikanthc made their first contribution in #15

Full Changelog: vbeta-0.2...v0.1.0

Release Beta-0.2

What's Changed

• Support alternative OpenAI API Spec services (ex: ollama) by @phillipjf in #10
• refactor(image): use multi-stage builds to reduce final image size by @phillipjf in #11

New Contributors

• @phillipjf made their first contribution in #10

Full Changelog: https://github.com/rishikanthc/Scriberr/commits/vbeta-0.2