omnilingual-asr Releases
81 releases of facebookresearch/omnilingual-asr
- 0.2.0unlimited audio length support
This release introduces unlimited audio length support for the inference pipeline and models, enabling processing of audio files of any duration. Model documentation has also been updated.
Dec 22, 2025
- 0.1.0
Dec 12, 2025
- v3.2.0
This release marks a significant accessibility change by open-sourcing all Voice-Pro code and removing all associated costs, making the component completely free to use.
Dec 5, 2025
- v7.0.0-rc.9
Nov 21, 2025
- v3.1.4
Oct 5, 2025
- v4.2.1
Fixes Workspace Cache Locking issues by improving lock retry delays, adding stale file cleanup, and implementing recursive retry mechanisms.
Aug 22, 2025
- v4.2.0
Adds VS Code Configuration Compatibility, including support for `.vscode/mcp.json`, JSON5 syntax, and specific VS Code environment variables.
Jul 20, 2025
- v4.1.1
Enhances workspace cache with real-time lifecycle tracking, adding states like 'active' and 'shutting_down' and updating via SSE events.
Jul 17, 2025
- v4.1.0
Introduces global workspace cache management, multi-configuration file support, and global environment variable injection via `MCP_HUB_ENV`.
Jul 15, 2025
- v4.0.0
Major version update migrating the marketplace from Cline marketplace API to MCP Registry system, altering data structures and cache settings.
Jul 9, 2025
- v3.1.3
May 28, 2025
- 2.0.4VvvebJs 2.0.4
Update to VvvebJs includes a Landing theme, various UI fixes for tabs, accordions, and editor interactions, alongside improvements to CSS variable display and zip file naming.
May 3, 2025
- v0.2.0
Adds channel concatenated control conditioning trainers for Wan2.1 and CogView4, introduces support for FLUX.1-dev and Wan2.1 I2V, and expands attention provider support.
Apr 25, 2025
- v3.1.2
Apr 15, 2025
- v3.1.1
Mar 30, 2025
- v0.1.0
Initial stable release supporting CogView4, Wan 2.1, and LTX Video models, with fully restored Accelerate support and remote dataset loading improvements.
Mar 21, 2025
- v135.0.1-beta.24
Beta update adds Playwright 1.51.0 compatibility and introduces new certificate management properties for trusted authorities.
Mar 15, 2025
- v0.0.1
Initial release of the FineTrainers library supporting Diffusers-based models like CogVideoX, LTX Video, and Hunyuan Video, with LoRA and full-rank finetuning capabilities.
Feb 24, 2025
- 0.4.0
Released evaluation tools for Chat Engine and added support for new LLM and embedding providers like Gitee AI, vLLM, and Amazon Bedrock.
Jan 3, 2025
- v0.3.00.3.0
Renamed the project to autoflow and introduced support for multiple knowledge bases alongside new embedding model providers.
Dec 9, 2024