Currently, Whisper works very nicely on the GPU. However, once it has taken up VRAM, it never releases it, even when it is no longer in use. It would be a nice feature if we could release VRAM after a few minutes of inactivity, similar to how Ollama handles this.
Originally created by @mbeltagy on GitHub (Feb 28, 2025).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/11003
Currently, Whisper works very nicely on the GPU. However, once it has taken up VRAM, it never releases it, even when it is no longer in use. It would be a nice feature if we could release VRAM after a few minutes of inactivity, similar to how Ollama handles this.
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Originally created by @mbeltagy on GitHub (Feb 28, 2025).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/11003
Currently, Whisper works very nicely on the GPU. However, once it has taken up VRAM, it never releases it, even when it is no longer in use. It would be a nice feature if we could release VRAM after a few minutes of inactivity, similar to how Ollama handles this.