Originally created by @thusinh1969 on GitHub (May 12, 2024).
Is your feature request related to a problem? Please describe. GGUF posed accuracy drop because of missing of bfloat16. We have to switch to vLLM and Triton.
Describe the solution you'd like Wishes to be able to run multiple OpenAI-compatible end points in the same front-end of Open WebUI likes what it does with ollama's multiple GGUF.
Additional context
This will be great as Open WebUI is one of the best for simplification, pre-prompting and all.
Thanks,
Steve
Originally created by @thusinh1969 on GitHub (May 12, 2024).
**Is your feature request related to a problem? Please describe.**
**GGUF posed accuracy drop because of missing of bfloat16. We have to switch to vLLM and Triton.**
**Describe the solution you'd like**
**Wishes to be able to run multiple OpenAI-compatible end points in the same front-end of Open WebUI likes what it does with ollama's multiple GGUF.**
**Additional context**
This will be great as Open WebUI is one of the best for simplification, pre-prompting and all.
Thanks,
Steve
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 @thusinh1969 on GitHub (May 12, 2024).
Is your feature request related to a problem? Please describe.
GGUF posed accuracy drop because of missing of bfloat16. We have to switch to vLLM and Triton.
Describe the solution you'd like
Wishes to be able to run multiple OpenAI-compatible end points in the same front-end of Open WebUI likes what it does with ollama's multiple GGUF.
Additional context
This will be great as Open WebUI is one of the best for simplification, pre-prompting and all.
Thanks,
Steve