[GH-ISSUE #1862] Awq mod support/awq-gguf #78741

Closed
opened 2026-05-09 01:36:03 -05:00 by GiteaMirror · 3 comments
Owner

Originally created by @sviknesh97 on GitHub (Jan 9, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/1862

Does ollama support awq formats instead of gguf the gguf inference seems to be alittle slow hence thinking about awq and if it doesnt support is there a way to convert awq to gguf

Originally created by @sviknesh97 on GitHub (Jan 9, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/1862 Does ollama support awq formats instead of gguf the gguf inference seems to be alittle slow hence thinking about awq and if it doesnt support is there a way to convert awq to gguf
Author
Owner

@easp commented on GitHub (Jan 9, 2024):

It doesn't support awq. I kind of doubt that there is a good way to convert from awq to gguf. Do you not have access to the unquantized model?

<!-- gh-comment-id:1883457993 --> @easp commented on GitHub (Jan 9, 2024): It doesn't support awq. I kind of doubt that there is a good way to convert from awq to gguf. Do you not have access to the unquantized model?
Author
Owner

@sviknesh97 commented on GitHub (Jan 9, 2024):

Thanks Eric,

I can use the unquantized model but i was just wondering if it was possible.

On Wed, 10 Jan 2024, 01:12 Erik S, @.***> wrote:

It doesn't support awq. I kind of doubt that there is a good way to
convert from awq to gguf. Do you not have access to the unquantized model?


Reply to this email directly, view it on GitHub
https://github.com/jmorganca/ollama/issues/1862#issuecomment-1883457993,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/AX4XCOBRRXFWIKYNCGHR6ALYNV3AXAVCNFSM6AAAAABBSSEYZOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQOBTGQ2TOOJZGM
.
You are receiving this because you authored the thread.Message ID:
@.***>

<!-- gh-comment-id:1883571531 --> @sviknesh97 commented on GitHub (Jan 9, 2024): Thanks Eric, I can use the unquantized model but i was just wondering if it was possible. On Wed, 10 Jan 2024, 01:12 Erik S, ***@***.***> wrote: > It doesn't support awq. I kind of doubt that there is a good way to > convert from awq to gguf. Do you not have access to the unquantized model? > > — > Reply to this email directly, view it on GitHub > <https://github.com/jmorganca/ollama/issues/1862#issuecomment-1883457993>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/AX4XCOBRRXFWIKYNCGHR6ALYNV3AXAVCNFSM6AAAAABBSSEYZOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQOBTGQ2TOOJZGM> > . > You are receiving this because you authored the thread.Message ID: > ***@***.***> >
Author
Owner

@jmorganca commented on GitHub (May 10, 2024):

Hi @sviknesh97 thanks for the issue. In this case would it be possible to open an issue with the performance improvements or feature sets AWQ brings? That way we can focus on improving those parts of Ollama :)

<!-- gh-comment-id:2103666465 --> @jmorganca commented on GitHub (May 10, 2024): Hi @sviknesh97 thanks for the issue. In this case would it be possible to open an issue with the performance improvements or feature sets AWQ brings? That way we can focus on improving those parts of Ollama :)
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#78741