Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Migrate FP6-LLM implementation from module to tensor subclass #402

Closed
gau-nernst opened this issue Jun 19, 2024 · 0 comments · Fixed by #399
Closed

Migrate FP6-LLM implementation from module to tensor subclass #402

gau-nernst opened this issue Jun 19, 2024 · 0 comments · Fixed by #399
Assignees
@gau-nernst

Comments

@gau-nernst
Copy link
Collaborator

With the tensor subclass guide published in #391, FP6-LLM implementation, which is currently a Linear module replacement, should migrate to tensor subclass too.
@gau-nernst gau-nernst self-assigned this Jun 19, 2024
@gau-nernst gau-nernst mentioned this issue Jun 24, 2024
Merged
yanbing-j pushed a commit to yanbing-j/ao that referenced this issue Dec 9, 2024
@mikekgfb @malfet
Embedding quantization per backend (pytorch#402)
c204150 
* ET or AOTI backend logic

* use args, not builder_args

* typo

* typo

* typo
yanbing-j pushed a commit to yanbing-j/ao that referenced this issue Dec 9, 2024
@malfet
Revert "Embedding quantization per backend (pytorch#402)" (pytorch#411)
0f69590 
This reverts commit 052fb1a3792123e291fb5f47084644913c8ecba7 to  fix [test-tinystories-executorch](https://github.com/pytorch/torchchat/actions/runs/8802811768/job/24159454773?pr=411#logs)
yanbing-j pushed a commit to yanbing-j/ao that referenced this issue Dec 9, 2024
@mikekgfb @malfet
Revert "Revert "Embedding quantization per backend"" (pytorch#415)
46bad0e 
* Revert "Revert "Embedding quantization per backend (pytorch#402)" (pytorch#411)"

This reverts commit 8b35acdff4fded779799ab8a419e55f885dd8918.

* 4b and 8b embedding table quantization

* minor changes

* remove extra et workflow
yanbing-j pushed a commit to yanbing-j/ao that referenced this issue Dec 9, 2024
@mikekgfb @malfet
Merge lowering of gguf files with ET tests into pull.yml (pytorch#417)
a99dc12 
* Revert "Revert "Embedding quantization per backend (pytorch#402)" (pytorch#411)"

This reverts commit 8b35acdff4fded779799ab8a419e55f885dd8918.

* merge GGUF tests into pull.yml
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@gau-nernst gau-nernst

Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging a pull request may close this issue.

Add FP5 E2M2 support from upstream gau-nernst/ao
1 participant
@gau-nernst