Skip to content

Navigation Menu

Sign in
Appearance settings

Search code, repositories, users, issues, pull requests...

Provide feedback

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly
Appearance settings

Pull requests: vllm-project/llm-compressor

New pull request New
46 Open 986 Closed
46 Open 986 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

[AWQ] Fix _accumulate_mean bug, move AWQ activation averaging off CPU, and improve logging
#2161 opened Dec 19, 2025 by ZewenShen-Cohere Loading…
6
[AWQ] Support for a module used in an AWQ mapping to be unquantized && other bug fixes
#2158 opened Dec 19, 2025 by ZewenShen-Cohere • Draft
11
Enhance Autoround to support multiple cards tuning autoround For any PR / issue related to autoround support
#2157 opened Dec 19, 2025 by yiliu30 Loading…
3 tasks done
10
[AWQ][nvfp4] adding support/tests awq For any issue / PR related to AWQ support nvfp4 For any PR / issue related to NVFP4 support
#2154 opened Dec 18, 2025 by HDCharles Loading…
6
[AWQ] mapping shouldn't use ignore awq For any issue / PR related to AWQ support bug Something isn't working ready When a PR is ready for review
#2152 opened Dec 18, 2025 by HDCharles Loading…
1
8
[AutoRound] Support w8a8 scheme in auto-round and add example autoround For any PR / issue related to autoround support fp8 For any issue / PR related to FP8 support
#2150 opened Dec 18, 2025 by mengniwang95 Loading…
12
[TorchOffloader] Switch to torch offloader from accelerate
#2148 opened Dec 18, 2025 by kylesayrs • Draft
4
[TorchOffloader] Code Cleanup
#2147 opened Dec 18, 2025 by kylesayrs • Draft
6
[Tracing] Dispatch after tracing ready When a PR is ready for review
#2146 opened Dec 17, 2025 by kylesayrs Loading…
2
[Args] Shuffle data samples by default
#2144 opened Dec 17, 2025 by kylesayrs Loading…
6
[Bugfix] Improve pipeline inference ready When a PR is ready for review
#2131 opened Dec 15, 2025 by kylesayrs Loading…
5
[Example] MedGemma ready When a PR is ready for review
#2126 opened Dec 14, 2025 by kylesayrs Loading…
1
5
[Examples] QwenOmni Example
#2125 opened Dec 14, 2025 by kylesayrs Loading…
1
12
[Misc] Better debugging and guards to autowrapping ready When a PR is ready for review
#2124 opened Dec 14, 2025 by kylesayrs Loading…
6
Add MSE vs MinMax observer comparison tests
#2110 opened Dec 11, 2025 by GOavi101 Loading…
9
Fix deprecated torch_dtype usage in transformers loading ready When a PR is ready for review
#2109 opened Dec 11, 2025 by jangel97 Loading…
14
[Bug fix] fix Qwen3VLMoe
#2104 opened Dec 9, 2025 by Wangzheee Loading…
9
add kv quant example autoround For any PR / issue related to autoround support
#2100 opened Dec 5, 2025 by mengniwang95 Loading…
12
[Utils] Delete deprecated utilities
#2098 opened Dec 5, 2025 by kylesayrs • Draft
4
feat: add importance-aware mixed-precision quantization
#2083 opened Dec 2, 2025 by wangwenmingaa Loading…
8
[test] add e2e test for qwen3 moe w4a16 ready When a PR is ready for review
#2071 opened Nov 25, 2025 by HDCharles • Draft
7
[Misc] Remove is_moe_model ready When a PR is ready for review
#2053 opened Nov 20, 2025 by kylesayrs Loading…
4
Testing Clean-up
#2045 opened Nov 18, 2025 by dsikka • Draft
2
Support wInt4aFp8 for moe
#2027 opened Nov 12, 2025 by Wangzheee Loading…
6
[Sequential Onloading] Support onloading and offloading frozen dataclasses
#2016 opened Nov 10, 2025 by kylesayrs Loading…
5
ProTip! Find all pull requests that aren't related to any open issues with -linked:issue.
Morty Proxy This is a proxified and sanitized view of the page, visit original site.