Skip to content

Pull requests: vllm-project/llm-compressor

New pull request New
71 Open 1,415 Closed
71 Open 1,415 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

[ZENDNN] Added support for qwen 3.6 W8A8
#2789 opened Jun 2, 2026 by Priyjain-amd • Draft
1
7
[Distributed] Module parallel calibration for QuantizationModifier
#2785 opened Jun 2, 2026 by kylesayrs Collaborator Loading…
5
[Lifecycle] Add modules argument to on_sequential_epoch_end
#2784 opened Jun 2, 2026 by kylesayrs Collaborator Loading…
3
[Lifecycle] Make calibration events first-class citizens awq For any issue / PR related to AWQ support gptq For any PR / issue related to GPTQ support ready When a PR is ready for review smoothquant For any issue / PR related to SmoothQuant support transforms Related to transforms-based modifiers like SpinQuant and Quip two-reviews When a PR requires two reviews
#2783 opened Jun 2, 2026 by kylesayrs Collaborator Loading…
12
[Misc] Remove oneshot output_dir warning enhancement New feature or request
#2782 opened Jun 2, 2026 by kylesayrs Collaborator Loading…
2
[Autowrapper] [Tracing] Unpin Gemma4, support tracing UserDict and IfExp enhancement New feature or request moe ready When a PR is ready for review tracing Issues related to model tracing
#2781 opened Jun 2, 2026 by kylesayrs Collaborator Loading…
10
[Tests] Sampling params for E2E tests enhancement New feature or request
#2780 opened Jun 1, 2026 by kylesayrs Collaborator • Draft
9
refactor: modernize type hints in modeling/ module (part of #1927) llama For any PR / issue related to Llama herd support moe Refactor Code cleanup and/or improvements to existing features two-reviews When a PR requires two reviews
#2779 opened Jun 1, 2026 by u7k4rs6 Loading…
3 of 4 tasks
4
refactor: modernize type hints in utils/ module (part of #1927) Refactor Code cleanup and/or improvements to existing features two-reviews When a PR requires two reviews
#2777 opened Jun 1, 2026 by u7k4rs6 Loading…
3 of 5 tasks
7
[XPU] Add torch.cuda linter Refactor Code cleanup and/or improvements to existing features
#2776 opened Jun 1, 2026 by kylesayrs Collaborator • Draft
25
refactor: modernize utils, pipeline, and observers modules with Python 3.10+ type hints Refactor Code cleanup and/or improvements to existing features two-reviews When a PR requires two reviews
#2775 opened Jun 1, 2026 by AsadShahid04 Contributor Loading…
1
9
[Docs] Minor fixes for sequential onloading docs documentation Improvements or additions to documentation ready When a PR is ready for review
#2771 opened May 29, 2026 by kylesayrs Collaborator Loading…
5
fix: Add AWQ model mappings for Step3p5 awq For any issue / PR related to AWQ support enhancement New feature or request requires-validate Indicates that a PR looks appropriate, but needs to be run before merging transforms Related to transforms-based modifiers like SpinQuant and Quip two-reviews When a PR requires two reviews
#2770 opened May 28, 2026 by wanadzhar913 Loading…
1
10
[Tests] Add comprehensive DDP smoke tests for compression modifiers autoround For any PR / issue related to autoround support awq For any issue / PR related to AWQ support gptq For any PR / issue related to GPTQ support smoothquant For any issue / PR related to SmoothQuant support transforms Related to transforms-based modifiers like SpinQuant and Quip
#2769 opened May 28, 2026 by HDCharles Collaborator Loading…
12
Add AWQ Mappings for Glm4MoeLiteForCausalLM awq For any issue / PR related to AWQ support enhancement New feature or request moe requires-validate Indicates that a PR looks appropriate, but needs to be run before merging transforms Related to transforms-based modifiers like SpinQuant and Quip two-reviews When a PR requires two reviews
#2768 opened May 28, 2026 by Priya95715 Loading…
1
5
minor docs: remove non existent training args ready When a PR is ready for review two-reviews When a PR requires two reviews
#2762 opened May 27, 2026 by JINO-ROHIT Contributor Loading…
7
add config for step 3.5 awq For any issue / PR related to AWQ support enhancement New feature or request moe transforms Related to transforms-based modifiers like SpinQuant and Quip two-reviews When a PR requires two reviews
#2760 opened May 27, 2026 by JINO-ROHIT Contributor Loading…
1
7
[Claude] create-tiny-model skill for creating tiny test models enhancement New feature or request transforms Related to transforms-based modifiers like SpinQuant and Quip
#2753 opened May 24, 2026 by kylesayrs Collaborator Loading…
19
[Examples] Add MR-GPTQ (QuIP + GPTQ + NVFP4A16) example documentation Improvements or additions to documentation enhancement New feature or request gptq For any PR / issue related to GPTQ support llama For any PR / issue related to Llama herd support nvfp4 For any PR / issue related to NVFP4 support transforms Related to transforms-based modifiers like SpinQuant and Quip two-reviews When a PR requires two reviews
#2751 opened May 23, 2026 by Yatimai Contributor Loading…
9
[Examples] remove gpt oss example documentation Improvements or additions to documentation moe Refactor Code cleanup and/or improvements to existing features w4a16
#2742 opened May 20, 2026 by brian-dellabetta Collaborator Loading…
1
4
[deepseek_v4] Extend ARCH_TO_2D_MAPPINGS for MTP block two-reviews When a PR requires two reviews
#2739 opened May 20, 2026 by pasta-paul Loading…
4
[model_free_ptq] Deprecate reindex_fused_weights model_free_ptq For any PR/issue related to the `model_free_ptq` pathway ready When a PR is ready for review Refactor Code cleanup and/or improvements to existing features two-reviews When a PR requires two reviews
#2737 opened May 20, 2026 by kylesayrs Collaborator Loading…
8
fix bug fused obs + sequential targets bug Something isn't working ready When a PR is ready for review two-reviews When a PR requires two reviews
#2732 opened May 20, 2026 by HDCharles Collaborator Loading…
25
Remove unnecessary torch.no_grad from deepseek_v32 example deepseek documentation Improvements or additions to documentation ready When a PR is ready for review Refactor Code cleanup and/or improvements to existing features
#2731 opened May 20, 2026 by kylesayrs Collaborator Loading…
1 task
5
fix: add Qwen2.5-VL and sync model mappings for AWQ and SmoothQuant awq For any issue / PR related to AWQ support bug Something isn't working qwen For any PR / issue related to Qwen support requires-validate Indicates that a PR looks appropriate, but needs to be run before merging smoothquant For any issue / PR related to SmoothQuant support transforms Related to transforms-based modifiers like SpinQuant and Quip two-reviews When a PR requires two reviews
#2727 opened May 19, 2026 by AsadShahid04 Contributor Loading…
1
29
ProTip! Type g p on any issue or pull request to go back to the pull request listing page.