Skip to content

Pull requests: ROCm/vllm

New pull request New
19 Open 589 Closed
19 Open 589 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

Integrate mxfp4 MoE native kernels
#632 opened Aug 15, 2025 by mawong-amd Loading…
Updated README.md for August 12 RC2 throughput results only
#631 opened Aug 13, 2025 by Mcirino1 Loading…
3
[Pending][Fix] Shuffle weight and refine the quark MXFP4 kernel dispatch routine
#630 opened Aug 12, 2025 by zejunchen-zejun Loading…
2
[Model] Add GPT-OSS model code and config
#625 opened Aug 7, 2025 by ashishtanwer Loading…
add Fused_rms_quant for deepseek_v2 model
#611 opened Jul 29, 2025 by ZJLi2013 Loading…
1
[FEAT] [ROCm] Shared Experts Aiter
#605 opened Jul 25, 2025 by tjtanaavllm Loading…
1
add fused fp8 bmm
#604 opened Jul 25, 2025 by k50112113 Loading…
1
Update fp8 paged attention
#592 opened Jul 9, 2025 by amd-xiaoyu12 • Draft
Update test-template.j2
#579 opened Jun 16, 2025 by okakarpa Loading…
Disable skynny gemms by default
#568 opened Jun 5, 2025 by k-artem Loading…
1
Patch to run AITER 0507 stale
#541 opened May 8, 2025 by qli88 Loading…
1
Remap fp8 kv-scale names for Deepseek stale
#535 opened May 1, 2025 by sstamenk Loading…
2
Updated README.md with April 29 results stale
#526 opened Apr 27, 2025 by Mcirino1 Loading…
8
BF16 Skinny Optimization stale
#520 opened Apr 22, 2025 by amd-hhashemi Loading…
1
Test Queues
#456 opened Feb 28, 2025 by dhonnappa-amd • Draft
Enable custom paged attention kernel for Navi 3/4
#446 opened Feb 24, 2025 by hyoon1 Loading…
1
updating dev-docker README 20250214
#426 opened Feb 14, 2025 by arakowsk-amd • Draft
[Bugfix] Deepseek v3 fix max_num_batched_tokens
#386 opened Jan 24, 2025 by Concurrensee • Draft
merge paged attention feature and moe feature into llama_fp8_12062024
#370 opened Jan 21, 2025 by yuzho-amd • Draft
ProTip! Add no:assignee to see everything that’s not assigned.