Skip to content

Pull requests: vllm-project/vllm-gaudi

New pull request New
43 Open 1,693 Closed
43 Open 1,693 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

Revert "Enable Nemotron-H FP8 on HPU: quant-aware Mamba in_proj + non-gated FP8 MoE (#1666)"
#1785 opened Sep 7, 2026 by rsmyrek Contributor Loading…
7
Port of: Handle None attn_bias in FusedSDPA tiling path- #1778
#1784 opened Sep 7, 2026 by iboiko-habana Collaborator Loading…
2
Bump github/codeql-action/upload-sarif from 4.37.7 to 4.37.9
#1783 opened Sep 7, 2026 by iboiko-habana Collaborator Loading…
2
GLM-5.2-FP8 with DSA enablement
#1777 opened Sep 2, 2026 by jkaniecki Contributor Loading…
18
Fix VLLM_PROFILE_DECODE profiling a coarser bucket than requested
#1772 opened Aug 30, 2026 by osavchenkox Contributor Loading…
temp pr for test
#1771 opened Aug 28, 2026 by shepark Collaborator • Draft
GLM-5.2-FP8 with DSA enablement
#1760 opened Aug 27, 2026 by jkaniecki Contributor Loading…
5
feat: custom gathered-expert MoE combine for silu/FP8 (specifically Qwen 3.5 family)
#1731 opened Aug 13, 2026 by NatTuck Contributor Loading…
14
Restore the TPC causal_conv1d update path, switchable by env var
#1714 opened Aug 8, 2026 by rsenaikh Loading…
feat(model) NemotronH support
#1645 opened Jul 27, 2026 by narain1 • Draft
4
[kimi2.6] enable model text only initial
#1574 opened Jun 28, 2026 by sureshnam Collaborator Loading…
4
Pin vLLM core version to VLLM_GAUDI_COMMIT in RHEL UBI Dockerfile
#1571 opened Jun 26, 2026 by ghandoura Contributor Loading…
2
[DRAFT][GAUDISW-248216] Flag compact-GDN per-layer states as independent for the Synapse bridge
#1563 opened Jun 25, 2026 by moshehoori • Draft
4
Port of #1558 Adapt multi_model_api_server to vLLM ServingTokenizatio…
#1559 opened Jun 24, 2026 by iboiko-habana Collaborator Loading…
2
Added Hi Cache CPU Offloading test to CI
#1533 opened Jun 8, 2026 by slokesha Contributor • Draft
5
[WIP][GDN] Integrate gdn_read_decayed_state TPC kernel into decode path
#1531 opened Jun 8, 2026 by osavchenkox Contributor Loading…
6 tasks done
11
dynamic_quant: 20% improvements in per-channel mode
#1512 opened Jun 1, 2026 by osavchenkox Contributor Loading…
6
[DO_NOT_MERGE] Skip gather/index_copy for sequential state indices in GDN decode
#1487 opened May 24, 2026 by osavchenkox Contributor Loading…
4
[DO_NOT_MERGE] Remove repeat_interleave from GDN recurrent path
#1483 opened May 22, 2026 by osavchenkox Contributor Loading…
5
fix(test): work around Gaudi3 sdpa_recomp_fwd GQA kernel bug in reference
#1467 opened May 20, 2026 by yangulei Collaborator • Draft
5
[DO_NOT_MERGE] Use l2Norm TPC kernel for HPU compile
#1457 opened May 19, 2026 by osavchenkox Contributor Loading…
9
fix unit_moe: use loaded weight scales, force activation scale=1.0
#1456 opened May 19, 2026 by yangulei Collaborator • Draft
4
Global patch for t.compile high warmup time on MoE models
#1452 opened May 17, 2026 by tvoas Contributor Loading…
19
[WIP] enable step3p5 for bf16 and fp8
#1382 opened Apr 22, 2026 by ranzhejiang • Draft
36
Enhance GDN compact mode with prefix caching support
#1369 opened Apr 17, 2026 by jinyouzhi Contributor • Draft
6
ProTip! Mix and match filters to narrow down what you’re looking for.