Skip to content

Issues: vllm-project/vllm

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

Clear current search query, filters, and sorts
58 Open 664 Closed
58 Open 664 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
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

[Not to be Submitted] [WIP] Force Unit tests to run with BlockManager V2 ready ONLY add when PR is ready to merge/full CI is needed
#8678 opened Sep 20, 2024 by sroy745 • Draft
2
[Core] Factor out common code in SequenceData and Sequence ready ONLY add when PR is ready to merge/full CI is needed
#8675 opened Sep 20, 2024 by DarkLight1337 Loading…
1
[Core] Rename PromptInputs to PromptType, and inputs to prompt ready ONLY add when PR is ready to merge/full CI is needed
#8673 opened Sep 20, 2024 by DarkLight1337 Loading…
1
[Kernel] Split Marlin MoE kernels into multiple files ready ONLY add when PR is ready to merge/full CI is needed
#8661 opened Sep 20, 2024 by ElizaWszola Loading…
2
[Kernel][Triton][AMD] Remove tl.atomic_add from awq_gemm_kernel, 2-5x speedup MI300, minor improvement for MI250 ready ONLY add when PR is ready to merge/full CI is needed
#8646 opened Sep 20, 2024 by rasmith Loading…
3
[Kernel][Bugfix] Delete some more useless code in marlin_moe_ops.cu ready ONLY add when PR is ready to merge/full CI is needed
#8643 opened Sep 19, 2024 by tlrmchlsmth Loading…
2
[Misc] Support FP8 MoE for compressed-tensors ready ONLY add when PR is ready to merge/full CI is needed
#8588 opened Sep 19, 2024 by mgoin Loading…
4 tasks done
4
[MISC] add support custom_op check ready ONLY add when PR is ready to merge/full CI is needed
#8557 opened Sep 18, 2024 by jikunshang Loading…
7
[Bugfix] fix OpenAI API server startup with --disable-frontend-multiprocessing ready ONLY add when PR is ready to merge/full CI is needed
#8537 opened Sep 17, 2024 by dtrifiro Loading…
10
[CI/Build][Misc] Comparing between block manager v1 and v2, under full prefix sharing and no prefix sharing case. ready ONLY add when PR is ready to merge/full CI is needed
#8528 opened Sep 16, 2024 by KuntaiDu Loading…
3
[dbrx] refactor dbrx experts to extend FusedMoe class ready ONLY add when PR is ready to merge/full CI is needed
#8518 opened Sep 16, 2024 by divakar-amd Loading…
4
[Core] Implementing disaggregated prefilling, and caching KV cache in CPU/disk/database. ready ONLY add when PR is ready to merge/full CI is needed
#8498 opened Sep 16, 2024 by KuntaiDu Loading…
9
[Model] Refactor BLIP/BLIP-2 to support composite model loading ready ONLY add when PR is ready to merge/full CI is needed
#8407 opened Sep 12, 2024 by DarkLight1337 Loading…
7
[Core] Multi-Step + Single Step Prefills via Chunked Prefill code path ready ONLY add when PR is ready to merge/full CI is needed
#8378 opened Sep 11, 2024 by varun-sundar-rabindranath Loading…
@comaniac
80
Add output streaming support to multi-step + async ready ONLY add when PR is ready to merge/full CI is needed
#8335 opened Sep 10, 2024 by alexm-neuralmagic Loading…
19
[do-not-merge][CI/Build] Buildkite pipeline generator ready ONLY add when PR is ready to merge/full CI is needed
#8324 opened Sep 10, 2024 by khluu Loading…
33
[Bugfix] Fix LongRoPE bug ready ONLY add when PR is ready to merge/full CI is needed
#8254 opened Sep 7, 2024 by garg-amit Loading…
1
11
[BugFix] Propagate 'trust_remote_code' setting in internvl and minicpmv ready ONLY add when PR is ready to merge/full CI is needed
#8250 opened Sep 6, 2024 by zifeitong Loading…
30
[Kernel] Build flash-attn from source ready ONLY add when PR is ready to merge/full CI is needed
#8245 opened Sep 6, 2024 by ProExpertProg Loading…
1
14
[BugFix] Fix metrics error for --num-scheduler-steps > 1 ready ONLY add when PR is ready to merge/full CI is needed
#8234 opened Sep 6, 2024 by yuleil Loading…
1
4
[Misc] Upgrade vllm-flash-attn to v2.6.2 ready ONLY add when PR is ready to merge/full CI is needed
#8211 opened Sep 5, 2024 by WoosukKwon Loading…
1
[Model] Adding Granite MoE. ready ONLY add when PR is ready to merge/full CI is needed
#8206 opened Sep 5, 2024 by shawntan Loading…
8
[Misc] add iteration_tokens metric ready ONLY add when PR is ready to merge/full CI is needed
#8140 opened Sep 4, 2024 by LucasWilkinson Loading…
2
[Bugfix] Fix bug in detokenizer.py ready ONLY add when PR is ready to merge/full CI is needed
#8112 opened Sep 3, 2024 by cafeii Loading…
4
[Core][Kernel][Misc] Support external swapper for vllm ready ONLY add when PR is ready to merge/full CI is needed
#8018 opened Aug 30, 2024 by zeroorhero Loading…
7
ProTip! Adding no:label will show everything without a label.