Workflow runs · vllm-project/vllm · GitHub

Actions

All workflows
Workflows
- Add label on auto-merge enabled Add label on auto-merge enabled
- Close inactive issues and PRs Close inactive issues and PRs
- CodeQL CodeQL
- Copilot code review Copilot code review
- Copilot coding agent Copilot coding agent
- Dependabot Updates Dependabot Updates
- Label issues based on keywords Label issues based on keywords
- macOS Apple Silicon Smoke Test macOS Apple Silicon Smoke Test
- New PR Bot New PR Bot
- pre-commit pre-commit
Management
- Caches

All workflows

Actions

Loading...
Loading

Showing runs from all workflows

303,625 workflow runs

303,625 workflow runs

[XPU] [Quant] rename mxfp8_e4m3_quantize and add xpu backend implementation pre-commit #110286: Pull request #38682 synchronize by zufangzhu

In progress zufangzhu:zufang/refactor_mxfp8_quant

zufangzhu:zufang/refactor_mxfp8_quant

In progress

[XPU] [Quant] rename mxfp8_e4m3_quantize and add xpu backend implementation pre-commit #110285: Pull request #38682 synchronize by zufangzhu

3m 23s zufangzhu:zufang/refactor_mxfp8_quant

zufangzhu:zufang/refactor_mxfp8_quant

3m 23s

[XPU] Enable sequence parallel support for XPU pre-commit #110284: Pull request #38608 synchronize by chaojun-zhang

5m 20s chaojun-zhang:seq_parallel

chaojun-zhang:seq_parallel

5m 20s

[XPU] Enable sequence parallel support for XPU pre-commit #110283: Pull request #38608 synchronize by chaojun-zhang

2m 59s chaojun-zhang:seq_parallel

chaojun-zhang:seq_parallel

2m 59s

[Bugfix] install failed on macbook(cpu) pre-commit #110282: Pull request #38768 opened by jesse996

6s jesse996:main

jesse996:main

6s

[Bugfix] install failed on macbook(cpu) New PR Bot #911: Pull request #38768 opened by jesse996

7s

7s

[Performance] Fuse RoPE + KV cache update for MLA backends pre-commit #110281: Pull request #35879 synchronize by ElizaWszola

3m 21s neuralmagic:fuse-mla-rope-kv-update

neuralmagic:fuse-mla-rope-kv-update

3m 21s

Support AsyncTP pattern matching for FlashInfer vllm.bmm_fp8 on Blackwell#27893 pre-commit #110280: Pull request #38177 synchronize by baonudesifeizhai

5m 55s baonudesifeizhai:AjAnubolu-fix-async-tp-flashinfer-blackwell

baonudesifeizhai:AjAnubolu-fix-async-tp-flashinfer-blackwell

5m 55s

fix: enable producer and consumer in ECConnectorBase for ec_both role pre-commit #110279: Pull request #36848 synchronize by benyebai

51s benyebai:fix/ec-connector-base

benyebai:fix/ec-connector-base

51s

Support AsyncTP pattern matching for FlashInfer vllm.bmm_fp8 on Blackwell#27893 pre-commit #110278: Pull request #38177 synchronize by baonudesifeizhai

18s baonudesifeizhai:AjAnubolu-fix-async-tp-flashinfer-blackwell

baonudesifeizhai:AjAnubolu-fix-async-tp-flashinfer-blackwell

18s

Support AsyncTP pattern matching for FlashInfer vllm.bmm_fp8 on Blackwell#27893 pre-commit #110277: Pull request #38177 synchronize by baonudesifeizhai

3m 18s baonudesifeizhai:AjAnubolu-fix-async-tp-flashinfer-blackwell

baonudesifeizhai:AjAnubolu-fix-async-tp-flashinfer-blackwell

3m 18s

[Transformers v5] Add SarvamMLAConfig to fix SarvamMLAForCausalLM (#38734) New PR Bot #910: Pull request #38767 opened by Zelys-DFKH

8s

8s

[Transformers v5] Add SarvamMLAConfig to fix SarvamMLAForCausalLM (#38734) pre-commit #110276: Pull request #38767 opened by Zelys-DFKH

8s Zelys-DFKH:fix/sarvam-mla-transformers-v5

Zelys-DFKH:fix/sarvam-mla-transformers-v5

8s

Support AsyncTP pattern matching for FlashInfer vllm.bmm_fp8 on Blackwell#27893 pre-commit #110275: Pull request #38177 synchronize by baonudesifeizhai

1m 58s baonudesifeizhai:AjAnubolu-fix-async-tp-flashinfer-blackwell

baonudesifeizhai:AjAnubolu-fix-async-tp-flashinfer-blackwell

1m 58s

Support AsyncTP pattern matching for FlashInfer vllm.bmm_fp8 on Blackwell#27893 pre-commit #110274: Pull request #38177 synchronize by baonudesifeizhai

3m 19s baonudesifeizhai:AjAnubolu-fix-async-tp-flashinfer-blackwell

baonudesifeizhai:AjAnubolu-fix-async-tp-flashinfer-blackwell

3m 19s

Support AsyncTP pattern matching for FlashInfer vllm.bmm_fp8 on Blackwell#27893 pre-commit #110273: Pull request #38177 synchronize by baonudesifeizhai

2m 52s baonudesifeizhai:AjAnubolu-fix-async-tp-flashinfer-blackwell

baonudesifeizhai:AjAnubolu-fix-async-tp-flashinfer-blackwell

2m 52s

[ROCM][Bug fix] fix aiter asm atten on hybrid models New PR Bot #909: Pull request #38766 opened by yuankaichen-amd

10s

10s

[Kernel] Fuse FP8 output quantization into merge_attn_states pre-commit #110272: Pull request #36518 synchronize by carlyou

3m 8s carlyou:fuse-fp8-quant-merge-attn-states

carlyou:fuse-fp8-quant-merge-attn-states

3m 8s

Fix multiline-format string for python 3.10 pre-commit #110271: Pull request #38739 synchronize by ProExpertProg

3m 4s neuralmagic:luka/vllm-ir/fix-format-string

neuralmagic:luka/vllm-ir/fix-format-string

3m 4s

[vLLM IR] OOT migration guide Label issues based on keywords #9124: Issue #38765 opened by ProExpertProg

8s

8s

[vLLM IR] 3/N fused_add_rms_norm and maybe_inplace pre-commit #110270: Pull request #36823 synchronize by ProExpertProg

3m 18s luka/vllm-ir/rms-norm-inplace

luka/vllm-ir/rms-norm-inplace

3m 18s

[WIP][Model Runner V2] Implement flash sampling for the draft model pre-commit #110269: Pull request #38399 synchronize by TheEpicDolphin

3m 16s TheEpicDolphin:gdelfin/mrv2-draft-flash-sample

TheEpicDolphin:gdelfin/mrv2-draft-flash-sample

3m 16s

[Kernel] Fuse FP8 output quantization into merge_attn_states pre-commit #110268: Pull request #36518 synchronize by carlyou

3m 14s carlyou:fuse-fp8-quant-merge-attn-states

carlyou:fuse-fp8-quant-merge-attn-states

3m 14s

Support AsyncTP pattern matching for FlashInfer vllm.bmm_fp8 on Blackwell#27893 pre-commit #110267: Pull request #38177 synchronize by baonudesifeizhai

3m 14s baonudesifeizhai:AjAnubolu-fix-async-tp-flashinfer-blackwell

baonudesifeizhai:AjAnubolu-fix-async-tp-flashinfer-blackwell

3m 14s

[Misc][LoRA] Add automerge weight merge for single-adapter LoRA serving pre-commit #110266: Pull request #38764 opened by bhoomit

6s bhoomit:lora-automerge-weight-merge

bhoomit:lora-automerge-weight-merge

6s