PR Test (AMD) · Workflow runs · sgl-project/sglang · GitHub

Actions

All workflows
Workflows
- Build and Push Development Docker Images Build and Push Development Docker Images
- CI Failure Monitor CI Failure Monitor
- List Active PR Runs List Active PR Runs
- Nightly Test (Nvidia) Nightly Test (Nvidia)
- PR Test PR Test
- PR Test (AMD) PR Test (AMD)
- .github/workflows/nightly-test-amd-rocm720.yml .github/workflows/nightly-test-amd-rocm720.yml
- .github/workflows/pr-gate.yml .github/workflows/pr-gate.yml
- .github/workflows/pr-test-amd-rocm720.yml .github/workflows/pr-test-amd-rocm720.yml
- 5090 GPU Test 5090 GPU Test
- AMD CI Job Monitor AMD CI Job Monitor
Management
- Caches
- Deployments

PR Test (AMD)

Actions

Loading...
Loading

pr-test-amd.yml

50,648 workflow runs

50,648 workflow runs

Add piecewise cuda graph for Qwen3-Next FP8 flashinfer_trtllm moe backend PR Test (AMD) #57337: Pull request #18184 synchronize by xiaoweiw-nv

13s xiaoweiw-nv:piecewise_cudagraph_q3n_fp8

xiaoweiw-nv:piecewise_cudagraph_q3n_fp8

13s

[Omni] Optimize Qwen3_Omni_Thinker PR Test (AMD) #57336: Pull request #18185 opened by yuan-luo

14s antgroup:optimize_qwen3_omni_thinker

antgroup:optimize_qwen3_omni_thinker

14s

[Feature]Introduce DeepEP's Per-Expert-overlap(PEO) capability into SGLang. PR Test (AMD) #57335: Pull request #13442 synchronize by zhihui1084

18s zhihui1084:support_PEO

zhihui1084:support_PEO

18s

Add piecewise cuda graph for Qwen3-Next FP8 flashinfer_trtllm moe backend PR Test (AMD) #57334: Pull request #18184 opened by xiaoweiw-nv

39s xiaoweiw-nv:piecewise_cudagraph_q3n_fp8

xiaoweiw-nv:piecewise_cudagraph_q3n_fp8

39s

Enhance Kimi K2.5 support with improved reasoning parser and function call detector (#18086) PR Test (AMD) #57333: Pull request #18178 synchronize by cynial

12s cynial:feat/k25-tool-call-streaming

cynial:feat/k25-tool-call-streaming

12s

[2/N] Quantization Refactor: Compressed tensors MoE schemes PR Test (AMD) #57332: Pull request #17503 synchronize by TamirBaydasov

In progress TamirBaydasov:compressed_tensors_moe_schemes

TamirBaydasov:compressed_tensors_moe_schemes

In progress

[Quantization] Support online FP8 to MXFP4 requantization on AMD GPUs PR Test (AMD) #57331: Pull request #18182 synchronize by fxmarty-amd

15s fxmarty-amd:fp8-to-mxfp4

fxmarty-amd:fp8-to-mxfp4

15s

[MXFP4] Support online MXFP4 quantization in quark_w4a4_mxfp4.py PR Test (AMD) #57330: Pull request #18005 synchronize by fxmarty-amd

17s fxmarty-amd:online-quark-mxfp4

fxmarty-amd:online-quark-mxfp4

17s

Enhance Kimi K2.5 support with improved reasoning parser and function call detector (#18086) PR Test (AMD) #57329: Pull request #18178 synchronize by cynial

15s cynial:feat/k25-tool-call-streaming

cynial:feat/k25-tool-call-streaming

15s

[Quantization] Support online FP8 to MXFP4 requantization on AMD GPUs PR Test (AMD) #57328: Pull request #18182 synchronize by fxmarty-amd

33s fxmarty-amd:fp8-to-mxfp4

fxmarty-amd:fp8-to-mxfp4

33s

Add Qwen3-TTS text-to-speech model support PR Test (AMD) #57327: Pull request #18183 opened by qianjun22

18s qianjun22:add-qwen3-tts-support

qianjun22:add-qwen3-tts-support

18s

[Quantization] Support online FP8 to MXFP4 requantization on AMD GPUs PR Test (AMD) #57326: Pull request #18182 opened by fxmarty-amd

13s fxmarty-amd:fp8-to-mxfp4

fxmarty-amd:fp8-to-mxfp4

13s

[DeepGemm] Add a flag for fast warmup PR Test (AMD) #57325: Pull request #18111 synchronize by Fridge003

46m 8s baizhou/deepgemm

baizhou/deepgemm

46m 8s

Add basic function testcases PR Test (AMD) #57324: Pull request #18180 opened by Sugar920

18s shun8686:pr-commit

shun8686:pr-commit

18s

[diffusion] feat: support parallel wan-vae encode PR Test (AMD) #57323: Pull request #18179 opened by nono-Sang

15s nono-Sang:parallel_wanvae_encode

nono-Sang:parallel_wanvae_encode

15s

Add Nemotron 3 Nano tests PR Test (AMD) #57322: Pull request #18119 synchronize by shaharmor98

16s shaharmor98:feat/add-nemotron-3-nano-tests

shaharmor98:feat/add-nemotron-3-nano-tests

16s

Add Nemotron 3 Nano tests PR Test (AMD) #57321: Pull request #18119 synchronize by shaharmor98

15s shaharmor98:feat/add-nemotron-3-nano-tests

shaharmor98:feat/add-nemotron-3-nano-tests

15s

Add Nemotron 3 Nano tests PR Test (AMD) #57320: Pull request #18119 synchronize by shaharmor98

15s shaharmor98:feat/add-nemotron-3-nano-tests

shaharmor98:feat/add-nemotron-3-nano-tests

15s

[Diffusion] [NPU] w8a8 Wan2.2-I2V-A14B-Diffusers modelslim quantization support PR Test (AMD) #57319: Pull request #17996 synchronize by OrangeRedeng

19s OrangeRedeng:diffusion_npu_w8a8_support

OrangeRedeng:diffusion_npu_w8a8_support

19s

[Diffusion] Downgrade prompt log from info to debug. PR Test (AMD) #57318: Pull request #17813 synchronize by mickqian

1h 39m 56s Evrard-Nil:fix/diffusion-do-not-log-prompt-in-info-logs

Evrard-Nil:fix/diffusion-do-not-log-prompt-in-info-logs

1h 39m 56s

Support for multi-layer Eagle3 in SGLang PR Test (AMD) #57317: Pull request #18112 synchronize by xiaoxi-s

13s xiaoxi-s:eagle3-draft-decoder-with-multiple-layers

xiaoxi-s:eagle3-draft-decoder-with-multiple-layers

13s

[NPU] [BUGFIX] Test ascend memory consumption.py fix PR Test (AMD) #57316: Pull request #17995 synchronize by OrangeRedeng

54m 14s OrangeRedeng:test_ascend_memory_consumption.py-bugfix

OrangeRedeng:test_ascend_memory_consumption.py-bugfix

54m 14s

Speculative Decoding: Implement basic decoder as the additional layers of the draft decoder PR Test (AMD) #57315: Pull request #18175 opened by xiaoxi-s

15s xiaoxi-s:eagle3-multiple-draft-layers-basic

xiaoxi-s:eagle3-multiple-draft-layers-basic

15s

[Bugfix] Catch errors when DeepSeek-V3.2 generates malformed JSON PR Test (AMD) #57314: Pull request #18174 opened by Muqi1029

18s Muqi1029:bugfix/catch_error

Muqi1029:bugfix/catch_error

18s

[diffusion] operator: unify rotary embedding impl PR Test (AMD) #57313: Pull request #18164 synchronize by triple-mu

16s triple-mu:triplemu/unified-apply-rope

triple-mu:triplemu/unified-apply-rope

16s