Skip to content

Actions: ROCm/vllm

Cleanup PR Body

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
96 workflow runs
96 workflow runs

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

Triton version in the base docker
Cleanup PR Body #46: Pull request #315 opened by gshtras
December 10, 2024 22:45 21s
December 10, 2024 22:45 21s
Upstream merge 24 12 09
Cleanup PR Body #45: Pull request #314 opened by gshtras
December 9, 2024 22:47 21s
December 9, 2024 22:47 21s
Setting the value for the scpecilative decoding worker class on rocm platform
Cleanup PR Body #44: Pull request #313 opened by gshtras
December 9, 2024 21:41 23s
December 9, 2024 21:41 23s
Fix max_seqlens_q/k initialization for Navi GPUs
Cleanup PR Body #43: Pull request #310 edited by hyoon1
December 7, 2024 09:27 24s
December 7, 2024 09:27 24s
Fix max_seqlens_q/k initialization for Navi GPUs
Cleanup PR Body #42: Pull request #310 edited by hyoon1
December 7, 2024 09:08 21s
December 7, 2024 09:08 21s
Update README.md
Cleanup PR Body #41: Pull request #309 opened by t-parry
December 6, 2024 18:44 20s
December 6, 2024 18:44 20s
Using ROCm6.3 release image as a base
Cleanup PR Body #40: Pull request #308 opened by gshtras
December 5, 2024 23:54 23s
December 5, 2024 23:54 23s
Fix vllm_test_utils install.
Cleanup PR Body #39: Pull request #307 opened by saienduri
December 5, 2024 18:46 23s
December 5, 2024 18:46 23s
(temp workaround for Triton bug)
Cleanup PR Body #38: Pull request #306 opened by ilia-cher
December 5, 2024 00:22 25s
December 5, 2024 00:22 25s
rm old moe tune file. Add bash script for tuning reference
Cleanup PR Body #37: Pull request #305 opened by divakar-amd
December 4, 2024 22:34 18s
December 4, 2024 22:34 18s
re-tune fp8 mixtral8x22B
Cleanup PR Body #36: Pull request #304 edited by divakar-amd
December 4, 2024 21:38 16s
December 4, 2024 21:38 16s
re-tune fp8 mixtral8x22B
Cleanup PR Body #35: Pull request #304 opened by divakar-amd
December 4, 2024 21:38 25s
December 4, 2024 21:38 25s
Always use 64 as the block size of moe_align kernel to avoid lds out of limit
Cleanup PR Body #34: Pull request #303 edited by charlifu
December 4, 2024 21:17 23s
December 4, 2024 21:17 23s
Always use 64 as the block size of moe_align kernel to avoid lds out of limit
Cleanup PR Body #33: Pull request #303 opened by charlifu
December 4, 2024 21:08 19s
December 4, 2024 21:08 19s
[vllm] Add support for FP8 in Triton FA kernel
Cleanup PR Body #32: Pull request #301 edited by ilia-cher
December 4, 2024 02:45 19s
December 4, 2024 02:45 19s
[vllm] Add support for FP8 in Triton FA kernel
Cleanup PR Body #31: Pull request #301 opened by ilia-cher
December 4, 2024 02:44 25s
December 4, 2024 02:44 25s
fused_moe configs for MI325X
Cleanup PR Body #30: Pull request #300 edited by JArnoldAMD
December 3, 2024 19:26 16s
December 3, 2024 19:26 16s
Fix type hints for cython
Cleanup PR Body #29: Pull request #299 opened by gshtras
December 3, 2024 16:44 25s
December 3, 2024 16:44 25s
Add usermarker to the develop branch
Cleanup PR Body #28: Pull request #298 opened by Lzy17
December 2, 2024 23:50 20s
December 2, 2024 23:50 20s
enable softcap and gemma2
Cleanup PR Body #27: Pull request #288 edited by hliuca
December 2, 2024 21:09 21s
December 2, 2024 21:09 21s
Upstream merge 24/11/25 and 24/12/2
Cleanup PR Body #26: Pull request #297 opened by gshtras
December 2, 2024 16:24 26s
December 2, 2024 16:24 26s
Run clang-format on develop
Cleanup PR Body #25: Pull request #296 opened by gshtras
November 27, 2024 15:24 19s
November 27, 2024 15:24 19s
add mi308 fp16|fp8 mixtral8x(7B,22B)TP=1,2,4,8
Cleanup PR Body #24: Pull request #295 opened by BruceXcluding
November 27, 2024 07:00 23s
November 27, 2024 07:00 23s
Fix correctness regression (from PR#258) in Llama-3.2-90B-Vision-Instruct-FP8-KV test
Cleanup PR Body #23: Pull request #294 opened by kkHuang-amd
November 27, 2024 03:06 25s
November 27, 2024 03:06 25s
Revert "[OPT] improve rms_norm kernel"
Cleanup PR Body #22: Pull request #293 edited by gshtras
November 26, 2024 23:37 19s
November 26, 2024 23:37 19s