Skip to content

[Draft] use tilelang mhc on rocm#84

Draft
tjtanaa wants to merge 34 commits into
mainfrom
tilelangmhc
Draft

[Draft] use tilelang mhc on rocm#84
tjtanaa wants to merge 34 commits into
mainfrom
tilelangmhc

Conversation

@tjtanaa
Copy link
Copy Markdown
Member

@tjtanaa tjtanaa commented May 26, 2026 •
edited by github-actions Bot
Loading

Purpose

Test Plan

Test Result

Essential Elements of an Effective PR Description Checklist
  • The purpose of the PR, such as "Fix some issue (link existing issues this PR will resolve)".
  • The test plan, such as providing test command.
  • The test results, such as pasting the results comparison before and after, or e2e results
  • (Optional) The necessary documentation update, such as updating supported_models.md and examples for a new model.

tjtanaa and others added 30 commits May 25, 2026 21:11
@tjtanaa
use tilelang mhc on rocm
9676182 
Signed-off-by: tjtanaa <tunjian.tan@embeddedllm.com>
@tjtanaa
clean up mhc test
5be8268 
Signed-off-by: tjtanaa <tunjian.tan@embeddedllm.com>
@tjtanaa
add tilelang to requirements.txt
122ad44 
Signed-off-by: tjtanaa <tunjian.tan@embeddedllm.com>
@tjtanaa
add tilelang_hc_pre_norm_gemm
63d5159 
Signed-off-by: tjtanaa <tunjian.tan@embeddedllm.com>
@tjtanaa
fix prefix commit; support fuse and unfused codepath
c966a7d 
Signed-off-by: tjtanaa <tunjian.tan@embeddedllm.com>
@tjtanaa
use tilelang mhc in mtp as well
d2f3c6d 
Signed-off-by: tjtanaa <tunjian.tan@embeddedllm.com>
@tjtanaa
clean up code
9111eee 
Signed-off-by: tjtanaa <tunjian.tan@embeddedllm.com>
@tjtanaa
clean up code
fbb43d7 
Signed-off-by: tjtanaa <tunjian.tan@embeddedllm.com>
@tjtanaa
clean up code
ed3db32 
Signed-off-by: tjtanaa <tunjian.tan@embeddedllm.com>
@tjtanaa
bugfix functionality
d0534f7 
Signed-off-by: tjtanaa <tunjian.tan@embeddedllm.com>
@tjtanaa
resolve merge conflict
4610096 
Signed-off-by: tjtanaa <tunjian.tan@embeddedllm.com>
@tjtanaa
Merge remote-tracking branch 'origin/main' into tilelangmhc
6656eb3
@tjtanaa
sync main
dc01802 
Signed-off-by: tjtanaa <tunjian.tan@embeddedllm.com>
@ashwing @chaunceyjiang
[Bugfix] Map reasoning_effort to enable_thinking in chat template kwa…
52a31cc 
…rgs (vllm-project#43401)

Signed-off-by: Ashwin Giridharan <girida@amazon.com>
Signed-off-by: Chauncey <chaunceyjiang@gmail.com>
Co-authored-by: Chauncey <chaunceyjiang@gmail.com>
@zyongye
[misc] Bump cutedsl version to 4.5.2 (vllm-project#43745)
03d9cc2 
Signed-off-by: Yongye Zhu <zyy1102000@gmail.com>
@sts07142
[BugFix] HFValidationError with cloud storage URIs when HF_HUB_OFFLIN…
1654609 
…E=1 (vllm-project#39155)

Signed-off-by: Injae Ryou <injaeryou@gmail.com>
@chunyang-wen
[Docs] Fix the duplicate doc icon issue (vllm-project#43546)
49a3510 
Signed-off-by: chunyang.wen <chunyang.wen@gmail.com>
@hmellor
Fix early CUDA init (vllm-project#43791)
41688e2 
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
@jatseng-ai
[ROCm] mori: add InterNodeV1LL inter-node kernel selection via VLLM_M…
05c50c7 
…ORI_INTERNODE_KERNEL (vllm-project#41751)

Signed-off-by: jatseng-ai <jatseng@amd.com>
@cleonard530 @mikaylagawarecki @Harry-Chen
[8/n] Migrate merge_attn_states, mamba, sampler to torch stable ABI (…
284e6f5 
…continued) (vllm-project#43361)

Signed-off-by: Mikayla Gawarecki <mikaylagawarecki@gmail.com>
Signed-off-by: Chris Leonard <chleonar@redhat.com>
Co-authored-by: Mikayla Gawarecki <mikaylagawarecki@gmail.com>
Co-authored-by: Shengqi Chen <harry-chen@outlook.com>
@tjtanaa
Merge branch 'main' into tilelangmhc
77e91d6
@fallintoplace @yewentao256
[Quantization] Fix Humming RoutedExperts import (vllm-project#43540)
206b72c 
Signed-off-by: Minh Vu <vuhoangminh97@gmail.com>
Co-authored-by: Wentao Ye <44945378+yewentao256@users.noreply.github.com>
@hmellor
Remove Transformers forward/backward compatibility tests (vllm-projec…
2616f67 
…t#43785)

Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
@hmellor
Validate against some config fields being set to 0 (vllm-project#43794)
2c2c966 
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
@benchislett @NickLucche
[Bugfix][DFlash]allocate the proper number of lookahead slots (vllm-p…
7fb9c01 
…roject#43733)

Signed-off-by: Benjamin Chislett <bchislett@nvidia.com>
Signed-off-by: Benjamin Chislett <chislett.ben@gmail.com>
Co-authored-by: Nicolò Lucchesi <nicolo.lucchesi@gmail.com>
@andakai
Fix Qwen3-VL and Qwen3-omni-thinker accuracy degradation from deepsta…
5963c19 
…ck inputs under torch.compile (vllm-project#43617)

Signed-off-by: Dakai An <dakaian108@gmail.com>
@AndreasKaratzas
Add @AndreasKaratzas to CODEOWNERS (vllm-project#43740)
094124a 
Signed-off-by: Andreas Karatzas <akaratza@amd.com>
@amitz-nv
[Bugfix][Kernel] TRTLLM NVFP4 MoE chunking (vllm-project#43599)
381edde 
Signed-off-by: amitz-nv <203509407+amitz-nv@users.noreply.github.com>
@tjtanaa
add adaptation citation
97c342a 
Signed-off-by: tjtanaa <tunjian.tan@embeddedllm.com>
@tjtanaa
remove unnecessary citation
5bc61a4 
Signed-off-by: tjtanaa <tunjian.tan@embeddedllm.com>
tjtanaa added 4 commits May 27, 2026 19:47
@tjtanaa
remove unnecessary citation
685bede 
Signed-off-by: tjtanaa <tunjian.tan@embeddedllm.com>
@tjtanaa
Merge branch 'main' into tilelangmhc
b1af57d
@tjtanaa
pin tilelang version rather than relax
8f40bc1 
Signed-off-by: tjtanaa <tunjian.tan@embeddedllm.com>
@tjtanaa
Merge branch 'tilelangmhc' of https://github.com/EmbeddedLLM/vllm int…
2c539fb 
…o tilelangmhc
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@DarkLight1337 DarkLight1337 Awaiting requested review from DarkLight1337 DarkLight1337 will be requested when the pull request is marked ready for review DarkLight1337 is a code owner

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.