Skip to content

Pull requests: ROCm/ATOM

New pull request New
141 Open 993 Closed
141 Open 993 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
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

fix(minimax): restore qkv<=256 shape guard + harden compile-cache frozen-path guard
#1300 opened Jun 20, 2026 by sunway513 Collaborator Loading…
1
fix(minimax): restore qkv<=256 shape guard + harden compile-cache frozen-path guard (v0.1.5)
#1299 opened Jun 20, 2026 by sunway513 Collaborator Loading…
Atom shuffle scale
#1297 opened Jun 20, 2026 by yadaish Loading…
1 task
Modify model cache mount direction for atom-sgl-accuracy and resume 'clean up contrainers and workspace'
#1293 opened Jun 19, 2026 by junyyang-amd Contributor Loading…
[Triton] [GFX12] Add gluon pa_decode_sparse from AITER
#1290 opened Jun 19, 2026 by k50112113 Collaborator Loading…
Feat: RTP-LLM plugin GLM5-FP8 support
#1289 opened Jun 18, 2026 by zhaoan12-prc Contributor Loading…
12
[ci][mesh] cancel slurm job before exit
#1288 opened Jun 18, 2026 by wanzhenchn Contributor Loading…
1 task
[atom-sgl-ci] Use workspace DOCKER_CONFIG and BUILDKIT_TMPDIR on self-hosted runner
#1285 opened Jun 18, 2026 by junyyang-amd Contributor Loading…
1
wip
#1284 opened Jun 18, 2026 by haoyangli0109 Contributor Loading…
[vllm-atom] support eagle3 for M3
#1283 opened Jun 18, 2026 by whx-sjtu Contributor • Draft
merge build sparse block into topk
#1282 opened Jun 18, 2026 by ganyi1996ppo Contributor Loading…
1 task
mla
#1280 opened Jun 18, 2026 by feifei14119 Loading…
1 task
10
Add mxfp8 x mxfp4 Triton MoE for DSv4
#1277 opened Jun 18, 2026 by azaidy Contributor Loading…
support TBO decode in Deepseek v4
#1275 opened Jun 18, 2026 by ZhangLirong-amd Collaborator Loading…
1 task
7
support norm fp32 and bf16 output
#1273 opened Jun 18, 2026 by XiaobingSuper Contributor • Draft
1 task
Fpz/mixed mla dispatch
#1271 opened Jun 18, 2026 by jiayyu Contributor • Draft
1 task
[Triton] DSV4 replace einsum with Triton BMM
#1270 opened Jun 17, 2026 by k50112113 Collaborator Loading…
fix(attn): enable PA ASM for MHA decode
#1269 opened Jun 17, 2026 by yhl-amd Contributor Loading…
[atom][M3] add gemma+allreduce fusion for m3
#1261 opened Jun 17, 2026 by zhuyuhua-v Collaborator • Draft
1 task
fix(sglang): skip sparse MLA fast metadata for unsupported heads
#1252 opened Jun 17, 2026 by qichu-yun Contributor Loading…
1 task
[ci][mesh] add model P/D benchmark workflow
#1244 opened Jun 17, 2026 by wanzhenchn Contributor Loading…
pa draft
#1242 opened Jun 17, 2026 by vgokhale • Draft
2
aiter asm pa have right acc
#1241 opened Jun 17, 2026 by ganyi1996ppo Contributor Loading…
1 task
Add MiniMax-M3 BF16 support
#1238 opened Jun 16, 2026 by wuhuikx Collaborator • Draft
feat(v4): prefix chunk and chunk long prefill related
#1236 opened Jun 16, 2026 by jiayyu Contributor Loading…
1 task
6
ProTip! Add no:assignee to see everything that’s not assigned.