Skip to content

Pull requests: ROCm/ATOM

New pull request New
127 Open 810 Closed
127 Open 810 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

[EP] fix EP for modular
#1087 opened Jun 5, 2026 by amd-ruitang3 Contributor Loading…
1 task
5
[refactor] Unify engine utility command dispatch into EngineUtilityHandler
#1086 opened Jun 4, 2026 by sijyang Contributor Loading…
[feat] Add hidden states extraction for TorchSpec training
#1085 opened Jun 4, 2026 by sijyang Contributor Loading…
[ci][SGL]change sglang-atom ci ds fp4 model to ds-fp4-v2
#1083 opened Jun 4, 2026 by zhuyuhua-v Collaborator • Draft
docker: install MoRI from PyPI nightly wheel instead of source build
#1081 opened Jun 4, 2026 by jhchouuu Loading…
3
enable index cache feature for dsa model
#1080 opened Jun 4, 2026 by XiaobingSuper Contributor Loading…
1 task done
2
fix: support DeepSeek-V4-Flash-FP8 on gfx942 device
#1077 opened Jun 4, 2026 by junna2016 Contributor Loading…
2
[ATOM SGLang] Fix DP EP fail
#1072 opened Jun 4, 2026 by ZhiweiYan-96 Contributor • Draft
1 task
[atom-vllm] remove benchmark override
#1070 opened Jun 4, 2026 by zejunchen-zejun Collaborator • Draft
gpt-oss WAs + moe a8w4 gemm support
#1067 opened Jun 4, 2026 by ahmed-bsod Loading…
4
[vLLM-ATOM] DPEP enablement sync main
#1066 opened Jun 4, 2026 by kliuae Contributor Loading…
1 task
CI: Use linux-atom-do-mi350x-8 in model accuracy tests
#1062 opened Jun 4, 2026 by gyohuangxin Member Loading…
2
[vLLM Plugin] Enable dsv4
#1060 opened Jun 3, 2026 by whx-sjtu Contributor • Draft
1 task
[ATOM SGL] fix v2 model launch issue
#1059 opened Jun 3, 2026 by ZhiweiYan-96 Contributor Loading…
2
perf: add preshuffle logic for HIP fused_moe
#1057 opened Jun 3, 2026 by ftyghome Loading…
4
[Unfinished] fix pd_disaggregation hang when enable TBO
#1056 opened Jun 3, 2026 by ZhangLirong-amd Contributor Loading…
1 task
2
[Feat] Support input_rmsnorm_quant & qknorm_quant fused for Atom dpsk v2 model
#1055 opened Jun 3, 2026 by qichu-yun Contributor Loading…
1 task
2
[atom-vllm MLA] add back missing concat+quant kernel from original vllm
#1054 opened Jun 3, 2026 by zejunchen-zejun Collaborator • Draft
[SGL][feat] upgrade sglang from v0.5.10 to v0.5.12
#1052 opened Jun 3, 2026 by zhuyuhua-v Collaborator Loading…
1 task
[atom-sglang][Fix] fix error that mtp+dp can't use speculative-num-steps> 1
#1046 opened Jun 3, 2026 by ZLkanyo009 Contributor Loading…
1
gpt-oss WAs + triton moe a8w4 support
#1045 opened Jun 3, 2026 by ahmed-bsod • Draft
MoE AITER Triton Kernels Integration
#1044 opened Jun 2, 2026 by amirumoAMD Contributor Loading…
1 task done
2
Remove Triton build stage from Dockerfile
#1035 opened Jun 2, 2026 by gyohuangxin Member • Draft
2
[feature] Support deepseekv4 online quantization
#1027 opened Jun 2, 2026 by haoyangli0109 Contributor Loading…
1
Fix LMCache offload reload handoff
#1026 opened Jun 2, 2026 by yhl-amd Contributor • Draft
ProTip! Follow long discussions with comments:>50.