Skip to content

Pull requests: SemiAnalysisAI/InferenceX

New pull request New
44 Open 698 Closed
44 Open 698 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

Enable shuffled KV cache layout for MiniMax vLLM
#1199 opened Apr 27, 2026 by jiacao-amd Loading…
1
[AMD/ROCM] Update gptoss-fp4-mi355x-atom config AMD
#1195 opened Apr 27, 2026 by seungrokj Collaborator Loading…
1 task
3
[AMD/ROCM] Update minimaxm2.5-fp8-mi355x-atom config AMD
#1194 opened Apr 27, 2026 by seungrokj Collaborator Loading…
1 task
2
Add vLLM DSv4 FP8 MI355X benchmark (vllm#40889) full-sweep-enabled
#1188 opened Apr 26, 2026 by Oseltamivir Collaborator Loading…
3 of 4 tasks
1
dsv4-b300-sglang: conc=2048 mega_moe deepep recipe
#1179 opened Apr 26, 2026 by yhyang201 Collaborator Loading…
3 tasks
@Qiaolin-Yu
32
gb300 1k1k sglang sweep-enabled
#1169 opened Apr 26, 2026 by Oseltamivir Collaborator Loading…
4 of 5 tasks
5
[DON'T MERGE] [NV] dsv4-fp4-gb200-dynamo-vllm sweep-enabled
#1163 opened Apr 26, 2026 by Ankur-singh Collaborator Loading…
2
Day 0 DeepSeek V4 Pro FP4 GB200 disaggregated SGLang benchmarks sweep-enabled
#1157 opened Apr 25, 2026 by Oseltamivir Collaborator Loading…
4 of 5 tasks
2
Day 0 GB300 DeepSeek-V4-Pro FP4 vLLM disagg sweep-enabled
#1150 opened Apr 25, 2026 by Oseltamivir Collaborator Loading…
3 tasks
3
[AMD/ROCM] Qwen3.5-397B-A17B BF16 MI355X Atom benchmarks
#1149 opened Apr 25, 2026 by seungrokj Collaborator Loading…
2
[NVIDIA] chore: B200 single node DeepSeek v4 SGLang MTP NVIDIA sweep-enabled
#1145 opened Apr 24, 2026 by cquil11 Collaborator Loading…
1 task
6
Add H100 config: dsv4-fp8-dynamo-vllm (DeepSeek-V4-Pro multinode disagg) sweep-enabled
#1142 opened Apr 24, 2026 by Oseltamivir Collaborator Loading…
1
[draft] minimax & kimi Amd/vllm disagg mvp dev
#1141 opened Apr 24, 2026 by ichbinblau Collaborator Loading…
Add DeepSeek-V4-Pro SGLang aggregated GB200 benchmarks (NVIDIA srt-slurm PR #69) sweep-enabled
#1137 opened Apr 24, 2026 by Oseltamivir Collaborator Loading…
3 of 5 tasks
4
[AMD/ROCM] atom qwen3.5 fp4 on mi355x AMD
#1133 opened Apr 24, 2026 by seungrokj Collaborator Loading…
1 task
5
[AMD/ROCM] atom glm5 fp8 on mi355x AMD
#1126 opened Apr 24, 2026 by seungrokj Collaborator Loading…
2 tasks
6
[AMD/ROCM] GLM5.1 FP8 MTP Support on MI355X AMD
#1122 opened Apr 23, 2026 by ajith-sirra-amd Contributor Loading…
5
[WIP] Allow overriding srt-slurm repo/ref at the launcher level
#1118 opened Apr 22, 2026 by Oseltamivir Collaborator Loading…
[AMD/ROCm] Add Kimi-K2.5 FP4 vLLM Eagle3 speculative decoding config for MI355X
#1116 opened Apr 22, 2026 by chunfangamd Collaborator • Draft
@chunfangamd
9
Add new feature for kimi k2.5 mtp support
#1115 opened Apr 22, 2026 by haic0 Collaborator • Draft
6
[AMD/Hyperloom] Tune dsr1-fp8-mi355x-sglang: --num-continuous-decode-steps 4 → 8
#1109 opened Apr 21, 2026 by lishuoshuo-amd Loading…
4 tasks done
3
Add haic0 patch for AMD kimi k2.5 MTP support
#1108 opened Apr 21, 2026 by haic0 Collaborator • Draft
6
[DISCLAIMER - EXPERIMENTAL][First iteration of MVP v0.1, shipping second iteration next week] [WIP] feat: agentic benchmark
#1103 opened Apr 20, 2026 by cquil11 Collaborator • Draft
19
[sglang broken] Add MI355X config: qwen3.5-fp4-sglang-mtp vllm/sglang release broken -need to wait
#1078 opened Apr 18, 2026 by functionstackx Contributor Loading…
3 of 4 tasks
5
[vllm broken - waiting for 0.20] Add B300 config: kimi-k2.5-int4-vllm vllm/sglang release broken -need to wait
#1071 opened Apr 17, 2026 by cquil11 Collaborator Loading…
2 tasks
8
ProTip! Follow long discussions with comments:>50.