-
Notifications
You must be signed in to change notification settings - Fork 229
Pull requests: SemiAnalysisAI/InferenceX
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
[LMCache] Add LMCache arm on tuned DSV4 FP4 B300 vLLM AgentX recipe / 在调优后的 DSV4 FP4 B300 vLLM AgentX 配方上新增 LMCache 分支
agentx
AgentX benchmarks, recipes, and infrastructure
full-sweep-enabled
#2232
opened Jul 16, 2026 by
ApostaC
Collaborator
Loading…
3 of 10 tasks
[LMCache] Add LMCache arm on tuned DSV4 FP4 B200 vLLM AgentX recipe
agentx
AgentX benchmarks, recipes, and infrastructure
full-sweep-enabled
#2231
opened Jul 16, 2026 by
ApostaC
Collaborator
Loading…
3 of 10 tasks
[WIP] feat(minimaxm3): add B200 disaggregated configs
full-sweep-fail-fast
#2230
opened Jul 15, 2026 by
xinli-sw
Collaborator
Loading…
Add Kimi K2.6 NVFP4 B300 EAGLE3 AgentX benchmark / 新增 Kimi K2.6 NVFP4 B300 EAGLE3 AgentX 基准测试
agentx
AgentX benchmarks, recipes, and infrastructure
full-sweep-enabled
NVIDIA
#2228
opened Jul 15, 2026 by
cquil11
Collaborator
Loading…
fix(agentx): rerun DeepSeek-V4 B300 SGLang with OOM headroom / 修复 AgentX：为 DeepSeek-V4 B300 SGLang 重跑预留 OOM 余量
agentx
AgentX benchmarks, recipes, and infrastructure
#2226
opened Jul 15, 2026 by
cquil11
Collaborator
Loading…
[AgentX] vLLM DeepSeek-V4 B200 aggregate / vLLM DeepSeek-V4 B200 聚合配置
agentx
AgentX benchmarks, recipes, and infrastructure
full-sweep-enabled
NVIDIA
#2224
opened Jul 15, 2026 by
cquil11
Collaborator
Loading…
[AgentX] vLLM DeepSeek-V4 B300 aggregate / vLLM DeepSeek-V4 B300 聚合配置
agentx
AgentX benchmarks, recipes, and infrastructure
full-sweep-enabled
NVIDIA
#2225
opened Jul 15, 2026 by
cquil11
Collaborator
Loading…
[NV] Add gb300 nvfp4 qwen3.5 MTP 8k1k recipes
full-sweep-enabled
#2218
opened Jul 15, 2026 by
richardhuo-nv
Collaborator
Loading…
[WIP][AMD][MI355x] dsv4-fp4-mi355x-vllm-agentic, install latest lmcache
agentx
AgentX benchmarks, recipes, and infrastructure
AMD
full-sweep-enabled
#2216
opened Jul 14, 2026 by
ruizhang0101
Collaborator
Loading…
[AMD] Tune Kimi-K2.5 MXFP4 MI355X vLLM toward ATOM parity / 调优 Kimi-K2.5 MXFP4 MI355X vLLM，缩小与 ATOM 的性能差距
#2213
opened Jul 14, 2026 by
Fangzhou-Ai
Collaborator
•
Draft
8 tasks
Align result metadata validation with Pydantic schemas / 统一结果元数据与 Pydantic 模式验证
#2208
opened Jul 14, 2026 by
cquil11
Collaborator
Loading…
Add DSV4 GB300 TRT disaggregated offline benchmark / 新增 DSV4 GB300 TRT 分离式离线基准测试
#2207
opened Jul 14, 2026 by
Ankur-singh
Collaborator
Loading…
[NV] Add dynamo trtllm gb300 qwen3.5 nvfp4 recipes for 8k1k and 1k1k
full-sweep-enabled
#2206
opened Jul 14, 2026 by
richardhuo-nv
Collaborator
Loading…
Switch Qwen3.5 FP4 SGLang configs to nvidia/Qwen3.5-397B-A17B-NVFP4-V2
full-sweep-enabled
#2205
opened Jul 14, 2026 by
hshrivastava-droid
Collaborator
Loading…
[AMD][MI35X] Qwen3.5-fp4 SGLang single-node benchmark env update
AMD
full-sweep-fail-fast
#2201
opened Jul 14, 2026 by
yichiche
Collaborator
Loading…
[DNM][AMD][MI355X] DSv4 SGLang agentic: v0.5.14 image, router, HiCache + MTP
agentx
AgentX benchmarks, recipes, and infrastructure
AMD
bug
Something isn't working
full-sweep-enabled
#2200
opened Jul 14, 2026 by
seungrokj
Collaborator
Loading…
3 tasks
[DSv4][bring-up] FP8 H100 Dynamo-vLLM multinode benchmark / DSv4 FP8 H100 Dynamo-vLLM 多节点基准测试（bring-up）
full-sweep-fail-fast
#2196
opened Jul 14, 2026 by
Oseltamivir
Collaborator
Loading…
[DSv4][bring-up] FP8 MI325X SGLang benchmark / DSv4 FP8 MI325X SGLang 基准（bring-up）
full-sweep-fail-fast
#2195
opened Jul 14, 2026 by
Oseltamivir
Collaborator
Loading…
[DSv4] Re-run GB300 dynamo-trt configs / 重跑 GB300 dynamo-trt 配置
#2193
opened Jul 14, 2026 by
Oseltamivir
Collaborator
Loading…
[AMD] MiniMax-M3 MXFP8 MI355X vLLM: gate AITER sparse PA to 8k1k high-conc + native linear default / MiniMax-M3 MXFP8 MI355X vLLM：AITER 稀疏分页注意力门控至 8k1k 高并发 + 默认原生 linear
full-sweep-fail-fast
#2187
opened Jul 14, 2026 by
hongxiayang
Collaborator
Loading…
3 of 5 tasks
[DoNotMerge] Add MiniMax M3 vLLM B300 disagg EAGLE
#2182
opened Jul 13, 2026 by
hjjq
Collaborator
Loading…
[NV] Kimi-K2.6 NVFP4 B300 disagg multinode dynamo-vllm benchmarks / [NV] Kimi-K2.6 NVFP4 B300 分离式多节点 dynamo-vllm 基准测试
full-sweep-enabled
#2181
opened Jul 13, 2026 by
RohitNagraj
Collaborator
Loading…
4 of 10 tasks
Previous Next
ProTip!
Updated in the last three days: updated:>2026-07-12.