-
Notifications
You must be signed in to change notification settings - Fork 242
Pull requests: SemiAnalysisAI/InferenceX
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
[experimental][AgentX]: B300 deterministic AIPerf
experimental
full-sweep-enabled
#2408
opened Jul 29, 2026 by
cquil11
Collaborator
Loading…
CollectiveX: one nccl-ep handle per group, and restore the low-latency rows
#2407
opened Jul 29, 2026 by
Oseltamivir
Collaborator
Loading…
deprcate even more stuff, cut compute cost by 60%
#2406
opened Jul 29, 2026 by
functionstackx
Collaborator
Loading…
6 tasks done
[KimiK3][AgentX]: GB200 Simple CPU KV offload / [KimiK3][AgentX]：GB200 Simple CPU KV 卸载
full-sweep-enabled
#2404
opened Jul 29, 2026 by
cquil11
Collaborator
Loading…
12 tasks done
[AMD][AgentX] Kimi-K3 FP4 MI355X agentX vLLM
agentx
AgentX benchmarks, recipes, and infrastructure
AMD
#2403
opened Jul 29, 2026 by
hyukjlee
Collaborator
Loading…
Add opt-in native two-node MI300X path for Kimi K3 AgentX / 新增 Kimi K3 AgentX 的 MI300X 双节点 native 路径
#2401
opened Jul 28, 2026 by
edwingao28
Collaborator
•
Draft
1 of 6 tasks
Kimi K3 B200 Day 0 AgentX Pipeline Parallelism (rebased onto main; non-DSpark as it doesn't work with Pipeline yet, offloading & TP16 and DEP8PP2 to be done in followup PR)
full-sweep-fail-fast
#2400
opened Jul 28, 2026 by
functionstackx
Collaborator
Loading…
[goldenAL] update golden AL for MiniMax-M3 Eagle head to Inferact/MiniMax-M3-EAGLE3-GQA- #2393
#2394
opened Jul 28, 2026 by
xinli-sw
Collaborator
Loading…
[AgentX] K3 b300 offload
full-sweep-enabled
#2392
opened Jul 28, 2026 by
xinli-sw
Collaborator
Loading…
Add DSv4-Pro FP4 B200 disaggregated Dynamo vLLM MTP config
full-sweep-fail-fast
#2390
opened Jul 28, 2026 by
hshrivastava-droid
Collaborator
Loading…
[AMD] [AGENTX] dsv4-fp4-mi355x-vllm-agentic: DEP tuning + LMCache refresh (MTP variant)
agentx
AgentX benchmarks, recipes, and infrastructure
AMD
full-sweep-fail-fast
#2381
opened Jul 28, 2026 by
seungrokj
Collaborator
Loading…
3 tasks
[Klaud Cold] kimik3-fp4-b300-vllm-agentic-dspark: Kimi-K3 B300 DSpark agentic recipe
non-canary-full-sweep-enabled
Run the full sweep without the canary gate (full search space, no trim)
skip_queue
#2380
opened Jul 28, 2026 by
Oseltamivir
Collaborator
Loading…
[AMD] [AGENTX] dsv4-fp4-mi355x-vllm-agentic: DEP tuning + LMCache refresh
agentx
AgentX benchmarks, recipes, and infrastructure
AMD
#2378
opened Jul 28, 2026 by
seungrokj
Collaborator
Loading…
3 tasks
[Klaud Cold][agentic experiment][Variant F] Kimi-K3 B200 agg TP8xPP2 agentic — SimpleCPU offload + prefix-cache retention / Kimi-K3 B200 聚合式 TP8xPP2 智能体实验——SimpleCPU 卸载 + 前缀缓存留存
full-sweep-fail-fast
#2372
opened Jul 28, 2026 by
functionstackx
Collaborator
Loading…
[NV] Add H200 DeepSeek-V4-Pro AgentX recipes
full-sweep-enabled
NVIDIA
#2364
opened Jul 27, 2026 by
csahithi
Collaborator
Loading…
feat: add DSV4 B300 Dynamo-SGLang MTP configuration / 新增 DSV4 B300 Dynamo-SGLang MTP 配置
full-sweep-enabled
#2363
opened Jul 27, 2026 by
RohitNagraj
Collaborator
Loading…
feat: add DSV4 FP4 B300 Dynamo-SGLang STP configuration / 新增 DSV4 B300 Dynamo-SGLang STP 配置
full-sweep-enabled
#2362
opened Jul 27, 2026 by
RohitNagraj
Collaborator
Loading…
Add Kimi K2.6 NVFP4 B200 Dynamo vLLM configuration / 新增 Kimi K2.6 NVFP4 B200 Dynamo vLLM 配置
full-sweep-enabled
#2360
opened Jul 27, 2026 by
RohitNagraj
Collaborator
Loading…
[KimiK3][AgentX]: GB200 KimiK3 Day 0 support / [KimiK3][AgentX]：GB200 KimiK3 Day 0 支持
full-sweep-enabled
#2354
opened Jul 27, 2026 by
cquil11
Collaborator
Loading…
[KimiK3][AgentX]: H200 KimiK3 Day 0 support / [KimiK3][AgentX]：H200 KimiK3 Day 0 支持
#2353
opened Jul 27, 2026 by
cquil11
Collaborator
Loading…
chore(deps): bump actions/checkout from 7.0.0 to 7.0.1 in the github-actions group
dependencies
Pull requests that update a dependency file
github_actions
Pull requests that update GitHub Actions code
#2347
opened Jul 27, 2026 by
dependabot
Bot
Loading…
[AMD][AgentX] Add DeepSeek-V4-Pro FP4 MI355X ATOM + MTP agentic-coding recipe
agentx
AgentX benchmarks, recipes, and infrastructure
AMD
#2346
opened Jul 27, 2026 by
seungrokj
Collaborator
Loading…
2 tasks
Previous Next
ProTip!
Exclude everything labeled
bug with -label:bug.