add B300 vLLM AgentX single-node MiniMax-M3 FP4 EAGLE3-GQA MTP / 新增 B300 vLLM AgentX 单节点 MiniMax-M3 FP4 EAGLE3-GQA MTP - #2328

Open

hshrivastava-droid wants to merge 20 commits into

mainfrom

minimaxm3-fp4-b300-vllm-agentic-mtp

Commits on Jul 30, 2026

Commits on Jul 31, 2026

chore(agentx): pin MiniMax-M3 sweep to AIPerf PR 31
RohitNagraj
committed
chore(agentx): pin MiniMax-M3 sweep to AIPerf ed05782 (globally anchor profiling handoff)

xinli-sw
and
claude
committed

Commits on Aug 2, 2026

chore(agentx): pin MiniMax-M3 sweep to AIPerf abf55f9 (keep idle watchdogs active across barriers)

xinli-sw
and
claude
committed
feat(minimaxm3-b300-mtp): expand AgentX search space for TP4/TP2/TP4-DRAM arms

xinli-sw
and
claude
committed
feat(minimaxm3-b300-mtp): prune search space — trim TP2, extend DRAM offload

xinli-sw
and
claude
committed