Add SGLang Qwen3-8B torch.profiler harness and per-component latency classifier#264

Closed

chhwang wants to merge 4 commits into

mainfrom

qwen3-q2-profile

Commits on Jun 12, 2026

ark-dev: Implement Q2: profile SGLang Qwen3-8B forward with torch.profiler for both prefill and decode phases; write PROFILE.md with a ranked per-component latency budget that re-ranks Q4–Q8.
ark-dev
committed
ark-dev: Implement Q2: profile SGLang Qwen3-8B forward with torch.profiler for both prefill and decode phases; write PROFILE.md with a ranked per-component latency budget that re-ranks Q4–Q8.
ark-dev
committed
ci: use --no-install-recommends for lcov to prevent OOM (exit 137)
chhwang
committed
ark-dev: Add SGLang Qwen3-8B torch.profiler harness: capture per-component latency budget for prefill and decode, write PROFILE.md with ranked breakdown and targets
ark-dev
committed