Commit History
-
feat: add gemm_n16_k2048 workloads and baseline solution
Verified signature## Summary - Add 15 workloads (JSONL, inline tensors) for `gemm_n16_k2048` - Baseline solution and reference test - Definition JSON mirrored from flashinfer-bench Kernel: Qwen3.5-35B-A3B GEMM (N=16, K=2048) Related: flashinfer-ai/flashinfer-bench#None11d8a22