The warmup custom op call hit cudaErrorIllegalAddress because our custom op GEMM implementation doesn't match the runner's call convention. Skip warmup for now — MoE kernel warmup handles CuTeDSL JIT cleanup.
The warmup custom op call hit cudaErrorIllegalAddress because our custom op GEMM implementation doesn't match the runner's call convention. Skip warmup for now — MoE kernel warmup handles CuTeDSL JIT cleanup.