Wenhao Zhang
Community@spacecat2002 · Shanghai
Shanghai Jiao Tong University
Agent Skills by Wenhao Zhang
Showing 13 vetted skills indexed across 1 GitHub repositories.
spacecat2002sglang-diffusion-benchmark-profile
Measures diffusion denoise bottlenecks in SGLang via CUDA hotspot ranking and perf-dump comparisons.
spacecat2002sglang-diffusion-add-model
Integrate diffusion models and Diffusers pipelines into the SGLang runtime.
spacecat2002sglang-diffusion-ako4all-kernel
Optimize SGLang diffusion kernels via AKO4ALL tuning and validation workflows.
spacecat2002sglang-diffusion-performance
Select SGLang Diffusion configuration flags for generation speed and GPU memory usage.
spacecat2002write-sglang-test
Create SGLang unit, integration, kernel, or performance tests with CI registration.
spacecat2002ci-workflow-guide
Orchestrate and troubleshoot SGLang CI workflows across staged test pipelines.
spacecat2002add-jit-kernel
Add and validate lightweight CUDA JIT kernels in SGLang's jit_kernel module.
spacecat2002debug-cuda-crash
Diagnose CUDA crashes in SGLang by logging tensors at kernel API boundaries.
spacecat2002sglang-torch-profiler-analysis
Analyze SGLang torch-profiler traces to rank GPU kernels and identify optimization opportunities.
spacecat2002generate-profile
Generate validated SGLang server profiling traces with Chrome-compatible output.
spacecat2002sglang-auto-benchmark
Search SGLang serving configurations for maximum throughput under latency SLAs.
spacecat2002add-sgl-kernel
Add AOT CUDA or C++ kernels to sgl-kernel with CMake and pytest.
spacecat2002debug-distributed-hang
Diagnose hanging SGLang distributed inference jobs by locating blocked NCCL collectives and divergent rank states.