sqjian
Community@sqjian
You know, madness is a lot like gravity… sometimes all you need is a little push.”
Agent Skills by sqjian
Showing 14 vetted skills indexed across 1 GitHub repositories.
sqjiansglang-diffusion-modelopt-quant
Quantize diffusion transformer models with NVIDIA ModelOpt FP8 and NVFP4 workflows.
sqjiansglang-diffusion-benchmark-profile
Measure diffusion denoise latency and collect perf dumps for SGLang backend validation.
sqjiansglang-diffusion-add-model
Generate SGLang integration templates for Diffusers-based diffusion models.
sqjiansglang-diffusion-ako4all-kernel
Automate AKO4ALL optimization loops for SGLang diffusion kernels with benchmarking and denoise validation.
sqjiansglang-diffusion-performance
Benchmark and tune SGLang diffusion performance configurations across GPUs and VRAM budgets.
sqjianwrite-sglang-test
Write and register SGLang CI/UT tests with CustomTestCase and server fixtures.
sqjianci-workflow-guide
Orchestrate and debug SGLang CI pipelines with multi-stage gating and fast-fail checks.
sqjianadd-jit-kernel
Automate adding lightweight JIT CUDA kernels to SGLang's jit_kernel module.
sqjiandebug-cuda-crash
Log CUDA kernel API calls and inputs to diagnose SGLang crashes.
sqjiansglang-torch-profiler-analysis
Analyze SGLang torch-profiler traces for kernel overlaps and fusion opportunities.
sqjiangenerate-profile
Launch an SGLang server, validate accuracy, and generate a Chrome/Perfetto trace.
sqjiansglang-auto-benchmark
Automates SGLang benchmarking by coordinating datasets, flag sweeps, and result CSV export.
sqjianadd-sgl-kernel
Add heavyweight AOT CUDA/C++ kernels to sgl-kernel with tests and benchmarks.
sqjiandebug-distributed-hang
Diagnose distributed SGLang hangs using py-spy traces and NCCL logs.