- π Software Engineering student at USTC
- βοΈ Currently exploring AI Infrastructure
- π Interested in GPU computing, LLM serving, and inference optimization
- π± Learning and building with CUDA, C++, Python, vLLM, SGLang, PyTorch, and TensorRT
Popular repositories Loading
-
-
-
-
nano-vllm-spec-decode
nano-vllm-spec-decode PublicSpeculative decoding (n-gram / draft model, dynamic k) and vLLM V1-style mixed batching on nano-vllm, benchmarked against vLLM on A100
Python
-
nano-vllm-triton-kernels
nano-vllm-triton-kernels PublicTriton kernels and a fused QK-norm + RoPE + paged-KV-write kernel for nano-vllm
Python
Something went wrong, please refresh the page to try again.
If the problem persists, check the GitHub status page or contact support.
If the problem persists, check the GitHub status page or contact support.