Pinned Loading

1. vllm-project/vllm vllm-project/vllm Public

   A high-throughput and memory-efficient inference and serving engine for LLMs

   Python 89k 20.7k

2. cuda-kernel-optimizer cuda-kernel-optimizer Public

   Evidence-driven CUDA, CUTLASS, Triton and GPU workload optimization for ChatGPT · 使用 ChatGPT 驱动 GPU workload 性能优化

   Python 7

3. Nekofish-L/mxfp6_sm120 Nekofish-L/mxfp6_sm120 Public

   High-performance MXFP6 GEMM implementation optimized for NVIDIA SM120 GPUs.

   Cuda 10 2

4. learn-inference learn-inference Public

   《大模型推理原理与优化》：面向系统/架构/后端研发工程师的模型原理入门课，目标是通俗易懂的解释推理过程，理解原理有助于系统开发/维护工作

   Ruby 1