High-speed Large Language Model Serving on PCs with Consumer-grade GPUs

hip gemm optimization

HIP: C Heterogeneous-Compute Interface for Portability

Examples for HIP

MindSpore is a new open source deep learning training/inference framework that could be used for mobile, edge and cloud scenarios.

Repository for the QUIK project, enabling the use of 4bit kernels for generative inference

Awesome LLM compression research papers and tools.