hkeee21

PCEngine

[MLSys'23] Exploiting Hardware Utilization and Adaptive Dataflow for Efficient Sparse Convolution in 3D Point Clouds.

Byte-GLM

Efficient implementation of LLM model GLM.

ByteEngine

An LLM engine based on ByteTransformer.

chatglm-throughput

A plugin to measure the throughput of LLMs like chatGLM.

FGMS

Efficient kernel implementation for fused gather-matmul-scatter operation.

FuseSage

Multistream accelerating strategy in GraphSage

gnn-acceleration-framework-with-FPGA

including compiler to encode DGL GNN model to instructions, runtime software to transfer data and control the accelerator, and hardware verilog code that can be implemented on FPGA

lightllm

LightLLM is a Python-based LLM (Large Language Model) inference and serving framework, notable for its lightweight design, easy scalability, and high-speed performance.

LLaMA-fast

PCEngine-AE

sparse-op-test

Some test files to measure the performance of spmm, sddmm and spgemm on GPU.

vllm_qwen

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

xformers-hacked

Hackable and optimized Transformers building blocks, supporting a composable construction.