LLM inference engine using flashinfer kernels with various SD techniques.
git clone https://github.com/mlsys-seo/multisd-llm-engine
cd SDProfiler
docker build -t multi_sd_engine:0.1 . # It tasks ~25 min
./run_docker.sh
pip install -e .
2 commits
Python
98.1%
LLM inference engine using flashinfer kernels with various SD techniques.
git clone https://github.com/mlsys-seo/multisd-llm-engine
cd SDProfiler
docker build -t multi_sd_engine:0.1 . # It tasks ~25 min
./run_docker.sh
pip install -e .
2 commits
Python
98.1%