Songwei Liu

Senior Engineer, ByteNN ByteNN Team

Graduated from Zhejiang University with a master degree, his research area focuses on full-stack optimisation of deep learning algorithms, covering model optimisation and N-card inference optimisation. In ByteDance ByteNN team, he has been responsible for the construction of server-side sparse acceleration/LLM inference optimisation capability, and supported the inference optimisation of beanbag vision multimodal large model project. Currently, he is responsible for model optimisation in ByteNN team, and is committed to reducing the cost of cloud-based reasoning for LLM/SD models through the collaborative optimisation at the reasoning engine and model level, and further promoting the end-side landing of AIGC models.