A powerful model compression framework for LLMs and LVLMs, adapted for NVIDIA GPUs and Huawei Ascend NPUs.
MAC-AutoML/MindPipe has added +504 stars since the first tracked point, with current momentum at 20.70.