Views
No views yet
1# 针对OpenCL后端优化版本 2python3 convert_mnn.py onnx_path mnn_save_path "--weightQuantBits=8 --transformerFuse"