ModelTC / United-Perception

United Perception
Apache License 2.0
426 stars 65 forks source link

YOLOX QAT成功,精度无下降,但是在deploy to tengine的时候出现多个问题 #39

Closed RedHandLM closed 1 year ago

RedHandLM commented 1 year ago

基于UP量化YOLOX,QAT训练成功,精度无下降,但是在deploy到tengine的时候出现多个问题, 1:miss key, 缺少量化节点 971cdb10062985847adad7eaf8c45e1

2:fake_quantize_per_tensor_affine() received an invalid combination of arguments 导出onnx时出现参数不对应的情况,以下为报错日志 deploy_tengine_error.txt

环境: env: Ubuntu 20.04 RTX3060TI CUDA: 11.4 Name: torch Version: 1.10.0+cu111 Name: MQBench Version: 0.0.6 onnx 1.7.0

RedHandLM commented 1 year ago

替换文件解决, 插入了错误的节点,需要进行过滤 tengine_u8_quantizer.txt