x86上预编译模型报错

不使用预编译没问题，是torch库损坏了吗
以下是报错日志

1. D Generate fake input
   
2. D Compiling start...
   
3. D gcc -Wall -std=c++0x -I. -I/home/chen/anaconda3/envs/rk/lib/python3.6/site-packages/rknn/simulator/LION/include/ -I/home/chen/anaconda3/envs/rk/lib/python3.6/site-packages/rknn/simulator/LION/include/CL -I/home/chen/anaconda3/envs/rk/lib/python3.6/site-packages/rknn/simulator/LION/include/VX -I/home/chen/anaconda3/envs/rk/lib/python3.6/site-packages/rknn/simulator/LION/include/ovxlib -D__linux__ -DLINUX -O3 -c rknn_torchjitexport.c
   
4. D /tmp/tmplmmm3adp/makefile.linux:44: recipe for target 'rknn_torchjitexport.o' failed
   
5. D Compiling finished.
   
6. D Generate pre-compile model data start...
   
7. D Generate pre-compile model data finished.
   
8. E Generate nbg failed.
   
9. E Catch exception when building RKNN model!

复制代码

版本不一致吧？

jefferyzhang 发表于 2023-11-21 10:52
版本不一致吧？

是一致的，之前能正常导出，系统文件被加密过，后来把环境和项目解密了，导出还是预编译的编译报错

已解决.