Can you please help me to know the reason!!
what do you mean by, the size increased?
When I loaded by Python code together with the .pth models (or at least they should be included automatically) into a .pt model for C++, the model actually got substantially smaller than the .pth model!
Thanks for the concern, Actual model size is 2.3MB. After the torch script procedure for c++, it became 4MB.Actually the problem is resolved…
Can you plz help me to cross compile the pytorch code for pine64…