Using TFLite model on Xtensa for acceleration of ML algorithms

186 views Asked by At

I was working on CNN acceleration on XTensa by using TFLite models.I used the convolution model of TFlite and imported it into xtensa for building and running the model. However, while building i encountered an error:

Assertion failure at line 4373 of /build/tree/RG-2017.7_kuma/p4root/Xtensa/Software/xcc/be/cg/whirl2ops.cxx:

Compiler Error in file U:/workspace/CNN/tensorflow/lite/core/api/flatbuffer_conversions.cc during Code_Expansion phase:

XGOTO block doesn't have goto's? (1315943)

xt-xcc ERROR: C:/usr/xtensa/XtDevTools/install/tools/RG-2017.7-win32/XtensaTools/libexec/xcc/be.exe returned non-zero status 3

I am not able to figure out the cause for this error.Can anyone help me out?

0

There are 0 answers