Aşkın Üçüncü

I guess there is no improvement in this regard. Because the system still produces negative values.

Is there an explanatory document on how to use the produced model file?

I have the same problem. Please help me.

As suggested, I removed the build folder. It didn't happen again

I made the CUDA version 10.1. However, the same error occurs again

open editor --> /myproject/env/lib/python3.7/site-packages/kur/backend/keras_backend.py -->line 195 -->keras.backend.set_image_dim_ordering('tf') change keras.backend.set_image_data_format('channels_last') save file and run -> kur -v train speech.yml

i have the same problem