CNN Training Error

AI & Data Science Deep Learning (Training & Inference) cuDNN
1

Fundamentals of Deep Learning

03_aasl_cnn.ipynb
model.fit(x_train, y_train, epochs=20, verbose=1, validation_data=(x_valid, y_valid))

I found the below error on AWS provided by NVIDIA.
Please solve it

UnknownError: Failed to get convolution algorithm. This is probably because cuDNN failed to initialize, so try looking to see if a warning log message was printed above.
[[node sequential_1/conv2d_3/Conv2D (defined at :1) ]] [Op:__inference_train_function_3683]

Function call stack:
train_function

2

Hi,

This could be due to incompatible platform (CUDA, Tensorflow, Drivers versions) or you’re running out of memory.
Please refer to the following support matrix doc and installation doc, make sure cuDNN is available correctly.

Thank you.