Could not create cudnn handle: CUDNN_STATUS_INTERNAL_ERROR

[Varat7v2]

When I run it on CPU, it works well but throws CUDNN related error message. I did not find any solution. Why is not running on GPU? The complete error message is as below.

tensorflow.python.framework.errors_impl.UnknownError: 2 root error(s) found. (0) Unknown: 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 conv2d_1/convolution}}]] [[boolean_mask_53/GatherV2/_2673]]

(1) Unknown: 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 conv2d_1/convolution}}]] 0 successful operations. 0 derived errors ignored.

Errors may have originated from an input operation. Input Source operations connected to node conv2d_1/convolution:

But when I try to train some model using GPU, it works then.

[fenglinlie]

tensorflow.python.framework.errors_impl.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.

i got this error,
solve:
after yolo.py import …… code,befoer class YOLO(obj)
add:
config = tf.ConfigProto()
config.gpu_options.allow_growth=True
sess = tf.Session(config=config)
K.set_session(sess)