Unknown: Failed to get convolution algorithm.

Using TensorFlow backend.
Epoch 1/2
Traceback (most recent call last):
  File "/tmp/keras-learn/keras_3.py", line 60, in <module>
    model.fit(X_train, y_train, epochs=2, batch_size=32)
  File "/root/anaconda3/envs/ncc/lib/python3.6/site-packages/keras/engine/training.py", line 1178, in fit
    validation_freq=validation_freq)
  File "/root/anaconda3/envs/ncc/lib/python3.6/site-packages/keras/engine/training_arrays.py", line 204, in fit_loop
    outs = fit_function(ins_batch)
  File "/root/anaconda3/envs/ncc/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py", line 2979, in __call__
    return self._call(inputs)
  File "/root/anaconda3/envs/ncc/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py", line 2937, in _call
    fetched = self._callable_fn(*array_vals)
  File "/root/anaconda3/envs/ncc/lib/python3.6/site-packages/tensorflow/python/client/session.py", line 1458, in __call__
    run_metadata_ptr)
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}}]]
	 [[loss/mul/_119]]
  (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.

Process finished with exit code 1

keras运行报错
我开始觉得很奇怪，为什么有时候有这个问题，有时候没有
后面才发现，原来是GPU被别人占了，而不是我的代码问题