Keras с бэкэндом Tensorflow: не удалось уничтожить дескриптор cudnn: CUDNN_STATUS_BAD_PARAM

Question

Получил эту ошибку вида ошибки в обычном приложении keras resnet50. Эти ошибки бесполезны. Можно как нибудь отладить?

    /usr/local/lib/python2.7/dist-packages/h5py/__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.
  from ._conv import register_converters as _register_converters
Using TensorFlow backend.
libdc1394 error: Failed to initialize libdc1394
(435, 'train samples')
(171, 'test samples')
2018-02-09 11:37:35.987262: I tensorflow/core/platform/cpu_feature_guard.cc:137] Your CPU supports instructions that this TensorFlow binary was not compiled to use: SSE4.1 SSE4.2 AVX AVX2 FMA
2018-02-09 11:37:39.228270: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:892] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2018-02-09 11:37:39.228695: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1030] Found device 0 with properties: 
name: Tesla K80 major: 3 minor: 7 memoryClockRate(GHz): 0.8235
pciBusID: 0000:00:04.0
totalMemory: 11.17GiB freeMemory: 11.10GiB
2018-02-09 11:37:39.228721: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1120] Creating TensorFlow device (/device:GPU:0) -> (device: 0, name: Tesla K80, pci bus id: 0000:00:04.0, compute capability: 3.7)

__________________________________________________________________________________________________
train_funcaptcha.py:90: UserWarning: The semantics of the Keras 2 argument `steps_per_epoch` is not the same as the Keras 1 argument `samples_per_epoch`. `steps_per_epoch` is the number of batches to draw from the generator at each epoch. Basically steps_per_epoch = samples_per_epoch/batch_size. Similarly `nb_val_samples`->`validation_steps` and `val_samples`->`steps` arguments have changed. Update your method calls accordingly.
  verbose=1
train_funcaptcha.py:90: UserWarning: Update your `fit_generator` call to the Keras 2 API: `fit_generator(<utils.Rot..., verbose=1, workers=10, validation_data=<utils.Rot..., steps_per_epoch=13, epochs=200, callbacks=[<keras.ca..., validation_steps=5, use_multiprocessing=True)`
  verbose=1
Epoch 1/200
2018-02-09 11:38:55.104696: E tensorflow/stream_executor/cuda/cuda_blas.cc:366] failed to create cublas handle: CUBLAS_STATUS_NOT_INITIALIZED
2018-02-09 11:38:55.132607: E tensorflow/stream_executor/cuda/cuda_blas.cc:366] failed to create cublas handle: CUBLAS_STATUS_NOT_INITIALIZED
2018-02-09 11:38:55.151822: E tensorflow/stream_executor/cuda/cuda_blas.cc:366] failed to create cublas handle: CUBLAS_STATUS_NOT_INITIALIZED
2018-02-09 11:38:55.278180: E tensorflow/stream_executor/cuda/cuda_dnn.cc:385] could not create cudnn handle: CUDNN_STATUS_INTERNAL_ERROR
2018-02-09 11:38:55.278239: E tensorflow/stream_executor/cuda/cuda_dnn.cc:352] could not destroy cudnn handle: CUDNN_STATUS_BAD_PARAM
2018-02-09 11:38:55.278268: F tensorflow/core/kernels/conv_ops.cc:667] Check failed: stream->parent()->GetConvolveAlgorithms( conv_parameters.ShouldIncludeWinogradNonfusedAlgo<T>(), &algorithms) 
Aborted (core dumped)