Блог

Я пытаюсь установить CUDA и cuDNN для работы с моим tensorflow. Я установил

visual studio 2019

CUDA 11.0,

cuDNN 8.0.4

tensorflow 2.4.0.

и Microsoft Visual C , распространяемый для Visual Studio 2015, 2017 и 2019 (vc_redist.x64.exe ) Я также пытался установить tf-nightly с существующей проблемой.

Я добавил полные пути к

CUDA v11.0bin

CUDA v11.0 extras CUPTI lib64

CUDA v11.0 include

и в bin, где я извлек cuDNN.

И это то, что я получаю, когда пытаюсь запустить код по ссылке. Я пытаюсь обучаться на наборе данных keras.

 tf.keras.datasets.cifar10
 

https://pastebin.com/ccbtZ8Bh

Все это .подтверждено, что библиотеки DLL находятся в подкаталогах CUDA.

 2021-03-02 03:21:38.544772: W tensorflow/stream_executor/platform/default/dso_loader.cc:60] Could not load dynamic library 'cudart64_110.dll'; dlerror: cudart64_110.dll not found
2021-03-02 03:21:38.544870: I tensorflow/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
2021-03-02 03:21:41.131356: I tensorflow/compiler/jit/xla_cpu_device.cc:41] Not creating XLA devices, tf_xla_enable_xla_devices not set
2021-03-02 03:21:41.132353: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library nvcuda.dll
2021-03-02 03:21:41.166235: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1720] Found device 0 with properties:
pciBusID: 0000:06:00.0 name: GeForce RTX 3090 computeCapability: 8.6
coreClock: 1.725GHz coreCount: 82 deviceMemorySize: 24.00GiB deviceMemoryBandwidth: 871.81GiB/s
2021-03-02 03:21:41.166421: W tensorflow/stream_executor/platform/default/dso_loader.cc:60] Could not load dynamic library 'cudart64_110.dll'; dlerror: cudart64_110.dll not found
2021-03-02 03:21:41.166849: W tensorflow/stream_executor/platform/default/dso_loader.cc:60] Could not load dynamic library 'cublas64_11.dll'; dlerror: cublas64_11.dll not found
2021-03-02 03:21:41.167105: W tensorflow/stream_executor/platform/default/dso_loader.cc:60] Could not load dynamic library 'cublasLt64_11.dll'; dlerror: cublasLt64_11.dll not found
2021-03-02 03:21:41.167431: W tensorflow/stream_executor/platform/default/dso_loader.cc:60] Could not load dynamic library 'cufft64_10.dll'; dlerror: cufft64_10.dll not found
2021-03-02 03:21:41.167737: W tensorflow/stream_executor/platform/default/dso_loader.cc:60] Could not load dynamic library 'curand64_10.dll'; dlerror: curand64_10.dll not found
2021-03-02 03:21:41.168048: W tensorflow/stream_executor/platform/default/dso_loader.cc:60] Could not load dynamic library 'cusolver64_10.dll'; dlerror: cusolver64_10.dll not found
2021-03-02 03:21:41.168364: W tensorflow/stream_executor/platform/default/dso_loader.cc:60] Could not load dynamic library 'cusparse64_11.dll'; dlerror: cusparse64_11.dll not found
2021-03-02 03:21:41.168676: W tensorflow/stream_executor/platform/default/dso_loader.cc:60] Could not load dynamic library 'cudnn64_8.dll'; dlerror: cudnn64_8.dll not found
2021-03-02 03:21:41.168913: W tensorflow/core/common_runtime/gpu/gpu_device.cc:1757] Cannot dlopen some GPU libraries. Please make sure the missing libraries mentioned above are installed properly if you would like to use GPU. Follow the guide at https://www.tensorflow.org/install/gpu for how to download and setup the required libraries for your platform.
Skipping registering GPU devices...
2021-03-02 03:21:41.169900: I tensorflow/core/platform/cpu_feature_guard.cc:142] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2
To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
2021-03-02 03:21:41.173494: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1261] Device interconnect StreamExecutor with strength 1 edge matrix:
2021-03-02 03:21:41.173550: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1267]
2021-03-02 03:21:41.173970: I tensorflow/compiler/jit/xla_gpu_device.cc:99] Not creating XLA devices, tf_xla_enable_xla_devices not set
2021-03-02 03:21:42.210887: I tensorflow/compiler/mlir/mlir_graph_optimization_pass.cc:116] None of the MLIR optimization passes are enabled (registered 2)