Question

在尝试使用启用 GPU 的 tensorflow 时，我遇到了一些错误消息，尽管日志中有许多成功消息，但这些消息似乎限制了我运行 GPU 的能力（见下文）。在软件方面，我使用的是 Anaconda Lite 环境。在硬件方面，我有一个 RTX 2080 gpu。以下是我的环境中安装的软件包。

我正在使用这个 tensorflow 批准的依赖组合：

这是重要的安装，但我在下面附上了整个 conda 列表。

TensorFlow-GPU 2.4.0
库德恩 8.0.5
Cudatoolkit 11.0.3

absl-py                   0.12.0                   pypi_0    pypi
astunparse                1.6.3                    pypi_0    pypi
ca-certificates           2020.12.5            h5b45459_0    conda-forge
cachetools                4.2.1                    pypi_0    pypi
certifi                   2020.12.5        py37h03978a9_1    conda-forge
chardet                   4.0.0                    pypi_0    pypi
cudatoolkit               11.0.3               h3f58a73_8    conda-forge
cudnn                     8.0.5.39             hfe7f257_1    conda-forge
cycler                    0.10.0                   pypi_0    pypi
filetype                  1.0.7                    pypi_0    pypi
flatbuffers               1.12                     pypi_0    pypi
gast                      0.3.3                    pypi_0    pypi
google-auth               1.28.0                   pypi_0    pypi
google-auth-oauthlib      0.4.4                    pypi_0    pypi
google-pasta              0.2.0                    pypi_0    pypi
grpcio                    1.32.0                   pypi_0    pypi
h5py                      2.10.0                   pypi_0    pypi
idna                      2.10                     pypi_0    pypi
imageai                   2.1.6                    pypi_0    pypi
importlib-metadata        3.10.0                   pypi_0    pypi
keras                     2.4.3                    pypi_0    pypi
keras-preprocessing       1.1.2                    pypi_0    pypi
keras-resnet              0.2.0                    pypi_0    pypi
kiwisolver                1.3.1                    pypi_0    pypi
markdown                  3.3.4                    pypi_0    pypi
matplotlib                3.3.2                    pypi_0    pypi
numpy                     1.19.3                   pypi_0    pypi
oauthlib                  3.1.0                    pypi_0    pypi
opencv-python             4.5.1.48                 pypi_0    pypi
openssl                   1.1.1k               h8ffe710_0    conda-forge
opt-einsum                3.3.0                    pypi_0    pypi
pillow                    7.0.0                    pypi_0    pypi
pip                       21.0.1           py37haa95532_0
protobuf                  3.15.7                   pypi_0    pypi
pyasn1                    0.4.8                    pypi_0    pypi
pyasn1-modules            0.2.8                    pypi_0    pypi
pyparsing                 2.4.7                    pypi_0    pypi
python                    3.7.6                h60c2a47_2
python-dateutil           2.8.1                    pypi_0    pypi
python_abi                3.7                     1_cp37m    conda-forge
pyyaml                    5.4.1                    pypi_0    pypi
requests                  2.25.1                   pypi_0    pypi
requests-oauthlib         1.3.0                    pypi_0    pypi
rsa                       4.7.2                    pypi_0    pypi
scipy                     1.4.1                    pypi_0    pypi
setuptools                52.0.0           py37haa95532_0
six                       1.15.0                   pypi_0    pypi
sqlite                    3.35.3               h2bbff1b_0
tensorboard               2.4.1                    pypi_0    pypi
tensorboard-plugin-wit    1.8.0                    pypi_0    pypi
tensorflow                2.4.0                    pypi_0    pypi
tensorflow-estimator      2.4.0                    pypi_0    pypi
tensorflow-gpu            2.4.0                    pypi_0    pypi
termcolor                 1.1.0                    pypi_0    pypi
typing-extensions         3.7.4.3                  pypi_0    pypi
urllib3                   1.26.4                   pypi_0    pypi
vc                        14.2                 h21ff451_1
vs2015_runtime            14.27.29016          h5e58377_2
werkzeug                  1.0.1                    pypi_0    pypi
wheel                     0.36.2             pyhd3eb1b0_0
wincertstore              0.2                      py37_0
wrapt                     1.12.1                   pypi_0    pypi
zipp                      3.4.1                    pypi_0    pypi

在执行与 ImageAI python 包相关的代码时，我遇到了以下打印输出。注意：有很多成功的 DLL 加载，但是需要注意三个重要的失败点。

无法调用 ptxas.exe --version
调用 CreateProcess 失败。错误代码：2
未创建 XLA 设备，未设置 tf_xla_enable_xla_devices

学分：尝试遵循其他一些线程并得出结论，此功能似乎正在开发中。 Tensorflow 2.4.1 - Couldn't invoke ptxas.exe

创建过程：我发现了有此错误的线程，但没有与 tensorflow 相关的线程。

tf_xla_enable_xla_devices：这是一个可以忽略的错误，每个线程：https://github.com/tensorflow/tensorflow/issues/44683

附加的是调用 tensorflow 脚本时的整个日志记录。

error:2021-04-04 19:16:49.100399: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library cudart64_110.dll

error:2021-04-04 19:16:50.654630: I tensorflow/compiler/jit/xla_cpu_device.cc:41] Not creating XLA devices, tf_xla_enable_xla_devices not set

error:2021-04-04 19:16:50.655534: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library nvcuda.dll

error:2021-04-04 19:16:50.685778: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1720] Found device 0 with properties:
pciBusID: 0000:01:00.0 name: NVIDIA GeForce RTX 2080 computeCapability: 7.5
coreClock: 1.71GHz coreCount: 46 deviceMemorySize: 8.00GiB deviceMemoryBandwidth: 417.23GiB/s
2021-04-04 19:16:50.685800: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library cudart64_110.dll

error:2021-04-04 19:16:50.696485: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library cublas64_11.dll
2021-04-04 19:16:50.696504: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library cublasLt64_11.dll

error:2021-04-04 19:16:50.700890: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library cufft64_10.dll

error:2021-04-04 19:16:50.702964: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library curand64_10.dll

error:2021-04-04 19:16:50.711373: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library cusolver64_10.dll

error:2021-04-04 19:16:50.714875: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library cusparse64_11.dll

error:2021-04-04 19:16:50.716294: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library cudnn64_8.dll

error:2021-04-04 19:16:50.716353: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1862] Adding visible gpu devices: 0

error:2021-04-04 19:16:50.716572: 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.
2021-04-04 19:16:50.716942: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1720] Found device 0 with properties:
pciBusID: 0000:01:00.0 name: NVIDIA GeForce RTX 2080 computeCapability: 7.5
coreClock: 1.71GHz coreCount: 46 deviceMemorySize: 8.00GiB deviceMemoryBandwidth: 417.23GiB/s
2021-04-04 19:16:50.716955: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library cudart64_110.dll

error:2021-04-04 19:16:50.716962: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library cublas64_11.dll
2021-04-04 19:16:50.716968: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library cublasLt64_11.dll
2021-04-04 19:16:50.716975: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library cufft64_10.dll
2021-04-04 19:16:50.716979: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library curand64_10.dll
2021-04-04 19:16:50.716983: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library cusolver64_10.dll
2021-04-04 19:16:50.716988: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library cusparse64_11.dll
2021-04-04 19:16:50.716992: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library cudnn64_8.dll
2021-04-04 19:16:50.717021: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1862] Adding visible gpu devices: 0

error:2021-04-04 19:16:51.182297: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1261] Device interconnect StreamExecutor with strength 1 edge matrix:
2021-04-04 19:16:51.182316: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1267]      0
2021-04-04 19:16:51.182320: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1280] 0:   N

error:2021-04-04 19:16:51.182515: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1406] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 6599 MB memory) -> physical GPU (device: 0, name: NVIDIA GeForce RTX 2080, pci bus id: 0000:01:00.0, compute capability: 7.5)

error:2021-04-04 19:16:51.184721: I tensorflow/compiler/jit/xla_gpu_device.cc:99] Not creating XLA devices, tf_xla_enable_xla_devices not set

error:2021-04-04 19:16:52.512582: I tensorflow/compiler/mlir/mlir_graph_optimization_pass.cc:116] None of the MLIR optimization passes are enabled (registered 2)

error:2021-04-04 19:16:52.527832: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library cudnn64_8.dll

error:2021-04-04 19:16:53.360829: E tensorflow/core/platform/windows/subprocess.cc:283] Call to CreateProcess failed. Error code: 2
2021-04-04 19:16:53.360847: W tensorflow/stream_executor/gpu/asm_compiler.cc:55] Couldn't invoke ptxas.exe --version

error:2021-04-04 19:16:53.369627: E tensorflow/core/platform/windows/subprocess.cc:283] Call to CreateProcess failed. Error code: 2

error:2021-04-04 19:16:53.369929: W tensorflow/stream_executor/gpu/redzone_allocator.cc:314] Internal: Failed to launch ptxas
Relying on driver to perform ptx compilation.
Modify $PATH to customize ptxas location.
This message will be only logged once.

error:2021-04-04 19:16:53.375630: E tensorflow/core/platform/windows/subprocess.cc:283] Call to CreateProcess failed. Error code: 2

error:2021-04-04 19:16:53.381209: E tensorflow/core/platform/windows/subprocess.cc:283] Call to CreateProcess failed. Error code: 2

error:2021-04-04 19:16:53.386810: E tensorflow/core/platform/windows/subprocess.cc:283] Call to CreateProcess failed. Error code: 2

error:2021-04-04 19:16:53.391792: E tensorflow/core/platform/windows/subprocess.cc:283] Call to CreateProcess failed. Error code: 2

error:2021-04-04 19:16:53.396262: E tensorflow/core/platform/windows/subprocess.cc:283] Call to CreateProcess failed. Error code: 2

error:2021-04-04 19:16:53.401723: E tensorflow/core/platform/windows/subprocess.cc:283] Call to CreateProcess failed. Error code: 2

There is a whole bunch more of the same print out that I have omitted.

Answer 1

我只安装了 Cudatoolkit 的运行时部分（在 Windows 上使用 GUI 安装程序）后遇到了类似的问题。

确保您安装了整个 CUDA 包（包括运行时和开发）。之后，ptxas.exe 应该会出现在您的 CUDA bin 文件夹中，Tensorflow 应该可以找到它。

调用 CreateProcess 失败。错误代码：2（TensorFlow）

1 个答案: