Skip to content

Instantly share code, notes, and snippets.

@xdhmoore
Created June 19, 2020 00:49
Show Gist options
  • Save xdhmoore/3b879edc723c32a9fa426a0097f6951c to your computer and use it in GitHub Desktop.
Save xdhmoore/3b879edc723c32a9fa426a0097f6951c to your computer and use it in GitHub Desktop.
Minimal test of tensorflow image on Windows
FROM tensorflow/tensorflow:1.15.0
COPY test-tf.py /tensorflow/test-tf.py
WORKDIR /tensorflow
#!/usr/bin/env python
import tensorflow as tf
hello = tf.constant('Hello, TensorFlow!')
sess = tf.Session()
@xdhmoore
Copy link
Author

Output is:

WARNING:tensorflow:From ./test-tf.py:5: The name tf.Session is deprecated. Please use tf.compat.v1.Session instead.

2020-06-19 00:48:56.905823: W tensorflow/stream_executor/platform/default/dso_loader.cc:55] Could not load dynamic library 'libcuda.so.1'; dlerror: libcuda.so.1: cannot open shared object file: No such file or directory
2020-06-19 00:48:56.905858: E tensorflow/stream_executor/cuda/cuda_driver.cc:318] failed call to cuInit: UNKNOWN ERROR (303)
2020-06-19 00:48:56.905987: I tensorflow/stream_executor/cuda/cuda_diagnostics.cc:156] kernel driver does not appear to be running on this host (c0e8a9783c35): /proc/driver/nvidia/version does not exist
2020-06-19 00:48:56.906459: I tensorflow/core/platform/cpu_feature_guard.cc:142] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA
2020-06-19 00:48:56.913014: I tensorflow/core/platform/profile_utils/cpu_utils.cc:94] CPU Frequency: 1992000000 Hz
2020-06-19 00:48:56.913329: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x5583d4015850 initialized for platform Host (this does not guarantee that XLA will be used). Devices:
2020-06-19 00:48:56.913358: I tensorflow/compiler/xla/service/service.cc:176]   StreamExecutor device (0): Host, Default Version

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment