Anaconda環境でのTensorFlowがGPUをうまく使ってくれない件

TensorFlowのword2vec_basic.pyをAnaconda環境下で動かしてみたら..

harahara@ada:~/notebooks$ python word2vec_basic.py
I tensorflow/stream_executor/dso_loader.cc:111] successfully opened CUDA library libcublas.so locally
I tensorflow/stream_executor/dso_loader.cc:111] successfully opened CUDA library libcudnn.so locally
I tensorflow/stream_executor/dso_loader.cc:111] successfully opened CUDA library libcufft.so locally
I tensorflow/stream_executor/dso_loader.cc:111] successfully opened CUDA library libcuda.so.1 locally
I tensorflow/stream_executor/dso_loader.cc:111] successfully opened CUDA library libcurand.so locally
Found and verified text8.zip
Data size 17005207
Most common words (+UNK) [['UNK', 418391], ('the', 1061396), ('of', 593677), ('and', 416629), ('one', 411764)]
Sample data [5239, 3082, 12, 6, 195, 2, 3135, 46, 59, 156] ['anarchism', 'originated', 'as', 'a', 'term', 'of', 'abuse', 'first', 'used', 'against']
3082 originated -> 5239 anarchism
3082 originated -> 12 as
12 as -> 6 a
12 as -> 3082 originated
6 a -> 195 term
6 a -> 12 as
195 term -> 2 of
195 term -> 6 a
I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:925] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
I tensorflow/core/common_runtime/gpu/gpu_device.cc:951] Found device 0 with properties:
name: GeForce GTX 750 Ti
major: 5 minor: 0 memoryClockRate (GHz) 1.0845
pciBusID 0000:01:00.0
Total memory: 1.95GiB
Free memory: 173.94MiB
I tensorflow/core/common_runtime/gpu/gpu_device.cc:972] DMA: 0
I tensorflow/core/common_runtime/gpu/gpu_device.cc:982] 0:   Y
I tensorflow/core/common_runtime/gpu/gpu_device.cc:1041] Creating TensorFlow device (/gpu:0) -> (device: 0, name: GeForce GTX 750 Ti, pci bus id: 0000:01:00.0)
E tensorflow/stream_executor/cuda/cuda_driver.cc:965] failed to allocate 173.94M (182386688 bytes) from device: CUDA_ERROR_OUT_OF_MEMORY

...（略、もうひとつExceptionが出て終了）

どうもうまく動かない..
上のログからすると、GPUは認識して動かそうとしているんだけど、空きメモリが173.94MiBしかなくてメモリが足りないとかのたまう..


しょうがないので同じマシンでAnaconda環境ではなく、docker/nvidia-docker環境を構築して、

$ sudo -b nohup nvidia-docker-plugin
$ sudo nvidia-docker run -it -p 8888:8888 gcr.io/tensorflow/tensorflow:latest-gpu

したコンテナ上のjupyter上で動かしてみた..

root@4df25c85f208:/notebooks# python word2vec_basic.py
I tensorflow/stream_executor/dso_loader.cc:111] successfully opened CUDA library libcublas.so locally
I tensorflow/stream_executor/dso_loader.cc:111] successfully opened CUDA library libcudnn.so locally
I tensorflow/stream_executor/dso_loader.cc:111] successfully opened CUDA library libcufft.so locally
I tensorflow/stream_executor/dso_loader.cc:111] successfully opened CUDA library libcuda.so.1 locally
I tensorflow/stream_executor/dso_loader.cc:111] successfully opened CUDA library libcurand.so locally
Found and verified text8.zip
Data size 17005207
Most common words (+UNK) [['UNK', 418391], ('the', 1061396), ('of', 593677), ('and', 416629), ('one', 411764)]
Sample data [5239, 3084, 12, 6, 195, 2, 3137, 46, 59, 156] ['anarchism', 'originated', 'as', 'a', 'term', 'of', 'abuse', 'first', 'used', 'against']
3084 originated -> 5239 anarchism
3084 originated -> 12 as
12 as -> 3084 originated
12 as -> 6 a
6 a -> 195 term
6 a -> 12 as
195 term -> 6 a
195 term -> 2 of
I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:925] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
I tensorflow/core/common_runtime/gpu/gpu_device.cc:951] Found device 0 with properties:
name: GeForce GTX 750 Ti
major: 5 minor: 0 memoryClockRate (GHz) 1.0845
pciBusID 0000:01:00.0
Total memory: 1.95GiB
Free memory: 1.92GiB
I tensorflow/core/common_runtime/gpu/gpu_device.cc:972] DMA: 0
I tensorflow/core/common_runtime/gpu/gpu_device.cc:982] 0:   Y
I tensorflow/core/common_runtime/gpu/gpu_device.cc:1041] Creating TensorFlow device (/gpu:0) -> (device: 0, name: GeForce GTX 750 Ti, pci bus id: 0000:01:00.0)
Initialized
Average loss at step  0 :  263.091247559
Nearest to often: moshiach, paratime, opacity, rumford, considered, ashkenazic, bias, ornithology,
Nearest to had: ftp, xv, jt, inexplicably, fock, savages, perfect, typhoons,
Nearest to an: cloak, carlin, siding, iala, infinitive, wilford, disseminate, patterned,
Nearest to one: irt, glencoe, part, formats, jpeg, decorate, koontz, yehudi,
Nearest to or: goetz, bequest, conflicting, invasive, biosphere, flat, shrews, tempted,
Nearest to as: negligent, militants, chaitin, blurring, comment, excitement, assured, lix,
..以下、正常終了。

..うまくいった...

よくよくドキュメント読むと..

TensorFlowのGET STARTEDのAnaconda installation/Using condaに

Only the CPU version of TensorFlow is available at the moment and can be installed in the conda environment for Python 2 or Python 3.

Pyhon2/3 conda環境内でインストールされたTensorFlowは現時点ではCPUのみが有効です。

..もっと前のほうに書こうよ、Google..