tensorflow-gpu==1.13.1

[alex-wey]

System information

• OS Platform and Distribution: Linux Ubuntu 18.04
• TensorFlow version: 1.13.1
• Python version: 3.6.9
• Installed using: Miniconda3, pip3
• GPU model and memory: Google Colab

Description of the problem
Hi Rowanz,

I'm trying to get the 'Quickstart' version of Grover working. As you already know, in order to run contextual_generate.py, TensorFlow-GPU and TensorBoard both version 1.13.1 are required. According to the log, the second code block does run successfully - however, in reality, there seems to be an installation issue since a ModuleNotFoundError is prompted.

Did you run into this issue at all? If so, how were you able to get around it? Some general assistance and explanation would be extremely helpful - perhaps there exists a direct link that can be used for these installations instead?

Sequence of commands executed before running into the problem

!wget -c https://repo.anaconda.com/miniconda/Miniconda3-4.5.4-Linux-x86_64.sh
!chmod +x Miniconda3-4.5.4-Linux-x86_64.sh
!bash ./Miniconda3-4.5.4-Linux-x86_64.sh -b -f -p /usr/local
!conda install -q -y --prefix /usr/local python=3.6 ujson
import sys
sys.path.append('/usr/local/lib/python3.6/site-packages')
# test it
import ujson

!pip3 install tensorboard==1.13.1
!pip3 install tensorflow-gpu==1.13.1

Unsuccessful command and traceback

!conda create -y -n grover python=3.6 && source activate grover && pip3 install -r /content/drive/My\ Drive/CS/GPT-2/grover/requirements-gpu.txt
!python /content/drive/My\ Drive/CS/GPT-2/grover/download_model.py base
!PYTHONPATH=$(pwd) python /content/drive/My\ Drive/CS/GPT-2/grover/sample/contextual_generate.py -model_config_fn lm/configs/base.json -model_ckpt models/base/model.ckpt -metadata_fn sample/april2019_set_mini.jsonl -out_fn april2019_set_mini_out.jsonl

Traceback (most recent call last):
  File "/content/drive/My Drive/CS/GPT-2/grover/sample/contextual_generate.py", line 1, in <module>
    import tensorflow as tf
ModuleNotFoundError: No module named 'tensorflow'