You can set up for the workshop in two different ways -- installing the necessary packages into a virtual environment, or running a docker container.
Anaconda is a Python distribution that includes a large number of standard numeric and scientific computing packages. Anaconda uses a package manager called "conda" that has its own environment system similar to Virtualenv.
Install the version of Conda that uses Python 3.5 by default. Follow the instructions here. The miniconda version should suffice.
Follow the instructions on the TensorFlow site to create a Conda environment, activate it, and use pip to install TensorFlow within it. When following these instructions, be sure to use the Python 3 variant for both environment creation and in grabbing the TensorFlow .whl file.
Remember to activate this environment in all the terminal windows you use during this workshop.
With your conda environment activated, install the following packages:
$ conda install numpy
$ conda install scipy
$ pip install sklearn
$ conda install matplotlib
$ conda install jupyterWe're also providing a Docker container image with all the necessary libraries included, for you to download.
To use it, you'll need to have Docker installed. To run some of the examples, you'll likely need to configure it with at least 4GB of memory.
Once Docker is installed and running, download the workshop image:
$ docker pull gcr.io/google-samples/tf-workshop:v1Here's the Dockerfile used to build this image.
Create a directory (called, say, workshop-data) to mount as a volume when you start up the docker container. This will let you download data files into that directory, to make them accessible to the running container.
Once you've downloaded the container image, you can run it like this:
$ docker run -v `pwd`/workshop-data:/workshop-data -it \
-p 6006:6006 -p 8888:8888 gcr.io/google-samples/tf-workshop:v1Edit the path to the directory you're mounting as appropriate. The first component of the -v arg is the local directory, and the second component is where you want to mount it in your running container.
If you later exit your container and then want to restart it again, you can find the container ID by running:
$ docker ps -aThen, run:
$ docker start <container_id>(docker ps should then show it running). Once the workshop container is running again, you can exec back into it like this:
$ docker exec -it <container_id> bashAt various stages in this workshop, we'll have you download some data files. For convenience, we list them here:
https://storage.googleapis.com/oscon-tf-workshop-materials/saved_word2vec_model.zip https://storage.googleapis.com/oscon-tf-workshop-materials/processed_reddit_data/reddit_post_title_words.zip https://storage.googleapis.com/oscon-tf-workshop-materials/processed_reddit_data/news_aww/reddit_data.zip https://storage.googleapis.com/oscon-tf-workshop-materials/learned_word_embeddings/reddit_embeds.zip
(Thanks to reddit, for allowing us to use some post data for a training corpus.)
If you're running the Docker image, and have mounted a data directory as a volume (as described above), then copy the downloads into that directory so that they're accessible inside the container.
We'll be looking at some examples based on code in the tensorflow repo. While it's not necessary, you might want to clone or download it here, or grab the 0.8 release here.
In one section of the workshop, we'll look at running the TensorFlow distributed runtime on a Kubernetes cluster. Kubernetes is Google's open-source container orchestration framework. It provides a useful framework to run microservice-based apps.
If you want to play along, create a Google Cloud Platform account and project ahead of time -- start with the 'try it free' button on this page if you don't already have an account. (Alternately, understand how to stand up a Kubernetes cluster on some other cloud provider).
Then, download and install the latest Kubernetes release.