new notebooks, cleanup & updates (#73)

* added notebooks, reverted to tf.train.AdamOptimizer
* readme work & various cleanup, better prediction output, updated docker image

Author: amygdala

INSTALL.md

@@ -1,4 +1,6 @@
 
+**Note**: you do not need to follow these instructions for the [O'Reilly AI Conf workshop](https://conferences.oreilly.com/artificial-intelligence/ai-ca/public/schedule/detail/60305).  For that workshop, you just need [TensorFlow installed](https://www.tensorflow.org/install/) on your laptop.
+
 # Installation instructions for the TensorFlow/Cloud ML workshop
 
   - [Project and Cloud ML setup](#project-and-cloud-ml-setup)
@@ -52,7 +54,7 @@ If you like, you can start up a Google Compute Engine (GCE) VM with docker insta
 Once Docker is installed and running, download the workshop image:
 
 ```sh
-$ docker pull gcr.io/google-samples/tf-workshop:v6
+$ docker pull gcr.io/google-samples/tf-workshop:v7
 ```
 
 [Here's the Dockerfile](https://github.com/amygdala/tensorflow-workshop/tree/master/workshop_image) used to build this image.
@@ -67,7 +69,7 @@ Once you've downloaded the container image, you can run it like this:
 
 ```sh
 docker run -v `pwd`/workshop-data:/root/tensorflow-workshop-master/workshop-data -it \
-    -p 6006:6006 -p 8888:8888 -p 5000:5000 gcr.io/google-samples/tf-workshop:v6
+    -p 6006:6006 -p 8888:8888 -p 5000:5000 gcr.io/google-samples/tf-workshop:v7
 ```
 
 Edit 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.
@@ -80,7 +82,7 @@ For the second two, you will get a URL to paste in your browser, to obtain an au
 ```shell
 gcloud config set project <your-project-name>
 gcloud auth login
-gcloud beta auth application-default login
+gcloud auth application-default login
 ```
 
 ### Restarting the container later
@@ -105,7 +107,7 @@ $ docker exec -it <container_id> bash
 
 ### Running the Docker container on a VM
 
-It is easy to set up a Google Compute Engine (GCE) VM on which to run the Docker container. We sketch the steps below, or see [TLDR_CLOUD_INSTALL.md](TLDR_CLOUD_INSTALL.md) for more detail.
+It is easy to set up a Google Compute Engine (GCE) VM on which to run the Docker container. We sketch the steps below.
 
 First, make sure that your project has the GCE API enabled. An easy way to do this is to go to the [Cloud Console](https://console.cloud.google.com/), and visit the Compute Engine panel.  It should display a button to enable the API.
 

README.md

@@ -14,6 +14,7 @@ This document points to more information for each workshop lab.
 
 - [Building a small starter TensorFlow graph](workshop_sections/getting_started/starter_tf_graph/README.md)
 - [XOR: A minimal training example](workshop_sections/getting_started/xor/README.md)
+- A [LinearRegressor example](workshop_sections/linear_regressor_datasets) that uses Datasets.
 
 ## The MNIST (& 'fashion MNIST') series
 

TLDR_CLOUD_INSTALL.md

@@ -22,13 +22,13 @@ RUN apt-get install -y python-scipy
 RUN pip install sklearn nltk pillow setuptools
 RUN pip install flask google-api-python-client
 RUN pip install pandas python-snappy scipy scikit-learn requests uritemplate
-RUN pip install --upgrade --force-reinstall https://storage.googleapis.com/cloud-ml/sdk/cloudml.latest.tar.gz
+# RUN pip install --upgrade --force-reinstall https://storage.googleapis.com/cloud-ml/sdk/cloudml.latest.tar.gz
 
 # RUN python -c "import nltk; nltk.download('punkt')"
 
-RUN curl https://dl.google.com/dl/cloudsdk/channels/rapid/downloads/google-cloud-sdk-159.0.0-linux-x86_64.tar.gz | tar xvz
+RUN curl https://dl.google.com/dl/cloudsdk/channels/rapid/downloads/google-cloud-sdk-171.0.0-linux-x86_64.tar.gz | tar xvz
 RUN ./google-cloud-sdk/install.sh -q
-RUN ./google-cloud-sdk/bin/gcloud components install beta
+# RUN ./google-cloud-sdk/bin/gcloud components install beta
 
 ADD download_git_repo.py download_git_repo.py
 ENV PATH="${PATH}:/root/google-cloud-sdk/bin"

workshop_image/Dockerfile

@@ -11,6 +11,8 @@ This directory contains the workshop labs.
 
 - [Building a small starter TensorFlow graph](getting_started/starter_tf_graph/README.md)
 - [XOR: A minimal training example](getting_started/xor/README.md)
+- A [LinearRegressor example](linear_regressor_datasets) that uses Datasets.
+
 
 ## The MNIST (& 'fashion MNIST') series
 

workshop_sections/README.md

@@ -14,7 +14,7 @@ To run locally, first choose either:
 python -m xor.xor_summaries --output-dir ${OUTPUT_DIR}
 ```
 
-or if you want to try using the `gcloud beta ml local train` command the following will work:
+or if you want to try using the `gcloud ml-engine local train` command, the following will work:
 
 ```
 gcloud ml-engine local train \
@@ -24,7 +24,7 @@ gcloud ml-engine local train \
     --output-dir ${OUTPUT_DIR}
 ```
 
-The `gcloud beta ml local train` command runs your python code locally in an environment that emulates that of the Google Cloud Machine Learning API. This is primarily useful for distributed execution, where the local tool can serve as validation, that your code will run properly.
+The `gcloud ml-engine local train` command runs your python code locally in an environment that emulates that of the Google Cloud Machine Learning API. This is primarily useful for distributed execution, where the local tool can serve as validation, that your code will run properly.
 
 ### Running in the Cloud
 

workshop_sections/getting_started/xor/README.md

@@ -8,5 +8,4 @@ Each builds conceptually on the previous ones.  Start at the README numbered '01
 
 - [02_README_mnist_estimator](./02_README_mnist_estimator.md): Use the high-level TensorFlow APIs in `tf.estimator` to easily build a `LinearClassifier` and a `DNNClassifier` with hidden layers. Introducing TensorBoard.
 
-**Details TBD**: Building a CNN Custom Estimator: both TensorFlow and [Keras](https://keras.io/) versions. Run locally or do (distributed) training on CMLE.
-[mnist_cnn_custom_estimator](mnist_cnn_custom_estimator).
+- [Building Custom CNN Estimators](mnist_cnn_custom_estimator): where 'canned' Estimators aren't available, you can build a custom one, to get all the advantages of using an Estimator, including support for distributed training. Examples show how to do this with both Keras and TF layers.