/tensorflow-intro

An Introduction to TensorFlow Basics for the HandsOn Data Science Meetup Group

Primary LanguageJupyter Notebook

intro-to-tensorflow

This notebook was made by me and was used as a resource to help members of the HandsOn Data Science Meetup Group (Malmo, Sweden) learn the basics of using the TenorFlow framework. This notebook provides knowledge needed to understand and write simple TensorFlow code in Python. This notebook does NOT cover deep learning. The techniques chosen for this notebook do not necessarily, and most likely, will NOT transer well to other problems. There are many different methods and approaches that could have been used to solve the simple problem of this notebook and the methods chosen for this notebook are not necessarily the best either; they were chosen to simplify the content. Therefore, try not to become too engulfed in all the details of this notebook and instead just try to get the overall picture of what's going on.

Content:

  1. Computational Graphs

  2. Tensors

    • variables
    • constants
    • placeholders
    • performing operations
  3. Initialize variables + run session

  4. Cost function, optimizer, and forward and backward propagation

  5. Excercise: Solving a Linear Regression Model

  6. Bonus: TensorBoard

I undid pushed commits. Vib2006's answer was the only one that worked for me https://stackoverflow.com/questions/22682870/git-undo-pushed-commits. Revert did not work. This moved the local head to the commit and cleans all the commits after it.

git log --oneline --decorate --graph
git reset --hard <commit id here>
git clean -f -d

This updates all the indexes and shows what's behind compared to the remote branch

git fetch --all

This pushes forcefully with the "+" in front of the branch to the remote

git push -u origin +<branch here>