Task2

README.md

@@ -1,30 +1,13 @@
-# Git Workshop Exercise
-In this exercise your team will create a mini-website (with simple markdown pages) documenting your project.
-There are several tasks to complete (listed as `GitLab Issues`), but take these with a pinch of salt.
-The questions should get you thinking about positive open-source documentation, but the priority here is to **practice conflict resolution with Git**.
 
-To complete the exercises you'll just need to edit markdown files (end with a `.md`) with any text-editor.
-The `#`'s are for headings (more `#` means smaller).
-Wrap text in `*` for italics or `**` for bold and `\`` for code-blocks (`\`\`\`` for multi-line code blocks).
 
-Think very carefully about your actions (creating files, editing files, where you create a branch from...) and discuss any doubts/questions with your teammates.
-If you stay conscious of what you're doing, you'll quickly become a Git lean mean machine!
-Please **work on merging and fixing conflicts together** where possible.
-Note that it may be useful to use a visual UI to show what changes have been made (VScode is a great choice).
+#Who are we
+We are the dynamic balancing team from DeepNeuron. We are a reinforcement learning focused group working towards the end goal of 
+having an AI dynamically balanced to always provide a reasonable challenge to the player.  Going forward this could be implemented by
+game developers to help keep players engaged in the growing games market with a perfect challenge provided.
 
-# Setup and Workflow
-Sart by forking and cloning this repository (`git clone YOUR_REPO_LINK`).
-There will be a fork button near the top right hand side of the screen (left of the blue clone button).
-**Only one person per-team should do this** (you can share screen with everyone else so they know how too though)!
-Once you've got your own GitLab repository clone it (hitting the clone button and then HTTPS will give you the URL) to use in `git clone REPO_LINK`.
 
-Once everyone knows what task they'll start with create a feature branch to work in (`git checkout -b some-feature-branch`, but use a meaningful name which doesn't include the words feature branch).
-Although it won't be immidiately useful, do recall that you by default branch from your current (**not `master`**) branch.
-
-> Remember, each task should be completed in a new branch!
-
-You should gradually commit (`git commit -m "ADDED COOL THING/CHANGED BLAH BLAH BLAH"`) your changes.
-Once you've completed a task make sure to push (`git push`) everything and create a pull request (through the website).
-
-Before creating a pull request remember to get the latest changes (`git pull`).
-This is one place where conflicts may occur (hint - this is when rebases are useful).
+#Moar details.
+This semester we are working our way up through reinforcement learning algorithms with Gym enviroments. We have implented a DQN to solve cartpole
+before moving on to ATARI games and building up to a DDQN algorithm. When moving to ATARI we increased our networks by introducing CNN layers to 
+extract features out of the screen. Moving forward we intend on developing actor critic and then PPO networks. Once these algorithms have been explored 
+and our agent can kick ass we will return to the goal of dynamic balancing.