Seattle Deep Learning Meetup Projects!

This repository contains all info about the current (and past) Seattle Deep Learning Meetup projects.

How to join a project group

Just follow the following list:

1. Look at the list of Active Projects below.
2. Once you've found a project group which strikes your fancy, take note of the project group leader's Slack contact info.
3. Join us on Slack!
4. Contact the project leader (via PM or in the team's Slack channel) and tell them you're interested in joining their group! They should be able to provide more information and help you contribute to the project.

Projects

Active Projects

• CNN Learning via Stanford CS231N course
• NLP Joke Project
• Deep learning on shallow hardware
• A self-driving model car
• Berkeley CS294: Deep Reinforcement Learning - Study Group
• NLP with Deep Learning - Stanford CS224n

In The Works!

The following projects are being considered or organized by members of our meetup group. Look for them at the next Meetup!

• Photo Captions: Build a network to produce captions for photos
  • Slack channel #photocap

Past Projects

How to start a project group

Thanks for your interest! We'd love to help you find some collaborators for your project! To start a group, just do the following:

1. Create a team channel on Slack.
2. Add your project group to this repository.
   • See below for an example template.
   • Contact the Github manager to have it uploaded.
3. Notify the meetup group in the #projects channel on Slack.
   • e.g. "I'm starting a new project group for X! It'll involve A, B, and C! We're on channel #myprojectgroup! Let me know if you're interested in joining!"
4. Give a ~5-10 minute talk at our next meetup.
   • Your talk should be to let everyone know about your project group and give them the opportunity to ask questions.
   • Feel free to bring slides! Pictures and examples are great.
   • Invite people who are interested to join your project group!
   • Contact one of the meetup organizers and they'll get you slotted to speak!
5. Grab a table after your talk and begin planning your project group with your new members.
   • We'd encourage you to decide on what tools you'd like to use as a team (e.g. Github, Trello, PyTorch, etc)
   • We'd also encourage you to meet up outside of the meetup! Meeting at coffee shops, etc.

my_project_group.md example

# Handwritten digit classifier project

|                               |                  |
| ----------------------------- | ---------------- |
| Project name                  | Deep Retro       |
| Project leader                | Benjamin Bastian |
| Project leader Slack username | @sagelywizard    |
| Project slack channel         | #retro           |

## Description

In this project, we're building the world's best handwritten digit detector.
We're ultimately going to use cutting-edge deep learning technology known as
LeNet-5, implemented in c64 assembler. We'll be using the MNIST dataset, which
contains 60,000 training examples of handwritten digits and 10,000 examples for
a test set. We're starting off with using simpler classifiers, like kNN and
shallow MLPs. If you'd like to join our group, reach out to Ben on Slack!

## Group meetup schedule

We meet at Zeitgeist in Pioneer Square once or twice per month (check our Slack
channel for more info!). 

## More info...

If you're interested in reading a bit more about LeNet-5, [here's a link to the
original paper](http://yann.lecun.com/exdb/publis/pdf/lecun-01a.pdf).

Here are some of the images we'll be classifying.

![MNIST digits](mnist.jpg)