• This is the companion repository supporting the MMLG meetup on 19th Jan, 2019: https://www.meetup.com/Mumbai-Machine-Learning-Group/events/256998069/
• You can attend the talks and take notes as you see fit. If you wish to code along with the speakers, please ensure you have setup your laptop with the instructions in the following sections.
• If you encounter any problems in setting up, please reach out in the meetup comments' section. Otherwise, you can create issues in this repo.

Setting Up

Getting this Repo

1. (Option 1) You can download this repo by clicking on this line
   • After downloading it, extract and rename the extracted folder to cnn-visualisation
2. (Option 2) You can clone this repo and change into it
   • (if you have set-up your SSH key) run git clone git@github.com:soumendra/cnn-visualisation.git; cd cnn-visualisation (run in terminal)
   • (if you have not set-up your SSH key) run git clone https://github.com/soumendra/cnn-visualisation.git; cd cnn-visualisation (run in terminal)

Linux (Ubuntu)

Assuming you are inside the cnn-visualisation folder:

1. Make the scripts executable: chmod +x *.sh (run in terminal)
2. Install Miniconda (skip if you have miniconda or anaconda installed): ./conda.sh (run in terminal)
3. Close your terminal application. Start it again.
4. Create the cnn environment: conda env create -f cnn.yml (run in terminal)
5. Activate the environment: source activate cnn (run in terminal)
6. Install keras-vis: pip install keras-vis (run in terminal)

To start using this environment:

1. source activate cnn (run in terminal, in case you haven't previously)
2. jupyter notebook (run in terminal)
   • This will start a Jupyter Notebook in your browser
   • Getting started with Jupyter Notebook

Note: You can also install conda using the official guide, but using the conda.sh script is likely to be more painless.

Windows

1. Install Anaconda using this guide: https://conda.io/docs/user-guide/install/windows.html
2. Start Anaconda Prompt (you can search for it). This will open a terminal.
3. In the terminal, navigate to the folder where you unzipped this repo (covered in Getting this Repo section).
4. Create the cnn environment: conda env create -f cnn.yml
5. Activate the environment: activate cnn (run in Anaconda Prompt)
6. Install keras-vis: pip install keras-vis (run in Anaconda Prompt)

To start using this environment:

1. activate cnn (run in Anaconda Prompt)
2. jupyter notebook (run in Anaconda Prompt)
   • This will start a Jupyter Notebook in your browser
   • Getting started with Jupyter Notebook

For more on using Anaconda on Windows: https://docs.anaconda.com/anaconda/user-guide/getting-started/#open-nav-win .

Mac

1. Install miniconda using the official guide: https://conda.io/docs/user-guide/install/macos.html
   • Alternatively, install using homebrew
     • Install homebrew: https://brew.sh/
     • Install miniconda: brew install miniconda (run in terminal)
2. Close your terminal application. Start it again.
3. Assuming you are inside the cnn-visualisation folder:
   • Create the cnn environment: conda env create -f cnn.yml (run in terminal)
4. Activate the environment: source activate cnn (run in terminal)
5. Install keras-vis: pip install keras-vis (run in terminal)

To start using this environment:

1. source activate cnn (run in terminal, in case you haven't previously)
2. jupyter notebook (run in terminal)
   • This will start a Jupyter Notebook in your browser
   • Getting started with Jupyter Notebook

(Extra) Working with Git and GitHub

1. Your first guide to Git
2. Your second guide to Git
3. Your guide to GitHub
4. Going deeper