therajtiwari / Auto-DL

Auto-DL helps you make Deep Learning models without writing a single line of code and giving as little input as possible.

Auto-DL

The interface section of Auto-DL contains front-end and back-end servers based on React and Django Rest Framework respectively.

Demo

Setup

1. Install all the necessary libraries and binaries

sudo ./scripts/install.sh

2. Follow the instructions to run BackEndApp and FrontEndApp locally or Let it RIP!

./scripts/run.sh
# or you can pass --install to perform both step 1 and 2
./scripts/run.sh --install

How to run

1. # clone the repo
   git clone https://github.com/Auto-DL/Auto-DL.git

2. Activate your environment (not necessary but highly recommended).

3. # install the requirements, this might take some time, be patient
   pip install -r requirements.txt

4. # If you think your machine can handle a simulatenous installation of node modules, open another terminal
   
   cd FrontEndApp
   npm install
   
   # go grab a cup of coffee (or tea), it takes an eternity XD

5. Place data in the ./data directory.

   Your data should be divided into classes for classification, for example, if you're classifying "Cats V/s Dogs", then your ./data directory would look like:

   data
   └───dogs_and_cats
       ├───test
       │   ├───cats
       │   └───dogs
       └───train
           ├───cats
           └───dogs

6. Clone the sample.env to create .env in both BackEndApp/ and FrontEndApp/v1-react/ and configure the necessary environment variables

7. # run the backend
   # only after all requriements from requirements.txt are installed
   cd BackEndApp
   mkdir logs
   python manage.py runserver
   # you can ignore any migration warnings

8. # finally, run the react frontend
   # on a new terminal tab
   cd FrontEndApp/v1-react
   npm start

Note: For detailed instruction on data directory (point 5) please read DLMML's User Guide.

Using Docker

Configure the necessary environment variables in docker-compose.yml and run docker-compose up. This will setup a development server, so instead if you want to setup a production server you can replace the dockerfile context in docker-compose.yml for each container to include the production Dockerfile instead of Dockerfile.dev.

Note: Before running the production docker containers modify the nginx configuration if needed in nginx/nginx.conf as the FrontEndApp docker container uses nginx in production

Where to go next?

To know more about the project and initiative, please visit our website

Note

• To know more about the technicalities of the project, read our developer guidelines.
• For more detailed instructions to run the Generator module, Read our User guidelines

Contributing

Please take a look at our contributing guidelines if you're interested in helping!

Features/Enhancements Planned

• Improve the UI and UX.

• Show model training realted stats on the frontend.

• Visualization and data preprocessing steps.

• Model Explainability.