dog classifier web app

• Build an end to end Deep learning application, which can classify 120 different dog breeds.

• Used models from tensoflow applications and tensorflow hub , later used the best model in my case mobilenet_v2_130_224 and later added costume layers for our prediction

• the front end of is created with Streamlit

Deployed Here

Live Link

Download the repository and install the required packages:

Clone this repo and install the required packages:

1. python -m venv my-env
2. my-env\Scripts\activate
3. pip install -r requirements.txt

To run streamlit app

streamlit run app.py

Docker commands for Development:

docker build -f Dockerfile.dev -t app:latest .

docker run -p 8000:8000 app:latest

For GCP Production:

gcloud auth login

gcloud config set project PROJECT_ID

gcloud builds submit --tag gcr.io/<ProjectName>/<AppName> --project=<ProjectName>

gcloud run deploy --image gcr.io/<ProjectName>/<AppName> --platform managed --project=<ProjectName> --allow-unauthenticated

Demo

dog_small.mp4