Using the Kaggle's dataset of SMS spam I will attempt to come up with a classifier to detect spam.
View the eda notebook to look at my eda and the model selection
- Install and activate conda environment
- launch api with Uvicorn
- Access at
localhost:8000/docs
or call api using curl
Build docker container with the following command at current directory
docker build -t <image name>:<tag> -f docker/simple_spam_classifer.dockerfile .
docker run -it <image name>:<tag> -p 8000
Access at localhost:8000/docs
Available here