• The predictor is powered by a time-series-forecasting Machine Learning Model.

• DATA-SOURCE : coinmarketcap

• The current real-time exchange rate data of Bitcoin in NGN and USD currencies are provided via the Coinbase Api

• Check out the live API via this or that link

• NB: these metrics improve over time as the model keeps learning from new data and hyperparameters are tweaked

MAPE=0.1580988278794064
MAE=4238.996757222961
RMSE=4585.15513690739

To get started with the project, ensure you have setup and activated a virtual environment, guides on that here

clone the repository via the command

git clone https://github.com/E-wave112/bitfast_2.0

install dependencies

python3 -m pip install -r requirements.txt

start the server by running the bash script below:

bash start.sh

Alternatively, you can start the server using the command below:

uvicorn application:app --reload

the server will be running on http://localhost:8000/docs

docker-compose up --build

To run the application, use the following command:

docker-compose up

• The app will be running on http://127.0.0.1:8000

• Access the docker image for this project on the cloud here

• It is not recommended to use alpine based images for this project(or most of any other python projects) and here's why

• A useful resource on how to push your docker image to DockerHub can be found here

• Ensure you have minikube installed on your machine

• Start the minikube cluster by running the command below:

minikube start

• To check the status of the cluster, run the command below:

minikube status

• Create a service discovery pattern for the application by running the command below:

kubectl apply -f k8s/services/service.yaml

• To deploy the application to the cluster, run the command below:

kubectl apply -f k8s/deployments/deployment.yaml

• To check the status of the pods, run the command below:

kubectl get pods

• To check the status of the services, run the command below:

kubectl get services

• To access the application, tunnel the service via the command below (we need to do this because our k8s service is of type LoadBalancer):

minikube tunnel

• Check the external IP (because you are running the cluster locally, your external IP address will be 127.0.0.1). More guides on that here

kubectl get services bitfast-service

• Open the url below in your browser (ensure the external IP is not pending )

http://<external-ip>:9500

• To delete the application, run the command below:

kubectl delete -f k8s/deployments/deployment.yaml

• To stop the minikube cluster, run the command below:

minikube stop

• To deploy your cluster to the cloud, check out the guides from the official kubernetes docs

Wanna check out my other machine learning projects and implementations? see them all here.