MLOPS_Project2

Getting Started

Follow these steps to set up your project:

1. Add Neptune Credentials:

   • Create a .env file in the root folder.
   • Add your Neptune credentials to the .env file:

     NEPTUNE_API_TOKEN=""
     NEPTUNE_PROJECT=""
     

     How to get my api token?

2. Create the Conda Environment:

   • Run the command: conda create -n mlopsp2 python=3.10.12

3. Activate Conda Environment:

   • Activate the environment: conda activate mlopsp2

4. Install Requirements:

   • Install necessary packages: pip install -r requirements.txt

5. Run the Script Locally:

   • Execute the script: python train.py --checkpoint_dir ./

Building & Running Docker Container

To build and run the Docker container, follow these steps:

1. Build the container:

   • docker build -t mlops-p2 .

2. Run the container:

   • docker run --name MLOPS-Project2 mlops-p2

Pushing the Docker Container

To push the Docker container to a registry, follow these instructions:

1. Modify .env for Security:

   • In the .env file, replace
     COPY .env .
     with
     #COPY .env .
     to prevent copying credentials.

2. Build the Container:

   • Build with your username: docker build -t {yourUsername}/mlops-p2 .
   • Replace {yourUsername} with your Docker Hub username.

3. Verify the Container:

   • Run and verify the container:

     docker run --name MLOPS-Project2 -e NEPTUNE_API_TOKEN="YOUR_API_KEY" -e NEPTUNE_PROJECT='YOUR_PROJECT' {yourUsername}/mlops-p2
     

   • Replace {yourUsername}, YOUR_API_KEY, and YOUR_PROJECT appropriately.

4. Push the Container:

   • Push to Docker Hub: docker push {yourUsername}/mlops-p2
   • Again, replace {yourUsername} with your username.

Pulling Your Container

To pull your previously pushed container, use the command:

• docker pull {yourUsername}/mlops-p2
• Remember to replace {yourUsername} with your Docker Hub username.