MLE-Project (Iris)

Description

This project involves the training and inference of a machine learning model using the Iris dataset. The repository is structured to separate the training and inference environments, ensuring a clear workflow and easy reproducibility.

Repository Structure

project_repository/
│
├── datafiles/
│   ├── train_set.csv      # Training data file
│   └── inference_set.csv  # Inference data file
│
├── training/
│   ├── Dockerfile         # Dockerfile for training environment
│   ├── train_script.py    # Python script for training the model
│   └── requirements_train.txt  # Dependencies for training
│
├── inference/
│   ├── Dockerfile         # Dockerfile for inference environment
│   ├── infer_script.py    # Python script for model inference
│   └── requirements_infer.txt  # Dependencies for inference
│
├── unittests/
│   └── unittests.py
│
└── README.md              # README file

Setup and Installation

Prerequisites

• Docker
• Python 3.x

Installation

1. Clone the repository:

   git clone [repository URL]
   

2. Navigate to the cloned directory:

   cd your_project_repository
   

Training the Model

1. Build the Docker image for training:

   docker build -t iris-train -f training/Dockerfile .
   

2. Run the Docker container:

   docker run iris-train
   

Running Inference

1. Build the Docker image for inference:

   docker build -t iris-infer -f inference/Dockerfile .
   

2. Run the Docker container:

   docker run iris-infer
   

Usage

Training

• The training script (train_script.py) is located in the training/ directory. This script trains the model using train_set.csv and saves the trained model for inference.

Inference

• The inference script (infer_script.py) is located in the inference/ directory. This script loads the trained model and performs inference on inference_set.csv.

Authors

• O. K.