Tena-ethiopia/PACE-Breast-Cancer-

# 🛠️ PACE 2025 – Final Model Containerization Submission 📅 **Final Submission Deadline: 24th August 2025 (23:59 GMT)** This repository contains the **final submission guidelines** for the PACE 2025 Challenge. Participants are required to submit their trained models as a **Docker container** to ensure reproducibility. --- ## 📜 Submission Policy Each team is allowed **only one submission**. Please ensure you submit your **best model**. Using GitHub for submission may prevent you from uploading large files. If you encounter this issue, please follow the official Git LFS guide: 👉 Managing Large Files with Git LFS If you still face challenges uploading your model, contact the organizers through the **official communication channel** **before the specified deadline**, or you will be disqualified. An alternative submission method will be provided. You are allowed to include a **README.md** file explaining the flow of your code and model. --- ## PACE2025 Docker Guide This guide explains how to build and run the inference for the **ultrasound multi-task segmentation and classification model** in a Docker environment with GPU/cpu support. --- ## 📌 Project Overview This Docker container runs a multi-task deep learning model for ultrasound image analysis that can perform: - **Segmentation** – Generates binary masks for ultrasound images - **Classification** – Classifies ultrasound images into predefined categories --- ## 📂 Project Structure ``` ├── Dockerfile ├── requirements.txt ├── main.py # Main inference script ├── model.py # Model architecture ├── checkpoints/ # Model weights directory │ └── best_multitask_model.pth # Trained model checkpoint └── tools/ # Preprocessing and postprocessing utilities ├── preprocess.py └── postprocess.py ``` ## 1. Install Docker First, install Docker Desktop (available for Windows, macOS, and Linux): - Download: docker.com After installation, …