README.md

---
title: YOLO Trainer & Detector
emoji: 🎯
colorFrom: blue
colorTo: green
sdk: gradio
sdk_version: 5.29.0
app_file: app.py
pinned: false
license: apache-2.0
short_description: Train YOLOv8 and run object detection
---

# 🎯 YOLO Trainer & Detector

Train a YOLOv8 object detection model and run inference — all from a web UI.

## Features

- **🏋️ Training Tab** — Train YOLOv8 on the [yolo-detection-dataset](https://huggingface.co/datasets/dharshanzeb/yolo-detection-dataset) with configurable hyperparameters. Real-time training logs and metrics charts.
- **🔍 Inference Tab** — Upload any image and detect objects using your trained model or a pretrained COCO model. Bounding boxes drawn with confidence scores.
- **📊 Live Metrics** — Loss curves and mAP charts update after each epoch.

## How to Use

1. Go to the **Train** tab
2. Pick model size, epochs, batch size, learning rate
3. Click **Start Training** — watch logs and metrics in real-time
4. Switch to the **Inference** tab
5. Upload an image → see detections!