README.md

metadata

title: ACCEPTIN - Telecom Site Quality Classification
emoji: 📡
colorFrom: blue
colorTo: purple
sdk: streamlit
sdk_version: 1.31.0
app_file: app.py
pinned: false
license: mit

📡 ACCEPTIN - Telecom Site Quality Classification

AI-powered telecom site inspection using ConvNeXt transfer learning.

---

🚀 Deploying ACCEPTIN on Hugging Face Spaces

1. Prepare Your Project Directory

Ensure your project has the following structure:

ACCEPTIN/
├── app.py
├── requirements.txt
├── README.md
├── models/
│   └── telecom_classifier.pth
├── utils/
│   ├── data_utils.py
│   └── model_utils.py
└── ... (other files)

2. Create a New Space on Hugging Face

1. Go to Hugging Face Spaces.
2. Click Create new Space.
3. Fill in:
   • Space name: acceptin (or your choice)
   • SDK: Streamlit
   • Hardware: CPU basic (free)
   • Visibility: Public or Private
4. Click Create Space.

3. Prepare Your Files

• requirements.txt: List all dependencies (see below for example).
• README.md: This file, with the YAML header above.
• Model File: Place telecom_classifier.pth in a models/ folder. If >10MB, use Git LFS.

Example requirements.txt

streamlit==1.31.0
torch==2.2.0
torchvision==0.17.0
Pillow==10.2.0
numpy==1.26.0
timm==0.9.12
opencv-python-headless==4.8.1.78
plotly==5.18.0
pandas==2.2.0
scikit-learn==1.4.0
matplotlib==3.8.0
seaborn==0.13.0
tqdm==4.66.0

4. (If Needed) Set Up Git LFS for Large Files

If your model file is large:

git lfs install
git lfs track "*.pth"
git add .gitattributes

5. Upload Your Files to the Space

A. Web Interface

• Go to your Space on Hugging Face.
• Click the Files tab.
• Upload all files and folders (app.py, requirements.txt, README.md, models/, utils/, etc.).

B. Git Method (Recommended)

git clone https://huggingface.co/spaces/YOUR_USERNAME/acceptin
cd acceptin
# Copy your files into this directory
# If using Git LFS:
git lfs install
git lfs track "*.pth"
git add .gitattributes
git add .
git commit -m "Initial ACCEPTIN deployment"
git push

6. Wait for Build & Test

• Hugging Face will automatically build your Space.
• Wait for the build to finish (watch the logs for errors).
• Test your app in the browser.

7. Troubleshooting

• If you see errors, check the build logs.
• Make sure all dependencies are in requirements.txt.
• Ensure your model path in app.py matches the uploaded file location.

8. Share Your Space

• Once working, share your Space URL (e.g., https://huggingface.co/spaces/YOUR_USERNAME/acceptin).

---

🏗️ Technical Overview

• Model: ConvNeXt Large (197M parameters, transfer learning from food detection)
• Task: Binary classification (good/bad telecom site)
• App: Streamlit web interface
• Data: Images of telecom sites, labeled as good or bad
• Deployment: Hugging Face Spaces (Streamlit SDK)

📋 Features

• Upload telecom site images for instant quality assessment
• Visual confidence scores and inspection breakdown
• Modern, responsive UI

📊 Model Performance

• Validation Accuracy: ~94%
• Model Size: ~750MB

📚 Data Requirements

• Images of telecom sites (good/bad)
• Recommended: 100+ images per class

---

For more details, see the in-app documentation or contact the author.