README.md

metadata

title: AgriLens AI FastAPI
emoji: 🌱
colorFrom: green
colorTo: blue
sdk: docker
app_port: 7860
tags:
  - fastapi
  - agriculture
  - ai
  - plant-disease
pinned: false
short_description: AI-powered plant disease diagnosis API using Gemma-3n model
license: mit

AgriLens AI FastAPI

AI-powered plant disease diagnosis API using Google's Gemma-3n model.

Features

• Plant disease diagnosis from leaf images
• RESTful API endpoints
• Model sharing with Streamlit application
• CORS enabled for cross-origin requests

API Endpoints

• GET / - API status and information
• GET /health - Health check
• POST /load - Load the AI model
• POST /diagnose - Diagnose plant disease from image

Usage

Health Check

curl https://huggingface.co/spaces/sido1991/apiagrilens/health

Load Model

curl -X POST https://huggingface.co/spaces/sido1991/apiagrilens/load

Diagnose Plant Disease

curl -X POST \
  -F "image=@plant_leaf.jpg" \
  -F "culture=tomato" \
  -F "notes=Yellow spots on leaves" \
  https://huggingface.co/spaces/sido1991/apiagrilens/diagnose

Configuration

Variables d'environnement

• Model: google/gemma-3n-E4B-it
• Device: CPU (optimisé pour Hugging Face Spaces)
• Max Tokens: 256
• Cache Directory: /tmp/hf_home

Variables configurables

• MODEL_ID: Model identifier (default: google/gemma-3n-E4B-it)
• DEVICE_MAP: Device mapping (default: cpu)
• MAX_NEW_TOKENS: Maximum new tokens for generation (default: 256)
• HF_HOME: Hugging Face cache directory (default: /tmp/hf_home)