Kahsi13/tomato-disease-bert

🍅 Tomato Disease Classifier – BERT Model

This project provides a fine-tuned Turkish BERT model for tomato plant disease diagnosis based on text input. It's designed for integration into real-time mobile apps and APIs, especially for agricultural use-cases.

---

📌 Model Overview

• Base model: dbmdz/bert-base-turkish-cased
• Framework: PyTorch + Hugging Face Transformers
• Task: Text Classification (multi-class)
• Languages: Turkish
• Classes: 27 tomato leaf diseases (fungal, bacterial, viral, abiotic, pest-related)
• Deployment: Compatible with FastAPI, Hugging Face Hub and Flutter (via backend API)

---

🧪 Example Usage

from transformers import AutoTokenizer, AutoModelForSequenceClassification
import torch

model_path = "Kahsi13/tomato-disease-bert"
tokenizer = AutoTokenizer.from_pretrained(model_path)
model = AutoModelForSequenceClassification.from_pretrained(model_path)

text = "Yapraklarda kahverengi lekeler görülüyor"
inputs = tokenizer(text, return_tensors="pt", truncation=True, padding=True)

with torch.no_grad():
    outputs = model(**inputs)
    predicted_class = torch.argmax(outputs.logits, dim=1).item()

# The predicted_class index can be converted to label using label_encoder.pkl (optional)