app.py

import gradio as gr
from transformers import BertTokenizer, BertForSequenceClassification
import torch
import pickle
from collections import defaultdict
import random
import os
from safetensors.torch import load_file

# Model və label_encoder yüklənməsi
def load_model():
    try:
        # Label encoder
        with open("best_model/label_encoder.pkl", "rb") as f:
            label_encoder = pickle.load(f)
        
        # Tokenizer
        tokenizer = BertTokenizer.from_pretrained("best_model")
        
        # Model (safetensors avtomatik dəstəklənir)
        model = BertForSequenceClassification.from_pretrained("best_model", use_safetensors=True)
        model.eval()
        
        print(f"Model uğurla yükləndi. Label sayı: {len(label_encoder.classes_)}")
        return tokenizer, model, label_encoder
    
    except Exception as e:
        print(f"Model yüklənmə xətası: {e}")
        # Faylları yoxla
        if os.path.exists("best_model"):
            files = os.listdir("best_model")
            print(f"best_model qovluğundakı fayllar: {files}")
        else:
            print("best_model qovluğu mövcud deyil")
        
        return None, None, None

# Model yükləmə
tokenizer, model, label_encoder = load_model()

# Prediction funksiyası
def predict_disease(text):
    if tokenizer is None or model is None or label_encoder is None:
        return "❌ Model yüklənməyib! Xəta var."
    
    if not text.strip():
        return "⚠️ Please enter some symptoms!"
    
    symptoms = [s.strip() for s in text.split(",") if s.strip()]
    if not symptoms:
        return "⚠️ Please enter valid symptoms separated by commas!"
    
    try:
        agg_probs = defaultdict(float)
        n_shuffles = 10
        
        for i in range(n_shuffles):
            random.shuffle(symptoms)
            shuffled_text = ", ".join(symptoms)
            
            inputs = tokenizer(
                shuffled_text, 
                return_tensors="pt", 
                truncation=True, 
                padding=True, 
                max_length=128
            )
            
            with torch.no_grad():
                outputs = model(**inputs)
                probs = torch.nn.functional.softmax(outputs.logits, dim=-1).squeeze()
            
            for idx, p in enumerate(probs):
                agg_probs[idx] += p.item()
        
        # Ortalama hesabla
        for k in agg_probs:
            agg_probs[k] /= n_shuffles
        
        # Top 3 nəticə
        top_3 = sorted(agg_probs.items(), key=lambda x: x[1], reverse=True)[:3]
        
        results = ["🏥 Top 3 Predicted Diseases:\n"]
        for idx, prob in top_3:
            label = label_encoder.classes_[idx]
            results.append(f"• **{label}** — Probability: {prob*100:.2f}%")
        
        return "\n".join(results)
    
    except Exception as e:
        return f"❌ Prediction xətası: {str(e)}"

# Gradio interface
iface = gr.Interface(
    fn=predict_disease,
    inputs=gr.Textbox(
        lines=2, 
        placeholder="fever, cough, headache, shortness of breath",
        label="Enter your symptoms (comma separated)"
    ),
    outputs=gr.Textbox(label="Predicted Diseases"),
    title="🏥 Disease NLP Classifier",
    description="Enter your symptoms separated by commas and get top 3 predicted diseases with confidence scores.",
    examples=[
        ["fever, cough, headache"],
        ["stomach pain, nausea, vomiting"],
        ["chest pain, shortness of breath"],
        ["dizziness, fatigue, weakness"],
        ["skin rash, itching, redness"]
    ]
)

# Launch
if __name__ == "__main__":
    if tokenizer and model and label_encoder:
        print("✅ Model hazırdır, Gradio başladılır...")
        iface.launch(
            server_name="0.0.0.0", 
            server_port=int(os.environ.get("PORT", 7860)),
            share=True  # Public link yaradır
        )
    else:
        print("❌ Model yüklənmədi, Gradio başladıla bilmir!")
        print("\nDebug məlumatları:")
        print(f"Hazırkı qovluq: {os.getcwd()}")
        print(f"Qovluq məzmunu: {os.listdir('.')}")
        
        # Sadə debug interface
        def debug_info():
            return f"Debug məlumatları:\nHazırkı qovluq: {os.getcwd()}\nFayllar: {os.listdir('.')}"
        
        debug_iface = gr.Interface(
            fn=debug_info,
            inputs=gr.Textbox(placeholder="Debug üçün hər hansı mətn yazın"),
            outputs=gr.Textbox(),
            title="🔧 Debug Interface"
        )
        
        debug_iface.launch(
            server_name="0.0.0.0",
            server_port=int(os.environ.get("PORT", 7860))
        )