Update app.py

app.py

@@ -3,21 +3,15 @@ import requests
 import pandas as pd
 import textstat
 import os
-from transformers import pipeline
 
-# Chargement du token Hugging Face
+# Récupération du token Hugging Face
 HF_TOKEN = os.getenv("HF_TOKEN")
 
-# Appel API Zephyr-7B
+# Fonction pour appeler l'API Zephyr-7B
 def call_zephyr_api(prompt, hf_token=HF_TOKEN):
     API_URL = "https://api-inference.huggingface.co/models/HuggingFaceH4/zephyr-7b-beta"
     headers = {"Authorization": f"Bearer {hf_token}"}
-    payload = {
-        "inputs": prompt,
-        "parameters": {
-            "max_new_tokens": 300
-        }
-    }
+    payload = {"inputs": prompt}
 
     try:
         response = requests.post(API_URL, headers=headers, json=payload, timeout=60)
@@ -26,33 +20,46 @@ def call_zephyr_api(prompt, hf_token=HF_TOKEN):
     except Exception as e:
         raise gr.Error(f"❌ Erreur d'appel API Hugging Face : {str(e)}")
 
-# Chargement du pipeline de sentiment
-classifier = pipeline("sentiment-analysis", model="mrm8488/distilroberta-finetuned-financial-news-sentiment-analysis")
-
 # Fonction principale d'analyse
 def full_analysis(text, history):
     if not text:
-        return "Entrez une phrase.", "", 0, history
+        return "Entrez une phrase.", "", 0, history, 0
+
+    # 1. Demander à Zephyr de détecter le sentiment
+    prompt_sentiment = f"""
+You are a financial news sentiment detector.
+
+Given the following news text:
+\"{text}\"
+
+Respond only with one word: positive, neutral, or negative.
 
-    # 1. Détection sentiment par pipeline
-    result = classifier(text)[0]
-    detected_sentiment = result['label'].lower()
+Do not add any explanation or extra text.
+"""
+    detected_sentiment = call_zephyr_api(prompt_sentiment).lower()
 
-    # 2. Appel Zephyr pour explication basée sur ce sentiment
+    if detected_sentiment not in ["positive", "neutral", "negative"]:
+        detected_sentiment = "neutral"
+
+    # 2. Demander à Zephyr d'expliquer
     prompt_explanation = f"""
 You are a financial analyst AI.
-Based on the following financial news:
+
+Given the following financial news:
 \"{text}\"
-Explain clearly why the sentiment is {detected_sentiment}.
+
+The detected sentiment is: {detected_sentiment}.
+
+Now explain clearly why the sentiment is {detected_sentiment}.
 Write a concise paragraph.
 """
     explanation = call_zephyr_api(prompt_explanation)
 
-    # 3. Calcul score de clarté
+    # 3. Calculer la clarté
     clarity_score = textstat.flesch_reading_ease(explanation)
-    clarity_score = max(0, min(clarity_score, 100))  # Limité 0-100
+    clarity_score = max(0, min(clarity_score, 100))  # Limité entre 0-100
 
-    # 4. Stocker historique
+    # 4. Sauvegarder dans l'historique
     history.append({
         "Texte": text,
         "Sentiment": detected_sentiment.capitalize(),
@@ -60,9 +67,9 @@ Write a concise paragraph.
         "Explication": explanation
     })
 
-    return detected_sentiment.capitalize(), explanation, int(clarity_score), history
+    return detected_sentiment.capitalize(), explanation, clarity_score, history, int(clarity_score)
 
-# Fonction pour télécharger historique
+# Fonction pour télécharger l'historique
 def download_history(history):
     if not history:
         return None
@@ -71,10 +78,10 @@ def download_history(history):
     df.to_csv(file_path, index=False)
     return file_path
 
-# Interface Gradio
+# Gradio Interface
 def launch_app():
     with gr.Blocks() as iface:
-        gr.Markdown("# 📈 Analyse Financière Premium - Zephyr7B (Sentiment Pipeline)")
+        gr.Markdown("# 📈 Analyse Financière Premium - Zephyr7B")
 
         with gr.Row():
             input_text = gr.Textbox(label="Entrez votre question financière", lines=3)
@@ -85,7 +92,8 @@ def launch_app():
 
         sentiment_output = gr.Textbox(label="Sentiment Détecté")
         explanation_output = gr.Textbox(label="Explication de l'IA", lines=5)
-        clarity_slider = gr.Slider(0, 100, label="Score de Clarté (%)", interactive=False)
+        clarity_score_text = gr.Textbox(label="Score de Clarté (%)")
+        clarity_slider = gr.Slider(0, 100, label="Clarté (%)", interactive=False)
         file_output = gr.File(label="Fichier CSV")
 
         history = gr.State([])
@@ -93,7 +101,7 @@ def launch_app():
         analyze_btn.click(
             full_analysis,
             inputs=[input_text, history],
-            outputs=[sentiment_output, explanation_output, clarity_slider, history]
+            outputs=[sentiment_output, explanation_output, clarity_score_text, history, clarity_slider]
         )
 
         download_btn.click(
@@ -106,4 +114,3 @@ def launch_app():
 
 if __name__ == "__main__":
     launch_app()
-