Update app.py

app.py

@@ -205,7 +205,8 @@ mistral_api_key = os.getenv("mistral_api_key")
 llm = ChatMistralAI(
     model="mistral-large-latest",
     mistral_api_key=mistral_api_key,
-    temperature=0.1  # Lower temperature for faster responses
+    temperature=0.1,
+    streaming=True,
 )
 
 rag_loader = OptimizedRAGLoader()
@@ -372,7 +373,7 @@ button.primary-button:hover {
 }
 """
 
-# Interface Gradio avec la correction
+# Interface Gradio avec streaming
 with gr.Blocks(css=custom_css) as iface:
     with gr.Column(elem_classes="container"):
         gr.Markdown(
@@ -395,33 +396,26 @@ with gr.Blocks(css=custom_css) as iface:
                     lines=5,
                     elem_classes="rtl-text textbox-container"
                 )
-            # with gr.Column(scale=1):
-            #     context_box = gr.Textbox(
-            #         label="السياق المستخدم",
-            #         lines=4,
-            #         elem_classes="rtl-text textbox-container"
-            #     )
         
         submit_btn = gr.Button(
             "إرسال السؤال",
             elem_classes="primary-button",
             variant="primary"
         )
-       
 
-        def on_submit(question):
-            for response, context in process_question(question):
-                yield response  # Yield À CHAQUE itération
+        def stream_response(question):
+            response_stream = process_question(question)
+            for response, _ in response_stream:
+                gr.update(value=response)
+                yield response
 
-        
         submit_btn.click(
-            fn=on_submit,
+            fn=stream_response,
             inputs=input_text,
             outputs=answer_box,
             api_name="predict",
-            queue=False,
-)
-
+            queue=False
+        )
 
 if __name__ == "__main__":
     iface.launch(
@@ -429,9 +423,69 @@ if __name__ == "__main__":
         server_name="0.0.0.0",
         server_port=7860,
         max_threads=3,
-        show_error=True,
+        show_error=True
     )
 
+# # Interface Gradio avec la correction
+# with gr.Blocks(css=custom_css) as iface:
+#     with gr.Column(elem_classes="container"):
+#         gr.Markdown(
+#             "# نظام الأسئلة والأجوبة الذكي",
+#             elem_classes="app-title rtl-text"
+#         )
+        
+#         with gr.Column(elem_classes="textbox-container"):
+#             input_text = gr.Textbox(
+#                 label="السؤال",
+#                 placeholder="اكتب سؤالك هنا...",
+#                 lines=1,
+#                 elem_classes="rtl-text"
+#             )
+        
+#         with gr.Row():
+#             with gr.Column():
+#                 answer_box = gr.Textbox(
+#                     label="الإجابة",
+#                     lines=5,
+#                     elem_classes="rtl-text textbox-container"
+#                 )
+#             # with gr.Column(scale=1):
+#             #     context_box = gr.Textbox(
+#             #         label="السياق المستخدم",
+#             #         lines=4,
+#             #         elem_classes="rtl-text textbox-container"
+#             #     )
+        
+#         submit_btn = gr.Button(
+#             "إرسال السؤال",
+#             elem_classes="primary-button",
+#             variant="primary"
+#         )
+       
+
+#         def on_submit(question):
+#             for response, context in process_question(question):
+#                 yield response  # Yield À CHAQUE itération
+
+        
+#         submit_btn.click(
+#             fn=on_submit,
+#             inputs=input_text,
+#             outputs=answer_box,
+#             api_name="predict",
+#             queue=False,
+# )
+
+
+# if __name__ == "__main__":
+#     iface.launch(
+#         share=True,
+#         server_name="0.0.0.0",
+#         server_port=7860,
+#         max_threads=3,
+#         show_error=True,
+#     )
+
 # def process_question(question: str):  
 #     """  
 #     Process the question and yield the answer progressively.