Update app.py

app.py

@@ -1,9 +1,7 @@
 import os
 import google.generativeai as genai
 import gradio as gr
-from gradio_multimodalchatbot import MultimodalChatbot
 from PIL import Image
-import io
 
 # Configure the API
 genai.configure(api_key=os.environ["GEMINI_API_KEY"])
@@ -17,7 +15,7 @@ generation_config = {
 }
 
 model = genai.GenerativeModel(
-    model_name="Gemini-1.5-Flash-8B-Exp-0924",
+    model_name="gemini-1.5-pro-latest",
     generation_config=generation_config,
 )
 
@@ -25,41 +23,38 @@ model = genai.GenerativeModel(
 chat_session = model.start_chat(history=[])
 
 def process_file(file):
-    if file.type.startswith('image'):
-        return Image.open(file.path)
-    elif file.type.startswith('audio') or file.type.startswith('video'):
-        return file.path
+    if file.name.lower().endswith(('.png', '.jpg', '.jpeg', '.gif', '.bmp')):
+        return Image.open(file.name)
+    elif file.name.lower().endswith(('.mp3', '.wav', '.ogg')):
+        return file.name  # Return the file path for audio
+    elif file.name.lower().endswith(('.mp4', '.avi', '.mov')):
+        return file.name  # Return the file path for video
     else:
         return None
 
-def respond(message, history):
-    files = []
-    for file in message.get('files', []):
-        processed_file = process_file(file['file'])
-        if processed_file:
-            files.append(processed_file)
-
-    prompt = message['text']
+def respond(message, chat_history, files):
+    processed_files = [process_file(file) for file in files if file is not None]
     
-    if files:
-        response = chat_session.send_message([prompt, *files])
+    if processed_files:
+        response = chat_session.send_message([message, *processed_files])
     else:
-        response = chat_session.send_message(prompt)
+        response = chat_session.send_message(message)
     
-    return {"text": response.text, "files": []}
+    chat_history.append((message, response.text))
+    return "", chat_history
 
 with gr.Blocks() as demo:
     gr.Markdown("# Gemini Multimodal Chatbot")
     gr.Markdown("Chat with the Gemini 1.5 Pro model. You can send text, images, audio, and video!")
     
-    chatbot = MultimodalChatbot(
-        height=600,
-        bubble_full_width=False,
-        avatar_images=(None, "https://lh3.googleusercontent.com/d/1pIo02xepBgqt9gMdFkJHSocJfH_A2dqj"),
-        render_markdown=True
-    )
+    chatbot = gr.Chatbot()
+    msg = gr.Textbox()
+    clear = gr.ClearButton([msg, chatbot])
     
-    chatbot.chat(respond, fill_height=False)
+    file_output = gr.File()
+    upload_button = gr.UploadButton("📁 Upload files", file_types=["image", "audio", "video"], file_count="multiple")
+
+    msg.submit(respond, [msg, chatbot, upload_button], [msg, chatbot])
 
 if __name__ == "__main__":
     demo.launch()