Merge branch 'main' of https://huggingface.co/spaces/TIGER-Lab/Mantis

README.md

@@ -8,6 +8,7 @@ sdk_version: 4.24.0
 app_file: app.py
 pinned: false
 license: apache-2.0
+short_description: Multimodal Language Model
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

app.py

@@ -3,6 +3,7 @@ import spaces
 import os
 import time
 from PIL import Image
+import functools
 from models.mllava import MLlavaProcessor, LlavaForConditionalGeneration, chat_mllava, MLlavaForConditionalGeneration
 from typing import List
 processor = MLlavaProcessor.from_pretrained("TIGER-Lab/Mantis-8B-siglip-llama3")
@@ -48,13 +49,15 @@ def get_chat_history(history):
                 chat_history.append({"role": "assistant", "text": ""})
     return chat_history
 
+
 def get_chat_images(history):
     images = []
     for message in history:
         if isinstance(message[0], tuple):
             images.extend(message[0])
     return images
-            
+
+
 def bot(history):
     print(history)
     cur_messages = {"text": "", "images": []}
@@ -93,6 +96,8 @@ def bot(history):
         history[-1][1] = _output
         time.sleep(0.05)
         yield history
+
+
         
 def build_demo():
     with gr.Blocks() as demo:
@@ -118,14 +123,35 @@ def build_demo():
         chat_input = gr.MultimodalTextbox(interactive=True, file_types=["image"], placeholder="Enter message or upload images. Please use <image> to indicate the position of uploaded images", show_label=True)
         
         chat_msg = chat_input.submit(add_message, [chatbot, chat_input], [chatbot, chat_input])
+        
+        """
+        with gr.Accordion(label='Advanced options', open=False):
+            temperature = gr.Slider(
+                label='Temperature',
+                minimum=0.1,
+                maximum=2.0,
+                step=0.1,
+                value=0.2,
+                interactive=True
+            )
+            top_p = gr.Slider(
+                label='Top-p',
+                minimum=0.05,
+                maximum=1.0,
+                step=0.05,
+                value=1.0,
+                interactive=True
+            )
+        """
+
         bot_msg = chat_msg.success(bot, chatbot, chatbot, api_name="bot_response")
         
         chatbot.like(print_like_dislike, None, None)
-        
+
         with gr.Row():
             send_button = gr.Button("Send")
             clear_button = gr.ClearButton([chatbot, chat_input])
-        
+
         send_button.click(
             add_message, [chatbot, chat_input], [chatbot, chat_input]
         ).then(
@@ -136,7 +162,7 @@ def build_demo():
             examples=[
                 {
                     "text": "<image> <image> How many dices are there in image 1 and image 2 respectively?",
-                    "files": ["./examples/image10.jpg", "./examples/image11.jpg"]
+                    "files": ["./examples/image10.jpg", "./examples/image15.jpg"]
                 },
                 {
                     "text": "<image> <image> <image> Which image shows a different mood of character from the others?",