Update app.py

app.py

@@ -39,6 +39,10 @@ from PIL import Image
 from io import BytesIO
 from transformers import TextStreamer, TextIteratorStreamer
 
+import hashlib
+import PIL
+import base64
+
 import gradio as gr
 import gradio_client
 import subprocess
@@ -226,25 +230,9 @@ def bot(history):
     assert len(images_this_term) > 0, "must have an image"
     # image_files = (args.image_file).split(',')
     # image = [load_image(f) for f in images_this_term if f]
-    image_list = []
-    for f in images_this_term:
-        if is_valid_video_filename(f):
-            image_list += sample_frames(f, our_chatbot.num_frames)
-        elif is_valid_image_filename(f):
-            image_list.append(load_image(f))
-        else:
-            raise ValueError("Invalid image file")
-    
-    image_tensor = [
-        our_chatbot.image_processor.preprocess(f, return_tensors="pt")["pixel_values"][
-            0
-        ]
-        .half()
-        .to(our_chatbot.model.device)
-        for f in image_list
-    ]
+
     all_image_hash = []
-    for image_path in image_list:
+    for image_path in images_this_term:
         with open(image_path, "rb") as image_file:
             image_data = image_file.read()
             image_hash = hashlib.md5(image_data).hexdigest()
@@ -261,6 +249,25 @@ def bot(history):
             if not os.path.isfile(filename):
                 os.makedirs(os.path.dirname(filename), exist_ok=True)
                 image.save(filename)
+    
+    image_list = []
+    for f in images_this_term:
+        if is_valid_video_filename(f):
+            image_list += sample_frames(f, our_chatbot.num_frames)
+        elif is_valid_image_filename(f):
+            image_list.append(load_image(f))
+        else:
+            raise ValueError("Invalid image file")
+    
+    image_tensor = [
+        our_chatbot.image_processor.preprocess(f, return_tensors="pt")["pixel_values"][
+            0
+        ]
+        .half()
+        .to(our_chatbot.model.device)
+        for f in image_list
+    ]
+
 
     image_tensor = torch.stack(image_tensor)
     image_token = DEFAULT_IMAGE_TOKEN * num_new_images