Update app.py

app.py

@@ -58,10 +58,12 @@ def bot_streaming(message, history):
   if len(image) == 1:
     if image[0].endswith(video_extensions):
         
-        image = sample_frames(image[0], 32)
+        video = sample_frames(image[0], 32)
+        image = None
         prompt = f"<|im_start|>user <video>\n{message.text}<|im_end|><|im_start|>assistant"
     elif image[0].endswith(image_extensions):
         image = Image.open(image[0]).convert("RGB")
+        video = None
         prompt = f"<|im_start|>user <image>\n{message.text}<|im_end|><|im_start|>assistant"
 
   elif len(image) > 1:
@@ -82,9 +84,10 @@ def bot_streaming(message, history):
     prompt = "<|im_start|>user"+ toks + f"\n{user_prompt}<|im_end|><|im_start|>assistant"
 
     image = image_list
+    video = None
 
 
-  inputs = processor(prompt, image, return_tensors="pt").to("cuda", torch.float16)
+  inputs = processor(text=prompt, images=image, videos=video, return_tensors="pt").to("cuda", torch.float16)
   streamer = TextIteratorStreamer(processor, **{"max_new_tokens": 200, "skip_special_tokens": True})
   generation_kwargs = dict(inputs, streamer=streamer, max_new_tokens=100)
   generated_text = ""