Update app.py

app.py

@@ -4,7 +4,7 @@ import torch
 from PIL import Image
 from pathlib import Path
 from threading import Thread
-from transformers import AutoTokenizer, AutoModelForCausalLM, AutoProcessor, Gemma3ForConditionalGeneration, TextIteratorStreamer
+from transformers import AutoModelForCausalLM, AutoProcessor, Gemma3ForConditionalGeneration, TextIteratorStreamer
 import spaces
 import time
 
@@ -14,17 +14,15 @@ DESCRIPTION= """
        """
 
 # model config
-model_270m_name = "google/gemma-3-270m-it"
-model_270m = AutoModelForCausalLM.from_pretrained(
-    model_270m_name,
+model_name = "google/gemma-3-270m-it"
+model = Gemma3ForConditionalGeneration.from_pretrained(
+    model_name,
     torch_dtype="auto",
     device_map="auto",
     attn_implementation="eager"
 ).eval()
 
-tokenizer = AutoTokenizer.from_pretrained(model_270m_name)
-
-processor_270m = AutoProcessor.from_pretrained(model_270m_name)
+processor = AutoProcessor.from_pretrained(model_name)
 # I will add timestamp later
 def extract_video_frames(video_path, num_frames=8):
     cap = cv2.VideoCapture(video_path)
@@ -104,9 +102,8 @@ def generate_response(input_data, chat_history, max_new_tokens, system_prompt, t
     if messages and messages[-1]["role"] == "user":
         messages[-1]["content"].extend(new_message["content"])
     else:
-        messages.append(new_message)
-    model = model_270m
-    processor = processor_270m
+        messages.append(new_message)    
+    
     inputs = processor.apply_chat_template(
         messages,
         add_generation_prompt=True,