Update app.py

app.py

@@ -21,9 +21,12 @@ MAX_MAX_NEW_TOKENS = 2048
 DEFAULT_MAX_NEW_TOKENS = 1024
 MAX_INPUT_TOKEN_LENGTH = int(os.getenv("MAX_INPUT_TOKEN_LENGTH", "4096"))
 
+auth_token = os.environ.get("HF_TOKEN")
+
 model_id = 'Qwen/Qwen2.5-VL-3B-Instruct'
 model = Qwen2_5_VLForConditionalGeneration.from_pretrained(
-    model_id,
+    'farrell236/test_model',
+    use_auth_token=auth_token,
     # torch_dtype=torch.bfloat16,
     # attn_implementation="flash_attention_2",
     device_map="auto"