Update app.py

app.py

@@ -1,4 +1,4 @@
-from smolagents import CodeAgent, DuckDuckGoSearchTool, HfApiModel, load_tool, tool, LiteLLMModel, LocalModel
+from smolagents import CodeAgent, DuckDuckGoSearchTool, HfApiModel, load_tool, tool, LiteLLMModel, TransformersModel
 import datetime
 import requests
 import pytz
@@ -72,11 +72,12 @@ ddg = DuckDuckGoSearchTool()
 #     temperature=0.5                           
 # )
 
-model = LocalModel(
-    model_id="TheBloke/vicuna-7B-1.1-HF",  # or another HF repo you prefer
-    trust_remote_code=True,               # allow custom model code
-    device="cuda",                        # or "cpu" if no GPU
-    quantize="4bit"                       # drastically reduce VRAM use
+model = TransformersModel(
+    model_id="TheBloke/vicuna-7B-1.1-HF",  # or any other Hugging Face repo
+    device="cuda",                        # or "cpu"
+    max_new_tokens=1024,                  # how many tokens to generate
+    temperature=0.5                       # same as before
+    # quantize="4bit"                     # optional if supported
 )
 
 # Import tool from Hub