Update app.py

app.py

@@ -17,7 +17,7 @@ import re
 from gradio_client import Client
 from simple_salesforce import Salesforce, SalesforceLogin
 from llama_index.llms.huggingface import HuggingFaceLLM
-from llama_index.llms.huggingface import HuggingFaceInferenceAPI
+# from llama_index.llms.huggingface import HuggingFaceInferenceAPI
 
 
 # Define Pydantic model for incoming request body
@@ -72,7 +72,7 @@ app.mount("/static", StaticFiles(directory="static"), name="static")
 
 templates = Jinja2Templates(directory="static")
 # Configure Llama index settings
-Settings.llm = HuggingFaceInferenceAPI(
+Settings.llm = HuggingFaceLLM(
     model_name="meta-llama/Meta-Llama-3-8B-Instruct",
     tokenizer_name="meta-llama/Meta-Llama-3-8B-Instruct",
     context_window=3000,