Update app.py

app.py

@@ -60,13 +60,15 @@ client = InferenceClient(HF_MODEL_ID, token=os.getenv("HF_TOKEN"))
 import os
 from langchain_huggingface import HuggingFaceEndpoint, ChatHuggingFace
 
+hf_token = os.getenv("HF_TOKEN")
+
 # 1. Create the endpoint with the 'conversational' task
 llm_endpoint = HuggingFaceEndpoint(
     #repo_id="mistralai/Mistral-7B-Instruct-v0.3",
     repo_id = "HuggingFaceH4/zephyr-7b-beta",
-    task="conversational",  # Change this from 'text-generation'
+    task="text-generation",
     max_new_tokens=512,
-    huggingfacehub_api_token=os.getenv("HF_TOKEN"),
+    huggingfacehub_api_token=hf_token,
 )
 
 # 2. Wrap it in ChatHuggingFace