Update utils/inference.py

utils/inference.py

@@ -1,6 +1,11 @@
 from huggingface_hub import InferenceClient
+import os
 
-client = InferenceClient("mistralai/Mistral-7B-Instruct-v0.2")
+client = InferenceClient(
+    model="mistralai/Mistral-7B-Instruct-v0.2",
+    token=os.environ.get("HF_TOKEN"),
+    provider="hf-inference"
+)
 
 def call_model(prompt: str) -> str:
     return client.text_generation(prompt, max_new_tokens=2048, temperature=0.3, return_full_text=False)