fix: route chat completions to free serverless Inference API

By default, recent huggingface_hub versions route InferenceClient.chat_completion()
to router.huggingface.co, which requires paid providers. Explicitly passing
base_url='https://api-inference.huggingface.co' forces it to use the free serverless
inference endpoint.

shared/inference_client.py

@@ -106,6 +106,7 @@ def _get_client():
     return InferenceClient(
         model=INFERENCE_MODEL,
         token=HF_TOKEN,
+        base_url="https://api-inference.huggingface.co",
     )
 def generate(
     project: str,