One Line Fix: missing text output with huggingface TGI LLM (#7972)

Small bug fix. The async _call method was missing a line to return the
generated text.

@baskaryan

langchain/llms/huggingface_text_gen_inference.py

@@ -255,4 +255,5 @@ class HuggingFaceTextGenInference(LLM):
                 if not token.special:
                     if text_callback:
                         await text_callback(token.text)
+                    text += token.text
         return text