From 227760d7ec988c0c246b5089977fa1e9e51ca2f9 Mon Sep 17 00:00:00 2001 From: Yuan Tang Date: Tue, 11 Feb 2025 19:32:26 -0500 Subject: [PATCH] Update openai_compat.py --- llama_stack/providers/utils/inference/openai_compat.py | 1 + 1 file changed, 1 insertion(+) diff --git a/llama_stack/providers/utils/inference/openai_compat.py b/llama_stack/providers/utils/inference/openai_compat.py index ac9d18312..8ee838d84 100644 --- a/llama_stack/providers/utils/inference/openai_compat.py +++ b/llama_stack/providers/utils/inference/openai_compat.py @@ -175,6 +175,7 @@ def process_chat_completion_response( choice = response.choices[0] # TODO: This does not work well with tool calls for vLLM remote provider + # Ref: https://github.com/meta-llama/llama-stack/issues/1058 raw_message = formatter.decode_assistant_message_from_content( text_from_choice(choice), get_stop_reason(choice.finish_reason) )