diff --git a/omniserve/engine/llm_engine.py b/omniserve/engine/llm_engine.py index 7937021..218ff9c 100644 --- a/omniserve/engine/llm_engine.py +++ b/omniserve/engine/llm_engine.py @@ -444,7 +444,7 @@ def _process_model_outputs( self.scheduler.free_seq(seqs[0]) request_outputs.append( { - "id": seqs[0].seq_id, + "id": seq_group.request_id, "text": seqs[0].output_text, "finished": True, } @@ -453,7 +453,7 @@ def _process_model_outputs( else: request_outputs.append( { - "id": seqs[0].seq_id, + "id": seq_group.request_id, "tokens": seqs[0].get_token_ids(), "finished": False, }