Sort the outputs before return (#402)

b6fbb9a5 · Woosuk Kwon · GitHub · 2179e4f4 · b6fbb9a5
Unverified Commit b6fbb9a5 authored Jul 08, 2023 by Woosuk Kwon Committed by GitHub Jul 08, 2023
Show whitespace changes
Inline Side-by-side

Showing with 4 additions and 0 deletions

vllm/entrypoints/llm.py vllm/entrypoints/llm.py +4 -0

No files found.
--- a/vllm/entrypoints/llm.py
+++ b/vllm/entrypoints/llm.py
@@ -155,4 +155,8 @@ class LLM:
                        pbar.update(1)
        if use_tqdm:
            pbar.close()
+        # Sort the outputs by request ID.
+        # This is necessary because some requests may be finished earlier than
+        # its previous requests.
+        outputs = sorted(outputs, key=lambda x: int(x.request_id))
        return outputs