From 4ae172231c633f00be275133ad91094396d3280c Mon Sep 17 00:00:00 2001 From: eligotts <78387377+eligotts@users.noreply.github.com> Date: Sun, 30 Aug 2026 23:29:53 -0700 Subject: [PATCH] [Frontend] Forward cache salt for content parts (#54315) Signed-off-by: Eli Gotts <78387377+eligotts@users.noreply.github.com> Co-authored-by: OpenAI Codex --- vllm/entrypoints/scale_out/token_in_token_out/serving.py | 2 ++ 1 file changed, 2 insertions(+) diff --git a/vllm/entrypoints/scale_out/token_in_token_out/serving.py b/vllm/entrypoints/scale_out/token_in_token_out/serving.py index 631393a43d..90d12f118e 100644 --- a/vllm/entrypoints/scale_out/token_in_token_out/serving.py +++ b/vllm/entrypoints/scale_out/token_in_token_out/serving.py @@ -163,6 +163,8 @@ class ServingTokens(GenerateBaseServing): mm_parser.parse_video(url, uuid) mm_data, mm_uuids = await tracker.resolve_items() prompt = TokensPrompt(prompt_token_ids=request.token_ids) + if request.cache_salt is not None: + prompt["cache_salt"] = request.cache_salt if mm_data: prompt["multi_modal_data"] = mm_data if mm_uuids: