From 2741eee1abf9f805948fa546b68aec4bad8f173a Mon Sep 17 00:00:00 2001 From: ahmedshakill Date: Mon, 14 Apr 2025 14:34:46 +0600 Subject: [PATCH] set b = ub when b > ub with embedding --- examples/server/server.cpp | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/examples/server/server.cpp b/examples/server/server.cpp index 1bf1ee876b40f..72a3e71d651b7 100644 --- a/examples/server/server.cpp +++ b/examples/server/server.cpp @@ -3002,6 +3002,12 @@ struct server_context { } if (slot.is_non_causal()) { + if (slot.task_type == SERVER_TASK_TYPE_EMBEDDING) { + if(n_batch > n_ubatch){ + SRV_WRN("%s","setting batch= ubatch \n"); + n_batch = n_ubatch; + } + } if (slot.n_prompt_tokens > n_ubatch) { slot.release(); send_error(slot, "input is too large to process. increase the physical batch size", ERROR_TYPE_SERVER);