From b8eb8ccbb5b0ba7d9f8e49985c7d43b0a11842bd Mon Sep 17 00:00:00 2001 From: firecoperana <18252262+firecoperana@users.noreply.github.com> Date: Sat, 2 May 2026 10:19:59 -0500 Subject: [PATCH] server: revert llama_decode stop (#1722) Co-authored-by: firecoperana --- src/llama.cpp | 5 ----- 1 file changed, 5 deletions(-) diff --git a/src/llama.cpp b/src/llama.cpp index f7b55bbf..12299766 100644 --- a/src/llama.cpp +++ b/src/llama.cpp @@ -8477,11 +8477,6 @@ int32_t llama_encode( int32_t llama_decode( struct llama_context * ctx, struct llama_batch batch) { - // Clear any leftover stop signal from a previous (already-returned) decode. llama_decode_stop() - // is intended to interrupt the decode that is currently in flight; without this reset, a stop - // that arrived after the interrupted call returned would bleed into the next decode and cause - // an immediate ret=-3, which servers interpret as a fatal decode failure. - stop_internal_decode = false; const int ret = llama_decode_internal(*ctx, batch); if (ret < 0) { LLAMA_LOG_ERROR("%s: failed to decode, ret = %d\n", __func__, ret);