From 6cef600d25a30ae73a7673c2f4843d46b2fb2f0a Mon Sep 17 00:00:00 2001 From: ginnoir Date: Sat, 27 Jun 2026 00:09:09 -0500 Subject: [PATCH] fix(llm): add --jinja so gpt-oss harmony template returns content Co-Authored-By: Claude Opus 4.8 --- stacks/llm/llama-swap-config.yaml | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/stacks/llm/llama-swap-config.yaml b/stacks/llm/llama-swap-config.yaml index d8caee7..1b1daec 100644 --- a/stacks/llm/llama-swap-config.yaml +++ b/stacks/llm/llama-swap-config.yaml @@ -15,7 +15,9 @@ healthCheckTimeout: 300 logLevel: info macros: - common: "-ngl 99 --parallel 1 --ctx-size 65536 --flash-attn on --cache-type-k q8_0 --cache-type-v q8_0" + # --jinja applies each model's embedded chat template (REQUIRED for gpt-oss' + # harmony format, else content comes back empty; harmless/correct for gemma4). + common: "-ngl 99 --parallel 1 --ctx-size 65536 --flash-attn on --cache-type-k q8_0 --cache-type-v q8_0 --jinja" models: "gpt-oss-20b":