From a80da2e182164bcde41799539244fc230e1985ec Mon Sep 17 00:00:00 2001 From: "Rodrigo Rodriguez (Pragmatismo)" Date: Wed, 12 Nov 2025 18:37:01 -0300 Subject: [PATCH] feat(llm): add jinja flag to local server args Added the --jinja flag to the LLM server startup arguments to enable Jinja template support. This allows for more flexible prompt formatting when using the local LLM server. The change maintains all existing functionality while adding the new feature. --- src/llm/local.rs | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/src/llm/local.rs b/src/llm/local.rs index cf395b6b0..cf5269dda 100644 --- a/src/llm/local.rs +++ b/src/llm/local.rs @@ -234,14 +234,18 @@ pub async fn start_llm_server( .unwrap_or("4096".to_string()); // TODO: Move flash-attn, temp, top_p, repeat-penalty to config as well. - + // TODO: Create --jinja. + let mut args = format!( - "-m {} --host 0.0.0.0 --port {} --top_p 0.95 --flash-attn on --temp 0.6 --repeat-penalty 1.2 --n-gpu-layers {}", + "-m {} --host 0.0.0.0 --port {} --top_p 0.95 --jinja --flash-attn on --temp 0.6 --repeat-penalty 1.2 --n-gpu-layers {}", model_path, port, gpu_layers ); if !reasoning_format.is_empty() { args.push_str(&format!(" --reasoning-format {}", reasoning_format)); } + + + if n_moe != "0" { args.push_str(&format!(" --n-cpu-moe {}", n_moe)); }