From 293e4e4b09c014ff8ae52a0ebd039b94ce522b29 Mon Sep 17 00:00:00 2001
From: Frank <frank@dankware>
Date: Mon, 21 Sep 2026 20:07:13 -0700
Subject: [PATCH] frank-large: wire Qwen-Sharp v22.4.0 template via
 --chat-template-file

The serve script only passed --chat-template-kwargs, so the binary fell
back to the built-in peg-native template. A 250-token request produced a
73,600-token prompt (280x the expected ~260). Wiring in --chat-template-file
drops the prompt to 62 tokens. Draft acceptance 0.79 (mean 4.14), model
answers correctly, reasoning lands in reasoning_content.
---
 scripts/serve-frank-large.sh | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/scripts/serve-frank-large.sh b/scripts/serve-frank-large.sh
index 2b18ac5..a114a0b 100755
--- a/scripts/serve-frank-large.sh
+++ b/scripts/serve-frank-large.sh
@@ -50,7 +50,7 @@ fi
 : "${PARALLEL_SLOTS:=1}"
 : "${SPEC_DRAFT_N_MAX:=4}"
 : "${COMPATIBILITY_MODE:=1}"
-: "${CACHE_RAM:=8192}"
+TEMPLATE_FILE="${TEMPLATE_FILE:-$SCRIPT_DIR/../dankserver/frank-medium/qwen-sharp-v22.4.0.jinja}"
 
 SERVER_BIN="$RUNTIME_DIR/bin/llama-server"
 MODEL_FILE="$MODEL_DIR/Qwen3.8-27B-IU4-Kairic-Edge.gguf"
@@ -113,6 +113,7 @@ ARGS=(
     -m "$MODEL_FILE" --alias frank-large
     --host "$HOST" --port "$PORT" --jinja
     --chat-template-kwargs '{"preserve_thinking":true}'
+    --chat-template-file "$TEMPLATE_FILE"
     -dev ROCm0 -ngl 999
     -c "$CONTEXT_SIZE" -b 2048 -ub 512 -fa on -ctk f16 -ctv f16
     -t 16 -tb 32 -np "$PARALLEL_SLOTS" -ctxcp 32 --cache-ram "$CACHE_RAM"