From 2c78bbe8c0687c864cbadfc3afa850c0f9171819 Mon Sep 17 00:00:00 2001
From: huni <chenweize1202@gmail.com>
Date: Mon, 24 Aug 2026 18:07:05 +0800
Subject: [PATCH] add entrypoint.sh

---
 entrypoint.sh | 39 +++++++++++++++++++++++++++++++++++++++
 1 file changed, 39 insertions(+)
 create mode 100644 entrypoint.sh

diff --git a/entrypoint.sh b/entrypoint.sh
new file mode 100644
index 0000000..07308d3
--- /dev/null
+++ b/entrypoint.sh
@@ -0,0 +1,39 @@
+#!/bin/bash
+set -e
+
+MODEL_DIR=${1:-/model}
+shift || true
+
+FIX_TOKENIZER_DIR=/tmp/fixed_tokenizer
+AUTO_FIX=${AUTO_FIX_TOKENIZER:-auto}
+
+echo "[entrypoint] model dir: $MODEL_DIR"
+
+NEED_FIX=0
+
+if [ "$AUTO_FIX" = "1" ] || [ "$AUTO_FIX" = "true" ]; then
+    NEED_FIX=1
+elif [ "$AUTO_FIX" = "auto" ]; then
+    if [ -f "$MODEL_DIR/tokenizer_config.json" ]; then
+        if grep -q "TokenizersBackend\|TiktokenTokenizer" "$MODEL_DIR/tokenizer_config.json"; then
+            NEED_FIX=1
+        fi
+        # 检测 extra_special_tokens 是否为 list 格式
+        if grep -q '"extra_special_tokens":\s*\[' "$MODEL_DIR/tokenizer_config.json"; then
+            NEED_FIX=1
+        fi
+    fi
+fi
+
+if [ $NEED_FIX -eq 1 ]; then
+    echo "[entrypoint] fixing tokenizer..."
+    python3 /opt/fix_tokenizer.py
+    TOKENIZER_ARG="--tokenizer $FIX_TOKENIZER_DIR"
+else
+    echo "[entrypoint] tokenizer OK, skip fix"
+    TOKENIZER_ARG=""
+fi
+
+echo "[entrypoint] starting vllm..."
+
+exec vllm serve "$MODEL_DIR" $TOKENIZER_ARG "$@"