From 83ed41ee761001b36d4b79a674db21e616e79750 Mon Sep 17 00:00:00 2001
From: peixingyu <i-peixingyu@4paradigm.com>
Date: Fri, 24 Apr 2026 07:43:54 +0000
Subject: [PATCH] Add Biren166M vllm patch-tokenizer image

---
 Dockerfile |  4 ++++
 patch.py   | 19 +++++++++++++++++++
 2 files changed, 23 insertions(+)
 create mode 100644 Dockerfile
 create mode 100644 patch.py

diff --git a/Dockerfile b/Dockerfile
new file mode 100644
index 0000000..b49a1c7
--- /dev/null
+++ b/Dockerfile
@@ -0,0 +1,4 @@
+FROM git.modelhub.org.cn:9443/enginex/xc-llm-biren166m:26.01
+
+COPY patch.py /tmp/patch.py
+RUN python3 /tmp/patch.py
diff --git a/patch.py b/patch.py
new file mode 100644
index 0000000..5aa6ea5
--- /dev/null
+++ b/patch.py
@@ -0,0 +1,19 @@
+path = '/usr/local/lib/python3.10/dist-packages/transformers/tokenization_utils_base.py'
+with open(path, 'r') as f:
+    content = f.read()
+
+old = ('        self.SPECIAL_TOKENS_ATTRIBUTES = self.SPECIAL_TOKENS_ATTRIBUTES'
+       ' + list(special_tokens.keys())')
+new = ('        # PATCH: some models have extra_special_tokens as list instead of dict\n'
+       '        if isinstance(special_tokens, list):\n'
+       '            special_tokens = {t: t for t in special_tokens}\n'
+       '        self.SPECIAL_TOKENS_ATTRIBUTES = self.SPECIAL_TOKENS_ATTRIBUTES'
+       ' + list(special_tokens.keys())')
+
+if old in content:
+    content = content.replace(old, new)
+    with open(path, 'w') as f:
+        f.write(content)
+    print('Patch applied successfully')
+else:
+    print('WARNING: pattern not found')