gguf-py : fail fast on nonsensical special token IDs (ggerganov#4489)

sealad886 · Dec 17, 2023 · f7f468a · f7f468a
1 parent 919c406
commit f7f468a
Showing 1 changed file with 3 additions and 1 deletion.
diff --git a/gguf-py/gguf/vocab.py b/gguf-py/gguf/vocab.py
@@ -109,8 +109,10 @@ def _try_load_merges_txt(self, path: Path) -> bool:
         return True
 
     def _set_special_token(self, typ: str, tid: Any) -> None:
-        if not isinstance(tid, int) or tid < 0:
+        if not isinstance(tid, int):
             return
+        if tid < 0:
+            raise ValueError(f'invalid value for special token type {typ}: {tid}')
         if self.n_vocab is None or tid < self.n_vocab:
             if typ in self.special_token_ids:
                 return