feat(LLM): Allow loading models that execute custom code

services/tokenizationService.py

@@ -37,7 +37,7 @@ class TokenizationService:
         model_config: ModelConfig,
     ) -> PreTrainedTokenizer | PreTrainedTokenizerFast:
         model_id = model_config.huggingface_id
-        return AutoTokenizer.from_pretrained(model_id)
+        return AutoTokenizer.from_pretrained(model_id, trust_remote_code=True)
 
     def tokenCount(self, text: str) -> int:
         """