Fix lint

python/mlc_llm/embeddings/embeddings.py

@@ -169,7 +169,7 @@ def embed(self, queries: List[str]) -> tvm.runtime.NDArray:
         return output
 
     def _tokenize_queries(self, queries: List[str]) -> Tuple[np.ndarray, np.ndarray]:
-        tokens = engine_utils.process_prompts(queries, self.tokenizer.encode)
+        tokens = engine_utils.process_prompts(queries, self.tokenizer.encode)  # type: ignore
         max_query_length = max(len(token_seq) for token_seq in tokens)
 
         token_inputs = np.zeros((len(tokens), max_query_length), dtype=np.int32)

python/mlc_llm/model/model.py

@@ -9,6 +9,7 @@
 from mlc_llm.quantization.quantization import Quantization
 
 from .baichuan import baichuan_loader, baichuan_model, baichuan_quantization
+from .bert import bert_loader, bert_model, bert_quantization
 from .chatglm3 import chatglm3_loader, chatglm3_model, chatglm3_quantization
 from .eagle import eagle_loader, eagle_model, eagle_quantization
 from .gemma import gemma_loader, gemma_model, gemma_quantization
@@ -27,7 +28,6 @@
 from .rwkv5 import rwkv5_loader, rwkv5_model, rwkv5_quantization
 from .rwkv6 import rwkv6_loader, rwkv6_model, rwkv6_quantization
 from .stable_lm import stablelm_loader, stablelm_model, stablelm_quantization
-from .bert import bert_loader, bert_model, bert_quantization
 
 ModelConfig = Any
 """A ModelConfig is an object that represents a model architecture. It is required to have