diff --git a/hezar/preprocessors/tokenizers/tokenizer.py b/hezar/preprocessors/tokenizers/tokenizer.py
index 3723fe20..6d5db78d 100644
--- a/hezar/preprocessors/tokenizers/tokenizer.py
+++ b/hezar/preprocessors/tokenizers/tokenizer.py
@@ -60,6 +60,7 @@ class Tokenizer(Preprocessor):
         "mask_token",
         "additional_special_tokens",
     ]
+    uncastable_keys = ["word_ids", "tokens", "offset_mapping"]
 
     def __init__(self, config: TokenizerConfig, **kwargs):
         super().__init__(config, **kwargs)
@@ -117,7 +118,6 @@ def pad_encoded_batch(
         Returns:
 
         """
-
         if isinstance(inputs, (list, tuple)) and isinstance(inputs[0], Mapping):
             inputs = {key: [example[key] for example in inputs] for key in inputs[0].keys()}
 
@@ -155,6 +155,7 @@ def pad_encoded_batch(
                 else:
                     inputs_length = max_length
 
+        skip_keys += self.uncastable_keys  # avoid possible errors
         inputs = convert_batch_dict_dtype(inputs, dtype="list", skip_keys=skip_keys)
 
         skip_keys = skip_keys or []
@@ -170,7 +171,7 @@ def pad_encoded_batch(
                 padded_batch.append(padded_x)
             inputs[key] = padded_batch
 
-        inputs = convert_batch_dict_dtype(inputs, dtype=return_tensors)
+        inputs = convert_batch_dict_dtype(inputs, dtype=return_tensors, skip_keys=skip_keys)
 
         return inputs
 
@@ -282,7 +283,7 @@ def __call__(
                 for key, value in sanitized_outputs.items()
             }
 
-        outputs = convert_batch_dict_dtype(sanitized_outputs, dtype=return_tensors)
+        outputs = convert_batch_dict_dtype(sanitized_outputs, dtype=return_tensors, skip_keys=self.uncastable_keys)
         if device and return_tensors == "pt":
             outputs = {k: v.to(device) if isinstance(v, torch.Tensor) else v for k, v in outputs.items()}