diff --git a/gliner/modeling/encoder.py b/gliner/modeling/encoder.py
index 2852e3c..1717053 100644
--- a/gliner/modeling/encoder.py
+++ b/gliner/modeling/encoder.py
@@ -943,8 +943,8 @@ def encode_labels(
             input_ids: Label token IDs of shape (batch_size, seq_len).
             attention_mask: Attention mask of shape (batch_size, seq_len).
             *args: Additional positional arguments.
-            **kwargs: Additional keyword arguments (packing_config and pair_attention_mask
-                are removed as they're not supported for labels).
+            **kwargs: Additional keyword arguments (packing_config, pair_attention_mask
+                and token_lengths are removed as they're not supported for labels).
 
         Returns:
             Pooled label embeddings of shape (batch_size, hidden_size).
@@ -952,6 +952,11 @@ def encode_labels(
         label_kwargs = dict(kwargs)
         label_kwargs.pop("packing_config", None)
         label_kwargs.pop("pair_attention_mask", None)
+        # token_lengths carries precomputed lengths for the *text* inputs (see
+        # collator); forwarding it to the labels encoder crashes plain HF
+        # backbones ("BertModel.forward() got an unexpected keyword argument
+        # 'token_lengths'", #370).
+        label_kwargs.pop("token_lengths", None)
         label_kwargs["attention_mask"] = attention_mask
         labels_embeddings = self.labels_encoder(input_ids, *args, **label_kwargs)
         if hasattr(self, "labels_projection"):