Update modeling_dbrx.py

gradient_checkpointing fixed

modeling_dbrx.py

@@ -1093,13 +1093,13 @@ class DbrxModel(DbrxPreTrainedModel):
                 block_outputs = self._gradient_checkpointing_func(
                     block.__call__,
                     hidden_states,
-                    attention_mask=causal_mask,
-                    position_ids=position_ids,
-                    past_key_values=past_key_values,
-                    output_attentions=output_attentions,
-                    output_router_logits=output_router_logits,
-                    use_cache=use_cache,
-                    cache_position=cache_position,
+                    causal_mask,
+                    position_ids,
+                    past_key_values,
+                    output_attentions,
+                    output_router_logits,
+                    use_cache,
+                    cache_position,
                 )
             else:
                 block_outputs = block(