Adding _set_gradient_checkpointing for compatibility

modeling_mixformer_sequential.py

@@ -712,6 +712,10 @@ class MixFormerSequentialPreTrainedModel(PreTrainedModel):
             "attention_mask": attention_mask,
         }
     
+    def _set_gradient_checkpointing(self, module, value=False):
+            if isinstance(module, MixFormerSequentialPreTrainedModel):
+                module.gradient_checkpointing = value
+
 
 class MixFormerSequentialForCausalLM(MixFormerSequentialPreTrainedModel):
     """MixFormer (sequential for DeepSpeed) for Causal Language Modeling."""