Fix gradient checkpointing bug in git (#21818)

Co-authored-by: Sylvain Gugger <35901082+sgugger@users.noreply.github.com>

src/transformers/models/git/modeling_git.py

@@ -431,6 +431,13 @@ class GitEncoder(nn.Module):
         pixel_values_present: Optional[bool] = False,
         return_dict: Optional[bool] = True,
     ) -> Union[Tuple[torch.Tensor], BaseModelOutputWithPast]:
+        if self.gradient_checkpointing and self.training:
+            if use_cache:
+                logger.warning_once(
+                    "`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`..."
+                )
+                use_cache = False
+
         all_hidden_states = () if output_hidden_states else None
         all_self_attentions = () if output_attentions else None
 
@@ -443,11 +450,6 @@ class GitEncoder(nn.Module):
             past_key_value = past_key_values[i] if past_key_values is not None else None
 
             if self.gradient_checkpointing and self.training:
-                if use_cache:
-                    logger.warning_once(
-                        "`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`..."
-                    )
-                    use_cache = False
 
                 def create_custom_forward(module):
                     def custom_forward(*inputs):