Added parallel device usage for GPT-J (#22713)

src/transformers/models/gptj/modeling_gptj.py

@@ -1012,6 +1012,7 @@ class GPTJForSequenceClassification(GPTJPreTrainedModel):
 
         loss = None
         if labels is not None:
+            labels = labels.to(pooled_logits.device)
             if self.config.problem_type is None:
                 if self.num_labels == 1:
                     self.config.problem_type = "regression"