Fix T5 incorrect weight decay in Trainer and official summarization example (#18002)

* Add ALL_LAYERNORM_LAYERS for LayerNorm

* fix bug of appending layer norm

examples/pytorch/summarization/run_summarization_no_trainer.py

@@ -526,7 +526,7 @@ def main():
 
     # Optimizer
     # Split weights in two groups, one with weight decay and the other not.
-    no_decay = ["bias", "LayerNorm.weight"]
+    no_decay = ["bias", "LayerNorm.weight", "layer_norm.weight"]
     optimizer_grouped_parameters = [
         {
             "params": [p for n, p in model.named_parameters() if not any(nd in n for nd in no_decay)],