Update modeling_tf_longformer.py (#7359)

correct a very small mistake

src/transformers/modeling_tf_longformer.py

@@ -348,7 +348,7 @@ class TFLongformerSelfAttention(tf.keras.layers.Layer):
         # matrix multipication
         # bcxd: batch_size * num_heads x chunks x 2window_overlap x head_dim
         # bcyd: batch_size * num_heads x chunks x 2window_overlap x head_dim
-        # bcxy: batch_size * num_heads x chunks x 2window_overlap x window_overlap
+        # bcxy: batch_size * num_heads x chunks x 2window_overlap x 2window_overlap
         chunked_attention_scores = tf.einsum("bcxd,bcyd->bcxy", chunked_query, chunked_key)  # multiply
 
         # convert diagonals into columns