Text lowering issue

I found an issue here https://github.com/google-research/big_vision/blob/main/big_vision/pp/ops_text.py#L165
When lowering UTF-8 non-latin text `encoding ='utf-8'` should be used as mentioned here https://www.tensorflow.org/api_docs/python/tf/strings/lower .

This at least can influence at i18n model.
But due to models already trained, i'm not sure if this issue should be fixed.

Start at big_vision/pp/ops_text.py around line 165 and compare the current lowering call with the TensorFlow tf.strings.lower documentation linked in the issue. Verify the behavior for UTF-8 non-Latin text and consider the stated compatibility concern for already-trained models; done means the intended behavior and whether a fix is appropriate are established.

Written by the indexing model from the issue text.