Fine-tune Language Models with Layer Freezing in run_clm.py

Hi,

I want to fine-tune a language model with run_clm.py script. Based on my understanding, the script does not provide layer freezing options.

Is there any suggestion of how to do it?

Many thanks in advance!