TOS are not protection against scraping. Is there any reason to not assume by default that Codeberg gets used for training just like every other online collection of content?