[behnamoh]

this is the same team that a few months ago here on hacker news talked about how to do fine-tuning on large language models, and then made it close source.