LoRA Learns Less and Forgets Less Updated | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

		LoRA Learns Less and Forgets Less Updated (openreview.net)
		1 point by danielhanchen on Aug 27, 2024 \| hide \| past \| favorite \| 1 comment

danielhanchen on Aug 27, 2024 [–]

Updated research paper showing alpha = 2*rank for LoRA finetuning of LLMs gets much better results + other cool results!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact