Lewis Tunstall
@_lewtun
๐ค post-training @huggingface
๊ฐ์
August 2018
549 ํ๋ก์ ์ค 20.4K ํฌ
If you're using GRPO in TRL, you should really switch to the new async trainer. In our benchmarks, it's ~2-4x faster ๐ฅ
