Incorrect paper link in summarize_rlhf example README.MD

[jboboyle]

📚 The doc issue

The example summarize_rlhf readme has an incorrect link to the Stiennon et al.'s Learning to Summarize from Human Feedback. It currently links to: https://arxiv.org/abs/2106.00987

Suggest a potential alternative/fix

Link to: https://arxiv.org/abs/2009.01325

[jon-tow]

Updated. Thank you!