Update README.md

Author: EmilyWebber

9_rlhf/README.md

@@ -1 +1,4 @@
-
+# Reinforcement learning with human feedback
+Hi there! This directory has a few examples for you. First is a notebook that has all of the PyTorch code running the entire process, end-to-end on a single instance. This is just called `RLHF_locally.ipynb`, and you should be able to run this largely without error start to finish.
+ 
+As you can imagine this introduces complexities around managing the software, distributed the models, and so forth. So, we're working on another example to containerize this and make it easier for you to work with. That's in the subdirectory below, `wip`.