Update README.md

Update README.md

Upload sample train val file for training reward models on relative feedback data

Upload relative gpt3.5-turbo feedback data

add absolute gpt3.5-turbo feedback data

initial commit