-
Notifications
You must be signed in to change notification settings - Fork 365
Flores 101 prompts #779
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Flores 101 prompts #779
Conversation
…pletion tasks (simpler?)
…pletion tasks (simpler?)
… to be done separately later. Also some slight modifs such as removing excess words and quotes
Adding @fyvo for info |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
looks good to me, thank you @rbawden !
(nice usage of anchors in yaml btw! :) )
I don't have a strong opinion on the format so will defer to the evaluation folks!
(Actually python seems to have dealt with that automatically since I created the templates automatically, so I won't take too much credit)! |
…tsource into flores-101-prompts
thank you @rbawden for the fixes! |
Automatically created prompts for MT using the Flores-101 dataset.
Contains prompts for all language directions using 31 BigScience languages (1 English prompt per language direction = 930 templates).
Question: is this format ok or should I separate out those that are for inclusion in the upcoming evaluation (i.e. only into and from English). This depends on how the notion of subtask is going defined and whether there is a possibility of selecting only certain templates.