-
Notifications
You must be signed in to change notification settings - Fork 544
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
Make mpt7b finetuning more obvious #101
Conversation
Note: #90 is merged |
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.
A couple minor suggestions to correct some issues with model configs but otherwise looks great!
BTW t5-small_dolly_sft.yaml
also has the device
instead of init_device
issue, but I couldn't suggest a change on it because it was just a rename. Please apply the same fix there.
* make mpt7b finetuning more obvious * change yaml structure and references to paths * needed to intall pre-commit * fix merge issue * Local dataset rework * Apply suggestions from code review * YAML touch ups --------- Co-authored-by: Vitaliy Chiley <6439018+vchiley@users.noreply.github.com> Co-authored-by: Alex Trott <alex@mosaicml.com>
Resolves RESEARCH-710
Also change the folder structure of the YAMLs to make it more clear which MPTs are pretrained from the HF Hub and which are an architecture for pretraining.
WARNING: This will not work until #90 is merged!