-
Notifications
You must be signed in to change notification settings - Fork 16
Migrate train_on_inputs to sft-specific params #297
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
base: main
Are you sure you want to change the base?
Migrate train_on_inputs to sft-specific params #297
Conversation
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.
formatting only
4b502d1
to
8ef8769
Compare
8ef8769
to
1ce13d6
Compare
) | ||
train_on_inputs = "auto" | ||
|
||
if dpo_beta is not None and training_method != "dpo": |
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.
Other option might be just a warning. What do you think?
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.
Don't forget version bump up: https://github.com/togethercomputer/together-python/blob/main/pyproject.toml#L15
|
||
if train_on_inputs is None and training_method == "sft": | ||
log_warn_once( | ||
"train_on_inputs is not set for SFT training, it will be set to 'auto' automatically" |
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.
"train_on_inputs is not set for SFT training, it will be set to 'auto' automatically" | |
"train_on_inputs is not set for SFT training, it will be set to 'auto'" |
this PR adjusts the behavior of train_on_inputs.
if train type is SFT, we include this parameter in the TrainingMethod and default it to auto.
if train type is DPO, we default to None and raise if parameter is supplied.