Skip to content
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.

Sign up for GitHub

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

Jump to bottom

allow overriding of model_config parameters from the YML #853

Merged
merged 4 commits into from
Nov 16, 2023
Merged

allow overriding of model_config parameters from the YML #853

merged 4 commits into from
Nov 16, 2023

Conversation

winglian
Copy link
Collaborator

This should allow users to update values in the model config without having to edit the JSON.

for example for the attention_dropout for #672

model_config:
    attention_dropout: 0.01

NanoCode012
NanoCode012 reviewed Nov 15, 2023
View reviewed changes
Copy link
Collaborator

@NanoCode012 NanoCode012 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Nice clean PR! I think you'd want to add this to the docs as well.

Should we deprecate the current rope configs and move them here also?

src/axolotl/utils/models.py Outdated Show resolved Hide resolved
NanoCode012
NanoCode012 reviewed Nov 15, 2023
View reviewed changes
Copy link
Collaborator

@NanoCode012 NanoCode012 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

One more point to update would be the config_validate for backward compatible..

src/axolotl/utils/models.py Outdated Show resolved Hide resolved
@winglian
Copy link
Collaborator Author

Should we deprecate the current rope configs and move them here also?

which rope configs are you referring to?

@NanoCode012
Copy link
Collaborator

Should we deprecate the current rope configs and move them here also?

which rope configs are you referring to?

Ah, I meant rope_scaling. Since we now expect them in the model_overide, we should add it in our config validation check, so that it doesn't break old configs and warns user of deprecation.

@winglian winglian merged commit 1bc1186 into main Nov 16, 2023
4 checks passed
@winglian winglian deleted the model-config-overrides branch November 16, 2023 04:47
mkeoliya pushed a commit to mkeoliya/axolotl that referenced this pull request Dec 15, 2023
@winglian
allow overriding of model_config parameters from the YML (axolotl-ai-…
3b6dcd8 
…cloud#853)

* allow overriding of model_config parameters from the YML

* remove old logging, update readme

* move the updating of model config to the load_model_config function

* add warning for deprecated rope_scaling in the root of the YML config
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@NanoCode012 NanoCode012 NanoCode012 left review comments

Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.

None yet
2 participants
@winglian @NanoCode012