Skip to content

Conversation

NathanHB
Copy link
Member

you should now be able to run a nanotron model with:

 torchrun --standalone --nnodes=1 --nproc-per-node=2  src/lighteval/__main__.py nanotron --checkpoint-config-path ../nanotron/checkpoints/10/config.yaml --lighteval-override examples/nanotron/lighteval_config_override_template.yaml

Copy link
Collaborator

@hynky1999 hynky1999 left a comment

Choose a reason for hiding this comment

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

LGTM 🚀.
I tested with fresh new checkpoint generated using nanotron with hash 4a2ddca23bcddb3fb35a1c7032cd6fc11f3ef54a.

But we really do need regression tests for nanotron as this will inevitably happen without them :(

Just small nitpick.

@hynky1999 hynky1999 self-requested a review September 2, 2024 13:07
@hynky1999
Copy link
Collaborator

lgtm

@NathanHB NathanHB merged commit cdeb6c2 into main Sep 2, 2024
2 checks passed
@clefourrier
Copy link
Member

Thanks a lot! Agree on the regression tests, a PR was started on this here #160 but I'm unsure where it stands now

hynky1999 added a commit that referenced this pull request May 22, 2025
* fix nanotron

* fix nanotron

* fix nanotron

* Update examples/nanotron/lighteval_config_override_template.yaml

Co-authored-by: Hynek Kydlíček <[email protected]>

---------

Co-authored-by: Hynek Kydlíček <[email protected]>
NathanHB added a commit that referenced this pull request Sep 19, 2025
* fix nanotron

* fix nanotron

* fix nanotron

* Update examples/nanotron/lighteval_config_override_template.yaml

Co-authored-by: Hynek Kydlíček <[email protected]>

---------

Co-authored-by: Hynek Kydlíček <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants