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.

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

feat(train): create single entrypoint #553

Merged
merged 5 commits into from
Nov 26, 2023

Conversation

akhileshh
Copy link
Contributor

No description provided.

Copy link

codecov bot commented Nov 24, 2023

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (38b1187) 100.00% compared to head (3f23966) 100.00%.

Additional details and impacted files
@@            Coverage Diff            @@
##              main      #553   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files          128       128           
  Lines         4284      4284           
=========================================
  Hits          4284      4284           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

num_nodes: int = 1,
nproc_per_node: int = 1,
retry_count: int = 3,
local_run: bool = False,
Copy link
Member

Choose a reason for hiding this comment

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

probably should be True by default

torch_launcher_api.elastic_launch(config, _parse_spec_and_train)()


def _lightning_train(
Copy link
Member

Choose a reason for hiding this comment

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

maybe rename to _lightning_train_local

"full_state_ckpt_path": full_state_ckpt_path,
}

_create_ddp_master_job(
Copy link
Member

Choose a reason for hiding this comment

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

maybe rename to _lightning_train_remote


train_spec = {
"@type": "lightning_train",
"regime": regime,
Copy link
Member

Choose a reason for hiding this comment

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

should prolly use [builder lookup here

@akhileshh akhileshh requested a review from nkemnitz November 25, 2023 16:36
"regime": builder.get_initial_builder_spec(regime),
"trainer": builder.get_initial_builder_spec(trainer),
"train_dataloader": builder.get_initial_builder_spec(train_dataloader),
"val_dataloader": builder.get_initial_builder_spec(val_dataloader),
Copy link
Member

Choose a reason for hiding this comment

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

I think we gotta check if any of these are None here

Copy link
Member

@supersergiy supersergiy left a comment

Choose a reason for hiding this comment

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

💯

@supersergiy supersergiy merged commit fb45713 into main Nov 26, 2023
11 checks passed
@akhileshh akhileshh deleted the akhilesh/single-training-entrypoint branch November 26, 2023 21:16
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.

2 participants