Skip to content

[ENH] Add multiloss support for v2 by adding one more dimension to x and y - #2356

Draft
phoeenniixx wants to merge 2 commits into
sktime:mainfrom
phoeenniixx:multiloss-v2
Draft

[ENH] Add multiloss support for v2 by adding one more dimension to x and y#2356
phoeenniixx wants to merge 2 commits into
sktime:mainfrom
phoeenniixx:multiloss-v2

Conversation

@phoeenniixx

Copy link
Copy Markdown
Member

This PR adds support for MultiLoss to ptf v2

Fixes #2308

Changes

  • Add weights to the collate_fn and __getitem__ (of private dataset class of datamodule) return for MultiLoss. The return would be (x, (y, weight)) in place of (x,y).
    Done to:
    • EncoderDecoderTimeSeriesDataModule
    • TslibDataModule
  • Add support to BaseModel.
  • Add tests and update the test framework - maybe should happen in a new PR instead - along with fixes if needed?

@phoeenniixx phoeenniixx added enhancement New feature or request module:models module:datasets&dataloaders ptf-v2 Related to `pytorch-forecasting` v2 labels Jul 27, 2026
@phoeenniixx

Copy link
Copy Markdown
Member Author

tests will fail as i still need to add support to BaseModel

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

enhancement New feature or request module:datasets&dataloaders module:models ptf-v2 Related to `pytorch-forecasting` v2

Projects

None yet

Development

Successfully merging this pull request may close these issues.

[ENH] Add MultiLoss and multi-target support to ptf-v2

2 participants