chore(main): Release training 0.5.0

[DeployDuck]

🤖 Automated Release PR

This PR was created by release-please to prepare the next release. Once merged:

1. A new version tag will be created
2. A GitHub release will be published
3. The changelog will be updated

Changes to be included in the next release:

0.5.0 (2025-05-30)

⚠ BREAKING CHANGES

• generalise activation function (#163)
• Rework Loss Scalings to provide better modularity (#52)

Features

• generalise activation function (#163) (98c4d06)
• graphs: add scale_resolutions arg (#188) (0ea0642)
• transformer mapper (#179) (2cea7db)

Bug Fixes

• Adapt ensemble configs to work with loss refactor. (#300) (c210478)
• Adapt predict_step in model interface to pass on arguments for model classes (#281) (a5b2643)
• Change output format in validation step to fix issue in plotting (#305) (cc941f3)
• clean params complement (#291) (4e9ca31)
• models,training: Remove unnecessary torch-geometric maximum version (#326) (fe93ea8)
• Move scaler to device in scale (#317) (1592d09)
• remove activation entry from MLP noise block (#340) (2d060f5)
• Rework Loss Scalings to provide better modularity (#52) (162b906)
• training: bump anemoi-graphs version to 0.5.2 (#276) (9b8ec13)
• training: Explicit Batch invariance (#318) (45f6e15)
• training: no logging variable metadata to mlflow (#304) (4f7c2c9)
• update defaults for mlflow logging (#333) (5560581)

Documentation

• downloading era5 o96 dataset (#307) (b9e2d24)
• Fix minor mistakes in CRPS user guide. (#264) (52d9a2e)

---

Important

Please do not change the PR title, manifest file, or any other automatically generated content in this PR unless you understand the implications. Changes here can break the release process.

⚠️ Merging this PR will:

• Create a new release
• Trigger deployment pipelines
• Update package versions

Before merging:

• Ensure all tests pass
• Review the changelog carefully
• Get required approvals

Release-please documentation