Add MatNova model

[zzz-sl]

Description

Hi Matbench-Discovery team,
We are submitting MatNova-mptrj model to the leaderboard.

Please let us know if any additional adjustments are required.

Best

Checklist

Please open PRs as draft and only mark as ready to review after checking off the following items:

• [√] I created a new folder and YAML metadata file models/<arch_name>/<model_variant>.yml for my submission. arch_name is the name of the architecture and model_variant.yml includes things like author details, training set names and important hyperparameters.
• [√] I added the new model as a new attribute Model.<model_variant> enum on the Model enum in enums.py.
• [√] I uploaded the energy/force/stress model prediction file for the WBM test set to Figshare or another cloud storage service (<yyyy-mm-dd>-<model_variant>-preds.csv.gz).
• [√] I uploaded the model-relaxed structures file to Figshare or another cloud storage service in JSON lines format (<yyyy-mm-dd>-wbm-IS2RE-FIRE.jsonl.gz). JSON Lines allows fast loading of small numbers of structures with pandas.read_json(lines=True, nrows=100) for inspection.
• [√] I uploaded the phonon predictions to Figshare or another cloud storage service (<yyyy-mm-dd>-kappa-103-FIRE-<values-of-dist|fmax|symprec>.gz).
• [√] I included the urls to the Figshare files in the YAML metadata file (models/<arch_name>/<model_variant>.yml). If not using Figshare I have included the urls to the cloud storage service in the description of the PR.
• [√] I included the test script (test_<arch_name>_<task>.py for task in discovery, kappa, diatomics) that generated the prediction files.

Marking your PR as ready to review will trigger an automated code review, please address any issues raised by the automated review. For the maintainers minimizing the final human review process enables us to merge your submissions much faster!

Additional Information (Optional)

• I included a training script (train_<arch_name>.py) if I trained a model specifically for this benchmark.
• I included a readme.md with additional details about my model.

Summary by CodeRabbit

• New Features

  • Added the matNova-mptrj model with comprehensive benchmarking for phonons, geometry optimization, and discovery.
  • Introduced thermal conductivity calculation workflows.
  • Added structure relaxation and discovery evaluation pipelines that save and merge prediction results for analysis.

• Documentation

  • Updated OC20 dataset resource link.

[CompRhys]

running just prepare-model-submission matnova_mptrj gives very different results for the kappa 103 metrics, see contributing.md

[zzz-sl]

Thanks for the reminder! There was an error in the kappa 103 JSON submission. Both the JSON file and the corresponding metrics results have now been updated.

[CompRhys]

can you please run just prepare-model-submission matnova_mptrj again and upload the figures and per element errors json that it creates as those are needed for the site.

[zzz-sl]

The required figures and per-element errors JSON have been uploaded.

[CompRhys]

populating these URLs feels like the last blocker to merge.

[CompRhys]

@zzz-sl Bump on populating this repo