lin ref in ase datasets

[lbluque]

I think the linear reference should be computed in the models as part of #556 or in the trainer, but for now we can match the oc22_dataset, and deprecate this later.

[codecov]

Codecov Report

Attention: Patch coverage is 57.14286% with 3 lines in your changes are missing coverage. Please review.

Project coverage is 57.29%. Comparing base (be0f727) to head (c8ada66).

Files	Patch %	Lines
ocpmodels/datasets/ase_datasets.py	57.14%	3 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #643      +/-   ##
==========================================
- Coverage   57.29%   57.29%   -0.01%     
==========================================
  Files         109      109              
  Lines       10300    10307       +7     
==========================================
+ Hits         5901     5905       +4     
- Misses       4399     4402       +3     

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

[anuroopsriram]

LGTM for now, but I agree that the linear referencing code should ideally be shared across all dataset and model types.