Added 2D TMDs NEB benchmark and docs updates

[Jad-J3far]

Pre-review checklist for PR author

• I've confirmed the contribution guidelines.

Summary

Adds a new NEB benchmark for oxygen adatom diffusion on 2D TMDs (MoS2, MoSe2, MoTe2, WS2, WSe2, WTe2).
Includes calculation workflows, analysis metrics/tables, and the Dash app entry, plus user guide documentation and model registry updates.

Linked issue

Resolves #303

Progress

• Calculations
• Analysis
• Application
• Documentation

Testing

Tested on CUDA with: mace-mp-0b3, orb-v3-consv-inf-omat, uma-s-1p1-omat

New decorators/callbacks

None.

[joehart2001]

in the calc, why is plt.close("all") used? is this related to janus?

[Jad-J3far]

in the calc, why is plt.close("all") used? is this related to janus?

I think its a Matplotlib default behaviour which Janus inherits when plot_band=True.

The problem was that figures for subsequents runs while looping through models and materials were accumulating previous profiles. As in, plots from earlier NEB runs were getting drawn onto subsequent figures. I thought this was related to a Matplotlib maintaining a global figure state or canvas by default.
I used plt.close("all") after each .run() to clears all open figures from Matplotlib's global state and ensure each NEB run starts with a clean canvas

[joehart2001]

in the calc, why is plt.close("all") used? is this related to janus?

I think its a Matplotlib default behaviour which Janus inherits when plot_band=True.

The problem was that figures for subsequents runs while looping through models and materials were accumulating previous profiles. As in, plots from earlier NEB runs were getting drawn onto subsequent figures. I thought this was related to a Matplotlib maintaining a global figure state or canvas by default. I used plt.close("all") after each .run() to clears all open figures from Matplotlib's global state and ensure each NEB run starts with a clean canvas

@ElliottKasoar