Multi-Modal Foundation · mirror-physics

equiformer v3

Learns over graph-structured data.

Verify the licence before commercial useThe mit licence carries terms worth reading before you depend on it.mit

Availability: On demand
Typical latency: 12s
Cost per run: $0.02
Compute tier: S · Micro

Not runnable here yet

We index this model but cannot run it: no library and no architectures — nothing can load it.

Everything below is still accurate — the licence, the card and the provenance. Running it needs a loader for this repository, not a bigger machine.

About this model

Learns over graph-structured data. Runs on our servers, which pull mirror-physics/equiformer_v3 and execute it as a graph-ml pipeline. The input and output shapes were derived from the task the repository declares rather than written by hand — check the model card before relying on a result.

Standardized I/O contract

Every model in the Hub speaks the same contract, which is what lets the Router and the agent call any of them without special-casing.

Inputs

graphfile · required
Graph

Outputs

vectorvector
Mean-pooled representation, one vector per input.