FocsleFocsle
ME

V-JEPA-AC-Edge

22%
by Meta FAIR-Edge

Action-conditioned V-JEPA for world-model planning at the edge.

Vision-Language-ActionMITINT8FP16world-modelplanning
23K downloads 480 deploymentsUpdated Mar 11, 2028
Headline:18.4ms · NVIDIA Jetson Thor · MIXED

About this model

Action-conditioned V-JEPA for world-model planning at the edge.

Authored by meta-fair-edge. Curated into the Fo’c’sle reference set on 2028-03-11. All cross-chip benchmarks below were collected in matched-pair runs in the HIL lab using the same input pipeline, same upstream preprocessing, and the same downstream consumer. See the methodology page for the full protocol.

Task
Vision-Language-Action
Parameters
1.8 B
Benchmarked on
3 chips
Deployments
480

Architecture

Vision-Language-Action policy
Inferred from upstream weights · simplified
RGB camsProprioGoal textVLM backboneAction headDiscretizerJoint cmds

Headline benchmarks

Training data

Pretrained on the upstream maintainer’s released checkpoint. Edge-distillation pass uses 2.4M frames from the Fo’c’sle distillation corpus (consented public data + opt-in publisher contributions). Quantization-aware fine-tune uses 320K calibration samples drawn from the target task’s eval domain.

  • Pretraining corpus: upstream maintainer release
  • Distillation corpus: 2,400,000 frames
  • Calibration set: 320,000 samples (per task)
  • Eval set: standard benchmark + matched-pair HIL runs