tao-train-oneformer

Pass

Audited by Gen Agent Trust Hub on Jun 30, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Fetches a pretrained model backbone from the official Swin Transformer repository on GitHub for model initialization.
  • [EXTERNAL_DOWNLOADS]: References official NVIDIA TAO Toolkit container images hosted on the nvcr.io registry.
  • [COMMAND_EXECUTION]: Provides standard CLI instructions for training, evaluation, and inference workflows using the oneformer tool within a Docker environment.
  • [SAFE]: The skill documentation includes security-aware guidance regarding the loading of model checkpoints, specifically advising against disabling weight-only loading for untrusted artifacts.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 30, 2026, 05:48 PM
Security Audit — agent-trust-hub — tao-train-oneformer