Train RT-DETR models for real-time object detection.
What it does
Trains and optimizes RT-DETR object detectors and produces evaluations, predictions, and exported models.
When to use it
Use for training, evaluating, distilling, quantizing, exporting, or running inference with TAO RT-DETR models.
How to use it
Give it datasets and any checkpoints needed for the requested operation; it produces model artifacts and results in an output folder.
What you provide
- Your files
Uses
Good to know
- a credential is hard-coded: encryption_key
What you need · 5
Requires Docker together with nvidia-container-toolkit.
Requires nvidia-container-toolkit together with Docker.
Requires at least one GPU with 16GB or more VRAM per GPU; recommends two GPUs and names V100 or A100.
Must run first when the session was not initialized by the TAO skill bank plugin, for host preflight, credentials, and cross-skill discovery.
Access · 0
No additional actions listed.