fiftyone-dataset-inference

What problem does it solve?

This Skill enables running ML model inference on FiftyOne datasets to generate predictions, embeddings, and related outputs for downstream analysis.

Core Features & Use Cases

• Inference for detection, classification, segmentation, and embeddings: Apply zoo models to datasets and produce a labeled field with predictions.
• End-to-end workflows: Import data with fiftyone-dataset-import and then run inference in a single flow.
• View results and compare fields: Inspect predictions alongside ground-truth or other labels.

Quick Start

Use the fiftyone-dataset-inference skill to run an inference pass on your dataset. For example:

• Ensure the dataset exists with list_datasets()
• set_context(dataset_name="my-dataset")
• launch_app(dataset_name="my-dataset")
• execute_operator( operator_uri="@voxel51/zoo/apply_zoo_model", params={"tab": "BUILTIN", "model": "yolov8n-coco-torch", "label_field": "predictions"} )
• set_view(exists=["predictions"])