datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
starlink-fleet-data
Starlink Constellation Fleet Data
Credit: NASA
Part of a dataset collection on Hugging Face.
Dataset description
Daily health snapshots of the SpaceX Starlink constellation, derived from CelesTrak GP (General Perturbations) data. Tracks satellite count, orbital shells, operational status, and ISL capability across five inclination-based shells.
Starlink is the largest satellite constellation ever built, representing a fundamental shift in how broadband… See the full description on the dataset page: https://huggingface.co/datasets/juliensimon/starlink-fleet-data.ast-spacemobile-fleet-data
AST SpaceMobile BlueBird Fleet Data
Credit: NASA
Part of a dataset collection on Hugging Face.
Dataset description
Daily health snapshots of the AST SpaceMobile BlueBird direct-to-cell satellite constellation, derived from CelesTrak GP (General Perturbations) data. Tracks the BlueWalker 3 prototype and all production BlueBird satellites as the constellation is built out.
AST SpaceMobile is developing a unique LEO constellation designed to provide… See the full description on the dataset page: https://huggingface.co/datasets/juliensimon/ast-spacemobile-fleet-data.oneweb-fleet-data
OneWeb Constellation Fleet Data
Credit: NASA
Part of a dataset collection on Hugging Face.
Dataset description
Daily health snapshots of the OneWeb (Eutelsat OneWeb) satellite constellation, derived from CelesTrak GP (General Perturbations) data. Tracks satellite count and lifecycle status across OneWeb's near-polar orbital planes.
OneWeb operates a ~650-satellite first-generation LEO broadband constellation at roughly 1,200 km altitude and 87.9 degrees… See the full description on the dataset page: https://huggingface.co/datasets/juliensimon/oneweb-fleet-data.globalstar-fleet-data
Globalstar Constellation Fleet Data
Credit: NASA
Part of a dataset collection on Hugging Face.
Dataset description
Daily health snapshots of the Globalstar satellite constellation, derived from CelesTrak GP (General Perturbations) data. Tracks satellite count, hardware generation, and lifecycle status for the full ~85-satellite fleet.
Globalstar operates a LEO satellite-messaging and voice constellation at roughly 1,414 km altitude and 52 degrees… See the full description on the dataset page: https://huggingface.co/datasets/juliensimon/globalstar-fleet-data.kuiper-fleet-data
Kuiper Constellation Fleet Data
Credit: NASA
Part of a dataset collection on Hugging Face.
Dataset description
Daily health snapshots of Amazon's Project Kuiper broadband constellation, derived from CelesTrak GP (General Perturbations) data. Tracks satellite count, orbital shell, and lifecycle status across the three FCC-authorized Kuiper shells.
Project Kuiper is Amazon's answer to SpaceX Starlink: a low-Earth-orbit broadband constellation of 3,236… See the full description on the dataset page: https://huggingface.co/datasets/juliensimon/kuiper-fleet-data.trajectories-thinking-toolstrajectories-thinkingtrajectories-tool_use-difficult-envs-splitstrajectories_thinking-tool_use-difficult-envs-rawtrajectories-tool_use-difficult-envs-rawtrajectories-thinking-tools-tool_use-difficult-envs-rawgsm8k-fleet-evaltrajectories-thinking-tools-tool_use-difficult-envs-splitstrajectories_thinking-tool_use-difficult-envs-splitsfleet-test-falcon-ocr
Document Processing using Falcon OCR (plain mode)
This dataset contains OCR results from images in davanstrien/ufo-ColPali using Falcon OCR, a 0.3B early-fusion vision-language model.
Processing Details
Source Dataset: davanstrien/ufo-ColPali
Model: tiiuae/Falcon-OCR
Task Mode: plain - Full-page text extraction
Number of Samples: 2
Processing Time: 1.2 min
Processing Date: 2026-06-05 09:47 UTC
Backend: falcon-perception (OCRInferenceEngine)… See the full description on the dataset page: https://huggingface.co/datasets/davanstrien/fleet-test-falcon-ocr.fleet-test-paddleocr-vl-1.6
Document Processing using PaddleOCR-VL-1.6 (OCR mode)
This dataset contains OCR results from images in davanstrien/ufo-ColPali using PaddleOCR-VL-1.6, an ultra-compact 0.9B OCR model (96.33% SOTA on OmniDocBench v1.6).
Processing Details
Source Dataset: davanstrien/ufo-ColPali
Model: PaddlePaddle/PaddleOCR-VL-1.6
Task Mode: ocr - General text extraction to markdown format
Number of Samples: 2
Processing Time: 1.9 min
Processing Date: 2026-06-05 09:51 UTC… See the full description on the dataset page: https://huggingface.co/datasets/davanstrien/fleet-test-paddleocr-vl-1.6.
