Instructions to use v13s/tancho-generator-v5 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Cosmos
How to use v13s/tancho-generator-v5 with Cosmos:
# No code snippets available yet for this library. # To use this model, check the repository files and the library's documentation. # Want to help? PRs adding snippets are welcome at: # https://github.com/huggingface/huggingface.js
- Notebooks
- Google Colab
- Kaggle
Download SOURCES.md from v13s/tancho-generator-v5: direct link, hf CLI and curl.
- Browser
- Download file 1.48 kB
-
https://huggingface.co/v13s/tancho-generator-v5/resolve/main/SOURCES.md
- Command line
-
hf download hf://v13s/tancho-generator-v5/SOURCES.md
-
curl -L -o SOURCES.md https://huggingface.co/v13s/tancho-generator-v5/resolve/main/SOURCES.md
Sources for this release
Both Generator training packs use reference-image tiles from the Geospatial Information Authority of Japan (GSI), Nagano City Omachi, October 13, 2019, Typhoon Hagibis. Source footage SHA-256 recorded locally: a112641a9335adc310f872963ce49a4e73d7a9451fe07fc06705286e32379ec2. This is a local provenance checksum, not one issued by GSI.
Official source: https://www.gsi.go.jp/BOUSAI/R1.taihuu19gou.html GSI content terms: https://www.gsi.go.jp/kikakuchousei/kikakuchousei40182.html
Attribution: 出典:国土地理院ウェブサイト。動画の切り出し・縮小と合成観測条件・数値教師・評価はTancho開発側で作成。国土地理院の公式評価ではありません。
Training and evaluation tiles are separated within this one flight. Images are reference inputs; synthetic target coordinates have no provided image-to-3D correspondence. mixed240 uses 480 numeric examples; balanced288 uses 576. The released saved answers include newly generated numerical geometry and explicitly marked reused regression cases; they reuse reference evaluation images. No footage or image tiles are included in this repository.
Earlier documents described Omachi as evaluation-only for a retired observation task. That statement does not apply to these Generator training packs. No client footage was used in these checkpoints. Aso, Sannoudani, and other events mentioned in older experiments are not these models' training imagery.