Oriskany/coco2017-mds-512px

COCO 2017 MosaicML MDS Dataset (512px Shortest-Edge) This dataset contains pre-sharded MosaicML Streaming (.mds) shards for monocular depth estimation pre-training: Image Resolution: Scaled with aspect-ratio preservation to min(H, W) = 512px (JPEG 95). Depth Labels: 16-bit unsigned PNG (uint16, $[0, 65535]$) generated via Depth-Anything-V2-Small-hf teacher inference. Format: Zero-copy chunked binary shards (.mds) with index.json manifests. Total Processed Samples: 287,360… See the full description on the dataset page: https://huggingface.co/datasets/Oriskany/coco2017-mds-512px.

<iframe src="https://silytics.si/embed/item/96cdca74-2040-423d-9abf-4971a207a444" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 17 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.