{"success":true,"data":{"author":"MosaicML / Databricks (Aaron Gokaslan et al.)","bucketAccessType":"public","bucketName":"common-canvas/commoncatalog-cc-by","bucketPrefix":"","category":"Computer Vision","cloudProvider":"huggingface","createdAt":"2026-11-14T14:15:13.308Z","description":"CommonCatalog is the image-caption dataset assembled by MosaicML to train CommonCanvas, a family of latent diffusion models built to match Stable Diffusion 2 using only openly licensed images of known provenance (Gokaslan et al., arXiv:2310.16825). The full catalogue covers roughly 100 million high-resolution photographs from Yahoo Flickr Creative Commons 100M (YFCC100M), collected from Flickr users around 2014, each given a synthetic caption generated with BLIP-2. The collection is split by licence into separate repositories; this one holds the images whose Flickr licence is CC BY. Each Parquet row carries the JPEG bytes, the BLIP-2 caption, the original Flickr title, user tags and machine tags, capture device, date taken, latitude and longitude where present, image size, the Flickr page and download URLs, and the licensename and licenseurl of the individual photo, so attribution can be generated per image. Images go up to 4K resolution. Measured 2026-11-14: ten top-level subsets (0 to 9), each partitioned by least_dim_range (256-512 up to 4096+) and aspect_ratio_bucket folders of Parquet files; the Hub reports about 21.5 TB of storage. Read with pyarrow, pandas or the datasets library in streaming mode. Uses: text-to-image and image-to-text training, captioning, data attribution research. Each image is CC BY under its Flickr licence; credit the photographer.","displayTitle":"CommonCatalog CC-BY: Flickr Creative Commons images with synthetic captions","id":"b5472514-e3a9-50e4-ace2-b18ebc804a13","license":"CC-BY-4.0","licenseEvidenceUrl":"https://huggingface.co/datasets/common-canvas/commoncatalog-cc-by","objectCountEstimate":5575,"probeStatus":"ok","referenceUrl":"https://arxiv.org/abs/2310.16825","region":"us","schemaCard":{"accessHints":[{"language":"python-datasets","snippet":"from datasets import load_dataset\nds = load_dataset(\"common-canvas/commoncatalog-cc-by\")"},{"language":"python-hub","snippet":"from huggingface_hub import snapshot_download\nsnapshot_download(repo_id=\"common-canvas/commoncatalog-cc-by\", repo_type=\"dataset\")"}],"caveats":["HuggingFace metadata: author=common-canvas, license:cc-by-4.0, size_categories:10M<n<100M, format:parquet, downloads=84521"],"format":"mixed","generatedAt":"2026-11-14T14:15:21.702Z","layout":{"example":"0/least_dim_range=1024-2048/aspect_ratio_bucket=1-1/part-00000-tid-6217004709699050820-27b7ba6e-e051-4949-a7d0-9573b234b3fd-52879-1-c000.parquet","pattern":"{repoId}/{filename}"},"mountPath":"/mnt/archive","sampleHeaders":[{"bytes":"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","encoding":"hex","key":"0/least_dim_range=1024-2048/aspect_ratio_bucket=1-1/part-00000-tid-6217004709699050820-27b7ba6e-e051-4949-a7d0-9573b234b3fd-52879-1-c000.parquet","truncatedAt":512}],"sampleObjects":["0/least_dim_range=1024-2048/aspect_ratio_bucket=1-1/part-00000-tid-6217004709699050820-27b7ba6e-e051-4949-a7d0-9573b234b3fd-52879-1-c000.parquet","0/least_dim_range=1024-2048/aspect_ratio_bucket=1-2/part-00000-tid-4428982189861888730-94e6e6c5-d072-47bd-930c-5ffccb2d147d-52903-1-c000.parquet","0/least_dim_range=1024-2048/aspect_ratio_bucket=11-21/part-00000-tid-5246289597322746979-0426e71f-6c06-4e0b-82eb-307f0d6df61b-53520-1-c000.parquet","0/least_dim_range=1024-2048/aspect_ratio_bucket=13-18/part-00000-tid-6788610333251177415-200b26fd-7ec7-4ac9-857a-cdfc7823dce5-56235-1-c000.parquet","0/least_dim_range=1024-2048/aspect_ratio_bucket=13-18/part-00001-tid-6788610333251177415-200b26fd-7ec7-4ac9-857a-cdfc7823dce5-56242-1-c000.parquet"],"sidecars":[{"bytes":"---\nlicense: cc-by-4.0\ndataset_info:\n  features:\n    - name: jpg\n      dtype: image\n    - name: blip2_caption\n      dtype: string\n    - name: caption\n      dtype: string\n    - name: licensename\n      dtype: string\n    - name: licenseurl\n      dtype: string\n    - name: width\n      dtype: int32\n    - name: height\n      dtype: int32\n    - name: original_width\n      dtype: int32\n    - name: original_height\n      dtype: int32\n    - name: photoid\n      dtype: int64\n    - name: uid\n      dtype: string\n    - name: unickname\n      dtype: string\n    - name: datetaken\n      dtype: timestamp[us]\n    - name: dateuploaded\n      dtype: int64\n    - name: capturedevice\n      dtype: string\n    - name: title\n      dtype: string\n    - name: usertags\n      dtype: string\n    - name: machinetags\n      dtype: string\n    - name: longitude\n      dtype: float64\n    - name: latitude\n      dtype: float64\n    - name: accuracy\n      dtype: int64\n    - name: pageurl\n      dtype: string\n    - name: downloadurl\n      dtype: string\n    - name: serverid\n      dtype: int64\n    - name: farmid\n      dtype: int64\n    - name: secret\n      dtype: string\n    - name: secretoriginal\n      dtype: string\n    - name: ext\n      dtype: string\n    - name: url\n      dtype: string\n    - name: key\n      dtype: string\n    - name: status\n      dtype: string\n    - name: error_message\n      dtype: string\n    - name: exif\n      dtype: string\n    - name: sha256\n      dtype: string\n    - name: description\n      dtype: string\ntask_categories:\n  - text-to-image\nlanguage:\n  - en\n---\n# Dataset Card for CommonCatalog CC-BY\n\nThis dataset is a large collection of high-resolution Creative Common images (composed of different licenses, see paper Table 1 in the Appendix) collected in 2014 from users of Yahoo Flickr. \nThe dataset contains images of up to 4k resolution, making this one of the highest resolution captioned image datasets.\n \n## Dataset Details\n\n### Dataset Description\n\nWe provide captions synthetic captions to approximately 100 million high resolution images collected from Yahoo Flickr Creative Commons (YFCC).\n\n- **Curated by:** Aaron Gokaslan \n- **Language(s) (NLP):** en\n- **License:** See relevant yaml tag / dataset name.\n\n### Dataset Sources\n\n<!-- Provide the basic links for the dataset. -->\n\n- **Repository:** https://github.com/mosaicml/diffusion  \n- **Paper:** https://arxiv.org/abs/2310.16825\n- **Demo:** See CommonCanvas Gradios\n\n## Uses\n\nWe use CommonCatalog to train a family latent diffusion models called CommonCanvas.\nThe goal is to produce a model that is competitive with Stable Diffusion 2, but to do so using an easily accessible dataset of known provenance. \nDoing so makes replicating the model significantly easier, and provides a clearer mechanism for applying training-data attribution techniques. \n\n\n### Direct Use\n\nTraining text-to-image models\nTraining image-to-text models\n\n\n### Out-of-Scope Use\n\n<!-- This section addresses misuse, malicious use, and uses that the dataset will not work well for. -->\n\n* Crafting content that is offensive or injurious towards individuals, including negative portrayals of their living conditions, cultural backgrounds, religious beliefs, etc.\n* Deliberately creating or spreading content that is discriminatory or reinforces harmful stereotypes.\n* Falsely representing individuals without their permission.\n* Generating sexual content that may be seen by individuals without their consent.\n* Producing or disseminating false or misleading information.\n* Creating content that depicts extreme violence or bloodshed.\n* Distributing content that modifies copyrighted or licensed material in a way that breaches its usage terms.\n\n## Dataset Structure\n\nThe dataset is divided into 10 subsets each containing parquets about 4GB each. Each subfolder within contains a resolution range of the images and their respective aspect ratios.\nThe dataset is also divided along images licensed for commercial use (C) and those that are not (NC).\n\n## Dataset Creation\n\n### Curation Rationale\n\nCreating a standardized, accessible dataset with synthetic caption and releas","encoding":"utf8","key":"README.md","name":"README.md","size":5522,"truncated":true}],"sourceObjectKey":"0/least_dim_range=1024-2048/aspect_ratio_bucket=1-1/part-00000-tid-6217004709699050820-27b7ba6e-e051-4949-a7d0-9573b234b3fd-52879-1-c000.parquet","starterCell":"from datasets import load_dataset\nds = load_dataset(\"common-canvas/commoncatalog-cc-by\")\nprint(ds)","starterSamplePath":"0/least_dim_range=1024-2048/aspect_ratio_bucket=1-1/part-00000-tid-6217004709699050820-27b7ba6e-e051-4949-a7d0-9573b234b3fd-52879-1-c000.parquet","totalObjects":5575},"summary":"CommonCatalog is the image-caption dataset assembled by MosaicML to train CommonCanvas, a family of latent diffusion models built to match Stable Diffusion 2 using only openly…","title":"CommonCatalog CC-BY: Flickr Creative Commons images with synthetic captions","updatedAt":"2026-11-14T17:32:13.586Z","visibility":"public","websiteUrl":"https://github.com/mosaicml/diffusion"}}