Unofficial mirror to build hype on the release of the Common Canvas line of models by MosaicML.( I am not the creator, don't credit me unless you want to, lol)Common Canvas models are trained on CC and Public domain works using the same architectures as SD2 and SDXL for the Small and XL types, respectively, using the unique training approach of PAG, or Perturbed Attention Guidance, allowing for significantly smaller datasets that achieve similar quality to models that until now only large companies could afford to train. There are two main models, <a target="_blank" rel="ugc" href="https://huggingface.co/common-canvas/CommonCanvas-XL-C">Common Canvas XL Commercial</a>, and <a target="_blank" rel="ugc" href="https://huggingface.co/common-canvas/CommonCanvas-S-C">Common Canvas Small Commercial</a>, along with two even larger, but noncommercial, trained on CC-NC, <a target="_blank" rel="ugc" href="https://huggingface.co/common-canvas/CommonCanvas-XL-NC">XL NC</a> and <a target="_blank" rel="ugc" href="https://huggingface.co/common-canvas/CommonCanvas-S-NC">S NC</a> models.Additional files and resources are available at the <a target="_blank" rel="ugc" href="https://huggingface.co/common-canvas">Hugging Face official repo</a>, including access to <a target="_blank" rel="ugc" href="https://huggingface.co/collections/common-canvas/commoncatalog-6530907589ffafffe87c31c5">the entire catalogue of training data</a>.Link to Hugging Face demo page:<a target="_blank" rel="ugc" href="https://huggingface.co/spaces/common-canvas/CommonCanvas">https://huggingface.co/spaces/common-canvas/CommonCanvas</a>Link to the original paper:<a target="_blank" rel="ugc" href="https://arxiv.org/pdf/2310.16825">https://arxiv.org/pdf/2310.16825</a>Link to the Github repo used for training, which is also recommended for further finetuning:<a target="_blank" rel="ugc" href="https://github.com/mosaicml/diffusion">https://github.com/mosaicml/diffusion</a>Licenses are <a target="_blank" rel="ugc" href="https://huggingface.co/common-canvas/CommonCanvas-XL-C/blob/main/LICENSE.md">Open RAIL++-M License</a>, for CCXL, and <a target="_blank" rel="ugc" href="https://huggingface.co/datasets/choosealicense/licenses/blob/main/markdown/cc-by-sa-4.0.md">Creative Commons Attribution Share Alike 4.0</a>, for CCS.(I've uploaded only the Commercial models here.)(I've selected the Civitai options to try and match, otherwise would just leave open on everything.)

Common Canvas (Unofficial Mirror)

XL equivalent, and copyrighted material free, model that is commercial use approved. Needs finetuning on similarly obtained image resources, with a database containing many available in the description.

commonCanvasUnofficial_commoncanvasxlv01.safetensors

CommonCanvasXLv0.1

CommonCanvasSv0.1

<a target="_blank" rel="ugc" href="https://huggingface.co/common-canvas/CommonCanvas-XL-C">https://huggingface.co/common-canvas/CommonCanvas-XL-C</a>A reupload since the original one didn't select the architecture type.CIVIT AI MODERATION: DO NOT MARK AS DUPLICATE; <a target="_blank" rel="ugc" href="https://civarchive.com/models/459119?modelVersionId=510986">https://civarchive.com/models/459119?modelVersionId=510986</a> IS NOT PROPERLY MARKED AS A SDXL MODEL AND THEREFORE CANNOT BE USED ON CIVITAI.CommonCanvas is a family of latent diffusion models capable of generating images from a given text prompt. The architecture is based off of Stable Diffusion XL. Different CommonCanvas models are trained exclusively on subsets of the CommonCatalog Dataset (See Data Card), a large dataset of Creative Commons licensed images with synthetic captions produced using a pre-trained BLIP-2 captioning model.

Files with SHA256: 727c0895273b9cc5d338d96cdce9812a4b1708ba0c7aa0a6f2a1b26658e10967