<h1 id="stable-diffusion-vae-anime-c05tz70p2">stable-diffusion-vae-anime</h1>huggingface: <a rel="ugc" href="https://huggingface.co/ttj/stable-diffusion-vae-anime">ttj/stable-diffusion-vae-anime</a><ul><li>License: <a target="_blank" rel="ugc" href="https://huggingface.co/spaces/CompVis/stable-diffusion-license">The CreativeML OpenRAIL M license</a> is an <a target="_blank" rel="ugc" href="https://www.licenses.ai/blog/2022/8/18/naming-convention-of-responsible-ai-licenses">Open RAIL M license</a>, adapted from the work that <a target="_blank" rel="ugc" href="https://bigscience.huggingface.co/">BigScience</a> and <a target="_blank" rel="ugc" href="https://www.licenses.ai/">the RAIL Initiative</a> are jointly carrying in the area of responsible AI licensing. See also <a target="_blank" rel="ugc" href="https://bigscience.huggingface.co/blog/the-bigscience-rail-license">the article about the BLOOM Open RAIL license</a> on which our license is based.</li><li>Training : This model is fine-tuned from the vae use in this stable-diffusion checkpoint <a target="_blank" rel="ugc" href="https://huggingface.co/CompVis/stable-diffusion-v1-4">CompVis/stable-diffusion-v1-4</a><ul><li>Dataset: a subset of Danbooru2017, can be downloaded from <a target="_blank" rel="ugc" href="https://www.kaggle.com/">kaggle</a>.</li><li>Compute: The training using only one RTX 3090. Training was stopped at about 17 hours. And the latest checkpoint is exported.</li><li>Training code: The code used for training can be found in this github repo: <a target="_blank" rel="ugc" href="https://github.com/cccntu/fine-tune-models">cccntu/fine-tune-models</a></li></ul></li></ul>

Files with SHA256: 758acb192d423ec095e945b4953ba24a0fe10d191dd0efe4dc5328eb86bc1f23