代码拉取完成,页面将自动刷新
I can't say what my data sources are because too many people downloading from them could cause them to block scraping. Danbooru2018 is a good starting point.
My training code is is modified from Forty-lock's project PEPSI, which is the official implementation of the paper PEPSI : Fast Image Inpainting With Parallel Decoding Network. PEPSI is licensed under the MIT license.
Black and white images contain screentone patterns which are difficult for neural networks to replicate.
It's on my to-do list, but it's not a high priority.
DeepCreamPy is not trained on real life porn, so it will not work with real life porn.
此处可能存在不合适展示的内容,页面不予展示。您可通过相关编辑功能自查并修改。
如您确认内容无涉及 不当用语 / 纯广告导流 / 暴力 / 低俗色情 / 侵权 / 盗版 / 虚假 / 无价值内容或违法国家有关法律法规的内容,可点击提交进行申诉,我们将尽快为您处理。