None-South1.5-PLUS

      With the rapid development of AI visual technology, we have established a research and development team with AI as the core driving force, aiming to break through the diversification of art and creativity, and liberate productivity bottlenecks as our vision ----Niuneng Technology

      Key members include former senior software engineer at Huawei Nanjing Research InstituteLIONEL, former senior algorithm engineer at Nanjing Algorithm Research InstituteShaolin Xiake, stable model developer of Moryou humanoid series, artist at Central Academy of Fine ArtsMoryou, author of None series models, LIBLIBAI training instructorNONE and many other top industry professionals.

      In the future, we will focus on the research and development of AI vision, work on practical scenario applications, uphold the original intention of enabling everyone to enjoy a new ecosystem of AI technology, and forge ahead!

Primary upgrade contents:

Imaging quality optimization, getting a little taste of XL

Limbs optimization, solving some issues with V1 hand breaks

Light and shadow optimization, enhanced light-dark contrast, enhanced light halo logic, probability of MJ foreground halo

Composition optimization, strengthening the concept of photographic composition, making the output more photographic

Optimizing adaptability for DPM++2M SDE karrs sampler

Subtract negative words and embed negative embeddings, use large common negative words: (worst quality:2), (low quality:2), (normal quality:2), lowres, watermark,

Starting tag: none  , stable output with a single tag, so no need to add quality words, just describe the content

Bayounetta series photography tags can continue to be used

Usage guide: built-in special VAE, no need for external plugins

               Recommended samplerDPM++2M karrs/ DPM++2M SDE karrs   Step 20-30   CFG: 5-9

                                restart             Step 15-25   CFG: 5-9

               Recommended use of high-definition restoration:

                   Recommended enlargement algorithm:4x-UltraSharp 2x redraw range 0.2

                                        Or 8x-NMKD 2x redraw range 0.2

                                            AfterDetailer for face fix plugin, do not use for facial repair!!!

Negative embeddings and plugins, enlargement algorithm files can be obtained from relevant groups

None's AI Community QQ Group: 820145842

Not for commercial use, do not modify models without permission, merge, etc.