【Realistic Photography】Dreams and bubbles

Model Thoughts: Throughout the entire process from the initial preparation of the model to subsequent fine-tuning effects, special thanks to Pailao for his technical support and meticulous care! Without him, this series wouldn't exist; he is truly the creator who gave life to this model! He is the one who made this 1.5 model take a qualitative leap! With the fine-tuning version coupled with the texture and style of the original version, the success rate of Asian faces and limbs has significantly increased, resulting in a very nice overall output! Hope everyone can pay attention to Pailao! Check out Pailao here:https://www.liblib.art/userpage/a16754deb59229950eee60efefd74793/publish

During the preparation and training of the model, the selection of training set materials, model training parameters, and the use of cloud training platforms have always been a headache for me! Many thanks to None for his selfless and careful guidance; it was also him who enlightened me with the idea of training large models! Hope everyone can pay attention to None! Check out None here:https://www.liblib.art/userpage/275954c1f9994660cdf62ccad717092c/publish

Also, during the model training process, I am very grateful for the technical and spiritual support from the big brothers in the group!

Not in any particular order:

Maiguo, None, Miaoka, Linhe, Liaowushi, Baize, Chaixi, Yinan, Laoxiatiao, YUNY.L, Jia Shu Martin, Big Brother's Mouse, Chonglou, Nanqiang, AiARTist Guangyuan, Cayden, Xuer, Moyo, Gray Wolf, Adam, Cyberdream, Nawusjia, Dao Dai, MEXX, Yuaner Wushang, Dianfen Uncle, Zhan Yexiao,

........... and many other great masters in the group that were not mentioned. (There are too many to list)

Model Introduction:【Realistic Photography】Dreams and bubbles is a large model primarily focused on photography themes

After comparison and testing, the progress of this version includes:

1. The character poses and composition are more rich, dynamic, and visually appealing;

2. Through extensive running and testing, this model shows ideal understanding in both character composition and natural language aspects.

Adding in the text (8k, RAW photo, best quality, masterpiece:1.2,), there will be a significant improvement in the output effect.

3.Sampling usedEuler a, DPM++ 2M Karras, DPM++ 2M SDE Karras, DPM2 a Karras, DPM++ 2S a Karras, DDIM series

 Steps (Sampling steps):30-60,CFG: 2.6-5, Redraw range 0.3-0.45

5. Remember to enable face repair, please use afterdetiler.

6. For comprehensive adjustment of filters/tone, please use sd-webui-bmab.

7. Also, enabling high-resolution effects will yield better results.

Pairing large models with dedicated camera lens TAG recommendations: Nighttime and low-light photography (Nikon AF-S NIKKOR 35mm f/1.4G) This 35mm lens with a large aperture of f/1.4 excels in low-light conditions and is ideal for nighttime and low-light photography. Portrait photography (Nikon AF-S 105mm f/1.4E ED) The 105mm lens is suitable for portrait photography, offering excellent clarity and color reproduction. Portrait photography (Canon RF 85mm f/1.2L  85mm) The 85mm lens is a classic choice for portrait photography, creating beautiful background blur effects. Large model paired with front TAG recommendations: cowboy shot Seven-tenths body lens  close-up    close-up lens    depth of field  Background blur

Legal Disclaimer:

Downloading and using (online or locally) this model represents your agreement and acceptance of the following:

1. Uphold the core socialist values, generate content that does not incite subversion of state power, overthrow the socialist system, harm national security and interests, damage the national image, incite separatism, disrupt national unity and social stability, advocate terrorism, extremism, national hatred, racial discrimination, violence, obscenity, and dissemination of false harmful information prohibited by laws and regulations;

2. Respect the legitimate rights and interests of others, do not harm the physical and mental health of others, do not infringe on the portrait rights, reputation rights, honor rights, privacy rights, and personal information rights of others;

3. Without the consent of others, do not generate images of real people or images generated based on the real portraits of others;

4. Do not generate any content that contains intellectual property or may infringe on intellectual property rights;

5. You guarantee that when using this model, you will not infringe on any third party's intellectual property rights, reputation rights, name rights, privacy rights, and other legitimate rights and interests, and will not use this model for any illegal purposes. Otherwise, you shall handle it on your own and bear all legal responsibilities that may arise from it;

6. If your infringement causes losses to the model author (including but not limited to economic, reputation losses), you shall fully compensate the model author for all losses.