Keltezaa/real-asian-women-zhen-shi-ya-zhou-nu-xing
Real Asian Women | 真实亚洲女性
<Gallery />
Model description
<p>v1.0</p><p>This model is a face model based on Flux dev 1.0</p><p>Recommended weights: 0.72-0.85 for standalone Use around 0.5 for mixed face lora. Weight reference look at the preview image.</p><p>Recommended size: 7681024,7681344,7201280</p><p>Recommended steps: 22-30 steps</p><p>Trigger word: asg</p><p></p><p>Customized lora can add WeChat: cjw0402 Note the intention.</p><p>Using completely artificial natural language manual marking training, cue word comprehension is very strong. It is not recommended that the cue word in the form of a word for out of the picture. And the chances of hand collapsing are very small, almost none.</p><p>Can be free to play, welcome to actively return to the map.</p><p></p><p><s> </s></p><ul><li><p><span style="color:rgb(255, 255, 255)">v1.0</span></p></li><li><p><span style="color:rgb(255, 255, 255)">本模型是一个基于Flux dev 1.0的脸型模型</span></p></li><li><p><span style="color:rgb(255, 255, 255)">推荐权重：单独使用0.72-0.85 混合脸型lora使用在0.5左右。权重参考看预览图。</span></p></li><li><p><span style="color:rgb(255, 255, 255)">推荐尺寸：7681024，7681344,7201280</span></p></li><li><p><span style="color:rgb(255, 255, 255)">推荐步数：22-30步</span></p></li><li><p><span style="color:rgb(255, 255, 255)">触发词：asg</span></p></li><li><p><em><u><span style="color:rgb(245, 219, 77)">定制lora可加V：cjw0402 备注来意。</span></u></em></p></li><li><p><span style="color:rgba(255, 255, 255, 0.8)">使用完全人工的自然语言手动打标训练，提示词理解能力非常强。不推荐单词形式的提示词进行出图。并且</span><strong><em><span style="color:rgb(225, 60, 57)">手部崩坏</span></em></strong><span style="color:rgba(255, 255, 255, 0.8)">的几率很小，几乎没有。</span></p></li><li><p><u><span style="color:rgba(255, 255, 255, 0.8)">可自由发挥，欢迎大家积极返图。</span></u></p></li></ul>
Trigger words
You should use asg to trigger the image generation.
Download model
Weights for this model are available in Safetensors format.
Download them in the Files & versions tab.
Use it with the 🧨 diffusers library
from diffusers import AutoPipelineForText2Image
import torch
device = "cuda" if torch.cuda.is_available() else "cpu"
pipeline = AutoPipelineForText2Image.from_pretrained('black-forest-labs/FLUX.1-dev', torch_dtype=torch.bfloat16).to(device)
pipeline.load_lora_weights('Keltezaa/real-asian-women-zhen-shi-ya-zhou-nu-xing', weight_name='asian-girl-000008.safetensors')
image = pipeline('`asg`').images[0]For more details, including weighting, merging and fusing LoRAs, check the documentation on loading LoRAs in diffusers