首先感謝大家的支持，讓一個幾百播放量的up破萬，其實這期畫的背景有些亂，鏈接也忘放到簡介了，本人也是個業(yè)余愛好者，并不是專業(yè)畫畫有什么問題請大家包涵。

鏈接：https://pan.quark.cn/s/16b4f516b5ec

提取碼：SBTM

anatomically correct, high details, award winning, 8k
1girl, solo, breasts, flower, long hair, thighhighs, night, hair ornament, dress, hair flower, black dress, white thighhighs, sitting, sky, cleavage, outdoors, night sky, large breasts, looking at viewer, lantern, water, (shooting star:0.9), sky, bare shoulders, east asian architecture, sleeveless dress, arm up, starry sky, white flower, very long hair, bangs, mountain, architecture, tree, lake, arm support, closed mouth，
<lora:dudou_623:0.7> <lyco:Dreamwave v2L:0.6>
Negative prompt: bad-artist bad-hands-5 bad_prompt bad-picture-chill-75v badhandv4 bad_prompt_version2 easynegative ng_deepnegative_v1_75t EasyNegativeV2
Steps: 40, Sampler: DDIM, CFG scale: 7, Seed: 3006389725, Size: 1344x704, Model hash: 70ba76a999, Model: starlake-4.0, Clip skip: 2, ControlNet 0: "preprocessor: tile_resample, model: control_v11f1e_sd15_tile [a371b31b], weight: 0.4, starting/ending: (0, 1), resize mode: Crop and Resize, pixel perfect: False, control mode: ControlNet is more important, preprocessor params: (512, 1, 64)", Lora hashes: "dudou_623: 005ab420cf07", AddNet Enabled: True, AddNet Module 1: LoRA, AddNet Model 1: Chinese style_20230607154437-000017(1eb2082f3da6), AddNet Weight A 1: 0.7, AddNet Weight B 1: 0.7, Version: v1.3.2

成片的圖都是基本用這個詞改動圖生圖來的。大模型用的starlake4.0和majicMIX lux_v2.civitai