林忆宁，塞纳河小偶像 (010, idol girl from SNH48)版本v2.0 (ID: 25363)

New Release: 林忆宁，古户绘梨花 (010, Furuki Edoka)

X/Y Tests:

Prompt test:

CFG test (w/010_v32-000016 model):

Generalization test (w/010_v32-000016 model):

Dataset:

Images chosen from 010's selfie set, specifically during her days as an acitve member of SNH48's Team X.

Training:

This LoRA model is trained with Kohya_ss GUI. Please check version descriptions for training parameters

Tips:

1. I suggest using this model at 0.6~0.9 weight for best results. Going above 0.9 usually causes distortion in face.

2. Low CFG scale around 2.5 to 5.0 is recommended. Turn it down if you are seeing oversaturated images.

3. Using this model to generate NSFW content is possible, but highly dissuaded.

Reach Out:

七海电报群：https://t.me/nanamivr

宁宁肉体厨：https://t.me/nanami_ai

Any questions, suggestions and ideas on improving the model are appreciated.

描述:

Retrained the model with 70 pics resized to 512×640 as oppose to 20pics resized to 512×512 in v1.0. 30 steps each image which makes 2,100 steps in total.

• v2.0 can do a better job of handling pictures with upper body, while v1.0 can only generate portraits properly.

• v2.0 is less 010-like, in the sense that the loss function is greater at the end of the training as to v1.0.

训练词语: