Hello everyone.
This is an extension of AI SM J V1. I tripled the dataset (used about 1000 images). Since I lack the necessary equipment and knowledge, all example images were created using pure Prompt and LORA (except of course AI SM J. V2), i.e. without post-processing, upscaling, etc. I hope you can use them to create beautiful images.
It would have to be suitable for many types of images (people, design, colors) except anime. Unfortunately I couldn't test everything.
Example without AI SM J V2
Example with AI SM J V2
Prompt: A disturbing cinematic scene of Darth Vader sitting on a toilet and holding a paper toilet roll. The toilet room is poorly lit and very dirty.
Guidance: 3.5, Steps 35 Flux(pro)
Seed 2411980994
描述:
训练词语:
名称: AI SM J V2_epoch_5.safetensors
大小 (KB): 18843
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success
