This LoRA is trained on a hyper-detailed fantasy illustration style blending anime aesthetics with Japanese woodcut influences, intricate realism, and vibrant compositions.

Hyperdetailed Illustration

This is one of the first Qwen LoRAs I am releasing here on civit and is trained in AI Toolkit (default setting in Q6). The LoRAs has some Artifacts present due to the Quantized training , so I would recommend either: 1. Start with lower strenght ~0.82. Use it with lighting 4/8 steps LoRA

HyperdetailedillustrationQwen.safetensors

Qwen

Klein4B

Klein9B

ZImageBase

ZImageTurbo

Illustrious

Flux

SDXL

Pony

Files with SHA256: e83abb49e5c656f6dcde73e939147389ee9a36b035622f43b184d4ea3c8c433b