
I find it hard to achieve photorealism in Pony Diffusion so I made my own LoRa. It may not come out that realistic if additional LoRas are used. Mix and match if you want a semirealistic style. I also found out that certain character prompts don't come out as realistic without additional work. Using quality tags score_9,score_8_up,score_7_up,score_6_up,score_5_up,score_4_up may decrease realism.
Recommended Prompts (Optional):
For human characters - human, light skin, pale skin, dark skin, tan skin, tan, curly hair, straight hair, wavy hair
For non-human characters - Specify the species, creature, animal, insect, no humans
Negative prompts - Traditional media, sketch, source_anime, source_cartoon, cartoon, anime, painting,
描述:
训练词语: Realistic
名称: 400450_training_data.zip
大小 (KB): 10774
类型: Training Data
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success
名称: Realistic_Photography_for_Pony_DIffusion.safetensors
大小 (KB): 223112
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success