This model produces a style inspired by an ancient indian epic, the Rāmāyaṇa (or Ramayana). That and the Mahābhārata (or Mahabharata) are foundational ancient texts that form part of the Itihasa.
Using a weight of 0.5 to 0.8 appears to work best; less than 0.5 might work for your purposes too. Here's how to bring in this model at 0.7 weight in your prompt: <lora:ramastyle:0.7>.
I'm adding images generated with no additional Loras or negative prompts embeddings you would have to seek out, made with just the prompt, negative prompt, and base model. However, this model does seem to respond well to enhancers like Wowifier and more_details.
Ramastyle doesn't produce great hands or feet by itself.
It's my first uploaded model, and I'm not convinced it is great, but it often creates appropriate images that I enjoy. Comments and feedback greatly appreciated!
描述:
Good results from Clip Skip 2 but have not tried other values.
训练词语:
名称: ramastyle.safetensors
大小 (KB): 147572
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success
