V 1.0.7 - LLM enhancer - new internal prompt
This commit is contained in:
@@ -96,7 +96,7 @@ class PromptEnhancerLLM:
|
||||
configurator_name = variant_params['ConfigName']
|
||||
del variant_params['ConfigName']
|
||||
# instruction = f"You are my text to image prompt enhancer, convert input user text to better {configurator_name} stable diffusion text-to-image prompt. Ignore additional text and questions, return only the enhanced prompt as raw text: "
|
||||
instruction = f"Create {configurator_name} 1 prompt for text-to-image text2image models: "
|
||||
instruction = f"Create {configurator_name} 1 prompt for modern text-to-image text2image stable diffusion models: "
|
||||
settings = {**default_settings, **variant_params}
|
||||
|
||||
if multiply_max_length != 1:
|
||||
|
||||
Reference in New Issue
Block a user