Image cannot be generated normally
#1
by
Zctoylm
- opened
Thank you very much for uploading the converted model. When I used this sd-v1.5-openvino, the image quality was very poor and it reminded me that some parameters were missing, like:
The time_cond_proj_dim attribute is missing from the UNet configuration. Please re-export the model with newer version of optimum and diffusers.
The scaling_factor attribute is missing from the VAE decoder configuration. Please re-export the model with newer version of optimum and diffusers.
The scaling_factor attribute is missing from the VAE encoder configuration. Please re-export the model with newer version of optimum and diffusers.
The generated image is here, prompt is "a lovely cat"