How to load lora trained with Stable Cascade?

[zengjie617789]

I finished a lora traning based on Stable Cascade with onetrainer, but I cannot find a solution to load the load in diffusers pipeline. Anyone who can help me will be appreciated.

[sayakpaul]

Support not yet landed :) Stay tuned. Also, give us some more details:

• The LoRA you are trying to load.
• What steps you have already tried, etc.

[zengjie617789]

I used onetrainer to train lora model based on stable cascade.

1. the lora model weight are constructed below like:

'lora_prior_te_text_model_encoder_layers_0_mlp_fc1.alpha', 'lora_prior_te_text_model_encoder_layers_0_mlp_fc1.lora_down.weight', 'lora_prior_te_text_model_encoder_layers_0_mlp_fc1.lora_up.weight', 'lora_prior_te_text_model_encoder_layers_0_mlp_fc2.alpha', 'lora_prior_te_text_model_encoder_layers_0_mlp_fc2.lora_down.weight',

2. I did my best try to load lora offered by the officia onetrainer code in which use WuerstchenLoRAModelSampler to sample, and it is much different from the sd pipeline.

[sayakpaul]

I did my best try to load lora offered by the officia onetrainer code in which use WuerstchenLoRAModelSampler to sample, and it is much different from the sd pipeline.

Do you have a link to the LoRA?

Also, let us know about the code you used from diffusers to load that LoRA.

[github-actions[bot]]

This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.

Please note that issues that do not follow the contributing guidelines are likely to be ignored.

[sayakpaul]

Closing because of inactivity.