how?

[zachysaur]

how can you add your own character and clothing lora

[maszhongming]

https://github.com/maszhongming/Multi-LoRA-Composition/blob/8321b9084d260ab927a42971d95afe1fcf7bf806/example.py#L40

You can modify the lora_path and weight_name during the initialization of LoRA to load your own LoRA weights.

[zachysaur]

this need to be from huggingface or you can just directly upload lora file

[maszhongming]

Both are possible. LoRA can be sourced from Hugging Face or your local files.

[zachysaur]

so downloading and generating pictures only required creating envirement running requirements putting model in checking point and use it?

[maszhongming]

The workflow, as described in the README, is:

1. Select a specific stable diffusion model.
2. Initialize LoRAs, which can be loaded from Hugging Face or from your local files.
3. Choose the method you want to use ("merge", "switch", "composite").
4. Provide the text prompts for the images you want to generate.
5. Generate the images.

Overall, the input consists of several LoRAs and prompts, without downloading images.