Stable-LoRA
Stable-LoRA copied to clipboard
ExponentialML
Train LoRA using Microsoft's official implementation with Stable Diffusion models.
When I used code like that (want to lora train the unet part as well): `train_unet_lora = add_lora_to( unet, target_module=unet_replace, search_class=[torch.nn.Linear], r=args.lora_rank, lora_bias=args.lora_bias )` But it raise error: `File "train_lora.py",...
The IDE reminds me that "train_lora.py: error: the following arguments are required: --json_path" @ExponentialML
Metadata
Owner
Metadata
Train LoRA using Microsoft's official implementation with Stable Diffusion models.