Hello everyone! This is a funky LoRA for creating illustration style images that have mostly bold lines, colour, and hint of graffiti. Great for designing t-shirt graphics. The prompts can be simple or complex. Tested only with SDXL in ComfyUI so not sure how it'll work in Automatic1111 without the refiner.
Assuming all the metadata is intact in the images, you should be able to drop this into your ComfyUI to see the workflow I made.
Please share your creations as I would love to see how this LoRA performs with different prompts.
The dataset mostly consists of images generated in SD1.5 using a previous LoRA I trained in a similar style. Hoping to share more. =)
Description
FAQ
Comments (12)
Freaking good! Well done!
Absolutely love the style! Doesn't work too well with image resolutions beyond 1024x1024. Tried a few images at 1920x1080 and you get a lot of doubling and random limbs.
This is a limitation of SDXL and not specific to this LoRA. If you want to end up with 1920x1080 images I would suggest generating them at a lower resolution then upscaling.
@Atteronious Yeah, I've realised that in the last few days. Time to get good at outpainting!
good job! would you tell me how much images using for this training? and parameters used. thx
very much.
There were 36 images total. I'm not sure which parameters you mean. The settings in Kohya?
@Rasali yes,setting in kohya for style training.
@tigerhe6866828 {
"LoRA_type": "Standard",
"adaptive_noise_scale": 0,
"additional_parameters": "--cache_text_encoder_outputs --network_train_unet_only",
"block_alphas": "",
"block_dims": "",
"block_lr_zero_threshold": "",
"bucket_no_upscale": true,
"bucket_reso_steps": 32,
"cache_latents": true,
"cache_latents_to_disk": true,
"caption_dropout_every_n_epochs": 0.0,
"caption_dropout_rate": 0,
"caption_extension": "",
"clip_skip": "1",
"color_aug": false,
"conv_alpha": 1,
"conv_block_alphas": "",
"conv_block_dims": "",
"conv_dim": 1,
"decompose_both": false,
"dim_from_weights": false,
"down_lr_weight": "",
"enable_bucket": true,
"epoch": 1,
"factor": -1,
"flip_aug": false,
"full_bf16": true,
"full_fp16": false,
"gradient_accumulation_steps": 1.0,
"gradient_checkpointing": true,
"keep_tokens": "0",
"learning_rate": 0.0001,
"logging_dir": " ",
"lora_network_weights": "",
"lr_scheduler": "constant_with_warmup",
"lr_scheduler_num_cycles": "",
"lr_scheduler_power": "",
"lr_warmup": 0,
"max_bucket_reso": 2048,
"max_data_loader_n_workers": "0",
"max_resolution": "1024,1024",
"max_timestep": 1000,
"max_token_length": "75",
"max_train_epochs": "",
"mem_eff_attn": false,
"mid_lr_weight": "",
"min_bucket_reso": 256,
"min_snr_gamma": 0,
"min_timestep": 0,
"mixed_precision": "bf16",
"model_list": "custom",
"module_dropout": 0,
"multires_noise_discount": 0,
"multires_noise_iterations": 0,
"network_alpha": 128,
"network_dim": 128,
"network_dropout": 0,
"no_token_padding": false,
"noise_offset": "0.0357",
"noise_offset_type": "Original",
"num_cpu_threads_per_process": 2,
"optimizer": "Adafactor",
"optimizer_args": "scale_parameter=False, relative_step=False, warmup_init=False",
"output_dir": " ",
"output_name": "DFunk_SDXL",
"persistent_data_loader_workers": false,
"pretrained_model_name_or_path": " ",
"prior_loss_weight": 1.0,
"random_crop": false,
"rank_dropout": 0,
"reg_data_dir": "",
"resume": "",
"sample_every_n_epochs": 0,
"sample_every_n_steps": 0,
"sample_prompts": "",
"sample_sampler": "euler_a",
"save_every_n_epochs": 1,
"save_every_n_steps": 0,
"save_last_n_steps": 0,
"save_last_n_steps_state": 0,
"save_model_as": "safetensors",
"save_precision": "bf16",
"save_state": false,
"scale_v_pred_loss_like_noise_pred": false,
"scale_weight_norms": 0,
"sdxl": true,
"sdxl_cache_text_encoder_outputs": false,
"sdxl_no_half_vae": true,
"seed": "1234",
"shuffle_caption": false,
"stop_text_encoder_training": 0,
"text_encoder_lr": 0.0001,
"train_batch_size": 1,
"train_data_dir": " ",
"train_on_input": true,
"training_comment": "",
"unet_lr": 0.0001,
"unit": 1,
"up_lr_weight": "",
"use_cp": false,
"use_wandb": false,
"v2": false,
"v_parameterization": false,
"vae_batch_size": 0,
"wandb_api_key": "",
"weighted_captions": false,
"xformers": true
}
@Rasali thx so much
I have had some exceptional UN-intended results, love it.
thank you sir, great work
Great love that comic look