Follow me on X/Twitter for exclusive content: <a target="_blank" rel="ugc" href="https://x.com/Citron_Legacy">https://x.com/Citron_Legacy</a> 😁Check out all my Style Loras: <a target="_blank" rel="ugc" href="https://civarchive.com/collections/10766871">https://civarchive.com/collections/10766871</a>I intentionally used "sketchy" art so that this Lora generates not rough art not professional tier art.

Sketch (Style) [ZIT & SDXL]

[sdxl_arguments]cache_text_encoder_outputs = falseno_half_vae = truemin_timestep = 0max_timestep = 1000shuffle_caption = truelowram = true[model_arguments]pretrained_model_name_or_path = "stabilityai/stable-diffusion-xl-base-1.0"vae = "/content/vae/sdxl_vae.safetensors"[dataset_arguments]debug_dataset = falsein_json = "/content/LoRA/meta_lat.json"train_data_dir = "/content/drive/MyDrive/lora_training/datasets/Anime_Sketch_SDXL"dataset_repeats = 2keep_tokens = 1resolution = "1024,1024"color_aug = falsetoken_warmup_min = 1token_warmup_step = 0[training_arguments]output_dir = "/content/drive/MyDrive/kohya-trainer/output/Anime_Sketch_SDXL"output_name = "Anime_Sketch_SDXL"save_precision = "fp16"save_every_n_epochs = 1train_batch_size = 4max_token_length = 225mem_eff_attn = falsesdpa = truexformers = falsemax_train_epochs = 10max_data_loader_n_workers = 8persistent_data_loader_workers = truegradient_checkpointing = truegradient_accumulation_steps = 1mixed_precision = "fp16"[logging_arguments]log_with = "tensorboard"logging_dir = "/content/LoRA/logs"log_prefix = "Anime_Sketch_SDXL"[sample_prompt_arguments]sample_sampler = "euler_a"[saving_arguments]save_model_as = "safetensors"[optimizer_arguments]optimizer_type = "AdaFactor"learning_rate = 0.0001max_grad_norm = 0optimizer_args = [ "scale_parameter=False", "relative_step=False", "warmup_init=False",]lr_scheduler = "constant_with_warmup"lr_warmup_steps = 100[additional_network_arguments]no_metadata = falsenetwork_module = "networks.lora"network_dim = 32network_alpha = 16network_args = []network_train_unet_only = true

Anime_Sketch_SDXL.safetensors

Anime Sketch (Style) SDXL & SD1.5

SDXL

SD1.5_v1

<h3 id="heading-2">Want to learn how to draw Anime sketches? Now you can use AI to help you practice!</h3>I intentionally used some "sketchy" art so that its not professional tier art but rather something a bit more rough and easy to use as examples. <hr />This is a little unstable but it produces nice results with the right prompts. Please refer to my examples to see how to use it.I intended to use the 10 Epochs version but Google Colab crashed after 75 minutes. I figured I'd test out the 5 Epoch version while I waited another 120 minutes for the 10 Epochs. I want to refine this to make it more sketchy and more like lineart.<hr />Check out all the Art styles I’ve made: <a target="_blank" rel="ugc" href="https://civitai.com/user/CitronLegacy/models?tag=style">https://civitai.com/user/CitronLegacy/models?tag=style</a><hr />Training DataTrained on: 218 imagesTraining Model: SD-v1-5-pruned-noema-fp16flip_aug: TrueNum of Repeats: 10Unit is Epochs or Steps: EpochsNumber of Epochs or Steps: 10Training Batch Size: 2Network Dim: Dim 64 and Alpha 32Lora Creation Process took: about 75 minutes

Files with SHA256: 3e96580405e162ae438f103a20f6cd488aa6ee6c567be29d4308af996f7be17e