Name: Flat Color - Style
Rating: 5 (5996 reviews)
Author: motimalu

We rebuilt the trainer. Watch your run as it trains, generate with an epoch before it's done, and compare epochs side by side. It's in beta, and you switch it on yourself in your preferences.

Training Studio: Watch It Train, Generate As It Goes

A training and content contest with over 2 Million Buzz in prizes and 50% off H3 generation for the duration of the contest.

The Civitai H3 Contest: 2 Million Buzz Up For Grabs

<h1 id="flat-color-style">Flat Color - Style</h1>Trained on images without visible lineart, flat colors, and little to no indication of depth.ℹ️ LoRA work best when applied to the base models on which they are trained. Please read the About This Version on the appropriate base models and workflow/training information.This is a small style LoRA I thought would be interesting to try with a v-pred model (noobai v-pred), for the reduced color bleeding and strong blacks in particular.The effect is quite nice and easy to evaluate in training, so I've extended the dataset with videos in following versions for text-to-video models like Wan and Hunyuan, and it is what I am generally using to test LoRA training on new models now.Recommended tags:<pre><code>flat color, no lineart, blending, negative space, {{color}} background</code></pre>

Trained on <a target="_blank" rel="ugc" href="https://civitai.red/models/2458426/anima?modelVersionId=2836417">Anima Preview 3 base</a>Updated dataset of some more recent works featuring flat colors and no line art.I think Anima has always been pretty good with flat color styles, but the outputs seem a bit saturated/high contrast so this LoRA I trained as a test seems to be worth releasing to help with that a little bit.Training config:<pre><code># trained using diffusion-pipe commit b0aa4f1e03169f3280c8518d37570a448420f8be

output_dir = '/mnt/d/anima/training_output/flat_color-v3'
dataset = 'dataset-anima-flat.toml'

# training settings
epochs = 5
# Per-resolution batch sizes
micro_batch_size_per_gpu = [[512, 64], [768, 32], [1024, 32]]
pipeline_stages = 1
gradient_accumulation_steps = 1
gradient_clipping = 1
warmup_steps = 100
lr_scheduler = 'cosine'

# misc settings
save_every_n_epochs = 1
activation_checkpointing = true
#reentrant_activation_checkpointing = true

partition_method = 'parameters'

save_dtype = 'bfloat16'
caching_batch_size = 1
map_num_proc = 8
steps_per_print = 1
compile = true

[model]
type = 'anima'
transformer_path = '/mnt/c/workspace/models/diffusion_models/anima-preview3-base.safetensors'
vae_path = '/mnt/c/workspace/models/vae/qwen_image_vae.safetensors'
llm_path = '/mnt/c/workspace/models/text_encoders/qwen_3_06b_base.safetensors'
dtype = 'bfloat16'
#cache_text_embeddings = false
llm_adapter_lr = 0
#timestep_sample_method = 'uniform'
flux_shift = true
multiscale_loss_weight = 0.5
sigmoid_scale = 1.3

[adapter]
type = 'lora'
rank = 32
dtype = 'bfloat16'

[optimizer]
type = 'adamw_optimi'
lr = 4e-5
betas = [0.9, 0.99]
weight_decay = 0.01
eps = 1e-8</code></pre>

v3.0 [anima-preview-3]

Trained on <a target="_blank" rel="ugc" href="https://huggingface.co/Tongyi-MAI/Z-Image">Z Image Base </a>with <a target="_blank" rel="ugc" href="https://github.com/tdrussell/diffusion-pipe">diffusion-pipe</a>Same dataset/training settings as used for version <code>2.1 [z-image-turbo]</code>Training Config:<pre><code># dataset-zimage.toml

# Resolution settings.
resolutions = [1024]

# Aspect ratio bucketing settings
enable_ar_bucket = true
min_ar = 0.5
max_ar = 2.0
num_ar_buckets = 7

[[directory]] # IMAGES
path = '/training_data/images'
num_repeats = 5
resolutions = [1024]

# config-zimage-base.toml

output_dir = '/mnt/d/zimage/training_output'
dataset = 'dataset-zimage.toml'

# training settings
epochs = 50
micro_batch_size_per_gpu = 1
pipeline_stages = 1
gradient_accumulation_steps = 1
gradient_clipping = 1

# eval settings
eval_every_n_epochs = 1
#eval_every_n_steps = 100
eval_before_first_step = true
eval_micro_batch_size_per_gpu = 1
eval_gradient_accumulation_steps = 1

# misc settings
save_every_n_epochs = 5
checkpoint_every_n_minutes = 120
activation_checkpointing = true
partition_method = 'parameters'
save_dtype = 'bfloat16'
caching_batch_size = 8
steps_per_print = 1

[model]
type = 'z_image'
diffusion_model = '/diffusion_models/z_image_bf16.safetensors'
vae = '/models/vae/ae.safetensors'
text_encoders = [
 {path = '/models/text_encoders/qwen_3_4b.safetensors', type = 'lumina2'}
]
dtype = 'bfloat16'
#diffusion_model_dtype = 'float8'

[adapter]
type = 'lora'
rank = 32
dtype = 'bfloat16'

[optimizer]
type = 'AdamW8bitKahan'
lr = 2e-5
betas = [0.9, 0.99]
weight_decay = 0.01</code></pre>

v2.1 [z-image-base]

Trained on <a target="_blank" rel="ugc" href="https://huggingface.co/Tongyi-MAI/Z-Image-Turbo">Z Image Turbo </a>with <a target="_blank" rel="ugc" href="https://github.com/tdrussell/diffusion-pipe">diffusion-pipe</a>Training Config:<pre><code># dataset-zimage.toml

# Resolution settings.
resolutions = [1024]

# Aspect ratio bucketing settings
enable_ar_bucket = true
min_ar = 0.5
max_ar = 2.0
num_ar_buckets = 7

[[directory]] # IMAGES
path = '/mnt/d/huanvideo/training_data/images'
num_repeats = 5
resolutions = [1024]

# config-zimage-base.toml

output_dir = '/mnt/d/zimage/training_output'
dataset = 'dataset-zimage.toml'

# training settings
epochs = 50
micro_batch_size_per_gpu = 1
pipeline_stages = 1
gradient_accumulation_steps = 1
gradient_clipping = 1

# eval settings
eval_every_n_epochs = 1
#eval_every_n_steps = 100
eval_before_first_step = true
eval_micro_batch_size_per_gpu = 1
eval_gradient_accumulation_steps = 1

# misc settings
save_every_n_epochs = 5
checkpoint_every_n_minutes = 120
activation_checkpointing = true
partition_method = 'parameters'
save_dtype = 'bfloat16'
caching_batch_size = 8
steps_per_print = 1

[model]
type = 'z_image'
diffusion_model = '/diffusion_models/z_image_bf16.safetensors'
vae = '/models/vae/ae.safetensors'
text_encoders = [
 {path = '/models/text_encoders/qwen_3_4b.safetensors', type = 'lumina2'}
]
# Use if training Z-Image-Turbo
merge_adapters = ['/models/loras/zimage_turbo_training_adapter_v2.safetensors']
dtype = 'bfloat16'
#diffusion_model_dtype = 'float8'

[adapter]
type = 'lora'
rank = 32
dtype = 'bfloat16'

[optimizer]
type = 'AdamW8bitKahan'
lr = 2e-5
betas = [0.9, 0.99]
weight_decay = 0.01</code></pre>

v2.1 [z-image-turbo]

Trained on <a target="_blank" rel="ugc" href="https://huggingface.co/black-forest-labs/FLUX.2-klein-base-4B">FLUX.2-klein-base-4B </a>with <a target="_blank" rel="ugc" href="https://github.com/tdrussell/diffusion-pipe">diffusion-pipe</a>

v2.1 [klein-4b]

Trained on <a target="_blank" rel="ugc" href="https://huggingface.co/Qwen/Qwen-Image">Qwen Image</a>Using the default <a target="_blank" rel="ugc" href="https://github.com/tdrussell/diffusion-pipe/blob/4a61bcfd82eb8d54a807ac7a15f548fab73ae5e1/examples/qwen_image_24gb_vram.toml">diffusion-pipe qwen config</a>Dataset resolution of 640Previews generated with the lightx2v Lighting 4step LoRA:<a target="_blank" rel="ugc" href="https://huggingface.co/lightx2v/Qwen-Image-Lightning/blob/main/Qwen-Image-Lightning-4steps-V1.0.safetensors">https://huggingface.co/lightx2v/Qwen-Image-Lightning/blob/main/Qwen-Image-Lightning-4steps-V1.0.safetensors</a>

v2.0 [qwen]

Trained on <a target="_blank" rel="ugc" href="https://civitai.com/models/795765/illustrious-xl">Illustrious XL 0.1</a>With the updated dataset for v2

v2.0 [Illustrious]

[WAN 14B] LoRA<ul><li>Trained with <a target="_blank" rel="ugc" href="https://github.com/tdrussell/diffusion-pipe">diffusion-pipe</a> on <a target="_blank" rel="ugc" href="https://huggingface.co/Wan-AI/Wan2.1-T2V-14B">Wan2.1-T2V-14B</a> with the same config as the v2.0 [hunyuan] version</li><li>Seems usable for both text-to-video and image-to-video workflows with Wan</li><li>Text to Video previews generated with <a target="_blank" rel="ugc" href="https://comfyanonymous.github.io/ComfyUI_examples/wan/#text-to-video">ComfyUI_examples/wan/#text-to-video</a><ul><li>Loading the LoRA with LoraLoaderModelOnly node and using the fp8 14B <a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/Wan_2.1_ComfyUI_repackaged/blob/main/split_files/diffusion_models/wan2.1_t2v_14B_fp8_e4m3fn.safetensors">wan2.1_t2v_14B_fp8_e4m3fn.safetensors</a></li><li>* it is recommended to use the newer scaled fp8 14B <a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/Wan_2.1_ComfyUI_repackaged/blob/main/split_files/diffusion_models/wan2.1_t2v_14B_fp8_scaled.safetensors">wan2.1_t2v_14B_fp8_scaled.safetensors</a></li></ul></li><li>Image to Video previews generated with <a target="_blank" rel="ugc" href="https://comfyanonymous.github.io/ComfyUI_examples/wan/#image-to-video">ComfyUI_examples/wan/#image-to-video</a><ul><li>Using image generations from the IL/Noobai versions of this model card</li></ul></li></ul>

v2.0 [wan-t2v-14b]

v2 changelog:- large dataset update to better target <code>flat color</code> and <code>no lineart</code> styles and techniques, including <code>blending</code>Trained on <a target="_blank" rel="ugc" href="https://civitai.com/models/833294/noobai-xl-nai-xl?modelVersionId=1190596">NoobAI-XL (NAI-XL) V-Pred 1.0-Version</a>With the kohya_ss dev branch, <code>v_paratemization</code> ,<code>zero_terminal_ssr</code> enabled, and noise offet set to 0.Previews are generated in Forge with DynamicThresholding (CFG-Fix) Integrated enabled, settings:<pre><code>dynthres_enabled: True, dynthres_mimic_scale: 5, dynthres_threshold_percentile: 1, dynthres_mimic_mode: Half Cosine Down, dynthres_mimic_scale_min: 1, dynthres_cfg_mode: Half Cosine Down, dynthres_cfg_scale_min: 3, dynthres_sched_val: 1, dynthres_separate_feature_channels: enable, dynthres_scaling_startpoint: ZERO, dynthres_variability_measure: STD, dynthres_interpolate_phi: 1</code></pre>A ComfyUI workflow reference with DynamicThresholding, Upscaling, and FaceDetailer can be found here: <a target="_blank" rel="ugc" href="https://civitai.com/posts/11457095">https://civitai.com/posts/11457095</a>

v2.0 [noobai-v-pred-1]

[WAN 2.2 TI2V 5B] LoRA<ul><li>Trained with <a target="_blank" rel="ugc" href="https://github.com/tdrussell/diffusion-pipe">diffusion-pipe</a> on <a target="_blank" rel="ugc" href="https://huggingface.co/Wan-AI/Wan2.2-TI2V-5B">Wan2.2-TI2V-5B</a><ul><li>Experimental - first test for Wan 2.2 training</li><li>Image dataset only</li><li>Less effect at longer framerates</li></ul></li></ul><ul><li>Text to Video previews generated with <a target="_blank" rel="ugc" href="https://comfyanonymous.github.io/ComfyUI_examples/wan22/">ComfyUI_examples/wan22/#text-to-video</a><ul><li>Loading the LoRA with LoraLoaderModelOnly node</li></ul></li></ul>dataset.toml<pre><code># Resolution settings.
resolutions = [1024]

# Aspect ratio bucketing settings
enable_ar_bucket = true
min_ar = 0.5
max_ar = 2.0
num_ar_buckets = 7

[[directory]] # IMAGES
# Path to the directory containing images and their corresponding caption files.
path = '/mnt/d/training_data/images'
num_repeats = 5
resolutions = [1024]</code></pre>config.toml<pre><code># Dataset config file.
output_dir = '/mnt/d/wan/training_output'
dataset = 'dataset.toml'

# Training settings
epochs = 50
micro_batch_size_per_gpu = 1
pipeline_stages = 1
gradient_accumulation_steps = 4
gradient_clipping = 1.0
warmup_steps = 100
# blocks_to_swap=32

# eval settings
eval_every_n_epochs = 5
eval_before_first_step = true
eval_micro_batch_size_per_gpu = 1
eval_gradient_accumulation_steps = 1

# misc settings
save_every_n_epochs = 5
checkpoint_every_n_minutes = 30
activation_checkpointing = true
partition_method = 'parameters'
save_dtype = 'bfloat16'
caching_batch_size = 1
steps_per_print = 1
video_clip_mode = 'single_middle'

[model]
type = 'wan'
ckpt_path = '../Wan2.2-TI2V-5B' 
dtype = 'bfloat16'
# You can use fp8 for the transformer when training LoRA.
transformer_dtype = 'float8'
timestep_sample_method = 'logit_normal'

[adapter]
type = 'lora'
rank = 32
dtype = 'bfloat16'

[optimizer]
type = 'adamw_optimi'
lr = 5e-5
betas = [0.9, 0.99]
weight_decay = 0.02
eps = 1e-8</code></pre>

v2.0 [wan-ti2v-5b]

v2.0 [wan-t2v-1.3b]

[HIDREAM] LoRA<ul><li>Trained with <a target="_blank" rel="ugc" href="https://github.com/tdrussell/diffusion-pipe">diffusion-pipe</a> on <a target="_blank" rel="ugc" href="https://huggingface.co/HiDream-ai/HiDream-I1-Full">HiDream-I1-Full</a><ul><li>1480 steps on a 4090 over ~7hrs</li></ul></li><li>Previews generated with <a target="_blank" rel="ugc" href="https://comfyanonymous.github.io/ComfyUI_examples/hidream/#hidream-dev-workflow">ComfyUI_examples/hidream/#hidream-dev-workflow</a><ul><li>Loading the LoRA with LoraLoaderModelOnly node and using the HiDream DEV fp8 model: <a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/HiDream-I1_ComfyUI/blob/main/split_files/diffusion_models/hidream_i1_dev_fp8.safetensors">hidream_i1_dev_fp8.safetensors</a></li></ul></li></ul>config.toml<pre><code># Dataset config file.
output_dir = '/mnt/d/hidream/training_output'
dataset = 'dataset-hidream.toml'

# Training settings
epochs = 50
micro_batch_size_per_gpu = 1
pipeline_stages = 1
gradient_accumulation_steps = 4
gradient_clipping = 1.0
warmup_steps = 100
blocks_to_swap = 20

# eval settings
eval_every_n_epochs = 5
eval_before_first_step = true
eval_micro_batch_size_per_gpu = 1
eval_gradient_accumulation_steps = 1

# misc settings
save_every_n_epochs = 10
checkpoint_every_n_minutes = 30
activation_checkpointing = true
partition_method = 'parameters'
save_dtype = 'bfloat16'
caching_batch_size = 1
steps_per_print = 1
video_clip_mode = 'single_beginning'

[model]
type = 'hidream'
diffusers_path = '../hidream-full'
llama3_path = '../llama-3.1'
llama3_4bit = true
dtype = 'bfloat16'
transformer_dtype = 'nf4'
max_llama3_sequence_length = 256
flux_shift = true

[adapter]
type = 'lora'
rank = 32
dtype = 'bfloat16'

[optimizer]
type = 'adamw_optimi'
lr = 5e-5
betas = [0.9, 0.99]
weight_decay = 0.02
eps = 1e-8</code></pre>dataset.toml<pre><code># Resolution settings.
resolutions = [1024]

# Aspect ratio bucketing settings
enable_ar_bucket = true
min_ar = 0.5
max_ar = 2.0
num_ar_buckets = 7

[[directory]] # IMAGES
# Path to the directory containing images and their corresponding caption files.
path = '/mnt/d/huanvideo/training_data/images'
num_repeats = 5
resolutions = [1024]</code></pre>

v2.0 [hidream]

Trained with <a target="_blank" rel="ugc" href="https://github.com/tdrussell/diffusion-pipe">https://github.com/tdrussell/diffusion-pipe</a>Training data consists of:<ul><li>42 images as a combination of<ul><li>Images used from other versions this model card</li><li>Images extracted as keyframes from several videos</li></ul></li><li>19 short video clips ~40 frames each</li></ul>Training configs:dataset.toml<pre><code># Aspect ratio bucketing settings
enable_ar_bucket = true
min_ar = 0.5
max_ar = 2.0
num_ar_buckets = 7

[[directory]] # IMAGES
# Path to the directory containing images and their corresponding caption files.
path = '/mnt/d/huanvideo/training_data/images'
num_repeats = 5
resolutions = [1024]
frame_buckets = [1] # Use 1 frame for images.


[[directory]] # VIDEOS
# Path to the directory containing videos and their corresponding caption files.
path = '/mnt/d/huanvideo/training_data/videos'
num_repeats = 5
resolutions = [256] # Set video resolution to 256 (e.g., 244p).
frame_buckets = [33, 49, 81] # Define frame buckets for videos.</code></pre>config.toml<pre><code># Dataset config file.
output_dir = '/mnt/d/huanvideo/training_output'
dataset = 'dataset.toml'

# Training settings
epochs = 50
micro_batch_size_per_gpu = 1
pipeline_stages = 1
gradient_accumulation_steps = 4
gradient_clipping = 1.0
warmup_steps = 100

# eval settings
eval_every_n_epochs = 5
eval_before_first_step = true
eval_micro_batch_size_per_gpu = 1
eval_gradient_accumulation_steps = 1

# misc settings
save_every_n_epochs = 15
checkpoint_every_n_minutes = 30
activation_checkpointing = true
partition_method = 'parameters'
save_dtype = 'bfloat16'
caching_batch_size = 1
steps_per_print = 1
video_clip_mode = 'single_middle'

[model]
type = 'hunyuan-video'

transformer_path = '/mnt/d/huanvideo/models/diffusion_models/hunyuan_video_720_cfgdistill_fp8_e4m3fn.safetensors'
vae_path = '/mnt/d/huanvideo/models/vae/hunyuan_video_vae_bf16.safetensors'
llm_path = '/mnt/d/huanvideo/models/llm'
clip_path = '/mnt/d/huanvideo/models/clip'

dtype = 'bfloat16'
transformer_dtype = 'float8'
timestep_sample_method = 'logit_normal'

[adapter]
type = 'lora'
rank = 32
dtype = 'bfloat16'

[optimizer]
type = 'adamw_optimi'
lr = 5e-5
betas = [0.9, 0.99]
weight_decay = 0.02
eps = 1e-8</code></pre>

v2.0 [hunyuan]

Trained as a small experiment with <a target="_blank" rel="ugc" href="https://github.com/tdrussell/diffusion-pipe">https://github.com/tdrussell/diffusion-pipe</a>Gives a charming flat color style to Hunyuan t2v generations

v1.0 [hunyuan]

Trained on <a target="_blank" rel="ugc" href="https://civitai.com/models/795765/illustrious-xl">Illustrious XL 0.1</a>

v1.0 [Illustrious]

Trained on <a target="_blank" rel="ugc" href="https://civitai.com/models/833294/noobai-xl-nai-xl?modelVersionId=1190596">NoobAI-XL (NAI-XL) V-Pred 1.0-Version</a>With the kohya_ss dev branch, <code>v_paratemization</code> ,<code>zero_terminal_ssr</code> enabled, and noise offet set to 0.