Doodle Toon Flux LoRA
Doodle Toon Flux was trained on approximately 44 AI generated images, the style combines the organic textures of markers and colored pencils with the whimsy of quirky doodle illustrations. The result is a fun Flux LoRA that feels like magic to use. I’m uploading both a v1 and v2 version that were trained with some different parameters, each has some distinct differences.
Each of the images in the gallery have a workflow and an example prompt attached to them, some used the model sampling node in ComfyUI, others did not. Explore and experiment to suit your taste. The dataset was auto captioned with Joy Caption Batch.
v1 – More flexible, more creative, better prompt adherence.
v2 – Enhanced textures but less flexibility, sometimes better sharpness and clarity.
Experiment with lowering the guidance scale from the default of 3.0-3.5 down to 2.5 to allow for more creative prompting, doing this will allow for more flexibility both in form and materials.
This LoRA was trained with 44 images captioned with Joy Caption Batch, 1 repeat, 2 batch, 150 epochs, 24 DIM/8 ALPHA for a total of 3250 Steps with a LR of 0.0005.
Trigger keyword: ‘d00dlet00n’ ‘illustration’
Suggested Tokens: colored pencil textures, alcohol marker, quirky doodle
Recommended strengths: 0.7 – 1
Link to Doodle Toon Flux LoRA on CivitAi
Link to Doodle Toon Flux LoRA on Hugging Face
Link to Doodle Toon Flux LoRA on Glif











