This LoRA is meant to stay as close as possible to the original art style and duo to the limited amount of (GOOD) images of this character, the quality suffers a bit and it may come out as sketchy/drawn sometimes. This is some-what intended to an extent.

Settings i usually use:

• LoRA Weight: 0.9 – 1.2

• Sampler: Euler a (or DPMPP_2M + Karras if backgrounds break)

• Scheduler: Normal or beta_1_1

• Steps: 20–40

• CFG Scale: 5–7

• Resolution: 768×1024 (good starting point, i'm using 832 x 1216)

Trigger Word: DriftPixel_SayeonLee

Do NOT use the trigger word alone, results will be half-baked. Be very specific and descriptive in your prompt for best quality.

Recommended Universal Negative Prompt:

(low quality, worst quality:1.2), blurry, jpeg artifacts, bad anatomy, bad hands, deformed, distorted, disfigured, extra limbs, missing limbs, malformed limbs, fused fingers, too many fingers, mutated hands, poorly drawn hands, poorly drawn face, long neck, gross proportions, cropped, out of frame, watermark, text, signature, logo, username

ComfyUI Workflow Included!

Download any sample image I uploaded and drag & drop it into ComfyUI — it should load the full workflow from metadata.

This represents many hours of work. If you enjoy it, please 👍like, 💬 comment , and feel free to ⚡tip!