LTX-2.3-skin-hair

This LoRA is designed to refine the texture and physical properties of skin and hair for LTX-2.3. It focuses on reducing "plastic" skin artifacts and improving the specular highlights and movement of hair strands.

Training Progression Comparison

The video below demonstrates the training impact across different intervals. Sequence: Base Model (No LoRA) $\rightarrow$ 1,000 steps $\rightarrow$ 2,000 steps $\rightarrow$ 3,000 steps $\rightarrow$ 4,000 steps.


Model Details

  • Developer: Burgstall
  • Base model: Lightricks/LTX-2.3 (LoRA adapter)
  • Checkpoint Sizes: 353 MB (.safetensors)
  • Training focus: High-fidelity human textures.

Usage Instructions

ComfyUI

  1. Download your preferred step checkpoint.
  2. Place it in your models/loras folder.
  3. Use the LoraLoader node and connect it between your Model and CLIP inputs.
  4. Recommended strength: 0.75 - 0.9 for best results without over-sharpening.

Trigger Words

While this is a fine-tune of the latent space, using descriptive prompts helps guide the LoRA:

  • extreme close up, pores, detailed skin, individual hair strands, photorealistic.

License

This LoRA is a Derivative of LTX-2 (as defined in the LTX-2 Community License Agreement) and is distributed under the LTX-2 Community License Agreement (Lightricks Ltd., January 5, 2026). A complete copy is included in this repository as LICENSE.

Use is subject to the Agreement's use-based restrictions (Section 4 and Attachment A). Entities with annual revenues of at least US$10,000,000 must obtain a Commercial Use Agreement from Lightricks before using LTX-2 or its Derivatives.

LoRA weights and this model card © 2026 Burgstall. No additional restrictions beyond the LTX-2 Community License Agreement are imposed on Burgstall's contribution.

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW

This task can take several minutes

Model tree for TheBurgstall/LTX-2.3-skin-hair

Adapter
(121)
this model