<h3 id="how-to-use">How to Use</h3>This is a style LoRA, designed to be the base layer of your LoRA stack. It creates the foundational aesthetic of realism, upon which you can add character or concept LoRAs.Note: The included ZIP archive contains both the high-noise and low-noise LoRA variants, along with our recommended ComfyUI workflows.<ul><li>Trigger Word: <code>Instacam</code></li><li>Recommended Strength: <code>1.0</code>. Start here and adjust in small increments.</li></ul>I would also like to thank Danrisi, who originally taught us how to train LoRAs and helped make our work possible.

Instagirl WAN 2.2

Instagirl v2.5 is a significant update focused on increasing model diversity and detail fidelity.<ul><li>Expanded Dataset: This version is trained on a new, highly-curated dataset of over 2,400 images. The primary goal of this expansion was to reduce the "signature face" effect of previous versions and increase the variety of generated characters.</li><li>Increased Network Rank: Instagirl v2.5 low noise model was trained at a network rank of 256. This provides the LoRA with a significantly greater capacity for learning fine details, which can result in improved skin textures and more nuanced lighting, particularly in close-up compositions. High noise we have kept as rank 32 for smaller download size as it does not benefit from the higher rank which only helps with details.</li></ul>

Files with SHA256: 2835129849b588c2d71129344651402357d439f53beb2c3ccf5aac6cd8835763