Very first credit for the idea comes from seeing <a target="_blank" rel="ugc" href="https://civarchive.com/user/Anaphiel">@Anaphiel</a> 's Art! Wow Just Wow. Please Post Results! I want to see if your art turned out as good as mine.Trained on <a target="_blank" rel="ugc" href="https://yorespot.com">https://yorespot.com</a> — where creators train faster, cheaper, and with minimal censorship (within legal limits). If you’re serious about building models, that’s where you should be.<hr />WAN 2.2 I2V LoRA========================================================<h2 id="trigger-word">Trigger Word</h2>ysp_ruftst12,<hr />OverviewThis is a WAN 2.2 I2V LoRA trained for high-quality portrait video generation. It uses a two-phase training approach to properly separate low-noise detail learning and high-noise motion structure — resulting in cleaner motion, stronger consistency, and better frame coherence than single-pass LoRAs.<hr />Training DetailsBase Model: • WAN 2.2 I2V 14B • (LOW + HIGH LoRA files)Method: • Musubi Tuner video LoRA workflow • Two-phase training (LOW timesteps → HIGH timesteps)Dataset: • 16 portrait video clips (512×512) • Hand-captioned (.txt per clip) • Full frame extraction (~30 FPS) • Up to 33 frames per clip • 4 repeats per clipTraining Config: • LoRA rank: 32 • Alpha: 32 • Learning rate: 2e-4 • Optimizer: AdamW (8-bit) • Scheduler: Cosine • Precision: bf16 + gradient checkpointingTraining Process: • 50 epochs (LOW noise model) • 50 epochs (HIGH noise model) • Total: 100 effective passes across both expertsTraining Time: • ~20 hours total runtime<hr />Recommended Usage (IMPORTANT)This is a dual-LoRA system. You MUST use both files together.HIGH noise LoRA: • Strength: 0.55LOW noise LoRA: • Strength: 0.90Best results come from using both simultaneously in a WAN 2.2 I2V workflow.Using only one will significantly reduce quality.<hr />Why This MattersMost LoRAs fail at motion consistency or overfit to static detail. This setup avoids that by splitting the learning problem:• LOW model → structure, identity, fine detail • HIGH model → motion, transitions, temporal coherenceThe result is smoother, more stable, and more realistic video output.<hr />If you want to train models like this yourself — faster, cheaper, and without the usual platform limitations — train at <a target="_blank" rel="ugc" href="https://yorespot.com">https://yorespot.com</a>.

Files with SHA256: 8b97d62f8a210e9197988e3abc7f5c33fda2068a5ee595340979544f82fad75b