Stable Diffusion · LoRA

Microverse Redmond WAN2 T2V 14B

LoRA for Wan2.2-T2V-A14B that generates animated time-lapse videos of microbial colonies forming shapes like clocks, trees, fish, and volcanoes inside a petr…

About this model

This LoRA fine-tunes the Wan2.2-T2V-A14B text-to-video base model to create animated time-lapse videos of microscopic life. Bacteria, fungi, and microorganisms grow and evolve into recognizable shapes—clocks with sweeping hands, upward-growing trees, swimming fish, beating hearts, and erupting volcanoes—all within the confines of a petri dish on agar. Each animation captures macro-detail with scientific precision and artistic flair, blending microbiology aesthetics with surreal visual outcomes.

The trigger phrase is "Microverse." For best results, begin your prompt with "Microverse" followed by a description of what you want the microorganisms to form. The model supports five built-in example prompts covering clocks, trees, fish, hearts, and volcanic eruptions, each emphasizing macro lens detail and time-lapse growth.

To use this LoRA, place the file `[WAN2.2]Microverse_Redmond_low_noise.safetensors` in your `ComfyUI/models/loras/` directory. Load your base Wan model, add a Load LoRA node, connect the model output to the KSampler, and set `strength_model` to 0.8–1.0 for a strong effect. The model is released under the cc0-1.0 license and was trained with GPU time sponsored by Redmond Ai.

Training is based on Wan-AI/Wan2.2-T2V-A14B. The LoRA file is approximately 458.65 MB. Sample generations demonstrate vibrant bacterial colony colors—electric blues shifting to neon greens—against stark black backgrounds, emphasizing every textured detail of the microbial formations.