metadata

license: apache-2.0
language:
  - en
base_model:
  - Wan-AI/Wan2.1-T2V-14B
pipeline_tag: text-to-video
tags:
  - text-to-video
  - text-to-image
  - lora
  - diffusers
  - template:diffusion-lora
widget:
  - text: >-
      A winding river flows through a lush valley, surrounded by towering
      mountains and vibrant flowers. In the distance, a large, floating island
      hovers in the sky. The sky is a mix of orange and blue.
    output:
      url: example_videos/1.mp4
  - text: >-
      A large, gothic castle stands on a cliff, with a large waterfall cascading
      down the rock face.  Four knights on horseback are in the foreground, on a
      small grassy outcropping, and the sky is grey and overcast.
    output:
      url: example_videos/2.mp4
  - text: >-
      f4nt4sy_sc3n3 fantasy landscape, A forest of massive, twisting trees with
      glowing silver leaves stretches out beneath a sky filled with twin moons.
      The camera follows a soft breeze as the leaves rustle, causing tiny
      glowing spores to drift into the air. In the distance, a waterfall flows
      in reverse, its water spiraling up into the sky before vanishing into the
      clouds.
    output:
      url: example_videos/3.mp4
  - text: >-
      f4nt4sy_sc3n3 fantasy landscape, A golden field of tall grass sways in the
      wind as the camera moves slowly forward. In the distance, a massive
      ancient statue, half-buried in the earth, stands against the backdrop of a
      glowing sunset. Birds take off from the statue’s shoulder, scattering into
      the sky as the last rays of light catch the cracks in the stone, revealing
      glowing veins of magic within.
    output:
      url: example_videos/4.mp4

Fantasy Landscape Style LoRA for Wan2.1 14B T2V

Overview

This LoRA is trained on the Wan2.1 14B T2V model and allows you to generate cinematic fantasy landscape videos!

Features

Trained on the Wan2.1 14B T2V base model
Consistent results across different object and scenes types
Simple prompt examples that are easy to adapt

Community

Discord: Join our community to generate videos with this LoRA for free
Request LoRAs: We're training and open-sourcing Wan2.1 LoRAs for free - join our Discord to make requests!

Prompt: A winding river flows through a lush valley, surrounded by towering mountains and vibrant flowers. In the distance, a large, floating island hovers in the sky. The sky is a mix of orange and blue.

Prompt: A large, gothic castle stands on a cliff, with a large waterfall cascading down the rock face. Four knights on horseback are in the foreground, on a small grassy outcropping, and the sky is grey and overcast.

Prompt: f4nt4sy_sc3n3 fantasy landscape, A forest of massive, twisting trees with glowing silver leaves stretches out beneath a sky filled with twin moons. The camera follows a soft breeze as the leaves rustle, causing tiny glowing spores to drift into the air. In the distance, a waterfall flows in reverse, its water spiraling up into the sky before vanishing into the clouds.

Prompt: f4nt4sy_sc3n3 fantasy landscape, A golden field of tall grass sways in the wind as the camera moves slowly forward. In the distance, a massive ancient statue, half-buried in the earth, stands against the backdrop of a glowing sunset. Birds take off from the statue’s shoulder, scattering into the sky as the last rays of light catch the cracks in the stone, revealing glowing veins of magic within.

Model File and Inference Workflow

📥 Download Links:

fantasy_50_epochs.safetensors - LoRA Model File
wan_txt2vid_lora_workflow.json - Wan T2V with LoRA Workflow for ComfyUI

Recommended Settings

LoRA Strength: 1.0
Embedded Guidance Scale: 6.0
Flow Shift: 5.0

Trigger Words

The key trigger phrase is: f4nt4sy_sc3n3 fantasy landscape

Prompt Template

For prompting, check out the example prompts; this way of prompting seems to work very well.

ComfyUI Workflow

This LoRA works with a modified version of Kijai's Wan Video Wrapper workflow. The main modification is adding a Wan LoRA node connected to the base model.

See the Downloads section above for the modified workflow.

Model Information

The model weights are available in Safetensors format. See the Downloads section above.

Training Details

Base Model: Wan2.1 14B T2V
Training Data: Trained on 1 minute of video comprised of 17 short clips (each clip captioned separately) of various fantasy landscapes.
Epochs: 50

Additional Information

Training was done using Diffusion Pipe for Training

Acknowledgments

Special thanks to Kijai for the ComfyUI Wan Video Wrapper and tdrussell for the training scripts!