import torch
from diffusers import DiffusionPipeline
from diffusers.utils import load_image, export_to_video
# switch to "mps" for apple devices
pipe = DiffusionPipeline.from_pretrained("Wan-Video/Wan2.1-I2V-14B-720P", dtype=torch.bfloat16, device_map="cuda")
pipe.load_lora_weights("mycoolusername509732/cool-mc")
prompt = "A man with short gray hair plays a red electric guitar."
input_image = load_image("https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/guitar-man.png")
image = pipe(image=input_image, prompt=prompt).frames[0]
export_to_video(output, "output.mp4")Wan 2.1 I2V LoRA - [Your Model Name]
This is a LoRA trained for the Wan 2.1 Image-to-Video 14B model using the WaveSpeed-AI Wan-2.1-I2V-LoRA-Trainer.
Model Details
- Trained by: [Your Name/Handle]
- Base Model: Wan 2.1 I2V 14B (720p)
- Training Tool: WaveSpeed-AI
- LoRA Type: Separate High/Low Noise LoRAs
Files Included
This repository contains two versions of the LoRA weights:
- High Noise LoRA: Optimized for the early stages of the diffusion process (setting the scene and movement).
- Low Noise LoRA: Optimized for the later stages of the diffusion process (refining details and textures).
Usage
To use these weights, you typically apply them to the Wan 2.1 Transformer.
Recommended Settings
- Trigger Words: [Insert trigger words used during training, e.g., "in the style of SKS"]
- LoRA Strength: 0.6 - 1.0
- Resolution: 720x1280 or 1280x720
Training Info
- Dataset Size: [Number of videos]
- Number of Steps: [Total steps]
- Precision: bf16
- Downloads last month
- 13