Instructions to use rzgar/minimax-h3_fl2v_8Step_motion_enhancer with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use rzgar/minimax-h3_fl2v_8Step_motion_enhancer with Diffusers:
pip install -U diffusers transformers accelerate
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("MiniMaxAI/MiniMax-H3", dtype=torch.bfloat16, device_map="cuda") pipe.load_lora_weights("rzgar/minimax-h3_fl2v_8Step_motion_enhancer") 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") - Inference
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- Draw Things
What settings should we use with it please?
strength? audio/video shift? sampler? schedular?
thank you.
sorry i downloaded your attached video and saw that you're not using audio/video shifts, should i do the same? or use it? and at what strengths, thanks again.
sorry i downloaded your attached video and saw that you're not using audio/video shifts, should i do the same? or use it? and at what strengths, thanks again.
Hi, you can use the same strength and settings as you would for the regular LightX2V 8-Step Turbo LoRA.
However, using a strength of 0.8 and increasing the steps to 10 or 12 tends to produce better animation and audio. (true for all speedup LoRAs)
If you use KSampler, it uses the default values for video and audio shift.