Date: 02.01.2026

WAN 2.1 1.3b diffusers CUDA PyTorch BNB4 Test

Test environment

• Workstation 40 GB RAM, 500GB SSD, 750W Power supply
• Ubuntu 24.04 LTS HWE Kernel
• Install python 3.12

My test environment: HP Z440 + NVIDIA RTX 3090

Ubuntu preparation

sudo apt-get install --install-recommends linux-generic-hwe-24.04
hwe-support-status --verbose
sudo apt dist-upgrade
sudo reboot

Driver setup

• Install drivers nvidia-driver-570

sudo apt install nvidia-driver-570 clinfo
sudo reboot

• Check installation

nvidia-smi
clinfo

Check CUDA in python

• Priparing PyTorch

mkdir -p ~/llm && cd ~/llm
python3 -m venv .venv_llm_wan
source ./.venv_llm_wan/bin/activate
python -m pip install --upgrade pip
pip install "torch==2.7.1" "torchvision==0.22.1" "torchaudio==2.7.1" --index-url https://download.pytorch.org/whl/cu128
pip install "bitsandbytes==0.46.1"
pip install transformers accelerate diffusers safetensors
pip install ftfy opencv-python imageio imageio-ffmpeg
python3 -c "import torch; print(torch.__version__); print(torch.cuda.is_available());print(torch.cuda.get_device_name(0));"

• Expected responce

2.7.0+cu128
True
NVIDIA GeForce RTX 3090

• Check BitsAndBytes installation

python -m bitsandbytes

Steps

Get the WAN 2.1

git lfs install
git clone https://huggingface.co/Wan-AI/Wan2.1-VACE-1.3B-diffusers wan2.1d

Create script test_cuda_wan2.1d_prompt.py:

import torch
from diffusers import AutoencoderKLWan, WanVACEPipeline, WanVACETransformer3DModel
from transformers import UMT5EncoderModel, AutoTokenizer

from diffusers.utils import load_image, export_to_video

model_dir = "/home/sysadmin/llm/wan2.1d"

vae = AutoencoderKLWan.from_pretrained(model_dir, subfolder="vae", torch_dtype=torch.float16)

pipe = WanVACEPipeline.from_pretrained(
    model_dir, 
    vae=vae, 
    torch_dtype=torch.float16,
    safety_checker=None,
).to("cuda")


prompt = "ginger cat sits on a chair"
negative = "text on screen, watermarks, blurry, distortion, low quality"

seed = 7155058141399808394 #torch.seed()
generator = torch.Generator("cuda").manual_seed(seed)

result = pipe(
    prompt=prompt,
    negative_prompt=negative,
    #reference_images=[ref], 
    height=512,
    width=512,
    num_frames=40, 
    num_inference_steps=20,
    guidance_scale=5.0,
    generator=generator
)

export_to_video(result.frames[0], f"test_video_wan2.1_{seed}.mp4", fps=16)

Create script test_cuda_bnb4_wan2.1_prompt.py:

import torch
from diffusers import AutoencoderKLWan, WanVACEPipeline, WanVACETransformer3DModel
from diffusers import BitsAndBytesConfig as DdiffusersBitsAndBytesConfig
from transformers import UMT5EncoderModel, AutoTokenizer
from transformers import BitsAndBytesConfig as TransformersBitsAndBytesConfig

from diffusers.utils import load_image, export_to_video

model_dir = "/home/sysadmin/llm/wan2.1d"

bnb_tx = DdiffusersBitsAndBytesConfig(
    load_in_4bit=True,
    bnb_4bit_quant_type="nf4",
    bnb_4bit_use_double_quant=True, 
    bnb_4bit_compute_dtype=torch.float16,
)
transformer = WanVACETransformer3DModel.from_pretrained(
    model_dir, 
    subfolder="transformer",
    quantization_config=bnb_tx, 
    torch_dtype=torch.float16
)

bnb_te = TransformersBitsAndBytesConfig(
    load_in_4bit=True,
    bnb_4bit_quant_type="nf4",
    bnb_4bit_use_double_quant=True, 
    bnb_4bit_compute_dtype=torch.float16,
)
text_encoder = UMT5EncoderModel.from_pretrained(
    model_dir,
    subfolder="text_encoder",
    quantization_config=bnb_te, 
    torch_dtype=torch.float16
)

vae = AutoencoderKLWan.from_pretrained(model_dir, subfolder="vae", torch_dtype=torch.float16)

pipe = WanVACEPipeline.from_pretrained(
    model_dir, 
    vae=vae, 
    transformer=transformer, 
    text_encoder=text_encoder,
    torch_dtype=torch.float16
).to("cuda")

prompt = "ginger cat sits on a chair"
negative = "text on screen, watermarks, blurry, distortion, low quality"

seed = 7155058141399808394 #torch.seed()
generator = torch.Generator("cuda").manual_seed(seed)

result = pipe(
    prompt=prompt,
    negative_prompt=negative,
    #reference_images=[ref], 
    height=512,
    width=512,
    num_frames=40, 
    num_inference_steps=20,
    guidance_scale=5.0,
    generator=generator
)

export_to_video(result.frames[0], f"test_video_wan2.1_bnb4_{seed}.mp4", fps=16)

Run test

Check nvidia-smi during the test while true; do nvidia-smi; sleep 1; done

• Without quantization

python ./test_cuda_wan2.1_prompt.py

• With quantization 4bit

python ./test_cuda_bnb4_wan2.1_prompt.py

Open test_video_wan2.1_7155058141399808394.mp4 and test_video_wan2.1_bnb4_7155058141399808394.mp4

• Enjoy the result!