theluyuan/DiffSynth-Studio
1
0
Fork 0
mirror of https://github.com/modelscope/DiffSynth-Studio.git synced 2026-03-22 00:38:11 +00:00
Code Issues Packages Projects Releases Wiki Activity

support wan2.2 A14B I2V&T2V

Browse Source
This commit is contained in:
mi804
2025-07-25 17:09:53 +08:00
parent 3aed244c6f
commit 9015d08927
6 changed files with 175 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files

32
examples/wanvideo/model_inference/Wan2.2-I2V-A14B.py Normal file
View File

@@ -0,0 +1,32 @@
import torch
from PIL import Image
from diffsynth import save_video
from diffsynth.pipelines.wan_video_new import WanVideoPipeline, ModelConfig
from modelscope import dataset_snapshot_download
pipe = WanVideoPipeline.from_pretrained(
torch_dtype=torch.bfloat16,
device="cuda",
model_configs=[
ModelConfig(model_id="Wan-AI/Wan2.2-I2V-A14B", origin_file_pattern="high_noise_model/diffusion_pytorch_model*.safetensors", offload_device="cpu"),
ModelConfig(model_id="Wan-AI/Wan2.2-I2V-A14B", origin_file_pattern="low_noise_model/diffusion_pytorch_model*.safetensors", offload_device="cpu"),
ModelConfig(model_id="Wan-AI/Wan2.2-I2V-A14B", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
ModelConfig(model_id="Wan-AI/Wan2.2-I2V-A14B", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
],
)
pipe.enable_vram_management()
dataset_snapshot_download(
dataset_id="DiffSynth-Studio/examples_in_diffsynth",
local_dir="./",
allow_file_pattern=["data/examples/wan/cat_fightning.jpg"]
)
input_image = Image.open("data/examples/wan/cat_fightning.jpg").resize((832, 480))
# Text-to-video
video = pipe(
prompt="Two anthropomorphic cats in comfy boxing gear and bright gloves fight intensely on a spotlighted stage.",
negative_prompt="色调艳丽，过曝，静态，细节模糊不清，字幕，风格，作品，画作，画面，静止，整体发灰，最差质量，低质量，JPEG压缩残留，丑陋的，残缺的，多余的手指，画得不好的手部，画得不好的脸部，畸形的，毁容的，形态畸形的肢体，手指融合，静止不动的画面，杂乱的背景，三条腿，背景人很多，倒着走",
seed=0, tiled=True,
input_image=input_image,
)
save_video(video, "video1.mp4", fps=15, quality=5)

27
examples/wanvideo/model_inference/Wan2.2-T2V-A14B.py Normal file
View File

@@ -0,0 +1,27 @@
import torch
from diffsynth import save_video
from diffsynth.pipelines.wan_video_new import WanVideoPipeline, ModelConfig
from modelscope import snapshot_download
snapshot_download("Wan-AI/Wan2.2-T2V-A14B", local_dir="models/Wan-AI/Wan2.2-T2V-A14B")
pipe = WanVideoPipeline.from_pretrained(
torch_dtype=torch.bfloat16,
device="cuda",
model_configs=[
ModelConfig(model_id="Wan-AI/Wan2.2-T2V-A14B", origin_file_pattern="high_noise_model/diffusion_pytorch_model*.safetensors", offload_device="cpu"),
ModelConfig(model_id="Wan-AI/Wan2.2-T2V-A14B", origin_file_pattern="low_noise_model/diffusion_pytorch_model*.safetensors", offload_device="cpu"),
ModelConfig(model_id="Wan-AI/Wan2.2-T2V-A14B", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
ModelConfig(model_id="Wan-AI/Wan2.2-T2V-A14B", origin_file_pattern="Wan2.1_VAE.pth", offload_device="cpu"),
],
)
pipe.enable_vram_management()
# Text-to-video
video = pipe(
prompt="Two anthropomorphic cats in comfy boxing gear and bright gloves fight intensely on a spotlighted stage.",
negative_prompt="色调艳丽，过曝，静态，细节模糊不清，字幕，风格，作品，画作，画面，静止，整体发灰，最差质量，低质量，JPEG压缩残留，丑陋的，残缺的，多余的手指，画得不好的手部，画得不好的脸部，畸形的，毁容的，形态畸形的肢体，手指融合，静止不动的画面，杂乱的背景，三条腿，背景人很多，倒着走",
seed=0, tiled=True,
)
save_video(video, "video1.mp4", fps=15, quality=5)

8
examples/wanvideo/model_inference/Wan2.2-TI2V-5B.py
View File

@@ -1,17 +1,15 @@
import torch
from PIL import Image
from diffsynth import save_video, VideoData
from diffsynth import save_video
from diffsynth.pipelines.wan_video_new import WanVideoPipeline, ModelConfig
from modelscope import snapshot_download
from diffsynth.models.utils import load_state_dict, hash_state_dict_keys
from modelscope import dataset_snapshot_download
pipe = WanVideoPipeline.from_pretrained(
torch_dtype=torch.bfloat16,
device="cuda",
model_configs=[
ModelConfig(model_id="Wan-AI/Wan2.1-T2V-14B", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
ModelConfig(model_id="Wan-AI/Wan2.2-TI2V-5B", origin_file_pattern="model_shards/model-*.safetensors", offload_device="cpu"),
ModelConfig(model_id="Wan-AI/Wan2.2-TI2V-5B", origin_file_pattern="models_t5_umt5-xxl-enc-bf16.pth", offload_device="cpu"),
ModelConfig(model_id="Wan-AI/Wan2.2-TI2V-5B", origin_file_pattern="diffusion_pytorch_model*.safetensors", offload_device="cpu"),
ModelConfig(model_id="Wan-AI/Wan2.2-TI2V-5B", origin_file_pattern="Wan2.2_VAE.safetensors", offload_device="cpu"),
],
)