Wan2.2-Dasiwa - 专注动漫及艺术类视频生成 支持NSFW 一键整合下载
Wan2.2-Dasiwa (模型如其名,大丝袜),这是是一个基于Wan2.2的优化版高级视频生成模型,它就像一个“超级会拍艺术类短视频的AI导演”。你给它一张图片,它就能自动生成一段动作自然、细节丰富、很感性的短视频。特别擅长把静态图片变成动态的、诱惑力强的艺术视频。
Wan2.2-Dasiwa 默认已包含 NSFW 功能,能够生成高质量的 NSFW 主题视频,尤其是动漫类视频生成,无需额外的 LoRA。且经过多次优化改进,无论是生成视频质量还是动画效果,都优于同类开源模型。
主要特点
动作更流畅、自然:比很多同类模型动得更多、更真实(提升20-30%动作量),不会看起来僵硬或慢动作过多。
细节超强:即使分辨率不高,皮肤、眼睛、嘴唇、胸部、私密部位等都画得很精致,几乎全程保持一致(很少出现“视频里突然变丑”的情况)。
听提示词能力强:你写什么它就容易做出什么(prompt adherence高),包括各种姿势、表情、特殊效果。
很适合NSFW:内置了丰富的性感理解,能轻松生成诱惑、性感动作等,但也不会过度偏向某个极端风格。
应用领域
主要用于艺术视频生成(NSFW短视频、动态图、OnlyFans风格内容等)。
也支持普通SFW(非艺术类)内容,比如动画风、美女动态肖像、时尚动作等。
适合做办公室OL、丝袜诱惑、慢挑逗、性感舞蹈这类带故事感的短视频(跟你之前写的那个视频提示词风格非常搭)。
使用教程:(建议N卡,显存8G起,支持50系显卡)
整合包包含所需所有节点,下载主程序和模型(ComfyUI文件夹),解压主程序一键包,将ComfyUI文件夹移动到主程序目录下即可。
ComfyUI模式:
双击启动ComfyUI,进入WebUI后,点击左侧的 工作流程,选择对应的工作流,根据需要选择图生视频和首尾帧视频生成,左侧有开关,根据需要开启和关闭。
上传需要生成的图像,输入提示词,设置相关参数,运行即可。
实测8G显存可以运行,但需要开启共享显存,且运存需要32G起。
WebUI模式:
双击启动Webui,根据需要切换图生视频和首尾帧视频生成选项卡,上传图像,输入提示词,设置相关参数,运行生成即可。
模型切换:(包含高噪和低噪模型,模型手动下载后,保存到 ComfyUI\models\unet 目录下,运行WebUI,在模型切换下拉选择更新的模型)
默认开启Sage加速,如报错,可手动关闭加速。注. 开启sage_attention 加速需要本地安装VC编译环境,具体安装教程参考 此贴
提示词(prompt)写得越详细越好,比如加上“流畅运动、自然光影、高细节、电影感”等,能让效果更好。
一键包内包含 提示词生成技巧教程,教程来自B站UP Work-Fisher
软件目录结构:
📂 ComfyUI/
├── 📂 models/
│ ├── 📂 unet/
│ │ └── Dasiwa_WAN22-I2V14BSynthseduction_q4High.gguf
│ │ └── Dasiwa_WAN22-I2V14BSynthseduction_q4Low.gguf
│ ├── 📂 text_encoders/
│ │ └── nsfw_wan_umt5-xxl_fp8_scaled.safetensors
│ └── 📂 vae/
│ └── wan_2.1_vae.safetensors
📂 deepface/
......
下载地址:
迅雷云盘:https://pan.xunlei.com/s/VOp5qP1f9G0LEce98jq71UUkA1?pwd=m7km
夸克网盘:
**** 本内容需购买 ****
百度网盘:
**** 本内容需购买 ****
你好,wan2.2和wan2gp哪个更好用?显卡是4060,8g显存 longli 发表于 2026-4-1 15:30
你好,wan2.2和wan2gp哪个更好用?显卡是4060,8g显存
wan2gp 无言以对 发表于 2026-4-1 15:32
wan2gp
好的,感谢解答 用迅雷下载的ComfyUI文件夹,里面只有models\configs一个文件夹呢 cc142006596 发表于 2026-4-7 21:42
用迅雷下载的ComfyUI文件夹,里面只有models\configs一个文件夹呢
被吞了,已补 大佬,求助,请教个问题,执行生成时报错,第一次是提交任务失败,说是加载vae问题,手动选择了pixel_space这个之后解决了,问题1:后面又遇到问题,查资料好像说是vae默认用的wan2.1,不兼容,要用wan2.2,问题2:我下载了wan2.2放在新建的vae目录下,vae加载器选择这个又报错了,好像还是说这个vae加载用的默认就是2.1的,新建vae加载器其实也是2.1的,和wan2.2的模型不兼容,不知道我前面解决的是否正确,所有报错都贴出来了,从最初的报错开始:报错1:
“ got prompt
Failed to validate prompt for output 51:
* VAELoader 18:
- Value not in list: vae_name: 'wan_2.1_vae.safetensors' not in ['pixel_space']
Output will be ignored
Failed to validate prompt for output 56:
Output will be ignored
Failed to validate prompt for output 48:
Output will be ignored
invalid prompt: {'type': 'prompt_outputs_failed_validation', 'message': 'Prompt outputs failed validation', 'details': '', 'extra_info': {}}
Traceback (most recent call last):
File "C:\AI\ComfyUI_Wan2.2-Dasiwa\deepface\Lib\site-packages\gradio\queueing.py", line 759, in process_events
response = await route_utils.call_process_api(
File "C:\AI\ComfyUI_Wan2.2-Dasiwa\deepface\Lib\site-packages\gradio\blocks.py", line 1623, in call_function
prediction = await anyio.to_thread.run_sync(# type: ignore
File "C:\AI\ComfyUI_Wan2.2-Dasiwa\deepface\Lib\site-packages\anyio\to_thread.py", line 56, in run_sync
return await get_async_backend().run_sync_in_worker_thread(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\AI\ComfyUI_Wan2.2-Dasiwa\deepface\Lib\site-packages\anyio\_backends\_asyncio.py", line 2485, in run_sync_in_worker_thread
return await future
^^^^^^^^^^^^
File "C:\AI\ComfyUI_Wan2.2-Dasiwa\deepface\Lib\site-packages\anyio\_backends\_asyncio.py", line 976, in run
result = context.run(func, *args)
^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\AI\ComfyUI_Wan2.2-Dasiwa\deepface\Lib\site-packages\gradio\utils.py", line 915, in wrapper
response = f(*args, **kwargs)
^^^^^^^^^^^^^^^^^^
File "<frozen app>", line 760, in generate_i2v
gradio.exceptions.Error: '提交任务失败'
”
报错2:
生成报错,“ got prompt
Requested to load PixelspaceConversionVAE
loaded completely 7591.3107421875 1.9073486328125e-06 True
0 models unloaded.
loaded partially 128.0 127.998046875 0
# 😺dzNodes: LayerStyle -> ImageScaleByAspectRatio V2 Processed 1 image(s).
Initializing block swap: 0%| | 0/40
Initializing block swap: 100%|██████████| 40/40
!!! Exception during processing !!! The size of tensor a (81) must match the size of tensor b (16) at non-singleton dimension 1
Traceback (most recent call last):
File "C:\AI\ComfyUI_Wan2.2-Dasiwa\ComfyUI\execution.py", line 498, in execute
output_data, output_ui, has_subgraph, has_pending_tasks = await get_output_data(prompt_id, unique_id, obj, input_data_all, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb, hidden_inputs=hidden_inputs)
File "C:\AI\ComfyUI_Wan2.2-Dasiwa\ComfyUI\execution.py", line 316, in get_output_data
return_values = await _async_map_node_over_list(prompt_id, unique_id, obj, input_data_all, obj.FUNCTION, allow_interrupt=True, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb, hidden_inputs=hidden_inputs)
File "C:\AI\ComfyUI_Wan2.2-Dasiwa\ComfyUI\comfy\latent_formats.py", line 453, in process_in
return (latent - latents_mean) * self.scale_factor / latents_std
~~~~~~~^~~~~~~~~~~~~~
RuntimeError: The size of tensor a (81) must match the size of tensor b (16) at non-singleton dimension 1
Prompt executed in 1.33 seconds
got prompt
0 models unloaded.
loaded partially 128.0 127.998046875 0
Initializing block swap: 0%| | 0/40
Initializing block swap: 100%|██████████| 40/40
!!! Exception during processing !!! The size of tensor a (81) must match the size of tensor b (16) at non-singleton dimension 1
Traceback (most recent call last):
File "C:\AI\ComfyUI_Wan2.2-Dasiwa\ComfyUI\execution.py", line 498, in execute
output_data, output_ui, has_subgraph, has_pending_tasks = await get_output_data(prompt_id, unique_id, obj, input_data_all, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb, hidden_inputs=hidden_inputs)
File "C:\AI\ComfyUI_Wan2.2-Dasiwa\ComfyUI\comfy\model_base.py", line 1125, in concat_cond
image[:, i: i + latent_dim] = self.process_latent_in(image[:, i: i + latent_dim])
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\AI\ComfyUI_Wan2.2-Dasiwa\ComfyUI\comfy\model_base.py", line 317, in process_latent_in
return self.latent_format.process_in(latent)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\AI\ComfyUI_Wan2.2-Dasiwa\ComfyUI\comfy\latent_formats.py", line 453, in process_in
return (latent - latents_mean) * self.scale_factor / latents_std
~~~~~~~^~~~~~~~~~~~~~
RuntimeError: The size of tensor a (81) must match the size of tensor b (16) at non-singleton dimension 1
Prompt executed in 0.43 seconds
”
报错3:
我在vae加载器中选择pixel_space还是报错,选择wan2.2vae可以,但是又出现以下报错“ got prompt
Attempting to release mmap (468)
!!! Exception during processing !!! The size of tensor a (81) must match the size of tensor b (16) at non-singleton dimension 1
Traceback (most recent call last):
File "C:\AI\ComfyUI_Wan2.2-Dasiwa\ComfyUI\execution.py", line 498, in execute
output_data, output_ui, has_subgraph, has_pending_tasks = await get_output_data(prompt_id, unique_id, obj, input_data_all, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb, hidden_inputs=hidden_inputs)
File "C:\AI\ComfyUI_Wan2.2-Dasiwa\ComfyUI\execution.py", line 316, in get_output_data
return_values = await _async_map_node_over_list(prompt_id, unique_id, obj, input_data_all, obj.FUNCTION, allow_interrupt=True, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb, hidden_inputs=hidden_inputs)
~~~~~~~^~~~~~~~~~~~~~
RuntimeError: The size of tensor a (81) must match the size of tensor b (16) at non-singleton dimension 1
Prompt executed in 40.68 seconds
got prompt
Using pytorch attention in VAE
Using pytorch attention in VAE
VAE load device: cuda:0, offload device: cpu, dtype: torch.bfloat16
Requested to load WanVAE
loaded completely 4394.8172607421875 1344.0869674682617 True
0 models unloaded.
!!! Exception during processing !!! Given groups=1, weight of size , expected input to have 36 channels, but got 64 channels instead
Traceback (most recent call last):
File "C:\AI\ComfyUI_Wan2.2-Dasiwa\ComfyUI\execution.py", line 498, in execute
output_data, output_ui, has_subgraph, has_pending_tasks = await get_output_data(prompt_id, unique_id, obj, input_data_all, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb, hidden_inputs=hidden_inputs)
File "C:\AI\ComfyUI_Wan2.2-Dasiwa\ComfyUI\comfy\ops.py", line 172, in _conv_forward
return super()._conv_forward(input, weight, bias, *args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\AI\ComfyUI_Wan2.2-Dasiwa\deepface\Lib\site-packages\torch\nn\modules\conv.py", line 720, in _conv_forward
return F.conv3d(
^^^^^^^^^
RuntimeError: Given groups=1, weight of size , expected input to have 36 channels, but got 64 channels instead
Prompt executed in 11.51 seconds
”
3153532099 发表于 2026-4-9 18:00
大佬,求助,请教个问题,执行生成时报错,第一次是提交任务失败,说是加载vae问题,手动选择了pixel_space ...
vae没问题,检查你的帧数设置和分辨率设置
帧数为16的倍数+1
用webui默认的参数生成试试
无言以对 发表于 2026-4-9 18:31
vae没问题,检查你的帧数设置和分辨率设置
帧数为16的倍数+1
用webui默认的参数生成试试
感谢大佬的回复,参数是用的默认的,帧数和分辨率都是。无意中,回头又看到第一个报错“Value not in list: vae_name: 'wan_2.1_vae.safetensors' not in ['pixel_space']”,我找了一个wan_2.1_vae.safetensors,把它放在vae目录下,vae加载器选择加载这个,结果好了。。。 3153532099 发表于 2026-4-9 19:05
感谢大佬的回复,参数是用的默认的,帧数和分辨率都是。无意中,回头又看到第一个报错“Value not in lis ...
在哪找的 wan_2.1_vae.safetensors ??求问 我也这问题
页:
[1]
2