实时数字人 InfiniteTalk 半身

实时数字人InfiniteTalk

AI视觉网奇

260人浏览 · 2025-12-04 00:23:08

AI视觉网奇 · 2025-12-04 00:23:08 发布

InfiniteTalk v0828整合包更新

实时数字人只有头：

lam

浪子之心实时数字人收费

AI 浪子之心科技各种商用实时智能数字人效果展示我们的商用实时数字人，支持英伟达显卡，华为显卡，其他各型号CPU等；支持2D，2.5D，3D数字人；支持云端部署，本地部署，信创改造；支持打断、唤醒、VAD；支持MCP、物联、智能体接入；支持720p、1080p、2K，4K等清晰度；支持PC端，手机端，大屏幕，全息仓，网页端等。各位朋友：我们提供LangQing：商用高清实时定制数字人系统及源码。包含数字人训练代码，推理代码，实时流式推理代码，及全套实时数字人解决方案。我们提供LangYuan：商用高清泛化实时数字人系统及源码。可以直接上传视频就能进行实时对话或者实时数字人直播。我们提供LangXin：商用高清泛化非实时数字人系统及源码。包含API部署调用，非实时数字人短视频生成等服务。我们提供LangYing：商用手机端实时推理数字人系统及源码。商务合作请加wei xin 洽谈：langzizhixinkeji shou ji ：一八一一六五八一七六零（wei xin 同shou ji 号一样）

报错信息：

Nunchaku version: 0.3.1
ComfyUI-nunchaku version: 0.3.4
ComfyUI-nunchaku 0.3.4 is not compatible with nunchaku 0.3.1. Please update nunchaku to a supported version in ['v0.3.2'].
Please 'pip install apex'
Nodes `NunchakuFluxIPAdapterApply` and `NunchakuIPAdapterLoader` import failed:
Traceback (most recent call last):
  File "D:\BaiduNetdiskDownload\ComfyUI_InfiniteTalk（最新）\custom_nodes\ComfyUI-nunchaku\__init__.py", line 78, in <module>
    from .nodes.models.ipadapter import NunchakuFluxIPAdapterApply, NunchakuIPAdapterLoader
  File "D:\BaiduNetdiskDownload\ComfyUI_InfiniteTalk（最新）\custom_nodes\ComfyUI-nunchaku\nodes\models\ipadapter.py", line 14, in <module>
    from nunchaku.models.ip_adapter.diffusers_adapters import apply_IPA_on_pipe
ModuleNotFoundError: No module named 'nunchaku.models.ip_adapter'

报错信息2：

  from nunchaku import NunchakuT5EncoderModel
ImportError: cannot import name 'NunchakuT5EncoderModel' from 'nunchaku' (D:\BaiduNetdiskDownload\ComfyUI_InfiniteTalk（最新）\python\lib\site-packages\nunchaku\__init__.py)
Nodes `NunchakuPulidApply`,`NunchakuPulidLoader`, `NunchakuPuLIDLoaderV2` and `NunchakuFluxPuLIDApplyV2` import failed:
Traceback (most recent call last):
  File "D:\BaiduNetdiskDownload\ComfyUI_InfiniteTalk（最新）\custom_nodes\ComfyUI-nunchaku\__init__.py", line 61, in <module>
    from .nodes.models.pulid import (
  File "D:\BaiduNetdiskDownload\ComfyUI_InfiniteTalk（最新）\custom_nodes\ComfyUI-nunchaku\nodes\models\pulid.py", line 21, in <module>
    from nunchaku.models.pulid.pulid_forward import pulid_forward
ModuleNotFoundError: No module named 'nunchaku.models'
Nodes `NunchakuFluxIPAdapterApply` and `NunchakuIPAdapterLoader` import failed:
Traceback (most recent call last):
  File "D:\BaiduNetdiskDownload\ComfyUI_InfiniteTalk（最新）\custom_nodes\ComfyUI-nunchaku\__init__.py", line 78, in <module>
    from .nodes.models.ipadapter import NunchakuFluxIPAdapterApply, NunchakuIPAdapterLoader
  File "D:\BaiduNetdiskDownload\ComfyUI_InfiniteTalk（最新）\custom_nodes\ComfyUI-nunchaku\nodes\models\ipadapter.py", line 14, in <module>
    from nunchaku.models.ip_adapter.diffusers_adapters import apply_IPA_on_pipe
ModuleNotFoundError: No module named 'nunchaku.models'
Node `NunchakuModelMerger` import failed:
Traceback (most recent call last):
  File "D:\BaiduNetdiskDownload\ComfyUI_InfiniteTalk（最新）\custom_nodes\ComfyUI-nunchaku\__init__.py", line 86, in <module>
    from .nodes.tools.merge_safetensors import NunchakuModelMerger
  File "D:\BaiduNetdiskDownload\ComfyUI_InfiniteTalk（最新）\custom_nodes\ComfyUI-nunchaku\nodes\tools\merge_safetensors.py", line 11, in <module>
    from nunchaku.merge_safetensors import merge_safetensors
ModuleNotFoundError: No module named 'nunchaku.merge_safetensors'

安装库：

git clone https://github.com/nunchaku-tech/nunchaku.git

cd nunchaku
git checkout v0.3.2


../python/python.exe -m pip install .

git 地址：

https://github.com/MeiGen-AI/InfiniteTalk

下载模型：

huggingface-cli download Wan-AI/Wan2.1-I2V-14B-480P --local-dir ./weights/Wan2.1-I2V-14B-480P
huggingface-cli download TencentGameMate/chinese-wav2vec2-base --local-dir ./weights/chinese-wav2vec2-base
huggingface-cli download TencentGameMate/chinese-wav2vec2-base model.safetensors --revision refs/pr/1 --local-dir ./weights/chinese-wav2vec2-base
huggingface-cli download MeiGen-AI/InfiniteTalk --local-dir ./weights/InfiniteTalk

python generate_infinitetalk.py \
    --ckpt_dir weights/Wan2.1-I2V-14B-480P \
    --wav2vec_dir 'weights/chinese-wav2vec2-base' \
    --infinitetalk_dir weights/InfiniteTalk/single/infinitetalk.safetensors \
    --input_json examples/single_example_image.json \
    --size infinitetalk-480 \
    --sample_steps 40 \
    --mode streaming \
    --motion_frame 9 \
    --save_file infinitetalk_res