实时数字人LiveTalking解压即用N卡整合版(兼容50卡)_251127

samcotes

Thanks a lot !!

sundl

求地址，非常感谢楼主分享

大酒鬼

回复一下学习学习

大酒鬼

希望老大多出几个人物模型，一个人物太单调了，不好玩，而且人物能放大，缩小全屏就更牛了，实时的数字人确实少

大酒鬼

0528可以用，这个用不了，我是16G P5200，出错如下：
INFO:logger:warmup model...
Traceback (most recent call last):
  File "K:\LiveTalking_251127\app.py", line 375, in <module>
    warm_up(opt.batch_size,model,256)
  File "K:\LiveTalking_251127\env\lib\site-packages\torch\utils\_contextlib.py", line 120, in decorate_context
    return func(*args, **kwargs)
  File "K:\LiveTalking_251127\lipreal.py", line 95, in warm_up
    model(mel_batch, img_batch)
  File "K:\LiveTalking_251127\env\lib\site-packages\torch\nn\modules\module.py", line 1775, in _wrapped_call_impl
    return self._call_impl(*args, **kwargs)
  File "K:\LiveTalking_251127\env\lib\site-packages\torch\nn\modules\module.py", line 1786, in _call_impl
    return forward_call(*args, **kwargs)
  File "K:\LiveTalking_251127\wav2lip\models\wav2lip_v2.py", line 132, in forward
    audio_embedding = self.audio_encoder(audio_sequences)  # [bz*5, 1, 80, 16]->[bz*5, 512, 1, 1]
  File "K:\LiveTalking_251127\env\lib\site-packages\torch\nn\modules\module.py", line 1775, in _wrapped_call_impl
    return self._call_impl(*args, **kwargs)
  File "K:\LiveTalking_251127\env\lib\site-packages\torch\nn\modules\module.py", line 1786, in _call_impl
    return forward_call(*args, **kwargs)
  File "K:\LiveTalking_251127\env\lib\site-packages\torch\nn\modules\container.py", line 250, in forward
    input = module(input)
  File "K:\LiveTalking_251127\env\lib\site-packages\torch\nn\modules\module.py", line 1775, in _wrapped_call_impl
    return self._call_impl(*args, **kwargs)
  File "K:\LiveTalking_251127\env\lib\site-packages\torch\nn\modules\module.py", line 1786, in _call_impl
    return forward_call(*args, **kwargs)
  File "K:\LiveTalking_251127\wav2lip\models\conv.py", line 16, in forward
    out = self.conv_block(x)
  File "K:\LiveTalking_251127\env\lib\site-packages\torch\nn\modules\module.py", line 1775, in _wrapped_call_impl
    return self._call_impl(*args, **kwargs)
  File "K:\LiveTalking_251127\env\lib\site-packages\torch\nn\modules\module.py", line 1786, in _call_impl
    return forward_call(*args, **kwargs)
  File "K:\LiveTalking_251127\env\lib\site-packages\torch\nn\modules\container.py", line 250, in forward
    input = module(input)
  File "K:\LiveTalking_251127\env\lib\site-packages\torch\nn\modules\module.py", line 1775, in _wrapped_call_impl
    return self._call_impl(*args, **kwargs)
  File "K:\LiveTalking_251127\env\lib\site-packages\torch\nn\modules\module.py", line 1786, in _call_impl
    return forward_call(*args, **kwargs)
  File "K:\LiveTalking_251127\env\lib\site-packages\torch\nn\modules\conv.py", line 548, in forward
    return self._conv_forward(input, self.weight, self.bias)
  File "K:\LiveTalking_251127\env\lib\site-packages\torch\nn\modules\conv.py", line 543, in _conv_forward
    return F.conv2d(
torch.AcceleratorError: CUDA error: no kernel image is available for execution on the device
Search for `cudaErrorNoKernelImageForDevice' in https://docs.nvidia.com/cuda/cud ... _CUDART__TYPES.html for more information.
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1
Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.

请按任意键继续. . .

星空

楼主的作品很不错，支持下！期待能发表更多好作品!