ashawkey / RAD-NeRF

Real-time Neural Radiance Talking Portrait Synthesis via Audio-spatial Decomposition
MIT License
862 stars 151 forks source link

"The name 'deepspeech/logits:0' refers to a Tensor which does not exist. The operation, 'deepspeech/logits', does not exist in the graph." #54

Closed 912602337 closed 1 year ago

912602337 commented 1 year ago

您好,https://github.com/osmr/deepspeech_features/releases/download/v0.0.1/deepspeech-0_1_0-b90017e8.pb.zip在这个地址手动下载的模型,为什么显示图里没有呢

OceanTan commented 1 year ago

您好,https://github.com/osmr/deepspeech_features/releases/download/v0.0.1/deepspeech-0_1_0-b90017e8.pb.zip在这个地址手动下载的模型,为什么显示图里没有呢

你好解决了吗?The name 'deepspeech/logits:0' refers to a Tensor which does not exist. The operation, 'deepspeech/logits', does not exist in the graph.

Suryavf commented 1 year ago

In the path 'data_utils/deepspeech_features/deepspeech_features.py' change deepspeech/logits:0 to logits:0.