问仅用pth文件加载pth Tacotron2模型的问题

EN

我使用Mozilla在自定义数据集上训练了一个Tacotron2模型。训练器输出一个pth文件和一个config.json文件。我很难将经过训练的模型加载到PyTorch中。

from torchaudio.models.tacotron2 import Tacotron2
tacotron2 =Tacotron2()
tacotron2.load_state_dict(torch.load('models/best_model.pth'))

RuntimeError:在为Tacotron2加载state_dict时出错:state_dict中缺少键："embedding.weight“、”encoder.卷积s.0.0.权重“、”encoder.卷积s.0.0.偏向“、”encoder.卷积s.0.1.权重“、”encoder.卷积s.0.1.偏向“、"encoder.convolutions.0.1.running_mean”、"encoder.convolutions.0.1.running_var“、”encoder.卷积s.1.0.权重“，"encoder.convolutions.1.1.running_mean“、"encoder.convolutions.1.1.running_var”、“encoder.卷积s.2.0.权重”、“Encoder.卷积s.2.0.偏倚”、“encoder.卷积s.2.1偏倚”、“encoder.卷积s.2.1偏倚”、“encoder.卷积s.2.1偏倚”，"encoder.convolutions.2.1.running_mean“、"encoder.convolutions.2.1.running_var”、"encoder.lstm.weight_ih_l0“、"encoder.lstm.weight_hh_l0”、"encoder.lstm.bias_ih_l0“、"encoder.lstm.bias_hh_l0”、"encoder.lstm.weight_ih_l0_reverse“、"encoder.lstm.weight_hh_l0_reverse”、"encoder.lstm.bias_ih_l0_reverse““"decoder.attention_layer.memory_layer.weight"，”、"decoder.attention_rnn.weight_ih“、"decoder.attention_rnn.weight_hh”、"decoder.attention_rnn.bias_ih“、"decoder.attention_rnn.bias_hh”、"decoder.attention_layer.query_layer.weight“、”decoder.attention_rnn.bias_ih“"decoder.attention_layer.v.weight"，"decoder.decoder_rnn.weight_ih"，"decoder.decoder_rnn.weight_hh"，"decoder.decoder_rnn.bias_ih"，"decoder.decoder_rnn.bias_hh"，"decoder.linear_projection.weight"，"decoder.linear_projection.bias"，"decoder.gate_layer.weight“、"decoder.gate_layer.bias”、“postnet.卷积s.0.0.权重”、“postnet.卷积s.0.0.偏向”、“postnet.卷积s.0.1.权重”、“postnet.卷积s.0.1.偏向”、"postnet.convolutions.0.1.running_mean“、"postnet.convolutions.0.1.running_var”、“postnet.卷积s.1.0.权重”、“postnet.卷积s.1.0偏见”，“postnet.卷积s.1.1.重量”、“postnet.卷积s.1.1.偏向”、"postnet.convolutions.1.1.running_mean“、"postnet.convolutions.1.1.running_var”、“postnet.卷积s.2.0.权重”、“postnet.卷积s.2.0.偏向”、“postnet.卷积s.2.1.权重”、“postnet.卷积s.2.1.偏向”、"postnet.convolutions.2.1.running_mean"，"postnet.convolutions.2.1.running_var“、”Postnet.卷积s.3.0.权重“、”postnet.卷积s.3.0.偏倚“、”postnet.卷积s.3.1重量“、”postnet.卷积s.3.1.偏向“、"postnet.convolutions.3.1.running_mean”、"postnet.convolutions.3.1.running_var“、”postnet.卷积s.4.0.权重“、”postnet.卷积s.4.0.偏向“，“postnet.卷积s.4.1权重”、“postnet.卷积s.4.1.偏向”、"postnet.convolutions.4.1.running_mean“、"postnet.convolutions.4.1.running_var”。State_dict中的意外键：“配置”、“模型”、“优化器”、“标量器”、“步骤”、“时代”、“日期”、“model_loss”。