o i@sXdZddlmZmZddlZddlmZddlmZddl m Z Gdddej j Z dS) z`Posterior encoder module in VITS. This code is based on https://github.com/jaywalnut310/vits. )OptionalTupleN)WaveNet)Conv1d)make_non_pad_maskcseZdZdZ          dd ed ed ed edededededededeffdd Z ddej dej de ej de ej ej ej ej ffddZ Z S) PosteriorEncoderaSPosterior encoder module in VITS. This is a module of posterior encoder described in `Conditional Variational Autoencoder with Adversarial Learning for End-to-End Text-to-Speech`_. .. _`Conditional Variational Autoencoder with Adversarial Learning for End-to-End Text-to-Speech`: https://arxiv.org/abs/2006.04558 T in_channels out_channelshidden_channels kernel_sizelayersstacks base_dilationglobal_channels dropout_ratebiasuse_weight_normc stt||d|_tdiddddd|d|d|d|d |d dd |d d |d|d| d| d| dddddddd|_t||d d|_dS)aInitilialize PosteriorEncoder module. Args: in_channels (int): Number of input channels. out_channels (int): Number of output channels. hidden_channels (int): Number of hidden channels. kernel_size (int): Kernel size in WaveNet. layers (int): Number of layers of WaveNet. stacks (int): Number of repeat stacking of WaveNet. base_dilation (int): Base dilation factor. global_channels (int): Number of global conditioning channels. dropout_rate (float): Dropout rate. bias (bool): Whether to use bias parameters in conv. use_weight_norm (bool): Whether to apply weight norm. r rr rrrrrresidual_channels aux_channels gate_channels skip_channelsrrrruse_first_convF use_last_convscale_residualscale_skip_connectTN)super__init__r input_convrencoderproj) selfrrrrrrrrrrr __class__r#Z/home/ubuntu/.local/lib/python3.10/site-packages/espnet2/gan_tts/vits/posterior_encoder.pyr%sP      zPosteriorEncoder.__init__Nx x_lengthsgreturnc Cst|dj|j|jd}|||}|j|||d}|||}|j| dddd\}}|t |t ||}||||fS)a7Calculate forward propagation. Args: x (Tensor): Input tensor (B, in_channels, T_feats). x_lengths (Tensor): Length tensor (B,). g (Optional[Tensor]): Global conditioning tensor (B, global_channels, 1). Returns: Tensor: Encoded hidden representation tensor (B, out_channels, T_feats). Tensor: Projected mean tensor (B, out_channels, T_feats). Tensor: Projected scale tensor (B, out_channels, T_feats). Tensor: Mask tensor for input tensor (B, 1, T_feats). r )dtypedevice)r/r)dim) r unsqueezetor1r2r&r'r(splitsizetorch randn_likeexp) r)r-r.r/x_maskstatsmlogszr#r#r,forwardUs zPosteriorEncoder.forward) rr r r r r r r rTT)N)__name__ __module__ __qualname____doc__intfloatboolr%r8Tensorrrr@ __classcell__r#r#r*r,rs\     9r)rDtypingrrr8espnet2.gan_tts.wavenetr&espnet2.gan_tts.wavenet.residual_blockr&espnet.nets.pytorch_backend.nets_utilsrnnModulerr#r#r#r,s