tortoise/models/diffusion_decoder.py

import math
import random
from abc import abstractmethod

import torch
import torch.nn as nn
import torch.nn.functional as F
from torch import autocast

from tortoise.models.arch_util import AttentionBlock, normalization


def is_latent(t):
    return t.dtype == torch.float


def is_sequence(t):
    return t.dtype == torch.long


def timestep_embedding(timesteps, dim, max_period=10000):
    """
    Create sinusoidal timestep embeddings.

    :param timesteps: a 1-D Tensor of N indices, one per batch element.
                      These may be fractional.
    :param dim: the dimension of the output.
    :param max_period: controls the minimum frequency of the embeddings.
    :return: an [N x dim] Tensor of positional embeddings.
    """
    half = dim // 2
    freqs = torch.exp(
        -math.log(max_period)
        * torch.arange(start=0, end=half, dtype=torch.float32)
        / half
    ).to(device=timesteps.device)
    args = timesteps[:, None].float() * freqs[None]
    embedding = torch.cat([torch.cos(args), torch.sin(args)], dim=-1)
    if dim % 2:
        embedding = torch.cat([embedding, torch.zeros_like(embedding[:, :1])], dim=-1)
    return embedding


class TimestepBlock(nn.Module):
    @abstractmethod
    def forward(self, x, emb):
        """
        Apply the module to `x` given `emb` timestep embeddings.
        """


class TimestepEmbedSequential(nn.Sequential, TimestepBlock):
    def forward(self, x, emb):
        for layer in self:
            if isinstance(layer, TimestepBlock):
                x = layer(x, emb)
            else:
                x = layer(x)
        return x


class ResBlock(TimestepBlock):
    def __init__(
        self,
        channels,
        emb_channels,
        dropout,
        out_channels=None,
        dims=2,
        kernel_size=3,
        efficient_config=True,
        use_scale_shift_norm=False,
    ):
        super().__init__()
        self.channels = channels
        self.emb_channels = emb_channels
        self.dropout = dropout
        self.out_channels = out_channels or channels
        self.use_scale_shift_norm = use_scale_shift_norm
        padding = {1: 0, 3: 1, 5: 2}[kernel_size]
        eff_kernel = 1 if efficient_config else 3
        eff_padding = 0 if efficient_config else 1

        self.in_layers = nn.Sequential(
            normalization(channels),
            nn.SiLU(),
            nn.Conv1d(channels, self.out_channels, eff_kernel, padding=eff_padding),
        )

        self.emb_layers = nn.Sequential(
            nn.SiLU(),
            nn.Linear(
                emb_channels,
                2 * self.out_channels if use_scale_shift_norm else self.out_channels,
            ),
        )
        self.out_layers = nn.Sequential(
            normalization(self.out_channels),
            nn.SiLU(),
            nn.Dropout(p=dropout),
            nn.Conv1d(
                self.out_channels, self.out_channels, kernel_size, padding=padding
            ),
        )

        if self.out_channels == channels:
            self.skip_connection = nn.Identity()
        else:
            self.skip_connection = nn.Conv1d(
                channels, self.out_channels, eff_kernel, padding=eff_padding
            )

    def forward(self, x, emb):
        h = self.in_layers(x)
        emb_out = self.emb_layers(emb).type(h.dtype)
        while len(emb_out.shape) < len(h.shape):
            emb_out = emb_out[..., None]
        if self.use_scale_shift_norm:
            out_norm, out_rest = self.out_layers[0], self.out_layers[1:]
            scale, shift = torch.chunk(emb_out, 2, dim=1)
            h = out_norm(h) * (1 + scale) + shift
            h = out_rest(h)
        else:
            h = h + emb_out
            h = self.out_layers(h)
        return self.skip_connection(x) + h


class DiffusionLayer(TimestepBlock):
    def __init__(self, model_channels, dropout, num_heads):
        super().__init__()
        self.resblk = ResBlock(
            model_channels,
            model_channels,
            dropout,
            model_channels,
            dims=1,
            use_scale_shift_norm=True,
        )
        self.attn = AttentionBlock(
            model_channels, num_heads, relative_pos_embeddings=True
        )

    def forward(self, x, time_emb):
        y = self.resblk(x, time_emb)
        return self.attn(y)


class DiffusionTts(nn.Module):
    def __init__(
        self,
        model_channels=512,
        num_layers=8,
        in_channels=100,
        in_latent_channels=512,
        in_tokens=8193,
        out_channels=200,  # mean and variance
        dropout=0,
        use_fp16=False,
        num_heads=16,
        # Parameters for regularization.
        layer_drop=0.1,
        unconditioned_percentage=0.1,  # This implements a mechanism similar to what is used in classifier-free training.
    ):
        super().__init__()

        self.in_channels = in_channels
        self.model_channels = model_channels
        self.out_channels = out_channels
        self.dropout = dropout
        self.num_heads = num_heads
        self.unconditioned_percentage = unconditioned_percentage
        self.enable_fp16 = use_fp16
        self.layer_drop = layer_drop

        self.inp_block = nn.Conv1d(in_channels, model_channels, 3, 1, 1)
        self.time_embed = nn.Sequential(
            nn.Linear(model_channels, model_channels),
            nn.SiLU(),
            nn.Linear(model_channels, model_channels),
        )

        # Either code_converter or latent_converter is used, depending on what type of conditioning data is fed.
        # This model is meant to be able to be trained on both for efficiency purposes - it is far less computationally
        # complex to generate tokens, while generating latents will normally mean propagating through a deep autoregressive
        # transformer network.
        self.code_embedding = nn.Embedding(in_tokens, model_channels)
        self.code_converter = nn.Sequential(
            AttentionBlock(model_channels, num_heads, relative_pos_embeddings=True),
            AttentionBlock(model_channels, num_heads, relative_pos_embeddings=True),
            AttentionBlock(model_channels, num_heads, relative_pos_embeddings=True),
        )
        self.code_norm = normalization(model_channels)
        self.latent_conditioner = nn.Sequential(
            nn.Conv1d(in_latent_channels, model_channels, 3, padding=1),
            AttentionBlock(model_channels, num_heads, relative_pos_embeddings=True),
            AttentionBlock(model_channels, num_heads, relative_pos_embeddings=True),
            AttentionBlock(model_channels, num_heads, relative_pos_embeddings=True),
            AttentionBlock(model_channels, num_heads, relative_pos_embeddings=True),
        )
        self.contextual_embedder = nn.Sequential(
            nn.Conv1d(in_channels, model_channels, 3, padding=1, stride=2),
            nn.Conv1d(model_channels, model_channels * 2, 3, padding=1, stride=2),
            AttentionBlock(
                model_channels * 2,
                num_heads,
                relative_pos_embeddings=True,
                do_checkpoint=False,
            ),
            AttentionBlock(
                model_channels * 2,
                num_heads,
                relative_pos_embeddings=True,
                do_checkpoint=False,
            ),
            AttentionBlock(
                model_channels * 2,
                num_heads,
                relative_pos_embeddings=True,
                do_checkpoint=False,
            ),
            AttentionBlock(
                model_channels * 2,
                num_heads,
                relative_pos_embeddings=True,
                do_checkpoint=False,
            ),
            AttentionBlock(
                model_channels * 2,
                num_heads,
                relative_pos_embeddings=True,
                do_checkpoint=False,
            ),
        )
        self.unconditioned_embedding = nn.Parameter(torch.randn(1, model_channels, 1))
        self.conditioning_timestep_integrator = TimestepEmbedSequential(
            DiffusionLayer(model_channels, dropout, num_heads),
            DiffusionLayer(model_channels, dropout, num_heads),
            DiffusionLayer(model_channels, dropout, num_heads),
        )

        self.integrating_conv = nn.Conv1d(
            model_channels * 2, model_channels, kernel_size=1
        )
        self.mel_head = nn.Conv1d(model_channels, in_channels, kernel_size=3, padding=1)

        self.layers = nn.ModuleList(
            [
                DiffusionLayer(model_channels, dropout, num_heads)
                for _ in range(num_layers)
            ]
            + [
                ResBlock(
                    model_channels,
                    model_channels,
                    dropout,
                    dims=1,
                    use_scale_shift_norm=True,
                )
                for _ in range(3)
            ]
        )

        self.out = nn.Sequential(
            normalization(model_channels),
            nn.SiLU(),
            nn.Conv1d(model_channels, out_channels, 3, padding=1),
        )

    def get_grad_norm_parameter_groups(self):
        groups = {
            "minicoder": list(self.contextual_embedder.parameters()),
            "layers": list(self.layers.parameters()),
            "code_converters": list(self.code_embedding.parameters())
            + list(self.code_converter.parameters())
            + list(self.latent_conditioner.parameters())
            + list(self.latent_conditioner.parameters()),
            "timestep_integrator": list(
                self.conditioning_timestep_integrator.parameters()
            )
            + list(self.integrating_conv.parameters()),
            "time_embed": list(self.time_embed.parameters()),
        }
        return groups

    def get_conditioning(self, conditioning_input):
        speech_conditioning_input = (
            conditioning_input.unsqueeze(1)
            if len(conditioning_input.shape) == 3
            else conditioning_input
        )
        conds = []
        for j in range(speech_conditioning_input.shape[1]):
            conds.append(self.contextual_embedder(speech_conditioning_input[:, j]))
        conds = torch.cat(conds, dim=-1)
        conds = conds.mean(dim=-1)
        return conds

    def timestep_independent(
        self,
        aligned_conditioning,
        conditioning_latent,
        expected_seq_len,
        return_code_pred,
    ):
        # Shuffle aligned_latent to BxCxS format
        if is_latent(aligned_conditioning):
            aligned_conditioning = aligned_conditioning.permute(0, 2, 1)

        cond_scale, cond_shift = torch.chunk(conditioning_latent, 2, dim=1)
        if is_latent(aligned_conditioning):
            code_emb = self.latent_conditioner(aligned_conditioning)
        else:
            code_emb = self.code_embedding(aligned_conditioning).permute(0, 2, 1)
            code_emb = self.code_converter(code_emb)
        code_emb = self.code_norm(code_emb) * (
            1 + cond_scale.unsqueeze(-1)
        ) + cond_shift.unsqueeze(-1)

        unconditioned_batches = torch.zeros(
            (code_emb.shape[0], 1, 1), device=code_emb.device
        )
        # Mask out the conditioning branch for whole batch elements, implementing something similar to classifier-free guidance.
        if self.training and self.unconditioned_percentage > 0:
            unconditioned_batches = (
                torch.rand((code_emb.shape[0], 1, 1), device=code_emb.device)
                < self.unconditioned_percentage
            )
            code_emb = torch.where(
                unconditioned_batches,
                self.unconditioned_embedding.repeat(
                    aligned_conditioning.shape[0], 1, 1
                ),
                code_emb,
            )
        expanded_code_emb = F.interpolate(
            code_emb, size=expected_seq_len, mode="nearest"
        )

        if not return_code_pred:
            return expanded_code_emb
        else:
            mel_pred = self.mel_head(expanded_code_emb)
            # Multiply mel_pred by !unconditioned_branches, which drops the gradient on unconditioned branches. This is because we don't want that gradient being used to train parameters through the codes_embedder as it unbalances contributions to that network from the MSE loss.
            mel_pred = mel_pred * unconditioned_batches.logical_not()
            return expanded_code_emb, mel_pred

    def forward(
        self,
        x,
        timesteps,
        aligned_conditioning=None,
        conditioning_latent=None,
        precomputed_aligned_embeddings=None,
        conditioning_free=False,
        return_code_pred=False,
    ):
        """
        Apply the model to an input batch.

        :param x: an [N x C x ...] Tensor of inputs.
        :param timesteps: a 1-D batch of timesteps.
        :param aligned_conditioning: an aligned latent or sequence of tokens providing useful data about the sample to be produced.
        :param conditioning_latent: a pre-computed conditioning latent; see get_conditioning().
        :param precomputed_aligned_embeddings: Embeddings returned from self.timestep_independent()
        :param conditioning_free: When set, all conditioning inputs (including tokens and conditioning_input) will not be considered.
        :return: an [N x C x ...] Tensor of outputs.
        """
        assert precomputed_aligned_embeddings is not None or (
            aligned_conditioning is not None and conditioning_latent is not None
        )
        assert not (
            return_code_pred and precomputed_aligned_embeddings is not None
        )  # These two are mutually exclusive.

        unused_params = []
        if conditioning_free:
            code_emb = self.unconditioned_embedding.repeat(x.shape[0], 1, x.shape[-1])
            unused_params.extend(
                list(self.code_converter.parameters())
                + list(self.code_embedding.parameters())
            )
            unused_params.extend(list(self.latent_conditioner.parameters()))
        else:
            if precomputed_aligned_embeddings is not None:
                code_emb = precomputed_aligned_embeddings
            else:
                code_emb, mel_pred = self.timestep_independent(
                    aligned_conditioning, conditioning_latent, x.shape[-1], True
                )
                if is_latent(aligned_conditioning):
                    unused_params.extend(
                        list(self.code_converter.parameters())
                        + list(self.code_embedding.parameters())
                    )
                else:
                    unused_params.extend(list(self.latent_conditioner.parameters()))

            unused_params.append(self.unconditioned_embedding)

        time_emb = self.time_embed(timestep_embedding(timesteps, self.model_channels))
        code_emb = self.conditioning_timestep_integrator(code_emb, time_emb)
        x = self.inp_block(x)
        x = torch.cat([x, code_emb], dim=1)
        x = self.integrating_conv(x)
        for i, lyr in enumerate(self.layers):
            # Do layer drop where applicable. Do not drop first and last layers.
            if (
                self.training
                and self.layer_drop > 0
                and i != 0
                and i != (len(self.layers) - 1)
                and random.random() < self.layer_drop
            ):
                unused_params.extend(list(lyr.parameters()))
            else:
                # First and last blocks will have autocast disabled for improved precision.
                with autocast(x.device.type, enabled=self.enable_fp16 and i != 0):
                    x = lyr(x, time_emb)

        x = x.float()
        out = self.out(x)

        # Involve probabilistic or possibly unused parameters in loss so we don't get DDP errors.
        extraneous_addition = 0
        for p in unused_params:
            extraneous_addition = extraneous_addition + p.mean()
        out = out + extraneous_addition * 0

        if return_code_pred:
            return out, mel_pred
        return out


if __name__ == "__main__":
    clip = torch.randn(2, 100, 400)
    aligned_latent = torch.randn(2, 388, 512)
    aligned_sequence = torch.randint(0, 8192, (2, 100))
    cond = torch.randn(2, 100, 400)
    ts = torch.LongTensor([600, 600])
    model = DiffusionTts(512, layer_drop=0.3, unconditioned_percentage=0.5)
    # Test with latent aligned conditioning
    # o = model(clip, ts, aligned_latent, cond)
    # Test with sequence aligned conditioning
    o = model(clip, ts, aligned_sequence, cond)