src/models/autoencoder.py

import torch
from einops import rearrange

import torch.nn.functional as F
import pytorch_lightning as pl

from src.modules.ae_modules import Encoder, Decoder
from src.distributions import DiagonalGaussianDistribution
from utils.common_utils import instantiate_from_config


class AutoencoderKL(pl.LightningModule):
    def __init__(
        self,
        ddconfig,
        lossconfig,
        embed_dim,
        use_quant_conv=True,
        ckpt_path=None,
        ignore_keys=[],
        image_key="image",
        colorize_nlabels=None,
        monitor=None,
        test=False,
        logdir=None,
        input_dim=4,
        test_args=None,
    ):
        super().__init__()
        self.image_key = image_key
        self.encoder = Encoder(**ddconfig)
        self.decoder = Decoder(**ddconfig)
        self.loss = instantiate_from_config(lossconfig)
        assert ddconfig["double_z"]

        if use_quant_conv:
            self.quant_conv = torch.nn.Conv2d(
                2 * ddconfig["z_channels"], 2 * embed_dim, 1
            )
            self.post_quant_conv = torch.nn.Conv2d(embed_dim, ddconfig["z_channels"], 1)
            self.embed_dim = embed_dim

        self.use_quant_conv = use_quant_conv

        self.input_dim = input_dim
        self.test = test
        self.test_args = test_args
        self.logdir = logdir
        if colorize_nlabels is not None:
            assert type(colorize_nlabels) == int
            self.register_buffer("colorize", torch.randn(3, colorize_nlabels, 1, 1))
        if monitor is not None:
            self.monitor = monitor
        if ckpt_path is not None:
            self.init_from_ckpt(ckpt_path, ignore_keys=ignore_keys)

    def init_from_ckpt(self, path, ignore_keys=list()):
        sd = torch.load(path, map_location="cpu")
        try:
            self._cur_epoch = sd["epoch"]
            sd = sd["state_dict"]
        except:
            self._cur_epoch = "null"
        keys = list(sd.keys())
        for k in keys:
            for ik in ignore_keys:
                if k.startswith(ik):
                    # print("Deleting key {} from state_dict.".format(k))
                    del sd[k]
        self.load_state_dict(sd, strict=False)
        # self.load_state_dict(sd, strict=True)
        print(f"Restored from {path}")

    def encode(self, x, **kwargs):

        h = self.encoder(x)
        moments = h
        if self.use_quant_conv:
            moments = self.quant_conv(h)
        posterior = DiagonalGaussianDistribution(moments)
        return posterior

    def decode(self, z, **kwargs):
        if self.use_quant_conv:
            z = self.post_quant_conv(z)
        dec = self.decoder(z)
        return dec

    def forward(self, input, sample_posterior=True):
        posterior = self.encode(input)
        if sample_posterior:
            z = posterior.sample()
        else:
            z = posterior.mode()
        dec = self.decode(z)
        return dec, posterior

    def get_input(self, batch, k):
        x = batch[k]
        if x.dim() == 5 and self.input_dim == 4:
            b, c, t, h, w = x.shape
            self.b = b
            self.t = t
            x = rearrange(x, "b c t h w -> (b t) c h w")

        return x

    def training_step(self, batch, batch_idx, optimizer_idx):
        inputs = self.get_input(batch, self.image_key)
        reconstructions, posterior = self(inputs)

        if optimizer_idx == 0:
            # train encoder+decoder+logvar
            aeloss, log_dict_ae = self.loss(
                inputs,
                reconstructions,
                posterior,
                optimizer_idx,
                self.global_step,
                last_layer=self.get_last_layer(),
                split="train",
            )
            self.log(
                "aeloss",
                aeloss,
                prog_bar=True,
                logger=True,
                on_step=True,
                on_epoch=True,
            )
            self.log_dict(
                log_dict_ae, prog_bar=False, logger=True, on_step=True, on_epoch=False
            )
            return aeloss

        if optimizer_idx == 1:
            # train the discriminator
            discloss, log_dict_disc = self.loss(
                inputs,
                reconstructions,
                posterior,
                optimizer_idx,
                self.global_step,
                last_layer=self.get_last_layer(),
                split="train",
            )

            self.log(
                "discloss",
                discloss,
                prog_bar=True,
                logger=True,
                on_step=True,
                on_epoch=True,
            )
            self.log_dict(
                log_dict_disc, prog_bar=False, logger=True, on_step=True, on_epoch=False
            )
            return discloss

    def validation_step(self, batch, batch_idx):
        inputs = self.get_input(batch, self.image_key)
        reconstructions, posterior = self(inputs)
        aeloss, log_dict_ae = self.loss(
            inputs,
            reconstructions,
            posterior,
            0,
            self.global_step,
            last_layer=self.get_last_layer(),
            split="val",
        )

        discloss, log_dict_disc = self.loss(
            inputs,
            reconstructions,
            posterior,
            1,
            self.global_step,
            last_layer=self.get_last_layer(),
            split="val",
        )

        recontructions = reconstructions.cpu().detach()

        self.log("val/rec_loss", log_dict_ae["val/rec_loss"])
        self.log_dict(log_dict_ae)
        self.log_dict(log_dict_disc)
        return self.log_dict

    def configure_optimizers(self):
        lr = self.learning_rate
        opt_ae = torch.optim.Adam(
            list(self.encoder.parameters())
            + list(self.decoder.parameters())
            + list(self.quant_conv.parameters())
            + list(self.post_quant_conv.parameters()),
            lr=lr,
            betas=(0.5, 0.9),
        )
        opt_disc = torch.optim.Adam(
            self.loss.discriminator.parameters(), lr=lr, betas=(0.5, 0.9)
        )
        return [opt_ae, opt_disc], []

    def get_last_layer(self):
        return self.decoder.conv_out.weight

    @torch.no_grad()
    def log_images(self, batch, only_inputs=False, **kwargs):
        log = dict()
        x = self.get_input(batch, self.image_key)
        x = x.to(self.device)
        if not only_inputs:
            xrec, posterior = self(x)
            if x.shape[1] > 3:
                # colorize with random projection
                assert xrec.shape[1] > 3
                x = self.to_rgb(x)
                xrec = self.to_rgb(xrec)

            log["samples"] = self.decode(torch.randn_like(posterior.sample()))
            xrec = xrec.cpu().detach()
            log["reconstructions"] = xrec

        x = x.cpu().detach()
        log["inputs"] = x
        return log

    def to_rgb(self, x):
        assert self.image_key == "segmentation"
        if not hasattr(self, "colorize"):
            self.register_buffer("colorize", torch.randn(3, x.shape[1], 1, 1).to(x))
        x = F.conv2d(x, weight=self.colorize)
        x = 2.0 * (x - x.min()) / (x.max() - x.min()) - 1.0
        return x


class IdentityFirstStage(torch.nn.Module):
    def __init__(self, *args, vq_interface=False, **kwargs):
        # TODO: Should be true by default but check to not break older stuff
        self.vq_interface = vq_interface
        super().__init__()

    def encode(self, x, *args, **kwargs):
        return x

    def decode(self, x, *args, **kwargs):
        return x

    def quantize(self, x, *args, **kwargs):
        if self.vq_interface:
            return x, None, [None, None, None]
        return x

    def forward(self, x, *args, **kwargs):
        return x