unit.py

import argparse
import os
import numpy as np
import math
import datetime
import time
import sys
import jittor.transform as transform
from models import *
from datasets import *
import jittor as jt

jt.flags.use_cuda = 1

parser = argparse.ArgumentParser()
parser.add_argument("--epoch", type=int, default=0, help="epoch to start training from")
parser.add_argument("--n_epochs", type=int, default=200, help="number of epochs of training")
parser.add_argument("--dataset_name", type=str, default="monet2photo", help="name of the dataset")
parser.add_argument("--batch_size", type=int, default=1, help="size of the batches")
parser.add_argument("--lr", type=float, default=0.0001, help="adam: learning rate")
parser.add_argument("--b1", type=float, default=0.5, help="adam: decay of first order momentum of gradient")
parser.add_argument("--b2", type=float, default=0.999, help="adam: decay of first order momentum of gradient")
parser.add_argument("--decay_epoch", type=int, default=100, help="epoch from which to start lr decay")
parser.add_argument("--n_cpu", type=int, default=8, help="number of cpu threads to use during batch generation")
parser.add_argument("--img_height", type=int, default=256, help="size of image height")
parser.add_argument("--img_width", type=int, default=256, help="size of image width")
parser.add_argument("--channels", type=int, default=3, help="number of image channels")
parser.add_argument("--sample_interval", type=int, default=100, help="interval between saving generator samples")
parser.add_argument("--checkpoint_interval", type=int, default=-1, help="interval between saving model checkpoints")
parser.add_argument("--n_downsample", type=int, default=2, help="number downsampling layers in encoder")
parser.add_argument("--dim", type=int, default=64, help="number of filters in first encoder layer")
opt = parser.parse_args()
print(opt)

# Create sample and checkpoint directories
os.makedirs("images/%s" % opt.dataset_name, exist_ok=True)
os.makedirs("saved_models/%s" % opt.dataset_name, exist_ok=True)

# Losses
criterion_GAN = nn.MSELoss()
criterion_pixel = nn.L1Loss()

input_shape = (opt.channels, opt.img_height, opt.img_width)

# Dimensionality (channel-wise) of image embedding
shared_dim = opt.dim * 2 ** opt.n_downsample

# Initialize generator and discriminator
shared_E = ResidualBlock(features=shared_dim)
E1 = Encoder(dim=opt.dim, n_downsample=opt.n_downsample, shared_block=shared_E)
E2 = Encoder(dim=opt.dim, n_downsample=opt.n_downsample, shared_block=shared_E)
shared_G = ResidualBlock(features=shared_dim)
G1 = Generator(dim=opt.dim, n_upsample=opt.n_downsample, shared_block=shared_G)
G2 = Generator(dim=opt.dim, n_upsample=opt.n_downsample, shared_block=shared_G)
D1 = Discriminator(input_shape)
D2 = Discriminator(input_shape)

# Loss weights
lambda_0 = 10  # GAN
lambda_1 = 0.1  # KL (encoded images)
lambda_2 = 100  # ID pixel-wise
lambda_3 = 0.1  # KL (encoded translated images)
lambda_4 = 100  # Cycle pixel-wise

# Optimizers
optimizer_G =  nn.Adam(
    E1.parameters() + E2.parameters() + G1.parameters() + G2.parameters(),
    lr=opt.lr,
    betas=(opt.b1, opt.b2),
)
optimizer_D1 = nn.Adam(D1.parameters(), lr=opt.lr, betas=(opt.b1, opt.b2))
optimizer_D2 = nn.Adam(D2.parameters(), lr=opt.lr, betas=(opt.b1, opt.b2))

# Image transformations
transform_ = [
    transform.Resize(int(opt.img_height * 1.12), Image.BICUBIC),
    transform.RandomCrop((opt.img_height, opt.img_width)),
    transform.RandomHorizontalFlip(),
    transform.ImageNormalize([0.5, 0.5, 0.5], [0.5, 0.5, 0.5]),
]

dataloader = ImageDataset("../../data/%s" % opt.dataset_name, transforms_=transform_, unaligned=True).set_attrs(batch_size=opt.batch_size, shuffle=True, num_workers=opt.n_cpu)

val_dataloader = ImageDataset("../../data/%s" % opt.dataset_name, transforms_=transform_, unaligned=True, mode="test").set_attrs(batch_size=5, shuffle=True, num_workers=1)

import cv2

def save_image(img, path, nrow=10, padding=5):
    N,C,W,H = img.shape
    if (N%nrow!=0):
        print("N%nrow!=0")
        return
    ncol=int(N/nrow)
    img_all = []
    for i in range(ncol):
        img_ = []
        for j in range(nrow):
            img_.append(img[i*nrow+j])
            img_.append(np.zeros((C,W,padding)))
        img_all.append(np.concatenate(img_, 2))
        img_all.append(np.zeros((C,padding,img_all[0].shape[2])))
    img = np.concatenate(img_all, 1)
    img = np.concatenate([np.zeros((C,padding,img.shape[2])), img], 1)
    img = np.concatenate([np.zeros((C,img.shape[1],padding)), img], 2)
    min_=img.min()
    max_=img.max()
    img=(img-min_)/(max_-min_)*255
    img=img.transpose((1,2,0))
    if C==3:
        img = img[:,:,::-1]
    cv2.imwrite(path,img)

def sample_images(batches_done):
    """Saves a generated sample from the test set"""
    imgs = next(iter(val_dataloader))
    X1 = imgs[0].stop_grad()
    X2 = imgs[1].stop_grad()
    E1.eval()
    E2.eval()
    G1.eval()
    G2.eval()
    _, Z1 = E1(X1)
    _, Z2 = E2(X2)
    fake_X1 = G1(Z2)
    fake_X2 = G2(Z1)
    img_sample = jt.contrib.concat((X1, fake_X2, X2, fake_X1), 0)
    save_image(img_sample.numpy(), "images/%s/%s.png" % (opt.dataset_name, batches_done), nrow=5)

def compute_kl(mu):
    mu_2 = mu.sqr()
    loss = jt.mean(mu_2)
    return loss


# ----------
#  Training
# ----------
prev_time = time.time()
for epoch in range(opt.epoch, opt.n_epochs):
    for i, batch in enumerate(dataloader):
        # Set model input
        X1 = batch[0].stop_grad()
        X2 = batch[1].stop_grad()

        # Adversarial ground truths
        valid = jt.ones((X1.size(0), *D1.output_shape)).stop_grad()
        fake = jt.zeros((X1.size(0), *D1.output_shape)).stop_grad()
        # -------------------------------
        #  Train Encoders and Generators
        # -------------------------------

        E1.train()
        E2.train()
        G1.train()
        G2.train()

        # Get shared latent representation
        mu1, Z1 = E1(X1)
        mu2, Z2 = E2(X2)

        # Reconstruct images
        recon_X1 = G1(Z1)
        recon_X2 = G2(Z2)

        # Translate images
        fake_X1 = G1(Z2)
        fake_X2 = G2(Z1)

        # Cycle translation
        mu1_, Z1_ = E1(fake_X1)
        mu2_, Z2_ = E2(fake_X2)
        cycle_X1 = G1(Z2_)
        cycle_X2 = G2(Z1_)

        # Losses
        loss_GAN_1 = lambda_0 * criterion_GAN(D1(fake_X1), valid)
        loss_GAN_2 = lambda_0 * criterion_GAN(D2(fake_X2), valid)
        loss_KL_1 = lambda_1 * compute_kl(mu1)
        loss_KL_2 = lambda_1 * compute_kl(mu2)
        loss_ID_1 = lambda_2 * criterion_pixel(recon_X1, X1)
        loss_ID_2 = lambda_2 * criterion_pixel(recon_X2, X2)
        loss_KL_1_ = lambda_3 * compute_kl(mu1_)
        loss_KL_2_ = lambda_3 * compute_kl(mu2_)
        loss_cyc_1 = lambda_4 * criterion_pixel(cycle_X1, X1)
        loss_cyc_2 = lambda_4 * criterion_pixel(cycle_X2, X2)

        # Total loss
        loss_G = (
            loss_KL_1
            + loss_KL_2
            + loss_ID_1
            + loss_ID_2
            + loss_GAN_1
            + loss_GAN_2
            + loss_KL_1_
            + loss_KL_2_
            + loss_cyc_1
            + loss_cyc_2
        )
        loss_G.sync()
        optimizer_G.step(loss_G)

        # -----------------------
        #  Train Discriminator 1
        # -----------------------


        loss_D1 = criterion_GAN(D1(X1), valid) + criterion_GAN(D1(fake_X1.detach()), fake)

        loss_D1.sync()
        optimizer_D1.step(loss_D1)

        # -----------------------
        #  Train Discriminator 2
        # -----------------------


        loss_D2 = criterion_GAN(D2(X2), valid) + criterion_GAN(D2(fake_X2.detach()), fake)

        loss_D2.sync()
        optimizer_D2.step(loss_D2)

        # --------------
        #  Log Progress
        # --------------

        # Determine approximate time left
        batches_done = epoch * len(dataloader) + i
        batches_left = opt.n_epochs * len(dataloader) - batches_done
        time_left = datetime.timedelta(seconds=batches_left * (time.time() - prev_time))
        prev_time = time.time()

        # Print log
        sys.stdout.write(
            "\r[Epoch %d/%d] [Batch %d/%d] [D loss: %f] [G loss: %f] ETA: %s"
            % (epoch, opt.n_epochs, i, len(dataloader), (loss_D1 + loss_D2).data[0], loss_G.data[0], time_left)
        )

        # If at sample interval save image
        if batches_done % opt.sample_interval == 0:
            sample_images(batches_done)

    if epoch >= opt.decay_epoch:
        optimizer_G.lr = opt.lr * (opt.n_epochs - epoch - 1) / (opt.n_epochs - opt.decay_epoch)
        optimizer_D1.lr = opt.lr * (opt.n_epochs - epoch - 1) / (opt.n_epochs - opt.decay_epoch)
        optimizer_D2.lr = opt.lr * (opt.n_epochs - epoch - 1) / (opt.n_epochs - opt.decay_epoch)