328afde9c0
SPSR_model really isn't that different from SRGAN_model. Rather than continuing to re-implement everything I've done in SRGAN_model, port the new stuff from SPSR over. This really demonstrates the need to refactor SRGAN_model a bit to make it cleaner. It is quite the beast these days..
458 lines
20 KiB
Python
458 lines
20 KiB
Python
import os
|
|
import logging
|
|
from collections import OrderedDict
|
|
|
|
import torch
|
|
import torch.nn as nn
|
|
from torch.optim import lr_scheduler
|
|
from apex import amp
|
|
|
|
import models.networks as networks
|
|
from .base_model import BaseModel
|
|
from models.loss import GANLoss
|
|
import torchvision.utils as utils
|
|
from .archs.SPSR_arch import ImageGradient, ImageGradientNoPadding
|
|
|
|
logger = logging.getLogger('base')
|
|
|
|
class SPSRModel(BaseModel):
|
|
def __init__(self, opt):
|
|
super(SPSRModel, self).__init__(opt)
|
|
train_opt = opt['train']
|
|
|
|
# define networks and load pretrained models
|
|
self.netG = networks.define_G(opt).to(self.device) # G
|
|
if self.is_train:
|
|
self.netD = networks.define_D(opt).to(self.device) # D
|
|
self.netD_grad = networks.define_D(opt).to(self.device) # D_grad
|
|
self.netG.train()
|
|
self.netD.train()
|
|
self.netD_grad.train()
|
|
self.mega_batch_factor = 1
|
|
self.load() # load G and D if needed
|
|
|
|
# define losses, optimizer and scheduler
|
|
if self.is_train:
|
|
self.mega_batch_factor = train_opt['mega_batch_factor']
|
|
|
|
# G pixel loss
|
|
if train_opt['pixel_weight'] > 0:
|
|
l_pix_type = train_opt['pixel_criterion']
|
|
if l_pix_type == 'l1':
|
|
self.cri_pix = nn.L1Loss().to(self.device)
|
|
elif l_pix_type == 'l2':
|
|
self.cri_pix = nn.MSELoss().to(self.device)
|
|
else:
|
|
raise NotImplementedError('Loss type [{:s}] not recognized.'.format(l_pix_type))
|
|
self.l_pix_w = train_opt['pixel_weight']
|
|
else:
|
|
logger.info('Remove pixel loss.')
|
|
self.cri_pix = None
|
|
|
|
# G feature loss
|
|
if train_opt['feature_weight'] > 0:
|
|
l_fea_type = train_opt['feature_criterion']
|
|
if l_fea_type == 'l1':
|
|
self.cri_fea = nn.L1Loss().to(self.device)
|
|
elif l_fea_type == 'l2':
|
|
self.cri_fea = nn.MSELoss().to(self.device)
|
|
else:
|
|
raise NotImplementedError('Loss type [{:s}] not recognized.'.format(l_fea_type))
|
|
self.l_fea_w = train_opt['feature_weight']
|
|
else:
|
|
logger.info('Remove feature loss.')
|
|
self.cri_fea = None
|
|
if self.cri_fea: # load VGG perceptual loss
|
|
self.netF = networks.define_F(opt, use_bn=False).to(self.device)
|
|
|
|
# GD gan loss
|
|
self.cri_gan = GANLoss(train_opt['gan_type'], 1.0, 0.0).to(self.device)
|
|
self.l_gan_w = train_opt['gan_weight']
|
|
# D_update_ratio and D_init_iters are for WGAN
|
|
self.D_update_ratio = train_opt['D_update_ratio'] if train_opt['D_update_ratio'] else 1
|
|
self.D_init_iters = train_opt['D_init_iters'] if train_opt['D_init_iters'] else 0
|
|
# Branch_init_iters
|
|
self.branch_pretrain = train_opt['branch_pretrain'] if train_opt['branch_pretrain'] else 0
|
|
self.branch_init_iters = train_opt['branch_init_iters'] if train_opt['branch_init_iters'] else 1
|
|
|
|
# gradient_pixel_loss
|
|
if train_opt['gradient_pixel_weight'] > 0:
|
|
self.cri_pix_grad = nn.MSELoss().to(self.device)
|
|
self.l_pix_grad_w = train_opt['gradient_pixel_weight']
|
|
else:
|
|
self.cri_pix_grad = None
|
|
|
|
# gradient_gan_loss
|
|
if train_opt['gradient_gan_weight'] > 0:
|
|
self.cri_grad_gan = GANLoss(train_opt['gan_type'], 1.0, 0.0).to(self.device)
|
|
self.l_gan_grad_w = train_opt['gradient_gan_weight']
|
|
else:
|
|
self.cri_grad_gan = None
|
|
|
|
# G_grad pixel loss
|
|
if train_opt['pixel_branch_weight'] > 0:
|
|
l_pix_type = train_opt['pixel_branch_criterion']
|
|
if l_pix_type == 'l1':
|
|
self.cri_pix_branch = nn.L1Loss().to(self.device)
|
|
elif l_pix_type == 'l2':
|
|
self.cri_pix_branch = nn.MSELoss().to(self.device)
|
|
else:
|
|
raise NotImplementedError('Loss type [{:s}] not recognized.'.format(l_pix_type))
|
|
self.l_pix_branch_w = train_opt['pixel_branch_weight']
|
|
else:
|
|
logger.info('Remove G_grad pixel loss.')
|
|
self.cri_pix_branch = None
|
|
|
|
# optimizers
|
|
# G
|
|
wd_G = train_opt['weight_decay_G'] if train_opt['weight_decay_G'] else 0
|
|
|
|
optim_params = []
|
|
for k, v in self.netG.named_parameters(): # optimize part of the model
|
|
|
|
if v.requires_grad:
|
|
optim_params.append(v)
|
|
else:
|
|
logger.warning('Params [{:s}] will not optimize.'.format(k))
|
|
self.optimizer_G = torch.optim.Adam(optim_params, lr=train_opt['lr_G'], \
|
|
weight_decay=wd_G, betas=(train_opt['beta1_G'], 0.999))
|
|
self.optimizers.append(self.optimizer_G)
|
|
|
|
# D
|
|
wd_D = train_opt['weight_decay_D'] if train_opt['weight_decay_D'] else 0
|
|
self.optimizer_D = torch.optim.Adam(self.netD.parameters(), lr=train_opt['lr_D'], \
|
|
weight_decay=wd_D, betas=(train_opt['beta1_D'], 0.999))
|
|
|
|
self.optimizers.append(self.optimizer_D)
|
|
|
|
# D_grad
|
|
wd_D_grad = train_opt['weight_decay_D'] if train_opt['weight_decay_D'] else 0
|
|
self.optimizer_D_grad = torch.optim.Adam(self.netD_grad.parameters(), lr=train_opt['lr_D'], \
|
|
weight_decay=wd_D, betas=(train_opt['beta1_D'], 0.999))
|
|
|
|
self.optimizers.append(self.optimizer_D_grad)
|
|
|
|
# AMP
|
|
[self.netG, self.netD, self.netD_grad], [self.optimizer_G, self.optimizer_D, self.optimizer_D_grad] = \
|
|
amp.initialize([self.netG, self.netD, self.netD_grad],
|
|
[self.optimizer_G, self.optimizer_D, self.optimizer_D_grad],
|
|
opt_level=self.amp_level, num_losses=3)
|
|
|
|
# schedulers
|
|
if train_opt['lr_scheme'] == 'MultiStepLR':
|
|
for optimizer in self.optimizers:
|
|
self.schedulers.append(lr_scheduler.MultiStepLR(optimizer, \
|
|
train_opt['lr_steps'], train_opt['lr_gamma']))
|
|
else:
|
|
raise NotImplementedError('MultiStepLR learning rate scheme is enough.')
|
|
|
|
self.log_dict = OrderedDict()
|
|
self.get_grad = ImageGradient()
|
|
self.get_grad_nopadding = ImageGradientNoPadding()
|
|
|
|
def feed_data(self, data, need_HR=True):
|
|
# LR
|
|
self.var_L = [t.to(self.device) for t in torch.chunk(data['LQ'], chunks=self.mega_batch_factor, dim=0)]
|
|
|
|
if need_HR: # train or val
|
|
self.var_H = [t.to(self.device) for t in torch.chunk(data['GT'], chunks=self.mega_batch_factor, dim=0)]
|
|
input_ref = data['ref'] if 'ref' in data else data['GT']
|
|
self.var_ref = [t.to(self.device) for t in torch.chunk(input_ref.to(self.device), chunks=self.mega_batch_factor, dim=0)]
|
|
|
|
|
|
|
|
def optimize_parameters(self, step):
|
|
# Some generators have variants depending on the current step.
|
|
if hasattr(self.netG.module, "update_for_step"):
|
|
self.netG.module.update_for_step(step, os.path.join(self.opt['path']['models'], ".."))
|
|
if hasattr(self.netD.module, "update_for_step"):
|
|
self.netD.module.update_for_step(step, os.path.join(self.opt['path']['models'], ".."))
|
|
|
|
# G
|
|
for p in self.netD.parameters():
|
|
p.requires_grad = False
|
|
|
|
for p in self.netD_grad.parameters():
|
|
p.requires_grad = False
|
|
|
|
if(self.branch_pretrain):
|
|
if(step < self.branch_init_iters):
|
|
for k,v in self.netG.named_parameters():
|
|
if 'f_' not in k :
|
|
v.requires_grad=False
|
|
else:
|
|
for k,v in self.netG.named_parameters():
|
|
if 'f_' not in k :
|
|
v.requires_grad=True
|
|
|
|
self.optimizer_G.zero_grad()
|
|
|
|
self.fake_H_branch = []
|
|
self.fake_H = []
|
|
self.grad_LR = []
|
|
for var_L, var_H, var_ref in zip(self.var_L, self.var_H, self.var_ref):
|
|
fake_H_branch, fake_H, grad_LR = self.netG(var_L)
|
|
self.fake_H_branch.append(fake_H_branch.detach())
|
|
self.fake_H.append(fake_H.detach())
|
|
self.grad_LR.append(grad_LR.detach())
|
|
|
|
fake_H_grad = self.get_grad(fake_H)
|
|
var_H_grad = self.get_grad(var_H)
|
|
var_ref_grad = self.get_grad(var_ref)
|
|
var_H_grad_nopadding = self.get_grad_nopadding(var_H)
|
|
|
|
l_g_total = 0
|
|
if step % self.D_update_ratio == 0 and step > self.D_init_iters:
|
|
if self.cri_pix: # pixel loss
|
|
l_g_pix = self.l_pix_w * self.cri_pix(fake_H, var_H)
|
|
l_g_total += l_g_pix
|
|
if self.cri_fea: # feature loss
|
|
real_fea = self.netF(var_H).detach()
|
|
fake_fea = self.netF(fake_H)
|
|
l_g_fea = self.l_fea_w * self.cri_fea(fake_fea, real_fea)
|
|
l_g_total += l_g_fea
|
|
|
|
if self.cri_pix_grad: #gradient pixel loss
|
|
l_g_pix_grad = self.l_pix_grad_w * self.cri_pix_grad(fake_H_grad, var_H_grad)
|
|
l_g_total += l_g_pix_grad
|
|
|
|
if self.cri_pix_branch: #branch pixel loss
|
|
l_g_pix_grad_branch = self.l_pix_branch_w * self.cri_pix_branch(fake_H_branch, var_H_grad_nopadding)
|
|
l_g_total += l_g_pix_grad_branch
|
|
|
|
if self.l_gan_w > 0:
|
|
# G gan + cls loss
|
|
pred_g_fake = self.netD(fake_H)
|
|
pred_d_real = self.netD(var_ref).detach()
|
|
|
|
l_g_gan = self.l_gan_w * (self.cri_gan(pred_d_real - torch.mean(pred_g_fake), False) +
|
|
self.cri_gan(pred_g_fake - torch.mean(pred_d_real), True)) / 2
|
|
l_g_total += l_g_gan
|
|
|
|
if self.cri_grad_gan:
|
|
# grad G gan + cls loss
|
|
pred_g_fake_grad = self.netD_grad(fake_H_grad)
|
|
pred_d_real_grad = self.netD_grad(var_ref_grad).detach()
|
|
|
|
l_g_gan_grad = self.l_gan_grad_w * (self.cri_grad_gan(pred_d_real_grad - torch.mean(pred_g_fake_grad), False) +
|
|
self.cri_grad_gan(pred_g_fake_grad - torch.mean(pred_d_real_grad), True)) /2
|
|
l_g_total += l_g_gan_grad
|
|
|
|
l_g_total /= self.mega_batch_factor
|
|
with amp.scale_loss(l_g_total, self.optimizer_G, loss_id=0) as l_g_total_scaled:
|
|
l_g_total_scaled.backward()
|
|
|
|
if step % self.D_update_ratio == 0 and step > self.D_init_iters:
|
|
self.optimizer_G.step()
|
|
|
|
|
|
if self.l_gan_w > 0:
|
|
# D
|
|
for p in self.netD.parameters():
|
|
p.requires_grad = True
|
|
|
|
self.optimizer_D.zero_grad()
|
|
for var_ref, fake_H in zip(self.var_ref, self.fake_H):
|
|
pred_d_real = self.netD(var_ref)
|
|
pred_d_fake = self.netD(fake_H) # detach to avoid BP to G
|
|
l_d_real = self.cri_gan(pred_d_real - torch.mean(pred_d_fake), True)
|
|
l_d_fake = self.cri_gan(pred_d_fake - torch.mean(pred_d_real), False)
|
|
|
|
l_d_total = (l_d_real + l_d_fake) / 2
|
|
|
|
l_d_total /= self.mega_batch_factor
|
|
with amp.scale_loss(l_d_total, self.optimizer_D, loss_id=1) as l_d_total_scaled:
|
|
l_d_total_scaled.backward()
|
|
|
|
self.optimizer_D.step()
|
|
|
|
if self.cri_grad_gan:
|
|
for p in self.netD_grad.parameters():
|
|
p.requires_grad = True
|
|
|
|
self.optimizer_D_grad.zero_grad()
|
|
for var_ref, fake_H in zip(self.var_ref, self.fake_H):
|
|
fake_H_grad = self.get_grad(fake_H)
|
|
var_ref_grad = self.get_grad(var_ref)
|
|
|
|
pred_d_real_grad = self.netD_grad(var_ref_grad)
|
|
pred_d_fake_grad = self.netD_grad(fake_H_grad.detach()) # detach to avoid BP to G
|
|
|
|
l_d_real_grad = self.cri_grad_gan(pred_d_real_grad - torch.mean(pred_d_fake_grad), True)
|
|
l_d_fake_grad = self.cri_grad_gan(pred_d_fake_grad - torch.mean(pred_d_real_grad), False)
|
|
|
|
l_d_total_grad = (l_d_real_grad + l_d_fake_grad) / 2
|
|
l_d_total_grad /= self.mega_batch_factor
|
|
|
|
with amp.scale_loss(l_d_total_grad, self.optimizer_D_grad, loss_id=2) as l_d_total_grad_scaled:
|
|
l_d_total_grad_scaled.backward()
|
|
|
|
self.optimizer_D_grad.step()
|
|
|
|
# Log sample images from first microbatch.
|
|
if step % 50 == 0:
|
|
sample_save_path = os.path.join(self.opt['path']['models'], "..", "temp")
|
|
os.makedirs(os.path.join(sample_save_path, "hr"), exist_ok=True)
|
|
os.makedirs(os.path.join(sample_save_path, "lr"), exist_ok=True)
|
|
os.makedirs(os.path.join(sample_save_path, "gen"), exist_ok=True)
|
|
os.makedirs(os.path.join(sample_save_path, "gen_grad"), exist_ok=True)
|
|
# fed_LQ is not chunked.
|
|
utils.save_image(self.var_H[0].cpu(), os.path.join(sample_save_path, "hr", "%05i.png" % (step,)))
|
|
utils.save_image(self.var_L[0].cpu(), os.path.join(sample_save_path, "lr", "%05i.png" % (step,)))
|
|
utils.save_image(self.fake_H[0].cpu(), os.path.join(sample_save_path, "gen", "%05i.png" % (step,)))
|
|
utils.save_image(self.grad_LR[0].cpu(), os.path.join(sample_save_path, "gen_grad", "%05i.png" % (step,)))
|
|
|
|
|
|
# set log
|
|
if step % self.D_update_ratio == 0 and step > self.D_init_iters:
|
|
# G
|
|
if self.cri_pix:
|
|
self.add_log_entry('l_g_pix', l_g_pix.item())
|
|
if self.cri_fea:
|
|
self.add_log_entry('l_g_fea', l_g_fea.item())
|
|
if self.l_gan_w > 0:
|
|
self.add_log_entry('l_g_gan', l_g_gan.item())
|
|
|
|
if self.cri_pix_branch: #branch pixel loss
|
|
self.add_log_entry('l_g_pix_grad_branch', l_g_pix_grad_branch.item())
|
|
|
|
if self.l_gan_w > 0:
|
|
self.add_log_entry('l_d_real', l_d_real.item())
|
|
self.add_log_entry('l_d_fake', l_d_fake.item())
|
|
self.add_log_entry('l_d_real_grad', l_d_real_grad.item())
|
|
self.add_log_entry('l_d_fake_grad', l_d_fake_grad.item())
|
|
self.add_log_entry('D_real', torch.mean(pred_d_real.detach()))
|
|
self.add_log_entry('D_fake', torch.mean(pred_d_fake.detach()))
|
|
self.add_log_entry('D_real_grad', torch.mean(pred_d_real_grad.detach()))
|
|
self.add_log_entry('D_fake_grad', torch.mean(pred_d_fake_grad.detach()))
|
|
|
|
# Allows the log to serve as an easy-to-use rotating buffer.
|
|
def add_log_entry(self, key, value):
|
|
key_it = "%s_it" % (key,)
|
|
log_rotating_buffer_size = 50
|
|
if key not in self.log_dict.keys():
|
|
self.log_dict[key] = []
|
|
self.log_dict[key_it] = 0
|
|
if len(self.log_dict[key]) < log_rotating_buffer_size:
|
|
self.log_dict[key].append(value)
|
|
else:
|
|
self.log_dict[key][self.log_dict[key_it] % log_rotating_buffer_size] = value
|
|
self.log_dict[key_it] += 1
|
|
|
|
def test(self):
|
|
self.netG.eval()
|
|
with torch.no_grad():
|
|
self.fake_H_branch = []
|
|
self.fake_H = []
|
|
self.grad_LR = []
|
|
for var_L in self.var_L:
|
|
fake_H_branch, fake_H, grad_LR = self.netG(var_L)
|
|
self.fake_H_branch.append(fake_H_branch)
|
|
self.fake_H.append(fake_H)
|
|
self.grad_LR.append(grad_LR)
|
|
|
|
self.netG.train()
|
|
|
|
# Fetches a summary of the log.
|
|
def get_current_log(self, step):
|
|
return_log = {}
|
|
for k in self.log_dict.keys():
|
|
if not isinstance(self.log_dict[k], list):
|
|
continue
|
|
return_log[k] = sum(self.log_dict[k]) / len(self.log_dict[k])
|
|
|
|
# Some generators can do their own metric logging.
|
|
if hasattr(self.netG.module, "get_debug_values"):
|
|
return_log.update(self.netG.module.get_debug_values(step))
|
|
if hasattr(self.netD.module, "get_debug_values"):
|
|
return_log.update(self.netD.module.get_debug_values(step))
|
|
|
|
return return_log
|
|
|
|
def get_current_visuals(self, need_HR=True):
|
|
out_dict = OrderedDict()
|
|
out_dict['LR'] = self.var_L[0].float().cpu()
|
|
|
|
out_dict['rlt'] = self.fake_H[0].float().cpu()
|
|
out_dict['SR_branch'] = self.fake_H_branch[0].float().cpu()
|
|
out_dict['LR_grad'] = self.grad_LR[0].float().cpu()
|
|
if need_HR:
|
|
out_dict['GT'] = self.var_H[0].float().cpu()
|
|
return out_dict
|
|
|
|
def print_network(self):
|
|
# Generator
|
|
s, n = self.get_network_description(self.netG)
|
|
if isinstance(self.netG, nn.DataParallel):
|
|
net_struc_str = '{} - {}'.format(self.netG.__class__.__name__,
|
|
self.netG.module.__class__.__name__)
|
|
else:
|
|
net_struc_str = '{}'.format(self.netG.__class__.__name__)
|
|
|
|
logger.info('Network G structure: {}, with parameters: {:,d}'.format(net_struc_str, n))
|
|
logger.info(s)
|
|
if self.is_train:
|
|
# Disriminator
|
|
s, n = self.get_network_description(self.netD)
|
|
if isinstance(self.netD, nn.DataParallel):
|
|
net_struc_str = '{} - {}'.format(self.netD.__class__.__name__,
|
|
self.netD.module.__class__.__name__)
|
|
else:
|
|
net_struc_str = '{}'.format(self.netD.__class__.__name__)
|
|
|
|
logger.info('Network D structure: {}, with parameters: {:,d}'.format(net_struc_str, n))
|
|
logger.info(s)
|
|
|
|
if self.cri_fea: # F, Perceptual Network
|
|
s, n = self.get_network_description(self.netF)
|
|
if isinstance(self.netF, nn.DataParallel):
|
|
net_struc_str = '{} - {}'.format(self.netF.__class__.__name__,
|
|
self.netF.module.__class__.__name__)
|
|
else:
|
|
net_struc_str = '{}'.format(self.netF.__class__.__name__)
|
|
|
|
logger.info('Network F structure: {}, with parameters: {:,d}'.format(net_struc_str, n))
|
|
logger.info(s)
|
|
|
|
def load(self):
|
|
load_path_G = self.opt['path']['pretrain_model_G']
|
|
if load_path_G is not None:
|
|
logger.info('Loading pretrained model for G [{:s}] ...'.format(load_path_G))
|
|
self.load_network(load_path_G, self.netG)
|
|
load_path_D = self.opt['path']['pretrain_model_D']
|
|
if self.opt['is_train'] and load_path_D is not None:
|
|
logger.info('Loading pretrained model for D [{:s}] ...'.format(load_path_D))
|
|
self.load_network(load_path_D, self.netD)
|
|
load_path_D_grad = self.opt['path']['pretrain_model_D_grad']
|
|
if self.opt['is_train'] and load_path_D_grad is not None:
|
|
logger.info('Loading pretrained model for D_grad [{:s}] ...'.format(load_path_D_grad))
|
|
self.load_network(load_path_D_grad, self.netD_grad)
|
|
|
|
def compute_fea_loss(self, real, fake):
|
|
if self.cri_fea is None:
|
|
return 0
|
|
with torch.no_grad():
|
|
real = real.unsqueeze(dim=0).to(self.device)
|
|
fake = fake.unsqueeze(dim=0).to(self.device)
|
|
real_fea = self.netF(real).detach()
|
|
fake_fea = self.netF(fake)
|
|
return self.cri_fea(fake_fea, real_fea).item()
|
|
|
|
def force_restore_swapout(self):
|
|
pass
|
|
|
|
def save(self, iter_step):
|
|
self.save_network(self.netG, 'G', iter_step)
|
|
self.save_network(self.netD, 'D', iter_step)
|
|
self.save_network(self.netD_grad, 'D_grad', iter_step)
|
|
|
|
# override of load_network that allows loading partial params (like RRDB_PSNR_x4)
|
|
def load_network(self, load_path, network, strict=True):
|
|
if isinstance(network, nn.DataParallel):
|
|
network = network.module
|
|
pretrained_dict = torch.load(load_path)
|
|
model_dict = network.state_dict()
|
|
pretrained_dict = {k: v for k, v in pretrained_dict.items() if k in model_dict}
|
|
|
|
model_dict.update(pretrained_dict)
|
|
network.load_state_dict(model_dict) |