# Copyright (c) 2021, NVIDIA CORPORATION. All rights reserved. # # NVIDIA CORPORATION and its licensors retain all intellectual property # and proprietary rights in and to this software, related documentation # and any modifications thereto. Any use, reproduction, disclosure or # distribution of this software and related documentation without an express # license agreement from NVIDIA CORPORATION is strictly prohibited. import numpy as np import torch from torch_utils import training_stats from torch_utils import misc from torch_utils.ops import conv2d_gradfix #---------------------------------------------------------------------------- class Loss: def accumulate_gradients(self, phase, real_img, real_c, gen_z, gen_c, sync, gain, lambda_sparse): # to be overridden by subclass raise NotImplementedError() #---------------------------------------------------------------------------- class StyleGAN2Loss(Loss): def __init__(self, device, G_mapping, G_synthesis, D, M, augment_pipe=None, style_mixing_prob=0.9, r1_gamma=10, pl_batch_shrink=2, pl_decay=0.01, pl_weight=2): super().__init__() self.device = device self.G_mapping = G_mapping self.G_synthesis = G_synthesis self.D = D self.M = M self.augment_pipe = augment_pipe self.style_mixing_prob = style_mixing_prob self.r1_gamma = r1_gamma self.pl_batch_shrink = pl_batch_shrink self.pl_decay = pl_decay self.pl_weight = pl_weight self.pl_mean = torch.zeros([], device=device) self.K = 4 def run_G(self, z, c, sync, mask_mode, sparse_loss=False, entropy_thr=0.5, temperature=1.0): with misc.ddp_sync(self.G_mapping, sync): mask = self.M(c) if sparse_loss: ws, loss_dict = self.G_mapping(z, mask, mask_mode=mask_mode, sparse_loss=True, entropy_thr=entropy_thr, temperature=temperature) else: ws = self.G_mapping(z, mask, mask_mode=mask_mode,entropy_thr=entropy_thr, temperature=temperature) if self.style_mixing_prob > 0: with torch.autograd.profiler.record_function('style_mixing'): cutoff = torch.empty([], dtype=torch.int64, device=ws.device).random_(1, ws.shape[1]) cutoff = torch.where(torch.rand([], device=ws.device) < self.style_mixing_prob, cutoff, torch.full_like(cutoff, ws.shape[1])) ws[:, cutoff:] = self.G_mapping(torch.randn_like(z), mask, mask_mode=mask_mode, skip_w_avg_update=True, sparse_loss=True, temperature=temperature, entropy_thr=entropy_thr )[0][:, cutoff:] with misc.ddp_sync(self.G_synthesis, sync): img = self.G_synthesis(ws) if sparse_loss: return img, ws, loss_dict else: return img, ws def run_D(self, img, c, sync): if self.augment_pipe is not None: img = self.augment_pipe(img) with misc.ddp_sync(self.D, sync): logits = self.D(img, c) return logits def accumulate_gradients(self, phase, real_img, real_c, gen_z, gen_c, sync, gain, lambda_sparse, lambda_entropy, lambda_ortho, lambda_path, lambda_epsilon, lambda_colvar, lambda_rowvar, lambda_equal, temperature, entropy_thr): assert phase in ['Gmain', 'Greg', 'Gboth', 'Dmain', 'Dreg', 'Dboth'] do_Gmain = (phase in ['Gmain', 'Gboth']) do_Dmain = (phase in ['Dmain', 'Dboth']) do_Gpl = (phase in ['Greg', 'Gboth']) and (self.pl_weight != 0) do_Dr1 = (phase in ['Dreg', 'Dboth']) and (self.r1_gamma != 0) loss_dict = {} self.mask_mode = 'gumbel_hard' # Gmain: Maximize logits for generated images. if do_Gmain: with torch.autograd.profiler.record_function('Gmain_forward'): gen_img, all_gen_ws, loss_dict = self.run_G(gen_z, gen_c, mask_mode=self.mask_mode,sync=(sync and not do_Gpl), sparse_loss=True, temperature=temperature, entropy_thr=entropy_thr) # May get synced by Gpl. gen_logits = self.run_D(gen_img, gen_c, sync=False) training_stats.report('Loss/scores/fake', gen_logits) training_stats.report('Loss/signs/fake', gen_logits.sign()) loss_Gmain = torch.nn.functional.softplus(-gen_logits) # -log(sigmoid(gen_logits)) training_stats.report('Loss/G/sparse', loss_dict['loss_sparse']) loss_Gmain = loss_Gmain + lambda_sparse * loss_dict['loss_sparse'] loss_Gmain = loss_Gmain + lambda_entropy * loss_dict['loss_entropy'] loss_Gmain = loss_Gmain + lambda_ortho * loss_dict['loss_ortho'] loss_Gmain = loss_Gmain + lambda_path * loss_dict['loss_path'] loss_Gmain = loss_Gmain + lambda_epsilon * loss_dict['loss_epsilon'] loss_Gmain = loss_Gmain + lambda_colvar * loss_dict['loss_colvar'] loss_Gmain = loss_Gmain + lambda_rowvar * loss_dict['loss_rowvar'] loss_Gmain = loss_Gmain + lambda_equal * loss_dict['loss_equal'] training_stats.report('Loss/G/loss', loss_Gmain) with torch.autograd.profiler.record_function('Gmain_backward'): loss_Gmain.mean().mul(gain).backward() # Gpl: Apply path length regularization. if do_Gpl: with torch.autograd.profiler.record_function('Gpl_forward'): batch_size = gen_z.shape[0] // self.pl_batch_shrink gen_img, gen_ws, tq_loss_dict = self.run_G(gen_z[:batch_size], gen_c[:batch_size], mask_mode=self.mask_mode,sync=sync, sparse_loss=True, temperature=temperature, ) pl_noise = torch.randn_like(gen_img) / np.sqrt(gen_img.shape[2] * gen_img.shape[3]) with torch.autograd.profiler.record_function('pl_grads'), conv2d_gradfix.no_weight_gradients(): pl_grads = torch.autograd.grad(outputs=[(gen_img * pl_noise).sum()], inputs=[gen_ws], create_graph=True, only_inputs=True)[0] pl_lengths = pl_grads.square().sum(2).mean(1).sqrt() pl_mean = self.pl_mean.lerp(pl_lengths.mean(), self.pl_decay) self.pl_mean.copy_(pl_mean.detach()) pl_penalty = (pl_lengths - pl_mean).square() training_stats.report('Loss/pl_penalty', pl_penalty) loss_Gpl = pl_penalty * self.pl_weight training_stats.report('Loss/G/reg', loss_Gpl) loss_Gpl = loss_Gpl + 0 * tq_loss_dict['loss_equal'] with torch.autograd.profiler.record_function('Gpl_backward'): (gen_img[:, 0, 0, 0] * 0 + loss_Gpl).mean().mul(gain).backward() # Dmain: Minimize logits for generated images. loss_Dgen = 0 if do_Dmain: with torch.autograd.profiler.record_function('Dgen_forward'): gen_img, all_gen_ws, dmain_loss_dict = self.run_G(gen_z, gen_c, mask_mode=self.mask_mode, sync=False, sparse_loss=True, temperature=temperature) gen_logits = self.run_D(gen_img, gen_c, sync=False) # Gets synced by loss_Dreal. training_stats.report('Loss/scores/fake', gen_logits) training_stats.report('Loss/signs/fake', gen_logits.sign()) loss_Dgen = torch.nn.functional.softplus(gen_logits) # -log(1 - sigmoid(gen_logits)) loss_Dgen = loss_Dgen + 0 * dmain_loss_dict['loss_equal'] with torch.autograd.profiler.record_function('Dgen_backward'): loss_Dgen.mean().mul(gain).backward() # Dmain: Maximize logits for real images. # Dr1: Apply R1 regularization. if do_Dmain or do_Dr1: name = 'Dreal_Dr1' if do_Dmain and do_Dr1 else 'Dreal' if do_Dmain else 'Dr1' with torch.autograd.profiler.record_function(name + '_forward'): real_img_tmp = real_img.detach().requires_grad_(do_Dr1) real_logits = self.run_D(real_img_tmp, real_c, sync=sync) training_stats.report('Loss/scores/real', real_logits) training_stats.report('Loss/signs/real', real_logits.sign()) loss_Dreal = 0 if do_Dmain: loss_Dreal = torch.nn.functional.softplus(-real_logits) # -log(sigmoid(real_logits)) training_stats.report('Loss/D/loss', loss_Dgen + loss_Dreal) loss_Dr1 = 0 if do_Dr1: with torch.autograd.profiler.record_function('r1_grads'), conv2d_gradfix.no_weight_gradients(): r1_grads = torch.autograd.grad(outputs=[real_logits.sum()], inputs=[real_img_tmp], create_graph=True, only_inputs=True)[0] r1_penalty = r1_grads.square().sum([1,2,3]) loss_Dr1 = r1_penalty * (self.r1_gamma / 2) training_stats.report('Loss/r1_penalty', r1_penalty) training_stats.report('Loss/D/reg', loss_Dr1) with torch.autograd.profiler.record_function(name + '_backward'): (real_logits * 0 + loss_Dreal + loss_Dr1).mean().mul(gain).backward() return loss_dict #----------------------------------------------------------------------------