search

SenHe / Flow-Style-VTON

Other
282 stars 45 forks source link

Very off result despite same packages version #48

Closed kayabutterkun closed 8 months ago

kayabutterkun commented 9 months ago

Hello! I am using the same package version but my results are very off. What am I missing?

Python 3.6.13 PyTorch version: 1.1.0 Torchvision version: 0.3.0 CV2 version: 3.4.3

image

Screenshot 2024-01-26 at 3 44 00 PM

Screenshot 2024-01-26 at 3 44 04 PM

Screenshot 2024-01-26 at 3 44 13 PM

Screenshot 2024-01-26 at 3 44 10 PM

kayabutterkun commented 9 months ago

Logs:

------------ Options ------------- batchSize: 1 data_type: 32 dataroot: /datasets/john/tryOn/Tests display_winsize: 512 fineSize: 512 gen_checkpoint: /datasets/john/tryOn/Flow-Style-VTON-Checkpoints/PFAFN_gen_epoch_101.pth gpu_ids: [0] input_nc: 3 isTrain: False loadSize: 512 max_dataset_size: inf nThreads: 1 name: demo no_flip: False norm: instance output: /work/output/2024-01-26-1434 output_nc: 3 phase: test resize_or_crop: None serial_batches: False test_pair: /datasets/john/tryOn/Tests/test_pairs.txt tf_log: False use_dropout: False verbose: False warp_checkpoint: /datasets/john/tryOn/Flow-Style-VTON-Checkpoints/PFAFN_warp_epoch_101.pth -------------- End ---------------- CustomDatasetDataLoader dataset [AlignedDataset] was created 6 AFWM( (image_features): FeatureEncoder( (encoders): ModuleList( (0): Sequential( (0): DownSample( (block): Sequential( (0): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (1): ReLU(inplace) (2): Conv2d(3, 64, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False) ) ) (1): ResBlock( (block): Sequential( (0): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (1): ReLU(inplace) (2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (3): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (4): ReLU(inplace) (5): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) ) ) (2): ResBlock( (block): Sequential( (0): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (1): ReLU(inplace) (2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (3): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (4): ReLU(inplace) (5): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) ) ) ) (1): Sequential( (0): DownSample( (block): Sequential( (0): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (1): ReLU(inplace) (2): Conv2d(64, 128, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False) ) ) (1): ResBlock( (block): Sequential( (0): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (1): ReLU(inplace) (2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (3): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (4): ReLU(inplace) (5): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) ) ) (2): ResBlock( (block): Sequential( (0): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (1): ReLU(inplace) (2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (3): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (4): ReLU(inplace) (5): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) ) ) ) (2): Sequential( (0): DownSample( (block): Sequential( (0): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (1): ReLU(inplace) (2): Conv2d(128, 256, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False) ) ) (1): ResBlock( (block): Sequential( (0): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (1): ReLU(inplace) (2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (4): ReLU(inplace) (5): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) ) ) (2): ResBlock( (block): Sequential( (0): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (1): ReLU(inplace) (2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (4): ReLU(inplace) (5): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) ) ) ) (3): Sequential( (0): DownSample( (block): Sequential( (0): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (1): ReLU(inplace) (2): Conv2d(256, 256, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False) ) ) (1): ResBlock( (block): Sequential( (0): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (1): ReLU(inplace) (2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (4): ReLU(inplace) (5): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) ) ) (2): ResBlock( (block): Sequential( (0): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (1): ReLU(inplace) (2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (4): ReLU(inplace) (5): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) ) ) ) (4): Sequential( (0): DownSample( (block): Sequential( (0): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (1): ReLU(inplace) (2): Conv2d(256, 256, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False) ) ) (1): ResBlock( (block): Sequential( (0): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (1): ReLU(inplace) (2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (4): ReLU(inplace) (5): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) ) ) (2): ResBlock( (block): Sequential( (0): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (1): ReLU(inplace) (2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (4): ReLU(inplace) (5): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) ) ) ) ) ) (cond_features): FeatureEncoder( (encoders): ModuleList( (0): Sequential( (0): DownSample( (block): Sequential( (0): BatchNorm2d(3, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (1): ReLU(inplace) (2): Conv2d(3, 64, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False) ) ) (1): ResBlock( (block): Sequential( (0): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (1): ReLU(inplace) (2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (3): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (4): ReLU(inplace) (5): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) ) ) (2): ResBlock( (block): Sequential( (0): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (1): ReLU(inplace) (2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (3): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (4): ReLU(inplace) (5): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) ) ) ) (1): Sequential( (0): DownSample( (block): Sequential( (0): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (1): ReLU(inplace) (2): Conv2d(64, 128, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False) ) ) (1): ResBlock( (block): Sequential( (0): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (1): ReLU(inplace) (2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (3): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (4): ReLU(inplace) (5): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) ) ) (2): ResBlock( (block): Sequential( (0): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (1): ReLU(inplace) (2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (3): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (4): ReLU(inplace) (5): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) ) ) ) (2): Sequential( (0): DownSample( (block): Sequential( (0): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (1): ReLU(inplace) (2): Conv2d(128, 256, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False) ) ) (1): ResBlock( (block): Sequential( (0): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (1): ReLU(inplace) (2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (4): ReLU(inplace) (5): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) ) ) (2): ResBlock( (block): Sequential( (0): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (1): ReLU(inplace) (2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (4): ReLU(inplace) (5): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) ) ) ) (3): Sequential( (0): DownSample( (block): Sequential( (0): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (1): ReLU(inplace) (2): Conv2d(256, 256, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False) ) ) (1): ResBlock( (block): Sequential( (0): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (1): ReLU(inplace) (2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (4): ReLU(inplace) (5): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) ) ) (2): ResBlock( (block): Sequential( (0): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (1): ReLU(inplace) (2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (4): ReLU(inplace) (5): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) ) ) ) (4): Sequential( (0): DownSample( (block): Sequential( (0): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (1): ReLU(inplace) (2): Conv2d(256, 256, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False) ) ) (1): ResBlock( (block): Sequential( (0): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (1): ReLU(inplace) (2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (4): ReLU(inplace) (5): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) ) ) (2): ResBlock( (block): Sequential( (0): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (1): ReLU(inplace) (2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) (3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True) (4): ReLU(inplace) (5): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False) ) ) ) ) ) (image_FPN): RefinePyramid( (adaptive): ModuleList( (0): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1)) (1): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1)) (2): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1)) (3): Conv2d(128, 256, kernel_size=(1, 1), stride=(1, 1)) (4): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1)) ) (smooth): ModuleList( (0): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) (1): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) (2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) (3): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) (4): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) ) ) (cond_FPN): RefinePyramid( (adaptive): ModuleList( (0): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1)) (1): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1)) (2): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1)) (3): Conv2d(128, 256, kernel_size=(1, 1), stride=(1, 1)) (4): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1)) ) (smooth): ModuleList( (0): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) (1): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) (2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) (3): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) (4): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) ) ) (aflow_net): AFlowNet( (netRefine): ModuleList( (0): Sequential( (0): Conv2d(512, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) (1): LeakyReLU(negative_slope=0.1) (2): Conv2d(128, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) (3): LeakyReLU(negative_slope=0.1) (4): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) (5): LeakyReLU(negative_slope=0.1) (6): Conv2d(32, 2, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) ) (1): Sequential( (0): Conv2d(512, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) (1): LeakyReLU(negative_slope=0.1) (2): Conv2d(128, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) (3): LeakyReLU(negative_slope=0.1) (4): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) (5): LeakyReLU(negative_slope=0.1) (6): Conv2d(32, 2, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) ) (2): Sequential( (0): Conv2d(512, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) (1): LeakyReLU(negative_slope=0.1) (2): Conv2d(128, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) (3): LeakyReLU(negative_slope=0.1) (4): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) (5): LeakyReLU(negative_slope=0.1) (6): Conv2d(32, 2, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) ) (3): Sequential( (0): Conv2d(512, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) (1): LeakyReLU(negative_slope=0.1) (2): Conv2d(128, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) (3): LeakyReLU(negative_slope=0.1) (4): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) (5): LeakyReLU(negative_slope=0.1) (6): Conv2d(32, 2, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) ) (4): Sequential( (0): Conv2d(512, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) (1): LeakyReLU(negative_slope=0.1) (2): Conv2d(128, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) (3): LeakyReLU(negative_slope=0.1) (4): Conv2d(64, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) (5): LeakyReLU(negative_slope=0.1) (6): Conv2d(32, 2, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1)) ) ) (netStyle): ModuleList( (0): StyledConvBlock( (conv0): ModulatedConv2d( (mlp_class_std): EqualLinear( (linear): Linear(in_features=256, out_features=256, bias=True) ) (padding): ZeroPad2d(padding=(1, 1, 1, 1), value=0.0) ) (actvn0): LeakyReLU(negative_slope=0.2, inplace) (conv1): ModulatedConv2d( (mlp_class_std): EqualLinear( (linear): Linear(in_features=256, out_features=49, bias=True) ) (padding): ZeroPad2d(padding=(1, 1, 1, 1), value=0.0) ) (actvn1): LeakyReLU(negative_slope=0.2, inplace) ) (1): StyledConvBlock( (conv0): ModulatedConv2d( (mlp_class_std): EqualLinear( (linear): Linear(in_features=256, out_features=256, bias=True) ) (padding): ZeroPad2d(padding=(1, 1, 1, 1), value=0.0) ) (actvn0): LeakyReLU(negative_slope=0.2, inplace) (conv1): ModulatedConv2d( (mlp_class_std): EqualLinear( (linear): Linear(in_features=256, out_features=49, bias=True) ) (padding): ZeroPad2d(padding=(1, 1, 1, 1), value=0.0) ) (actvn1): LeakyReLU(negative_slope=0.2, inplace) ) (2): StyledConvBlock( (conv0): ModulatedConv2d( (mlp_class_std): EqualLinear( (linear): Linear(in_features=256, out_features=256, bias=True) ) (padding): ZeroPad2d(padding=(1, 1, 1, 1), value=0.0) ) (actvn0): LeakyReLU(negative_slope=0.2, inplace) (conv1): ModulatedConv2d( (mlp_class_std): EqualLinear( (linear): Linear(in_features=256, out_features=49, bias=True) ) (padding): ZeroPad2d(padding=(1, 1, 1, 1), value=0.0) ) (actvn1): LeakyReLU(negative_slope=0.2, inplace) ) (3): StyledConvBlock( (conv0): ModulatedConv2d( (mlp_class_std): EqualLinear( (linear): Linear(in_features=256, out_features=256, bias=True) ) (padding): ZeroPad2d(padding=(1, 1, 1, 1), value=0.0) ) (actvn0): LeakyReLU(negative_slope=0.2, inplace) (conv1): ModulatedConv2d( (mlp_class_std): EqualLinear( (linear): Linear(in_features=256, out_features=49, bias=True) ) (padding): ZeroPad2d(padding=(1, 1, 1, 1), value=0.0) ) (actvn1): LeakyReLU(negative_slope=0.2, inplace) ) (4): StyledConvBlock( (conv0): ModulatedConv2d( (mlp_class_std): EqualLinear( (linear): Linear(in_features=256, out_features=256, bias=True) ) (padding): ZeroPad2d(padding=(1, 1, 1, 1), value=0.0) ) (actvn0): LeakyReLU(negative_slope=0.2, inplace) (conv1): ModulatedConv2d( (mlp_class_std): EqualLinear( (linear): Linear(in_features=256, out_features=49, bias=True) ) (padding): ZeroPad2d(padding=(1, 1, 1, 1), value=0.0) ) (actvn1): LeakyReLU(negative_slope=0.2, inplace) ) ) (netF): ModuleList( (0): Styled_F_ConvBlock( (conv0): ModulatedConv2d( (mlp_class_std): EqualLinear( (linear): Linear(in_features=256, out_features=49, bias=True) ) (padding): ZeroPad2d(padding=(1, 1, 1, 1), value=0.0) ) (actvn0): LeakyReLU(negative_slope=0.2, inplace) (conv1): ModulatedConv2d( (mlp_class_std): EqualLinear( (linear): Linear(in_features=256, out_features=128, bias=True) ) (padding): ZeroPad2d(padding=(1, 1, 1, 1), value=0.0) ) ) (1): Styled_F_ConvBlock( (conv0): ModulatedConv2d( (mlp_class_std): EqualLinear( (linear): Linear(in_features=256, out_features=49, bias=True) ) (padding): ZeroPad2d(padding=(1, 1, 1, 1), value=0.0) ) (actvn0): LeakyReLU(negative_slope=0.2, inplace) (conv1): ModulatedConv2d( (mlp_class_std): EqualLinear( (linear): Linear(in_features=256, out_features=128, bias=True) ) (padding): ZeroPad2d(padding=(1, 1, 1, 1), value=0.0) ) ) (2): Styled_F_ConvBlock( (conv0): ModulatedConv2d( (mlp_class_std): EqualLinear( (linear): Linear(in_features=256, out_features=49, bias=True) ) (padding): ZeroPad2d(padding=(1, 1, 1, 1), value=0.0) ) (actvn0): LeakyReLU(negative_slope=0.2, inplace) (conv1): ModulatedConv2d( (mlp_class_std): EqualLinear( (linear): Linear(in_features=256, out_features=128, bias=True) ) (padding): ZeroPad2d(padding=(1, 1, 1, 1), value=0.0) ) ) (3): Styled_F_ConvBlock( (conv0): ModulatedConv2d( (mlp_class_std): EqualLinear( (linear): Linear(in_features=256, out_features=49, bias=True) ) (padding): ZeroPad2d(padding=(1, 1, 1, 1), value=0.0) ) (actvn0): LeakyReLU(negative_slope=0.2, inplace) (conv1): ModulatedConv2d( (mlp_class_std): EqualLinear( (linear): Linear(in_features=256, out_features=128, bias=True) ) (padding): ZeroPad2d(padding=(1, 1, 1, 1), value=0.0) ) ) (4): Styled_F_ConvBlock( (conv0): ModulatedConv2d( (mlp_class_std): EqualLinear( (linear): Linear(in_features=256, out_features=49, bias=True) ) (padding): ZeroPad2d(padding=(1, 1, 1, 1), value=0.0) ) (actvn0): LeakyReLU(negative_slope=0.2, inplace) (conv1): ModulatedConv2d( (mlp_class_std): EqualLinear( (linear): Linear(in_features=256, out_features=128, bias=True) ) (padding): ZeroPad2d(padding=(1, 1, 1, 1), value=0.0) ) ) ) (cond_style): Sequential( (0): Conv2d(256, 128, kernel_size=(8, 6), stride=(1, 1)) (1): LeakyReLU(negative_slope=0.1) ) (image_style): Sequential( (0): Conv2d(256, 128, kernel_size=(8, 6), stride=(1, 1)) (1): LeakyReLU(negative_slope=0.1) ) ) ) ############################### /datasets/john/tryOn/Flow-Style-VTON-Checkpoints/PFAFN_warp_epoch_101.pth /datasets/john/tryOn/Flow-Style-VTON-Checkpoints/PFAFN_warp_epoch_101.pth No checkpoint! /datasets/john/tryOn/Flow-Style-VTON-Checkpoints/PFAFN_gen_epoch_101.pth No checkpoint! /usr/local/lib/python3.6/site-packages/torch/nn/functional.py:2539: UserWarning: Default upsampling behavior when mode=bilinear is changed to align_corners=False since 0.4.0. Please specify align_corners=True if the old behavior is desired. See the documentation of nn.Upsample for details. "See the documentation of nn.Upsample for details.".format(mode)) ['016962_0.jpg'] tensor([[[[ 0.0123, -0.0173, -0.0079, ..., -0.0007, -0.0003, -0.0209], [ 0.0056, -0.0065, -0.0189, ..., -0.0412, -0.0389, -0.0472], [ 0.0180, -0.0155, -0.0281, ..., -0.0524, -0.0492, -0.0567], ..., [ 0.0518, 0.0434, 0.0337, ..., 0.0086, 0.0104, -0.0182], [ 0.0531, 0.0435, 0.0346, ..., 0.0108, 0.0124, -0.0167], [ 0.0602, 0.0384, 0.0341, ..., 0.0229, 0.0233, 0.0078]],

     [[-0.0131, -0.0698, -0.0633,  ..., -0.0459, -0.0438, -0.0486],
      [-0.0237, -0.0781, -0.0694,  ..., -0.0725, -0.0693, -0.0836],
      [-0.0076, -0.0751, -0.0741,  ..., -0.0856, -0.0838, -0.1058],
      ...,
      [-0.0147, -0.0902, -0.0975,  ..., -0.1009, -0.1019, -0.1087],
      [-0.0158, -0.0910, -0.0983,  ..., -0.0993, -0.1004, -0.1077],
      [-0.0067, -0.0219, -0.0362,  ..., -0.0487, -0.0509, -0.0395]],

     [[ 0.0171,  0.0278,  0.0342,  ...,  0.0439,  0.0455,  0.0540],
      [ 0.0395,  0.0132,  0.0324,  ...,  0.0316,  0.0324,  0.0061],
      [ 0.0399,  0.0409,  0.0494,  ...,  0.0409,  0.0419,  0.0016],
      ...,
      [ 0.0323,  0.0158,  0.0168,  ...,  0.0182,  0.0189, -0.0224],
      [ 0.0332,  0.0159,  0.0177,  ...,  0.0212,  0.0221, -0.0211],
      [ 0.0193,  0.0035,  0.0128,  ...,  0.0075,  0.0080, -0.0325]]]],
   grad_fn=<AddBackward0>)

['015794_0.jpg'] tensor([[[[ 0.0138, -0.0189, -0.0087, ..., -0.0007, 0.0001, -0.0188], [ 0.0072, -0.0074, -0.0213, ..., -0.0388, -0.0366, -0.0436], [ 0.0200, -0.0164, -0.0309, ..., -0.0511, -0.0481, -0.0533], ..., [ 0.0583, 0.0483, 0.0376, ..., 0.0088, 0.0108, -0.0151], [ 0.0597, 0.0485, 0.0386, ..., 0.0108, 0.0125, -0.0137], [ 0.0678, 0.0427, 0.0381, ..., 0.0221, 0.0225, 0.0084]],

     [[-0.0137, -0.0761, -0.0686,  ..., -0.0445, -0.0423, -0.0455],
      [-0.0259, -0.0841, -0.0741,  ..., -0.0708, -0.0675, -0.0795],
      [-0.0089, -0.0811, -0.0799,  ..., -0.0839, -0.0820, -0.1008],
      ...,
      [-0.0166, -0.1014, -0.1097,  ..., -0.0946, -0.0955, -0.0997],
      [-0.0178, -0.1023, -0.1106,  ..., -0.0931, -0.0942, -0.0989],
      [-0.0077, -0.0247, -0.0409,  ..., -0.0468, -0.0487, -0.0371]],

     [[ 0.0183,  0.0300,  0.0383,  ...,  0.0402,  0.0417,  0.0500],
      [ 0.0418,  0.0134,  0.0349,  ...,  0.0286,  0.0297,  0.0059],
      [ 0.0432,  0.0444,  0.0539,  ...,  0.0369,  0.0380,  0.0013],
      ...,
      [ 0.0363,  0.0175,  0.0187,  ...,  0.0150,  0.0158, -0.0206],
      [ 0.0372,  0.0175,  0.0197,  ...,  0.0176,  0.0185, -0.0195],
      [ 0.0216,  0.0036,  0.0143,  ...,  0.0057,  0.0065, -0.0288]]]],
   grad_fn=<AddBackward0>)

['014834_0.jpg'] tensor([[[[ 0.0143, -0.0199, -0.0092, ..., -0.0011, -0.0005, -0.0217], [ 0.0074, -0.0084, -0.0230, ..., -0.0445, -0.0419, -0.0498], [ 0.0200, -0.0173, -0.0327, ..., -0.0576, -0.0540, -0.0607], ..., [ 0.0537, 0.0449, 0.0349, ..., 0.0082, 0.0103, -0.0197], [ 0.0551, 0.0454, 0.0362, ..., 0.0108, 0.0127, -0.0178], [ 0.0628, 0.0391, 0.0353, ..., 0.0232, 0.0238, 0.0079]],

     [[-0.0141, -0.0793, -0.0714,  ..., -0.0498, -0.0477, -0.0523],
      [-0.0271, -0.0874, -0.0768,  ..., -0.0790, -0.0755, -0.0901],
      [-0.0106, -0.0851, -0.0838,  ..., -0.0941, -0.0920, -0.1142],
      ...,
      [-0.0166, -0.0966, -0.1052,  ..., -0.1093, -0.1103, -0.1160],
      [-0.0177, -0.0972, -0.1059,  ..., -0.1075, -0.1086, -0.1149],
      [-0.0082, -0.0247, -0.0400,  ..., -0.0534, -0.0556, -0.0423]],

     [[ 0.0188,  0.0310,  0.0403,  ...,  0.0465,  0.0481,  0.0570],
      [ 0.0426,  0.0134,  0.0358,  ...,  0.0340,  0.0347,  0.0067],
      [ 0.0444,  0.0457,  0.0557,  ...,  0.0433,  0.0443,  0.0018],
      ...,
      [ 0.0334,  0.0160,  0.0172,  ...,  0.0184,  0.0190, -0.0246],
      [ 0.0342,  0.0159,  0.0178,  ...,  0.0213,  0.0222, -0.0233],
      [ 0.0196,  0.0029,  0.0127,  ...,  0.0069,  0.0075, -0.0349]]]],
   grad_fn=<AddBackward0>)

['005510_0.jpg'] tensor([[[[ 0.0143, -0.0194, -0.0088, ..., -0.0004, 0.0001, -0.0230], [ 0.0066, -0.0074, -0.0213, ..., -0.0447, -0.0423, -0.0513], [ 0.0209, -0.0171, -0.0314, ..., -0.0580, -0.0544, -0.0616], ..., [ 0.0599, 0.0501, 0.0401, ..., 0.0107, 0.0132, -0.0185], [ 0.0614, 0.0502, 0.0412, ..., 0.0131, 0.0152, -0.0168], [ 0.0703, 0.0444, 0.0405, ..., 0.0262, 0.0270, 0.0098]],

     [[-0.0146, -0.0791, -0.0715,  ..., -0.0514, -0.0491, -0.0536],
      [-0.0267, -0.0885, -0.0782,  ..., -0.0819, -0.0785, -0.0926],
      [-0.0083, -0.0849, -0.0833,  ..., -0.0957, -0.0937, -0.1169],
      ...,
      [-0.0170, -0.1036, -0.1119,  ..., -0.1128, -0.1140, -0.1196],
      [-0.0181, -0.1044, -0.1128,  ..., -0.1111, -0.1124, -0.1186],
      [-0.0079, -0.0260, -0.0426,  ..., -0.0559, -0.0581, -0.0444]],

     [[ 0.0194,  0.0316,  0.0392,  ...,  0.0481,  0.0499,  0.0595],
      [ 0.0447,  0.0149,  0.0370,  ...,  0.0343,  0.0352,  0.0061],
      [ 0.0454,  0.0467,  0.0567,  ...,  0.0436,  0.0447,  0.0009],
      ...,
      [ 0.0363,  0.0177,  0.0190,  ...,  0.0183,  0.0190, -0.0246],
      [ 0.0372,  0.0177,  0.0199,  ...,  0.0214,  0.0224, -0.0234],
      [ 0.0214,  0.0036,  0.0148,  ...,  0.0074,  0.0081, -0.0348]]]],
   grad_fn=<AddBackward0>)

['004912_0.jpg'] tensor([[[[ 0.0142, -0.0192, -0.0088, ..., -0.0007, -0.0003, -0.0202], [ 0.0075, -0.0076, -0.0219, ..., -0.0399, -0.0378, -0.0458], [ 0.0204, -0.0163, -0.0313, ..., -0.0508, -0.0477, -0.0550], ..., [ 0.0533, 0.0447, 0.0347, ..., 0.0072, 0.0088, -0.0161], [ 0.0547, 0.0448, 0.0356, ..., 0.0092, 0.0105, -0.0148], [ 0.0620, 0.0395, 0.0351, ..., 0.0203, 0.0205, 0.0070]],

     [[-0.0138, -0.0774, -0.0697,  ..., -0.0445, -0.0424, -0.0471],
      [-0.0263, -0.0855, -0.0751,  ..., -0.0702, -0.0672, -0.0811],
      [-0.0092, -0.0825, -0.0811,  ..., -0.0830, -0.0812, -0.1026],
      ...,
      [-0.0152, -0.0930, -0.1005,  ..., -0.0899, -0.0907, -0.0965],
      [-0.0163, -0.0937, -0.1013,  ..., -0.0885, -0.0894, -0.0956],
      [-0.0070, -0.0226, -0.0373,  ..., -0.0431, -0.0451, -0.0350]],

     [[ 0.0185,  0.0305,  0.0393,  ...,  0.0426,  0.0442,  0.0524],
      [ 0.0423,  0.0135,  0.0356,  ...,  0.0306,  0.0315,  0.0059],
      [ 0.0440,  0.0452,  0.0551,  ...,  0.0396,  0.0406,  0.0015],
      ...,
      [ 0.0333,  0.0163,  0.0173,  ...,  0.0162,  0.0170, -0.0198],
      [ 0.0342,  0.0163,  0.0183,  ...,  0.0188,  0.0198, -0.0187],
      [ 0.0199,  0.0036,  0.0132,  ...,  0.0064,  0.0069, -0.0288]]]],
   grad_fn=<AddBackward0>)

['000066_0.jpg'] tensor([[[[ 0.0146, -0.0198, -0.0091, ..., -0.0012, -0.0006, -0.0234], [ 0.0077, -0.0079, -0.0225, ..., -0.0468, -0.0441, -0.0529], [ 0.0209, -0.0168, -0.0322, ..., -0.0595, -0.0558, -0.0635], ..., [ 0.0544, 0.0455, 0.0354, ..., 0.0099, 0.0119, -0.0188], [ 0.0557, 0.0456, 0.0363, ..., 0.0123, 0.0140, -0.0171], [ 0.0632, 0.0403, 0.0358, ..., 0.0256, 0.0260, 0.0091]],

     [[-0.0142, -0.0797, -0.0717,  ..., -0.0515, -0.0493, -0.0546],
      [-0.0271, -0.0879, -0.0772,  ..., -0.0818, -0.0782, -0.0938],
      [-0.0095, -0.0849, -0.0834,  ..., -0.0967, -0.0945, -0.1186],
      ...,
      [-0.0155, -0.0948, -0.1024,  ..., -0.1093, -0.1103, -0.1171],
      [-0.0166, -0.0956, -0.1032,  ..., -0.1076, -0.1087, -0.1161],
      [-0.0071, -0.0230, -0.0380,  ..., -0.0529, -0.0552, -0.0429]],

     [[ 0.0190,  0.0314,  0.0404,  ...,  0.0496,  0.0512,  0.0605],
      [ 0.0435,  0.0139,  0.0366,  ...,  0.0358,  0.0366,  0.0067],
      [ 0.0452,  0.0465,  0.0566,  ...,  0.0459,  0.0470,  0.0017],
      ...,
      [ 0.0340,  0.0166,  0.0176,  ...,  0.0194,  0.0203, -0.0240],
      [ 0.0349,  0.0166,  0.0186,  ...,  0.0226,  0.0237, -0.0226],
      [ 0.0202,  0.0036,  0.0135,  ...,  0.0079,  0.0086, -0.0346]]]],
   grad_fn=<AddBackward0>)