Config updates

codes/options/test/test_ESRGAN_adrianna_full.yml

@@ -13,7 +13,7 @@ datasets:
     mode: LQ
     batch_size: 13
     dataroot_LQ: ..\..\datasets\kayden\images
-    start_at: 10000
+    start_at: 0
 
 #### network structures
 network_G:

codes/options/test/test_vix_corrupt.yml

@@ -0,0 +1,28 @@
+name: vix_corrupt
+suffix: ~  # add suffix to saved images
+model: sr
+distortion: sr
+scale: 4
+crop_border: ~  # crop border when evaluation. If None(~), crop the scale pixels
+gpu_ids: [0]
+amp_opt_level: O3
+
+datasets:
+  test_1:  # the 1st test dataset
+    name: vix_lr
+    mode: LQ
+    batch_size: 64
+    dataroot_LQ: K:\\4k6k\\4k_closeup\\lr_corrupted
+    start_at: 35000 #ready to go.
+
+#### network structures
+network_G:
+  which_model_G: ResGenV2
+  nf: 192
+  nb_denoiser: 20
+  nb_upsampler: 0
+  upscale_applications: 0
+
+#### path
+path:
+  pretrain_model_G: ../experiments/pretrained_corruptors/decolorize_nonoise.pth

codes/options/train/train_ESRGAN_blacked_colab.yml

@@ -1,84 +0,0 @@
-#### general settings
-name: blacked_fix_and_upconv
-use_tb_logger: true
-model: srgan
-distortion: sr
-scale: 4
-gpu_ids: [0]
-amp_opt_level: O1
-
-#### datasets
-datasets:
-  train:
-    name: vixcloseup
-    mode: LQGT
-    dataroot_GT: /content/hr
-    dataroot_LQ: /content/lr_corrupted
-    doCrop: false
-    use_shuffle: true
-    n_workers: 12  # per GPU
-    batch_size: 12
-    target_size: 256
-    color: RGB
-  val:
-    name: adrianna_val
-    mode: LQGT
-    dataroot_GT: /content/val/hhq
-    dataroot_LQ: /content/val/hr
-
-#### network structures
-network_G:
-  which_model_G: RRDBNet
-  in_nc: 3
-  out_nc: 3
-  nf: 48
-  nb: 23
-network_D:
-  which_model_D: discriminator_resnet
-  in_nc: 3
-  nf: 48
-
-#### path
-path:
-  pretrain_model_G: experiments/7000_G.pth
-  pretrain_model_D: experiments/7000_D.pth
-  strict_load: true
-  resume_state: ~
-
-#### training settings: learning rate scheme, loss
-train:
-  lr_G: !!float 6e-5
-  weight_decay_G: 0
-  beta1_G: 0.9
-  beta2_G: 0.99
-  lr_D: !!float 8e-5
-  weight_decay_D: 0
-  beta1_D: 0.9
-  beta2_D: 0.99
-  lr_scheme: MultiStepLR
-
-  niter: 400000
-  warmup_iter: -1  # no warm up
-  lr_steps: [20000, 40000, 60000, 80000]
-  lr_gamma: 0.5
-
-  pixel_criterion: l1
-  pixel_weight: !!float 1e-2
-  feature_criterion: l1
-  feature_weight: 1
-  feature_weight_decay: 1
-  feature_weight_decay_steps: 500
-  feature_weight_minimum: 1
-  gan_type: gan  # gan | ragan
-  gan_weight: !!float 1e-2
-
-  D_update_ratio: 1
-  D_init_iters: 0
-
-  manual_seed: 10
-  val_freq: !!float 5e2
-
-#### logger
-logger:
-  print_freq: 50
-  save_checkpoint_freq: !!float 5e2

codes/options/train/train_GAN_blacked_corrupt_via_transform.yml

@@ -1,83 +0,0 @@
-#### general settings
-name: corruptGAN_4k_lqprn_closeup_hq_to_hq
-use_tb_logger: true
-model: corruptgan
-distortion: downsample
-scale: 1
-gpu_ids: [0]
-amp_opt_level: O0
-
-#### datasets
-datasets:
-  train:
-    name: blacked
-    mode: downsample
-    dataroot_GT: K:\\4k6k\\4k_closeup\\hr
-    dataroot_LQ: E:\\4k6k\\adrianna\\for_training\\hr
-    mismatched_Data_OK: true
-    doCrop: false
-    use_shuffle: true
-    n_workers: 4  # per GPU
-    batch_size: 16
-    target_size: 256
-    use_flip: false
-    use_rot: false
-    color: RGB
-  val:
-    name: blacked_val
-    mode: downsample
-    target_size: 256
-    dataroot_GT: ../datasets/blacked/val/hr
-    dataroot_LQ: ../datasets/blacked/val/hr
-
-#### network structures
-network_G:
-  which_model_G: HighToLowResNet
-  in_nc: 3
-  out_nc: 3
-  nf: 16
-network_D:
-  which_model_D: discriminator_vgg_128
-  in_nc: 3
-  nf: 96
-
-#### path
-path:
-  pretrain_model_G: ~
-  resume_state: ~
-  strict_load: true
-
-#### training settings: learning rate scheme, loss
-train:
-  lr_G: !!float 1e-4
-  weight_decay_G: 0
-  beta1_G: 0.9
-  beta2_G: 0.99
-  lr_D: !!float 1e-4
-  weight_decay_D: 0
-  beta1_D: 0.9
-  beta2_D: 0.99
-  lr_scheme: MultiStepLR
-
-  niter: 400000
-  warmup_iter: -1  # no warm up
-  lr_steps: [4000, 8000, 12000, 15000, 20000]
-  lr_gamma: 0.5
-
-  pixel_criterion: l1
-  pixel_weight: !!float 1e-2
-  feature_criterion: l1
-  feature_weight: 0
-  gan_type: ragan  # gan | ragan
-  gan_weight: !!float 1e-1
-
-  D_update_ratio: 1
-  D_init_iters: 0
-
-  manual_seed: 10
-  val_freq: !!float 5e2
-
-#### logger
-logger:
-  print_freq: 50
-  save_checkpoint_freq: !!float 5e2

codes/options/train/train_vix_corrupt.yml

@@ -5,19 +5,19 @@ model: corruptgan
 distortion: downsample
 scale: 1
 gpu_ids: [0]
-amp_opt_level: O1
+amp_opt_level: O0
 
 #### datasets
 datasets:
   train:
     name: blacked
     mode: downsample
-    dataroot_GT: K:\\4k6k\\4k_closeup\\lr_corrupted
+    dataroot_GT: E:\\4k6k\\datasets\\vixen\\4k_closeup\\lr_corrupted
     dataroot_LQ: G:\\data\\pr_upsample\\ultra_lowq\\for_training\\lr
     mismatched_Data_OK: true
     use_shuffle: true
     n_workers: 0 # per GPU
-    batch_size: 1
+    batch_size: 24
     target_size: 64
     use_flip: false
     use_rot: false
@@ -33,26 +33,27 @@ datasets:
 #### network structures
 network_G:
   which_model_G: ResGenV2
-  nf: 16
+  nf: 192
-  nb_denoiser: 5
+  nb_denoiser: 20
   nb_upsampler: 0
   upscale_applications: 0
+  inject_noise: False
 
 network_D:
   which_model_D: discriminator_vgg_128
   in_nc: 3
-  nf: 16
+  nf: 64
 
 #### path
 path:
   pretrain_model_G: ~
   pretrain_model_D: ~
-  resume_state: ~
+  resume_state: ../experiments/train_vix_corrupt/training_state/19000.state
   strict_load: true
 
 #### training settings: learning rate scheme, loss
 train:
-  lr_G: !!float 5e-5
+  lr_G: !!float 1e-4
   weight_decay_G: 0
   beta1_G: 0.9
   beta2_G: 0.99
@@ -60,19 +61,25 @@ train:
   weight_decay_D: 0
   beta1_D: 0.9
   beta2_D: 0.99
+  D_noise_theta_init: .01
+  D_noise_final_it: 20000
+  D_noise_theta_floor: .005
   lr_scheme: MultiStepLR
 
   niter: 400000
   warmup_iter: -1  # no warm up
-  lr_steps: [30000, 60000, 100000, 200000]
+  lr_steps: [15000, 50000, 100000, 200000]
   lr_gamma: 0.5
 
   pixel_criterion: l2
   pixel_weight: !!float 1e-2
   feature_criterion: l1
-  feature_weight: 0
+  feature_weight: 1.0
-  gan_type: gan  # gan | ragan
+  feature_weight_decay: .98
-  gan_weight: !!float 1e-1
+  feature_weight_decay_steps: 1000
+  feature_weight_minimum: .5
+  gan_type: ragan  # gan | ragan
+  gan_weight: .1
   mega_batch_factor: 1
 
   D_update_ratio: 1

codes/test.py

@@ -15,7 +15,7 @@ if __name__ == "__main__":
     #### options
     want_just_images = True
     parser = argparse.ArgumentParser()
-    parser.add_argument('-opt', type=str, help='Path to options YMAL file.', default='options/test/test_ESRGAN_adrianna_full.yml')
+    parser.add_argument('-opt', type=str, help='Path to options YMAL file.', default='options/test/test_vix_corrupt.yml')
     opt = option.parse(parser.parse_args().opt, is_train=False)
     opt = option.dict_to_nonedict(opt)
 

codes/train.py

@@ -30,7 +30,7 @@ def init_dist(backend='nccl', **kwargs):
 def main():
     #### options
     parser = argparse.ArgumentParser()
-    parser.add_argument('-opt', type=str, help='Path to option YAML file.', default='options/train/train_ESRGAN_blacked_xl.yml')
+    parser.add_argument('-opt', type=str, help='Path to option YAML file.', default='options/train/train_vix_corrupt.yml')
     parser.add_argument('--launcher', choices=['none', 'pytorch'], default='none',
                         help='job launcher')
     parser.add_argument('--local_rank', type=int, default=0)