Delete files model/training.log with huggingface_hub

model/training.log

@@ -1,1818 +0,0 @@
-_meta_: {}
-acc: null
-amp: false
-anisotropic_scales: true
-auto_scale_allowed: true
-auto_scale_batch: true
-auto_scale_filters: false
-auto_scale_roi: false
-batch_size: 1
-bundle_root: /Users/sakshirathi/neurotk/bundles/segresnet
-cache_class_indices: null
-cache_rate: null
-calc_val_loss: false
-channels_last: true
-ckpt_path: /Users/sakshirathi/neurotk/bundles/segresnet/model
-ckpt_save: true
-class_index: null
-class_names:
-- acc_0
-crop_add_background: true
-crop_foreground: true
-crop_mode: ratio
-crop_ratios: null
-cuda: false
-data_file_base_dir: /Users/sakshirathi/neurotk/bundles
-data_list_file_path: /Users/sakshirathi/Downloads/work_dir/dataset_local.json
-debug: false
-determ: false
-early_stopping_fraction: 0.001
-extra_modalities: {}
-finetune:
-  ckpt_name: /Users/sakshirathi/neurotk/bundles/segresnet/model/model.pt
-  enabled: false
-float32_precision: null
-fold: 0
-fork: true
-global_rank: 0
-image_size:
-- 544
-- 544
-- 69
-image_size_mm_90:
-- 265.61599121093747
-- 265.6159922216141
-- 190.12765338720757
-image_size_mm_median:
-- 249.68374495589455
-- 249.68375462603575
-- 168.30083390623668
-infer:
-  ckpt_name: /Users/sakshirathi/neurotk/bundles/segresnet/model/model.pt
-  data_list_key: testing
-  enabled: true
-  output_path: /Users/sakshirathi/neurotk/bundles/segresnet/prediction_testing
-input_channels: 1
-intensity_bounds:
-- 39.63595217750186
-- 97.59593563988095
-learning_rate: 0.0002
-log_output_file: /Users/sakshirathi/neurotk/bundles/segresnet/model/training.log
-loss:
-  _target_: DiceCELoss
-  include_background: true
-  sigmoid: false
-  smooth_dr: 1.0e-05
-  smooth_nr: 0
-  softmax: true
-  squared_pred: true
-  to_onehot_y: true
-max_samples_per_class: 12500
-mlflow_experiment_name: Auto3DSeg
-mlflow_tracking_uri: /Users/sakshirathi/neurotk/bundles/segresnet/model/mlruns/
-modality: ct
-network:
-  _target_: SegResNetDS
-  blocks_down:
-  - 1
-  - 2
-  - 2
-  - 4
-  - 4
-  dsdepth: 4
-  in_channels: 1
-  init_filters: 32
-  norm: INSTANCE_NVFUSER
-  out_channels: 2
-normalize_mode: range
-notf32: false
-num_crops_per_image: 2
-num_epochs: 1250
-num_epochs_per_saving: 1
-num_epochs_per_validation: null
-num_images_per_batch: 1
-num_steps_per_image: null
-num_warmup_epochs: 3
-num_workers: 4
-optimizer:
-  _target_: torch.optim.AdamW
-  lr: 0.0002
-  weight_decay: 1.0e-05
-orientation_ras: true
-output_classes: 2
-pretrained_ckpt_name: /Users/sakshirathi/neurotk/bundles/segresnet/model/model.pt
-quick: false
-rank: 0
-resample: true
-resample_resolution:
-- 0.48766356436698155
-- 0.4876635832539761
-- 2.748479210553717
-roi_size:
-- 384
-- 384
-- 60
-sigmoid: false
-spacing_lower:
-- 0.42813486948609353
-- 0.428134856247896
-- 2.499999978382533
-spacing_median:
-- 0.48766356436698155
-- 0.4876635832539761
-- 4.770811902267695
-spacing_upper:
-- 0.5859375
-- 0.5859375004856939
-- 5.012642938162783
-start_epoch: 0
-stop_on_lowacc: true
-validate:
-  ckpt_name: /Users/sakshirathi/neurotk/bundles/segresnet/model/model.pt
-  enabled: false
-  invert: true
-  output_path: /Users/sakshirathi/neurotk/bundles/segresnet/prediction_validation
-  save_mask: false
-validate_final_original_res: true
-
-auto_adjust_network_settings no distributed global_rank 0
-GPU device memory min: 16
-base_numel 7225344 gpu_factor 1 gpu_factor_init 1
-input roi [224 224 144] image_size [ 512.000  512.000  61.000] numel  7225344
-increasing roi step [ 257.600  257.600  61.000]
-increasing roi result 1 [ 257.600  257.600  61.000]
-increasing roi step [ 296.240  296.240  61.000]
-increasing roi result 1 [ 296.240  296.240  61.000]
-increasing roi step [ 340.676  340.676  61.000]
-increasing roi result 1 [ 340.676  340.676  61.000]
-increasing roi step [ 391.777  391.777  61.000]
-increasing roi result 1 [ 391.777  391.777  61.000]
-roi_size factored [ 384.000  384.000  60.000] factor [ 16.000  16.000  4.000] extra_levels [ 0.000  0.000  2.000]
-kept filters the same base_numel 7225344,  gpu_factor 1
-kept batch the same base_numel 7225344,  gpu_factor 1,  gpu_factor_init 1
-Suggested network parameters: 
-Batch size 1 => 1 
-ROI size [224, 224, 144] => [384, 384, 60] 
-init_filters 32 => 32 
-aniso: True image_size_mm: [249.68374495589455, 249.68375462603575, 168.30083390623668] spacing: [0.48766356436698155, 0.4876635832539761, 2.748479210553717] levels: 5 
-
-Using anisotropic scales {'_target_': 'SegResNetDS', 'init_filters': 32, 'blocks_down': [1, 2, 2, 4, 4], 'norm': 'INSTANCE', 'in_channels': 1, 'out_channels': 2, 'dsdepth': 4, 'resolution': [0.48766356436698155, 0.4876635832539761, 2.748479210553717]}
-SegResNetDS(
-  (encoder): SegResEncoder(
-    (conv_init): Conv3d(1, 32, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-    (layers): ModuleList(
-      (0): ModuleDict(
-        (blocks): Sequential(
-          (0): SegResBlock(
-            (norm1): InstanceNorm3d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(32, 32, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-            (norm2): InstanceNorm3d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(32, 32, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-          )
-        )
-        (downsample): Conv3d(32, 64, kernel_size=(3, 3, 1), stride=(np.int64(2), np.int64(2), np.int64(1)), padding=(1, 1, 0), bias=False)
-      )
-      (1): ModuleDict(
-        (blocks): Sequential(
-          (0): SegResBlock(
-            (norm1): InstanceNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(64, 64, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-            (norm2): InstanceNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(64, 64, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-          )
-          (1): SegResBlock(
-            (norm1): InstanceNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(64, 64, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-            (norm2): InstanceNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(64, 64, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-          )
-        )
-        (downsample): Conv3d(64, 128, kernel_size=(3, 3, 1), stride=(np.int64(2), np.int64(2), np.int64(1)), padding=(1, 1, 0), bias=False)
-      )
-      (2): ModuleDict(
-        (blocks): Sequential(
-          (0): SegResBlock(
-            (norm1): InstanceNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (1): SegResBlock(
-            (norm1): InstanceNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-        )
-        (downsample): Conv3d(128, 256, kernel_size=(3, 3, 3), stride=(2, 2, 2), padding=(1, 1, 1), bias=False)
-      )
-      (3): ModuleDict(
-        (blocks): Sequential(
-          (0): SegResBlock(
-            (norm1): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (1): SegResBlock(
-            (norm1): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (2): SegResBlock(
-            (norm1): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (3): SegResBlock(
-            (norm1): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-        )
-        (downsample): Conv3d(256, 512, kernel_size=(3, 3, 3), stride=(2, 2, 2), padding=(1, 1, 1), bias=False)
-      )
-      (4): ModuleDict(
-        (blocks): Sequential(
-          (0): SegResBlock(
-            (norm1): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (1): SegResBlock(
-            (norm1): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (2): SegResBlock(
-            (norm1): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (3): SegResBlock(
-            (norm1): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-        )
-        (downsample): Identity()
-      )
-    )
-  )
-  (up_layers): ModuleList(
-    (0): ModuleDict(
-      (upsample): UpSample(
-        (deconv): ConvTranspose3d(512, 256, kernel_size=(3, 3, 3), stride=(2, 2, 2), padding=(1, 1, 1), output_padding=(1, 1, 1), bias=False)
-      )
-      (blocks): Sequential(
-        (0): SegResBlock(
-          (norm1): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act1): ReLU(inplace=True)
-          (conv1): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          (norm2): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act2): ReLU(inplace=True)
-          (conv2): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-        )
-      )
-      (head): Conv3d(256, 2, kernel_size=(1, 1, 1), stride=(1, 1, 1))
-    )
-    (1): ModuleDict(
-      (upsample): UpSample(
-        (deconv): ConvTranspose3d(256, 128, kernel_size=(3, 3, 3), stride=(2, 2, 2), padding=(1, 1, 1), output_padding=(1, 1, 1), bias=False)
-      )
-      (blocks): Sequential(
-        (0): SegResBlock(
-          (norm1): InstanceNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act1): ReLU(inplace=True)
-          (conv1): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          (norm2): InstanceNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act2): ReLU(inplace=True)
-          (conv2): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-        )
-      )
-      (head): Conv3d(128, 2, kernel_size=(1, 1, 1), stride=(1, 1, 1))
-    )
-    (2): ModuleDict(
-      (upsample): UpSample(
-        (deconv): ConvTranspose3d(128, 64, kernel_size=(3, 3, 1), stride=(np.int64(2), np.int64(2), np.int64(1)), padding=(1, 1, 0), output_padding=(np.int64(1), np.int64(1), np.int64(0)), bias=False)
-      )
-      (blocks): Sequential(
-        (0): SegResBlock(
-          (norm1): InstanceNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act1): ReLU(inplace=True)
-          (conv1): Conv3d(64, 64, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-          (norm2): InstanceNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act2): ReLU(inplace=True)
-          (conv2): Conv3d(64, 64, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-        )
-      )
-      (head): Conv3d(64, 2, kernel_size=(1, 1, 1), stride=(1, 1, 1))
-    )
-    (3): ModuleDict(
-      (upsample): UpSample(
-        (deconv): ConvTranspose3d(64, 32, kernel_size=(3, 3, 1), stride=(np.int64(2), np.int64(2), np.int64(1)), padding=(1, 1, 0), output_padding=(np.int64(1), np.int64(1), np.int64(0)), bias=False)
-      )
-      (blocks): Sequential(
-        (0): SegResBlock(
-          (norm1): InstanceNorm3d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act1): ReLU(inplace=True)
-          (conv1): Conv3d(32, 32, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-          (norm2): InstanceNorm3d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act2): ReLU(inplace=True)
-          (conv2): Conv3d(32, 32, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-        )
-      )
-      (head): Conv3d(32, 2, kernel_size=(1, 1, 1), stride=(1, 1, 1))
-    )
-  )
-)
-=> loaded checkpoint /Users/sakshirathi/neurotk/bundles/segresnet/model/model.pt (epoch 1122) (best_metric 0.843817412853241) setting start_epoch 0
-Total parameters count: 86278888 distributed: False
-Inference complete, time 234.85s shape torch.Size([2, 512, 512, 40]) {'image': 'sample_data/images/TBI_INVAC184NYT.nii'}
-_meta_: {}
-acc: null
-amp: false
-anisotropic_scales: true
-auto_scale_allowed: true
-auto_scale_batch: true
-auto_scale_filters: false
-auto_scale_roi: false
-batch_size: 1
-bundle_root: /Users/sakshirathi/neurotk/bundles/segresnet
-cache_class_indices: null
-cache_rate: null
-calc_val_loss: false
-channels_last: true
-ckpt_path: /Users/sakshirathi/neurotk/bundles/segresnet/model
-ckpt_save: true
-class_index: null
-class_names:
-- acc_0
-crop_add_background: true
-crop_foreground: true
-crop_mode: ratio
-crop_ratios: null
-cuda: false
-data_file_base_dir: /Users/sakshirathi/neurotk/bundles
-data_list_file_path: /Users/sakshirathi/Downloads/work_dir/dataset_local.json
-debug: false
-determ: false
-early_stopping_fraction: 0.001
-extra_modalities: {}
-finetune:
-  ckpt_name: /Users/sakshirathi/neurotk/bundles/segresnet/model/model.pt
-  enabled: false
-float32_precision: null
-fold: 0
-fork: true
-global_rank: 0
-image_size:
-- 544
-- 544
-- 69
-image_size_mm_90:
-- 265.61599121093747
-- 265.6159922216141
-- 190.12765338720757
-image_size_mm_median:
-- 249.68374495589455
-- 249.68375462603575
-- 168.30083390623668
-infer:
-  ckpt_name: /Users/sakshirathi/neurotk/bundles/segresnet/model/model.pt
-  data_list_key: testing
-  enabled: true
-  output_path: /Users/sakshirathi/neurotk/bundles/segresnet/prediction_testing
-input_channels: 1
-intensity_bounds:
-- 39.63595217750186
-- 97.59593563988095
-learning_rate: 0.0002
-log_output_file: /Users/sakshirathi/neurotk/bundles/segresnet/model/training.log
-loss:
-  _target_: DiceCELoss
-  include_background: true
-  sigmoid: false
-  smooth_dr: 1.0e-05
-  smooth_nr: 0
-  softmax: true
-  squared_pred: true
-  to_onehot_y: true
-max_samples_per_class: 12500
-mlflow_experiment_name: Auto3DSeg
-mlflow_tracking_uri: /Users/sakshirathi/neurotk/bundles/segresnet/model/mlruns/
-modality: ct
-network:
-  _target_: SegResNetDS
-  blocks_down:
-  - 1
-  - 2
-  - 2
-  - 4
-  - 4
-  dsdepth: 4
-  in_channels: 1
-  init_filters: 32
-  norm: INSTANCE_NVFUSER
-  out_channels: 2
-normalize_mode: range
-notf32: false
-num_crops_per_image: 2
-num_epochs: 1250
-num_epochs_per_saving: 1
-num_epochs_per_validation: null
-num_images_per_batch: 1
-num_steps_per_image: null
-num_warmup_epochs: 3
-num_workers: 4
-optimizer:
-  _target_: torch.optim.AdamW
-  lr: 0.0002
-  weight_decay: 1.0e-05
-orientation_ras: true
-output_classes: 2
-pretrained_ckpt_name: /Users/sakshirathi/neurotk/bundles/segresnet/model/model.pt
-quick: false
-rank: 0
-resample: true
-resample_resolution:
-- 0.48766356436698155
-- 0.4876635832539761
-- 2.748479210553717
-roi_size:
-- 384
-- 384
-- 60
-sigmoid: false
-spacing_lower:
-- 0.42813486948609353
-- 0.428134856247896
-- 2.499999978382533
-spacing_median:
-- 0.48766356436698155
-- 0.4876635832539761
-- 4.770811902267695
-spacing_upper:
-- 0.5859375
-- 0.5859375004856939
-- 5.012642938162783
-start_epoch: 0
-stop_on_lowacc: true
-validate:
-  ckpt_name: /Users/sakshirathi/neurotk/bundles/segresnet/model/model.pt
-  enabled: false
-  invert: true
-  output_path: /Users/sakshirathi/neurotk/bundles/segresnet/prediction_validation
-  save_mask: false
-validate_final_original_res: true
-
-auto_adjust_network_settings no distributed global_rank 0
-GPU device memory min: 16
-base_numel 7225344 gpu_factor 1 gpu_factor_init 1
-input roi [224 224 144] image_size [ 512.000  512.000  61.000] numel  7225344
-increasing roi step [ 257.600  257.600  61.000]
-increasing roi result 1 [ 257.600  257.600  61.000]
-increasing roi step [ 296.240  296.240  61.000]
-increasing roi result 1 [ 296.240  296.240  61.000]
-increasing roi step [ 340.676  340.676  61.000]
-increasing roi result 1 [ 340.676  340.676  61.000]
-increasing roi step [ 391.777  391.777  61.000]
-increasing roi result 1 [ 391.777  391.777  61.000]
-roi_size factored [ 384.000  384.000  60.000] factor [ 16.000  16.000  4.000] extra_levels [ 0.000  0.000  2.000]
-kept filters the same base_numel 7225344,  gpu_factor 1
-kept batch the same base_numel 7225344,  gpu_factor 1,  gpu_factor_init 1
-Suggested network parameters: 
-Batch size 1 => 1 
-ROI size [224, 224, 144] => [384, 384, 60] 
-init_filters 32 => 32 
-aniso: True image_size_mm: [249.68374495589455, 249.68375462603575, 168.30083390623668] spacing: [0.48766356436698155, 0.4876635832539761, 2.748479210553717] levels: 5 
-
-Using anisotropic scales {'_target_': 'SegResNetDS', 'init_filters': 32, 'blocks_down': [1, 2, 2, 4, 4], 'norm': 'INSTANCE', 'in_channels': 1, 'out_channels': 2, 'dsdepth': 4, 'resolution': [0.48766356436698155, 0.4876635832539761, 2.748479210553717]}
-SegResNetDS(
-  (encoder): SegResEncoder(
-    (conv_init): Conv3d(1, 32, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-    (layers): ModuleList(
-      (0): ModuleDict(
-        (blocks): Sequential(
-          (0): SegResBlock(
-            (norm1): InstanceNorm3d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(32, 32, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-            (norm2): InstanceNorm3d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(32, 32, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-          )
-        )
-        (downsample): Conv3d(32, 64, kernel_size=(3, 3, 1), stride=(np.int64(2), np.int64(2), np.int64(1)), padding=(1, 1, 0), bias=False)
-      )
-      (1): ModuleDict(
-        (blocks): Sequential(
-          (0): SegResBlock(
-            (norm1): InstanceNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(64, 64, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-            (norm2): InstanceNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(64, 64, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-          )
-          (1): SegResBlock(
-            (norm1): InstanceNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(64, 64, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-            (norm2): InstanceNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(64, 64, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-          )
-        )
-        (downsample): Conv3d(64, 128, kernel_size=(3, 3, 1), stride=(np.int64(2), np.int64(2), np.int64(1)), padding=(1, 1, 0), bias=False)
-      )
-      (2): ModuleDict(
-        (blocks): Sequential(
-          (0): SegResBlock(
-            (norm1): InstanceNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (1): SegResBlock(
-            (norm1): InstanceNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-        )
-        (downsample): Conv3d(128, 256, kernel_size=(3, 3, 3), stride=(2, 2, 2), padding=(1, 1, 1), bias=False)
-      )
-      (3): ModuleDict(
-        (blocks): Sequential(
-          (0): SegResBlock(
-            (norm1): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (1): SegResBlock(
-            (norm1): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (2): SegResBlock(
-            (norm1): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (3): SegResBlock(
-            (norm1): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-        )
-        (downsample): Conv3d(256, 512, kernel_size=(3, 3, 3), stride=(2, 2, 2), padding=(1, 1, 1), bias=False)
-      )
-      (4): ModuleDict(
-        (blocks): Sequential(
-          (0): SegResBlock(
-            (norm1): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (1): SegResBlock(
-            (norm1): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (2): SegResBlock(
-            (norm1): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (3): SegResBlock(
-            (norm1): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-        )
-        (downsample): Identity()
-      )
-    )
-  )
-  (up_layers): ModuleList(
-    (0): ModuleDict(
-      (upsample): UpSample(
-        (deconv): ConvTranspose3d(512, 256, kernel_size=(3, 3, 3), stride=(2, 2, 2), padding=(1, 1, 1), output_padding=(1, 1, 1), bias=False)
-      )
-      (blocks): Sequential(
-        (0): SegResBlock(
-          (norm1): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act1): ReLU(inplace=True)
-          (conv1): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          (norm2): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act2): ReLU(inplace=True)
-          (conv2): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-        )
-      )
-      (head): Conv3d(256, 2, kernel_size=(1, 1, 1), stride=(1, 1, 1))
-    )
-    (1): ModuleDict(
-      (upsample): UpSample(
-        (deconv): ConvTranspose3d(256, 128, kernel_size=(3, 3, 3), stride=(2, 2, 2), padding=(1, 1, 1), output_padding=(1, 1, 1), bias=False)
-      )
-      (blocks): Sequential(
-        (0): SegResBlock(
-          (norm1): InstanceNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act1): ReLU(inplace=True)
-          (conv1): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          (norm2): InstanceNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act2): ReLU(inplace=True)
-          (conv2): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-        )
-      )
-      (head): Conv3d(128, 2, kernel_size=(1, 1, 1), stride=(1, 1, 1))
-    )
-    (2): ModuleDict(
-      (upsample): UpSample(
-        (deconv): ConvTranspose3d(128, 64, kernel_size=(3, 3, 1), stride=(np.int64(2), np.int64(2), np.int64(1)), padding=(1, 1, 0), output_padding=(np.int64(1), np.int64(1), np.int64(0)), bias=False)
-      )
-      (blocks): Sequential(
-        (0): SegResBlock(
-          (norm1): InstanceNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act1): ReLU(inplace=True)
-          (conv1): Conv3d(64, 64, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-          (norm2): InstanceNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act2): ReLU(inplace=True)
-          (conv2): Conv3d(64, 64, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-        )
-      )
-      (head): Conv3d(64, 2, kernel_size=(1, 1, 1), stride=(1, 1, 1))
-    )
-    (3): ModuleDict(
-      (upsample): UpSample(
-        (deconv): ConvTranspose3d(64, 32, kernel_size=(3, 3, 1), stride=(np.int64(2), np.int64(2), np.int64(1)), padding=(1, 1, 0), output_padding=(np.int64(1), np.int64(1), np.int64(0)), bias=False)
-      )
-      (blocks): Sequential(
-        (0): SegResBlock(
-          (norm1): InstanceNorm3d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act1): ReLU(inplace=True)
-          (conv1): Conv3d(32, 32, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-          (norm2): InstanceNorm3d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act2): ReLU(inplace=True)
-          (conv2): Conv3d(32, 32, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-        )
-      )
-      (head): Conv3d(32, 2, kernel_size=(1, 1, 1), stride=(1, 1, 1))
-    )
-  )
-)
-=> loaded checkpoint /Users/sakshirathi/neurotk/bundles/segresnet/model/model.pt (epoch 1122) (best_metric 0.843817412853241) setting start_epoch 0
-Total parameters count: 86278888 distributed: False
-Inference complete, time 233.93s shape torch.Size([2, 512, 512, 40]) {'image': 'sample_data/images/TBI_INVAC184NYT.nii'}
-_meta_: {}
-acc: null
-amp: false
-anisotropic_scales: true
-auto_scale_allowed: true
-auto_scale_batch: true
-auto_scale_filters: false
-auto_scale_roi: false
-batch_size: 1
-bundle_root: /Users/sakshirathi/neurotk/bundles/segresnet
-cache_class_indices: null
-cache_rate: null
-calc_val_loss: false
-channels_last: true
-ckpt_path: /Users/sakshirathi/neurotk/bundles/segresnet/model
-ckpt_save: true
-class_index: null
-class_names:
-- acc_0
-crop_add_background: true
-crop_foreground: true
-crop_mode: ratio
-crop_ratios: null
-cuda: false
-data_file_base_dir: /Users/sakshirathi/neurotk/bundles
-data_list_file_path: /Users/sakshirathi/Downloads/work_dir/dataset_local.json
-debug: false
-determ: false
-early_stopping_fraction: 0.001
-extra_modalities: {}
-finetune:
-  ckpt_name: /Users/sakshirathi/neurotk/bundles/segresnet/model/model.pt
-  enabled: false
-float32_precision: null
-fold: 0
-fork: true
-global_rank: 0
-image_size:
-- 544
-- 544
-- 69
-image_size_mm_90:
-- 265.61599121093747
-- 265.6159922216141
-- 190.12765338720757
-image_size_mm_median:
-- 249.68374495589455
-- 249.68375462603575
-- 168.30083390623668
-infer:
-  ckpt_name: /Users/sakshirathi/neurotk/bundles/segresnet/model/model.pt
-  data_list_key: testing
-  enabled: true
-  output_path: /Users/sakshirathi/neurotk/bundles/segresnet/prediction_testing
-input_channels: 1
-intensity_bounds:
-- 39.63595217750186
-- 97.59593563988095
-learning_rate: 0.0002
-log_output_file: /Users/sakshirathi/neurotk/bundles/segresnet/model/training.log
-loss:
-  _target_: DiceCELoss
-  include_background: true
-  sigmoid: false
-  smooth_dr: 1.0e-05
-  smooth_nr: 0
-  softmax: true
-  squared_pred: true
-  to_onehot_y: true
-max_samples_per_class: 12500
-mlflow_experiment_name: Auto3DSeg
-mlflow_tracking_uri: /Users/sakshirathi/neurotk/bundles/segresnet/model/mlruns/
-modality: ct
-network:
-  _target_: SegResNetDS
-  blocks_down:
-  - 1
-  - 2
-  - 2
-  - 4
-  - 4
-  dsdepth: 4
-  in_channels: 1
-  init_filters: 32
-  norm: INSTANCE_NVFUSER
-  out_channels: 2
-normalize_mode: range
-notf32: false
-num_crops_per_image: 2
-num_epochs: 1250
-num_epochs_per_saving: 1
-num_epochs_per_validation: null
-num_images_per_batch: 1
-num_steps_per_image: null
-num_warmup_epochs: 3
-num_workers: 4
-optimizer:
-  _target_: torch.optim.AdamW
-  lr: 0.0002
-  weight_decay: 1.0e-05
-orientation_ras: true
-output_classes: 2
-pretrained_ckpt_name: /Users/sakshirathi/neurotk/bundles/segresnet/model/model.pt
-quick: false
-rank: 0
-resample: true
-resample_resolution:
-- 0.48766356436698155
-- 0.4876635832539761
-- 2.748479210553717
-roi_size:
-- 384
-- 384
-- 60
-sigmoid: false
-spacing_lower:
-- 0.42813486948609353
-- 0.428134856247896
-- 2.499999978382533
-spacing_median:
-- 0.48766356436698155
-- 0.4876635832539761
-- 4.770811902267695
-spacing_upper:
-- 0.5859375
-- 0.5859375004856939
-- 5.012642938162783
-start_epoch: 0
-stop_on_lowacc: true
-validate:
-  ckpt_name: /Users/sakshirathi/neurotk/bundles/segresnet/model/model.pt
-  enabled: false
-  invert: true
-  output_path: /Users/sakshirathi/neurotk/bundles/segresnet/prediction_validation
-  save_mask: false
-validate_final_original_res: true
-
-auto_adjust_network_settings no distributed global_rank 0
-GPU device memory min: 16
-base_numel 7225344 gpu_factor 1 gpu_factor_init 1
-input roi [224 224 144] image_size [ 512.000  512.000  61.000] numel  7225344
-increasing roi step [ 257.600  257.600  61.000]
-increasing roi result 1 [ 257.600  257.600  61.000]
-increasing roi step [ 296.240  296.240  61.000]
-increasing roi result 1 [ 296.240  296.240  61.000]
-increasing roi step [ 340.676  340.676  61.000]
-increasing roi result 1 [ 340.676  340.676  61.000]
-increasing roi step [ 391.777  391.777  61.000]
-increasing roi result 1 [ 391.777  391.777  61.000]
-roi_size factored [ 384.000  384.000  60.000] factor [ 16.000  16.000  4.000] extra_levels [ 0.000  0.000  2.000]
-kept filters the same base_numel 7225344,  gpu_factor 1
-kept batch the same base_numel 7225344,  gpu_factor 1,  gpu_factor_init 1
-Suggested network parameters: 
-Batch size 1 => 1 
-ROI size [224, 224, 144] => [384, 384, 60] 
-init_filters 32 => 32 
-aniso: True image_size_mm: [249.68374495589455, 249.68375462603575, 168.30083390623668] spacing: [0.48766356436698155, 0.4876635832539761, 2.748479210553717] levels: 5 
-
-Using anisotropic scales {'_target_': 'SegResNetDS', 'init_filters': 32, 'blocks_down': [1, 2, 2, 4, 4], 'norm': 'INSTANCE', 'in_channels': 1, 'out_channels': 2, 'dsdepth': 4, 'resolution': [0.48766356436698155, 0.4876635832539761, 2.748479210553717]}
-SegResNetDS(
-  (encoder): SegResEncoder(
-    (conv_init): Conv3d(1, 32, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-    (layers): ModuleList(
-      (0): ModuleDict(
-        (blocks): Sequential(
-          (0): SegResBlock(
-            (norm1): InstanceNorm3d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(32, 32, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-            (norm2): InstanceNorm3d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(32, 32, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-          )
-        )
-        (downsample): Conv3d(32, 64, kernel_size=(3, 3, 1), stride=(np.int64(2), np.int64(2), np.int64(1)), padding=(1, 1, 0), bias=False)
-      )
-      (1): ModuleDict(
-        (blocks): Sequential(
-          (0): SegResBlock(
-            (norm1): InstanceNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(64, 64, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-            (norm2): InstanceNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(64, 64, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-          )
-          (1): SegResBlock(
-            (norm1): InstanceNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(64, 64, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-            (norm2): InstanceNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(64, 64, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-          )
-        )
-        (downsample): Conv3d(64, 128, kernel_size=(3, 3, 1), stride=(np.int64(2), np.int64(2), np.int64(1)), padding=(1, 1, 0), bias=False)
-      )
-      (2): ModuleDict(
-        (blocks): Sequential(
-          (0): SegResBlock(
-            (norm1): InstanceNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (1): SegResBlock(
-            (norm1): InstanceNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-        )
-        (downsample): Conv3d(128, 256, kernel_size=(3, 3, 3), stride=(2, 2, 2), padding=(1, 1, 1), bias=False)
-      )
-      (3): ModuleDict(
-        (blocks): Sequential(
-          (0): SegResBlock(
-            (norm1): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (1): SegResBlock(
-            (norm1): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (2): SegResBlock(
-            (norm1): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (3): SegResBlock(
-            (norm1): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-        )
-        (downsample): Conv3d(256, 512, kernel_size=(3, 3, 3), stride=(2, 2, 2), padding=(1, 1, 1), bias=False)
-      )
-      (4): ModuleDict(
-        (blocks): Sequential(
-          (0): SegResBlock(
-            (norm1): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (1): SegResBlock(
-            (norm1): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (2): SegResBlock(
-            (norm1): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (3): SegResBlock(
-            (norm1): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-        )
-        (downsample): Identity()
-      )
-    )
-  )
-  (up_layers): ModuleList(
-    (0): ModuleDict(
-      (upsample): UpSample(
-        (deconv): ConvTranspose3d(512, 256, kernel_size=(3, 3, 3), stride=(2, 2, 2), padding=(1, 1, 1), output_padding=(1, 1, 1), bias=False)
-      )
-      (blocks): Sequential(
-        (0): SegResBlock(
-          (norm1): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act1): ReLU(inplace=True)
-          (conv1): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          (norm2): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act2): ReLU(inplace=True)
-          (conv2): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-        )
-      )
-      (head): Conv3d(256, 2, kernel_size=(1, 1, 1), stride=(1, 1, 1))
-    )
-    (1): ModuleDict(
-      (upsample): UpSample(
-        (deconv): ConvTranspose3d(256, 128, kernel_size=(3, 3, 3), stride=(2, 2, 2), padding=(1, 1, 1), output_padding=(1, 1, 1), bias=False)
-      )
-      (blocks): Sequential(
-        (0): SegResBlock(
-          (norm1): InstanceNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act1): ReLU(inplace=True)
-          (conv1): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          (norm2): InstanceNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act2): ReLU(inplace=True)
-          (conv2): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-        )
-      )
-      (head): Conv3d(128, 2, kernel_size=(1, 1, 1), stride=(1, 1, 1))
-    )
-    (2): ModuleDict(
-      (upsample): UpSample(
-        (deconv): ConvTranspose3d(128, 64, kernel_size=(3, 3, 1), stride=(np.int64(2), np.int64(2), np.int64(1)), padding=(1, 1, 0), output_padding=(np.int64(1), np.int64(1), np.int64(0)), bias=False)
-      )
-      (blocks): Sequential(
-        (0): SegResBlock(
-          (norm1): InstanceNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act1): ReLU(inplace=True)
-          (conv1): Conv3d(64, 64, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-          (norm2): InstanceNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act2): ReLU(inplace=True)
-          (conv2): Conv3d(64, 64, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-        )
-      )
-      (head): Conv3d(64, 2, kernel_size=(1, 1, 1), stride=(1, 1, 1))
-    )
-    (3): ModuleDict(
-      (upsample): UpSample(
-        (deconv): ConvTranspose3d(64, 32, kernel_size=(3, 3, 1), stride=(np.int64(2), np.int64(2), np.int64(1)), padding=(1, 1, 0), output_padding=(np.int64(1), np.int64(1), np.int64(0)), bias=False)
-      )
-      (blocks): Sequential(
-        (0): SegResBlock(
-          (norm1): InstanceNorm3d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act1): ReLU(inplace=True)
-          (conv1): Conv3d(32, 32, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-          (norm2): InstanceNorm3d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act2): ReLU(inplace=True)
-          (conv2): Conv3d(32, 32, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-        )
-      )
-      (head): Conv3d(32, 2, kernel_size=(1, 1, 1), stride=(1, 1, 1))
-    )
-  )
-)
-=> loaded checkpoint /Users/sakshirathi/neurotk/bundles/segresnet/model/model.pt (epoch 1122) (best_metric 0.843817412853241) setting start_epoch 0
-Total parameters count: 86278888 distributed: False
-Inference complete, time 226.94s shape torch.Size([2, 512, 512, 40]) {'image': 'sample_data/images/TBI_INVAC184NYT.nii'}
-_meta_: {}
-acc: null
-amp: false
-anisotropic_scales: true
-auto_scale_allowed: true
-auto_scale_batch: true
-auto_scale_filters: false
-auto_scale_roi: false
-batch_size: 1
-bundle_root: /Users/sakshirathi/neurotk/bundles/segresnet
-cache_class_indices: null
-cache_rate: null
-calc_val_loss: false
-channels_last: true
-ckpt_path: /Users/sakshirathi/neurotk/bundles/segresnet/model
-ckpt_save: true
-class_index: null
-class_names:
-- acc_0
-crop_add_background: true
-crop_foreground: true
-crop_mode: ratio
-crop_ratios: null
-cuda: false
-data_file_base_dir: /Users/sakshirathi/neurotk/bundles
-data_list_file_path: /Users/sakshirathi/Downloads/work_dir/dataset_local.json
-debug: false
-determ: false
-early_stopping_fraction: 0.001
-extra_modalities: {}
-finetune:
-  ckpt_name: /Users/sakshirathi/neurotk/bundles/segresnet/model/model.pt
-  enabled: false
-float32_precision: null
-fold: 0
-fork: true
-global_rank: 0
-image_size:
-- 544
-- 544
-- 69
-image_size_mm_90:
-- 265.61599121093747
-- 265.6159922216141
-- 190.12765338720757
-image_size_mm_median:
-- 249.68374495589455
-- 249.68375462603575
-- 168.30083390623668
-infer:
-  ckpt_name: /Users/sakshirathi/neurotk/bundles/segresnet/model/model.pt
-  data_list_key: testing
-  enabled: true
-  output_path: /Users/sakshirathi/neurotk/bundles/segresnet/prediction_testing
-input_channels: 1
-intensity_bounds:
-- 39.63595217750186
-- 97.59593563988095
-learning_rate: 0.0002
-log_output_file: /Users/sakshirathi/neurotk/bundles/segresnet/model/training.log
-loss:
-  _target_: DiceCELoss
-  include_background: true
-  sigmoid: false
-  smooth_dr: 1.0e-05
-  smooth_nr: 0
-  softmax: true
-  squared_pred: true
-  to_onehot_y: true
-max_samples_per_class: 12500
-mlflow_experiment_name: Auto3DSeg
-mlflow_tracking_uri: /Users/sakshirathi/neurotk/bundles/segresnet/model/mlruns/
-modality: ct
-network:
-  _target_: SegResNetDS
-  blocks_down:
-  - 1
-  - 2
-  - 2
-  - 4
-  - 4
-  dsdepth: 4
-  in_channels: 1
-  init_filters: 32
-  norm: INSTANCE_NVFUSER
-  out_channels: 2
-normalize_mode: range
-notf32: false
-num_crops_per_image: 2
-num_epochs: 1250
-num_epochs_per_saving: 1
-num_epochs_per_validation: null
-num_images_per_batch: 1
-num_steps_per_image: null
-num_warmup_epochs: 3
-num_workers: 4
-optimizer:
-  _target_: torch.optim.AdamW
-  lr: 0.0002
-  weight_decay: 1.0e-05
-orientation_ras: true
-output_classes: 2
-pretrained_ckpt_name: /Users/sakshirathi/neurotk/bundles/segresnet/model/model.pt
-quick: false
-rank: 0
-resample: true
-resample_resolution:
-- 0.48766356436698155
-- 0.4876635832539761
-- 2.748479210553717
-roi_size:
-- 384
-- 384
-- 60
-sigmoid: false
-spacing_lower:
-- 0.42813486948609353
-- 0.428134856247896
-- 2.499999978382533
-spacing_median:
-- 0.48766356436698155
-- 0.4876635832539761
-- 4.770811902267695
-spacing_upper:
-- 0.5859375
-- 0.5859375004856939
-- 5.012642938162783
-start_epoch: 0
-stop_on_lowacc: true
-validate:
-  ckpt_name: /Users/sakshirathi/neurotk/bundles/segresnet/model/model.pt
-  enabled: false
-  invert: true
-  output_path: /Users/sakshirathi/neurotk/bundles/segresnet/prediction_validation
-  save_mask: false
-validate_final_original_res: true
-
-auto_adjust_network_settings no distributed global_rank 0
-GPU device memory min: 16
-base_numel 7225344 gpu_factor 1 gpu_factor_init 1
-input roi [224 224 144] image_size [ 512.000  512.000  61.000] numel  7225344
-increasing roi step [ 257.600  257.600  61.000]
-increasing roi result 1 [ 257.600  257.600  61.000]
-increasing roi step [ 296.240  296.240  61.000]
-increasing roi result 1 [ 296.240  296.240  61.000]
-increasing roi step [ 340.676  340.676  61.000]
-increasing roi result 1 [ 340.676  340.676  61.000]
-increasing roi step [ 391.777  391.777  61.000]
-increasing roi result 1 [ 391.777  391.777  61.000]
-roi_size factored [ 384.000  384.000  60.000] factor [ 16.000  16.000  4.000] extra_levels [ 0.000  0.000  2.000]
-kept filters the same base_numel 7225344,  gpu_factor 1
-kept batch the same base_numel 7225344,  gpu_factor 1,  gpu_factor_init 1
-Suggested network parameters: 
-Batch size 1 => 1 
-ROI size [224, 224, 144] => [384, 384, 60] 
-init_filters 32 => 32 
-aniso: True image_size_mm: [249.68374495589455, 249.68375462603575, 168.30083390623668] spacing: [0.48766356436698155, 0.4876635832539761, 2.748479210553717] levels: 5 
-
-Using anisotropic scales {'_target_': 'SegResNetDS', 'init_filters': 32, 'blocks_down': [1, 2, 2, 4, 4], 'norm': 'INSTANCE', 'in_channels': 1, 'out_channels': 2, 'dsdepth': 4, 'resolution': [0.48766356436698155, 0.4876635832539761, 2.748479210553717]}
-SegResNetDS(
-  (encoder): SegResEncoder(
-    (conv_init): Conv3d(1, 32, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-    (layers): ModuleList(
-      (0): ModuleDict(
-        (blocks): Sequential(
-          (0): SegResBlock(
-            (norm1): InstanceNorm3d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(32, 32, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-            (norm2): InstanceNorm3d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(32, 32, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-          )
-        )
-        (downsample): Conv3d(32, 64, kernel_size=(3, 3, 1), stride=(np.int64(2), np.int64(2), np.int64(1)), padding=(1, 1, 0), bias=False)
-      )
-      (1): ModuleDict(
-        (blocks): Sequential(
-          (0): SegResBlock(
-            (norm1): InstanceNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(64, 64, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-            (norm2): InstanceNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(64, 64, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-          )
-          (1): SegResBlock(
-            (norm1): InstanceNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(64, 64, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-            (norm2): InstanceNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(64, 64, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-          )
-        )
-        (downsample): Conv3d(64, 128, kernel_size=(3, 3, 1), stride=(np.int64(2), np.int64(2), np.int64(1)), padding=(1, 1, 0), bias=False)
-      )
-      (2): ModuleDict(
-        (blocks): Sequential(
-          (0): SegResBlock(
-            (norm1): InstanceNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (1): SegResBlock(
-            (norm1): InstanceNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-        )
-        (downsample): Conv3d(128, 256, kernel_size=(3, 3, 3), stride=(2, 2, 2), padding=(1, 1, 1), bias=False)
-      )
-      (3): ModuleDict(
-        (blocks): Sequential(
-          (0): SegResBlock(
-            (norm1): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (1): SegResBlock(
-            (norm1): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (2): SegResBlock(
-            (norm1): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (3): SegResBlock(
-            (norm1): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-        )
-        (downsample): Conv3d(256, 512, kernel_size=(3, 3, 3), stride=(2, 2, 2), padding=(1, 1, 1), bias=False)
-      )
-      (4): ModuleDict(
-        (blocks): Sequential(
-          (0): SegResBlock(
-            (norm1): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (1): SegResBlock(
-            (norm1): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (2): SegResBlock(
-            (norm1): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (3): SegResBlock(
-            (norm1): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-        )
-        (downsample): Identity()
-      )
-    )
-  )
-  (up_layers): ModuleList(
-    (0): ModuleDict(
-      (upsample): UpSample(
-        (deconv): ConvTranspose3d(512, 256, kernel_size=(3, 3, 3), stride=(2, 2, 2), padding=(1, 1, 1), output_padding=(1, 1, 1), bias=False)
-      )
-      (blocks): Sequential(
-        (0): SegResBlock(
-          (norm1): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act1): ReLU(inplace=True)
-          (conv1): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          (norm2): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act2): ReLU(inplace=True)
-          (conv2): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-        )
-      )
-      (head): Conv3d(256, 2, kernel_size=(1, 1, 1), stride=(1, 1, 1))
-    )
-    (1): ModuleDict(
-      (upsample): UpSample(
-        (deconv): ConvTranspose3d(256, 128, kernel_size=(3, 3, 3), stride=(2, 2, 2), padding=(1, 1, 1), output_padding=(1, 1, 1), bias=False)
-      )
-      (blocks): Sequential(
-        (0): SegResBlock(
-          (norm1): InstanceNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act1): ReLU(inplace=True)
-          (conv1): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          (norm2): InstanceNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act2): ReLU(inplace=True)
-          (conv2): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-        )
-      )
-      (head): Conv3d(128, 2, kernel_size=(1, 1, 1), stride=(1, 1, 1))
-    )
-    (2): ModuleDict(
-      (upsample): UpSample(
-        (deconv): ConvTranspose3d(128, 64, kernel_size=(3, 3, 1), stride=(np.int64(2), np.int64(2), np.int64(1)), padding=(1, 1, 0), output_padding=(np.int64(1), np.int64(1), np.int64(0)), bias=False)
-      )
-      (blocks): Sequential(
-        (0): SegResBlock(
-          (norm1): InstanceNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act1): ReLU(inplace=True)
-          (conv1): Conv3d(64, 64, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-          (norm2): InstanceNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act2): ReLU(inplace=True)
-          (conv2): Conv3d(64, 64, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-        )
-      )
-      (head): Conv3d(64, 2, kernel_size=(1, 1, 1), stride=(1, 1, 1))
-    )
-    (3): ModuleDict(
-      (upsample): UpSample(
-        (deconv): ConvTranspose3d(64, 32, kernel_size=(3, 3, 1), stride=(np.int64(2), np.int64(2), np.int64(1)), padding=(1, 1, 0), output_padding=(np.int64(1), np.int64(1), np.int64(0)), bias=False)
-      )
-      (blocks): Sequential(
-        (0): SegResBlock(
-          (norm1): InstanceNorm3d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act1): ReLU(inplace=True)
-          (conv1): Conv3d(32, 32, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-          (norm2): InstanceNorm3d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act2): ReLU(inplace=True)
-          (conv2): Conv3d(32, 32, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-        )
-      )
-      (head): Conv3d(32, 2, kernel_size=(1, 1, 1), stride=(1, 1, 1))
-    )
-  )
-)
-=> loaded checkpoint /Users/sakshirathi/neurotk/bundles/segresnet/model/model.pt (epoch 1122) (best_metric 0.843817412853241) setting start_epoch 0
-Total parameters count: 86278888 distributed: False
-_meta_: {}
-acc: null
-amp: false
-anisotropic_scales: true
-auto_scale_allowed: true
-auto_scale_batch: true
-auto_scale_filters: false
-auto_scale_roi: false
-batch_size: 1
-bundle_root: /Users/sakshirathi/neurotk/bundles/segresnet
-cache_class_indices: null
-cache_rate: null
-calc_val_loss: false
-channels_last: true
-ckpt_path: /Users/sakshirathi/neurotk/bundles/segresnet/model
-ckpt_save: true
-class_index: null
-class_names:
-- acc_0
-crop_add_background: true
-crop_foreground: true
-crop_mode: ratio
-crop_ratios: null
-cuda: false
-data_file_base_dir: /Users/sakshirathi/neurotk/bundles
-data_list_file_path: /Users/sakshirathi/Downloads/work_dir/dataset_local.json
-debug: false
-determ: false
-early_stopping_fraction: 0.001
-extra_modalities: {}
-finetune:
-  ckpt_name: /Users/sakshirathi/neurotk/bundles/segresnet/model/model.pt
-  enabled: false
-float32_precision: null
-fold: 0
-fork: true
-global_rank: 0
-image_size:
-- 544
-- 544
-- 69
-image_size_mm_90:
-- 265.61599121093747
-- 265.6159922216141
-- 190.12765338720757
-image_size_mm_median:
-- 249.68374495589455
-- 249.68375462603575
-- 168.30083390623668
-infer:
-  ckpt_name: /Users/sakshirathi/neurotk/bundles/segresnet/model/model.pt
-  data_list_key: testing
-  enabled: true
-  output_path: /Users/sakshirathi/neurotk/bundles/segresnet/prediction_testing
-input_channels: 1
-intensity_bounds:
-- 39.63595217750186
-- 97.59593563988095
-learning_rate: 0.0002
-log_output_file: /Users/sakshirathi/neurotk/bundles/segresnet/model/training.log
-loss:
-  _target_: DiceCELoss
-  include_background: true
-  sigmoid: false
-  smooth_dr: 1.0e-05
-  smooth_nr: 0
-  softmax: true
-  squared_pred: true
-  to_onehot_y: true
-max_samples_per_class: 12500
-mlflow_experiment_name: Auto3DSeg
-mlflow_tracking_uri: /Users/sakshirathi/neurotk/bundles/segresnet/model/mlruns/
-modality: ct
-network:
-  _target_: SegResNetDS
-  blocks_down:
-  - 1
-  - 2
-  - 2
-  - 4
-  - 4
-  dsdepth: 4
-  in_channels: 1
-  init_filters: 32
-  norm: INSTANCE_NVFUSER
-  out_channels: 2
-normalize_mode: range
-notf32: false
-num_crops_per_image: 2
-num_epochs: 1250
-num_epochs_per_saving: 1
-num_epochs_per_validation: null
-num_images_per_batch: 1
-num_steps_per_image: null
-num_warmup_epochs: 3
-num_workers: 4
-optimizer:
-  _target_: torch.optim.AdamW
-  lr: 0.0002
-  weight_decay: 1.0e-05
-orientation_ras: true
-output_classes: 2
-pretrained_ckpt_name: /Users/sakshirathi/neurotk/bundles/segresnet/model/model.pt
-quick: false
-rank: 0
-resample: true
-resample_resolution:
-- 0.48766356436698155
-- 0.4876635832539761
-- 2.748479210553717
-roi_size:
-- 384
-- 384
-- 60
-sigmoid: false
-spacing_lower:
-- 0.42813486948609353
-- 0.428134856247896
-- 2.499999978382533
-spacing_median:
-- 0.48766356436698155
-- 0.4876635832539761
-- 4.770811902267695
-spacing_upper:
-- 0.5859375
-- 0.5859375004856939
-- 5.012642938162783
-start_epoch: 0
-stop_on_lowacc: true
-validate:
-  ckpt_name: /Users/sakshirathi/neurotk/bundles/segresnet/model/model.pt
-  enabled: false
-  invert: true
-  output_path: /Users/sakshirathi/neurotk/bundles/segresnet/prediction_validation
-  save_mask: false
-validate_final_original_res: true
-
-auto_adjust_network_settings no distributed global_rank 0
-GPU device memory min: 16
-base_numel 7225344 gpu_factor 1 gpu_factor_init 1
-input roi [224 224 144] image_size [ 512.000  512.000  61.000] numel  7225344
-increasing roi step [ 257.600  257.600  61.000]
-increasing roi result 1 [ 257.600  257.600  61.000]
-increasing roi step [ 296.240  296.240  61.000]
-increasing roi result 1 [ 296.240  296.240  61.000]
-increasing roi step [ 340.676  340.676  61.000]
-increasing roi result 1 [ 340.676  340.676  61.000]
-increasing roi step [ 391.777  391.777  61.000]
-increasing roi result 1 [ 391.777  391.777  61.000]
-roi_size factored [ 384.000  384.000  60.000] factor [ 16.000  16.000  4.000] extra_levels [ 0.000  0.000  2.000]
-kept filters the same base_numel 7225344,  gpu_factor 1
-kept batch the same base_numel 7225344,  gpu_factor 1,  gpu_factor_init 1
-Suggested network parameters: 
-Batch size 1 => 1 
-ROI size [224, 224, 144] => [384, 384, 60] 
-init_filters 32 => 32 
-aniso: True image_size_mm: [249.68374495589455, 249.68375462603575, 168.30083390623668] spacing: [0.48766356436698155, 0.4876635832539761, 2.748479210553717] levels: 5 
-
-Using anisotropic scales {'_target_': 'SegResNetDS', 'init_filters': 32, 'blocks_down': [1, 2, 2, 4, 4], 'norm': 'INSTANCE', 'in_channels': 1, 'out_channels': 2, 'dsdepth': 4, 'resolution': [0.48766356436698155, 0.4876635832539761, 2.748479210553717]}
-SegResNetDS(
-  (encoder): SegResEncoder(
-    (conv_init): Conv3d(1, 32, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-    (layers): ModuleList(
-      (0): ModuleDict(
-        (blocks): Sequential(
-          (0): SegResBlock(
-            (norm1): InstanceNorm3d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(32, 32, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-            (norm2): InstanceNorm3d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(32, 32, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-          )
-        )
-        (downsample): Conv3d(32, 64, kernel_size=(3, 3, 1), stride=(np.int64(2), np.int64(2), np.int64(1)), padding=(1, 1, 0), bias=False)
-      )
-      (1): ModuleDict(
-        (blocks): Sequential(
-          (0): SegResBlock(
-            (norm1): InstanceNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(64, 64, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-            (norm2): InstanceNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(64, 64, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-          )
-          (1): SegResBlock(
-            (norm1): InstanceNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(64, 64, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-            (norm2): InstanceNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(64, 64, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-          )
-        )
-        (downsample): Conv3d(64, 128, kernel_size=(3, 3, 1), stride=(np.int64(2), np.int64(2), np.int64(1)), padding=(1, 1, 0), bias=False)
-      )
-      (2): ModuleDict(
-        (blocks): Sequential(
-          (0): SegResBlock(
-            (norm1): InstanceNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (1): SegResBlock(
-            (norm1): InstanceNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-        )
-        (downsample): Conv3d(128, 256, kernel_size=(3, 3, 3), stride=(2, 2, 2), padding=(1, 1, 1), bias=False)
-      )
-      (3): ModuleDict(
-        (blocks): Sequential(
-          (0): SegResBlock(
-            (norm1): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (1): SegResBlock(
-            (norm1): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (2): SegResBlock(
-            (norm1): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (3): SegResBlock(
-            (norm1): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-        )
-        (downsample): Conv3d(256, 512, kernel_size=(3, 3, 3), stride=(2, 2, 2), padding=(1, 1, 1), bias=False)
-      )
-      (4): ModuleDict(
-        (blocks): Sequential(
-          (0): SegResBlock(
-            (norm1): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (1): SegResBlock(
-            (norm1): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (2): SegResBlock(
-            (norm1): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-          (3): SegResBlock(
-            (norm1): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act1): ReLU(inplace=True)
-            (conv1): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-            (norm2): InstanceNorm3d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-            (act2): ReLU(inplace=True)
-            (conv2): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          )
-        )
-        (downsample): Identity()
-      )
-    )
-  )
-  (up_layers): ModuleList(
-    (0): ModuleDict(
-      (upsample): UpSample(
-        (deconv): ConvTranspose3d(512, 256, kernel_size=(3, 3, 3), stride=(2, 2, 2), padding=(1, 1, 1), output_padding=(1, 1, 1), bias=False)
-      )
-      (blocks): Sequential(
-        (0): SegResBlock(
-          (norm1): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act1): ReLU(inplace=True)
-          (conv1): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          (norm2): InstanceNorm3d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act2): ReLU(inplace=True)
-          (conv2): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-        )
-      )
-      (head): Conv3d(256, 2, kernel_size=(1, 1, 1), stride=(1, 1, 1))
-    )
-    (1): ModuleDict(
-      (upsample): UpSample(
-        (deconv): ConvTranspose3d(256, 128, kernel_size=(3, 3, 3), stride=(2, 2, 2), padding=(1, 1, 1), output_padding=(1, 1, 1), bias=False)
-      )
-      (blocks): Sequential(
-        (0): SegResBlock(
-          (norm1): InstanceNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act1): ReLU(inplace=True)
-          (conv1): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-          (norm2): InstanceNorm3d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act2): ReLU(inplace=True)
-          (conv2): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), bias=False)
-        )
-      )
-      (head): Conv3d(128, 2, kernel_size=(1, 1, 1), stride=(1, 1, 1))
-    )
-    (2): ModuleDict(
-      (upsample): UpSample(
-        (deconv): ConvTranspose3d(128, 64, kernel_size=(3, 3, 1), stride=(np.int64(2), np.int64(2), np.int64(1)), padding=(1, 1, 0), output_padding=(np.int64(1), np.int64(1), np.int64(0)), bias=False)
-      )
-      (blocks): Sequential(
-        (0): SegResBlock(
-          (norm1): InstanceNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act1): ReLU(inplace=True)
-          (conv1): Conv3d(64, 64, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-          (norm2): InstanceNorm3d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act2): ReLU(inplace=True)
-          (conv2): Conv3d(64, 64, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-        )
-      )
-      (head): Conv3d(64, 2, kernel_size=(1, 1, 1), stride=(1, 1, 1))
-    )
-    (3): ModuleDict(
-      (upsample): UpSample(
-        (deconv): ConvTranspose3d(64, 32, kernel_size=(3, 3, 1), stride=(np.int64(2), np.int64(2), np.int64(1)), padding=(1, 1, 0), output_padding=(np.int64(1), np.int64(1), np.int64(0)), bias=False)
-      )
-      (blocks): Sequential(
-        (0): SegResBlock(
-          (norm1): InstanceNorm3d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act1): ReLU(inplace=True)
-          (conv1): Conv3d(32, 32, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-          (norm2): InstanceNorm3d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=False)
-          (act2): ReLU(inplace=True)
-          (conv2): Conv3d(32, 32, kernel_size=(3, 3, 1), stride=(1, 1, 1), padding=(1, 1, 0), bias=False)
-        )
-      )
-      (head): Conv3d(32, 2, kernel_size=(1, 1, 1), stride=(1, 1, 1))
-    )
-  )
-)
-=> loaded checkpoint /Users/sakshirathi/neurotk/bundles/segresnet/model/model.pt (epoch 1122) (best_metric 0.843817412853241) setting start_epoch 0
-Total parameters count: 86278888 distributed: False