| { | |
| "_best_ema": "0.8976196782396484", | |
| "batch_size": "2", | |
| "configuration_manager": "{'data_identifier': 'nnUNetPlans_3d_fullres', 'preprocessor_name': 'DefaultPreprocessor', 'batch_size': 2, 'patch_size': [192, 192, 192], 'median_image_size_in_voxels': [235.0, 240.0, 240.0], 'spacing': [1.5, 1.5, 1.5], 'normalization_schemes': ['CTNormalization'], 'use_mask_for_norm': [False], 'resampling_fn_data': 'resample_data_or_seg_to_shape', 'resampling_fn_seg': 'resample_data_or_seg_to_shape', 'resampling_fn_data_kwargs': {'is_seg': False, 'order': 3, 'order_z': 0, 'force_separate_z': None}, 'resampling_fn_seg_kwargs': {'is_seg': True, 'order': 1, 'order_z': 0, 'force_separate_z': None}, 'resampling_fn_probabilities': 'resample_data_or_seg_to_shape', 'resampling_fn_probabilities_kwargs': {'is_seg': False, 'order': 1, 'order_z': 0, 'force_separate_z': None}, 'architecture': {'network_class_name': 'dynamic_network_architectures.architectures.unet.ResidualEncoderUNet', 'arch_kwargs': {'n_stages': 6, 'features_per_stage': [32, 64, 128, 256, 320, 320], 'conv_op': 'torch.nn.modules.conv.Conv3d', 'kernel_sizes': [[3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3]], 'strides': [[1, 1, 1], [2, 2, 2], [2, 2, 2], [2, 2, 2], [2, 2, 2], [2, 2, 2]], 'n_blocks_per_stage': [1, 3, 4, 6, 6, 6], 'n_conv_per_stage_decoder': [1, 1, 1, 1, 1], 'conv_bias': True, 'norm_op': 'torch.nn.modules.instancenorm.InstanceNorm3d', 'norm_op_kwargs': {'eps': 1e-05, 'affine': True}, 'dropout_op': None, 'dropout_op_kwargs': None, 'nonlin': 'torch.nn.LeakyReLU', 'nonlin_kwargs': {'inplace': True}, 'deep_supervision': True}, '_kw_requires_import': ['conv_op', 'norm_op', 'dropout_op', 'nonlin']}, 'batch_dice': False}", | |
| "configuration_name": "3d_fullres", | |
| "cudnn_version": 8902, | |
| "current_epoch": "900", | |
| "dataloader_train": "<nnunetv2.training.data_augmentation.custom_transforms.limited_length_multithreaded_augmenter.LimitedLenWrapper object at 0x7f1a6d313160>", | |
| "dataloader_train.generator": "<nnunetv2.training.dataloading.data_loader_3d.nnUNetDataLoader3D object at 0x7f1a6d313310>", | |
| "dataloader_train.num_processes": "16", | |
| "dataloader_train.transform": "Compose ( [SpatialTransform( independent_scale_for_each_axis = False, p_rot_per_sample = 0.2, p_scale_per_sample = 0.2, p_el_per_sample = 0, data_key = 'data', label_key = 'seg', patch_size = [192, 192, 192], patch_center_dist_from_border = None, do_elastic_deform = False, alpha = (0, 0), sigma = (0, 0), do_rotation = True, angle_x = (-0.5235987755982988, 0.5235987755982988), angle_y = (-0.5235987755982988, 0.5235987755982988), angle_z = (-0.5235987755982988, 0.5235987755982988), do_scale = True, scale = (0.7, 1.4), border_mode_data = 'constant', border_cval_data = 0, order_data = 3, border_mode_seg = 'constant', border_cval_seg = -1, order_seg = 1, random_crop = False, p_rot_per_axis = 1, p_independent_scale_per_axis = 1 ), GaussianNoiseTransform( p_per_sample = 0.1, data_key = 'data', noise_variance = (0, 0.1), p_per_channel = 1, per_channel = False ), GaussianBlurTransform( p_per_sample = 0.2, different_sigma_per_channel = True, p_per_channel = 0.5, data_key = 'data', blur_sigma = (0.5, 1.0), different_sigma_per_axis = False, p_isotropic = 0 ), BrightnessMultiplicativeTransform( p_per_sample = 0.15, data_key = 'data', multiplier_range = (0.75, 1.25), per_channel = True ), ContrastAugmentationTransform( p_per_sample = 0.15, data_key = 'data', contrast_range = (0.75, 1.25), preserve_range = True, per_channel = True, p_per_channel = 1 ), SimulateLowResolutionTransform( order_upsample = 3, order_downsample = 0, channels = None, per_channel = True, p_per_channel = 0.5, p_per_sample = 0.25, data_key = 'data', zoom_range = (0.5, 1), ignore_axes = None ), GammaTransform( p_per_sample = 0.1, retain_stats = True, per_channel = True, data_key = 'data', gamma_range = (0.7, 1.5), invert_image = True ), GammaTransform( p_per_sample = 0.3, retain_stats = True, per_channel = True, data_key = 'data', gamma_range = (0.7, 1.5), invert_image = False ), MirrorTransform( p_per_sample = 1, data_key = 'data', label_key = 'seg', axes = (0, 1, 2) ), RemoveLabelTransform( output_key = 'seg', input_key = 'seg', replace_with = 0, remove_label = -1 ), RenameTransform( delete_old = True, out_key = 'target', in_key = 'seg' ), DownsampleSegForDSTransform2( axes = None, output_key = 'target', input_key = 'target', order = 0, ds_scales = [[1.0, 1.0, 1.0], [0.5, 0.5, 0.5], [0.25, 0.25, 0.25], [0.125, 0.125, 0.125], [0.0625, 0.0625, 0.0625]] ), NumpyToTensor( keys = ['data', 'target'], cast_to = 'float' )] )", | |
| "dataloader_val": "<nnunetv2.training.data_augmentation.custom_transforms.limited_length_multithreaded_augmenter.LimitedLenWrapper object at 0x7f1a6d312fb0>", | |
| "dataloader_val.generator": "<nnunetv2.training.dataloading.data_loader_3d.nnUNetDataLoader3D object at 0x7f1a6d3131c0>", | |
| "dataloader_val.num_processes": "8", | |
| "dataloader_val.transform": "Compose ( [RemoveLabelTransform( output_key = 'seg', input_key = 'seg', replace_with = 0, remove_label = -1 ), RenameTransform( delete_old = True, out_key = 'target', in_key = 'seg' ), DownsampleSegForDSTransform2( axes = None, output_key = 'target', input_key = 'target', order = 0, ds_scales = [[1.0, 1.0, 1.0], [0.5, 0.5, 0.5], [0.25, 0.25, 0.25], [0.125, 0.125, 0.125], [0.0625, 0.0625, 0.0625]] ), NumpyToTensor( keys = ['data', 'target'], cast_to = 'float' )] )", | |
| "dataset_json": "{'channel_names': {'0': 'CT'}, 'labels': {'background': 0, 'T1': 1, 'T2': 2, 'T3': 3, 'T4': 4, 'T5': 5, 'T6': 6, 'T7': 7, 'T8': 8, 'T9': 9, 'T10': 10, 'T11': 11, 'T12': 12, 'L1': 13, 'L2': 14, 'L3': 15, 'L4': 16, 'L5': 17, 'L6': 18, 'sacrum': 19, 'coccyx': 20, 'T13': 21}, 'numTraining': 1216, 'file_ending': '.nii.gz', 'dataset_name': 'VertebralBodies Thorax/Abdomen/Sacrum', 'reference': 'https://huggingface.co./datasets/fhofmann/VertebralBodiesCT-Labels', 'release': '2024-08-29', 'license': 'CC BY-SA 4.0', 'description': '1216 CT scans, with segmentations of the vertebral bodies in the thorax, abdomen, and sacrum. CT scans and initial labels derive from the VerSe (https://github.com/anjany/verse, license CC BY-SA 4.0) and TotalSegmentator (license CC BY 4.0 DEED) dataset. Files were processed, and labels were adapted as described in the README.md. A validation set is available separately.'}", | |
| "device": "cuda:0", | |
| "disable_checkpointing": "False", | |
| "enable_deep_supervision": "True", | |
| "fold": "2", | |
| "folder_with_segs_from_previous_stage": "None", | |
| "gpu_name": "NVIDIA A100-SXM4-40GB", | |
| "grad_scaler": "<torch.cuda.amp.grad_scaler.GradScaler object at 0x7f1b1fa6f550>", | |
| "hostname": "srvdrai1", | |
| "inference_allowed_mirroring_axes": "(0, 1, 2)", | |
| "initial_lr": "0.01", | |
| "is_cascaded": "False", | |
| "is_ddp": "False", | |
| "label_manager": "<nnunetv2.utilities.label_handling.label_handling.LabelManager object at 0x7f19a90daec0>", | |
| "local_rank": "0", | |
| "log_file": "/workspace/data/nnUNet_results/Dataset601_VertebralBodies/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres/fold_2/training_log_2024_9_3_10_35_47.txt", | |
| "logger": "<nnunetv2.training.logging.nnunet_logger.nnUNetLogger object at 0x7f19a88a35b0>", | |
| "loss": "DeepSupervisionWrapper(\n (loss): DC_and_CE_loss(\n (ce): RobustCrossEntropyLoss()\n (dc): MemoryEfficientSoftDiceLoss()\n )\n)", | |
| "lr_scheduler": "<nnunetv2.training.lr_scheduler.polylr.PolyLRScheduler object at 0x7f19a9a55270>", | |
| "my_init_kwargs": "{'plans': {'dataset_name': 'Dataset601_VertebralBodies', 'plans_name': 'nnUNetResEncUNetLPlans', 'original_median_spacing_after_transp': [1.5, 1.5, 1.5], 'original_median_shape_after_transp': [239, 252, 252], 'image_reader_writer': 'SimpleITKIO', 'transpose_forward': [0, 1, 2], 'transpose_backward': [0, 1, 2], 'configurations': {'2d': {'data_identifier': 'nnUNetPlans_2d', 'preprocessor_name': 'DefaultPreprocessor', 'batch_size': 136, 'patch_size': [256, 256], 'median_image_size_in_voxels': [240.0, 240.0], 'spacing': [1.5, 1.5], 'normalization_schemes': ['CTNormalization'], 'use_mask_for_norm': [False], 'resampling_fn_data': 'resample_data_or_seg_to_shape', 'resampling_fn_seg': 'resample_data_or_seg_to_shape', 'resampling_fn_data_kwargs': {'is_seg': False, 'order': 3, 'order_z': 0, 'force_separate_z': None}, 'resampling_fn_seg_kwargs': {'is_seg': True, 'order': 1, 'order_z': 0, 'force_separate_z': None}, 'resampling_fn_probabilities': 'resample_data_or_seg_to_shape', 'resampling_fn_probabilities_kwargs': {'is_seg': False, 'order': 1, 'order_z': 0, 'force_separate_z': None}, 'architecture': {'network_class_name': 'dynamic_network_architectures.architectures.unet.ResidualEncoderUNet', 'arch_kwargs': {'n_stages': 7, 'features_per_stage': [32, 64, 128, 256, 512, 512, 512], 'conv_op': 'torch.nn.modules.conv.Conv2d', 'kernel_sizes': [[3, 3], [3, 3], [3, 3], [3, 3], [3, 3], [3, 3], [3, 3]], 'strides': [[1, 1], [2, 2], [2, 2], [2, 2], [2, 2], [2, 2], [2, 2]], 'n_blocks_per_stage': [1, 3, 4, 6, 6, 6, 6], 'n_conv_per_stage_decoder': [1, 1, 1, 1, 1, 1], 'conv_bias': True, 'norm_op': 'torch.nn.modules.instancenorm.InstanceNorm2d', 'norm_op_kwargs': {'eps': 1e-05, 'affine': True}, 'dropout_op': None, 'dropout_op_kwargs': None, 'nonlin': 'torch.nn.LeakyReLU', 'nonlin_kwargs': {'inplace': True}}, '_kw_requires_import': ['conv_op', 'norm_op', 'dropout_op', 'nonlin']}, 'batch_dice': True}, '3d_fullres': {'data_identifier': 'nnUNetPlans_3d_fullres', 'preprocessor_name': 'DefaultPreprocessor', 'batch_size': 2, 'patch_size': [192, 192, 192], 'median_image_size_in_voxels': [235.0, 240.0, 240.0], 'spacing': [1.5, 1.5, 1.5], 'normalization_schemes': ['CTNormalization'], 'use_mask_for_norm': [False], 'resampling_fn_data': 'resample_data_or_seg_to_shape', 'resampling_fn_seg': 'resample_data_or_seg_to_shape', 'resampling_fn_data_kwargs': {'is_seg': False, 'order': 3, 'order_z': 0, 'force_separate_z': None}, 'resampling_fn_seg_kwargs': {'is_seg': True, 'order': 1, 'order_z': 0, 'force_separate_z': None}, 'resampling_fn_probabilities': 'resample_data_or_seg_to_shape', 'resampling_fn_probabilities_kwargs': {'is_seg': False, 'order': 1, 'order_z': 0, 'force_separate_z': None}, 'architecture': {'network_class_name': 'dynamic_network_architectures.architectures.unet.ResidualEncoderUNet', 'arch_kwargs': {'n_stages': 6, 'features_per_stage': [32, 64, 128, 256, 320, 320], 'conv_op': 'torch.nn.modules.conv.Conv3d', 'kernel_sizes': [[3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3]], 'strides': [[1, 1, 1], [2, 2, 2], [2, 2, 2], [2, 2, 2], [2, 2, 2], [2, 2, 2]], 'n_blocks_per_stage': [1, 3, 4, 6, 6, 6], 'n_conv_per_stage_decoder': [1, 1, 1, 1, 1], 'conv_bias': True, 'norm_op': 'torch.nn.modules.instancenorm.InstanceNorm3d', 'norm_op_kwargs': {'eps': 1e-05, 'affine': True}, 'dropout_op': None, 'dropout_op_kwargs': None, 'nonlin': 'torch.nn.LeakyReLU', 'nonlin_kwargs': {'inplace': True}}, '_kw_requires_import': ['conv_op', 'norm_op', 'dropout_op', 'nonlin']}, 'batch_dice': False}}, 'experiment_planner_used': 'nnUNetPlannerResEncL', 'label_manager': 'LabelManager', 'foreground_intensity_properties_per_channel': {'0': {'max': 10076.0, 'mean': 251.18613451837237, 'median': 216.0, 'min': -2048.0, 'percentile_00_5': -158.0, 'percentile_99_5': 1100.0, 'std': 205.1041738973246}}}, 'configuration': '3d_fullres', 'fold': 2, 'dataset_json': {'channel_names': {'0': 'CT'}, 'labels': {'background': 0, 'T1': 1, 'T2': 2, 'T3': 3, 'T4': 4, 'T5': 5, 'T6': 6, 'T7': 7, 'T8': 8, 'T9': 9, 'T10': 10, 'T11': 11, 'T12': 12, 'L1': 13, 'L2': 14, 'L3': 15, 'L4': 16, 'L5': 17, 'L6': 18, 'sacrum': 19, 'coccyx': 20, 'T13': 21}, 'numTraining': 1216, 'file_ending': '.nii.gz', 'dataset_name': 'VertebralBodies Thorax/Abdomen/Sacrum', 'reference': 'https://huggingface.co./datasets/fhofmann/VertebralBodiesCT-Labels', 'release': '2024-08-29', 'license': 'CC BY-SA 4.0', 'description': '1216 CT scans, with segmentations of the vertebral bodies in the thorax, abdomen, and sacrum. CT scans and initial labels derive from the VerSe (https://github.com/anjany/verse, license CC BY-SA 4.0) and TotalSegmentator (license CC BY 4.0 DEED) dataset. Files were processed, and labels were adapted as described in the README.md. A validation set is available separately.'}, 'unpack_dataset': False, 'device': device(type='cuda')}", | |
| "network": "OptimizedModule", | |
| "num_epochs": "1000", | |
| "num_input_channels": "1", | |
| "num_iterations_per_epoch": "250", | |
| "num_val_iterations_per_epoch": "50", | |
| "optimizer": "SGD (\nParameter Group 0\n dampening: 0\n differentiable: False\n foreach: None\n fused: None\n initial_lr: 0.01\n lr: 0.0012702500959998477\n maximize: False\n momentum: 0.99\n nesterov: True\n weight_decay: 3e-05\n)", | |
| "output_folder": "/workspace/data/nnUNet_results/Dataset601_VertebralBodies/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres/fold_2", | |
| "output_folder_base": "/workspace/data/nnUNet_results/Dataset601_VertebralBodies/nnUNetTrainer__nnUNetResEncUNetLPlans__3d_fullres", | |
| "oversample_foreground_percent": "0.33", | |
| "plans_manager": "{'dataset_name': 'Dataset601_VertebralBodies', 'plans_name': 'nnUNetResEncUNetLPlans', 'original_median_spacing_after_transp': [1.5, 1.5, 1.5], 'original_median_shape_after_transp': [239, 252, 252], 'image_reader_writer': 'SimpleITKIO', 'transpose_forward': [0, 1, 2], 'transpose_backward': [0, 1, 2], 'configurations': {'2d': {'data_identifier': 'nnUNetPlans_2d', 'preprocessor_name': 'DefaultPreprocessor', 'batch_size': 136, 'patch_size': [256, 256], 'median_image_size_in_voxels': [240.0, 240.0], 'spacing': [1.5, 1.5], 'normalization_schemes': ['CTNormalization'], 'use_mask_for_norm': [False], 'resampling_fn_data': 'resample_data_or_seg_to_shape', 'resampling_fn_seg': 'resample_data_or_seg_to_shape', 'resampling_fn_data_kwargs': {'is_seg': False, 'order': 3, 'order_z': 0, 'force_separate_z': None}, 'resampling_fn_seg_kwargs': {'is_seg': True, 'order': 1, 'order_z': 0, 'force_separate_z': None}, 'resampling_fn_probabilities': 'resample_data_or_seg_to_shape', 'resampling_fn_probabilities_kwargs': {'is_seg': False, 'order': 1, 'order_z': 0, 'force_separate_z': None}, 'architecture': {'network_class_name': 'dynamic_network_architectures.architectures.unet.ResidualEncoderUNet', 'arch_kwargs': {'n_stages': 7, 'features_per_stage': [32, 64, 128, 256, 512, 512, 512], 'conv_op': 'torch.nn.modules.conv.Conv2d', 'kernel_sizes': [[3, 3], [3, 3], [3, 3], [3, 3], [3, 3], [3, 3], [3, 3]], 'strides': [[1, 1], [2, 2], [2, 2], [2, 2], [2, 2], [2, 2], [2, 2]], 'n_blocks_per_stage': [1, 3, 4, 6, 6, 6, 6], 'n_conv_per_stage_decoder': [1, 1, 1, 1, 1, 1], 'conv_bias': True, 'norm_op': 'torch.nn.modules.instancenorm.InstanceNorm2d', 'norm_op_kwargs': {'eps': 1e-05, 'affine': True}, 'dropout_op': None, 'dropout_op_kwargs': None, 'nonlin': 'torch.nn.LeakyReLU', 'nonlin_kwargs': {'inplace': True}}, '_kw_requires_import': ['conv_op', 'norm_op', 'dropout_op', 'nonlin']}, 'batch_dice': True}, '3d_fullres': {'data_identifier': 'nnUNetPlans_3d_fullres', 'preprocessor_name': 'DefaultPreprocessor', 'batch_size': 2, 'patch_size': [192, 192, 192], 'median_image_size_in_voxels': [235.0, 240.0, 240.0], 'spacing': [1.5, 1.5, 1.5], 'normalization_schemes': ['CTNormalization'], 'use_mask_for_norm': [False], 'resampling_fn_data': 'resample_data_or_seg_to_shape', 'resampling_fn_seg': 'resample_data_or_seg_to_shape', 'resampling_fn_data_kwargs': {'is_seg': False, 'order': 3, 'order_z': 0, 'force_separate_z': None}, 'resampling_fn_seg_kwargs': {'is_seg': True, 'order': 1, 'order_z': 0, 'force_separate_z': None}, 'resampling_fn_probabilities': 'resample_data_or_seg_to_shape', 'resampling_fn_probabilities_kwargs': {'is_seg': False, 'order': 1, 'order_z': 0, 'force_separate_z': None}, 'architecture': {'network_class_name': 'dynamic_network_architectures.architectures.unet.ResidualEncoderUNet', 'arch_kwargs': {'n_stages': 6, 'features_per_stage': [32, 64, 128, 256, 320, 320], 'conv_op': 'torch.nn.modules.conv.Conv3d', 'kernel_sizes': [[3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3], [3, 3, 3]], 'strides': [[1, 1, 1], [2, 2, 2], [2, 2, 2], [2, 2, 2], [2, 2, 2], [2, 2, 2]], 'n_blocks_per_stage': [1, 3, 4, 6, 6, 6], 'n_conv_per_stage_decoder': [1, 1, 1, 1, 1], 'conv_bias': True, 'norm_op': 'torch.nn.modules.instancenorm.InstanceNorm3d', 'norm_op_kwargs': {'eps': 1e-05, 'affine': True}, 'dropout_op': None, 'dropout_op_kwargs': None, 'nonlin': 'torch.nn.LeakyReLU', 'nonlin_kwargs': {'inplace': True}}, '_kw_requires_import': ['conv_op', 'norm_op', 'dropout_op', 'nonlin']}, 'batch_dice': False}}, 'experiment_planner_used': 'nnUNetPlannerResEncL', 'label_manager': 'LabelManager', 'foreground_intensity_properties_per_channel': {'0': {'max': 10076.0, 'mean': 251.18613451837237, 'median': 216.0, 'min': -2048.0, 'percentile_00_5': -158.0, 'percentile_99_5': 1100.0, 'std': 205.1041738973246}}}", | |
| "preprocessed_dataset_folder": "/workspace/data2/nnUNet_preprocessed/Dataset601_VertebralBodies/nnUNetPlans_3d_fullres", | |
| "preprocessed_dataset_folder_base": "/workspace/data2/nnUNet_preprocessed/Dataset601_VertebralBodies", | |
| "save_every": "50", | |
| "torch_version": "2.3.0+cu121", | |
| "unpack_dataset": "False", | |
| "was_initialized": "True", | |
| "weight_decay": "3e-05" | |
| } |