diff --git a/MedNeXt_Base/fold_1/nnUNet_results/nnUNet/3d_fullres/Task001_Brain2026/nnUNetTrainerV2_MedNeXt_B_kernel3__nnUNetPlansv2.1_trgSp_1x1x1/fold_1/debug.json b/MedNeXt_Base/fold_1/nnUNet_results/nnUNet/3d_fullres/Task001_Brain2026/nnUNetTrainerV2_MedNeXt_B_kernel3__nnUNetPlansv2.1_trgSp_1x1x1/fold_1/debug.json
index 9670f5f38e8641370f6079e467f422ef57846f21..cb0cd6294bf3bb3ce91be7a715a4250cffe44536 100644
--- a/MedNeXt_Base/fold_1/nnUNet_results/nnUNet/3d_fullres/Task001_Brain2026/nnUNetTrainerV2_MedNeXt_B_kernel3__nnUNetPlansv2.1_trgSp_1x1x1/fold_1/debug.json
+++ b/MedNeXt_Base/fold_1/nnUNet_results/nnUNet/3d_fullres/Task001_Brain2026/nnUNetTrainerV2_MedNeXt_B_kernel3__nnUNetPlansv2.1_trgSp_1x1x1/fold_1/debug.json
@@ -1,28 +1,28 @@
 {
-    "all_tr_losses": "[]",
-    "all_val_eval_metrics": "[]",
-    "all_val_losses": "[]",
+    "all_tr_losses": "[np.float32(0.021483444), np.float32(-0.11146875), np.float32(-0.14440492), np.float32(-0.24443859), np.float32(-0.36445296), np.float32(-0.39437535), np.float32(-0.40792683), np.float32(-0.4200531), np.float32(-0.4325301), np.float32(-0.43809718), np.float32(-0.4546987), np.float32(-0.47324157), np.float32(-0.4904157), np.float32(-0.50101864), np.float32(-0.49778777), np.float32(-0.5230691), np.float32(-0.5323168), np.float32(-0.53293663), np.float32(-0.5560165), np.float32(-0.5554745), np.float32(-0.5598833), np.float32(-0.55329734), np.float32(-0.56848294), np.float32(-0.57045275), np.float32(-0.5853508), np.float32(-0.5702357), np.float32(-0.5773311), np.float32(-0.58233035), np.float32(-0.58506304), np.float32(-0.5965259), np.float32(-0.6032619), np.float32(-0.5900096), np.float32(-0.60379994), np.float32(-0.59322286), np.float32(-0.5969801), np.float32(-0.5948246), np.float32(-0.5960668), np.float32(-0.58386534), np.float32(-0.5938346), np.float32(-0.6074265), np.float32(-0.60020286), np.float32(-0.5970148), np.float32(-0.618273), np.float32(-0.6189575), np.float32(-0.6259663), np.float32(-0.59553635), np.float32(-0.6034077), np.float32(-0.6237228), np.float32(-0.6130356), np.float32(-0.6210732), np.float32(-0.627766), np.float32(-0.60405076), np.float32(-0.6267225), np.float32(-0.61242104), np.float32(-0.6308109), np.float32(-0.63114446), np.float32(-0.62956655), np.float32(-0.61518896), np.float32(-0.60810804), np.float32(-0.6202561), np.float32(-0.6277382), np.float32(-0.61818093), np.float32(-0.6260043), np.float32(-0.636891), np.float32(-0.6390251), np.float32(-0.63386804), np.float32(-0.6507668), np.float32(-0.62312436), np.float32(-0.6337467), np.float32(-0.6208022), np.float32(-0.6419503), np.float32(-0.6441393), np.float32(-0.64534456), np.float32(-0.651346), np.float32(-0.64969504), np.float32(-0.65728396), np.float32(-0.6617566), np.float32(-0.64839345), np.float32(-0.6385542), np.float32(-0.64664906), np.float32(-0.64798605), np.float32(-0.6399866), np.float32(-0.63610846), np.float32(-0.6543623), np.float32(-0.65984875), np.float32(-0.6527612), np.float32(-0.65395296), np.float32(-0.66415167), np.float32(-0.6626488), np.float32(-0.646508), np.float32(-0.66523826), np.float32(-0.6693098), np.float32(-0.6723793), np.float32(-0.6632665), np.float32(-0.66382325), np.float32(-0.6615557), np.float32(-0.6461383), np.float32(-0.6647244), np.float32(-0.6684059), np.float32(-0.6716227), np.float32(-0.64190996), np.float32(-0.6654336), np.float32(-0.6652307), np.float32(-0.6748037), np.float32(-0.6762971), np.float32(-0.66714954), np.float32(-0.6728248), np.float32(-0.6769093), np.float32(-0.66296697), np.float32(-0.6822878), np.float32(-0.67795974), np.float32(-0.66137075), np.float32(-0.68105775), np.float32(-0.68930906), np.float32(-0.67588466), np.float32(-0.66726613), np.float32(-0.6848649), np.float32(-0.6913262), np.float32(-0.6915221), np.float32(-0.69724184), np.float32(-0.6908957), np.float32(-0.70058465), np.float32(-0.7034883), np.float32(-0.6994904), np.float32(-0.7013602), np.float32(-0.70121604), np.float32(-0.6986172), np.float32(-0.69292444), np.float32(-0.7007394), np.float32(-0.6957776), np.float32(-0.6948666), np.float32(-0.71042305), np.float32(-0.7064369), np.float32(-0.6841923), np.float32(-0.6728657), np.float32(-0.6900254), np.float32(-0.6836837), np.float32(-0.702558), np.float32(-0.7069059), np.float32(-0.7163717), np.float32(-0.71552795), np.float32(-0.711771), np.float32(-0.7113856), np.float32(-0.70909387), np.float32(-0.6973179), np.float32(-0.72050416), np.float32(-0.7106394), np.float32(-0.7252629), np.float32(-0.7311752), np.float32(-0.73185265)]",
+    "all_val_eval_metrics": "[np.float32(0.00017466318), np.float32(0.021189876), np.float32(0.105327815), np.float32(0.17012873), np.float32(0.18574187), np.float32(0.18313855), np.float32(0.21423414), np.float32(0.25053057), np.float32(0.16533911), np.float32(0.1638105), np.float32(0.28763068), np.float32(0.29555872), np.float32(0.2795966), np.float32(0.23340246), np.float32(0.28143787), np.float32(0.30510038), np.float32(0.2975163), np.float32(0.22506697), np.float32(0.22173904), np.float32(0.2702406), np.float32(0.2119721), np.float32(0.31513065), np.float32(0.2769431), np.float32(0.35048038), np.float32(0.3656213), np.float32(0.23531525), np.float32(0.30753246), np.float32(0.3020315), np.float32(0.1313487), np.float32(0.34889838), np.float32(0.37516946), np.float32(0.3173487), np.float32(0.31577137), np.float32(0.25885922), np.float32(0.3185648), np.float32(0.2643943), np.float32(0.24780495), np.float32(0.2867595), np.float32(0.23933488), np.float32(0.11698307), np.float32(0.16100371), np.float32(0.1960939), np.float32(0.2927131), np.float32(0.2512934), np.float32(0.111189045), np.float32(0.20392255), np.float32(0.12993309), np.float32(0.16949207), np.float32(0.38088605), np.float32(0.28633344), np.float32(0.28359026), np.float32(0.26677704), np.float32(0.16854693), np.float32(0.23974474), np.float32(0.2150013), np.float32(0.17441328), np.float32(0.19348724), np.float32(0.1532411), np.float32(0.20337732), np.float32(0.15911943), np.float32(0.16662686), np.float32(0.33801007), np.float32(0.28375345), np.float32(0.2029703), np.float32(0.20145899), np.float32(0.3452817), np.float32(0.19129445), np.float32(0.31553423), np.float32(0.13373956), np.float32(0.38760865), np.float32(0.25676152), np.float32(0.2815446), np.float32(0.23934764), np.float32(0.43698105), np.float32(0.15255079), np.float32(0.21667838), np.float32(0.19700779), np.float32(0.15372533), np.float32(0.15868105), np.float32(0.22163062), np.float32(0.21999615), np.float32(0.22635971), np.float32(0.29574317), np.float32(0.26083583), np.float32(0.408208), np.float32(0.2779527), np.float32(0.2896338), np.float32(0.22233504), np.float32(0.22392589), np.float32(0.37460062), np.float32(0.18932822), np.float32(0.25886774), np.float32(0.22375137), np.float32(0.31505838), np.float32(0.26592386), np.float32(0.25933105), np.float32(0.14926209), np.float32(0.27998605), np.float32(0.3852249), np.float32(0.30196607), np.float32(0.2992246), np.float32(0.40994778), np.float32(0.317711), np.float32(0.14161514), np.float32(0.44781718), np.float32(0.16054086), np.float32(0.25375354), np.float32(0.40127105), np.float32(0.28625977), np.float32(0.30086374), np.float32(0.3788962), np.float32(0.28472254), np.float32(0.22942284), np.float32(0.26602206), np.float32(0.31659067), np.float32(0.33532563), np.float32(0.29273283), np.float32(0.3182021), np.float32(0.21795726), np.float32(0.28415138), np.float32(0.19119187), np.float32(0.3267253), np.float32(0.27092722), np.float32(0.18949632), np.float32(0.10407389), np.float32(0.26047018), np.float32(0.41401702), np.float32(0.19356596), np.float32(0.19335772), np.float32(0.28217435), np.float32(0.4017335), np.float32(0.27499413), np.float32(0.22865179), np.float32(0.38296163), np.float32(0.23218809), np.float32(0.33916786), np.float32(0.3674019), np.float32(0.2745694), np.float32(0.36488357), np.float32(0.40118495), np.float32(0.31819054), np.float32(0.33008382), np.float32(0.2493332), np.float32(0.23046756), np.float32(0.2740649), np.float32(0.42349446), np.float32(0.32972053), np.float32(0.2776941), np.float32(0.29219943), np.float32(0.2649039)]",
+    "all_val_losses": "[np.float32(-0.087872855), np.float32(-0.12076438), np.float32(-0.18725796), np.float32(-0.33754387), np.float32(-0.38727674), np.float32(-0.3948473), np.float32(-0.39974877), np.float32(-0.42445156), np.float32(-0.36016887), np.float32(-0.36184213), np.float32(-0.4620918), np.float32(-0.46901485), np.float32(-0.45250285), np.float32(-0.4167744), np.float32(-0.4505102), np.float32(-0.47962704), np.float32(-0.47389808), np.float32(-0.40671393), np.float32(-0.4008667), np.float32(-0.43510395), np.float32(-0.3981306), np.float32(-0.49479046), np.float32(-0.44799885), np.float32(-0.51504916), np.float32(-0.5302984), np.float32(-0.40885323), np.float32(-0.47965544), np.float32(-0.4842287), np.float32(-0.3270911), np.float32(-0.52332246), np.float32(-0.5439542), np.float32(-0.5093156), np.float32(-0.49581707), np.float32(-0.43374917), np.float32(-0.48983356), np.float32(-0.4491955), np.float32(-0.43180606), np.float32(-0.47305256), np.float32(-0.41450375), np.float32(-0.30283067), np.float32(-0.35823444), np.float32(-0.38787922), np.float32(-0.46296555), np.float32(-0.42216137), np.float32(-0.29296693), np.float32(-0.3872464), np.float32(-0.30031645), np.float32(-0.35371518), np.float32(-0.5583941), np.float32(-0.45297974), np.float32(-0.4295419), np.float32(-0.45082733), np.float32(-0.3483701), np.float32(-0.40606102), np.float32(-0.40274486), np.float32(-0.35274565), np.float32(-0.3701127), np.float32(-0.3245539), np.float32(-0.38491765), np.float32(-0.34316313), np.float32(-0.35335147), np.float32(-0.50617456), np.float32(-0.45387906), np.float32(-0.3894697), np.float32(-0.38632268), np.float32(-0.52456903), np.float32(-0.35850555), np.float32(-0.5004133), np.float32(-0.29260185), np.float32(-0.5649532), np.float32(-0.4304586), np.float32(-0.4512955), np.float32(-0.40949437), np.float32(-0.6165686), np.float32(-0.3309762), np.float32(-0.37366322), np.float32(-0.36406085), np.float32(-0.30982864), np.float32(-0.32582405), np.float32(-0.39785114), np.float32(-0.38767096), np.float32(-0.39887384), np.float32(-0.46052358), np.float32(-0.44831702), np.float32(-0.5873836), np.float32(-0.46554738), np.float32(-0.4751033), np.float32(-0.3944073), np.float32(-0.38925168), np.float32(-0.54067004), np.float32(-0.35715508), np.float32(-0.41771403), np.float32(-0.38553268), np.float32(-0.4806604), np.float32(-0.43360373), np.float32(-0.40782943), np.float32(-0.31121215), np.float32(-0.4377116), np.float32(-0.5537004), np.float32(-0.47155255), np.float32(-0.4625329), np.float32(-0.58026963), np.float32(-0.4771339), np.float32(-0.31337082), np.float32(-0.60579973), np.float32(-0.30740076), np.float32(-0.40346217), np.float32(-0.56743747), np.float32(-0.43498597), np.float32(-0.4449178), np.float32(-0.53000396), np.float32(-0.42801893), np.float32(-0.3883855), np.float32(-0.41310275), np.float32(-0.47026372), np.float32(-0.4916751), np.float32(-0.44708663), np.float32(-0.428909), np.float32(-0.3365038), np.float32(-0.40743464), np.float32(-0.30969992), np.float32(-0.47231537), np.float32(-0.3916529), np.float32(-0.32978204), np.float32(-0.2376054), np.float32(-0.39156032), np.float32(-0.56153387), np.float32(-0.33491096), np.float32(-0.2988381), np.float32(-0.41964844), np.float32(-0.53775907), np.float32(-0.41336337), np.float32(-0.36073086), np.float32(-0.5129612), np.float32(-0.3589186), np.float32(-0.46774712), np.float32(-0.50133955), np.float32(-0.40900335), np.float32(-0.48906574), np.float32(-0.530339), np.float32(-0.45159504), np.float32(-0.44846395), np.float32(-0.37163335), np.float32(-0.3672916), np.float32(-0.40524933), np.float32(-0.5695139), np.float32(-0.4430707), np.float32(-0.39538956), np.float32(-0.40970996), np.float32(-0.37603873)]",
     "all_val_losses_tr_mode": "[]",
     "also_val_in_tr_mode": "False",
-    "amp_grad_scaler": "None",
+    "amp_grad_scaler": "<torch.cuda.amp.grad_scaler.GradScaler object at 0x7f06d4850ec0>",
     "base_num_features": "30",
     "basic_generator_patch_size": "[205 205 205]",
     "batch_dice": "False",
     "batch_size": "2",
-    "best_MA_tr_loss_for_patience": "None",
-    "best_epoch_based_on_MA_tr_loss": "None",
-    "best_val_eval_criterion_MA": "None",
+    "best_MA_tr_loss_for_patience": "-0.70342416",
+    "best_epoch_based_on_MA_tr_loss": "148",
+    "best_val_eval_criterion_MA": "0.3105384",
     "classes": "[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36]",
     "conv_per_stage": "2",
     "data_aug_params": "{'selected_data_channels': None, 'selected_seg_channels': [0], 'do_elastic': False, 'elastic_deform_alpha': (0.0, 900.0), 'elastic_deform_sigma': (9.0, 13.0), 'p_eldef': 0.2, 'do_scaling': True, 'scale_range': (0.7, 1.4), 'independent_scale_factor_for_each_axis': False, 'p_independent_scale_per_axis': 1, 'p_scale': 0.2, 'do_rotation': True, 'rotation_x': (-0.5235987755982988, 0.5235987755982988), 'rotation_y': (-0.5235987755982988, 0.5235987755982988), 'rotation_z': (-0.5235987755982988, 0.5235987755982988), 'rotation_p_per_axis': 1, 'p_rot': 0.2, 'random_crop': False, 'random_crop_dist_to_border': None, 'do_gamma': True, 'gamma_retain_stats': True, 'gamma_range': (0.7, 1.5), 'p_gamma': 0.3, 'do_mirror': True, 'mirror_axes': (0, 1, 2), 'dummy_2D': False, 'mask_was_used_for_normalization': OrderedDict({0: False}), 'border_mode_data': 'constant', 'all_segmentation_labels': None, 'move_last_seg_chanel_to_data': False, 'cascade_do_cascade_augmentations': False, 'cascade_random_binary_transform_p': 0.4, 'cascade_random_binary_transform_p_per_label': 1, 'cascade_random_binary_transform_size': (1, 8), 'cascade_remove_conn_comp_p': 0.2, 'cascade_remove_conn_comp_max_size_percent_threshold': 0.15, 'cascade_remove_conn_comp_fill_with_other_class_p': 0.0, 'do_additive_brightness': False, 'additive_brightness_p_per_sample': 0.15, 'additive_brightness_p_per_channel': 0.5, 'additive_brightness_mu': 0.0, 'additive_brightness_sigma': 0.1, 'num_threads': 8, 'num_cached_per_thread': 2, 'patch_size_for_spatialtransform': array([128, 128, 128])}",
     "dataset_directory": "/marimo/TopBrain2026/nnUNet_preprocessed/Task001_Brain2026",
     "deep_supervision_scales": "[[1, 1, 1], [np.float64(0.5), np.float64(0.5), np.float64(0.5)], [np.float64(0.25), np.float64(0.25), np.float64(0.25)], [np.float64(0.125), np.float64(0.125), np.float64(0.125)], [np.float64(0.0625), np.float64(0.0625), np.float64(0.0625)], [np.float64(0.03125), np.float64(0.03125), np.float64(0.03125)]]",
     "deterministic": "False",
-    "dl_tr": "<nnunet_mednext.training.dataloading.dataset_loading.DataLoader3D object at 0x7f02c2c5a270>",
-    "dl_val": "<nnunet_mednext.training.dataloading.dataset_loading.DataLoader3D object at 0x7f02c2a7ce10>",
+    "dl_tr": "<nnunet_mednext.training.dataloading.dataset_loading.DataLoader3D object at 0x7f06d4a4e3c0>",
+    "dl_val": "<nnunet_mednext.training.dataloading.dataset_loading.DataLoader3D object at 0x7f06d487ce10>",
     "do_dummy_2D_aug": "False",
     "ds_loss_weights": "[0.51612903 0.25806452 0.12903226 0.06451613 0.03225806 0.        ]",
-    "epoch": "0",
+    "epoch": "150",
     "experiment_name": "nnUNetTrainerV2_MedNeXt_B_kernel3",
     "fold": "1",
     "folder_with_preprocessed_data": "/marimo/TopBrain2026/nnUNet_preprocessed/Task001_Brain2026/nnUNetData_plans_v2.1_trgSp_1x1x1_stage0",
@@ -32,7 +32,7 @@
     "inference_pad_kwargs": "{'constant_values': 0}",
     "init_args": "('/marimo/TopBrain2026/nnUNet_preprocessed/Task001_Brain2026/nnUNetPlansv2.1_trgSp_1x1x1_plans_3D.pkl', 1, '/marimo/TopBrain2026/nnUNet_results/nnUNet/3d_fullres/Task001_Brain2026/nnUNetTrainerV2_MedNeXt_B_kernel3__nnUNetPlansv2.1_trgSp_1x1x1', '/marimo/TopBrain2026/nnUNet_preprocessed/Task001_Brain2026', False, 0, True, False, True)",
     "initial_lr": "0.001",
-    "log_file": "/marimo/TopBrain2026/nnUNet_results/nnUNet/3d_fullres/Task001_Brain2026/nnUNetTrainerV2_MedNeXt_B_kernel3__nnUNetPlansv2.1_trgSp_1x1x1/fold_1/training_log_2026_10_16_18_36_02.txt",
+    "log_file": "/marimo/TopBrain2026/nnUNet_results/nnUNet/3d_fullres/Task001_Brain2026/nnUNetTrainerV2_MedNeXt_B_kernel3__nnUNetPlansv2.1_trgSp_1x1x1/fold_1/training_log_2026_10_17_06_02_05.txt",
     "lr_scheduler": "None",
     "lr_scheduler_eps": "0.001",
     "lr_scheduler_patience": "30",
@@ -53,7 +53,7 @@
     "online_eval_fp": "[]",
     "online_eval_tp": "[]",
     "only_keep_largest_connected_component": "None",
-    "optimizer": "AdamW (\nParameter Group 0\n    amsgrad: False\n    betas: (0.9, 0.999)\n    capturable: False\n    decoupled_weight_decay: True\n    differentiable: False\n    eps: 0.0001\n    foreach: None\n    fused: None\n    lr: 0.001\n    maximize: False\n    weight_decay: 3e-05\n)",
+    "optimizer": "AdamW (\nParameter Group 0\n    amsgrad: False\n    betas: (0.9, 0.999)\n    capturable: False\n    decoupled_weight_decay: True\n    differentiable: False\n    eps: 0.0001\n    foreach: None\n    fused: None\n    lr: 0.0008639269723028191\n    maximize: False\n    weight_decay: 3e-05\n)",
     "output_folder": "/marimo/TopBrain2026/nnUNet_results/nnUNet/3d_fullres/Task001_Brain2026/nnUNetTrainerV2_MedNeXt_B_kernel3__nnUNetPlansv2.1_trgSp_1x1x1/fold_1",
     "output_folder_base": "/marimo/TopBrain2026/nnUNet_results/nnUNet/3d_fullres/Task001_Brain2026/nnUNetTrainerV2_MedNeXt_B_kernel3__nnUNetPlansv2.1_trgSp_1x1x1",
     "oversample_foreground_percent": "0.33",
@@ -71,7 +71,7 @@
     "save_latest_only": "True",
     "stage": "0",
     "threeD": "True",
-    "tr_gen": "<batchgenerators.dataloading.multi_threaded_augmenter.MultiThreadedAugmenter object at 0x7f02c2c5b8c0>",
+    "tr_gen": "<batchgenerators.dataloading.multi_threaded_augmenter.MultiThreadedAugmenter object at 0x7f06d4a4fa10>",
     "train_loss_MA": "None",
     "train_loss_MA_alpha": "0.93",
     "train_loss_MA_eps": "0.0005",
@@ -82,7 +82,7 @@
     "use_progress_bar": "False",
     "val_eval_criterion_MA": "None",
     "val_eval_criterion_alpha": "0.9",
-    "val_gen": "<batchgenerators.dataloading.multi_threaded_augmenter.MultiThreadedAugmenter object at 0x7f02c2a7dbd0>",
+    "val_gen": "<batchgenerators.dataloading.multi_threaded_augmenter.MultiThreadedAugmenter object at 0x7f06d487dbd0>",
     "was_initialized": "True",
     "weight_decay": "3e-05"
 }
\ No newline at end of file
diff --git a/MedNeXt_Base/fold_1/nnUNet_results/nnUNet/3d_fullres/Task001_Brain2026/nnUNetTrainerV2_MedNeXt_B_kernel3__nnUNetPlansv2.1_trgSp_1x1x1/fold_1/model_latest.model b/MedNeXt_Base/fold_1/nnUNet_results/nnUNet/3d_fullres/Task001_Brain2026/nnUNetTrainerV2_MedNeXt_B_kernel3__nnUNetPlansv2.1_trgSp_1x1x1/fold_1/model_latest.model
index 7469ba59a16a779c0754367ca30471596403adf6..3d55d93fe4a7210fc10346e7ffa1b43f1f5dbd5c 100644
--- a/MedNeXt_Base/fold_1/nnUNet_results/nnUNet/3d_fullres/Task001_Brain2026/nnUNetTrainerV2_MedNeXt_B_kernel3__nnUNetPlansv2.1_trgSp_1x1x1/fold_1/model_latest.model
+++ b/MedNeXt_Base/fold_1/nnUNet_results/nnUNet/3d_fullres/Task001_Brain2026/nnUNetTrainerV2_MedNeXt_B_kernel3__nnUNetPlansv2.1_trgSp_1x1x1/fold_1/model_latest.model
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f198d0f2b306fb28987952ed6690304d3f978f005022daa2b5198f856946e0b8
-size 127068464
+oid sha256:587877add181ff571a1bd37d13194a5549fd8b5d541c7f8a68cd6419c3f01422
+size 127095472
diff --git a/MedNeXt_Base/fold_1/nnUNet_results/nnUNet/3d_fullres/Task001_Brain2026/nnUNetTrainerV2_MedNeXt_B_kernel3__nnUNetPlansv2.1_trgSp_1x1x1/fold_1/progress.png b/MedNeXt_Base/fold_1/nnUNet_results/nnUNet/3d_fullres/Task001_Brain2026/nnUNetTrainerV2_MedNeXt_B_kernel3__nnUNetPlansv2.1_trgSp_1x1x1/fold_1/progress.png
index 96b099f565878f64006e36f5333fe11312e75dd6..cdd1b54bc4429872a12009d7454d721e61944f0e 100644
--- a/MedNeXt_Base/fold_1/nnUNet_results/nnUNet/3d_fullres/Task001_Brain2026/nnUNetTrainerV2_MedNeXt_B_kernel3__nnUNetPlansv2.1_trgSp_1x1x1/fold_1/progress.png
+++ b/MedNeXt_Base/fold_1/nnUNet_results/nnUNet/3d_fullres/Task001_Brain2026/nnUNetTrainerV2_MedNeXt_B_kernel3__nnUNetPlansv2.1_trgSp_1x1x1/fold_1/progress.png
@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d4f5ae65bd6346508d9b956a4962732a05dfb67c92caa29fb65f9abe37808f2e
-size 607004
+oid sha256:b50596d26bca0a249fea0fd3c545baad6bf8b7cc7fdfbe45215b3f0626071ccc
+size 826898
diff --git a/MedNeXt_Base/fold_1/nnUNet_results/nnUNet/3d_fullres/Task001_Brain2026/nnUNetTrainerV2_MedNeXt_B_kernel3__nnUNetPlansv2.1_trgSp_1x1x1/fold_1/training_log_2026_10_17_06_02_05.txt b/MedNeXt_Base/fold_1/nnUNet_results/nnUNet/3d_fullres/Task001_Brain2026/nnUNetTrainerV2_MedNeXt_B_kernel3__nnUNetPlansv2.1_trgSp_1x1x1/fold_1/training_log_2026_10_17_06_02_05.txt
new file mode 100644
index 0000000000000000000000000000000000000000..8bf336649004464da54d9c41f26f395540f67be2
--- /dev/null
+++ b/MedNeXt_Base/fold_1/nnUNet_results/nnUNet/3d_fullres/Task001_Brain2026/nnUNetTrainerV2_MedNeXt_B_kernel3__nnUNetPlansv2.1_trgSp_1x1x1/fold_1/training_log_2026_10_17_06_02_05.txt
@@ -0,0 +1,1625 @@
+Starting... 
+2026-10-17 06:02:05.030477: Creating new 5-fold cross-validation split... 
+2026-10-17 06:02:05.031372: Desired fold for training: 1 
+2026-10-17 06:02:05.031445: This split has 20 training and 5 validation cases. 
+2026-10-17 06:02:05.685766: TRAINING KEYS:
+ odict_keys([np.str_('topcow_ct_001'), np.str_('topcow_ct_002'), np.str_('topcow_ct_003'), np.str_('topcow_ct_005'), np.str_('topcow_ct_006'), np.str_('topcow_ct_007'), np.str_('topcow_ct_008'), np.str_('topcow_ct_010'), np.str_('topcow_ct_011'), np.str_('topcow_ct_013'), np.str_('topcow_ct_015'), np.str_('topcow_ct_016'), np.str_('topcow_ct_018'), np.str_('topcow_ct_021'), np.str_('topcow_ct_022'), np.str_('topcow_ct_023'), np.str_('topcow_ct_024'), np.str_('topcow_ct_025'), np.str_('topcow_ct_026'), np.str_('topcow_ct_027')]) 
+2026-10-17 06:02:05.686106: VALIDATION KEYS:
+ odict_keys([np.str_('topcow_ct_004'), np.str_('topcow_ct_012'), np.str_('topcow_ct_014'), np.str_('topcow_ct_017'), np.str_('topcow_ct_020')]) 
+2026-10-17 06:02:07.347900: loading checkpoint /marimo/TopBrain2026/nnUNet_results/nnUNet/3d_fullres/Task001_Brain2026/nnUNetTrainerV2_MedNeXt_B_kernel3__nnUNetPlansv2.1_trgSp_1x1x1/fold_1/model_latest.model train= True 
+2026-10-17 06:02:07.500826: lr: 0.000864 
+2026-10-17 06:02:10.915132: Unable to plot network architecture: 
+2026-10-17 06:02:10.915696: No module named 'hiddenlayer' 
+2026-10-17 06:02:10.916665: 
+printing the network instead:
+ 
+2026-10-17 06:02:10.917547: MedNeXt(
+  (stem): Conv3d(1, 32, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+  (enc_block_0): Sequential(
+    (0): MedNeXtBlock(
+      (conv1): Conv3d(32, 32, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), groups=32)
+      (norm): GroupNorm(32, 32, eps=1e-05, affine=True)
+      (conv2): Conv3d(32, 64, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+      (act): GELU(approximate='none')
+      (conv3): Conv3d(64, 32, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    )
+    (1): MedNeXtBlock(
+      (conv1): Conv3d(32, 32, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), groups=32)
+      (norm): GroupNorm(32, 32, eps=1e-05, affine=True)
+      (conv2): Conv3d(32, 64, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+      (act): GELU(approximate='none')
+      (conv3): Conv3d(64, 32, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    )
+  )
+  (down_0): MedNeXtDownBlock(
+    (conv1): Conv3d(32, 32, kernel_size=(3, 3, 3), stride=(2, 2, 2), padding=(1, 1, 1), groups=32)
+    (norm): GroupNorm(32, 32, eps=1e-05, affine=True)
+    (conv2): Conv3d(32, 96, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    (act): GELU(approximate='none')
+    (conv3): Conv3d(96, 64, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    (res_conv): Conv3d(32, 64, kernel_size=(1, 1, 1), stride=(2, 2, 2))
+  )
+  (enc_block_1): Sequential(
+    (0): MedNeXtBlock(
+      (conv1): Conv3d(64, 64, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), groups=64)
+      (norm): GroupNorm(64, 64, eps=1e-05, affine=True)
+      (conv2): Conv3d(64, 192, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+      (act): GELU(approximate='none')
+      (conv3): Conv3d(192, 64, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    )
+    (1): MedNeXtBlock(
+      (conv1): Conv3d(64, 64, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), groups=64)
+      (norm): GroupNorm(64, 64, eps=1e-05, affine=True)
+      (conv2): Conv3d(64, 192, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+      (act): GELU(approximate='none')
+      (conv3): Conv3d(192, 64, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    )
+  )
+  (down_1): MedNeXtDownBlock(
+    (conv1): Conv3d(64, 64, kernel_size=(3, 3, 3), stride=(2, 2, 2), padding=(1, 1, 1), groups=64)
+    (norm): GroupNorm(64, 64, eps=1e-05, affine=True)
+    (conv2): Conv3d(64, 256, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    (act): GELU(approximate='none')
+    (conv3): Conv3d(256, 128, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    (res_conv): Conv3d(64, 128, kernel_size=(1, 1, 1), stride=(2, 2, 2))
+  )
+  (enc_block_2): Sequential(
+    (0): MedNeXtBlock(
+      (conv1): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), groups=128)
+      (norm): GroupNorm(128, 128, eps=1e-05, affine=True)
+      (conv2): Conv3d(128, 512, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+      (act): GELU(approximate='none')
+      (conv3): Conv3d(512, 128, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    )
+    (1): MedNeXtBlock(
+      (conv1): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), groups=128)
+      (norm): GroupNorm(128, 128, eps=1e-05, affine=True)
+      (conv2): Conv3d(128, 512, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+      (act): GELU(approximate='none')
+      (conv3): Conv3d(512, 128, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    )
+  )
+  (down_2): MedNeXtDownBlock(
+    (conv1): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(2, 2, 2), padding=(1, 1, 1), groups=128)
+    (norm): GroupNorm(128, 128, eps=1e-05, affine=True)
+    (conv2): Conv3d(128, 512, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    (act): GELU(approximate='none')
+    (conv3): Conv3d(512, 256, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    (res_conv): Conv3d(128, 256, kernel_size=(1, 1, 1), stride=(2, 2, 2))
+  )
+  (enc_block_3): Sequential(
+    (0): MedNeXtBlock(
+      (conv1): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), groups=256)
+      (norm): GroupNorm(256, 256, eps=1e-05, affine=True)
+      (conv2): Conv3d(256, 1024, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+      (act): GELU(approximate='none')
+      (conv3): Conv3d(1024, 256, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    )
+    (1): MedNeXtBlock(
+      (conv1): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), groups=256)
+      (norm): GroupNorm(256, 256, eps=1e-05, affine=True)
+      (conv2): Conv3d(256, 1024, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+      (act): GELU(approximate='none')
+      (conv3): Conv3d(1024, 256, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    )
+  )
+  (down_3): MedNeXtDownBlock(
+    (conv1): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(2, 2, 2), padding=(1, 1, 1), groups=256)
+    (norm): GroupNorm(256, 256, eps=1e-05, affine=True)
+    (conv2): Conv3d(256, 1024, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    (act): GELU(approximate='none')
+    (conv3): Conv3d(1024, 512, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    (res_conv): Conv3d(256, 512, kernel_size=(1, 1, 1), stride=(2, 2, 2))
+  )
+  (bottleneck): Sequential(
+    (0): MedNeXtBlock(
+      (conv1): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), groups=512)
+      (norm): GroupNorm(512, 512, eps=1e-05, affine=True)
+      (conv2): Conv3d(512, 2048, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+      (act): GELU(approximate='none')
+      (conv3): Conv3d(2048, 512, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    )
+    (1): MedNeXtBlock(
+      (conv1): Conv3d(512, 512, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), groups=512)
+      (norm): GroupNorm(512, 512, eps=1e-05, affine=True)
+      (conv2): Conv3d(512, 2048, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+      (act): GELU(approximate='none')
+      (conv3): Conv3d(2048, 512, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    )
+  )
+  (up_3): MedNeXtUpBlock(
+    (conv1): ConvTranspose3d(512, 512, kernel_size=(3, 3, 3), stride=(2, 2, 2), padding=(1, 1, 1), groups=512)
+    (norm): GroupNorm(512, 512, eps=1e-05, affine=True)
+    (conv2): Conv3d(512, 2048, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    (act): GELU(approximate='none')
+    (conv3): Conv3d(2048, 256, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    (res_conv): ConvTranspose3d(512, 256, kernel_size=(1, 1, 1), stride=(2, 2, 2))
+  )
+  (dec_block_3): Sequential(
+    (0): MedNeXtBlock(
+      (conv1): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), groups=256)
+      (norm): GroupNorm(256, 256, eps=1e-05, affine=True)
+      (conv2): Conv3d(256, 1024, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+      (act): GELU(approximate='none')
+      (conv3): Conv3d(1024, 256, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    )
+    (1): MedNeXtBlock(
+      (conv1): Conv3d(256, 256, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), groups=256)
+      (norm): GroupNorm(256, 256, eps=1e-05, affine=True)
+      (conv2): Conv3d(256, 1024, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+      (act): GELU(approximate='none')
+      (conv3): Conv3d(1024, 256, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    )
+  )
+  (up_2): MedNeXtUpBlock(
+    (conv1): ConvTranspose3d(256, 256, kernel_size=(3, 3, 3), stride=(2, 2, 2), padding=(1, 1, 1), groups=256)
+    (norm): GroupNorm(256, 256, eps=1e-05, affine=True)
+    (conv2): Conv3d(256, 1024, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    (act): GELU(approximate='none')
+    (conv3): Conv3d(1024, 128, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    (res_conv): ConvTranspose3d(256, 128, kernel_size=(1, 1, 1), stride=(2, 2, 2))
+  )
+  (dec_block_2): Sequential(
+    (0): MedNeXtBlock(
+      (conv1): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), groups=128)
+      (norm): GroupNorm(128, 128, eps=1e-05, affine=True)
+      (conv2): Conv3d(128, 512, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+      (act): GELU(approximate='none')
+      (conv3): Conv3d(512, 128, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    )
+    (1): MedNeXtBlock(
+      (conv1): Conv3d(128, 128, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), groups=128)
+      (norm): GroupNorm(128, 128, eps=1e-05, affine=True)
+      (conv2): Conv3d(128, 512, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+      (act): GELU(approximate='none')
+      (conv3): Conv3d(512, 128, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    )
+  )
+  (up_1): MedNeXtUpBlock(
+    (conv1): ConvTranspose3d(128, 128, kernel_size=(3, 3, 3), stride=(2, 2, 2), padding=(1, 1, 1), groups=128)
+    (norm): GroupNorm(128, 128, eps=1e-05, affine=True)
+    (conv2): Conv3d(128, 384, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    (act): GELU(approximate='none')
+    (conv3): Conv3d(384, 64, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    (res_conv): ConvTranspose3d(128, 64, kernel_size=(1, 1, 1), stride=(2, 2, 2))
+  )
+  (dec_block_1): Sequential(
+    (0): MedNeXtBlock(
+      (conv1): Conv3d(64, 64, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), groups=64)
+      (norm): GroupNorm(64, 64, eps=1e-05, affine=True)
+      (conv2): Conv3d(64, 192, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+      (act): GELU(approximate='none')
+      (conv3): Conv3d(192, 64, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    )
+    (1): MedNeXtBlock(
+      (conv1): Conv3d(64, 64, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), groups=64)
+      (norm): GroupNorm(64, 64, eps=1e-05, affine=True)
+      (conv2): Conv3d(64, 192, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+      (act): GELU(approximate='none')
+      (conv3): Conv3d(192, 64, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    )
+  )
+  (up_0): MedNeXtUpBlock(
+    (conv1): ConvTranspose3d(64, 64, kernel_size=(3, 3, 3), stride=(2, 2, 2), padding=(1, 1, 1), groups=64)
+    (norm): GroupNorm(64, 64, eps=1e-05, affine=True)
+    (conv2): Conv3d(64, 128, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    (act): GELU(approximate='none')
+    (conv3): Conv3d(128, 32, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    (res_conv): ConvTranspose3d(64, 32, kernel_size=(1, 1, 1), stride=(2, 2, 2))
+  )
+  (dec_block_0): Sequential(
+    (0): MedNeXtBlock(
+      (conv1): Conv3d(32, 32, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), groups=32)
+      (norm): GroupNorm(32, 32, eps=1e-05, affine=True)
+      (conv2): Conv3d(32, 64, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+      (act): GELU(approximate='none')
+      (conv3): Conv3d(64, 32, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    )
+    (1): MedNeXtBlock(
+      (conv1): Conv3d(32, 32, kernel_size=(3, 3, 3), stride=(1, 1, 1), padding=(1, 1, 1), groups=32)
+      (norm): GroupNorm(32, 32, eps=1e-05, affine=True)
+      (conv2): Conv3d(32, 64, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+      (act): GELU(approximate='none')
+      (conv3): Conv3d(64, 32, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+    )
+  )
+  (out_0): OutBlock(
+    (conv_out): ConvTranspose3d(32, 37, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+  )
+  (out_1): OutBlock(
+    (conv_out): ConvTranspose3d(64, 37, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+  )
+  (out_2): OutBlock(
+    (conv_out): ConvTranspose3d(128, 37, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+  )
+  (out_3): OutBlock(
+    (conv_out): ConvTranspose3d(256, 37, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+  )
+  (out_4): OutBlock(
+    (conv_out): ConvTranspose3d(512, 37, kernel_size=(1, 1, 1), stride=(1, 1, 1))
+  )
+) 
+2026-10-17 06:02:10.948024: 
+ 
+2026-10-17 06:02:10.948387: 
+epoch:  150 
+2026-10-17 06:06:16.409323: train loss : -0.7348 
+2026-10-17 06:06:21.785041: validation loss: -0.3910 
+2026-10-17 06:06:21.786124: Average global foreground Dice: [np.float32(0.8606), np.float32(0.336), np.float32(0.3301), np.float32(0.3803), np.float32(0.2749), np.float32(0.3626), np.float32(0.3388), np.float32(0.0), np.float32(0.0), np.float32(0.5837), np.float32(0.3223), np.float32(0.3015), np.float32(0.3417), np.float32(0.3498), np.float32(0.0), np.float32(0.0), np.float32(0.2712), np.float32(0.2145), np.float32(0.3256), np.float32(0.3044), np.float32(0.2909), np.float32(0.2926), np.float32(0.3736), np.float32(0.4138), np.float32(0.1882), np.float32(0.2318), np.float32(0.0), np.float32(0.0), np.float32(0.1103), np.float32(0.3632), np.float32(0.0), np.float32(0.0), np.float32(0.1542), np.float32(0.1396), np.float32(0.3242), np.float32(0.332)] 
+2026-10-17 06:06:21.786544: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 06:06:22.174720: lr: 0.000863 
+2026-10-17 06:06:22.175024: This epoch took 251.226304 s
+ 
+2026-10-17 06:06:22.175122: 
+epoch:  151 
+2026-10-17 06:10:39.575393: train loss : -0.7208 
+2026-10-17 06:10:44.917264: validation loss: -0.3883 
+2026-10-17 06:10:44.918327: Average global foreground Dice: [np.float32(0.8668), np.float32(0.3746), np.float32(0.3851), np.float32(0.429), np.float32(0.3632), np.float32(0.4325), np.float32(0.3678), np.float32(0.0), np.float32(0.0), np.float32(0.5568), np.float32(0.3183), np.float32(0.377), np.float32(0.2216), np.float32(0.2514), np.float32(0.0), np.float32(0.0), np.float32(0.3678), np.float32(0.2592), np.float32(0.3263), np.float32(0.3195), np.float32(0.0395), np.float32(0.248), np.float32(0.6441), np.float32(0.4392), np.float32(0.1575), np.float32(0.1767), np.float32(0.0), np.float32(0.0), np.float32(0.2002), np.float32(0.1725), np.float32(0.0), np.float32(0.0), np.float32(0.1627), np.float32(0.1963), np.float32(0.4779), np.float32(0.412)] 
+2026-10-17 06:10:44.918523: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 06:10:45.290642: lr: 0.000862 
+2026-10-17 06:10:45.290958: This epoch took 263.115765 s
+ 
+2026-10-17 06:10:45.291256: 
+epoch:  152 
+2026-10-17 06:14:38.149299: train loss : -0.7314 
+2026-10-17 06:14:43.494579: validation loss: -0.2506 
+2026-10-17 06:14:43.495478: Average global foreground Dice: [np.float32(0.8595), np.float32(0.1237), np.float32(0.1532), np.float32(0.1769), np.float32(0.1218), np.float32(0.1276), np.float32(0.1445), np.float32(0.0), np.float32(0.0), np.float32(0.6566), np.float32(0.1795), np.float32(0.1323), np.float32(0.1878), np.float32(0.1713), np.float32(0.0), np.float32(0.0), np.float32(0.1437), np.float32(0.115), np.float32(0.1258), np.float32(0.118), np.float32(0.0447), np.float32(0.101), np.float32(0.131), np.float32(0.2857), np.float32(0.0421), np.float32(0.0466), np.float32(0.0), np.float32(0.0), np.float32(0.0457), np.float32(0.0304), np.float32(0.0), np.float32(0.0), np.float32(0.1084), np.float32(0.0839), np.float32(0.1515), np.float32(0.1495)] 
+2026-10-17 06:14:43.495640: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 06:14:43.810411: lr: 0.000861 
+2026-10-17 06:14:43.810714: This epoch took 238.519241 s
+ 
+2026-10-17 06:14:43.810809: 
+epoch:  153 
+2026-10-17 06:19:00.235448: train loss : -0.7426 
+2026-10-17 06:19:05.589733: validation loss: -0.2582 
+2026-10-17 06:19:05.590931: Average global foreground Dice: [np.float32(0.8347), np.float32(0.1813), np.float32(0.1968), np.float32(0.2344), np.float32(0.153), np.float32(0.2102), np.float32(0.1772), np.float32(0.0), np.float32(0.0), np.float32(0.6373), np.float32(0.1967), np.float32(0.135), np.float32(0.192), np.float32(0.1804), np.float32(0.0), np.float32(0.0), np.float32(0.1418), np.float32(0.0941), np.float32(0.1597), np.float32(0.1306), np.float32(0.0676), np.float32(0.1322), np.float32(0.1395), np.float32(0.2753), np.float32(0.0591), np.float32(0.0471), np.float32(0.0), np.float32(0.0), np.float32(0.1268), np.float32(0.1118), np.float32(0.0), np.float32(0.0), np.float32(0.1564), np.float32(0.1161), np.float32(0.1962), np.float32(0.1756)] 
+2026-10-17 06:19:05.591291: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 06:19:05.912377: lr: 0.00086 
+2026-10-17 06:19:05.912670: This epoch took 262.101787 s
+ 
+2026-10-17 06:19:05.912774: 
+epoch:  154 
+2026-10-17 06:23:17.035105: train loss : -0.7414 
+2026-10-17 06:23:22.384313: validation loss: -0.4313 
+2026-10-17 06:23:22.385293: Average global foreground Dice: [np.float32(0.8785), np.float32(0.4538), np.float32(0.4231), np.float32(0.4556), np.float32(0.3701), np.float32(0.4239), np.float32(0.4281), np.float32(0.0), np.float32(0.0), np.float32(0.5574), np.float32(0.2652), np.float32(0.3313), np.float32(0.3172), np.float32(0.4007), np.float32(0.0), np.float32(0.0), np.float32(0.4167), np.float32(0.3573), np.float32(0.3925), np.float32(0.3565), np.float32(0.2902), np.float32(0.2792), np.float32(0.5239), np.float32(0.4984), np.float32(0.1706), np.float32(0.2053), np.float32(0.0), np.float32(0.0), np.float32(0.1047), np.float32(0.2887), np.float32(0.0), np.float32(0.0), np.float32(0.2305), np.float32(0.258), np.float32(0.4565), np.float32(0.452)] 
+2026-10-17 06:23:22.385523: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 06:23:22.724416: lr: 0.000859 
+2026-10-17 06:23:22.724882: This epoch took 256.812032 s
+ 
+2026-10-17 06:23:22.725194: 
+epoch:  155 
+2026-10-17 06:27:38.601108: train loss : -0.7305 
+2026-10-17 06:27:43.960680: validation loss: -0.3711 
+2026-10-17 06:27:43.961805: Average global foreground Dice: [np.float32(0.8163), np.float32(0.4315), np.float32(0.3843), np.float32(0.4031), np.float32(0.3337), np.float32(0.2607), np.float32(0.3421), np.float32(0.0), np.float32(0.0), np.float32(0.6191), np.float32(0.3404), np.float32(0.2348), np.float32(0.311), np.float32(0.1525), np.float32(0.0), np.float32(0.0), np.float32(0.3174), np.float32(0.2417), np.float32(0.3313), np.float32(0.2852), np.float32(0.1363), np.float32(0.2234), np.float32(0.4656), np.float32(0.46), np.float32(0.1735), np.float32(0.1959), np.float32(0.0), np.float32(0.0), np.float32(0.0911), np.float32(0.2124), np.float32(0.0), np.float32(0.0), np.float32(0.3091), np.float32(0.1872), np.float32(0.3592), np.float32(0.3738)] 
+2026-10-17 06:27:43.962076: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 06:27:44.289197: lr: 0.000858 
+2026-10-17 06:27:44.289461: This epoch took 261.564005 s
+ 
+2026-10-17 06:27:44.289559: 
+epoch:  156 
+2026-10-17 06:31:55.401352: train loss : -0.7281 
+2026-10-17 06:32:00.742525: validation loss: -0.5386 
+2026-10-17 06:32:00.743478: Average global foreground Dice: [np.float32(0.8651), np.float32(0.6832), np.float32(0.6417), np.float32(0.6997), np.float32(0.5953), np.float32(0.6993), np.float32(0.6134), np.float32(0.0), np.float32(0.0), np.float32(0.5863), np.float32(0.5211), np.float32(0.5201), np.float32(0.4909), np.float32(0.504), np.float32(0.0), np.float32(0.0), np.float32(0.6031), np.float32(0.455), np.float32(0.5817), np.float32(0.4814), np.float32(0.5783), np.float32(0.4606), np.float32(0.5718), np.float32(0.5157), np.float32(0.4376), np.float32(0.2974), np.float32(0.0), np.float32(0.0), np.float32(0.4161), np.float32(0.0399), np.float32(0.0), np.float32(0.0), np.float32(0.4862), np.float32(0.5559), np.float32(0.6707), np.float32(0.6606)] 
+2026-10-17 06:32:00.743832: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 06:32:01.062039: lr: 0.000858 
+2026-10-17 06:32:01.062262: This epoch took 256.772648 s
+ 
+2026-10-17 06:32:01.062343: 
+epoch:  157 
+2026-10-17 06:36:02.364673: train loss : -0.7382 
+2026-10-17 06:36:07.730305: validation loss: -0.3548 
+2026-10-17 06:36:07.732489: Average global foreground Dice: [np.float32(0.8565), np.float32(0.3442), np.float32(0.3554), np.float32(0.4106), np.float32(0.2961), np.float32(0.3791), np.float32(0.3178), np.float32(0.0), np.float32(0.0), np.float32(0.6791), np.float32(0.2425), np.float32(0.2531), np.float32(0.2426), np.float32(0.2471), np.float32(0.0), np.float32(0.0), np.float32(0.2483), np.float32(0.1678), np.float32(0.2913), np.float32(0.172), np.float32(0.3535), np.float32(0.2733), np.float32(0.3685), np.float32(0.4244), np.float32(0.1794), np.float32(0.1312), np.float32(0.0), np.float32(0.0), np.float32(0.4196), np.float32(0.014), np.float32(0.0), np.float32(0.0), np.float32(0.2479), np.float32(0.3248), np.float32(0.4053), np.float32(0.3665)] 
+2026-10-17 06:36:07.733024: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 06:36:08.138644: lr: 0.000857 
+2026-10-17 06:36:08.138905: This epoch took 247.076509 s
+ 
+2026-10-17 06:36:08.138999: 
+epoch:  158 
+2026-10-17 06:40:16.824345: train loss : -0.7482 
+2026-10-17 06:40:22.182832: validation loss: -0.3592 
+2026-10-17 06:40:22.183964: Average global foreground Dice: [np.float32(0.8541), np.float32(0.3639), np.float32(0.3447), np.float32(0.389), np.float32(0.2777), np.float32(0.3713), np.float32(0.3017), np.float32(0.0), np.float32(0.0), np.float32(0.6319), np.float32(0.302), np.float32(0.2292), np.float32(0.2555), np.float32(0.2547), np.float32(0.0), np.float32(0.0), np.float32(0.2787), np.float32(0.1927), np.float32(0.282), np.float32(0.1815), np.float32(0.2803), np.float32(0.2514), np.float32(0.3735), np.float32(0.3803), np.float32(0.2112), np.float32(0.1236), np.float32(0.0), np.float32(0.0), np.float32(0.2975), np.float32(0.2091), np.float32(0.0), np.float32(0.0), np.float32(0.2115), np.float32(0.237), np.float32(0.3624), np.float32(0.3724)] 
+2026-10-17 06:40:22.184414: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 06:40:22.555295: lr: 0.000856 
+2026-10-17 06:40:22.555545: This epoch took 254.416481 s
+ 
+2026-10-17 06:40:22.555629: 
+epoch:  159 
+2026-10-17 06:44:35.543308: train loss : -0.7362 
+2026-10-17 06:44:40.901512: validation loss: -0.3670 
+2026-10-17 06:44:40.902511: Average global foreground Dice: [np.float32(0.8747), np.float32(0.38), np.float32(0.3431), np.float32(0.4273), np.float32(0.321), np.float32(0.3796), np.float32(0.3204), np.float32(0.0), np.float32(0.0), np.float32(0.6442), np.float32(0.325), np.float32(0.2733), np.float32(0.3123), np.float32(0.2648), np.float32(0.0), np.float32(0.0), np.float32(0.3627), np.float32(0.2785), np.float32(0.3484), np.float32(0.223), np.float32(0.3155), np.float32(0.325), np.float32(0.2231), np.float32(0.2981), np.float32(0.2361), np.float32(0.0576), np.float32(0.0), np.float32(0.0), np.float32(0.5061), np.float32(0.0593), np.float32(0.0), np.float32(0.0), np.float32(0.3419), np.float32(0.336), np.float32(0.3954), np.float32(0.3491)] 
+2026-10-17 06:44:40.902697: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 06:44:41.230386: lr: 0.000855 
+2026-10-17 06:44:41.230613: This epoch took 258.674848 s
+ 
+2026-10-17 06:44:41.230681: 
+epoch:  160 
+2026-10-17 06:48:48.611195: train loss : -0.7395 
+2026-10-17 06:48:53.978384: validation loss: -0.3622 
+2026-10-17 06:48:53.979425: Average global foreground Dice: [np.float32(0.8576), np.float32(0.3555), np.float32(0.3826), np.float32(0.366), np.float32(0.2943), np.float32(0.357), np.float32(0.3731), np.float32(0.0), np.float32(0.0), np.float32(0.4187), np.float32(0.3033), np.float32(0.2924), np.float32(0.25), np.float32(0.1712), np.float32(0.0), np.float32(0.0), np.float32(0.2607), np.float32(0.174), np.float32(0.3457), np.float32(0.2277), np.float32(0.2508), np.float32(0.2683), np.float32(0.3481), np.float32(0.3454), np.float32(0.1608), np.float32(0.0767), np.float32(0.0), np.float32(0.0), np.float32(0.3636), np.float32(0.2387), np.float32(0.0), np.float32(0.0), np.float32(0.255), np.float32(0.3087), np.float32(0.4115), np.float32(0.4092)] 
+2026-10-17 06:48:53.979851: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 06:48:54.306677: lr: 0.000854 
+2026-10-17 06:48:54.307214: This epoch took 253.076472 s
+ 
+2026-10-17 06:48:54.307442: 
+epoch:  161 
+2026-10-17 06:53:12.491819: train loss : -0.7495 
+2026-10-17 06:53:17.844555: validation loss: -0.3137 
+2026-10-17 06:53:17.845652: Average global foreground Dice: [np.float32(0.8907), np.float32(0.3158), np.float32(0.2971), np.float32(0.2633), np.float32(0.2318), np.float32(0.2971), np.float32(0.295), np.float32(0.0), np.float32(0.0), np.float32(0.5834), np.float32(0.2362), np.float32(0.2014), np.float32(0.1645), np.float32(0.1486), np.float32(0.0), np.float32(0.0), np.float32(0.2257), np.float32(0.1753), np.float32(0.2849), np.float32(0.188), np.float32(0.2233), np.float32(0.2635), np.float32(0.2766), np.float32(0.2967), np.float32(0.141), np.float32(0.0494), np.float32(0.0), np.float32(0.0), np.float32(0.3097), np.float32(0.1575), np.float32(0.0), np.float32(0.0), np.float32(0.16), np.float32(0.2742), np.float32(0.3253), np.float32(0.2861)] 
+2026-10-17 06:53:17.846154: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 06:53:18.214818: lr: 0.000853 
+2026-10-17 06:53:18.215057: This epoch took 263.907479 s
+ 
+2026-10-17 06:53:18.215137: 
+epoch:  162 
+2026-10-17 06:57:31.285486: train loss : -0.7492 
+2026-10-17 06:57:36.635881: validation loss: -0.2273 
+2026-10-17 06:57:36.636970: Average global foreground Dice: [np.float32(0.8813), np.float32(0.1196), np.float32(0.1292), np.float32(0.1577), np.float32(0.0906), np.float32(0.1208), np.float32(0.1212), np.float32(0.0), np.float32(0.0), np.float32(0.6271), np.float32(0.1621), np.float32(0.1365), np.float32(0.1625), np.float32(0.0727), np.float32(0.0), np.float32(0.0), np.float32(0.1129), np.float32(0.0862), np.float32(0.123), np.float32(0.0777), np.float32(0.0391), np.float32(0.1058), np.float32(0.167), np.float32(0.2433), np.float32(0.0462), np.float32(0.0364), np.float32(0.0), np.float32(0.0), np.float32(0.3208), np.float32(0.0529), np.float32(0.0), np.float32(0.0), np.float32(0.1044), np.float32(0.0894), np.float32(0.1333), np.float32(0.141)] 
+2026-10-17 06:57:36.637137: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 06:57:36.959741: lr: 0.000852 
+2026-10-17 06:57:36.960302: This epoch took 258.745107 s
+ 
+2026-10-17 06:57:36.960489: 
+epoch:  163 
+2026-10-17 07:01:44.308696: train loss : -0.7493 
+2026-10-17 07:01:49.694173: validation loss: -0.4041 
+2026-10-17 07:01:49.695259: Average global foreground Dice: [np.float32(0.8768), np.float32(0.3556), np.float32(0.39), np.float32(0.497), np.float32(0.3623), np.float32(0.4286), np.float32(0.3869), np.float32(0.0), np.float32(0.0), np.float32(0.5901), np.float32(0.3905), np.float32(0.3541), np.float32(0.4178), np.float32(0.3677), np.float32(0.0), np.float32(0.0), np.float32(0.3614), np.float32(0.2416), np.float32(0.3759), np.float32(0.2148), np.float32(0.2297), np.float32(0.2587), np.float32(0.2338), np.float32(0.3816), np.float32(0.1657), np.float32(0.1153), np.float32(0.0), np.float32(0.0), np.float32(0.4478), np.float32(0.3094), np.float32(0.0), np.float32(0.0), np.float32(0.3278), np.float32(0.3495), np.float32(0.3824), np.float32(0.4023)] 
+2026-10-17 07:01:49.695502: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 07:01:50.029391: lr: 0.000851 
+2026-10-17 07:01:50.029739: This epoch took 253.069156 s
+ 
+2026-10-17 07:01:50.029959: 
+epoch:  164 
+2026-10-17 07:05:56.307322: train loss : -0.7551 
+2026-10-17 07:06:01.663316: validation loss: -0.3287 
+2026-10-17 07:06:01.664256: Average global foreground Dice: [np.float32(0.864), np.float32(0.3021), np.float32(0.2776), np.float32(0.3303), np.float32(0.2565), np.float32(0.2908), np.float32(0.3071), np.float32(0.0), np.float32(0.0), np.float32(0.4787), np.float32(0.3257), np.float32(0.2135), np.float32(0.2855), np.float32(0.161), np.float32(0.0), np.float32(0.0), np.float32(0.2204), np.float32(0.1557), np.float32(0.2795), np.float32(0.2018), np.float32(0.1119), np.float32(0.14), np.float32(0.259), np.float32(0.2949), np.float32(0.0949), np.float32(0.0436), np.float32(0.0), np.float32(0.0), np.float32(0.2393), np.float32(0.0845), np.float32(0.0), np.float32(0.0), np.float32(0.1933), np.float32(0.325), np.float32(0.2827), np.float32(0.324)] 
+2026-10-17 07:06:01.664588: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 07:06:02.039431: lr: 0.00085 
+2026-10-17 07:06:02.039927: This epoch took 252.009744 s
+ 
+2026-10-17 07:06:02.040130: 
+epoch:  165 
+2026-10-17 07:10:33.200590: train loss : -0.7538 
+2026-10-17 07:10:38.549544: validation loss: -0.3504 
+2026-10-17 07:10:38.550549: Average global foreground Dice: [np.float32(0.8919), np.float32(0.3189), np.float32(0.343), np.float32(0.4323), np.float32(0.2669), np.float32(0.3818), np.float32(0.3714), np.float32(0.0), np.float32(0.0), np.float32(0.5645), np.float32(0.3597), np.float32(0.276), np.float32(0.3178), np.float32(0.2783), np.float32(0.0), np.float32(0.0), np.float32(0.2851), np.float32(0.194), np.float32(0.3416), np.float32(0.2095), np.float32(0.2289), np.float32(0.2107), np.float32(0.2454), np.float32(0.3325), np.float32(0.1278), np.float32(0.0901), np.float32(0.0), np.float32(0.0), np.float32(0.2681), np.float32(0.2203), np.float32(0.0), np.float32(0.0), np.float32(0.1668), np.float32(0.3942), np.float32(0.2707), np.float32(0.3782)] 
+2026-10-17 07:10:38.550765: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 07:10:39.020323: lr: 0.000849 
+2026-10-17 07:10:39.020614: This epoch took 276.980391 s
+ 
+2026-10-17 07:10:39.020712: 
+epoch:  166 
+2026-10-17 07:14:42.676638: train loss : -0.7554 
+2026-10-17 07:14:48.024058: validation loss: -0.3598 
+2026-10-17 07:14:48.025192: Average global foreground Dice: [np.float32(0.8584), np.float32(0.3108), np.float32(0.3445), np.float32(0.4446), np.float32(0.286), np.float32(0.4129), np.float32(0.3475), np.float32(0.0), np.float32(0.0), np.float32(0.5281), np.float32(0.3762), np.float32(0.358), np.float32(0.3186), np.float32(0.3044), np.float32(0.0), np.float32(0.0), np.float32(0.2909), np.float32(0.1924), np.float32(0.3333), np.float32(0.2102), np.float32(0.3852), np.float32(0.2418), np.float32(0.191), np.float32(0.2785), np.float32(0.1704), np.float32(0.0476), np.float32(0.0), np.float32(0.0), np.float32(0.3515), np.float32(0.2632), np.float32(0.0), np.float32(0.0), np.float32(0.3428), np.float32(0.4081), np.float32(0.3345), np.float32(0.4075)] 
+2026-10-17 07:14:48.025619: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 07:14:48.350241: lr: 0.000848 
+2026-10-17 07:14:48.350612: This epoch took 249.329828 s
+ 
+2026-10-17 07:14:48.350823: 
+epoch:  167 
+2026-10-17 07:18:53.582796: train loss : -0.7566 
+2026-10-17 07:18:58.938741: validation loss: -0.3731 
+2026-10-17 07:18:58.939819: Average global foreground Dice: [np.float32(0.8678), np.float32(0.3411), np.float32(0.3578), np.float32(0.4569), np.float32(0.311), np.float32(0.3964), np.float32(0.3515), np.float32(0.0), np.float32(0.0), np.float32(0.6695), np.float32(0.3382), np.float32(0.3004), np.float32(0.3881), np.float32(0.3389), np.float32(0.0), np.float32(0.0), np.float32(0.3396), np.float32(0.2451), np.float32(0.3563), np.float32(0.2199), np.float32(0.244), np.float32(0.2555), np.float32(0.2824), np.float32(0.3557), np.float32(0.1761), np.float32(0.0636), np.float32(0.0), np.float32(0.0), np.float32(0.4219), np.float32(0.1775), np.float32(0.0), np.float32(0.0), np.float32(0.4293), np.float32(0.4575), np.float32(0.3773), np.float32(0.3531)] 
+2026-10-17 07:18:58.940197: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 07:18:59.257122: lr: 0.000847 
+2026-10-17 07:18:59.257598: This epoch took 250.906672 s
+ 
+2026-10-17 07:18:59.257827: 
+epoch:  168 
+2026-10-17 07:22:58.480798: train loss : -0.7504 
+2026-10-17 07:23:03.832159: validation loss: -0.3262 
+2026-10-17 07:23:03.833198: Average global foreground Dice: [np.float32(0.8565), np.float32(0.2438), np.float32(0.2826), np.float32(0.3595), np.float32(0.2536), np.float32(0.3227), np.float32(0.2791), np.float32(0.0), np.float32(0.0), np.float32(0.5546), np.float32(0.2698), np.float32(0.226), np.float32(0.2271), np.float32(0.2514), np.float32(0.0), np.float32(0.0), np.float32(0.2474), np.float32(0.1521), np.float32(0.2594), np.float32(0.1623), np.float32(0.1186), np.float32(0.2015), np.float32(0.2303), np.float32(0.2482), np.float32(0.1268), np.float32(0.0607), np.float32(0.0), np.float32(0.0), np.float32(0.3289), np.float32(0.2566), np.float32(0.0), np.float32(0.0), np.float32(0.2119), np.float32(0.2676), np.float32(0.289), np.float32(0.2636)] 
+2026-10-17 07:23:03.833355: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 07:23:04.143052: lr: 0.000847 
+2026-10-17 07:23:04.143544: This epoch took 244.885498 s
+ 
+2026-10-17 07:23:04.143959: 
+epoch:  169 
+2026-10-17 07:27:24.665972: train loss : -0.7382 
+2026-10-17 07:27:30.021341: validation loss: -0.4751 
+2026-10-17 07:27:30.023185: Average global foreground Dice: [np.float32(0.8354), np.float32(0.5148), np.float32(0.5375), np.float32(0.6108), np.float32(0.4882), np.float32(0.6154), np.float32(0.5332), np.float32(0.0), np.float32(0.0), np.float32(0.6098), np.float32(0.5283), np.float32(0.5068), np.float32(0.4785), np.float32(0.3704), np.float32(0.0), np.float32(0.0), np.float32(0.4681), np.float32(0.3197), np.float32(0.5072), np.float32(0.3359), np.float32(0.3445), np.float32(0.4082), np.float32(0.4295), np.float32(0.4821), np.float32(0.2774), np.float32(0.1228), np.float32(0.0), np.float32(0.0), np.float32(0.5638), np.float32(0.2869), np.float32(0.0), np.float32(0.0), np.float32(0.533), np.float32(0.4703), np.float32(0.5847), np.float32(0.5675)] 
+2026-10-17 07:27:30.023461: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 07:27:30.347486: lr: 0.000846 
+2026-10-17 07:27:30.347943: This epoch took 266.203714 s
+ 
+2026-10-17 07:27:30.348144: 
+epoch:  170 
+2026-10-17 07:31:51.764603: train loss : -0.7331 
+2026-10-17 07:31:57.119354: validation loss: -0.5248 
+2026-10-17 07:31:57.120343: Average global foreground Dice: [np.float32(0.8597), np.float32(0.6392), np.float32(0.6237), np.float32(0.6879), np.float32(0.5933), np.float32(0.7139), np.float32(0.6179), np.float32(0.0), np.float32(0.0), np.float32(0.6527), np.float32(0.5456), np.float32(0.5375), np.float32(0.3278), np.float32(0.3548), np.float32(0.0), np.float32(0.0), np.float32(0.5531), np.float32(0.3853), np.float32(0.5294), np.float32(0.4051), np.float32(0.5634), np.float32(0.4482), np.float32(0.715), np.float32(0.6083), np.float32(0.2443), np.float32(0.231), np.float32(0.0), np.float32(0.0), np.float32(0.4118), np.float32(0.3293), np.float32(0.0), np.float32(0.0), np.float32(0.5718), np.float32(0.5077), np.float32(0.7212), np.float32(0.7103)] 
+2026-10-17 07:31:57.120650: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 07:31:57.443828: lr: 0.000845 
+2026-10-17 07:31:57.444254: This epoch took 267.096024 s
+ 
+2026-10-17 07:31:57.444587: 
+epoch:  171 
+2026-10-17 07:36:11.414617: train loss : -0.7528 
+2026-10-17 07:36:16.770320: validation loss: -0.5041 
+2026-10-17 07:36:16.771547: Average global foreground Dice: [np.float32(0.8452), np.float32(0.5497), np.float32(0.5747), np.float32(0.6392), np.float32(0.5526), np.float32(0.6626), np.float32(0.6185), np.float32(0.0), np.float32(0.0), np.float32(0.6541), np.float32(0.5227), np.float32(0.5077), np.float32(0.3311), np.float32(0.3241), np.float32(0.0), np.float32(0.0), np.float32(0.5876), np.float32(0.4509), np.float32(0.5462), np.float32(0.4834), np.float32(0.5187), np.float32(0.4216), np.float32(0.5768), np.float32(0.5757), np.float32(0.219), np.float32(0.1822), np.float32(0.0), np.float32(0.0), np.float32(0.4658), np.float32(0.1663), np.float32(0.0), np.float32(0.0), np.float32(0.4841), np.float32(0.5163), np.float32(0.6537), np.float32(0.6713)] 
+2026-10-17 07:36:16.772089: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 07:36:17.091102: lr: 0.000844 
+2026-10-17 07:36:17.091691: This epoch took 259.646880 s
+ 
+2026-10-17 07:36:17.091801: 
+epoch:  172 
+2026-10-17 07:40:17.869224: train loss : -0.7549 
+2026-10-17 07:40:23.223936: validation loss: -0.4472 
+2026-10-17 07:40:23.225061: Average global foreground Dice: [np.float32(0.8793), np.float32(0.465), np.float32(0.5214), np.float32(0.5038), np.float32(0.4341), np.float32(0.5366), np.float32(0.5093), np.float32(0.0), np.float32(0.0), np.float32(0.0647), np.float32(0.4372), np.float32(0.4356), np.float32(0.2487), np.float32(0.2924), np.float32(0.0), np.float32(0.0), np.float32(0.4076), np.float32(0.3115), np.float32(0.4799), np.float32(0.4314), np.float32(0.3285), np.float32(0.3804), np.float32(0.6594), np.float32(0.4663), np.float32(0.2241), np.float32(0.2149), np.float32(0.0), np.float32(0.0), np.float32(0.4624), np.float32(0.2018), np.float32(0.0), np.float32(0.0), np.float32(0.3565), np.float32(0.4119), np.float32(0.5408), np.float32(0.5902)] 
+2026-10-17 07:40:23.225681: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 07:40:23.537402: lr: 0.000843 
+2026-10-17 07:40:23.537663: This epoch took 246.445795 s
+ 
+2026-10-17 07:40:23.537748: 
+epoch:  173 
+2026-10-17 07:44:41.169990: train loss : -0.7495 
+2026-10-17 07:44:46.518299: validation loss: -0.4317 
+2026-10-17 07:44:46.519302: Average global foreground Dice: [np.float32(0.8439), np.float32(0.4204), np.float32(0.4454), np.float32(0.474), np.float32(0.3772), np.float32(0.487), np.float32(0.4656), np.float32(0.0), np.float32(0.0), np.float32(0.5666), np.float32(0.4007), np.float32(0.3766), np.float32(0.3013), np.float32(0.3339), np.float32(0.0), np.float32(0.0), np.float32(0.3974), np.float32(0.3079), np.float32(0.4446), np.float32(0.3616), np.float32(0.4546), np.float32(0.3431), np.float32(0.4756), np.float32(0.3893), np.float32(0.1821), np.float32(0.1543), np.float32(0.0), np.float32(0.0), np.float32(0.4916), np.float32(0.3353), np.float32(0.0), np.float32(0.0), np.float32(0.3874), np.float32(0.43), np.float32(0.4472), np.float32(0.5011)] 
+2026-10-17 07:44:46.519569: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 07:44:46.859524: lr: 0.000842 
+2026-10-17 07:44:46.859987: This epoch took 263.322130 s
+ 
+2026-10-17 07:44:46.860085: 
+epoch:  174 
+2026-10-17 07:49:00.856331: train loss : -0.7617 
+2026-10-17 07:49:06.211604: validation loss: -0.4253 
+2026-10-17 07:49:06.212839: Average global foreground Dice: [np.float32(0.886), np.float32(0.418), np.float32(0.4379), np.float32(0.4614), np.float32(0.4089), np.float32(0.5046), np.float32(0.4795), np.float32(0.0), np.float32(0.0), np.float32(0.6118), np.float32(0.3824), np.float32(0.3384), np.float32(0.2698), np.float32(0.3316), np.float32(0.0), np.float32(0.0), np.float32(0.4175), np.float32(0.3259), np.float32(0.4392), np.float32(0.3471), np.float32(0.3948), np.float32(0.3105), np.float32(0.5528), np.float32(0.4431), np.float32(0.1714), np.float32(0.1399), np.float32(0.0), np.float32(0.0), np.float32(0.4036), np.float32(0.0), np.float32(0.0), np.float32(0.0), np.float32(0.3247), np.float32(0.4939), np.float32(0.4474), np.float32(0.4908)] 
+2026-10-17 07:49:06.213178: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 07:49:06.530528: lr: 0.000841 
+2026-10-17 07:49:06.530820: This epoch took 259.670670 s
+ 
+2026-10-17 07:49:06.530994: 
+epoch:  175 
+2026-10-17 07:53:29.568169: train loss : -0.7535 
+2026-10-17 07:53:34.936864: validation loss: -0.4687 
+2026-10-17 07:53:34.937769: Average global foreground Dice: [np.float32(0.8813), np.float32(0.4939), np.float32(0.4854), np.float32(0.5142), np.float32(0.4719), np.float32(0.5525), np.float32(0.5456), np.float32(0.0), np.float32(0.0), np.float32(0.6843), np.float32(0.4498), np.float32(0.4208), np.float32(0.3707), np.float32(0.3284), np.float32(0.0), np.float32(0.0), np.float32(0.4935), np.float32(0.3567), np.float32(0.4972), np.float32(0.3957), np.float32(0.4249), np.float32(0.3992), np.float32(0.5395), np.float32(0.4617), np.float32(0.2071), np.float32(0.2123), np.float32(0.0), np.float32(0.0), np.float32(0.4824), np.float32(0.2027), np.float32(0.0), np.float32(0.0), np.float32(0.4683), np.float32(0.4748), np.float32(0.5679), np.float32(0.5779)] 
+2026-10-17 07:53:34.937938: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 07:53:35.250794: lr: 0.00084 
+2026-10-17 07:53:35.251165: This epoch took 268.720083 s
+ 
+2026-10-17 07:53:35.251419: 
+epoch:  176 
+2026-10-17 07:57:46.176572: train loss : -0.7644 
+2026-10-17 07:57:51.539873: validation loss: -0.4717 
+2026-10-17 07:57:51.541046: Average global foreground Dice: [np.float32(0.8535), np.float32(0.5267), np.float32(0.4471), np.float32(0.5049), np.float32(0.5029), np.float32(0.3836), np.float32(0.4557), np.float32(0.0), np.float32(0.0), np.float32(0.6203), np.float32(0.4059), np.float32(0.3652), np.float32(0.3785), np.float32(0.2968), np.float32(0.0), np.float32(0.0), np.float32(0.5025), np.float32(0.3826), np.float32(0.45), np.float32(0.4028), np.float32(0.432), np.float32(0.383), np.float32(0.6573), np.float32(0.628), np.float32(0.213), np.float32(0.2361), np.float32(0.0), np.float32(0.0), np.float32(0.3246), np.float32(0.4234), np.float32(0.0), np.float32(0.0), np.float32(0.4085), np.float32(0.228), np.float32(0.5861), np.float32(0.4638)] 
+2026-10-17 07:57:51.541268: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 07:57:51.877391: lr: 0.000839 
+2026-10-17 07:57:51.878158: This epoch took 256.626559 s
+ 
+2026-10-17 07:57:51.878468: 
+epoch:  177 
+2026-10-17 08:02:09.404104: train loss : -0.7391 
+2026-10-17 08:02:14.776244: validation loss: -0.4603 
+2026-10-17 08:02:14.777214: Average global foreground Dice: [np.float32(0.8131), np.float32(0.4465), np.float32(0.4387), np.float32(0.5184), np.float32(0.4611), np.float32(0.4857), np.float32(0.5472), np.float32(0.0), np.float32(0.0), np.float32(0.4758), np.float32(0.4348), np.float32(0.422), np.float32(0.3455), np.float32(0.2215), np.float32(0.0), np.float32(0.0), np.float32(0.4303), np.float32(0.3823), np.float32(0.4595), np.float32(0.4031), np.float32(0.148), np.float32(0.2604), np.float32(0.5572), np.float32(0.6432), np.float32(0.18), np.float32(0.2606), np.float32(0.0), np.float32(0.0), np.float32(0.0088), np.float32(0.1556), np.float32(0.0), np.float32(0.0), np.float32(0.2781), np.float32(0.007), np.float32(0.5159), np.float32(0.5484)] 
+2026-10-17 08:02:14.777366: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 08:02:15.100347: lr: 0.000838 
+2026-10-17 08:02:15.100887: This epoch took 263.222233 s
+ 
+2026-10-17 08:02:15.101182: 
+epoch:  178 
+2026-10-17 08:06:48.095416: train loss : -0.7338 
+2026-10-17 08:06:53.446087: validation loss: -0.4100 
+2026-10-17 08:06:53.447245: Average global foreground Dice: [np.float32(0.8587), np.float32(0.3686), np.float32(0.4444), np.float32(0.442), np.float32(0.4045), np.float32(0.4627), np.float32(0.4941), np.float32(0.0), np.float32(0.0), np.float32(0.5762), np.float32(0.3403), np.float32(0.361), np.float32(0.2909), np.float32(0.3387), np.float32(0.0), np.float32(0.0), np.float32(0.4209), np.float32(0.3684), np.float32(0.4212), np.float32(0.3624), np.float32(0.1252), np.float32(0.3298), np.float32(0.3815), np.float32(0.3414), np.float32(0.1396), np.float32(0.204), np.float32(0.0), np.float32(0.0), np.float32(0.2662), np.float32(0.0692), np.float32(0.0), np.float32(0.0), np.float32(0.476), np.float32(0.3698), np.float32(0.3811), np.float32(0.4578)] 
+2026-10-17 08:06:53.447620: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 08:06:53.765342: lr: 0.000837 
+2026-10-17 08:06:53.765684: This epoch took 278.664313 s
+ 
+2026-10-17 08:06:53.765908: 
+epoch:  179 
+2026-10-17 08:11:15.017953: train loss : -0.7540 
+2026-10-17 08:11:20.369202: validation loss: -0.3144 
+2026-10-17 08:11:20.370317: Average global foreground Dice: [np.float32(0.848), np.float32(0.1848), np.float32(0.2542), np.float32(0.2998), np.float32(0.2113), np.float32(0.267), np.float32(0.2535), np.float32(0.0), np.float32(0.0), np.float32(0.6276), np.float32(0.1881), np.float32(0.2227), np.float32(0.2707), np.float32(0.272), np.float32(0.0), np.float32(0.0), np.float32(0.2492), np.float32(0.2105), np.float32(0.2459), np.float32(0.1728), np.float32(0.035), np.float32(0.184), np.float32(0.1681), np.float32(0.3191), np.float32(0.0786), np.float32(0.0969), np.float32(0.0), np.float32(0.0), np.float32(0.1902), np.float32(0.1347), np.float32(0.0), np.float32(0.0), np.float32(0.2454), np.float32(0.1487), np.float32(0.2376), np.float32(0.2181)] 
+2026-10-17 08:11:20.370670: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 08:11:20.744165: lr: 0.000836 
+2026-10-17 08:11:20.744546: This epoch took 266.978552 s
+ 
+2026-10-17 08:11:20.744754: 
+epoch:  180 
+2026-10-17 08:15:46.505948: train loss : -0.7617 
+2026-10-17 08:15:51.873948: validation loss: -0.2993 
+2026-10-17 08:15:51.874953: Average global foreground Dice: [np.float32(0.8589), np.float32(0.1941), np.float32(0.2382), np.float32(0.2868), np.float32(0.2151), np.float32(0.217), np.float32(0.2323), np.float32(0.0), np.float32(0.0), np.float32(0.5471), np.float32(0.2332), np.float32(0.2599), np.float32(0.2144), np.float32(0.2247), np.float32(0.0), np.float32(0.0), np.float32(0.2145), np.float32(0.1312), np.float32(0.1976), np.float32(0.134), np.float32(0.155), np.float32(0.2175), np.float32(0.2023), np.float32(0.2394), np.float32(0.0663), np.float32(0.0947), np.float32(0.0), np.float32(0.0), np.float32(0.2184), np.float32(0.1761), np.float32(0.0), np.float32(0.0), np.float32(0.2262), np.float32(0.2198), np.float32(0.2058), np.float32(0.2273)] 
+2026-10-17 08:15:51.875152: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 08:15:52.194577: lr: 0.000836 
+2026-10-17 08:15:52.194834: This epoch took 271.449801 s
+ 
+2026-10-17 08:15:52.194926: 
+epoch:  181 
+2026-10-17 08:19:47.950997: train loss : -0.7597 
+2026-10-17 08:19:53.297942: validation loss: -0.4121 
+2026-10-17 08:19:53.299351: Average global foreground Dice: [np.float32(0.8551), np.float32(0.4417), np.float32(0.4347), np.float32(0.5455), np.float32(0.4607), np.float32(0.5591), np.float32(0.4633), np.float32(0.0), np.float32(0.0), np.float32(0.6427), np.float32(0.4318), np.float32(0.4564), np.float32(0.339), np.float32(0.3862), np.float32(0.0), np.float32(0.0), np.float32(0.4788), np.float32(0.3556), np.float32(0.4358), np.float32(0.2799), np.float32(0.4024), np.float32(0.3334), np.float32(0.3873), np.float32(0.2481), np.float32(0.2226), np.float32(0.1881), np.float32(0.0), np.float32(0.0), np.float32(0.3954), np.float32(0.236), np.float32(0.0), np.float32(0.0), np.float32(0.4941), np.float32(0.4985), np.float32(0.4868), np.float32(0.5226)] 
+2026-10-17 08:19:53.299679: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 08:19:53.670416: lr: 0.000835 
+2026-10-17 08:19:53.670707: This epoch took 241.475711 s
+ 
+2026-10-17 08:19:53.670929: 
+epoch:  182 
+2026-10-17 08:24:00.434831: train loss : -0.7609 
+2026-10-17 08:24:05.795561: validation loss: -0.3473 
+2026-10-17 08:24:05.796881: Average global foreground Dice: [np.float32(0.8455), np.float32(0.3194), np.float32(0.3528), np.float32(0.3802), np.float32(0.3293), np.float32(0.386), np.float32(0.3778), np.float32(0.0), np.float32(0.0), np.float32(0.6073), np.float32(0.3275), np.float32(0.3563), np.float32(0.2794), np.float32(0.249), np.float32(0.0), np.float32(0.0), np.float32(0.3359), np.float32(0.2471), np.float32(0.3252), np.float32(0.1998), np.float32(0.2341), np.float32(0.2762), np.float32(0.2481), np.float32(0.3339), np.float32(0.14), np.float32(0.1348), np.float32(0.0), np.float32(0.0), np.float32(0.188), np.float32(0.2429), np.float32(0.0), np.float32(0.0), np.float32(0.3357), np.float32(0.2841), np.float32(0.3004), np.float32(0.2635)] 
+2026-10-17 08:24:05.797086: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 08:24:06.167436: lr: 0.000834 
+2026-10-17 08:24:06.167802: This epoch took 252.496689 s
+ 
+2026-10-17 08:24:06.168033: 
+epoch:  183 
+2026-10-17 08:28:14.773026: train loss : -0.7672 
+2026-10-17 08:28:20.122079: validation loss: -0.2613 
+2026-10-17 08:28:20.123118: Average global foreground Dice: [np.float32(0.8414), np.float32(0.1504), np.float32(0.1793), np.float32(0.2058), np.float32(0.1688), np.float32(0.2109), np.float32(0.1976), np.float32(0.0), np.float32(0.0), np.float32(0.6165), np.float32(0.196), np.float32(0.1801), np.float32(0.1976), np.float32(0.1811), np.float32(0.0), np.float32(0.0), np.float32(0.1995), np.float32(0.1454), np.float32(0.1684), np.float32(0.1047), np.float32(0.0732), np.float32(0.1312), np.float32(0.2426), np.float32(0.266), np.float32(0.0335), np.float32(0.0584), np.float32(0.0), np.float32(0.0), np.float32(0.1837), np.float32(0.1734), np.float32(0.0), np.float32(0.0), np.float32(0.2379), np.float32(0.2189), np.float32(0.1915), np.float32(0.1249)] 
+2026-10-17 08:28:20.123274: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 08:28:20.458944: lr: 0.000833 
+2026-10-17 08:28:20.459189: This epoch took 254.290926 s
+ 
+2026-10-17 08:28:20.459271: 
+epoch:  184 
+2026-10-17 08:32:59.920256: train loss : -0.7539 
+2026-10-17 08:33:05.273206: validation loss: -0.2903 
+2026-10-17 08:33:05.274071: Average global foreground Dice: [np.float32(0.8653), np.float32(0.208), np.float32(0.2351), np.float32(0.313), np.float32(0.1861), np.float32(0.2833), np.float32(0.2074), np.float32(0.0), np.float32(0.0), np.float32(0.6443), np.float32(0.2302), np.float32(0.2406), np.float32(0.2776), np.float32(0.2517), np.float32(0.0), np.float32(0.0), np.float32(0.2055), np.float32(0.1253), np.float32(0.183), np.float32(0.1343), np.float32(0.1201), np.float32(0.1839), np.float32(0.1329), np.float32(0.3147), np.float32(0.1413), np.float32(0.1687), np.float32(0.0), np.float32(0.0), np.float32(0.0459), np.float32(0.0994), np.float32(0.0), np.float32(0.0), np.float32(0.2247), np.float32(0.1574), np.float32(0.2304), np.float32(0.2122)] 
+2026-10-17 08:33:05.274462: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 08:33:05.595156: lr: 0.000832 
+2026-10-17 08:33:05.595442: This epoch took 285.136050 s
+ 
+2026-10-17 08:33:05.595659: 
+epoch:  185 
+2026-10-17 08:37:12.899539: train loss : -0.7591 
+2026-10-17 08:37:18.256011: validation loss: -0.2899 
+2026-10-17 08:37:18.257087: Average global foreground Dice: [np.float32(0.8701), np.float32(0.2029), np.float32(0.2334), np.float32(0.3017), np.float32(0.1918), np.float32(0.2706), np.float32(0.2418), np.float32(0.0), np.float32(0.0), np.float32(0.6663), np.float32(0.2204), np.float32(0.1982), np.float32(0.2194), np.float32(0.2441), np.float32(0.0027), np.float32(0.0), np.float32(0.1623), np.float32(0.1057), np.float32(0.2151), np.float32(0.1802), np.float32(0.0739), np.float32(0.1654), np.float32(0.1721), np.float32(0.4084), np.float32(0.1204), np.float32(0.1417), np.float32(0.0), np.float32(0.0), np.float32(0.1047), np.float32(0.1074), np.float32(0.0), np.float32(0.0), np.float32(0.2247), np.float32(0.1566), np.float32(0.2391), np.float32(0.2196)] 
+2026-10-17 08:37:18.257495: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 08:37:18.603790: lr: 0.000831 
+2026-10-17 08:37:18.604036: This epoch took 253.008298 s
+ 
+2026-10-17 08:37:18.604112: 
+epoch:  186 
+2026-10-17 08:41:16.361338: train loss : -0.7645 
+2026-10-17 08:41:21.712104: validation loss: -0.4275 
+2026-10-17 08:41:21.713062: Average global foreground Dice: [np.float32(0.8711), np.float32(0.4626), np.float32(0.3777), np.float32(0.4898), np.float32(0.3986), np.float32(0.503), np.float32(0.4639), np.float32(0.0), np.float32(0.0), np.float32(0.6562), np.float32(0.3689), np.float32(0.408), np.float32(0.2673), np.float32(0.3671), np.float32(0.0), np.float32(0.0), np.float32(0.4484), np.float32(0.3252), np.float32(0.3859), np.float32(0.3873), np.float32(0.2782), np.float32(0.2822), np.float32(0.5555), np.float32(0.5069), np.float32(0.2912), np.float32(0.3623), np.float32(0.0), np.float32(0.0), np.float32(0.0962), np.float32(0.2426), np.float32(0.0), np.float32(0.0), np.float32(0.3216), np.float32(0.3248), np.float32(0.4475), np.float32(0.5007)] 
+2026-10-17 08:41:21.713255: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 08:41:22.042501: lr: 0.00083 
+2026-10-17 08:41:22.042840: This epoch took 243.438670 s
+ 
+2026-10-17 08:41:22.042962: 
+epoch:  187 
+2026-10-17 08:45:54.029599: train loss : -0.7593 
+2026-10-17 08:45:59.388455: validation loss: -0.3142 
+2026-10-17 08:45:59.389447: Average global foreground Dice: [np.float32(0.8528), np.float32(0.2514), np.float32(0.2208), np.float32(0.3358), np.float32(0.2453), np.float32(0.2931), np.float32(0.2456), np.float32(0.0), np.float32(0.0), np.float32(0.5549), np.float32(0.2955), np.float32(0.2883), np.float32(0.2387), np.float32(0.159), np.float32(0.0), np.float32(0.0), np.float32(0.2955), np.float32(0.2805), np.float32(0.2176), np.float32(0.1415), np.float32(0.3486), np.float32(0.2173), np.float32(0.2533), np.float32(0.2913), np.float32(0.1651), np.float32(0.0727), np.float32(0.0), np.float32(0.0), np.float32(0.3021), np.float32(0.0084), np.float32(0.0), np.float32(0.0), np.float32(0.262), np.float32(0.3593), np.float32(0.2807), np.float32(0.2354)] 
+2026-10-17 08:45:59.389753: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 08:45:59.710037: lr: 0.000829 
+2026-10-17 08:45:59.710442: This epoch took 277.667422 s
+ 
+2026-10-17 08:45:59.710577: 
+epoch:  188 
+2026-10-17 08:50:18.846434: train loss : -0.7463 
+2026-10-17 08:50:24.207547: validation loss: -0.2653 
+2026-10-17 08:50:24.208621: Average global foreground Dice: [np.float32(0.8584), np.float32(0.2015), np.float32(0.21), np.float32(0.2448), np.float32(0.1894), np.float32(0.1569), np.float32(0.1676), np.float32(0.0), np.float32(0.0), np.float32(0.6429), np.float32(0.1647), np.float32(0.1957), np.float32(0.1944), np.float32(0.0996), np.float32(0.0), np.float32(0.0), np.float32(0.2002), np.float32(0.1329), np.float32(0.1374), np.float32(0.1013), np.float32(0.0325), np.float32(0.1385), np.float32(0.1905), np.float32(0.327), np.float32(0.0737), np.float32(0.0734), np.float32(0.0), np.float32(0.0), np.float32(0.0636), np.float32(0.1761), np.float32(0.0), np.float32(0.0), np.float32(0.1973), np.float32(0.1789), np.float32(0.1986), np.float32(0.1882)] 
+2026-10-17 08:50:24.208918: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 08:50:24.582751: lr: 0.000828 
+2026-10-17 08:50:24.583230: This epoch took 264.872588 s
+ 
+2026-10-17 08:50:24.583474: 
+epoch:  189 
+2026-10-17 08:54:29.207516: train loss : -0.7561 
+2026-10-17 08:54:34.588585: validation loss: -0.2600 
+2026-10-17 08:54:34.589680: Average global foreground Dice: [np.float32(0.857), np.float32(0.1107), np.float32(0.1776), np.float32(0.2112), np.float32(0.1594), np.float32(0.1723), np.float32(0.1749), np.float32(0.0), np.float32(0.0), np.float32(0.6692), np.float32(0.1416), np.float32(0.1049), np.float32(0.2698), np.float32(0.075), np.float32(0.0), np.float32(0.0), np.float32(0.1759), np.float32(0.1127), np.float32(0.1469), np.float32(0.1275), np.float32(0.0242), np.float32(0.1455), np.float32(0.166), np.float32(0.2677), np.float32(0.0609), np.float32(0.0736), np.float32(0.0), np.float32(0.0), np.float32(0.084), np.float32(0.0807), np.float32(0.0), np.float32(0.0), np.float32(0.1419), np.float32(0.1313), np.float32(0.1522), np.float32(0.1597)] 
+2026-10-17 08:54:34.589893: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 08:54:34.912995: lr: 0.000827 
+2026-10-17 08:54:34.913277: This epoch took 250.329700 s
+ 
+2026-10-17 08:54:34.913376: 
+epoch:  190 
+2026-10-17 08:59:08.557680: train loss : -0.7440 
+2026-10-17 08:59:13.915220: validation loss: -0.4479 
+2026-10-17 08:59:13.916555: Average global foreground Dice: [np.float32(0.8798), np.float32(0.5023), np.float32(0.5127), np.float32(0.5672), np.float32(0.4095), np.float32(0.5934), np.float32(0.5388), np.float32(0.0), np.float32(0.0), np.float32(0.5685), np.float32(0.4021), np.float32(0.463), np.float32(0.3568), np.float32(0.4387), np.float32(0.0), np.float32(0.0), np.float32(0.4462), np.float32(0.3878), np.float32(0.4845), np.float32(0.4213), np.float32(0.4094), np.float32(0.3853), np.float32(0.315), np.float32(0.3669), np.float32(0.2914), np.float32(0.2008), np.float32(0.0), np.float32(0.0), np.float32(0.2982), np.float32(0.3416), np.float32(0.0), np.float32(0.0), np.float32(0.3605), np.float32(0.4518), np.float32(0.4948), np.float32(0.4891)] 
+2026-10-17 08:59:13.916831: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 08:59:14.248610: lr: 0.000826 
+2026-10-17 08:59:14.249083: This epoch took 279.335633 s
+ 
+2026-10-17 08:59:14.249350: 
+epoch:  191 
+2026-10-17 09:03:11.046732: train loss : -0.7659 
+2026-10-17 09:03:16.403488: validation loss: -0.2337 
+2026-10-17 09:03:16.404427: Average global foreground Dice: [np.float32(0.8512), np.float32(0.1824), np.float32(0.1514), np.float32(0.2321), np.float32(0.1619), np.float32(0.1912), np.float32(0.1433), np.float32(0.0), np.float32(0.0), np.float32(0.5697), np.float32(0.143), np.float32(0.1269), np.float32(0.1283), np.float32(0.1177), np.float32(0.0), np.float32(0.0), np.float32(0.182), np.float32(0.1148), np.float32(0.1131), np.float32(0.0685), np.float32(0.1363), np.float32(0.0948), np.float32(0.167), np.float32(0.1764), np.float32(0.0781), np.float32(0.0761), np.float32(0.0), np.float32(0.0), np.float32(0.2432), np.float32(0.2369), np.float32(0.0), np.float32(0.0), np.float32(0.1289), np.float32(0.0855), np.float32(0.2168), np.float32(0.136)] 
+2026-10-17 09:03:16.404691: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 09:03:16.734806: lr: 0.000825 
+2026-10-17 09:03:16.735103: This epoch took 242.485460 s
+ 
+2026-10-17 09:03:16.735199: 
+epoch:  192 
+2026-10-17 09:07:20.055724: train loss : -0.7555 
+2026-10-17 09:07:25.422786: validation loss: -0.4733 
+2026-10-17 09:07:25.424067: Average global foreground Dice: [np.float32(0.8295), np.float32(0.5389), np.float32(0.5013), np.float32(0.6219), np.float32(0.4585), np.float32(0.5869), np.float32(0.5313), np.float32(0.0), np.float32(0.0), np.float32(0.6814), np.float32(0.5674), np.float32(0.5391), np.float32(0.4265), np.float32(0.3925), np.float32(0.0058), np.float32(0.0), np.float32(0.5115), np.float32(0.3683), np.float32(0.5217), np.float32(0.3918), np.float32(0.465), np.float32(0.4725), np.float32(0.3167), np.float32(0.3772), np.float32(0.3226), np.float32(0.3905), np.float32(0.0), np.float32(0.0), np.float32(0.4716), np.float32(0.4064), np.float32(0.0), np.float32(0.0), np.float32(0.458), np.float32(0.5254), np.float32(0.5416), np.float32(0.534)] 
+2026-10-17 09:07:25.424340: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 09:07:25.749408: lr: 0.000824 
+2026-10-17 09:07:25.749702: This epoch took 249.014444 s
+ 
+2026-10-17 09:07:25.749913: 
+epoch:  193 
+2026-10-17 09:11:41.057214: train loss : -0.7617 
+2026-10-17 09:11:46.458698: validation loss: -0.3386 
+2026-10-17 09:11:46.459676: Average global foreground Dice: [np.float32(0.8483), np.float32(0.2893), np.float32(0.3404), np.float32(0.4253), np.float32(0.3035), np.float32(0.4029), np.float32(0.3259), np.float32(0.0), np.float32(0.0), np.float32(0.6344), np.float32(0.3396), np.float32(0.3314), np.float32(0.2707), np.float32(0.2279), np.float32(0.0025), np.float32(0.0), np.float32(0.3186), np.float32(0.2264), np.float32(0.3416), np.float32(0.2316), np.float32(0.3859), np.float32(0.3199), np.float32(0.1908), np.float32(0.2713), np.float32(0.1546), np.float32(0.0646), np.float32(0.0), np.float32(0.0), np.float32(0.3714), np.float32(0.2821), np.float32(0.0), np.float32(0.0), np.float32(0.386), np.float32(0.3686), np.float32(0.3569), np.float32(0.3409)] 
+2026-10-17 09:11:46.460051: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 09:11:46.836667: lr: 0.000824 
+2026-10-17 09:11:46.837327: This epoch took 261.087317 s
+ 
+2026-10-17 09:11:46.837485: 
+epoch:  194 
+2026-10-17 09:15:50.049263: train loss : -0.7664 
+2026-10-17 09:15:55.399652: validation loss: -0.2997 
+2026-10-17 09:15:55.400960: Average global foreground Dice: [np.float32(0.8678), np.float32(0.21), np.float32(0.2496), np.float32(0.2876), np.float32(0.1649), np.float32(0.2695), np.float32(0.2415), np.float32(0.0), np.float32(0.0), np.float32(0.6141), np.float32(0.1784), np.float32(0.1976), np.float32(0.1928), np.float32(0.1067), np.float32(0.0098), np.float32(0.0), np.float32(0.1891), np.float32(0.1277), np.float32(0.186), np.float32(0.1116), np.float32(0.086), np.float32(0.1769), np.float32(0.1542), np.float32(0.3113), np.float32(0.0884), np.float32(0.0673), np.float32(0.0), np.float32(0.0), np.float32(0.2752), np.float32(0.3364), np.float32(0.0), np.float32(0.0), np.float32(0.1464), np.float32(0.1765), np.float32(0.238), np.float32(0.2608)] 
+2026-10-17 09:15:55.401193: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 09:15:55.764326: lr: 0.000823 
+2026-10-17 09:15:55.764680: This epoch took 248.927127 s
+ 
+2026-10-17 09:15:55.764841: 
+epoch:  195 
+2026-10-17 09:20:11.797556: train loss : -0.7751 
+2026-10-17 09:20:17.149807: validation loss: -0.3327 
+2026-10-17 09:20:17.150762: Average global foreground Dice: [np.float32(0.868), np.float32(0.2593), np.float32(0.3076), np.float32(0.3598), np.float32(0.2488), np.float32(0.3302), np.float32(0.2909), np.float32(0.0), np.float32(0.0), np.float32(0.6483), np.float32(0.2721), np.float32(0.2294), np.float32(0.2561), np.float32(0.2444), np.float32(0.0619), np.float32(0.0), np.float32(0.2633), np.float32(0.1622), np.float32(0.2852), np.float32(0.1936), np.float32(0.2141), np.float32(0.2557), np.float32(0.1775), np.float32(0.327), np.float32(0.107), np.float32(0.0848), np.float32(0.0), np.float32(0.0), np.float32(0.2523), np.float32(0.2958), np.float32(0.0), np.float32(0.0), np.float32(0.2401), np.float32(0.2648), np.float32(0.2954), np.float32(0.2905)] 
+2026-10-17 09:20:17.151188: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 09:20:17.487202: lr: 0.000822 
+2026-10-17 09:20:17.487677: This epoch took 261.722732 s
+ 
+2026-10-17 09:20:17.487876: 
+epoch:  196 
+2026-10-17 09:24:49.685977: train loss : -0.7613 
+2026-10-17 09:24:55.039424: validation loss: -0.2870 
+2026-10-17 09:24:55.040479: Average global foreground Dice: [np.float32(0.842), np.float32(0.2145), np.float32(0.2117), np.float32(0.2735), np.float32(0.1968), np.float32(0.1858), np.float32(0.1868), np.float32(0.0), np.float32(0.0), np.float32(0.6409), np.float32(0.1871), np.float32(0.1767), np.float32(0.2481), np.float32(0.2264), np.float32(0.0781), np.float32(0.0), np.float32(0.2069), np.float32(0.1543), np.float32(0.1608), np.float32(0.1281), np.float32(0.0632), np.float32(0.1581), np.float32(0.277), np.float32(0.32), np.float32(0.0879), np.float32(0.0731), np.float32(0.0), np.float32(0.0), np.float32(0.0176), np.float32(0.2476), np.float32(0.0), np.float32(0.0), np.float32(0.1671), np.float32(0.1267), np.float32(0.2486), np.float32(0.1677)] 
+2026-10-17 09:24:55.040934: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 09:24:55.373411: lr: 0.000821 
+2026-10-17 09:24:55.373716: This epoch took 277.885755 s
+ 
+2026-10-17 09:24:55.373811: 
+epoch:  197 
+2026-10-17 09:29:35.373563: train loss : -0.7657 
+2026-10-17 09:29:40.736571: validation loss: -0.3313 
+2026-10-17 09:29:40.737529: Average global foreground Dice: [np.float32(0.8434), np.float32(0.3484), np.float32(0.2487), np.float32(0.3897), np.float32(0.2859), np.float32(0.2876), np.float32(0.2643), np.float32(0.0), np.float32(0.0), np.float32(0.6155), np.float32(0.3107), np.float32(0.3022), np.float32(0.2452), np.float32(0.25), np.float32(0.0186), np.float32(0.0), np.float32(0.3461), np.float32(0.1898), np.float32(0.236), np.float32(0.1495), np.float32(0.1435), np.float32(0.1797), np.float32(0.2193), np.float32(0.323), np.float32(0.2831), np.float32(0.2187), np.float32(0.0), np.float32(0.0), np.float32(0.0671), np.float32(0.2985), np.float32(0.0), np.float32(0.0), np.float32(0.233), np.float32(0.2067), np.float32(0.3184), np.float32(0.2925)] 
+2026-10-17 09:29:40.737820: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 09:29:41.078707: lr: 0.00082 
+2026-10-17 09:29:41.079177: This epoch took 285.705297 s
+ 
+2026-10-17 09:29:41.079467: 
+epoch:  198 
+2026-10-17 09:33:44.096551: train loss : -0.7729 
+2026-10-17 09:33:49.446696: validation loss: -0.5273 
+2026-10-17 09:33:49.447797: Average global foreground Dice: [np.float32(0.8408), np.float32(0.6196), np.float32(0.5857), np.float32(0.7148), np.float32(0.5286), np.float32(0.7057), np.float32(0.6286), np.float32(0.0), np.float32(0.0), np.float32(0.668), np.float32(0.5921), np.float32(0.5964), np.float32(0.5026), np.float32(0.5301), np.float32(0.0), np.float32(0.0), np.float32(0.6188), np.float32(0.4802), np.float32(0.6145), np.float32(0.5201), np.float32(0.5503), np.float32(0.4335), np.float32(0.4559), np.float32(0.532), np.float32(0.3875), np.float32(0.3456), np.float32(0.0), np.float32(0.0), np.float32(0.3472), np.float32(0.4326), np.float32(0.0), np.float32(0.0), np.float32(0.6007), np.float32(0.5552), np.float32(0.6363), np.float32(0.6455)] 
+2026-10-17 09:33:49.448210: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 09:33:49.767237: lr: 0.000819 
+2026-10-17 09:33:49.767578: This epoch took 248.687893 s
+ 
+2026-10-17 09:33:49.767753: 
+epoch:  199 
+2026-10-17 09:38:24.981528: train loss : -0.7652 
+2026-10-17 09:38:30.351945: validation loss: -0.3025 
+2026-10-17 09:38:30.353269: Average global foreground Dice: [np.float32(0.8507), np.float32(0.2209), np.float32(0.2962), np.float32(0.3048), np.float32(0.1968), np.float32(0.2893), np.float32(0.2599), np.float32(0.0), np.float32(0.0), np.float32(0.55), np.float32(0.2058), np.float32(0.22), np.float32(0.2485), np.float32(0.2169), np.float32(0.0084), np.float32(0.0), np.float32(0.2622), np.float32(0.1839), np.float32(0.2714), np.float32(0.2371), np.float32(0.1236), np.float32(0.2401), np.float32(0.1042), np.float32(0.2675), np.float32(0.133), np.float32(0.0644), np.float32(0.0), np.float32(0.0), np.float32(0.2277), np.float32(0.2767), np.float32(0.0), np.float32(0.0), np.float32(0.2674), np.float32(0.1623), np.float32(0.2274), np.float32(0.2216)] 
+2026-10-17 09:38:30.353744: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 09:38:30.699862: lr: 0.000818 
+2026-10-17 09:38:30.700130: saving scheduled checkpoint file... 
+2026-10-17 09:38:30.709150: saving checkpoint... 
+2026-10-17 09:38:30.845191: done, saving took 0.14 seconds 
+2026-10-17 09:38:30.847120: done 
+2026-10-17 09:38:30.847369: This epoch took 281.079428 s
+ 
+2026-10-17 09:38:30.847480: 
+epoch:  200 
+2026-10-17 09:42:25.776106: train loss : -0.7550 
+2026-10-17 09:42:31.134304: validation loss: -0.4181 
+2026-10-17 09:42:31.135230: Average global foreground Dice: [np.float32(0.8411), np.float32(0.431), np.float32(0.4694), np.float32(0.5175), np.float32(0.4044), np.float32(0.4761), np.float32(0.4285), np.float32(0.0), np.float32(0.0), np.float32(0.5985), np.float32(0.4), np.float32(0.3672), np.float32(0.3501), np.float32(0.3595), np.float32(0.0), np.float32(0.0), np.float32(0.4051), np.float32(0.2916), np.float32(0.4207), np.float32(0.2856), np.float32(0.4049), np.float32(0.3696), np.float32(0.2821), np.float32(0.2754), np.float32(0.2314), np.float32(0.0965), np.float32(0.0), np.float32(0.0), np.float32(0.3826), np.float32(0.2991), np.float32(0.0), np.float32(0.0), np.float32(0.4718), np.float32(0.4105), np.float32(0.4288), np.float32(0.4038)] 
+2026-10-17 09:42:31.135760: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 09:42:31.474032: lr: 0.000817 
+2026-10-17 09:42:31.474397: This epoch took 240.626780 s
+ 
+2026-10-17 09:42:31.474639: 
+epoch:  201 
+2026-10-17 09:46:43.916255: train loss : -0.7634 
+2026-10-17 09:46:49.281657: validation loss: -0.3857 
+2026-10-17 09:46:49.282482: Average global foreground Dice: [np.float32(0.8522), np.float32(0.358), np.float32(0.4287), np.float32(0.4968), np.float32(0.3548), np.float32(0.4815), np.float32(0.4048), np.float32(0.0), np.float32(0.0), np.float32(0.6338), np.float32(0.3956), np.float32(0.3869), np.float32(0.2833), np.float32(0.3687), np.float32(0.0016), np.float32(0.0), np.float32(0.3324), np.float32(0.1972), np.float32(0.3969), np.float32(0.2869), np.float32(0.3685), np.float32(0.3228), np.float32(0.248), np.float32(0.2624), np.float32(0.1862), np.float32(0.2008), np.float32(0.0), np.float32(0.0), np.float32(0.3512), np.float32(0.3431), np.float32(0.0), np.float32(0.0), np.float32(0.4311), np.float32(0.4109), np.float32(0.3791), np.float32(0.3727)] 
+2026-10-17 09:46:49.282631: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 09:46:49.621428: lr: 0.000816 
+2026-10-17 09:46:49.621970: This epoch took 258.147166 s
+ 
+2026-10-17 09:46:49.622235: 
+epoch:  202 
+2026-10-17 09:51:21.214432: train loss : -0.7742 
+2026-10-17 09:51:26.568824: validation loss: -0.3882 
+2026-10-17 09:51:26.569875: Average global foreground Dice: [np.float32(0.8607), np.float32(0.3114), np.float32(0.3952), np.float32(0.4782), np.float32(0.3219), np.float32(0.3924), np.float32(0.3439), np.float32(0.0), np.float32(0.0), np.float32(0.6036), np.float32(0.3622), np.float32(0.351), np.float32(0.3693), np.float32(0.3817), np.float32(0.0), np.float32(0.0), np.float32(0.3229), np.float32(0.236), np.float32(0.3621), np.float32(0.2086), np.float32(0.3788), np.float32(0.2913), np.float32(0.2446), np.float32(0.3294), np.float32(0.1578), np.float32(0.1276), np.float32(0.0), np.float32(0.0), np.float32(0.2732), np.float32(0.361), np.float32(0.0), np.float32(0.0), np.float32(0.3468), np.float32(0.4224), np.float32(0.3684), np.float32(0.3738)] 
+2026-10-17 09:51:26.570162: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 09:51:26.939241: lr: 0.000815 
+2026-10-17 09:51:26.939548: This epoch took 277.317098 s
+ 
+2026-10-17 09:51:26.939641: 
+epoch:  203 
+2026-10-17 09:55:31.722827: train loss : -0.7675 
+2026-10-17 09:55:37.091948: validation loss: -0.2729 
+2026-10-17 09:55:37.092900: Average global foreground Dice: [np.float32(0.8769), np.float32(0.1509), np.float32(0.2037), np.float32(0.2598), np.float32(0.1638), np.float32(0.2918), np.float32(0.2026), np.float32(0.0), np.float32(0.0), np.float32(0.5866), np.float32(0.2246), np.float32(0.2229), np.float32(0.1794), np.float32(0.2169), np.float32(0.0438), np.float32(0.0), np.float32(0.1444), np.float32(0.1168), np.float32(0.1862), np.float32(0.1503), np.float32(0.0611), np.float32(0.1696), np.float32(0.1407), np.float32(0.2203), np.float32(0.0734), np.float32(0.0798), np.float32(0.0), np.float32(0.0), np.float32(0.1295), np.float32(0.2084), np.float32(0.0), np.float32(0.0), np.float32(0.1906), np.float32(0.2235), np.float32(0.1881), np.float32(0.2256)] 
+2026-10-17 09:55:37.093059: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 09:55:37.461319: lr: 0.000814 
+2026-10-17 09:55:37.461603: This epoch took 250.521900 s
+ 
+2026-10-17 09:55:37.461687: 
+epoch:  204 
+2026-10-17 09:59:38.690956: train loss : -0.7707 
+2026-10-17 09:59:44.074872: validation loss: -0.2987 
+2026-10-17 09:59:44.076096: Average global foreground Dice: [np.float32(0.8566), np.float32(0.2511), np.float32(0.2993), np.float32(0.3962), np.float32(0.2523), np.float32(0.3035), np.float32(0.2445), np.float32(0.0), np.float32(0.0), np.float32(0.5444), np.float32(0.3102), np.float32(0.2107), np.float32(0.2802), np.float32(0.2609), np.float32(0.0267), np.float32(0.0), np.float32(0.229), np.float32(0.1289), np.float32(0.2951), np.float32(0.2069), np.float32(0.2261), np.float32(0.2565), np.float32(0.1524), np.float32(0.2229), np.float32(0.1411), np.float32(0.048), np.float32(0.0), np.float32(0.0), np.float32(0.2995), np.float32(0.2132), np.float32(0.0), np.float32(0.0), np.float32(0.2915), np.float32(0.2542), np.float32(0.2845), np.float32(0.2744)] 
+2026-10-17 09:59:44.076392: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 09:59:44.414579: lr: 0.000813 
+2026-10-17 09:59:44.414856: This epoch took 246.953107 s
+ 
+2026-10-17 09:59:44.414938: 
+epoch:  205 
+2026-10-17 10:03:36.110251: train loss : -0.7738 
+2026-10-17 10:03:41.459015: validation loss: -0.2564 
+2026-10-17 10:03:41.460100: Average global foreground Dice: [np.float32(0.8591), np.float32(0.1535), np.float32(0.1854), np.float32(0.2352), np.float32(0.1522), np.float32(0.2043), np.float32(0.1645), np.float32(0.0), np.float32(0.0), np.float32(0.57), np.float32(0.1792), np.float32(0.1861), np.float32(0.1533), np.float32(0.1535), np.float32(0.0), np.float32(0.0), np.float32(0.1539), np.float32(0.1025), np.float32(0.1652), np.float32(0.1173), np.float32(0.0834), np.float32(0.1635), np.float32(0.1222), np.float32(0.2453), np.float32(0.0872), np.float32(0.0654), np.float32(0.0), np.float32(0.0), np.float32(0.1837), np.float32(0.1632), np.float32(0.0), np.float32(0.0), np.float32(0.1902), np.float32(0.1615), np.float32(0.1802), np.float32(0.1923)] 
+2026-10-17 10:03:41.460429: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 10:03:41.780065: lr: 0.000813 
+2026-10-17 10:03:41.780506: This epoch took 237.365513 s
+ 
+2026-10-17 10:03:41.780813: 
+epoch:  206 
+2026-10-17 10:07:52.000932: train loss : -0.7715 
+2026-10-17 10:07:57.360732: validation loss: -0.3547 
+2026-10-17 10:07:57.361791: Average global foreground Dice: [np.float32(0.8597), np.float32(0.3254), np.float32(0.3962), np.float32(0.4452), np.float32(0.3157), np.float32(0.4109), np.float32(0.3772), np.float32(0.0), np.float32(0.0), np.float32(0.593), np.float32(0.3431), np.float32(0.2861), np.float32(0.2203), np.float32(0.2069), np.float32(0.0456), np.float32(0.0), np.float32(0.3192), np.float32(0.1947), np.float32(0.3532), np.float32(0.2479), np.float32(0.3894), np.float32(0.3058), np.float32(0.1636), np.float32(0.3123), np.float32(0.1121), np.float32(0.065), np.float32(0.0), np.float32(0.0), np.float32(0.3877), np.float32(0.283), np.float32(0.0), np.float32(0.0), np.float32(0.18), np.float32(0.365), np.float32(0.3483), np.float32(0.3638)] 
+2026-10-17 10:07:57.361972: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 10:07:57.738744: lr: 0.000812 
+2026-10-17 10:07:57.739559: This epoch took 255.958514 s
+ 
+2026-10-17 10:07:57.739772: 
+epoch:  207 
+2026-10-17 10:11:56.778133: train loss : -0.7711 
+2026-10-17 10:12:02.126241: validation loss: -0.5051 
+2026-10-17 10:12:02.127486: Average global foreground Dice: [np.float32(0.8655), np.float32(0.567), np.float32(0.5745), np.float32(0.6336), np.float32(0.5079), np.float32(0.6165), np.float32(0.571), np.float32(0.0), np.float32(0.0), np.float32(0.61), np.float32(0.4272), np.float32(0.4555), np.float32(0.4203), np.float32(0.4542), np.float32(0.0028), np.float32(0.0), np.float32(0.4744), np.float32(0.3474), np.float32(0.5355), np.float32(0.4568), np.float32(0.4722), np.float32(0.4217), np.float32(0.3198), np.float32(0.4734), np.float32(0.2852), np.float32(0.2938), np.float32(0.0), np.float32(0.0), np.float32(0.3182), np.float32(0.3573), np.float32(0.0), np.float32(0.0), np.float32(0.4575), np.float32(0.4368), np.float32(0.5418), np.float32(0.5211)] 
+2026-10-17 10:12:02.127692: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 10:12:02.442441: lr: 0.000811 
+2026-10-17 10:12:02.442814: This epoch took 244.702952 s
+ 
+2026-10-17 10:12:02.442921: 
+epoch:  208 
+2026-10-17 10:16:31.033176: train loss : -0.7665 
+2026-10-17 10:16:36.378696: validation loss: -0.4121 
+2026-10-17 10:16:36.379964: Average global foreground Dice: [np.float32(0.8498), np.float32(0.4229), np.float32(0.4473), np.float32(0.5525), np.float32(0.4102), np.float32(0.5167), np.float32(0.4697), np.float32(0.0), np.float32(0.0), np.float32(0.6418), np.float32(0.431), np.float32(0.4205), np.float32(0.3696), np.float32(0.363), np.float32(0.0746), np.float32(0.0), np.float32(0.407), np.float32(0.2588), np.float32(0.4606), np.float32(0.3333), np.float32(0.4747), np.float32(0.3189), np.float32(0.2489), np.float32(0.3581), np.float32(0.2232), np.float32(0.1393), np.float32(0.0), np.float32(0.0), np.float32(0.4713), np.float32(0.2709), np.float32(0.0), np.float32(0.0), np.float32(0.3588), np.float32(0.4208), np.float32(0.4709), np.float32(0.4301)] 
+2026-10-17 10:16:36.380355: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 10:16:36.721676: lr: 0.00081 
+2026-10-17 10:16:36.722044: This epoch took 274.279055 s
+ 
+2026-10-17 10:16:36.722226: 
+epoch:  209 
+2026-10-17 10:20:51.068808: train loss : -0.7748 
+2026-10-17 10:20:56.418410: validation loss: -0.2889 
+2026-10-17 10:20:56.420240: Average global foreground Dice: [np.float32(0.867), np.float32(0.328), np.float32(0.2689), np.float32(0.3398), np.float32(0.2918), np.float32(0.2688), np.float32(0.1924), np.float32(0.0), np.float32(0.0), np.float32(0.5939), np.float32(0.2419), np.float32(0.2067), np.float32(0.1554), np.float32(0.2154), np.float32(0.0), np.float32(0.0), np.float32(0.2354), np.float32(0.1318), np.float32(0.1562), np.float32(0.1573), np.float32(0.238), np.float32(0.246), np.float32(0.1991), np.float32(0.3119), np.float32(0.1016), np.float32(0.1136), np.float32(0.0), np.float32(0.0), np.float32(0.1868), np.float32(0.1965), np.float32(0.0), np.float32(0.0), np.float32(0.1956), np.float32(0.1894), np.float32(0.27), np.float32(0.2579)] 
+2026-10-17 10:20:56.420966: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 10:20:56.791103: lr: 0.000809 
+2026-10-17 10:20:56.791350: This epoch took 260.069034 s
+ 
+2026-10-17 10:20:56.791432: 
+epoch:  210 
+2026-10-17 10:25:01.261956: train loss : -0.7768 
+2026-10-17 10:25:06.620561: validation loss: -0.3042 
+2026-10-17 10:25:06.621541: Average global foreground Dice: [np.float32(0.8608), np.float32(0.3185), np.float32(0.2839), np.float32(0.3684), np.float32(0.2515), np.float32(0.3245), np.float32(0.2127), np.float32(0.0), np.float32(0.0), np.float32(0.5922), np.float32(0.2501), np.float32(0.2889), np.float32(0.2149), np.float32(0.2561), np.float32(0.0), np.float32(0.0), np.float32(0.2408), np.float32(0.132), np.float32(0.1703), np.float32(0.1291), np.float32(0.3317), np.float32(0.2533), np.float32(0.18), np.float32(0.2694), np.float32(0.1948), np.float32(0.1503), np.float32(0.0), np.float32(0.0), np.float32(0.326), np.float32(0.239), np.float32(0.0), np.float32(0.0), np.float32(0.2193), np.float32(0.256), np.float32(0.2623), np.float32(0.3043)] 
+2026-10-17 10:25:06.621734: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 10:25:06.947665: lr: 0.000808 
+2026-10-17 10:25:06.948021: This epoch took 250.156534 s
+ 
+2026-10-17 10:25:06.948249: 
+epoch:  211 
+2026-10-17 10:29:25.903313: train loss : -0.7785 
+2026-10-17 10:29:31.252349: validation loss: -0.3240 
+2026-10-17 10:29:31.253432: Average global foreground Dice: [np.float32(0.8639), np.float32(0.2755), np.float32(0.3304), np.float32(0.4036), np.float32(0.2914), np.float32(0.3131), np.float32(0.2523), np.float32(0.0), np.float32(0.0), np.float32(0.5673), np.float32(0.2831), np.float32(0.2625), np.float32(0.2129), np.float32(0.2789), np.float32(0.0), np.float32(0.0), np.float32(0.2738), np.float32(0.1622), np.float32(0.229), np.float32(0.1622), np.float32(0.2776), np.float32(0.2851), np.float32(0.1281), np.float32(0.2441), np.float32(0.1441), np.float32(0.1297), np.float32(0.0), np.float32(0.0), np.float32(0.2412), np.float32(0.2359), np.float32(0.0), np.float32(0.0), np.float32(0.2503), np.float32(0.2738), np.float32(0.3088), np.float32(0.2842)] 
+2026-10-17 10:29:31.253855: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 10:29:31.579520: lr: 0.000807 
+2026-10-17 10:29:31.579814: This epoch took 264.631485 s
+ 
+2026-10-17 10:29:31.579911: 
+epoch:  212 
+2026-10-17 10:33:46.833471: train loss : -0.7753 
+2026-10-17 10:33:52.193593: validation loss: -0.3508 
+2026-10-17 10:33:52.194575: Average global foreground Dice: [np.float32(0.8473), np.float32(0.3385), np.float32(0.3689), np.float32(0.4398), np.float32(0.3268), np.float32(0.3314), np.float32(0.2547), np.float32(0.0), np.float32(0.0), np.float32(0.5357), np.float32(0.3241), np.float32(0.2596), np.float32(0.2924), np.float32(0.2916), np.float32(0.0), np.float32(0.0), np.float32(0.2932), np.float32(0.178), np.float32(0.2297), np.float32(0.1821), np.float32(0.3341), np.float32(0.2895), np.float32(0.1589), np.float32(0.3575), np.float32(0.208), np.float32(0.1184), np.float32(0.0), np.float32(0.0), np.float32(0.2055), np.float32(0.2324), np.float32(0.0), np.float32(0.0), np.float32(0.282), np.float32(0.281), np.float32(0.3459), np.float32(0.3463)] 
+2026-10-17 10:33:52.194719: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 10:33:52.521295: lr: 0.000806 
+2026-10-17 10:33:52.521750: This epoch took 260.941771 s
+ 
+2026-10-17 10:33:52.521979: 
+epoch:  213 
+2026-10-17 10:38:12.046849: train loss : -0.7447 
+2026-10-17 10:38:17.387727: validation loss: -0.5663 
+2026-10-17 10:38:17.388709: Average global foreground Dice: [np.float32(0.8701), np.float32(0.6686), np.float32(0.6524), np.float32(0.7716), np.float32(0.601), np.float32(0.7582), np.float32(0.6786), np.float32(0.0), np.float32(0.0), np.float32(0.5883), np.float32(0.6865), np.float32(0.6734), np.float32(0.6116), np.float32(0.4449), np.float32(0.1556), np.float32(0.0), np.float32(0.6811), np.float32(0.5052), np.float32(0.6805), np.float32(0.5144), np.float32(0.5405), np.float32(0.4338), np.float32(0.6932), np.float32(0.5888), np.float32(0.4952), np.float32(0.3689), np.float32(0.0), np.float32(0.0), np.float32(0.4144), np.float32(0.4239), np.float32(0.0), np.float32(0.0), np.float32(0.6689), np.float32(0.6243), np.float32(0.7221), np.float32(0.7268)] 
+2026-10-17 10:38:17.388888: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 10:38:17.765995: lr: 0.000805 
+2026-10-17 10:38:17.766248: This epoch took 265.244160 s
+ 
+2026-10-17 10:38:17.766345: 
+epoch:  214 
+2026-10-17 10:42:21.403182: train loss : -0.7567 
+2026-10-17 10:42:26.776671: validation loss: -0.3433 
+2026-10-17 10:42:26.777954: Average global foreground Dice: [np.float32(0.8702), np.float32(0.3196), np.float32(0.3451), np.float32(0.4477), np.float32(0.3535), np.float32(0.38), np.float32(0.2909), np.float32(0.0), np.float32(0.0), np.float32(0.623), np.float32(0.3789), np.float32(0.3155), np.float32(0.3006), np.float32(0.2611), np.float32(0.0), np.float32(0.0), np.float32(0.4251), np.float32(0.2989), np.float32(0.3183), np.float32(0.2346), np.float32(0.3176), np.float32(0.2969), np.float32(0.2018), np.float32(0.2229), np.float32(0.2178), np.float32(0.0819), np.float32(0.0), np.float32(0.0), np.float32(0.466), np.float32(0.2382), np.float32(0.0), np.float32(0.0), np.float32(0.3404), np.float32(0.2646), np.float32(0.3162), np.float32(0.3569)] 
+2026-10-17 10:42:26.778174: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 10:42:27.119284: lr: 0.000804 
+2026-10-17 10:42:27.119672: This epoch took 249.353259 s
+ 
+2026-10-17 10:42:27.119760: 
+epoch:  215 
+2026-10-17 10:46:09.256581: train loss : -0.7476 
+2026-10-17 10:46:14.626942: validation loss: -0.5884 
+2026-10-17 10:46:14.627879: Average global foreground Dice: [np.float32(0.8486), np.float32(0.7007), np.float32(0.6868), np.float32(0.7651), np.float32(0.6461), np.float32(0.7559), np.float32(0.7073), np.float32(0.0), np.float32(0.0), np.float32(0.6319), np.float32(0.5956), np.float32(0.6226), np.float32(0.5136), np.float32(0.5417), np.float32(0.0149), np.float32(0.0), np.float32(0.6528), np.float32(0.5035), np.float32(0.6553), np.float32(0.5542), np.float32(0.5558), np.float32(0.503), np.float32(0.7936), np.float32(0.784), np.float32(0.3747), np.float32(0.3268), np.float32(0.0), np.float32(0.0), np.float32(0.415), np.float32(0.239), np.float32(0.0), np.float32(0.0), np.float32(0.5612), np.float32(0.5318), np.float32(0.7849), np.float32(0.7875)] 
+2026-10-17 10:46:14.628151: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 10:46:14.952134: lr: 0.000803 
+2026-10-17 10:46:14.952400: This epoch took 227.832572 s
+ 
+2026-10-17 10:46:14.952496: 
+epoch:  216 
+2026-10-17 10:50:50.414821: train loss : -0.7556 
+2026-10-17 10:50:55.785800: validation loss: -0.3375 
+2026-10-17 10:50:55.787000: Average global foreground Dice: [np.float32(0.8364), np.float32(0.2541), np.float32(0.3497), np.float32(0.3462), np.float32(0.2631), np.float32(0.2992), np.float32(0.3013), np.float32(0.0), np.float32(0.0), np.float32(0.5997), np.float32(0.2544), np.float32(0.26), np.float32(0.294), np.float32(0.2827), np.float32(0.0914), np.float32(0.0), np.float32(0.2593), np.float32(0.1803), np.float32(0.2635), np.float32(0.2126), np.float32(0.2593), np.float32(0.3037), np.float32(0.1899), np.float32(0.379), np.float32(0.1181), np.float32(0.1501), np.float32(0.0), np.float32(0.0), np.float32(0.3212), np.float32(0.3203), np.float32(0.0), np.float32(0.0), np.float32(0.2672), np.float32(0.2106), np.float32(0.2811), np.float32(0.3148)] 
+2026-10-17 10:50:55.787167: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 10:50:56.157693: lr: 0.000802 
+2026-10-17 10:50:56.157956: This epoch took 281.205173 s
+ 
+2026-10-17 10:50:56.158036: 
+epoch:  217 
+2026-10-17 10:55:01.681407: train loss : -0.7598 
+2026-10-17 10:55:07.038844: validation loss: -0.3441 
+2026-10-17 10:55:07.039953: Average global foreground Dice: [np.float32(0.8459), np.float32(0.3943), np.float32(0.3047), np.float32(0.3866), np.float32(0.3387), np.float32(0.3274), np.float32(0.2519), np.float32(0.0), np.float32(0.0), np.float32(0.549), np.float32(0.2989), np.float32(0.211), np.float32(0.1854), np.float32(0.1525), np.float32(0.011), np.float32(0.0), np.float32(0.397), np.float32(0.2387), np.float32(0.2172), np.float32(0.1718), np.float32(0.3533), np.float32(0.3514), np.float32(0.1133), np.float32(0.3531), np.float32(0.327), np.float32(0.0342), np.float32(0.0), np.float32(0.0), np.float32(0.3084), np.float32(0.2868), np.float32(0.0), np.float32(0.0), np.float32(0.2439), np.float32(0.36), np.float32(0.4429), np.float32(0.3493)] 
+2026-10-17 10:55:07.040426: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 10:55:07.370183: lr: 0.000801 
+2026-10-17 10:55:07.370656: This epoch took 251.212564 s
+ 
+2026-10-17 10:55:07.370918: 
+epoch:  218 
+2026-10-17 10:59:20.762016: train loss : -0.7710 
+2026-10-17 10:59:26.125268: validation loss: -0.3692 
+2026-10-17 10:59:26.126229: Average global foreground Dice: [np.float32(0.8599), np.float32(0.3736), np.float32(0.3833), np.float32(0.4634), np.float32(0.3221), np.float32(0.4142), np.float32(0.3504), np.float32(0.0), np.float32(0.0), np.float32(0.6025), np.float32(0.3499), np.float32(0.2919), np.float32(0.3226), np.float32(0.2837), np.float32(0.2056), np.float32(0.0), np.float32(0.3309), np.float32(0.2181), np.float32(0.3417), np.float32(0.2541), np.float32(0.288), np.float32(0.335), np.float32(0.1505), np.float32(0.3766), np.float32(0.2557), np.float32(0.1174), np.float32(0.0), np.float32(0.0), np.float32(0.3154), np.float32(0.3205), np.float32(0.0), np.float32(0.0), np.float32(0.3716), np.float32(0.3492), np.float32(0.3475), np.float32(0.4013)] 
+2026-10-17 10:59:26.126494: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 10:59:26.462479: lr: 0.000801 
+2026-10-17 10:59:26.462982: This epoch took 259.091956 s
+ 
+2026-10-17 10:59:26.463304: 
+epoch:  219 
+2026-10-17 11:03:30.741621: train loss : -0.7760 
+2026-10-17 11:03:36.102202: validation loss: -0.4317 
+2026-10-17 11:03:36.103143: Average global foreground Dice: [np.float32(0.8667), np.float32(0.4714), np.float32(0.4738), np.float32(0.572), np.float32(0.4119), np.float32(0.5193), np.float32(0.4265), np.float32(0.0), np.float32(0.0), np.float32(0.5954), np.float32(0.4323), np.float32(0.3404), np.float32(0.3593), np.float32(0.316), np.float32(0.0105), np.float32(0.0), np.float32(0.46), np.float32(0.3193), np.float32(0.4438), np.float32(0.3165), np.float32(0.4054), np.float32(0.3631), np.float32(0.2658), np.float32(0.3412), np.float32(0.2825), np.float32(0.2179), np.float32(0.0), np.float32(0.0), np.float32(0.3886), np.float32(0.2946), np.float32(0.0), np.float32(0.0), np.float32(0.4972), np.float32(0.4719), np.float32(0.5172), np.float32(0.4597)] 
+2026-10-17 11:03:36.103461: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 11:03:36.425519: lr: 0.0008 
+2026-10-17 11:03:36.425802: This epoch took 249.962270 s
+ 
+2026-10-17 11:03:36.426037: 
+epoch:  220 
+2026-10-17 11:08:04.879031: train loss : -0.7732 
+2026-10-17 11:08:10.233008: validation loss: -0.4208 
+2026-10-17 11:08:10.234146: Average global foreground Dice: [np.float32(0.881), np.float32(0.3572), np.float32(0.4306), np.float32(0.5167), np.float32(0.3669), np.float32(0.4534), np.float32(0.4018), np.float32(0.0), np.float32(0.0), np.float32(0.6352), np.float32(0.3688), np.float32(0.2999), np.float32(0.4221), np.float32(0.4005), np.float32(0.1037), np.float32(0.0), np.float32(0.4157), np.float32(0.282), np.float32(0.4095), np.float32(0.2933), np.float32(0.3657), np.float32(0.3582), np.float32(0.1875), np.float32(0.3591), np.float32(0.2271), np.float32(0.1873), np.float32(0.0), np.float32(0.0), np.float32(0.4742), np.float32(0.4264), np.float32(0.0), np.float32(0.0), np.float32(0.4319), np.float32(0.3603), np.float32(0.4241), np.float32(0.3828)] 
+2026-10-17 11:08:10.234666: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 11:08:10.579873: lr: 0.000799 
+2026-10-17 11:08:10.580133: This epoch took 274.154021 s
+ 
+2026-10-17 11:08:10.580229: 
+epoch:  221 
+2026-10-17 11:12:12.494385: train loss : -0.7597 
+2026-10-17 11:12:17.858883: validation loss: -0.3508 
+2026-10-17 11:12:17.859880: Average global foreground Dice: [np.float32(0.8548), np.float32(0.3135), np.float32(0.3391), np.float32(0.4509), np.float32(0.3195), np.float32(0.3932), np.float32(0.3383), np.float32(0.0), np.float32(0.0), np.float32(0.6866), np.float32(0.342), np.float32(0.2867), np.float32(0.3677), np.float32(0.3336), np.float32(0.0), np.float32(0.0), np.float32(0.3307), np.float32(0.214), np.float32(0.354), np.float32(0.2338), np.float32(0.3543), np.float32(0.2893), np.float32(0.1814), np.float32(0.2795), np.float32(0.2683), np.float32(0.061), np.float32(0.0), np.float32(0.0), np.float32(0.3746), np.float32(0.2447), np.float32(0.0), np.float32(0.0), np.float32(0.3706), np.float32(0.4102), np.float32(0.3638), np.float32(0.3387)] 
+2026-10-17 11:12:17.860048: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 11:12:18.190901: lr: 0.000798 
+2026-10-17 11:12:18.191437: This epoch took 247.611137 s
+ 
+2026-10-17 11:12:18.191883: 
+epoch:  222 
+2026-10-17 11:16:39.525859: train loss : -0.7716 
+2026-10-17 11:16:44.890891: validation loss: -0.3983 
+2026-10-17 11:16:44.891755: Average global foreground Dice: [np.float32(0.862), np.float32(0.3626), np.float32(0.3943), np.float32(0.5081), np.float32(0.3671), np.float32(0.4871), np.float32(0.402), np.float32(0.0), np.float32(0.0), np.float32(0.617), np.float32(0.4312), np.float32(0.3906), np.float32(0.4523), np.float32(0.4146), np.float32(0.0), np.float32(0.0), np.float32(0.3857), np.float32(0.2436), np.float32(0.4088), np.float32(0.2864), np.float32(0.3241), np.float32(0.3204), np.float32(0.1837), np.float32(0.3018), np.float32(0.1868), np.float32(0.1019), np.float32(0.0), np.float32(0.0), np.float32(0.3751), np.float32(0.2741), np.float32(0.0), np.float32(0.0), np.float32(0.4528), np.float32(0.437), np.float32(0.3906), np.float32(0.3739)] 
+2026-10-17 11:16:44.892004: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 11:16:45.220405: lr: 0.000797 
+2026-10-17 11:16:45.220814: This epoch took 267.028788 s
+ 
+2026-10-17 11:16:45.221083: 
+epoch:  223 
+2026-10-17 11:21:03.555805: train loss : -0.7796 
+2026-10-17 11:21:08.911892: validation loss: -0.3457 
+2026-10-17 11:21:08.912780: Average global foreground Dice: [np.float32(0.8428), np.float32(0.2593), np.float32(0.3305), np.float32(0.4395), np.float32(0.3015), np.float32(0.3853), np.float32(0.3404), np.float32(0.0), np.float32(0.0), np.float32(0.5797), np.float32(0.3447), np.float32(0.2974), np.float32(0.3479), np.float32(0.3173), np.float32(0.0), np.float32(0.0), np.float32(0.3103), np.float32(0.2077), np.float32(0.3231), np.float32(0.2099), np.float32(0.2594), np.float32(0.2847), np.float32(0.1637), np.float32(0.3306), np.float32(0.2094), np.float32(0.07), np.float32(0.0), np.float32(0.0), np.float32(0.3306), np.float32(0.275), np.float32(0.0), np.float32(0.0), np.float32(0.4123), np.float32(0.3741), np.float32(0.3425), np.float32(0.3158)] 
+2026-10-17 11:21:08.912934: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 11:21:09.275819: lr: 0.000796 
+2026-10-17 11:21:09.276034: This epoch took 264.054737 s
+ 
+2026-10-17 11:21:09.276094: 
+epoch:  224 
+2026-10-17 11:25:15.391815: train loss : -0.7783 
+2026-10-17 11:25:20.755509: validation loss: -0.3446 
+2026-10-17 11:25:20.756589: Average global foreground Dice: [np.float32(0.8487), np.float32(0.2916), np.float32(0.3443), np.float32(0.4455), np.float32(0.2996), np.float32(0.4093), np.float32(0.3436), np.float32(0.0), np.float32(0.0), np.float32(0.6601), np.float32(0.3516), np.float32(0.3096), np.float32(0.4086), np.float32(0.3568), np.float32(0.0688), np.float32(0.0), np.float32(0.3529), np.float32(0.2413), np.float32(0.3521), np.float32(0.2314), np.float32(0.2427), np.float32(0.2666), np.float32(0.1885), np.float32(0.3075), np.float32(0.1752), np.float32(0.0888), np.float32(0.0), np.float32(0.0), np.float32(0.3731), np.float32(0.3121), np.float32(0.0), np.float32(0.0), np.float32(0.396), np.float32(0.4593), np.float32(0.3516), np.float32(0.3238)] 
+2026-10-17 11:25:20.756961: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 11:25:21.082183: lr: 0.000795 
+2026-10-17 11:25:21.082652: This epoch took 251.806506 s
+ 
+2026-10-17 11:25:21.082985: 
+epoch:  225 
+2026-10-17 11:29:13.634843: train loss : -0.7803 
+2026-10-17 11:29:19.000459: validation loss: -0.3009 
+2026-10-17 11:29:19.001551: Average global foreground Dice: [np.float32(0.8625), np.float32(0.2547), np.float32(0.2904), np.float32(0.3157), np.float32(0.2389), np.float32(0.3295), np.float32(0.2592), np.float32(0.0), np.float32(0.0), np.float32(0.615), np.float32(0.2138), np.float32(0.2078), np.float32(0.2408), np.float32(0.2154), np.float32(0.0304), np.float32(0.0), np.float32(0.2512), np.float32(0.1949), np.float32(0.2145), np.float32(0.1397), np.float32(0.1758), np.float32(0.2511), np.float32(0.1137), np.float32(0.264), np.float32(0.1527), np.float32(0.0657), np.float32(0.0), np.float32(0.0), np.float32(0.2334), np.float32(0.043), np.float32(0.0), np.float32(0.0), np.float32(0.2296), np.float32(0.2948), np.float32(0.2616), np.float32(0.2908)] 
+2026-10-17 11:29:19.002017: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 11:29:19.340804: lr: 0.000794 
+2026-10-17 11:29:19.341314: This epoch took 238.258092 s
+ 
+2026-10-17 11:29:19.341712: 
+epoch:  226 
+2026-10-17 11:33:14.592959: train loss : -0.7706 
+2026-10-17 11:33:19.974738: validation loss: -0.3071 
+2026-10-17 11:33:19.975687: Average global foreground Dice: [np.float32(0.8522), np.float32(0.2555), np.float32(0.3191), np.float32(0.3424), np.float32(0.2563), np.float32(0.3534), np.float32(0.2792), np.float32(0.0), np.float32(0.0), np.float32(0.6163), np.float32(0.2905), np.float32(0.2361), np.float32(0.2436), np.float32(0.1913), np.float32(0.1178), np.float32(0.0), np.float32(0.2687), np.float32(0.1868), np.float32(0.2855), np.float32(0.1702), np.float32(0.3073), np.float32(0.2457), np.float32(0.1604), np.float32(0.1716), np.float32(0.1426), np.float32(0.068), np.float32(0.0), np.float32(0.0), np.float32(0.2442), np.float32(0.2553), np.float32(0.0), np.float32(0.0), np.float32(0.2135), np.float32(0.1987), np.float32(0.3303), np.float32(0.2825)] 
+2026-10-17 11:33:19.976066: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 11:33:20.356102: lr: 0.000793 
+2026-10-17 11:33:20.356609: This epoch took 241.014659 s
+ 
+2026-10-17 11:33:20.356824: 
+epoch:  227 
+2026-10-17 11:37:32.059076: train loss : -0.7760 
+2026-10-17 11:37:37.439505: validation loss: -0.2782 
+2026-10-17 11:37:37.440786: Average global foreground Dice: [np.float32(0.833), np.float32(0.2472), np.float32(0.2279), np.float32(0.3012), np.float32(0.271), np.float32(0.2262), np.float32(0.1851), np.float32(0.0), np.float32(0.0), np.float32(0.5874), np.float32(0.2142), np.float32(0.1828), np.float32(0.2041), np.float32(0.1513), np.float32(0.0331), np.float32(0.0), np.float32(0.29), np.float32(0.1919), np.float32(0.1771), np.float32(0.0943), np.float32(0.2726), np.float32(0.2173), np.float32(0.2158), np.float32(0.2192), np.float32(0.1592), np.float32(0.0482), np.float32(0.0), np.float32(0.0), np.float32(0.3008), np.float32(0.2103), np.float32(0.0), np.float32(0.0), np.float32(0.2169), np.float32(0.2015), np.float32(0.2916), np.float32(0.2337)] 
+2026-10-17 11:37:37.440980: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 11:37:37.955734: lr: 0.000792 
+2026-10-17 11:37:37.956092: This epoch took 257.599183 s
+ 
+2026-10-17 11:37:37.956291: 
+epoch:  228 
+2026-10-17 11:42:02.428746: train loss : -0.7763 
+2026-10-17 11:42:07.802951: validation loss: -0.3258 
+2026-10-17 11:42:07.804302: Average global foreground Dice: [np.float32(0.8639), np.float32(0.3108), np.float32(0.3434), np.float32(0.3912), np.float32(0.2827), np.float32(0.3581), np.float32(0.2949), np.float32(0.0), np.float32(0.0), np.float32(0.5644), np.float32(0.2803), np.float32(0.2579), np.float32(0.2549), np.float32(0.2326), np.float32(0.0), np.float32(0.0), np.float32(0.2879), np.float32(0.2119), np.float32(0.2958), np.float32(0.178), np.float32(0.2719), np.float32(0.2592), np.float32(0.1792), np.float32(0.3277), np.float32(0.1294), np.float32(0.082), np.float32(0.0), np.float32(0.0), np.float32(0.3123), np.float32(0.3159), np.float32(0.0), np.float32(0.0), np.float32(0.1793), np.float32(0.2791), np.float32(0.3322), np.float32(0.3224)] 
+2026-10-17 11:42:07.804547: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 11:42:08.133214: lr: 0.000791 
+2026-10-17 11:42:08.133485: This epoch took 270.177105 s
+ 
+2026-10-17 11:42:08.133579: 
+epoch:  229 
+2026-10-17 11:46:47.635360: train loss : -0.7819 
+2026-10-17 11:46:52.990843: validation loss: -0.3365 
+2026-10-17 11:46:52.991820: Average global foreground Dice: [np.float32(0.8481), np.float32(0.293), np.float32(0.3254), np.float32(0.3987), np.float32(0.2871), np.float32(0.3816), np.float32(0.3084), np.float32(0.0), np.float32(0.0), np.float32(0.6176), np.float32(0.2987), np.float32(0.2652), np.float32(0.2315), np.float32(0.2185), np.float32(0.1733), np.float32(0.0), np.float32(0.3278), np.float32(0.2058), np.float32(0.3359), np.float32(0.2064), np.float32(0.3596), np.float32(0.2777), np.float32(0.1538), np.float32(0.2154), np.float32(0.2543), np.float32(0.0734), np.float32(0.0), np.float32(0.0), np.float32(0.4033), np.float32(0.1977), np.float32(0.0), np.float32(0.0), np.float32(0.1836), np.float32(0.3211), np.float32(0.3329), np.float32(0.3263)] 
+2026-10-17 11:46:52.992256: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 11:46:53.322957: lr: 0.00079 
+2026-10-17 11:46:53.323339: This epoch took 285.189696 s
+ 
+2026-10-17 11:46:53.323462: 
+epoch:  230 
+2026-10-17 11:51:18.534161: train loss : -0.7820 
+2026-10-17 11:51:23.884445: validation loss: -0.3093 
+2026-10-17 11:51:23.885547: Average global foreground Dice: [np.float32(0.8541), np.float32(0.2567), np.float32(0.2999), np.float32(0.3595), np.float32(0.2443), np.float32(0.3391), np.float32(0.2738), np.float32(0.0), np.float32(0.0), np.float32(0.6333), np.float32(0.2577), np.float32(0.2267), np.float32(0.2543), np.float32(0.1965), np.float32(0.2111), np.float32(0.0), np.float32(0.2319), np.float32(0.15), np.float32(0.272), np.float32(0.1597), np.float32(0.2533), np.float32(0.2299), np.float32(0.2446), np.float32(0.2635), np.float32(0.1747), np.float32(0.0458), np.float32(0.0), np.float32(0.0), np.float32(0.4261), np.float32(0.2533), np.float32(0.0), np.float32(0.0), np.float32(0.21), np.float32(0.2739), np.float32(0.3049), np.float32(0.3224)] 
+2026-10-17 11:51:23.885743: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 11:51:24.217245: lr: 0.000789 
+2026-10-17 11:51:24.217514: This epoch took 270.893984 s
+ 
+2026-10-17 11:51:24.217831: 
+epoch:  231 
+2026-10-17 11:54:57.032085: train loss : -0.7888 
+2026-10-17 11:55:02.391240: validation loss: -0.3162 
+2026-10-17 11:55:02.392426: Average global foreground Dice: [np.float32(0.8389), np.float32(0.2469), np.float32(0.3138), np.float32(0.3248), np.float32(0.2335), np.float32(0.2999), np.float32(0.2766), np.float32(0.0), np.float32(0.0), np.float32(0.6397), np.float32(0.2329), np.float32(0.2175), np.float32(0.3051), np.float32(0.2881), np.float32(0.1524), np.float32(0.0), np.float32(0.2201), np.float32(0.1392), np.float32(0.2588), np.float32(0.1563), np.float32(0.2609), np.float32(0.2695), np.float32(0.131), np.float32(0.2917), np.float32(0.1228), np.float32(0.0621), np.float32(0.0), np.float32(0.0), np.float32(0.3445), np.float32(0.2831), np.float32(0.0), np.float32(0.0), np.float32(0.2083), np.float32(0.2164), np.float32(0.2678), np.float32(0.2792)] 
+2026-10-17 11:55:02.392866: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 11:55:02.718043: lr: 0.000789 
+2026-10-17 11:55:02.718299: This epoch took 218.500360 s
+ 
+2026-10-17 11:55:02.718378: 
+epoch:  232 
+2026-10-17 11:59:03.777189: train loss : -0.7800 
+2026-10-17 11:59:09.164135: validation loss: -0.3455 
+2026-10-17 11:59:09.165164: Average global foreground Dice: [np.float32(0.8203), np.float32(0.335), np.float32(0.3471), np.float32(0.3979), np.float32(0.2814), np.float32(0.3783), np.float32(0.3298), np.float32(0.0), np.float32(0.0), np.float32(0.5775), np.float32(0.2801), np.float32(0.2669), np.float32(0.2511), np.float32(0.2742), np.float32(0.0077), np.float32(0.0), np.float32(0.3224), np.float32(0.2283), np.float32(0.2763), np.float32(0.1843), np.float32(0.2012), np.float32(0.3234), np.float32(0.177), np.float32(0.3615), np.float32(0.1346), np.float32(0.1658), np.float32(0.0), np.float32(0.0), np.float32(0.2775), np.float32(0.1449), np.float32(0.0), np.float32(0.0), np.float32(0.1899), np.float32(0.307), np.float32(0.3339), np.float32(0.3508)] 
+2026-10-17 11:59:09.165573: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 11:59:09.494602: lr: 0.000788 
+2026-10-17 11:59:09.495106: This epoch took 246.776442 s
+ 
+2026-10-17 11:59:09.495486: 
+epoch:  233 
+2026-10-17 12:03:39.035271: train loss : -0.7752 
+2026-10-17 12:03:44.385506: validation loss: -0.3390 
+2026-10-17 12:03:44.386685: Average global foreground Dice: [np.float32(0.8584), np.float32(0.2641), np.float32(0.2929), np.float32(0.3411), np.float32(0.2318), np.float32(0.3037), np.float32(0.2707), np.float32(0.0), np.float32(0.0), np.float32(0.6306), np.float32(0.2544), np.float32(0.1685), np.float32(0.2662), np.float32(0.2387), np.float32(0.3054), np.float32(0.0), np.float32(0.3045), np.float32(0.218), np.float32(0.2723), np.float32(0.1751), np.float32(0.1232), np.float32(0.2511), np.float32(0.143), np.float32(0.3897), np.float32(0.1615), np.float32(0.0805), np.float32(0.0), np.float32(0.0), np.float32(0.1492), np.float32(0.3541), np.float32(0.0), np.float32(0.0), np.float32(0.1894), np.float32(0.2286), np.float32(0.2856), np.float32(0.2761)] 
+2026-10-17 12:03:44.386943: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 12:03:44.716661: lr: 0.000787 
+2026-10-17 12:03:44.717195: This epoch took 275.221466 s
+ 
+2026-10-17 12:03:44.717446: 
+epoch:  234 
+2026-10-17 12:07:55.803776: train loss : -0.7841 
+2026-10-17 12:08:01.160936: validation loss: -0.2874 
+2026-10-17 12:08:01.162177: Average global foreground Dice: [np.float32(0.8625), np.float32(0.2131), np.float32(0.2477), np.float32(0.2837), np.float32(0.1794), np.float32(0.2616), np.float32(0.2199), np.float32(0.0), np.float32(0.0), np.float32(0.6525), np.float32(0.2049), np.float32(0.1751), np.float32(0.2228), np.float32(0.1848), np.float32(0.1299), np.float32(0.0), np.float32(0.1811), np.float32(0.1343), np.float32(0.2158), np.float32(0.1289), np.float32(0.0741), np.float32(0.1992), np.float32(0.0918), np.float32(0.3256), np.float32(0.0893), np.float32(0.049), np.float32(0.0), np.float32(0.0), np.float32(0.0941), np.float32(0.2065), np.float32(0.0), np.float32(0.0), np.float32(0.1768), np.float32(0.1939), np.float32(0.2005), np.float32(0.2331)] 
+2026-10-17 12:08:01.162679: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 12:08:01.497728: lr: 0.000786 
+2026-10-17 12:08:01.498271: This epoch took 256.780636 s
+ 
+2026-10-17 12:08:01.498612: 
+epoch:  235 
+2026-10-17 12:12:10.315067: train loss : -0.7808 
+2026-10-17 12:12:15.684023: validation loss: -0.2901 
+2026-10-17 12:12:15.685203: Average global foreground Dice: [np.float32(0.8455), np.float32(0.1944), np.float32(0.2442), np.float32(0.2598), np.float32(0.1928), np.float32(0.258), np.float32(0.2003), np.float32(0.0), np.float32(0.0), np.float32(0.5256), np.float32(0.1768), np.float32(0.1284), np.float32(0.2188), np.float32(0.1999), np.float32(0.0), np.float32(0.0), np.float32(0.1765), np.float32(0.1212), np.float32(0.1964), np.float32(0.0983), np.float32(0.1362), np.float32(0.1561), np.float32(0.1541), np.float32(0.3255), np.float32(0.1111), np.float32(0.081), np.float32(0.0), np.float32(0.0), np.float32(0.1234), np.float32(0.2975), np.float32(0.0), np.float32(0.0), np.float32(0.1495), np.float32(0.1783), np.float32(0.205), np.float32(0.2741)] 
+2026-10-17 12:12:15.685548: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 12:12:16.051988: lr: 0.000785 
+2026-10-17 12:12:16.052412: This epoch took 254.553630 s
+ 
+2026-10-17 12:12:16.052682: 
+epoch:  236 
+2026-10-17 12:16:51.602349: train loss : -0.7770 
+2026-10-17 12:16:56.963286: validation loss: -0.2975 
+2026-10-17 12:16:56.964211: Average global foreground Dice: [np.float32(0.8634), np.float32(0.2101), np.float32(0.2518), np.float32(0.3153), np.float32(0.221), np.float32(0.2599), np.float32(0.2303), np.float32(0.0), np.float32(0.0), np.float32(0.5545), np.float32(0.1612), np.float32(0.1516), np.float32(0.3209), np.float32(0.3314), np.float32(0.0145), np.float32(0.0), np.float32(0.2374), np.float32(0.1627), np.float32(0.2089), np.float32(0.1322), np.float32(0.0661), np.float32(0.1944), np.float32(0.1425), np.float32(0.3429), np.float32(0.0859), np.float32(0.0845), np.float32(0.0), np.float32(0.0), np.float32(0.095), np.float32(0.3082), np.float32(0.0), np.float32(0.0), np.float32(0.1912), np.float32(0.196), np.float32(0.2356), np.float32(0.2104)] 
+2026-10-17 12:16:56.964392: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 12:16:57.300123: lr: 0.000784 
+2026-10-17 12:16:57.300600: This epoch took 281.247694 s
+ 
+2026-10-17 12:16:57.300924: 
+epoch:  237 
+2026-10-17 12:21:33.613116: train loss : -0.7834 
+2026-10-17 12:21:38.977780: validation loss: -0.5087 
+2026-10-17 12:21:38.978857: Average global foreground Dice: [np.float32(0.8562), np.float32(0.5548), np.float32(0.5741), np.float32(0.6759), np.float32(0.5644), np.float32(0.6719), np.float32(0.5876), np.float32(0.0), np.float32(0.0), np.float32(0.5976), np.float32(0.539), np.float32(0.5451), np.float32(0.3418), np.float32(0.23), np.float32(0.0297), np.float32(0.0), np.float32(0.5996), np.float32(0.4477), np.float32(0.5988), np.float32(0.4745), np.float32(0.3814), np.float32(0.3682), np.float32(0.4893), np.float32(0.5268), np.float32(0.3198), np.float32(0.3542), np.float32(0.0), np.float32(0.0), np.float32(0.3777), np.float32(0.4832), np.float32(0.0), np.float32(0.0), np.float32(0.5042), np.float32(0.4596), np.float32(0.6375), np.float32(0.6622)] 
+2026-10-17 12:21:38.979081: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 12:21:39.329831: lr: 0.000783 
+2026-10-17 12:21:39.330056: This epoch took 282.029021 s
+ 
+2026-10-17 12:21:39.330134: 
+epoch:  238 
+2026-10-17 12:25:41.571717: train loss : -0.7658 
+2026-10-17 12:25:46.948677: validation loss: -0.3636 
+2026-10-17 12:25:46.949894: Average global foreground Dice: [np.float32(0.8363), np.float32(0.2874), np.float32(0.3874), np.float32(0.4307), np.float32(0.3331), np.float32(0.4339), np.float32(0.3649), np.float32(0.0), np.float32(0.0), np.float32(0.5747), np.float32(0.2842), np.float32(0.2655), np.float32(0.245), np.float32(0.2409), np.float32(0.0302), np.float32(0.0), np.float32(0.3413), np.float32(0.2333), np.float32(0.3139), np.float32(0.1902), np.float32(0.2531), np.float32(0.2578), np.float32(0.2404), np.float32(0.2871), np.float32(0.1862), np.float32(0.066), np.float32(0.0), np.float32(0.0), np.float32(0.382), np.float32(0.2678), np.float32(0.0), np.float32(0.0), np.float32(0.3035), np.float32(0.331), np.float32(0.3827), np.float32(0.391)] 
+2026-10-17 12:25:46.950356: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 12:25:47.301913: lr: 0.000782 
+2026-10-17 12:25:47.302513: This epoch took 247.972324 s
+ 
+2026-10-17 12:25:47.302711: 
+epoch:  239 
+2026-10-17 12:30:35.595731: train loss : -0.7587 
+2026-10-17 12:30:40.961785: validation loss: -0.2359 
+2026-10-17 12:30:40.963102: Average global foreground Dice: [np.float32(0.8575), np.float32(0.1384), np.float32(0.2144), np.float32(0.1822), np.float32(0.1197), np.float32(0.1993), np.float32(0.1766), np.float32(0.0), np.float32(0.0), np.float32(0.5949), np.float32(0.1541), np.float32(0.1141), np.float32(0.1679), np.float32(0.1078), np.float32(0.1098), np.float32(0.0), np.float32(0.1348), np.float32(0.0903), np.float32(0.1641), np.float32(0.117), np.float32(0.0873), np.float32(0.1366), np.float32(0.1526), np.float32(0.1991), np.float32(0.0788), np.float32(0.0693), np.float32(0.0), np.float32(0.0), np.float32(0.1906), np.float32(0.1937), np.float32(0.0), np.float32(0.0), np.float32(0.1267), np.float32(0.137), np.float32(0.1903), np.float32(0.167)] 
+2026-10-17 12:30:40.963392: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 12:30:41.297706: lr: 0.000781 
+2026-10-17 12:30:41.298131: This epoch took 293.995333 s
+ 
+2026-10-17 12:30:41.298424: 
+epoch:  240 
+2026-10-17 12:34:41.330333: train loss : -0.7819 
+2026-10-17 12:34:46.724653: validation loss: -0.1770 
+2026-10-17 12:34:46.725555: Average global foreground Dice: [np.float32(0.839), np.float32(0.0777), np.float32(0.1003), np.float32(0.1216), np.float32(0.0797), np.float32(0.1163), np.float32(0.1019), np.float32(0.0), np.float32(0.0), np.float32(0.6213), np.float32(0.0826), np.float32(0.0857), np.float32(0.1423), np.float32(0.1102), np.float32(0.1219), np.float32(0.0), np.float32(0.0804), np.float32(0.056), np.float32(0.0891), np.float32(0.0761), np.float32(0.0156), np.float32(0.0661), np.float32(0.1044), np.float32(0.1486), np.float32(0.0279), np.float32(0.0246), np.float32(0.0), np.float32(0.0), np.float32(0.1122), np.float32(0.1368), np.float32(0.0), np.float32(0.0), np.float32(0.0772), np.float32(0.0946), np.float32(0.1094), np.float32(0.0696)] 
+2026-10-17 12:34:46.725861: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 12:34:47.060434: lr: 0.00078 
+2026-10-17 12:34:47.060690: This epoch took 245.762141 s
+ 
+2026-10-17 12:34:47.060782: 
+epoch:  241 
+2026-10-17 12:39:23.024341: train loss : -0.7495 
+2026-10-17 12:39:28.382060: validation loss: -0.4289 
+2026-10-17 12:39:28.383131: Average global foreground Dice: [np.float32(0.8544), np.float32(0.4001), np.float32(0.3697), np.float32(0.464), np.float32(0.3678), np.float32(0.5754), np.float32(0.4594), np.float32(0.0), np.float32(0.0), np.float32(0.5712), np.float32(0.3133), np.float32(0.3925), np.float32(0.2503), np.float32(0.3768), np.float32(0.0), np.float32(0.0), np.float32(0.3494), np.float32(0.2513), np.float32(0.3377), np.float32(0.3238), np.float32(0.3485), np.float32(0.2928), np.float32(0.5162), np.float32(0.5964), np.float32(0.2075), np.float32(0.2438), np.float32(0.0), np.float32(0.0), np.float32(0.1646), np.float32(0.3525), np.float32(0.0), np.float32(0.0), np.float32(0.36), np.float32(0.5257), np.float32(0.472), np.float32(0.564)] 
+2026-10-17 12:39:28.383314: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 12:39:28.713478: lr: 0.000779 
+2026-10-17 12:39:28.713939: This epoch took 281.653090 s
+ 
+2026-10-17 12:39:28.714208: 
+epoch:  242 
+2026-10-17 12:43:43.815756: train loss : -0.7731 
+2026-10-17 12:43:49.187187: validation loss: -0.3173 
+2026-10-17 12:43:49.188386: Average global foreground Dice: [np.float32(0.857), np.float32(0.2056), np.float32(0.2864), np.float32(0.2414), np.float32(0.2453), np.float32(0.3113), np.float32(0.2713), np.float32(0.0), np.float32(0.0), np.float32(0.6411), np.float32(0.1755), np.float32(0.1977), np.float32(0.1123), np.float32(0.1894), np.float32(0.0), np.float32(0.0), np.float32(0.2576), np.float32(0.185), np.float32(0.2327), np.float32(0.2326), np.float32(0.1208), np.float32(0.2159), np.float32(0.5334), np.float32(0.4795), np.float32(0.0896), np.float32(0.1457), np.float32(0.0), np.float32(0.0), np.float32(0.2726), np.float32(0.2052), np.float32(0.0), np.float32(0.0), np.float32(0.2479), np.float32(0.238), np.float32(0.3733), np.float32(0.3512)] 
+2026-10-17 12:43:49.189000: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 12:43:49.555660: lr: 0.000778 
+2026-10-17 12:43:49.555953: This epoch took 260.841566 s
+ 
+2026-10-17 12:43:49.556278: 
+epoch:  243 
+2026-10-17 12:48:13.112153: train loss : -0.7870 
+2026-10-17 12:48:18.465651: validation loss: -0.2758 
+2026-10-17 12:48:18.466632: Average global foreground Dice: [np.float32(0.852), np.float32(0.1795), np.float32(0.2252), np.float32(0.2099), np.float32(0.1402), np.float32(0.2452), np.float32(0.2211), np.float32(0.0), np.float32(0.0), np.float32(0.5239), np.float32(0.154), np.float32(0.1592), np.float32(0.1612), np.float32(0.1599), np.float32(0.0046), np.float32(0.0), np.float32(0.1301), np.float32(0.0888), np.float32(0.1976), np.float32(0.1645), np.float32(0.0505), np.float32(0.2263), np.float32(0.3454), np.float32(0.3622), np.float32(0.0851), np.float32(0.1057), np.float32(0.0), np.float32(0.0), np.float32(0.1251), np.float32(0.2982), np.float32(0.0), np.float32(0.0), np.float32(0.1654), np.float32(0.2384), np.float32(0.2257), np.float32(0.2614)] 
+2026-10-17 12:48:18.467016: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 12:48:18.796537: lr: 0.000777 
+2026-10-17 12:48:18.796818: This epoch took 269.240316 s
+ 
+2026-10-17 12:48:18.796920: 
+epoch:  244 
+2026-10-17 12:52:45.312079: train loss : -0.7830 
+2026-10-17 12:52:50.666387: validation loss: -0.2722 
+2026-10-17 12:52:50.667434: Average global foreground Dice: [np.float32(0.8712), np.float32(0.168), np.float32(0.2257), np.float32(0.1793), np.float32(0.1243), np.float32(0.2052), np.float32(0.1466), np.float32(0.0), np.float32(0.0), np.float32(0.6093), np.float32(0.1057), np.float32(0.109), np.float32(0.1408), np.float32(0.1251), np.float32(0.0915), np.float32(0.0), np.float32(0.1661), np.float32(0.1272), np.float32(0.1516), np.float32(0.0927), np.float32(0.0448), np.float32(0.2724), np.float32(0.223), np.float32(0.2939), np.float32(0.0665), np.float32(0.0899), np.float32(0.0), np.float32(0.0), np.float32(0.1838), np.float32(0.3175), np.float32(0.0), np.float32(0.0), np.float32(0.1154), np.float32(0.2143), np.float32(0.2051), np.float32(0.2132)] 
+2026-10-17 12:52:50.667819: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 12:52:50.998833: lr: 0.000777 
+2026-10-17 12:52:50.999071: This epoch took 272.202069 s
+ 
+2026-10-17 12:52:50.999149: 
+epoch:  245 
+2026-10-17 12:57:02.109952: train loss : -0.7783 
+2026-10-17 12:57:07.471693: validation loss: -0.1919 
+2026-10-17 12:57:07.472865: Average global foreground Dice: [np.float32(0.8317), np.float32(0.0779), np.float32(0.1346), np.float32(0.0667), np.float32(0.0594), np.float32(0.0799), np.float32(0.0551), np.float32(0.0), np.float32(0.0), np.float32(0.5629), np.float32(0.096), np.float32(0.0296), np.float32(0.0688), np.float32(0.0347), np.float32(0.0496), np.float32(0.0), np.float32(0.0605), np.float32(0.0407), np.float32(0.0656), np.float32(0.0625), np.float32(0.016), np.float32(0.115), np.float32(0.1055), np.float32(0.2638), np.float32(0.0397), np.float32(0.0438), np.float32(0.0), np.float32(0.0), np.float32(0.1487), np.float32(0.1756), np.float32(0.0), np.float32(0.0), np.float32(0.0191), np.float32(0.0533), np.float32(0.1215), np.float32(0.1103)] 
+2026-10-17 12:57:07.473104: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 12:57:07.810917: lr: 0.000776 
+2026-10-17 12:57:07.811347: This epoch took 256.812143 s
+ 
+2026-10-17 12:57:07.811486: 
+epoch:  246 
+2026-10-17 13:01:06.522590: train loss : -0.7876 
+2026-10-17 13:01:11.872640: validation loss: -0.2478 
+2026-10-17 13:01:11.873758: Average global foreground Dice: [np.float32(0.8516), np.float32(0.1413), np.float32(0.1891), np.float32(0.2391), np.float32(0.135), np.float32(0.2006), np.float32(0.1811), np.float32(0.0), np.float32(0.0), np.float32(0.6807), np.float32(0.1459), np.float32(0.1243), np.float32(0.1503), np.float32(0.1125), np.float32(0.0), np.float32(0.0), np.float32(0.1501), np.float32(0.109), np.float32(0.1824), np.float32(0.117), np.float32(0.028), np.float32(0.1932), np.float32(0.0716), np.float32(0.2573), np.float32(0.0658), np.float32(0.062), np.float32(0.0), np.float32(0.0), np.float32(0.0597), np.float32(0.2398), np.float32(0.0), np.float32(0.0), np.float32(0.118), np.float32(0.14), np.float32(0.1657), np.float32(0.1731)] 
+2026-10-17 13:01:11.874260: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 13:01:12.215180: lr: 0.000775 
+2026-10-17 13:01:12.215415: This epoch took 244.403845 s
+ 
+2026-10-17 13:01:12.215495: 
+epoch:  247 
+2026-10-17 13:05:32.453957: train loss : -0.7829 
+2026-10-17 13:05:37.821101: validation loss: -0.3857 
+2026-10-17 13:05:37.822085: Average global foreground Dice: [np.float32(0.8571), np.float32(0.3302), np.float32(0.4183), np.float32(0.474), np.float32(0.4014), np.float32(0.4874), np.float32(0.3992), np.float32(0.0), np.float32(0.0), np.float32(0.5396), np.float32(0.3269), np.float32(0.3582), np.float32(0.2252), np.float32(0.3406), np.float32(0.0204), np.float32(0.0), np.float32(0.4006), np.float32(0.297), np.float32(0.4032), np.float32(0.3007), np.float32(0.2548), np.float32(0.369), np.float32(0.2768), np.float32(0.395), np.float32(0.1856), np.float32(0.1667), np.float32(0.0), np.float32(0.0), np.float32(0.2363), np.float32(0.1912), np.float32(0.0), np.float32(0.0), np.float32(0.2438), np.float32(0.3848), np.float32(0.4391), np.float32(0.4237)] 
+2026-10-17 13:05:37.822512: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 13:05:38.154186: lr: 0.000774 
+2026-10-17 13:05:38.154413: This epoch took 265.938858 s
+ 
+2026-10-17 13:05:38.154490: 
+epoch:  248 
+2026-10-17 13:09:50.201580: train loss : -0.7864 
+2026-10-17 13:09:55.556836: validation loss: -0.2916 
+2026-10-17 13:09:55.557823: Average global foreground Dice: [np.float32(0.8765), np.float32(0.2406), np.float32(0.2898), np.float32(0.3272), np.float32(0.2768), np.float32(0.2609), np.float32(0.2396), np.float32(0.0), np.float32(0.0), np.float32(0.6604), np.float32(0.2291), np.float32(0.181), np.float32(0.2303), np.float32(0.1736), np.float32(0.0), np.float32(0.0), np.float32(0.2917), np.float32(0.1962), np.float32(0.2413), np.float32(0.1516), np.float32(0.2025), np.float32(0.2764), np.float32(0.1131), np.float32(0.2095), np.float32(0.1636), np.float32(0.0851), np.float32(0.0), np.float32(0.0), np.float32(0.1086), np.float32(0.2436), np.float32(0.0), np.float32(0.0), np.float32(0.2657), np.float32(0.209), np.float32(0.2356), np.float32(0.2617)] 
+2026-10-17 13:09:55.558110: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 13:09:55.917474: lr: 0.000773 
+2026-10-17 13:09:55.917763: This epoch took 257.763209 s
+ 
+2026-10-17 13:09:55.917863: 
+epoch:  249 
+2026-10-17 13:13:45.976468: train loss : -0.7920 
+2026-10-17 13:13:51.346775: validation loss: -0.2370 
+2026-10-17 13:13:51.347774: Average global foreground Dice: [np.float32(0.8515), np.float32(0.1602), np.float32(0.2188), np.float32(0.2118), np.float32(0.1142), np.float32(0.207), np.float32(0.1832), np.float32(0.0), np.float32(0.0), np.float32(0.6565), np.float32(0.1363), np.float32(0.1048), np.float32(0.1429), np.float32(0.132), np.float32(0.0667), np.float32(0.0), np.float32(0.1313), np.float32(0.0979), np.float32(0.1881), np.float32(0.1369), np.float32(0.0585), np.float32(0.2066), np.float32(0.0635), np.float32(0.1867), np.float32(0.1026), np.float32(0.0367), np.float32(0.0), np.float32(0.0), np.float32(0.0775), np.float32(0.2506), np.float32(0.0), np.float32(0.0), np.float32(0.1019), np.float32(0.2388), np.float32(0.1467), np.float32(0.2096)] 
+2026-10-17 13:13:51.348199: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 13:13:51.681257: lr: 0.000772 
+2026-10-17 13:13:51.681774: saving scheduled checkpoint file... 
+2026-10-17 13:13:51.689131: saving checkpoint... 
+2026-10-17 13:13:51.822397: done, saving took 0.14 seconds 
+2026-10-17 13:13:51.824207: done 
+2026-10-17 13:13:51.824424: This epoch took 235.906494 s
+ 
+2026-10-17 13:13:51.824530: 
+epoch:  250 
+2026-10-17 13:17:52.591253: train loss : -0.7862 
+2026-10-17 13:17:57.946355: validation loss: -0.2696 
+2026-10-17 13:17:57.947403: Average global foreground Dice: [np.float32(0.8713), np.float32(0.1622), np.float32(0.2454), np.float32(0.2159), np.float32(0.1629), np.float32(0.2135), np.float32(0.2116), np.float32(0.0), np.float32(0.0), np.float32(0.6307), np.float32(0.1225), np.float32(0.1361), np.float32(0.1365), np.float32(0.0546), np.float32(0.0166), np.float32(0.0), np.float32(0.1776), np.float32(0.1226), np.float32(0.2046), np.float32(0.1269), np.float32(0.1115), np.float32(0.1804), np.float32(0.0961), np.float32(0.3299), np.float32(0.1118), np.float32(0.0752), np.float32(0.0), np.float32(0.0), np.float32(0.0821), np.float32(0.2583), np.float32(0.0), np.float32(0.0), np.float32(0.1103), np.float32(0.0945), np.float32(0.1894), np.float32(0.1862)] 
+2026-10-17 13:17:57.947864: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 13:17:58.280156: lr: 0.000771 
+2026-10-17 13:17:58.280573: This epoch took 246.455954 s
+ 
+2026-10-17 13:17:58.280862: 
+epoch:  251 
+2026-10-17 13:22:13.824841: train loss : -0.7871 
+2026-10-17 13:22:19.189923: validation loss: -0.3003 
+2026-10-17 13:22:19.191204: Average global foreground Dice: [np.float32(0.851), np.float32(0.1913), np.float32(0.2727), np.float32(0.2951), np.float32(0.1725), np.float32(0.2534), np.float32(0.2394), np.float32(0.0), np.float32(0.0), np.float32(0.656), np.float32(0.2013), np.float32(0.1731), np.float32(0.2153), np.float32(0.146), np.float32(0.1248), np.float32(0.0), np.float32(0.222), np.float32(0.1621), np.float32(0.2221), np.float32(0.1585), np.float32(0.0874), np.float32(0.2289), np.float32(0.0758), np.float32(0.2989), np.float32(0.1033), np.float32(0.1041), np.float32(0.0), np.float32(0.0), np.float32(0.0325), np.float32(0.3254), np.float32(0.0), np.float32(0.0), np.float32(0.1411), np.float32(0.0623), np.float32(0.1951), np.float32(0.1816)] 
+2026-10-17 13:22:19.191786: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 13:22:19.528209: lr: 0.00077 
+2026-10-17 13:22:19.528675: This epoch took 261.247585 s
+ 
+2026-10-17 13:22:19.528995: 
+epoch:  252 
+2026-10-17 13:26:20.921498: train loss : -0.7957 
+2026-10-17 13:26:26.292627: validation loss: -0.2072 
+2026-10-17 13:26:26.293810: Average global foreground Dice: [np.float32(0.8357), np.float32(0.1331), np.float32(0.1751), np.float32(0.1783), np.float32(0.1151), np.float32(0.0986), np.float32(0.1318), np.float32(0.0), np.float32(0.0), np.float32(0.6094), np.float32(0.1058), np.float32(0.0492), np.float32(0.0587), np.float32(0.0166), np.float32(0.0709), np.float32(0.0), np.float32(0.1278), np.float32(0.0876), np.float32(0.115), np.float32(0.0926), np.float32(0.0613), np.float32(0.1455), np.float32(0.1225), np.float32(0.3028), np.float32(0.0769), np.float32(0.0719), np.float32(0.0), np.float32(0.0), np.float32(0.0628), np.float32(0.2052), np.float32(0.0), np.float32(0.0), np.float32(0.0822), np.float32(0.0104), np.float32(0.1487), np.float32(0.0873)] 
+2026-10-17 13:26:26.294020: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 13:26:26.634663: lr: 0.000769 
+2026-10-17 13:26:26.634962: This epoch took 247.105682 s
+ 
+2026-10-17 13:26:26.635060: 
+epoch:  253 
+2026-10-17 13:30:41.590121: train loss : -0.7868 
+2026-10-17 13:30:46.950225: validation loss: -0.2675 
+2026-10-17 13:30:46.951670: Average global foreground Dice: [np.float32(0.8501), np.float32(0.2245), np.float32(0.2061), np.float32(0.2636), np.float32(0.2053), np.float32(0.2162), np.float32(0.1947), np.float32(0.0), np.float32(0.0), np.float32(0.5894), np.float32(0.2235), np.float32(0.1298), np.float32(0.1305), np.float32(0.0685), np.float32(0.0), np.float32(0.0), np.float32(0.2428), np.float32(0.1688), np.float32(0.1863), np.float32(0.1223), np.float32(0.3174), np.float32(0.2358), np.float32(0.0707), np.float32(0.2331), np.float32(0.1348), np.float32(0.0713), np.float32(0.0), np.float32(0.0), np.float32(0.1225), np.float32(0.2558), np.float32(0.0), np.float32(0.0), np.float32(0.099), np.float32(0.2097), np.float32(0.1594), np.float32(0.2311)] 
+2026-10-17 13:30:46.951895: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 13:30:47.295077: lr: 0.000768 
+2026-10-17 13:30:47.295560: This epoch took 260.660435 s
+ 
+2026-10-17 13:30:47.295913: 
+epoch:  254 
+2026-10-17 13:35:12.721276: train loss : -0.7911 
+2026-10-17 13:35:18.086622: validation loss: -0.2879 
+2026-10-17 13:35:18.087769: Average global foreground Dice: [np.float32(0.8799), np.float32(0.1917), np.float32(0.2388), np.float32(0.2797), np.float32(0.177), np.float32(0.2649), np.float32(0.2352), np.float32(0.0), np.float32(0.0), np.float32(0.6294), np.float32(0.2347), np.float32(0.1867), np.float32(0.234), np.float32(0.2009), np.float32(0.1851), np.float32(0.0), np.float32(0.2144), np.float32(0.149), np.float32(0.2402), np.float32(0.157), np.float32(0.2172), np.float32(0.2334), np.float32(0.1253), np.float32(0.2668), np.float32(0.1145), np.float32(0.053), np.float32(0.0), np.float32(0.0), np.float32(0.1305), np.float32(0.2677), np.float32(0.0), np.float32(0.0), np.float32(0.1867), np.float32(0.1766), np.float32(0.2086), np.float32(0.2152)] 
+2026-10-17 13:35:18.088156: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 13:35:18.424628: lr: 0.000767 
+2026-10-17 13:35:18.424893: This epoch took 271.128864 s
+ 
+2026-10-17 13:35:18.425162: 
+epoch:  255 
+2026-10-17 13:39:52.038481: train loss : -0.7862 
+2026-10-17 13:39:57.403883: validation loss: -0.2555 
+2026-10-17 13:39:57.405087: Average global foreground Dice: [np.float32(0.8302), np.float32(0.1685), np.float32(0.2555), np.float32(0.2556), np.float32(0.164), np.float32(0.247), np.float32(0.1952), np.float32(0.0), np.float32(0.0), np.float32(0.6005), np.float32(0.1533), np.float32(0.1444), np.float32(0.1381), np.float32(0.0744), np.float32(0.1501), np.float32(0.0), np.float32(0.1922), np.float32(0.1385), np.float32(0.2232), np.float32(0.1585), np.float32(0.1117), np.float32(0.2842), np.float32(0.1245), np.float32(0.3177), np.float32(0.081), np.float32(0.0847), np.float32(0.0), np.float32(0.0), np.float32(0.2208), np.float32(0.3076), np.float32(0.0), np.float32(0.0), np.float32(0.1351), np.float32(0.0845), np.float32(0.1948), np.float32(0.228)] 
+2026-10-17 13:39:57.405375: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 13:39:57.748140: lr: 0.000766 
+2026-10-17 13:39:57.748383: This epoch took 279.323057 s
+ 
+2026-10-17 13:39:57.748464: 
+epoch:  256 
+2026-10-17 13:44:01.540547: train loss : -0.7886 
+2026-10-17 13:44:06.898014: validation loss: -0.2086 
+2026-10-17 13:44:06.899229: Average global foreground Dice: [np.float32(0.8348), np.float32(0.1145), np.float32(0.1498), np.float32(0.1831), np.float32(0.0982), np.float32(0.1435), np.float32(0.1169), np.float32(0.0), np.float32(0.0), np.float32(0.6654), np.float32(0.1375), np.float32(0.1237), np.float32(0.1397), np.float32(0.0894), np.float32(0.0), np.float32(0.0), np.float32(0.1158), np.float32(0.0918), np.float32(0.1201), np.float32(0.083), np.float32(0.0477), np.float32(0.18), np.float32(0.0589), np.float32(0.2108), np.float32(0.0488), np.float32(0.0448), np.float32(0.0), np.float32(0.0), np.float32(0.109), np.float32(0.1708), np.float32(0.0), np.float32(0.0), np.float32(0.1072), np.float32(0.0998), np.float32(0.1307), np.float32(0.1467)] 
+2026-10-17 13:44:06.899691: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 13:44:07.225262: lr: 0.000765 
+2026-10-17 13:44:07.225638: This epoch took 249.477121 s
+ 
+2026-10-17 13:44:07.225857: 
+epoch:  257 
+2026-10-17 13:48:32.174179: train loss : -0.7886 
+2026-10-17 13:48:37.540166: validation loss: -0.2348 
+2026-10-17 13:48:37.541250: Average global foreground Dice: [np.float32(0.8265), np.float32(0.1379), np.float32(0.2303), np.float32(0.219), np.float32(0.1519), np.float32(0.1703), np.float32(0.1865), np.float32(0.0), np.float32(0.0), np.float32(0.6618), np.float32(0.1138), np.float32(0.1088), np.float32(0.0559), np.float32(0.0628), np.float32(0.0029), np.float32(0.0), np.float32(0.1505), np.float32(0.103), np.float32(0.1815), np.float32(0.1423), np.float32(0.0647), np.float32(0.1805), np.float32(0.1009), np.float32(0.2596), np.float32(0.0644), np.float32(0.0905), np.float32(0.0), np.float32(0.0), np.float32(0.2365), np.float32(0.2677), np.float32(0.0), np.float32(0.0), np.float32(0.1101), np.float32(0.0181), np.float32(0.1904), np.float32(0.1291)] 
+2026-10-17 13:48:37.541637: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 13:48:37.875869: lr: 0.000764 
+2026-10-17 13:48:37.876191: This epoch took 270.650248 s
+ 
+2026-10-17 13:48:37.876312: 
+epoch:  258 
+2026-10-17 13:52:47.229070: train loss : -0.7909 
+2026-10-17 13:52:52.600936: validation loss: -0.2414 
+2026-10-17 13:52:52.602184: Average global foreground Dice: [np.float32(0.8585), np.float32(0.1295), np.float32(0.1804), np.float32(0.2187), np.float32(0.1282), np.float32(0.1676), np.float32(0.1581), np.float32(0.0), np.float32(0.0), np.float32(0.5903), np.float32(0.1357), np.float32(0.0984), np.float32(0.1933), np.float32(0.0795), np.float32(0.0025), np.float32(0.0), np.float32(0.1512), np.float32(0.1054), np.float32(0.158), np.float32(0.0953), np.float32(0.0938), np.float32(0.1483), np.float32(0.0855), np.float32(0.2817), np.float32(0.076), np.float32(0.0554), np.float32(0.0), np.float32(0.0), np.float32(0.0595), np.float32(0.2178), np.float32(0.0), np.float32(0.0), np.float32(0.116), np.float32(0.0869), np.float32(0.1572), np.float32(0.1365)] 
+2026-10-17 13:52:52.602861: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 13:52:52.957252: lr: 0.000764 
+2026-10-17 13:52:52.957755: This epoch took 255.081385 s
+ 
+2026-10-17 13:52:52.958058: 
+epoch:  259 
+2026-10-17 13:56:36.895233: train loss : -0.7981 
+2026-10-17 13:56:42.262258: validation loss: -0.3842 
+2026-10-17 13:56:42.263273: Average global foreground Dice: [np.float32(0.8217), np.float32(0.407), np.float32(0.4249), np.float32(0.4841), np.float32(0.4006), np.float32(0.489), np.float32(0.4521), np.float32(0.0), np.float32(0.0), np.float32(0.6263), np.float32(0.3664), np.float32(0.3749), np.float32(0.2398), np.float32(0.2722), np.float32(0.0362), np.float32(0.0), np.float32(0.3803), np.float32(0.2787), np.float32(0.4033), np.float32(0.2896), np.float32(0.4259), np.float32(0.3369), np.float32(0.3314), np.float32(0.2722), np.float32(0.1846), np.float32(0.1958), np.float32(0.0), np.float32(0.0), np.float32(0.4011), np.float32(0.2773), np.float32(0.0), np.float32(0.0), np.float32(0.3669), np.float32(0.4493), np.float32(0.4194), np.float32(0.3997)] 
+2026-10-17 13:56:42.263528: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 13:56:42.585894: lr: 0.000763 
+2026-10-17 13:56:42.586250: This epoch took 229.627922 s
+ 
+2026-10-17 13:56:42.586391: 
+epoch:  260 
+2026-10-17 14:00:50.129986: train loss : -0.7938 
+2026-10-17 14:00:55.476695: validation loss: -0.3642 
+2026-10-17 14:00:55.477710: Average global foreground Dice: [np.float32(0.825), np.float32(0.3514), np.float32(0.4006), np.float32(0.4455), np.float32(0.3328), np.float32(0.4066), np.float32(0.3694), np.float32(0.0), np.float32(0.0), np.float32(0.5127), np.float32(0.3173), np.float32(0.2752), np.float32(0.1987), np.float32(0.205), np.float32(0.0363), np.float32(0.0), np.float32(0.3461), np.float32(0.2474), np.float32(0.3558), np.float32(0.2358), np.float32(0.3313), np.float32(0.3521), np.float32(0.2411), np.float32(0.3656), np.float32(0.1859), np.float32(0.1173), np.float32(0.0), np.float32(0.0), np.float32(0.4134), np.float32(0.3241), np.float32(0.0), np.float32(0.0), np.float32(0.3676), np.float32(0.3226), np.float32(0.3764), np.float32(0.3623)] 
+2026-10-17 14:00:55.477893: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 14:00:55.839617: lr: 0.000762 
+2026-10-17 14:00:55.839884: This epoch took 253.253413 s
+ 
+2026-10-17 14:00:55.839980: 
+epoch:  261 
+2026-10-17 14:05:01.717327: train loss : -0.7880 
+2026-10-17 14:05:07.072401: validation loss: -0.3554 
+2026-10-17 14:05:07.073315: Average global foreground Dice: [np.float32(0.8726), np.float32(0.3461), np.float32(0.3808), np.float32(0.3488), np.float32(0.3484), np.float32(0.3436), np.float32(0.3757), np.float32(0.0), np.float32(0.0), np.float32(0.627), np.float32(0.2576), np.float32(0.2562), np.float32(0.1818), np.float32(0.151), np.float32(0.1024), np.float32(0.0), np.float32(0.3269), np.float32(0.225), np.float32(0.3601), np.float32(0.2849), np.float32(0.2774), np.float32(0.3112), np.float32(0.2132), np.float32(0.3775), np.float32(0.1194), np.float32(0.1191), np.float32(0.0), np.float32(0.0), np.float32(0.1015), np.float32(0.2333), np.float32(0.0), np.float32(0.0), np.float32(0.2896), np.float32(0.1807), np.float32(0.2961), np.float32(0.3381)] 
+2026-10-17 14:05:07.073833: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 14:05:07.390865: lr: 0.000761 
+2026-10-17 14:05:07.391284: This epoch took 251.551241 s
+ 
+2026-10-17 14:05:07.391592: 
+epoch:  262 
+2026-10-17 14:09:17.281857: train loss : -0.7892 
+2026-10-17 14:09:22.650298: validation loss: -0.2847 
+2026-10-17 14:09:22.651409: Average global foreground Dice: [np.float32(0.8627), np.float32(0.1661), np.float32(0.2039), np.float32(0.2486), np.float32(0.1455), np.float32(0.2032), np.float32(0.1902), np.float32(0.0), np.float32(0.0), np.float32(0.6689), np.float32(0.1478), np.float32(0.1447), np.float32(0.2491), np.float32(0.2008), np.float32(0.2329), np.float32(0.0), np.float32(0.1736), np.float32(0.131), np.float32(0.1692), np.float32(0.1307), np.float32(0.0368), np.float32(0.1538), np.float32(0.1301), np.float32(0.2888), np.float32(0.0941), np.float32(0.0649), np.float32(0.0), np.float32(0.0), np.float32(0.0712), np.float32(0.2883), np.float32(0.0), np.float32(0.0), np.float32(0.1551), np.float32(0.1402), np.float32(0.1863), np.float32(0.1599)] 
+2026-10-17 14:09:22.651711: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 14:09:22.982412: lr: 0.00076 
+2026-10-17 14:09:22.982652: This epoch took 255.590835 s
+ 
+2026-10-17 14:09:22.982731: 
+epoch:  263 
+2026-10-17 14:13:35.491299: train loss : -0.7853 
+2026-10-17 14:13:40.851440: validation loss: -0.2706 
+2026-10-17 14:13:40.852600: Average global foreground Dice: [np.float32(0.8693), np.float32(0.1566), np.float32(0.2151), np.float32(0.2603), np.float32(0.1647), np.float32(0.2148), np.float32(0.2043), np.float32(0.0), np.float32(0.0), np.float32(0.6291), np.float32(0.1666), np.float32(0.1273), np.float32(0.2674), np.float32(0.2756), np.float32(0.036), np.float32(0.0), np.float32(0.198), np.float32(0.1385), np.float32(0.1801), np.float32(0.1479), np.float32(0.0209), np.float32(0.1463), np.float32(0.1491), np.float32(0.3277), np.float32(0.0623), np.float32(0.0617), np.float32(0.0), np.float32(0.0), np.float32(0.0506), np.float32(0.2602), np.float32(0.0), np.float32(0.0), np.float32(0.279), np.float32(0.1272), np.float32(0.1853), np.float32(0.1628)] 
+2026-10-17 14:13:40.853043: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 14:13:41.231924: lr: 0.000759 
+2026-10-17 14:13:41.232200: This epoch took 258.249418 s
+ 
+2026-10-17 14:13:41.232292: 
+epoch:  264 
+2026-10-17 14:17:54.727742: train loss : -0.7838 
+2026-10-17 14:18:00.093045: validation loss: -0.2630 
+2026-10-17 14:18:00.094144: Average global foreground Dice: [np.float32(0.8621), np.float32(0.1694), np.float32(0.2004), np.float32(0.2478), np.float32(0.1406), np.float32(0.2), np.float32(0.1696), np.float32(0.0), np.float32(0.0), np.float32(0.6458), np.float32(0.1857), np.float32(0.1307), np.float32(0.2473), np.float32(0.168), np.float32(0.0), np.float32(0.0), np.float32(0.1851), np.float32(0.1611), np.float32(0.1703), np.float32(0.1141), np.float32(0.0413), np.float32(0.1674), np.float32(0.0997), np.float32(0.2729), np.float32(0.067), np.float32(0.0946), np.float32(0.0), np.float32(0.0), np.float32(0.0748), np.float32(0.2634), np.float32(0.0), np.float32(0.0), np.float32(0.0622), np.float32(0.188), np.float32(0.1816), np.float32(0.1493)] 
+2026-10-17 14:18:00.094583: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 14:18:00.462085: lr: 0.000758 
+2026-10-17 14:18:00.462338: This epoch took 259.229983 s
+ 
+2026-10-17 14:18:00.462520: 
+epoch:  265 
+2026-10-17 14:22:02.396331: train loss : -0.7840 
+2026-10-17 14:22:07.744929: validation loss: -0.3088 
+2026-10-17 14:22:07.745889: Average global foreground Dice: [np.float32(0.8413), np.float32(0.2017), np.float32(0.2863), np.float32(0.2282), np.float32(0.1705), np.float32(0.2929), np.float32(0.3437), np.float32(0.0), np.float32(0.0), np.float32(0.5962), np.float32(0.1711), np.float32(0.2121), np.float32(0.1438), np.float32(0.1386), np.float32(0.0), np.float32(0.0), np.float32(0.1587), np.float32(0.1387), np.float32(0.2725), np.float32(0.1633), np.float32(0.1663), np.float32(0.2458), np.float32(0.1368), np.float32(0.3972), np.float32(0.0977), np.float32(0.1184), np.float32(0.0), np.float32(0.0), np.float32(0.1562), np.float32(0.2739), np.float32(0.0), np.float32(0.0), np.float32(0.0884), np.float32(0.2272), np.float32(0.1985), np.float32(0.2479)] 
+2026-10-17 14:22:07.746101: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 14:22:08.076822: lr: 0.000757 
+2026-10-17 14:22:08.077273: This epoch took 247.614662 s
+ 
+2026-10-17 14:22:08.077595: 
+epoch:  266 
+2026-10-17 14:26:11.480971: train loss : -0.7889 
+2026-10-17 14:26:16.824259: validation loss: -0.2672 
+2026-10-17 14:26:16.825334: Average global foreground Dice: [np.float32(0.8818), np.float32(0.2165), np.float32(0.2526), np.float32(0.2786), np.float32(0.1753), np.float32(0.2582), np.float32(0.2386), np.float32(0.0), np.float32(0.0), np.float32(0.2855), np.float32(0.1793), np.float32(0.1481), np.float32(0.2065), np.float32(0.1905), np.float32(0.0), np.float32(0.0), np.float32(0.1843), np.float32(0.1411), np.float32(0.2253), np.float32(0.1672), np.float32(0.11), np.float32(0.2151), np.float32(0.2081), np.float32(0.2761), np.float32(0.106), np.float32(0.1446), np.float32(0.0), np.float32(0.0), np.float32(0.1073), np.float32(0.273), np.float32(0.0), np.float32(0.0), np.float32(0.1256), np.float32(0.2478), np.float32(0.1826), np.float32(0.2171)] 
+2026-10-17 14:26:16.825759: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 14:26:17.149789: lr: 0.000756 
+2026-10-17 14:26:17.150222: This epoch took 249.072425 s
+ 
+2026-10-17 14:26:17.150694: 
+epoch:  267 
+2026-10-17 14:30:13.497987: train loss : -0.7777 
+2026-10-17 14:30:18.845776: validation loss: -0.4339 
+2026-10-17 14:30:18.847160: Average global foreground Dice: [np.float32(0.8526), np.float32(0.4762), np.float32(0.4357), np.float32(0.5519), np.float32(0.4275), np.float32(0.5062), np.float32(0.4936), np.float32(0.0), np.float32(0.0), np.float32(0.1319), np.float32(0.4489), np.float32(0.4647), np.float32(0.4057), np.float32(0.399), np.float32(0.0), np.float32(0.0), np.float32(0.4542), np.float32(0.3146), np.float32(0.4554), np.float32(0.3534), np.float32(0.3632), np.float32(0.3574), np.float32(0.3779), np.float32(0.4742), np.float32(0.2846), np.float32(0.2743), np.float32(0.0), np.float32(0.0), np.float32(0.3431), np.float32(0.4763), np.float32(0.0), np.float32(0.0), np.float32(0.3981), np.float32(0.4028), np.float32(0.5285), np.float32(0.5284)] 
+2026-10-17 14:30:18.847692: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 14:30:19.225935: lr: 0.000755 
+2026-10-17 14:30:19.226281: This epoch took 242.075267 s
+ 
+2026-10-17 14:30:19.226390: 
+epoch:  268 
+2026-10-17 14:34:59.236715: train loss : -0.7875 
+2026-10-17 14:35:04.581914: validation loss: -0.3086 
+2026-10-17 14:35:04.582838: Average global foreground Dice: [np.float32(0.8054), np.float32(0.2772), np.float32(0.2643), np.float32(0.3227), np.float32(0.2121), np.float32(0.3307), np.float32(0.2993), np.float32(0.0), np.float32(0.0), np.float32(0.5903), np.float32(0.2681), np.float32(0.2661), np.float32(0.273), np.float32(0.2659), np.float32(0.0524), np.float32(0.0), np.float32(0.2259), np.float32(0.1433), np.float32(0.2181), np.float32(0.1656), np.float32(0.2424), np.float32(0.2333), np.float32(0.2118), np.float32(0.3015), np.float32(0.1693), np.float32(0.2223), np.float32(0.0), np.float32(0.0), np.float32(0.2101), np.float32(0.1851), np.float32(0.0), np.float32(0.0), np.float32(0.1653), np.float32(0.2017), np.float32(0.254), np.float32(0.4029)] 
+2026-10-17 14:35:04.583077: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 14:35:04.915347: lr: 0.000754 
+2026-10-17 14:35:04.915629: This epoch took 285.689169 s
+ 
+2026-10-17 14:35:04.915849: 
+epoch:  269 
+2026-10-17 14:39:13.981709: train loss : -0.7827 
+2026-10-17 14:39:19.334367: validation loss: -0.4700 
+2026-10-17 14:39:19.335482: Average global foreground Dice: [np.float32(0.8559), np.float32(0.4831), np.float32(0.4949), np.float32(0.6343), np.float32(0.4601), np.float32(0.6119), np.float32(0.5208), np.float32(0.0), np.float32(0.0), np.float32(0.5912), np.float32(0.5387), np.float32(0.5142), np.float32(0.5647), np.float32(0.4916), np.float32(0.2752), np.float32(0.0), np.float32(0.5305), np.float32(0.386), np.float32(0.5289), np.float32(0.3933), np.float32(0.4194), np.float32(0.3467), np.float32(0.3164), np.float32(0.4376), np.float32(0.3304), np.float32(0.29), np.float32(0.0), np.float32(0.0), np.float32(0.3182), np.float32(0.492), np.float32(0.0), np.float32(0.0), np.float32(0.4915), np.float32(0.5305), np.float32(0.519), np.float32(0.5296)] 
+2026-10-17 14:39:19.335801: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 14:39:19.655724: lr: 0.000753 
+2026-10-17 14:39:19.656060: This epoch took 254.740114 s
+ 
+2026-10-17 14:39:19.656235: 
+epoch:  270 
+2026-10-17 14:43:34.206811: train loss : -0.7875 
+2026-10-17 14:43:39.563069: validation loss: -0.5613 
+2026-10-17 14:43:39.564104: Average global foreground Dice: [np.float32(0.8469), np.float32(0.6294), np.float32(0.6387), np.float32(0.7834), np.float32(0.637), np.float32(0.7995), np.float32(0.735), np.float32(0.0), np.float32(0.0), np.float32(0.5971), np.float32(0.6382), np.float32(0.6447), np.float32(0.5377), np.float32(0.6022), np.float32(0.0119), np.float32(0.0), np.float32(0.7046), np.float32(0.5261), np.float32(0.6772), np.float32(0.52), np.float32(0.5527), np.float32(0.4087), np.float32(0.4239), np.float32(0.4272), np.float32(0.2986), np.float32(0.3913), np.float32(0.0), np.float32(0.0), np.float32(0.319), np.float32(0.343), np.float32(0.0), np.float32(0.0), np.float32(0.5246), np.float32(0.5926), np.float32(0.7042), np.float32(0.6952)] 
+2026-10-17 14:43:39.564446: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 14:43:39.891674: lr: 0.000752 
+2026-10-17 14:43:39.891926: This epoch took 260.235623 s
+ 
+2026-10-17 14:43:39.892009: 
+epoch:  271 
+2026-10-17 14:47:52.741823: train loss : -0.7854 
+2026-10-17 14:47:58.089662: validation loss: -0.4138 
+2026-10-17 14:47:58.090858: Average global foreground Dice: [np.float32(0.8691), np.float32(0.4124), np.float32(0.4132), np.float32(0.5314), np.float32(0.402), np.float32(0.4992), np.float32(0.459), np.float32(0.0), np.float32(0.0), np.float32(0.4657), np.float32(0.4449), np.float32(0.3972), np.float32(0.3702), np.float32(0.3599), np.float32(0.0082), np.float32(0.0), np.float32(0.4609), np.float32(0.343), np.float32(0.4632), np.float32(0.3498), np.float32(0.4169), np.float32(0.3736), np.float32(0.2578), np.float32(0.2986), np.float32(0.3371), np.float32(0.2147), np.float32(0.0), np.float32(0.0), np.float32(0.3925), np.float32(0.3131), np.float32(0.0), np.float32(0.0), np.float32(0.4493), np.float32(0.4602), np.float32(0.47), np.float32(0.4719)] 
+2026-10-17 14:47:58.091123: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 14:47:58.491575: lr: 0.000751 
+2026-10-17 14:47:58.491827: This epoch took 258.599762 s
+ 
+2026-10-17 14:47:58.491907: 
+epoch:  272 
+2026-10-17 14:52:15.904960: train loss : -0.7922 
+2026-10-17 14:52:21.267146: validation loss: -0.4135 
+2026-10-17 14:52:21.268204: Average global foreground Dice: [np.float32(0.8473), np.float32(0.3805), np.float32(0.4218), np.float32(0.5376), np.float32(0.3835), np.float32(0.4922), np.float32(0.4191), np.float32(0.0), np.float32(0.0), np.float32(0.4726), np.float32(0.4105), np.float32(0.3845), np.float32(0.4853), np.float32(0.4288), np.float32(0.1626), np.float32(0.0), np.float32(0.4633), np.float32(0.3189), np.float32(0.4573), np.float32(0.3238), np.float32(0.4159), np.float32(0.3416), np.float32(0.2338), np.float32(0.4042), np.float32(0.2848), np.float32(0.2001), np.float32(0.0), np.float32(0.0), np.float32(0.4033), np.float32(0.3966), np.float32(0.0), np.float32(0.0), np.float32(0.4366), np.float32(0.4465), np.float32(0.4462), np.float32(0.4335)] 
+2026-10-17 14:52:21.268819: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 14:52:21.584723: lr: 0.000751 
+2026-10-17 14:52:21.585047: This epoch took 263.093087 s
+ 
+2026-10-17 14:52:21.585142: 
+epoch:  273 
+2026-10-17 14:56:13.731735: train loss : -0.7985 
+2026-10-17 14:56:19.087853: validation loss: -0.4673 
+2026-10-17 14:56:19.088793: Average global foreground Dice: [np.float32(0.8566), np.float32(0.4683), np.float32(0.4865), np.float32(0.6193), np.float32(0.4796), np.float32(0.5714), np.float32(0.5255), np.float32(0.0), np.float32(0.0), np.float32(0.5781), np.float32(0.5099), np.float32(0.4818), np.float32(0.4881), np.float32(0.4863), np.float32(0.2833), np.float32(0.0), np.float32(0.524), np.float32(0.3746), np.float32(0.5499), np.float32(0.404), np.float32(0.4424), np.float32(0.418), np.float32(0.3655), np.float32(0.3657), np.float32(0.3172), np.float32(0.2187), np.float32(0.0), np.float32(0.0), np.float32(0.4606), np.float32(0.402), np.float32(0.0), np.float32(0.0), np.float32(0.5843), np.float32(0.5322), np.float32(0.5391), np.float32(0.5435)] 
+2026-10-17 14:56:19.089246: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 14:56:19.425242: lr: 0.00075 
+2026-10-17 14:56:19.425488: This epoch took 237.840275 s
+ 
+2026-10-17 14:56:19.425581: 
+epoch:  274 
+2026-10-17 15:00:22.649830: train loss : -0.7909 
+2026-10-17 15:00:27.997798: validation loss: -0.3252 
+2026-10-17 15:00:27.998781: Average global foreground Dice: [np.float32(0.8614), np.float32(0.2886), np.float32(0.3361), np.float32(0.381), np.float32(0.2749), np.float32(0.3499), np.float32(0.2965), np.float32(0.0), np.float32(0.0), np.float32(0.6028), np.float32(0.291), np.float32(0.2769), np.float32(0.281), np.float32(0.2845), np.float32(0.1544), np.float32(0.0), np.float32(0.2685), np.float32(0.1633), np.float32(0.301), np.float32(0.1804), np.float32(0.2795), np.float32(0.2831), np.float32(0.2363), np.float32(0.2899), np.float32(0.1236), np.float32(0.1142), np.float32(0.0), np.float32(0.0), np.float32(0.3766), np.float32(0.274), np.float32(0.0), np.float32(0.0), np.float32(0.1778), np.float32(0.2431), np.float32(0.3387), np.float32(0.318)] 
+2026-10-17 15:00:27.999114: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 15:00:28.407027: lr: 0.000749 
+2026-10-17 15:00:28.407305: This epoch took 248.981668 s
+ 
+2026-10-17 15:00:28.407397: 
+epoch:  275 
+2026-10-17 15:04:26.763058: train loss : -0.7945 
+2026-10-17 15:04:32.134960: validation loss: -0.4336 
+2026-10-17 15:04:32.136038: Average global foreground Dice: [np.float32(0.8498), np.float32(0.4389), np.float32(0.4796), np.float32(0.5852), np.float32(0.416), np.float32(0.5599), np.float32(0.4671), np.float32(0.0), np.float32(0.0), np.float32(0.5776), np.float32(0.479), np.float32(0.4614), np.float32(0.386), np.float32(0.4342), np.float32(0.2017), np.float32(0.0), np.float32(0.4578), np.float32(0.3258), np.float32(0.4753), np.float32(0.3526), np.float32(0.4383), np.float32(0.3999), np.float32(0.281), np.float32(0.3345), np.float32(0.2856), np.float32(0.301), np.float32(0.0), np.float32(0.0), np.float32(0.5037), np.float32(0.3201), np.float32(0.0), np.float32(0.0), np.float32(0.4556), np.float32(0.4526), np.float32(0.5063), np.float32(0.5062)] 
+2026-10-17 15:04:32.136412: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 15:04:32.461863: lr: 0.000748 
+2026-10-17 15:04:32.462365: This epoch took 244.054895 s
+ 
+2026-10-17 15:04:32.462766: 
+epoch:  276 
+2026-10-17 15:08:36.750576: train loss : -0.7985 
+2026-10-17 15:08:42.101082: validation loss: -0.4514 
+2026-10-17 15:08:42.102352: Average global foreground Dice: [np.float32(0.8685), np.float32(0.4741), np.float32(0.5101), np.float32(0.6164), np.float32(0.4721), np.float32(0.5789), np.float32(0.5378), np.float32(0.0), np.float32(0.0), np.float32(0.6689), np.float32(0.4843), np.float32(0.4315), np.float32(0.3276), np.float32(0.3716), np.float32(0.0545), np.float32(0.0), np.float32(0.4776), np.float32(0.3426), np.float32(0.5009), np.float32(0.3914), np.float32(0.4443), np.float32(0.4224), np.float32(0.2455), np.float32(0.3661), np.float32(0.3236), np.float32(0.201), np.float32(0.0), np.float32(0.0), np.float32(0.3359), np.float32(0.302), np.float32(0.0), np.float32(0.0), np.float32(0.4792), np.float32(0.4804), np.float32(0.4987), np.float32(0.503)] 
+2026-10-17 15:08:42.102603: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 15:08:42.445882: lr: 0.000747 
+2026-10-17 15:08:42.446281: This epoch took 249.983175 s
+ 
+2026-10-17 15:08:42.446445: 
+epoch:  277 
+2026-10-17 15:12:54.659257: train loss : -0.7940 
+2026-10-17 15:13:00.014928: validation loss: -0.4111 
+2026-10-17 15:13:00.016220: Average global foreground Dice: [np.float32(0.833), np.float32(0.4102), np.float32(0.4459), np.float32(0.5362), np.float32(0.3824), np.float32(0.4997), np.float32(0.4169), np.float32(0.0), np.float32(0.0), np.float32(0.6477), np.float32(0.413), np.float32(0.358), np.float32(0.455), np.float32(0.4308), np.float32(0.0985), np.float32(0.0), np.float32(0.3854), np.float32(0.2571), np.float32(0.4199), np.float32(0.287), np.float32(0.3005), np.float32(0.3152), np.float32(0.3084), np.float32(0.3454), np.float32(0.2848), np.float32(0.1716), np.float32(0.0), np.float32(0.0), np.float32(0.379), np.float32(0.3096), np.float32(0.0), np.float32(0.0), np.float32(0.4058), np.float32(0.4131), np.float32(0.4478), np.float32(0.4452)] 
+2026-10-17 15:13:00.016725: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 15:13:00.353887: lr: 0.000746 
+2026-10-17 15:13:00.354480: This epoch took 257.907865 s
+ 
+2026-10-17 15:13:00.354644: 
+epoch:  278 
+2026-10-17 15:17:26.890486: train loss : -0.7892 
+2026-10-17 15:17:32.236644: validation loss: -0.3865 
+2026-10-17 15:17:32.237823: Average global foreground Dice: [np.float32(0.85), np.float32(0.3786), np.float32(0.4283), np.float32(0.5393), np.float32(0.3949), np.float32(0.4963), np.float32(0.4309), np.float32(0.0), np.float32(0.0), np.float32(0.3367), np.float32(0.4224), np.float32(0.3908), np.float32(0.354), np.float32(0.3391), np.float32(0.0183), np.float32(0.0), np.float32(0.4321), np.float32(0.3144), np.float32(0.4453), np.float32(0.3047), np.float32(0.4261), np.float32(0.3478), np.float32(0.2509), np.float32(0.3187), np.float32(0.231), np.float32(0.0914), np.float32(0.0), np.float32(0.0), np.float32(0.4358), np.float32(0.3313), np.float32(0.0), np.float32(0.0), np.float32(0.4967), np.float32(0.4886), np.float32(0.4666), np.float32(0.4555)] 
+2026-10-17 15:17:32.238172: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 15:17:32.617415: lr: 0.000745 
+2026-10-17 15:17:32.617756: This epoch took 272.263034 s
+ 
+2026-10-17 15:17:32.617854: 
+epoch:  279 
+2026-10-17 15:21:26.570564: train loss : -0.7971 
+2026-10-17 15:21:31.929607: validation loss: -0.3249 
+2026-10-17 15:21:31.930752: Average global foreground Dice: [np.float32(0.8489), np.float32(0.2623), np.float32(0.3726), np.float32(0.3936), np.float32(0.2574), np.float32(0.3513), np.float32(0.3317), np.float32(0.0), np.float32(0.0), np.float32(0.5266), np.float32(0.2476), np.float32(0.223), np.float32(0.2865), np.float32(0.3018), np.float32(0.0247), np.float32(0.0), np.float32(0.2856), np.float32(0.1802), np.float32(0.3208), np.float32(0.2192), np.float32(0.24), np.float32(0.2841), np.float32(0.1268), np.float32(0.315), np.float32(0.0976), np.float32(0.1289), np.float32(0.0), np.float32(0.0), np.float32(0.2388), np.float32(0.2943), np.float32(0.0), np.float32(0.0), np.float32(0.2866), np.float32(0.3336), np.float32(0.2998), np.float32(0.3341)] 
+2026-10-17 15:21:31.931318: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 15:21:32.260982: lr: 0.000744 
+2026-10-17 15:21:32.261540: This epoch took 239.643612 s
+ 
+2026-10-17 15:21:32.261867: 
+epoch:  280 
+2026-10-17 15:25:47.461556: train loss : -0.7875 
+2026-10-17 15:25:52.816441: validation loss: -0.3626 
+2026-10-17 15:25:52.817416: Average global foreground Dice: [np.float32(0.8585), np.float32(0.3418), np.float32(0.3909), np.float32(0.4661), np.float32(0.3531), np.float32(0.4036), np.float32(0.3895), np.float32(0.0), np.float32(0.0), np.float32(0.4994), np.float32(0.3266), np.float32(0.2656), np.float32(0.274), np.float32(0.2522), np.float32(0.0152), np.float32(0.0), np.float32(0.393), np.float32(0.2845), np.float32(0.4023), np.float32(0.2902), np.float32(0.3363), np.float32(0.3419), np.float32(0.2007), np.float32(0.2325), np.float32(0.2053), np.float32(0.1073), np.float32(0.0), np.float32(0.0), np.float32(0.46), np.float32(0.3009), np.float32(0.0), np.float32(0.0), np.float32(0.2831), np.float32(0.3695), np.float32(0.3982), np.float32(0.3884)] 
+2026-10-17 15:25:52.817654: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 15:25:53.145742: lr: 0.000743 
+2026-10-17 15:25:53.146256: This epoch took 260.884139 s
+ 
+2026-10-17 15:25:53.146559: 
+epoch:  281 
+2026-10-17 15:30:18.253684: train loss : -0.7889 
+2026-10-17 15:30:23.637352: validation loss: -0.2909 
+2026-10-17 15:30:23.638759: Average global foreground Dice: [np.float32(0.8977), np.float32(0.2358), np.float32(0.3091), np.float32(0.2495), np.float32(0.2324), np.float32(0.2392), np.float32(0.2249), np.float32(0.0), np.float32(0.0012), np.float32(0.5771), np.float32(0.2325), np.float32(0.1569), np.float32(0.2813), np.float32(0.208), np.float32(0.0456), np.float32(0.0), np.float32(0.3363), np.float32(0.2481), np.float32(0.2645), np.float32(0.2162), np.float32(0.1777), np.float32(0.3597), np.float32(0.1399), np.float32(0.2846), np.float32(0.0897), np.float32(0.094), np.float32(0.0), np.float32(0.0), np.float32(0.0291), np.float32(0.1681), np.float32(0.0), np.float32(0.0), np.float32(0.1624), np.float32(0.1194), np.float32(0.2286), np.float32(0.2358)] 
+2026-10-17 15:30:23.638975: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 15:30:23.982999: lr: 0.000742 
+2026-10-17 15:30:23.983233: This epoch took 270.836429 s
+ 
+2026-10-17 15:30:23.983316: 
+epoch:  282 
+2026-10-17 15:34:48.406004: train loss : -0.7881 
+2026-10-17 15:34:53.755682: validation loss: -0.2761 
+2026-10-17 15:34:53.758127: Average global foreground Dice: [np.float32(0.8645), np.float32(0.1798), np.float32(0.2739), np.float32(0.2848), np.float32(0.2031), np.float32(0.2545), np.float32(0.2184), np.float32(0.0), np.float32(0.0), np.float32(0.6126), np.float32(0.2269), np.float32(0.1606), np.float32(0.2664), np.float32(0.2407), np.float32(0.0149), np.float32(0.0), np.float32(0.2454), np.float32(0.1929), np.float32(0.2405), np.float32(0.1896), np.float32(0.3104), np.float32(0.3119), np.float32(0.0887), np.float32(0.2457), np.float32(0.119), np.float32(0.0572), np.float32(0.0), np.float32(0.0), np.float32(0.1272), np.float32(0.1454), np.float32(0.0), np.float32(0.0), np.float32(0.1998), np.float32(0.2858), np.float32(0.2146), np.float32(0.2038)] 
+2026-10-17 15:34:53.758334: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 15:34:54.153762: lr: 0.000741 
+2026-10-17 15:34:54.154024: This epoch took 270.170655 s
+ 
+2026-10-17 15:34:54.154111: 
+epoch:  283 
+2026-10-17 15:39:19.366744: train loss : -0.7957 
+2026-10-17 15:39:24.720773: validation loss: -0.3373 
+2026-10-17 15:39:24.721663: Average global foreground Dice: [np.float32(0.8709), np.float32(0.3303), np.float32(0.3969), np.float32(0.4461), np.float32(0.337), np.float32(0.4152), np.float32(0.3511), np.float32(0.0), np.float32(0.0), np.float32(0.5199), np.float32(0.3112), np.float32(0.2768), np.float32(0.2709), np.float32(0.2713), np.float32(0.0379), np.float32(0.0), np.float32(0.3785), np.float32(0.2824), np.float32(0.3897), np.float32(0.261), np.float32(0.2993), np.float32(0.2854), np.float32(0.1442), np.float32(0.2351), np.float32(0.2451), np.float32(0.1897), np.float32(0.0), np.float32(0.0), np.float32(0.4452), np.float32(0.2649), np.float32(0.0), np.float32(0.0), np.float32(0.3081), np.float32(0.3768), np.float32(0.3743), np.float32(0.3468)] 
+2026-10-17 15:39:24.721829: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 15:39:25.100337: lr: 0.00074 
+2026-10-17 15:39:25.100588: This epoch took 270.946415 s
+ 
+2026-10-17 15:39:25.100667: 
+epoch:  284 
+2026-10-17 15:43:49.728390: train loss : -0.7937 
+2026-10-17 15:43:55.080662: validation loss: -0.3385 
+2026-10-17 15:43:55.081656: Average global foreground Dice: [np.float32(0.8457), np.float32(0.3043), np.float32(0.3534), np.float32(0.4419), np.float32(0.3144), np.float32(0.4015), np.float32(0.3321), np.float32(0.0), np.float32(0.0), np.float32(0.5117), np.float32(0.3485), np.float32(0.3047), np.float32(0.2837), np.float32(0.2738), np.float32(0.0), np.float32(0.0), np.float32(0.3528), np.float32(0.2454), np.float32(0.3556), np.float32(0.2131), np.float32(0.4036), np.float32(0.3202), np.float32(0.1342), np.float32(0.2982), np.float32(0.1687), np.float32(0.136), np.float32(0.0), np.float32(0.0), np.float32(0.2789), np.float32(0.2556), np.float32(0.0), np.float32(0.0), np.float32(0.2843), np.float32(0.3322), np.float32(0.3241), np.float32(0.3181)] 
+2026-10-17 15:43:55.081950: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 15:43:55.409222: lr: 0.000739 
+2026-10-17 15:43:55.409667: This epoch took 270.308943 s
+ 
+2026-10-17 15:43:55.409942: 
+epoch:  285 
+2026-10-17 15:47:56.860422: train loss : -0.7982 
+2026-10-17 15:48:02.217934: validation loss: -0.4974 
+2026-10-17 15:48:02.218947: Average global foreground Dice: [np.float32(0.8318), np.float32(0.5517), np.float32(0.5787), np.float32(0.6329), np.float32(0.577), np.float32(0.6313), np.float32(0.6499), np.float32(0.0), np.float32(0.0), np.float32(0.618), np.float32(0.4555), np.float32(0.4404), np.float32(0.3934), np.float32(0.4754), np.float32(0.0198), np.float32(0.0), np.float32(0.6027), np.float32(0.4386), np.float32(0.6008), np.float32(0.4283), np.float32(0.5559), np.float32(0.3383), np.float32(0.3717), np.float32(0.4416), np.float32(0.2108), np.float32(0.2572), np.float32(0.0), np.float32(0.0), np.float32(0.4469), np.float32(0.252), np.float32(0.0), np.float32(0.0), np.float32(0.4901), np.float32(0.5192), np.float32(0.5527), np.float32(0.6075)] 
+2026-10-17 15:48:02.219339: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 15:48:02.558052: lr: 0.000738 
+2026-10-17 15:48:02.558290: This epoch took 247.148098 s
+ 
+2026-10-17 15:48:02.558373: 
+epoch:  286 
+2026-10-17 15:52:05.783615: train loss : -0.7766 
+2026-10-17 15:52:11.130623: validation loss: -0.2344 
+2026-10-17 15:52:11.131587: Average global foreground Dice: [np.float32(0.8413), np.float32(0.1243), np.float32(0.1905), np.float32(0.1956), np.float32(0.1593), np.float32(0.0564), np.float32(0.0371), np.float32(0.0), np.float32(0.0), np.float32(0.5693), np.float32(0.1352), np.float32(0.0167), np.float32(0.1), np.float32(0.0323), np.float32(0.0), np.float32(0.0), np.float32(0.1314), np.float32(0.073), np.float32(0.0802), np.float32(0.1392), np.float32(0.035), np.float32(0.1479), np.float32(0.4321), np.float32(0.3256), np.float32(0.0873), np.float32(0.1071), np.float32(0.0), np.float32(0.0), np.float32(0.1165), np.float32(0.2312), np.float32(0.0), np.float32(0.0), np.float32(0.2445), np.float32(0.0), np.float32(0.2935), np.float32(0.0831)] 
+2026-10-17 15:52:11.131771: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 15:52:11.480292: lr: 0.000738 
+2026-10-17 15:52:11.480713: This epoch took 248.922284 s
+ 
+2026-10-17 15:52:11.480922: 
+epoch:  287 
+2026-10-17 15:56:19.699027: train loss : -0.7825 
+2026-10-17 15:56:25.044030: validation loss: -0.3989 
+2026-10-17 15:56:25.044964: Average global foreground Dice: [np.float32(0.8496), np.float32(0.4159), np.float32(0.4459), np.float32(0.4953), np.float32(0.3485), np.float32(0.4545), np.float32(0.4131), np.float32(0.0), np.float32(0.0), np.float32(0.5224), np.float32(0.3394), np.float32(0.3438), np.float32(0.291), np.float32(0.2974), np.float32(0.1304), np.float32(0.0), np.float32(0.402), np.float32(0.2983), np.float32(0.4099), np.float32(0.3125), np.float32(0.4325), np.float32(0.3963), np.float32(0.3541), np.float32(0.448), np.float32(0.308), np.float32(0.1747), np.float32(0.0), np.float32(0.0), np.float32(0.2798), np.float32(0.4078), np.float32(0.0), np.float32(0.0), np.float32(0.2681), np.float32(0.3126), np.float32(0.4486), np.float32(0.4502)] 
+2026-10-17 15:56:25.045126: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 15:56:25.374753: lr: 0.000737 
+2026-10-17 15:56:25.375007: This epoch took 253.893972 s
+ 
+2026-10-17 15:56:25.375093: 
+epoch:  288 
+2026-10-17 16:00:10.896546: train loss : -0.7902 
+2026-10-17 16:00:16.273703: validation loss: -0.3590 
+2026-10-17 16:00:16.274719: Average global foreground Dice: [np.float32(0.8612), np.float32(0.4387), np.float32(0.4434), np.float32(0.345), np.float32(0.3503), np.float32(0.2738), np.float32(0.2783), np.float32(0.0), np.float32(0.0), np.float32(0.612), np.float32(0.2095), np.float32(0.1489), np.float32(0.2301), np.float32(0.131), np.float32(0.1347), np.float32(0.0), np.float32(0.3999), np.float32(0.3504), np.float32(0.2856), np.float32(0.3043), np.float32(0.4394), np.float32(0.3992), np.float32(0.6246), np.float32(0.452), np.float32(0.1279), np.float32(0.1551), np.float32(0.0), np.float32(0.0), np.float32(0.2234), np.float32(0.1859), np.float32(0.0), np.float32(0.0), np.float32(0.2126), np.float32(0.1297), np.float32(0.4638), np.float32(0.329)] 
+2026-10-17 16:00:16.274904: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 16:00:16.603597: lr: 0.000736 
+2026-10-17 16:00:16.603865: This epoch took 231.228715 s
+ 
+2026-10-17 16:00:16.603959: 
+epoch:  289 
+2026-10-17 16:04:40.259503: train loss : -0.7920 
+2026-10-17 16:04:45.621255: validation loss: -0.3672 
+2026-10-17 16:04:45.622736: Average global foreground Dice: [np.float32(0.8373), np.float32(0.3457), np.float32(0.4036), np.float32(0.3725), np.float32(0.2638), np.float32(0.3633), np.float32(0.3543), np.float32(0.0), np.float32(0.0), np.float32(0.6301), np.float32(0.2468), np.float32(0.2109), np.float32(0.3148), np.float32(0.2273), np.float32(0.0173), np.float32(0.0), np.float32(0.3201), np.float32(0.2412), np.float32(0.3716), np.float32(0.29), np.float32(0.2683), np.float32(0.3518), np.float32(0.181), np.float32(0.444), np.float32(0.1984), np.float32(0.2414), np.float32(0.0), np.float32(0.0), np.float32(0.1778), np.float32(0.385), np.float32(0.0), np.float32(0.0), np.float32(0.2221), np.float32(0.3261), np.float32(0.2883), np.float32(0.412)] 
+2026-10-17 16:04:45.623241: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 16:04:46.019184: lr: 0.000735 
+2026-10-17 16:04:46.019524: This epoch took 269.415496 s
+ 
+2026-10-17 16:04:46.019666: 
+epoch:  290 
+2026-10-17 16:09:02.496967: train loss : -0.7949 
+2026-10-17 16:09:07.862409: validation loss: -0.4116 
+2026-10-17 16:09:07.863547: Average global foreground Dice: [np.float32(0.8439), np.float32(0.438), np.float32(0.4975), np.float32(0.5028), np.float32(0.4229), np.float32(0.4751), np.float32(0.4406), np.float32(0.0), np.float32(0.0), np.float32(0.5372), np.float32(0.3718), np.float32(0.2968), np.float32(0.3197), np.float32(0.2875), np.float32(0.0), np.float32(0.0), np.float32(0.513), np.float32(0.4022), np.float32(0.5103), np.float32(0.3945), np.float32(0.3432), np.float32(0.4327), np.float32(0.2315), np.float32(0.4095), np.float32(0.2208), np.float32(0.2611), np.float32(0.0), np.float32(0.0), np.float32(0.3016), np.float32(0.2709), np.float32(0.0), np.float32(0.0), np.float32(0.3085), np.float32(0.336), np.float32(0.3815), np.float32(0.4495)] 
+2026-10-17 16:09:07.863803: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 16:09:08.194721: lr: 0.000734 
+2026-10-17 16:09:08.195195: This epoch took 262.175458 s
+ 
+2026-10-17 16:09:08.195503: 
+epoch:  291 
+2026-10-17 16:13:23.319598: train loss : -0.8000 
+2026-10-17 16:13:28.675880: validation loss: -0.3916 
+2026-10-17 16:13:28.677554: Average global foreground Dice: [np.float32(0.8326), np.float32(0.3853), np.float32(0.4215), np.float32(0.4656), np.float32(0.3723), np.float32(0.4232), np.float32(0.4058), np.float32(0.0), np.float32(0.0), np.float32(0.5448), np.float32(0.3438), np.float32(0.2955), np.float32(0.3135), np.float32(0.3093), np.float32(0.0), np.float32(0.0), np.float32(0.4695), np.float32(0.3595), np.float32(0.4558), np.float32(0.3778), np.float32(0.41), np.float32(0.4042), np.float32(0.253), np.float32(0.3645), np.float32(0.1717), np.float32(0.1886), np.float32(0.0), np.float32(0.0), np.float32(0.2622), np.float32(0.2579), np.float32(0.0), np.float32(0.0), np.float32(0.3289), np.float32(0.3262), np.float32(0.3484), np.float32(0.3862)] 
+2026-10-17 16:13:28.677733: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 16:13:29.020587: lr: 0.000733 
+2026-10-17 16:13:29.020860: This epoch took 260.825078 s
+ 
+2026-10-17 16:13:29.021080: 
+epoch:  292 
+2026-10-17 16:17:38.478028: train loss : -0.8054 
+2026-10-17 16:17:43.837625: validation loss: -0.3868 
+2026-10-17 16:17:43.838772: Average global foreground Dice: [np.float32(0.8211), np.float32(0.3669), np.float32(0.4437), np.float32(0.4705), np.float32(0.3475), np.float32(0.4276), np.float32(0.3908), np.float32(0.0), np.float32(0.0), np.float32(0.5695), np.float32(0.3322), np.float32(0.3095), np.float32(0.3832), np.float32(0.3573), np.float32(0.0257), np.float32(0.0), np.float32(0.4334), np.float32(0.3202), np.float32(0.4392), np.float32(0.3384), np.float32(0.4078), np.float32(0.3976), np.float32(0.2423), np.float32(0.4048), np.float32(0.1838), np.float32(0.1462), np.float32(0.0), np.float32(0.0), np.float32(0.3021), np.float32(0.2495), np.float32(0.0), np.float32(0.0), np.float32(0.3866), np.float32(0.3938), np.float32(0.3757), np.float32(0.3994)] 
+2026-10-17 16:17:43.839016: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 16:17:44.190007: lr: 0.000732 
+2026-10-17 16:17:44.190382: This epoch took 255.169076 s
+ 
+2026-10-17 16:17:44.190613: 
+epoch:  293 
+2026-10-17 16:21:47.943055: train loss : -0.8007 
+2026-10-17 16:21:53.283043: validation loss: -0.2960 
+2026-10-17 16:21:53.284343: Average global foreground Dice: [np.float32(0.8715), np.float32(0.2082), np.float32(0.2823), np.float32(0.2925), np.float32(0.2123), np.float32(0.2449), np.float32(0.2318), np.float32(0.0), np.float32(0.0), np.float32(0.6135), np.float32(0.1767), np.float32(0.1968), np.float32(0.2268), np.float32(0.1888), np.float32(0.1857), np.float32(0.0), np.float32(0.2331), np.float32(0.1659), np.float32(0.2353), np.float32(0.1995), np.float32(0.0869), np.float32(0.2092), np.float32(0.1828), np.float32(0.4098), np.float32(0.1228), np.float32(0.1238), np.float32(0.0), np.float32(0.0), np.float32(0.1205), np.float32(0.2872), np.float32(0.0), np.float32(0.0), np.float32(0.1621), np.float32(0.1768), np.float32(0.2376), np.float32(0.2544)] 
+2026-10-17 16:21:53.284720: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 16:21:53.811407: lr: 0.000731 
+2026-10-17 16:21:53.811648: This epoch took 249.620954 s
+ 
+2026-10-17 16:21:53.811732: 
+epoch:  294 
+2026-10-17 16:26:16.272982: train loss : -0.7984 
+2026-10-17 16:26:21.631894: validation loss: -0.3428 
+2026-10-17 16:26:21.632956: Average global foreground Dice: [np.float32(0.8536), np.float32(0.3055), np.float32(0.3717), np.float32(0.4313), np.float32(0.317), np.float32(0.3525), np.float32(0.3251), np.float32(0.0), np.float32(0.0345), np.float32(0.5832), np.float32(0.2708), np.float32(0.2441), np.float32(0.3182), np.float32(0.2758), np.float32(0.1205), np.float32(0.0), np.float32(0.3814), np.float32(0.2713), np.float32(0.3552), np.float32(0.2995), np.float32(0.3049), np.float32(0.3372), np.float32(0.2025), np.float32(0.3637), np.float32(0.1426), np.float32(0.1687), np.float32(0.0), np.float32(0.0), np.float32(0.3281), np.float32(0.3638), np.float32(0.0), np.float32(0.0), np.float32(0.2153), np.float32(0.2436), np.float32(0.3479), np.float32(0.362)] 
+2026-10-17 16:26:21.633141: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 16:26:21.963803: lr: 0.00073 
+2026-10-17 16:26:21.964147: This epoch took 268.152363 s
+ 
+2026-10-17 16:26:21.964368: 
+epoch:  295 
+2026-10-17 16:30:46.756046: train loss : -0.7966 
+2026-10-17 16:30:52.134622: validation loss: -0.3895 
+2026-10-17 16:30:52.135577: Average global foreground Dice: [np.float32(0.8634), np.float32(0.3203), np.float32(0.4051), np.float32(0.4372), np.float32(0.3442), np.float32(0.3544), np.float32(0.3872), np.float32(0.0), np.float32(0.0), np.float32(0.5222), np.float32(0.3239), np.float32(0.3662), np.float32(0.338), np.float32(0.3962), np.float32(0.0334), np.float32(0.0), np.float32(0.3906), np.float32(0.2859), np.float32(0.3464), np.float32(0.3209), np.float32(0.1748), np.float32(0.2731), np.float32(0.5132), np.float32(0.4663), np.float32(0.2664), np.float32(0.3141), np.float32(0.0), np.float32(0.0), np.float32(0.2242), np.float32(0.2097), np.float32(0.0), np.float32(0.0), np.float32(0.2756), np.float32(0.3203), np.float32(0.4103), np.float32(0.4174)] 
+2026-10-17 16:30:52.135763: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 16:30:52.481456: lr: 0.000729 
+2026-10-17 16:30:52.481743: This epoch took 270.517297 s
+ 
+2026-10-17 16:30:52.481833: 
+epoch:  296 
+2026-10-17 16:34:58.940445: train loss : -0.7577 
+2026-10-17 16:35:04.303693: validation loss: -0.2691 
+2026-10-17 16:35:04.304856: Average global foreground Dice: [np.float32(0.7826), np.float32(0.2236), np.float32(0.2822), np.float32(0.3272), np.float32(0.2343), np.float32(0.2727), np.float32(0.2392), np.float32(0.0), np.float32(0.0), np.float32(0.6034), np.float32(0.2326), np.float32(0.1918), np.float32(0.2932), np.float32(0.1917), np.float32(0.0807), np.float32(0.0), np.float32(0.2616), np.float32(0.1964), np.float32(0.2297), np.float32(0.1536), np.float32(0.1469), np.float32(0.2936), np.float32(0.1179), np.float32(0.2554), np.float32(0.058), np.float32(0.0685), np.float32(0.0), np.float32(0.0), np.float32(0.2497), np.float32(0.034), np.float32(0.0), np.float32(0.0), np.float32(0.2089), np.float32(0.2186), np.float32(0.194), np.float32(0.238)] 
+2026-10-17 16:35:04.305076: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 16:35:04.644801: lr: 0.000728 
+2026-10-17 16:35:04.645133: This epoch took 252.163233 s
+ 
+2026-10-17 16:35:04.645228: 
+epoch:  297 
+2026-10-17 16:39:45.034849: train loss : -0.7785 
+2026-10-17 16:39:50.389602: validation loss: -0.4085 
+2026-10-17 16:39:50.390796: Average global foreground Dice: [np.float32(0.8751), np.float32(0.4944), np.float32(0.4943), np.float32(0.5916), np.float32(0.5161), np.float32(0.5179), np.float32(0.4692), np.float32(0.0), np.float32(0.0), np.float32(0.5537), np.float32(0.4201), np.float32(0.3758), np.float32(0.3317), np.float32(0.3982), np.float32(0.0), np.float32(0.0), np.float32(0.4266), np.float32(0.274), np.float32(0.4448), np.float32(0.3432), np.float32(0.2596), np.float32(0.3612), np.float32(0.239), np.float32(0.3576), np.float32(0.2076), np.float32(0.1659), np.float32(0.0), np.float32(0.0), np.float32(0.3346), np.float32(0.2249), np.float32(0.0), np.float32(0.0), np.float32(0.4392), np.float32(0.3037), np.float32(0.4831), np.float32(0.4601)] 
+2026-10-17 16:39:50.391252: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 16:39:50.766573: lr: 0.000727 
+2026-10-17 16:39:50.766927: This epoch took 286.121635 s
+ 
+2026-10-17 16:39:50.767180: 
+epoch:  298 
+2026-10-17 16:44:03.601560: train loss : -0.7861 
+2026-10-17 16:44:09.004823: validation loss: -0.4070 
+2026-10-17 16:44:09.005751: Average global foreground Dice: [np.float32(0.8374), np.float32(0.5278), np.float32(0.4679), np.float32(0.6133), np.float32(0.5127), np.float32(0.4511), np.float32(0.3851), np.float32(0.0), np.float32(0.0), np.float32(0.3774), np.float32(0.5), np.float32(0.3952), np.float32(0.3802), np.float32(0.3622), np.float32(0.0), np.float32(0.0), np.float32(0.5107), np.float32(0.3818), np.float32(0.4741), np.float32(0.3422), np.float32(0.4109), np.float32(0.3991), np.float32(0.3358), np.float32(0.2928), np.float32(0.3609), np.float32(0.133), np.float32(0.0), np.float32(0.0), np.float32(0.4393), np.float32(0.0507), np.float32(0.0), np.float32(0.0), np.float32(0.4783), np.float32(0.402), np.float32(0.5033), np.float32(0.4389)] 
+2026-10-17 16:44:09.006013: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 16:44:09.359117: lr: 0.000726 
+2026-10-17 16:44:09.359630: This epoch took 258.592259 s
+ 
+2026-10-17 16:44:09.359888: 
+epoch:  299 
+2026-10-17 16:48:02.892851: train loss : -0.7964 
+2026-10-17 16:48:08.244909: validation loss: -0.3995 
+2026-10-17 16:48:08.246054: Average global foreground Dice: [np.float32(0.837), np.float32(0.4638), np.float32(0.4205), np.float32(0.5449), np.float32(0.4647), np.float32(0.3758), np.float32(0.3919), np.float32(0.0), np.float32(0.0), np.float32(0.3563), np.float32(0.3932), np.float32(0.3309), np.float32(0.3806), np.float32(0.4058), np.float32(0.0), np.float32(0.0), np.float32(0.4482), np.float32(0.2848), np.float32(0.4086), np.float32(0.2969), np.float32(0.3519), np.float32(0.3491), np.float32(0.2031), np.float32(0.3541), np.float32(0.2403), np.float32(0.1846), np.float32(0.0), np.float32(0.0), np.float32(0.4818), np.float32(0.0824), np.float32(0.0), np.float32(0.0), np.float32(0.4122), np.float32(0.3964), np.float32(0.4707), np.float32(0.3962)] 
+2026-10-17 16:48:08.246294: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 16:48:08.588828: lr: 0.000725 
+2026-10-17 16:48:08.589136: saving scheduled checkpoint file... 
+2026-10-17 16:48:08.598914: saving checkpoint... 
+2026-10-17 16:48:08.729890: done, saving took 0.14 seconds 
+2026-10-17 16:48:08.731807: done 
+2026-10-17 16:48:08.732037: This epoch took 239.372055 s
+ 
+2026-10-17 16:48:08.732292: 
+epoch:  300 
+2026-10-17 16:52:34.934958: train loss : -0.7982 
+2026-10-17 16:52:40.288918: validation loss: -0.4393 
+2026-10-17 16:52:40.289901: Average global foreground Dice: [np.float32(0.8391), np.float32(0.4671), np.float32(0.5073), np.float32(0.6026), np.float32(0.5027), np.float32(0.5554), np.float32(0.5022), np.float32(0.0), np.float32(0.0), np.float32(0.4078), np.float32(0.4464), np.float32(0.413), np.float32(0.3991), np.float32(0.3654), np.float32(0.0923), np.float32(0.0), np.float32(0.4632), np.float32(0.3265), np.float32(0.5108), np.float32(0.3987), np.float32(0.4014), np.float32(0.3943), np.float32(0.4105), np.float32(0.4436), np.float32(0.2784), np.float32(0.1749), np.float32(0.0), np.float32(0.0156), np.float32(0.5558), np.float32(0.2804), np.float32(0.0), np.float32(0.0), np.float32(0.4003), np.float32(0.5052), np.float32(0.5518), np.float32(0.5335)] 
+2026-10-17 16:52:40.290206: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 16:52:40.659366: lr: 0.000724 
+2026-10-17 16:52:40.659914: This epoch took 271.927427 s
+ 
+2026-10-17 16:52:40.660105: 
+epoch:  301 
+2026-10-17 16:57:16.045217: train loss : -0.7903 
+2026-10-17 16:57:21.403599: validation loss: -0.4516 
+2026-10-17 16:57:21.404721: Average global foreground Dice: [np.float32(0.8332), np.float32(0.4951), np.float32(0.5228), np.float32(0.6152), np.float32(0.4953), np.float32(0.6085), np.float32(0.5294), np.float32(0.0), np.float32(0.0), np.float32(0.4666), np.float32(0.458), np.float32(0.4502), np.float32(0.4211), np.float32(0.4596), np.float32(0.0822), np.float32(0.0), np.float32(0.5013), np.float32(0.3535), np.float32(0.5027), np.float32(0.3974), np.float32(0.3935), np.float32(0.3937), np.float32(0.3219), np.float32(0.4537), np.float32(0.2522), np.float32(0.2375), np.float32(0.0), np.float32(0.0175), np.float32(0.5136), np.float32(0.3468), np.float32(0.0), np.float32(0.0), np.float32(0.4749), np.float32(0.5248), np.float32(0.5119), np.float32(0.522)] 
+2026-10-17 16:57:21.405115: (interpret this as an estimate for the Dice of the different classes. This is not exact.) 
+2026-10-17 16:57:21.785625: lr: 0.000724 
+2026-10-17 16:57:21.786087: This epoch took 281.125906 s
+ 
+2026-10-17 16:57:21.786510: 
+epoch:  302