What do the results below actually mean? https://www.kaggle.com/c/tgs-salt-identification-challenge#evaluation

(6400, 101, 101, 1) (800, 101, 101, 1) Stage 1 __________________________________________________________________________________________________ Layer (type) Output Shape Param # Connected to ================================================================================================== input_1 (InputLayer) (None, 101, 101, 1) 0 __________________________________________________________________________________________________ conv2d_1 (Conv2D) (None, 101, 101, 16) 160 input_1[0][0] __________________________________________________________________________________________________ batch_normalization_1 (BatchNor (None, 101, 101, 16) 64 conv2d_1[0][0] __________________________________________________________________________________________________ activation_1 (Activation) (None, 101, 101, 16) 0 batch_normalization_1[0][0] __________________________________________________________________________________________________ conv2d_2 (Conv2D) (None, 101, 101, 16) 2320 activation_1[0][0] __________________________________________________________________________________________________ batch_normalization_2 (BatchNor (None, 101, 101, 16) 64 conv2d_2[0][0] __________________________________________________________________________________________________ activation_2 (Activation) (None, 101, 101, 16) 0 batch_normalization_2[0][0] __________________________________________________________________________________________________ conv2d_3 (Conv2D) (None, 101, 101, 16) 2320 activation_2[0][0] __________________________________________________________________________________________________ add_1 (Add) (None, 101, 101, 16) 0 conv2d_3[0][0] conv2d_1[0][0] __________________________________________________________________________________________________ batch_normalization_3 (BatchNor (None, 101, 101, 16) 64 add_1[0][0] __________________________________________________________________________________________________ activation_3 (Activation) (None, 101, 101, 16) 0 batch_normalization_3[0][0] _______________________________________________________________________________

conv2d_44 (Conv2D) (None, 101, 101, 16) 2320 activation_43[0][0] __________________________________________________________________________________________________ batch_normalization_44 (BatchNo (None, 101, 101, 16) 64 conv2d_44[0][0] __________________________________________________________________________________________________ activation_44 (Activation) (None, 101, 101, 16) 0 batch_normalization_44[0][0] __________________________________________________________________________________________________ conv2d_45 (Conv2D) (None, 101, 101, 16) 2320 activation_44[0][0] __________________________________________________________________________________________________ add_18 (Add) (None, 101, 101, 16) 0 conv2d_45[0][0] add_17[0][0] __________________________________________________________________________________________________ batch_normalization_45 (BatchNo (None, 101, 101, 16) 64 add_18[0][0] __________________________________________________________________________________________________ activation_45 (Activation) (None, 101, 101, 16) 0 batch_normalization_45[0][0] __________________________________________________________________________________________________ conv2d_46 (Conv2D) (None, 101, 101, 1) 17 activation_45[0][0] __________________________________________________________________________________________________ activation_46 (Activation) (None, 101, 101, 1) 0 conv2d_46[0][0] ================================================================================================== Total params: 5,119,857 Trainable params: 5,112,497 Non-trainable params: 7,360 __________________________________________________________________________________________________ Train on 6400 samples, validate on 800 samples Epoch 1/10 - 2307s - loss: 0.3595 - my_iou_metric: 0.3674 - val_loss: 0.6009 - val_my_iou_metric: 0.2154

Epoch 00001: val_my_iou_metric improved from -inf to 0.21537, saving model to Unet_resnet_v1_1.model Epoch 2/10 - 2424s - loss: 0.2766 - my_iou_metric: 0.5058 - val_loss: 0.3745 - val_my_iou_metric: 0.3724

Epoch 00002: val_my_iou_metric improved from 0.21537 to 0.37237, saving model to Unet_resnet_v1_1.model Epoch 3/10 - 2403s - loss: 0.2452 - my_iou_metric: 0.5494 - val_loss: 0.8235 - val_my_iou_metric: 0.3622

Epoch 00003: val_my_iou_metric did not improve from 0.37237 Epoch 4/10 - 2424s - loss: 0.2355 - my_iou_metric: 0.5611 - val_loss: 0.3038 - val_my_iou_metric: 0.4791

Epoch 00004: val_my_iou_metric improved from 0.37237 to 0.47913, saving model to Unet_resnet_v1_1.model Epoch 5/10 - 2436s - loss: 0.2145 - my_iou_metric: 0.5933 - val_loss: 0.2974 - val_my_iou_metric: 0.6087

Epoch 00005: val_my_iou_metric improved from 0.47913 to 0.60875, saving model to Unet_resnet_v1_1.model Epoch 6/10 - 2415s - loss: 0.1941 - my_iou_metric: 0.6188 - val_loss: 0.2150 - val_my_iou_metric: 0.5763

Epoch 00006: val_my_iou_metric did not improve from 0.60875 Epoch 7/10 - 2547s - loss: 0.1929 - my_iou_metric: 0.6248 - val_loss: 0.3102 - val_my_iou_metric: 0.5329

Epoch 00007: val_my_iou_metric did not improve from 0.60875 Epoch 8/10 - 2645s - loss: 0.1891 - my_iou_metric: 0.6360 - val_loss: 0.1830 - val_my_iou_metric: 0.6756

Epoch 00008: val_my_iou_metric improved from 0.60875 to 0.67562, saving model to Unet_resnet_v1_1.model Epoch 9/10 - 2634s - loss: 0.1817 - my_iou_metric: 0.6476 - val_loss: 0.1895 - val_my_iou_metric: 0.6669

Epoch 00009: val_my_iou_metric did not improve from 0.67562 Epoch 10/10 - 2547s - loss: 0.1661 - my_iou_metric: 0.6683 - val_loss: 0.1879 - val_my_iou_metric: 0.6193

Epoch 00010: val_my_iou_metric did not improve from 0.67562 Stage 2 Train on 6400 samples, validate on 800 samples Epoch 1/10 - 2699s - loss: -5.3927e-01 - my_iou_metric: 0.5045 - val_loss: -6.2262e-01 - val_my_iou_metric: 0.6154

Epoch 00001: val_my_iou_metric did not improve from 0.67562 Epoch 2/10 - 2563s - loss: -6.3637e-01 - my_iou_metric: 0.5750 - val_loss: -6.7379e-01 - val_my_iou_metric: 0.6295

Epoch 00002: val_my_iou_metric did not improve from 0.67562 Epoch 3/10 - 2548s - loss: -6.6867e-01 - my_iou_metric: 0.5982 - val_loss: -5.7147e-01 - val_my_iou_metric: 0.5976

Epoch 00003: val_my_iou_metric did not improve from 0.67562 Epoch 4/10 - 2533s - loss: -7.0277e-01 - my_iou_metric: 0.6389 - val_loss: -6.4307e-01 - val_my_iou_metric: 0.6221

Epoch 00004: val_my_iou_metric did not improve from 0.67562 Epoch 5/10 - 2436s - loss: -7.1246e-01 - my_iou_metric: 0.6401 - val_loss: -7.0912e-01 - val_my_iou_metric: 0.5839

Epoch 00005: val_my_iou_metric did not improve from 0.67562 Epoch 6/10 - 2496s - loss: -7.3510e-01 - my_iou_metric: 0.6671 - val_loss: -6.0356e-01 - val_my_iou_metric: 0.6135

Epoch 00006: val_my_iou_metric did not improve from 0.67562 Epoch 7/10 - 2507s - loss: -7.3554e-01 - my_iou_metric: 0.6723 - val_loss: -7.0727e-01 - val_my_iou_metric: 0.6504

Epoch 00007: val_my_iou_metric did not improve from 0.67562 Epoch 8/10 - 2503s - loss: -7.5242e-01 - my_iou_metric: 0.6820 - val_loss: -7.3483e-01 - val_my_iou_metric: 0.7004

Epoch 00008: val_my_iou_metric improved from 0.67562 to 0.70038, saving model to Unet_resnet_v1_1.model Epoch 9/10 - 2483s - loss: -7.5873e-01 - my_iou_metric: 0.6945 - val_loss: -6.6214e-01 - val_my_iou_metric: 0.6593

Epoch 00009: val_my_iou_metric did not improve from 0.70038 Epoch 10/10 - 2492s - loss: -7.6327e-01 - my_iou_metric: 0.7058 - val_loss: -7.2479e-01 - val_my_iou_metric: 0.6996

Epoch 00010: val_my_iou_metric did not improve from 0.70038 Stage 3 Train on 6400 samples, validate on 800 samples Epoch 1/10 - 2527s - loss: 0.1249 - my_iou_metric: 0.5980 - val_loss: 0.3356 - val_my_iou_metric: 0.6039

Epoch 00001: val_my_iou_metric did not improve from 0.70038 Epoch 2/10 - 2593s - loss: -5.3444e-02 - my_iou_metric: 0.6258 - val_loss: -1.0343e-01 - val_my_iou_metric: 0.6231

Epoch 00002: val_my_iou_metric did not improve from 0.70038 Epoch 3/10 - 2534s - loss: -1.2991e-01 - my_iou_metric: 0.6390 - val_loss: -2.4771e-01 - val_my_iou_metric: 0.6951

Epoch 00003: val_my_iou_metric did not improve from 0.70038 Epoch 4/10 - 2141s - loss: -1.7572e-01 - my_iou_metric: 0.6627 - val_loss: -1.1991e-01 - val_my_iou_metric: 0.6806

Epoch 00004: val_my_iou_metric did not improve from 0.70038 Epoch 5/10 - 2137s - loss: -2.3527e-01 - my_iou_metric: 0.6767 - val_loss: -2.2720e-01 - val_my_iou_metric: 0.6901

Epoch 00005: val_my_iou_metric did not improve from 0.70038 Epoch 6/10 - 2130s - loss: -2.4228e-01 - my_iou_metric: 0.6800 - val_loss: -5.1456e-02 - val_my_iou_metric: 0.6451

Epoch 00006: val_my_iou_metric did not improve from 0.70038 Epoch 7/10 - 2137s - loss: -2.6496e-01 - my_iou_metric: 0.6882 - val_loss: -2.1933e-01 - val_my_iou_metric: 0.6926

Epoch 00007: val_my_iou_metric did not improve from 0.70038 Epoch 8/10 - 2139s - loss: -2.8349e-01 - my_iou_metric: 0.6957 - val_loss: -2.6460e-01 - val_my_iou_metric: 0.7121

Epoch 00008: val_my_iou_metric improved from 0.70038 to 0.71212, saving model to Unet_resnet_v1_1.model Epoch 9/10 - 2147s - loss: -3.0860e-01 - my_iou_metric: 0.7040 - val_loss: -3.1754e-01 - val_my_iou_metric: 0.7159

Epoch 00009: val_my_iou_metric improved from 0.71212 to 0.71588, saving model to Unet_resnet_v1_1.model Epoch 10/10 - 2177s - loss: -2.9721e-01 - my_iou_metric: 0.6990 - val_loss: -3.3557e-01 - val_my_iou_metric: 0.7260

Epoch 00010: val_my_iou_metric improved from 0.71588 to 0.72600, saving model to Unet_resnet_v1_1.model Intersection-over-Union values for different thresholds are listed below [0.133125 0.133125 0.133125 0.133125 0.133125 0.133125 0.133125 0.133125 0.133125 0.133125 0.133125 0.133125 0.133125 0.133125 0.133125 0.133125 0.51525 0.606 0.631 0.637125 0.64075 0.648375 0.65275 0.653625 0.65325 0.655375 0.657875 0.653625 0.653875 0.651375 0.645875] Usedtime = 8.400719404220581 s Total run time = 21.081168919139436 hours