diff --git "a/README.md" "b/README.md"
--- "a/README.md"
+++ "b/README.md"
@@ -43,416 +43,116 @@ IoU metric: bbox
 - lr: 0.0001
 - dropout_rate: 0.1
 - weight_decay: 0.0001
-- max_epochs: 400
+- max_epochs: 100
 - train samples: 2
 
 ## Logging
 ### Training process
 ```
-{'validation_loss': tensor(7.5952, device='cuda:0'), 'validation_loss_ce': tensor(2.2881, device='cuda:0'), 'validation_loss_bbox': tensor(0.6367, device='cuda:0'), 'validation_loss_giou': tensor(1.0618, device='cuda:0'), 'validation_cardinality_error': tensor(99., device='cuda:0')}
-{'training_loss': tensor(7.5952, device='cuda:0'), 'train_loss_ce': tensor(2.2881, device='cuda:0'), 'train_loss_bbox': tensor(0.6367, device='cuda:0'), 'train_loss_giou': tensor(1.0618, device='cuda:0'), 'train_cardinality_error': tensor(99., device='cuda:0'), 'validation_loss': tensor(5.2760, device='cuda:0'), 'validation_loss_ce': tensor(2.1397, device='cuda:0'), 'validation_loss_bbox': tensor(0.2398, device='cuda:0'), 'validation_loss_giou': tensor(0.9687, device='cuda:0'), 'validation_cardinality_error': tensor(98.5000, device='cuda:0')}
-{'training_loss': tensor(5.2760, device='cuda:0'), 'train_loss_ce': tensor(2.1397, device='cuda:0'), 'train_loss_bbox': tensor(0.2398, device='cuda:0'), 'train_loss_giou': tensor(0.9687, device='cuda:0'), 'train_cardinality_error': tensor(98.5000, device='cuda:0'), 'validation_loss': tensor(8.4159, device='cuda:0'), 'validation_loss_ce': tensor(2.0334, device='cuda:0'), 'validation_loss_bbox': tensor(0.8392, device='cuda:0'), 'validation_loss_giou': tensor(1.0932, device='cuda:0'), 'validation_cardinality_error': tensor(76.5000, device='cuda:0')}
-{'training_loss': tensor(8.4159, device='cuda:0'), 'train_loss_ce': tensor(2.0334, device='cuda:0'), 'train_loss_bbox': tensor(0.8392, device='cuda:0'), 'train_loss_giou': tensor(1.0932, device='cuda:0'), 'train_cardinality_error': tensor(76.5000, device='cuda:0'), 'validation_loss': tensor(8.2033, device='cuda:0'), 'validation_loss_ce': tensor(2.0125, device='cuda:0'), 'validation_loss_bbox': tensor(0.5671, device='cuda:0'), 'validation_loss_giou': tensor(1.6776, device='cuda:0'), 'validation_cardinality_error': tensor(67.5000, device='cuda:0')}
-{'training_loss': tensor(8.2033, device='cuda:0'), 'train_loss_ce': tensor(2.0125, device='cuda:0'), 'train_loss_bbox': tensor(0.5671, device='cuda:0'), 'train_loss_giou': tensor(1.6776, device='cuda:0'), 'train_cardinality_error': tensor(67.5000, device='cuda:0'), 'validation_loss': tensor(8.1704, device='cuda:0'), 'validation_loss_ce': tensor(1.9750, device='cuda:0'), 'validation_loss_bbox': tensor(0.5571, device='cuda:0'), 'validation_loss_giou': tensor(1.7049, device='cuda:0'), 'validation_cardinality_error': tensor(54., device='cuda:0')}
-{'training_loss': tensor(8.1704, device='cuda:0'), 'train_loss_ce': tensor(1.9750, device='cuda:0'), 'train_loss_bbox': tensor(0.5571, device='cuda:0'), 'train_loss_giou': tensor(1.7049, device='cuda:0'), 'train_cardinality_error': tensor(54., device='cuda:0'), 'validation_loss': tensor(8.0808, device='cuda:0'), 'validation_loss_ce': tensor(1.9083, device='cuda:0'), 'validation_loss_bbox': tensor(0.5557, device='cuda:0'), 'validation_loss_giou': tensor(1.6971, device='cuda:0'), 'validation_cardinality_error': tensor(29.5000, device='cuda:0')}
-{'training_loss': tensor(8.0808, device='cuda:0'), 'train_loss_ce': tensor(1.9083, device='cuda:0'), 'train_loss_bbox': tensor(0.5557, device='cuda:0'), 'train_loss_giou': tensor(1.6971, device='cuda:0'), 'train_cardinality_error': tensor(29.5000, device='cuda:0'), 'validation_loss': tensor(7.8634, device='cuda:0'), 'validation_loss_ce': tensor(1.8297, device='cuda:0'), 'validation_loss_bbox': tensor(0.5691, device='cuda:0'), 'validation_loss_giou': tensor(1.5941, device='cuda:0'), 'validation_cardinality_error': tensor(20.5000, device='cuda:0')}
-{'training_loss': tensor(7.8634, device='cuda:0'), 'train_loss_ce': tensor(1.8297, device='cuda:0'), 'train_loss_bbox': tensor(0.5691, device='cuda:0'), 'train_loss_giou': tensor(1.5941, device='cuda:0'), 'train_cardinality_error': tensor(20.5000, device='cuda:0'), 'validation_loss': tensor(7.6190, device='cuda:0'), 'validation_loss_ce': tensor(1.7556, device='cuda:0'), 'validation_loss_bbox': tensor(0.5273, device='cuda:0'), 'validation_loss_giou': tensor(1.6134, device='cuda:0'), 'validation_cardinality_error': tensor(14.5000, device='cuda:0')}
-{'training_loss': tensor(7.6190, device='cuda:0'), 'train_loss_ce': tensor(1.7556, device='cuda:0'), 'train_loss_bbox': tensor(0.5273, device='cuda:0'), 'train_loss_giou': tensor(1.6134, device='cuda:0'), 'train_cardinality_error': tensor(14.5000, device='cuda:0'), 'validation_loss': tensor(6.8056, device='cuda:0'), 'validation_loss_ce': tensor(1.6719, device='cuda:0'), 'validation_loss_bbox': tensor(0.4739, device='cuda:0'), 'validation_loss_giou': tensor(1.3821, device='cuda:0'), 'validation_cardinality_error': tensor(11., device='cuda:0')}
-{'training_loss': tensor(6.8056, device='cuda:0'), 'train_loss_ce': tensor(1.6719, device='cuda:0'), 'train_loss_bbox': tensor(0.4739, device='cuda:0'), 'train_loss_giou': tensor(1.3821, device='cuda:0'), 'train_cardinality_error': tensor(11., device='cuda:0'), 'validation_loss': tensor(6.5120, device='cuda:0'), 'validation_loss_ce': tensor(1.6247, device='cuda:0'), 'validation_loss_bbox': tensor(0.4546, device='cuda:0'), 'validation_loss_giou': tensor(1.3072, device='cuda:0'), 'validation_cardinality_error': tensor(12., device='cuda:0')}
-{'training_loss': tensor(6.5120, device='cuda:0'), 'train_loss_ce': tensor(1.6247, device='cuda:0'), 'train_loss_bbox': tensor(0.4546, device='cuda:0'), 'train_loss_giou': tensor(1.3072, device='cuda:0'), 'train_cardinality_error': tensor(12., device='cuda:0'), 'validation_loss': tensor(6.4916, device='cuda:0'), 'validation_loss_ce': tensor(1.5904, device='cuda:0'), 'validation_loss_bbox': tensor(0.4385, device='cuda:0'), 'validation_loss_giou': tensor(1.3544, device='cuda:0'), 'validation_cardinality_error': tensor(6.5000, device='cuda:0')}
-{'training_loss': tensor(6.4916, device='cuda:0'), 'train_loss_ce': tensor(1.5904, device='cuda:0'), 'train_loss_bbox': tensor(0.4385, device='cuda:0'), 'train_loss_giou': tensor(1.3544, device='cuda:0'), 'train_cardinality_error': tensor(6.5000, device='cuda:0'), 'validation_loss': tensor(4.7066, device='cuda:0'), 'validation_loss_ce': tensor(1.6010, device='cuda:0'), 'validation_loss_bbox': tensor(0.2114, device='cuda:0'), 'validation_loss_giou': tensor(1.0243, device='cuda:0'), 'validation_cardinality_error': tensor(10.5000, device='cuda:0')}
-{'training_loss': tensor(4.7066, device='cuda:0'), 'train_loss_ce': tensor(1.6010, device='cuda:0'), 'train_loss_bbox': tensor(0.2114, device='cuda:0'), 'train_loss_giou': tensor(1.0243, device='cuda:0'), 'train_cardinality_error': tensor(10.5000, device='cuda:0'), 'validation_loss': tensor(6.0679, device='cuda:0'), 'validation_loss_ce': tensor(1.4556, device='cuda:0'), 'validation_loss_bbox': tensor(0.4071, device='cuda:0'), 'validation_loss_giou': tensor(1.2885, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(6.0679, device='cuda:0'), 'train_loss_ce': tensor(1.4556, device='cuda:0'), 'train_loss_bbox': tensor(0.4071, device='cuda:0'), 'train_loss_giou': tensor(1.2885, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(6.0086, device='cuda:0'), 'validation_loss_ce': tensor(1.4175, device='cuda:0'), 'validation_loss_bbox': tensor(0.4350, device='cuda:0'), 'validation_loss_giou': tensor(1.2081, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(6.0086, device='cuda:0'), 'train_loss_ce': tensor(1.4175, device='cuda:0'), 'train_loss_bbox': tensor(0.4350, device='cuda:0'), 'train_loss_giou': tensor(1.2081, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(5.6130, device='cuda:0'), 'validation_loss_ce': tensor(1.3818, device='cuda:0'), 'validation_loss_bbox': tensor(0.3950, device='cuda:0'), 'validation_loss_giou': tensor(1.1281, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(5.6130, device='cuda:0'), 'train_loss_ce': tensor(1.3818, device='cuda:0'), 'train_loss_bbox': tensor(0.3950, device='cuda:0'), 'train_loss_giou': tensor(1.1281, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(5.2649, device='cuda:0'), 'validation_loss_ce': tensor(1.3391, device='cuda:0'), 'validation_loss_bbox': tensor(0.3671, device='cuda:0'), 'validation_loss_giou': tensor(1.0451, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(5.2649, device='cuda:0'), 'train_loss_ce': tensor(1.3391, device='cuda:0'), 'train_loss_bbox': tensor(0.3671, device='cuda:0'), 'train_loss_giou': tensor(1.0451, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(5.3221, device='cuda:0'), 'validation_loss_ce': tensor(1.2945, device='cuda:0'), 'validation_loss_bbox': tensor(0.3751, device='cuda:0'), 'validation_loss_giou': tensor(1.0761, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(5.3221, device='cuda:0'), 'train_loss_ce': tensor(1.2945, device='cuda:0'), 'train_loss_bbox': tensor(0.3751, device='cuda:0'), 'train_loss_giou': tensor(1.0761, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(5.0296, device='cuda:0'), 'validation_loss_ce': tensor(1.2710, device='cuda:0'), 'validation_loss_bbox': tensor(0.3553, device='cuda:0'), 'validation_loss_giou': tensor(0.9911, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(5.0296, device='cuda:0'), 'train_loss_ce': tensor(1.2710, device='cuda:0'), 'train_loss_bbox': tensor(0.3553, device='cuda:0'), 'train_loss_giou': tensor(0.9911, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.9502, device='cuda:0'), 'validation_loss_ce': tensor(1.2432, device='cuda:0'), 'validation_loss_bbox': tensor(0.3481, device='cuda:0'), 'validation_loss_giou': tensor(0.9832, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(4.9502, device='cuda:0'), 'train_loss_ce': tensor(1.2432, device='cuda:0'), 'train_loss_bbox': tensor(0.3481, device='cuda:0'), 'train_loss_giou': tensor(0.9832, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.7995, device='cuda:0'), 'validation_loss_ce': tensor(1.2057, device='cuda:0'), 'validation_loss_bbox': tensor(0.3536, device='cuda:0'), 'validation_loss_giou': tensor(0.9128, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(4.7995, device='cuda:0'), 'train_loss_ce': tensor(1.2057, device='cuda:0'), 'train_loss_bbox': tensor(0.3536, device='cuda:0'), 'train_loss_giou': tensor(0.9128, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(5.0234, device='cuda:0'), 'validation_loss_ce': tensor(1.1686, device='cuda:0'), 'validation_loss_bbox': tensor(0.3727, device='cuda:0'), 'validation_loss_giou': tensor(0.9955, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(5.0234, device='cuda:0'), 'train_loss_ce': tensor(1.1686, device='cuda:0'), 'train_loss_bbox': tensor(0.3727, device='cuda:0'), 'train_loss_giou': tensor(0.9955, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.7973, device='cuda:0'), 'validation_loss_ce': tensor(1.1335, device='cuda:0'), 'validation_loss_bbox': tensor(0.3595, device='cuda:0'), 'validation_loss_giou': tensor(0.9333, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(4.7973, device='cuda:0'), 'train_loss_ce': tensor(1.1335, device='cuda:0'), 'train_loss_bbox': tensor(0.3595, device='cuda:0'), 'train_loss_giou': tensor(0.9333, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.7624, device='cuda:0'), 'validation_loss_ce': tensor(1.1001, device='cuda:0'), 'validation_loss_bbox': tensor(0.3463, device='cuda:0'), 'validation_loss_giou': tensor(0.9655, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(4.7624, device='cuda:0'), 'train_loss_ce': tensor(1.1001, device='cuda:0'), 'train_loss_bbox': tensor(0.3463, device='cuda:0'), 'train_loss_giou': tensor(0.9655, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.9176, device='cuda:0'), 'validation_loss_ce': tensor(1.0683, device='cuda:0'), 'validation_loss_bbox': tensor(0.3584, device='cuda:0'), 'validation_loss_giou': tensor(1.0287, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(4.9176, device='cuda:0'), 'train_loss_ce': tensor(1.0683, device='cuda:0'), 'train_loss_bbox': tensor(0.3584, device='cuda:0'), 'train_loss_giou': tensor(1.0287, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.5064, device='cuda:0'), 'validation_loss_ce': tensor(1.0346, device='cuda:0'), 'validation_loss_bbox': tensor(0.2908, device='cuda:0'), 'validation_loss_giou': tensor(1.0090, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(4.5064, device='cuda:0'), 'train_loss_ce': tensor(1.0346, device='cuda:0'), 'train_loss_bbox': tensor(0.2908, device='cuda:0'), 'train_loss_giou': tensor(1.0090, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.3008, device='cuda:0'), 'validation_loss_ce': tensor(0.9959, device='cuda:0'), 'validation_loss_bbox': tensor(0.2638, device='cuda:0'), 'validation_loss_giou': tensor(0.9930, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(4.3008, device='cuda:0'), 'train_loss_ce': tensor(0.9959, device='cuda:0'), 'train_loss_bbox': tensor(0.2638, device='cuda:0'), 'train_loss_giou': tensor(0.9930, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.6145, device='cuda:0'), 'validation_loss_ce': tensor(0.9614, device='cuda:0'), 'validation_loss_bbox': tensor(0.3392, device='cuda:0'), 'validation_loss_giou': tensor(0.9787, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(4.6145, device='cuda:0'), 'train_loss_ce': tensor(0.9614, device='cuda:0'), 'train_loss_bbox': tensor(0.3392, device='cuda:0'), 'train_loss_giou': tensor(0.9787, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.4039, device='cuda:0'), 'validation_loss_ce': tensor(0.9249, device='cuda:0'), 'validation_loss_bbox': tensor(0.3214, device='cuda:0'), 'validation_loss_giou': tensor(0.9359, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(4.4039, device='cuda:0'), 'train_loss_ce': tensor(0.9249, device='cuda:0'), 'train_loss_bbox': tensor(0.3214, device='cuda:0'), 'train_loss_giou': tensor(0.9359, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.5377, device='cuda:0'), 'validation_loss_ce': tensor(0.8932, device='cuda:0'), 'validation_loss_bbox': tensor(0.3316, device='cuda:0'), 'validation_loss_giou': tensor(0.9933, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(4.5377, device='cuda:0'), 'train_loss_ce': tensor(0.8932, device='cuda:0'), 'train_loss_bbox': tensor(0.3316, device='cuda:0'), 'train_loss_giou': tensor(0.9933, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.3868, device='cuda:0'), 'validation_loss_ce': tensor(0.8632, device='cuda:0'), 'validation_loss_bbox': tensor(0.3171, device='cuda:0'), 'validation_loss_giou': tensor(0.9690, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(4.3868, device='cuda:0'), 'train_loss_ce': tensor(0.8632, device='cuda:0'), 'train_loss_bbox': tensor(0.3171, device='cuda:0'), 'train_loss_giou': tensor(0.9690, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.2354, device='cuda:0'), 'validation_loss_ce': tensor(0.8325, device='cuda:0'), 'validation_loss_bbox': tensor(0.3248, device='cuda:0'), 'validation_loss_giou': tensor(0.8894, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(4.2354, device='cuda:0'), 'train_loss_ce': tensor(0.8325, device='cuda:0'), 'train_loss_bbox': tensor(0.3248, device='cuda:0'), 'train_loss_giou': tensor(0.8894, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.1366, device='cuda:0'), 'validation_loss_ce': tensor(0.8096, device='cuda:0'), 'validation_loss_bbox': tensor(0.3125, device='cuda:0'), 'validation_loss_giou': tensor(0.8823, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(4.1366, device='cuda:0'), 'train_loss_ce': tensor(0.8096, device='cuda:0'), 'train_loss_bbox': tensor(0.3125, device='cuda:0'), 'train_loss_giou': tensor(0.8823, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.0733, device='cuda:0'), 'validation_loss_ce': tensor(0.7833, device='cuda:0'), 'validation_loss_bbox': tensor(0.3110, device='cuda:0'), 'validation_loss_giou': tensor(0.8676, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(4.0733, device='cuda:0'), 'train_loss_ce': tensor(0.7833, device='cuda:0'), 'train_loss_bbox': tensor(0.3110, device='cuda:0'), 'train_loss_giou': tensor(0.8676, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.0966, device='cuda:0'), 'validation_loss_ce': tensor(0.7600, device='cuda:0'), 'validation_loss_bbox': tensor(0.3066, device='cuda:0'), 'validation_loss_giou': tensor(0.9018, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(4.0966, device='cuda:0'), 'train_loss_ce': tensor(0.7600, device='cuda:0'), 'train_loss_bbox': tensor(0.3066, device='cuda:0'), 'train_loss_giou': tensor(0.9018, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.0793, device='cuda:0'), 'validation_loss_ce': tensor(0.7322, device='cuda:0'), 'validation_loss_bbox': tensor(0.3120, device='cuda:0'), 'validation_loss_giou': tensor(0.8934, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(4.0793, device='cuda:0'), 'train_loss_ce': tensor(0.7322, device='cuda:0'), 'train_loss_bbox': tensor(0.3120, device='cuda:0'), 'train_loss_giou': tensor(0.8934, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.1925, device='cuda:0'), 'validation_loss_ce': tensor(0.7148, device='cuda:0'), 'validation_loss_bbox': tensor(0.3127, device='cuda:0'), 'validation_loss_giou': tensor(0.9571, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(4.1925, device='cuda:0'), 'train_loss_ce': tensor(0.7148, device='cuda:0'), 'train_loss_bbox': tensor(0.3127, device='cuda:0'), 'train_loss_giou': tensor(0.9571, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.1689, device='cuda:0'), 'validation_loss_ce': tensor(0.6909, device='cuda:0'), 'validation_loss_bbox': tensor(0.3186, device='cuda:0'), 'validation_loss_giou': tensor(0.9426, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(4.1689, device='cuda:0'), 'train_loss_ce': tensor(0.6909, device='cuda:0'), 'train_loss_bbox': tensor(0.3186, device='cuda:0'), 'train_loss_giou': tensor(0.9426, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.1089, device='cuda:0'), 'validation_loss_ce': tensor(0.6678, device='cuda:0'), 'validation_loss_bbox': tensor(0.3087, device='cuda:0'), 'validation_loss_giou': tensor(0.9487, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(4.1089, device='cuda:0'), 'train_loss_ce': tensor(0.6678, device='cuda:0'), 'train_loss_bbox': tensor(0.3087, device='cuda:0'), 'train_loss_giou': tensor(0.9487, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.0811, device='cuda:0'), 'validation_loss_ce': tensor(0.6494, device='cuda:0'), 'validation_loss_bbox': tensor(0.3144, device='cuda:0'), 'validation_loss_giou': tensor(0.9299, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(4.0811, device='cuda:0'), 'train_loss_ce': tensor(0.6494, device='cuda:0'), 'train_loss_bbox': tensor(0.3144, device='cuda:0'), 'train_loss_giou': tensor(0.9299, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.9771, device='cuda:0'), 'validation_loss_ce': tensor(0.6317, device='cuda:0'), 'validation_loss_bbox': tensor(0.3040, device='cuda:0'), 'validation_loss_giou': tensor(0.9128, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(3.9771, device='cuda:0'), 'train_loss_ce': tensor(0.6317, device='cuda:0'), 'train_loss_bbox': tensor(0.3040, device='cuda:0'), 'train_loss_giou': tensor(0.9128, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.8654, device='cuda:0'), 'validation_loss_ce': tensor(0.6171, device='cuda:0'), 'validation_loss_bbox': tensor(0.2980, device='cuda:0'), 'validation_loss_giou': tensor(0.8791, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(3.8654, device='cuda:0'), 'train_loss_ce': tensor(0.6171, device='cuda:0'), 'train_loss_bbox': tensor(0.2980, device='cuda:0'), 'train_loss_giou': tensor(0.8791, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.7580, device='cuda:0'), 'validation_loss_ce': tensor(0.6039, device='cuda:0'), 'validation_loss_bbox': tensor(0.2968, device='cuda:0'), 'validation_loss_giou': tensor(0.8351, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(3.7580, device='cuda:0'), 'train_loss_ce': tensor(0.6039, device='cuda:0'), 'train_loss_bbox': tensor(0.2968, device='cuda:0'), 'train_loss_giou': tensor(0.8351, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.9037, device='cuda:0'), 'validation_loss_ce': tensor(0.5912, device='cuda:0'), 'validation_loss_bbox': tensor(0.3089, device='cuda:0'), 'validation_loss_giou': tensor(0.8840, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(3.9037, device='cuda:0'), 'train_loss_ce': tensor(0.5912, device='cuda:0'), 'train_loss_bbox': tensor(0.3089, device='cuda:0'), 'train_loss_giou': tensor(0.8840, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.8569, device='cuda:0'), 'validation_loss_ce': tensor(0.5805, device='cuda:0'), 'validation_loss_bbox': tensor(0.3027, device='cuda:0'), 'validation_loss_giou': tensor(0.8816, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(3.8569, device='cuda:0'), 'train_loss_ce': tensor(0.5805, device='cuda:0'), 'train_loss_bbox': tensor(0.3027, device='cuda:0'), 'train_loss_giou': tensor(0.8816, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.7900, device='cuda:0'), 'validation_loss_ce': tensor(0.5685, device='cuda:0'), 'validation_loss_bbox': tensor(0.3083, device='cuda:0'), 'validation_loss_giou': tensor(0.8400, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(3.7900, device='cuda:0'), 'train_loss_ce': tensor(0.5685, device='cuda:0'), 'train_loss_bbox': tensor(0.3083, device='cuda:0'), 'train_loss_giou': tensor(0.8400, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.7441, device='cuda:0'), 'validation_loss_ce': tensor(0.5645, device='cuda:0'), 'validation_loss_bbox': tensor(0.2991, device='cuda:0'), 'validation_loss_giou': tensor(0.8421, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(3.7441, device='cuda:0'), 'train_loss_ce': tensor(0.5645, device='cuda:0'), 'train_loss_bbox': tensor(0.2991, device='cuda:0'), 'train_loss_giou': tensor(0.8421, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.7900, device='cuda:0'), 'validation_loss_ce': tensor(0.5624, device='cuda:0'), 'validation_loss_bbox': tensor(0.3031, device='cuda:0'), 'validation_loss_giou': tensor(0.8562, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(3.7900, device='cuda:0'), 'train_loss_ce': tensor(0.5624, device='cuda:0'), 'train_loss_bbox': tensor(0.3031, device='cuda:0'), 'train_loss_giou': tensor(0.8562, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.8771, device='cuda:0'), 'validation_loss_ce': tensor(0.5557, device='cuda:0'), 'validation_loss_bbox': tensor(0.3262, device='cuda:0'), 'validation_loss_giou': tensor(0.8453, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(3.8771, device='cuda:0'), 'train_loss_ce': tensor(0.5557, device='cuda:0'), 'train_loss_bbox': tensor(0.3262, device='cuda:0'), 'train_loss_giou': tensor(0.8453, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.6650, device='cuda:0'), 'validation_loss_ce': tensor(0.5487, device='cuda:0'), 'validation_loss_bbox': tensor(0.3064, device='cuda:0'), 'validation_loss_giou': tensor(0.7922, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(3.6650, device='cuda:0'), 'train_loss_ce': tensor(0.5487, device='cuda:0'), 'train_loss_bbox': tensor(0.3064, device='cuda:0'), 'train_loss_giou': tensor(0.7922, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.5456, device='cuda:0'), 'validation_loss_ce': tensor(0.5411, device='cuda:0'), 'validation_loss_bbox': tensor(0.4130, device='cuda:0'), 'validation_loss_giou': tensor(0.9698, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(4.5456, device='cuda:0'), 'train_loss_ce': tensor(0.5411, device='cuda:0'), 'train_loss_bbox': tensor(0.4130, device='cuda:0'), 'train_loss_giou': tensor(0.9698, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.6485, device='cuda:0'), 'validation_loss_ce': tensor(0.5304, device='cuda:0'), 'validation_loss_bbox': tensor(0.4590, device='cuda:0'), 'validation_loss_giou': tensor(0.9116, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(4.6485, device='cuda:0'), 'train_loss_ce': tensor(0.5304, device='cuda:0'), 'train_loss_bbox': tensor(0.4590, device='cuda:0'), 'train_loss_giou': tensor(0.9116, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.5504, device='cuda:0'), 'validation_loss_ce': tensor(0.5244, device='cuda:0'), 'validation_loss_bbox': tensor(0.4588, device='cuda:0'), 'validation_loss_giou': tensor(0.8659, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(4.5504, device='cuda:0'), 'train_loss_ce': tensor(0.5244, device='cuda:0'), 'train_loss_bbox': tensor(0.4588, device='cuda:0'), 'train_loss_giou': tensor(0.8659, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.5681, device='cuda:0'), 'validation_loss_ce': tensor(0.5229, device='cuda:0'), 'validation_loss_bbox': tensor(0.4541, device='cuda:0'), 'validation_loss_giou': tensor(0.8874, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(4.5681, device='cuda:0'), 'train_loss_ce': tensor(0.5229, device='cuda:0'), 'train_loss_bbox': tensor(0.4541, device='cuda:0'), 'train_loss_giou': tensor(0.8874, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.5211, device='cuda:0'), 'validation_loss_ce': tensor(0.5187, device='cuda:0'), 'validation_loss_bbox': tensor(0.4401, device='cuda:0'), 'validation_loss_giou': tensor(0.9010, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(4.5211, device='cuda:0'), 'train_loss_ce': tensor(0.5187, device='cuda:0'), 'train_loss_bbox': tensor(0.4401, device='cuda:0'), 'train_loss_giou': tensor(0.9010, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.3318, device='cuda:0'), 'validation_loss_ce': tensor(0.5132, device='cuda:0'), 'validation_loss_bbox': tensor(0.4153, device='cuda:0'), 'validation_loss_giou': tensor(0.8711, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(4.3318, device='cuda:0'), 'train_loss_ce': tensor(0.5132, device='cuda:0'), 'train_loss_bbox': tensor(0.4153, device='cuda:0'), 'train_loss_giou': tensor(0.8711, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.0188, device='cuda:0'), 'validation_loss_ce': tensor(0.5055, device='cuda:0'), 'validation_loss_bbox': tensor(0.3876, device='cuda:0'), 'validation_loss_giou': tensor(0.7876, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(4.0188, device='cuda:0'), 'train_loss_ce': tensor(0.5055, device='cuda:0'), 'train_loss_bbox': tensor(0.3876, device='cuda:0'), 'train_loss_giou': tensor(0.7876, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.1366, device='cuda:0'), 'validation_loss_ce': tensor(0.4989, device='cuda:0'), 'validation_loss_bbox': tensor(0.3734, device='cuda:0'), 'validation_loss_giou': tensor(0.8854, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(4.1366, device='cuda:0'), 'train_loss_ce': tensor(0.4989, device='cuda:0'), 'train_loss_bbox': tensor(0.3734, device='cuda:0'), 'train_loss_giou': tensor(0.8854, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.1246, device='cuda:0'), 'validation_loss_ce': tensor(0.4929, device='cuda:0'), 'validation_loss_bbox': tensor(0.3589, device='cuda:0'), 'validation_loss_giou': tensor(0.9187, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(4.1246, device='cuda:0'), 'train_loss_ce': tensor(0.4929, device='cuda:0'), 'train_loss_bbox': tensor(0.3589, device='cuda:0'), 'train_loss_giou': tensor(0.9187, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.0014, device='cuda:0'), 'validation_loss_ce': tensor(0.4860, device='cuda:0'), 'validation_loss_bbox': tensor(0.3432, device='cuda:0'), 'validation_loss_giou': tensor(0.8996, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(4.0014, device='cuda:0'), 'train_loss_ce': tensor(0.4860, device='cuda:0'), 'train_loss_bbox': tensor(0.3432, device='cuda:0'), 'train_loss_giou': tensor(0.8996, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.7879, device='cuda:0'), 'validation_loss_ce': tensor(0.4811, device='cuda:0'), 'validation_loss_bbox': tensor(0.3114, device='cuda:0'), 'validation_loss_giou': tensor(0.8748, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(3.7879, device='cuda:0'), 'train_loss_ce': tensor(0.4811, device='cuda:0'), 'train_loss_bbox': tensor(0.3114, device='cuda:0'), 'train_loss_giou': tensor(0.8748, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.5049, device='cuda:0'), 'validation_loss_ce': tensor(0.4802, device='cuda:0'), 'validation_loss_bbox': tensor(0.2632, device='cuda:0'), 'validation_loss_giou': tensor(0.8545, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(3.5049, device='cuda:0'), 'train_loss_ce': tensor(0.4802, device='cuda:0'), 'train_loss_bbox': tensor(0.2632, device='cuda:0'), 'train_loss_giou': tensor(0.8545, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.4778, device='cuda:0'), 'validation_loss_ce': tensor(0.4773, device='cuda:0'), 'validation_loss_bbox': tensor(0.2375, device='cuda:0'), 'validation_loss_giou': tensor(0.9065, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(3.4778, device='cuda:0'), 'train_loss_ce': tensor(0.4773, device='cuda:0'), 'train_loss_bbox': tensor(0.2375, device='cuda:0'), 'train_loss_giou': tensor(0.9065, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.4224, device='cuda:0'), 'validation_loss_ce': tensor(0.4733, device='cuda:0'), 'validation_loss_bbox': tensor(0.2288, device='cuda:0'), 'validation_loss_giou': tensor(0.9026, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(3.4224, device='cuda:0'), 'train_loss_ce': tensor(0.4733, device='cuda:0'), 'train_loss_bbox': tensor(0.2288, device='cuda:0'), 'train_loss_giou': tensor(0.9026, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.2998, device='cuda:0'), 'validation_loss_ce': tensor(0.4627, device='cuda:0'), 'validation_loss_bbox': tensor(0.2316, device='cuda:0'), 'validation_loss_giou': tensor(0.8397, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(3.2998, device='cuda:0'), 'train_loss_ce': tensor(0.4627, device='cuda:0'), 'train_loss_bbox': tensor(0.2316, device='cuda:0'), 'train_loss_giou': tensor(0.8397, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.1477, device='cuda:0'), 'validation_loss_ce': tensor(0.4610, device='cuda:0'), 'validation_loss_bbox': tensor(0.2228, device='cuda:0'), 'validation_loss_giou': tensor(0.7864, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(3.1477, device='cuda:0'), 'train_loss_ce': tensor(0.4610, device='cuda:0'), 'train_loss_bbox': tensor(0.2228, device='cuda:0'), 'train_loss_giou': tensor(0.7864, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.1142, device='cuda:0'), 'validation_loss_ce': tensor(0.4583, device='cuda:0'), 'validation_loss_bbox': tensor(0.2191, device='cuda:0'), 'validation_loss_giou': tensor(0.7803, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(3.1142, device='cuda:0'), 'train_loss_ce': tensor(0.4583, device='cuda:0'), 'train_loss_bbox': tensor(0.2191, device='cuda:0'), 'train_loss_giou': tensor(0.7803, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.0002, device='cuda:0'), 'validation_loss_ce': tensor(0.4554, device='cuda:0'), 'validation_loss_bbox': tensor(0.2115, device='cuda:0'), 'validation_loss_giou': tensor(0.7435, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(3.0002, device='cuda:0'), 'train_loss_ce': tensor(0.4554, device='cuda:0'), 'train_loss_bbox': tensor(0.2115, device='cuda:0'), 'train_loss_giou': tensor(0.7435, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.9943, device='cuda:0'), 'validation_loss_ce': tensor(0.4482, device='cuda:0'), 'validation_loss_bbox': tensor(0.2124, device='cuda:0'), 'validation_loss_giou': tensor(0.7420, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.9943, device='cuda:0'), 'train_loss_ce': tensor(0.4482, device='cuda:0'), 'train_loss_bbox': tensor(0.2124, device='cuda:0'), 'train_loss_giou': tensor(0.7420, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.0455, device='cuda:0'), 'validation_loss_ce': tensor(0.4422, device='cuda:0'), 'validation_loss_bbox': tensor(0.2170, device='cuda:0'), 'validation_loss_giou': tensor(0.7592, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(3.0455, device='cuda:0'), 'train_loss_ce': tensor(0.4422, device='cuda:0'), 'train_loss_bbox': tensor(0.2170, device='cuda:0'), 'train_loss_giou': tensor(0.7592, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.9776, device='cuda:0'), 'validation_loss_ce': tensor(0.4369, device='cuda:0'), 'validation_loss_bbox': tensor(0.2112, device='cuda:0'), 'validation_loss_giou': tensor(0.7422, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.9776, device='cuda:0'), 'train_loss_ce': tensor(0.4369, device='cuda:0'), 'train_loss_bbox': tensor(0.2112, device='cuda:0'), 'train_loss_giou': tensor(0.7422, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.9267, device='cuda:0'), 'validation_loss_ce': tensor(0.4307, device='cuda:0'), 'validation_loss_bbox': tensor(0.2064, device='cuda:0'), 'validation_loss_giou': tensor(0.7320, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.9267, device='cuda:0'), 'train_loss_ce': tensor(0.4307, device='cuda:0'), 'train_loss_bbox': tensor(0.2064, device='cuda:0'), 'train_loss_giou': tensor(0.7320, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.7626, device='cuda:0'), 'validation_loss_ce': tensor(0.4253, device='cuda:0'), 'validation_loss_bbox': tensor(0.1968, device='cuda:0'), 'validation_loss_giou': tensor(0.6767, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.7626, device='cuda:0'), 'train_loss_ce': tensor(0.4253, device='cuda:0'), 'train_loss_bbox': tensor(0.1968, device='cuda:0'), 'train_loss_giou': tensor(0.6767, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.7814, device='cuda:0'), 'validation_loss_ce': tensor(0.4684, device='cuda:0'), 'validation_loss_bbox': tensor(0.1021, device='cuda:0'), 'validation_loss_giou': tensor(0.4014, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.7814, device='cuda:0'), 'train_loss_ce': tensor(0.4684, device='cuda:0'), 'train_loss_bbox': tensor(0.1021, device='cuda:0'), 'train_loss_giou': tensor(0.4014, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.5182, device='cuda:0'), 'validation_loss_ce': tensor(0.5261, device='cuda:0'), 'validation_loss_bbox': tensor(0.1545, device='cuda:0'), 'validation_loss_giou': tensor(0.6097, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.5182, device='cuda:0'), 'train_loss_ce': tensor(0.5261, device='cuda:0'), 'train_loss_bbox': tensor(0.1545, device='cuda:0'), 'train_loss_giou': tensor(0.6097, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.1803, device='cuda:0'), 'validation_loss_ce': tensor(0.5506, device='cuda:0'), 'validation_loss_bbox': tensor(0.2151, device='cuda:0'), 'validation_loss_giou': tensor(0.7771, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(3.1803, device='cuda:0'), 'train_loss_ce': tensor(0.5506, device='cuda:0'), 'train_loss_bbox': tensor(0.2151, device='cuda:0'), 'train_loss_giou': tensor(0.7771, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.4451, device='cuda:0'), 'validation_loss_ce': tensor(0.5370, device='cuda:0'), 'validation_loss_bbox': tensor(0.1379, device='cuda:0'), 'validation_loss_giou': tensor(0.6094, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.4451, device='cuda:0'), 'train_loss_ce': tensor(0.5370, device='cuda:0'), 'train_loss_bbox': tensor(0.1379, device='cuda:0'), 'train_loss_giou': tensor(0.6094, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.8647, device='cuda:0'), 'validation_loss_ce': tensor(0.5240, device='cuda:0'), 'validation_loss_bbox': tensor(0.1929, device='cuda:0'), 'validation_loss_giou': tensor(0.6882, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.8647, device='cuda:0'), 'train_loss_ce': tensor(0.5240, device='cuda:0'), 'train_loss_bbox': tensor(0.1929, device='cuda:0'), 'train_loss_giou': tensor(0.6882, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.7498, device='cuda:0'), 'validation_loss_ce': tensor(0.5096, device='cuda:0'), 'validation_loss_bbox': tensor(0.1838, device='cuda:0'), 'validation_loss_giou': tensor(0.6606, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.7498, device='cuda:0'), 'train_loss_ce': tensor(0.5096, device='cuda:0'), 'train_loss_bbox': tensor(0.1838, device='cuda:0'), 'train_loss_giou': tensor(0.6606, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.2658, device='cuda:0'), 'validation_loss_ce': tensor(0.4970, device='cuda:0'), 'validation_loss_bbox': tensor(0.1442, device='cuda:0'), 'validation_loss_giou': tensor(0.5240, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.2658, device='cuda:0'), 'train_loss_ce': tensor(0.4970, device='cuda:0'), 'train_loss_bbox': tensor(0.1442, device='cuda:0'), 'train_loss_giou': tensor(0.5240, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.1836, device='cuda:0'), 'validation_loss_ce': tensor(0.4837, device='cuda:0'), 'validation_loss_bbox': tensor(0.1189, device='cuda:0'), 'validation_loss_giou': tensor(0.5527, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.1836, device='cuda:0'), 'train_loss_ce': tensor(0.4837, device='cuda:0'), 'train_loss_bbox': tensor(0.1189, device='cuda:0'), 'train_loss_giou': tensor(0.5527, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.7796, device='cuda:0'), 'validation_loss_ce': tensor(0.4814, device='cuda:0'), 'validation_loss_bbox': tensor(0.0791, device='cuda:0'), 'validation_loss_giou': tensor(0.4514, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.7796, device='cuda:0'), 'train_loss_ce': tensor(0.4814, device='cuda:0'), 'train_loss_bbox': tensor(0.0791, device='cuda:0'), 'train_loss_giou': tensor(0.4514, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.5771, device='cuda:0'), 'validation_loss_ce': tensor(0.4849, device='cuda:0'), 'validation_loss_bbox': tensor(0.0657, device='cuda:0'), 'validation_loss_giou': tensor(0.3818, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.5771, device='cuda:0'), 'train_loss_ce': tensor(0.4849, device='cuda:0'), 'train_loss_bbox': tensor(0.0657, device='cuda:0'), 'train_loss_giou': tensor(0.3818, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.3797, device='cuda:0'), 'validation_loss_ce': tensor(0.4945, device='cuda:0'), 'validation_loss_bbox': tensor(0.0573, device='cuda:0'), 'validation_loss_giou': tensor(0.2994, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.3797, device='cuda:0'), 'train_loss_ce': tensor(0.4945, device='cuda:0'), 'train_loss_bbox': tensor(0.0573, device='cuda:0'), 'train_loss_giou': tensor(0.2994, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.5624, device='cuda:0'), 'validation_loss_ce': tensor(0.4785, device='cuda:0'), 'validation_loss_bbox': tensor(0.0731, device='cuda:0'), 'validation_loss_giou': tensor(0.3593, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.5624, device='cuda:0'), 'train_loss_ce': tensor(0.4785, device='cuda:0'), 'train_loss_bbox': tensor(0.0731, device='cuda:0'), 'train_loss_giou': tensor(0.3593, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.8937, device='cuda:0'), 'validation_loss_ce': tensor(0.4943, device='cuda:0'), 'validation_loss_bbox': tensor(0.0838, device='cuda:0'), 'validation_loss_giou': tensor(0.4903, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.8937, device='cuda:0'), 'train_loss_ce': tensor(0.4943, device='cuda:0'), 'train_loss_bbox': tensor(0.0838, device='cuda:0'), 'train_loss_giou': tensor(0.4903, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.5233, device='cuda:0'), 'validation_loss_ce': tensor(0.5102, device='cuda:0'), 'validation_loss_bbox': tensor(0.0664, device='cuda:0'), 'validation_loss_giou': tensor(0.3406, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.5233, device='cuda:0'), 'train_loss_ce': tensor(0.5102, device='cuda:0'), 'train_loss_bbox': tensor(0.0664, device='cuda:0'), 'train_loss_giou': tensor(0.3406, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.8535, device='cuda:0'), 'validation_loss_ce': tensor(0.5116, device='cuda:0'), 'validation_loss_bbox': tensor(0.0938, device='cuda:0'), 'validation_loss_giou': tensor(0.4365, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.8535, device='cuda:0'), 'train_loss_ce': tensor(0.5116, device='cuda:0'), 'train_loss_bbox': tensor(0.0938, device='cuda:0'), 'train_loss_giou': tensor(0.4365, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.7981, device='cuda:0'), 'validation_loss_ce': tensor(0.5058, device='cuda:0'), 'validation_loss_bbox': tensor(0.0915, device='cuda:0'), 'validation_loss_giou': tensor(0.4174, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.7981, device='cuda:0'), 'train_loss_ce': tensor(0.5058, device='cuda:0'), 'train_loss_bbox': tensor(0.0915, device='cuda:0'), 'train_loss_giou': tensor(0.4174, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.5423, device='cuda:0'), 'validation_loss_ce': tensor(0.4984, device='cuda:0'), 'validation_loss_bbox': tensor(0.0670, device='cuda:0'), 'validation_loss_giou': tensor(0.3543, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.5423, device='cuda:0'), 'train_loss_ce': tensor(0.4984, device='cuda:0'), 'train_loss_bbox': tensor(0.0670, device='cuda:0'), 'train_loss_giou': tensor(0.3543, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.4708, device='cuda:0'), 'validation_loss_ce': tensor(0.4898, device='cuda:0'), 'validation_loss_bbox': tensor(0.0620, device='cuda:0'), 'validation_loss_giou': tensor(0.3355, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.4708, device='cuda:0'), 'train_loss_ce': tensor(0.4898, device='cuda:0'), 'train_loss_bbox': tensor(0.0620, device='cuda:0'), 'train_loss_giou': tensor(0.3355, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.2020, device='cuda:0'), 'validation_loss_ce': tensor(0.4936, device='cuda:0'), 'validation_loss_bbox': tensor(0.0431, device='cuda:0'), 'validation_loss_giou': tensor(0.2464, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.2020, device='cuda:0'), 'train_loss_ce': tensor(0.4936, device='cuda:0'), 'train_loss_bbox': tensor(0.0431, device='cuda:0'), 'train_loss_giou': tensor(0.2464, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.6945, device='cuda:0'), 'validation_loss_ce': tensor(0.5133, device='cuda:0'), 'validation_loss_bbox': tensor(0.0717, device='cuda:0'), 'validation_loss_giou': tensor(0.4114, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.6945, device='cuda:0'), 'train_loss_ce': tensor(0.5133, device='cuda:0'), 'train_loss_bbox': tensor(0.0717, device='cuda:0'), 'train_loss_giou': tensor(0.4114, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.3169, device='cuda:0'), 'validation_loss_ce': tensor(0.5157, device='cuda:0'), 'validation_loss_bbox': tensor(0.0449, device='cuda:0'), 'validation_loss_giou': tensor(0.2882, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.3169, device='cuda:0'), 'train_loss_ce': tensor(0.5157, device='cuda:0'), 'train_loss_bbox': tensor(0.0449, device='cuda:0'), 'train_loss_giou': tensor(0.2882, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.1922, device='cuda:0'), 'validation_loss_ce': tensor(0.5193, device='cuda:0'), 'validation_loss_bbox': tensor(0.0421, device='cuda:0'), 'validation_loss_giou': tensor(0.2312, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.1922, device='cuda:0'), 'train_loss_ce': tensor(0.5193, device='cuda:0'), 'train_loss_bbox': tensor(0.0421, device='cuda:0'), 'train_loss_giou': tensor(0.2312, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.4203, device='cuda:0'), 'validation_loss_ce': tensor(0.4967, device='cuda:0'), 'validation_loss_bbox': tensor(0.0586, device='cuda:0'), 'validation_loss_giou': tensor(0.3154, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.4203, device='cuda:0'), 'train_loss_ce': tensor(0.4967, device='cuda:0'), 'train_loss_bbox': tensor(0.0586, device='cuda:0'), 'train_loss_giou': tensor(0.3154, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.4091, device='cuda:0'), 'validation_loss_ce': tensor(0.5036, device='cuda:0'), 'validation_loss_bbox': tensor(0.0634, device='cuda:0'), 'validation_loss_giou': tensor(0.2943, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.4091, device='cuda:0'), 'train_loss_ce': tensor(0.5036, device='cuda:0'), 'train_loss_bbox': tensor(0.0634, device='cuda:0'), 'train_loss_giou': tensor(0.2943, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.4067, device='cuda:0'), 'validation_loss_ce': tensor(0.4804, device='cuda:0'), 'validation_loss_bbox': tensor(0.0409, device='cuda:0'), 'validation_loss_giou': tensor(0.3608, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.4067, device='cuda:0'), 'train_loss_ce': tensor(0.4804, device='cuda:0'), 'train_loss_bbox': tensor(0.0409, device='cuda:0'), 'train_loss_giou': tensor(0.3608, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.1247, device='cuda:0'), 'validation_loss_ce': tensor(0.4851, device='cuda:0'), 'validation_loss_bbox': tensor(0.0284, device='cuda:0'), 'validation_loss_giou': tensor(0.2489, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.1247, device='cuda:0'), 'train_loss_ce': tensor(0.4851, device='cuda:0'), 'train_loss_bbox': tensor(0.0284, device='cuda:0'), 'train_loss_giou': tensor(0.2489, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.9934, device='cuda:0'), 'validation_loss_ce': tensor(0.4870, device='cuda:0'), 'validation_loss_bbox': tensor(0.0277, device='cuda:0'), 'validation_loss_giou': tensor(0.1839, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.9934, device='cuda:0'), 'train_loss_ce': tensor(0.4870, device='cuda:0'), 'train_loss_bbox': tensor(0.0277, device='cuda:0'), 'train_loss_giou': tensor(0.1839, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.1130, device='cuda:0'), 'validation_loss_ce': tensor(0.4643, device='cuda:0'), 'validation_loss_bbox': tensor(0.0427, device='cuda:0'), 'validation_loss_giou': tensor(0.2176, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.1130, device='cuda:0'), 'train_loss_ce': tensor(0.4643, device='cuda:0'), 'train_loss_bbox': tensor(0.0427, device='cuda:0'), 'train_loss_giou': tensor(0.2176, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.9284, device='cuda:0'), 'validation_loss_ce': tensor(0.4571, device='cuda:0'), 'validation_loss_bbox': tensor(0.0195, device='cuda:0'), 'validation_loss_giou': tensor(0.1869, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.9284, device='cuda:0'), 'train_loss_ce': tensor(0.4571, device='cuda:0'), 'train_loss_bbox': tensor(0.0195, device='cuda:0'), 'train_loss_giou': tensor(0.1869, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.1390, device='cuda:0'), 'validation_loss_ce': tensor(0.4492, device='cuda:0'), 'validation_loss_bbox': tensor(0.0426, device='cuda:0'), 'validation_loss_giou': tensor(0.2385, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.1390, device='cuda:0'), 'train_loss_ce': tensor(0.4492, device='cuda:0'), 'train_loss_bbox': tensor(0.0426, device='cuda:0'), 'train_loss_giou': tensor(0.2385, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.9522, device='cuda:0'), 'validation_loss_ce': tensor(0.4392, device='cuda:0'), 'validation_loss_bbox': tensor(0.0280, device='cuda:0'), 'validation_loss_giou': tensor(0.1864, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.9522, device='cuda:0'), 'train_loss_ce': tensor(0.4392, device='cuda:0'), 'train_loss_bbox': tensor(0.0280, device='cuda:0'), 'train_loss_giou': tensor(0.1864, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.7214, device='cuda:0'), 'validation_loss_ce': tensor(0.4303, device='cuda:0'), 'validation_loss_bbox': tensor(0.0166, device='cuda:0'), 'validation_loss_giou': tensor(0.1040, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.7214, device='cuda:0'), 'train_loss_ce': tensor(0.4303, device='cuda:0'), 'train_loss_bbox': tensor(0.0166, device='cuda:0'), 'train_loss_giou': tensor(0.1040, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.1696, device='cuda:0'), 'validation_loss_ce': tensor(0.4353, device='cuda:0'), 'validation_loss_bbox': tensor(0.0340, device='cuda:0'), 'validation_loss_giou': tensor(0.2823, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.1696, device='cuda:0'), 'train_loss_ce': tensor(0.4353, device='cuda:0'), 'train_loss_bbox': tensor(0.0340, device='cuda:0'), 'train_loss_giou': tensor(0.2823, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.8795, device='cuda:0'), 'validation_loss_ce': tensor(0.4209, device='cuda:0'), 'validation_loss_bbox': tensor(0.0209, device='cuda:0'), 'validation_loss_giou': tensor(0.1769, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.8795, device='cuda:0'), 'train_loss_ce': tensor(0.4209, device='cuda:0'), 'train_loss_bbox': tensor(0.0209, device='cuda:0'), 'train_loss_giou': tensor(0.1769, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.6846, device='cuda:0'), 'validation_loss_ce': tensor(0.4147, device='cuda:0'), 'validation_loss_bbox': tensor(0.0149, device='cuda:0'), 'validation_loss_giou': tensor(0.0977, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.6846, device='cuda:0'), 'train_loss_ce': tensor(0.4147, device='cuda:0'), 'train_loss_bbox': tensor(0.0149, device='cuda:0'), 'train_loss_giou': tensor(0.0977, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.8974, device='cuda:0'), 'validation_loss_ce': tensor(0.4082, device='cuda:0'), 'validation_loss_bbox': tensor(0.0214, device='cuda:0'), 'validation_loss_giou': tensor(0.1910, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.8974, device='cuda:0'), 'train_loss_ce': tensor(0.4082, device='cuda:0'), 'train_loss_bbox': tensor(0.0214, device='cuda:0'), 'train_loss_giou': tensor(0.1910, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.6828, device='cuda:0'), 'validation_loss_ce': tensor(0.4125, device='cuda:0'), 'validation_loss_bbox': tensor(0.0122, device='cuda:0'), 'validation_loss_giou': tensor(0.1046, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.6828, device='cuda:0'), 'train_loss_ce': tensor(0.4125, device='cuda:0'), 'train_loss_bbox': tensor(0.0122, device='cuda:0'), 'train_loss_giou': tensor(0.1046, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.6780, device='cuda:0'), 'validation_loss_ce': tensor(0.4130, device='cuda:0'), 'validation_loss_bbox': tensor(0.0589, device='cuda:0'), 'validation_loss_giou': tensor(0.4852, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.6780, device='cuda:0'), 'train_loss_ce': tensor(0.4130, device='cuda:0'), 'train_loss_bbox': tensor(0.0589, device='cuda:0'), 'train_loss_giou': tensor(0.4852, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.9580, device='cuda:0'), 'validation_loss_ce': tensor(0.4035, device='cuda:0'), 'validation_loss_bbox': tensor(0.0331, device='cuda:0'), 'validation_loss_giou': tensor(0.1945, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.9580, device='cuda:0'), 'train_loss_ce': tensor(0.4035, device='cuda:0'), 'train_loss_bbox': tensor(0.0331, device='cuda:0'), 'train_loss_giou': tensor(0.1945, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.9431, device='cuda:0'), 'validation_loss_ce': tensor(0.4002, device='cuda:0'), 'validation_loss_bbox': tensor(0.0324, device='cuda:0'), 'validation_loss_giou': tensor(0.1903, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.9431, device='cuda:0'), 'train_loss_ce': tensor(0.4002, device='cuda:0'), 'train_loss_bbox': tensor(0.0324, device='cuda:0'), 'train_loss_giou': tensor(0.1903, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.8074, device='cuda:0'), 'validation_loss_ce': tensor(0.3932, device='cuda:0'), 'validation_loss_bbox': tensor(0.0227, device='cuda:0'), 'validation_loss_giou': tensor(0.1503, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.8074, device='cuda:0'), 'train_loss_ce': tensor(0.3932, device='cuda:0'), 'train_loss_bbox': tensor(0.0227, device='cuda:0'), 'train_loss_giou': tensor(0.1503, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.7362, device='cuda:0'), 'validation_loss_ce': tensor(0.3862, device='cuda:0'), 'validation_loss_bbox': tensor(0.0151, device='cuda:0'), 'validation_loss_giou': tensor(0.1373, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.7362, device='cuda:0'), 'train_loss_ce': tensor(0.3862, device='cuda:0'), 'train_loss_bbox': tensor(0.0151, device='cuda:0'), 'train_loss_giou': tensor(0.1373, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.6583, device='cuda:0'), 'validation_loss_ce': tensor(0.3798, device='cuda:0'), 'validation_loss_bbox': tensor(0.0157, device='cuda:0'), 'validation_loss_giou': tensor(0.1000, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.6583, device='cuda:0'), 'train_loss_ce': tensor(0.3798, device='cuda:0'), 'train_loss_bbox': tensor(0.0157, device='cuda:0'), 'train_loss_giou': tensor(0.1000, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.6930, device='cuda:0'), 'validation_loss_ce': tensor(0.3756, device='cuda:0'), 'validation_loss_bbox': tensor(0.0173, device='cuda:0'), 'validation_loss_giou': tensor(0.1154, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.6930, device='cuda:0'), 'train_loss_ce': tensor(0.3756, device='cuda:0'), 'train_loss_bbox': tensor(0.0173, device='cuda:0'), 'train_loss_giou': tensor(0.1154, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.7621, device='cuda:0'), 'validation_loss_ce': tensor(0.3726, device='cuda:0'), 'validation_loss_bbox': tensor(0.0193, device='cuda:0'), 'validation_loss_giou': tensor(0.1464, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.7621, device='cuda:0'), 'train_loss_ce': tensor(0.3726, device='cuda:0'), 'train_loss_bbox': tensor(0.0193, device='cuda:0'), 'train_loss_giou': tensor(0.1464, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.7702, device='cuda:0'), 'validation_loss_ce': tensor(0.3717, device='cuda:0'), 'validation_loss_bbox': tensor(0.0212, device='cuda:0'), 'validation_loss_giou': tensor(0.1462, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.7702, device='cuda:0'), 'train_loss_ce': tensor(0.3717, device='cuda:0'), 'train_loss_bbox': tensor(0.0212, device='cuda:0'), 'train_loss_giou': tensor(0.1462, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.7285, device='cuda:0'), 'validation_loss_ce': tensor(0.3720, device='cuda:0'), 'validation_loss_bbox': tensor(0.0185, device='cuda:0'), 'validation_loss_giou': tensor(0.1320, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.7285, device='cuda:0'), 'train_loss_ce': tensor(0.3720, device='cuda:0'), 'train_loss_bbox': tensor(0.0185, device='cuda:0'), 'train_loss_giou': tensor(0.1320, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.7187, device='cuda:0'), 'validation_loss_ce': tensor(0.3734, device='cuda:0'), 'validation_loss_bbox': tensor(0.0154, device='cuda:0'), 'validation_loss_giou': tensor(0.1342, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.7187, device='cuda:0'), 'train_loss_ce': tensor(0.3734, device='cuda:0'), 'train_loss_bbox': tensor(0.0154, device='cuda:0'), 'train_loss_giou': tensor(0.1342, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.6206, device='cuda:0'), 'validation_loss_ce': tensor(0.3688, device='cuda:0'), 'validation_loss_bbox': tensor(0.0102, device='cuda:0'), 'validation_loss_giou': tensor(0.1003, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.6206, device='cuda:0'), 'train_loss_ce': tensor(0.3688, device='cuda:0'), 'train_loss_bbox': tensor(0.0102, device='cuda:0'), 'train_loss_giou': tensor(0.1003, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.5728, device='cuda:0'), 'validation_loss_ce': tensor(0.3646, device='cuda:0'), 'validation_loss_bbox': tensor(0.0091, device='cuda:0'), 'validation_loss_giou': tensor(0.0812, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.5728, device='cuda:0'), 'train_loss_ce': tensor(0.3646, device='cuda:0'), 'train_loss_bbox': tensor(0.0091, device='cuda:0'), 'train_loss_giou': tensor(0.0812, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.6276, device='cuda:0'), 'validation_loss_ce': tensor(0.3640, device='cuda:0'), 'validation_loss_bbox': tensor(0.0126, device='cuda:0'), 'validation_loss_giou': tensor(0.1003, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.6276, device='cuda:0'), 'train_loss_ce': tensor(0.3640, device='cuda:0'), 'train_loss_bbox': tensor(0.0126, device='cuda:0'), 'train_loss_giou': tensor(0.1003, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.6354, device='cuda:0'), 'validation_loss_ce': tensor(0.3612, device='cuda:0'), 'validation_loss_bbox': tensor(0.0135, device='cuda:0'), 'validation_loss_giou': tensor(0.1033, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.6354, device='cuda:0'), 'train_loss_ce': tensor(0.3612, device='cuda:0'), 'train_loss_bbox': tensor(0.0135, device='cuda:0'), 'train_loss_giou': tensor(0.1033, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.6197, device='cuda:0'), 'validation_loss_ce': tensor(0.3586, device='cuda:0'), 'validation_loss_bbox': tensor(0.0122, device='cuda:0'), 'validation_loss_giou': tensor(0.1000, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.6197, device='cuda:0'), 'train_loss_ce': tensor(0.3586, device='cuda:0'), 'train_loss_bbox': tensor(0.0122, device='cuda:0'), 'train_loss_giou': tensor(0.1000, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.6223, device='cuda:0'), 'validation_loss_ce': tensor(0.3570, device='cuda:0'), 'validation_loss_bbox': tensor(0.0140, device='cuda:0'), 'validation_loss_giou': tensor(0.0977, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.6223, device='cuda:0'), 'train_loss_ce': tensor(0.3570, device='cuda:0'), 'train_loss_bbox': tensor(0.0140, device='cuda:0'), 'train_loss_giou': tensor(0.0977, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4793, device='cuda:0'), 'validation_loss_ce': tensor(0.3574, device='cuda:0'), 'validation_loss_bbox': tensor(0.0064, device='cuda:0'), 'validation_loss_giou': tensor(0.0451, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4793, device='cuda:0'), 'train_loss_ce': tensor(0.3574, device='cuda:0'), 'train_loss_bbox': tensor(0.0064, device='cuda:0'), 'train_loss_giou': tensor(0.0451, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.7346, device='cuda:0'), 'validation_loss_ce': tensor(0.3589, device='cuda:0'), 'validation_loss_bbox': tensor(0.0139, device='cuda:0'), 'validation_loss_giou': tensor(0.1530, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.7346, device='cuda:0'), 'train_loss_ce': tensor(0.3589, device='cuda:0'), 'train_loss_bbox': tensor(0.0139, device='cuda:0'), 'train_loss_giou': tensor(0.1530, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.7305, device='cuda:0'), 'validation_loss_ce': tensor(0.3606, device='cuda:0'), 'validation_loss_bbox': tensor(0.0168, device='cuda:0'), 'validation_loss_giou': tensor(0.1429, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.7305, device='cuda:0'), 'train_loss_ce': tensor(0.3606, device='cuda:0'), 'train_loss_bbox': tensor(0.0168, device='cuda:0'), 'train_loss_giou': tensor(0.1429, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.5383, device='cuda:0'), 'validation_loss_ce': tensor(0.3602, device='cuda:0'), 'validation_loss_bbox': tensor(0.0096, device='cuda:0'), 'validation_loss_giou': tensor(0.0650, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.5383, device='cuda:0'), 'train_loss_ce': tensor(0.3602, device='cuda:0'), 'train_loss_bbox': tensor(0.0096, device='cuda:0'), 'train_loss_giou': tensor(0.0650, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.6690, device='cuda:0'), 'validation_loss_ce': tensor(0.3584, device='cuda:0'), 'validation_loss_bbox': tensor(0.0129, device='cuda:0'), 'validation_loss_giou': tensor(0.1229, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.6690, device='cuda:0'), 'train_loss_ce': tensor(0.3584, device='cuda:0'), 'train_loss_bbox': tensor(0.0129, device='cuda:0'), 'train_loss_giou': tensor(0.1229, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.6236, device='cuda:0'), 'validation_loss_ce': tensor(0.3541, device='cuda:0'), 'validation_loss_bbox': tensor(0.0140, device='cuda:0'), 'validation_loss_giou': tensor(0.0997, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.6236, device='cuda:0'), 'train_loss_ce': tensor(0.3541, device='cuda:0'), 'train_loss_bbox': tensor(0.0140, device='cuda:0'), 'train_loss_giou': tensor(0.0997, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.5408, device='cuda:0'), 'validation_loss_ce': tensor(0.3513, device='cuda:0'), 'validation_loss_bbox': tensor(0.0083, device='cuda:0'), 'validation_loss_giou': tensor(0.0739, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.5408, device='cuda:0'), 'train_loss_ce': tensor(0.3513, device='cuda:0'), 'train_loss_bbox': tensor(0.0083, device='cuda:0'), 'train_loss_giou': tensor(0.0739, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.6693, device='cuda:0'), 'validation_loss_ce': tensor(0.3501, device='cuda:0'), 'validation_loss_bbox': tensor(0.0110, device='cuda:0'), 'validation_loss_giou': tensor(0.1320, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.6693, device='cuda:0'), 'train_loss_ce': tensor(0.3501, device='cuda:0'), 'train_loss_bbox': tensor(0.0110, device='cuda:0'), 'train_loss_giou': tensor(0.1320, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.6217, device='cuda:0'), 'validation_loss_ce': tensor(0.3491, device='cuda:0'), 'validation_loss_bbox': tensor(0.0090, device='cuda:0'), 'validation_loss_giou': tensor(0.1137, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.6217, device='cuda:0'), 'train_loss_ce': tensor(0.3491, device='cuda:0'), 'train_loss_bbox': tensor(0.0090, device='cuda:0'), 'train_loss_giou': tensor(0.1137, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4850, device='cuda:0'), 'validation_loss_ce': tensor(0.3479, device='cuda:0'), 'validation_loss_bbox': tensor(0.0055, device='cuda:0'), 'validation_loss_giou': tensor(0.0549, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4850, device='cuda:0'), 'train_loss_ce': tensor(0.3479, device='cuda:0'), 'train_loss_bbox': tensor(0.0055, device='cuda:0'), 'train_loss_giou': tensor(0.0549, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.5365, device='cuda:0'), 'validation_loss_ce': tensor(0.3471, device='cuda:0'), 'validation_loss_bbox': tensor(0.0078, device='cuda:0'), 'validation_loss_giou': tensor(0.0751, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.5365, device='cuda:0'), 'train_loss_ce': tensor(0.3471, device='cuda:0'), 'train_loss_bbox': tensor(0.0078, device='cuda:0'), 'train_loss_giou': tensor(0.0751, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.5671, device='cuda:0'), 'validation_loss_ce': tensor(0.3466, device='cuda:0'), 'validation_loss_bbox': tensor(0.0101, device='cuda:0'), 'validation_loss_giou': tensor(0.0849, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.5671, device='cuda:0'), 'train_loss_ce': tensor(0.3466, device='cuda:0'), 'train_loss_bbox': tensor(0.0101, device='cuda:0'), 'train_loss_giou': tensor(0.0849, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.5174, device='cuda:0'), 'validation_loss_ce': tensor(0.3462, device='cuda:0'), 'validation_loss_bbox': tensor(0.0092, device='cuda:0'), 'validation_loss_giou': tensor(0.0627, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.5174, device='cuda:0'), 'train_loss_ce': tensor(0.3462, device='cuda:0'), 'train_loss_bbox': tensor(0.0092, device='cuda:0'), 'train_loss_giou': tensor(0.0627, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.5524, device='cuda:0'), 'validation_loss_ce': tensor(0.3454, device='cuda:0'), 'validation_loss_bbox': tensor(0.0113, device='cuda:0'), 'validation_loss_giou': tensor(0.0753, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.5524, device='cuda:0'), 'train_loss_ce': tensor(0.3454, device='cuda:0'), 'train_loss_bbox': tensor(0.0113, device='cuda:0'), 'train_loss_giou': tensor(0.0753, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4886, device='cuda:0'), 'validation_loss_ce': tensor(0.3434, device='cuda:0'), 'validation_loss_bbox': tensor(0.0071, device='cuda:0'), 'validation_loss_giou': tensor(0.0548, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4886, device='cuda:0'), 'train_loss_ce': tensor(0.3434, device='cuda:0'), 'train_loss_bbox': tensor(0.0071, device='cuda:0'), 'train_loss_giou': tensor(0.0548, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4750, device='cuda:0'), 'validation_loss_ce': tensor(0.3416, device='cuda:0'), 'validation_loss_bbox': tensor(0.0062, device='cuda:0'), 'validation_loss_giou': tensor(0.0513, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4750, device='cuda:0'), 'train_loss_ce': tensor(0.3416, device='cuda:0'), 'train_loss_bbox': tensor(0.0062, device='cuda:0'), 'train_loss_giou': tensor(0.0513, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.5415, device='cuda:0'), 'validation_loss_ce': tensor(0.3400, device='cuda:0'), 'validation_loss_bbox': tensor(0.0100, device='cuda:0'), 'validation_loss_giou': tensor(0.0758, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.5415, device='cuda:0'), 'train_loss_ce': tensor(0.3400, device='cuda:0'), 'train_loss_bbox': tensor(0.0100, device='cuda:0'), 'train_loss_giou': tensor(0.0758, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4985, device='cuda:0'), 'validation_loss_ce': tensor(0.3392, device='cuda:0'), 'validation_loss_bbox': tensor(0.0061, device='cuda:0'), 'validation_loss_giou': tensor(0.0645, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4985, device='cuda:0'), 'train_loss_ce': tensor(0.3392, device='cuda:0'), 'train_loss_bbox': tensor(0.0061, device='cuda:0'), 'train_loss_giou': tensor(0.0645, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4817, device='cuda:0'), 'validation_loss_ce': tensor(0.3387, device='cuda:0'), 'validation_loss_bbox': tensor(0.0078, device='cuda:0'), 'validation_loss_giou': tensor(0.0519, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4817, device='cuda:0'), 'train_loss_ce': tensor(0.3387, device='cuda:0'), 'train_loss_bbox': tensor(0.0078, device='cuda:0'), 'train_loss_giou': tensor(0.0519, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.6237, device='cuda:0'), 'validation_loss_ce': tensor(0.3385, device='cuda:0'), 'validation_loss_bbox': tensor(0.0140, device='cuda:0'), 'validation_loss_giou': tensor(0.1076, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.6237, device='cuda:0'), 'train_loss_ce': tensor(0.3385, device='cuda:0'), 'train_loss_bbox': tensor(0.0140, device='cuda:0'), 'train_loss_giou': tensor(0.1076, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.6418, device='cuda:0'), 'validation_loss_ce': tensor(0.3376, device='cuda:0'), 'validation_loss_bbox': tensor(0.0136, device='cuda:0'), 'validation_loss_giou': tensor(0.1181, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.6418, device='cuda:0'), 'train_loss_ce': tensor(0.3376, device='cuda:0'), 'train_loss_bbox': tensor(0.0136, device='cuda:0'), 'train_loss_giou': tensor(0.1181, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.5515, device='cuda:0'), 'validation_loss_ce': tensor(0.3360, device='cuda:0'), 'validation_loss_bbox': tensor(0.0099, device='cuda:0'), 'validation_loss_giou': tensor(0.0831, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.5515, device='cuda:0'), 'train_loss_ce': tensor(0.3360, device='cuda:0'), 'train_loss_bbox': tensor(0.0099, device='cuda:0'), 'train_loss_giou': tensor(0.0831, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4512, device='cuda:0'), 'validation_loss_ce': tensor(0.3340, device='cuda:0'), 'validation_loss_bbox': tensor(0.0079, device='cuda:0'), 'validation_loss_giou': tensor(0.0390, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4512, device='cuda:0'), 'train_loss_ce': tensor(0.3340, device='cuda:0'), 'train_loss_bbox': tensor(0.0079, device='cuda:0'), 'train_loss_giou': tensor(0.0390, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.6148, device='cuda:0'), 'validation_loss_ce': tensor(0.3322, device='cuda:0'), 'validation_loss_bbox': tensor(0.0155, device='cuda:0'), 'validation_loss_giou': tensor(0.1025, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.6148, device='cuda:0'), 'train_loss_ce': tensor(0.3322, device='cuda:0'), 'train_loss_bbox': tensor(0.0155, device='cuda:0'), 'train_loss_giou': tensor(0.1025, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.6107, device='cuda:0'), 'validation_loss_ce': tensor(0.3308, device='cuda:0'), 'validation_loss_bbox': tensor(0.0140, device='cuda:0'), 'validation_loss_giou': tensor(0.1050, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.6107, device='cuda:0'), 'train_loss_ce': tensor(0.3308, device='cuda:0'), 'train_loss_bbox': tensor(0.0140, device='cuda:0'), 'train_loss_giou': tensor(0.1050, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.5029, device='cuda:0'), 'validation_loss_ce': tensor(0.3296, device='cuda:0'), 'validation_loss_bbox': tensor(0.0093, device='cuda:0'), 'validation_loss_giou': tensor(0.0635, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.5029, device='cuda:0'), 'train_loss_ce': tensor(0.3296, device='cuda:0'), 'train_loss_bbox': tensor(0.0093, device='cuda:0'), 'train_loss_giou': tensor(0.0635, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.5343, device='cuda:0'), 'validation_loss_ce': tensor(0.3294, device='cuda:0'), 'validation_loss_bbox': tensor(0.0117, device='cuda:0'), 'validation_loss_giou': tensor(0.0733, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.5343, device='cuda:0'), 'train_loss_ce': tensor(0.3294, device='cuda:0'), 'train_loss_bbox': tensor(0.0117, device='cuda:0'), 'train_loss_giou': tensor(0.0733, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.5938, device='cuda:0'), 'validation_loss_ce': tensor(0.3275, device='cuda:0'), 'validation_loss_bbox': tensor(0.0104, device='cuda:0'), 'validation_loss_giou': tensor(0.1070, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.5938, device='cuda:0'), 'train_loss_ce': tensor(0.3275, device='cuda:0'), 'train_loss_bbox': tensor(0.0104, device='cuda:0'), 'train_loss_giou': tensor(0.1070, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.5726, device='cuda:0'), 'validation_loss_ce': tensor(0.3267, device='cuda:0'), 'validation_loss_bbox': tensor(0.0116, device='cuda:0'), 'validation_loss_giou': tensor(0.0940, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.5726, device='cuda:0'), 'train_loss_ce': tensor(0.3267, device='cuda:0'), 'train_loss_bbox': tensor(0.0116, device='cuda:0'), 'train_loss_giou': tensor(0.0940, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4901, device='cuda:0'), 'validation_loss_ce': tensor(0.3256, device='cuda:0'), 'validation_loss_bbox': tensor(0.0084, device='cuda:0'), 'validation_loss_giou': tensor(0.0612, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4901, device='cuda:0'), 'train_loss_ce': tensor(0.3256, device='cuda:0'), 'train_loss_bbox': tensor(0.0084, device='cuda:0'), 'train_loss_giou': tensor(0.0612, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4965, device='cuda:0'), 'validation_loss_ce': tensor(0.3239, device='cuda:0'), 'validation_loss_bbox': tensor(0.0088, device='cuda:0'), 'validation_loss_giou': tensor(0.0643, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4965, device='cuda:0'), 'train_loss_ce': tensor(0.3239, device='cuda:0'), 'train_loss_bbox': tensor(0.0088, device='cuda:0'), 'train_loss_giou': tensor(0.0643, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.5424, device='cuda:0'), 'validation_loss_ce': tensor(0.3222, device='cuda:0'), 'validation_loss_bbox': tensor(0.0082, device='cuda:0'), 'validation_loss_giou': tensor(0.0895, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.5424, device='cuda:0'), 'train_loss_ce': tensor(0.3222, device='cuda:0'), 'train_loss_bbox': tensor(0.0082, device='cuda:0'), 'train_loss_giou': tensor(0.0895, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.5243, device='cuda:0'), 'validation_loss_ce': tensor(0.3209, device='cuda:0'), 'validation_loss_bbox': tensor(0.0101, device='cuda:0'), 'validation_loss_giou': tensor(0.0765, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.5243, device='cuda:0'), 'train_loss_ce': tensor(0.3209, device='cuda:0'), 'train_loss_bbox': tensor(0.0101, device='cuda:0'), 'train_loss_giou': tensor(0.0765, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.5208, device='cuda:0'), 'validation_loss_ce': tensor(0.3197, device='cuda:0'), 'validation_loss_bbox': tensor(0.0104, device='cuda:0'), 'validation_loss_giou': tensor(0.0745, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.5208, device='cuda:0'), 'train_loss_ce': tensor(0.3197, device='cuda:0'), 'train_loss_bbox': tensor(0.0104, device='cuda:0'), 'train_loss_giou': tensor(0.0745, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4340, device='cuda:0'), 'validation_loss_ce': tensor(0.3186, device='cuda:0'), 'validation_loss_bbox': tensor(0.0056, device='cuda:0'), 'validation_loss_giou': tensor(0.0436, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4340, device='cuda:0'), 'train_loss_ce': tensor(0.3186, device='cuda:0'), 'train_loss_bbox': tensor(0.0056, device='cuda:0'), 'train_loss_giou': tensor(0.0436, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4344, device='cuda:0'), 'validation_loss_ce': tensor(0.3181, device='cuda:0'), 'validation_loss_bbox': tensor(0.0054, device='cuda:0'), 'validation_loss_giou': tensor(0.0447, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4344, device='cuda:0'), 'train_loss_ce': tensor(0.3181, device='cuda:0'), 'train_loss_bbox': tensor(0.0054, device='cuda:0'), 'train_loss_giou': tensor(0.0447, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4758, device='cuda:0'), 'validation_loss_ce': tensor(0.3172, device='cuda:0'), 'validation_loss_bbox': tensor(0.0082, device='cuda:0'), 'validation_loss_giou': tensor(0.0588, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4758, device='cuda:0'), 'train_loss_ce': tensor(0.3172, device='cuda:0'), 'train_loss_bbox': tensor(0.0082, device='cuda:0'), 'train_loss_giou': tensor(0.0588, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4397, device='cuda:0'), 'validation_loss_ce': tensor(0.3159, device='cuda:0'), 'validation_loss_bbox': tensor(0.0062, device='cuda:0'), 'validation_loss_giou': tensor(0.0463, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4397, device='cuda:0'), 'train_loss_ce': tensor(0.3159, device='cuda:0'), 'train_loss_bbox': tensor(0.0062, device='cuda:0'), 'train_loss_giou': tensor(0.0463, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4343, device='cuda:0'), 'validation_loss_ce': tensor(0.3145, device='cuda:0'), 'validation_loss_bbox': tensor(0.0051, device='cuda:0'), 'validation_loss_giou': tensor(0.0472, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4343, device='cuda:0'), 'train_loss_ce': tensor(0.3145, device='cuda:0'), 'train_loss_bbox': tensor(0.0051, device='cuda:0'), 'train_loss_giou': tensor(0.0472, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4379, device='cuda:0'), 'validation_loss_ce': tensor(0.3131, device='cuda:0'), 'validation_loss_bbox': tensor(0.0056, device='cuda:0'), 'validation_loss_giou': tensor(0.0483, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4379, device='cuda:0'), 'train_loss_ce': tensor(0.3131, device='cuda:0'), 'train_loss_bbox': tensor(0.0056, device='cuda:0'), 'train_loss_giou': tensor(0.0483, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4278, device='cuda:0'), 'validation_loss_ce': tensor(0.3116, device='cuda:0'), 'validation_loss_bbox': tensor(0.0056, device='cuda:0'), 'validation_loss_giou': tensor(0.0441, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4278, device='cuda:0'), 'train_loss_ce': tensor(0.3116, device='cuda:0'), 'train_loss_bbox': tensor(0.0056, device='cuda:0'), 'train_loss_giou': tensor(0.0441, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4685, device='cuda:0'), 'validation_loss_ce': tensor(0.3104, device='cuda:0'), 'validation_loss_bbox': tensor(0.0063, device='cuda:0'), 'validation_loss_giou': tensor(0.0632, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4685, device='cuda:0'), 'train_loss_ce': tensor(0.3104, device='cuda:0'), 'train_loss_bbox': tensor(0.0063, device='cuda:0'), 'train_loss_giou': tensor(0.0632, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4206, device='cuda:0'), 'validation_loss_ce': tensor(0.3092, device='cuda:0'), 'validation_loss_bbox': tensor(0.0045, device='cuda:0'), 'validation_loss_giou': tensor(0.0444, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4206, device='cuda:0'), 'train_loss_ce': tensor(0.3092, device='cuda:0'), 'train_loss_bbox': tensor(0.0045, device='cuda:0'), 'train_loss_giou': tensor(0.0444, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3999, device='cuda:0'), 'validation_loss_ce': tensor(0.3082, device='cuda:0'), 'validation_loss_bbox': tensor(0.0039, device='cuda:0'), 'validation_loss_giou': tensor(0.0362, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3999, device='cuda:0'), 'train_loss_ce': tensor(0.3082, device='cuda:0'), 'train_loss_bbox': tensor(0.0039, device='cuda:0'), 'train_loss_giou': tensor(0.0362, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4333, device='cuda:0'), 'validation_loss_ce': tensor(0.3074, device='cuda:0'), 'validation_loss_bbox': tensor(0.0054, device='cuda:0'), 'validation_loss_giou': tensor(0.0495, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4333, device='cuda:0'), 'train_loss_ce': tensor(0.3074, device='cuda:0'), 'train_loss_bbox': tensor(0.0054, device='cuda:0'), 'train_loss_giou': tensor(0.0495, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4274, device='cuda:0'), 'validation_loss_ce': tensor(0.3062, device='cuda:0'), 'validation_loss_bbox': tensor(0.0058, device='cuda:0'), 'validation_loss_giou': tensor(0.0461, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4274, device='cuda:0'), 'train_loss_ce': tensor(0.3062, device='cuda:0'), 'train_loss_bbox': tensor(0.0058, device='cuda:0'), 'train_loss_giou': tensor(0.0461, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4432, device='cuda:0'), 'validation_loss_ce': tensor(0.3051, device='cuda:0'), 'validation_loss_bbox': tensor(0.0072, device='cuda:0'), 'validation_loss_giou': tensor(0.0509, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4432, device='cuda:0'), 'train_loss_ce': tensor(0.3051, device='cuda:0'), 'train_loss_bbox': tensor(0.0072, device='cuda:0'), 'train_loss_giou': tensor(0.0509, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3974, device='cuda:0'), 'validation_loss_ce': tensor(0.3041, device='cuda:0'), 'validation_loss_bbox': tensor(0.0053, device='cuda:0'), 'validation_loss_giou': tensor(0.0334, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3974, device='cuda:0'), 'train_loss_ce': tensor(0.3041, device='cuda:0'), 'train_loss_bbox': tensor(0.0053, device='cuda:0'), 'train_loss_giou': tensor(0.0334, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4090, device='cuda:0'), 'validation_loss_ce': tensor(0.3029, device='cuda:0'), 'validation_loss_bbox': tensor(0.0060, device='cuda:0'), 'validation_loss_giou': tensor(0.0379, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4090, device='cuda:0'), 'train_loss_ce': tensor(0.3029, device='cuda:0'), 'train_loss_bbox': tensor(0.0060, device='cuda:0'), 'train_loss_giou': tensor(0.0379, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3797, device='cuda:0'), 'validation_loss_ce': tensor(0.3014, device='cuda:0'), 'validation_loss_bbox': tensor(0.0038, device='cuda:0'), 'validation_loss_giou': tensor(0.0297, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3797, device='cuda:0'), 'train_loss_ce': tensor(0.3014, device='cuda:0'), 'train_loss_bbox': tensor(0.0038, device='cuda:0'), 'train_loss_giou': tensor(0.0297, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3533, device='cuda:0'), 'validation_loss_ce': tensor(0.2998, device='cuda:0'), 'validation_loss_bbox': tensor(0.0030, device='cuda:0'), 'validation_loss_giou': tensor(0.0192, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3533, device='cuda:0'), 'train_loss_ce': tensor(0.2998, device='cuda:0'), 'train_loss_bbox': tensor(0.0030, device='cuda:0'), 'train_loss_giou': tensor(0.0192, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4473, device='cuda:0'), 'validation_loss_ce': tensor(0.2985, device='cuda:0'), 'validation_loss_bbox': tensor(0.0073, device='cuda:0'), 'validation_loss_giou': tensor(0.0562, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4473, device='cuda:0'), 'train_loss_ce': tensor(0.2985, device='cuda:0'), 'train_loss_bbox': tensor(0.0073, device='cuda:0'), 'train_loss_giou': tensor(0.0562, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4429, device='cuda:0'), 'validation_loss_ce': tensor(0.2971, device='cuda:0'), 'validation_loss_bbox': tensor(0.0081, device='cuda:0'), 'validation_loss_giou': tensor(0.0525, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4429, device='cuda:0'), 'train_loss_ce': tensor(0.2971, device='cuda:0'), 'train_loss_bbox': tensor(0.0081, device='cuda:0'), 'train_loss_giou': tensor(0.0525, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4153, device='cuda:0'), 'validation_loss_ce': tensor(0.2957, device='cuda:0'), 'validation_loss_bbox': tensor(0.0059, device='cuda:0'), 'validation_loss_giou': tensor(0.0450, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4153, device='cuda:0'), 'train_loss_ce': tensor(0.2957, device='cuda:0'), 'train_loss_bbox': tensor(0.0059, device='cuda:0'), 'train_loss_giou': tensor(0.0450, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4302, device='cuda:0'), 'validation_loss_ce': tensor(0.2948, device='cuda:0'), 'validation_loss_bbox': tensor(0.0063, device='cuda:0'), 'validation_loss_giou': tensor(0.0519, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4302, device='cuda:0'), 'train_loss_ce': tensor(0.2948, device='cuda:0'), 'train_loss_bbox': tensor(0.0063, device='cuda:0'), 'train_loss_giou': tensor(0.0519, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4024, device='cuda:0'), 'validation_loss_ce': tensor(0.2941, device='cuda:0'), 'validation_loss_bbox': tensor(0.0060, device='cuda:0'), 'validation_loss_giou': tensor(0.0391, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4024, device='cuda:0'), 'train_loss_ce': tensor(0.2941, device='cuda:0'), 'train_loss_bbox': tensor(0.0060, device='cuda:0'), 'train_loss_giou': tensor(0.0391, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4266, device='cuda:0'), 'validation_loss_ce': tensor(0.2929, device='cuda:0'), 'validation_loss_bbox': tensor(0.0048, device='cuda:0'), 'validation_loss_giou': tensor(0.0548, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4266, device='cuda:0'), 'train_loss_ce': tensor(0.2929, device='cuda:0'), 'train_loss_bbox': tensor(0.0048, device='cuda:0'), 'train_loss_giou': tensor(0.0548, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4652, device='cuda:0'), 'validation_loss_ce': tensor(0.2913, device='cuda:0'), 'validation_loss_bbox': tensor(0.0054, device='cuda:0'), 'validation_loss_giou': tensor(0.0734, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4652, device='cuda:0'), 'train_loss_ce': tensor(0.2913, device='cuda:0'), 'train_loss_bbox': tensor(0.0054, device='cuda:0'), 'train_loss_giou': tensor(0.0734, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4440, device='cuda:0'), 'validation_loss_ce': tensor(0.2899, device='cuda:0'), 'validation_loss_bbox': tensor(0.0063, device='cuda:0'), 'validation_loss_giou': tensor(0.0613, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4440, device='cuda:0'), 'train_loss_ce': tensor(0.2899, device='cuda:0'), 'train_loss_bbox': tensor(0.0063, device='cuda:0'), 'train_loss_giou': tensor(0.0613, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3999, device='cuda:0'), 'validation_loss_ce': tensor(0.2889, device='cuda:0'), 'validation_loss_bbox': tensor(0.0057, device='cuda:0'), 'validation_loss_giou': tensor(0.0412, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3999, device='cuda:0'), 'train_loss_ce': tensor(0.2889, device='cuda:0'), 'train_loss_bbox': tensor(0.0057, device='cuda:0'), 'train_loss_giou': tensor(0.0412, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4217, device='cuda:0'), 'validation_loss_ce': tensor(0.2874, device='cuda:0'), 'validation_loss_bbox': tensor(0.0071, device='cuda:0'), 'validation_loss_giou': tensor(0.0495, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4217, device='cuda:0'), 'train_loss_ce': tensor(0.2874, device='cuda:0'), 'train_loss_bbox': tensor(0.0071, device='cuda:0'), 'train_loss_giou': tensor(0.0495, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3994, device='cuda:0'), 'validation_loss_ce': tensor(0.2858, device='cuda:0'), 'validation_loss_bbox': tensor(0.0056, device='cuda:0'), 'validation_loss_giou': tensor(0.0428, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3994, device='cuda:0'), 'train_loss_ce': tensor(0.2858, device='cuda:0'), 'train_loss_bbox': tensor(0.0056, device='cuda:0'), 'train_loss_giou': tensor(0.0428, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3585, device='cuda:0'), 'validation_loss_ce': tensor(0.2841, device='cuda:0'), 'validation_loss_bbox': tensor(0.0038, device='cuda:0'), 'validation_loss_giou': tensor(0.0278, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3585, device='cuda:0'), 'train_loss_ce': tensor(0.2841, device='cuda:0'), 'train_loss_bbox': tensor(0.0038, device='cuda:0'), 'train_loss_giou': tensor(0.0278, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3458, device='cuda:0'), 'validation_loss_ce': tensor(0.2824, device='cuda:0'), 'validation_loss_bbox': tensor(0.0042, device='cuda:0'), 'validation_loss_giou': tensor(0.0212, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3458, device='cuda:0'), 'train_loss_ce': tensor(0.2824, device='cuda:0'), 'train_loss_bbox': tensor(0.0042, device='cuda:0'), 'train_loss_giou': tensor(0.0212, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3684, device='cuda:0'), 'validation_loss_ce': tensor(0.2813, device='cuda:0'), 'validation_loss_bbox': tensor(0.0047, device='cuda:0'), 'validation_loss_giou': tensor(0.0318, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3684, device='cuda:0'), 'train_loss_ce': tensor(0.2813, device='cuda:0'), 'train_loss_bbox': tensor(0.0047, device='cuda:0'), 'train_loss_giou': tensor(0.0318, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3771, device='cuda:0'), 'validation_loss_ce': tensor(0.2798, device='cuda:0'), 'validation_loss_bbox': tensor(0.0057, device='cuda:0'), 'validation_loss_giou': tensor(0.0345, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3771, device='cuda:0'), 'train_loss_ce': tensor(0.2798, device='cuda:0'), 'train_loss_bbox': tensor(0.0057, device='cuda:0'), 'train_loss_giou': tensor(0.0345, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4019, device='cuda:0'), 'validation_loss_ce': tensor(0.2783, device='cuda:0'), 'validation_loss_bbox': tensor(0.0053, device='cuda:0'), 'validation_loss_giou': tensor(0.0486, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4019, device='cuda:0'), 'train_loss_ce': tensor(0.2783, device='cuda:0'), 'train_loss_bbox': tensor(0.0053, device='cuda:0'), 'train_loss_giou': tensor(0.0486, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3581, device='cuda:0'), 'validation_loss_ce': tensor(0.2770, device='cuda:0'), 'validation_loss_bbox': tensor(0.0040, device='cuda:0'), 'validation_loss_giou': tensor(0.0306, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3581, device='cuda:0'), 'train_loss_ce': tensor(0.2770, device='cuda:0'), 'train_loss_bbox': tensor(0.0040, device='cuda:0'), 'train_loss_giou': tensor(0.0306, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4301, device='cuda:0'), 'validation_loss_ce': tensor(0.2757, device='cuda:0'), 'validation_loss_bbox': tensor(0.0054, device='cuda:0'), 'validation_loss_giou': tensor(0.0636, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4301, device='cuda:0'), 'train_loss_ce': tensor(0.2757, device='cuda:0'), 'train_loss_bbox': tensor(0.0054, device='cuda:0'), 'train_loss_giou': tensor(0.0636, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4152, device='cuda:0'), 'validation_loss_ce': tensor(0.2741, device='cuda:0'), 'validation_loss_bbox': tensor(0.0048, device='cuda:0'), 'validation_loss_giou': tensor(0.0586, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4152, device='cuda:0'), 'train_loss_ce': tensor(0.2741, device='cuda:0'), 'train_loss_bbox': tensor(0.0048, device='cuda:0'), 'train_loss_giou': tensor(0.0586, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3867, device='cuda:0'), 'validation_loss_ce': tensor(0.2726, device='cuda:0'), 'validation_loss_bbox': tensor(0.0065, device='cuda:0'), 'validation_loss_giou': tensor(0.0409, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3867, device='cuda:0'), 'train_loss_ce': tensor(0.2726, device='cuda:0'), 'train_loss_bbox': tensor(0.0065, device='cuda:0'), 'train_loss_giou': tensor(0.0409, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4053, device='cuda:0'), 'validation_loss_ce': tensor(0.2714, device='cuda:0'), 'validation_loss_bbox': tensor(0.0066, device='cuda:0'), 'validation_loss_giou': tensor(0.0504, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4053, device='cuda:0'), 'train_loss_ce': tensor(0.2714, device='cuda:0'), 'train_loss_bbox': tensor(0.0066, device='cuda:0'), 'train_loss_giou': tensor(0.0504, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4142, device='cuda:0'), 'validation_loss_ce': tensor(0.2702, device='cuda:0'), 'validation_loss_bbox': tensor(0.0055, device='cuda:0'), 'validation_loss_giou': tensor(0.0583, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4142, device='cuda:0'), 'train_loss_ce': tensor(0.2702, device='cuda:0'), 'train_loss_bbox': tensor(0.0055, device='cuda:0'), 'train_loss_giou': tensor(0.0583, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3579, device='cuda:0'), 'validation_loss_ce': tensor(0.2691, device='cuda:0'), 'validation_loss_bbox': tensor(0.0039, device='cuda:0'), 'validation_loss_giou': tensor(0.0347, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3579, device='cuda:0'), 'train_loss_ce': tensor(0.2691, device='cuda:0'), 'train_loss_bbox': tensor(0.0039, device='cuda:0'), 'train_loss_giou': tensor(0.0347, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3694, device='cuda:0'), 'validation_loss_ce': tensor(0.2679, device='cuda:0'), 'validation_loss_bbox': tensor(0.0056, device='cuda:0'), 'validation_loss_giou': tensor(0.0367, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3694, device='cuda:0'), 'train_loss_ce': tensor(0.2679, device='cuda:0'), 'train_loss_bbox': tensor(0.0056, device='cuda:0'), 'train_loss_giou': tensor(0.0367, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3928, device='cuda:0'), 'validation_loss_ce': tensor(0.2665, device='cuda:0'), 'validation_loss_bbox': tensor(0.0054, device='cuda:0'), 'validation_loss_giou': tensor(0.0496, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3928, device='cuda:0'), 'train_loss_ce': tensor(0.2665, device='cuda:0'), 'train_loss_bbox': tensor(0.0054, device='cuda:0'), 'train_loss_giou': tensor(0.0496, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3568, device='cuda:0'), 'validation_loss_ce': tensor(0.2650, device='cuda:0'), 'validation_loss_bbox': tensor(0.0049, device='cuda:0'), 'validation_loss_giou': tensor(0.0336, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3568, device='cuda:0'), 'train_loss_ce': tensor(0.2650, device='cuda:0'), 'train_loss_bbox': tensor(0.0049, device='cuda:0'), 'train_loss_giou': tensor(0.0336, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3655, device='cuda:0'), 'validation_loss_ce': tensor(0.2632, device='cuda:0'), 'validation_loss_bbox': tensor(0.0052, device='cuda:0'), 'validation_loss_giou': tensor(0.0381, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3655, device='cuda:0'), 'train_loss_ce': tensor(0.2632, device='cuda:0'), 'train_loss_bbox': tensor(0.0052, device='cuda:0'), 'train_loss_giou': tensor(0.0381, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4024, device='cuda:0'), 'validation_loss_ce': tensor(0.2616, device='cuda:0'), 'validation_loss_bbox': tensor(0.0053, device='cuda:0'), 'validation_loss_giou': tensor(0.0572, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4024, device='cuda:0'), 'train_loss_ce': tensor(0.2616, device='cuda:0'), 'train_loss_bbox': tensor(0.0053, device='cuda:0'), 'train_loss_giou': tensor(0.0572, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3881, device='cuda:0'), 'validation_loss_ce': tensor(0.2605, device='cuda:0'), 'validation_loss_bbox': tensor(0.0054, device='cuda:0'), 'validation_loss_giou': tensor(0.0502, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3881, device='cuda:0'), 'train_loss_ce': tensor(0.2605, device='cuda:0'), 'train_loss_bbox': tensor(0.0054, device='cuda:0'), 'train_loss_giou': tensor(0.0502, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3261, device='cuda:0'), 'validation_loss_ce': tensor(0.2596, device='cuda:0'), 'validation_loss_bbox': tensor(0.0030, device='cuda:0'), 'validation_loss_giou': tensor(0.0257, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3261, device='cuda:0'), 'train_loss_ce': tensor(0.2596, device='cuda:0'), 'train_loss_bbox': tensor(0.0030, device='cuda:0'), 'train_loss_giou': tensor(0.0257, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3776, device='cuda:0'), 'validation_loss_ce': tensor(0.2589, device='cuda:0'), 'validation_loss_bbox': tensor(0.0047, device='cuda:0'), 'validation_loss_giou': tensor(0.0476, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3776, device='cuda:0'), 'train_loss_ce': tensor(0.2589, device='cuda:0'), 'train_loss_bbox': tensor(0.0047, device='cuda:0'), 'train_loss_giou': tensor(0.0476, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4269, device='cuda:0'), 'validation_loss_ce': tensor(0.2581, device='cuda:0'), 'validation_loss_bbox': tensor(0.0070, device='cuda:0'), 'validation_loss_giou': tensor(0.0669, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4269, device='cuda:0'), 'train_loss_ce': tensor(0.2581, device='cuda:0'), 'train_loss_bbox': tensor(0.0070, device='cuda:0'), 'train_loss_giou': tensor(0.0669, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4229, device='cuda:0'), 'validation_loss_ce': tensor(0.2568, device='cuda:0'), 'validation_loss_bbox': tensor(0.0072, device='cuda:0'), 'validation_loss_giou': tensor(0.0650, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4229, device='cuda:0'), 'train_loss_ce': tensor(0.2568, device='cuda:0'), 'train_loss_bbox': tensor(0.0072, device='cuda:0'), 'train_loss_giou': tensor(0.0650, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3840, device='cuda:0'), 'validation_loss_ce': tensor(0.2553, device='cuda:0'), 'validation_loss_bbox': tensor(0.0072, device='cuda:0'), 'validation_loss_giou': tensor(0.0464, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3840, device='cuda:0'), 'train_loss_ce': tensor(0.2553, device='cuda:0'), 'train_loss_bbox': tensor(0.0072, device='cuda:0'), 'train_loss_giou': tensor(0.0464, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3690, device='cuda:0'), 'validation_loss_ce': tensor(0.2535, device='cuda:0'), 'validation_loss_bbox': tensor(0.0044, device='cuda:0'), 'validation_loss_giou': tensor(0.0467, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3690, device='cuda:0'), 'train_loss_ce': tensor(0.2535, device='cuda:0'), 'train_loss_bbox': tensor(0.0044, device='cuda:0'), 'train_loss_giou': tensor(0.0467, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4179, device='cuda:0'), 'validation_loss_ce': tensor(0.2516, device='cuda:0'), 'validation_loss_bbox': tensor(0.0065, device='cuda:0'), 'validation_loss_giou': tensor(0.0670, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4179, device='cuda:0'), 'train_loss_ce': tensor(0.2516, device='cuda:0'), 'train_loss_bbox': tensor(0.0065, device='cuda:0'), 'train_loss_giou': tensor(0.0670, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3683, device='cuda:0'), 'validation_loss_ce': tensor(0.2502, device='cuda:0'), 'validation_loss_bbox': tensor(0.0054, device='cuda:0'), 'validation_loss_giou': tensor(0.0457, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3683, device='cuda:0'), 'train_loss_ce': tensor(0.2502, device='cuda:0'), 'train_loss_bbox': tensor(0.0054, device='cuda:0'), 'train_loss_giou': tensor(0.0457, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3372, device='cuda:0'), 'validation_loss_ce': tensor(0.2489, device='cuda:0'), 'validation_loss_bbox': tensor(0.0028, device='cuda:0'), 'validation_loss_giou': tensor(0.0372, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3372, device='cuda:0'), 'train_loss_ce': tensor(0.2489, device='cuda:0'), 'train_loss_bbox': tensor(0.0028, device='cuda:0'), 'train_loss_giou': tensor(0.0372, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4015, device='cuda:0'), 'validation_loss_ce': tensor(0.2481, device='cuda:0'), 'validation_loss_bbox': tensor(0.0063, device='cuda:0'), 'validation_loss_giou': tensor(0.0610, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4015, device='cuda:0'), 'train_loss_ce': tensor(0.2481, device='cuda:0'), 'train_loss_bbox': tensor(0.0063, device='cuda:0'), 'train_loss_giou': tensor(0.0610, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3654, device='cuda:0'), 'validation_loss_ce': tensor(0.2468, device='cuda:0'), 'validation_loss_bbox': tensor(0.0055, device='cuda:0'), 'validation_loss_giou': tensor(0.0457, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3654, device='cuda:0'), 'train_loss_ce': tensor(0.2468, device='cuda:0'), 'train_loss_bbox': tensor(0.0055, device='cuda:0'), 'train_loss_giou': tensor(0.0457, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3836, device='cuda:0'), 'validation_loss_ce': tensor(0.2453, device='cuda:0'), 'validation_loss_bbox': tensor(0.0043, device='cuda:0'), 'validation_loss_giou': tensor(0.0584, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3836, device='cuda:0'), 'train_loss_ce': tensor(0.2453, device='cuda:0'), 'train_loss_bbox': tensor(0.0043, device='cuda:0'), 'train_loss_giou': tensor(0.0584, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3906, device='cuda:0'), 'validation_loss_ce': tensor(0.2436, device='cuda:0'), 'validation_loss_bbox': tensor(0.0055, device='cuda:0'), 'validation_loss_giou': tensor(0.0597, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3906, device='cuda:0'), 'train_loss_ce': tensor(0.2436, device='cuda:0'), 'train_loss_bbox': tensor(0.0055, device='cuda:0'), 'train_loss_giou': tensor(0.0597, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3406, device='cuda:0'), 'validation_loss_ce': tensor(0.2419, device='cuda:0'), 'validation_loss_bbox': tensor(0.0045, device='cuda:0'), 'validation_loss_giou': tensor(0.0381, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3406, device='cuda:0'), 'train_loss_ce': tensor(0.2419, device='cuda:0'), 'train_loss_bbox': tensor(0.0045, device='cuda:0'), 'train_loss_giou': tensor(0.0381, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4259, device='cuda:0'), 'validation_loss_ce': tensor(0.2409, device='cuda:0'), 'validation_loss_bbox': tensor(0.0076, device='cuda:0'), 'validation_loss_giou': tensor(0.0736, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4259, device='cuda:0'), 'train_loss_ce': tensor(0.2409, device='cuda:0'), 'train_loss_bbox': tensor(0.0076, device='cuda:0'), 'train_loss_giou': tensor(0.0736, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4189, device='cuda:0'), 'validation_loss_ce': tensor(0.2396, device='cuda:0'), 'validation_loss_bbox': tensor(0.0064, device='cuda:0'), 'validation_loss_giou': tensor(0.0737, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4189, device='cuda:0'), 'train_loss_ce': tensor(0.2396, device='cuda:0'), 'train_loss_bbox': tensor(0.0064, device='cuda:0'), 'train_loss_giou': tensor(0.0737, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3400, device='cuda:0'), 'validation_loss_ce': tensor(0.2380, device='cuda:0'), 'validation_loss_bbox': tensor(0.0051, device='cuda:0'), 'validation_loss_giou': tensor(0.0382, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3400, device='cuda:0'), 'train_loss_ce': tensor(0.2380, device='cuda:0'), 'train_loss_bbox': tensor(0.0051, device='cuda:0'), 'train_loss_giou': tensor(0.0382, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4960, device='cuda:0'), 'validation_loss_ce': tensor(0.2369, device='cuda:0'), 'validation_loss_bbox': tensor(0.0114, device='cuda:0'), 'validation_loss_giou': tensor(0.1011, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4960, device='cuda:0'), 'train_loss_ce': tensor(0.2369, device='cuda:0'), 'train_loss_bbox': tensor(0.0114, device='cuda:0'), 'train_loss_giou': tensor(0.1011, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.5265, device='cuda:0'), 'validation_loss_ce': tensor(0.2358, device='cuda:0'), 'validation_loss_bbox': tensor(0.0120, device='cuda:0'), 'validation_loss_giou': tensor(0.1153, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.5265, device='cuda:0'), 'train_loss_ce': tensor(0.2358, device='cuda:0'), 'train_loss_bbox': tensor(0.0120, device='cuda:0'), 'train_loss_giou': tensor(0.1153, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3918, device='cuda:0'), 'validation_loss_ce': tensor(0.2347, device='cuda:0'), 'validation_loss_bbox': tensor(0.0067, device='cuda:0'), 'validation_loss_giou': tensor(0.0617, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3918, device='cuda:0'), 'train_loss_ce': tensor(0.2347, device='cuda:0'), 'train_loss_bbox': tensor(0.0067, device='cuda:0'), 'train_loss_giou': tensor(0.0617, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3887, device='cuda:0'), 'validation_loss_ce': tensor(0.2337, device='cuda:0'), 'validation_loss_bbox': tensor(0.0078, device='cuda:0'), 'validation_loss_giou': tensor(0.0580, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3887, device='cuda:0'), 'train_loss_ce': tensor(0.2337, device='cuda:0'), 'train_loss_bbox': tensor(0.0078, device='cuda:0'), 'train_loss_giou': tensor(0.0580, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4944, device='cuda:0'), 'validation_loss_ce': tensor(0.2321, device='cuda:0'), 'validation_loss_bbox': tensor(0.0109, device='cuda:0'), 'validation_loss_giou': tensor(0.1039, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4944, device='cuda:0'), 'train_loss_ce': tensor(0.2321, device='cuda:0'), 'train_loss_bbox': tensor(0.0109, device='cuda:0'), 'train_loss_giou': tensor(0.1039, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4368, device='cuda:0'), 'validation_loss_ce': tensor(0.2303, device='cuda:0'), 'validation_loss_bbox': tensor(0.0078, device='cuda:0'), 'validation_loss_giou': tensor(0.0838, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4368, device='cuda:0'), 'train_loss_ce': tensor(0.2303, device='cuda:0'), 'train_loss_bbox': tensor(0.0078, device='cuda:0'), 'train_loss_giou': tensor(0.0838, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3772, device='cuda:0'), 'validation_loss_ce': tensor(0.2288, device='cuda:0'), 'validation_loss_bbox': tensor(0.0063, device='cuda:0'), 'validation_loss_giou': tensor(0.0586, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3772, device='cuda:0'), 'train_loss_ce': tensor(0.2288, device='cuda:0'), 'train_loss_bbox': tensor(0.0063, device='cuda:0'), 'train_loss_giou': tensor(0.0586, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3994, device='cuda:0'), 'validation_loss_ce': tensor(0.2275, device='cuda:0'), 'validation_loss_bbox': tensor(0.0073, device='cuda:0'), 'validation_loss_giou': tensor(0.0678, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3994, device='cuda:0'), 'train_loss_ce': tensor(0.2275, device='cuda:0'), 'train_loss_bbox': tensor(0.0073, device='cuda:0'), 'train_loss_giou': tensor(0.0678, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3371, device='cuda:0'), 'validation_loss_ce': tensor(0.2260, device='cuda:0'), 'validation_loss_bbox': tensor(0.0046, device='cuda:0'), 'validation_loss_giou': tensor(0.0439, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3371, device='cuda:0'), 'train_loss_ce': tensor(0.2260, device='cuda:0'), 'train_loss_bbox': tensor(0.0046, device='cuda:0'), 'train_loss_giou': tensor(0.0439, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3244, device='cuda:0'), 'validation_loss_ce': tensor(0.2244, device='cuda:0'), 'validation_loss_bbox': tensor(0.0046, device='cuda:0'), 'validation_loss_giou': tensor(0.0386, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3244, device='cuda:0'), 'train_loss_ce': tensor(0.2244, device='cuda:0'), 'train_loss_bbox': tensor(0.0046, device='cuda:0'), 'train_loss_giou': tensor(0.0386, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3663, device='cuda:0'), 'validation_loss_ce': tensor(0.2227, device='cuda:0'), 'validation_loss_bbox': tensor(0.0048, device='cuda:0'), 'validation_loss_giou': tensor(0.0599, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3663, device='cuda:0'), 'train_loss_ce': tensor(0.2227, device='cuda:0'), 'train_loss_bbox': tensor(0.0048, device='cuda:0'), 'train_loss_giou': tensor(0.0599, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3038, device='cuda:0'), 'validation_loss_ce': tensor(0.2211, device='cuda:0'), 'validation_loss_bbox': tensor(0.0033, device='cuda:0'), 'validation_loss_giou': tensor(0.0331, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3038, device='cuda:0'), 'train_loss_ce': tensor(0.2211, device='cuda:0'), 'train_loss_bbox': tensor(0.0033, device='cuda:0'), 'train_loss_giou': tensor(0.0331, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3736, device='cuda:0'), 'validation_loss_ce': tensor(0.2199, device='cuda:0'), 'validation_loss_bbox': tensor(0.0069, device='cuda:0'), 'validation_loss_giou': tensor(0.0597, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3736, device='cuda:0'), 'train_loss_ce': tensor(0.2199, device='cuda:0'), 'train_loss_bbox': tensor(0.0069, device='cuda:0'), 'train_loss_giou': tensor(0.0597, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3623, device='cuda:0'), 'validation_loss_ce': tensor(0.2185, device='cuda:0'), 'validation_loss_bbox': tensor(0.0063, device='cuda:0'), 'validation_loss_giou': tensor(0.0562, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3623, device='cuda:0'), 'train_loss_ce': tensor(0.2185, device='cuda:0'), 'train_loss_bbox': tensor(0.0063, device='cuda:0'), 'train_loss_giou': tensor(0.0562, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3253, device='cuda:0'), 'validation_loss_ce': tensor(0.2168, device='cuda:0'), 'validation_loss_bbox': tensor(0.0043, device='cuda:0'), 'validation_loss_giou': tensor(0.0434, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3253, device='cuda:0'), 'train_loss_ce': tensor(0.2168, device='cuda:0'), 'train_loss_bbox': tensor(0.0043, device='cuda:0'), 'train_loss_giou': tensor(0.0434, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3179, device='cuda:0'), 'validation_loss_ce': tensor(0.2151, device='cuda:0'), 'validation_loss_bbox': tensor(0.0054, device='cuda:0'), 'validation_loss_giou': tensor(0.0379, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3179, device='cuda:0'), 'train_loss_ce': tensor(0.2151, device='cuda:0'), 'train_loss_bbox': tensor(0.0054, device='cuda:0'), 'train_loss_giou': tensor(0.0379, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3873, device='cuda:0'), 'validation_loss_ce': tensor(0.2133, device='cuda:0'), 'validation_loss_bbox': tensor(0.0081, device='cuda:0'), 'validation_loss_giou': tensor(0.0668, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3873, device='cuda:0'), 'train_loss_ce': tensor(0.2133, device='cuda:0'), 'train_loss_bbox': tensor(0.0081, device='cuda:0'), 'train_loss_giou': tensor(0.0668, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3520, device='cuda:0'), 'validation_loss_ce': tensor(0.2117, device='cuda:0'), 'validation_loss_bbox': tensor(0.0064, device='cuda:0'), 'validation_loss_giou': tensor(0.0543, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3520, device='cuda:0'), 'train_loss_ce': tensor(0.2117, device='cuda:0'), 'train_loss_bbox': tensor(0.0064, device='cuda:0'), 'train_loss_giou': tensor(0.0543, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3594, device='cuda:0'), 'validation_loss_ce': tensor(0.2103, device='cuda:0'), 'validation_loss_bbox': tensor(0.0058, device='cuda:0'), 'validation_loss_giou': tensor(0.0600, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3594, device='cuda:0'), 'train_loss_ce': tensor(0.2103, device='cuda:0'), 'train_loss_bbox': tensor(0.0058, device='cuda:0'), 'train_loss_giou': tensor(0.0600, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4005, device='cuda:0'), 'validation_loss_ce': tensor(0.2090, device='cuda:0'), 'validation_loss_bbox': tensor(0.0072, device='cuda:0'), 'validation_loss_giou': tensor(0.0779, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4005, device='cuda:0'), 'train_loss_ce': tensor(0.2090, device='cuda:0'), 'train_loss_bbox': tensor(0.0072, device='cuda:0'), 'train_loss_giou': tensor(0.0779, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3650, device='cuda:0'), 'validation_loss_ce': tensor(0.2066, device='cuda:0'), 'validation_loss_bbox': tensor(0.0065, device='cuda:0'), 'validation_loss_giou': tensor(0.0629, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3650, device='cuda:0'), 'train_loss_ce': tensor(0.2066, device='cuda:0'), 'train_loss_bbox': tensor(0.0065, device='cuda:0'), 'train_loss_giou': tensor(0.0629, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3101, device='cuda:0'), 'validation_loss_ce': tensor(0.2047, device='cuda:0'), 'validation_loss_bbox': tensor(0.0056, device='cuda:0'), 'validation_loss_giou': tensor(0.0386, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3101, device='cuda:0'), 'train_loss_ce': tensor(0.2047, device='cuda:0'), 'train_loss_bbox': tensor(0.0056, device='cuda:0'), 'train_loss_giou': tensor(0.0386, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3566, device='cuda:0'), 'validation_loss_ce': tensor(0.2036, device='cuda:0'), 'validation_loss_bbox': tensor(0.0067, device='cuda:0'), 'validation_loss_giou': tensor(0.0597, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3566, device='cuda:0'), 'train_loss_ce': tensor(0.2036, device='cuda:0'), 'train_loss_bbox': tensor(0.0067, device='cuda:0'), 'train_loss_giou': tensor(0.0597, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3795, device='cuda:0'), 'validation_loss_ce': tensor(0.2023, device='cuda:0'), 'validation_loss_bbox': tensor(0.0079, device='cuda:0'), 'validation_loss_giou': tensor(0.0689, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3795, device='cuda:0'), 'train_loss_ce': tensor(0.2023, device='cuda:0'), 'train_loss_bbox': tensor(0.0079, device='cuda:0'), 'train_loss_giou': tensor(0.0689, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3640, device='cuda:0'), 'validation_loss_ce': tensor(0.2011, device='cuda:0'), 'validation_loss_bbox': tensor(0.0076, device='cuda:0'), 'validation_loss_giou': tensor(0.0626, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3640, device='cuda:0'), 'train_loss_ce': tensor(0.2011, device='cuda:0'), 'train_loss_bbox': tensor(0.0076, device='cuda:0'), 'train_loss_giou': tensor(0.0626, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.2940, device='cuda:0'), 'validation_loss_ce': tensor(0.1998, device='cuda:0'), 'validation_loss_bbox': tensor(0.0054, device='cuda:0'), 'validation_loss_giou': tensor(0.0336, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.2940, device='cuda:0'), 'train_loss_ce': tensor(0.1998, device='cuda:0'), 'train_loss_bbox': tensor(0.0054, device='cuda:0'), 'train_loss_giou': tensor(0.0336, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4443, device='cuda:0'), 'validation_loss_ce': tensor(0.1987, device='cuda:0'), 'validation_loss_bbox': tensor(0.0083, device='cuda:0'), 'validation_loss_giou': tensor(0.1021, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4443, device='cuda:0'), 'train_loss_ce': tensor(0.1987, device='cuda:0'), 'train_loss_bbox': tensor(0.0083, device='cuda:0'), 'train_loss_giou': tensor(0.1021, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4852, device='cuda:0'), 'validation_loss_ce': tensor(0.1974, device='cuda:0'), 'validation_loss_bbox': tensor(0.0099, device='cuda:0'), 'validation_loss_giou': tensor(0.1192, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.4852, device='cuda:0'), 'train_loss_ce': tensor(0.1974, device='cuda:0'), 'train_loss_bbox': tensor(0.0099, device='cuda:0'), 'train_loss_giou': tensor(0.1192, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3820, device='cuda:0'), 'validation_loss_ce': tensor(0.1957, device='cuda:0'), 'validation_loss_bbox': tensor(0.0063, device='cuda:0'), 'validation_loss_giou': tensor(0.0774, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3820, device='cuda:0'), 'train_loss_ce': tensor(0.1957, device='cuda:0'), 'train_loss_bbox': tensor(0.0063, device='cuda:0'), 'train_loss_giou': tensor(0.0774, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.2904, device='cuda:0'), 'validation_loss_ce': tensor(0.1939, device='cuda:0'), 'validation_loss_bbox': tensor(0.0061, device='cuda:0'), 'validation_loss_giou': tensor(0.0331, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.2904, device='cuda:0'), 'train_loss_ce': tensor(0.1939, device='cuda:0'), 'train_loss_bbox': tensor(0.0061, device='cuda:0'), 'train_loss_giou': tensor(0.0331, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3496, device='cuda:0'), 'validation_loss_ce': tensor(0.1924, device='cuda:0'), 'validation_loss_bbox': tensor(0.0078, device='cuda:0'), 'validation_loss_giou': tensor(0.0590, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3496, device='cuda:0'), 'train_loss_ce': tensor(0.1924, device='cuda:0'), 'train_loss_bbox': tensor(0.0078, device='cuda:0'), 'train_loss_giou': tensor(0.0590, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3211, device='cuda:0'), 'validation_loss_ce': tensor(0.1906, device='cuda:0'), 'validation_loss_bbox': tensor(0.0070, device='cuda:0'), 'validation_loss_giou': tensor(0.0478, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3211, device='cuda:0'), 'train_loss_ce': tensor(0.1906, device='cuda:0'), 'train_loss_bbox': tensor(0.0070, device='cuda:0'), 'train_loss_giou': tensor(0.0478, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.2748, device='cuda:0'), 'validation_loss_ce': tensor(0.1891, device='cuda:0'), 'validation_loss_bbox': tensor(0.0043, device='cuda:0'), 'validation_loss_giou': tensor(0.0321, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.2748, device='cuda:0'), 'train_loss_ce': tensor(0.1891, device='cuda:0'), 'train_loss_bbox': tensor(0.0043, device='cuda:0'), 'train_loss_giou': tensor(0.0321, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.2585, device='cuda:0'), 'validation_loss_ce': tensor(0.1877, device='cuda:0'), 'validation_loss_bbox': tensor(0.0033, device='cuda:0'), 'validation_loss_giou': tensor(0.0272, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.2585, device='cuda:0'), 'train_loss_ce': tensor(0.1877, device='cuda:0'), 'train_loss_bbox': tensor(0.0033, device='cuda:0'), 'train_loss_giou': tensor(0.0272, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.2615, device='cuda:0'), 'validation_loss_ce': tensor(0.1868, device='cuda:0'), 'validation_loss_bbox': tensor(0.0048, device='cuda:0'), 'validation_loss_giou': tensor(0.0252, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.2615, device='cuda:0'), 'train_loss_ce': tensor(0.1868, device='cuda:0'), 'train_loss_bbox': tensor(0.0048, device='cuda:0'), 'train_loss_giou': tensor(0.0252, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.2647, device='cuda:0'), 'validation_loss_ce': tensor(0.1853, device='cuda:0'), 'validation_loss_bbox': tensor(0.0054, device='cuda:0'), 'validation_loss_giou': tensor(0.0263, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.2647, device='cuda:0'), 'train_loss_ce': tensor(0.1853, device='cuda:0'), 'train_loss_bbox': tensor(0.0054, device='cuda:0'), 'train_loss_giou': tensor(0.0263, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.2732, device='cuda:0'), 'validation_loss_ce': tensor(0.1832, device='cuda:0'), 'validation_loss_bbox': tensor(0.0046, device='cuda:0'), 'validation_loss_giou': tensor(0.0336, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.2732, device='cuda:0'), 'train_loss_ce': tensor(0.1832, device='cuda:0'), 'train_loss_bbox': tensor(0.0046, device='cuda:0'), 'train_loss_giou': tensor(0.0336, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.2604, device='cuda:0'), 'validation_loss_ce': tensor(0.1814, device='cuda:0'), 'validation_loss_bbox': tensor(0.0044, device='cuda:0'), 'validation_loss_giou': tensor(0.0285, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.2604, device='cuda:0'), 'train_loss_ce': tensor(0.1814, device='cuda:0'), 'train_loss_bbox': tensor(0.0044, device='cuda:0'), 'train_loss_giou': tensor(0.0285, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.2703, device='cuda:0'), 'validation_loss_ce': tensor(0.1799, device='cuda:0'), 'validation_loss_bbox': tensor(0.0051, device='cuda:0'), 'validation_loss_giou': tensor(0.0326, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.2703, device='cuda:0'), 'train_loss_ce': tensor(0.1799, device='cuda:0'), 'train_loss_bbox': tensor(0.0051, device='cuda:0'), 'train_loss_giou': tensor(0.0326, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.2654, device='cuda:0'), 'validation_loss_ce': tensor(0.1787, device='cuda:0'), 'validation_loss_bbox': tensor(0.0048, device='cuda:0'), 'validation_loss_giou': tensor(0.0315, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.2654, device='cuda:0'), 'train_loss_ce': tensor(0.1787, device='cuda:0'), 'train_loss_bbox': tensor(0.0048, device='cuda:0'), 'train_loss_giou': tensor(0.0315, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.2268, device='cuda:0'), 'validation_loss_ce': tensor(0.1771, device='cuda:0'), 'validation_loss_bbox': tensor(0.0034, device='cuda:0'), 'validation_loss_giou': tensor(0.0164, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.2268, device='cuda:0'), 'train_loss_ce': tensor(0.1771, device='cuda:0'), 'train_loss_bbox': tensor(0.0034, device='cuda:0'), 'train_loss_giou': tensor(0.0164, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.2521, device='cuda:0'), 'validation_loss_ce': tensor(0.1749, device='cuda:0'), 'validation_loss_bbox': tensor(0.0034, device='cuda:0'), 'validation_loss_giou': tensor(0.0300, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.2521, device='cuda:0'), 'train_loss_ce': tensor(0.1749, device='cuda:0'), 'train_loss_bbox': tensor(0.0034, device='cuda:0'), 'train_loss_giou': tensor(0.0300, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.2550, device='cuda:0'), 'validation_loss_ce': tensor(0.1734, device='cuda:0'), 'validation_loss_bbox': tensor(0.0053, device='cuda:0'), 'validation_loss_giou': tensor(0.0276, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.2550, device='cuda:0'), 'train_loss_ce': tensor(0.1734, device='cuda:0'), 'train_loss_bbox': tensor(0.0053, device='cuda:0'), 'train_loss_giou': tensor(0.0276, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.2990, device='cuda:0'), 'validation_loss_ce': tensor(0.1733, device='cuda:0'), 'validation_loss_bbox': tensor(0.0057, device='cuda:0'), 'validation_loss_giou': tensor(0.0486, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.2990, device='cuda:0'), 'train_loss_ce': tensor(0.1733, device='cuda:0'), 'train_loss_bbox': tensor(0.0057, device='cuda:0'), 'train_loss_giou': tensor(0.0486, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.2564, device='cuda:0'), 'validation_loss_ce': tensor(0.1716, device='cuda:0'), 'validation_loss_bbox': tensor(0.0050, device='cuda:0'), 'validation_loss_giou': tensor(0.0299, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.2564, device='cuda:0'), 'train_loss_ce': tensor(0.1716, device='cuda:0'), 'train_loss_bbox': tensor(0.0050, device='cuda:0'), 'train_loss_giou': tensor(0.0299, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3220, device='cuda:0'), 'validation_loss_ce': tensor(0.1685, device='cuda:0'), 'validation_loss_bbox': tensor(0.0054, device='cuda:0'), 'validation_loss_giou': tensor(0.0631, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3220, device='cuda:0'), 'train_loss_ce': tensor(0.1685, device='cuda:0'), 'train_loss_bbox': tensor(0.0054, device='cuda:0'), 'train_loss_giou': tensor(0.0631, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3351, device='cuda:0'), 'validation_loss_ce': tensor(0.1658, device='cuda:0'), 'validation_loss_bbox': tensor(0.0046, device='cuda:0'), 'validation_loss_giou': tensor(0.0732, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3351, device='cuda:0'), 'train_loss_ce': tensor(0.1658, device='cuda:0'), 'train_loss_bbox': tensor(0.0046, device='cuda:0'), 'train_loss_giou': tensor(0.0732, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3223, device='cuda:0'), 'validation_loss_ce': tensor(0.1642, device='cuda:0'), 'validation_loss_bbox': tensor(0.0047, device='cuda:0'), 'validation_loss_giou': tensor(0.0673, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.3223, device='cuda:0'), 'train_loss_ce': tensor(0.1642, device='cuda:0'), 'train_loss_bbox': tensor(0.0047, device='cuda:0'), 'train_loss_giou': tensor(0.0673, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.2442, device='cuda:0'), 'validation_loss_ce': tensor(0.1630, device='cuda:0'), 'validation_loss_bbox': tensor(0.0037, device='cuda:0'), 'validation_loss_giou': tensor(0.0313, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.2442, device='cuda:0'), 'train_loss_ce': tensor(0.1630, device='cuda:0'), 'train_loss_bbox': tensor(0.0037, device='cuda:0'), 'train_loss_giou': tensor(0.0313, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.2644, device='cuda:0'), 'validation_loss_ce': tensor(0.1618, device='cuda:0'), 'validation_loss_bbox': tensor(0.0037, device='cuda:0'), 'validation_loss_giou': tensor(0.0422, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.2644, device='cuda:0'), 'train_loss_ce': tensor(0.1618, device='cuda:0'), 'train_loss_bbox': tensor(0.0037, device='cuda:0'), 'train_loss_giou': tensor(0.0422, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.2532, device='cuda:0'), 'validation_loss_ce': tensor(0.1604, device='cuda:0'), 'validation_loss_bbox': tensor(0.0047, device='cuda:0'), 'validation_loss_giou': tensor(0.0346, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.2532, device='cuda:0'), 'train_loss_ce': tensor(0.1604, device='cuda:0'), 'train_loss_bbox': tensor(0.0047, device='cuda:0'), 'train_loss_giou': tensor(0.0346, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.2619, device='cuda:0'), 'validation_loss_ce': tensor(0.1590, device='cuda:0'), 'validation_loss_bbox': tensor(0.0049, device='cuda:0'), 'validation_loss_giou': tensor(0.0393, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.2619, device='cuda:0'), 'train_loss_ce': tensor(0.1590, device='cuda:0'), 'train_loss_bbox': tensor(0.0049, device='cuda:0'), 'train_loss_giou': tensor(0.0393, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.2812, device='cuda:0'), 'validation_loss_ce': tensor(0.1573, device='cuda:0'), 'validation_loss_bbox': tensor(0.0051, device='cuda:0'), 'validation_loss_giou': tensor(0.0492, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.2812, device='cuda:0'), 'train_loss_ce': tensor(0.1573, device='cuda:0'), 'train_loss_bbox': tensor(0.0051, device='cuda:0'), 'train_loss_giou': tensor(0.0492, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.2471, device='cuda:0'), 'validation_loss_ce': tensor(0.1553, device='cuda:0'), 'validation_loss_bbox': tensor(0.0043, device='cuda:0'), 'validation_loss_giou': tensor(0.0352, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.2471, device='cuda:0'), 'train_loss_ce': tensor(0.1553, device='cuda:0'), 'train_loss_bbox': tensor(0.0043, device='cuda:0'), 'train_loss_giou': tensor(0.0352, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.2768, device='cuda:0'), 'validation_loss_ce': tensor(0.1538, device='cuda:0'), 'validation_loss_bbox': tensor(0.0047, device='cuda:0'), 'validation_loss_giou': tensor(0.0496, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.2768, device='cuda:0'), 'train_loss_ce': tensor(0.1538, device='cuda:0'), 'train_loss_bbox': tensor(0.0047, device='cuda:0'), 'train_loss_giou': tensor(0.0496, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.2507, device='cuda:0'), 'validation_loss_ce': tensor(0.1530, device='cuda:0'), 'validation_loss_bbox': tensor(0.0049, device='cuda:0'), 'validation_loss_giou': tensor(0.0366, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.2507, device='cuda:0'), 'train_loss_ce': tensor(0.1530, device='cuda:0'), 'train_loss_bbox': tensor(0.0049, device='cuda:0'), 'train_loss_giou': tensor(0.0366, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.2696, device='cuda:0'), 'validation_loss_ce': tensor(0.1518, device='cuda:0'), 'validation_loss_bbox': tensor(0.0052, device='cuda:0'), 'validation_loss_giou': tensor(0.0458, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.2696, device='cuda:0'), 'train_loss_ce': tensor(0.1518, device='cuda:0'), 'train_loss_bbox': tensor(0.0052, device='cuda:0'), 'train_loss_giou': tensor(0.0458, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.2683, device='cuda:0'), 'validation_loss_ce': tensor(0.1503, device='cuda:0'), 'validation_loss_bbox': tensor(0.0057, device='cuda:0'), 'validation_loss_giou': tensor(0.0446, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.2683, device='cuda:0'), 'train_loss_ce': tensor(0.1503, device='cuda:0'), 'train_loss_bbox': tensor(0.0057, device='cuda:0'), 'train_loss_giou': tensor(0.0446, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.2487, device='cuda:0'), 'validation_loss_ce': tensor(0.1483, device='cuda:0'), 'validation_loss_bbox': tensor(0.0060, device='cuda:0'), 'validation_loss_giou': tensor(0.0351, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.2487, device='cuda:0'), 'train_loss_ce': tensor(0.1483, device='cuda:0'), 'train_loss_bbox': tensor(0.0060, device='cuda:0'), 'train_loss_giou': tensor(0.0351, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.2512, device='cuda:0'), 'validation_loss_ce': tensor(0.1459, device='cuda:0'), 'validation_loss_bbox': tensor(0.0064, device='cuda:0'), 'validation_loss_giou': tensor(0.0365, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(0.2512, device='cuda:0'), 'train_loss_ce': tensor(0.1459, device='cuda:0'), 'train_loss_bbox': tensor(0.0064, device='cuda:0'), 'train_loss_giou': tensor(0.0365, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.2583, device='cuda:0'), 'validation_loss_ce': tensor(0.1439, device='cuda:0'), 'validation_loss_bbox': tensor(0.0060, device='cuda:0'), 'validation_loss_giou': tensor(0.0423, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.2583, device='cuda:0'), 'train_loss_ce': tensor(0.1439, device='cuda:0'), 'train_loss_bbox': tensor(0.0060, device='cuda:0'), 'train_loss_giou': tensor(0.0423, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.2555, device='cuda:0'), 'validation_loss_ce': tensor(0.1424, device='cuda:0'), 'validation_loss_bbox': tensor(0.0051, device='cuda:0'), 'validation_loss_giou': tensor(0.0439, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.2555, device='cuda:0'), 'train_loss_ce': tensor(0.1424, device='cuda:0'), 'train_loss_bbox': tensor(0.0051, device='cuda:0'), 'train_loss_giou': tensor(0.0439, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.2234, device='cuda:0'), 'validation_loss_ce': tensor(0.1409, device='cuda:0'), 'validation_loss_bbox': tensor(0.0036, device='cuda:0'), 'validation_loss_giou': tensor(0.0322, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.2234, device='cuda:0'), 'train_loss_ce': tensor(0.1409, device='cuda:0'), 'train_loss_bbox': tensor(0.0036, device='cuda:0'), 'train_loss_giou': tensor(0.0322, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.2025, device='cuda:0'), 'validation_loss_ce': tensor(0.1396, device='cuda:0'), 'validation_loss_bbox': tensor(0.0033, device='cuda:0'), 'validation_loss_giou': tensor(0.0231, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.2025, device='cuda:0'), 'train_loss_ce': tensor(0.1396, device='cuda:0'), 'train_loss_bbox': tensor(0.0033, device='cuda:0'), 'train_loss_giou': tensor(0.0231, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.2653, device='cuda:0'), 'validation_loss_ce': tensor(0.1383, device='cuda:0'), 'validation_loss_bbox': tensor(0.0043, device='cuda:0'), 'validation_loss_giou': tensor(0.0527, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.2653, device='cuda:0'), 'train_loss_ce': tensor(0.1383, device='cuda:0'), 'train_loss_bbox': tensor(0.0043, device='cuda:0'), 'train_loss_giou': tensor(0.0527, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.2622, device='cuda:0'), 'validation_loss_ce': tensor(0.1370, device='cuda:0'), 'validation_loss_bbox': tensor(0.0053, device='cuda:0'), 'validation_loss_giou': tensor(0.0494, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.2622, device='cuda:0'), 'train_loss_ce': tensor(0.1370, device='cuda:0'), 'train_loss_bbox': tensor(0.0053, device='cuda:0'), 'train_loss_giou': tensor(0.0494, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.2316, device='cuda:0'), 'validation_loss_ce': tensor(0.1355, device='cuda:0'), 'validation_loss_bbox': tensor(0.0056, device='cuda:0'), 'validation_loss_giou': tensor(0.0340, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.2316, device='cuda:0'), 'train_loss_ce': tensor(0.1355, device='cuda:0'), 'train_loss_bbox': tensor(0.0056, device='cuda:0'), 'train_loss_giou': tensor(0.0340, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.3128, device='cuda:0'), 'validation_loss_ce': tensor(0.1339, device='cuda:0'), 'validation_loss_bbox': tensor(0.0075, device='cuda:0'), 'validation_loss_giou': tensor(0.0707, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.3128, device='cuda:0'), 'train_loss_ce': tensor(0.1339, device='cuda:0'), 'train_loss_bbox': tensor(0.0075, device='cuda:0'), 'train_loss_giou': tensor(0.0707, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.3186, device='cuda:0'), 'validation_loss_ce': tensor(0.1322, device='cuda:0'), 'validation_loss_bbox': tensor(0.0066, device='cuda:0'), 'validation_loss_giou': tensor(0.0767, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.3186, device='cuda:0'), 'train_loss_ce': tensor(0.1322, device='cuda:0'), 'train_loss_bbox': tensor(0.0066, device='cuda:0'), 'train_loss_giou': tensor(0.0767, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.2534, device='cuda:0'), 'validation_loss_ce': tensor(0.1312, device='cuda:0'), 'validation_loss_bbox': tensor(0.0048, device='cuda:0'), 'validation_loss_giou': tensor(0.0490, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.2534, device='cuda:0'), 'train_loss_ce': tensor(0.1312, device='cuda:0'), 'train_loss_bbox': tensor(0.0048, device='cuda:0'), 'train_loss_giou': tensor(0.0490, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.2030, device='cuda:0'), 'validation_loss_ce': tensor(0.1291, device='cuda:0'), 'validation_loss_bbox': tensor(0.0040, device='cuda:0'), 'validation_loss_giou': tensor(0.0270, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.2030, device='cuda:0'), 'train_loss_ce': tensor(0.1291, device='cuda:0'), 'train_loss_bbox': tensor(0.0040, device='cuda:0'), 'train_loss_giou': tensor(0.0270, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.2516, device='cuda:0'), 'validation_loss_ce': tensor(0.1270, device='cuda:0'), 'validation_loss_bbox': tensor(0.0048, device='cuda:0'), 'validation_loss_giou': tensor(0.0504, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.2516, device='cuda:0'), 'train_loss_ce': tensor(0.1270, device='cuda:0'), 'train_loss_bbox': tensor(0.0048, device='cuda:0'), 'train_loss_giou': tensor(0.0504, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.2556, device='cuda:0'), 'validation_loss_ce': tensor(0.1252, device='cuda:0'), 'validation_loss_bbox': tensor(0.0064, device='cuda:0'), 'validation_loss_giou': tensor(0.0492, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.2556, device='cuda:0'), 'train_loss_ce': tensor(0.1252, device='cuda:0'), 'train_loss_bbox': tensor(0.0064, device='cuda:0'), 'train_loss_giou': tensor(0.0492, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.2389, device='cuda:0'), 'validation_loss_ce': tensor(0.1241, device='cuda:0'), 'validation_loss_bbox': tensor(0.0059, device='cuda:0'), 'validation_loss_giou': tensor(0.0426, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.2389, device='cuda:0'), 'train_loss_ce': tensor(0.1241, device='cuda:0'), 'train_loss_bbox': tensor(0.0059, device='cuda:0'), 'train_loss_giou': tensor(0.0426, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.2394, device='cuda:0'), 'validation_loss_ce': tensor(0.1227, device='cuda:0'), 'validation_loss_bbox': tensor(0.0061, device='cuda:0'), 'validation_loss_giou': tensor(0.0432, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.2394, device='cuda:0'), 'train_loss_ce': tensor(0.1227, device='cuda:0'), 'train_loss_bbox': tensor(0.0061, device='cuda:0'), 'train_loss_giou': tensor(0.0432, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.2464, device='cuda:0'), 'validation_loss_ce': tensor(0.1210, device='cuda:0'), 'validation_loss_bbox': tensor(0.0059, device='cuda:0'), 'validation_loss_giou': tensor(0.0480, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.2464, device='cuda:0'), 'train_loss_ce': tensor(0.1210, device='cuda:0'), 'train_loss_bbox': tensor(0.0059, device='cuda:0'), 'train_loss_giou': tensor(0.0480, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.2313, device='cuda:0'), 'validation_loss_ce': tensor(0.1198, device='cuda:0'), 'validation_loss_bbox': tensor(0.0050, device='cuda:0'), 'validation_loss_giou': tensor(0.0432, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.2313, device='cuda:0'), 'train_loss_ce': tensor(0.1198, device='cuda:0'), 'train_loss_bbox': tensor(0.0050, device='cuda:0'), 'train_loss_giou': tensor(0.0432, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.2748, device='cuda:0'), 'validation_loss_ce': tensor(0.1190, device='cuda:0'), 'validation_loss_bbox': tensor(0.0079, device='cuda:0'), 'validation_loss_giou': tensor(0.0581, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.2748, device='cuda:0'), 'train_loss_ce': tensor(0.1190, device='cuda:0'), 'train_loss_bbox': tensor(0.0079, device='cuda:0'), 'train_loss_giou': tensor(0.0581, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.2519, device='cuda:0'), 'validation_loss_ce': tensor(0.1179, device='cuda:0'), 'validation_loss_bbox': tensor(0.0076, device='cuda:0'), 'validation_loss_giou': tensor(0.0480, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.2519, device='cuda:0'), 'train_loss_ce': tensor(0.1179, device='cuda:0'), 'train_loss_bbox': tensor(0.0076, device='cuda:0'), 'train_loss_giou': tensor(0.0480, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.2228, device='cuda:0'), 'validation_loss_ce': tensor(0.1161, device='cuda:0'), 'validation_loss_bbox': tensor(0.0062, device='cuda:0'), 'validation_loss_giou': tensor(0.0379, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.2228, device='cuda:0'), 'train_loss_ce': tensor(0.1161, device='cuda:0'), 'train_loss_bbox': tensor(0.0062, device='cuda:0'), 'train_loss_giou': tensor(0.0379, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.2016, device='cuda:0'), 'validation_loss_ce': tensor(0.1149, device='cuda:0'), 'validation_loss_bbox': tensor(0.0044, device='cuda:0'), 'validation_loss_giou': tensor(0.0323, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.2016, device='cuda:0'), 'train_loss_ce': tensor(0.1149, device='cuda:0'), 'train_loss_bbox': tensor(0.0044, device='cuda:0'), 'train_loss_giou': tensor(0.0323, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.1942, device='cuda:0'), 'validation_loss_ce': tensor(0.1138, device='cuda:0'), 'validation_loss_bbox': tensor(0.0030, device='cuda:0'), 'validation_loss_giou': tensor(0.0326, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.1942, device='cuda:0'), 'train_loss_ce': tensor(0.1138, device='cuda:0'), 'train_loss_bbox': tensor(0.0030, device='cuda:0'), 'train_loss_giou': tensor(0.0326, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.1876, device='cuda:0'), 'validation_loss_ce': tensor(0.1116, device='cuda:0'), 'validation_loss_bbox': tensor(0.0036, device='cuda:0'), 'validation_loss_giou': tensor(0.0290, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.1876, device='cuda:0'), 'train_loss_ce': tensor(0.1116, device='cuda:0'), 'train_loss_bbox': tensor(0.0036, device='cuda:0'), 'train_loss_giou': tensor(0.0290, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.1975, device='cuda:0'), 'validation_loss_ce': tensor(0.1107, device='cuda:0'), 'validation_loss_bbox': tensor(0.0044, device='cuda:0'), 'validation_loss_giou': tensor(0.0323, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.1975, device='cuda:0'), 'train_loss_ce': tensor(0.1107, device='cuda:0'), 'train_loss_bbox': tensor(0.0044, device='cuda:0'), 'train_loss_giou': tensor(0.0323, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.1931, device='cuda:0'), 'validation_loss_ce': tensor(0.1101, device='cuda:0'), 'validation_loss_bbox': tensor(0.0044, device='cuda:0'), 'validation_loss_giou': tensor(0.0306, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.1931, device='cuda:0'), 'train_loss_ce': tensor(0.1101, device='cuda:0'), 'train_loss_bbox': tensor(0.0044, device='cuda:0'), 'train_loss_giou': tensor(0.0306, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.1748, device='cuda:0'), 'validation_loss_ce': tensor(0.1089, device='cuda:0'), 'validation_loss_bbox': tensor(0.0039, device='cuda:0'), 'validation_loss_giou': tensor(0.0233, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.1748, device='cuda:0'), 'train_loss_ce': tensor(0.1089, device='cuda:0'), 'train_loss_bbox': tensor(0.0039, device='cuda:0'), 'train_loss_giou': tensor(0.0233, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.1657, device='cuda:0'), 'validation_loss_ce': tensor(0.1067, device='cuda:0'), 'validation_loss_bbox': tensor(0.0025, device='cuda:0'), 'validation_loss_giou': tensor(0.0231, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.1657, device='cuda:0'), 'train_loss_ce': tensor(0.1067, device='cuda:0'), 'train_loss_bbox': tensor(0.0025, device='cuda:0'), 'train_loss_giou': tensor(0.0231, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.1763, device='cuda:0'), 'validation_loss_ce': tensor(0.1050, device='cuda:0'), 'validation_loss_bbox': tensor(0.0040, device='cuda:0'), 'validation_loss_giou': tensor(0.0258, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.1763, device='cuda:0'), 'train_loss_ce': tensor(0.1050, device='cuda:0'), 'train_loss_bbox': tensor(0.0040, device='cuda:0'), 'train_loss_giou': tensor(0.0258, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.2391, device='cuda:0'), 'validation_loss_ce': tensor(0.1037, device='cuda:0'), 'validation_loss_bbox': tensor(0.0058, device='cuda:0'), 'validation_loss_giou': tensor(0.0532, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.2391, device='cuda:0'), 'train_loss_ce': tensor(0.1037, device='cuda:0'), 'train_loss_bbox': tensor(0.0058, device='cuda:0'), 'train_loss_giou': tensor(0.0532, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.2123, device='cuda:0'), 'validation_loss_ce': tensor(0.1028, device='cuda:0'), 'validation_loss_bbox': tensor(0.0053, device='cuda:0'), 'validation_loss_giou': tensor(0.0415, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.2123, device='cuda:0'), 'train_loss_ce': tensor(0.1028, device='cuda:0'), 'train_loss_bbox': tensor(0.0053, device='cuda:0'), 'train_loss_giou': tensor(0.0415, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.1701, device='cuda:0'), 'validation_loss_ce': tensor(0.1017, device='cuda:0'), 'validation_loss_bbox': tensor(0.0034, device='cuda:0'), 'validation_loss_giou': tensor(0.0256, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.1701, device='cuda:0'), 'train_loss_ce': tensor(0.1017, device='cuda:0'), 'train_loss_bbox': tensor(0.0034, device='cuda:0'), 'train_loss_giou': tensor(0.0256, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.1911, device='cuda:0'), 'validation_loss_ce': tensor(0.1006, device='cuda:0'), 'validation_loss_bbox': tensor(0.0046, device='cuda:0'), 'validation_loss_giou': tensor(0.0338, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.1911, device='cuda:0'), 'train_loss_ce': tensor(0.1006, device='cuda:0'), 'train_loss_bbox': tensor(0.0046, device='cuda:0'), 'train_loss_giou': tensor(0.0338, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.1820, device='cuda:0'), 'validation_loss_ce': tensor(0.0990, device='cuda:0'), 'validation_loss_bbox': tensor(0.0036, device='cuda:0'), 'validation_loss_giou': tensor(0.0326, device='cuda:0'), 'validation_cardinality_error': tensor(0.5000, device='cuda:0')}
-{'training_loss': tensor(0.1820, device='cuda:0'), 'train_loss_ce': tensor(0.0990, device='cuda:0'), 'train_loss_bbox': tensor(0.0036, device='cuda:0'), 'train_loss_giou': tensor(0.0326, device='cuda:0'), 'train_cardinality_error': tensor(0.5000, device='cuda:0'), 'validation_loss': tensor(0.2153, device='cuda:0'), 'validation_loss_ce': tensor(0.0981, device='cuda:0'), 'validation_loss_bbox': tensor(0.0038, device='cuda:0'), 'validation_loss_giou': tensor(0.0489, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.2153, device='cuda:0'), 'train_loss_ce': tensor(0.0981, device='cuda:0'), 'train_loss_bbox': tensor(0.0038, device='cuda:0'), 'train_loss_giou': tensor(0.0489, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1920, device='cuda:0'), 'validation_loss_ce': tensor(0.0971, device='cuda:0'), 'validation_loss_bbox': tensor(0.0032, device='cuda:0'), 'validation_loss_giou': tensor(0.0395, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1920, device='cuda:0'), 'train_loss_ce': tensor(0.0971, device='cuda:0'), 'train_loss_bbox': tensor(0.0032, device='cuda:0'), 'train_loss_giou': tensor(0.0395, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1752, device='cuda:0'), 'validation_loss_ce': tensor(0.0957, device='cuda:0'), 'validation_loss_bbox': tensor(0.0036, device='cuda:0'), 'validation_loss_giou': tensor(0.0307, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1752, device='cuda:0'), 'train_loss_ce': tensor(0.0957, device='cuda:0'), 'train_loss_bbox': tensor(0.0036, device='cuda:0'), 'train_loss_giou': tensor(0.0307, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1762, device='cuda:0'), 'validation_loss_ce': tensor(0.0947, device='cuda:0'), 'validation_loss_bbox': tensor(0.0035, device='cuda:0'), 'validation_loss_giou': tensor(0.0319, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1762, device='cuda:0'), 'train_loss_ce': tensor(0.0947, device='cuda:0'), 'train_loss_bbox': tensor(0.0035, device='cuda:0'), 'train_loss_giou': tensor(0.0319, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1624, device='cuda:0'), 'validation_loss_ce': tensor(0.0937, device='cuda:0'), 'validation_loss_bbox': tensor(0.0028, device='cuda:0'), 'validation_loss_giou': tensor(0.0274, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1624, device='cuda:0'), 'train_loss_ce': tensor(0.0937, device='cuda:0'), 'train_loss_bbox': tensor(0.0028, device='cuda:0'), 'train_loss_giou': tensor(0.0274, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1555, device='cuda:0'), 'validation_loss_ce': tensor(0.0925, device='cuda:0'), 'validation_loss_bbox': tensor(0.0022, device='cuda:0'), 'validation_loss_giou': tensor(0.0260, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1555, device='cuda:0'), 'train_loss_ce': tensor(0.0925, device='cuda:0'), 'train_loss_bbox': tensor(0.0022, device='cuda:0'), 'train_loss_giou': tensor(0.0260, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1715, device='cuda:0'), 'validation_loss_ce': tensor(0.0912, device='cuda:0'), 'validation_loss_bbox': tensor(0.0036, device='cuda:0'), 'validation_loss_giou': tensor(0.0310, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1715, device='cuda:0'), 'train_loss_ce': tensor(0.0912, device='cuda:0'), 'train_loss_bbox': tensor(0.0036, device='cuda:0'), 'train_loss_giou': tensor(0.0310, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1425, device='cuda:0'), 'validation_loss_ce': tensor(0.0900, device='cuda:0'), 'validation_loss_bbox': tensor(0.0017, device='cuda:0'), 'validation_loss_giou': tensor(0.0220, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1425, device='cuda:0'), 'train_loss_ce': tensor(0.0900, device='cuda:0'), 'train_loss_bbox': tensor(0.0017, device='cuda:0'), 'train_loss_giou': tensor(0.0220, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1540, device='cuda:0'), 'validation_loss_ce': tensor(0.0887, device='cuda:0'), 'validation_loss_bbox': tensor(0.0033, device='cuda:0'), 'validation_loss_giou': tensor(0.0245, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1540, device='cuda:0'), 'train_loss_ce': tensor(0.0887, device='cuda:0'), 'train_loss_bbox': tensor(0.0033, device='cuda:0'), 'train_loss_giou': tensor(0.0245, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1737, device='cuda:0'), 'validation_loss_ce': tensor(0.0874, device='cuda:0'), 'validation_loss_bbox': tensor(0.0045, device='cuda:0'), 'validation_loss_giou': tensor(0.0320, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1737, device='cuda:0'), 'train_loss_ce': tensor(0.0874, device='cuda:0'), 'train_loss_bbox': tensor(0.0045, device='cuda:0'), 'train_loss_giou': tensor(0.0320, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1747, device='cuda:0'), 'validation_loss_ce': tensor(0.0861, device='cuda:0'), 'validation_loss_bbox': tensor(0.0036, device='cuda:0'), 'validation_loss_giou': tensor(0.0352, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1747, device='cuda:0'), 'train_loss_ce': tensor(0.0861, device='cuda:0'), 'train_loss_bbox': tensor(0.0036, device='cuda:0'), 'train_loss_giou': tensor(0.0352, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1523, device='cuda:0'), 'validation_loss_ce': tensor(0.0854, device='cuda:0'), 'validation_loss_bbox': tensor(0.0030, device='cuda:0'), 'validation_loss_giou': tensor(0.0259, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1523, device='cuda:0'), 'train_loss_ce': tensor(0.0854, device='cuda:0'), 'train_loss_bbox': tensor(0.0030, device='cuda:0'), 'train_loss_giou': tensor(0.0259, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1906, device='cuda:0'), 'validation_loss_ce': tensor(0.0842, device='cuda:0'), 'validation_loss_bbox': tensor(0.0043, device='cuda:0'), 'validation_loss_giou': tensor(0.0426, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1906, device='cuda:0'), 'train_loss_ce': tensor(0.0842, device='cuda:0'), 'train_loss_bbox': tensor(0.0043, device='cuda:0'), 'train_loss_giou': tensor(0.0426, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1742, device='cuda:0'), 'validation_loss_ce': tensor(0.0829, device='cuda:0'), 'validation_loss_bbox': tensor(0.0033, device='cuda:0'), 'validation_loss_giou': tensor(0.0374, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1742, device='cuda:0'), 'train_loss_ce': tensor(0.0829, device='cuda:0'), 'train_loss_bbox': tensor(0.0033, device='cuda:0'), 'train_loss_giou': tensor(0.0374, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.2211, device='cuda:0'), 'validation_loss_ce': tensor(0.0816, device='cuda:0'), 'validation_loss_bbox': tensor(0.0061, device='cuda:0'), 'validation_loss_giou': tensor(0.0545, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.2211, device='cuda:0'), 'train_loss_ce': tensor(0.0816, device='cuda:0'), 'train_loss_bbox': tensor(0.0061, device='cuda:0'), 'train_loss_giou': tensor(0.0545, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.2272, device='cuda:0'), 'validation_loss_ce': tensor(0.0809, device='cuda:0'), 'validation_loss_bbox': tensor(0.0062, device='cuda:0'), 'validation_loss_giou': tensor(0.0577, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.2272, device='cuda:0'), 'train_loss_ce': tensor(0.0809, device='cuda:0'), 'train_loss_bbox': tensor(0.0062, device='cuda:0'), 'train_loss_giou': tensor(0.0577, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.2108, device='cuda:0'), 'validation_loss_ce': tensor(0.0799, device='cuda:0'), 'validation_loss_bbox': tensor(0.0066, device='cuda:0'), 'validation_loss_giou': tensor(0.0488, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.2108, device='cuda:0'), 'train_loss_ce': tensor(0.0799, device='cuda:0'), 'train_loss_bbox': tensor(0.0066, device='cuda:0'), 'train_loss_giou': tensor(0.0488, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.2418, device='cuda:0'), 'validation_loss_ce': tensor(0.0795, device='cuda:0'), 'validation_loss_bbox': tensor(0.0077, device='cuda:0'), 'validation_loss_giou': tensor(0.0620, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.2418, device='cuda:0'), 'train_loss_ce': tensor(0.0795, device='cuda:0'), 'train_loss_bbox': tensor(0.0077, device='cuda:0'), 'train_loss_giou': tensor(0.0620, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1713, device='cuda:0'), 'validation_loss_ce': tensor(0.0782, device='cuda:0'), 'validation_loss_bbox': tensor(0.0047, device='cuda:0'), 'validation_loss_giou': tensor(0.0348, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1713, device='cuda:0'), 'train_loss_ce': tensor(0.0782, device='cuda:0'), 'train_loss_bbox': tensor(0.0047, device='cuda:0'), 'train_loss_giou': tensor(0.0348, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1889, device='cuda:0'), 'validation_loss_ce': tensor(0.0771, device='cuda:0'), 'validation_loss_bbox': tensor(0.0044, device='cuda:0'), 'validation_loss_giou': tensor(0.0448, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1889, device='cuda:0'), 'train_loss_ce': tensor(0.0771, device='cuda:0'), 'train_loss_bbox': tensor(0.0044, device='cuda:0'), 'train_loss_giou': tensor(0.0448, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.2010, device='cuda:0'), 'validation_loss_ce': tensor(0.0762, device='cuda:0'), 'validation_loss_bbox': tensor(0.0062, device='cuda:0'), 'validation_loss_giou': tensor(0.0468, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.2010, device='cuda:0'), 'train_loss_ce': tensor(0.0762, device='cuda:0'), 'train_loss_bbox': tensor(0.0062, device='cuda:0'), 'train_loss_giou': tensor(0.0468, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.2380, device='cuda:0'), 'validation_loss_ce': tensor(0.0758, device='cuda:0'), 'validation_loss_bbox': tensor(0.0076, device='cuda:0'), 'validation_loss_giou': tensor(0.0621, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.2380, device='cuda:0'), 'train_loss_ce': tensor(0.0758, device='cuda:0'), 'train_loss_bbox': tensor(0.0076, device='cuda:0'), 'train_loss_giou': tensor(0.0621, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.2395, device='cuda:0'), 'validation_loss_ce': tensor(0.0751, device='cuda:0'), 'validation_loss_bbox': tensor(0.0073, device='cuda:0'), 'validation_loss_giou': tensor(0.0640, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.2395, device='cuda:0'), 'train_loss_ce': tensor(0.0751, device='cuda:0'), 'train_loss_bbox': tensor(0.0073, device='cuda:0'), 'train_loss_giou': tensor(0.0640, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1901, device='cuda:0'), 'validation_loss_ce': tensor(0.0744, device='cuda:0'), 'validation_loss_bbox': tensor(0.0056, device='cuda:0'), 'validation_loss_giou': tensor(0.0438, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1901, device='cuda:0'), 'train_loss_ce': tensor(0.0744, device='cuda:0'), 'train_loss_bbox': tensor(0.0056, device='cuda:0'), 'train_loss_giou': tensor(0.0438, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1685, device='cuda:0'), 'validation_loss_ce': tensor(0.0747, device='cuda:0'), 'validation_loss_bbox': tensor(0.0043, device='cuda:0'), 'validation_loss_giou': tensor(0.0363, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1685, device='cuda:0'), 'train_loss_ce': tensor(0.0747, device='cuda:0'), 'train_loss_bbox': tensor(0.0043, device='cuda:0'), 'train_loss_giou': tensor(0.0363, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.2369, device='cuda:0'), 'validation_loss_ce': tensor(0.0752, device='cuda:0'), 'validation_loss_bbox': tensor(0.0073, device='cuda:0'), 'validation_loss_giou': tensor(0.0625, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.2369, device='cuda:0'), 'train_loss_ce': tensor(0.0752, device='cuda:0'), 'train_loss_bbox': tensor(0.0073, device='cuda:0'), 'train_loss_giou': tensor(0.0625, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.2553, device='cuda:0'), 'validation_loss_ce': tensor(0.0749, device='cuda:0'), 'validation_loss_bbox': tensor(0.0094, device='cuda:0'), 'validation_loss_giou': tensor(0.0668, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.2553, device='cuda:0'), 'train_loss_ce': tensor(0.0749, device='cuda:0'), 'train_loss_bbox': tensor(0.0094, device='cuda:0'), 'train_loss_giou': tensor(0.0668, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.2394, device='cuda:0'), 'validation_loss_ce': tensor(0.0737, device='cuda:0'), 'validation_loss_bbox': tensor(0.0090, device='cuda:0'), 'validation_loss_giou': tensor(0.0604, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.2394, device='cuda:0'), 'train_loss_ce': tensor(0.0737, device='cuda:0'), 'train_loss_bbox': tensor(0.0090, device='cuda:0'), 'train_loss_giou': tensor(0.0604, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1787, device='cuda:0'), 'validation_loss_ce': tensor(0.0727, device='cuda:0'), 'validation_loss_bbox': tensor(0.0062, device='cuda:0'), 'validation_loss_giou': tensor(0.0375, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1787, device='cuda:0'), 'train_loss_ce': tensor(0.0727, device='cuda:0'), 'train_loss_bbox': tensor(0.0062, device='cuda:0'), 'train_loss_giou': tensor(0.0375, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1819, device='cuda:0'), 'validation_loss_ce': tensor(0.0716, device='cuda:0'), 'validation_loss_bbox': tensor(0.0047, device='cuda:0'), 'validation_loss_giou': tensor(0.0434, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1819, device='cuda:0'), 'train_loss_ce': tensor(0.0716, device='cuda:0'), 'train_loss_bbox': tensor(0.0047, device='cuda:0'), 'train_loss_giou': tensor(0.0434, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1459, device='cuda:0'), 'validation_loss_ce': tensor(0.0706, device='cuda:0'), 'validation_loss_bbox': tensor(0.0036, device='cuda:0'), 'validation_loss_giou': tensor(0.0288, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1459, device='cuda:0'), 'train_loss_ce': tensor(0.0706, device='cuda:0'), 'train_loss_bbox': tensor(0.0036, device='cuda:0'), 'train_loss_giou': tensor(0.0288, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1681, device='cuda:0'), 'validation_loss_ce': tensor(0.0699, device='cuda:0'), 'validation_loss_bbox': tensor(0.0051, device='cuda:0'), 'validation_loss_giou': tensor(0.0364, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1681, device='cuda:0'), 'train_loss_ce': tensor(0.0699, device='cuda:0'), 'train_loss_bbox': tensor(0.0051, device='cuda:0'), 'train_loss_giou': tensor(0.0364, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1715, device='cuda:0'), 'validation_loss_ce': tensor(0.0696, device='cuda:0'), 'validation_loss_bbox': tensor(0.0046, device='cuda:0'), 'validation_loss_giou': tensor(0.0395, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1715, device='cuda:0'), 'train_loss_ce': tensor(0.0696, device='cuda:0'), 'train_loss_bbox': tensor(0.0046, device='cuda:0'), 'train_loss_giou': tensor(0.0395, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1331, device='cuda:0'), 'validation_loss_ce': tensor(0.0688, device='cuda:0'), 'validation_loss_bbox': tensor(0.0039, device='cuda:0'), 'validation_loss_giou': tensor(0.0224, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1331, device='cuda:0'), 'train_loss_ce': tensor(0.0688, device='cuda:0'), 'train_loss_bbox': tensor(0.0039, device='cuda:0'), 'train_loss_giou': tensor(0.0224, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1515, device='cuda:0'), 'validation_loss_ce': tensor(0.0683, device='cuda:0'), 'validation_loss_bbox': tensor(0.0047, device='cuda:0'), 'validation_loss_giou': tensor(0.0298, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1515, device='cuda:0'), 'train_loss_ce': tensor(0.0683, device='cuda:0'), 'train_loss_bbox': tensor(0.0047, device='cuda:0'), 'train_loss_giou': tensor(0.0298, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1779, device='cuda:0'), 'validation_loss_ce': tensor(0.0679, device='cuda:0'), 'validation_loss_bbox': tensor(0.0063, device='cuda:0'), 'validation_loss_giou': tensor(0.0392, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1779, device='cuda:0'), 'train_loss_ce': tensor(0.0679, device='cuda:0'), 'train_loss_bbox': tensor(0.0063, device='cuda:0'), 'train_loss_giou': tensor(0.0392, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1967, device='cuda:0'), 'validation_loss_ce': tensor(0.0673, device='cuda:0'), 'validation_loss_bbox': tensor(0.0068, device='cuda:0'), 'validation_loss_giou': tensor(0.0476, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1967, device='cuda:0'), 'train_loss_ce': tensor(0.0673, device='cuda:0'), 'train_loss_bbox': tensor(0.0068, device='cuda:0'), 'train_loss_giou': tensor(0.0476, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1919, device='cuda:0'), 'validation_loss_ce': tensor(0.0657, device='cuda:0'), 'validation_loss_bbox': tensor(0.0049, device='cuda:0'), 'validation_loss_giou': tensor(0.0510, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1919, device='cuda:0'), 'train_loss_ce': tensor(0.0657, device='cuda:0'), 'train_loss_bbox': tensor(0.0049, device='cuda:0'), 'train_loss_giou': tensor(0.0510, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1943, device='cuda:0'), 'validation_loss_ce': tensor(0.0644, device='cuda:0'), 'validation_loss_bbox': tensor(0.0062, device='cuda:0'), 'validation_loss_giou': tensor(0.0493, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1943, device='cuda:0'), 'train_loss_ce': tensor(0.0644, device='cuda:0'), 'train_loss_bbox': tensor(0.0062, device='cuda:0'), 'train_loss_giou': tensor(0.0493, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.2108, device='cuda:0'), 'validation_loss_ce': tensor(0.0635, device='cuda:0'), 'validation_loss_bbox': tensor(0.0078, device='cuda:0'), 'validation_loss_giou': tensor(0.0540, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.2108, device='cuda:0'), 'train_loss_ce': tensor(0.0635, device='cuda:0'), 'train_loss_bbox': tensor(0.0078, device='cuda:0'), 'train_loss_giou': tensor(0.0540, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1728, device='cuda:0'), 'validation_loss_ce': tensor(0.0626, device='cuda:0'), 'validation_loss_bbox': tensor(0.0060, device='cuda:0'), 'validation_loss_giou': tensor(0.0402, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1728, device='cuda:0'), 'train_loss_ce': tensor(0.0626, device='cuda:0'), 'train_loss_bbox': tensor(0.0060, device='cuda:0'), 'train_loss_giou': tensor(0.0402, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1853, device='cuda:0'), 'validation_loss_ce': tensor(0.0619, device='cuda:0'), 'validation_loss_bbox': tensor(0.0053, device='cuda:0'), 'validation_loss_giou': tensor(0.0484, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1853, device='cuda:0'), 'train_loss_ce': tensor(0.0619, device='cuda:0'), 'train_loss_bbox': tensor(0.0053, device='cuda:0'), 'train_loss_giou': tensor(0.0484, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1878, device='cuda:0'), 'validation_loss_ce': tensor(0.0608, device='cuda:0'), 'validation_loss_bbox': tensor(0.0047, device='cuda:0'), 'validation_loss_giou': tensor(0.0516, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1878, device='cuda:0'), 'train_loss_ce': tensor(0.0608, device='cuda:0'), 'train_loss_bbox': tensor(0.0047, device='cuda:0'), 'train_loss_giou': tensor(0.0516, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1642, device='cuda:0'), 'validation_loss_ce': tensor(0.0600, device='cuda:0'), 'validation_loss_bbox': tensor(0.0049, device='cuda:0'), 'validation_loss_giou': tensor(0.0400, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1642, device='cuda:0'), 'train_loss_ce': tensor(0.0600, device='cuda:0'), 'train_loss_bbox': tensor(0.0049, device='cuda:0'), 'train_loss_giou': tensor(0.0400, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1760, device='cuda:0'), 'validation_loss_ce': tensor(0.0595, device='cuda:0'), 'validation_loss_bbox': tensor(0.0053, device='cuda:0'), 'validation_loss_giou': tensor(0.0449, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1760, device='cuda:0'), 'train_loss_ce': tensor(0.0595, device='cuda:0'), 'train_loss_bbox': tensor(0.0053, device='cuda:0'), 'train_loss_giou': tensor(0.0449, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1807, device='cuda:0'), 'validation_loss_ce': tensor(0.0588, device='cuda:0'), 'validation_loss_bbox': tensor(0.0039, device='cuda:0'), 'validation_loss_giou': tensor(0.0511, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1807, device='cuda:0'), 'train_loss_ce': tensor(0.0588, device='cuda:0'), 'train_loss_bbox': tensor(0.0039, device='cuda:0'), 'train_loss_giou': tensor(0.0511, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1372, device='cuda:0'), 'validation_loss_ce': tensor(0.0580, device='cuda:0'), 'validation_loss_bbox': tensor(0.0047, device='cuda:0'), 'validation_loss_giou': tensor(0.0279, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1372, device='cuda:0'), 'train_loss_ce': tensor(0.0580, device='cuda:0'), 'train_loss_bbox': tensor(0.0047, device='cuda:0'), 'train_loss_giou': tensor(0.0279, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1723, device='cuda:0'), 'validation_loss_ce': tensor(0.0573, device='cuda:0'), 'validation_loss_bbox': tensor(0.0053, device='cuda:0'), 'validation_loss_giou': tensor(0.0441, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1723, device='cuda:0'), 'train_loss_ce': tensor(0.0573, device='cuda:0'), 'train_loss_bbox': tensor(0.0053, device='cuda:0'), 'train_loss_giou': tensor(0.0441, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1298, device='cuda:0'), 'validation_loss_ce': tensor(0.0565, device='cuda:0'), 'validation_loss_bbox': tensor(0.0032, device='cuda:0'), 'validation_loss_giou': tensor(0.0286, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1298, device='cuda:0'), 'train_loss_ce': tensor(0.0565, device='cuda:0'), 'train_loss_bbox': tensor(0.0032, device='cuda:0'), 'train_loss_giou': tensor(0.0286, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1863, device='cuda:0'), 'validation_loss_ce': tensor(0.0560, device='cuda:0'), 'validation_loss_bbox': tensor(0.0051, device='cuda:0'), 'validation_loss_giou': tensor(0.0523, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1863, device='cuda:0'), 'train_loss_ce': tensor(0.0560, device='cuda:0'), 'train_loss_bbox': tensor(0.0051, device='cuda:0'), 'train_loss_giou': tensor(0.0523, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.2272, device='cuda:0'), 'validation_loss_ce': tensor(0.0554, device='cuda:0'), 'validation_loss_bbox': tensor(0.0066, device='cuda:0'), 'validation_loss_giou': tensor(0.0693, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.2272, device='cuda:0'), 'train_loss_ce': tensor(0.0554, device='cuda:0'), 'train_loss_bbox': tensor(0.0066, device='cuda:0'), 'train_loss_giou': tensor(0.0693, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1638, device='cuda:0'), 'validation_loss_ce': tensor(0.0544, device='cuda:0'), 'validation_loss_bbox': tensor(0.0049, device='cuda:0'), 'validation_loss_giou': tensor(0.0424, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1638, device='cuda:0'), 'train_loss_ce': tensor(0.0544, device='cuda:0'), 'train_loss_bbox': tensor(0.0049, device='cuda:0'), 'train_loss_giou': tensor(0.0424, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1463, device='cuda:0'), 'validation_loss_ce': tensor(0.0536, device='cuda:0'), 'validation_loss_bbox': tensor(0.0037, device='cuda:0'), 'validation_loss_giou': tensor(0.0372, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1463, device='cuda:0'), 'train_loss_ce': tensor(0.0536, device='cuda:0'), 'train_loss_bbox': tensor(0.0037, device='cuda:0'), 'train_loss_giou': tensor(0.0372, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1664, device='cuda:0'), 'validation_loss_ce': tensor(0.0530, device='cuda:0'), 'validation_loss_bbox': tensor(0.0054, device='cuda:0'), 'validation_loss_giou': tensor(0.0433, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1664, device='cuda:0'), 'train_loss_ce': tensor(0.0530, device='cuda:0'), 'train_loss_bbox': tensor(0.0054, device='cuda:0'), 'train_loss_giou': tensor(0.0433, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1509, device='cuda:0'), 'validation_loss_ce': tensor(0.0524, device='cuda:0'), 'validation_loss_bbox': tensor(0.0043, device='cuda:0'), 'validation_loss_giou': tensor(0.0385, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1509, device='cuda:0'), 'train_loss_ce': tensor(0.0524, device='cuda:0'), 'train_loss_bbox': tensor(0.0043, device='cuda:0'), 'train_loss_giou': tensor(0.0385, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1813, device='cuda:0'), 'validation_loss_ce': tensor(0.0518, device='cuda:0'), 'validation_loss_bbox': tensor(0.0048, device='cuda:0'), 'validation_loss_giou': tensor(0.0529, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1813, device='cuda:0'), 'train_loss_ce': tensor(0.0518, device='cuda:0'), 'train_loss_bbox': tensor(0.0048, device='cuda:0'), 'train_loss_giou': tensor(0.0529, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.2015, device='cuda:0'), 'validation_loss_ce': tensor(0.0513, device='cuda:0'), 'validation_loss_bbox': tensor(0.0060, device='cuda:0'), 'validation_loss_giou': tensor(0.0602, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.2015, device='cuda:0'), 'train_loss_ce': tensor(0.0513, device='cuda:0'), 'train_loss_bbox': tensor(0.0060, device='cuda:0'), 'train_loss_giou': tensor(0.0602, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1461, device='cuda:0'), 'validation_loss_ce': tensor(0.0510, device='cuda:0'), 'validation_loss_bbox': tensor(0.0046, device='cuda:0'), 'validation_loss_giou': tensor(0.0360, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1461, device='cuda:0'), 'train_loss_ce': tensor(0.0510, device='cuda:0'), 'train_loss_bbox': tensor(0.0046, device='cuda:0'), 'train_loss_giou': tensor(0.0360, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1633, device='cuda:0'), 'validation_loss_ce': tensor(0.0506, device='cuda:0'), 'validation_loss_bbox': tensor(0.0065, device='cuda:0'), 'validation_loss_giou': tensor(0.0400, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1633, device='cuda:0'), 'train_loss_ce': tensor(0.0506, device='cuda:0'), 'train_loss_bbox': tensor(0.0065, device='cuda:0'), 'train_loss_giou': tensor(0.0400, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.2247, device='cuda:0'), 'validation_loss_ce': tensor(0.0508, device='cuda:0'), 'validation_loss_bbox': tensor(0.0084, device='cuda:0'), 'validation_loss_giou': tensor(0.0660, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.2247, device='cuda:0'), 'train_loss_ce': tensor(0.0508, device='cuda:0'), 'train_loss_bbox': tensor(0.0084, device='cuda:0'), 'train_loss_giou': tensor(0.0660, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1867, device='cuda:0'), 'validation_loss_ce': tensor(0.0520, device='cuda:0'), 'validation_loss_bbox': tensor(0.0074, device='cuda:0'), 'validation_loss_giou': tensor(0.0489, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1867, device='cuda:0'), 'train_loss_ce': tensor(0.0520, device='cuda:0'), 'train_loss_bbox': tensor(0.0074, device='cuda:0'), 'train_loss_giou': tensor(0.0489, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1368, device='cuda:0'), 'validation_loss_ce': tensor(0.0526, device='cuda:0'), 'validation_loss_bbox': tensor(0.0047, device='cuda:0'), 'validation_loss_giou': tensor(0.0303, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1368, device='cuda:0'), 'train_loss_ce': tensor(0.0526, device='cuda:0'), 'train_loss_bbox': tensor(0.0047, device='cuda:0'), 'train_loss_giou': tensor(0.0303, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1527, device='cuda:0'), 'validation_loss_ce': tensor(0.0491, device='cuda:0'), 'validation_loss_bbox': tensor(0.0052, device='cuda:0'), 'validation_loss_giou': tensor(0.0388, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1527, device='cuda:0'), 'train_loss_ce': tensor(0.0491, device='cuda:0'), 'train_loss_bbox': tensor(0.0052, device='cuda:0'), 'train_loss_giou': tensor(0.0388, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1806, device='cuda:0'), 'validation_loss_ce': tensor(0.0475, device='cuda:0'), 'validation_loss_bbox': tensor(0.0078, device='cuda:0'), 'validation_loss_giou': tensor(0.0470, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1806, device='cuda:0'), 'train_loss_ce': tensor(0.0475, device='cuda:0'), 'train_loss_bbox': tensor(0.0078, device='cuda:0'), 'train_loss_giou': tensor(0.0470, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1694, device='cuda:0'), 'validation_loss_ce': tensor(0.0472, device='cuda:0'), 'validation_loss_bbox': tensor(0.0067, device='cuda:0'), 'validation_loss_giou': tensor(0.0443, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1694, device='cuda:0'), 'train_loss_ce': tensor(0.0472, device='cuda:0'), 'train_loss_bbox': tensor(0.0067, device='cuda:0'), 'train_loss_giou': tensor(0.0443, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1294, device='cuda:0'), 'validation_loss_ce': tensor(0.0467, device='cuda:0'), 'validation_loss_bbox': tensor(0.0039, device='cuda:0'), 'validation_loss_giou': tensor(0.0315, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1294, device='cuda:0'), 'train_loss_ce': tensor(0.0467, device='cuda:0'), 'train_loss_bbox': tensor(0.0039, device='cuda:0'), 'train_loss_giou': tensor(0.0315, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1349, device='cuda:0'), 'validation_loss_ce': tensor(0.0466, device='cuda:0'), 'validation_loss_bbox': tensor(0.0049, device='cuda:0'), 'validation_loss_giou': tensor(0.0319, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1349, device='cuda:0'), 'train_loss_ce': tensor(0.0466, device='cuda:0'), 'train_loss_bbox': tensor(0.0049, device='cuda:0'), 'train_loss_giou': tensor(0.0319, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1814, device='cuda:0'), 'validation_loss_ce': tensor(0.0464, device='cuda:0'), 'validation_loss_bbox': tensor(0.0072, device='cuda:0'), 'validation_loss_giou': tensor(0.0494, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1814, device='cuda:0'), 'train_loss_ce': tensor(0.0464, device='cuda:0'), 'train_loss_bbox': tensor(0.0072, device='cuda:0'), 'train_loss_giou': tensor(0.0494, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1913, device='cuda:0'), 'validation_loss_ce': tensor(0.0464, device='cuda:0'), 'validation_loss_bbox': tensor(0.0067, device='cuda:0'), 'validation_loss_giou': tensor(0.0558, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1913, device='cuda:0'), 'train_loss_ce': tensor(0.0464, device='cuda:0'), 'train_loss_bbox': tensor(0.0067, device='cuda:0'), 'train_loss_giou': tensor(0.0558, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1542, device='cuda:0'), 'validation_loss_ce': tensor(0.0465, device='cuda:0'), 'validation_loss_bbox': tensor(0.0049, device='cuda:0'), 'validation_loss_giou': tensor(0.0416, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1542, device='cuda:0'), 'train_loss_ce': tensor(0.0465, device='cuda:0'), 'train_loss_bbox': tensor(0.0049, device='cuda:0'), 'train_loss_giou': tensor(0.0416, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1515, device='cuda:0'), 'validation_loss_ce': tensor(0.0469, device='cuda:0'), 'validation_loss_bbox': tensor(0.0054, device='cuda:0'), 'validation_loss_giou': tensor(0.0388, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1515, device='cuda:0'), 'train_loss_ce': tensor(0.0469, device='cuda:0'), 'train_loss_bbox': tensor(0.0054, device='cuda:0'), 'train_loss_giou': tensor(0.0388, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1680, device='cuda:0'), 'validation_loss_ce': tensor(0.0471, device='cuda:0'), 'validation_loss_bbox': tensor(0.0049, device='cuda:0'), 'validation_loss_giou': tensor(0.0481, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1680, device='cuda:0'), 'train_loss_ce': tensor(0.0471, device='cuda:0'), 'train_loss_bbox': tensor(0.0049, device='cuda:0'), 'train_loss_giou': tensor(0.0481, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1445, device='cuda:0'), 'validation_loss_ce': tensor(0.0475, device='cuda:0'), 'validation_loss_bbox': tensor(0.0049, device='cuda:0'), 'validation_loss_giou': tensor(0.0361, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1445, device='cuda:0'), 'train_loss_ce': tensor(0.0475, device='cuda:0'), 'train_loss_bbox': tensor(0.0049, device='cuda:0'), 'train_loss_giou': tensor(0.0361, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1210, device='cuda:0'), 'validation_loss_ce': tensor(0.0470, device='cuda:0'), 'validation_loss_bbox': tensor(0.0035, device='cuda:0'), 'validation_loss_giou': tensor(0.0282, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1210, device='cuda:0'), 'train_loss_ce': tensor(0.0470, device='cuda:0'), 'train_loss_bbox': tensor(0.0035, device='cuda:0'), 'train_loss_giou': tensor(0.0282, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1835, device='cuda:0'), 'validation_loss_ce': tensor(0.0455, device='cuda:0'), 'validation_loss_bbox': tensor(0.0056, device='cuda:0'), 'validation_loss_giou': tensor(0.0549, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1835, device='cuda:0'), 'train_loss_ce': tensor(0.0455, device='cuda:0'), 'train_loss_bbox': tensor(0.0056, device='cuda:0'), 'train_loss_giou': tensor(0.0549, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1472, device='cuda:0'), 'validation_loss_ce': tensor(0.0447, device='cuda:0'), 'validation_loss_bbox': tensor(0.0046, device='cuda:0'), 'validation_loss_giou': tensor(0.0398, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1472, device='cuda:0'), 'train_loss_ce': tensor(0.0447, device='cuda:0'), 'train_loss_bbox': tensor(0.0046, device='cuda:0'), 'train_loss_giou': tensor(0.0398, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.2155, device='cuda:0'), 'validation_loss_ce': tensor(0.0445, device='cuda:0'), 'validation_loss_bbox': tensor(0.0073, device='cuda:0'), 'validation_loss_giou': tensor(0.0672, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.2155, device='cuda:0'), 'train_loss_ce': tensor(0.0445, device='cuda:0'), 'train_loss_bbox': tensor(0.0073, device='cuda:0'), 'train_loss_giou': tensor(0.0672, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.2256, device='cuda:0'), 'validation_loss_ce': tensor(0.0443, device='cuda:0'), 'validation_loss_bbox': tensor(0.0073, device='cuda:0'), 'validation_loss_giou': tensor(0.0723, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.2256, device='cuda:0'), 'train_loss_ce': tensor(0.0443, device='cuda:0'), 'train_loss_bbox': tensor(0.0073, device='cuda:0'), 'train_loss_giou': tensor(0.0723, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.2005, device='cuda:0'), 'validation_loss_ce': tensor(0.0440, device='cuda:0'), 'validation_loss_bbox': tensor(0.0065, device='cuda:0'), 'validation_loss_giou': tensor(0.0620, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.2005, device='cuda:0'), 'train_loss_ce': tensor(0.0440, device='cuda:0'), 'train_loss_bbox': tensor(0.0065, device='cuda:0'), 'train_loss_giou': tensor(0.0620, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1726, device='cuda:0'), 'validation_loss_ce': tensor(0.0433, device='cuda:0'), 'validation_loss_bbox': tensor(0.0065, device='cuda:0'), 'validation_loss_giou': tensor(0.0484, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1726, device='cuda:0'), 'train_loss_ce': tensor(0.0433, device='cuda:0'), 'train_loss_bbox': tensor(0.0065, device='cuda:0'), 'train_loss_giou': tensor(0.0484, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.2005, device='cuda:0'), 'validation_loss_ce': tensor(0.0424, device='cuda:0'), 'validation_loss_bbox': tensor(0.0058, device='cuda:0'), 'validation_loss_giou': tensor(0.0646, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.2005, device='cuda:0'), 'train_loss_ce': tensor(0.0424, device='cuda:0'), 'train_loss_bbox': tensor(0.0058, device='cuda:0'), 'train_loss_giou': tensor(0.0646, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.2282, device='cuda:0'), 'validation_loss_ce': tensor(0.0420, device='cuda:0'), 'validation_loss_bbox': tensor(0.0074, device='cuda:0'), 'validation_loss_giou': tensor(0.0747, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.2282, device='cuda:0'), 'train_loss_ce': tensor(0.0420, device='cuda:0'), 'train_loss_bbox': tensor(0.0074, device='cuda:0'), 'train_loss_giou': tensor(0.0747, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.2553, device='cuda:0'), 'validation_loss_ce': tensor(0.0422, device='cuda:0'), 'validation_loss_bbox': tensor(0.0083, device='cuda:0'), 'validation_loss_giou': tensor(0.0858, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.2553, device='cuda:0'), 'train_loss_ce': tensor(0.0422, device='cuda:0'), 'train_loss_bbox': tensor(0.0083, device='cuda:0'), 'train_loss_giou': tensor(0.0858, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.1922, device='cuda:0'), 'validation_loss_ce': tensor(0.0424, device='cuda:0'), 'validation_loss_bbox': tensor(0.0061, device='cuda:0'), 'validation_loss_giou': tensor(0.0597, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
-{'training_loss': tensor(0.1922, device='cuda:0'), 'train_loss_ce': tensor(0.0424, device='cuda:0'), 'train_loss_bbox': tensor(0.0061, device='cuda:0'), 'train_loss_giou': tensor(0.0597, device='cuda:0'), 'train_cardinality_error': tensor(0., device='cuda:0'), 'validation_loss': tensor(0.0993, device='cuda:0'), 'validation_loss_ce': tensor(0.0419, device='cuda:0'), 'validation_loss_bbox': tensor(0.0020, device='cuda:0'), 'validation_loss_giou': tensor(0.0237, device='cuda:0'), 'validation_cardinality_error': tensor(0., device='cuda:0')}
+{'validation_loss': tensor(5.6017, device='cuda:0'), 'validation_loss_ce': tensor(2.0849, device='cuda:0'), 'validation_loss_bbox': tensor(0.4302, device='cuda:0'), 'validation_loss_giou': tensor(0.6829, device='cuda:0'), 'validation_cardinality_error': tensor(87., device='cuda:0')}
+{'training_loss': tensor(5.6017, device='cuda:0'), 'train_loss_ce': tensor(2.0849, device='cuda:0'), 'train_loss_bbox': tensor(0.4302, device='cuda:0'), 'train_loss_giou': tensor(0.6829, device='cuda:0'), 'train_cardinality_error': tensor(87., device='cuda:0'), 'validation_loss': tensor(6.3627, device='cuda:0'), 'validation_loss_ce': tensor(1.8613, device='cuda:0'), 'validation_loss_bbox': tensor(0.5789, device='cuda:0'), 'validation_loss_giou': tensor(0.8034, device='cuda:0'), 'validation_cardinality_error': tensor(43.5000, device='cuda:0')}
+{'training_loss': tensor(6.3627, device='cuda:0'), 'train_loss_ce': tensor(1.8613, device='cuda:0'), 'train_loss_bbox': tensor(0.5789, device='cuda:0'), 'train_loss_giou': tensor(0.8034, device='cuda:0'), 'train_cardinality_error': tensor(43.5000, device='cuda:0'), 'validation_loss': tensor(7.0474, device='cuda:0'), 'validation_loss_ce': tensor(1.9586, device='cuda:0'), 'validation_loss_bbox': tensor(0.6317, device='cuda:0'), 'validation_loss_giou': tensor(0.9652, device='cuda:0'), 'validation_cardinality_error': tensor(73.5000, device='cuda:0')}
+{'training_loss': tensor(7.0474, device='cuda:0'), 'train_loss_ce': tensor(1.9586, device='cuda:0'), 'train_loss_bbox': tensor(0.6317, device='cuda:0'), 'train_loss_giou': tensor(0.9652, device='cuda:0'), 'train_cardinality_error': tensor(73.5000, device='cuda:0'), 'validation_loss': tensor(6.2814, device='cuda:0'), 'validation_loss_ce': tensor(1.8006, device='cuda:0'), 'validation_loss_bbox': tensor(0.5896, device='cuda:0'), 'validation_loss_giou': tensor(0.7664, device='cuda:0'), 'validation_cardinality_error': tensor(25.5000, device='cuda:0')}
+{'training_loss': tensor(6.2814, device='cuda:0'), 'train_loss_ce': tensor(1.8006, device='cuda:0'), 'train_loss_bbox': tensor(0.5896, device='cuda:0'), 'train_loss_giou': tensor(0.7664, device='cuda:0'), 'train_cardinality_error': tensor(25.5000, device='cuda:0'), 'validation_loss': tensor(5.7531, device='cuda:0'), 'validation_loss_ce': tensor(1.7510, device='cuda:0'), 'validation_loss_bbox': tensor(0.5194, device='cuda:0'), 'validation_loss_giou': tensor(0.7026, device='cuda:0'), 'validation_cardinality_error': tensor(5., device='cuda:0')}
+{'training_loss': tensor(5.7531, device='cuda:0'), 'train_loss_ce': tensor(1.7510, device='cuda:0'), 'train_loss_bbox': tensor(0.5194, device='cuda:0'), 'train_loss_giou': tensor(0.7026, device='cuda:0'), 'train_cardinality_error': tensor(5., device='cuda:0'), 'validation_loss': tensor(5.7845, device='cuda:0'), 'validation_loss_ce': tensor(1.7435, device='cuda:0'), 'validation_loss_bbox': tensor(0.5269, device='cuda:0'), 'validation_loss_giou': tensor(0.7032, device='cuda:0'), 'validation_cardinality_error': tensor(5.5000, device='cuda:0')}
+{'training_loss': tensor(5.7845, device='cuda:0'), 'train_loss_ce': tensor(1.7435, device='cuda:0'), 'train_loss_bbox': tensor(0.5269, device='cuda:0'), 'train_loss_giou': tensor(0.7032, device='cuda:0'), 'train_cardinality_error': tensor(5.5000, device='cuda:0'), 'validation_loss': tensor(5.2789, device='cuda:0'), 'validation_loss_ce': tensor(1.6650, device='cuda:0'), 'validation_loss_bbox': tensor(0.4057, device='cuda:0'), 'validation_loss_giou': tensor(0.7927, device='cuda:0'), 'validation_cardinality_error': tensor(2.5000, device='cuda:0')}
+{'training_loss': tensor(5.2789, device='cuda:0'), 'train_loss_ce': tensor(1.6650, device='cuda:0'), 'train_loss_bbox': tensor(0.4057, device='cuda:0'), 'train_loss_giou': tensor(0.7927, device='cuda:0'), 'train_cardinality_error': tensor(2.5000, device='cuda:0'), 'validation_loss': tensor(4.2995, device='cuda:0'), 'validation_loss_ce': tensor(1.6013, device='cuda:0'), 'validation_loss_bbox': tensor(0.3041, device='cuda:0'), 'validation_loss_giou': tensor(0.5890, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(4.2995, device='cuda:0'), 'train_loss_ce': tensor(1.6013, device='cuda:0'), 'train_loss_bbox': tensor(0.3041, device='cuda:0'), 'train_loss_giou': tensor(0.5890, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.6465, device='cuda:0'), 'validation_loss_ce': tensor(1.5378, device='cuda:0'), 'validation_loss_bbox': tensor(0.3922, device='cuda:0'), 'validation_loss_giou': tensor(0.5738, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(4.6465, device='cuda:0'), 'train_loss_ce': tensor(1.5378, device='cuda:0'), 'train_loss_bbox': tensor(0.3922, device='cuda:0'), 'train_loss_giou': tensor(0.5738, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.6222, device='cuda:0'), 'validation_loss_ce': tensor(1.5296, device='cuda:0'), 'validation_loss_bbox': tensor(0.3489, device='cuda:0'), 'validation_loss_giou': tensor(0.6740, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(4.6222, device='cuda:0'), 'train_loss_ce': tensor(1.5296, device='cuda:0'), 'train_loss_bbox': tensor(0.3489, device='cuda:0'), 'train_loss_giou': tensor(0.6740, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.0194, device='cuda:0'), 'validation_loss_ce': tensor(1.5352, device='cuda:0'), 'validation_loss_bbox': tensor(0.2939, device='cuda:0'), 'validation_loss_giou': tensor(0.5074, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(4.0194, device='cuda:0'), 'train_loss_ce': tensor(1.5352, device='cuda:0'), 'train_loss_bbox': tensor(0.2939, device='cuda:0'), 'train_loss_giou': tensor(0.5074, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.5375, device='cuda:0'), 'validation_loss_ce': tensor(1.4859, device='cuda:0'), 'validation_loss_bbox': tensor(0.2289, device='cuda:0'), 'validation_loss_giou': tensor(0.4536, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(3.5375, device='cuda:0'), 'train_loss_ce': tensor(1.4859, device='cuda:0'), 'train_loss_bbox': tensor(0.2289, device='cuda:0'), 'train_loss_giou': tensor(0.4536, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.1891, device='cuda:0'), 'validation_loss_ce': tensor(1.3903, device='cuda:0'), 'validation_loss_bbox': tensor(0.1878, device='cuda:0'), 'validation_loss_giou': tensor(0.4300, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(3.1891, device='cuda:0'), 'train_loss_ce': tensor(1.3903, device='cuda:0'), 'train_loss_bbox': tensor(0.1878, device='cuda:0'), 'train_loss_giou': tensor(0.4300, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.7035, device='cuda:0'), 'validation_loss_ce': tensor(1.3643, device='cuda:0'), 'validation_loss_bbox': tensor(0.1410, device='cuda:0'), 'validation_loss_giou': tensor(0.3170, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(2.7035, device='cuda:0'), 'train_loss_ce': tensor(1.3643, device='cuda:0'), 'train_loss_bbox': tensor(0.1410, device='cuda:0'), 'train_loss_giou': tensor(0.3170, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.3029, device='cuda:0'), 'validation_loss_ce': tensor(1.3261, device='cuda:0'), 'validation_loss_bbox': tensor(0.0884, device='cuda:0'), 'validation_loss_giou': tensor(0.2675, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(2.3029, device='cuda:0'), 'train_loss_ce': tensor(1.3261, device='cuda:0'), 'train_loss_bbox': tensor(0.0884, device='cuda:0'), 'train_loss_giou': tensor(0.2675, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.7946, device='cuda:0'), 'validation_loss_ce': tensor(1.3143, device='cuda:0'), 'validation_loss_bbox': tensor(0.1497, device='cuda:0'), 'validation_loss_giou': tensor(0.3658, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(2.7946, device='cuda:0'), 'train_loss_ce': tensor(1.3143, device='cuda:0'), 'train_loss_bbox': tensor(0.1497, device='cuda:0'), 'train_loss_giou': tensor(0.3658, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.5422, device='cuda:0'), 'validation_loss_ce': tensor(1.2696, device='cuda:0'), 'validation_loss_bbox': tensor(0.1339, device='cuda:0'), 'validation_loss_giou': tensor(0.3015, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(2.5422, device='cuda:0'), 'train_loss_ce': tensor(1.2696, device='cuda:0'), 'train_loss_bbox': tensor(0.1339, device='cuda:0'), 'train_loss_giou': tensor(0.3015, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.6529, device='cuda:0'), 'validation_loss_ce': tensor(1.2525, device='cuda:0'), 'validation_loss_bbox': tensor(0.1608, device='cuda:0'), 'validation_loss_giou': tensor(0.2982, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(2.6529, device='cuda:0'), 'train_loss_ce': tensor(1.2525, device='cuda:0'), 'train_loss_bbox': tensor(0.1608, device='cuda:0'), 'train_loss_giou': tensor(0.2982, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.1509, device='cuda:0'), 'validation_loss_ce': tensor(1.1918, device='cuda:0'), 'validation_loss_bbox': tensor(0.0941, device='cuda:0'), 'validation_loss_giou': tensor(0.2443, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(2.1509, device='cuda:0'), 'train_loss_ce': tensor(1.1918, device='cuda:0'), 'train_loss_bbox': tensor(0.0941, device='cuda:0'), 'train_loss_giou': tensor(0.2443, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.6681, device='cuda:0'), 'validation_loss_ce': tensor(1.0913, device='cuda:0'), 'validation_loss_bbox': tensor(0.0635, device='cuda:0'), 'validation_loss_giou': tensor(0.1297, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(1.6681, device='cuda:0'), 'train_loss_ce': tensor(1.0913, device='cuda:0'), 'train_loss_bbox': tensor(0.0635, device='cuda:0'), 'train_loss_giou': tensor(0.1297, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.8172, device='cuda:0'), 'validation_loss_ce': tensor(1.0121, device='cuda:0'), 'validation_loss_bbox': tensor(0.0878, device='cuda:0'), 'validation_loss_giou': tensor(0.1830, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(1.8172, device='cuda:0'), 'train_loss_ce': tensor(1.0121, device='cuda:0'), 'train_loss_bbox': tensor(0.0878, device='cuda:0'), 'train_loss_giou': tensor(0.1830, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.7235, device='cuda:0'), 'validation_loss_ce': tensor(0.9522, device='cuda:0'), 'validation_loss_bbox': tensor(0.0711, device='cuda:0'), 'validation_loss_giou': tensor(0.2078, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(1.7235, device='cuda:0'), 'train_loss_ce': tensor(0.9522, device='cuda:0'), 'train_loss_bbox': tensor(0.0711, device='cuda:0'), 'train_loss_giou': tensor(0.2078, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.5285, device='cuda:0'), 'validation_loss_ce': tensor(0.9022, device='cuda:0'), 'validation_loss_bbox': tensor(0.0596, device='cuda:0'), 'validation_loss_giou': tensor(0.1642, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(1.5285, device='cuda:0'), 'train_loss_ce': tensor(0.9022, device='cuda:0'), 'train_loss_bbox': tensor(0.0596, device='cuda:0'), 'train_loss_giou': tensor(0.1642, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.2834, device='cuda:0'), 'validation_loss_ce': tensor(0.8606, device='cuda:0'), 'validation_loss_bbox': tensor(0.0364, device='cuda:0'), 'validation_loss_giou': tensor(0.1205, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(1.2834, device='cuda:0'), 'train_loss_ce': tensor(0.8606, device='cuda:0'), 'train_loss_bbox': tensor(0.0364, device='cuda:0'), 'train_loss_giou': tensor(0.1205, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.3022, device='cuda:0'), 'validation_loss_ce': tensor(0.8215, device='cuda:0'), 'validation_loss_bbox': tensor(0.0422, device='cuda:0'), 'validation_loss_giou': tensor(0.1348, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(1.3022, device='cuda:0'), 'train_loss_ce': tensor(0.8215, device='cuda:0'), 'train_loss_bbox': tensor(0.0422, device='cuda:0'), 'train_loss_giou': tensor(0.1348, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.0249, device='cuda:0'), 'validation_loss_ce': tensor(0.7992, device='cuda:0'), 'validation_loss_bbox': tensor(0.0205, device='cuda:0'), 'validation_loss_giou': tensor(0.0614, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(1.0249, device='cuda:0'), 'train_loss_ce': tensor(0.7992, device='cuda:0'), 'train_loss_bbox': tensor(0.0205, device='cuda:0'), 'train_loss_giou': tensor(0.0614, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.0823, device='cuda:0'), 'validation_loss_ce': tensor(0.7821, device='cuda:0'), 'validation_loss_bbox': tensor(0.0313, device='cuda:0'), 'validation_loss_giou': tensor(0.0718, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(1.0823, device='cuda:0'), 'train_loss_ce': tensor(0.7821, device='cuda:0'), 'train_loss_bbox': tensor(0.0313, device='cuda:0'), 'train_loss_giou': tensor(0.0718, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.9961, device='cuda:0'), 'validation_loss_ce': tensor(0.7615, device='cuda:0'), 'validation_loss_bbox': tensor(0.0242, device='cuda:0'), 'validation_loss_giou': tensor(0.0567, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.9961, device='cuda:0'), 'train_loss_ce': tensor(0.7615, device='cuda:0'), 'train_loss_bbox': tensor(0.0242, device='cuda:0'), 'train_loss_giou': tensor(0.0567, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.9793, device='cuda:0'), 'validation_loss_ce': tensor(0.7294, device='cuda:0'), 'validation_loss_bbox': tensor(0.0255, device='cuda:0'), 'validation_loss_giou': tensor(0.0611, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.9793, device='cuda:0'), 'train_loss_ce': tensor(0.7294, device='cuda:0'), 'train_loss_bbox': tensor(0.0255, device='cuda:0'), 'train_loss_giou': tensor(0.0611, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.9178, device='cuda:0'), 'validation_loss_ce': tensor(0.7077, device='cuda:0'), 'validation_loss_bbox': tensor(0.0225, device='cuda:0'), 'validation_loss_giou': tensor(0.0488, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.9178, device='cuda:0'), 'train_loss_ce': tensor(0.7077, device='cuda:0'), 'train_loss_bbox': tensor(0.0225, device='cuda:0'), 'train_loss_giou': tensor(0.0488, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.8359, device='cuda:0'), 'validation_loss_ce': tensor(0.6958, device='cuda:0'), 'validation_loss_bbox': tensor(0.0142, device='cuda:0'), 'validation_loss_giou': tensor(0.0345, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.8359, device='cuda:0'), 'train_loss_ce': tensor(0.6958, device='cuda:0'), 'train_loss_bbox': tensor(0.0142, device='cuda:0'), 'train_loss_giou': tensor(0.0345, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.8585, device='cuda:0'), 'validation_loss_ce': tensor(0.6812, device='cuda:0'), 'validation_loss_bbox': tensor(0.0170, device='cuda:0'), 'validation_loss_giou': tensor(0.0462, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.8585, device='cuda:0'), 'train_loss_ce': tensor(0.6812, device='cuda:0'), 'train_loss_bbox': tensor(0.0170, device='cuda:0'), 'train_loss_giou': tensor(0.0462, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.7617, device='cuda:0'), 'validation_loss_ce': tensor(0.6592, device='cuda:0'), 'validation_loss_bbox': tensor(0.0102, device='cuda:0'), 'validation_loss_giou': tensor(0.0257, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.7617, device='cuda:0'), 'train_loss_ce': tensor(0.6592, device='cuda:0'), 'train_loss_bbox': tensor(0.0102, device='cuda:0'), 'train_loss_giou': tensor(0.0257, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.8044, device='cuda:0'), 'validation_loss_ce': tensor(0.6329, device='cuda:0'), 'validation_loss_bbox': tensor(0.0130, device='cuda:0'), 'validation_loss_giou': tensor(0.0533, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.8044, device='cuda:0'), 'train_loss_ce': tensor(0.6329, device='cuda:0'), 'train_loss_bbox': tensor(0.0130, device='cuda:0'), 'train_loss_giou': tensor(0.0533, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.8279, device='cuda:0'), 'validation_loss_ce': tensor(0.6079, device='cuda:0'), 'validation_loss_bbox': tensor(0.0221, device='cuda:0'), 'validation_loss_giou': tensor(0.0547, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.8279, device='cuda:0'), 'train_loss_ce': tensor(0.6079, device='cuda:0'), 'train_loss_bbox': tensor(0.0221, device='cuda:0'), 'train_loss_giou': tensor(0.0547, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.8013, device='cuda:0'), 'validation_loss_ce': tensor(0.5928, device='cuda:0'), 'validation_loss_bbox': tensor(0.0228, device='cuda:0'), 'validation_loss_giou': tensor(0.0472, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.8013, device='cuda:0'), 'train_loss_ce': tensor(0.5928, device='cuda:0'), 'train_loss_bbox': tensor(0.0228, device='cuda:0'), 'train_loss_giou': tensor(0.0472, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.7321, device='cuda:0'), 'validation_loss_ce': tensor(0.5778, device='cuda:0'), 'validation_loss_bbox': tensor(0.0156, device='cuda:0'), 'validation_loss_giou': tensor(0.0382, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.7321, device='cuda:0'), 'train_loss_ce': tensor(0.5778, device='cuda:0'), 'train_loss_bbox': tensor(0.0156, device='cuda:0'), 'train_loss_giou': tensor(0.0382, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.6580, device='cuda:0'), 'validation_loss_ce': tensor(0.5646, device='cuda:0'), 'validation_loss_bbox': tensor(0.0091, device='cuda:0'), 'validation_loss_giou': tensor(0.0239, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.6580, device='cuda:0'), 'train_loss_ce': tensor(0.5646, device='cuda:0'), 'train_loss_bbox': tensor(0.0091, device='cuda:0'), 'train_loss_giou': tensor(0.0239, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.6818, device='cuda:0'), 'validation_loss_ce': tensor(0.5514, device='cuda:0'), 'validation_loss_bbox': tensor(0.0147, device='cuda:0'), 'validation_loss_giou': tensor(0.0285, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.6818, device='cuda:0'), 'train_loss_ce': tensor(0.5514, device='cuda:0'), 'train_loss_bbox': tensor(0.0147, device='cuda:0'), 'train_loss_giou': tensor(0.0285, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.6764, device='cuda:0'), 'validation_loss_ce': tensor(0.5379, device='cuda:0'), 'validation_loss_bbox': tensor(0.0143, device='cuda:0'), 'validation_loss_giou': tensor(0.0334, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.6764, device='cuda:0'), 'train_loss_ce': tensor(0.5379, device='cuda:0'), 'train_loss_bbox': tensor(0.0143, device='cuda:0'), 'train_loss_giou': tensor(0.0334, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.6775, device='cuda:0'), 'validation_loss_ce': tensor(0.5208, device='cuda:0'), 'validation_loss_bbox': tensor(0.0129, device='cuda:0'), 'validation_loss_giou': tensor(0.0462, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.6775, device='cuda:0'), 'train_loss_ce': tensor(0.5208, device='cuda:0'), 'train_loss_bbox': tensor(0.0129, device='cuda:0'), 'train_loss_giou': tensor(0.0462, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.6520, device='cuda:0'), 'validation_loss_ce': tensor(0.5099, device='cuda:0'), 'validation_loss_bbox': tensor(0.0125, device='cuda:0'), 'validation_loss_giou': tensor(0.0399, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.6520, device='cuda:0'), 'train_loss_ce': tensor(0.5099, device='cuda:0'), 'train_loss_bbox': tensor(0.0125, device='cuda:0'), 'train_loss_giou': tensor(0.0399, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.6051, device='cuda:0'), 'validation_loss_ce': tensor(0.5030, device='cuda:0'), 'validation_loss_bbox': tensor(0.0071, device='cuda:0'), 'validation_loss_giou': tensor(0.0334, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.6051, device='cuda:0'), 'train_loss_ce': tensor(0.5030, device='cuda:0'), 'train_loss_bbox': tensor(0.0071, device='cuda:0'), 'train_loss_giou': tensor(0.0334, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.6969, device='cuda:0'), 'validation_loss_ce': tensor(0.4982, device='cuda:0'), 'validation_loss_bbox': tensor(0.0175, device='cuda:0'), 'validation_loss_giou': tensor(0.0556, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.6969, device='cuda:0'), 'train_loss_ce': tensor(0.4982, device='cuda:0'), 'train_loss_bbox': tensor(0.0175, device='cuda:0'), 'train_loss_giou': tensor(0.0556, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.6910, device='cuda:0'), 'validation_loss_ce': tensor(0.4916, device='cuda:0'), 'validation_loss_bbox': tensor(0.0197, device='cuda:0'), 'validation_loss_giou': tensor(0.0506, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.6910, device='cuda:0'), 'train_loss_ce': tensor(0.4916, device='cuda:0'), 'train_loss_bbox': tensor(0.0197, device='cuda:0'), 'train_loss_giou': tensor(0.0506, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.7090, device='cuda:0'), 'validation_loss_ce': tensor(0.4834, device='cuda:0'), 'validation_loss_bbox': tensor(0.0232, device='cuda:0'), 'validation_loss_giou': tensor(0.0549, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.7090, device='cuda:0'), 'train_loss_ce': tensor(0.4834, device='cuda:0'), 'train_loss_bbox': tensor(0.0232, device='cuda:0'), 'train_loss_giou': tensor(0.0549, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.7374, device='cuda:0'), 'validation_loss_ce': tensor(0.4753, device='cuda:0'), 'validation_loss_bbox': tensor(0.0242, device='cuda:0'), 'validation_loss_giou': tensor(0.0707, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.7374, device='cuda:0'), 'train_loss_ce': tensor(0.4753, device='cuda:0'), 'train_loss_bbox': tensor(0.0242, device='cuda:0'), 'train_loss_giou': tensor(0.0707, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.6690, device='cuda:0'), 'validation_loss_ce': tensor(0.4680, device='cuda:0'), 'validation_loss_bbox': tensor(0.0166, device='cuda:0'), 'validation_loss_giou': tensor(0.0589, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.6690, device='cuda:0'), 'train_loss_ce': tensor(0.4680, device='cuda:0'), 'train_loss_bbox': tensor(0.0166, device='cuda:0'), 'train_loss_giou': tensor(0.0589, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.6594, device='cuda:0'), 'validation_loss_ce': tensor(0.4621, device='cuda:0'), 'validation_loss_bbox': tensor(0.0226, device='cuda:0'), 'validation_loss_giou': tensor(0.0421, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.6594, device='cuda:0'), 'train_loss_ce': tensor(0.4621, device='cuda:0'), 'train_loss_bbox': tensor(0.0226, device='cuda:0'), 'train_loss_giou': tensor(0.0421, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.7248, device='cuda:0'), 'validation_loss_ce': tensor(0.4574, device='cuda:0'), 'validation_loss_bbox': tensor(0.0245, device='cuda:0'), 'validation_loss_giou': tensor(0.0726, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.7248, device='cuda:0'), 'train_loss_ce': tensor(0.4574, device='cuda:0'), 'train_loss_bbox': tensor(0.0245, device='cuda:0'), 'train_loss_giou': tensor(0.0726, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.7158, device='cuda:0'), 'validation_loss_ce': tensor(0.4522, device='cuda:0'), 'validation_loss_bbox': tensor(0.0206, device='cuda:0'), 'validation_loss_giou': tensor(0.0804, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.7158, device='cuda:0'), 'train_loss_ce': tensor(0.4522, device='cuda:0'), 'train_loss_bbox': tensor(0.0206, device='cuda:0'), 'train_loss_giou': tensor(0.0804, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.6459, device='cuda:0'), 'validation_loss_ce': tensor(0.4469, device='cuda:0'), 'validation_loss_bbox': tensor(0.0133, device='cuda:0'), 'validation_loss_giou': tensor(0.0662, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.6459, device='cuda:0'), 'train_loss_ce': tensor(0.4469, device='cuda:0'), 'train_loss_bbox': tensor(0.0133, device='cuda:0'), 'train_loss_giou': tensor(0.0662, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.5799, device='cuda:0'), 'validation_loss_ce': tensor(0.4413, device='cuda:0'), 'validation_loss_bbox': tensor(0.0099, device='cuda:0'), 'validation_loss_giou': tensor(0.0447, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.5799, device='cuda:0'), 'train_loss_ce': tensor(0.4413, device='cuda:0'), 'train_loss_bbox': tensor(0.0099, device='cuda:0'), 'train_loss_giou': tensor(0.0447, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.5798, device='cuda:0'), 'validation_loss_ce': tensor(0.4356, device='cuda:0'), 'validation_loss_bbox': tensor(0.0144, device='cuda:0'), 'validation_loss_giou': tensor(0.0361, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.5798, device='cuda:0'), 'train_loss_ce': tensor(0.4356, device='cuda:0'), 'train_loss_bbox': tensor(0.0144, device='cuda:0'), 'train_loss_giou': tensor(0.0361, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.6000, device='cuda:0'), 'validation_loss_ce': tensor(0.4313, device='cuda:0'), 'validation_loss_bbox': tensor(0.0143, device='cuda:0'), 'validation_loss_giou': tensor(0.0487, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.6000, device='cuda:0'), 'train_loss_ce': tensor(0.4313, device='cuda:0'), 'train_loss_bbox': tensor(0.0143, device='cuda:0'), 'train_loss_giou': tensor(0.0487, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.5141, device='cuda:0'), 'validation_loss_ce': tensor(0.4280, device='cuda:0'), 'validation_loss_bbox': tensor(0.0066, device='cuda:0'), 'validation_loss_giou': tensor(0.0265, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.5141, device='cuda:0'), 'train_loss_ce': tensor(0.4280, device='cuda:0'), 'train_loss_bbox': tensor(0.0066, device='cuda:0'), 'train_loss_giou': tensor(0.0265, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.5365, device='cuda:0'), 'validation_loss_ce': tensor(0.4249, device='cuda:0'), 'validation_loss_bbox': tensor(0.0093, device='cuda:0'), 'validation_loss_giou': tensor(0.0325, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.5365, device='cuda:0'), 'train_loss_ce': tensor(0.4249, device='cuda:0'), 'train_loss_bbox': tensor(0.0093, device='cuda:0'), 'train_loss_giou': tensor(0.0325, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.5261, device='cuda:0'), 'validation_loss_ce': tensor(0.4211, device='cuda:0'), 'validation_loss_bbox': tensor(0.0074, device='cuda:0'), 'validation_loss_giou': tensor(0.0339, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.5261, device='cuda:0'), 'train_loss_ce': tensor(0.4211, device='cuda:0'), 'train_loss_bbox': tensor(0.0074, device='cuda:0'), 'train_loss_giou': tensor(0.0339, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4900, device='cuda:0'), 'validation_loss_ce': tensor(0.4172, device='cuda:0'), 'validation_loss_bbox': tensor(0.0063, device='cuda:0'), 'validation_loss_giou': tensor(0.0206, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4900, device='cuda:0'), 'train_loss_ce': tensor(0.4172, device='cuda:0'), 'train_loss_bbox': tensor(0.0063, device='cuda:0'), 'train_loss_giou': tensor(0.0206, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4777, device='cuda:0'), 'validation_loss_ce': tensor(0.4140, device='cuda:0'), 'validation_loss_bbox': tensor(0.0062, device='cuda:0'), 'validation_loss_giou': tensor(0.0164, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4777, device='cuda:0'), 'train_loss_ce': tensor(0.4140, device='cuda:0'), 'train_loss_bbox': tensor(0.0062, device='cuda:0'), 'train_loss_giou': tensor(0.0164, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4761, device='cuda:0'), 'validation_loss_ce': tensor(0.4111, device='cuda:0'), 'validation_loss_bbox': tensor(0.0055, device='cuda:0'), 'validation_loss_giou': tensor(0.0188, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4761, device='cuda:0'), 'train_loss_ce': tensor(0.4111, device='cuda:0'), 'train_loss_bbox': tensor(0.0055, device='cuda:0'), 'train_loss_giou': tensor(0.0188, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4847, device='cuda:0'), 'validation_loss_ce': tensor(0.4080, device='cuda:0'), 'validation_loss_bbox': tensor(0.0061, device='cuda:0'), 'validation_loss_giou': tensor(0.0230, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4847, device='cuda:0'), 'train_loss_ce': tensor(0.4080, device='cuda:0'), 'train_loss_bbox': tensor(0.0061, device='cuda:0'), 'train_loss_giou': tensor(0.0230, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4779, device='cuda:0'), 'validation_loss_ce': tensor(0.4051, device='cuda:0'), 'validation_loss_bbox': tensor(0.0056, device='cuda:0'), 'validation_loss_giou': tensor(0.0224, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4779, device='cuda:0'), 'train_loss_ce': tensor(0.4051, device='cuda:0'), 'train_loss_bbox': tensor(0.0056, device='cuda:0'), 'train_loss_giou': tensor(0.0224, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4429, device='cuda:0'), 'validation_loss_ce': tensor(0.4020, device='cuda:0'), 'validation_loss_bbox': tensor(0.0031, device='cuda:0'), 'validation_loss_giou': tensor(0.0126, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4429, device='cuda:0'), 'train_loss_ce': tensor(0.4020, device='cuda:0'), 'train_loss_bbox': tensor(0.0031, device='cuda:0'), 'train_loss_giou': tensor(0.0126, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4499, device='cuda:0'), 'validation_loss_ce': tensor(0.3991, device='cuda:0'), 'validation_loss_bbox': tensor(0.0042, device='cuda:0'), 'validation_loss_giou': tensor(0.0148, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4499, device='cuda:0'), 'train_loss_ce': tensor(0.3991, device='cuda:0'), 'train_loss_bbox': tensor(0.0042, device='cuda:0'), 'train_loss_giou': tensor(0.0148, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4548, device='cuda:0'), 'validation_loss_ce': tensor(0.3966, device='cuda:0'), 'validation_loss_bbox': tensor(0.0055, device='cuda:0'), 'validation_loss_giou': tensor(0.0153, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4548, device='cuda:0'), 'train_loss_ce': tensor(0.3966, device='cuda:0'), 'train_loss_bbox': tensor(0.0055, device='cuda:0'), 'train_loss_giou': tensor(0.0153, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4586, device='cuda:0'), 'validation_loss_ce': tensor(0.3941, device='cuda:0'), 'validation_loss_bbox': tensor(0.0055, device='cuda:0'), 'validation_loss_giou': tensor(0.0184, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4586, device='cuda:0'), 'train_loss_ce': tensor(0.3941, device='cuda:0'), 'train_loss_bbox': tensor(0.0055, device='cuda:0'), 'train_loss_giou': tensor(0.0184, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4583, device='cuda:0'), 'validation_loss_ce': tensor(0.3914, device='cuda:0'), 'validation_loss_bbox': tensor(0.0063, device='cuda:0'), 'validation_loss_giou': tensor(0.0178, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4583, device='cuda:0'), 'train_loss_ce': tensor(0.3914, device='cuda:0'), 'train_loss_bbox': tensor(0.0063, device='cuda:0'), 'train_loss_giou': tensor(0.0178, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4632, device='cuda:0'), 'validation_loss_ce': tensor(0.3887, device='cuda:0'), 'validation_loss_bbox': tensor(0.0076, device='cuda:0'), 'validation_loss_giou': tensor(0.0182, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4632, device='cuda:0'), 'train_loss_ce': tensor(0.3887, device='cuda:0'), 'train_loss_bbox': tensor(0.0076, device='cuda:0'), 'train_loss_giou': tensor(0.0182, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4547, device='cuda:0'), 'validation_loss_ce': tensor(0.3862, device='cuda:0'), 'validation_loss_bbox': tensor(0.0067, device='cuda:0'), 'validation_loss_giou': tensor(0.0174, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4547, device='cuda:0'), 'train_loss_ce': tensor(0.3862, device='cuda:0'), 'train_loss_bbox': tensor(0.0067, device='cuda:0'), 'train_loss_giou': tensor(0.0174, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4658, device='cuda:0'), 'validation_loss_ce': tensor(0.3836, device='cuda:0'), 'validation_loss_bbox': tensor(0.0068, device='cuda:0'), 'validation_loss_giou': tensor(0.0241, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4658, device='cuda:0'), 'train_loss_ce': tensor(0.3836, device='cuda:0'), 'train_loss_bbox': tensor(0.0068, device='cuda:0'), 'train_loss_giou': tensor(0.0241, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4300, device='cuda:0'), 'validation_loss_ce': tensor(0.3811, device='cuda:0'), 'validation_loss_bbox': tensor(0.0034, device='cuda:0'), 'validation_loss_giou': tensor(0.0161, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4300, device='cuda:0'), 'train_loss_ce': tensor(0.3811, device='cuda:0'), 'train_loss_bbox': tensor(0.0034, device='cuda:0'), 'train_loss_giou': tensor(0.0161, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4717, device='cuda:0'), 'validation_loss_ce': tensor(0.3787, device='cuda:0'), 'validation_loss_bbox': tensor(0.0076, device='cuda:0'), 'validation_loss_giou': tensor(0.0275, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4717, device='cuda:0'), 'train_loss_ce': tensor(0.3787, device='cuda:0'), 'train_loss_bbox': tensor(0.0076, device='cuda:0'), 'train_loss_giou': tensor(0.0275, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4389, device='cuda:0'), 'validation_loss_ce': tensor(0.3766, device='cuda:0'), 'validation_loss_bbox': tensor(0.0055, device='cuda:0'), 'validation_loss_giou': tensor(0.0173, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4389, device='cuda:0'), 'train_loss_ce': tensor(0.3766, device='cuda:0'), 'train_loss_bbox': tensor(0.0055, device='cuda:0'), 'train_loss_giou': tensor(0.0173, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4551, device='cuda:0'), 'validation_loss_ce': tensor(0.3746, device='cuda:0'), 'validation_loss_bbox': tensor(0.0055, device='cuda:0'), 'validation_loss_giou': tensor(0.0265, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4551, device='cuda:0'), 'train_loss_ce': tensor(0.3746, device='cuda:0'), 'train_loss_bbox': tensor(0.0055, device='cuda:0'), 'train_loss_giou': tensor(0.0265, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4469, device='cuda:0'), 'validation_loss_ce': tensor(0.3724, device='cuda:0'), 'validation_loss_bbox': tensor(0.0053, device='cuda:0'), 'validation_loss_giou': tensor(0.0241, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4469, device='cuda:0'), 'train_loss_ce': tensor(0.3724, device='cuda:0'), 'train_loss_bbox': tensor(0.0053, device='cuda:0'), 'train_loss_giou': tensor(0.0241, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4077, device='cuda:0'), 'validation_loss_ce': tensor(0.3700, device='cuda:0'), 'validation_loss_bbox': tensor(0.0032, device='cuda:0'), 'validation_loss_giou': tensor(0.0107, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4077, device='cuda:0'), 'train_loss_ce': tensor(0.3700, device='cuda:0'), 'train_loss_bbox': tensor(0.0032, device='cuda:0'), 'train_loss_giou': tensor(0.0107, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4331, device='cuda:0'), 'validation_loss_ce': tensor(0.3681, device='cuda:0'), 'validation_loss_bbox': tensor(0.0058, device='cuda:0'), 'validation_loss_giou': tensor(0.0181, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4331, device='cuda:0'), 'train_loss_ce': tensor(0.3681, device='cuda:0'), 'train_loss_bbox': tensor(0.0058, device='cuda:0'), 'train_loss_giou': tensor(0.0181, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4463, device='cuda:0'), 'validation_loss_ce': tensor(0.3662, device='cuda:0'), 'validation_loss_bbox': tensor(0.0077, device='cuda:0'), 'validation_loss_giou': tensor(0.0208, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4463, device='cuda:0'), 'train_loss_ce': tensor(0.3662, device='cuda:0'), 'train_loss_bbox': tensor(0.0077, device='cuda:0'), 'train_loss_giou': tensor(0.0208, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4259, device='cuda:0'), 'validation_loss_ce': tensor(0.3639, device='cuda:0'), 'validation_loss_bbox': tensor(0.0058, device='cuda:0'), 'validation_loss_giou': tensor(0.0165, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4259, device='cuda:0'), 'train_loss_ce': tensor(0.3639, device='cuda:0'), 'train_loss_bbox': tensor(0.0058, device='cuda:0'), 'train_loss_giou': tensor(0.0165, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4593, device='cuda:0'), 'validation_loss_ce': tensor(0.3618, device='cuda:0'), 'validation_loss_bbox': tensor(0.0071, device='cuda:0'), 'validation_loss_giou': tensor(0.0310, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4593, device='cuda:0'), 'train_loss_ce': tensor(0.3618, device='cuda:0'), 'train_loss_bbox': tensor(0.0071, device='cuda:0'), 'train_loss_giou': tensor(0.0310, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4432, device='cuda:0'), 'validation_loss_ce': tensor(0.3599, device='cuda:0'), 'validation_loss_bbox': tensor(0.0076, device='cuda:0'), 'validation_loss_giou': tensor(0.0227, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4432, device='cuda:0'), 'train_loss_ce': tensor(0.3599, device='cuda:0'), 'train_loss_bbox': tensor(0.0076, device='cuda:0'), 'train_loss_giou': tensor(0.0227, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4709, device='cuda:0'), 'validation_loss_ce': tensor(0.3582, device='cuda:0'), 'validation_loss_bbox': tensor(0.0087, device='cuda:0'), 'validation_loss_giou': tensor(0.0345, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4709, device='cuda:0'), 'train_loss_ce': tensor(0.3582, device='cuda:0'), 'train_loss_bbox': tensor(0.0087, device='cuda:0'), 'train_loss_giou': tensor(0.0345, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.5017, device='cuda:0'), 'validation_loss_ce': tensor(0.3562, device='cuda:0'), 'validation_loss_bbox': tensor(0.0117, device='cuda:0'), 'validation_loss_giou': tensor(0.0435, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.5017, device='cuda:0'), 'train_loss_ce': tensor(0.3562, device='cuda:0'), 'train_loss_bbox': tensor(0.0117, device='cuda:0'), 'train_loss_giou': tensor(0.0435, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4748, device='cuda:0'), 'validation_loss_ce': tensor(0.3539, device='cuda:0'), 'validation_loss_bbox': tensor(0.0120, device='cuda:0'), 'validation_loss_giou': tensor(0.0304, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4748, device='cuda:0'), 'train_loss_ce': tensor(0.3539, device='cuda:0'), 'train_loss_bbox': tensor(0.0120, device='cuda:0'), 'train_loss_giou': tensor(0.0304, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4469, device='cuda:0'), 'validation_loss_ce': tensor(0.3517, device='cuda:0'), 'validation_loss_bbox': tensor(0.0061, device='cuda:0'), 'validation_loss_giou': tensor(0.0324, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4469, device='cuda:0'), 'train_loss_ce': tensor(0.3517, device='cuda:0'), 'train_loss_bbox': tensor(0.0061, device='cuda:0'), 'train_loss_giou': tensor(0.0324, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4906, device='cuda:0'), 'validation_loss_ce': tensor(0.3500, device='cuda:0'), 'validation_loss_bbox': tensor(0.0099, device='cuda:0'), 'validation_loss_giou': tensor(0.0456, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4906, device='cuda:0'), 'train_loss_ce': tensor(0.3500, device='cuda:0'), 'train_loss_bbox': tensor(0.0099, device='cuda:0'), 'train_loss_giou': tensor(0.0456, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4591, device='cuda:0'), 'validation_loss_ce': tensor(0.3482, device='cuda:0'), 'validation_loss_bbox': tensor(0.0088, device='cuda:0'), 'validation_loss_giou': tensor(0.0335, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4591, device='cuda:0'), 'train_loss_ce': tensor(0.3482, device='cuda:0'), 'train_loss_bbox': tensor(0.0088, device='cuda:0'), 'train_loss_giou': tensor(0.0335, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4272, device='cuda:0'), 'validation_loss_ce': tensor(0.3466, device='cuda:0'), 'validation_loss_bbox': tensor(0.0064, device='cuda:0'), 'validation_loss_giou': tensor(0.0242, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4272, device='cuda:0'), 'train_loss_ce': tensor(0.3466, device='cuda:0'), 'train_loss_bbox': tensor(0.0064, device='cuda:0'), 'train_loss_giou': tensor(0.0242, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4671, device='cuda:0'), 'validation_loss_ce': tensor(0.3447, device='cuda:0'), 'validation_loss_bbox': tensor(0.0106, device='cuda:0'), 'validation_loss_giou': tensor(0.0348, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4671, device='cuda:0'), 'train_loss_ce': tensor(0.3447, device='cuda:0'), 'train_loss_bbox': tensor(0.0106, device='cuda:0'), 'train_loss_giou': tensor(0.0348, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4188, device='cuda:0'), 'validation_loss_ce': tensor(0.3429, device='cuda:0'), 'validation_loss_bbox': tensor(0.0088, device='cuda:0'), 'validation_loss_giou': tensor(0.0160, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4188, device='cuda:0'), 'train_loss_ce': tensor(0.3429, device='cuda:0'), 'train_loss_bbox': tensor(0.0088, device='cuda:0'), 'train_loss_giou': tensor(0.0160, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4382, device='cuda:0'), 'validation_loss_ce': tensor(0.3411, device='cuda:0'), 'validation_loss_bbox': tensor(0.0072, device='cuda:0'), 'validation_loss_giou': tensor(0.0306, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4382, device='cuda:0'), 'train_loss_ce': tensor(0.3411, device='cuda:0'), 'train_loss_bbox': tensor(0.0072, device='cuda:0'), 'train_loss_giou': tensor(0.0306, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4394, device='cuda:0'), 'validation_loss_ce': tensor(0.3396, device='cuda:0'), 'validation_loss_bbox': tensor(0.0085, device='cuda:0'), 'validation_loss_giou': tensor(0.0286, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4394, device='cuda:0'), 'train_loss_ce': tensor(0.3396, device='cuda:0'), 'train_loss_bbox': tensor(0.0085, device='cuda:0'), 'train_loss_giou': tensor(0.0286, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4583, device='cuda:0'), 'validation_loss_ce': tensor(0.3380, device='cuda:0'), 'validation_loss_bbox': tensor(0.0111, device='cuda:0'), 'validation_loss_giou': tensor(0.0323, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4583, device='cuda:0'), 'train_loss_ce': tensor(0.3380, device='cuda:0'), 'train_loss_bbox': tensor(0.0111, device='cuda:0'), 'train_loss_giou': tensor(0.0323, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.5158, device='cuda:0'), 'validation_loss_ce': tensor(0.3363, device='cuda:0'), 'validation_loss_bbox': tensor(0.0167, device='cuda:0'), 'validation_loss_giou': tensor(0.0480, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.5158, device='cuda:0'), 'train_loss_ce': tensor(0.3363, device='cuda:0'), 'train_loss_bbox': tensor(0.0167, device='cuda:0'), 'train_loss_giou': tensor(0.0480, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4768, device='cuda:0'), 'validation_loss_ce': tensor(0.3341, device='cuda:0'), 'validation_loss_bbox': tensor(0.0148, device='cuda:0'), 'validation_loss_giou': tensor(0.0342, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4768, device='cuda:0'), 'train_loss_ce': tensor(0.3341, device='cuda:0'), 'train_loss_bbox': tensor(0.0148, device='cuda:0'), 'train_loss_giou': tensor(0.0342, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4106, device='cuda:0'), 'validation_loss_ce': tensor(0.3318, device='cuda:0'), 'validation_loss_bbox': tensor(0.0076, device='cuda:0'), 'validation_loss_giou': tensor(0.0206, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4106, device='cuda:0'), 'train_loss_ce': tensor(0.3318, device='cuda:0'), 'train_loss_bbox': tensor(0.0076, device='cuda:0'), 'train_loss_giou': tensor(0.0206, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.3909, device='cuda:0'), 'validation_loss_ce': tensor(0.3296, device='cuda:0'), 'validation_loss_bbox': tensor(0.0057, device='cuda:0'), 'validation_loss_giou': tensor(0.0164, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.3909, device='cuda:0'), 'train_loss_ce': tensor(0.3296, device='cuda:0'), 'train_loss_bbox': tensor(0.0057, device='cuda:0'), 'train_loss_giou': tensor(0.0164, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4470, device='cuda:0'), 'validation_loss_ce': tensor(0.3276, device='cuda:0'), 'validation_loss_bbox': tensor(0.0112, device='cuda:0'), 'validation_loss_giou': tensor(0.0316, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4470, device='cuda:0'), 'train_loss_ce': tensor(0.3276, device='cuda:0'), 'train_loss_bbox': tensor(0.0112, device='cuda:0'), 'train_loss_giou': tensor(0.0316, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4419, device='cuda:0'), 'validation_loss_ce': tensor(0.3257, device='cuda:0'), 'validation_loss_bbox': tensor(0.0101, device='cuda:0'), 'validation_loss_giou': tensor(0.0328, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(0.4419, device='cuda:0'), 'train_loss_ce': tensor(0.3257, device='cuda:0'), 'train_loss_bbox': tensor(0.0101, device='cuda:0'), 'train_loss_giou': tensor(0.0328, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(0.4000, device='cuda:0'), 'validation_loss_ce': tensor(0.3241, device='cuda:0'), 'validation_loss_bbox': tensor(0.0051, device='cuda:0'), 'validation_loss_giou': tensor(0.0252, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
 ```
 
 ## Examples
-{'size': tensor([800, 800]), 'image_id': tensor([89]), 'class_labels': tensor([4]), 'boxes': tensor([[0.2217, 0.6439, 0.2169, 0.0921]]), 'area': tensor([12781.7246]), 'iscrowd': tensor([0]), 'orig_size': tensor([1024, 1024])}
+{'size': tensor([800, 800]), 'image_id': tensor([282]), 'class_labels': tensor([6]), 'boxes': tensor([[0.5497, 0.6339, 0.5344, 0.4466]]), 'area': tensor([152728.5312]), 'iscrowd': tensor([0]), 'orig_size': tensor([1024, 1024])}
 
 ![Example](./example.png)