Update README file

README.md

@@ -23,83 +23,83 @@ IoU metric: bbox
 ## After training result
 ```
 IoU metric: bbox
- Average Precision  (AP) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.013
- Average Precision  (AP) @[ IoU=0.50      | area=   all | maxDets=100 ] = 0.030
- Average Precision  (AP) @[ IoU=0.75      | area=   all | maxDets=100 ] = 0.006
+ Average Precision  (AP) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.000
+ Average Precision  (AP) @[ IoU=0.50      | area=   all | maxDets=100 ] = 0.001
+ Average Precision  (AP) @[ IoU=0.75      | area=   all | maxDets=100 ] = 0.000
  Average Precision  (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
  Average Precision  (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.000
- Average Precision  (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.013
- Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=  1 ] = 0.071
- Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets= 10 ] = 0.141
- Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.159
+ Average Precision  (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.000
+ Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=  1 ] = 0.001
+ Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets= 10 ] = 0.019
+ Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.021
  Average Recall     (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = -1.000
  Average Recall     (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.000
- Average Recall     (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.164
+ Average Recall     (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.021
 ```
 
 ## Config
 - dataset: NIH
 - original model: hustvl/yolos-tiny
-- lr: 0.0001
+- lr: 1e-06
 - dropout_rate: 0.1
-- weight_decay: 0.001
+- weight_decay: 0.05
 - max_epochs: 50
 - train samples: 885
 
 ## Logging
 ### Training process
 ```
-{'validation_loss': tensor(6.9980, device='cuda:0'), 'validation_loss_ce': tensor(2.0774, device='cuda:0'), 'validation_loss_bbox': tensor(0.5764, device='cuda:0'), 'validation_loss_giou': tensor(1.0194, device='cuda:0'), 'validation_cardinality_error': tensor(96.1875, device='cuda:0')}
-{'training_loss': tensor(1.9591, device='cuda:0'), 'train_loss_ce': tensor(0.3168, device='cuda:0'), 'train_loss_bbox': tensor(0.1506, device='cuda:0'), 'train_loss_giou': tensor(0.4447, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.3060, device='cuda:0'), 'validation_loss_ce': tensor(0.4328, device='cuda:0'), 'validation_loss_bbox': tensor(0.1669, device='cuda:0'), 'validation_loss_giou': tensor(0.5192, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.6421, device='cuda:0'), 'train_loss_ce': tensor(0.4176, device='cuda:0'), 'train_loss_bbox': tensor(0.2296, device='cuda:0'), 'train_loss_giou': tensor(0.5382, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.3129, device='cuda:0'), 'validation_loss_ce': tensor(0.4220, device='cuda:0'), 'validation_loss_bbox': tensor(0.1653, device='cuda:0'), 'validation_loss_giou': tensor(0.5321, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.0230, device='cuda:0'), 'train_loss_ce': tensor(0.3812, device='cuda:0'), 'train_loss_bbox': tensor(0.1228, device='cuda:0'), 'train_loss_giou': tensor(0.5139, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.5272, device='cuda:0'), 'validation_loss_ce': tensor(0.4293, device='cuda:0'), 'validation_loss_bbox': tensor(0.1835, device='cuda:0'), 'validation_loss_giou': tensor(0.5901, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.9008, device='cuda:0'), 'train_loss_ce': tensor(0.4246, device='cuda:0'), 'train_loss_bbox': tensor(0.1294, device='cuda:0'), 'train_loss_giou': tensor(0.4146, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.3039, device='cuda:0'), 'validation_loss_ce': tensor(0.4134, device='cuda:0'), 'validation_loss_bbox': tensor(0.1584, device='cuda:0'), 'validation_loss_giou': tensor(0.5492, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.1614, device='cuda:0'), 'train_loss_ce': tensor(0.4317, device='cuda:0'), 'train_loss_bbox': tensor(0.1515, device='cuda:0'), 'train_loss_giou': tensor(0.4861, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.3038, device='cuda:0'), 'validation_loss_ce': tensor(0.4126, device='cuda:0'), 'validation_loss_bbox': tensor(0.1583, device='cuda:0'), 'validation_loss_giou': tensor(0.5497, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.5595, device='cuda:0'), 'train_loss_ce': tensor(0.4053, device='cuda:0'), 'train_loss_bbox': tensor(0.0866, device='cuda:0'), 'train_loss_giou': tensor(0.3605, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.1651, device='cuda:0'), 'validation_loss_ce': tensor(0.4196, device='cuda:0'), 'validation_loss_bbox': tensor(0.1478, device='cuda:0'), 'validation_loss_giou': tensor(0.5031, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.0434, device='cuda:0'), 'train_loss_ce': tensor(0.3087, device='cuda:0'), 'train_loss_bbox': tensor(0.1553, device='cuda:0'), 'train_loss_giou': tensor(0.4792, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.2701, device='cuda:0'), 'validation_loss_ce': tensor(0.3932, device='cuda:0'), 'validation_loss_bbox': tensor(0.1609, device='cuda:0'), 'validation_loss_giou': tensor(0.5361, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.3151, device='cuda:0'), 'train_loss_ce': tensor(0.3850, device='cuda:0'), 'train_loss_bbox': tensor(0.1603, device='cuda:0'), 'train_loss_giou': tensor(0.5642, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.7141, device='cuda:0'), 'validation_loss_ce': tensor(0.4040, device='cuda:0'), 'validation_loss_bbox': tensor(0.2126, device='cuda:0'), 'validation_loss_giou': tensor(0.6236, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.7767, device='cuda:0'), 'train_loss_ce': tensor(0.4482, device='cuda:0'), 'train_loss_bbox': tensor(0.2168, device='cuda:0'), 'train_loss_giou': tensor(0.6224, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.4795, device='cuda:0'), 'validation_loss_ce': tensor(0.4138, device='cuda:0'), 'validation_loss_bbox': tensor(0.1788, device='cuda:0'), 'validation_loss_giou': tensor(0.5857, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.5112, device='cuda:0'), 'train_loss_ce': tensor(0.4653, device='cuda:0'), 'train_loss_bbox': tensor(0.1600, device='cuda:0'), 'train_loss_giou': tensor(0.6229, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.5765, device='cuda:0'), 'validation_loss_ce': tensor(0.4042, device='cuda:0'), 'validation_loss_bbox': tensor(0.1957, device='cuda:0'), 'validation_loss_giou': tensor(0.5970, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.2503, device='cuda:0'), 'train_loss_ce': tensor(0.4681, device='cuda:0'), 'train_loss_bbox': tensor(0.1526, device='cuda:0'), 'train_loss_giou': tensor(0.5096, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.5902, device='cuda:0'), 'validation_loss_ce': tensor(0.4213, device='cuda:0'), 'validation_loss_bbox': tensor(0.1939, device='cuda:0'), 'validation_loss_giou': tensor(0.5997, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.9702, device='cuda:0'), 'train_loss_ce': tensor(0.4538, device='cuda:0'), 'train_loss_bbox': tensor(0.2257, device='cuda:0'), 'train_loss_giou': tensor(0.6939, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.5045, device='cuda:0'), 'validation_loss_ce': tensor(0.4158, device='cuda:0'), 'validation_loss_bbox': tensor(0.1966, device='cuda:0'), 'validation_loss_giou': tensor(0.5528, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.1300, device='cuda:0'), 'train_loss_ce': tensor(0.3665, device='cuda:0'), 'train_loss_bbox': tensor(0.1390, device='cuda:0'), 'train_loss_giou': tensor(0.5341, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.3374, device='cuda:0'), 'validation_loss_ce': tensor(0.4036, device='cuda:0'), 'validation_loss_bbox': tensor(0.1654, device='cuda:0'), 'validation_loss_giou': tensor(0.5534, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.2715, device='cuda:0'), 'train_loss_ce': tensor(0.4282, device='cuda:0'), 'train_loss_bbox': tensor(0.1883, device='cuda:0'), 'train_loss_giou': tensor(0.4509, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.3349, device='cuda:0'), 'validation_loss_ce': tensor(0.3914, device='cuda:0'), 'validation_loss_bbox': tensor(0.1709, device='cuda:0'), 'validation_loss_giou': tensor(0.5445, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.7526, device='cuda:0'), 'train_loss_ce': tensor(0.3276, device='cuda:0'), 'train_loss_bbox': tensor(0.1172, device='cuda:0'), 'train_loss_giou': tensor(0.4195, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.2411, device='cuda:0'), 'validation_loss_ce': tensor(0.3867, device='cuda:0'), 'validation_loss_bbox': tensor(0.1625, device='cuda:0'), 'validation_loss_giou': tensor(0.5210, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.7417, device='cuda:0'), 'train_loss_ce': tensor(0.3457, device='cuda:0'), 'train_loss_bbox': tensor(0.1130, device='cuda:0'), 'train_loss_giou': tensor(0.4154, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.3407, device='cuda:0'), 'validation_loss_ce': tensor(0.3910, device='cuda:0'), 'validation_loss_bbox': tensor(0.1646, device='cuda:0'), 'validation_loss_giou': tensor(0.5634, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.7487, device='cuda:0'), 'train_loss_ce': tensor(0.3850, device='cuda:0'), 'train_loss_bbox': tensor(0.1692, device='cuda:0'), 'train_loss_giou': tensor(0.7589, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.4066, device='cuda:0'), 'validation_loss_ce': tensor(0.4094, device='cuda:0'), 'validation_loss_bbox': tensor(0.1719, device='cuda:0'), 'validation_loss_giou': tensor(0.5690, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.7448, device='cuda:0'), 'train_loss_ce': tensor(0.4563, device='cuda:0'), 'train_loss_bbox': tensor(0.2001, device='cuda:0'), 'train_loss_giou': tensor(0.6441, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.2312, device='cuda:0'), 'validation_loss_ce': tensor(0.3842, device='cuda:0'), 'validation_loss_bbox': tensor(0.1588, device='cuda:0'), 'validation_loss_giou': tensor(0.5265, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.9861, device='cuda:0'), 'train_loss_ce': tensor(0.4062, device='cuda:0'), 'train_loss_bbox': tensor(0.1622, device='cuda:0'), 'train_loss_giou': tensor(0.3844, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.2770, device='cuda:0'), 'validation_loss_ce': tensor(0.3843, device='cuda:0'), 'validation_loss_bbox': tensor(0.1617, device='cuda:0'), 'validation_loss_giou': tensor(0.5421, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.4871, device='cuda:0'), 'train_loss_ce': tensor(0.3550, device='cuda:0'), 'train_loss_bbox': tensor(0.1734, device='cuda:0'), 'train_loss_giou': tensor(0.6325, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.3588, device='cuda:0'), 'validation_loss_ce': tensor(0.3915, device='cuda:0'), 'validation_loss_bbox': tensor(0.1721, device='cuda:0'), 'validation_loss_giou': tensor(0.5533, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.4463, device='cuda:0'), 'train_loss_ce': tensor(0.4212, device='cuda:0'), 'train_loss_bbox': tensor(0.1563, device='cuda:0'), 'train_loss_giou': tensor(0.6217, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.1789, device='cuda:0'), 'validation_loss_ce': tensor(0.3907, device='cuda:0'), 'validation_loss_bbox': tensor(0.1505, device='cuda:0'), 'validation_loss_giou': tensor(0.5179, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.8296, device='cuda:0'), 'train_loss_ce': tensor(0.3541, device='cuda:0'), 'train_loss_bbox': tensor(0.1274, device='cuda:0'), 'train_loss_giou': tensor(0.4192, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.2105, device='cuda:0'), 'validation_loss_ce': tensor(0.3775, device='cuda:0'), 'validation_loss_bbox': tensor(0.1559, device='cuda:0'), 'validation_loss_giou': tensor(0.5268, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.3388, device='cuda:0'), 'train_loss_ce': tensor(0.3709, device='cuda:0'), 'train_loss_bbox': tensor(0.1459, device='cuda:0'), 'train_loss_giou': tensor(0.6191, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.1570, device='cuda:0'), 'validation_loss_ce': tensor(0.3723, device='cuda:0'), 'validation_loss_bbox': tensor(0.1484, device='cuda:0'), 'validation_loss_giou': tensor(0.5215, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.2575, device='cuda:0'), 'train_loss_ce': tensor(0.4642, device='cuda:0'), 'train_loss_bbox': tensor(0.1563, device='cuda:0'), 'train_loss_giou': tensor(0.5059, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.2539, device='cuda:0'), 'validation_loss_ce': tensor(0.3751, device='cuda:0'), 'validation_loss_bbox': tensor(0.1614, device='cuda:0'), 'validation_loss_giou': tensor(0.5358, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.3852, device='cuda:0'), 'train_loss_ce': tensor(0.4486, device='cuda:0'), 'train_loss_bbox': tensor(0.1840, device='cuda:0'), 'train_loss_giou': tensor(0.5083, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.2098, device='cuda:0'), 'validation_loss_ce': tensor(0.3731, device='cuda:0'), 'validation_loss_bbox': tensor(0.1589, device='cuda:0'), 'validation_loss_giou': tensor(0.5212, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.0163, device='cuda:0'), 'train_loss_ce': tensor(0.3895, device='cuda:0'), 'train_loss_bbox': tensor(0.1363, device='cuda:0'), 'train_loss_giou': tensor(0.4726, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.1935, device='cuda:0'), 'validation_loss_ce': tensor(0.3829, device='cuda:0'), 'validation_loss_bbox': tensor(0.1517, device='cuda:0'), 'validation_loss_giou': tensor(0.5262, device='cuda:0'), 'validation_cardinality_error': tensor(0.8788, device='cuda:0')}
-{'training_loss': tensor(1.8352, device='cuda:0'), 'train_loss_ce': tensor(0.3180, device='cuda:0'), 'train_loss_bbox': tensor(0.1492, device='cuda:0'), 'train_loss_giou': tensor(0.3855, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.2605, device='cuda:0'), 'validation_loss_ce': tensor(0.3676, device='cuda:0'), 'validation_loss_bbox': tensor(0.1609, device='cuda:0'), 'validation_loss_giou': tensor(0.5442, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.3223, device='cuda:0'), 'train_loss_ce': tensor(0.3044, device='cuda:0'), 'train_loss_bbox': tensor(0.1013, device='cuda:0'), 'train_loss_giou': tensor(0.2556, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.2185, device='cuda:0'), 'validation_loss_ce': tensor(0.3744, device='cuda:0'), 'validation_loss_bbox': tensor(0.1587, device='cuda:0'), 'validation_loss_giou': tensor(0.5252, device='cuda:0'), 'validation_cardinality_error': tensor(0.9495, device='cuda:0')}
-{'training_loss': tensor(1.9661, device='cuda:0'), 'train_loss_ce': tensor(0.3483, device='cuda:0'), 'train_loss_bbox': tensor(0.1175, device='cuda:0'), 'train_loss_giou': tensor(0.5150, device='cuda:0'), 'train_cardinality_error': tensor(0.8000, device='cuda:0'), 'validation_loss': tensor(2.2522, device='cuda:0'), 'validation_loss_ce': tensor(0.3802, device='cuda:0'), 'validation_loss_bbox': tensor(0.1586, device='cuda:0'), 'validation_loss_giou': tensor(0.5394, device='cuda:0'), 'validation_cardinality_error': tensor(0.7778, device='cuda:0')}
-{'training_loss': tensor(2.3235, device='cuda:0'), 'train_loss_ce': tensor(0.3712, device='cuda:0'), 'train_loss_bbox': tensor(0.1939, device='cuda:0'), 'train_loss_giou': tensor(0.4915, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.1720, device='cuda:0'), 'validation_loss_ce': tensor(0.3719, device='cuda:0'), 'validation_loss_bbox': tensor(0.1527, device='cuda:0'), 'validation_loss_giou': tensor(0.5184, device='cuda:0'), 'validation_cardinality_error': tensor(0.8687, device='cuda:0')}
-{'training_loss': tensor(1.5666, device='cuda:0'), 'train_loss_ce': tensor(0.3918, device='cuda:0'), 'train_loss_bbox': tensor(0.0777, device='cuda:0'), 'train_loss_giou': tensor(0.3931, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.0891, device='cuda:0'), 'validation_loss_ce': tensor(0.3705, device='cuda:0'), 'validation_loss_bbox': tensor(0.1437, device='cuda:0'), 'validation_loss_giou': tensor(0.4999, device='cuda:0'), 'validation_cardinality_error': tensor(0.9899, device='cuda:0')}
-{'training_loss': tensor(2.0268, device='cuda:0'), 'train_loss_ce': tensor(0.3945, device='cuda:0'), 'train_loss_bbox': tensor(0.1465, device='cuda:0'), 'train_loss_giou': tensor(0.4499, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.1654, device='cuda:0'), 'validation_loss_ce': tensor(0.3704, device='cuda:0'), 'validation_loss_bbox': tensor(0.1474, device='cuda:0'), 'validation_loss_giou': tensor(0.5291, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.0776, device='cuda:0'), 'train_loss_ce': tensor(0.4210, device='cuda:0'), 'train_loss_bbox': tensor(0.1471, device='cuda:0'), 'train_loss_giou': tensor(0.4607, device='cuda:0'), 'train_cardinality_error': tensor(0.4000, device='cuda:0'), 'validation_loss': tensor(2.1170, device='cuda:0'), 'validation_loss_ce': tensor(0.3763, device='cuda:0'), 'validation_loss_bbox': tensor(0.1431, device='cuda:0'), 'validation_loss_giou': tensor(0.5127, device='cuda:0'), 'validation_cardinality_error': tensor(0.3232, device='cuda:0')}
-{'training_loss': tensor(1.4932, device='cuda:0'), 'train_loss_ce': tensor(0.3657, device='cuda:0'), 'train_loss_bbox': tensor(0.0990, device='cuda:0'), 'train_loss_giou': tensor(0.3163, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.0869, device='cuda:0'), 'validation_loss_ce': tensor(0.3713, device='cuda:0'), 'validation_loss_bbox': tensor(0.1457, device='cuda:0'), 'validation_loss_giou': tensor(0.4936, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.8174, device='cuda:0'), 'train_loss_ce': tensor(0.3460, device='cuda:0'), 'train_loss_bbox': tensor(0.1162, device='cuda:0'), 'train_loss_giou': tensor(0.4452, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.0891, device='cuda:0'), 'validation_loss_ce': tensor(0.3620, device='cuda:0'), 'validation_loss_bbox': tensor(0.1402, device='cuda:0'), 'validation_loss_giou': tensor(0.5130, device='cuda:0'), 'validation_cardinality_error': tensor(0.9596, device='cuda:0')}
-{'training_loss': tensor(2.1190, device='cuda:0'), 'train_loss_ce': tensor(0.3552, device='cuda:0'), 'train_loss_bbox': tensor(0.1349, device='cuda:0'), 'train_loss_giou': tensor(0.5447, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.0378, device='cuda:0'), 'validation_loss_ce': tensor(0.3590, device='cuda:0'), 'validation_loss_bbox': tensor(0.1388, device='cuda:0'), 'validation_loss_giou': tensor(0.4923, device='cuda:0'), 'validation_cardinality_error': tensor(0.9899, device='cuda:0')}
-{'training_loss': tensor(1.8194, device='cuda:0'), 'train_loss_ce': tensor(0.3136, device='cuda:0'), 'train_loss_bbox': tensor(0.1044, device='cuda:0'), 'train_loss_giou': tensor(0.4920, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.0059, device='cuda:0'), 'validation_loss_ce': tensor(0.3623, device='cuda:0'), 'validation_loss_bbox': tensor(0.1370, device='cuda:0'), 'validation_loss_giou': tensor(0.4794, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(2.0199, device='cuda:0'), 'train_loss_ce': tensor(0.3735, device='cuda:0'), 'train_loss_bbox': tensor(0.1110, device='cuda:0'), 'train_loss_giou': tensor(0.5458, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.0734, device='cuda:0'), 'validation_loss_ce': tensor(0.3659, device='cuda:0'), 'validation_loss_bbox': tensor(0.1376, device='cuda:0'), 'validation_loss_giou': tensor(0.5096, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.5889, device='cuda:0'), 'train_loss_ce': tensor(0.3230, device='cuda:0'), 'train_loss_bbox': tensor(0.0800, device='cuda:0'), 'train_loss_giou': tensor(0.4330, device='cuda:0'), 'train_cardinality_error': tensor(0.8000, device='cuda:0'), 'validation_loss': tensor(1.9917, device='cuda:0'), 'validation_loss_ce': tensor(0.3621, device='cuda:0'), 'validation_loss_bbox': tensor(0.1349, device='cuda:0'), 'validation_loss_giou': tensor(0.4775, device='cuda:0'), 'validation_cardinality_error': tensor(0.8788, device='cuda:0')}
-{'training_loss': tensor(1.5126, device='cuda:0'), 'train_loss_ce': tensor(0.3984, device='cuda:0'), 'train_loss_bbox': tensor(0.0810, device='cuda:0'), 'train_loss_giou': tensor(0.3545, device='cuda:0'), 'train_cardinality_error': tensor(0.2000, device='cuda:0'), 'validation_loss': tensor(2.0611, device='cuda:0'), 'validation_loss_ce': tensor(0.3687, device='cuda:0'), 'validation_loss_bbox': tensor(0.1368, device='cuda:0'), 'validation_loss_giou': tensor(0.5041, device='cuda:0'), 'validation_cardinality_error': tensor(0.7576, device='cuda:0')}
-{'training_loss': tensor(1.8280, device='cuda:0'), 'train_loss_ce': tensor(0.4029, device='cuda:0'), 'train_loss_bbox': tensor(0.0999, device='cuda:0'), 'train_loss_giou': tensor(0.4629, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.0638, device='cuda:0'), 'validation_loss_ce': tensor(0.3683, device='cuda:0'), 'validation_loss_bbox': tensor(0.1420, device='cuda:0'), 'validation_loss_giou': tensor(0.4927, device='cuda:0'), 'validation_cardinality_error': tensor(0.9899, device='cuda:0')}
-{'training_loss': tensor(1.7057, device='cuda:0'), 'train_loss_ce': tensor(0.4136, device='cuda:0'), 'train_loss_bbox': tensor(0.0915, device='cuda:0'), 'train_loss_giou': tensor(0.4172, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.0228, device='cuda:0'), 'validation_loss_ce': tensor(0.3651, device='cuda:0'), 'validation_loss_bbox': tensor(0.1387, device='cuda:0'), 'validation_loss_giou': tensor(0.4822, device='cuda:0'), 'validation_cardinality_error': tensor(0.9596, device='cuda:0')}
-{'training_loss': tensor(1.5449, device='cuda:0'), 'train_loss_ce': tensor(0.3997, device='cuda:0'), 'train_loss_bbox': tensor(0.0768, device='cuda:0'), 'train_loss_giou': tensor(0.3806, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.0353, device='cuda:0'), 'validation_loss_ce': tensor(0.3622, device='cuda:0'), 'validation_loss_bbox': tensor(0.1349, device='cuda:0'), 'validation_loss_giou': tensor(0.4992, device='cuda:0'), 'validation_cardinality_error': tensor(0.7879, device='cuda:0')}
-{'training_loss': tensor(1.7316, device='cuda:0'), 'train_loss_ce': tensor(0.4420, device='cuda:0'), 'train_loss_bbox': tensor(0.0924, device='cuda:0'), 'train_loss_giou': tensor(0.4137, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.0308, device='cuda:0'), 'validation_loss_ce': tensor(0.3487, device='cuda:0'), 'validation_loss_bbox': tensor(0.1383, device='cuda:0'), 'validation_loss_giou': tensor(0.4952, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
-{'training_loss': tensor(1.7057, device='cuda:0'), 'train_loss_ce': tensor(0.4140, device='cuda:0'), 'train_loss_bbox': tensor(0.1117, device='cuda:0'), 'train_loss_giou': tensor(0.3667, device='cuda:0'), 'train_cardinality_error': tensor(0.6000, device='cuda:0'), 'validation_loss': tensor(1.9358, device='cuda:0'), 'validation_loss_ce': tensor(0.3484, device='cuda:0'), 'validation_loss_bbox': tensor(0.1280, device='cuda:0'), 'validation_loss_giou': tensor(0.4737, device='cuda:0'), 'validation_cardinality_error': tensor(0.4545, device='cuda:0')}
-{'training_loss': tensor(2.1335, device='cuda:0'), 'train_loss_ce': tensor(0.4015, device='cuda:0'), 'train_loss_bbox': tensor(0.1885, device='cuda:0'), 'train_loss_giou': tensor(0.3948, device='cuda:0'), 'train_cardinality_error': tensor(0.6000, device='cuda:0'), 'validation_loss': tensor(1.9711, device='cuda:0'), 'validation_loss_ce': tensor(0.3577, device='cuda:0'), 'validation_loss_bbox': tensor(0.1309, device='cuda:0'), 'validation_loss_giou': tensor(0.4795, device='cuda:0'), 'validation_cardinality_error': tensor(0.8990, device='cuda:0')}
-{'training_loss': tensor(2.0193, device='cuda:0'), 'train_loss_ce': tensor(0.2849, device='cuda:0'), 'train_loss_bbox': tensor(0.1473, device='cuda:0'), 'train_loss_giou': tensor(0.4989, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.9930, device='cuda:0'), 'validation_loss_ce': tensor(0.3653, device='cuda:0'), 'validation_loss_bbox': tensor(0.1354, device='cuda:0'), 'validation_loss_giou': tensor(0.4753, device='cuda:0'), 'validation_cardinality_error': tensor(0.9899, device='cuda:0')}
-{'training_loss': tensor(1.7410, device='cuda:0'), 'train_loss_ce': tensor(0.3670, device='cuda:0'), 'train_loss_bbox': tensor(0.1155, device='cuda:0'), 'train_loss_giou': tensor(0.3983, device='cuda:0'), 'train_cardinality_error': tensor(0.8000, device='cuda:0'), 'validation_loss': tensor(2.0600, device='cuda:0'), 'validation_loss_ce': tensor(0.3666, device='cuda:0'), 'validation_loss_bbox': tensor(0.1421, device='cuda:0'), 'validation_loss_giou': tensor(0.4914, device='cuda:0'), 'validation_cardinality_error': tensor(0.8687, device='cuda:0')}
-{'training_loss': tensor(1.7164, device='cuda:0'), 'train_loss_ce': tensor(0.3726, device='cuda:0'), 'train_loss_bbox': tensor(0.1148, device='cuda:0'), 'train_loss_giou': tensor(0.3849, device='cuda:0'), 'train_cardinality_error': tensor(0.4000, device='cuda:0'), 'validation_loss': tensor(1.9918, device='cuda:0'), 'validation_loss_ce': tensor(0.3566, device='cuda:0'), 'validation_loss_bbox': tensor(0.1360, device='cuda:0'), 'validation_loss_giou': tensor(0.4777, device='cuda:0'), 'validation_cardinality_error': tensor(0.5354, device='cuda:0')}
-{'training_loss': tensor(2.4173, device='cuda:0'), 'train_loss_ce': tensor(0.2800, device='cuda:0'), 'train_loss_bbox': tensor(0.1449, device='cuda:0'), 'train_loss_giou': tensor(0.7064, device='cuda:0'), 'train_cardinality_error': tensor(0.6000, device='cuda:0'), 'validation_loss': tensor(2.0160, device='cuda:0'), 'validation_loss_ce': tensor(0.3560, device='cuda:0'), 'validation_loss_bbox': tensor(0.1355, device='cuda:0'), 'validation_loss_giou': tensor(0.4911, device='cuda:0'), 'validation_cardinality_error': tensor(0.5859, device='cuda:0')}
+{'validation_loss': tensor(7.1565, device='cuda:0'), 'validation_loss_ce': tensor(2.6168, device='cuda:0'), 'validation_loss_bbox': tensor(0.5211, device='cuda:0'), 'validation_loss_giou': tensor(0.9671, device='cuda:0'), 'validation_cardinality_error': tensor(99., device='cuda:0')}
+{'training_loss': tensor(6.0705, device='cuda:0'), 'train_loss_ce': tensor(2.5302, device='cuda:0'), 'train_loss_bbox': tensor(0.4304, device='cuda:0'), 'train_loss_giou': tensor(0.6943, device='cuda:0'), 'train_cardinality_error': tensor(99., device='cuda:0'), 'validation_loss': tensor(6.3058, device='cuda:0'), 'validation_loss_ce': tensor(2.4960, device='cuda:0'), 'validation_loss_bbox': tensor(0.4287, device='cuda:0'), 'validation_loss_giou': tensor(0.8332, device='cuda:0'), 'validation_cardinality_error': tensor(98.8889, device='cuda:0')}
+{'training_loss': tensor(5.3459, device='cuda:0'), 'train_loss_ce': tensor(2.4056, device='cuda:0'), 'train_loss_bbox': tensor(0.3181, device='cuda:0'), 'train_loss_giou': tensor(0.6750, device='cuda:0'), 'train_cardinality_error': tensor(98.6000, device='cuda:0'), 'validation_loss': tensor(5.6501, device='cuda:0'), 'validation_loss_ce': tensor(2.3928, device='cuda:0'), 'validation_loss_bbox': tensor(0.3429, device='cuda:0'), 'validation_loss_giou': tensor(0.7713, device='cuda:0'), 'validation_cardinality_error': tensor(98.2525, device='cuda:0')}
+{'training_loss': tensor(5.3122, device='cuda:0'), 'train_loss_ce': tensor(2.2708, device='cuda:0'), 'train_loss_bbox': tensor(0.3198, device='cuda:0'), 'train_loss_giou': tensor(0.7211, device='cuda:0'), 'train_cardinality_error': tensor(97.8000, device='cuda:0'), 'validation_loss': tensor(5.2498, device='cuda:0'), 'validation_loss_ce': tensor(2.2861, device='cuda:0'), 'validation_loss_bbox': tensor(0.3017, device='cuda:0'), 'validation_loss_giou': tensor(0.7276, device='cuda:0'), 'validation_cardinality_error': tensor(96.0101, device='cuda:0')}
+{'training_loss': tensor(3.9929, device='cuda:0'), 'train_loss_ce': tensor(2.1223, device='cuda:0'), 'train_loss_bbox': tensor(0.2062, device='cuda:0'), 'train_loss_giou': tensor(0.4198, device='cuda:0'), 'train_cardinality_error': tensor(84., device='cuda:0'), 'validation_loss': tensor(5.0342, device='cuda:0'), 'validation_loss_ce': tensor(2.1912, device='cuda:0'), 'validation_loss_bbox': tensor(0.2814, device='cuda:0'), 'validation_loss_giou': tensor(0.7179, device='cuda:0'), 'validation_cardinality_error': tensor(91., device='cuda:0')}
+{'training_loss': tensor(4.1406, device='cuda:0'), 'train_loss_ce': tensor(2.1059, device='cuda:0'), 'train_loss_bbox': tensor(0.2164, device='cuda:0'), 'train_loss_giou': tensor(0.4762, device='cuda:0'), 'train_cardinality_error': tensor(84., device='cuda:0'), 'validation_loss': tensor(4.8572, device='cuda:0'), 'validation_loss_ce': tensor(2.0894, device='cuda:0'), 'validation_loss_bbox': tensor(0.2727, device='cuda:0'), 'validation_loss_giou': tensor(0.7021, device='cuda:0'), 'validation_cardinality_error': tensor(78.2929, device='cuda:0')}
+{'training_loss': tensor(4.5332, device='cuda:0'), 'train_loss_ce': tensor(1.9598, device='cuda:0'), 'train_loss_bbox': tensor(0.2270, device='cuda:0'), 'train_loss_giou': tensor(0.7191, device='cuda:0'), 'train_cardinality_error': tensor(50.2000, device='cuda:0'), 'validation_loss': tensor(4.6922, device='cuda:0'), 'validation_loss_ce': tensor(1.9838, device='cuda:0'), 'validation_loss_bbox': tensor(0.2630, device='cuda:0'), 'validation_loss_giou': tensor(0.6967, device='cuda:0'), 'validation_cardinality_error': tensor(55.5758, device='cuda:0')}
+{'training_loss': tensor(4.4722, device='cuda:0'), 'train_loss_ce': tensor(1.7903, device='cuda:0'), 'train_loss_bbox': tensor(0.2521, device='cuda:0'), 'train_loss_giou': tensor(0.7108, device='cuda:0'), 'train_cardinality_error': tensor(8.2000, device='cuda:0'), 'validation_loss': tensor(4.5388, device='cuda:0'), 'validation_loss_ce': tensor(1.8704, device='cuda:0'), 'validation_loss_bbox': tensor(0.2569, device='cuda:0'), 'validation_loss_giou': tensor(0.6921, device='cuda:0'), 'validation_cardinality_error': tensor(30.3838, device='cuda:0')}
+{'training_loss': tensor(4.5265, device='cuda:0'), 'train_loss_ce': tensor(1.6432, device='cuda:0'), 'train_loss_bbox': tensor(0.2652, device='cuda:0'), 'train_loss_giou': tensor(0.7785, device='cuda:0'), 'train_cardinality_error': tensor(3., device='cuda:0'), 'validation_loss': tensor(4.3567, device='cuda:0'), 'validation_loss_ce': tensor(1.7606, device='cuda:0'), 'validation_loss_bbox': tensor(0.2432, device='cuda:0'), 'validation_loss_giou': tensor(0.6902, device='cuda:0'), 'validation_cardinality_error': tensor(13.5960, device='cuda:0')}
+{'training_loss': tensor(4.7480, device='cuda:0'), 'train_loss_ce': tensor(1.6500, device='cuda:0'), 'train_loss_bbox': tensor(0.2890, device='cuda:0'), 'train_loss_giou': tensor(0.8264, device='cuda:0'), 'train_cardinality_error': tensor(5., device='cuda:0'), 'validation_loss': tensor(4.2126, device='cuda:0'), 'validation_loss_ce': tensor(1.6448, device='cuda:0'), 'validation_loss_bbox': tensor(0.2399, device='cuda:0'), 'validation_loss_giou': tensor(0.6841, device='cuda:0'), 'validation_cardinality_error': tensor(4.7879, device='cuda:0')}
+{'training_loss': tensor(3.8881, device='cuda:0'), 'train_loss_ce': tensor(1.5876, device='cuda:0'), 'train_loss_bbox': tensor(0.2124, device='cuda:0'), 'train_loss_giou': tensor(0.6194, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(4.0765, device='cuda:0'), 'validation_loss_ce': tensor(1.5303, device='cuda:0'), 'validation_loss_bbox': tensor(0.2369, device='cuda:0'), 'validation_loss_giou': tensor(0.6807, device='cuda:0'), 'validation_cardinality_error': tensor(1.6667, device='cuda:0')}
+{'training_loss': tensor(4.0165, device='cuda:0'), 'train_loss_ce': tensor(1.3808, device='cuda:0'), 'train_loss_bbox': tensor(0.2712, device='cuda:0'), 'train_loss_giou': tensor(0.6398, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.9377, device='cuda:0'), 'validation_loss_ce': tensor(1.4131, device='cuda:0'), 'validation_loss_bbox': tensor(0.2350, device='cuda:0'), 'validation_loss_giou': tensor(0.6749, device='cuda:0'), 'validation_cardinality_error': tensor(0.9697, device='cuda:0')}
+{'training_loss': tensor(4.0245, device='cuda:0'), 'train_loss_ce': tensor(1.2671, device='cuda:0'), 'train_loss_bbox': tensor(0.2208, device='cuda:0'), 'train_loss_giou': tensor(0.8266, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.7913, device='cuda:0'), 'validation_loss_ce': tensor(1.2989, device='cuda:0'), 'validation_loss_bbox': tensor(0.2306, device='cuda:0'), 'validation_loss_giou': tensor(0.6698, device='cuda:0'), 'validation_cardinality_error': tensor(0.9697, device='cuda:0')}
+{'training_loss': tensor(3.4285, device='cuda:0'), 'train_loss_ce': tensor(1.1165, device='cuda:0'), 'train_loss_bbox': tensor(0.1885, device='cuda:0'), 'train_loss_giou': tensor(0.6848, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.6641, device='cuda:0'), 'validation_loss_ce': tensor(1.1870, device='cuda:0'), 'validation_loss_bbox': tensor(0.2271, device='cuda:0'), 'validation_loss_giou': tensor(0.6708, device='cuda:0'), 'validation_cardinality_error': tensor(0.9798, device='cuda:0')}
+{'training_loss': tensor(3.7639, device='cuda:0'), 'train_loss_ce': tensor(1.1571, device='cuda:0'), 'train_loss_bbox': tensor(0.2647, device='cuda:0'), 'train_loss_giou': tensor(0.6416, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.5421, device='cuda:0'), 'validation_loss_ce': tensor(1.0872, device='cuda:0'), 'validation_loss_bbox': tensor(0.2242, device='cuda:0'), 'validation_loss_giou': tensor(0.6668, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(2.9899, device='cuda:0'), 'train_loss_ce': tensor(0.9906, device='cuda:0'), 'train_loss_bbox': tensor(0.1988, device='cuda:0'), 'train_loss_giou': tensor(0.5028, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.4229, device='cuda:0'), 'validation_loss_ce': tensor(0.9928, device='cuda:0'), 'validation_loss_bbox': tensor(0.2208, device='cuda:0'), 'validation_loss_giou': tensor(0.6629, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(2.8417, device='cuda:0'), 'train_loss_ce': tensor(0.9287, device='cuda:0'), 'train_loss_bbox': tensor(0.1494, device='cuda:0'), 'train_loss_giou': tensor(0.5831, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.3267, device='cuda:0'), 'validation_loss_ce': tensor(0.9071, device='cuda:0'), 'validation_loss_bbox': tensor(0.2219, device='cuda:0'), 'validation_loss_giou': tensor(0.6550, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(2.9143, device='cuda:0'), 'train_loss_ce': tensor(0.8071, device='cuda:0'), 'train_loss_bbox': tensor(0.1854, device='cuda:0'), 'train_loss_giou': tensor(0.5901, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.2317, device='cuda:0'), 'validation_loss_ce': tensor(0.8366, device='cuda:0'), 'validation_loss_bbox': tensor(0.2197, device='cuda:0'), 'validation_loss_giou': tensor(0.6482, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(3.2213, device='cuda:0'), 'train_loss_ce': tensor(0.8347, device='cuda:0'), 'train_loss_bbox': tensor(0.2265, device='cuda:0'), 'train_loss_giou': tensor(0.6270, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.1624, device='cuda:0'), 'validation_loss_ce': tensor(0.7770, device='cuda:0'), 'validation_loss_bbox': tensor(0.2183, device='cuda:0'), 'validation_loss_giou': tensor(0.6469, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(2.1193, device='cuda:0'), 'train_loss_ce': tensor(0.6932, device='cuda:0'), 'train_loss_bbox': tensor(0.1298, device='cuda:0'), 'train_loss_giou': tensor(0.3885, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.0898, device='cuda:0'), 'validation_loss_ce': tensor(0.7267, device='cuda:0'), 'validation_loss_bbox': tensor(0.2196, device='cuda:0'), 'validation_loss_giou': tensor(0.6326, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(3.1823, device='cuda:0'), 'train_loss_ce': tensor(0.6994, device='cuda:0'), 'train_loss_bbox': tensor(0.2211, device='cuda:0'), 'train_loss_giou': tensor(0.6887, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.0247, device='cuda:0'), 'validation_loss_ce': tensor(0.6842, device='cuda:0'), 'validation_loss_bbox': tensor(0.2153, device='cuda:0'), 'validation_loss_giou': tensor(0.6321, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(2.7933, device='cuda:0'), 'train_loss_ce': tensor(0.6156, device='cuda:0'), 'train_loss_bbox': tensor(0.1934, device='cuda:0'), 'train_loss_giou': tensor(0.6054, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.9765, device='cuda:0'), 'validation_loss_ce': tensor(0.6504, device='cuda:0'), 'validation_loss_bbox': tensor(0.2157, device='cuda:0'), 'validation_loss_giou': tensor(0.6237, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(3.2350, device='cuda:0'), 'train_loss_ce': tensor(0.6573, device='cuda:0'), 'train_loss_bbox': tensor(0.2416, device='cuda:0'), 'train_loss_giou': tensor(0.6849, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.9092, device='cuda:0'), 'validation_loss_ce': tensor(0.6215, device='cuda:0'), 'validation_loss_bbox': tensor(0.2121, device='cuda:0'), 'validation_loss_giou': tensor(0.6135, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(3.0423, device='cuda:0'), 'train_loss_ce': tensor(0.6132, device='cuda:0'), 'train_loss_bbox': tensor(0.1892, device='cuda:0'), 'train_loss_giou': tensor(0.7415, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.9006, device='cuda:0'), 'validation_loss_ce': tensor(0.5973, device='cuda:0'), 'validation_loss_bbox': tensor(0.2138, device='cuda:0'), 'validation_loss_giou': tensor(0.6170, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(2.2000, device='cuda:0'), 'train_loss_ce': tensor(0.5179, device='cuda:0'), 'train_loss_bbox': tensor(0.1494, device='cuda:0'), 'train_loss_giou': tensor(0.4675, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.8636, device='cuda:0'), 'validation_loss_ce': tensor(0.5780, device='cuda:0'), 'validation_loss_bbox': tensor(0.2105, device='cuda:0'), 'validation_loss_giou': tensor(0.6167, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(2.4896, device='cuda:0'), 'train_loss_ce': tensor(0.5899, device='cuda:0'), 'train_loss_bbox': tensor(0.2030, device='cuda:0'), 'train_loss_giou': tensor(0.4422, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.8427, device='cuda:0'), 'validation_loss_ce': tensor(0.5624, device='cuda:0'), 'validation_loss_bbox': tensor(0.2102, device='cuda:0'), 'validation_loss_giou': tensor(0.6146, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(2.3007, device='cuda:0'), 'train_loss_ce': tensor(0.4799, device='cuda:0'), 'train_loss_bbox': tensor(0.1701, device='cuda:0'), 'train_loss_giou': tensor(0.4853, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.8161, device='cuda:0'), 'validation_loss_ce': tensor(0.5492, device='cuda:0'), 'validation_loss_bbox': tensor(0.2090, device='cuda:0'), 'validation_loss_giou': tensor(0.6110, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(2.4182, device='cuda:0'), 'train_loss_ce': tensor(0.5711, device='cuda:0'), 'train_loss_bbox': tensor(0.1410, device='cuda:0'), 'train_loss_giou': tensor(0.5709, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.8062, device='cuda:0'), 'validation_loss_ce': tensor(0.5394, device='cuda:0'), 'validation_loss_bbox': tensor(0.2077, device='cuda:0'), 'validation_loss_giou': tensor(0.6141, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(2.0376, device='cuda:0'), 'train_loss_ce': tensor(0.5342, device='cuda:0'), 'train_loss_bbox': tensor(0.1438, device='cuda:0'), 'train_loss_giou': tensor(0.3921, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.7929, device='cuda:0'), 'validation_loss_ce': tensor(0.5284, device='cuda:0'), 'validation_loss_bbox': tensor(0.2054, device='cuda:0'), 'validation_loss_giou': tensor(0.6187, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(1.9113, device='cuda:0'), 'train_loss_ce': tensor(0.5603, device='cuda:0'), 'train_loss_bbox': tensor(0.1336, device='cuda:0'), 'train_loss_giou': tensor(0.3416, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.7774, device='cuda:0'), 'validation_loss_ce': tensor(0.5212, device='cuda:0'), 'validation_loss_bbox': tensor(0.2079, device='cuda:0'), 'validation_loss_giou': tensor(0.6084, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(1.9175, device='cuda:0'), 'train_loss_ce': tensor(0.5524, device='cuda:0'), 'train_loss_bbox': tensor(0.1297, device='cuda:0'), 'train_loss_giou': tensor(0.3582, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.7511, device='cuda:0'), 'validation_loss_ce': tensor(0.5131, device='cuda:0'), 'validation_loss_bbox': tensor(0.2043, device='cuda:0'), 'validation_loss_giou': tensor(0.6082, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(3.0097, device='cuda:0'), 'train_loss_ce': tensor(0.4589, device='cuda:0'), 'train_loss_bbox': tensor(0.1780, device='cuda:0'), 'train_loss_giou': tensor(0.8303, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.7464, device='cuda:0'), 'validation_loss_ce': tensor(0.5085, device='cuda:0'), 'validation_loss_bbox': tensor(0.2042, device='cuda:0'), 'validation_loss_giou': tensor(0.6085, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(1.6472, device='cuda:0'), 'train_loss_ce': tensor(0.5455, device='cuda:0'), 'train_loss_bbox': tensor(0.1061, device='cuda:0'), 'train_loss_giou': tensor(0.2856, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.7289, device='cuda:0'), 'validation_loss_ce': tensor(0.5024, device='cuda:0'), 'validation_loss_bbox': tensor(0.2017, device='cuda:0'), 'validation_loss_giou': tensor(0.6090, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(2.2923, device='cuda:0'), 'train_loss_ce': tensor(0.5256, device='cuda:0'), 'train_loss_bbox': tensor(0.1532, device='cuda:0'), 'train_loss_giou': tensor(0.5003, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.7224, device='cuda:0'), 'validation_loss_ce': tensor(0.4985, device='cuda:0'), 'validation_loss_bbox': tensor(0.2026, device='cuda:0'), 'validation_loss_giou': tensor(0.6054, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(2.0831, device='cuda:0'), 'train_loss_ce': tensor(0.5228, device='cuda:0'), 'train_loss_bbox': tensor(0.1292, device='cuda:0'), 'train_loss_giou': tensor(0.4571, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.7133, device='cuda:0'), 'validation_loss_ce': tensor(0.4948, device='cuda:0'), 'validation_loss_bbox': tensor(0.2020, device='cuda:0'), 'validation_loss_giou': tensor(0.6043, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(2.8135, device='cuda:0'), 'train_loss_ce': tensor(0.4272, device='cuda:0'), 'train_loss_bbox': tensor(0.2240, device='cuda:0'), 'train_loss_giou': tensor(0.6332, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.7171, device='cuda:0'), 'validation_loss_ce': tensor(0.4908, device='cuda:0'), 'validation_loss_bbox': tensor(0.2027, device='cuda:0'), 'validation_loss_giou': tensor(0.6064, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(2.2130, device='cuda:0'), 'train_loss_ce': tensor(0.4784, device='cuda:0'), 'train_loss_bbox': tensor(0.1390, device='cuda:0'), 'train_loss_giou': tensor(0.5198, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.6983, device='cuda:0'), 'validation_loss_ce': tensor(0.4877, device='cuda:0'), 'validation_loss_bbox': tensor(0.2003, device='cuda:0'), 'validation_loss_giou': tensor(0.6044, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(2.7309, device='cuda:0'), 'train_loss_ce': tensor(0.5104, device='cuda:0'), 'train_loss_bbox': tensor(0.1930, device='cuda:0'), 'train_loss_giou': tensor(0.6278, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.6918, device='cuda:0'), 'validation_loss_ce': tensor(0.4846, device='cuda:0'), 'validation_loss_bbox': tensor(0.2006, device='cuda:0'), 'validation_loss_giou': tensor(0.6022, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(1.6741, device='cuda:0'), 'train_loss_ce': tensor(0.5058, device='cuda:0'), 'train_loss_bbox': tensor(0.0880, device='cuda:0'), 'train_loss_giou': tensor(0.3642, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.6956, device='cuda:0'), 'validation_loss_ce': tensor(0.4832, device='cuda:0'), 'validation_loss_bbox': tensor(0.2020, device='cuda:0'), 'validation_loss_giou': tensor(0.6012, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(3.0059, device='cuda:0'), 'train_loss_ce': tensor(0.4803, device='cuda:0'), 'train_loss_bbox': tensor(0.2021, device='cuda:0'), 'train_loss_giou': tensor(0.7575, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.6876, device='cuda:0'), 'validation_loss_ce': tensor(0.4818, device='cuda:0'), 'validation_loss_bbox': tensor(0.2015, device='cuda:0'), 'validation_loss_giou': tensor(0.5990, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(2.7421, device='cuda:0'), 'train_loss_ce': tensor(0.5130, device='cuda:0'), 'train_loss_bbox': tensor(0.1896, device='cuda:0'), 'train_loss_giou': tensor(0.6407, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.6769, device='cuda:0'), 'validation_loss_ce': tensor(0.4791, device='cuda:0'), 'validation_loss_bbox': tensor(0.2004, device='cuda:0'), 'validation_loss_giou': tensor(0.5979, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(1.7125, device='cuda:0'), 'train_loss_ce': tensor(0.4362, device='cuda:0'), 'train_loss_bbox': tensor(0.1277, device='cuda:0'), 'train_loss_giou': tensor(0.3188, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.6743, device='cuda:0'), 'validation_loss_ce': tensor(0.4768, device='cuda:0'), 'validation_loss_bbox': tensor(0.1980, device='cuda:0'), 'validation_loss_giou': tensor(0.6038, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(2.2567, device='cuda:0'), 'train_loss_ce': tensor(0.5142, device='cuda:0'), 'train_loss_bbox': tensor(0.1576, device='cuda:0'), 'train_loss_giou': tensor(0.4773, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.6748, device='cuda:0'), 'validation_loss_ce': tensor(0.4765, device='cuda:0'), 'validation_loss_bbox': tensor(0.2005, device='cuda:0'), 'validation_loss_giou': tensor(0.5980, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(1.5372, device='cuda:0'), 'train_loss_ce': tensor(0.5373, device='cuda:0'), 'train_loss_bbox': tensor(0.0994, device='cuda:0'), 'train_loss_giou': tensor(0.2516, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.6640, device='cuda:0'), 'validation_loss_ce': tensor(0.4742, device='cuda:0'), 'validation_loss_bbox': tensor(0.1982, device='cuda:0'), 'validation_loss_giou': tensor(0.5993, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(2.5040, device='cuda:0'), 'train_loss_ce': tensor(0.4272, device='cuda:0'), 'train_loss_bbox': tensor(0.1789, device='cuda:0'), 'train_loss_giou': tensor(0.5912, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.6623, device='cuda:0'), 'validation_loss_ce': tensor(0.4728, device='cuda:0'), 'validation_loss_bbox': tensor(0.1981, device='cuda:0'), 'validation_loss_giou': tensor(0.5996, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(2.0120, device='cuda:0'), 'train_loss_ce': tensor(0.4059, device='cuda:0'), 'train_loss_bbox': tensor(0.1301, device='cuda:0'), 'train_loss_giou': tensor(0.4778, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.6473, device='cuda:0'), 'validation_loss_ce': tensor(0.4709, device='cuda:0'), 'validation_loss_bbox': tensor(0.1967, device='cuda:0'), 'validation_loss_giou': tensor(0.5964, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(2.0679, device='cuda:0'), 'train_loss_ce': tensor(0.5134, device='cuda:0'), 'train_loss_bbox': tensor(0.1250, device='cuda:0'), 'train_loss_giou': tensor(0.4649, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.6562, device='cuda:0'), 'validation_loss_ce': tensor(0.4703, device='cuda:0'), 'validation_loss_bbox': tensor(0.1983, device='cuda:0'), 'validation_loss_giou': tensor(0.5971, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(2.2338, device='cuda:0'), 'train_loss_ce': tensor(0.5023, device='cuda:0'), 'train_loss_bbox': tensor(0.1235, device='cuda:0'), 'train_loss_giou': tensor(0.5569, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.6502, device='cuda:0'), 'validation_loss_ce': tensor(0.4681, device='cuda:0'), 'validation_loss_bbox': tensor(0.1989, device='cuda:0'), 'validation_loss_giou': tensor(0.5938, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(3.3818, device='cuda:0'), 'train_loss_ce': tensor(0.2822, device='cuda:0'), 'train_loss_bbox': tensor(0.2465, device='cuda:0'), 'train_loss_giou': tensor(0.9336, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.6623, device='cuda:0'), 'validation_loss_ce': tensor(0.4676, device='cuda:0'), 'validation_loss_bbox': tensor(0.1998, device='cuda:0'), 'validation_loss_giou': tensor(0.5978, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(2.7525, device='cuda:0'), 'train_loss_ce': tensor(0.4899, device='cuda:0'), 'train_loss_bbox': tensor(0.1936, device='cuda:0'), 'train_loss_giou': tensor(0.6474, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.6486, device='cuda:0'), 'validation_loss_ce': tensor(0.4659, device='cuda:0'), 'validation_loss_bbox': tensor(0.1986, device='cuda:0'), 'validation_loss_giou': tensor(0.5948, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
+{'training_loss': tensor(2.1041, device='cuda:0'), 'train_loss_ce': tensor(0.4949, device='cuda:0'), 'train_loss_bbox': tensor(0.1062, device='cuda:0'), 'train_loss_giou': tensor(0.5390, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.6575, device='cuda:0'), 'validation_loss_ce': tensor(0.4652, device='cuda:0'), 'validation_loss_bbox': tensor(0.2000, device='cuda:0'), 'validation_loss_giou': tensor(0.5961, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
 ```
 
 ## Examples