metadata
library_name: transformers
tags: []
Original result
IoU metric: bbox
Average Precision (AP) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.000
Average Precision (AP) @[ IoU=0.50 | area= all | maxDets=100 ] = 0.000
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 ] = 0.000
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 1 ] = 0.000
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 10 ] = 0.000
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.000
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 ] = 0.000
After training result
IoU metric: bbox
Average Precision (AP) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.053
Average Precision (AP) @[ IoU=0.50 | area= all | maxDets=100 ] = 0.096
Average Precision (AP) @[ IoU=0.75 | area= all | maxDets=100 ] = 0.063
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 ] = 0.053
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 1 ] = 0.100
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 10 ] = 0.179
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.211
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 ] = 0.232
Config
- dataset: NIH
- original model: hustvl/yolos-tiny
- lr: 0.0001
- dropout_rate: 0.1
- weight_decay: 0.05
- max_epochs: 30
- train samples: 885
Logging
Training process
{'validation_loss': tensor(6.9559, device='cuda:0'), 'validation_loss_ce': tensor(2.4853, device='cuda:0'), 'validation_loss_bbox': tensor(0.5189, device='cuda:0'), 'validation_loss_giou': tensor(0.9380, device='cuda:0'), 'validation_cardinality_error': tensor(99., device='cuda:0')}
{'training_loss': tensor(2.5294, device='cuda:0'), 'train_loss_ce': tensor(0.5418, device='cuda:0'), 'train_loss_bbox': tensor(0.1860, device='cuda:0'), 'train_loss_giou': tensor(0.5287, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.5128, device='cuda:0'), 'validation_loss_ce': tensor(0.4620, device='cuda:0'), 'validation_loss_bbox': tensor(0.1888, device='cuda:0'), 'validation_loss_giou': tensor(0.5533, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.2589, device='cuda:0'), 'train_loss_ce': tensor(0.4520, device='cuda:0'), 'train_loss_bbox': tensor(0.1275, device='cuda:0'), 'train_loss_giou': tensor(0.5847, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.1865, device='cuda:0'), 'validation_loss_ce': tensor(0.4225, device='cuda:0'), 'validation_loss_bbox': tensor(0.1510, device='cuda:0'), 'validation_loss_giou': tensor(0.5045, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.2895, device='cuda:0'), 'train_loss_ce': tensor(0.5192, device='cuda:0'), 'train_loss_bbox': tensor(0.1477, device='cuda:0'), 'train_loss_giou': tensor(0.5158, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.1751, device='cuda:0'), 'validation_loss_ce': tensor(0.4340, device='cuda:0'), 'validation_loss_bbox': tensor(0.1528, device='cuda:0'), 'validation_loss_giou': tensor(0.4887, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.1921, device='cuda:0'), 'train_loss_ce': tensor(0.4626, device='cuda:0'), 'train_loss_bbox': tensor(0.1593, device='cuda:0'), 'train_loss_giou': tensor(0.4666, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.3439, device='cuda:0'), 'validation_loss_ce': tensor(0.4168, device='cuda:0'), 'validation_loss_bbox': tensor(0.1764, device='cuda:0'), 'validation_loss_giou': tensor(0.5226, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(1.6664, device='cuda:0'), 'train_loss_ce': tensor(0.4768, device='cuda:0'), 'train_loss_bbox': tensor(0.0885, device='cuda:0'), 'train_loss_giou': tensor(0.3736, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.1216, device='cuda:0'), 'validation_loss_ce': tensor(0.4001, device='cuda:0'), 'validation_loss_bbox': tensor(0.1456, device='cuda:0'), 'validation_loss_giou': tensor(0.4969, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.5535, device='cuda:0'), 'train_loss_ce': tensor(0.4144, device='cuda:0'), 'train_loss_bbox': tensor(0.1866, device='cuda:0'), 'train_loss_giou': tensor(0.6031, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.3127, device='cuda:0'), 'validation_loss_ce': tensor(0.3946, device='cuda:0'), 'validation_loss_bbox': tensor(0.1638, device='cuda:0'), 'validation_loss_giou': tensor(0.5496, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.1711, device='cuda:0'), 'train_loss_ce': tensor(0.4254, device='cuda:0'), 'train_loss_bbox': tensor(0.1532, device='cuda:0'), 'train_loss_giou': tensor(0.4897, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.0774, device='cuda:0'), 'validation_loss_ce': tensor(0.4161, device='cuda:0'), 'validation_loss_bbox': tensor(0.1417, device='cuda:0'), 'validation_loss_giou': tensor(0.4764, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(1.7319, device='cuda:0'), 'train_loss_ce': tensor(0.4199, device='cuda:0'), 'train_loss_bbox': tensor(0.0986, device='cuda:0'), 'train_loss_giou': tensor(0.4096, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.0720, device='cuda:0'), 'validation_loss_ce': tensor(0.3979, device='cuda:0'), 'validation_loss_bbox': tensor(0.1357, device='cuda:0'), 'validation_loss_giou': tensor(0.4978, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(1.5806, device='cuda:0'), 'train_loss_ce': tensor(0.3604, device='cuda:0'), 'train_loss_bbox': tensor(0.1325, device='cuda:0'), 'train_loss_giou': tensor(0.2790, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.9735, device='cuda:0'), 'validation_loss_ce': tensor(0.3799, device='cuda:0'), 'validation_loss_bbox': tensor(0.1335, device='cuda:0'), 'validation_loss_giou': tensor(0.4631, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.0937, device='cuda:0'), 'train_loss_ce': tensor(0.4021, device='cuda:0'), 'train_loss_bbox': tensor(0.1257, device='cuda:0'), 'train_loss_giou': tensor(0.5315, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.9569, device='cuda:0'), 'validation_loss_ce': tensor(0.3864, device='cuda:0'), 'validation_loss_bbox': tensor(0.1335, device='cuda:0'), 'validation_loss_giou': tensor(0.4515, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(1.1840, device='cuda:0'), 'train_loss_ce': tensor(0.3635, device='cuda:0'), 'train_loss_bbox': tensor(0.0550, device='cuda:0'), 'train_loss_giou': tensor(0.2729, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.9329, device='cuda:0'), 'validation_loss_ce': tensor(0.3740, device='cuda:0'), 'validation_loss_bbox': tensor(0.1274, device='cuda:0'), 'validation_loss_giou': tensor(0.4610, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.0942, device='cuda:0'), 'train_loss_ce': tensor(0.3967, device='cuda:0'), 'train_loss_bbox': tensor(0.1509, device='cuda:0'), 'train_loss_giou': tensor(0.4715, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.9473, device='cuda:0'), 'validation_loss_ce': tensor(0.3690, device='cuda:0'), 'validation_loss_bbox': tensor(0.1283, device='cuda:0'), 'validation_loss_giou': tensor(0.4683, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(1.6510, device='cuda:0'), 'train_loss_ce': tensor(0.2956, device='cuda:0'), 'train_loss_bbox': tensor(0.0785, device='cuda:0'), 'train_loss_giou': tensor(0.4814, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.8971, device='cuda:0'), 'validation_loss_ce': tensor(0.3623, device='cuda:0'), 'validation_loss_bbox': tensor(0.1229, device='cuda:0'), 'validation_loss_giou': tensor(0.4603, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(1.3376, device='cuda:0'), 'train_loss_ce': tensor(0.2663, device='cuda:0'), 'train_loss_bbox': tensor(0.0985, device='cuda:0'), 'train_loss_giou': tensor(0.2893, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.8048, device='cuda:0'), 'validation_loss_ce': tensor(0.3550, device='cuda:0'), 'validation_loss_bbox': tensor(0.1180, device='cuda:0'), 'validation_loss_giou': tensor(0.4298, device='cuda:0'), 'validation_cardinality_error': tensor(0.9899, device='cuda:0')}
{'training_loss': tensor(1.8508, device='cuda:0'), 'train_loss_ce': tensor(0.4634, device='cuda:0'), 'train_loss_bbox': tensor(0.1160, device='cuda:0'), 'train_loss_giou': tensor(0.4036, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.8110, device='cuda:0'), 'validation_loss_ce': tensor(0.3658, device='cuda:0'), 'validation_loss_bbox': tensor(0.1133, device='cuda:0'), 'validation_loss_giou': tensor(0.4393, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.2132, device='cuda:0'), 'train_loss_ce': tensor(0.3865, device='cuda:0'), 'train_loss_bbox': tensor(0.1733, device='cuda:0'), 'train_loss_giou': tensor(0.4802, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.8956, device='cuda:0'), 'validation_loss_ce': tensor(0.3703, device='cuda:0'), 'validation_loss_bbox': tensor(0.1241, device='cuda:0'), 'validation_loss_giou': tensor(0.4525, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(1.4466, device='cuda:0'), 'train_loss_ce': tensor(0.4195, device='cuda:0'), 'train_loss_bbox': tensor(0.0736, device='cuda:0'), 'train_loss_giou': tensor(0.3295, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.9727, device='cuda:0'), 'validation_loss_ce': tensor(0.3733, device='cuda:0'), 'validation_loss_bbox': tensor(0.1318, device='cuda:0'), 'validation_loss_giou': tensor(0.4702, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(1.6351, device='cuda:0'), 'train_loss_ce': tensor(0.3461, device='cuda:0'), 'train_loss_bbox': tensor(0.0929, device='cuda:0'), 'train_loss_giou': tensor(0.4122, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.8800, device='cuda:0'), 'validation_loss_ce': tensor(0.3603, device='cuda:0'), 'validation_loss_bbox': tensor(0.1224, device='cuda:0'), 'validation_loss_giou': tensor(0.4539, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.1954, device='cuda:0'), 'train_loss_ce': tensor(0.3990, device='cuda:0'), 'train_loss_bbox': tensor(0.1613, device='cuda:0'), 'train_loss_giou': tensor(0.4950, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.8715, device='cuda:0'), 'validation_loss_ce': tensor(0.3536, device='cuda:0'), 'validation_loss_bbox': tensor(0.1234, device='cuda:0'), 'validation_loss_giou': tensor(0.4504, device='cuda:0'), 'validation_cardinality_error': tensor(0.9192, device='cuda:0')}
{'training_loss': tensor(1.8643, device='cuda:0'), 'train_loss_ce': tensor(0.3877, device='cuda:0'), 'train_loss_bbox': tensor(0.0971, device='cuda:0'), 'train_loss_giou': tensor(0.4955, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.7371, device='cuda:0'), 'validation_loss_ce': tensor(0.3362, device='cuda:0'), 'validation_loss_bbox': tensor(0.1148, device='cuda:0'), 'validation_loss_giou': tensor(0.4134, device='cuda:0'), 'validation_cardinality_error': tensor(0.9394, device='cuda:0')}
{'training_loss': tensor(1.3830, device='cuda:0'), 'train_loss_ce': tensor(0.1839, device='cuda:0'), 'train_loss_bbox': tensor(0.0825, device='cuda:0'), 'train_loss_giou': tensor(0.3932, device='cuda:0'), 'train_cardinality_error': tensor(0.6000, device='cuda:0'), 'validation_loss': tensor(1.8157, device='cuda:0'), 'validation_loss_ce': tensor(0.3523, device='cuda:0'), 'validation_loss_bbox': tensor(0.1163, device='cuda:0'), 'validation_loss_giou': tensor(0.4410, device='cuda:0'), 'validation_cardinality_error': tensor(0.9293, device='cuda:0')}
{'training_loss': tensor(1.6100, device='cuda:0'), 'train_loss_ce': tensor(0.3392, device='cuda:0'), 'train_loss_bbox': tensor(0.1034, device='cuda:0'), 'train_loss_giou': tensor(0.3770, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.7169, device='cuda:0'), 'validation_loss_ce': tensor(0.3482, device='cuda:0'), 'validation_loss_bbox': tensor(0.1110, device='cuda:0'), 'validation_loss_giou': tensor(0.4070, device='cuda:0'), 'validation_cardinality_error': tensor(0.9192, device='cuda:0')}
{'training_loss': tensor(1.5429, device='cuda:0'), 'train_loss_ce': tensor(0.3210, device='cuda:0'), 'train_loss_bbox': tensor(0.0945, device='cuda:0'), 'train_loss_giou': tensor(0.3747, device='cuda:0'), 'train_cardinality_error': tensor(0.8000, device='cuda:0'), 'validation_loss': tensor(1.7941, device='cuda:0'), 'validation_loss_ce': tensor(0.3513, device='cuda:0'), 'validation_loss_bbox': tensor(0.1161, device='cuda:0'), 'validation_loss_giou': tensor(0.4312, device='cuda:0'), 'validation_cardinality_error': tensor(0.8990, device='cuda:0')}
{'training_loss': tensor(1.4975, device='cuda:0'), 'train_loss_ce': tensor(0.3537, device='cuda:0'), 'train_loss_bbox': tensor(0.0965, device='cuda:0'), 'train_loss_giou': tensor(0.3307, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(1.7279, device='cuda:0'), 'validation_loss_ce': tensor(0.3444, device='cuda:0'), 'validation_loss_bbox': tensor(0.1122, device='cuda:0'), 'validation_loss_giou': tensor(0.4112, device='cuda:0'), 'validation_cardinality_error': tensor(0.8081, device='cuda:0')}
{'training_loss': tensor(1.6512, device='cuda:0'), 'train_loss_ce': tensor(0.3698, device='cuda:0'), 'train_loss_bbox': tensor(0.1170, device='cuda:0'), 'train_loss_giou': tensor(0.3481, device='cuda:0'), 'train_cardinality_error': tensor(0.4000, device='cuda:0'), 'validation_loss': tensor(1.8547, device='cuda:0'), 'validation_loss_ce': tensor(0.3486, device='cuda:0'), 'validation_loss_bbox': tensor(0.1224, device='cuda:0'), 'validation_loss_giou': tensor(0.4470, device='cuda:0'), 'validation_cardinality_error': tensor(0.7576, device='cuda:0')}
{'training_loss': tensor(2.2053, device='cuda:0'), 'train_loss_ce': tensor(0.3561, device='cuda:0'), 'train_loss_bbox': tensor(0.1675, device='cuda:0'), 'train_loss_giou': tensor(0.5059, device='cuda:0'), 'train_cardinality_error': tensor(0.8000, device='cuda:0'), 'validation_loss': tensor(1.8177, device='cuda:0'), 'validation_loss_ce': tensor(0.3480, device='cuda:0'), 'validation_loss_bbox': tensor(0.1206, device='cuda:0'), 'validation_loss_giou': tensor(0.4333, device='cuda:0'), 'validation_cardinality_error': tensor(0.7071, device='cuda:0')}
{'training_loss': tensor(1.6742, device='cuda:0'), 'train_loss_ce': tensor(0.3054, device='cuda:0'), 'train_loss_bbox': tensor(0.1093, device='cuda:0'), 'train_loss_giou': tensor(0.4112, device='cuda:0'), 'train_cardinality_error': tensor(0.8000, device='cuda:0'), 'validation_loss': tensor(1.7886, device='cuda:0'), 'validation_loss_ce': tensor(0.3294, device='cuda:0'), 'validation_loss_bbox': tensor(0.1209, device='cuda:0'), 'validation_loss_giou': tensor(0.4274, device='cuda:0'), 'validation_cardinality_error': tensor(0.6364, device='cuda:0')}
{'training_loss': tensor(0.8524, device='cuda:0'), 'train_loss_ce': tensor(0.1265, device='cuda:0'), 'train_loss_bbox': tensor(0.0692, device='cuda:0'), 'train_loss_giou': tensor(0.1898, device='cuda:0'), 'train_cardinality_error': tensor(0.8000, device='cuda:0'), 'validation_loss': tensor(1.7781, device='cuda:0'), 'validation_loss_ce': tensor(0.3431, device='cuda:0'), 'validation_loss_bbox': tensor(0.1189, device='cuda:0'), 'validation_loss_giou': tensor(0.4203, device='cuda:0'), 'validation_cardinality_error': tensor(0.6869, device='cuda:0')}
{'training_loss': tensor(1.4292, device='cuda:0'), 'train_loss_ce': tensor(0.3210, device='cuda:0'), 'train_loss_bbox': tensor(0.0939, device='cuda:0'), 'train_loss_giou': tensor(0.3193, device='cuda:0'), 'train_cardinality_error': tensor(0.6000, device='cuda:0'), 'validation_loss': tensor(1.8680, device='cuda:0'), 'validation_loss_ce': tensor(0.3525, device='cuda:0'), 'validation_loss_bbox': tensor(0.1307, device='cuda:0'), 'validation_loss_giou': tensor(0.4311, device='cuda:0'), 'validation_cardinality_error': tensor(0.8788, device='cuda:0')}
{'training_loss': tensor(1.2447, device='cuda:0'), 'train_loss_ce': tensor(0.2517, device='cuda:0'), 'train_loss_bbox': tensor(0.0609, device='cuda:0'), 'train_loss_giou': tensor(0.3442, device='cuda:0'), 'train_cardinality_error': tensor(0.6000, device='cuda:0'), 'validation_loss': tensor(1.7613, device='cuda:0'), 'validation_loss_ce': tensor(0.3511, device='cuda:0'), 'validation_loss_bbox': tensor(0.1156, device='cuda:0'), 'validation_loss_giou': tensor(0.4160, device='cuda:0'), 'validation_cardinality_error': tensor(0.5455, device='cuda:0')}
Examples
{'size': tensor([512, 512]), 'image_id': tensor([1]), 'class_labels': tensor([4]), 'boxes': tensor([[0.2622, 0.5729, 0.0847, 0.0773]]), 'area': tensor([1717.9431]), 'iscrowd': tensor([0]), 'orig_size': tensor([1024, 1024])}
