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 ] = 0.004
Average Precision (AP) @[ IoU=0.50 | area= all | maxDets=100 ] = 0.011
Average Precision (AP) @[ IoU=0.75 | area= all | maxDets=100 ] = 0.002
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 ] = 0.004
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 1 ] = 0.046
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 10 ] = 0.092
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.107
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 ] = 0.108
Config
- dataset: NIH
- original model: hustvl/yolos-tiny
- lr: 0.0001
- dropout_rate: 0.1
- weight_decay: 5.0
- max_epochs: 30
- train samples: 885
Logging
Training process
{'validation_loss': tensor(6.5564, device='cuda:0'), 'validation_loss_ce': tensor(2.0633, device='cuda:0'), 'validation_loss_bbox': tensor(0.5405, device='cuda:0'), 'validation_loss_giou': tensor(0.8954, device='cuda:0'), 'validation_cardinality_error': tensor(92.5312, device='cuda:0')}
{'training_loss': tensor(3.7441, device='cuda:0'), 'train_loss_ce': tensor(0.4851, device='cuda:0'), 'train_loss_bbox': tensor(0.2741, device='cuda:0'), 'train_loss_giou': tensor(0.9444, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.8539, device='cuda:0'), 'validation_loss_ce': tensor(0.4760, device='cuda:0'), 'validation_loss_bbox': tensor(0.2163, device='cuda:0'), 'validation_loss_giou': tensor(0.6482, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.7504, device='cuda:0'), 'train_loss_ce': tensor(0.5047, device='cuda:0'), 'train_loss_bbox': tensor(0.1953, device='cuda:0'), 'train_loss_giou': tensor(0.6345, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.6948, device='cuda:0'), 'validation_loss_ce': tensor(0.4437, device='cuda:0'), 'validation_loss_bbox': tensor(0.2097, device='cuda:0'), 'validation_loss_giou': tensor(0.6013, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.0392, device='cuda:0'), 'train_loss_ce': tensor(0.4076, device='cuda:0'), 'train_loss_bbox': tensor(0.1439, device='cuda:0'), 'train_loss_giou': tensor(0.4562, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.4453, device='cuda:0'), 'validation_loss_ce': tensor(0.4250, device='cuda:0'), 'validation_loss_bbox': tensor(0.1845, device='cuda:0'), 'validation_loss_giou': tensor(0.5490, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.2227, device='cuda:0'), 'train_loss_ce': tensor(0.4271, device='cuda:0'), 'train_loss_bbox': tensor(0.1462, device='cuda:0'), 'train_loss_giou': tensor(0.5323, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.4683, device='cuda:0'), 'validation_loss_ce': tensor(0.4277, device='cuda:0'), 'validation_loss_bbox': tensor(0.1854, device='cuda:0'), 'validation_loss_giou': tensor(0.5569, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.3469, device='cuda:0'), 'train_loss_ce': tensor(0.4323, device='cuda:0'), 'train_loss_bbox': tensor(0.1472, device='cuda:0'), 'train_loss_giou': tensor(0.5894, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.4288, device='cuda:0'), 'validation_loss_ce': tensor(0.4206, device='cuda:0'), 'validation_loss_bbox': tensor(0.1770, device='cuda:0'), 'validation_loss_giou': tensor(0.5616, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.7726, device='cuda:0'), 'train_loss_ce': tensor(0.4058, device='cuda:0'), 'train_loss_bbox': tensor(0.2171, device='cuda:0'), 'train_loss_giou': tensor(0.6406, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.4031, device='cuda:0'), 'validation_loss_ce': tensor(0.4243, device='cuda:0'), 'validation_loss_bbox': tensor(0.1760, device='cuda:0'), 'validation_loss_giou': tensor(0.5495, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.4850, device='cuda:0'), 'train_loss_ce': tensor(0.4841, device='cuda:0'), 'train_loss_bbox': tensor(0.1897, device='cuda:0'), 'train_loss_giou': tensor(0.5262, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.3094, device='cuda:0'), 'validation_loss_ce': tensor(0.4155, device='cuda:0'), 'validation_loss_bbox': tensor(0.1648, device='cuda:0'), 'validation_loss_giou': tensor(0.5348, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(1.7707, device='cuda:0'), 'train_loss_ce': tensor(0.4562, device='cuda:0'), 'train_loss_bbox': tensor(0.0906, device='cuda:0'), 'train_loss_giou': tensor(0.4307, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.3606, device='cuda:0'), 'validation_loss_ce': tensor(0.4136, device='cuda:0'), 'validation_loss_bbox': tensor(0.1676, device='cuda:0'), 'validation_loss_giou': tensor(0.5546, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(1.8267, device='cuda:0'), 'train_loss_ce': tensor(0.4556, device='cuda:0'), 'train_loss_bbox': tensor(0.1383, device='cuda:0'), 'train_loss_giou': tensor(0.3398, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.3544, device='cuda:0'), 'validation_loss_ce': tensor(0.4203, device='cuda:0'), 'validation_loss_bbox': tensor(0.1730, device='cuda:0'), 'validation_loss_giou': tensor(0.5346, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.6995, device='cuda:0'), 'train_loss_ce': tensor(0.4739, device='cuda:0'), 'train_loss_bbox': tensor(0.2123, device='cuda:0'), 'train_loss_giou': tensor(0.5820, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.3790, device='cuda:0'), 'validation_loss_ce': tensor(0.4226, device='cuda:0'), 'validation_loss_bbox': tensor(0.1763, device='cuda:0'), 'validation_loss_giou': tensor(0.5375, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(1.9513, device='cuda:0'), 'train_loss_ce': tensor(0.4720, device='cuda:0'), 'train_loss_bbox': tensor(0.0936, device='cuda:0'), 'train_loss_giou': tensor(0.5057, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.4871, device='cuda:0'), 'validation_loss_ce': tensor(0.3954, device='cuda:0'), 'validation_loss_bbox': tensor(0.1899, device='cuda:0'), 'validation_loss_giou': tensor(0.5711, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.8887, device='cuda:0'), 'train_loss_ce': tensor(0.4222, device='cuda:0'), 'train_loss_bbox': tensor(0.1872, device='cuda:0'), 'train_loss_giou': tensor(0.7651, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.3490, device='cuda:0'), 'validation_loss_ce': tensor(0.4116, device='cuda:0'), 'validation_loss_bbox': tensor(0.1671, device='cuda:0'), 'validation_loss_giou': tensor(0.5511, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(3.0653, device='cuda:0'), 'train_loss_ce': tensor(0.3967, device='cuda:0'), 'train_loss_bbox': tensor(0.2608, device='cuda:0'), 'train_loss_giou': tensor(0.6824, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.3907, device='cuda:0'), 'validation_loss_ce': tensor(0.3930, device='cuda:0'), 'validation_loss_bbox': tensor(0.1780, device='cuda:0'), 'validation_loss_giou': tensor(0.5538, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.6032, device='cuda:0'), 'train_loss_ce': tensor(0.4456, device='cuda:0'), 'train_loss_bbox': tensor(0.1778, device='cuda:0'), 'train_loss_giou': tensor(0.6342, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.5283, device='cuda:0'), 'validation_loss_ce': tensor(0.3733, device='cuda:0'), 'validation_loss_bbox': tensor(0.1893, device='cuda:0'), 'validation_loss_giou': tensor(0.6041, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.2535, device='cuda:0'), 'train_loss_ce': tensor(0.4025, device='cuda:0'), 'train_loss_bbox': tensor(0.1915, device='cuda:0'), 'train_loss_giou': tensor(0.4469, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.5700, device='cuda:0'), 'validation_loss_ce': tensor(0.3973, device='cuda:0'), 'validation_loss_bbox': tensor(0.2018, device='cuda:0'), 'validation_loss_giou': tensor(0.5819, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.3217, device='cuda:0'), 'train_loss_ce': tensor(0.3088, device='cuda:0'), 'train_loss_bbox': tensor(0.1950, device='cuda:0'), 'train_loss_giou': tensor(0.5188, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.4046, device='cuda:0'), 'validation_loss_ce': tensor(0.3842, device='cuda:0'), 'validation_loss_bbox': tensor(0.1765, device='cuda:0'), 'validation_loss_giou': tensor(0.5688, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.0773, device='cuda:0'), 'train_loss_ce': tensor(0.3404, device='cuda:0'), 'train_loss_bbox': tensor(0.1290, device='cuda:0'), 'train_loss_giou': tensor(0.5459, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.3971, device='cuda:0'), 'validation_loss_ce': tensor(0.3989, device='cuda:0'), 'validation_loss_bbox': tensor(0.1737, device='cuda:0'), 'validation_loss_giou': tensor(0.5648, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.3780, device='cuda:0'), 'train_loss_ce': tensor(0.4421, device='cuda:0'), 'train_loss_bbox': tensor(0.2007, device='cuda:0'), 'train_loss_giou': tensor(0.4663, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.5065, device='cuda:0'), 'validation_loss_ce': tensor(0.3807, device='cuda:0'), 'validation_loss_bbox': tensor(0.1850, device='cuda:0'), 'validation_loss_giou': tensor(0.6004, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(1.9553, device='cuda:0'), 'train_loss_ce': tensor(0.2818, device='cuda:0'), 'train_loss_bbox': tensor(0.1281, device='cuda:0'), 'train_loss_giou': tensor(0.5166, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.6497, device='cuda:0'), 'validation_loss_ce': tensor(0.4114, device='cuda:0'), 'validation_loss_bbox': tensor(0.2036, device='cuda:0'), 'validation_loss_giou': tensor(0.6101, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(3.0033, device='cuda:0'), 'train_loss_ce': tensor(0.4743, device='cuda:0'), 'train_loss_bbox': tensor(0.2637, device='cuda:0'), 'train_loss_giou': tensor(0.6052, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.7365, device='cuda:0'), 'validation_loss_ce': tensor(0.3921, device='cuda:0'), 'validation_loss_bbox': tensor(0.2165, device='cuda:0'), 'validation_loss_giou': tensor(0.6309, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.4115, device='cuda:0'), 'train_loss_ce': tensor(0.4218, device='cuda:0'), 'train_loss_bbox': tensor(0.1878, device='cuda:0'), 'train_loss_giou': tensor(0.5255, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.7406, device='cuda:0'), 'validation_loss_ce': tensor(0.3620, device='cuda:0'), 'validation_loss_bbox': tensor(0.2215, device='cuda:0'), 'validation_loss_giou': tensor(0.6356, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.6330, device='cuda:0'), 'train_loss_ce': tensor(0.3572, device='cuda:0'), 'train_loss_bbox': tensor(0.2481, device='cuda:0'), 'train_loss_giou': tensor(0.5176, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.5378, device='cuda:0'), 'validation_loss_ce': tensor(0.3836, device='cuda:0'), 'validation_loss_bbox': tensor(0.1972, device='cuda:0'), 'validation_loss_giou': tensor(0.5842, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.4576, device='cuda:0'), 'train_loss_ce': tensor(0.3233, device='cuda:0'), 'train_loss_bbox': tensor(0.1875, device='cuda:0'), 'train_loss_giou': tensor(0.5983, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.4691, device='cuda:0'), 'validation_loss_ce': tensor(0.3929, device='cuda:0'), 'validation_loss_bbox': tensor(0.1909, device='cuda:0'), 'validation_loss_giou': tensor(0.5609, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.4825, device='cuda:0'), 'train_loss_ce': tensor(0.4494, device='cuda:0'), 'train_loss_bbox': tensor(0.1926, device='cuda:0'), 'train_loss_giou': tensor(0.5351, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.6155, device='cuda:0'), 'validation_loss_ce': tensor(0.4034, device='cuda:0'), 'validation_loss_bbox': tensor(0.2007, device='cuda:0'), 'validation_loss_giou': tensor(0.6043, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.6028, device='cuda:0'), 'train_loss_ce': tensor(0.4207, device='cuda:0'), 'train_loss_bbox': tensor(0.1980, device='cuda:0'), 'train_loss_giou': tensor(0.5962, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.6668, device='cuda:0'), 'validation_loss_ce': tensor(0.3779, device='cuda:0'), 'validation_loss_bbox': tensor(0.2057, device='cuda:0'), 'validation_loss_giou': tensor(0.6303, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(1.9037, device='cuda:0'), 'train_loss_ce': tensor(0.2973, device='cuda:0'), 'train_loss_bbox': tensor(0.1633, device='cuda:0'), 'train_loss_giou': tensor(0.3951, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.5248, device='cuda:0'), 'validation_loss_ce': tensor(0.3994, device='cuda:0'), 'validation_loss_bbox': tensor(0.1943, device='cuda:0'), 'validation_loss_giou': tensor(0.5770, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.6755, device='cuda:0'), 'train_loss_ce': tensor(0.3837, device='cuda:0'), 'train_loss_bbox': tensor(0.2018, device='cuda:0'), 'train_loss_giou': tensor(0.6414, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.7245, device='cuda:0'), 'validation_loss_ce': tensor(0.4018, device='cuda:0'), 'validation_loss_bbox': tensor(0.2082, device='cuda:0'), 'validation_loss_giou': tensor(0.6410, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.8687, device='cuda:0'), 'train_loss_ce': tensor(0.4472, device='cuda:0'), 'train_loss_bbox': tensor(0.2041, device='cuda:0'), 'train_loss_giou': tensor(0.7005, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.4158, device='cuda:0'), 'validation_loss_ce': tensor(0.4104, device='cuda:0'), 'validation_loss_bbox': tensor(0.1761, device='cuda:0'), 'validation_loss_giou': tensor(0.5626, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.8324, device='cuda:0'), 'train_loss_ce': tensor(0.4533, device='cuda:0'), 'train_loss_bbox': tensor(0.1971, device='cuda:0'), 'train_loss_giou': tensor(0.6968, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.3753, device='cuda:0'), 'validation_loss_ce': tensor(0.4029, device='cuda:0'), 'validation_loss_bbox': tensor(0.1723, device='cuda:0'), 'validation_loss_giou': tensor(0.5555, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.8085, device='cuda:0'), 'train_loss_ce': tensor(0.3939, device='cuda:0'), 'train_loss_bbox': tensor(0.2750, device='cuda:0'), 'train_loss_giou': tensor(0.5198, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.3772, device='cuda:0'), 'validation_loss_ce': tensor(0.4097, device='cuda:0'), 'validation_loss_bbox': tensor(0.1783, device='cuda:0'), 'validation_loss_giou': tensor(0.5380, device='cuda:0'), 'validation_cardinality_error': tensor(1., 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])}
