---

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 ] = 0.007
Average Precision (AP) @[ IoU=0.50 | area= all | maxDets=100 ] = 0.019
Average Precision (AP) @[ IoU=0.75 | area= all | maxDets=100 ] = 0.003
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 ] = 0.008
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 1 ] = 0.060
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 10 ] = 0.147
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.160
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 ] = 0.167
```

## Config
- dataset: NIH
- original model: facebook/detr-resnet-50
- lr: 0.0001
- max_epochs: 100

## Logging
### Training process
```
{'validation_loss': tensor(6.3390, device='cuda:0'), 'validation_loss_ce': tensor(1.9257, device='cuda:0'), 'validation_loss_bbox': tensor(0.5244, device='cuda:0'), 'validation_loss_giou': tensor(0.8958, device='cuda:0'), 'validation_cardinality_error': tensor(60.5938, device='cuda:0')}
{'training_loss': tensor(3.2533, device='cuda:0'), 'train_loss_ce': tensor(0.6592, device='cuda:0'), 'train_loss_bbox': tensor(0.2086, device='cuda:0'), 'train_loss_giou': tensor(0.7757, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(3.0453, device='cuda:0'), 'validation_loss_ce': tensor(0.5371, device='cuda:0'), 'validation_loss_bbox': tensor(0.2398, device='cuda:0'), 'validation_loss_giou': tensor(0.6546, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.6515, device='cuda:0'), 'train_loss_ce': tensor(0.3552, device='cuda:0'), 'train_loss_bbox': tensor(0.1901, device='cuda:0'), 'train_loss_giou': tensor(0.6729, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.3760, device='cuda:0'), 'validation_loss_ce': tensor(0.4638, device='cuda:0'), 'validation_loss_bbox': tensor(0.1636, device='cuda:0'), 'validation_loss_giou': tensor(0.5470, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.2251, device='cuda:0'), 'train_loss_ce': tensor(0.4881, device='cuda:0'), 'train_loss_bbox': tensor(0.1581, device='cuda:0'), 'train_loss_giou': tensor(0.4732, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.3159, device='cuda:0'), 'validation_loss_ce': tensor(0.4545, device='cuda:0'), 'validation_loss_bbox': tensor(0.1584, device='cuda:0'), 'validation_loss_giou': tensor(0.5348, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.6269, device='cuda:0'), 'train_loss_ce': tensor(0.4099, device='cuda:0'), 'train_loss_bbox': tensor(0.2217, device='cuda:0'), 'train_loss_giou': tensor(0.5543, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.3354, device='cuda:0'), 'validation_loss_ce': tensor(0.4496, device='cuda:0'), 'validation_loss_bbox': tensor(0.1670, device='cuda:0'), 'validation_loss_giou': tensor(0.5253, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.0376, device='cuda:0'), 'train_loss_ce': tensor(0.4754, device='cuda:0'), 'train_loss_bbox': tensor(0.1529, device='cuda:0'), 'train_loss_giou': tensor(0.3987, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.2204, device='cuda:0'), 'validation_loss_ce': tensor(0.4507, device='cuda:0'), 'validation_loss_bbox': tensor(0.1517, device='cuda:0'), 'validation_loss_giou': tensor(0.5057, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(1.9224, device='cuda:0'), 'train_loss_ce': tensor(0.4734, device='cuda:0'), 'train_loss_bbox': tensor(0.1257, device='cuda:0'), 'train_loss_giou': tensor(0.4103, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.3003, device='cuda:0'), 'validation_loss_ce': tensor(0.4434, device='cuda:0'), 'validation_loss_bbox': tensor(0.1589, device='cuda:0'), 'validation_loss_giou': tensor(0.5311, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.0460, device='cuda:0'), 'train_loss_ce': tensor(0.3925, device='cuda:0'), 'train_loss_bbox': tensor(0.1550, device='cuda:0'), 'train_loss_giou': tensor(0.4393, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.2038, device='cuda:0'), 'validation_loss_ce': tensor(0.4365, device='cuda:0'), 'validation_loss_bbox': tensor(0.1544, device='cuda:0'), 'validation_loss_giou': tensor(0.4977, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(1.9657, device='cuda:0'), 'train_loss_ce': tensor(0.3771, device='cuda:0'), 'train_loss_bbox': tensor(0.1054, device='cuda:0'), 'train_loss_giou': tensor(0.5309, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.2134, device='cuda:0'), 'validation_loss_ce': tensor(0.4344, device='cuda:0'), 'validation_loss_bbox': tensor(0.1532, device='cuda:0'), 'validation_loss_giou': tensor(0.5065, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.5808, device='cuda:0'), 'train_loss_ce': tensor(0.4870, device='cuda:0'), 'train_loss_bbox': tensor(0.1566, device='cuda:0'), 'train_loss_giou': tensor(0.6554, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.1047, device='cuda:0'), 'validation_loss_ce': tensor(0.4241, device='cuda:0'), 'validation_loss_bbox': tensor(0.1445, device='cuda:0'), 'validation_loss_giou': tensor(0.4789, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(1.9521, device='cuda:0'), 'train_loss_ce': tensor(0.4557, device='cuda:0'), 'train_loss_bbox': tensor(0.1401, device='cuda:0'), 'train_loss_giou': tensor(0.3981, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.2028, device='cuda:0'), 'validation_loss_ce': tensor(0.4253, device='cuda:0'), 'validation_loss_bbox': tensor(0.1533, device='cuda:0'), 'validation_loss_giou': tensor(0.5055, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.1985, device='cuda:0'), 'train_loss_ce': tensor(0.4123, device='cuda:0'), 'train_loss_bbox': tensor(0.1583, device='cuda:0'), 'train_loss_giou': tensor(0.4974, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.2105, device='cuda:0'), 'validation_loss_ce': tensor(0.4224, device='cuda:0'), 'validation_loss_bbox': tensor(0.1548, device='cuda:0'), 'validation_loss_giou': tensor(0.5069, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(1.9210, device='cuda:0'), 'train_loss_ce': tensor(0.4587, device='cuda:0'), 'train_loss_bbox': tensor(0.1236, device='cuda:0'), 'train_loss_giou': tensor(0.4223, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.1568, device='cuda:0'), 'validation_loss_ce': tensor(0.4179, device='cuda:0'), 'validation_loss_bbox': tensor(0.1463, device='cuda:0'), 'validation_loss_giou': tensor(0.5037, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.2657, device='cuda:0'), 'train_loss_ce': tensor(0.4704, device='cuda:0'), 'train_loss_bbox': tensor(0.1318, device='cuda:0'), 'train_loss_giou': tensor(0.5682, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.1193, device='cuda:0'), 'validation_loss_ce': tensor(0.4051, device='cuda:0'), 'validation_loss_bbox': tensor(0.1438, device='cuda:0'), 'validation_loss_giou': tensor(0.4977, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.3029, device='cuda:0'), 'train_loss_ce': tensor(0.4030, device='cuda:0'), 'train_loss_bbox': tensor(0.1263, device='cuda:0'), 'train_loss_giou': tensor(0.6342, device='cuda:0'), 'train_cardinality_error': tensor(1., device='cuda:0'), 'validation_loss': tensor(2.2088, device='cuda:0'), 'validation_loss_ce': tensor(0.4041, device='cuda:0'), 'validation_loss_bbox': tensor(0.1504, device='cuda:0'), 'validation_loss_giou': tensor(0.5263, device='cuda:0'), 'validation_cardinality_error': tensor(1., device='cuda:0')}
{'training_loss': tensor(2.8812, device='cuda:0'), 'train_loss_ce': tensor(0.3058, device='cuda:0'), 'train_loss_bbox': tensor(0.2438, device='cuda:0'

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c6fafa66dd72743e910a015372c8cea6ce6aa2da61ecda01bff236b70ddd1be1
+oid sha256:b21cd620b3bf67531318d6449b26f1bc8b815810f7380ba5682d16d68f6d24e4
 size 166510924