Hugging Face's logo

varcoder
/
segformer-b0-DeepCrack

Transformers
PyTorch
TensorBoard
segformer
Generated from Trainer
Model card Files Metrics Community
1
varcoder commited on
Commit
f97995a
·
1 Parent(s): 4a5dcf3

update model card README.md

Browse files
Files changed (1) hide show
  1. README.md +41 -18
README.md CHANGED
@@ -12,16 +12,16 @@ should probably proofread and complete it, then remove this comment. -->
12
 
13
  # segformer-b0-DeepCrack
14
 
15
- This model is a fine-tuned version of [nvidia/mit-b4](https://huggingface.co/nvidia/mit-b4) on the None dataset.
16
  It achieves the following results on the evaluation set:
17
- - Loss: 0.0121
18
- - Mean Iou: 0.0096
19
- - Mean Accuracy: 0.0192
20
- - Overall Accuracy: 0.0192
21
- - Accuracy Background: nan
22
- - Accuracy Cracked: 0.0192
23
- - Iou Background: 0.0
24
- - Iou Cracked: 0.0192
25
 
26
  ## Model description
27
 
@@ -50,15 +50,38 @@ The following hyperparameters were used during training:
50
 
51
  ### Training results
52
 
53
- | Training Loss | Epoch | Step | Validation Loss | Mean Iou | Mean Accuracy | Overall Accuracy | Accuracy Background | Accuracy Cracked | Iou Background | Iou Cracked |
54
- |:-------------:|:-----:|:----:|:---------------:|:--------:|:-------------:|:----------------:|:-------------------:|:----------------:|:--------------:|:-----------:|
55
- | 0.2068 | 0.13 | 20 | 0.1942 | 0.0041 | 0.0082 | 0.0082 | nan | 0.0082 | 0.0 | 0.0082 |
56
- | 0.0778 | 0.27 | 40 | 0.0679 | 0.0306 | 0.0612 | 0.0612 | nan | 0.0612 | 0.0 | 0.0612 |
57
- | 0.0325 | 0.4 | 60 | 0.0328 | 0.0167 | 0.0334 | 0.0334 | nan | 0.0334 | 0.0 | 0.0334 |
58
- | 0.0203 | 0.53 | 80 | 0.0191 | 0.0074 | 0.0148 | 0.0148 | nan | 0.0148 | 0.0 | 0.0148 |
59
- | 0.015 | 0.67 | 100 | 0.0152 | 0.0144 | 0.0288 | 0.0288 | nan | 0.0288 | 0.0 | 0.0288 |
60
- | 0.0139 | 0.8 | 120 | 0.0130 | 0.0074 | 0.0148 | 0.0148 | nan | 0.0148 | 0.0 | 0.0148 |
61
- | 0.016 | 0.93 | 140 | 0.0121 | 0.0096 | 0.0192 | 0.0192 | nan | 0.0192 | 0.0 | 0.0192 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
62
 
63
 
64
  ### Framework versions
 
12
 
13
  # segformer-b0-DeepCrack
14
 
15
+ This model is a fine-tuned version of [nvidia/mit-b0](https://huggingface.co/nvidia/mit-b0) on the None dataset.
16
  It achieves the following results on the evaluation set:
17
+ - Loss: 0.3347
18
+ - Mean Iou: 0.6839
19
+ - Mean Accuracy: 0.7408
20
+ - Overall Accuracy: 0.9681
21
+ - Accuracy Background: 0.9897
22
+ - Accuracy Crack: 0.4918
23
+ - Iou Background: 0.9674
24
+ - Iou Crack: 0.4003
25
 
26
  ## Model description
27
 
 
50
 
51
  ### Training results
52
 
53
+ | Training Loss | Epoch | Step | Validation Loss | Mean Iou | Mean Accuracy | Overall Accuracy | Accuracy Background | Accuracy Crack | Iou Background | Iou Crack |
54
+ |:-------------:|:-----:|:----:|:---------------:|:--------:|:-------------:|:----------------:|:-------------------:|:--------------:|:--------------:|:---------:|
55
+ | 0.8203 | 0.03 | 5 | 0.6973 | 0.3317 | 0.7410 | 0.5924 | 0.5783 | 0.9037 | 0.5758 | 0.0876 |
56
+ | 0.7469 | 0.07 | 10 | 0.6930 | 0.3533 | 0.7185 | 0.6325 | 0.6244 | 0.8125 | 0.6192 | 0.0873 |
57
+ | 0.7324 | 0.1 | 15 | 0.6884 | 0.3545 | 0.6605 | 0.6436 | 0.6421 | 0.6788 | 0.6329 | 0.0762 |
58
+ | 0.7079 | 0.13 | 20 | 0.6910 | 0.2537 | 0.5518 | 0.4726 | 0.4650 | 0.6386 | 0.4576 | 0.0498 |
59
+ | 0.6472 | 0.17 | 25 | 0.6831 | 0.2972 | 0.5734 | 0.5519 | 0.5498 | 0.5969 | 0.5400 | 0.0545 |
60
+ | 0.6344 | 0.2 | 30 | 0.6630 | 0.4652 | 0.7477 | 0.8045 | 0.8099 | 0.6854 | 0.7985 | 0.1318 |
61
+ | 0.6264 | 0.23 | 35 | 0.6389 | 0.5567 | 0.7850 | 0.8977 | 0.9084 | 0.6617 | 0.8947 | 0.2187 |
62
+ | 0.5811 | 0.27 | 40 | 0.6087 | 0.6070 | 0.8069 | 0.9279 | 0.9394 | 0.6745 | 0.9257 | 0.2882 |
63
+ | 0.5928 | 0.3 | 45 | 0.5584 | 0.6469 | 0.7851 | 0.9503 | 0.9660 | 0.6042 | 0.9490 | 0.3448 |
64
+ | 0.5312 | 0.33 | 50 | 0.5476 | 0.6508 | 0.7789 | 0.9527 | 0.9692 | 0.5886 | 0.9515 | 0.3502 |
65
+ | 0.5209 | 0.37 | 55 | 0.5423 | 0.6561 | 0.7665 | 0.9564 | 0.9744 | 0.5586 | 0.9553 | 0.3568 |
66
+ | 0.4675 | 0.4 | 60 | 0.5332 | 0.6470 | 0.7529 | 0.9553 | 0.9745 | 0.5313 | 0.9543 | 0.3397 |
67
+ | 0.4831 | 0.43 | 65 | 0.4772 | 0.6746 | 0.7502 | 0.9644 | 0.9847 | 0.5157 | 0.9636 | 0.3855 |
68
+ | 0.4512 | 0.47 | 70 | 0.4624 | 0.6734 | 0.7830 | 0.9598 | 0.9765 | 0.5895 | 0.9587 | 0.3881 |
69
+ | 0.426 | 0.5 | 75 | 0.4589 | 0.6688 | 0.7912 | 0.9572 | 0.9730 | 0.6094 | 0.9561 | 0.3815 |
70
+ | 0.4147 | 0.53 | 80 | 0.4529 | 0.6769 | 0.7846 | 0.9606 | 0.9773 | 0.5918 | 0.9596 | 0.3942 |
71
+ | 0.4144 | 0.57 | 85 | 0.4160 | 0.6767 | 0.7616 | 0.9635 | 0.9827 | 0.5405 | 0.9627 | 0.3908 |
72
+ | 0.4192 | 0.6 | 90 | 0.3747 | 0.6612 | 0.7271 | 0.9639 | 0.9863 | 0.4680 | 0.9631 | 0.3593 |
73
+ | 0.4294 | 0.63 | 95 | 0.3649 | 0.6495 | 0.7064 | 0.9637 | 0.9880 | 0.4247 | 0.9630 | 0.3359 |
74
+ | 0.3609 | 0.67 | 100 | 0.3730 | 0.6480 | 0.7003 | 0.9642 | 0.9893 | 0.4113 | 0.9636 | 0.3324 |
75
+ | 0.3782 | 0.7 | 105 | 0.3699 | 0.6584 | 0.7229 | 0.9637 | 0.9865 | 0.4592 | 0.9630 | 0.3538 |
76
+ | 0.3594 | 0.73 | 110 | 0.3505 | 0.6638 | 0.7161 | 0.9662 | 0.9899 | 0.4423 | 0.9656 | 0.3619 |
77
+ | 0.3966 | 0.77 | 115 | 0.3474 | 0.6720 | 0.7263 | 0.9670 | 0.9898 | 0.4627 | 0.9663 | 0.3776 |
78
+ | 0.3365 | 0.8 | 120 | 0.3598 | 0.6710 | 0.7185 | 0.9678 | 0.9915 | 0.4456 | 0.9672 | 0.3748 |
79
+ | 0.3497 | 0.83 | 125 | 0.3530 | 0.6752 | 0.7161 | 0.9692 | 0.9932 | 0.4389 | 0.9686 | 0.3817 |
80
+ | 0.3303 | 0.87 | 130 | 0.3424 | 0.6792 | 0.7247 | 0.9690 | 0.9922 | 0.4572 | 0.9684 | 0.3899 |
81
+ | 0.3702 | 0.9 | 135 | 0.3379 | 0.6823 | 0.7341 | 0.9686 | 0.9908 | 0.4774 | 0.9679 | 0.3967 |
82
+ | 0.3199 | 0.93 | 140 | 0.3317 | 0.6858 | 0.7468 | 0.9678 | 0.9888 | 0.5048 | 0.9671 | 0.4044 |
83
+ | 0.304 | 0.97 | 145 | 0.3189 | 0.6854 | 0.7408 | 0.9685 | 0.9900 | 0.4916 | 0.9678 | 0.4030 |
84
+ | 0.3392 | 1.0 | 150 | 0.3347 | 0.6839 | 0.7408 | 0.9681 | 0.9897 | 0.4918 | 0.9674 | 0.4003 |
85
 
86
 
87
  ### Framework versions