Hugging Face's logo

izzako
/
detr-resnet-50-finetuned-IDD_Detection

Object Detection
Transformers
Safetensors
conditional_detr
vision
Generated from Trainer
Model card Files Community
izzako commited on
Commit
0ca3f7f
·
verified ·
1 Parent(s): f1d8037

End of training

Browse files
Files changed (4) hide show
  1. README.md +47 -3
  2. model.safetensors +1 -1
  3. preprocessor_config.json +6 -3
  4. training_args.bin +1 -1
README.md CHANGED
@@ -17,6 +17,38 @@ should probably proofread and complete it, then remove this comment. -->
17
  # detr-resnet-50-finetuned-IDD_Detection
18
 
19
  This model is a fine-tuned version of [microsoft/conditional-detr-resnet-50](https://huggingface.co/microsoft/conditional-detr-resnet-50) on the IDD 40K Object Detection Dataset dataset.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
20
 
21
  ## Model description
22
 
@@ -36,16 +68,28 @@ More information needed
36
 
37
  The following hyperparameters were used during training:
38
  - learning_rate: 5e-05
39
- - train_batch_size: 6
40
- - eval_batch_size: 2
41
  - seed: 42
42
  - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
43
  - lr_scheduler_type: cosine
44
- - num_epochs: 2
45
  - mixed_precision_training: Native AMP
46
 
47
  ### Training results
48
 
 
 
 
 
 
 
 
 
 
 
 
 
49
 
50
 
51
  ### Framework versions
 
17
  # detr-resnet-50-finetuned-IDD_Detection
18
 
19
  This model is a fine-tuned version of [microsoft/conditional-detr-resnet-50](https://huggingface.co/microsoft/conditional-detr-resnet-50) on the IDD 40K Object Detection Dataset dataset.
20
+ It achieves the following results on the evaluation set:
21
+ - Loss: 1.3330
22
+ - Map: 0.1783
23
+ - Map 50: 0.3077
24
+ - Map 75: 0.1751
25
+ - Map Small: 0.009
26
+ - Map Medium: 0.1185
27
+ - Map Large: 0.4695
28
+ - Mar 1: 0.1438
29
+ - Mar 10: 0.2377
30
+ - Mar 100: 0.2512
31
+ - Mar Small: 0.0376
32
+ - Mar Medium: 0.2328
33
+ - Mar Large: 0.6055
34
+ - Map Traffic sign: 0.0334
35
+ - Mar 100 Traffic sign: 0.0823
36
+ - Map Motorcycle: 0.2083
37
+ - Mar 100 Motorcycle: 0.2868
38
+ - Map Car: 0.2712
39
+ - Mar 100 Car: 0.3321
40
+ - Map Rider: 0.1622
41
+ - Mar 100 Rider: 0.2486
42
+ - Map Person: 0.1099
43
+ - Mar 100 Person: 0.2018
44
+ - Map Truck: 0.2245
45
+ - Mar 100 Truck: 0.3299
46
+ - Map Autorickshaw: 0.2524
47
+ - Mar 100 Autorickshaw: 0.3135
48
+ - Map Vehicle fallback: 0.0061
49
+ - Mar 100 Vehicle fallback: 0.0592
50
+ - Map Bus: 0.3368
51
+ - Mar 100 Bus: 0.407
52
 
53
  ## Model description
54
 
 
68
 
69
  The following hyperparameters were used during training:
70
  - learning_rate: 5e-05
71
+ - train_batch_size: 4
72
+ - eval_batch_size: 4
73
  - seed: 42
74
  - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
75
  - lr_scheduler_type: cosine
76
+ - num_epochs: 10
77
  - mixed_precision_training: Native AMP
78
 
79
  ### Training results
80
 
81
+ | Training Loss | Epoch | Step | Validation Loss | Map | Map 50 | Map 75 | Map Small | Map Medium | Map Large | Mar 1 | Mar 10 | Mar 100 | Mar Small | Mar Medium | Mar Large | Map Traffic sign | Mar 100 Traffic sign | Map Motorcycle | Mar 100 Motorcycle | Map Car | Mar 100 Car | Map Rider | Mar 100 Rider | Map Person | Mar 100 Person | Map Truck | Mar 100 Truck | Map Autorickshaw | Mar 100 Autorickshaw | Map Vehicle fallback | Mar 100 Vehicle fallback | Map Bus | Mar 100 Bus |
82
+ |:-------------:|:-----:|:-----:|:---------------:|:------:|:------:|:------:|:---------:|:----------:|:---------:|:------:|:------:|:-------:|:---------:|:----------:|:---------:|:----------------:|:--------------------:|:--------------:|:------------------:|:-------:|:-----------:|:---------:|:-------------:|:----------:|:--------------:|:---------:|:-------------:|:----------------:|:--------------------:|:--------------------:|:------------------------:|:-------:|:-----------:|
83
+ | 1.6274 | 1.0 | 7781 | 1.6594 | 0.0827 | 0.1591 | 0.0753 | 0.0016 | 0.0332 | 0.25 | 0.0893 | 0.1532 | 0.1635 | 0.0102 | 0.1057 | 0.4353 | 0.0015 | 0.0058 | 0.1068 | 0.1821 | 0.1663 | 0.2353 | 0.0914 | 0.1761 | 0.0479 | 0.1358 | 0.0757 | 0.2095 | 0.1084 | 0.2127 | 0.0004 | 0.0069 | 0.1461 | 0.307 |
84
+ | 1.5632 | 2.0 | 15562 | 1.5677 | 0.1108 | 0.2045 | 0.1037 | 0.0034 | 0.057 | 0.3197 | 0.1065 | 0.1784 | 0.1904 | 0.016 | 0.1461 | 0.4907 | 0.0117 | 0.0328 | 0.1375 | 0.2216 | 0.205 | 0.2525 | 0.1161 | 0.2156 | 0.0725 | 0.158 | 0.1007 | 0.2404 | 0.1531 | 0.2402 | 0.0004 | 0.0208 | 0.1997 | 0.332 |
85
+ | 1.4935 | 3.0 | 23343 | 1.5396 | 0.1167 | 0.2215 | 0.1064 | 0.0028 | 0.058 | 0.3394 | 0.1089 | 0.1799 | 0.1915 | 0.0222 | 0.1513 | 0.4898 | 0.0168 | 0.039 | 0.128 | 0.2075 | 0.2018 | 0.268 | 0.1156 | 0.2008 | 0.0755 | 0.1518 | 0.127 | 0.2524 | 0.1666 | 0.2503 | 0.0034 | 0.0316 | 0.2156 | 0.3223 |
86
+ | 1.4442 | 4.0 | 31124 | 1.4694 | 0.1381 | 0.249 | 0.1337 | 0.0049 | 0.0766 | 0.3827 | 0.123 | 0.2018 | 0.2131 | 0.0214 | 0.1776 | 0.5429 | 0.0163 | 0.052 | 0.1579 | 0.2365 | 0.2293 | 0.2896 | 0.1257 | 0.2101 | 0.0843 | 0.1708 | 0.1647 | 0.2867 | 0.2039 | 0.2688 | 0.0024 | 0.0442 | 0.2585 | 0.3591 |
87
+ | 1.3898 | 5.0 | 38905 | 1.4330 | 0.1491 | 0.2664 | 0.1432 | 0.0054 | 0.0849 | 0.4121 | 0.1255 | 0.2096 | 0.2229 | 0.0265 | 0.191 | 0.562 | 0.0222 | 0.0566 | 0.167 | 0.2472 | 0.2381 | 0.3032 | 0.1379 | 0.2335 | 0.0919 | 0.1759 | 0.1854 | 0.2939 | 0.2153 | 0.2812 | 0.002 | 0.0486 | 0.2818 | 0.366 |
88
+ | 1.3433 | 6.0 | 46686 | 1.3940 | 0.1596 | 0.2795 | 0.1558 | 0.0071 | 0.0989 | 0.4314 | 0.1332 | 0.2194 | 0.2319 | 0.0296 | 0.2061 | 0.5715 | 0.0245 | 0.0667 | 0.1876 | 0.2666 | 0.2513 | 0.3144 | 0.1425 | 0.2292 | 0.0952 | 0.1819 | 0.1991 | 0.2967 | 0.2351 | 0.2998 | 0.0037 | 0.0495 | 0.297 | 0.3821 |
89
+ | 1.3258 | 7.0 | 54467 | 1.3651 | 0.169 | 0.295 | 0.1636 | 0.0073 | 0.1059 | 0.4511 | 0.1391 | 0.2291 | 0.2424 | 0.0342 | 0.2198 | 0.592 | 0.0266 | 0.0738 | 0.1972 | 0.2787 | 0.2586 | 0.319 | 0.1509 | 0.2361 | 0.1016 | 0.1929 | 0.2129 | 0.3184 | 0.2417 | 0.3096 | 0.0054 | 0.0571 | 0.3264 | 0.3963 |
90
+ | 1.2718 | 8.0 | 62248 | 1.3452 | 0.1749 | 0.3018 | 0.1718 | 0.0088 | 0.1145 | 0.4622 | 0.1417 | 0.2325 | 0.2456 | 0.0361 | 0.224 | 0.5943 | 0.0321 | 0.0784 | 0.2052 | 0.2836 | 0.2666 | 0.3264 | 0.1579 | 0.2417 | 0.107 | 0.1962 | 0.22 | 0.3192 | 0.2488 | 0.3113 | 0.0055 | 0.0559 | 0.3311 | 0.3977 |
91
+ | 1.2798 | 9.0 | 70029 | 1.3349 | 0.1784 | 0.3074 | 0.1754 | 0.0091 | 0.1179 | 0.4695 | 0.1438 | 0.2375 | 0.2511 | 0.0373 | 0.2327 | 0.6057 | 0.0332 | 0.0817 | 0.2073 | 0.2846 | 0.2704 | 0.3311 | 0.1618 | 0.2479 | 0.1108 | 0.2005 | 0.2235 | 0.3288 | 0.253 | 0.3137 | 0.0074 | 0.0608 | 0.338 | 0.4105 |
92
+ | 1.2695 | 10.0 | 77810 | 1.3330 | 0.1783 | 0.3077 | 0.1751 | 0.009 | 0.1185 | 0.4695 | 0.1438 | 0.2377 | 0.2512 | 0.0376 | 0.2328 | 0.6055 | 0.0334 | 0.0823 | 0.2083 | 0.2868 | 0.2712 | 0.3321 | 0.1622 | 0.2486 | 0.1099 | 0.2018 | 0.2245 | 0.3299 | 0.2524 | 0.3135 | 0.0061 | 0.0592 | 0.3368 | 0.407 |
93
 
94
 
95
  ### Framework versions
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:078cb785cba45dbdac57d31c22115ad5e8b4b8ce06b28858f84f3b9b8066625e
3
  size 174083908
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3aa2d041a400cf8dac73cd7b965fc614aec3d44148cd80ff61ffecb5a3f269e0
3
  size 174083908
preprocessor_config.json CHANGED
@@ -16,11 +16,14 @@
16
  0.224,
17
  0.225
18
  ],
19
- "pad_size": null,
 
 
 
20
  "resample": 2,
21
  "rescale_factor": 0.00392156862745098,
22
  "size": {
23
- "longest_edge": 1333,
24
- "shortest_edge": 800
25
  }
26
  }
 
16
  0.224,
17
  0.225
18
  ],
19
+ "pad_size": {
20
+ "height": 480,
21
+ "width": 480
22
+ },
23
  "resample": 2,
24
  "rescale_factor": 0.00392156862745098,
25
  "size": {
26
+ "max_height": 480,
27
+ "max_width": 480
28
  }
29
  }
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8030b9d5a5febbab4f27b1e1063f54946e68386b679c9ef7c3f87bf8fca65226
3
  size 5304
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8a93710266b559e5d3ea225e36dbf8f9ef64dd69206ca4b0fd8937c938b1cbae
3
  size 5304