Hugging Face's logo

anastasispk
/
law-game-evidence-replacement-finetune

Object Detection
Transformers
Safetensors
rt_detr
Generated from Trainer
Inference Endpoints
Model card Files Community
anastasispk commited on
Commit
1578d03
•
1 Parent(s): d442101

anastasispk/law-game-rt-detr

Browse files
Files changed (5) hide show
  1. README.md +109 -0
  2. config.json +110 -0
  3. model.safetensors +3 -0
  4. preprocessor_config.json +26 -0
  5. training_args.bin +3 -0
README.md ADDED
@@ -0,0 +1,109 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ base_model: PekingU/rtdetr_r50vd_coco_o365
4
+ tags:
5
+ - generated_from_trainer
6
+ model-index:
7
+ - name: law-game-evidence-replacement-finetune
8
+ results: []
9
+ ---
10
+
11
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
12
+ should probably proofread and complete it, then remove this comment. -->
13
+
14
+ # law-game-evidence-replacement-finetune
15
+
16
+ This model is a fine-tuned version of [PekingU/rtdetr_r50vd_coco_o365](https://huggingface.co/PekingU/rtdetr_r50vd_coco_o365) on an unknown dataset.
17
+ It achieves the following results on the evaluation set:
18
+ - Loss: 3.5533
19
+ - Map: 0.9339
20
+ - Map 50: 0.9616
21
+ - Map 75: 0.9575
22
+ - Map Small: 0.5574
23
+ - Map Medium: 0.9423
24
+ - Map Large: 0.9699
25
+ - Mar 1: 0.6597
26
+ - Mar 10: 0.9522
27
+ - Mar 100: 0.9722
28
+ - Mar Small: 0.7411
29
+ - Mar Medium: 0.9806
30
+ - Mar Large: 0.9908
31
+ - Map Evidence: -1.0
32
+ - Mar 100 Evidence: -1.0
33
+ - Map Ambulance: 0.9802
34
+ - Mar 100 Ambulance: 0.9899
35
+ - Map Artificial Target: 0.9245
36
+ - Mar 100 Artificial Target: 0.9611
37
+ - Map Cartridge: 0.9759
38
+ - Mar 100 Cartridge: 0.9937
39
+ - Map Gun: 0.9225
40
+ - Mar 100 Gun: 0.9542
41
+ - Map Knife: 0.8562
42
+ - Mar 100 Knife: 0.9404
43
+ - Map Police: 0.9495
44
+ - Mar 100 Police: 0.999
45
+ - Map Traffic Cone: 0.9285
46
+ - Mar 100 Traffic Cone: 0.9673
47
+
48
+ ## Model description
49
+
50
+ More information needed
51
+
52
+ ## Intended uses & limitations
53
+
54
+ More information needed
55
+
56
+ ## Training and evaluation data
57
+
58
+ More information needed
59
+
60
+ ## Training procedure
61
+
62
+ ### Training hyperparameters
63
+
64
+ The following hyperparameters were used during training:
65
+ - learning_rate: 5e-05
66
+ - train_batch_size: 16
67
+ - eval_batch_size: 8
68
+ - seed: 42
69
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
70
+ - lr_scheduler_type: linear
71
+ - lr_scheduler_warmup_steps: 300
72
+ - num_epochs: 25
73
+
74
+ ### Training results
75
+
76
+ | 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 Evidence | Mar 100 Evidence | Map Ambulance | Mar 100 Ambulance | Map Artificial Target | Mar 100 Artificial Target | Map Cartridge | Mar 100 Cartridge | Map Gun | Mar 100 Gun | Map Knife | Mar 100 Knife | Map Police | Mar 100 Police | Map Traffic Cone | Mar 100 Traffic Cone |
77
+ |:-------------:|:-----:|:----:|:---------------:|:------:|:------:|:------:|:---------:|:----------:|:---------:|:------:|:------:|:-------:|:---------:|:----------:|:---------:|:------------:|:----------------:|:-------------:|:-----------------:|:---------------------:|:-------------------------:|:-------------:|:-----------------:|:-------:|:-----------:|:---------:|:-------------:|:----------:|:--------------:|:----------------:|:--------------------:|
78
+ | No log | 1.0 | 183 | 17.1925 | 0.553 | 0.61 | 0.584 | 0.1918 | 0.3235 | 0.6555 | 0.5467 | 0.8763 | 0.8964 | 0.3142 | 0.8206 | 0.9705 | -1.0 | -1.0 | 0.9057 | 0.9848 | 0.5233 | 0.7299 | 0.9125 | 0.9647 | 0.1841 | 0.9194 | 0.6003 | 0.8687 | 0.518 | 0.9286 | 0.2268 | 0.8789 |
79
+ | No log | 2.0 | 366 | 7.1301 | 0.7763 | 0.8536 | 0.8146 | 0.2855 | 0.6006 | 0.875 | 0.6198 | 0.9116 | 0.9359 | 0.62 | 0.876 | 0.9781 | -1.0 | -1.0 | 0.9418 | 0.9707 | 0.7052 | 0.8648 | 0.9529 | 0.9733 | 0.5436 | 0.9667 | 0.7831 | 0.9172 | 0.8516 | 0.9398 | 0.656 | 0.9191 |
80
+ | 37.3669 | 3.0 | 549 | 5.7075 | 0.848 | 0.9115 | 0.8936 | 0.3256 | 0.7543 | 0.9317 | 0.6317 | 0.9274 | 0.9486 | 0.6783 | 0.9289 | 0.9849 | -1.0 | -1.0 | 0.9687 | 0.9879 | 0.7575 | 0.8761 | 0.9619 | 0.9822 | 0.8187 | 0.9653 | 0.8076 | 0.9172 | 0.9181 | 0.9827 | 0.7032 | 0.9287 |
81
+ | 37.3669 | 4.0 | 732 | 5.8395 | 0.8232 | 0.8809 | 0.8653 | 0.3221 | 0.7104 | 0.8994 | 0.642 | 0.9362 | 0.9536 | 0.688 | 0.9333 | 0.9884 | -1.0 | -1.0 | 0.9718 | 0.9899 | 0.8061 | 0.8878 | 0.9676 | 0.9854 | 0.8731 | 0.9778 | 0.7678 | 0.9162 | 0.6454 | 0.9867 | 0.7303 | 0.9317 |
82
+ | 37.3669 | 5.0 | 915 | 5.2081 | 0.8722 | 0.924 | 0.9156 | 0.3818 | 0.7789 | 0.951 | 0.6457 | 0.9406 | 0.9593 | 0.6963 | 0.9663 | 0.9887 | -1.0 | -1.0 | 0.976 | 0.9899 | 0.8077 | 0.9071 | 0.973 | 0.9869 | 0.7967 | 0.9611 | 0.8391 | 0.9313 | 0.8822 | 0.9908 | 0.8309 | 0.9482 |
83
+ | 4.4127 | 6.0 | 1098 | 5.4515 | 0.8848 | 0.9339 | 0.9262 | 0.5118 | 0.8295 | 0.9572 | 0.6538 | 0.9446 | 0.9624 | 0.6997 | 0.9621 | 0.9903 | -1.0 | -1.0 | 0.9686 | 0.9889 | 0.7937 | 0.9057 | 0.9784 | 0.9886 | 0.8982 | 0.9722 | 0.8491 | 0.9434 | 0.8521 | 0.9888 | 0.8534 | 0.9495 |
84
+ | 4.4127 | 7.0 | 1281 | 4.9756 | 0.9019 | 0.9468 | 0.9396 | 0.5037 | 0.8805 | 0.9631 | 0.6476 | 0.9443 | 0.9666 | 0.703 | 0.9692 | 0.9932 | -1.0 | -1.0 | 0.9754 | 0.9889 | 0.821 | 0.9129 | 0.9753 | 0.9907 | 0.9017 | 0.9833 | 0.8054 | 0.9414 | 0.9414 | 0.9949 | 0.8933 | 0.9541 |
85
+ | 4.4127 | 8.0 | 1464 | 4.4998 | 0.9119 | 0.9554 | 0.9432 | 0.5098 | 0.9047 | 0.9619 | 0.6482 | 0.9436 | 0.9641 | 0.7091 | 0.9786 | 0.9876 | -1.0 | -1.0 | 0.9726 | 0.9899 | 0.869 | 0.9248 | 0.9738 | 0.9902 | 0.8528 | 0.9556 | 0.8667 | 0.9384 | 0.9631 | 0.9939 | 0.8853 | 0.9561 |
86
+ | 3.3287 | 9.0 | 1647 | 4.5378 | 0.9107 | 0.9472 | 0.9424 | 0.5347 | 0.91 | 0.9625 | 0.6551 | 0.9498 | 0.9694 | 0.7243 | 0.984 | 0.9908 | -1.0 | -1.0 | 0.9802 | 0.9899 | 0.8691 | 0.9281 | 0.9785 | 0.9939 | 0.9128 | 0.9736 | 0.8646 | 0.9424 | 0.8889 | 0.9929 | 0.8805 | 0.965 |
87
+ | 3.3287 | 10.0 | 1830 | 5.0033 | 0.8831 | 0.9264 | 0.9202 | 0.5206 | 0.8887 | 0.9369 | 0.6497 | 0.9456 | 0.9661 | 0.7104 | 0.9714 | 0.9895 | -1.0 | -1.0 | 0.9404 | 0.9899 | 0.8686 | 0.929 | 0.9741 | 0.9917 | 0.92 | 0.9722 | 0.8131 | 0.9323 | 0.7768 | 0.9908 | 0.889 | 0.9568 |
88
+ | 2.8465 | 11.0 | 2013 | 4.1896 | 0.9183 | 0.9522 | 0.9491 | 0.4507 | 0.8926 | 0.9704 | 0.6595 | 0.9497 | 0.9676 | 0.7033 | 0.9677 | 0.9884 | -1.0 | -1.0 | 0.9786 | 0.9899 | 0.8902 | 0.9333 | 0.9745 | 0.993 | 0.9182 | 0.9583 | 0.8633 | 0.9424 | 0.9004 | 0.9929 | 0.9031 | 0.963 |
89
+ | 2.8465 | 12.0 | 2196 | 4.3806 | 0.9118 | 0.9486 | 0.9445 | 0.5313 | 0.8959 | 0.9574 | 0.6545 | 0.9487 | 0.9701 | 0.688 | 0.9741 | 0.9929 | -1.0 | -1.0 | 0.9791 | 0.9899 | 0.8856 | 0.935 | 0.9736 | 0.9924 | 0.9151 | 0.975 | 0.8429 | 0.9384 | 0.8852 | 0.998 | 0.9008 | 0.9617 |
90
+ | 2.8465 | 13.0 | 2379 | 4.3575 | 0.9131 | 0.9471 | 0.9419 | 0.5419 | 0.9126 | 0.9643 | 0.6576 | 0.9531 | 0.9717 | 0.7239 | 0.9875 | 0.9909 | -1.0 | -1.0 | 0.9677 | 0.9899 | 0.8731 | 0.9358 | 0.9774 | 0.9951 | 0.9226 | 0.9708 | 0.8794 | 0.9545 | 0.8601 | 0.9939 | 0.9114 | 0.9617 |
91
+ | 2.5085 | 14.0 | 2562 | 4.0609 | 0.9277 | 0.9619 | 0.9539 | 0.5802 | 0.9195 | 0.9659 | 0.6566 | 0.9518 | 0.9703 | 0.7168 | 0.9791 | 0.9913 | -1.0 | -1.0 | 0.9697 | 0.9899 | 0.902 | 0.9451 | 0.9819 | 0.9958 | 0.9273 | 0.9667 | 0.8392 | 0.9374 | 0.954 | 0.9939 | 0.9199 | 0.9634 |
92
+ | 2.5085 | 15.0 | 2745 | 4.2034 | 0.9284 | 0.961 | 0.9559 | 0.541 | 0.9483 | 0.9743 | 0.6606 | 0.9502 | 0.9695 | 0.7169 | 0.9792 | 0.9902 | -1.0 | -1.0 | 0.979 | 0.9899 | 0.9022 | 0.9469 | 0.9781 | 0.9947 | 0.9251 | 0.9611 | 0.8495 | 0.9404 | 0.9481 | 0.9918 | 0.9167 | 0.9614 |
93
+ | 2.5085 | 16.0 | 2928 | 4.1849 | 0.9283 | 0.9599 | 0.9559 | 0.5591 | 0.9323 | 0.9644 | 0.6575 | 0.9493 | 0.9697 | 0.7267 | 0.9795 | 0.988 | -1.0 | -1.0 | 0.9716 | 0.9899 | 0.9033 | 0.948 | 0.9754 | 0.9949 | 0.9108 | 0.9583 | 0.8469 | 0.9404 | 0.9675 | 0.9929 | 0.9222 | 0.9634 |
94
+ | 2.2183 | 17.0 | 3111 | 4.0696 | 0.9222 | 0.9556 | 0.9503 | 0.5517 | 0.9288 | 0.9634 | 0.6572 | 0.9523 | 0.9726 | 0.7348 | 0.9863 | 0.992 | -1.0 | -1.0 | 0.9707 | 0.9899 | 0.9052 | 0.9496 | 0.9784 | 0.9949 | 0.9309 | 0.9694 | 0.8074 | 0.9465 | 0.9525 | 0.9959 | 0.91 | 0.962 |
95
+ | 2.2183 | 18.0 | 3294 | 4.3283 | 0.9126 | 0.9461 | 0.9414 | 0.5422 | 0.9246 | 0.9498 | 0.6564 | 0.9502 | 0.9698 | 0.7138 | 0.9805 | 0.9896 | -1.0 | -1.0 | 0.9723 | 0.9899 | 0.901 | 0.9483 | 0.9815 | 0.9952 | 0.9204 | 0.9528 | 0.7964 | 0.9394 | 0.9043 | 0.9969 | 0.9125 | 0.966 |
96
+ | 2.2183 | 19.0 | 3477 | 3.7839 | 0.9209 | 0.9518 | 0.9477 | 0.5608 | 0.9475 | 0.9583 | 0.6562 | 0.9512 | 0.9701 | 0.7414 | 0.9806 | 0.9885 | -1.0 | -1.0 | 0.9566 | 0.9899 | 0.9131 | 0.9531 | 0.9779 | 0.9949 | 0.9132 | 0.9486 | 0.833 | 0.9404 | 0.9407 | 0.9959 | 0.9117 | 0.9677 |
97
+ | 2.009 | 20.0 | 3660 | 3.7275 | 0.9287 | 0.958 | 0.9542 | 0.5558 | 0.9078 | 0.9681 | 0.6586 | 0.951 | 0.9709 | 0.7422 | 0.979 | 0.9892 | -1.0 | -1.0 | 0.9802 | 0.9899 | 0.9239 | 0.96 | 0.9761 | 0.9944 | 0.9222 | 0.9514 | 0.8345 | 0.9364 | 0.9389 | 0.998 | 0.9248 | 0.966 |
98
+ | 2.009 | 21.0 | 3843 | 3.8496 | 0.93 | 0.9592 | 0.9554 | 0.5552 | 0.9187 | 0.9664 | 0.6581 | 0.9508 | 0.9708 | 0.7373 | 0.9788 | 0.9904 | -1.0 | -1.0 | 0.9802 | 0.9899 | 0.9148 | 0.9565 | 0.9778 | 0.9949 | 0.9227 | 0.9556 | 0.853 | 0.9364 | 0.9443 | 1.0 | 0.9167 | 0.9624 |
99
+ | 1.8494 | 22.0 | 4026 | 3.6452 | 0.9309 | 0.9592 | 0.9551 | 0.5561 | 0.929 | 0.9664 | 0.6595 | 0.9495 | 0.9709 | 0.723 | 0.9801 | 0.9902 | -1.0 | -1.0 | 0.9802 | 0.9899 | 0.9176 | 0.9593 | 0.9764 | 0.9935 | 0.9212 | 0.95 | 0.8459 | 0.9374 | 0.9471 | 0.999 | 0.928 | 0.9673 |
100
+ | 1.8494 | 23.0 | 4209 | 3.6352 | 0.9299 | 0.9587 | 0.9546 | 0.5524 | 0.9155 | 0.9681 | 0.659 | 0.9509 | 0.9708 | 0.7217 | 0.98 | 0.9902 | -1.0 | -1.0 | 0.9802 | 0.9899 | 0.9175 | 0.9589 | 0.9756 | 0.9935 | 0.9217 | 0.9514 | 0.8458 | 0.9364 | 0.9448 | 0.999 | 0.9236 | 0.9667 |
101
+ | 1.8494 | 24.0 | 4392 | 3.6526 | 0.9298 | 0.9577 | 0.9535 | 0.5572 | 0.9119 | 0.9666 | 0.6593 | 0.9518 | 0.972 | 0.725 | 0.9794 | 0.9911 | -1.0 | -1.0 | 0.9802 | 0.9899 | 0.9229 | 0.9609 | 0.976 | 0.9934 | 0.9217 | 0.9542 | 0.8493 | 0.9394 | 0.936 | 1.0 | 0.9226 | 0.9663 |
102
+ | 1.7025 | 25.0 | 4575 | 3.5533 | 0.9339 | 0.9616 | 0.9575 | 0.5574 | 0.9423 | 0.9699 | 0.6597 | 0.9522 | 0.9722 | 0.7411 | 0.9806 | 0.9908 | -1.0 | -1.0 | 0.9802 | 0.9899 | 0.9245 | 0.9611 | 0.9759 | 0.9937 | 0.9225 | 0.9542 | 0.8562 | 0.9404 | 0.9495 | 0.999 | 0.9285 | 0.9673 |
103
+
104
+
105
+ ### Framework versions
106
+
107
+ - Transformers 4.44.0.dev0
108
+ - Pytorch 2.3.1+cu121
109
+ - Tokenizers 0.19.1
config.json ADDED
@@ -0,0 +1,110 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "PekingU/rtdetr_r50vd_coco_o365",
3
+ "activation_dropout": 0.0,
4
+ "activation_function": "silu",
5
+ "anchor_image_size": null,
6
+ "architectures": [
7
+ "RTDetrForObjectDetection"
8
+ ],
9
+ "attention_dropout": 0.0,
10
+ "auxiliary_loss": true,
11
+ "backbone": null,
12
+ "backbone_config": {
13
+ "model_type": "rt_detr_resnet",
14
+ "out_features": [
15
+ "stage2",
16
+ "stage3",
17
+ "stage4"
18
+ ],
19
+ "out_indices": [
20
+ 2,
21
+ 3,
22
+ 4
23
+ ]
24
+ },
25
+ "backbone_kwargs": null,
26
+ "batch_norm_eps": 1e-05,
27
+ "box_noise_scale": 1.0,
28
+ "d_model": 256,
29
+ "decoder_activation_function": "relu",
30
+ "decoder_attention_heads": 8,
31
+ "decoder_ffn_dim": 1024,
32
+ "decoder_in_channels": [
33
+ 256,
34
+ 256,
35
+ 256
36
+ ],
37
+ "decoder_layers": 6,
38
+ "decoder_n_points": 4,
39
+ "disable_custom_kernels": true,
40
+ "dropout": 0.0,
41
+ "encode_proj_layers": [
42
+ 2
43
+ ],
44
+ "encoder_activation_function": "gelu",
45
+ "encoder_attention_heads": 8,
46
+ "encoder_ffn_dim": 1024,
47
+ "encoder_hidden_dim": 256,
48
+ "encoder_in_channels": [
49
+ 512,
50
+ 1024,
51
+ 2048
52
+ ],
53
+ "encoder_layers": 1,
54
+ "eos_coefficient": 0.0001,
55
+ "eval_size": null,
56
+ "feat_strides": [
57
+ 8,
58
+ 16,
59
+ 32
60
+ ],
61
+ "focal_loss_alpha": 0.75,
62
+ "focal_loss_gamma": 2.0,
63
+ "hidden_expansion": 1.0,
64
+ "id2label": {
65
+ "0": "evidence",
66
+ "1": "ambulance",
67
+ "2": "artificial_target",
68
+ "3": "cartridge",
69
+ "4": "gun",
70
+ "5": "knife",
71
+ "6": "police",
72
+ "7": "traffic_cone"
73
+ },
74
+ "initializer_bias_prior_prob": null,
75
+ "initializer_range": 0.01,
76
+ "is_encoder_decoder": true,
77
+ "label2id": {
78
+ "ambulance": 1,
79
+ "artificial_target": 2,
80
+ "cartridge": 3,
81
+ "evidence": 0,
82
+ "gun": 4,
83
+ "knife": 5,
84
+ "police": 6,
85
+ "traffic_cone": 7
86
+ },
87
+ "label_noise_ratio": 0.5,
88
+ "layer_norm_eps": 1e-05,
89
+ "learn_initial_query": false,
90
+ "matcher_alpha": 0.25,
91
+ "matcher_bbox_cost": 5.0,
92
+ "matcher_class_cost": 2.0,
93
+ "matcher_gamma": 2.0,
94
+ "matcher_giou_cost": 2.0,
95
+ "model_type": "rt_detr",
96
+ "normalize_before": false,
97
+ "num_denoising": 100,
98
+ "num_feature_levels": 3,
99
+ "num_queries": 300,
100
+ "positional_encoding_temperature": 10000,
101
+ "torch_dtype": "float32",
102
+ "transformers_version": "4.44.0.dev0",
103
+ "use_focal_loss": true,
104
+ "use_pretrained_backbone": false,
105
+ "use_timm_backbone": false,
106
+ "weight_loss_bbox": 5.0,
107
+ "weight_loss_giou": 2.0,
108
+ "weight_loss_vfl": 1.0,
109
+ "with_box_refine": true
110
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0256d13a4c8cc36f9b56dfc7fa559d2beb24afddb71f26a0c3927cc5f21596e3
3
+ size 171584000
preprocessor_config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_convert_annotations": true,
3
+ "do_normalize": false,
4
+ "do_pad": false,
5
+ "do_rescale": true,
6
+ "do_resize": true,
7
+ "format": "coco_detection",
8
+ "image_mean": [
9
+ 0.485,
10
+ 0.456,
11
+ 0.406
12
+ ],
13
+ "image_processor_type": "RTDetrImageProcessor",
14
+ "image_std": [
15
+ 0.229,
16
+ 0.224,
17
+ 0.225
18
+ ],
19
+ "pad_size": null,
20
+ "resample": 2,
21
+ "rescale_factor": 0.00392156862745098,
22
+ "size": {
23
+ "height": 480,
24
+ "width": 480
25
+ }
26
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:21687de6b9b240ae5d7bb9bf07c02292a2b9fdba776a0ffb0b4ad3ba8412aef9
3
+ size 5240