Hugging Face's logo

mrdbourke
/
detr_finetuned_trashify_box_detector

Object Detection
Transformers
Safetensors
conditional_detr
Generated from Trainer
Inference Endpoints
Model card Files Community
mrdbourke commited on
Commit
730b969
·
verified ·
1 Parent(s): 60b75cd

trashify-v1

Browse files
Files changed (5) hide show
  1. README.md +158 -0
  2. config.json +67 -0
  3. model.safetensors +3 -0
  4. preprocessor_config.json +29 -0
  5. training_args.bin +3 -0
README.md ADDED
@@ -0,0 +1,158 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ base_model: microsoft/conditional-detr-resnet-50
4
+ tags:
5
+ - generated_from_trainer
6
+ model-index:
7
+ - name: detr_finetuned_trashify_box_detector
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
+ # detr_finetuned_trashify_box_detector
15
+
16
+ This model is a fine-tuned version of [microsoft/conditional-detr-resnet-50](https://huggingface.co/microsoft/conditional-detr-resnet-50) on an unknown dataset.
17
+ It achieves the following results on the evaluation set:
18
+ - Loss: 1.2103
19
+
20
+ ## Model description
21
+
22
+ More information needed
23
+
24
+ ## Intended uses & limitations
25
+
26
+ More information needed
27
+
28
+ ## Training and evaluation data
29
+
30
+ More information needed
31
+
32
+ ## Training procedure
33
+
34
+ ### Training hyperparameters
35
+
36
+ The following hyperparameters were used during training:
37
+ - learning_rate: 0.0001
38
+ - train_batch_size: 32
39
+ - eval_batch_size: 32
40
+ - seed: 42
41
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
42
+ - lr_scheduler_type: linear
43
+ - lr_scheduler_warmup_steps: 50
44
+ - num_epochs: 100
45
+ - mixed_precision_training: Native AMP
46
+
47
+ ### Training results
48
+
49
+ | Training Loss | Epoch | Step | Validation Loss |
50
+ |:-------------:|:-----:|:----:|:---------------:|
51
+ | 59.9289 | 1.0 | 9 | 59.2281 |
52
+ | 54.8011 | 2.0 | 18 | 43.7291 |
53
+ | 34.6414 | 3.0 | 27 | 22.8508 |
54
+ | 15.8995 | 4.0 | 36 | 8.1439 |
55
+ | 6.2262 | 5.0 | 45 | 3.7740 |
56
+ | 3.6279 | 6.0 | 54 | 2.9067 |
57
+ | 3.0063 | 7.0 | 63 | 2.9085 |
58
+ | 2.9371 | 8.0 | 72 | 2.5180 |
59
+ | 2.6431 | 9.0 | 81 | 2.3684 |
60
+ | 2.6074 | 10.0 | 90 | 2.2390 |
61
+ | 2.4797 | 11.0 | 99 | 2.0779 |
62
+ | 2.4703 | 12.0 | 108 | 2.1778 |
63
+ | 2.484 | 13.0 | 117 | 2.1163 |
64
+ | 2.3445 | 14.0 | 126 | 2.0291 |
65
+ | 2.3304 | 15.0 | 135 | 1.9298 |
66
+ | 2.3748 | 16.0 | 144 | 2.0053 |
67
+ | 2.2942 | 17.0 | 153 | 1.9890 |
68
+ | 2.2322 | 18.0 | 162 | 1.9092 |
69
+ | 2.2244 | 19.0 | 171 | 1.8906 |
70
+ | 2.1804 | 20.0 | 180 | 1.8215 |
71
+ | 2.1202 | 21.0 | 189 | 1.8159 |
72
+ | 2.1076 | 22.0 | 198 | 1.7617 |
73
+ | 2.0875 | 23.0 | 207 | 1.7749 |
74
+ | 2.0771 | 24.0 | 216 | 1.7343 |
75
+ | 2.0842 | 25.0 | 225 | 1.7500 |
76
+ | 1.9582 | 26.0 | 234 | 1.7164 |
77
+ | 1.9865 | 27.0 | 243 | 1.6550 |
78
+ | 1.9712 | 28.0 | 252 | 1.6415 |
79
+ | 1.9536 | 29.0 | 261 | 1.5604 |
80
+ | 1.8901 | 30.0 | 270 | 1.5556 |
81
+ | 1.8435 | 31.0 | 279 | 1.5524 |
82
+ | 1.8589 | 32.0 | 288 | 1.5356 |
83
+ | 1.848 | 33.0 | 297 | 1.4709 |
84
+ | 1.8287 | 34.0 | 306 | 1.5041 |
85
+ | 1.7871 | 35.0 | 315 | 1.4433 |
86
+ | 1.7822 | 36.0 | 324 | 1.4514 |
87
+ | 1.7421 | 37.0 | 333 | 1.3894 |
88
+ | 1.701 | 38.0 | 342 | 1.4114 |
89
+ | 1.6385 | 39.0 | 351 | 1.3454 |
90
+ | 1.6397 | 40.0 | 360 | 1.3693 |
91
+ | 1.6628 | 41.0 | 369 | 1.4018 |
92
+ | 1.6068 | 42.0 | 378 | 1.3678 |
93
+ | 1.6293 | 43.0 | 387 | 1.4055 |
94
+ | 1.655 | 44.0 | 396 | 1.3835 |
95
+ | 1.5538 | 45.0 | 405 | 1.3675 |
96
+ | 1.5304 | 46.0 | 414 | 1.3579 |
97
+ | 1.564 | 47.0 | 423 | 1.3538 |
98
+ | 1.5554 | 48.0 | 432 | 1.3381 |
99
+ | 1.5507 | 49.0 | 441 | 1.2798 |
100
+ | 1.5419 | 50.0 | 450 | 1.2863 |
101
+ | 1.5634 | 51.0 | 459 | 1.2955 |
102
+ | 1.5321 | 52.0 | 468 | 1.2939 |
103
+ | 1.5152 | 53.0 | 477 | 1.2924 |
104
+ | 1.5063 | 54.0 | 486 | 1.2978 |
105
+ | 1.4461 | 55.0 | 495 | 1.2925 |
106
+ | 1.3762 | 56.0 | 504 | 1.2707 |
107
+ | 1.4025 | 57.0 | 513 | 1.2640 |
108
+ | 1.4257 | 58.0 | 522 | 1.2618 |
109
+ | 1.4425 | 59.0 | 531 | 1.2675 |
110
+ | 1.3617 | 60.0 | 540 | 1.2676 |
111
+ | 1.4396 | 61.0 | 549 | 1.2623 |
112
+ | 1.3183 | 62.0 | 558 | 1.2347 |
113
+ | 1.3449 | 63.0 | 567 | 1.2279 |
114
+ | 1.3784 | 64.0 | 576 | 1.2135 |
115
+ | 1.3639 | 65.0 | 585 | 1.2468 |
116
+ | 1.3633 | 66.0 | 594 | 1.2355 |
117
+ | 1.2905 | 67.0 | 603 | 1.2114 |
118
+ | 1.3364 | 68.0 | 612 | 1.2301 |
119
+ | 1.2687 | 69.0 | 621 | 1.2220 |
120
+ | 1.2859 | 70.0 | 630 | 1.2227 |
121
+ | 1.2576 | 71.0 | 639 | 1.2194 |
122
+ | 1.3423 | 72.0 | 648 | 1.2285 |
123
+ | 1.2954 | 73.0 | 657 | 1.2470 |
124
+ | 1.3408 | 74.0 | 666 | 1.2514 |
125
+ | 1.2383 | 75.0 | 675 | 1.2120 |
126
+ | 1.2192 | 76.0 | 684 | 1.2318 |
127
+ | 1.221 | 77.0 | 693 | 1.2252 |
128
+ | 1.2352 | 78.0 | 702 | 1.2074 |
129
+ | 1.2214 | 79.0 | 711 | 1.2175 |
130
+ | 1.1694 | 80.0 | 720 | 1.2202 |
131
+ | 1.2178 | 81.0 | 729 | 1.2277 |
132
+ | 1.1687 | 82.0 | 738 | 1.2280 |
133
+ | 1.1424 | 83.0 | 747 | 1.2266 |
134
+ | 1.1668 | 84.0 | 756 | 1.2227 |
135
+ | 1.1986 | 85.0 | 765 | 1.2313 |
136
+ | 1.1763 | 86.0 | 774 | 1.2127 |
137
+ | 1.1623 | 87.0 | 783 | 1.2150 |
138
+ | 1.1303 | 88.0 | 792 | 1.2078 |
139
+ | 1.1504 | 89.0 | 801 | 1.2033 |
140
+ | 1.1061 | 90.0 | 810 | 1.2065 |
141
+ | 1.1025 | 91.0 | 819 | 1.2076 |
142
+ | 1.0971 | 92.0 | 828 | 1.2012 |
143
+ | 1.137 | 93.0 | 837 | 1.2228 |
144
+ | 1.1341 | 94.0 | 846 | 1.2145 |
145
+ | 1.0625 | 95.0 | 855 | 1.2141 |
146
+ | 1.1355 | 96.0 | 864 | 1.2167 |
147
+ | 1.1636 | 97.0 | 873 | 1.2121 |
148
+ | 1.1306 | 98.0 | 882 | 1.2106 |
149
+ | 1.1104 | 99.0 | 891 | 1.2111 |
150
+ | 1.1377 | 100.0 | 900 | 1.2103 |
151
+
152
+
153
+ ### Framework versions
154
+
155
+ - Transformers 4.44.0
156
+ - Pytorch 2.2.0+cu121
157
+ - Datasets 2.20.0
158
+ - Tokenizers 0.19.1
config.json ADDED
@@ -0,0 +1,67 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "microsoft/conditional-detr-resnet-50",
3
+ "activation_dropout": 0.0,
4
+ "activation_function": "relu",
5
+ "architectures": [
6
+ "ConditionalDetrForObjectDetection"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "auxiliary_loss": false,
10
+ "backbone": "resnet50",
11
+ "backbone_config": null,
12
+ "backbone_kwargs": {
13
+ "in_chans": 3,
14
+ "out_indices": [
15
+ 1,
16
+ 2,
17
+ 3,
18
+ 4
19
+ ]
20
+ },
21
+ "bbox_cost": 5,
22
+ "bbox_loss_coefficient": 5,
23
+ "class_cost": 2,
24
+ "cls_loss_coefficient": 2,
25
+ "d_model": 256,
26
+ "decoder_attention_heads": 8,
27
+ "decoder_ffn_dim": 2048,
28
+ "decoder_layerdrop": 0.0,
29
+ "decoder_layers": 6,
30
+ "dice_loss_coefficient": 1,
31
+ "dilation": false,
32
+ "dropout": 0.1,
33
+ "encoder_attention_heads": 8,
34
+ "encoder_ffn_dim": 2048,
35
+ "encoder_layerdrop": 0.0,
36
+ "encoder_layers": 6,
37
+ "focal_alpha": 0.25,
38
+ "giou_cost": 2,
39
+ "giou_loss_coefficient": 2,
40
+ "id2label": {
41
+ "0": "bin",
42
+ "1": "hand",
43
+ "2": "not_trash",
44
+ "3": "trash"
45
+ },
46
+ "init_std": 0.02,
47
+ "init_xavier_std": 1.0,
48
+ "is_encoder_decoder": true,
49
+ "label2id": {
50
+ "bin": 0,
51
+ "hand": 1,
52
+ "not_trash": 2,
53
+ "trash": 3
54
+ },
55
+ "mask_loss_coefficient": 1,
56
+ "max_position_embeddings": 1024,
57
+ "model_type": "conditional_detr",
58
+ "num_channels": 3,
59
+ "num_hidden_layers": 6,
60
+ "num_queries": 300,
61
+ "position_embedding_type": "sine",
62
+ "scale_embedding": false,
63
+ "torch_dtype": "float32",
64
+ "transformers_version": "4.44.0",
65
+ "use_pretrained_backbone": true,
66
+ "use_timm_backbone": true
67
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c08e8c1c5eeee2f1f8c51d24aaedfe23940b3274a188d972e65173892d06d5a4
3
+ size 174078768
preprocessor_config.json ADDED
@@ -0,0 +1,29 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_convert_annotations": true,
3
+ "do_normalize": true,
4
+ "do_pad": true,
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": "ConditionalDetrImageProcessor",
14
+ "image_std": [
15
+ 0.229,
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
+ "height": 480,
27
+ "width": 480
28
+ }
29
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ef6343e02276377f4f562acbd0a04f4a041885db3a533cd9e09fc453a025fb1d
3
+ size 5176