upload trashify box detection model trained on purely synthetic data
Browse files- README.md +108 -0
- config.json +71 -0
- model.safetensors +3 -0
- preprocessor_config.json +29 -0
- training_args.bin +3 -0
README.md
ADDED
@@ -0,0 +1,108 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
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: trashify-box-detection-model-synthetic-data-only
|
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 |
+
# trashify-box-detection-model-synthetic-data-only
|
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: 0.9981
|
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_ratio: 0.05
|
44 |
+
- num_epochs: 50
|
45 |
+
- mixed_precision_training: Native AMP
|
46 |
+
|
47 |
+
### Training results
|
48 |
+
|
49 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
50 |
+
|:-------------:|:-----:|:----:|:---------------:|
|
51 |
+
| 61.6636 | 1.0 | 83 | 4.7558 |
|
52 |
+
| 3.6677 | 2.0 | 166 | 2.7089 |
|
53 |
+
| 2.8053 | 3.0 | 249 | 2.4394 |
|
54 |
+
| 2.5214 | 4.0 | 332 | 2.0961 |
|
55 |
+
| 2.3314 | 5.0 | 415 | 1.9080 |
|
56 |
+
| 2.1984 | 6.0 | 498 | 1.7241 |
|
57 |
+
| 2.1242 | 7.0 | 581 | 1.6452 |
|
58 |
+
| 2.0397 | 8.0 | 664 | 1.5851 |
|
59 |
+
| 1.9768 | 9.0 | 747 | 1.6050 |
|
60 |
+
| 1.9331 | 10.0 | 830 | 1.4683 |
|
61 |
+
| 1.9034 | 11.0 | 913 | 1.4983 |
|
62 |
+
| 1.8388 | 12.0 | 996 | 1.4433 |
|
63 |
+
| 1.8016 | 13.0 | 1079 | 1.3840 |
|
64 |
+
| 1.7536 | 14.0 | 1162 | 1.3470 |
|
65 |
+
| 1.7232 | 15.0 | 1245 | 1.3208 |
|
66 |
+
| 1.7181 | 16.0 | 1328 | 1.3386 |
|
67 |
+
| 1.6767 | 17.0 | 1411 | 1.2681 |
|
68 |
+
| 1.657 | 18.0 | 1494 | 1.2348 |
|
69 |
+
| 1.6222 | 19.0 | 1577 | 1.2102 |
|
70 |
+
| 1.5906 | 20.0 | 1660 | 1.1885 |
|
71 |
+
| 1.546 | 21.0 | 1743 | 1.1645 |
|
72 |
+
| 1.5524 | 22.0 | 1826 | 1.1788 |
|
73 |
+
| 1.5119 | 23.0 | 1909 | 1.1606 |
|
74 |
+
| 1.4963 | 24.0 | 1992 | 1.1689 |
|
75 |
+
| 1.4932 | 25.0 | 2075 | 1.1395 |
|
76 |
+
| 1.4651 | 26.0 | 2158 | 1.1246 |
|
77 |
+
| 1.4358 | 27.0 | 2241 | 1.1250 |
|
78 |
+
| 1.4294 | 28.0 | 2324 | 1.1090 |
|
79 |
+
| 1.4277 | 29.0 | 2407 | 1.1183 |
|
80 |
+
| 1.4061 | 30.0 | 2490 | 1.1300 |
|
81 |
+
| 1.3885 | 31.0 | 2573 | 1.0943 |
|
82 |
+
| 1.3731 | 32.0 | 2656 | 1.0680 |
|
83 |
+
| 1.3811 | 33.0 | 2739 | 1.0820 |
|
84 |
+
| 1.3436 | 34.0 | 2822 | 1.0585 |
|
85 |
+
| 1.3314 | 35.0 | 2905 | 1.0574 |
|
86 |
+
| 1.347 | 36.0 | 2988 | 1.0766 |
|
87 |
+
| 1.3341 | 37.0 | 3071 | 1.0249 |
|
88 |
+
| 1.3188 | 38.0 | 3154 | 1.0250 |
|
89 |
+
| 1.3057 | 39.0 | 3237 | 1.0263 |
|
90 |
+
| 1.2932 | 40.0 | 3320 | 1.0307 |
|
91 |
+
| 1.2703 | 41.0 | 3403 | 1.0137 |
|
92 |
+
| 1.286 | 42.0 | 3486 | 1.0072 |
|
93 |
+
| 1.2864 | 43.0 | 3569 | 1.0097 |
|
94 |
+
| 1.2493 | 44.0 | 3652 | 0.9997 |
|
95 |
+
| 1.2511 | 45.0 | 3735 | 0.9994 |
|
96 |
+
| 1.2428 | 46.0 | 3818 | 1.0030 |
|
97 |
+
| 1.2306 | 47.0 | 3901 | 0.9964 |
|
98 |
+
| 1.2258 | 48.0 | 3984 | 1.0002 |
|
99 |
+
| 1.2307 | 49.0 | 4067 | 0.9977 |
|
100 |
+
| 1.2251 | 50.0 | 4150 | 0.9981 |
|
101 |
+
|
102 |
+
|
103 |
+
### Framework versions
|
104 |
+
|
105 |
+
- Transformers 4.44.0
|
106 |
+
- Pytorch 2.2.0+cu121
|
107 |
+
- Datasets 2.20.0
|
108 |
+
- Tokenizers 0.19.1
|
config.json
ADDED
@@ -0,0 +1,71 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
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_bin",
|
44 |
+
"3": "not_hand",
|
45 |
+
"4": "not_trash",
|
46 |
+
"5": "trash"
|
47 |
+
},
|
48 |
+
"init_std": 0.02,
|
49 |
+
"init_xavier_std": 1.0,
|
50 |
+
"is_encoder_decoder": true,
|
51 |
+
"label2id": {
|
52 |
+
"bin": 0,
|
53 |
+
"hand": 1,
|
54 |
+
"not_bin": 2,
|
55 |
+
"not_hand": 3,
|
56 |
+
"not_trash": 4,
|
57 |
+
"trash": 5
|
58 |
+
},
|
59 |
+
"mask_loss_coefficient": 1,
|
60 |
+
"max_position_embeddings": 1024,
|
61 |
+
"model_type": "conditional_detr",
|
62 |
+
"num_channels": 3,
|
63 |
+
"num_hidden_layers": 6,
|
64 |
+
"num_queries": 300,
|
65 |
+
"position_embedding_type": "sine",
|
66 |
+
"scale_embedding": false,
|
67 |
+
"torch_dtype": "float32",
|
68 |
+
"transformers_version": "4.44.0",
|
69 |
+
"use_pretrained_backbone": true,
|
70 |
+
"use_timm_backbone": true
|
71 |
+
}
|
model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:aaeaa56157c65f7fc5b845f3e14329b0c378dabe54c8bd40bfd06619b1c62d91
|
3 |
+
size 174080824
|
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 sha256:4a1af0f5f89b76bb6adbbc6e1a718eed986fcd2ad35bf19d3ec425ca7ea29d2e
|
3 |
+
size 5240
|