End of training
Browse files- README.md +53 -0
- config.json +232 -0
- model.safetensors +3 -0
- preprocessor_config.json +24 -0
- training_args.bin +3 -0
README.md
ADDED
@@ -0,0 +1,53 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: apache-2.0
|
3 |
+
base_model: facebook/detr-resnet-50
|
4 |
+
tags:
|
5 |
+
- generated_from_trainer
|
6 |
+
model-index:
|
7 |
+
- name: detr-resnet-50_finetuned_road_traffic
|
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-resnet-50_finetuned_road_traffic
|
15 |
+
|
16 |
+
This model is a fine-tuned version of [facebook/detr-resnet-50](https://huggingface.co/facebook/detr-resnet-50) on an unknown dataset.
|
17 |
+
|
18 |
+
## Model description
|
19 |
+
|
20 |
+
More information needed
|
21 |
+
|
22 |
+
## Intended uses & limitations
|
23 |
+
|
24 |
+
More information needed
|
25 |
+
|
26 |
+
## Training and evaluation data
|
27 |
+
|
28 |
+
More information needed
|
29 |
+
|
30 |
+
## Training procedure
|
31 |
+
|
32 |
+
### Training hyperparameters
|
33 |
+
|
34 |
+
The following hyperparameters were used during training:
|
35 |
+
- learning_rate: 1e-05
|
36 |
+
- train_batch_size: 8
|
37 |
+
- eval_batch_size: 8
|
38 |
+
- seed: 42
|
39 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
40 |
+
- lr_scheduler_type: linear
|
41 |
+
- num_epochs: 1
|
42 |
+
- mixed_precision_training: Native AMP
|
43 |
+
|
44 |
+
### Training results
|
45 |
+
|
46 |
+
|
47 |
+
|
48 |
+
### Framework versions
|
49 |
+
|
50 |
+
- Transformers 4.35.2
|
51 |
+
- Pytorch 2.1.1+cu118
|
52 |
+
- Datasets 2.15.0
|
53 |
+
- Tokenizers 0.15.0
|
config.json
ADDED
@@ -0,0 +1,232 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "facebook/detr-resnet-50",
|
3 |
+
"activation_dropout": 0.0,
|
4 |
+
"activation_function": "relu",
|
5 |
+
"architectures": [
|
6 |
+
"DetrForObjectDetection"
|
7 |
+
],
|
8 |
+
"attention_dropout": 0.0,
|
9 |
+
"auxiliary_loss": false,
|
10 |
+
"backbone": "resnet50",
|
11 |
+
"backbone_config": null,
|
12 |
+
"bbox_cost": 5,
|
13 |
+
"bbox_loss_coefficient": 5,
|
14 |
+
"class_cost": 1,
|
15 |
+
"classifier_dropout": 0.0,
|
16 |
+
"d_model": 256,
|
17 |
+
"decoder_attention_heads": 8,
|
18 |
+
"decoder_ffn_dim": 2048,
|
19 |
+
"decoder_layerdrop": 0.0,
|
20 |
+
"decoder_layers": 6,
|
21 |
+
"dice_loss_coefficient": 1,
|
22 |
+
"dilation": false,
|
23 |
+
"dropout": 0.1,
|
24 |
+
"encoder_attention_heads": 8,
|
25 |
+
"encoder_ffn_dim": 2048,
|
26 |
+
"encoder_layerdrop": 0.0,
|
27 |
+
"encoder_layers": 6,
|
28 |
+
"eos_coefficient": 0.1,
|
29 |
+
"giou_cost": 2,
|
30 |
+
"giou_loss_coefficient": 2,
|
31 |
+
"id2label": {
|
32 |
+
"0": "N/A",
|
33 |
+
"1": "person",
|
34 |
+
"2": "bicycle",
|
35 |
+
"3": "car",
|
36 |
+
"4": "motorcycle",
|
37 |
+
"5": "airplane",
|
38 |
+
"6": "bus",
|
39 |
+
"7": "train",
|
40 |
+
"8": "truck",
|
41 |
+
"9": "boat",
|
42 |
+
"10": "traffic light",
|
43 |
+
"11": "fire hydrant",
|
44 |
+
"12": "street sign",
|
45 |
+
"13": "stop sign",
|
46 |
+
"14": "parking meter",
|
47 |
+
"15": "bench",
|
48 |
+
"16": "bird",
|
49 |
+
"17": "cat",
|
50 |
+
"18": "dog",
|
51 |
+
"19": "horse",
|
52 |
+
"20": "sheep",
|
53 |
+
"21": "cow",
|
54 |
+
"22": "elephant",
|
55 |
+
"23": "bear",
|
56 |
+
"24": "zebra",
|
57 |
+
"25": "giraffe",
|
58 |
+
"26": "hat",
|
59 |
+
"27": "backpack",
|
60 |
+
"28": "umbrella",
|
61 |
+
"29": "shoe",
|
62 |
+
"30": "eye glasses",
|
63 |
+
"31": "handbag",
|
64 |
+
"32": "tie",
|
65 |
+
"33": "suitcase",
|
66 |
+
"34": "frisbee",
|
67 |
+
"35": "skis",
|
68 |
+
"36": "snowboard",
|
69 |
+
"37": "sports ball",
|
70 |
+
"38": "kite",
|
71 |
+
"39": "baseball bat",
|
72 |
+
"40": "baseball glove",
|
73 |
+
"41": "skateboard",
|
74 |
+
"42": "surfboard",
|
75 |
+
"43": "tennis racket",
|
76 |
+
"44": "bottle",
|
77 |
+
"45": "plate",
|
78 |
+
"46": "wine glass",
|
79 |
+
"47": "cup",
|
80 |
+
"48": "fork",
|
81 |
+
"49": "knife",
|
82 |
+
"50": "spoon",
|
83 |
+
"51": "bowl",
|
84 |
+
"52": "banana",
|
85 |
+
"53": "apple",
|
86 |
+
"54": "sandwich",
|
87 |
+
"55": "orange",
|
88 |
+
"56": "broccoli",
|
89 |
+
"57": "carrot",
|
90 |
+
"58": "hot dog",
|
91 |
+
"59": "pizza",
|
92 |
+
"60": "donut",
|
93 |
+
"61": "cake",
|
94 |
+
"62": "chair",
|
95 |
+
"63": "couch",
|
96 |
+
"64": "potted plant",
|
97 |
+
"65": "bed",
|
98 |
+
"66": "mirror",
|
99 |
+
"67": "dining table",
|
100 |
+
"68": "window",
|
101 |
+
"69": "desk",
|
102 |
+
"70": "toilet",
|
103 |
+
"71": "door",
|
104 |
+
"72": "tv",
|
105 |
+
"73": "laptop",
|
106 |
+
"74": "mouse",
|
107 |
+
"75": "remote",
|
108 |
+
"76": "keyboard",
|
109 |
+
"77": "cell phone",
|
110 |
+
"78": "microwave",
|
111 |
+
"79": "oven",
|
112 |
+
"80": "toaster",
|
113 |
+
"81": "sink",
|
114 |
+
"82": "refrigerator",
|
115 |
+
"83": "blender",
|
116 |
+
"84": "book",
|
117 |
+
"85": "clock",
|
118 |
+
"86": "vase",
|
119 |
+
"87": "scissors",
|
120 |
+
"88": "teddy bear",
|
121 |
+
"89": "hair drier",
|
122 |
+
"90": "toothbrush"
|
123 |
+
},
|
124 |
+
"init_std": 0.02,
|
125 |
+
"init_xavier_std": 1.0,
|
126 |
+
"is_encoder_decoder": true,
|
127 |
+
"label2id": {
|
128 |
+
"N/A": 0,
|
129 |
+
"airplane": 5,
|
130 |
+
"apple": 53,
|
131 |
+
"backpack": 27,
|
132 |
+
"banana": 52,
|
133 |
+
"baseball bat": 39,
|
134 |
+
"baseball glove": 40,
|
135 |
+
"bear": 23,
|
136 |
+
"bed": 65,
|
137 |
+
"bench": 15,
|
138 |
+
"bicycle": 2,
|
139 |
+
"bird": 16,
|
140 |
+
"blender": 83,
|
141 |
+
"boat": 9,
|
142 |
+
"book": 84,
|
143 |
+
"bottle": 44,
|
144 |
+
"bowl": 51,
|
145 |
+
"broccoli": 56,
|
146 |
+
"bus": 6,
|
147 |
+
"cake": 61,
|
148 |
+
"car": 3,
|
149 |
+
"carrot": 57,
|
150 |
+
"cat": 17,
|
151 |
+
"cell phone": 77,
|
152 |
+
"chair": 62,
|
153 |
+
"clock": 85,
|
154 |
+
"couch": 63,
|
155 |
+
"cow": 21,
|
156 |
+
"cup": 47,
|
157 |
+
"desk": 69,
|
158 |
+
"dining table": 67,
|
159 |
+
"dog": 18,
|
160 |
+
"donut": 60,
|
161 |
+
"door": 71,
|
162 |
+
"elephant": 22,
|
163 |
+
"eye glasses": 30,
|
164 |
+
"fire hydrant": 11,
|
165 |
+
"fork": 48,
|
166 |
+
"frisbee": 34,
|
167 |
+
"giraffe": 25,
|
168 |
+
"hair drier": 89,
|
169 |
+
"handbag": 31,
|
170 |
+
"hat": 26,
|
171 |
+
"horse": 19,
|
172 |
+
"hot dog": 58,
|
173 |
+
"keyboard": 76,
|
174 |
+
"kite": 38,
|
175 |
+
"knife": 49,
|
176 |
+
"laptop": 73,
|
177 |
+
"microwave": 78,
|
178 |
+
"mirror": 66,
|
179 |
+
"motorcycle": 4,
|
180 |
+
"mouse": 74,
|
181 |
+
"orange": 55,
|
182 |
+
"oven": 79,
|
183 |
+
"parking meter": 14,
|
184 |
+
"person": 1,
|
185 |
+
"pizza": 59,
|
186 |
+
"plate": 45,
|
187 |
+
"potted plant": 64,
|
188 |
+
"refrigerator": 82,
|
189 |
+
"remote": 75,
|
190 |
+
"sandwich": 54,
|
191 |
+
"scissors": 87,
|
192 |
+
"sheep": 20,
|
193 |
+
"shoe": 29,
|
194 |
+
"sink": 81,
|
195 |
+
"skateboard": 41,
|
196 |
+
"skis": 35,
|
197 |
+
"snowboard": 36,
|
198 |
+
"spoon": 50,
|
199 |
+
"sports ball": 37,
|
200 |
+
"stop sign": 13,
|
201 |
+
"street sign": 12,
|
202 |
+
"suitcase": 33,
|
203 |
+
"surfboard": 42,
|
204 |
+
"teddy bear": 88,
|
205 |
+
"tennis racket": 43,
|
206 |
+
"tie": 32,
|
207 |
+
"toaster": 80,
|
208 |
+
"toilet": 70,
|
209 |
+
"toothbrush": 90,
|
210 |
+
"traffic light": 10,
|
211 |
+
"train": 7,
|
212 |
+
"truck": 8,
|
213 |
+
"tv": 72,
|
214 |
+
"umbrella": 28,
|
215 |
+
"vase": 86,
|
216 |
+
"window": 68,
|
217 |
+
"wine glass": 46,
|
218 |
+
"zebra": 24
|
219 |
+
},
|
220 |
+
"mask_loss_coefficient": 1,
|
221 |
+
"max_position_embeddings": 1024,
|
222 |
+
"model_type": "detr",
|
223 |
+
"num_channels": 3,
|
224 |
+
"num_hidden_layers": 6,
|
225 |
+
"num_queries": 100,
|
226 |
+
"position_embedding_type": "sine",
|
227 |
+
"scale_embedding": false,
|
228 |
+
"torch_dtype": "float32",
|
229 |
+
"transformers_version": "4.35.2",
|
230 |
+
"use_pretrained_backbone": true,
|
231 |
+
"use_timm_backbone": true
|
232 |
+
}
|
model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7c84f937f6d5f5620fc6c7fdafecf43b236a972205c8b69d84f5a64931bf0f8e
|
3 |
+
size 166587352
|
preprocessor_config.json
ADDED
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"do_normalize": true,
|
3 |
+
"do_pad": true,
|
4 |
+
"do_rescale": true,
|
5 |
+
"do_resize": true,
|
6 |
+
"format": "coco_detection",
|
7 |
+
"image_mean": [
|
8 |
+
0.485,
|
9 |
+
0.456,
|
10 |
+
0.406
|
11 |
+
],
|
12 |
+
"image_processor_type": "DetrImageProcessor",
|
13 |
+
"image_std": [
|
14 |
+
0.229,
|
15 |
+
0.224,
|
16 |
+
0.225
|
17 |
+
],
|
18 |
+
"resample": 2,
|
19 |
+
"rescale_factor": 0.00392156862745098,
|
20 |
+
"size": {
|
21 |
+
"longest_edge": 1333,
|
22 |
+
"shortest_edge": 800
|
23 |
+
}
|
24 |
+
}
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:55b68f26de166d8aa9b50ebae1b4552bc2d95667d353a2db15a6843d3fde9e2e
|
3 |
+
size 4600
|