Hugging Face's logo

rathi2023
/
owlvit-base-patch32_FT_cppe5

Zero-Shot Object Detection
Transformers
TensorBoard
Safetensors
owlvit
Inference Endpoints
Model card Files Metrics Community
rathi2023 commited on
Commit
b2f71b6
•
1 Parent(s): b39932b

Training in progress, step 200

Browse files
Files changed (20) hide show
  1. config.json +106 -0
  2. merges.txt +0 -0
  3. model.safetensors +3 -0
  4. preprocessor_config.json +29 -0
  5. runs/Apr15_20-52-34_e2ec414f6ac2/events.out.tfevents.1713215319.e2ec414f6ac2.453.0 +3 -0
  6. runs/Apr15_20-52-34_e2ec414f6ac2/events.out.tfevents.1713215470.e2ec414f6ac2.453.1 +3 -0
  7. runs/Apr15_20-52-34_e2ec414f6ac2/events.out.tfevents.1713215505.e2ec414f6ac2.453.2 +3 -0
  8. runs/Apr15_20-52-34_e2ec414f6ac2/events.out.tfevents.1713215619.e2ec414f6ac2.453.3 +3 -0
  9. runs/Apr15_20-52-34_e2ec414f6ac2/events.out.tfevents.1713215696.e2ec414f6ac2.453.4 +3 -0
  10. runs/Apr15_20-52-34_e2ec414f6ac2/events.out.tfevents.1713215851.e2ec414f6ac2.453.5 +3 -0
  11. runs/Apr15_20-52-34_e2ec414f6ac2/events.out.tfevents.1713215983.e2ec414f6ac2.453.6 +3 -0
  12. runs/Apr15_20-52-34_e2ec414f6ac2/events.out.tfevents.1713216044.e2ec414f6ac2.453.7 +3 -0
  13. runs/Apr15_20-52-34_e2ec414f6ac2/events.out.tfevents.1713216099.e2ec414f6ac2.453.8 +3 -0
  14. runs/Apr15_20-52-34_e2ec414f6ac2/events.out.tfevents.1713216164.e2ec414f6ac2.453.9 +3 -0
  15. runs/Apr15_21-23-57_e2ec414f6ac2/events.out.tfevents.1713216251.e2ec414f6ac2.453.10 +3 -0
  16. special_tokens_map.json +24 -0
  17. tokenizer.json +0 -0
  18. tokenizer_config.json +39 -0
  19. training_args.bin +3 -0
  20. vocab.json +0 -0
config.json ADDED
@@ -0,0 +1,106 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "google/owlvit-base-patch32",
3
+ "architectures": [
4
+ "OwlViTForObjectDetection"
5
+ ],
6
+ "id2label": {
7
+ "0": 6,
8
+ "1": 7,
9
+ "10": 16,
10
+ "11": 17,
11
+ "12": 18,
12
+ "13": 19,
13
+ "14": 20,
14
+ "15": 24,
15
+ "16": 25,
16
+ "17": 26,
17
+ "18": 27,
18
+ "19": 28,
19
+ "2": 8,
20
+ "20": 29,
21
+ "21": 30,
22
+ "22": 31,
23
+ "23": 32,
24
+ "24": 33,
25
+ "25": 34,
26
+ "26": 36,
27
+ "27": 37,
28
+ "28": 38,
29
+ "29": 40,
30
+ "3": 9,
31
+ "30": 41,
32
+ "31": 42,
33
+ "32": 43,
34
+ "33": 44,
35
+ "34": 45,
36
+ "35": 46,
37
+ "36": 47,
38
+ "37": 48,
39
+ "4": 10,
40
+ "5": 11,
41
+ "6": 12,
42
+ "7": 13,
43
+ "8": 14,
44
+ "9": 15
45
+ },
46
+ "initializer_factor": 1.0,
47
+ "label2id": {
48
+ "6": "0",
49
+ "7": "1",
50
+ "8": "2",
51
+ "9": "3",
52
+ "10": "4",
53
+ "11": "5",
54
+ "12": "6",
55
+ "13": "7",
56
+ "14": "8",
57
+ "15": "9",
58
+ "16": "10",
59
+ "17": "11",
60
+ "18": "12",
61
+ "19": "13",
62
+ "20": "14",
63
+ "24": "15",
64
+ "25": "16",
65
+ "26": "17",
66
+ "27": "18",
67
+ "28": "19",
68
+ "29": "20",
69
+ "30": "21",
70
+ "31": "22",
71
+ "32": "23",
72
+ "33": "24",
73
+ "34": "25",
74
+ "36": "26",
75
+ "37": "27",
76
+ "38": "28",
77
+ "40": "29",
78
+ "41": "30",
79
+ "42": "31",
80
+ "43": "32",
81
+ "44": "33",
82
+ "45": "34",
83
+ "46": "35",
84
+ "47": "36",
85
+ "48": "37"
86
+ },
87
+ "logit_scale_init_value": 2.6592,
88
+ "model_type": "owlvit",
89
+ "projection_dim": 512,
90
+ "text_config": {
91
+ "bos_token_id": 0,
92
+ "dropout": 0.0,
93
+ "eos_token_id": 2,
94
+ "max_length": 16,
95
+ "model_type": "owlvit_text_model",
96
+ "pad_token_id": 1
97
+ },
98
+ "text_config_dict": null,
99
+ "torch_dtype": "float32",
100
+ "transformers_version": "4.38.2",
101
+ "vision_config": {
102
+ "dropout": 0.0,
103
+ "model_type": "owlvit_vision_model"
104
+ },
105
+ "vision_config_dict": null
106
+ }
merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:40b0fbb7b18a468417dd1786d035d993a8414706bb4b16098cfc1edbdba6031e
3
+ size 612978988
preprocessor_config.json ADDED
@@ -0,0 +1,29 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "crop_size": {
3
+ "height": 768,
4
+ "width": 768
5
+ },
6
+ "do_center_crop": false,
7
+ "do_convert_rgb": true,
8
+ "do_normalize": true,
9
+ "do_rescale": true,
10
+ "do_resize": true,
11
+ "image_mean": [
12
+ 0.48145466,
13
+ 0.4578275,
14
+ 0.40821073
15
+ ],
16
+ "image_processor_type": "OwlViTImageProcessor",
17
+ "image_std": [
18
+ 0.26862954,
19
+ 0.26130258,
20
+ 0.27577711
21
+ ],
22
+ "processor_class": "OwlViTProcessor",
23
+ "resample": 3,
24
+ "rescale_factor": 0.00392156862745098,
25
+ "size": {
26
+ "height": 768,
27
+ "width": 768
28
+ }
29
+ }
runs/Apr15_20-52-34_e2ec414f6ac2/events.out.tfevents.1713215319.e2ec414f6ac2.453.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3a2c6978b919f237b5a54830cdbc24c4352fb74815f0840421e4ae6f9d615026
3
+ size 6264
runs/Apr15_20-52-34_e2ec414f6ac2/events.out.tfevents.1713215470.e2ec414f6ac2.453.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9dabed3e62ecc406c6844e2bd6c06711896fc7dfdbf7801b585491291be4e69f
3
+ size 5709
runs/Apr15_20-52-34_e2ec414f6ac2/events.out.tfevents.1713215505.e2ec414f6ac2.453.2 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4d10e0d0ef09f8f2c20f7dcdd8b37db0895ed217fcde789a4cb2bd5f22e61946
3
+ size 6264
runs/Apr15_20-52-34_e2ec414f6ac2/events.out.tfevents.1713215619.e2ec414f6ac2.453.3 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:371e6de22a67f61fa7556c94321060a9fbf86da5250221c7e2f3783471c7a8bd
3
+ size 6264
runs/Apr15_20-52-34_e2ec414f6ac2/events.out.tfevents.1713215696.e2ec414f6ac2.453.4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d655503c8ddaa3de67a51c201759a1ea8b04a84ae4e5fd2e120dc388bb2cdb50
3
+ size 6264
runs/Apr15_20-52-34_e2ec414f6ac2/events.out.tfevents.1713215851.e2ec414f6ac2.453.5 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:12e95853a2d8389c8c9662c32078c50bfea2a8e78526f39f616878e72669f109
3
+ size 6264
runs/Apr15_20-52-34_e2ec414f6ac2/events.out.tfevents.1713215983.e2ec414f6ac2.453.6 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4ae9df2316ab4b23d4fc669c74eaca53c44a71cdf6d758844c9e81964e21288d
3
+ size 6264
runs/Apr15_20-52-34_e2ec414f6ac2/events.out.tfevents.1713216044.e2ec414f6ac2.453.7 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1499df79a9c5150a6f2bd5eb323a598db1badd2b0c9842d8cd08a8c295d987c0
3
+ size 6264
runs/Apr15_20-52-34_e2ec414f6ac2/events.out.tfevents.1713216099.e2ec414f6ac2.453.8 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c6da9e859a0dd782de8f1c5082e41edc6264991396d520175a3b8265285860bd
3
+ size 6264
runs/Apr15_20-52-34_e2ec414f6ac2/events.out.tfevents.1713216164.e2ec414f6ac2.453.9 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a727f63234f495e551959529076186c741b6bcce7098ebaf5d5c2b169d66e508
3
+ size 6264
runs/Apr15_21-23-57_e2ec414f6ac2/events.out.tfevents.1713216251.e2ec414f6ac2.453.10 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8b0692f37c4a8b60ce58dab4783b1b97e86963f23437cb4ec2a420851495ef65
3
+ size 6546
special_tokens_map.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<|startoftext|>",
4
+ "lstrip": false,
5
+ "normalized": true,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "<|endoftext|>",
11
+ "lstrip": false,
12
+ "normalized": true,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": "!",
17
+ "unk_token": {
18
+ "content": "<|endoftext|>",
19
+ "lstrip": false,
20
+ "normalized": true,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ }
24
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,39 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "added_tokens_decoder": {
4
+ "0": {
5
+ "content": "!",
6
+ "lstrip": false,
7
+ "normalized": false,
8
+ "rstrip": false,
9
+ "single_word": false,
10
+ "special": true
11
+ },
12
+ "49406": {
13
+ "content": "<|startoftext|>",
14
+ "lstrip": false,
15
+ "normalized": true,
16
+ "rstrip": false,
17
+ "single_word": false,
18
+ "special": true
19
+ },
20
+ "49407": {
21
+ "content": "<|endoftext|>",
22
+ "lstrip": false,
23
+ "normalized": true,
24
+ "rstrip": false,
25
+ "single_word": false,
26
+ "special": true
27
+ }
28
+ },
29
+ "bos_token": "<|startoftext|>",
30
+ "clean_up_tokenization_spaces": true,
31
+ "do_lower_case": true,
32
+ "eos_token": "<|endoftext|>",
33
+ "errors": "replace",
34
+ "model_max_length": 16,
35
+ "pad_token": "!",
36
+ "processor_class": "OwlViTProcessor",
37
+ "tokenizer_class": "CLIPTokenizer",
38
+ "unk_token": "<|endoftext|>"
39
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6097db6a8f4d68c524c821ebd95e9b57374e3eebe977ec386d4f75b55ff295d4
3
+ size 4920
vocab.json ADDED
The diff for this file is too large to render. See raw diff