Hugging Face's logo

abdelstark
/
vjepa2-vitl-img16-256-onnx

Image Feature Extraction
ONNX
onnxruntime
vision
video
self-supervised
feature-extraction
vit
jepa
Model card Files
xet
 Community
abdelstark commited on
Commit
5ea037f
·
verified ·
1 Parent(s): 4255752

Upload model.report.json with huggingface_hub

Browse files
Files changed (1) hide show
  1. model.report.json +79 -0
model.report.json ADDED
@@ -0,0 +1,79 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model_id": "facebook/vjepa2-vitl-fpc64-256",
3
+ "onnx_path": "/private/tmp/vjepa2-vitl-img16-256/model.onnx",
4
+ "opset": 17,
5
+ "image_size": 256,
6
+ "repeat_frames": 16,
7
+ "temporal_groups": 8,
8
+ "spatial_tokens": 256,
9
+ "thresholds": {
10
+ "min_patch_cosine": 0.999,
11
+ "max_patch_mean_abs_diff": 0.01,
12
+ "max_patch_max_abs_diff": 0.5
13
+ },
14
+ "validation_records": [
15
+ {
16
+ "image": "buffalo_sample_image.jpg",
17
+ "patch": {
18
+ "cosine": 0.9999999998024888,
19
+ "max_abs_diff": 0.0016198158264160156,
20
+ "mean_abs_diff": 2.3574510123580694e-05
21
+ },
22
+ "allclose_atol": 0.001,
23
+ "allclose_rtol": 0.001,
24
+ "allclose_pass": false,
25
+ "threshold_pass": true
26
+ },
27
+ {
28
+ "image": "cat.jpeg",
29
+ "patch": {
30
+ "cosine": 0.9999999932862402,
31
+ "max_abs_diff": 0.0064334869384765625,
32
+ "mean_abs_diff": 0.00013696696260012686
33
+ },
34
+ "allclose_atol": 0.001,
35
+ "allclose_rtol": 0.001,
36
+ "allclose_pass": false,
37
+ "threshold_pass": true
38
+ },
39
+ {
40
+ "image": "elephant_sample_image.jpg",
41
+ "patch": {
42
+ "cosine": 0.9999999997950355,
43
+ "max_abs_diff": 0.0012044906616210938,
44
+ "mean_abs_diff": 2.6018260541604832e-05
45
+ },
46
+ "allclose_atol": 0.001,
47
+ "allclose_rtol": 0.001,
48
+ "allclose_pass": false,
49
+ "threshold_pass": true
50
+ },
51
+ {
52
+ "image": "rhino_sample_image.jpg",
53
+ "patch": {
54
+ "cosine": 0.9999999998663357,
55
+ "max_abs_diff": 0.0006375312805175781,
56
+ "mean_abs_diff": 2.2421299945563078e-05
57
+ },
58
+ "allclose_atol": 0.001,
59
+ "allclose_rtol": 0.001,
60
+ "allclose_pass": true,
61
+ "threshold_pass": true
62
+ },
63
+ {
64
+ "image": "zebra_sample_image.jpg",
65
+ "patch": {
66
+ "cosine": 0.9999999998178878,
67
+ "max_abs_diff": 0.00168609619140625,
68
+ "mean_abs_diff": 2.2568987333215773e-05
69
+ },
70
+ "allclose_atol": 0.001,
71
+ "allclose_rtol": 0.001,
72
+ "allclose_pass": false,
73
+ "threshold_pass": true
74
+ }
75
+ ],
76
+ "validation_passed": true,
77
+ "input_independence_cosine": 0.31675689132085366,
78
+ "input_independence_threshold": 0.85
79
+ }