Instructions to use Chizzsam/shape-models-checkpoints with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Keras
How to use Chizzsam/shape-models-checkpoints with Keras:
# !pip install -U keras tensorflow huggingface_hub # Keras needs TensorFlow installed to read "hf://" paths, so the tensorflow backend is selected here; # "jax" and "torch" also work for computation once TensorFlow is installed. import os os.environ["KERAS_BACKEND"] = "tensorflow" import keras model = keras.saving.load_model("hf://Chizzsam/shape-models-checkpoints") - Notebooks
- Google Colab
- Kaggle
Upload body/body_shape_norm.json with huggingface_hub
Browse files- body/body_shape_norm.json +33 -21
body/body_shape_norm.json
CHANGED
|
@@ -1,5 +1,6 @@
|
|
| 1 |
{
|
| 2 |
-
"model_version": "3.
|
|
|
|
| 3 |
"feature_names": [
|
| 4 |
"shoulder_hip_ratio",
|
| 5 |
"waist_hip_ratio",
|
|
@@ -10,15 +11,26 @@
|
|
| 10 |
"hip_shoulder_diff",
|
| 11 |
"hip_waist_diff"
|
| 12 |
],
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 13 |
"mean": [
|
| 14 |
0.37780097126960754,
|
| 15 |
0.8161380887031555,
|
| 16 |
2.200714111328125,
|
| 17 |
0.9687443375587463,
|
| 18 |
1.1990960836410522,
|
| 19 |
-
-
|
| 20 |
-
|
| 21 |
-
|
| 22 |
],
|
| 23 |
"std": [
|
| 24 |
0.05251329392194748,
|
|
@@ -26,9 +38,9 @@
|
|
| 26 |
0.3728339374065399,
|
| 27 |
0.09726245701313019,
|
| 28 |
0.17114204168319702,
|
| 29 |
-
|
| 30 |
-
|
| 31 |
-
|
| 32 |
],
|
| 33 |
"raw_min": [
|
| 34 |
0.2409999966621399,
|
|
@@ -36,9 +48,9 @@
|
|
| 36 |
1.417099952697754,
|
| 37 |
0.6786999702453613,
|
| 38 |
0.8282999992370605,
|
| 39 |
-
-
|
| 40 |
-
|
| 41 |
-
-
|
| 42 |
],
|
| 43 |
"raw_max": [
|
| 44 |
0.5447999835014343,
|
|
@@ -46,9 +58,9 @@
|
|
| 46 |
3.4370999336242676,
|
| 47 |
1.2482999563217163,
|
| 48 |
1.700600028038025,
|
| 49 |
-
-
|
| 50 |
-
|
| 51 |
-
|
| 52 |
],
|
| 53 |
"normalized_min": [
|
| 54 |
-2.6050732135772705,
|
|
@@ -56,9 +68,9 @@
|
|
| 56 |
-2.101778030395508,
|
| 57 |
-2.982079267501831,
|
| 58 |
-2.1665985584259033,
|
| 59 |
-
-2.
|
| 60 |
-
-
|
| 61 |
-
-2.
|
| 62 |
],
|
| 63 |
"normalized_max": [
|
| 64 |
3.180128335952759,
|
|
@@ -66,9 +78,9 @@
|
|
| 66 |
3.3161838054656982,
|
| 67 |
2.874239683151245,
|
| 68 |
2.930337429046631,
|
| 69 |
-
2.
|
| 70 |
-
2.
|
| 71 |
-
2.
|
| 72 |
],
|
| 73 |
"class_names": [
|
| 74 |
"Hourglass",
|
|
@@ -83,7 +95,7 @@
|
|
| 83 |
8
|
| 84 |
],
|
| 85 |
"dtype": "float32",
|
| 86 |
-
"note": "standardize raw
|
| 87 |
},
|
| 88 |
"output": {
|
| 89 |
"shape": [
|
|
@@ -93,5 +105,5 @@
|
|
| 93 |
"dtype": "float32",
|
| 94 |
"type": "softmax"
|
| 95 |
},
|
| 96 |
-
"note_app_side": "
|
| 97 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"model_version": "3.3.0",
|
| 3 |
+
"feature_schema_version": 2,
|
| 4 |
"feature_names": [
|
| 5 |
"shoulder_hip_ratio",
|
| 6 |
"waist_hip_ratio",
|
|
|
|
| 11 |
"hip_shoulder_diff",
|
| 12 |
"hip_waist_diff"
|
| 13 |
],
|
| 14 |
+
"feature_units": {
|
| 15 |
+
"shoulder_hip_ratio": "dimensionless (ratio of two widths)",
|
| 16 |
+
"waist_hip_ratio": "dimensionless (ratio of two widths)",
|
| 17 |
+
"waist_shoulder_ratio": "dimensionless (ratio of two widths)",
|
| 18 |
+
"bust_hip_ratio": "dimensionless (ratio of two widths)",
|
| 19 |
+
"bust_waist_ratio": "dimensionless (ratio of two widths)",
|
| 20 |
+
"shoulder_waist_diff": "dimensionless (width diff / hip width)",
|
| 21 |
+
"hip_shoulder_diff": "dimensionless (width diff / hip width)",
|
| 22 |
+
"hip_waist_diff": "dimensionless (width diff / hip width)"
|
| 23 |
+
},
|
| 24 |
+
"hip_cm_reference": 1.0,
|
| 25 |
"mean": [
|
| 26 |
0.37780097126960754,
|
| 27 |
0.8161380887031555,
|
| 28 |
2.200714111328125,
|
| 29 |
0.9687443375587463,
|
| 30 |
1.1990960836410522,
|
| 31 |
+
-0.43833690881729126,
|
| 32 |
+
0.6221989393234253,
|
| 33 |
+
0.18386207520961761
|
| 34 |
],
|
| 35 |
"std": [
|
| 36 |
0.05251329392194748,
|
|
|
|
| 38 |
0.3728339374065399,
|
| 39 |
0.09726245701313019,
|
| 40 |
0.17114204168319702,
|
| 41 |
+
0.0974234789609909,
|
| 42 |
+
0.05251329392194748,
|
| 43 |
+
0.08545612543821335
|
| 44 |
],
|
| 45 |
"raw_min": [
|
| 46 |
0.2409999966621399,
|
|
|
|
| 48 |
1.417099952697754,
|
| 49 |
0.6786999702453613,
|
| 50 |
0.8282999992370605,
|
| 51 |
+
-0.7000498175621033,
|
| 52 |
+
0.45520004630088806,
|
| 53 |
+
-0.04578861966729164
|
| 54 |
],
|
| 55 |
"raw_max": [
|
| 56 |
0.5447999835014343,
|
|
|
|
| 58 |
3.4370999336242676,
|
| 59 |
1.2482999563217163,
|
| 60 |
1.700600028038025,
|
| 61 |
+
-0.20338687300682068,
|
| 62 |
+
0.7590000033378601,
|
| 63 |
+
0.38873669505119324
|
| 64 |
],
|
| 65 |
"normalized_min": [
|
| 66 |
-2.6050732135772705,
|
|
|
|
| 68 |
-2.101778030395508,
|
| 69 |
-2.982079267501831,
|
| 70 |
-2.1665985584259033,
|
| 71 |
+
-2.686343193054199,
|
| 72 |
+
-3.1801259517669678,
|
| 73 |
+
-2.687351942062378
|
| 74 |
],
|
| 75 |
"normalized_max": [
|
| 76 |
3.180128335952759,
|
|
|
|
| 78 |
3.3161838054656982,
|
| 79 |
2.874239683151245,
|
| 80 |
2.930337429046631,
|
| 81 |
+
2.4116365909576416,
|
| 82 |
+
2.605074882507324,
|
| 83 |
+
2.3974246978759766
|
| 84 |
],
|
| 85 |
"class_names": [
|
| 86 |
"Hourglass",
|
|
|
|
| 95 |
8
|
| 96 |
],
|
| 97 |
"dtype": "float32",
|
| 98 |
+
"note": "standardize raw values as (x - mean) / std BEFORE inference"
|
| 99 |
},
|
| 100 |
"output": {
|
| 101 |
"shape": [
|
|
|
|
| 105 |
"dtype": "float32",
|
| 106 |
"type": "softmax"
|
| 107 |
},
|
| 108 |
+
"note_app_side": "All 8 features are DIMENSIONLESS. Compute them from pose keypoint distances in pixels: ratios as the width ratio directly, and diffs as (widthA - widthB) / hipWidth. Then standardize with mean/std. hip_cm_reference should be ignored (it is 1.0)."
|
| 109 |
}
|