Hugging Face's logo

qualcomm
/
MNASNet05

Image Classification
PyTorch
backbone
bu_auto
android
Model card Files
xet
 Community
qaihm-bot commited on
Commit
cac203e
·
verified ·
1 Parent(s): 83aeaef

v0.49.1

Browse files

See https://github.com/qualcomm/ai-hub-models/releases/v0.49.1 for changelog.

Files changed (1) hide show
  1. README.md +52 -62
README.md CHANGED
@@ -16,7 +16,7 @@ pipeline_tag: image-classification
16
  MNASNet05 is a machine learning model that can classify images from the Imagenet dataset. It can also be used as a backbone in building more complex models for specific use cases.
17
 
18
  This is based on the implementation of MNASNet05 found [here](https://github.com/pytorch/vision/blob/main/torchvision/models/mnasnet.py).
19
- This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/main/qai_hub_models/models/mnasnet05) library to export with custom configurations. More details on model performance across various devices, can be found [here](#performance-summary).
20
 
21
  Qualcomm AI Hub Models uses [Qualcomm AI Hub Workbench](https://workbench.aihub.qualcomm.com) to compile, profile, and evaluate this model. [Sign up](https://myaccount.qualcomm.com/signup) to run these models on a hosted Qualcomm® device.
22
 
@@ -25,29 +25,19 @@ There are two ways to deploy this model on your device:
25
 
26
  ### Option 1: Download Pre-Exported Models
27
 
28
- Below are pre-exported model assets ready for deployment.
29
-
30
- | Runtime | Precision | Chipset | SDK Versions | Download |
31
- |---|---|---|---|---|
32
- | ONNX | float | Universal | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/mnasnet05/releases/v0.48.0/mnasnet05-onnx-float.zip)
33
- | ONNX | w8a16 | Universal | QAIRT 2.42, ONNX Runtime 1.24.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/mnasnet05/releases/v0.48.0/mnasnet05-onnx-w8a16.zip)
34
- | QNN_DLC | float | Universal | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/mnasnet05/releases/v0.48.0/mnasnet05-qnn_dlc-float.zip)
35
- | QNN_DLC | w8a16 | Universal | QAIRT 2.43 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/mnasnet05/releases/v0.48.0/mnasnet05-qnn_dlc-w8a16.zip)
36
- | TFLITE | float | Universal | QAIRT 2.43, TFLite 2.17.0 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/mnasnet05/releases/v0.48.0/mnasnet05-tflite-float.zip)
37
-
38
- For more device-specific assets and performance metrics, visit **[MNASNet05 on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/mnasnet05)**.
39
 
40
 
41
  ### Option 2: Export with Custom Configurations
42
 
43
- Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/main/qai_hub_models/models/mnasnet05) Python library to compile and export the model with your own:
44
  - Custom weights (e.g., fine-tuned checkpoints)
45
  - Custom input shapes
46
  - Target device and runtime configurations
47
 
48
  This option is ideal if you need to customize the model beyond the default configuration provided here.
49
 
50
- See our repository for [MNASNet05 on GitHub](https://github.com/qualcomm/ai-hub-models/blob/main/qai_hub_models/models/mnasnet05) for usage instructions.
51
 
52
  ## Model Details
53
 
@@ -63,60 +53,60 @@ See our repository for [MNASNet05 on GitHub](https://github.com/qualcomm/ai-hub-
63
  ## Performance Summary
64
  | Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
65
  |---|---|---|---|---|---|---
66
- | MNASNet05 | ONNX | float | Snapdragon® X2 Elite | 0.227 ms | 5 - 5 MB | NPU
67
- | MNASNet05 | ONNX | float | Snapdragon® X Elite | 0.609 ms | 5 - 5 MB | NPU
68
- | MNASNet05 | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 0.335 ms | 0 - 47 MB | NPU
69
- | MNASNet05 | ONNX | float | Qualcomm® QCS8550 (Proxy) | 0.49 ms | 0 - 2 MB | NPU
70
- | MNASNet05 | ONNX | float | Qualcomm® QCS9075 | 0.761 ms | 1 - 3 MB | NPU
71
- | MNASNet05 | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.266 ms | 0 - 28 MB | NPU
72
  | MNASNet05 | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.217 ms | 0 - 32 MB | NPU
73
- | MNASNet05 | ONNX | w8a16 | Snapdragon® X2 Elite | 0.23 ms | 0 - 0 MB | NPU
74
- | MNASNet05 | ONNX | w8a16 | Snapdragon® X Elite | 0.638 ms | 2 - 2 MB | NPU
75
- | MNASNet05 | ONNX | w8a16 | Snapdragon® 8 Gen 3 Mobile | 0.345 ms | 0 - 41 MB | NPU
76
- | MNASNet05 | ONNX | w8a16 | Qualcomm® QCS6490 | 29.042 ms | 9 - 13 MB | CPU
77
- | MNASNet05 | ONNX | w8a16 | Qualcomm® QCS8550 (Proxy) | 0.518 ms | 0 - 5 MB | NPU
78
- | MNASNet05 | ONNX | w8a16 | Qualcomm® QCS9075 | 0.682 ms | 0 - 3 MB | NPU
79
- | MNASNet05 | ONNX | w8a16 | Qualcomm® QCM6690 | 10.227 ms | 10 - 18 MB | CPU
80
- | MNASNet05 | ONNX | w8a16 | Snapdragon® 8 Elite For Galaxy Mobile | 0.262 ms | 0 - 27 MB | NPU
81
- | MNASNet05 | ONNX | w8a16 | Snapdragon® 7 Gen 4 Mobile | 7.798 ms | 12 - 19 MB | CPU
82
- | MNASNet05 | ONNX | w8a16 | Snapdragon® 8 Elite Gen 5 Mobile | 0.218 ms | 0 - 33 MB | NPU
83
- | MNASNet05 | QNN_DLC | float | Snapdragon® X2 Elite | 0.46 ms | 1 - 1 MB | NPU
84
- | MNASNet05 | QNN_DLC | float | Snapdragon® X Elite | 0.982 ms | 1 - 1 MB | NPU
85
- | MNASNet05 | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 0.517 ms | 0 - 46 MB | NPU
86
- | MNASNet05 | QNN_DLC | float | Qualcomm® QCS8275 (Proxy) | 2.361 ms | 1 - 30 MB | NPU
87
- | MNASNet05 | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 0.819 ms | 1 - 19 MB | NPU
88
- | MNASNet05 | QNN_DLC | float | Qualcomm® SA8775P | 1.121 ms | 1 - 32 MB | NPU
89
- | MNASNet05 | QNN_DLC | float | Qualcomm® QCS9075 | 0.977 ms | 1 - 3 MB | NPU
90
- | MNASNet05 | QNN_DLC | float | Qualcomm® QCS8450 (Proxy) | 1.57 ms | 0 - 48 MB | NPU
91
- | MNASNet05 | QNN_DLC | float | Qualcomm® SA7255P | 2.361 ms | 1 - 30 MB | NPU
92
- | MNASNet05 | QNN_DLC | float | Qualcomm® SA8295P | 1.446 ms | 0 - 28 MB | NPU
93
- | MNASNet05 | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.401 ms | 0 - 33 MB | NPU
94
- | MNASNet05 | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.302 ms | 1 - 33 MB | NPU
95
- | MNASNet05 | QNN_DLC | w8a16 | Snapdragon® X2 Elite | 0.424 ms | 0 - 0 MB | NPU
96
- | MNASNet05 | QNN_DLC | w8a16 | Snapdragon® X Elite | 0.92 ms | 0 - 0 MB | NPU
 
 
 
 
 
 
 
97
  | MNASNet05 | QNN_DLC | w8a16 | Snapdragon® 8 Gen 3 Mobile | 0.527 ms | 0 - 36 MB | NPU
98
- | MNASNet05 | QNN_DLC | w8a16 | Qualcomm® QCS6490 | 2.214 ms | 0 - 2 MB | NPU
99
- | MNASNet05 | QNN_DLC | w8a16 | Qualcomm® QCS8275 (Proxy) | 1.655 ms | 0 - 25 MB | NPU
100
- | MNASNet05 | QNN_DLC | w8a16 | Qualcomm® QCS8550 (Proxy) | 0.785 ms | 0 - 12 MB | NPU
101
- | MNASNet05 | QNN_DLC | w8a16 | Qualcomm® SA8775P | 0.993 ms | 0 - 27 MB | NPU
102
- | MNASNet05 | QNN_DLC | w8a16 | Qualcomm® QCS9075 | 0.927 ms | 2 - 4 MB | NPU
103
- | MNASNet05 | QNN_DLC | w8a16 | Qualcomm® QCM6690 | 3.031 ms | 0 - 138 MB | NPU
104
- | MNASNet05 | QNN_DLC | w8a16 | Qualcomm® QCS8450 (Proxy) | 0.957 ms | 0 - 38 MB | NPU
105
- | MNASNet05 | QNN_DLC | w8a16 | Qualcomm® SA7255P | 1.655 ms | 0 - 25 MB | NPU
106
- | MNASNet05 | QNN_DLC | w8a16 | Qualcomm® SA8295P | 1.273 ms | 0 - 23 MB | NPU
107
- | MNASNet05 | QNN_DLC | w8a16 | Snapdragon® 8 Elite For Galaxy Mobile | 0.362 ms | 0 - 28 MB | NPU
108
  | MNASNet05 | QNN_DLC | w8a16 | Snapdragon® 7 Gen 4 Mobile | 0.795 ms | 0 - 25 MB | NPU
109
- | MNASNet05 | QNN_DLC | w8a16 | Snapdragon® 8 Elite Gen 5 Mobile | 0.292 ms | 0 - 28 MB | NPU
110
- | MNASNet05 | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 0.525 ms | 0 - 48 MB | NPU
111
  | MNASNet05 | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 2.342 ms | 0 - 30 MB | NPU
112
- | MNASNet05 | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 0.804 ms | 0 - 1 MB | NPU
113
- | MNASNet05 | TFLITE | float | Qualcomm® SA8775P | 1.136 ms | 0 - 33 MB | NPU
114
- | MNASNet05 | TFLITE | float | Qualcomm® QCS9075 | 0.986 ms | 0 - 8 MB | NPU
115
- | MNASNet05 | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 1.58 ms | 0 - 50 MB | NPU
116
  | MNASNet05 | TFLITE | float | Qualcomm® SA7255P | 2.342 ms | 0 - 30 MB | NPU
117
- | MNASNet05 | TFLITE | float | Qualcomm® SA8295P | 1.439 ms | 0 - 29 MB | NPU
118
  | MNASNet05 | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.398 ms | 0 - 30 MB | NPU
119
- | MNASNet05 | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.306 ms | 0 - 33 MB | NPU
120
 
121
  ## License
122
  * The license for the original implementation of MNASNet05 can be found
 
16
  MNASNet05 is a machine learning model that can classify images from the Imagenet dataset. It can also be used as a backbone in building more complex models for specific use cases.
17
 
18
  This is based on the implementation of MNASNet05 found [here](https://github.com/pytorch/vision/blob/main/torchvision/models/mnasnet.py).
19
+ This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/tree/v0.49.1/qai_hub_models/models/mnasnet05) library to export with custom configurations. More details on model performance across various devices, can be found [here](#performance-summary).
20
 
21
  Qualcomm AI Hub Models uses [Qualcomm AI Hub Workbench](https://workbench.aihub.qualcomm.com) to compile, profile, and evaluate this model. [Sign up](https://myaccount.qualcomm.com/signup) to run these models on a hosted Qualcomm® device.
22
 
 
25
 
26
  ### Option 1: Download Pre-Exported Models
27
 
28
+ Download pre-exported model assets from **[MNASNet05 on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/mnasnet05)**.
 
 
 
 
 
 
 
 
 
 
29
 
30
 
31
  ### Option 2: Export with Custom Configurations
32
 
33
+ Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/tree/v0.49.1/qai_hub_models/models/mnasnet05) Python library to compile and export the model with your own:
34
  - Custom weights (e.g., fine-tuned checkpoints)
35
  - Custom input shapes
36
  - Target device and runtime configurations
37
 
38
  This option is ideal if you need to customize the model beyond the default configuration provided here.
39
 
40
+ See our repository for [MNASNet05 on GitHub](https://github.com/qualcomm/ai-hub-models/tree/v0.49.1/qai_hub_models/models/mnasnet05) for usage instructions.
41
 
42
  ## Model Details
43
 
 
53
  ## Performance Summary
54
  | Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
55
  |---|---|---|---|---|---|---
 
 
 
 
 
 
56
  | MNASNet05 | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.217 ms | 0 - 32 MB | NPU
57
+ | MNASNet05 | ONNX | float | Snapdragon® X2 Elite | 0.232 ms | 5 - 5 MB | NPU
58
+ | MNASNet05 | ONNX | float | Snapdragon® X Elite | 0.621 ms | 5 - 5 MB | NPU
59
+ | MNASNet05 | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 0.337 ms | 0 - 47 MB | NPU
60
+ | MNASNet05 | ONNX | float | Qualcomm® QCS8550 (Proxy) | 0.488 ms | 0 - 2 MB | NPU
61
+ | MNASNet05 | ONNX | float | Qualcomm® QCS9075 | 0.762 ms | 1 - 3 MB | NPU
62
+ | MNASNet05 | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.28 ms | 0 - 28 MB | NPU
63
+ | MNASNet05 | ONNX | w8a16 | Snapdragon® 8 Elite Gen 5 Mobile | 0.217 ms | 0 - 32 MB | NPU
64
+ | MNASNet05 | ONNX | w8a16 | Snapdragon® X2 Elite | 0.216 ms | 0 - 0 MB | NPU
65
+ | MNASNet05 | ONNX | w8a16 | Snapdragon® X Elite | 0.644 ms | 2 - 2 MB | NPU
66
+ | MNASNet05 | ONNX | w8a16 | Snapdragon® 8 Gen 3 Mobile | 0.344 ms | 0 - 40 MB | NPU
67
+ | MNASNet05 | ONNX | w8a16 | Qualcomm® QCS6490 | 29.144 ms | 10 - 13 MB | CPU
68
+ | MNASNet05 | ONNX | w8a16 | Qualcomm® QCS8550 (Proxy) | 0.519 ms | 0 - 7 MB | NPU
69
+ | MNASNet05 | ONNX | w8a16 | Qualcomm® QCS9075 | 0.711 ms | 0 - 3 MB | NPU
70
+ | MNASNet05 | ONNX | w8a16 | Qualcomm® QCM6690 | 10.406 ms | 9 - 17 MB | CPU
71
+ | MNASNet05 | ONNX | w8a16 | Snapdragon® 8 Elite For Galaxy Mobile | 0.261 ms | 0 - 27 MB | NPU
72
+ | MNASNet05 | ONNX | w8a16 | Snapdragon® 7 Gen 4 Mobile | 7.673 ms | 10 - 18 MB | CPU
73
+ | MNASNet05 | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.301 ms | 1 - 33 MB | NPU
74
+ | MNASNet05 | QNN_DLC | float | Snapdragon® X2 Elite | 0.446 ms | 1 - 1 MB | NPU
75
+ | MNASNet05 | QNN_DLC | float | Snapdragon® X Elite | 0.989 ms | 1 - 1 MB | NPU
76
+ | MNASNet05 | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 0.522 ms | 0 - 46 MB | NPU
77
+ | MNASNet05 | QNN_DLC | float | Qualcomm® QCS8275 (Proxy) | 2.331 ms | 1 - 29 MB | NPU
78
+ | MNASNet05 | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 0.821 ms | 1 - 7 MB | NPU
79
+ | MNASNet05 | QNN_DLC | float | Qualcomm® SA8775P | 1.121 ms | 1 - 31 MB | NPU
80
+ | MNASNet05 | QNN_DLC | float | Qualcomm® QCS9075 | 0.979 ms | 3 - 5 MB | NPU
81
+ | MNASNet05 | QNN_DLC | float | Qualcomm® QCS8450 (Proxy) | 1.573 ms | 0 - 48 MB | NPU
82
+ | MNASNet05 | QNN_DLC | float | Qualcomm® SA7255P | 2.331 ms | 1 - 29 MB | NPU
83
+ | MNASNet05 | QNN_DLC | float | Qualcomm® SA8295P | 1.443 ms | 1 - 29 MB | NPU
84
+ | MNASNet05 | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.401 ms | 0 - 34 MB | NPU
85
+ | MNASNet05 | QNN_DLC | w8a16 | Snapdragon® 8 Elite Gen 5 Mobile | 0.296 ms | 0 - 28 MB | NPU
86
+ | MNASNet05 | QNN_DLC | w8a16 | Snapdragon® X2 Elite | 0.422 ms | 0 - 0 MB | NPU
87
+ | MNASNet05 | QNN_DLC | w8a16 | Snapdragon® X Elite | 0.941 ms | 0 - 0 MB | NPU
88
  | MNASNet05 | QNN_DLC | w8a16 | Snapdragon® 8 Gen 3 Mobile | 0.527 ms | 0 - 36 MB | NPU
89
+ | MNASNet05 | QNN_DLC | w8a16 | Qualcomm® QCS6490 | 2.252 ms | 0 - 2 MB | NPU
90
+ | MNASNet05 | QNN_DLC | w8a16 | Qualcomm® QCS8275 (Proxy) | 1.675 ms | 0 - 26 MB | NPU
91
+ | MNASNet05 | QNN_DLC | w8a16 | Qualcomm® QCS8550 (Proxy) | 0.787 ms | 0 - 2 MB | NPU
92
+ | MNASNet05 | QNN_DLC | w8a16 | Qualcomm® SA8775P | 0.992 ms | 0 - 27 MB | NPU
93
+ | MNASNet05 | QNN_DLC | w8a16 | Qualcomm® QCS9075 | 0.923 ms | 0 - 2 MB | NPU
94
+ | MNASNet05 | QNN_DLC | w8a16 | Qualcomm® QCM6690 | 3.035 ms | 0 - 138 MB | NPU
95
+ | MNASNet05 | QNN_DLC | w8a16 | Qualcomm® QCS8450 (Proxy) | 0.951 ms | 0 - 38 MB | NPU
96
+ | MNASNet05 | QNN_DLC | w8a16 | Qualcomm® SA7255P | 1.675 ms | 0 - 26 MB | NPU
97
+ | MNASNet05 | QNN_DLC | w8a16 | Qualcomm® SA8295P | 1.247 ms | 0 - 23 MB | NPU
98
+ | MNASNet05 | QNN_DLC | w8a16 | Snapdragon® 8 Elite For Galaxy Mobile | 0.363 ms | 0 - 25 MB | NPU
99
  | MNASNet05 | QNN_DLC | w8a16 | Snapdragon® 7 Gen 4 Mobile | 0.795 ms | 0 - 25 MB | NPU
100
+ | MNASNet05 | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.306 ms | 0 - 34 MB | NPU
101
+ | MNASNet05 | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 0.531 ms | 0 - 47 MB | NPU
102
  | MNASNet05 | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 2.342 ms | 0 - 30 MB | NPU
103
+ | MNASNet05 | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 0.801 ms | 0 - 9 MB | NPU
104
+ | MNASNet05 | TFLITE | float | Qualcomm® SA8775P | 4.757 ms | 0 - 31 MB | NPU
105
+ | MNASNet05 | TFLITE | float | Qualcomm® QCS9075 | 0.987 ms | 0 - 8 MB | NPU
106
+ | MNASNet05 | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 1.587 ms | 0 - 49 MB | NPU
107
  | MNASNet05 | TFLITE | float | Qualcomm® SA7255P | 2.342 ms | 0 - 30 MB | NPU
108
+ | MNASNet05 | TFLITE | float | Qualcomm® SA8295P | 1.455 ms | 0 - 29 MB | NPU
109
  | MNASNet05 | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 0.398 ms | 0 - 30 MB | NPU
 
110
 
111
  ## License
112
  * The license for the original implementation of MNASNet05 can be found