Hugging Face's logo

qualcomm
/
MobileNet-v3-Large-Quantized

Image Classification
PyTorch
TF Lite
ONNX
quantized
backbone
real_time
android
Model card Files Community
qaihm-bot commited on
Commit
faa70e3
·
verified ·
1 Parent(s): 051dfa0

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +33 -32
README.md CHANGED
@@ -37,42 +37,42 @@ More details on model performance across various devices, can be found
37
 
38
  | Model | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
39
  |---|---|---|---|---|---|---|---|---|
40
- | MobileNet-v3-Large-Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 0.345 ms | 0 - 26 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.tflite) |
41
- | MobileNet-v3-Large-Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 0.631 ms | 0 - 25 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.so](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.so) |
42
- | MobileNet-v3-Large-Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 0.687 ms | 0 - 19 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.onnx](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.onnx) |
43
- | MobileNet-v3-Large-Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 0.243 ms | 0 - 31 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.tflite) |
44
- | MobileNet-v3-Large-Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 0.454 ms | 0 - 29 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.so](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.so) |
45
- | MobileNet-v3-Large-Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 0.462 ms | 0 - 35 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.onnx](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.onnx) |
46
- | MobileNet-v3-Large-Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 0.257 ms | 0 - 21 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.tflite) |
47
- | MobileNet-v3-Large-Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 0.465 ms | 0 - 21 MB | INT8 | NPU | Use Export Script |
48
- | MobileNet-v3-Large-Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 0.53 ms | 0 - 22 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.onnx](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.onnx) |
49
- | MobileNet-v3-Large-Quantized | RB3 Gen 2 (Proxy) | QCS6490 Proxy | TFLITE | 1.118 ms | 0 - 19 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.tflite) |
50
- | MobileNet-v3-Large-Quantized | RB3 Gen 2 (Proxy) | QCS6490 Proxy | QNN | 1.698 ms | 0 - 9 MB | INT8 | NPU | Use Export Script |
51
- | MobileNet-v3-Large-Quantized | RB5 (Proxy) | QCS8250 Proxy | TFLITE | 7.094 ms | 0 - 3 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.tflite) |
52
- | MobileNet-v3-Large-Quantized | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 0.355 ms | 0 - 26 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.tflite) |
53
- | MobileNet-v3-Large-Quantized | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 0.58 ms | 0 - 3 MB | INT8 | NPU | Use Export Script |
54
- | MobileNet-v3-Large-Quantized | SA7255P ADP | SA7255P | TFLITE | 2.559 ms | 0 - 14 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.tflite) |
55
- | MobileNet-v3-Large-Quantized | SA7255P ADP | SA7255P | QNN | 3.134 ms | 0 - 9 MB | INT8 | NPU | Use Export Script |
56
- | MobileNet-v3-Large-Quantized | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 0.345 ms | 0 - 25 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.tflite) |
57
- | MobileNet-v3-Large-Quantized | SA8255 (Proxy) | SA8255P Proxy | QNN | 0.585 ms | 0 - 3 MB | INT8 | NPU | Use Export Script |
58
- | MobileNet-v3-Large-Quantized | SA8295P ADP | SA8295P | TFLITE | 0.776 ms | 0 - 20 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.tflite) |
59
- | MobileNet-v3-Large-Quantized | SA8295P ADP | SA8295P | QNN | 1.368 ms | 0 - 14 MB | INT8 | NPU | Use Export Script |
60
- | MobileNet-v3-Large-Quantized | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 0.349 ms | 0 - 26 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.tflite) |
61
- | MobileNet-v3-Large-Quantized | SA8650 (Proxy) | SA8650P Proxy | QNN | 0.583 ms | 0 - 3 MB | INT8 | NPU | Use Export Script |
62
- | MobileNet-v3-Large-Quantized | SA8775P ADP | SA8775P | TFLITE | 0.68 ms | 0 - 14 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.tflite) |
63
- | MobileNet-v3-Large-Quantized | SA8775P ADP | SA8775P | QNN | 1.125 ms | 0 - 10 MB | INT8 | NPU | Use Export Script |
64
- | MobileNet-v3-Large-Quantized | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 0.446 ms | 0 - 23 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.tflite) |
65
- | MobileNet-v3-Large-Quantized | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 0.758 ms | 0 - 19 MB | INT8 | NPU | Use Export Script |
66
- | MobileNet-v3-Large-Quantized | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 0.713 ms | 0 - 0 MB | INT8 | NPU | Use Export Script |
67
- | MobileNet-v3-Large-Quantized | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 0.715 ms | 10 - 10 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.onnx](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.onnx) |
68
 
69
 
70
 
71
 
72
  ## Installation
73
 
74
- This model can be installed as a Python package via pip.
75
 
 
76
  ```bash
77
  pip install qai-hub-models
78
  ```
@@ -128,7 +128,7 @@ Profiling Results
128
  MobileNet-v3-Large-Quantized
129
  Device : Samsung Galaxy S23 (13)
130
  Runtime : TFLITE
131
- Estimated inference time (ms) : 0.3
132
  Estimated peak memory usage (MB): [0, 26]
133
  Total # Ops : 137
134
  Compute Unit(s) : NPU (137 ops)
@@ -156,7 +156,7 @@ from qai_hub_models.models.mobilenet_v3_large_quantized import Model
156
  torch_model = Model.from_pretrained()
157
 
158
  # Device
159
- device = hub.Device("Samsung Galaxy S23")
160
 
161
  # Trace model
162
  input_shape = torch_model.get_input_spec()
@@ -248,7 +248,8 @@ Explore all available models on [Qualcomm® AI Hub](https://aihub.qualcomm.com/)
248
 
249
 
250
  ## License
251
- * The license for the original implementation of MobileNet-v3-Large-Quantized can be found [here](https://github.com/pytorch/vision/blob/main/LICENSE).
 
252
  * The license for the compiled assets for on-device deployment can be found [here](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/Qualcomm+AI+Hub+Proprietary+License.pdf)
253
 
254
 
 
37
 
38
  | Model | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
39
  |---|---|---|---|---|---|---|---|---|
40
+ | MobileNet-v3-Large-Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 0.353 ms | 0 - 26 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.tflite) |
41
+ | MobileNet-v3-Large-Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 0.632 ms | 0 - 25 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.so](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.so) |
42
+ | MobileNet-v3-Large-Quantized | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 0.661 ms | 0 - 31 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.onnx](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.onnx) |
43
+ | MobileNet-v3-Large-Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 0.248 ms | 0 - 24 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.tflite) |
44
+ | MobileNet-v3-Large-Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 0.455 ms | 0 - 29 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.so](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.so) |
45
+ | MobileNet-v3-Large-Quantized | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 0.468 ms | 0 - 29 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.onnx](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.onnx) |
46
+ | MobileNet-v3-Large-Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 0.277 ms | 0 - 21 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.tflite) |
47
+ | MobileNet-v3-Large-Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 0.508 ms | 0 - 21 MB | INT8 | NPU | Use Export Script |
48
+ | MobileNet-v3-Large-Quantized | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 0.51 ms | 0 - 24 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.onnx](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.onnx) |
49
+ | MobileNet-v3-Large-Quantized | RB3 Gen 2 (Proxy) | QCS6490 Proxy | TFLITE | 1.131 ms | 0 - 19 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.tflite) |
50
+ | MobileNet-v3-Large-Quantized | RB3 Gen 2 (Proxy) | QCS6490 Proxy | QNN | 1.753 ms | 0 - 11 MB | INT8 | NPU | Use Export Script |
51
+ | MobileNet-v3-Large-Quantized | RB5 (Proxy) | QCS8250 Proxy | TFLITE | 6.989 ms | 0 - 3 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.tflite) |
52
+ | MobileNet-v3-Large-Quantized | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 0.354 ms | 0 - 25 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.tflite) |
53
+ | MobileNet-v3-Large-Quantized | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 0.581 ms | 0 - 3 MB | INT8 | NPU | Use Export Script |
54
+ | MobileNet-v3-Large-Quantized | SA7255P ADP | SA7255P | TFLITE | 2.488 ms | 0 - 14 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.tflite) |
55
+ | MobileNet-v3-Large-Quantized | SA7255P ADP | SA7255P | QNN | 3.021 ms | 0 - 9 MB | INT8 | NPU | Use Export Script |
56
+ | MobileNet-v3-Large-Quantized | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 0.347 ms | 0 - 26 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.tflite) |
57
+ | MobileNet-v3-Large-Quantized | SA8255 (Proxy) | SA8255P Proxy | QNN | 0.582 ms | 0 - 3 MB | INT8 | NPU | Use Export Script |
58
+ | MobileNet-v3-Large-Quantized | SA8295P ADP | SA8295P | TFLITE | 0.77 ms | 0 - 20 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.tflite) |
59
+ | MobileNet-v3-Large-Quantized | SA8295P ADP | SA8295P | QNN | 1.337 ms | 0 - 15 MB | INT8 | NPU | Use Export Script |
60
+ | MobileNet-v3-Large-Quantized | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 0.348 ms | 0 - 25 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.tflite) |
61
+ | MobileNet-v3-Large-Quantized | SA8650 (Proxy) | SA8650P Proxy | QNN | 0.582 ms | 0 - 3 MB | INT8 | NPU | Use Export Script |
62
+ | MobileNet-v3-Large-Quantized | SA8775P ADP | SA8775P | TFLITE | 0.69 ms | 0 - 15 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.tflite) |
63
+ | MobileNet-v3-Large-Quantized | SA8775P ADP | SA8775P | QNN | 1.126 ms | 0 - 10 MB | INT8 | NPU | Use Export Script |
64
+ | MobileNet-v3-Large-Quantized | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 0.456 ms | 0 - 23 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.tflite](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.tflite) |
65
+ | MobileNet-v3-Large-Quantized | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 0.778 ms | 0 - 22 MB | INT8 | NPU | Use Export Script |
66
+ | MobileNet-v3-Large-Quantized | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 0.702 ms | 0 - 0 MB | INT8 | NPU | Use Export Script |
67
+ | MobileNet-v3-Large-Quantized | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 0.678 ms | 9 - 9 MB | INT8 | NPU | [MobileNet-v3-Large-Quantized.onnx](https://huggingface.co/qualcomm/MobileNet-v3-Large-Quantized/blob/main/MobileNet-v3-Large-Quantized.onnx) |
68
 
69
 
70
 
71
 
72
  ## Installation
73
 
 
74
 
75
+ Install the package via pip:
76
  ```bash
77
  pip install qai-hub-models
78
  ```
 
128
  MobileNet-v3-Large-Quantized
129
  Device : Samsung Galaxy S23 (13)
130
  Runtime : TFLITE
131
+ Estimated inference time (ms) : 0.4
132
  Estimated peak memory usage (MB): [0, 26]
133
  Total # Ops : 137
134
  Compute Unit(s) : NPU (137 ops)
 
156
  torch_model = Model.from_pretrained()
157
 
158
  # Device
159
+ device = hub.Device("Samsung Galaxy S24")
160
 
161
  # Trace model
162
  input_shape = torch_model.get_input_spec()
 
248
 
249
 
250
  ## License
251
+ * The license for the original implementation of MobileNet-v3-Large-Quantized can be found
252
+ [here](https://github.com/pytorch/vision/blob/main/LICENSE).
253
  * The license for the compiled assets for on-device deployment can be found [here](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/Qualcomm+AI+Hub+Proprietary+License.pdf)
254
 
255