Hugging Face's logo

qualcomm
/
EfficientNet-V2-s

Image Classification
PyTorch
backbone
bu_auto
android
Model card Files
xet
Community
qaihm-bot commited on
Commit
9d80703
·
verified ·
1 Parent(s): 1d092ac

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +41 -28
README.md CHANGED
@@ -1,6 +1,6 @@
1
  ---
2
  library_name: pytorch
3
- license: bsd-3-clause
4
  tags:
5
  - backbone
6
  - android
@@ -26,34 +26,47 @@ More details on model performance across various devices, can be found
26
 
27
  ### Model Details
28
 
29
- - **Model Type:** Image classification
30
  - **Model Stats:**
31
  - Model checkpoint: Imagenet
32
  - Input resolution: 384x384
33
  - Number of parameters: 21.45M
34
  - Model size: 82.7 MB
35
 
36
- | Model | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
37
  |---|---|---|---|---|---|---|---|---|
38
- | EfficientNet-V2-s | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 2.679 ms | 0 - 324 MB | FP16 | NPU | [EfficientNet-V2-s.tflite](https://huggingface.co/qualcomm/EfficientNet-V2-s/blob/main/EfficientNet-V2-s.tflite) |
39
- | EfficientNet-V2-s | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 2.693 ms | 0 - 144 MB | FP16 | NPU | [EfficientNet-V2-s.so](https://huggingface.co/qualcomm/EfficientNet-V2-s/blob/main/EfficientNet-V2-s.so) |
40
- | EfficientNet-V2-s | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 2.596 ms | 0 - 98 MB | FP16 | NPU | [EfficientNet-V2-s.onnx](https://huggingface.co/qualcomm/EfficientNet-V2-s/blob/main/EfficientNet-V2-s.onnx) |
41
- | EfficientNet-V2-s | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 2.0 ms | 0 - 99 MB | FP16 | NPU | [EfficientNet-V2-s.tflite](https://huggingface.co/qualcomm/EfficientNet-V2-s/blob/main/EfficientNet-V2-s.tflite) |
42
- | EfficientNet-V2-s | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 2.022 ms | 1 - 59 MB | FP16 | NPU | [EfficientNet-V2-s.so](https://huggingface.co/qualcomm/EfficientNet-V2-s/blob/main/EfficientNet-V2-s.so) |
43
- | EfficientNet-V2-s | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 1.906 ms | 0 - 74 MB | FP16 | NPU | [EfficientNet-V2-s.onnx](https://huggingface.co/qualcomm/EfficientNet-V2-s/blob/main/EfficientNet-V2-s.onnx) |
44
- | EfficientNet-V2-s | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 1.852 ms | 0 - 75 MB | FP16 | NPU | [EfficientNet-V2-s.tflite](https://huggingface.co/qualcomm/EfficientNet-V2-s/blob/main/EfficientNet-V2-s.tflite) |
45
- | EfficientNet-V2-s | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 1.574 ms | 1 - 30 MB | FP16 | NPU | Use Export Script |
46
- | EfficientNet-V2-s | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 1.854 ms | 0 - 37 MB | FP16 | NPU | [EfficientNet-V2-s.onnx](https://huggingface.co/qualcomm/EfficientNet-V2-s/blob/main/EfficientNet-V2-s.onnx) |
47
- | EfficientNet-V2-s | QCS8275 (Proxy) | QCS8275 Proxy | TFLITE | 60.037 ms | 0 - 71 MB | FP16 | NPU | [EfficientNet-V2-s.tflite](https://huggingface.co/qualcomm/EfficientNet-V2-s/blob/main/EfficientNet-V2-s.tflite) |
48
- | EfficientNet-V2-s | QCS8275 (Proxy) | QCS8275 Proxy | QNN | 60.03 ms | 1 - 10 MB | FP16 | NPU | Use Export Script |
49
- | EfficientNet-V2-s | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 2.667 ms | 0 - 331 MB | FP16 | NPU | [EfficientNet-V2-s.tflite](https://huggingface.co/qualcomm/EfficientNet-V2-s/blob/main/EfficientNet-V2-s.tflite) |
50
- | EfficientNet-V2-s | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 2.519 ms | 1 - 4 MB | FP16 | NPU | Use Export Script |
51
- | EfficientNet-V2-s | QCS9075 (Proxy) | QCS9075 Proxy | TFLITE | 4.61 ms | 0 - 71 MB | FP16 | NPU | [EfficientNet-V2-s.tflite](https://huggingface.co/qualcomm/EfficientNet-V2-s/blob/main/EfficientNet-V2-s.tflite) |
52
- | EfficientNet-V2-s | QCS9075 (Proxy) | QCS9075 Proxy | QNN | 4.365 ms | 1 - 10 MB | FP16 | NPU | Use Export Script |
53
- | EfficientNet-V2-s | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 5.081 ms | 0 - 89 MB | FP16 | NPU | [EfficientNet-V2-s.tflite](https://huggingface.co/qualcomm/EfficientNet-V2-s/blob/main/EfficientNet-V2-s.tflite) |
54
- | EfficientNet-V2-s | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 5.162 ms | 1 - 42 MB | FP16 | NPU | Use Export Script |
55
- | EfficientNet-V2-s | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 2.796 ms | 1 - 1 MB | FP16 | NPU | Use Export Script |
56
- | EfficientNet-V2-s | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 2.716 ms | 48 - 48 MB | FP16 | NPU | [EfficientNet-V2-s.onnx](https://huggingface.co/qualcomm/EfficientNet-V2-s/blob/main/EfficientNet-V2-s.onnx) |
 
 
 
 
 
 
 
 
 
 
 
 
 
57
 
58
 
59
 
@@ -115,12 +128,12 @@ python -m qai_hub_models.models.efficientnet_v2_s.export
115
  Profiling Results
116
  ------------------------------------------------------------
117
  EfficientNet-V2-s
118
- Device : Samsung Galaxy S23 (13)
119
- Runtime : TFLITE
120
- Estimated inference time (ms) : 2.7
121
- Estimated peak memory usage (MB): [0, 324]
122
- Total # Ops : 508
123
- Compute Unit(s) : NPU (508 ops)
124
  ```
125
 
126
 
 
1
  ---
2
  library_name: pytorch
3
+ license: other
4
  tags:
5
  - backbone
6
  - android
 
26
 
27
  ### Model Details
28
 
29
+ - **Model Type:** Model_use_case.image_classification
30
  - **Model Stats:**
31
  - Model checkpoint: Imagenet
32
  - Input resolution: 384x384
33
  - Number of parameters: 21.45M
34
  - Model size: 82.7 MB
35
 
36
+ | Model | Precision | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit | Target Model
37
  |---|---|---|---|---|---|---|---|---|
38
+ | EfficientNet-V2-s | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 10.972 ms | 0 - 71 MB | NPU | [EfficientNet-V2-s.tflite](https://huggingface.co/qualcomm/EfficientNet-V2-s/blob/main/EfficientNet-V2-s.tflite) |
39
+ | EfficientNet-V2-s | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN | 10.795 ms | 1 - 10 MB | NPU | Use Export Script |
40
+ | EfficientNet-V2-s | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 5.227 ms | 0 - 86 MB | NPU | [EfficientNet-V2-s.tflite](https://huggingface.co/qualcomm/EfficientNet-V2-s/blob/main/EfficientNet-V2-s.tflite) |
41
+ | EfficientNet-V2-s | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN | 5.754 ms | 1 - 43 MB | NPU | Use Export Script |
42
+ | EfficientNet-V2-s | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 2.666 ms | 0 - 328 MB | NPU | [EfficientNet-V2-s.tflite](https://huggingface.co/qualcomm/EfficientNet-V2-s/blob/main/EfficientNet-V2-s.tflite) |
43
+ | EfficientNet-V2-s | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN | 2.516 ms | 1 - 9 MB | NPU | Use Export Script |
44
+ | EfficientNet-V2-s | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 3.858 ms | 0 - 72 MB | NPU | [EfficientNet-V2-s.tflite](https://huggingface.co/qualcomm/EfficientNet-V2-s/blob/main/EfficientNet-V2-s.tflite) |
45
+ | EfficientNet-V2-s | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN | 3.606 ms | 1 - 15 MB | NPU | Use Export Script |
46
+ | EfficientNet-V2-s | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | TFLITE | 2.679 ms | 0 - 327 MB | NPU | [EfficientNet-V2-s.tflite](https://huggingface.co/qualcomm/EfficientNet-V2-s/blob/main/EfficientNet-V2-s.tflite) |
47
+ | EfficientNet-V2-s | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | QNN | 2.544 ms | 0 - 25 MB | NPU | Use Export Script |
48
+ | EfficientNet-V2-s | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | ONNX | 2.55 ms | 0 - 92 MB | NPU | [EfficientNet-V2-s.onnx](https://huggingface.co/qualcomm/EfficientNet-V2-s/blob/main/EfficientNet-V2-s.onnx) |
49
+ | EfficientNet-V2-s | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 1.999 ms | 0 - 99 MB | NPU | [EfficientNet-V2-s.tflite](https://huggingface.co/qualcomm/EfficientNet-V2-s/blob/main/EfficientNet-V2-s.tflite) |
50
+ | EfficientNet-V2-s | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN | 1.862 ms | 1 - 59 MB | NPU | Use Export Script |
51
+ | EfficientNet-V2-s | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 1.878 ms | 0 - 69 MB | NPU | [EfficientNet-V2-s.onnx](https://huggingface.co/qualcomm/EfficientNet-V2-s/blob/main/EfficientNet-V2-s.onnx) |
52
+ | EfficientNet-V2-s | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | TFLITE | 1.853 ms | 0 - 75 MB | NPU | [EfficientNet-V2-s.tflite](https://huggingface.co/qualcomm/EfficientNet-V2-s/blob/main/EfficientNet-V2-s.tflite) |
53
+ | EfficientNet-V2-s | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | QNN | 1.494 ms | 1 - 33 MB | NPU | Use Export Script |
54
+ | EfficientNet-V2-s | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | ONNX | 1.841 ms | 0 - 35 MB | NPU | [EfficientNet-V2-s.onnx](https://huggingface.co/qualcomm/EfficientNet-V2-s/blob/main/EfficientNet-V2-s.onnx) |
55
+ | EfficientNet-V2-s | float | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 2.818 ms | 1 - 1 MB | NPU | Use Export Script |
56
+ | EfficientNet-V2-s | float | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 2.736 ms | 47 - 47 MB | NPU | [EfficientNet-V2-s.onnx](https://huggingface.co/qualcomm/EfficientNet-V2-s/blob/main/EfficientNet-V2-s.onnx) |
57
+ | EfficientNet-V2-s | w8a16 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN | 5.323 ms | 0 - 10 MB | NPU | Use Export Script |
58
+ | EfficientNet-V2-s | w8a16 | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN | 3.367 ms | 0 - 63 MB | NPU | Use Export Script |
59
+ | EfficientNet-V2-s | w8a16 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN | 2.618 ms | 0 - 3 MB | NPU | Use Export Script |
60
+ | EfficientNet-V2-s | w8a16 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN | 2.984 ms | 0 - 15 MB | NPU | Use Export Script |
61
+ | EfficientNet-V2-s | w8a16 | RB3 Gen 2 (Proxy) | Qualcomm® QCS6490 (Proxy) | QNN | 8.869 ms | 0 - 14 MB | NPU | Use Export Script |
62
+ | EfficientNet-V2-s | w8a16 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | QNN | 2.644 ms | 0 - 18 MB | NPU | Use Export Script |
63
+ | EfficientNet-V2-s | w8a16 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | ONNX | 2.828 ms | 0 - 46 MB | NPU | [EfficientNet-V2-s.onnx](https://huggingface.co/qualcomm/EfficientNet-V2-s/blob/main/EfficientNet-V2-s_w8a16.onnx) |
64
+ | EfficientNet-V2-s | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN | 1.777 ms | 0 - 62 MB | NPU | Use Export Script |
65
+ | EfficientNet-V2-s | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 1.912 ms | 0 - 90 MB | NPU | [EfficientNet-V2-s.onnx](https://huggingface.co/qualcomm/EfficientNet-V2-s/blob/main/EfficientNet-V2-s_w8a16.onnx) |
66
+ | EfficientNet-V2-s | w8a16 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | QNN | 1.5 ms | 0 - 58 MB | NPU | Use Export Script |
67
+ | EfficientNet-V2-s | w8a16 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | ONNX | 1.816 ms | 0 - 73 MB | NPU | [EfficientNet-V2-s.onnx](https://huggingface.co/qualcomm/EfficientNet-V2-s/blob/main/EfficientNet-V2-s_w8a16.onnx) |
68
+ | EfficientNet-V2-s | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 2.856 ms | 0 - 0 MB | NPU | Use Export Script |
69
+ | EfficientNet-V2-s | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 3.007 ms | 25 - 25 MB | NPU | [EfficientNet-V2-s.onnx](https://huggingface.co/qualcomm/EfficientNet-V2-s/blob/main/EfficientNet-V2-s_w8a16.onnx) |
70
 
71
 
72
 
 
128
  Profiling Results
129
  ------------------------------------------------------------
130
  EfficientNet-V2-s
131
+ Device : cs_8275 (ANDROID 14)
132
+ Runtime : TFLITE
133
+ Estimated inference time (ms) : 11.0
134
+ Estimated peak memory usage (MB): [0, 71]
135
+ Total # Ops : 508
136
+ Compute Unit(s) : npu (508 ops) gpu (0 ops) cpu (0 ops)
137
  ```
138
 
139