Hugging Face's logo

qualcomm
/
EfficientNet-B0

Image Classification
PyTorch
LiteRT
backbone
bu_auto
android
Model card Files
xet
Community
qaihm-bot commited on
Commit
34b4028
·
verified ·
1 Parent(s): 4414548

v0.42.0

Browse files

See https://github.com/quic/ai-hub-models/releases/v0.42.0 for changelog.

Files changed (5) hide show
  1. EfficientNet-B0_float.dlc +2 -2
  2. EfficientNet-B0_float.onnx.zip +1 -1
  3. EfficientNet-B0_w8a16.dlc +2 -2
  4. EfficientNet-B0_w8a16.onnx.zip +2 -2
  5. README.md +57 -54
EfficientNet-B0_float.dlc CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:6ab376af4f62249458cb6c5cad16a35a9e1f22d855f5b6d9e85c9e6831881d67
3
- size 21333908
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:19f04ac978c8169ad377d7f6ff6370ef5ab27d8cf25670af2f8dd284a61d3ba8
3
+ size 21334012
EfficientNet-B0_float.onnx.zip CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7f0f0e45fdf20043efb54a7f62e49e663d7cd6b4e77699b3ec4081331609c547
3
  size 19669241
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ea2b962cc201457bb158dd51bd6ce2e728901a2d4b37d520232164e7ad3df1e8
3
  size 19669241
EfficientNet-B0_w8a16.dlc CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:687ec56312d12534736b17ea89640a71d83867fa9e2af1b5a75a59f1f5a85945
3
- size 7341060
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:eb627c68d8dac2496d1216dfdf55564ccb9b32ecde0c30e3ec94b4269a4a5df1
3
+ size 7341164
EfficientNet-B0_w8a16.onnx.zip CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:36900b0c75150b06cdb7dedb52f1e596e0c39aecd1ba9882a0d2512a3bd4a900
3
- size 11214634
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:73f55b929d42372cfb4b116e3ebbcef7b0f389689d73407f1af993e6060447fd
3
+ size 11215237
README.md CHANGED
@@ -37,56 +37,59 @@ More details on model performance across various devices, can be found
37
 
38
  | Model | Precision | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit | Target Model
39
  |---|---|---|---|---|---|---|---|---|
40
- | EfficientNet-B0 | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 4.863 ms | 0 - 29 MB | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
41
- | EfficientNet-B0 | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC | 4.714 ms | 0 - 23 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.dlc) |
42
- | EfficientNet-B0 | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 2.915 ms | 0 - 44 MB | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
43
- | EfficientNet-B0 | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 3.455 ms | 1 - 39 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.dlc) |
44
- | EfficientNet-B0 | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 1.478 ms | 0 - 117 MB | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
45
- | EfficientNet-B0 | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 1.443 ms | 1 - 9 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.dlc) |
46
- | EfficientNet-B0 | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | ONNX | 1.441 ms | 0 - 94 MB | NPU | [EfficientNet-B0.onnx.zip](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.onnx.zip) |
47
- | EfficientNet-B0 | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 1.949 ms | 0 - 29 MB | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
48
- | EfficientNet-B0 | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC | 1.938 ms | 1 - 24 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.dlc) |
49
- | EfficientNet-B0 | float | SA7255P ADP | Qualcomm® SA7255P | TFLITE | 4.863 ms | 0 - 29 MB | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
50
- | EfficientNet-B0 | float | SA7255P ADP | Qualcomm® SA7255P | QNN_DLC | 4.714 ms | 0 - 23 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.dlc) |
51
- | EfficientNet-B0 | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | TFLITE | 1.476 ms | 0 - 117 MB | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
52
- | EfficientNet-B0 | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN_DLC | 1.448 ms | 1 - 70 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.dlc) |
53
- | EfficientNet-B0 | float | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 3.582 ms | 0 - 36 MB | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
54
- | EfficientNet-B0 | float | SA8295P ADP | Qualcomm® SA8295P | QNN_DLC | 3.52 ms | 1 - 32 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.dlc) |
55
- | EfficientNet-B0 | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | TFLITE | 1.472 ms | 0 - 117 MB | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
56
- | EfficientNet-B0 | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN_DLC | 1.446 ms | 0 - 92 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.dlc) |
57
- | EfficientNet-B0 | float | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 1.949 ms | 0 - 29 MB | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
58
- | EfficientNet-B0 | float | SA8775P ADP | Qualcomm® SA8775P | QNN_DLC | 1.938 ms | 1 - 24 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.dlc) |
59
- | EfficientNet-B0 | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 1.024 ms | 0 - 42 MB | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
60
- | EfficientNet-B0 | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 1.013 ms | 1 - 42 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.dlc) |
61
- | EfficientNet-B0 | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 1.001 ms | 0 - 40 MB | NPU | [EfficientNet-B0.onnx.zip](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.onnx.zip) |
62
- | EfficientNet-B0 | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | TFLITE | 0.822 ms | 0 - 30 MB | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
63
- | EfficientNet-B0 | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | QNN_DLC | 0.783 ms | 1 - 32 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.dlc) |
64
- | EfficientNet-B0 | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | ONNX | 0.811 ms | 0 - 28 MB | NPU | [EfficientNet-B0.onnx.zip](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.onnx.zip) |
65
- | EfficientNet-B0 | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | TFLITE | 0.635 ms | 0 - 36 MB | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
66
- | EfficientNet-B0 | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | QNN_DLC | 0.608 ms | 1 - 32 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.dlc) |
67
- | EfficientNet-B0 | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | ONNX | 0.687 ms | 1 - 31 MB | NPU | [EfficientNet-B0.onnx.zip](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.onnx.zip) |
68
- | EfficientNet-B0 | float | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_DLC | 1.652 ms | 79 - 79 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.dlc) |
69
- | EfficientNet-B0 | float | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 1.395 ms | 13 - 13 MB | NPU | [EfficientNet-B0.onnx.zip](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.onnx.zip) |
70
- | EfficientNet-B0 | w8a16 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC | 3.328 ms | 0 - 29 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.dlc) |
71
- | EfficientNet-B0 | w8a16 | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 2.034 ms | 0 - 40 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.dlc) |
72
- | EfficientNet-B0 | w8a16 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 1.685 ms | 0 - 39 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.dlc) |
73
- | EfficientNet-B0 | w8a16 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | ONNX | 1.728 ms | 0 - 52 MB | NPU | [EfficientNet-B0.onnx.zip](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.onnx.zip) |
74
- | EfficientNet-B0 | w8a16 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC | 2.001 ms | 0 - 29 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.dlc) |
75
- | EfficientNet-B0 | w8a16 | RB3 Gen 2 (Proxy) | Qualcomm® QCS6490 (Proxy) | ONNX | 61.267 ms | 39 - 54 MB | CPU | [EfficientNet-B0.onnx.zip](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.onnx.zip) |
76
- | EfficientNet-B0 | w8a16 | RB5 (Proxy) | Qualcomm® QCS8250 (Proxy) | ONNX | 48.16 ms | 35 - 45 MB | CPU | [EfficientNet-B0.onnx.zip](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.onnx.zip) |
77
- | EfficientNet-B0 | w8a16 | SA7255P ADP | Qualcomm® SA7255P | QNN_DLC | 3.328 ms | 0 - 29 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.dlc) |
78
- | EfficientNet-B0 | w8a16 | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN_DLC | 1.673 ms | 0 - 39 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.dlc) |
79
- | EfficientNet-B0 | w8a16 | SA8295P ADP | Qualcomm® SA8295P | QNN_DLC | 2.424 ms | 0 - 34 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.dlc) |
80
- | EfficientNet-B0 | w8a16 | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN_DLC | 1.685 ms | 0 - 39 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.dlc) |
81
- | EfficientNet-B0 | w8a16 | SA8775P ADP | Qualcomm® SA8775P | QNN_DLC | 2.001 ms | 0 - 29 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.dlc) |
82
- | EfficientNet-B0 | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 1.151 ms | 0 - 41 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.dlc) |
83
- | EfficientNet-B0 | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 1.131 ms | 0 - 48 MB | NPU | [EfficientNet-B0.onnx.zip](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.onnx.zip) |
84
- | EfficientNet-B0 | w8a16 | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | QNN_DLC | 0.78 ms | 0 - 35 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.dlc) |
85
- | EfficientNet-B0 | w8a16 | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | ONNX | 0.853 ms | 0 - 42 MB | NPU | [EfficientNet-B0.onnx.zip](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.onnx.zip) |
86
- | EfficientNet-B0 | w8a16 | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | QNN_DLC | 0.641 ms | 0 - 33 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.dlc) |
87
- | EfficientNet-B0 | w8a16 | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | ONNX | 0.768 ms | 0 - 42 MB | NPU | [EfficientNet-B0.onnx.zip](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.onnx.zip) |
88
- | EfficientNet-B0 | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_DLC | 1.905 ms | 18 - 18 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.dlc) |
89
- | EfficientNet-B0 | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 1.668 ms | 8 - 8 MB | NPU | [EfficientNet-B0.onnx.zip](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.onnx.zip) |
 
 
 
90
 
91
 
92
 
@@ -100,9 +103,9 @@ pip install qai-hub-models
100
  ```
101
 
102
 
103
- ## Configure Qualcomm® AI Hub to run this model on a cloud-hosted device
104
 
105
- Sign-in to [Qualcomm® AI Hub](https://app.aihub.qualcomm.com/) with your
106
  Qualcomm® ID. Once signed in navigate to `Account -> Settings -> API Token`.
107
 
108
  With this API token, you can configure your client to run models on the cloud
@@ -110,7 +113,7 @@ hosted devices.
110
  ```bash
111
  qai-hub configure --api_token API_TOKEN
112
  ```
113
- Navigate to [docs](https://app.aihub.qualcomm.com/docs/) for more information.
114
 
115
 
116
 
@@ -221,7 +224,7 @@ With the output of the model, you can compute like PSNR, relative errors or
221
  spot check the output with expected output.
222
 
223
  **Note**: This on-device profiling and inference requires access to Qualcomm®
224
- AI Hub. [Sign up for access](https://myaccount.qualcomm.com/signup).
225
 
226
 
227
 
 
37
 
38
  | Model | Precision | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit | Target Model
39
  |---|---|---|---|---|---|---|---|---|
40
+ | EfficientNet-B0 | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 4.88 ms | 0 - 31 MB | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
41
+ | EfficientNet-B0 | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC | 4.749 ms | 1 - 25 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.dlc) |
42
+ | EfficientNet-B0 | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 2.906 ms | 0 - 47 MB | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
43
+ | EfficientNet-B0 | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 3.438 ms | 1 - 39 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.dlc) |
44
+ | EfficientNet-B0 | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 1.48 ms | 0 - 117 MB | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
45
+ | EfficientNet-B0 | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 1.465 ms | 0 - 69 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.dlc) |
46
+ | EfficientNet-B0 | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | ONNX | 1.467 ms | 0 - 72 MB | NPU | [EfficientNet-B0.onnx.zip](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.onnx.zip) |
47
+ | EfficientNet-B0 | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 1.967 ms | 0 - 32 MB | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
48
+ | EfficientNet-B0 | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC | 1.965 ms | 1 - 25 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.dlc) |
49
+ | EfficientNet-B0 | float | SA7255P ADP | Qualcomm® SA7255P | TFLITE | 4.88 ms | 0 - 31 MB | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
50
+ | EfficientNet-B0 | float | SA7255P ADP | Qualcomm® SA7255P | QNN_DLC | 4.749 ms | 1 - 25 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.dlc) |
51
+ | EfficientNet-B0 | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | TFLITE | 1.482 ms | 0 - 117 MB | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
52
+ | EfficientNet-B0 | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN_DLC | 1.463 ms | 1 - 70 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.dlc) |
53
+ | EfficientNet-B0 | float | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 3.598 ms | 0 - 40 MB | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
54
+ | EfficientNet-B0 | float | SA8295P ADP | Qualcomm® SA8295P | QNN_DLC | 3.517 ms | 1 - 35 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.dlc) |
55
+ | EfficientNet-B0 | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | TFLITE | 1.484 ms | 0 - 117 MB | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
56
+ | EfficientNet-B0 | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN_DLC | 1.465 ms | 1 - 8 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.dlc) |
57
+ | EfficientNet-B0 | float | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 1.967 ms | 0 - 32 MB | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
58
+ | EfficientNet-B0 | float | SA8775P ADP | Qualcomm® SA8775P | QNN_DLC | 1.965 ms | 1 - 25 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.dlc) |
59
+ | EfficientNet-B0 | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 1.028 ms | 0 - 46 MB | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
60
+ | EfficientNet-B0 | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 1.02 ms | 1 - 43 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.dlc) |
61
+ | EfficientNet-B0 | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 1.003 ms | 0 - 39 MB | NPU | [EfficientNet-B0.onnx.zip](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.onnx.zip) |
62
+ | EfficientNet-B0 | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | TFLITE | 0.824 ms | 0 - 37 MB | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
63
+ | EfficientNet-B0 | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | QNN_DLC | 0.791 ms | 1 - 34 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.dlc) |
64
+ | EfficientNet-B0 | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | ONNX | 0.81 ms | 0 - 29 MB | NPU | [EfficientNet-B0.onnx.zip](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.onnx.zip) |
65
+ | EfficientNet-B0 | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | TFLITE | 0.636 ms | 1 - 32 MB | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
66
+ | EfficientNet-B0 | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | QNN_DLC | 0.614 ms | 1 - 33 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.dlc) |
67
+ | EfficientNet-B0 | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | ONNX | 0.69 ms | 1 - 31 MB | NPU | [EfficientNet-B0.onnx.zip](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.onnx.zip) |
68
+ | EfficientNet-B0 | float | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_DLC | 1.669 ms | 78 - 78 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.dlc) |
69
+ | EfficientNet-B0 | float | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 1.385 ms | 14 - 14 MB | NPU | [EfficientNet-B0.onnx.zip](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.onnx.zip) |
70
+ | EfficientNet-B0 | w8a16 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC | 3.285 ms | 0 - 28 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.dlc) |
71
+ | EfficientNet-B0 | w8a16 | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 1.99 ms | 0 - 37 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.dlc) |
72
+ | EfficientNet-B0 | w8a16 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 1.675 ms | 0 - 41 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.dlc) |
73
+ | EfficientNet-B0 | w8a16 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | ONNX | 1.708 ms | 0 - 53 MB | NPU | [EfficientNet-B0.onnx.zip](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.onnx.zip) |
74
+ | EfficientNet-B0 | w8a16 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC | 1.974 ms | 0 - 29 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.dlc) |
75
+ | EfficientNet-B0 | w8a16 | RB3 Gen 2 (Proxy) | Qualcomm® QCS6490 (Proxy) | QNN_DLC | 5.237 ms | 0 - 37 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.dlc) |
76
+ | EfficientNet-B0 | w8a16 | RB3 Gen 2 (Proxy) | Qualcomm® QCS6490 (Proxy) | ONNX | 61.07 ms | 39 - 54 MB | CPU | [EfficientNet-B0.onnx.zip](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.onnx.zip) |
77
+ | EfficientNet-B0 | w8a16 | RB5 (Proxy) | Qualcomm® QCS8250 (Proxy) | ONNX | 45.977 ms | 35 - 45 MB | CPU | [EfficientNet-B0.onnx.zip](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.onnx.zip) |
78
+ | EfficientNet-B0 | w8a16 | SA7255P ADP | Qualcomm® SA7255P | QNN_DLC | 3.285 ms | 0 - 28 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.dlc) |
79
+ | EfficientNet-B0 | w8a16 | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN_DLC | 1.671 ms | 0 - 42 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.dlc) |
80
+ | EfficientNet-B0 | w8a16 | SA8295P ADP | Qualcomm® SA8295P | QNN_DLC | 2.378 ms | 0 - 34 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.dlc) |
81
+ | EfficientNet-B0 | w8a16 | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN_DLC | 1.673 ms | 0 - 41 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.dlc) |
82
+ | EfficientNet-B0 | w8a16 | SA8775P ADP | Qualcomm® SA8775P | QNN_DLC | 1.974 ms | 0 - 29 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.dlc) |
83
+ | EfficientNet-B0 | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 1.135 ms | 0 - 43 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.dlc) |
84
+ | EfficientNet-B0 | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 1.132 ms | 0 - 53 MB | NPU | [EfficientNet-B0.onnx.zip](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.onnx.zip) |
85
+ | EfficientNet-B0 | w8a16 | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | QNN_DLC | 0.775 ms | 0 - 34 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.dlc) |
86
+ | EfficientNet-B0 | w8a16 | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | ONNX | 0.848 ms | 0 - 38 MB | NPU | [EfficientNet-B0.onnx.zip](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.onnx.zip) |
87
+ | EfficientNet-B0 | w8a16 | Snapdragon 7 Gen 4 QRD | Snapdragon® 7 Gen 4 Mobile | QNN_DLC | 1.691 ms | 0 - 34 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.dlc) |
88
+ | EfficientNet-B0 | w8a16 | Snapdragon 7 Gen 4 QRD | Snapdragon® 7 Gen 4 Mobile | ONNX | 58.667 ms | 42 - 59 MB | CPU | [EfficientNet-B0.onnx.zip](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.onnx.zip) |
89
+ | EfficientNet-B0 | w8a16 | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | QNN_DLC | 0.644 ms | 0 - 33 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.dlc) |
90
+ | EfficientNet-B0 | w8a16 | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | ONNX | 0.777 ms | 0 - 42 MB | NPU | [EfficientNet-B0.onnx.zip](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.onnx.zip) |
91
+ | EfficientNet-B0 | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_DLC | 1.898 ms | 30 - 30 MB | NPU | [EfficientNet-B0.dlc](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.dlc) |
92
+ | EfficientNet-B0 | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 1.672 ms | 8 - 8 MB | NPU | [EfficientNet-B0.onnx.zip](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0_w8a16.onnx.zip) |
93
 
94
 
95
 
 
103
  ```
104
 
105
 
106
+ ## Configure Qualcomm® AI Hub Workbench to run this model on a cloud-hosted device
107
 
108
+ Sign-in to [Qualcomm® AI Hub Workbench](https://workbench.aihub.qualcomm.com/) with your
109
  Qualcomm® ID. Once signed in navigate to `Account -> Settings -> API Token`.
110
 
111
  With this API token, you can configure your client to run models on the cloud
 
113
  ```bash
114
  qai-hub configure --api_token API_TOKEN
115
  ```
116
+ Navigate to [docs](https://workbench.aihub.qualcomm.com/docs/) for more information.
117
 
118
 
119
 
 
224
  spot check the output with expected output.
225
 
226
  **Note**: This on-device profiling and inference requires access to Qualcomm®
227
+ AI Hub Workbench. [Sign up for access](https://myaccount.qualcomm.com/signup).
228
 
229
 
230