v0.42.0
Browse filesSee https://github.com/quic/ai-hub-models/releases/v0.42.0 for changelog.
- GPUNet_float.dlc +2 -2
- GPUNet_float.onnx.zip +1 -1
- GPUNet_w8a16.dlc +2 -2
- GPUNet_w8a16.onnx.zip +2 -2
- GPUNet_w8a8.dlc +2 -2
- GPUNet_w8a8.onnx.zip +2 -2
- GPUNet_w8a8.tflite +1 -1
- README.md +94 -89
GPUNet_float.dlc
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:72454492ef7d2456a21a97449f797a3a7f289ed524224ca3e0acb7f8376a584a
|
| 3 |
+
size 47563724
|
GPUNet_float.onnx.zip
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 44302131
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:86465738c403fa7c621aac93fdc12ce4219f3ee5046ee9a2f546bee6c5e0d27b
|
| 3 |
size 44302131
|
GPUNet_w8a16.dlc
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14a3bc57c4a05e6b2b803393b7518821c06c946d2b499c927098d3f9d453124c
|
| 3 |
+
size 13470572
|
GPUNet_w8a16.onnx.zip
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:60d2cad665c72d37dec9c57b20102dd01b2df045002c1690dd285b9bb8e7fbea
|
| 3 |
+
size 22147598
|
GPUNet_w8a8.dlc
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0d41d64340fbf00f023a1dba8044e98bc3857127077e28a42c4ecd9b71422225
|
| 3 |
+
size 13470564
|
GPUNet_w8a8.onnx.zip
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3a30e826a4014bc84babc7c8d1d97d903b420cee7361a5874b3eb6055032c8d6
|
| 3 |
+
size 22332846
|
GPUNet_w8a8.tflite
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 12578264
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c0f151583accfe859ab05757cf9f9fa8430db42e8d63600b282bd03f8c60ca16
|
| 3 |
size 12578264
|
README.md
CHANGED
|
@@ -37,92 +37,97 @@ 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 |
-
| GPUNet | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 4.
|
| 41 |
-
| GPUNet | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC | 4.
|
| 42 |
-
| GPUNet | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 1.
|
| 43 |
-
| GPUNet | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 2.
|
| 44 |
-
| GPUNet | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 1.
|
| 45 |
-
| GPUNet | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 1.
|
| 46 |
-
| GPUNet | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | ONNX | 1.
|
| 47 |
-
| GPUNet | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 1.
|
| 48 |
-
| GPUNet | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC | 1.
|
| 49 |
-
| GPUNet | float | SA7255P ADP | Qualcomm® SA7255P | TFLITE | 4.
|
| 50 |
-
| GPUNet | float | SA7255P ADP | Qualcomm® SA7255P | QNN_DLC | 4.
|
| 51 |
-
| GPUNet | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | TFLITE | 1.
|
| 52 |
-
| GPUNet | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN_DLC | 1.
|
| 53 |
-
| GPUNet | float | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 2.
|
| 54 |
-
| GPUNet | float | SA8295P ADP | Qualcomm® SA8295P | QNN_DLC | 2.
|
| 55 |
-
| GPUNet | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | TFLITE | 1.
|
| 56 |
-
| GPUNet | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN_DLC | 1.
|
| 57 |
-
| GPUNet | float | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 1.
|
| 58 |
-
| GPUNet | float | SA8775P ADP | Qualcomm® SA8775P | QNN_DLC | 1.
|
| 59 |
-
| GPUNet | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 0.
|
| 60 |
-
| GPUNet | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 0.
|
| 61 |
-
| GPUNet | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 0.
|
| 62 |
-
| GPUNet | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | TFLITE | 0.
|
| 63 |
-
| GPUNet | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | QNN_DLC | 0.
|
| 64 |
| GPUNet | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | ONNX | 0.714 ms | 0 - 27 MB | NPU | [GPUNet.onnx.zip](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.onnx.zip) |
|
| 65 |
-
| GPUNet | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | TFLITE | 0.
|
| 66 |
-
| GPUNet | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | QNN_DLC | 0.
|
| 67 |
-
| GPUNet | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | ONNX | 0.
|
| 68 |
-
| GPUNet | float | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_DLC | 1.
|
| 69 |
-
| GPUNet | float | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 1.
|
| 70 |
-
| GPUNet | w8a16 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC | 2.
|
| 71 |
-
| GPUNet | w8a16 | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 1.
|
| 72 |
-
| GPUNet | w8a16 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 1.
|
| 73 |
-
| GPUNet | w8a16 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | ONNX | 1.
|
| 74 |
-
| GPUNet | w8a16 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC |
|
| 75 |
-
| GPUNet | w8a16 | RB3 Gen 2 (Proxy) | Qualcomm® QCS6490 (Proxy) | QNN_DLC | 4.
|
| 76 |
-
| GPUNet | w8a16 | RB3 Gen 2 (Proxy) | Qualcomm® QCS6490 (Proxy) | ONNX | 49.
|
| 77 |
-
| GPUNet | w8a16 | RB5 (Proxy) | Qualcomm® QCS8250 (Proxy) | ONNX |
|
| 78 |
-
| GPUNet | w8a16 | SA7255P ADP | Qualcomm® SA7255P | QNN_DLC | 2.
|
| 79 |
-
| GPUNet | w8a16 | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN_DLC | 1.
|
| 80 |
-
| GPUNet | w8a16 | SA8295P ADP | Qualcomm® SA8295P | QNN_DLC | 1.
|
| 81 |
-
| GPUNet | w8a16 | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN_DLC | 1.
|
| 82 |
-
| GPUNet | w8a16 | SA8775P ADP | Qualcomm® SA8775P | QNN_DLC |
|
| 83 |
-
| GPUNet | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 0.
|
| 84 |
-
| GPUNet | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 0.
|
| 85 |
-
| GPUNet | w8a16 | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | QNN_DLC | 0.
|
| 86 |
-
| GPUNet | w8a16 | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | ONNX | 0.
|
| 87 |
-
| GPUNet | w8a16 | Snapdragon
|
| 88 |
-
| GPUNet | w8a16 | Snapdragon
|
| 89 |
-
| GPUNet | w8a16 | Snapdragon
|
| 90 |
-
| GPUNet | w8a16 | Snapdragon
|
| 91 |
-
| GPUNet |
|
| 92 |
-
| GPUNet |
|
| 93 |
-
| GPUNet | w8a8 |
|
| 94 |
-
| GPUNet | w8a8 |
|
| 95 |
-
| GPUNet | w8a8 |
|
| 96 |
-
| GPUNet | w8a8 |
|
| 97 |
-
| GPUNet | w8a8 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) |
|
| 98 |
-
| GPUNet | w8a8 |
|
| 99 |
-
| GPUNet | w8a8 |
|
| 100 |
-
| GPUNet | w8a8 |
|
| 101 |
-
| GPUNet | w8a8 |
|
| 102 |
-
| GPUNet | w8a8 | RB3 Gen 2 (Proxy) | Qualcomm® QCS6490 (Proxy) |
|
| 103 |
-
| GPUNet | w8a8 |
|
| 104 |
-
| GPUNet | w8a8 |
|
| 105 |
-
| GPUNet | w8a8 |
|
| 106 |
-
| GPUNet | w8a8 |
|
| 107 |
-
| GPUNet | w8a8 |
|
| 108 |
-
| GPUNet | w8a8 |
|
| 109 |
-
| GPUNet | w8a8 |
|
| 110 |
-
| GPUNet | w8a8 |
|
| 111 |
-
| GPUNet | w8a8 |
|
| 112 |
-
| GPUNet | w8a8 |
|
| 113 |
-
| GPUNet | w8a8 |
|
| 114 |
-
| GPUNet | w8a8 |
|
| 115 |
-
| GPUNet | w8a8 |
|
| 116 |
-
| GPUNet | w8a8 |
|
| 117 |
-
| GPUNet | w8a8 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile |
|
| 118 |
-
| GPUNet | w8a8 | Samsung Galaxy
|
| 119 |
-
| GPUNet | w8a8 | Samsung Galaxy
|
| 120 |
-
| GPUNet | w8a8 | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile |
|
| 121 |
-
| GPUNet | w8a8 |
|
| 122 |
-
| GPUNet | w8a8 |
|
| 123 |
-
| GPUNet | w8a8 | Snapdragon
|
| 124 |
-
| GPUNet | w8a8 | Snapdragon
|
| 125 |
-
| GPUNet | w8a8 | Snapdragon
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 126 |
|
| 127 |
|
| 128 |
|
|
@@ -136,9 +141,9 @@ pip install qai-hub-models
|
|
| 136 |
```
|
| 137 |
|
| 138 |
|
| 139 |
-
## Configure Qualcomm® AI Hub to run this model on a cloud-hosted device
|
| 140 |
|
| 141 |
-
Sign-in to [Qualcomm® AI Hub](https://
|
| 142 |
Qualcomm® ID. Once signed in navigate to `Account -> Settings -> API Token`.
|
| 143 |
|
| 144 |
With this API token, you can configure your client to run models on the cloud
|
|
@@ -146,7 +151,7 @@ hosted devices.
|
|
| 146 |
```bash
|
| 147 |
qai-hub configure --api_token API_TOKEN
|
| 148 |
```
|
| 149 |
-
Navigate to [docs](https://
|
| 150 |
|
| 151 |
|
| 152 |
|
|
@@ -257,7 +262,7 @@ With the output of the model, you can compute like PSNR, relative errors or
|
|
| 257 |
spot check the output with expected output.
|
| 258 |
|
| 259 |
**Note**: This on-device profiling and inference requires access to Qualcomm®
|
| 260 |
-
AI Hub. [Sign up for access](https://myaccount.qualcomm.com/signup).
|
| 261 |
|
| 262 |
|
| 263 |
|
|
|
|
| 37 |
|
| 38 |
| Model | Precision | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit | Target Model
|
| 39 |
|---|---|---|---|---|---|---|---|---|
|
| 40 |
+
| GPUNet | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 4.713 ms | 0 - 51 MB | NPU | [GPUNet.tflite](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.tflite) |
|
| 41 |
+
| GPUNet | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC | 4.614 ms | 1 - 24 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.dlc) |
|
| 42 |
+
| GPUNet | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 1.799 ms | 0 - 63 MB | NPU | [GPUNet.tflite](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.tflite) |
|
| 43 |
+
| GPUNet | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 2.256 ms | 1 - 36 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.dlc) |
|
| 44 |
+
| GPUNet | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 1.226 ms | 0 - 181 MB | NPU | [GPUNet.tflite](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.tflite) |
|
| 45 |
+
| GPUNet | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 1.246 ms | 0 - 78 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.dlc) |
|
| 46 |
+
| GPUNet | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | ONNX | 1.242 ms | 0 - 118 MB | NPU | [GPUNet.onnx.zip](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.onnx.zip) |
|
| 47 |
+
| GPUNet | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 1.69 ms | 0 - 52 MB | NPU | [GPUNet.tflite](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.tflite) |
|
| 48 |
+
| GPUNet | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC | 1.705 ms | 1 - 24 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.dlc) |
|
| 49 |
+
| GPUNet | float | SA7255P ADP | Qualcomm® SA7255P | TFLITE | 4.713 ms | 0 - 51 MB | NPU | [GPUNet.tflite](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.tflite) |
|
| 50 |
+
| GPUNet | float | SA7255P ADP | Qualcomm® SA7255P | QNN_DLC | 4.614 ms | 1 - 24 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.dlc) |
|
| 51 |
+
| GPUNet | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | TFLITE | 1.233 ms | 0 - 180 MB | NPU | [GPUNet.tflite](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.tflite) |
|
| 52 |
+
| GPUNet | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN_DLC | 1.244 ms | 0 - 86 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.dlc) |
|
| 53 |
+
| GPUNet | float | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 2.211 ms | 0 - 59 MB | NPU | [GPUNet.tflite](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.tflite) |
|
| 54 |
+
| GPUNet | float | SA8295P ADP | Qualcomm® SA8295P | QNN_DLC | 2.203 ms | 1 - 31 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.dlc) |
|
| 55 |
+
| GPUNet | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | TFLITE | 1.233 ms | 0 - 180 MB | NPU | [GPUNet.tflite](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.tflite) |
|
| 56 |
+
| GPUNet | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN_DLC | 1.245 ms | 0 - 78 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.dlc) |
|
| 57 |
+
| GPUNet | float | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 1.69 ms | 0 - 52 MB | NPU | [GPUNet.tflite](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.tflite) |
|
| 58 |
+
| GPUNet | float | SA8775P ADP | Qualcomm® SA8775P | QNN_DLC | 1.705 ms | 1 - 24 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.dlc) |
|
| 59 |
+
| GPUNet | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 0.883 ms | 0 - 67 MB | NPU | [GPUNet.tflite](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.tflite) |
|
| 60 |
+
| GPUNet | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 0.894 ms | 1 - 36 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.dlc) |
|
| 61 |
+
| GPUNet | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 0.875 ms | 0 - 33 MB | NPU | [GPUNet.onnx.zip](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.onnx.zip) |
|
| 62 |
+
| GPUNet | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | TFLITE | 0.696 ms | 0 - 58 MB | NPU | [GPUNet.tflite](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.tflite) |
|
| 63 |
+
| GPUNet | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | QNN_DLC | 0.697 ms | 1 - 31 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.dlc) |
|
| 64 |
| GPUNet | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | ONNX | 0.714 ms | 0 - 27 MB | NPU | [GPUNet.onnx.zip](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.onnx.zip) |
|
| 65 |
+
| GPUNet | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | TFLITE | 0.581 ms | 0 - 58 MB | NPU | [GPUNet.tflite](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.tflite) |
|
| 66 |
+
| GPUNet | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | QNN_DLC | 0.568 ms | 1 - 33 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.dlc) |
|
| 67 |
+
| GPUNet | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | ONNX | 0.633 ms | 1 - 27 MB | NPU | [GPUNet.onnx.zip](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.onnx.zip) |
|
| 68 |
+
| GPUNet | float | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_DLC | 1.345 ms | 106 - 106 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.dlc) |
|
| 69 |
+
| GPUNet | float | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 1.105 ms | 25 - 25 MB | NPU | [GPUNet.onnx.zip](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet.onnx.zip) |
|
| 70 |
+
| GPUNet | w8a16 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC | 2.489 ms | 0 - 31 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a16.dlc) |
|
| 71 |
+
| GPUNet | w8a16 | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 1.456 ms | 0 - 38 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a16.dlc) |
|
| 72 |
+
| GPUNet | w8a16 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 1.068 ms | 0 - 58 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a16.dlc) |
|
| 73 |
+
| GPUNet | w8a16 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | ONNX | 1.084 ms | 0 - 56 MB | NPU | [GPUNet.onnx.zip](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a16.onnx.zip) |
|
| 74 |
+
| GPUNet | w8a16 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC | 5.52 ms | 0 - 31 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a16.dlc) |
|
| 75 |
+
| GPUNet | w8a16 | RB3 Gen 2 (Proxy) | Qualcomm® QCS6490 (Proxy) | QNN_DLC | 4.175 ms | 0 - 45 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a16.dlc) |
|
| 76 |
+
| GPUNet | w8a16 | RB3 Gen 2 (Proxy) | Qualcomm® QCS6490 (Proxy) | ONNX | 49.582 ms | 29 - 41 MB | CPU | [GPUNet.onnx.zip](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a16.onnx.zip) |
|
| 77 |
+
| GPUNet | w8a16 | RB5 (Proxy) | Qualcomm® QCS8250 (Proxy) | ONNX | 44.813 ms | 21 - 38 MB | CPU | [GPUNet.onnx.zip](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a16.onnx.zip) |
|
| 78 |
+
| GPUNet | w8a16 | SA7255P ADP | Qualcomm® SA7255P | QNN_DLC | 2.489 ms | 0 - 31 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a16.dlc) |
|
| 79 |
+
| GPUNet | w8a16 | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN_DLC | 1.07 ms | 0 - 57 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a16.dlc) |
|
| 80 |
+
| GPUNet | w8a16 | SA8295P ADP | Qualcomm® SA8295P | QNN_DLC | 1.661 ms | 0 - 37 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a16.dlc) |
|
| 81 |
+
| GPUNet | w8a16 | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN_DLC | 1.073 ms | 0 - 57 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a16.dlc) |
|
| 82 |
+
| GPUNet | w8a16 | SA8775P ADP | Qualcomm® SA8775P | QNN_DLC | 5.52 ms | 0 - 31 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a16.dlc) |
|
| 83 |
+
| GPUNet | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 0.768 ms | 0 - 40 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a16.dlc) |
|
| 84 |
+
| GPUNet | w8a16 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 0.738 ms | 0 - 46 MB | NPU | [GPUNet.onnx.zip](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a16.onnx.zip) |
|
| 85 |
+
| GPUNet | w8a16 | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | QNN_DLC | 0.526 ms | 0 - 39 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a16.dlc) |
|
| 86 |
+
| GPUNet | w8a16 | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | ONNX | 0.572 ms | 0 - 41 MB | NPU | [GPUNet.onnx.zip](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a16.onnx.zip) |
|
| 87 |
+
| GPUNet | w8a16 | Snapdragon 7 Gen 4 QRD | Snapdragon® 7 Gen 4 Mobile | QNN_DLC | 1.307 ms | 0 - 40 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a16.dlc) |
|
| 88 |
+
| GPUNet | w8a16 | Snapdragon 7 Gen 4 QRD | Snapdragon® 7 Gen 4 Mobile | ONNX | 51.509 ms | 29 - 45 MB | CPU | [GPUNet.onnx.zip](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a16.onnx.zip) |
|
| 89 |
+
| GPUNet | w8a16 | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | QNN_DLC | 0.438 ms | 0 - 39 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a16.dlc) |
|
| 90 |
+
| GPUNet | w8a16 | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | ONNX | 0.507 ms | 0 - 44 MB | NPU | [GPUNet.onnx.zip](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a16.onnx.zip) |
|
| 91 |
+
| GPUNet | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_DLC | 1.226 ms | 55 - 55 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a16.dlc) |
|
| 92 |
+
| GPUNet | w8a16 | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 1.033 ms | 14 - 14 MB | NPU | [GPUNet.onnx.zip](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a16.onnx.zip) |
|
| 93 |
+
| GPUNet | w8a8 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 1.118 ms | 0 - 31 MB | NPU | [GPUNet.tflite](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.tflite) |
|
| 94 |
+
| GPUNet | w8a8 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC | 1.397 ms | 0 - 30 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.dlc) |
|
| 95 |
+
| GPUNet | w8a8 | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 0.578 ms | 0 - 39 MB | NPU | [GPUNet.tflite](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.tflite) |
|
| 96 |
+
| GPUNet | w8a8 | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 0.844 ms | 0 - 41 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.dlc) |
|
| 97 |
+
| GPUNet | w8a8 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 0.449 ms | 0 - 62 MB | NPU | [GPUNet.tflite](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.tflite) |
|
| 98 |
+
| GPUNet | w8a8 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 0.619 ms | 0 - 62 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.dlc) |
|
| 99 |
+
| GPUNet | w8a8 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | ONNX | 0.856 ms | 0 - 48 MB | NPU | [GPUNet.onnx.zip](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.onnx.zip) |
|
| 100 |
+
| GPUNet | w8a8 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 2.286 ms | 0 - 31 MB | NPU | [GPUNet.tflite](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.tflite) |
|
| 101 |
+
| GPUNet | w8a8 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC | 0.802 ms | 0 - 30 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.dlc) |
|
| 102 |
+
| GPUNet | w8a8 | RB3 Gen 2 (Proxy) | Qualcomm® QCS6490 (Proxy) | TFLITE | 1.615 ms | 0 - 38 MB | NPU | [GPUNet.tflite](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.tflite) |
|
| 103 |
+
| GPUNet | w8a8 | RB3 Gen 2 (Proxy) | Qualcomm® QCS6490 (Proxy) | QNN_DLC | 2.257 ms | 0 - 38 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.dlc) |
|
| 104 |
+
| GPUNet | w8a8 | RB3 Gen 2 (Proxy) | Qualcomm® QCS6490 (Proxy) | ONNX | 12.323 ms | 12 - 24 MB | CPU | [GPUNet.onnx.zip](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.onnx.zip) |
|
| 105 |
+
| GPUNet | w8a8 | RB5 (Proxy) | Qualcomm® QCS8250 (Proxy) | TFLITE | 8.075 ms | 0 - 2 MB | NPU | [GPUNet.tflite](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.tflite) |
|
| 106 |
+
| GPUNet | w8a8 | RB5 (Proxy) | Qualcomm® QCS8250 (Proxy) | ONNX | 10.54 ms | 5 - 22 MB | CPU | [GPUNet.onnx.zip](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.onnx.zip) |
|
| 107 |
+
| GPUNet | w8a8 | SA7255P ADP | Qualcomm® SA7255P | TFLITE | 1.118 ms | 0 - 31 MB | NPU | [GPUNet.tflite](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.tflite) |
|
| 108 |
+
| GPUNet | w8a8 | SA7255P ADP | Qualcomm® SA7255P | QNN_DLC | 1.397 ms | 0 - 30 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.dlc) |
|
| 109 |
+
| GPUNet | w8a8 | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | TFLITE | 0.45 ms | 0 - 61 MB | NPU | [GPUNet.tflite](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.tflite) |
|
| 110 |
+
| GPUNet | w8a8 | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN_DLC | 0.612 ms | 0 - 61 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.dlc) |
|
| 111 |
+
| GPUNet | w8a8 | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 0.864 ms | 0 - 37 MB | NPU | [GPUNet.tflite](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.tflite) |
|
| 112 |
+
| GPUNet | w8a8 | SA8295P ADP | Qualcomm® SA8295P | QNN_DLC | 1.067 ms | 0 - 36 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.dlc) |
|
| 113 |
+
| GPUNet | w8a8 | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | TFLITE | 0.444 ms | 0 - 62 MB | NPU | [GPUNet.tflite](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.tflite) |
|
| 114 |
+
| GPUNet | w8a8 | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN_DLC | 0.619 ms | 0 - 62 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.dlc) |
|
| 115 |
+
| GPUNet | w8a8 | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 2.286 ms | 0 - 31 MB | NPU | [GPUNet.tflite](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.tflite) |
|
| 116 |
+
| GPUNet | w8a8 | SA8775P ADP | Qualcomm® SA8775P | QNN_DLC | 0.802 ms | 0 - 30 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.dlc) |
|
| 117 |
+
| GPUNet | w8a8 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 0.34 ms | 0 - 42 MB | NPU | [GPUNet.tflite](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.tflite) |
|
| 118 |
+
| GPUNet | w8a8 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 0.467 ms | 0 - 43 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.dlc) |
|
| 119 |
+
| GPUNet | w8a8 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 0.61 ms | 0 - 44 MB | NPU | [GPUNet.onnx.zip](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.onnx.zip) |
|
| 120 |
+
| GPUNet | w8a8 | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | TFLITE | 0.27 ms | 0 - 34 MB | NPU | [GPUNet.tflite](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.tflite) |
|
| 121 |
+
| GPUNet | w8a8 | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | QNN_DLC | 0.34 ms | 0 - 34 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.dlc) |
|
| 122 |
+
| GPUNet | w8a8 | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | ONNX | 0.525 ms | 0 - 42 MB | NPU | [GPUNet.onnx.zip](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.onnx.zip) |
|
| 123 |
+
| GPUNet | w8a8 | Snapdragon 7 Gen 4 QRD | Snapdragon® 7 Gen 4 Mobile | TFLITE | 0.635 ms | 0 - 38 MB | NPU | [GPUNet.tflite](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.tflite) |
|
| 124 |
+
| GPUNet | w8a8 | Snapdragon 7 Gen 4 QRD | Snapdragon® 7 Gen 4 Mobile | QNN_DLC | 0.8 ms | 0 - 42 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.dlc) |
|
| 125 |
+
| GPUNet | w8a8 | Snapdragon 7 Gen 4 QRD | Snapdragon® 7 Gen 4 Mobile | ONNX | 9.943 ms | 13 - 30 MB | CPU | [GPUNet.onnx.zip](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.onnx.zip) |
|
| 126 |
+
| GPUNet | w8a8 | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | TFLITE | 0.249 ms | 0 - 34 MB | NPU | [GPUNet.tflite](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.tflite) |
|
| 127 |
+
| GPUNet | w8a8 | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | QNN_DLC | 0.296 ms | 0 - 39 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.dlc) |
|
| 128 |
+
| GPUNet | w8a8 | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | ONNX | 0.499 ms | 0 - 43 MB | NPU | [GPUNet.onnx.zip](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.onnx.zip) |
|
| 129 |
+
| GPUNet | w8a8 | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_DLC | 0.731 ms | 63 - 63 MB | NPU | [GPUNet.dlc](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.dlc) |
|
| 130 |
+
| GPUNet | w8a8 | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 0.757 ms | 13 - 13 MB | NPU | [GPUNet.onnx.zip](https://huggingface.co/qualcomm/GPUNet/blob/main/GPUNet_w8a8.onnx.zip) |
|
| 131 |
|
| 132 |
|
| 133 |
|
|
|
|
| 141 |
```
|
| 142 |
|
| 143 |
|
| 144 |
+
## Configure Qualcomm® AI Hub Workbench to run this model on a cloud-hosted device
|
| 145 |
|
| 146 |
+
Sign-in to [Qualcomm® AI Hub Workbench](https://workbench.aihub.qualcomm.com/) with your
|
| 147 |
Qualcomm® ID. Once signed in navigate to `Account -> Settings -> API Token`.
|
| 148 |
|
| 149 |
With this API token, you can configure your client to run models on the cloud
|
|
|
|
| 151 |
```bash
|
| 152 |
qai-hub configure --api_token API_TOKEN
|
| 153 |
```
|
| 154 |
+
Navigate to [docs](https://workbench.aihub.qualcomm.com/docs/) for more information.
|
| 155 |
|
| 156 |
|
| 157 |
|
|
|
|
| 262 |
spot check the output with expected output.
|
| 263 |
|
| 264 |
**Note**: This on-device profiling and inference requires access to Qualcomm®
|
| 265 |
+
AI Hub Workbench. [Sign up for access](https://myaccount.qualcomm.com/signup).
|
| 266 |
|
| 267 |
|
| 268 |
|