Hugging Face's logo

qualcomm
/
ConvNext-Base

Image Classification
PyTorch
bu_auto
android
Model card Files
xet
Community
qaihm-bot commited on
Commit
7f2d655
·
verified ·
1 Parent(s): 1165831

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +20 -19
README.md CHANGED
@@ -34,29 +34,29 @@ More details on model performance across various devices, can be found
34
 
35
  | Model | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
36
  |---|---|---|---|---|---|---|---|---|
37
- | ConvNext-Base | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 7.748 ms | 0 - 22 MB | FP16 | NPU | [ConvNext-Base.tflite](https://huggingface.co/qualcomm/ConvNext-Base/blob/main/ConvNext-Base.tflite) |
38
- | ConvNext-Base | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 8.472 ms | 0 - 25 MB | FP16 | NPU | [ConvNext-Base.so](https://huggingface.co/qualcomm/ConvNext-Base/blob/main/ConvNext-Base.so) |
39
- | ConvNext-Base | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 37.333 ms | 0 - 295 MB | FP16 | NPU | [ConvNext-Base.onnx](https://huggingface.co/qualcomm/ConvNext-Base/blob/main/ConvNext-Base.onnx) |
40
- | ConvNext-Base | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 5.802 ms | 0 - 60 MB | FP16 | NPU | [ConvNext-Base.tflite](https://huggingface.co/qualcomm/ConvNext-Base/blob/main/ConvNext-Base.tflite) |
41
- | ConvNext-Base | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 6.165 ms | 1 - 60 MB | FP16 | NPU | [ConvNext-Base.so](https://huggingface.co/qualcomm/ConvNext-Base/blob/main/ConvNext-Base.so) |
42
- | ConvNext-Base | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 26.475 ms | 1 - 141 MB | FP16 | NPU | [ConvNext-Base.onnx](https://huggingface.co/qualcomm/ConvNext-Base/blob/main/ConvNext-Base.onnx) |
43
- | ConvNext-Base | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 5.023 ms | 0 - 73 MB | FP16 | NPU | [ConvNext-Base.tflite](https://huggingface.co/qualcomm/ConvNext-Base/blob/main/ConvNext-Base.tflite) |
44
- | ConvNext-Base | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 5.511 ms | 0 - 74 MB | FP16 | NPU | Use Export Script |
45
- | ConvNext-Base | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 23.22 ms | 1 - 154 MB | FP16 | NPU | [ConvNext-Base.onnx](https://huggingface.co/qualcomm/ConvNext-Base/blob/main/ConvNext-Base.onnx) |
46
- | ConvNext-Base | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 7.761 ms | 0 - 25 MB | FP16 | NPU | [ConvNext-Base.tflite](https://huggingface.co/qualcomm/ConvNext-Base/blob/main/ConvNext-Base.tflite) |
47
- | ConvNext-Base | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 8.189 ms | 1 - 2 MB | FP16 | NPU | Use Export Script |
48
- | ConvNext-Base | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 20.119 ms | 0 - 53 MB | FP16 | NPU | [ConvNext-Base.tflite](https://huggingface.co/qualcomm/ConvNext-Base/blob/main/ConvNext-Base.tflite) |
49
- | ConvNext-Base | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 21.257 ms | 0 - 53 MB | FP16 | NPU | Use Export Script |
50
- | ConvNext-Base | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 8.576 ms | 1 - 1 MB | FP16 | NPU | Use Export Script |
51
- | ConvNext-Base | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 41.04 ms | 176 - 176 MB | FP16 | NPU | [ConvNext-Base.onnx](https://huggingface.co/qualcomm/ConvNext-Base/blob/main/ConvNext-Base.onnx) |
52
 
53
 
54
 
55
 
56
  ## Installation
57
 
58
- This model can be installed as a Python package via pip.
59
 
 
60
  ```bash
61
  pip install qai-hub-models
62
  ```
@@ -113,7 +113,7 @@ ConvNext-Base
113
  Device : Samsung Galaxy S23 (13)
114
  Runtime : TFLITE
115
  Estimated inference time (ms) : 7.7
116
- Estimated peak memory usage (MB): [0, 22]
117
  Total # Ops : 598
118
  Compute Unit(s) : NPU (598 ops)
119
  ```
@@ -140,7 +140,7 @@ from qai_hub_models.models.convnext_base import Model
140
  torch_model = Model.from_pretrained()
141
 
142
  # Device
143
- device = hub.Device("Samsung Galaxy S23")
144
 
145
  # Trace model
146
  input_shape = torch_model.get_input_spec()
@@ -232,7 +232,8 @@ Explore all available models on [Qualcomm® AI Hub](https://aihub.qualcomm.com/)
232
 
233
 
234
  ## License
235
- * The license for the original implementation of ConvNext-Base can be found [here](https://github.com/pytorch/vision/blob/main/LICENSE).
 
236
  * 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)
237
 
238
 
 
34
 
35
  | Model | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
36
  |---|---|---|---|---|---|---|---|---|
37
+ | ConvNext-Base | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 7.743 ms | 0 - 27 MB | FP16 | NPU | [ConvNext-Base.tflite](https://huggingface.co/qualcomm/ConvNext-Base/blob/main/ConvNext-Base.tflite) |
38
+ | ConvNext-Base | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 8.489 ms | 1 - 22 MB | FP16 | NPU | [ConvNext-Base.so](https://huggingface.co/qualcomm/ConvNext-Base/blob/main/ConvNext-Base.so) |
39
+ | ConvNext-Base | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 37.604 ms | 0 - 413 MB | FP16 | NPU | [ConvNext-Base.onnx](https://huggingface.co/qualcomm/ConvNext-Base/blob/main/ConvNext-Base.onnx) |
40
+ | ConvNext-Base | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 5.804 ms | 0 - 61 MB | FP16 | NPU | [ConvNext-Base.tflite](https://huggingface.co/qualcomm/ConvNext-Base/blob/main/ConvNext-Base.tflite) |
41
+ | ConvNext-Base | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 6.163 ms | 1 - 61 MB | FP16 | NPU | [ConvNext-Base.so](https://huggingface.co/qualcomm/ConvNext-Base/blob/main/ConvNext-Base.so) |
42
+ | ConvNext-Base | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 26.61 ms | 1 - 142 MB | FP16 | NPU | [ConvNext-Base.onnx](https://huggingface.co/qualcomm/ConvNext-Base/blob/main/ConvNext-Base.onnx) |
43
+ | ConvNext-Base | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 4.334 ms | 0 - 73 MB | FP16 | NPU | [ConvNext-Base.tflite](https://huggingface.co/qualcomm/ConvNext-Base/blob/main/ConvNext-Base.tflite) |
44
+ | ConvNext-Base | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 4.688 ms | 1 - 74 MB | FP16 | NPU | Use Export Script |
45
+ | ConvNext-Base | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 23.177 ms | 1 - 153 MB | FP16 | NPU | [ConvNext-Base.onnx](https://huggingface.co/qualcomm/ConvNext-Base/blob/main/ConvNext-Base.onnx) |
46
+ | ConvNext-Base | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 7.77 ms | 0 - 22 MB | FP16 | NPU | [ConvNext-Base.tflite](https://huggingface.co/qualcomm/ConvNext-Base/blob/main/ConvNext-Base.tflite) |
47
+ | ConvNext-Base | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 8.227 ms | 1 - 3 MB | FP16 | NPU | Use Export Script |
48
+ | ConvNext-Base | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 20.143 ms | 0 - 54 MB | FP16 | NPU | [ConvNext-Base.tflite](https://huggingface.co/qualcomm/ConvNext-Base/blob/main/ConvNext-Base.tflite) |
49
+ | ConvNext-Base | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 21.227 ms | 0 - 52 MB | FP16 | NPU | Use Export Script |
50
+ | ConvNext-Base | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 8.566 ms | 1 - 1 MB | FP16 | NPU | Use Export Script |
51
+ | ConvNext-Base | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 40.983 ms | 175 - 175 MB | FP16 | NPU | [ConvNext-Base.onnx](https://huggingface.co/qualcomm/ConvNext-Base/blob/main/ConvNext-Base.onnx) |
52
 
53
 
54
 
55
 
56
  ## Installation
57
 
 
58
 
59
+ Install the package via pip:
60
  ```bash
61
  pip install qai-hub-models
62
  ```
 
113
  Device : Samsung Galaxy S23 (13)
114
  Runtime : TFLITE
115
  Estimated inference time (ms) : 7.7
116
+ Estimated peak memory usage (MB): [0, 27]
117
  Total # Ops : 598
118
  Compute Unit(s) : NPU (598 ops)
119
  ```
 
140
  torch_model = Model.from_pretrained()
141
 
142
  # Device
143
+ device = hub.Device("Samsung Galaxy S24")
144
 
145
  # Trace model
146
  input_shape = torch_model.get_input_spec()
 
232
 
233
 
234
  ## License
235
+ * The license for the original implementation of ConvNext-Base can be found
236
+ [here](https://github.com/pytorch/vision/blob/main/LICENSE).
237
  * 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)
238
 
239