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
b155e1a
·
verified ·
1 Parent(s): adce755

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +43 -25
README.md CHANGED
@@ -35,29 +35,31 @@ More details on model performance across various devices, can be found
35
 
36
  | Model | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
37
  |---|---|---|---|---|---|---|---|---|
38
- | EfficientNet-B0 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 1.607 ms | 0 - 282 MB | FP16 | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
39
- | EfficientNet-B0 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 1.666 ms | 0 - 215 MB | FP16 | NPU | [EfficientNet-B0.so](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.so) |
40
- | EfficientNet-B0 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 1.594 ms | 0 - 17 MB | FP16 | NPU | [EfficientNet-B0.onnx](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.onnx) |
41
- | EfficientNet-B0 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 1.124 ms | 0 - 76 MB | FP16 | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
42
- | EfficientNet-B0 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 1.172 ms | 0 - 20 MB | FP16 | NPU | [EfficientNet-B0.so](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.so) |
43
- | EfficientNet-B0 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 1.141 ms | 0 - 82 MB | FP16 | NPU | [EfficientNet-B0.onnx](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.onnx) |
44
- | EfficientNet-B0 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 1.124 ms | 0 - 32 MB | FP16 | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
45
- | EfficientNet-B0 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 0.953 ms | 0 - 15 MB | FP16 | NPU | Use Export Script |
46
- | EfficientNet-B0 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 1.178 ms | 0 - 34 MB | FP16 | NPU | [EfficientNet-B0.onnx](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.onnx) |
47
- | EfficientNet-B0 | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 1.599 ms | 0 - 246 MB | FP16 | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
48
- | EfficientNet-B0 | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 1.563 ms | 1 - 2 MB | FP16 | NPU | Use Export Script |
49
- | EfficientNet-B0 | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 1.604 ms | 0 - 1 MB | FP16 | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
50
- | EfficientNet-B0 | SA8255 (Proxy) | SA8255P Proxy | QNN | 1.579 ms | 1 - 2 MB | FP16 | NPU | Use Export Script |
51
- | EfficientNet-B0 | SA8775 (Proxy) | SA8775P Proxy | TFLITE | 1.605 ms | 0 - 3 MB | FP16 | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
52
- | EfficientNet-B0 | SA8775 (Proxy) | SA8775P Proxy | QNN | 1.575 ms | 1 - 2 MB | FP16 | NPU | Use Export Script |
53
- | EfficientNet-B0 | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 1.607 ms | 0 - 1 MB | FP16 | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
54
- | EfficientNet-B0 | SA8650 (Proxy) | SA8650P Proxy | QNN | 1.572 ms | 1 - 2 MB | FP16 | NPU | Use Export Script |
55
- | EfficientNet-B0 | SA8295P ADP | SA8295P | TFLITE | 3.77 ms | 0 - 38 MB | FP16 | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
56
- | EfficientNet-B0 | SA8295P ADP | SA8295P | QNN | 3.772 ms | 0 - 6 MB | FP16 | NPU | Use Export Script |
57
- | EfficientNet-B0 | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 3.089 ms | 0 - 84 MB | FP16 | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
58
- | EfficientNet-B0 | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 3.168 ms | 1 - 26 MB | FP16 | NPU | Use Export Script |
59
- | EfficientNet-B0 | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 1.754 ms | 1 - 1 MB | FP16 | NPU | Use Export Script |
60
- | EfficientNet-B0 | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 1.694 ms | 14 - 14 MB | FP16 | NPU | [EfficientNet-B0.onnx](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.onnx) |
 
 
61
 
62
 
63
 
@@ -122,7 +124,7 @@ EfficientNet-B0
122
  Device : Samsung Galaxy S23 (13)
123
  Runtime : TFLITE
124
  Estimated inference time (ms) : 1.6
125
- Estimated peak memory usage (MB): [0, 282]
126
  Total # Ops : 245
127
  Compute Unit(s) : NPU (245 ops)
128
  ```
@@ -143,13 +145,29 @@ in memory using the `jit.trace` and then call the `submit_compile_job` API.
143
  import torch
144
 
145
  import qai_hub as hub
146
- from qai_hub_models.models.efficientnet_b0 import
147
 
148
  # Load the model
 
149
 
150
  # Device
151
  device = hub.Device("Samsung Galaxy S23")
152
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
153
 
154
  ```
155
 
 
35
 
36
  | Model | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
37
  |---|---|---|---|---|---|---|---|---|
38
+ | EfficientNet-B0 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 1.613 ms | 0 - 81 MB | FP16 | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
39
+ | EfficientNet-B0 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 1.686 ms | 0 - 328 MB | FP16 | NPU | [EfficientNet-B0.so](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.so) |
40
+ | EfficientNet-B0 | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 1.58 ms | 0 - 237 MB | FP16 | NPU | [EfficientNet-B0.onnx](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.onnx) |
41
+ | EfficientNet-B0 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 1.129 ms | 0 - 18 MB | FP16 | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
42
+ | EfficientNet-B0 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 1.176 ms | 0 - 20 MB | FP16 | NPU | [EfficientNet-B0.so](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.so) |
43
+ | EfficientNet-B0 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 1.158 ms | 0 - 82 MB | FP16 | NPU | [EfficientNet-B0.onnx](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.onnx) |
44
+ | EfficientNet-B0 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 1.123 ms | 0 - 15 MB | FP16 | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
45
+ | EfficientNet-B0 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 1.179 ms | 0 - 15 MB | FP16 | NPU | Use Export Script |
46
+ | EfficientNet-B0 | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 1.178 ms | 0 - 35 MB | FP16 | NPU | [EfficientNet-B0.onnx](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.onnx) |
47
+ | EfficientNet-B0 | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 1.602 ms | 0 - 92 MB | FP16 | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
48
+ | EfficientNet-B0 | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 1.561 ms | 1 - 2 MB | FP16 | NPU | Use Export Script |
49
+ | EfficientNet-B0 | SA7255P ADP | SA7255P | TFLITE | 22.567 ms | 0 - 15 MB | FP16 | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
50
+ | EfficientNet-B0 | SA7255P ADP | SA7255P | QNN | 22.504 ms | 1 - 6 MB | FP16 | NPU | Use Export Script |
51
+ | EfficientNet-B0 | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 1.603 ms | 0 - 92 MB | FP16 | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
52
+ | EfficientNet-B0 | SA8255 (Proxy) | SA8255P Proxy | QNN | 1.574 ms | 1 - 2 MB | FP16 | NPU | Use Export Script |
53
+ | EfficientNet-B0 | SA8295P ADP | SA8295P | TFLITE | 3.784 ms | 0 - 15 MB | FP16 | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
54
+ | EfficientNet-B0 | SA8295P ADP | SA8295P | QNN | 3.675 ms | 0 - 6 MB | FP16 | NPU | Use Export Script |
55
+ | EfficientNet-B0 | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 1.613 ms | 0 - 91 MB | FP16 | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
56
+ | EfficientNet-B0 | SA8650 (Proxy) | SA8650P Proxy | QNN | 1.57 ms | 1 - 2 MB | FP16 | NPU | Use Export Script |
57
+ | EfficientNet-B0 | SA8775P ADP | SA8775P | TFLITE | 2.556 ms | 0 - 16 MB | FP16 | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
58
+ | EfficientNet-B0 | SA8775P ADP | SA8775P | QNN | 2.683 ms | 1 - 6 MB | FP16 | NPU | Use Export Script |
59
+ | EfficientNet-B0 | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 3.083 ms | 0 - 23 MB | FP16 | NPU | [EfficientNet-B0.tflite](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.tflite) |
60
+ | EfficientNet-B0 | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 3.163 ms | 1 - 25 MB | FP16 | NPU | Use Export Script |
61
+ | EfficientNet-B0 | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 1.782 ms | 1 - 1 MB | FP16 | NPU | Use Export Script |
62
+ | EfficientNet-B0 | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 1.694 ms | 15 - 15 MB | FP16 | NPU | [EfficientNet-B0.onnx](https://huggingface.co/qualcomm/EfficientNet-B0/blob/main/EfficientNet-B0.onnx) |
63
 
64
 
65
 
 
124
  Device : Samsung Galaxy S23 (13)
125
  Runtime : TFLITE
126
  Estimated inference time (ms) : 1.6
127
+ Estimated peak memory usage (MB): [0, 81]
128
  Total # Ops : 245
129
  Compute Unit(s) : NPU (245 ops)
130
  ```
 
145
  import torch
146
 
147
  import qai_hub as hub
148
+ from qai_hub_models.models.efficientnet_b0 import Model
149
 
150
  # Load the model
151
+ torch_model = Model.from_pretrained()
152
 
153
  # Device
154
  device = hub.Device("Samsung Galaxy S23")
155
 
156
+ # Trace model
157
+ input_shape = torch_model.get_input_spec()
158
+ sample_inputs = torch_model.sample_inputs()
159
+
160
+ pt_model = torch.jit.trace(torch_model, [torch.tensor(data[0]) for _, data in sample_inputs.items()])
161
+
162
+ # Compile model on a specific device
163
+ compile_job = hub.submit_compile_job(
164
+ model=pt_model,
165
+ device=device,
166
+ input_specs=torch_model.get_input_spec(),
167
+ )
168
+
169
+ # Get target model to run on-device
170
+ target_model = compile_job.get_target_model()
171
 
172
  ```
173