Hugging Face's logo

qualcomm
/
ResNet-2Plus1D

Video Classification
PyTorch
backbone
android
Model card Files
xet
 Community
qaihm-bot commited on
Commit
0eb525b
·
verified ·
1 Parent(s): 4d87200

v0.46.1

Browse files

See https://github.com/quic/ai-hub-models/releases/v0.46.1 for changelog.

Files changed (8) hide show
  1. README.md +113 -249
  2. ResNet-2Plus1D_float.dlc +0 -3
  3. ResNet-2Plus1D_float.onnx.zip +0 -3
  4. ResNet-2Plus1D_float.tflite +0 -3
  5. ResNet-2Plus1D_w8a8.dlc +0 -3
  6. ResNet-2Plus1D_w8a8.onnx.zip +0 -3
  7. ResNet-2Plus1D_w8a8.tflite +0 -3
  8. tool-versions.yaml +0 -4
README.md CHANGED
@@ -10,267 +10,131 @@ pipeline_tag: video-classification
10
 
11
  ![](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/resnet_2plus1d/web-assets/model_demo.png)
12
 
13
- # ResNet-2Plus1D: Optimized for Mobile Deployment
14
- ## Sports and human action recognition in videos
15
-
16
 
17
  ResNet (2+1)D Convolutions is a network which explicitly factorizes 3D convolution into two separate and successive operations, a 2D spatial convolution and a 1D temporal convolution. It used for video understanding applications.
18
 
19
- This model is an implementation of ResNet-2Plus1D found [here](https://github.com/pytorch/vision/blob/main/torchvision/models/video/resnet.py).
20
-
21
-
22
- This repository provides scripts to run ResNet-2Plus1D on Qualcomm® devices.
23
- More details on model performance across various devices, can be found
24
- [here](https://aihub.qualcomm.com/models/resnet_2plus1d).
25
-
26
-
27
-
28
- ### Model Details
29
-
30
- - **Model Type:** Model_use_case.video_classification
31
- - **Model Stats:**
32
- - Model checkpoint: Kinetics-400
33
- - Input resolution: 112x112
34
- - Number of parameters: 31.5M
35
- - Model size (float): 120 MB
36
- - Model size (w8a8): 30.8 MB
37
-
38
- | Model | Precision | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit | Target Model
39
- |---|---|---|---|---|---|---|---|---|
40
- | ResNet-2Plus1D | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 736.652 ms | 0 - 195 MB | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.tflite) |
41
- | ResNet-2Plus1D | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC | 81.891 ms | 0 - 181 MB | NPU | [ResNet-2Plus1D.dlc](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.dlc) |
42
- | ResNet-2Plus1D | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 459.478 ms | 0 - 272 MB | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.tflite) |
43
- | ResNet-2Plus1D | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 26.954 ms | 2 - 247 MB | NPU | [ResNet-2Plus1D.dlc](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.dlc) |
44
- | ResNet-2Plus1D | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 398.418 ms | 0 - 3 MB | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.tflite) |
45
- | ResNet-2Plus1D | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 12.863 ms | 2 - 4 MB | NPU | [ResNet-2Plus1D.dlc](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.dlc) |
46
- | ResNet-2Plus1D | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | ONNX | 11.931 ms | 0 - 64 MB | NPU | [ResNet-2Plus1D.onnx.zip](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.onnx.zip) |
47
- | ResNet-2Plus1D | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 392.239 ms | 0 - 195 MB | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.tflite) |
48
- | ResNet-2Plus1D | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC | 21.602 ms | 0 - 192 MB | NPU | [ResNet-2Plus1D.dlc](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.dlc) |
49
- | ResNet-2Plus1D | float | SA7255P ADP | Qualcomm® SA7255P | TFLITE | 736.652 ms | 0 - 195 MB | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.tflite) |
50
- | ResNet-2Plus1D | float | SA7255P ADP | Qualcomm® SA7255P | QNN_DLC | 81.891 ms | 0 - 181 MB | NPU | [ResNet-2Plus1D.dlc](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.dlc) |
51
- | ResNet-2Plus1D | float | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 475.165 ms | 0 - 188 MB | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.tflite) |
52
- | ResNet-2Plus1D | float | SA8295P ADP | Qualcomm® SA8295P | QNN_DLC | 22.955 ms | 0 - 168 MB | NPU | [ResNet-2Plus1D.dlc](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.dlc) |
53
- | ResNet-2Plus1D | float | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 392.239 ms | 0 - 195 MB | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.tflite) |
54
- | ResNet-2Plus1D | float | SA8775P ADP | Qualcomm® SA8775P | QNN_DLC | 21.602 ms | 0 - 192 MB | NPU | [ResNet-2Plus1D.dlc](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.dlc) |
55
- | ResNet-2Plus1D | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 292.093 ms | 0 - 281 MB | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.tflite) |
56
- | ResNet-2Plus1D | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 9.306 ms | 2 - 269 MB | NPU | [ResNet-2Plus1D.dlc](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.dlc) |
57
- | ResNet-2Plus1D | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 8.731 ms | 2 - 233 MB | NPU | [ResNet-2Plus1D.onnx.zip](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.onnx.zip) |
58
- | ResNet-2Plus1D | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | TFLITE | 266.815 ms | 0 - 193 MB | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.tflite) |
59
- | ResNet-2Plus1D | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | QNN_DLC | 7.356 ms | 2 - 184 MB | NPU | [ResNet-2Plus1D.dlc](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.dlc) |
60
- | ResNet-2Plus1D | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | ONNX | 7.199 ms | 1 - 144 MB | NPU | [ResNet-2Plus1D.onnx.zip](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.onnx.zip) |
61
- | ResNet-2Plus1D | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen 5 Mobile | TFLITE | 259.756 ms | 0 - 203 MB | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.tflite) |
62
- | ResNet-2Plus1D | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen 5 Mobile | QNN_DLC | 5.418 ms | 2 - 191 MB | NPU | [ResNet-2Plus1D.dlc](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.dlc) |
63
- | ResNet-2Plus1D | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen 5 Mobile | ONNX | 5.51 ms | 2 - 147 MB | NPU | [ResNet-2Plus1D.onnx.zip](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.onnx.zip) |
64
- | ResNet-2Plus1D | float | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_DLC | 13.285 ms | 2 - 2 MB | NPU | [ResNet-2Plus1D.dlc](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.dlc) |
65
- | ResNet-2Plus1D | float | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 12.322 ms | 60 - 60 MB | NPU | [ResNet-2Plus1D.onnx.zip](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D.onnx.zip) |
66
- | ResNet-2Plus1D | w8a8 | Dragonwing Q-6690 MTP | Qualcomm® QCM6690 | TFLITE | 1621.204 ms | 321 - 478 MB | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.tflite) |
67
- | ResNet-2Plus1D | w8a8 | Dragonwing Q-6690 MTP | Qualcomm® QCM6690 | QNN_DLC | 68.331 ms | 1 - 167 MB | NPU | [ResNet-2Plus1D.dlc](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.dlc) |
68
- | ResNet-2Plus1D | w8a8 | Dragonwing Q-6690 MTP | Qualcomm® QCM6690 | ONNX | 299.164 ms | 99 - 112 MB | CPU | [ResNet-2Plus1D.onnx.zip](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.onnx.zip) |
69
- | ResNet-2Plus1D | w8a8 | Dragonwing RB3 Gen 2 Vision Kit | Qualcomm® QCS6490 | TFLITE | 1797.052 ms | 267 - 428 MB | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.tflite) |
70
- | ResNet-2Plus1D | w8a8 | Dragonwing RB3 Gen 2 Vision Kit | Qualcomm® QCS6490 | QNN_DLC | 17.317 ms | 1 - 3 MB | NPU | [ResNet-2Plus1D.dlc](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.dlc) |
71
- | ResNet-2Plus1D | w8a8 | Dragonwing RB3 Gen 2 Vision Kit | Qualcomm® QCS6490 | ONNX | 317.409 ms | 98 - 129 MB | CPU | [ResNet-2Plus1D.onnx.zip](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.onnx.zip) |
72
- | ResNet-2Plus1D | w8a8 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 1292.279 ms | 0 - 394 MB | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.tflite) |
73
- | ResNet-2Plus1D | w8a8 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC | 13.087 ms | 1 - 164 MB | NPU | [ResNet-2Plus1D.dlc](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.dlc) |
74
- | ResNet-2Plus1D | w8a8 | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 919.714 ms | 0 - 371 MB | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.tflite) |
75
- | ResNet-2Plus1D | w8a8 | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 7.175 ms | 1 - 202 MB | NPU | [ResNet-2Plus1D.dlc](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.dlc) |
76
- | ResNet-2Plus1D | w8a8 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 699.325 ms | 0 - 3 MB | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.tflite) |
77
- | ResNet-2Plus1D | w8a8 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 4.426 ms | 1 - 3 MB | NPU | [ResNet-2Plus1D.dlc](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.dlc) |
78
- | ResNet-2Plus1D | w8a8 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | ONNX | 4.359 ms | 0 - 35 MB | NPU | [ResNet-2Plus1D.onnx.zip](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.onnx.zip) |
79
- | ResNet-2Plus1D | w8a8 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 672.511 ms | 0 - 339 MB | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.tflite) |
80
- | ResNet-2Plus1D | w8a8 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC | 4.653 ms | 1 - 163 MB | NPU | [ResNet-2Plus1D.dlc](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.dlc) |
81
- | ResNet-2Plus1D | w8a8 | SA7255P ADP | Qualcomm® SA7255P | TFLITE | 1292.279 ms | 0 - 394 MB | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.tflite) |
82
- | ResNet-2Plus1D | w8a8 | SA7255P ADP | Qualcomm® SA7255P | QNN_DLC | 13.087 ms | 1 - 164 MB | NPU | [ResNet-2Plus1D.dlc](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.dlc) |
83
- | ResNet-2Plus1D | w8a8 | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 763.639 ms | 0 - 340 MB | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.tflite) |
84
- | ResNet-2Plus1D | w8a8 | SA8295P ADP | Qualcomm® SA8295P | QNN_DLC | 7.567 ms | 1 - 166 MB | NPU | [ResNet-2Plus1D.dlc](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.dlc) |
85
- | ResNet-2Plus1D | w8a8 | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 672.511 ms | 0 - 339 MB | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.tflite) |
86
- | ResNet-2Plus1D | w8a8 | SA8775P ADP | Qualcomm® SA8775P | QNN_DLC | 4.653 ms | 1 - 163 MB | NPU | [ResNet-2Plus1D.dlc](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.dlc) |
87
- | ResNet-2Plus1D | w8a8 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 512.136 ms | 0 - 438 MB | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.tflite) |
88
- | ResNet-2Plus1D | w8a8 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 3.222 ms | 1 - 204 MB | NPU | [ResNet-2Plus1D.dlc](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.dlc) |
89
- | ResNet-2Plus1D | w8a8 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 3.205 ms | 0 - 184 MB | NPU | [ResNet-2Plus1D.onnx.zip](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.onnx.zip) |
90
- | ResNet-2Plus1D | w8a8 | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | TFLITE | 396.944 ms | 0 - 419 MB | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.tflite) |
91
- | ResNet-2Plus1D | w8a8 | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | QNN_DLC | 2.411 ms | 1 - 161 MB | NPU | [ResNet-2Plus1D.dlc](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.dlc) |
92
- | ResNet-2Plus1D | w8a8 | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | ONNX | 2.615 ms | 0 - 130 MB | NPU | [ResNet-2Plus1D.onnx.zip](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.onnx.zip) |
93
- | ResNet-2Plus1D | w8a8 | Snapdragon 7 Gen 4 QRD | Snapdragon® 7 Gen 4 Mobile | TFLITE | 1244.845 ms | 334 - 510 MB | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.tflite) |
94
- | ResNet-2Plus1D | w8a8 | Snapdragon 7 Gen 4 QRD | Snapdragon® 7 Gen 4 Mobile | QNN_DLC | 7.607 ms | 1 - 158 MB | NPU | [ResNet-2Plus1D.dlc](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.dlc) |
95
- | ResNet-2Plus1D | w8a8 | Snapdragon 7 Gen 4 QRD | Snapdragon® 7 Gen 4 Mobile | ONNX | 286.964 ms | 95 - 110 MB | CPU | [ResNet-2Plus1D.onnx.zip](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.onnx.zip) |
96
- | ResNet-2Plus1D | w8a8 | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen 5 Mobile | TFLITE | 582.234 ms | 0 - 372 MB | NPU | [ResNet-2Plus1D.tflite](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.tflite) |
97
- | ResNet-2Plus1D | w8a8 | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen 5 Mobile | QNN_DLC | 1.782 ms | 1 - 165 MB | NPU | [ResNet-2Plus1D.dlc](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.dlc) |
98
- | ResNet-2Plus1D | w8a8 | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen 5 Mobile | ONNX | 1.906 ms | 0 - 131 MB | NPU | [ResNet-2Plus1D.onnx.zip](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.onnx.zip) |
99
- | ResNet-2Plus1D | w8a8 | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_DLC | 4.749 ms | 1 - 1 MB | NPU | [ResNet-2Plus1D.dlc](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.dlc) |
100
- | ResNet-2Plus1D | w8a8 | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 4.487 ms | 31 - 31 MB | NPU | [ResNet-2Plus1D.onnx.zip](https://huggingface.co/qualcomm/ResNet-2Plus1D/blob/main/ResNet-2Plus1D_w8a8.onnx.zip) |
101
-
102
-
103
-
104
-
105
- ## Installation
106
-
107
-
108
- Install the package via pip:
109
- ```bash
110
- # NOTE: 3.10 <= PYTHON_VERSION < 3.14 is supported.
111
- pip install "qai-hub-models[resnet-2plus1d]"
112
- ```
113
-
114
-
115
- ## Configure Qualcomm® AI Hub Workbench to run this model on a cloud-hosted device
116
-
117
- Sign-in to [Qualcomm® AI Hub Workbench](https://workbench.aihub.qualcomm.com/) with your
118
- Qualcomm® ID. Once signed in navigate to `Account -> Settings -> API Token`.
119
-
120
- With this API token, you can configure your client to run models on the cloud
121
- hosted devices.
122
- ```bash
123
- qai-hub configure --api_token API_TOKEN
124
- ```
125
- Navigate to [docs](https://workbench.aihub.qualcomm.com/docs/) for more information.
126
-
127
-
128
-
129
- ## Demo off target
130
-
131
- The package contains a simple end-to-end demo that downloads pre-trained
132
- weights and runs this model on a sample input.
133
-
134
- ```bash
135
- python -m qai_hub_models.models.resnet_2plus1d.demo
136
- ```
137
-
138
- The above demo runs a reference implementation of pre-processing, model
139
- inference, and post processing.
140
-
141
- **NOTE**: If you want running in a Jupyter Notebook or Google Colab like
142
- environment, please add the following to your cell (instead of the above).
143
- ```
144
- %run -m qai_hub_models.models.resnet_2plus1d.demo
145
- ```
146
-
147
-
148
- ### Run model on a cloud-hosted device
149
-
150
- In addition to the demo, you can also run the model on a cloud-hosted Qualcomm®
151
- device. This script does the following:
152
- * Performance check on-device on a cloud-hosted device
153
- * Downloads compiled assets that can be deployed on-device for Android.
154
- * Accuracy check between PyTorch and on-device outputs.
155
-
156
- ```bash
157
- python -m qai_hub_models.models.resnet_2plus1d.export
158
- ```
159
-
160
-
161
-
162
- ## How does this work?
163
-
164
- This [export script](https://aihub.qualcomm.com/models/resnet_2plus1d/qai_hub_models/models/ResNet-2Plus1D/export.py)
165
- leverages [Qualcomm® AI Hub](https://aihub.qualcomm.com/) to optimize, validate, and deploy this model
166
- on-device. Lets go through each step below in detail:
167
-
168
- Step 1: **Compile model for on-device deployment**
169
-
170
- To compile a PyTorch model for on-device deployment, we first trace the model
171
- in memory using the `jit.trace` and then call the `submit_compile_job` API.
172
-
173
- ```python
174
- import torch
175
-
176
- import qai_hub as hub
177
- from qai_hub_models.models.resnet_2plus1d import Model
178
-
179
- # Load the model
180
- torch_model = Model.from_pretrained()
181
-
182
- # Device
183
- device = hub.Device("Samsung Galaxy S25")
184
-
185
- # Trace model
186
- input_shape = torch_model.get_input_spec()
187
- sample_inputs = torch_model.sample_inputs()
188
-
189
- pt_model = torch.jit.trace(torch_model, [torch.tensor(data[0]) for _, data in sample_inputs.items()])
190
-
191
- # Compile model on a specific device
192
- compile_job = hub.submit_compile_job(
193
- model=pt_model,
194
- device=device,
195
- input_specs=torch_model.get_input_spec(),
196
- )
197
-
198
- # Get target model to run on-device
199
- target_model = compile_job.get_target_model()
200
-
201
- ```
202
-
203
-
204
- Step 2: **Performance profiling on cloud-hosted device**
205
-
206
- After compiling models from step 1. Models can be profiled model on-device using the
207
- `target_model`. Note that this scripts runs the model on a device automatically
208
- provisioned in the cloud. Once the job is submitted, you can navigate to a
209
- provided job URL to view a variety of on-device performance metrics.
210
- ```python
211
- profile_job = hub.submit_profile_job(
212
- model=target_model,
213
- device=device,
214
- )
215
-
216
- ```
217
-
218
- Step 3: **Verify on-device accuracy**
219
-
220
- To verify the accuracy of the model on-device, you can run on-device inference
221
- on sample input data on the same cloud hosted device.
222
- ```python
223
- input_data = torch_model.sample_inputs()
224
- inference_job = hub.submit_inference_job(
225
- model=target_model,
226
- device=device,
227
- inputs=input_data,
228
- )
229
- on_device_output = inference_job.download_output_data()
230
-
231
- ```
232
- With the output of the model, you can compute like PSNR, relative errors or
233
- spot check the output with expected output.
234
-
235
- **Note**: This on-device profiling and inference requires access to Qualcomm®
236
- AI Hub Workbench. [Sign up for access](https://myaccount.qualcomm.com/signup).
237
-
238
-
239
-
240
-
241
- ## Deploying compiled model to Android
242
-
243
-
244
- The models can be deployed using multiple runtimes:
245
- - TensorFlow Lite (`.tflite` export): [This
246
- tutorial](https://www.tensorflow.org/lite/android/quickstart) provides a
247
- guide to deploy the .tflite model in an Android application.
248
-
249
-
250
- - QNN (`.so` export ): This [sample
251
- app](https://docs.qualcomm.com/bundle/publicresource/topics/80-63442-50/sample_app.html)
252
- provides instructions on how to use the `.so` shared library in an Android application.
253
-
254
-
255
- ## View on Qualcomm® AI Hub
256
- Get more details on ResNet-2Plus1D's performance across various devices [here](https://aihub.qualcomm.com/models/resnet_2plus1d).
257
- Explore all available models on [Qualcomm® AI Hub](https://aihub.qualcomm.com/)
258
-
259
 
260
  ## License
261
  * The license for the original implementation of ResNet-2Plus1D can be found
262
  [here](https://github.com/pytorch/vision/blob/main/LICENSE).
263
 
264
-
265
-
266
  ## References
267
  * [A Closer Look at Spatiotemporal Convolutions for Action Recognition](https://arxiv.org/abs/1711.11248)
268
  * [Source Model Implementation](https://github.com/pytorch/vision/blob/main/torchvision/models/video/resnet.py)
269
 
270
-
271
-
272
  ## Community
273
  * Join [our AI Hub Slack community](https://aihub.qualcomm.com/community/slack) to collaborate, post questions and learn more about on-device AI.
274
  * For questions or feedback please [reach out to us](mailto:ai-hub-support@qti.qualcomm.com).
275
-
276
-
 
10
 
11
  ![](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/resnet_2plus1d/web-assets/model_demo.png)
12
 
13
+ # ResNet-2Plus1D: Optimized for Qualcomm Devices
 
 
14
 
15
  ResNet (2+1)D Convolutions is a network which explicitly factorizes 3D convolution into two separate and successive operations, a 2D spatial convolution and a 1D temporal convolution. It used for video understanding applications.
16
 
17
+ This is based on the implementation of ResNet-2Plus1D found [here](https://github.com/pytorch/vision/blob/main/torchvision/models/video/resnet.py).
18
+ This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the [Qualcomm® AI Hub Models](https://github.com/quic/ai-hub-models/blob/main/qai_hub_models/models/resnet_2plus1d) library to export with custom configurations. More details on model performance across various devices, can be found [here](#performance-summary).
19
+
20
+ Qualcomm AI Hub Models uses [Qualcomm AI Hub Workbench](https://workbench.aihub.qualcomm.com) to compile, profile, and evaluate this model. [Sign up](https://myaccount.qualcomm.com/signup) to run these models on a hosted Qualcomm® device.
21
+
22
+ ## Getting Started
23
+ There are two ways to deploy this model on your device:
24
+
25
+ ### Option 1: Download Pre-Exported Models
26
+
27
+ Below are pre-exported model assets ready for deployment.
28
+
29
+ | Runtime | Precision | Chipset | SDK Versions | Download |
30
+ |---|---|---|---|---|
31
+ | ONNX | float | Universal | QAIRT 2.37, ONNX Runtime 1.23.0 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/resnet_2plus1d/releases/v0.46.1/resnet_2plus1d-onnx-float.zip)
32
+ | ONNX | w8a8 | Universal | QAIRT 2.37, ONNX Runtime 1.23.0 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/resnet_2plus1d/releases/v0.46.1/resnet_2plus1d-onnx-w8a8.zip)
33
+ | QNN_DLC | float | Universal | QAIRT 2.42 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/resnet_2plus1d/releases/v0.46.1/resnet_2plus1d-qnn_dlc-float.zip)
34
+ | QNN_DLC | w8a8 | Universal | QAIRT 2.42 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/resnet_2plus1d/releases/v0.46.1/resnet_2plus1d-qnn_dlc-w8a8.zip)
35
+ | TFLITE | float | Universal | QAIRT 2.42, TFLite 2.17.0 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/resnet_2plus1d/releases/v0.46.1/resnet_2plus1d-tflite-float.zip)
36
+ | TFLITE | w8a8 | Universal | QAIRT 2.42, TFLite 2.17.0 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/resnet_2plus1d/releases/v0.46.1/resnet_2plus1d-tflite-w8a8.zip)
37
+
38
+ For more device-specific assets and performance metrics, visit **[ResNet-2Plus1D on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/resnet_2plus1d)**.
39
+
40
+
41
+ ### Option 2: Export with Custom Configurations
42
+
43
+ Use the [Qualcomm® AI Hub Models](https://github.com/quic/ai-hub-models/blob/main/qai_hub_models/models/resnet_2plus1d) Python library to compile and export the model with your own:
44
+ - Custom weights (e.g., fine-tuned checkpoints)
45
+ - Custom input shapes
46
+ - Target device and runtime configurations
47
+
48
+ This option is ideal if you need to customize the model beyond the default configuration provided here.
49
+
50
+ See our repository for [ResNet-2Plus1D on GitHub](https://github.com/quic/ai-hub-models/blob/main/qai_hub_models/models/resnet_2plus1d) for usage instructions.
51
+
52
+ ## Model Details
53
+
54
+ **Model Type:** Model_use_case.video_classification
55
+
56
+ **Model Stats:**
57
+ - Model checkpoint: Kinetics-400
58
+ - Input resolution: 112x112
59
+ - Number of parameters: 31.5M
60
+ - Model size (float): 120 MB
61
+ - Model size (w8a8): 30.8 MB
62
+
63
+ ## Performance Summary
64
+ | Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
65
+ |---|---|---|---|---|---|---
66
+ | ResNet-2Plus1D | ONNX | float | Snapdragon® X Elite | 12.314 ms | 60 - 60 MB | NPU
67
+ | ResNet-2Plus1D | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 8.743 ms | 0 - 229 MB | NPU
68
+ | ResNet-2Plus1D | ONNX | float | Qualcomm® QCS8550 (Proxy) | 12.02 ms | 0 - 64 MB | NPU
69
+ | ResNet-2Plus1D | ONNX | float | Qualcomm® QCS9075 | 22.247 ms | 2 - 7 MB | NPU
70
+ | ResNet-2Plus1D | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 7.188 ms | 0 - 144 MB | NPU
71
+ | ResNet-2Plus1D | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 5.493 ms | 2 - 150 MB | NPU
72
+ | ResNet-2Plus1D | ONNX | w8a8 | Snapdragon® X Elite | 4.492 ms | 31 - 31 MB | NPU
73
+ | ResNet-2Plus1D | ONNX | w8a8 | Snapdragon® 8 Gen 3 Mobile | 3.21 ms | 0 - 184 MB | NPU
74
+ | ResNet-2Plus1D | ONNX | w8a8 | Qualcomm® QCS6490 | 316.355 ms | 97 - 129 MB | CPU
75
+ | ResNet-2Plus1D | ONNX | w8a8 | Qualcomm® QCS8550 (Proxy) | 4.359 ms | 0 - 37 MB | NPU
76
+ | ResNet-2Plus1D | ONNX | w8a8 | Qualcomm® QCS9075 | 4.338 ms | 1 - 3 MB | NPU
77
+ | ResNet-2Plus1D | ONNX | w8a8 | Qualcomm® QCM6690 | 300.353 ms | 98 - 106 MB | CPU
78
+ | ResNet-2Plus1D | ONNX | w8a8 | Snapdragon® 8 Elite For Galaxy Mobile | 2.618 ms | 0 - 130 MB | NPU
79
+ | ResNet-2Plus1D | ONNX | w8a8 | Snapdragon® 7 Gen 4 Mobile | 266.261 ms | 67 - 74 MB | CPU
80
+ | ResNet-2Plus1D | ONNX | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 1.909 ms | 0 - 133 MB | NPU
81
+ | ResNet-2Plus1D | QNN_DLC | float | Snapdragon® X Elite | 12.962 ms | 2 - 2 MB | NPU
82
+ | ResNet-2Plus1D | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 9.292 ms | 0 - 301 MB | NPU
83
+ | ResNet-2Plus1D | QNN_DLC | float | Qualcomm® QCS8275 (Proxy) | 81.966 ms | 1 - 216 MB | NPU
84
+ | ResNet-2Plus1D | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 12.555 ms | 2 - 4 MB | NPU
85
+ | ResNet-2Plus1D | QNN_DLC | float | Qualcomm® SA8775P | 21.332 ms | 0 - 214 MB | NPU
86
+ | ResNet-2Plus1D | QNN_DLC | float | Qualcomm® QCS9075 | 22.91 ms | 2 - 6 MB | NPU
87
+ | ResNet-2Plus1D | QNN_DLC | float | Qualcomm® QCS8450 (Proxy) | 28.82 ms | 1 - 278 MB | NPU
88
+ | ResNet-2Plus1D | QNN_DLC | float | Qualcomm® SA7255P | 81.966 ms | 1 - 216 MB | NPU
89
+ | ResNet-2Plus1D | QNN_DLC | float | Qualcomm® SA8295P | 22.718 ms | 0 - 198 MB | NPU
90
+ | ResNet-2Plus1D | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 7.243 ms | 0 - 233 MB | NPU
91
+ | ResNet-2Plus1D | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 5.598 ms | 2 - 230 MB | NPU
92
+ | ResNet-2Plus1D | QNN_DLC | w8a8 | Snapdragon® X Elite | 4.897 ms | 1 - 1 MB | NPU
93
+ | ResNet-2Plus1D | QNN_DLC | w8a8 | Snapdragon® 8 Gen 3 Mobile | 3.334 ms | 1 - 221 MB | NPU
94
+ | ResNet-2Plus1D | QNN_DLC | w8a8 | Qualcomm® QCS6490 | 19.711 ms | 1 - 3 MB | NPU
95
+ | ResNet-2Plus1D | QNN_DLC | w8a8 | Qualcomm® QCS8275 (Proxy) | 13.435 ms | 1 - 182 MB | NPU
96
+ | ResNet-2Plus1D | QNN_DLC | w8a8 | Qualcomm® QCS8550 (Proxy) | 4.565 ms | 1 - 28 MB | NPU
97
+ | ResNet-2Plus1D | QNN_DLC | w8a8 | Qualcomm® SA8775P | 20.06 ms | 1 - 181 MB | NPU
98
+ | ResNet-2Plus1D | QNN_DLC | w8a8 | Qualcomm® QCS9075 | 4.778 ms | 3 - 5 MB | NPU
99
+ | ResNet-2Plus1D | QNN_DLC | w8a8 | Qualcomm® QCM6690 | 72.289 ms | 1 - 197 MB | NPU
100
+ | ResNet-2Plus1D | QNN_DLC | w8a8 | Qualcomm® QCS8450 (Proxy) | 7.878 ms | 0 - 218 MB | NPU
101
+ | ResNet-2Plus1D | QNN_DLC | w8a8 | Qualcomm® SA7255P | 13.435 ms | 1 - 182 MB | NPU
102
+ | ResNet-2Plus1D | QNN_DLC | w8a8 | Qualcomm® SA8295P | 7.834 ms | 1 - 180 MB | NPU
103
+ | ResNet-2Plus1D | QNN_DLC | w8a8 | Snapdragon® 8 Elite For Galaxy Mobile | 2.574 ms | 1 - 181 MB | NPU
104
+ | ResNet-2Plus1D | QNN_DLC | w8a8 | Snapdragon® 7 Gen 4 Mobile | 7.789 ms | 1 - 189 MB | NPU
105
+ | ResNet-2Plus1D | QNN_DLC | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 1.874 ms | 1 - 185 MB | NPU
106
+ | ResNet-2Plus1D | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 287.649 ms | 0 - 323 MB | NPU
107
+ | ResNet-2Plus1D | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 726.044 ms | 0 - 236 MB | NPU
108
+ | ResNet-2Plus1D | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 405.049 ms | 0 - 3 MB | NPU
109
+ | ResNet-2Plus1D | TFLITE | float | Qualcomm® SA8775P | 369.49 ms | 0 - 236 MB | NPU
110
+ | ResNet-2Plus1D | TFLITE | float | Qualcomm® QCS9075 | 391.464 ms | 0 - 66 MB | NPU
111
+ | ResNet-2Plus1D | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 468.721 ms | 0 - 313 MB | NPU
112
+ | ResNet-2Plus1D | TFLITE | float | Qualcomm® SA7255P | 726.044 ms | 0 - 236 MB | NPU
113
+ | ResNet-2Plus1D | TFLITE | float | Qualcomm® SA8295P | 476.42 ms | 0 - 229 MB | NPU
114
+ | ResNet-2Plus1D | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 270.467 ms | 0 - 241 MB | NPU
115
+ | ResNet-2Plus1D | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 253.274 ms | 0 - 246 MB | NPU
116
+ | ResNet-2Plus1D | TFLITE | w8a8 | Snapdragon® 8 Gen 3 Mobile | 550.687 ms | 0 - 507 MB | NPU
117
+ | ResNet-2Plus1D | TFLITE | w8a8 | Qualcomm® QCS6490 | 1789.997 ms | 269 - 430 MB | NPU
118
+ | ResNet-2Plus1D | TFLITE | w8a8 | Qualcomm® QCS8275 (Proxy) | 1427.743 ms | 0 - 437 MB | NPU
119
+ | ResNet-2Plus1D | TFLITE | w8a8 | Qualcomm® QCS8550 (Proxy) | 757.743 ms | 0 - 2 MB | NPU
120
+ | ResNet-2Plus1D | TFLITE | w8a8 | Qualcomm® SA8775P | 754.27 ms | 0 - 434 MB | NPU
121
+ | ResNet-2Plus1D | TFLITE | w8a8 | Qualcomm® QCS9075 | 569.83 ms | 0 - 65 MB | NPU
122
+ | ResNet-2Plus1D | TFLITE | w8a8 | Qualcomm® QCM6690 | 1614.51 ms | 258 - 440 MB | NPU
123
+ | ResNet-2Plus1D | TFLITE | w8a8 | Qualcomm® QCS8450 (Proxy) | 858.539 ms | 0 - 428 MB | NPU
124
+ | ResNet-2Plus1D | TFLITE | w8a8 | Qualcomm® SA7255P | 1427.743 ms | 0 - 437 MB | NPU
125
+ | ResNet-2Plus1D | TFLITE | w8a8 | Qualcomm® SA8295P | 829.459 ms | 0 - 400 MB | NPU
126
+ | ResNet-2Plus1D | TFLITE | w8a8 | Snapdragon® 8 Elite For Galaxy Mobile | 478.279 ms | 0 - 546 MB | NPU
127
+ | ResNet-2Plus1D | TFLITE | w8a8 | Snapdragon® 7 Gen 4 Mobile | 1250.05 ms | 260 - 363 MB | NPU
128
+ | ResNet-2Plus1D | TFLITE | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 565.356 ms | 0 - 476 MB | NPU
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
129
 
130
  ## License
131
  * The license for the original implementation of ResNet-2Plus1D can be found
132
  [here](https://github.com/pytorch/vision/blob/main/LICENSE).
133
 
 
 
134
  ## References
135
  * [A Closer Look at Spatiotemporal Convolutions for Action Recognition](https://arxiv.org/abs/1711.11248)
136
  * [Source Model Implementation](https://github.com/pytorch/vision/blob/main/torchvision/models/video/resnet.py)
137
 
 
 
138
  ## Community
139
  * Join [our AI Hub Slack community](https://aihub.qualcomm.com/community/slack) to collaborate, post questions and learn more about on-device AI.
140
  * For questions or feedback please [reach out to us](mailto:ai-hub-support@qti.qualcomm.com).
 
 
ResNet-2Plus1D_float.dlc DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:5cc3a966d5f9ccc9fe76a6e491657571172338dcfcf452903f95bab682de6434
3
- size 126074004
 
 
 
 
ResNet-2Plus1D_float.onnx.zip DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:c754715e3601577fd7ab7030dc114a5cf9c8078140d668ffad8dfc5d20161664
3
- size 117305806
 
 
 
 
ResNet-2Plus1D_float.tflite DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:78eeb08f495b6c5740f5dbe62192d56569ab3bcc4a81a324599b00c81ca0ea80
3
- size 125994436
 
 
 
 
ResNet-2Plus1D_w8a8.dlc DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:838d36a33731963644a0515f294c2d1302716da68fafeb342de6e28bf8a7f27a
3
- size 32350620
 
 
 
 
ResNet-2Plus1D_w8a8.onnx.zip DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:1579045117c7cf9d045811444b81ecd0e134fca29ed67f99fee632237532b4c3
3
- size 53368255
 
 
 
 
ResNet-2Plus1D_w8a8.tflite DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:88159048c0eceba72c075be7b6b4a8ad282deb1f5f68a602e35818764af48201
3
- size 125399456
 
 
 
 
tool-versions.yaml DELETED
@@ -1,4 +0,0 @@
1
- tool_versions:
2
- onnx:
3
- qairt: 2.37.1.250807093845_124904
4
- onnx_runtime: 1.23.0