Skip to content

Commit 6fb3a32

Browse files
jack-majack-ma
authored
Merge pull request #1034 from EESN-W/fw702
feat: airbox q900 docs
2 parents c76f88b + 6930c3c commit 6fb3a32

File tree

416 files changed

+15950
-787
lines changed

Some content is hidden

Large Commits have some content hidden by default. Use the searchbox below for content that may be hidden.

416 files changed

+15950
-787
lines changed

‎docs/Home.md‎

Lines changed: 23 additions & 7 deletions
Original file line numberDiff line numberDiff line change
@@ -621,20 +621,20 @@ sidebar_custom_props:
621621
],
622622
},
623623
{
624-
series_en: "SG Family",
625-
series_zh: "SG 系列",
626-
series_introduction_en: "SG series is a series of computing modules and edge computing boxes developed by Radxa based on SOPHGO SG2300X. It has up to 32TOPS@INT8 computing power, powerful and complete deep learning toolchain, and up to 32-channel video processing capability. It is widely used and highly reliable.",
627-
series_introduction_zh: "SG 系列是瑞莎基于 SOPHGO SG2300X 研发的一系列计算模块和边缘智算盒子。它有高达32TOPS@INT8的算力， 具备强大完备的深度学习工具链、最高达32路视频处理能力。用途广泛，可靠性高。",
624+
series_en: "Fogwise® Family",
625+
series_zh: "Fogwise® 系列",
626+
series_introduction_en: "The FogWise® Edge Computing Series by Radxa is built on an open ecosystem and industrial-grade design, delivering integrated capabilities from device-side inference to cloud-edge collaboration. The product lineup includes edge AI computing boxes, industrial gateways, and edge servers, empowering enterprises to rapidly deploy stable and scalable edge intelligence across diverse scenarios.",
627+
series_introduction_zh: "瑞莎 FogWise® 边缘计算系列以开放生态和工业级设计为底座，提供从设备端推理到云边协同的一体化能力；产品覆盖边缘智算盒、工业网关、边缘服务器，助力企业在多场景快速落地稳定、可规模化的边缘智能。",
628628
products:
629629
[
630630
{
631631
products_name: "Fogwise® AirBox",
632-
products_photo_url: "/home/product-pictures/airbox.webp",
633-
products_link: "/sophon/airbox",
632+
products_photo_url: "/en/home/product-pictures/airbox.webp",
633+
products_link: "/fogwise/airbox",
634634
docs:
635635
[
636636
{
637-
docs_link: "/sophon/airbox",
637+
docs_link: "/fogwise/airbox",
638638
docs_photo_type: "Overview",
639639
docs_name_en: "Overview",
640640
docs_name_zh: "Radxa Fogwise® AirBox 概览",
@@ -643,6 +643,22 @@ sidebar_custom_props:
643643
},
644644
],
645645
},
646+
{
647+
products_name: "Fogwise® AIRbox Q900",
648+
products_photo_url: "/en/home/product-pictures/airbox-q900.webp",
649+
products_link: "/fogwise/airbox-q900",
650+
docs:
651+
[
652+
{
653+
docs_link: "/fogwise/airbox-q900",
654+
docs_photo_type: "Overview",
655+
docs_name_en: "Overview",
656+
docs_name_zh: "Radxa Fogwise® AIRbox Q900 概览",
657+
docs_info_en: "Overview of the Radxa Fogwise® AIRbox Q900",
658+
docs_info_zh: "Radxa Fogwise® AIRbox Q900 概览",
659+
},
660+
],
661+
},
646662
],
647663
},
648664
{

‎docs/aicore/dx-m1/dx-sdk/dx-npu-driver.md‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -37,7 +37,7 @@ lspci
3737
<NewCodeBlock tip="Host" type="device">
3838

3939
```bash
40-
tar -xvf dx_rt_npu_linux_driver_v1.3.1.tar.gz && cd dx_rt_npu_linux_driver
40+
tar -xvf dx_rt_npu_linux_driver_v1.3.1.tar.gz && cd dx_rt_npu_linux_driver/modules
4141
```
4242

4343
</NewCodeBlock>

‎docs/common/ai/_aotgan_qai-app-builder.mdx‎

Lines changed: 153 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,153 @@
1+
此文档讲述使用 [QAI AppBuilder](../qai-appbuilder) Python API 利用 Qualcomm® Hexagon™ Processor (NPU) 推理 [AOT-GAN](https://aihub.qualcomm.com/models/aotgan)
2+
图像修复模型。
3+
4+
**示例支持设备**
5+
6+
| 设备 | SoC |
7+
| --------------------- | ------- |
8+
| Fogwise® AIRbox Q900 | QCS9075 |
9+
10+
## 安装 QAI AppBuilder
11+
12+
:::tip
13+
14+
1. 请根据 [**QAI AppBuilder 安装方法**](../qai-appbuilder) 安装 QAI AppBuilder。
15+
16+
2. 请根据 [**配置 QAIRT 环境变量**](../qai-appbuilder#配置-qairt-环境变量) 配置 QAIRT 环境变量。
17+
:::
18+
19+
## 运行示例
20+
21+
### 安装依赖
22+
23+
<NewCodeBlock tip="Device" type="device">
24+
25+
```bash
26+
pip3 install requests tqdm qai-hub py3_wget opencv-python torch torchvision
27+
```
28+
29+
</NewCodeBlock>
30+
31+
### 运行脚本
32+
33+
- 进入示例目录
34+
35+
<Tabs>
36+
37+
<TabItem value="QCS9075">
38+
39+
<NewCodeBlock tip="Device" type="device">
40+
41+
```bash
42+
cd ai-engine-direct-helper/samples/python
43+
```
44+
45+
</NewCodeBlock>
46+
47+
</TabItem>
48+
49+
</Tabs>
50+
51+
- 准备输入图片，这里以以下图片为输入示例
52+
53+
{" "}
54+
55+
<div style={{ textAlign: "center" }}>
56+
<img
57+
src="/img/fogwise/airbox-q900/qairt-aotgan_input.webp"
58+
style={{ width: "65%" }}
59+
/>
60+
input image
61+
</div>
62+
63+
<div style={{ textAlign: "center" }}>
64+
<img
65+
src="/img/fogwise/airbox-q900/qairt-aotgan_mask.webp"
66+
style={{ width: "65%" }}
67+
/>
68+
mask image
69+
</div>
70+
71+
- 执行推理
72+
73+
<NewCodeBlock tip="Device" type="device">
74+
75+
```bash
76+
python3 aotgan/aotgan.py
77+
```
78+
79+
</NewCodeBlock>
80+
81+
```bash
82+
$ python3 aotgan/aotgan.py
83+
0.0ms [WARNING] <W> Initializing HtpProvider
84+
85+
/prj/qct/webtech_scratch20/mlg_user_admin/qaisw_source_repo/rel/qairt-2.37.1/point_release/SNPE_SRC/avante-tools/prebuilt/dsp/hexagon-sdk-5.5.5/ipc/fastrpc/rpcmem/src/rpcmem_android.c:38:dummy call to rpcmem_init, rpcmem APIs will be used from libxdsprpc
86+
0.0ms [WARNING] <W> This META does not have Alloc2 Support
87+
88+
0.0ms [WARNING] <W> This META does not have Alloc2 Support
89+
90+
0.0ms [WARNING] <W> This META does not have Alloc2 Support
91+
92+
0.0ms [WARNING] <W> This META does not have Alloc2 Support
93+
94+
136.6ms [WARNING] Time: Read model file to memory. 21.49
95+
96+
0.0ms [WARNING] <W> This META does not have Alloc2 Support
97+
98+
0.0ms [WARNING] <W> This META does not have Alloc2 Support
99+
100+
0.0ms [WARNING] <W> This META does not have Alloc2 Support
101+
102+
0.0ms [WARNING] <W> This META does not have Alloc2 Support
103+
104+
0.0ms [WARNING] <W> This META does not have Alloc2 Support
105+
106+
0.0ms [WARNING] <W> This META does not have Alloc2 Support
107+
108+
0.0ms [WARNING] <W> This META does not have Alloc2 Support
109+
110+
205.5ms [WARNING] Time: contextCreateFromBinary. 68.83
111+
112+
205.6ms [WARNING] Time: UnmapViewOfFile. 0.00
113+
114+
207.5ms [WARNING] Time: model_initialize aotgan 207.45
115+
116+
415.4ms [WARNING] Time: model_inference aotgan 154.66
117+
118+
0.0ms [WARNING] <W> This META does not have Alloc2 Support
119+
120+
0.0ms [WARNING] <W> This META does not have Alloc2 Support
121+
122+
/usr/bin/xdg-open: 882: www-browser: not found
123+
/usr/bin/xdg-open: 882: links2: not found
124+
/usr/bin/xdg-open: 882: elinks: not found
125+
/usr/bin/xdg-open: 882: links: not found
126+
0.0ms [WARNING] <W> This META does not have Alloc2 Support
127+
128+
0.0ms [WARNING] <W> This META does not have Alloc2 Support
129+
130+
/usr/bin/xdg-open: 882: lynx: not found
131+
/usr/bin/xdg-open: 882: w3m: not found
132+
xdg-open: no method available for opening '/tmp/tmpjvwp3wzi.PNG'
133+
/usr/bin/xdg-open: 882: www-browser: not found
134+
/usr/bin/xdg-open: 882: links2: not found
135+
/usr/bin/xdg-open: 882: elinks: not found
136+
/usr/bin/xdg-open: 882: links: not found
137+
/prj/qct/webtech_scratch20/mlg_user_admin/qaisw_source_repo/rel/qairt-2.37.1/point_release/SNPE_SRC/avante-tools/prebuilt/dsp/hexagon-sdk-5.5.5/ipc/fastrpc/rpcmem/src/rpcmem_android.c:42:dummy call to rpcmem_deinit, rpcmem APIs will be used from libxdsprpc
138+
801.6ms [WARNING] Time: model_destroy aotgan 67.93
139+
140+
/usr/bin/xdg-open: 882: lynx: not found
141+
/usr/bin/xdg-open: 882: w3m: not found
142+
xdg-open: no method available for opening '/tmp/tmp0z61xxdt.PNG'
143+
```
144+
145+
{" "}
146+
147+
<div style={{ textAlign: "center" }}>
148+
<img
149+
src="/img/fogwise/airbox-q900/qairt-aotgan_output.webp"
150+
style={{ width: "65%" }}
151+
/>
152+
output image
153+
</div>

‎docs/common/ai/_beit_qai-app-builder.mdx‎

Lines changed: 126 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,126 @@
1+
此文档讲述使用 [QAI AppBuilder](../qai-appbuilder) Python API 利用 Qualcomm® Hexagon™ Processor (NPU) 推理 [Beit](https://aihub.qualcomm.com/models/beit)
2+
目标识别模型。
3+
4+
**示例支持设备**
5+
6+
| 设备 | SoC |
7+
| --------------------- | ------- |
8+
| Fogwise® AIRbox Q900 | QCS9075 |
9+
10+
## 安装 QAI AppBuilder
11+
12+
:::tip
13+
14+
1. 请根据 [**QAI AppBuilder 安装方法**](../qai-appbuilder) 安装 QAI AppBuilder。
15+
16+
2. 请根据 [**配置 QAIRT 环境变量**](../qai-appbuilder#配置-qairt-环境变量) 配置 QAIRT 环境变量。
17+
:::
18+
19+
## 运行示例
20+
21+
### 安装依赖
22+
23+
<NewCodeBlock tip="Device" type="device">
24+
25+
```bash
26+
pip3 install requests tqdm qai-hub py3_wget opencv-python torch torchvision
27+
```
28+
29+
</NewCodeBlock>
30+
31+
### 运行脚本
32+
33+
- 进入示例目录
34+
35+
<Tabs>
36+
37+
<TabItem value="QCS9075">
38+
39+
<NewCodeBlock tip="Device" type="device">
40+
41+
```bash
42+
cd ai-engine-direct-helper/samples/python
43+
```
44+
45+
</NewCodeBlock>
46+
47+
</TabItem>
48+
49+
</Tabs>
50+
51+
- 准备输入图片，这里以以下图片为输入示例
52+
53+
{" "}
54+
55+
<div style={{ textAlign: "center" }}>
56+
<img src="/img/dragon/q6a/test_image.webp" style={{ width: "65%" }} />
57+
input image
58+
</div>
59+
60+
- 执行推理
61+
62+
<NewCodeBlock tip="Device" type="device">
63+
64+
```bash
65+
python3 beit/beit.py
66+
```
67+
68+
</NewCodeBlock>
69+
70+
```bash
71+
$ python3 beit/beit.py
72+
0.0ms [WARNING] <W> Initializing HtpProvider
73+
74+
/prj/qct/webtech_scratch20/mlg_user_admin/qaisw_source_repo/rel/qairt-2.37.1/point_release/SNPE_SRC/avante-tools/prebuilt/dsp/hexagon-sdk-5.5.5/ipc/fastrpc/rpcmem/src/rpcmem_android.c:38:dummy call to rpcmem_init, rpcmem APIs will be used from libxdsprpc
75+
0.0ms [WARNING] <W> This META does not have Alloc2 Support
76+
77+
0.0ms [WARNING] <W> This META does not have Alloc2 Support
78+
79+
0.0ms [WARNING] <W> This META does not have Alloc2 Support
80+
81+
0.0ms [WARNING] <W> This META does not have Alloc2 Support
82+
83+
179.0ms [WARNING] Time: Read model file to memory. 60.05
84+
85+
0.0ms [WARNING] <W> This META does not have Alloc2 Support
86+
87+
0.0ms [WARNING] <W> This META does not have Alloc2 Support
88+
89+
0.0ms [WARNING] <W> This META does not have Alloc2 Support
90+
91+
0.0ms [WARNING] <W> This META does not have Alloc2 Support
92+
93+
0.0ms [WARNING] <W> This META does not have Alloc2 Support
94+
95+
0.0ms [WARNING] <W> This META does not have Alloc2 Support
96+
97+
0.0ms [WARNING] <W> This META does not have Alloc2 Support
98+
99+
284.7ms [WARNING] Time: contextCreateFromBinary. 105.62
100+
101+
284.8ms [WARNING] Time: UnmapViewOfFile. 0.00
102+
103+
288.1ms [WARNING] Time: model_initialize beit 288.01
104+
105+
365.3ms [WARNING] Time: model_inference beit 17.72
106+
107+
Top 5 predictions for image:
108+
109+
"Samoyed", 0.6322221756
110+
"Pomeranian", 0.1098636836
111+
"Keeshond", 0.0410530604
112+
"Japanese Chin", 0.0093044275
113+
"Chow Chow", 0.0058453484
114+
0.0ms [WARNING] <W> This META does not have Alloc2 Support
115+
116+
0.0ms [WARNING] <W> This META does not have Alloc2 Support
117+
118+
0.0ms [WARNING] <W> This META does not have Alloc2 Support
119+
120+
0.0ms [WARNING] <W> This META does not have Alloc2 Support
121+
122+
/prj/qct/webtech_scratch20/mlg_user_admin/qaisw_source_repo/rel/qairt-2.37.1/point_release/SNPE_SRC/avante-tools/prebuilt/dsp/hexagon-sdk-5.5.5/ipc/fastrpc/rpcmem/src/rpcmem_android.c:42:dummy call to rpcmem_deinit, rpcmem APIs will be used from libxdsprpc
123+
483.8ms [WARNING] Time: model_destroy beit 113.14
124+
```
125+
126+
打印结果可以看到 `Samoyed` 置信度最高，这与输入图片内容吻合。

0 commit comments

Comments
 (0)