你好
1.我想在VIM3上使用gpio与其他MCU进行交互，更新到了最新的系统，如图所示。但是没有bash 无gpio的指令，请问是怎么回事？
2.我参考了HowToTestGPIOIRQ，但是我没法知道PIN Number是多少？我又参考了HowToAccessGpio，这里说的在UBUNTU系统下的gpio list可以怎么知道，我使用命令cat出来的pin范围是[410-496],[0-85]，和这里不对应？
简而言之，我手上的A311D开发板，上面的gpio对应的Number是多少？比如板子丝印1-40pin，分别对应的号应该是多少？
试试手动安装下wiringpi：
$ wget https://dl.khadas.com/repos/vim3/pool/main/w/wiringpi/wiringpi_2.46-6_arm64.deb
$ sudo dpkg -i wiringpi_2.46-6_arm64.deb
然后查看：
$ khadas@Khadas:~$ gpio readall
+------+-----+----------+------+---+----+---- Model Khadas VIM3 --+----+---+------+----------+-----+------+
| GPIO | wPi | Name | Mode | V | DS | PU/PD | Physical | PU/PD | DS | V | Mode | Name | wPi | GPIO |
+------+-----+----------+------+---+----+-------+----++----+-------+----+---+------+----------+-----+------+
| | | 5V | | | | | 1 || 21 | | | | | GND | | |
| | | 5V | | | | | 2 || 22 | P/U | | 1 | ALT1 | PIN.A15 | 6 | 475 |
| | | USB_DM | | | | | 3 || 23 | P/U | | 1 | ALT1 | PIN.A14 | 7 | 474 |
| | | USB_DP | | | | | 4 || 24 | | | | | GND | | |
| | | GND | | | | | 5 || 25 | P/U | | 1 | ALT0 | PIN.AO2 | 8 | 498 |
| | | MCU3V3 | | | | | 6 || 26 | P/U | | 1 | ALT0 | PIN.AO3 | 9 | 499 |
| | | MCUNRST | | | | | 7 || 27 | | | | | 3V3 | | |
| | | MCUSWIM | | | | | 8 || 28 | | | | | GND | | |
| | | GND | | | | | 9 || 29 | P/D | | 0 | ALT0 | PIN.A1 | 10 | 461 |
| | | ADC0 | | | | | 10 || 30 | P/D | | 0 | ALT0 | PIN.A0 | 11 | 460 |
| | | 1V8 | | | | | 11 || 31 | P/D | | 0 | ALT0 | PIN.A3 | 12 | 463 |
| | | ADC1 | | | | | 12 || 32 | P/D | | 0 | ALT0 | PIN.A2 | 13 | 462 |
| 506 | 1 | PIN.H5 | ALT3 | 0 | | P/U | 13 || 33 | P/D | | 0 | ALT1 | PIN.A4 | 14 | 464 |
| | | GND3 | | | | | 14 || 34 | | | | | GND | | |
| 433 | 2 | PIN.H6 | ALT1 | 0 | | P/D | 15 || 35 | P/D | | 0 | ALT3 | PWM-F | 15 | 432 |
| 434 | 3 | PIN.H7 | ALT1 | 1 | | P/D | 16 || 36 | | | | | RTC | | |
| | | GND | | | | | 17 || 37 | P/D | | 0 | IN | PIN.H4 | 16 | 431 |
| 497 | 4 | PIN.AO1 | ALT0 | 1 | | P/U | 18 || 38 | | | | | MCU-FA1 | | |
| 496 | 5 | PIN.AO0 | ALT0 | 0 | | P/U | 19 || 39 | P/D | | 0 | IN | PIN.Z15 | 17 | 426 |
| | | 3V3 | | | | | 20 || 40 | | | | | GND | | |
+------+-----+----------+------+---+----+-------+----++----+-------+----+---+------+----------+-----+------+
@Frank好的，感谢！是不是我因为我执行了sudo apt-get update && sudo apt-get upgrade才会出现这个需要手动安装的情况？ 因为我之前装了202005月份那个系统，用官方编好的inceptionv3 npu demo可以直接执行，现在不行了，得自己再编译一遍才可以。
@Frank 好的。谢谢。
如上图所示，我根据’01’那个pdf，看到了支持keras和pytorch框架，我在’02’里找到其他框架支持的Op，想请问keras和pytorch支持的算子可以在哪里查看到?
@Frank 你好，我使用tf2自己训练了一下mobilenetv1的网络，按照step1 0_import_model.sh 转成了***.json，***.data ，然后我执行step4，仿真在pc上的非量化结果，然后报了如下的错误：
好像说是命名有问题，可是我换了其他还是一样的问题？您能帮我看一下吗？这里好像发不了模型文件
***.json
{
"MetaData": {
"Name": "main",
"AcuityVersion": "2.0",
"Platform": "tensorflow",
"Org_Platform": "tensorflow"
},
"Layers": {
"input_0": {
"name": "input",
"op": "input",
"parameters": {
"size": "",
"channels": 1,
"shape": [
1,
224,
224,
1
]
},
"inputs": [],
"outputs": [
"out0"
]
},
"functional_1/conv1_relu/Relu6;functional_1/conv1_bn/FusedBatchNormV3;functional_1/conv_dw_1_bn/FusedBatchNormV3;functional_1/conv_dw_1/depthwise;functional_1/conv1/Conv2D_2": {
"name": "functional_1/conv1_relu/Relu6;functional_1/conv1_bn/FusedBatchNormV3;functional_1/conv_dw_1_bn/FusedBatchNormV3;functional_1/conv_dw_1/depthwise;functional_1/conv1/Conv2D",
"op": "convolution",
"parameters": {
"weights": 32,
"padding": "VALID",
"bias": true,
"group_number": 1,
"regularize": false,
"ksize_h": 3,
"ksize_w": 3,
"stride_h": 2,
"stride_w": 2,
"pad_h": 0,
"pad_w": 0,
"dilation": 1,
"pad_method": "padding_const",
"pad": [
0,
1,
0,
1
]
},
"inputs": [
"@input_0:out0"
],
"outputs": [
"out0"
]
},
"functional_1/conv1_relu/Relu6;functional_1/conv1_bn/FusedBatchNormV3;functional_1/conv_dw_1_bn/FusedBatchNormV3;functional_1/conv_dw_1/depthwise;functional_1/conv1/Conv2D_3": {
"name": "functional_1/conv1_relu/Relu6;functional_1/conv1_bn/FusedBatchNormV3;functional_1/conv_dw_1_bn/FusedBatchNormV3;functional_1/conv_dw_1/depthwise;functional_1/conv1/Conv2D",
"op": "relun",
"parameters": {
"relu_clamp_top": 6,
"relu_clamp_bottom": "0"
},
"inputs": [
"@functional_1/conv1_relu/Relu6;functional_1/conv1_bn/FusedBatchNormV3;functional_1/conv_dw_1_bn/FusedBatchNormV3;functional_1/conv_dw_1/depthwise;functional_1/conv1/Conv2D_2:out0"
],
"outputs": [
"out0"
]
},
"functional_1/conv_dw_1_relu/Relu6;functional_1/conv_dw_1_bn/FusedBatchNormV3;functional_1/conv_dw_1/depthwise_4": {
"name": "functional_1/conv_dw_1_relu/Relu6;functional_1/conv_dw_1_bn/FusedBatchNormV3;functional_1/conv_dw_1/depthwise",
"op": "convolution",
"parameters": {
"weights": 32,
"padding": "SAME",
"bias": true,
"group_number": 32,
"regularize": false,
"ksize_h": 3,
"ksize_w": 3,
"stride_h": 1,
"stride_w": 1,
"pad_h": 1,
"pad_w": 1,
"dilation": 1,
"pad_method": "padding_const",
"pad": [
1,
1,
1,
1
]
},
"inputs": [
"@functional_1/conv1_relu/Relu6;functional_1/conv1_bn/FusedBatchNormV3;functional_1/conv_dw_1_bn/FusedBatchNormV3;functional_1/conv_dw_1/depthwise;functional_1/conv1/Conv2D_3:out0"
],
"outputs": [
"out0"
]
},
"functional_1/conv_dw_1_relu/Relu6;functional_1/conv_dw_1_bn/FusedBatchNormV3;functional_1/conv_dw_1/depthwise_5": {
"name": "functional_1/conv_dw_1_relu/Relu6;functional_1/conv_dw_1_bn/FusedBatchNormV3;functional_1/conv_dw_1/depthwise",
"op": "relun",
"parameters": {
"relu_clamp_top": 6,
"relu_clamp_bottom": "0"
},
"inputs": [
"@functional_1/conv_dw_1_relu/Relu6;functional_1/conv_dw_1_bn/FusedBatchNormV3;functional_1/conv_dw_1/depthwise_4:out0"
],
"outputs": [
"out0"
]
},
…
@librazxc 这边建议你用tf1.4的或者tf2.0的,工具只支持2.0的版本,如果你的工具版本超过了这个版本,可能会导致出问题的.一般这种命名问题可能是版本更新以后,命名方式改变了,如果要等我这边验证的话,需要等待一段时间.要排进计划里面