-
Notifications
You must be signed in to change notification settings - Fork 1
/
install2.txt
259 lines (187 loc) · 11.3 KB
/
install2.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
nvidia-driver 450.51.05
CUDA Version: 11.0
cuDNN8.0.5
python 3.7
pytorch==1.7.1
torchvision==0.8.2
torchaudio==0.7.2
conda install pytorch==1.7.1 torchvision==0.8.2 torchaudio==0.7.2 cudatoolkit=11.0 -c pytorch
1. 解决Intel AX200 Wifi6无线网卡无法驱动的问题
到Intel官网下载驱动,20.04Kernel版本符合ax200的驱动要求
解压后拷贝到/lib/firmware
然后
sudo add-apt-repository ppa:canonical-hwe-team/backport-iwlwifi
sudo apt-get update
sudo apt-get install backport-iwlwifi-dkms
安装完毕重启,无线网卡应该可以使用了
Linux配置v2ray详细教程-Ubuntu为例
https://mahongfei.com/1776.html
Ubuntu 18.04 LTS操作系统的Conda安装及配置
https://blog.csdn.net/fhxy_xzw/article/details/98595878
【Linux】conda: command not found解决办法
https://blog.csdn.net/weixin_38705903/article/details/86533863
ubuntu18.04 安装 Google Chrome
https://blog.csdn.net/qq_39418067/article/details/89647002
选pytorch
https://pytorch.org/get-started/previous-versions/#conda-1
ORG14重装机
从win中卸载原来的ubuntu双系统：
ubuntu18.04装载
配翻墙
org14内核
terminator
google-Chrome
中文输入法
cudatoolkit
测试：查看环境变量是否配置成功：nvcc -V。成功会显示CUDA信息。
cuDNN
测试：
cp -r /usr/src/cudnn_samples_v8/ ~
cd ~/cudnn_samples_v8/mnistCUDNN/
make clean && make
./mnistCUDNN
if test passed then successed
pytorch （failed） && torchvision
测试：
import torch
x = torch.rand(5, 3)
print(x)
output:
tensor([[0.3380, 0.3845, 0.3217],
[0.8337, 0.9050, 0.2650],
[0.2979, 0.7141, 0.9069],
[0.1449, 0.1132, 0.1375],
[0.4675, 0.3947, 0.1426]])
验证显卡驱动和CUDA是否已被启用并可被Pytorch访问：
测试：
import torch # 可以import说明Pytorch安装成功
torch.cuda.is_available() #返回True则说明Pytorch可以在当前GPU上使用
tensorflow
tensorflow single 测试：
TensorFlow 2.0 test:
import tensorflow as tf
tf.compat.v1.disable_eager_execution() ＃保证sess.run()能够正常运行
hello = tf.constant('hello,tensorflow')
sess= tf.compat.v1.Session()＃版本2.0的函数
print(sess.run(hello))
tensorflow and cuda 测试：(failed )
/home/twilight/project/test tensorflow_test.py
##############
InvalidArchiveError('Error with archive /home/twilight/anaconda3/pkgs/pytorch-1.7.1-py3.7_cuda11.0.221_cudnn8.0.5_0.tar.bz2.
You probably need to delete and re-download or re-create this file. Message from libarchive was:\n\ntruncated bzip2 input (errno=-1, retcode=-30, archive_p=93831085324288)')
pytorch-1.7.1-py3.7_cuda11.0.221_cudnn8.0.5_0.tar.bz2.
(base) twilight@ORG: ~$ pip3 install tensorflow-gpu==2.4.0 -i https://pypi.douban.com/simple
ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
astroid 2.2.5 requires typed-ast>=1.3.0; implementation_name == "cpython", which is not installed.
(base) twilight@ORG: ~$ pip3 install typed-ast==2.4.0 -i https://pypi.douban.com/simple
Looking in indexes: https://pypi.douban.com/simple
ERROR: Could not find a version that satisfies the requirement typed-ast==2.4.0
ERROR: No matching distribution found for typed-ast==2.4.0
WARNING: You are using pip version 21.0.1; however, version 21.2.1 is available.
You should consider upgrading via the '/home/twilight/anaconda3/bin/
python3 -m pip install --user --upgrade pip
(base) twilight@ORG: ~/project/test$ python tensorflow_test.py
2021-07-29 22:10:56.760106: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcudart.so.11.0
2021-07-29 22:10:57.726623: I tensorflow/compiler/jit/xla_gpu_device.cc:99] Not creating XLA devices, tf_xla_enable_xla_devices not set
2021-07-29 22:10:57.727443: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcuda.so.1
2021-07-29 22:10:57.752628: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:941] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2021-07-29 22:10:57.752952: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1720] Found device 0 with properties:
pciBusID: 0000:01:00.0 name: GeForce RTX 2060 with Max-Q Design computeCapability: 7.5
coreClock: 1.185GHz coreCount: 30 deviceMemorySize: 5.79GiB deviceMemoryBandwidth: 245.91GiB/s
2021-07-29 22:10:57.752996: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcudart.so.11.0
2021-07-29 22:10:57.754789: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcublas.so.11
2021-07-29 22:10:57.754887: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcublasLt.so.11
2021-07-29 22:10:57.755736: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcufft.so.10
2021-07-29 22:10:57.756181: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcurand.so.10
2021-07-29 22:10:57.758173: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcusolver.so.10
2021-07-29 22:10:57.758824: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcusparse.so.11
2021-07-29 22:10:57.759158: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcudnn.so.8
2021-07-29 22:10:57.759373: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:941] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2021-07-29 22:10:57.759825: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:941] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2021-07-29 22:10:57.760101: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1862] Adding visible gpu devices: 0
2021-07-29 22:10:57.760164: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcudart.so.11.0
Traceback (most recent call last):
File "tensorflow_test.py", line 4, in <module>
device_name = tf.test.gpu_device_name()
File "/home/twilight/anaconda3/lib/python3.7/site-packages/tensorflow/python/framework/test_util.py", line 132, in gpu_device_name
for x in device_lib.list_local_devices():
File "/home/twilight/anaconda3/lib/python3.7/site-packages/tensorflow/python/client/device_lib.py", line 43, in list_local_devices
_convert(s) for s in _pywrap_device_lib.list_devices(serialized_config)
RuntimeError: CUDA runtime implicit initialization on GPU:0 failed. Status: out of memory
RuntimeError: CUDA runtime implicit initialization on GPU:0 failed. Status: out of memory
https://blog.csdn.net/ayiya_Oese/article/details/114022167
centerpoint :
git 加速
APEX
spconv
det3D环境部署
再bash setup.sh
spconv
python setup.py bdist_wheel
CMake install
https://www.cnblogs.com/yanqingyang/p/12731855.html
-- Found cuDNN: v? (include: /usr/local/cuda/include, library: /usr/local/cuda/lib64/libcudnn.so)
CMake Error at /home/twilight/.conda/envs/cp2/lib/python3.7/site-packages/torch/share/cmake/Caffe2/public/cuda.cmake:174 (message):
PyTorch requires cuDNN 7 and above.
Call Stack (most recent call first):
/home/twilight/.conda/envs/cp2/lib/python3.7/site-packages/torch/share/cmake/Caffe2/Caffe2Config.cmake:88 (include)
/home/twilight/.conda/envs/cp2/lib/python3.7/site-packages/torch/share/cmake/Torch/TorchConfig.cmake:40 (find_package)
CMakeLists.txt:23 (find_package)
https://blog.csdn.net/baidu_38505667/article/details/114011051
确保你已经把cudnn中的cudnn_version.h复制到了cuda目录（新版本cudnn的版本信息包含在cudnn_version.h而不是cudnn.h，安装cudnn时把所有cudnn开头的都复制过去）
修改libtorch/share/cmake/Caffe2/public/cuda.cmake at line 148
替换 : file(READ ${CUDNN_INCLUDE_PATH}/cudnn.h CUDNN_HEADER_CONTENTS)
为 : file(READ ${CUDNN_INCLUDE_PATH}/cudnn_version.h CUDNN_HEADER_CONTENTS)
tar -xvf cmake-3.18.0.tar.gz
./bootstrap
make
make install
訓練 推理 測試？
先用pip list看到我所安裝的包
遇到的setup.sh的問題，直接google 在git上面找，一般遇到的問題都能在git 社區上找到
看懂他安裝的庫就可以
setup.sh就是安裝了ded3d的庫，只用在這兩個就可以
英文叫什麼會叫
pp模型 實時性好，準確率低（什麼是效果），大概2 3 鎮出效果，肉眼可見的低，檢車人估計不大行
v效果好，實時性低
一般點雲都是100ms一幀
pp模型一般用第三個，第二個第三個一樣
cp只是做dec，tracking是離線的，就是把所有的dec出了結果以後再生成tracking
需要自己寫一部分代碼
在測試的時候，先測試dec，dec沒問題了，tracking就可以了（it has）
在single_infernece中，直接寫了ros的藉口，可以直接在ros上用，他輸出電暈，和bonging box
如果做tracking的話，需要在rviz中添加對mark的追蹤，
single_infernece這個文件寫的比較規範，以後也可以用，x y點顯示，這個代碼要看
測試數據的時候是用train數據做的，因爲test數據集是沒有標註的，如果測試的話需要上傳到他的服務器上進行測試
mmdet中的tools有一些工具可以看看，在tools裡面
如果可視化tracking的結果，可以用open3D，要看他的，藥接口，學一陣子
nuScenes detection task排行榜,tasks：
第一名是Centerpoint-Fusion，它做的是融合，自然效果好
CenterPoint v2是centerpoint
https://www.nuscenes.org/object-detection?externalData=all&mapData=all&modalities=Any
模型文件：https://drive.google.com/drive/folders/1K_wHrBo6yRSG7H7UUjKI4rPnyEA8HvOp
用mini數據集，改一下執行test的名稱
baseline centerpoint的baseline
mmdetection3d/tools/有很多可以學的
mmdetection3d/tools/misc/ 有可視化
mmdetection3d/tools/misc/visualize_results.py
mmdetection3d/mmdet3d/core/visualizer/open3d_vis.py
https://github.com/open-mmlab/mmdetection3d
‘AT_CHECK’ was not declared in this scope AT_CHECK(weight.ndimension() == 4'
https://github.com/open-mmlab/mmediting/issues/84
https://github.com/poodarchu/Det3D
open3d可視化3d檢測結果
3D可视化神器之Open3D
https://zhuanlan.zhihu.com/p/57215172
KITTI 目标检测结果的点云可视化
https://www.pianshen.com/article/73621695434/
OPEN3D 可视化使用
https://blog.csdn.net/weixin_43847196/article/details/106956427
为什么不翻墙就不能上网
为什么网卡驱动安装了但是不能连接wifi
显示屏输出
python ./tools/dist_test.py /home/twilight/project/cp2/CenterPoint/configs/nusc/pp
--work_dir /home/twilight/project/cp2/CenterPoint/configs/nusc/pp/nusc_centerpoint_pp_02voxel_two_pfn_10sweep_demo.py
--checkpoint /home/twilight/project/cp2/CenterPoint/configs/nusc/pp/nusc_centerpoint_pp_02voxel_two_pfn_10sweep_demo.py /latest.pth --speed_test
python ./tools/dist_test.py CONFIG_PATH --work_dir work_dirs/CONFIG_NAME --checkpoint work_dirs/CONFIG_NAME/latest.pth --speed_test
Python 3.7.10