• paddlepaddle2.6,paddleorc2.8,cuda12,cudnn,nccl,python10环境

    1.安装英伟达显卡驱动
    首先需要到NAVIDIA官网去查自己的电脑是不是支持GPU运算。
    网址是：CUDA GPUs | NVIDIA Developer。打开后的界面大致如下，只要里边有对应的型号就可以用GPU运算，并且每一款设备都列出来相关的计算能力（Compute Capability）。

    系统层面查看当前安装的显卡型号：

    1. (py10_paddleocr2.8_gpu_wubo) root@jettech-WS-C621E-SAGE-Series:~# lspci | grep nvida
    2. (py10_paddleocr2.8_gpu_wubo) root@jettech-WS-C621E-SAGE-Series:~# lspci | grep VGA
    3. 3b:00.0 VGA compatible controller: NVIDIA Corporation GV104 [GeForce GTX 1180] (rev a1)
    4. 5e:00.0 VGA compatible controller: NVIDIA Corporation GV104 [GeForce GTX 1180] (rev a1)
    5. 86:00.0 VGA compatible controller: NVIDIA Corporation GV104 [GeForce GTX 1180] (rev a1)
    6. af:00.0 VGA compatible controller: NVIDIA Corporation GV104 [GeForce GTX 1180] (rev a1)

    如果是ubuntu系统：明确了显卡性能后，接下来就开始在ubuntu系统安装对应的显卡驱动。

    首先，检测NVIDIA图形卡和推荐的驱动程序的模型，在终端输入：

    1. (py10_paddleocr2.8_gpu_wubo) root@jettech-WS-C621E-SAGE-Series:~# ubuntu-drivers devices
    2. WARNING:root:_pkg_get_support nvidia-driver-530: package has invalid Support PBheader, cannot determine support level
    3. WARNING:root:_pkg_get_support nvidia-driver-515-server: package has invalid Support PBheader, cannot determine support level
    4. WARNING:root:_pkg_get_support nvidia-driver-525-server: package has invalid Support PBheader, cannot determine support level
    5. == /sys/devices/pci0000:3a/0000:3a:00.0/0000:3b:00.0 ==
    6. modalias : pci:v000010DEd00001E87sv00001458sd000037A8bc03sc00i00
    7. vendor   : NVIDIA Corporation
    8. driver   : nvidia-driver-530 - distro non-free recommended
    9. driver   : nvidia-driver-470-server - distro non-free
    10. driver   : nvidia-driver-440 - third-party non-free
    11. driver   : nvidia-driver-515 - third-party non-free
    12. driver   : nvidia-driver-450-server - distro non-free
    13. driver   : nvidia-driver-515-server - distro non-free
    14. driver   : nvidia-driver-418-server - distro non-free
    15. driver   : nvidia-driver-418 - third-party non-free
    16. driver   : nvidia-driver-460 - third-party non-free
    17. driver   : nvidia-driver-450 - third-party non-free
    18. driver   : nvidia-driver-470 - third-party non-free
    19. driver   : nvidia-driver-455 - third-party non-free
    20. driver   : nvidia-driver-495 - third-party non-free
    21. driver   : nvidia-driver-525 - third-party non-free
    22. driver   : nvidia-driver-465 - third-party non-free
    23. driver   : nvidia-driver-525-server - distro non-free
    24. driver   : nvidia-driver-410 - third-party non-free
    25. driver   : nvidia-driver-520 - third-party non-free
    26. driver   : nvidia-driver-510 - third-party non-free
    27. driver   : xserver-xorg-video-nouveau - distro free builtin

    具体可以使用下面的命令安装：

    (py10_paddleocr2.8_gpu_wubo) root@jettech-WS-C621E-SAGE-Series:~#  ubuntu-drivers autoinstall

    或者去官网下载驱动再手动安装的方式，命令官网上有。

    下载 NVIDIA 官方驱动 | NVIDIA

    NVIDIA GeForce 驱动程序 - N 卡驱动 | NVIDIA

    安装完成后重启系统，然后在终端中输入命令检测是否安装成功：

    1. (py10_paddleocr2.8_gpu_wubo) root@jettech-WS-C621E-SAGE-Series:~# nvidia-smi
    2. (py10_paddleocr2.8_gpu_wubo) root@jettech-WS-C621E-SAGE-Series:~# nvidia-smi
    3. Fri Jul 12 15:43:58 2024       
    4. +---------------------------------------------------------------------------------------+
    5. | NVIDIA-SMI 530.41.03              Driver Version: 530.41.03    CUDA Version: 12.1     |
    6. |-----------------------------------------+----------------------+----------------------+
    7. | GPU  Name                  Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
    8. | Fan  Temp  Perf            Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
    9. |                                         |                      |               MIG M. |
    10. |=========================================+======================+======================|
    11. |   0  NVIDIA GeForce RTX 2080         Off| 00000000:3B:00.0 Off |                  N/A |
    12. | 32%   41C    P8                3W / 225W|      8MiB /  8192MiB |      0%      Default |
    13. |                                         |                      |                  N/A |
    14. +-----------------------------------------+----------------------+----------------------+
    15. |   1  NVIDIA GeForce RTX 2080         Off| 00000000:5E:00.0 Off |                  N/A |
    16. | 27%   41C    P8                4W / 225W|      8MiB /  8192MiB |      0%      Default |
    17. |                                         |                      |                  N/A |
    18. +-----------------------------------------+----------------------+----------------------+
    19. |   2  NVIDIA GeForce RTX 2080         Off| 00000000:86:00.0 Off |                  N/A |
    20. | 27%   36C    P8                1W / 225W|      8MiB /  8192MiB |      0%      Default |
    21. |                                         |                      |                  N/A |
    22. +-----------------------------------------+----------------------+----------------------+
    23. |   3  NVIDIA GeForce RTX 2080         Off| 00000000:AF:00.0 Off |                  N/A |
    24. | 31%   43C    P8                9W / 225W|     80MiB /  8192MiB |      0%      Default |
    25. |                                         |                      |                  N/A |
    26. +-----------------------------------------+----------------------+----------------------+
    27.                                                                                          
    28. +---------------------------------------------------------------------------------------+
    29. | Processes:                                                                            |
    30. |  GPU   GI   CI        PID   Type   Process name                            GPU Memory |
    31. |        ID   ID                                                             Usage      |
    32. |=======================================================================================|
    33. |    0   N/A  N/A     52177      G   /usr/lib/xorg/Xorg                            4MiB |
    34. |    1   N/A  N/A     52177      G   /usr/lib/xorg/Xorg                            4MiB |
    35. |    2   N/A  N/A     52177      G   /usr/lib/xorg/Xorg                            4MiB |
    36. |    3   N/A  N/A     52177      G   /usr/lib/xorg/Xorg                           28MiB |
    37. |    3   N/A  N/A     52282      G   /usr/bin/gnome-shell                         46MiB |
    38. +---------------------------------------------------------------------------------------+
    39. (py10_paddleocr2.8_gpu_wubo) root@jettech-WS-C621E-SAGE-Series:~#

    上图显示cuda最高支持12.1版本

    驱动版本Driver Version: 530.41.03

    显卡型号：NVIDIA GeForce RTX 2080

    显卡num：共计4个 每个显存大小8G

    2.安装CUDA
    首先要知道硬件支持的CUDA版本：
    在上图右上角我们看到“CUDA Version：12.1”，这个表明对于这款显卡，我们后面要装的CUDA版本最高不能超过12.1。

    其次要明确CUDA版本需求：

    本文最终的目的是装好深度学习环境，这里指的是最终能够正常的使用pytorch[facebook公司]和paddlepaddle【百度公司】或TensorFlow【google公司】。这三款是当前使用比较多的深度学习框架，pytorch[facebook]侧重于科研和模型验证，paddlepaddle更适合工业级深度学习开发部署（当然也可以使用tensorflow）。

    为了能够使用他们，我们接下来需要按照顺序安装CUDA、cuDNN、nccl、paddlepaddle、pytorch【省略】安装paddleocr。

    在正式安装前我们首先要来确定当前的版本一致性，否则装到后面就会发现各种版本问题了。

    接下来我们先看paddlepaddle和pytorch官网目前稳定版所支持的cuda。

    paddlepaddle目前官网安装界面如下图所示：

    pytorch官网安装界面：

    尽量选择两个框架都支持的了，并且本机驱动也支持的CUDA版本。

    接下来开始安装：

    首先在英伟达官网下载cuda12进行安装即可。

    照runfile（local）安装的方式简单，只需要在终端输入图中下方的两条NVIDIA推荐的命令就好了。

    1. 2中方式
    2. 1）交互
    3. ./cuda_xxxxxxx_linux.run
    4. 2）静默
    5. ./cuda_xxxxxxx_linux.run --silent --toolkit --samples
    1. (py10_paddleocr2.8_gpu_wubo) root@jettech-WS-C621E-SAGE-Series:~#  vim ~/.bashrc
    2.  
    3. export PATH=/usr/local/cuda-12.0/bin${PATH:+:${PATH}}
    4. export LD_LIBRARY_PATH=/usr/local/cuda-12.0/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}

    最后，更新环境变量配置:

    source ~/.bashrc

    至此cuda安装完成，输入nvcc -V命令查看cuda信息。

    1. (py10_paddleocr2.8_gpu_wubo) root@jettech-WS-C621E-SAGE-Series:~# nvcc -V
    2. nvcc: NVIDIA (R) Cuda compiler driver
    3. Copyright (c) 2005-2022 NVIDIA Corporation
    4. Built on Mon_Oct_24_19:12:58_PDT_2022
    5. Cuda compilation tools, release 12.0, V12.0.76
    6. Build cuda_12.0.r12.0/compiler.31968024_0

    如果想要卸载CUDA（例如重新安装了驱动等情况），需要使用下面的命令：

    1. cd /usr/local/cuda-xx.x/bin/
    2. sudo ./cuda-uninstaller
    3. sudo rm -rf /usr/local/cuda-xx.x

    3.安装CUDNN

    cuDNN（CUDA Deep Neural Network library） 是由NVIDIA开发的一个深度学习GPU加速库。

    目的和功能： cuDNN旨在提供高效、标准化的原语（基本操作）来加速深度学习框架（例如TensorFlow、PyTorch）在NVIDIA GPU上的运算。

    专门为深度学习设计：cuDNN提供了为深度学习任务高度优化的函数，如：

    • 卷积操作
    • 池化操作
    • 激活函数
    • 归一化等

    安装CUDNN的过程相对比较简单。上官网进行下载。

    选择对应的CUDA版本，单击后选择cuDNN Library for Linux（x86_64）下载安装包。

    然后打开终端输入类似下面的命令进行解压并拷贝安装：

    1. cp -Pcudnn*/include/cudnn*.h cuda/include/
    2. cp -P cudnn*/lib/libcudnn*  cuda/lib64/
    3. chmod a+r cuda/include/cudnn*.h cuda/lib64/libcudnn*

    其实，cuDNN的安装本质上就是复制一堆的文件到CUDA中去。

    我们可以使用如下的命令查看cuDNN的信息：

    CUDN + cuDNN安装完成，我们可以监控一下gpu状态：

    watch -n 1 nvidia-smi

    4.安装NCCL

    由于深度学习分布式训练需要nccl支持，可以调用多张显卡计算，因此本小节来安装nccl。

    首先从官网下载对应版本的nccl.

    1. (py10_paddleocr2.8_gpu_wubo) root@jettech-WS-C621E-SAGE-Series:/usr/local# tar -xf nccl_2.19.3-1+cuda12.0_x86_64.txz  
    2. (py10_paddleocr2.8_gpu_wubo) root@jettech-WS-C621E-SAGE-Series:/usr/local# ln -sf nccl_2.19.3-1+cuda12.0_x86_64 nccl 
    3. (py10_paddleocr2.8_gpu_wubo) root@jettech-WS-C621E-SAGE-Series:/usr/local# cd include/^C
    4. (py10_paddleocr2.8_gpu_wubo) root@jettech-WS-C621E-SAGE-Series:/usr/local# cat /etc/ld.so.conf.d/nccl_2.19.3-1+cuda12.0.conf 
    5. /usr/local/nccl/lib
    6. (py10_paddleocr2.8_gpu_wubo) root@jettech-WS-C621E-SAGE-Series:/usr/local/include# ln -sf ../nccl/include nccl

    没安装之前报错： 

    安装之后：

    1. >>> import paddle
    2. >>> paddle.utils.run_check()
    3. Running verify PaddlePaddle program ... 
    4. I0712 17:30:32.906308 16653 program_interpreter.cc:212] New Executor is Running.
    5. W0712 17:30:32.906838 16653 gpu_resources.cc:119] Please NOTE: device: 0, GPU Compute Capability: 7.5, Driver API Version: 12.1, Runtime API Version: 12.0
    6. W0712 17:30:32.940363 16653 gpu_resources.cc:164] device: 0, cuDNN Version: 8.0.
    7. I0712 17:30:35.770787 16653 interpreter_util.cc:624] Standalone Executor is Used.
    8. PaddlePaddle works well on 1 GPU.
    9. ======================= Modified FLAGS detected =======================
    10. FLAGS(name='FLAGS_selected_gpus', current_value='2', default_value='')
    11. =======================================================================
    12. I0712 17:30:38.527948 17096 tcp_utils.cc:107] Retry to connect to 127.0.0.1:40265 while the server is not yet listening.
    13. ======================= Modified FLAGS detected =======================
    14. FLAGS(name='FLAGS_selected_gpus', current_value='3', default_value='')
    15. =======================================================================
    16. I0712 17:30:38.738694 17097 tcp_utils.cc:107] Retry to connect to 127.0.0.1:40265 while the server is not yet listening.
    17. ======================= Modified FLAGS detected =======================
    18. FLAGS(name='FLAGS_selected_gpus', current_value='1', default_value='')
    19. =======================================================================
    20. I0712 17:30:38.817551 17095 tcp_utils.cc:107] Retry to connect to 127.0.0.1:40265 while the server is not yet listening.
    21. ======================= Modified FLAGS detected =======================
    22. FLAGS(name='FLAGS_selected_gpus', current_value='0', default_value='')
    23. =======================================================================
    24. I0712 17:30:39.014600 17094 tcp_utils.cc:181] The server starts to listen on IP_ANY:40265
    25. I0712 17:30:39.014768 17094 tcp_utils.cc:130] Successfully connected to 127.0.0.1:40265
    26. I0712 17:30:41.528342 17096 tcp_utils.cc:130] Successfully connected to 127.0.0.1:40265
    27. I0712 17:30:41.528888 17096 process_group_nccl.cc:129] ProcessGroupNCCL pg_timeout_ 1800000
    28. I0712 17:30:41.739022 17097 tcp_utils.cc:130] Successfully connected to 127.0.0.1:40265
    29. I0712 17:30:41.776871 17097 process_group_nccl.cc:129] ProcessGroupNCCL pg_timeout_ 1800000
    30. I0712 17:30:41.817867 17095 tcp_utils.cc:130] Successfully connected to 127.0.0.1:40265
    31. I0712 17:30:41.840788 17095 process_group_nccl.cc:129] ProcessGroupNCCL pg_timeout_ 1800000
    32. I0712 17:30:41.851110 17094 process_group_nccl.cc:129] ProcessGroupNCCL pg_timeout_ 1800000
    33. W0712 17:30:43.391786 17096 gpu_resources.cc:119] Please NOTE: device: 2, GPU Compute Capability: 7.5, Driver API Version: 12.1, Runtime API Version: 12.0
    34. W0712 17:30:43.394407 17096 gpu_resources.cc:164] device: 2, cuDNN Version: 8.0.
    35. W0712 17:30:43.564615 17097 gpu_resources.cc:119] Please NOTE: device: 3, GPU Compute Capability: 7.5, Driver API Version: 12.1, Runtime API Version: 12.0
    36. W0712 17:30:43.566882 17097 gpu_resources.cc:164] device: 3, cuDNN Version: 8.0.
    37. W0712 17:30:43.627422 17095 gpu_resources.cc:119] Please NOTE: device: 1, GPU Compute Capability: 7.5, Driver API Version: 12.1, Runtime API Version: 12.0
    38. W0712 17:30:43.629004 17095 gpu_resources.cc:164] device: 1, cuDNN Version: 8.0.
    39. W0712 17:30:43.656805 17094 gpu_resources.cc:119] Please NOTE: device: 0, GPU Compute Capability: 7.5, Driver API Version: 12.1, Runtime API Version: 12.0
    40. W0712 17:30:43.659112 17094 gpu_resources.cc:164] device: 0, cuDNN Version: 8.0.
    41. I0712 17:30:46.433609 17096 process_group_nccl.cc:132] ProcessGroupNCCL destruct 
    42. I0712 17:30:46.433516 17095 process_group_nccl.cc:132] ProcessGroupNCCL destruct 
    43. I0712 17:30:46.435761 17097 process_group_nccl.cc:132] ProcessGroupNCCL destruct 
    44. I0712 17:30:46.437583 17094 process_group_nccl.cc:132] ProcessGroupNCCL destruct 
    45. I0712 17:30:46.843884 17168 tcp_store.cc:289] receive shutdown event and so quit from MasterDaemon run loop
    46. PaddlePaddle works well on 4 GPUs.
    47. PaddlePaddle is installed successfully! Let's start deep learning with PaddlePaddle now.

     

    验证NCCL

    1.  https://github.com/NVIDIA/nccl-tests
    2. (py10_paddlepaddle2.6_paddleocr2.8_cuda12_gpu_wubo) root@jettech-WS-C621E-SAGE-Series:/data/wubo/paddleocr/env/nccl# ls
    3. nccl-tests-2.13.9  nccl-tests-2.13.9.tar.gz
    4. (py10_paddlepaddle2.6_paddleocr2.8_cuda12_gpu_wubo) root@jettech-WS-C621E-SAGE-Series:/data/wubo/paddleocr/env/nccl# cd nccl-tests-2.13.9/
    5. (py10_paddlepaddle2.6_paddleocr2.8_cuda12_gpu_wubo) root@jettech-WS-C621E-SAGE-Series:/data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9# ls^C
    6. (py10_paddlepaddle2.6_paddleocr2.8_cuda12_gpu_wubo) root@jettech-WS-C621E-SAGE-Series:/data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9# 
    7. (py10_paddlepaddle2.6_paddleocr2.8_cuda12_gpu_wubo) root@jettech-WS-C621E-SAGE-Series:/data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9# ls
    8. doc  LICENSE.txt  Makefile  README.md  src  verifiable
    9. (py10_paddlepaddle2.6_paddleocr2.8_cuda12_gpu_wubo) root@jettech-WS-C621E-SAGE-Series:/data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9# make
    10. make -C src build BUILDDIR=/data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build
    11. make[1]: 进入目录“/data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/src”
    12. Compiling  timer.cc                            > /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/timer.o
    13. Compiling /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/verifiable/verifiable.o
    14. Compiling  all_reduce.cu                       > /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/all_reduce.o
    15. Compiling  common.cu                           > /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/common.o
    16. Linking  /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/all_reduce.o > /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/all_reduce_perf
    17. Compiling  all_gather.cu                       > /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/all_gather.o
    18. Linking  /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/all_gather.o > /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/all_gather_perf
    19. Compiling  broadcast.cu                        > /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/broadcast.o
    20. Linking  /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/broadcast.o > /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/broadcast_perf
    21. Compiling  reduce_scatter.cu                   > /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/reduce_scatter.o
    22. Linking  /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/reduce_scatter.o > /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/reduce_scatter_perf
    23. Compiling  reduce.cu                           > /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/reduce.o
    24. Linking  /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/reduce.o > /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/reduce_perf
    25. Compiling  alltoall.cu                         > /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/alltoall.o
    26. Linking  /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/alltoall.o > /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/alltoall_perf
    27. Compiling  scatter.cu                          > /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/scatter.o
    28. Linking  /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/scatter.o > /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/scatter_perf
    29. Compiling  gather.cu                           > /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/gather.o
    30. Linking  /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/gather.o > /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/gather_perf
    31. Compiling  sendrecv.cu                         > /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/sendrecv.o
    32. Linking  /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/sendrecv.o > /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/sendrecv_perf
    33. Compiling  hypercube.cu                        > /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/hypercube.o
    34. Linking  /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/hypercube.o > /data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/build/hypercube_perf
    35. make[1]: 离开目录“/data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9/src”
    36. (py10_paddlepaddle2.6_paddleocr2.8_cuda12_gpu_wubo) root@jettech-WS-C621E-SAGE-Series:/data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9# ./build/all_reduce_perf -b 8 -e 128M -f 2 -g 8
    37. # nThread 1 nGpus 8 minBytes 8 maxBytes 134217728 step: 2(factor) warmup iters: 5 iters: 20 agg iters: 1 validation: 1 graph: 0
    38. #
    39. # Using devices
    40. jettech-WS-C621E-SAGE-Series: Test CUDA failure common.cu:894 'invalid device ordinal'
    41.  .. jettech-WS-C621E-SAGE-Series pid 24945: Test failure common.cu:844
    42. (py10_paddlepaddle2.6_paddleocr2.8_cuda12_gpu_wubo) root@jettech-WS-C621E-SAGE-Series:/data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9# ls build/all_reduce_perf ^C
    43. (py10_paddlepaddle2.6_paddleocr2.8_cuda12_gpu_wubo) root@jettech-WS-C621E-SAGE-Series:/data/wubo/paddleocr/env/nccl/nccl-tests-2.13.9# ./build/all_reduce_perf -b 8 -e 256M -f 2 -g4
    44. # nThread 1 nGpus 4 minBytes 8 maxBytes 268435456 step: 2(factor) warmup iters: 5 iters: 20 agg iters: 1 validation: 1 graph: 0
    45. #
    46. # Using devices
    47. #  Rank  0 Group  0 Pid  25570 on jettech-WS-C621E-SAGE-Series device  0 [0x3b] NVIDIA GeForce RTX 2080
    48. #  Rank  1 Group  0 Pid  25570 on jettech-WS-C621E-SAGE-Series device  1 [0x5e] NVIDIA GeForce RTX 2080
    49. #  Rank  2 Group  0 Pid  25570 on jettech-WS-C621E-SAGE-Series device  2 [0x86] NVIDIA GeForce RTX 2080
    50. #  Rank  3 Group  0 Pid  25570 on jettech-WS-C621E-SAGE-Series device  3 [0xaf] NVIDIA GeForce RTX 2080
    51. #
    52. #                                                              out-of-place                       in-place          
    53. #       size         count      type   redop    root     time   algbw   busbw #wrong     time   algbw   busbw #wrong
    54. #        (B)    (elements)                               (us)  (GB/s)  (GB/s)            (us)  (GB/s)  (GB/s)       
    55.            8             2     float     sum      -1    15.71    0.00    0.00      0    15.63    0.00    0.00      0
    56.           16             4     float     sum      -1    17.28    0.00    0.00      0    15.91    0.00    0.00      0
    57.           32             8     float     sum      -1    17.18    0.00    0.00      0    16.18    0.00    0.00      0
    58.           64            16     float     sum      -1    17.14    0.00    0.01      0    15.87    0.00    0.01      0
    59.          128            32     float     sum      -1    17.09    0.01    0.01      0    16.30    0.01    0.01      0
    60.          256            64     float     sum      -1    17.23    0.01    0.02      0    15.90    0.02    0.02      0
    61.          512           128     float     sum      -1    17.28    0.03    0.04      0    16.38    0.03    0.05      0
    62.         1024           256     float     sum      -1    17.13    0.06    0.09      0    15.81    0.06    0.10      0
    63.         2048           512     float     sum      -1    17.63    0.12    0.17      0    15.80    0.13    0.19      0
    64.         4096          1024     float     sum      -1    17.22    0.24    0.36      0    15.99    0.26    0.38      0
    65.         8192          2048     float     sum      -1    16.61    0.49    0.74      0    16.11    0.51    0.76      0
    66.        16384          4096     float     sum      -1    18.69    0.88    1.31      0    18.36    0.89    1.34      0
    67.        32768          8192     float     sum      -1    23.44    1.40    2.10      0    23.02    1.42    2.14      0
    68.        65536         16384     float     sum      -1    34.72    1.89    2.83      0    34.55    1.90    2.85      0
    69.       131072         32768     float     sum      -1    63.00    2.08    3.12      0    62.87    2.08    3.13      0
    70.       262144         65536     float     sum      -1    93.22    2.81    4.22      0    93.98    2.79    4.18      0
    71.       524288        131072     float     sum      -1    148.2    3.54    5.31      0    148.1    3.54    5.31      0
    72.      1048576        262144     float     sum      -1    294.1    3.57    5.35      0    289.8    3.62    5.43      0
    73.      2097152        524288     float     sum      -1    595.3    3.52    5.28      0    592.2    3.54    5.31      0
    74.      4194304       1048576     float     sum      -1   1319.9    3.18    4.77      0   1317.6    3.18    4.77      0
    75.      8388608       2097152     float     sum      -1   3014.5    2.78    4.17      0   3100.5    2.71    4.06      0
    76.     16777216       4194304     float     sum      -1   6966.1    2.41    3.61      0   7025.2    2.39    3.58      0
    77.     33554432       8388608     float     sum      -1    13814    2.43    3.64      0    13829    2.43    3.64      0
    78.     67108864      16777216     float     sum      -1    28272    2.37    3.56      0    28100    2.39    3.58      0
    79.    134217728      33554432     float     sum      -1    55028    2.44    3.66      0    55975    2.40    3.60      0
    80.    268435456      67108864     float     sum      -1   111871    2.40    3.60      0   111223    2.41    3.62      0
    81. # Out of bounds values : 0 OK
    82. # Avg bus bandwidth    : 2.23175 
    83. #

    5.安装anconda

    首先下载Anaconda3
    在[清华镜像]下载Linux版本的anaconda
    清华镜像官网Anaconda下载

    里选择的是Anaconda3-5.0.0-Linux-x86_64.sh

    在用户文件夹下新建一个名为anaconda的文件夹，并将刚刚下载的文件放在此文件夹中，执行以下命令：

    bash Anaconda3-5.0.0-Linux-x86_64.sh

    需要都很多页协议，不断按回车键跳过。
    出现询问时就输入yes
    之后选择默认的安装目录，按回车确定。
    出现询问是否初始化或配置环境变量就输入yes
    安装完成。
    创建虚拟环境

    (py10_paddleocr2.8_gpu_wubo) root@jettech-WS-C621E-SAGE-Series:/data/wubo/paddleocr/env# conda create --name py10_paddleocr2.8_gpu_wubo python=3.10

    6. 安装PaddlePaddle

    这里参照官网进行安装即可：

    (py10_paddleocr2.8_gpu_wubo) root@jettech-WS-C621E-SAGE-Series:/data/wubo/paddleocr/env# python -m pip install paddlepaddle-gpu==2.6.1.post120 -f https://www.paddlepaddle.org.cn/whl/linux/mkl/avx/stable.html

    最后进行验证。

    使用 python 或 python3 进入python解释器，输入：

    GPU版本

    1. import paddle 
    2. paddle.utils.run_check()

    如果出现PaddlePaddle is installed successfully!，说明您已成功安装。同时会显示当前可以并行使用的GPU数量。

    7.安装Pytorch

    参照官网命令进行安装：

    最后验证安装是否成功。

    打开Python,输入以下命令：

    1. import torch
    2. print(torch.cuda.is_available())

    8.安装paddleocr客户端 命令行模式

  • 相关阅读:
    shopee选品软件：解决你店铺选品难题的神器-shopee选品软件知虾
    软件开发中，做好需求管理，这4点很关键。
    C++ STL库 list(链表)
    力扣labuladong一刷day14天翻转单链表共2题
    ExtJS - Deployment Application（部署应用）
    Web前端 Promise 入门 【尚硅谷】
    Camunda工作流引擎简记
    付费？是不可能的！20行Python代码实现一款永久免费PDF编辑工具
    如何使用Postman调试HMS Core推送接口？
    【C++】C++ 入门
  • 原文地址：https://blog.csdn.net/Michaelwubo/article/details/140380986