연습장 끄적끄적

tensorflow 개발환경 만들기 (docker 사용)

2018-05-05T00:00:00+00:00

docker 이미지

tensorflow/tensorflow:latest-py3
python3 사용 https://hub.docker.com/r/tensorflow/tensorflow/
docker를 사용한 golang 개발환경 만들기 참고

hello world

  # 현재 디렉토리에 example download
  $ git clone https://github.com/jinikr/tensorflow-excersice .
  Cloning into '.'...
  remote: Counting objects: 6, done.
  remote: Compressing objects: 100% (4/4), done.
  remote: Total 6 (delta 0), reused 6 (delta 0), pack-reused 0
  Unpacking objects: 100% (6/6), done.

  # docker 실행 (docker-compose 사용)
  $ docker-compose up -d
  Creating network "tensorflow-excersice_default" with the default driver
  Building golang-dev
  Step 1/2 : FROM tensorflow/tensorflow:latest-py3
  latest-py3: Pulling from tensorflow/tensorflow
  297061f60c36: Pull complete
  e9ccef17b516: Pull complete
  dbc33716854d: Pull complete
  8fe36b178d25: Pull complete
  686596545a94: Pull complete
  11251ff8ca17: Pull complete
  632ad5221740: Pull complete
  c61bc0eaf8d9: Pull complete
  f5e02d7ae67a: Pull complete
  f29a5775b72a: Pull complete
  c9bf16bc9d22: Pull complete
  d4423e297b68: Pull complete
  418a4217796d: Pull complete
  02b18d984060: Pull complete
  Digest: sha256:24b533d6db08369375f85b7d8da0e7d8c35c105337a7f6fa71947195b900eec9
  Status: Downloaded newer image for tensorflow/tensorflow:latest-py3
  ---> a83a3dd79ff9
  Step 2/2 : RUN pip install "h5py==2.8.0rc1"     && export TF_CPP_MIN_LOG_LEVEL=2
  ---> Running in 7d944f6ed0e3
  Collecting h5py==2.8.0rc1
    Downloading https://files.pythonhosted.org/packages/00/26/4bbf836d3a32d91b0bec3a7533d27abcd13aeafb68fc7cbfdf3ff792758b/h5py-2.8.0rc1-cp35-cp35m-manylinux1_x86_64.whl (2.4MB)
  ^@Requirement already satisfied: numpy>=1.7 in /usr/local/lib/python3.5/dist-packages (from h5py==2.8.0rc1) (1.14.2)
  Requirement already satisfied: six in /usr/local/lib/python3.5/dist-packages (from h5py==2.8.0rc1) (1.11.0)
  Installing collected packages: h5py
    Found existing installation: h5py 2.7.1
      Uninstalling h5py-2.7.1:
        Successfully uninstalled h5py-2.7.1
  Successfully installed h5py-2.8.0rc1
  Removing intermediate container 7d944f6ed0e3
  ---> 57bc8e411f5d
  Successfully built 57bc8e411f5d
  Successfully tagged tf-dev:0.1
  WARNING: Image for service golang-dev was built because it did not already exist. To rebuild this image you must use `docker-compose build` or `docker-compose up --build`.
  Creating tf-dev ... done

  # docker 컨테이너 확인
  $ docker ps
  CONTAINER ID        IMAGE                   COMMAND                  CREATED             STATUS              PORTS                NAMES
  95fa37fdd8a4        tf-dev:0.1              "/run_jupyter.sh --a…"   About an hour ago   Up About an hour    6006/tcp, 8888/tcp   tf-dev

  # docker 컨테이너의 bash 접속
  $ docker exec -it tf-dev bash
  root@95fa37fdd8a4:/var/app#

  # hello world 테스트
  root@95fa37fdd8a4:/var/app# cd 01.\ hello-world/
  root@95fa37fdd8a4:/var/app/01. hello-world# python hello-world.py
  Tensor("Const:0", shape=(), dtype=string)
  2018-05-05 12:11:27.978539: I tensorflow/core/platform/cpu_feature_guard.cc:140] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA
  b'Hello, Tensor.'

부록

경고메세지 보기싫으면..
```
export TF_CPP_MIN_LOG_LEVEL=2
```

golang 개발환경 만들기 (docker 사용)

2018-05-05T00:00:00+00:00

docker 설치.

커뮤니티에디션 설치
https://www.docker.com/community-edition#/download

docker 이미지 고르기

https://hub.docker.com/_/golang/
stretch(Debian 9) 을 사용하자.
golang:1.10.1-stretch
alpine 계열은 가볍고… 가볍지만, 개발환경으로 사용하기엔 불편.
패키지관리자도 익숙하지 않고, bash도 없고.. 물론 설치하고 익숙해지면 되지만, 일단 익숙한걸로 사용하자.
https://www.lesstif.com/pages/viewpage.action?pageId=35356819

docker 실행하기

이미지를 받아서 실행하자.

$ docker pull golang:1.10.1-stretch
$ docker run -w /var/app -v `pwd`:/var/app --name golang-dev -i -t golang:1.10.1-stretch bash

컨테이너 삭제 명령
```
$ docker rm golang-dev
```

docker-compose 활용

docker-compose를 사용하자 (docker에 포함되어 이미 설치되어 있을것임.)
https://docs.docker.com/compose/
실행
```
$ docker-compose up -d
```
docker-compose로 실행된 docker container에 접속하자.
```
$ docker exec -it golang-dev bash
```
종료
```
$ docker-compose down
```

hello world

  # 현재 디렉토리에 example download
  $ git clone https://github.com/jinikr/golang-exercise .
  Cloning into '.'...
  remote: Counting objects: 13, done.
  remote: Compressing objects: 100% (7/7), done.
  remote: Total 13 (delta 0), reused 7 (delta 0), pack-reused 0
  Unpacking objects: 100% (13/13), done.

  # docker 실행 (docker-compose 사용)
  $ docker-compose up -d
  Pulling golang-dev (golang:1.10.1-stretch)...
  1.10.1-stretch: Pulling from library/golang
  cc1a78bfd46b: Pull complete
  53c14872d997: Pull complete
  99ae159b9cae: Pull complete
  66cbf2b79699: Pull complete
  a16f46d95485: Pull complete
  f235c5ef7891: Pull complete
  cdd98e20e590: Pull complete
  Digest: sha256:e7331dea607f66adb0d597f43d7cb9cdc36ac73cff56536bde3311b46c5186e9
  Status: Downloaded newer image for golang:1.10.1-stretch
  Creating golang-dev ... done

  # docker 컨테이너 확인
  $ docker ps
  CONTAINER ID        IMAGE                   COMMAND             CREATED             STATUS              PORTS               NAMES
  5a40fb644fbd        golang:1.10.1-stretch   "bash"              43 seconds ago      Up 42 seconds                           golang-dev

  # docker 컨테이너의 bash 접속
  $ docker exec -it golang-dev bash
  root@5a40fb644fbd:/var/app#

  # hello world 테스트
  root@5a40fb644fbd:/var/app# cd 01.\ hello-world/
  root@5a40fb644fbd:/var/app/01. hello-world# go run hello-world.go
  hello, world

부록

# docker container 조회
$ docker ps
# 종료된 컨테이너도 확인하려면 -a 옵션을 붙인다.)
$ docker ps -a

# docker image 확인
$ docker images

# docker image 삭제
$ docker rmi {image name}

# 미사용 이미지 정리
$ docker image prune

test!!!

2018-05-02T00:00:00+00:00