Shakerato

(Reference) https://www.youtube.com/watch?v=ghTb2kZSpZE&ab_channel=IrrelevantVoice 1. Download any video1 and video2, ffmepg.exe(https://www.ffmpeg.org/) 2. Create folder c:\fakes\data_A 3. Create folder c:\fakes\data_B 4. Create folder c:\fakes\model 5. Copy ffmpeg.exe and video2 file to c:\fakes\data_A 6. ffmpeg -i video2.mp4 -vf fps=1 "a%04d.png" 7. Copy ffmpeg.exe and video1 file to c:\fakes..

[Create Envirnment for tensorflow 1.4 in Anaconda 3]1. open 'cmd' as administrator 2. conda update -n base conda 3. conda create -n tensorflow1.4 python=3.5 anaconda 4. [install tensorflow-gpu 1.4] in the environment 'tensorflow1.4' pip install Iv tensorflow-gpu==1.4.0 [> tensorflow-gpu 1.5 needs cuda 9.0], install latest version pip install tensorflow-gpu 5. conda install -c anaconda cudnn 6. c..

0. nvidia-docker should be installed. 1. https://hub.docker.com/r/ppc64le/theano/ (1) docker pull ppc64le/theano (2) docker create -i -t --name mytheano ppc64le/theano:latest (3) docker start mytheano (4) docker attach mytheano 2. cd / ; cd tmp 3. apt-get install python-numpy python-scipy python-dev python-pip python-nose g++ libblas-dev git 4. https://github.com/marekrei/theano-tutorial (1) apt..

These links are about multi-gpu based deep learning environment construction.If someone is interested in this, just refer to the links below. [스파크와 분산 GPU 시스템에서 심층 신경망 학습하기]https://deeplearning4j.org/kr/spark-gpus [Building a Deep Learning Box with Multiple GPUs]https://medium.com/@acrosson/building-a-deep-learning-box-d17d97e2905c [Machine Learning on VMware vSphere 6 with NVIDIA GPUs]https://b..

unRAID Server OS (Lime Tech)는 USB 플래시 장치에서 부팅하고 디지털 미디어 저장을 위해 특별히 설계된 Network Attached Storage 서버 운영 체제 [unRAID (LIME TECH)] What is unRAID?: https://lime-technology.com/what-is-unraid/ 한 컴퓨터에 설치된 cpu, gpu, 하드디스크, 네트워크, 키보드, 마우스 등이 개별 할당된 여러개의 가상 환경을 구축 가능GPU path-through를 제공하기 때문에 각 독립된 가상환경에서 자신이 할당받은 gpu로 서로 다른 작업 (예를들어 게임)을 할 수 있음.아래는 이와 관련한 사례를 정리함. 대표적으로 Linus Tech Tips에서 한 컴퓨터 자원을 서로 다른 ..

Kitematic setup error: ~ ./pipe/docker_engine ~ Solution (Windows 10)1. Install latest Docker version (Install or not) (2017-09-04: version 17.06.1-ce) https://download.docker.com/win/stable/InstallDocker.msi2. Hyper-V service on (important)3. Restart Windows (important)