목록전체 글 (77)

Shakerato

Google colab에서 한국어 NLP 처리 환경 구축

What is Colab (Colaboratory) ?- https://colab.research.google.com- Colaboratory는 기계 학습 교육 및 연구 보급을 돕기 위해 만든 Google 연구 프로젝트입니다.- Jupyter 노트북 환경에서 클라우드 환경에서의 설치 및 실행이 전혀 필요하지 않습니다. - Colaboratory 노트는 Google 드라이브에 저장되며 Google 문서 또는 스프레드 시트에서와 마찬가지로 공유 할 수 있습니다.- Colaboratory는 무료입니다.- Now you can develop deep learning applications with Google Colaboratory on the free Tesla K80 GPU- using Keras, Tens..
Research 2018. 3. 31. 20:35
Tensorflow Object Detection Tutorial

[Code and Explain] https://github.com/tensorflow/models/tree/master/research/object_detection [Download] https://github.com/tensorflow/models [Related Youtube: sentdex] https://www.youtube.com/watch?v=COlbP62-B-U [Anaconda 3, Create Envirnment for tensorflow 1.4] 1. open 'cmd' as administrator 2. conda update -n base conda 3. conda create -n tensorflow1.4 python=3.5 anaconda 4. [tensorflow-gpu 1..
Research 2018. 3. 29. 17:51
How to make video clips of the long video using ffmpeg

1. https://www.ffmpeg.org/download.html2. Click Windows icon3. Click 'Windows Builds' text4. Download version 3.4.2, 64bit or 32bit, Static5. Extract zip file, copy /bin/ffmpeg.exe to video folder6. Making video clips using this command ffmpeg -i [Input file name] -ss [start time] -to [end time] [output file name] (example)ffmpeg -i video.mp4 -ss 00:00:03.000 -to 00:00:12.000 clip1.mp4
Research 2018. 3. 27. 16:56
Deep Learning - GPU memory limitation, How to overcome it?

Limited GPU Memory GPU usually has lesser device memory than host memoryThe latest high-end GPU (such as NVIDIA GPU P100)12–16 GB device memoryHost system memory256GBTrend for deep learning models is to have a “deeper and wider” architectureEspecially, RNN needs a lot of memory 1. First Solution: distributed Deep LearningSource: M. Cho et al., "PowerAI DDL", 2017PowerAI DDL provides a unified in..
Research 2018. 3. 26. 22:36
FakeApp 1.1 Tutorial

(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..
Research 2018. 2. 12. 13:28
Create environment for tensorflow 1.4 in Anaconda 3

[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..
Research 2018. 2. 10. 23:07
Prev 1 ··· 9 10 11 12 13 Next