Skip to content

Latest commit

 

History

History
14 lines (12 loc) · 689 Bytes

README.md

File metadata and controls

14 lines (12 loc) · 689 Bytes

Deep Learning with Defect : Intro

introduction of defect/aoi classification with common NN model.

Usage

學員容器打開後，啟動Jupyter，先開一個terminal用來下載資料:

  1. $git clone https://github.com/vscv/DLwDefectIntro
  2. $cd DLwDefectIntro/Lab_NEU/
  3. $sh getdata_NEU.sh
  4. 回到Jupyter 開啟NEU Surface Defect (InceptionV3)_20191211_v1.ipynb

TensorBoard in Jupyter

訓練結束後，在TensorBoard視覺化階段，再回到terminal開啟TensorBoard server:

  1. $tensorboard --logdir logs --port 5000 #目標埠: 5000
  2. 再參照個人容器的關聯port，每個人port數不同。在notebook中 %tensorboard --logdir="logs" --port=xxxxx #對外埠: xxxxx