꿈꾸는 아기토깡이 블로그♡

꿈꾸는 아기토깡이 블로그♡

  • 분류 전체보기 (14)
    • Hello TensorBoard (12)
    • Hello Spark (0)
    • Hello Python (1)
    • Hello WebCrawling (1)
    • Hello Pandas (0)
  • 홈
  • 태그
  • 미디어로그
  • 위치로그
  • 방명록
RSS 피드
로그인
로그아웃 글쓰기 관리

꿈꾸는 아기토깡이 블로그♡

컨텐츠 검색

태그

최근글

댓글

공지사항

아카이브

전체 글(14)

  • 09. Fix model by TensorBoard Histogram

    1. Seeing "Histogram", all values are set to zero. 2. Go to source code and fix the initial value of W, b. 3. Rerun the code and check the histogram.

    2018.11.01
  • 08. Collect summary

    1. tf.summary.scalar 2. tf.summary.image 3. tf.summary.histogram

    2018.11.01
  • 07. Cleaning the graph (node name, name scope)

    1. Node Name 2. Name Scope 3. Clean graph!

    2018.11.01
  • 06. Fail in training??

    1. Failed in training 2. See the graph, but way too complicated..

    2018.11.01
  • 05. Create a network

    1. Simple network for MNIST dataset training 2. Define layers (conv_layer, fc_layer) 3. Create a network 4. Adam Optimizer, Accuracy 5. Initialize variables 6. training and evaluate accuracy

    2018.11.01
  • 04. Placeholder (on TensorBoard)

    1. Setup placeholders, and reshape the data2. Activate TensorBoard (writer) 3. (Open terminal: Alt + F12) $ tensorboard --logdir graphs (or, --logdir==graphs, --logdir=="graphs", --logdir=="C:\~\graphs")4. (Open Chrome) http://localhost:6006

    2018.11.01
이전
1 2 3
다음
티스토리
© 2018 TISTORY. All rights reserved.

티스토리툴바