https://github.com/wookayin/tensorflowkr-2017-talk-bestpractice
π¬ Slides and Tutorial Codes for the talk 'Toward Best Practices of TensorFlow Code Patterns' (2017)
https://github.com/wookayin/tensorflowkr-2017-talk-bestpractice
best-practices slides tensorflow
Last synced: over 1 year ago
JSON representation
π¬ Slides and Tutorial Codes for the talk 'Toward Best Practices of TensorFlow Code Patterns' (2017)
- Host: GitHub
- URL: https://github.com/wookayin/tensorflowkr-2017-talk-bestpractice
- Owner: wookayin
- Created: 2017-01-04T09:32:08.000Z (over 9 years ago)
- Default Branch: master
- Last Pushed: 2019-04-05T22:24:25.000Z (about 7 years ago)
- Last Synced: 2025-02-27T12:15:22.938Z (over 1 year ago)
- Topics: best-practices, slides, tensorflow
- Language: CSS
- Homepage: https://wookayin.github.io/TensorFlowKR-2017-talk-bestpractice
- Size: 3.08 MB
- Stars: 153
- Watchers: 26
- Forks: 19
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
Toward Best Practices of TensorFlow Code Patterns
==================================================
**Jongwook Choi** ([@wookayin][wookayin]) and
**Beomjun Shin** ([@shastakr][shastakr])
A talk in [the 2nd meetup][tfkr-meetup] of [TensorFlow Korea Group][tfkr-facebook]
January 14th, 2017
π§ Warning!! π§
Many contents in this talk are outdated (and thus less recommended) as they were based on older versions of TensorFlow.
TensorFlow now has more handy APIs and is moving toward TF 2.0
that encourages many better practices.
This repository is kept only for archival purposes.
[https://wookayin.github.io/TensorFlowKR-2017-talk-bestpractice/][github-pages]
Slides
------
- :kr: [**λ°νμλ£ (In Korean)**](https://wookayin.github.io/TensorFlowKR-2017-talk-bestpractice/ko/)
- :us: Lecture Slide (In English): ~_Coming Soon!_~
Abstract (In Korean)
-----------------
μ΄λ² λ°νμμλ (μ£Όλ‘ μ°κ΅¬ νλ‘μ νΈλ₯Ό μν) TensorFlow μ½λλ₯Ό μμ±ν λμ
Best Practice κ·Έλ¦¬κ³ Code Patterns μ λν΄ μ΄μΌκΈ°ν©λλ€.
μ μ μ μΌλ‘ λ€μκ³Ό κ°μ λ΄μ©λ€μ λ€λ£° μμ μ
λλ€:
- ꡬν λ° μ€κ³ μμ κ°μ’
μꡬμ¬ν/κ³ λ €μ¬ν
- TensorFlowμ λͺ¨λΈ λͺ¨λν λ° μΆμν ν¨ν΄ (+ν‘λ¨κ΄μ¬μ μ²λ¦¬ λ°©λ²)
- High-level API μ κ·Ήμ μΌλ‘ μ¬μ©νκΈ° (tf.contrib.layers, tf.slim, tf.learn λ±)
- (i) λͺ¨λΈ μ½λ μμ± (ii) λ°μ΄ν°μ
λ° νΈλ μ΄λ μμ©κ΅¬ (iii) μ€ν λ° μ€μ κ΄λ¦¬ (iv) κΈ°ν μμν ν λ±μ κ΄ν ν¨ν΄κ³Ό μ½λ μμ
Abstract
--------
In this talk, we aim to deliver several best practices, guidelines, and code patterns for writing TensorFlow codes,
mainly when conducting research projects.
A tentative and incomplete list of topics:
- Common requirements and frequent design concerns in implementation
- Patterns of model modularization and abstraction in TensorFlow (+ cross-cutting concerns)
- Using modern and high-level APIs (tf.contrib.layers, tf.slim, tf.learn, etc.)
- Patterns and code examples on: (i) styles of model writing, (ii) dataset loader and training boilerplates, (iii) management of experiments and configuration,
and (iv) other general and miscellaneous tips.
Note
----
The talk will given in Korean, but I will self-translate the material into English shortly after the talk.
The topics are mainly based on the authors' personal experience, so may contain some opinionated suggestions.
Of course, there cannot be the only answer:
Please contact me if you have a suggestion, a question, or an idea that can improve the contents of this talk!
[tfkr-meetup]: http://onoffmix.com/event/86620
[tfkr-facebook]: https://www.facebook.com/groups/TensorFlowKR/
[github-pages]: https://wookayin.github.io/TensorFlowKR-2017-talk-bestpractice/
[wookayin]: https://github.com/wookayin
[shastakr]: https://github.com/shastakr