Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
Lightning Talk @ Chainer Meetup
Search
Sinhrks
October 14, 2015
Programming
0
53k
Lightning Talk @ Chainer Meetup
Sinhrks
October 14, 2015
Tweet
Share
More Decks by Sinhrks
See All by Sinhrks
daskperiment: Reproducibility for Humans
sinhrks
1
410
PythonとApache Arrow
sinhrks
6
1.9k
大規模データの機械学習におけるDaskの活用
sinhrks
10
3.2k
機械学習と解釈可能性
sinhrks
7
5.7k
LIME
sinhrks
2
1.4k
データ分析言語R 1年の振り返り
sinhrks
5
2.5k
pandasでのOSS活動事例と最初の一歩
sinhrks
2
19k
Dask Distributedによる分散機械学習
sinhrks
4
1.5k
Data processing using pandas and Dask
sinhrks
1
270
Other Decks in Programming
See All in Programming
スタートアップを支える技術戦略と組織づくり
pospome
8
11k
全員アーキテクトで挑む、 巨大で高密度なドメインの紐解き方
agatan
7
8.9k
例外処理を理解して、設計段階からエラーを見つけやすく、起こりにくく #phpconfuk
kajitack
12
6.4k
r2-image-worker
yusukebe
1
180
Building AI with AI
inesmontani
PRO
1
260
AIを駆使して新しい技術を効率的に理解する方法
nogu66
1
660
CloudflareのSandbox SDKを試してみた
syumai
0
180
ソフトウェア設計の課題・原則・実践技法
masuda220
PRO
21
16k
Module Harmony
petamoriken
2
540
Claude Code on the Web を超える!? Codex Cloud の実践テク5選
sunagaku
0
600
AIエージェントでのJava開発がはかどるMCPをAIを使って開発してみた / java mcp for jjug
kishida
4
770
[堅牢.py #1] テストを書かない研究者に送る、最初にテストを書く実験コード入門 / Let's start your ML project by writing tests
shunk031
11
5.7k
Featured
See All Featured
ReactJS: Keep Simple. Everything can be a component!
pedronauck
666
130k
RailsConf 2023
tenderlove
30
1.3k
Keith and Marios Guide to Fast Websites
keithpitt
413
23k
The Straight Up "How To Draw Better" Workshop
denniskardys
239
140k
The Art of Delivering Value - GDevCon NA Keynote
reverentgeek
16
1.8k
Mobile First: as difficult as doing things right
swwweet
225
10k
A Tale of Four Properties
chriscoyier
162
23k
Building an army of robots
kneath
306
46k
個人開発の失敗を避けるイケてる考え方 / tips for indie hackers
panda_program
118
20k
Practical Orchestrator
shlominoach
190
11k
How STYLIGHT went responsive
nonsquared
100
5.9k
Building a Modern Day E-commerce SEO Strategy
aleyda
45
8.1k
Transcript
Lightning Talk@ Chainer Meetup by sinhrks
ࣗݾհ • ϝʔΧʔʹͯσʔλੳ • OSS ׆ಈ: • PyData Development Team
(pandasίϛολ) • Blaze Development Team (Daskίϛολ) • NumPyαϒηοτͷฒྻࢄndarrayͷ࣮͋Γ • GitHub: https://github.com/sinhrks
Chainer ͷར༻໘ • ֶࣗࣗश • ۀͰͷ Deep Learning ׆༻ͳ͠
Chainer ͷྑ͍ͱ͜Ζ • ڧ͍ํ͕࡞͍ͬͯΔ • ϨΠϠఆ͕ٛతͰ໌ྎ • ෦ͷ࣮͕ಡΈ͍͢
Chainer ͷվળ • ֶशશମͷॲཧΛهड़͢Δඞཁ͕͋Δ • ҎԼͷΑ͏ͳఆ൪ॲཧͷͨΊͷ؆қه๏(ؔ/Ϋ ϥε)͕ཉ͍͠ • ॳճͷΈάϥϑग़ྗ •
epoch ͝ͱͷଛࣦ • early stopping • ՄࢹԽ
Chainer ͷ ՄࢹԽπʔϧ࡞Γ͍ͨ
֓ཁ • ֶशͷϞχλϦϯά • Computational Graph • ॏΈߦྻ (ະணख) •
ͳͲΛɺϦϞʔτ (Jupyter / Webϒϥβ) ͔Β ֬ೝ͍ͨ͠ • →Bokeh
ֶशͷϞχλϦϯά • ͪ͜ΒͷϒϩάΤϯτϦͷ༰ import chainer_vis.visualizer as visualizer iv = visualizer.JupyterVisualizer()
monitor = iv.get_classification_monitor() for … monitor.update(test_loss=loss, test_accuracy=acc)