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
Performance Tuning Tips of TensorFlow Inference
Search
chie8842
December 20, 2018
Technology
1
740
Performance Tuning Tips of TensorFlow Inference
https://m3-engineer.connpass.com/event/111354/
2018/12/20 Cookpad×merpay×m3 機械学習 Nightの登壇資料です。
chie8842
December 20, 2018
Tweet
Share
More Decks by chie8842
See All by chie8842
MongoDB Atlas Search のご紹介
chie8842
2
1.6k
MongoDB Atlas Vectorsearchではじめる生成AIアプリ開発
chie8842
3
1.6k
AWS GlueとAWS Lake Formationではじめるデータマネジメント
chie8842
0
1k
Distributed Processing in Python
chie8842
2
700
クックパッドにおける推薦(と検索)の取り組み
chie8842
20
8k
Understanding distributed processing in Python
chie8842
2
2k
クックパッドにおけるCloud AutoML事例
chie8842
9
7.9k
Cookpad_Internship_MLOps_Lecture_2018
chie8842
35
16k
機械学習デプロイを支えるコンテナ技術(Machine Learning on Docker)
chie8842
14
8.4k
Other Decks in Technology
See All in Technology
フロントエンドも盛り上げたい!フロントエンドCBとAmplifyの軌跡
mkdev10
2
270
生成AIによるCloud Native基盤構築の可能性と実践的ガードレールの敷設について
nwiizo
6
490
LLM とプロンプトエンジニアリング/チューターをビルドする / LLM, Prompt Engineering and Building Tutors
ks91
PRO
1
250
LangfuseでAIエージェントの 可観測性を高めよう!/Enhancing AI Agent Observability with Langfuse!
jnymyk
1
220
Running JavaScript within Ruby
hmsk
3
320
DuckDB MCPサーバーを使ってAWSコストを分析させてみた / AWS cost analysis with DuckDB MCP server
masahirokawahara
0
1.2k
AIコーディングの最前線 〜活用のコツと課題〜
pharma_x_tech
2
960
4/16/25 - SFJug - Java meets AI: Build LLM-Powered Apps with LangChain4j
edeandrea
PRO
1
100
YOLOv10~v12
tenten0727
4
940
ここはMCPの夜明けまえ
nwiizo
2
2.1k
MCPを活用した検索システムの作り方/How to implement search systems with MCP #catalks
quiver
12
6.4k
大AI時代で輝くために今こそドメインにディープダイブしよう / Deep Dive into Domain in AI-Agent-Era
yuitosato
1
360
Featured
See All Featured
Documentation Writing (for coders)
carmenintech
69
4.7k
Helping Users Find Their Own Way: Creating Modern Search Experiences
danielanewman
29
2.5k
Build The Right Thing And Hit Your Dates
maggiecrowley
35
2.6k
Building Flexible Design Systems
yeseniaperezcruz
329
38k
VelocityConf: Rendering Performance Case Studies
addyosmani
328
24k
YesSQL, Process and Tooling at Scale
rocio
172
14k
Building an army of robots
kneath
304
45k
ReactJS: Keep Simple. Everything can be a component!
pedronauck
666
120k
The World Runs on Bad Software
bkeepers
PRO
67
11k
A Tale of Four Properties
chriscoyier
158
23k
Visualizing Your Data: Incorporating Mongo into Loggly Infrastructure
mongodb
45
9.5k
Designing for Performance
lara
608
69k
Transcript
3
) ( • s W ei Vi h • F
PRP RS X c J • m 2E 12G7C: 1 4G :D :BCn • a .+ + 2 .44 o l • A:) D: C ( 08 : 8E r
• %*+5 !# • .6+5 !# +51/$"# 4'$
• %*)+5/2 &3,0.6 +5/2 & -(
F Tae P eU e p C
2 s • e t • Fnw C i I Tae • c 13 1 • + 25 1 5+ • lo V U • e h • r C
• / .. / .
/ . / :
J ca C F • •
/ Ie • P Ih SU S e • l J C ca C F • P _ • / / . / / . Ifi J C F • _ _ T I d
041 /16 l l 8 :
T o : • 041 /16 : T ( : T F e p . 0 o g n ( n a ) 2 7/ . n
• ) : : () ) ) )
) 1 () ) ) ) 1 ) ) ) () ) ) ) 1 () ) ) 1 S ) . ) =
• 5 .10 8 7 :
2 6/ 3 S
• /1423 025 . F6 l
TU 6on . C 9T P n 8 C 2 . 4 ) ( ) ( ) ( e % . n
clL 488 Ie w MI ng ) 8
5 K K ) 8 5 ( ) 8 5 .83 5 4/ . t iL Th Ly b r x x nga ( 0 :5.
• : FC CIP5/ TX S fe OL. TX
S N C ) C 2 C C C = C C F C A : C C F C : : ) 1 A : 1 C F 10 F:FF ) 3:FF C C ) C g ba c c _ ( J :
/ • dh c a o : • )(4 (
) ) ( ) _a t e gr m • (4 ( ) ) ( ) m • fp = l s SPm C _c i grn 4 . 4 . 4 4 4 . )(4 ( ) ) ( ) 4 . (4 ( ) ) ( ) 4 4 . 4 .
/ • pu alo nde Cm P • 7: 21
1 2 288487 91 6 423 m nde s fCe • 7: 4 1 1 2 288487 91 6 423 P fCe • r C U _ Phi C o s U .452 8 P ( 0 () t /0 % D 7: 21 1 2 288487 91 6 423 7: 4 1 1 2 288487 91 6 423 P
• 8 - n gch y de
Gf ib oz m N • OM rgchw p ʼ ul T s J F 8 . D / : 8 IL : :1 / t B B : 1
• a n 1 D in F 2 P
o F 2 3 • q la t6 n n z 2 D ( ) a 2 e b 2u a
• T • i 1 •