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
710
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
3
1.1k
MongoDB Atlas Vectorsearchではじめる生成AIアプリ開発
chie8842
3
1.2k
AWS GlueとAWS Lake Formationではじめるデータマネジメント
chie8842
0
880
Distributed Processing in Python
chie8842
2
620
クックパッドにおける推薦(と検索)の取り組み
chie8842
21
7.8k
Understanding distributed processing in Python
chie8842
2
1.8k
クックパッドにおけるCloud AutoML事例
chie8842
9
7.7k
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
8.2k
Other Decks in Technology
See All in Technology
PlaywrightによるE2Eテスト入門 / Introduction to E2E Testing with Playwright
rhumie
3
870
ログラスが面白いと思う理由をマネージャーがエモく語ってみる / 20240829 vs LT
yoshikiiida
1
520
Datadogの便利な使い方 - 意外と知らない?CoScreenとCloudcraft
taijihagino
PRO
1
600
Chunk computing
matyo91
0
270
Dive Deep in Cloud Run: Automatic Base Image update
aoto
PRO
0
900
[RSJ24] Object Retrieval in Large-Scale Indoor Environments Using Dense Text with a Multi-Modal Large Language Model
keio_smilab
PRO
0
220
Azure Cosmos DB での時系列ログの運用と改善
sansantech
PRO
0
190
Hyperledger Fabricの成長、成熟を振り返る / Looking back history of Hyperledger Fabric
gakumura
0
180
RAGHack: Kickoff and RAG 101
pamelafox
0
190
EitherT_with_Future
aoiroaoino
0
230
エンジニア向け会社紹介資料
caddi_eng
15
250k
歴史と背景から改めて振り返るVPC
shotashiratori
2
220
Featured
See All Featured
It's Worth the Effort
3n
182
27k
Thoughts on Productivity
jonyablonski
65
4.2k
Visualizing Your Data: Incorporating Mongo into Loggly Infrastructure
mongodb
38
9.1k
Templates, Plugins, & Blocks: Oh My! Creating the theme that thinks of everything
marktimemedia
24
2k
Visualization
eitanlees
142
15k
The World Runs on Bad Software
bkeepers
PRO
64
11k
Building Applications with DynamoDB
mza
89
5.9k
Music & Morning Musume
bryan
46
6k
The Illustrated Children's Guide to Kubernetes
chrisshort
46
48k
Designing with Data
zakiwarfel
98
5k
The Straight Up "How To Draw Better" Workshop
denniskardys
230
130k
Mobile First: as difficult as doing things right
swwweet
220
8.8k
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 •