第2回 Designing Deep Learning Systems 読書会 (2024/08/17 13:00〜) (original) (raw)

募集内容 無料リモート参加(discord) 無料 先着順 6/50人
申込者 Naoyuki Kakuda tomita DAI+=3 TakashiTsuruta ythink KanSAKAMOTO 申込者一覧を見る
開催日時 2024/08/17(土) 13:00 ~ 18:00 Googleカレンダー icsファイル
募集期間 2024/07/20(土) 17:17 〜 2024/08/17(土) 18:00まで
会場 オンライン オンライン
出席登録 (イベント開始時間の2時間前から終了時間まで、参加者のみに公開されます)

イベントの説明

内容

Discordサーバーへの参加

準備

進行

時間 内容
~13:00 入室,準備
13:00~13:10 開始の挨拶
13:10~14:00 自己紹介
14:10~15:00 読書会
15:00~15:20 長休憩
15:20~17:50 読書会
17:50~18:00 ふりかえり,退室
18:00~ 雑談、飲み会 (自由参加)

目次

目次
front matter
1 An introduction to deep learning systems
1.1 The deep learning development cycle
1.2 Deep learning system design overview
1.3 Building a deep learning system vs. developing a model
Chapter 1 Summary
2 Dataset management service
2.1 Understanding dataset management service
2.2 Touring a sample dataset management service
2.3 Open source approaches
Chapter 2 Summary
3 Model training service
3.1 Model training service: Design overview
3.2 Deep learning training code pattern
3.3 A sample model training service
3.4 Kubeflow training operators: An open source approach
3.5 When to use the public cloud
Chapter 3 Summary
4 Distributed training
4.1 Types of distributed training methods
4.2 Data parallelism
4.3 A sample service supporting data parallel–distributed training
4.4 Training large models that can’t load on one GPU
Chapter 4 Summary
5 Hyperparameter optimization service
5.1 Understanding hyperparameters
5.2 Understanding hyperparameter optimization
5.3 Designing an HPO service
5.4 Open source HPO libraries
Chapter 5 Summary
6 Model serving design
6.1 Explaining model serving
6.2 Common model serving strategies
6.3 Designing a prediction service
Chapter 6 Summary
7 Model serving in practice
7.1 A model service sample
7.2 TorchServe model server sample
7.3 Model server vs. model service
7.4 Touring open source model serving tools
7.5 Releasing models
7.6 Postproduction model monitoring
Chapter 7 Summary
8 Metadata and artifact store
8.1 Introducing artifacts
8.2 Metadata in a deep learning context
8.3 Designing a metadata and artifacts store
8.4 Open source solutions
Chapter 8 Summary
9 Workflow orchestration
9.1 Introducing workflow orchestration
9.2 Designing a workflow orchestration system
9.3 Touring open source workflow orchestration systems
Chapter 9 Summary
10 Path to production
10.1 Preparing for productionization
10.2 Model productionization
10.3 Model deployment strategies
Chapter 10 Summary
Appendix A. A “hello world” deep learning system
Appendix B. Survey of existing solutions
Appendix C. Creating an HPO service with Kubeflow Katib

フィード

グループ

終了

2024/08/17(土)

13:00 〜 18:00

募集期間
2024/07/20(土) 17:17 〜
2024/08/17(土) 18:00

イベントへのお問い合わせ

会場

オンライン

オンライン

オンライン

管理者