[B! æ©Ÿæ¢°å¦ç¿’][AWS] clavierã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯

clavier id:clavier

æ©Ÿæ¢°å¦ç¿’ã¨AWSã«é–¢ã™ã‚‹clavierã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯ (6)

${{author_name}}$

{{{comment_expanded}}}

{{label}}

{{#is_bookmark}}ãƒªã‚¹ãƒˆ{{/is_bookmark}}{{^is_bookmark}}ãƒªãƒ³ã‚¯{{/is_bookmark}}

${{author_name}}$
{{author_name}}{{created}}
{{ #comment }}{{ comment }}{{ /comment }}
- {{ label }}

${{author_name}}$

{{{comment_expanded}}}

{{label}}

{{#is_bookmark}}ãƒªã‚¹ãƒˆ{{/is_bookmark}}{{^is_bookmark}}ãƒªãƒ³ã‚¯{{/is_bookmark}}

æ©Ÿæ¢°å¦ç¿’ãƒ—ãƒã‚¸ã‚§ã‚¯ãƒˆã«ãŠã‘ã‚‹SageMaker Processingã®ä½¿ã„æ‰€ - ã‚³ãƒãƒ’ãƒˆé–‹ç™ºè€…ãƒ–ãƒã‚°
ã¿ãªã•ã‚“ã“ã‚“ã«ã¡ã¯ã€‚æ©Ÿæ¢°å¦ç¿’ãƒãƒ¼ãƒ ã®ãŸã‹ã±ã„ï¼ˆ@takapy0210ï¼‰ã§ã™ã€‚ 2021å¹´ã‚‚ã‚ã¨1ãƒ¶æœˆã¨ãªã‚Šã¾ã—ãŸãã€‚çš†æ§˜ã„ã‹ãŒãŠéŽã”ã—ã§ã—ã‚‡ã†ã‹ã€‚ ...ã•ã¦12æœˆã¨ã„ãˆã°ã€æ¯Žå¹´æ’ä¾‹ã®ã‚¢ãƒ‰ãƒ™ãƒ³ãƒˆã‚«ãƒ¬ãƒ³ãƒ€ãƒ¼ã®å£ç¯€ã§ã™ãï¼ ã¨ã„ã†ã‚ã‘ã§ã€2021å¹´ã‚‚ã‚³ãƒãƒ’ãƒˆ Advent CalendarãŒå§‹ã¾ã‚Šã¾ã™ï¼ðŸŽ‰ åˆæ—¥ã¨ãªã‚‹æœ¬ã‚¨ãƒ³ãƒˆãƒªã§ã¯ã€æ©Ÿæ¢°å¦ç¿’ãƒãƒ¼ãƒ ã§ä½¿ç”¨ã—ã¦ã„ã‚‹SageMaker*1ã®æ©Ÿèƒ½ã§ã‚ã‚‹ã€Processing*2ã«ã¤ã„ã¦ã€æ´»ç”¨äº‹ä¾‹ã¨ã¨ã‚‚ã«ã”ç´¹ä»‹ã—ã‚ˆã†ã¨æ€ã„ã¾ã™ã€‚ ç›®æ¬¡ SageMaker Processingã¨ã¯? SKLearnProcessor / PySparkProcessor Processor / ScriptProcessor ã“ã‚Œã¾ã§ã®èª²é¡Œæ„Ÿ SageMaker Processorã®æ´»ç”¨æ–¹æ³• SageMakerã‹ã‚‰ECRã«ã‚ã‚‹ã‚³ãƒ³ãƒ†ãƒŠã‚’æŒ‡å®šã—ã¦Processor Job
clavier 2021/12/03
machinelearning

aws

SageMaker

MLOps

æ©Ÿæ¢°å¦ç¿’

python
ãƒªãƒ³ã‚¯
æ–°ã—ã„AWS Well-Architected Machine Learning Lensã®ç´¹ä»‹ | Amazon Web Services
Amazon Web Services ãƒ–ãƒã‚° æ–°ã—ã„AWS Well-Architected Machine Learning Lensã®ç´¹ä»‹ ã“ã®è¨˜äº‹ã¯ã€ã€ŒIntroducing the new AWS Well-Architected Machine Learning Lensã€ã‚’ç¿»è¨³ã—ãŸã‚‚ã®ã§ã™ã€‚ AWS Well-Architected Frameworkã¯ã€ãŠå®¢æ§˜ã®ãƒ¯ãƒ¼ã‚¯ãƒãƒ¼ãƒ‰ã‚’ãƒ™ã‚¹ãƒˆãƒ—ãƒ©ã‚¯ãƒ†ã‚£ã‚¹ã¨æ¯”è¼ƒã™ã‚‹ãŸã‚ã®ä¸€è²«ã—ãŸã‚¢ãƒ—ãƒãƒ¼ãƒã‚’æä¾›ã—ã¾ã™ã€‚ã¾ãŸã€ã©ã®ã‚ˆã†ã«æ”¹å–„ã™ã¹ãã‹ã®æŒ‡é‡ã‚‚å«ã¾ã‚Œã¦ã„ã¾ã™ã€‚ æ©Ÿæ¢°å¦ç¿’ï¼ˆMLï¼‰ã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã¯ã€ãƒ‡ãƒ¼ã‚¿ã®ãƒ‘ã‚¿ãƒ¼ãƒ³ã‚’ç™ºè¦‹ãŠã‚ˆã³å¦ç¿’ã—ã€å°†æ¥ã®ãƒ‡ãƒ¼ã‚¿ã‚’äºˆæ¸¬ã™ã‚‹æ•°ç†ãƒ¢ãƒ‡ãƒ«ã‚’æ§‹ç¯‰ã—ã¾ã™ã€‚ã“ã‚Œã‚‰ã®ã‚½ãƒªãƒ¥ãƒ¼ã‚·ãƒ§ãƒ³ã¯ã€ç—…æ°—ã®è¨ºæ–ã®æ”¹å–„ã€ç’°å¢ƒä¿è·ã€è£½å“ã‚„ã‚µãƒ¼ãƒ“ã‚¹ã®å¤‰é©ãªã©ã€ç”Ÿæ´»ã«é©å‘½ã‚’ã‚‚ãŸã‚‰ã—ã¾ã™ã€‚ MLãƒ¢ãƒ‡ãƒ«ãŒæ£ç¢ºãªçµæžœã‚’å‡ºã™ãŸã‚ã«ã¯ã€å…¥åŠ›ãƒ‡ãƒ¼ã‚¿ã®è³ªãŒé‡è¦ã§ã™ã€‚
clavier 2021/11/04
aws

æ©Ÿæ¢°å¦ç¿’

machinelearning
ãƒªãƒ³ã‚¯
Apache Airflowã§ã‚¨ãƒ³ãƒ‰ãƒ¦ãƒ¼ã‚¶ãƒ¼ã®ãŸã‚ã®æ©Ÿæ¢°å¦ç¿’ãƒ‘ã‚¤ãƒ—ãƒ©ã‚¤ãƒ³ã‚’æ§‹ç¯‰ã™ã‚‹ Part5 (End)
We organized Japanese financial reports to encourage applying NLP techniques to financial analytics. You can downloadâ€¦ Part4ã‹ã‚‰ãšã„ã¶ã‚“é–“ãŒç©ºãã¾ã—ãŸãŒã€ãã®é–“ã«ã¯ãƒ‡ãƒ¼ã‚¿å…¬é–‹ã«ã¾ã¤ã‚ã‚‹ã‚‚ã‚ã‚‚ã‚ã®èª¿æ•´ãªã©ãŒã‚ã£ãŸã¨ã„ã†ã€‚ Airflowã‚’æŽ¡ç”¨ã—ãªã‹ã£ãŸç†ç”±æœ€çµ‚çš„ã«Airflowã‚’æŽ¡ç”¨ã—ãªã‹ã£ãŸç†ç”±ã¯2ã¤ã‚ã‚Šã¾ã™ã€‚ é‹ç”¨ã‚³ã‚¹ãƒˆé–‹ç™ºã‚³ã‚¹ãƒˆé‹ç”¨ã‚³ã‚¹ãƒˆ Part3ã§ã‚‚è§¦ã‚Œã¾ã—ãŸãŒã€Airflowã®ãƒ›ã‚¹ãƒ†ã‚£ãƒ³ã‚°ã¯çµæ§‹é«˜ãã¤ãã¾ã™ã€‚ãƒ›ã‚¹ãƒ†ã‚£ãƒ³ã‚°ã‚µãƒ¼ãƒ“ã‚¹ã‚’æä¾›ã—ã¦ã„ã‚‹ã®ã¯GCPã®Cloud Composerã¨Astronomerã®2ã¤ãŒä¸»ã§ã™ã€‚Astronomerã®å ´åˆã¯æœˆé¡$100ã¾ã§æŠ‘ãˆã‚‹ã“ã¨ãŒå¯èƒ½ã§ã™ãŒã€å›ºå®šã§æ¯Žæœˆã‹ã‹ã‚‹ã¨ãªã‚‹ã¨ãã“ãã“ã®é‡‘é¡ã§ã™ã€‚ ã‚¹ã‚±ã‚¸ãƒ¥ãƒ¼ãƒ©ãƒ¼ã¯ã€ã‚¹ã‚±ã‚¸ãƒ¥ãƒ¼
clavier 2019/09/27
æ©Ÿæ¢°å¦ç¿’

AWS

NLP

GCP

pipeline

data

Airflow

kaggle
ãƒªãƒ³ã‚¯
å¤ç‹¬ã§ã¤ã‚‰ã„æ©Ÿæ¢°å¦ç¿’ãƒ‡ãƒ¼ã‚¿ã®ä½œæˆã‚’ã€ãƒ¯ã‚¤ãƒ¯ã‚¤æ¥½ã—ãã™ã‚‹ç’°å¢ƒã‚’3æ—¥ã§ä½œã‚‹ - Qiita
21ä¸–ç´€ã§ã‚‚ã£ã¨ã‚‚ã‚»ã‚¯ã‚·ãƒ¼ã¨è¨€ã‚ã‚Œã‚‹ãƒ‡ãƒ¼ã‚¿ã‚µã‚¤ã‚¨ãƒ³ã‚¹ã¯ã€18ä¸–ç´€ã‚’å½·å½¿ã¨ã•ã›ã‚‹å¥´éš·çš„ãªä½œæ¥ã«ã‚ˆã£ã¦æ”¯ãˆã‚‰ã‚Œã¦ã„ã¾ã™(è¦å‡ºå…¸)ã€‚ãã®ä½œæ¥ã¨ã¯ã€ãƒ‡ãƒ¼ã‚¿ã‚’ä½œã‚‹ä½œæ¥(=ã‚¢ãƒŽãƒ†ãƒ¼ã‚·ãƒ§ãƒ³)ã§ã™ã€‚å¤šãã®å ´åˆã€ã‚¢ãƒŽãƒ†ãƒ¼ã‚·ãƒ§ãƒ³ã¯å¤ç‹¬ãªå˜èª¿ä½œæ¥ã®ç¹°ã‚Šè¿”ã—ã§ã™ã€‚å»¶ã€…ã¨ç¶šã‘ã¦ã„ã‚‹ã¨ã€å…¨ã¦ã®å˜èªžã«unkã¨ã¤ã‘ã‚‹ã‚ˆã†ã«ãªã‚‹äº‹ä¾‹ã‚‚å ±å‘Šã•ã‚Œã¦ã„ã¾ã™ã€‚ã¤ã¾ã‚Šã¤ã‚‰ã„ï¼ã®ã§ã™ã€‚ æœ¬è¨˜äº‹ã§ã¯ã€ã€Œå¤ç‹¬ã§è¾›ã„ã€ã‚¢ãƒŽãƒ†ãƒ¼ã‚·ãƒ§ãƒ³ã‚’ã€Œã¿ã‚“ãªã§æ¥½ã—ãã€è¡Œãˆã‚‹ç’°å¢ƒã‚’ä½œã‚‹æ–¹æ³•ã‚’ç´¹ä»‹ã—ã¾ã™ã€‚ãã®ãƒã‚¤ãƒ³ãƒˆã¯ä»¥ä¸‹3ç‚¹ã§ã™ã€‚ Easy: ã‚¢ãƒŽãƒ†ãƒ¼ã‚·ãƒ§ãƒ³ã‚’æ¥½ã« Feedback: ã‚¢ãƒŽãƒ†ãƒ¼ã‚·ãƒ§ãƒ³ã—ãŸçµæžœãŒã™ãã«ã‚ã‹ã‚‹ã‚ˆã†ã« Gamification: ã‚¢ãƒŽãƒ†ãƒ¼ã‚·ãƒ§ãƒ³çµæžœã‚’ã¿ã‚“ãªã§ç«¶ã„åˆã† Overview ã‚¢ãƒŽãƒ†ãƒ¼ã‚·ãƒ§ãƒ³ã‚’è¡Œã†æµã‚Œã¯ä»¥ä¸‹ã®ã‚ˆã†ã«ãªã‚Šã¾ã™ã€‚ Annotation Tool for Easy: ã‚¢ãƒŽãƒ†ãƒ¼ã‚·ãƒ§ãƒ³ã‚’æ¥½ã«ã™ã‚‹ãŸã‚ã®ãƒ„ãƒ¼ãƒ«ã‚’ç”¨æ„ In
clavier 2019/09/02
æ©Ÿæ¢°å¦ç¿’

lambda

ecs

AWS
ãƒªãƒ³ã‚¯
è¶£å‘³ã§ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã™ã‚‹ãŸã‚ã® GPU ç’°å¢ƒã‚’å®‰ä¸ŠãŒã‚Šã«ä½œã‚‹æ–¹æ³• - Qiita
è¶£å‘³ã§ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã§éŠã¶ãŸã‚ã« GPU ãƒžã‚·ãƒ³ã‚’ä½¿ã„ãŸã„ã€‚ GPU ã¯æœ¬å½“ã«ã™ã”ãã¦ã€è‡ªåˆ†ã® MacBook Air ã§ 2 æ™‚é–“ã‹ã‹ã‚‹ã‚ˆã†ãªå¦ç¿’ãŒ GPU ã‚’ä½¿ã†ã¨ 5 åˆ†è¶³ã‚‰ãšã§çµ‚ã‚ã‚‹ã€‚CPU ã ã‘ã§ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ã™ã‚‹ã®ã¯é¦¬é¹¿é¦¬é¹¿ã—ã„ã¨ã„ã†æ°—æŒã¡ã«ãªã‚‹ã€‚ ã—ã‹ã—è‡ªå®…ã« GPU ãƒžã‚·ãƒ³ã‚’çµ„ã‚“ã ã¨ã—ã¦ã‚‚å››å…æ™‚ä¸è¨ˆç®—ã—ç¶šã‘ã‚‹ã‚ã‘ã§ã‚‚ãªã„ã‹ã‚‰ã‚‚ã£ãŸã„ãªã„ã—ã€ã“ã“ã¯ã‚¯ãƒ©ã‚¦ãƒ‰ã‚µãƒ¼ãƒ“ã‚¹ã‚’æ´»ç”¨ã—ã¦å®‰ãæ¸ˆã¾ã›ãŸã„ã€‚1 ã•ã‚‰ã«ã€æœ€è¿‘ã§ã¯ Docker ã‚³ãƒ³ãƒ†ãƒŠå†…ã‹ã‚‰ GPU ã‚’åˆ©ç”¨ã™ã‚‹ã“ã¨ãŒã§ãã‚‹ NVIDIA Docker ã¨ã„ã† Docker ãƒ—ãƒ©ã‚°ã‚¤ãƒ³ãŒã‚ã‚Šã€ã“ã‚Œã‚’åˆ©ç”¨ã™ã‚‹ã“ã¨ã§ GPU ãƒžã‚·ãƒ³ã®ç’°å¢ƒã‚’æ±šã™ã“ã¨ãªãå¥½ããªã ã‘å¦ç¿’ç’°å¢ƒã‚’ç«‹ã¡ä¸Šã’ã‚‹ã“ã¨ãŒã§ãã‚‹ã€‚ ä»Šå›žã¯ Amazon EC2 ã® GPU ã‚¤ãƒ³ã‚¹ã‚¿ãƒ³ã‚¹ã¨ NVIDIA Docker ã‚’ä½¿ã£ã¦è¶£å‘³ç”¨ã®ãƒ‡ã‚£ãƒ¼ãƒ—ãƒ©ãƒ¼ãƒ‹ãƒ³ã‚°ç’°å¢ƒã‚’ä½œã£ãŸ
clavier 2017/06/08
AWS

deeplearning

æ©Ÿæ¢°å¦ç¿’
ãƒªãƒ³ã‚¯
[ãƒ¬ãƒãƒ¼ãƒˆ] ä»Šæ—¥ã‹ã‚‰å§‹ã‚ã‚‰ã‚Œã‚‹ã€æ©Ÿæ¢°å¦ç¿’ï¼ Amazon Machine Learningã®ã”ç´¹ä»‹ #AWSSummit | DevelopersIO
å¹³ç”°ã§ã™ã€‚ AWS Summit Tokyo ã§ Amazon Machine Learning ã«ã¤ã„ã¦ã®ãƒ©ãƒ³ãƒã‚»ãƒƒã‚·ãƒ§ãƒ³ã«å‚åŠ ã—ã¦ããŸã®ã§ã€ãã®æ§˜åã‚’ãƒ¬ãƒãƒ¼ãƒˆã—ã¾ã™ã€‚ å¾ŒåŠã¯ãƒ‡ãƒ¢ãŒãƒ¡ã‚¤ãƒ³ã ã£ãŸã®ã§ã€å‰åŠã®å†…å®¹ã«ã¤ã„ã¦ã®ãƒ¬ãƒãƒ¼ãƒˆã¨ãªã‚Šã¾ã™ã€‚ ã‚»ãƒƒã‚·ãƒ§ãƒ³æƒ…å ± ã‚»ãƒƒã‚·ãƒ§ãƒ³å : ä»Šæ—¥ã‹ã‚‰å§‹ã‚ã‚‰ã‚Œã‚‹ã€æ©Ÿæ¢°å¦ç¿’ï¼ Amazon Machine Learningã®ã”ç´¹ä»‹ ã‚¹ãƒ”ãƒ¼ã‚« ä»Šäº• é›„å¤ªï¼ˆã‚¢ãƒžã‚¾ãƒ³ ãƒ‡ãƒ¼ã‚¿ ã‚µãƒ¼ãƒ“ã‚¹ ã‚¸ãƒ£ãƒ‘ãƒ³æ ªå¼ä¼šç¤¾ ã‚½ãƒªãƒ¥ãƒ¼ã‚·ãƒ§ãƒ³ã‚¢ãƒ¼ã‚ãƒ†ã‚¯ãƒˆï¼‰ æ¦Žä¸¦ åˆ©æ™ƒï¼ˆã‚¢ãƒžã‚¾ãƒ³ ãƒ‡ãƒ¼ã‚¿ ã‚µãƒ¼ãƒ“ã‚¹ ã‚¸ãƒ£ãƒ‘ãƒ³æ ªå¼ä¼šç¤¾ ã‚½ãƒªãƒ¥ãƒ¼ã‚·ãƒ§ãƒ³ã‚¢ãƒ¼ã‚ãƒ†ã‚¯ãƒˆï¼‰ Amazon Machine Learning ã«ã¤ã„ã¦ æ©Ÿæ¢°å¦ç¿’ã¯ãªã«ã‚’ã™ã‚‹ã®ã‹ï¼Ÿ éŽåŽ»ã®ãƒ‡ãƒ¼ã‚¿ã‚’ä½¿ã£ã¦çµæžœã‚’äºˆæ¸¬ã™ã‚‹ãƒ¢ãƒ‡ãƒ«ã‚’æ§‹ç¯‰ã™ã‚‹ æ§‹ç¯‰ã—ãŸãƒ¢ãƒ‡ãƒ«ã‚’ä½¿ã£ã¦ã€æœªæ¥ã®äºˆæ¸¬ã‚’è¡Œã† Amazon Machine Learning ãŒã§ãã‚‹ã“ã¨ æ©Ÿæ¢°å¦
clavier 2015/06/04
AWS

æ©Ÿæ¢°å¦ç¿’
ãƒªãƒ³ã‚¯
1