投稿時間:2022-03-30 01:40:27 RSSフィード2022-03-30 01:00 分まとめ(47件)

カテゴリー等	サイト名等	記事タイトル・トレンドワード等	リンクURL	頻出ワード・要約等/検索ボリューム	登録日
AWS	AWS Japan Blog	クラウドの調達と、その導入。両面に特化した新資格の普及をAWSは支援。	https://aws.amazon.com/jp/blogs/news/new-axelos-itil-4-extension-module-develops-cloud-procurement-and-technology-integration-skills/	今回のブログでは、AWSジャパン・パブリックセクターより、「クラウドの調達にフォーカスした新しい認証資格」について紹介します。	2022-03-29 15:04:48
AWS	AWS	Toyota Motor NA: Lake House Architecture for Data and Analytics	https://www.youtube.com/watch?v=Kp51k6LY-2c	Toyota Motor NA Lake House Architecture for Data and AnalyticsToyota manufactures and distributes cars worldwide that are meticulously logged in vehicle data warehouses and data marts to provide insights on service checks and sales reports Toyota migrated and modernized its on premises vehicle data warehouse and data marts by leveraging a Lake House architecture on AWS Data is securely migrated from the source systems on premises to AWS using AWS Direct Connect and ingested in data integration tools running on Amazon EC such as Kafka Talend and Databricks Data is pushed into different layers for transformation with Amazon S AWS Glue and Amazon RDS To obtain read optimized data data is ingested into Amazon Redshift Amazon Redshift Spectrum and Amazon Athena are leveraged by Toyota s Data Scientists Analysts and Business user teams to produce self service reports With this Lake House architecture Toyota has improved their time to market and eliminated the duplication of data Check out more resources for architecting in the AWS​​​cloud ​ AWS AmazonWebServices CloudComputing ThisIsMyArchitecture	2022-03-29 15:31:43
python	Pythonタグが付けられた新着投稿 - Qiita	ユーザーがアップロードした画像をAWS S3に保存して、ユーザー本人だけアクセスできるようにする	https://qiita.com/yagrush/items/d0834003c705528a9c2b	その他、とりあえずこれで一旦CloudFront作成。	2022-03-30 00:57:13
python	Pythonタグが付けられた新着投稿 - Qiita	Jupyter NotebookのSnippetを作成してライブラリのimportを楽にする試み(Jupyter @ Dockerでの方法)	https://qiita.com/y-vectorfield/items/49692a5501e630f8fedc	そこで、NotebookのテンプレートSnippetを設定して、import文の記載を楽にする試みをやってみることにしました。	2022-03-30 00:29:53
python	Pythonタグが付けられた新着投稿 - Qiita	一問一答の自作問題集を作成する。（試験対策）	https://qiita.com/pyama2022/items/ad6fde9774e5319e921e	一問一答の自作問題集を作成する。	2022-03-30 00:12:06
python	Pythonタグが付けられた新着投稿 - Qiita	AtCoder ABC245 F - Endless Walk を SCC+トポロジカルソートで解く	https://qiita.com/recuraki/items/4366e3c29f487bde1f6d	そのような連結成分の点aに移動してから、のように移動を繰り返すことができますつまり、Gscc上で考えたときサイズ以上の連結成分に含まれる元の頂点たちはすべて条件を満たすサイズの頂点はサイズ以上の頂点に到達可能のとき、元の頂点は条件を満たすサイズの頂点はサイズの頂点にしか到達できないとき、元の頂点は条件を満たさない次の移動で必ずほかの連結成分に移動しなければならないが、GsccはDAGであるため、いずれサイズで出次の点に達し、それ以上の移動はできないため、条件を満たせない。	2022-03-30 00:08:56
AWS	AWSタグが付けられた新着投稿 - Qiita	ユーザーがアップロードした画像をAWS S3に保存して、ユーザー本人だけアクセスできるようにする	https://qiita.com/yagrush/items/d0834003c705528a9c2b	その他、とりあえずこれで一旦CloudFront作成。	2022-03-30 00:57:13
Ruby	Railsタグが付けられた新着投稿 - Qiita	RailsにVue3を入れて、webpack-dev-serverで失敗　"export 'default' (imported as 'Vue') was not found in 'vue'	https://qiita.com/Annmann3/items/9c99b0f9df9861da548b	RailsにVueを入れて、webpackdevserverで失敗quotexportxdefaultximportedasxVuexwasnotfoundinxvuex環境RailsVuejs背景binwebpackdevserverを実行して以下のエラーが出た。	2022-03-30 00:04:45
海外TECH	Ars Technica	Asus will lower some GPU prices up to 25% following tariff changes	https://arstechnica.com/?p=1844092	promising	2022-03-29 15:46:58
海外TECH	MakeUseOf	How to Extract Images From a PDF and Use Them Anywhere	https://www.makeuseof.com/tag/extract-images-pdf-files-save-windows/	anywherehere	2022-03-29 16:00:15
海外TECH	MakeUseOf	How to Cancel Your YouTube Music Subscription	https://www.makeuseof.com/how-to-cancel-youtube-music-subscription/	cancel	2022-03-29 16:00:14
海外TECH	MakeUseOf	Microsoft Just Made Clipchamp Suck a Little Less	https://www.makeuseof.com/microsoft-clipchamp-price-tier-changes/	videos	2022-03-29 15:51:15
海外TECH	MakeUseOf	How to Search for Packages in Ubuntu	https://www.makeuseof.com/how-to-search-for-packages-ubuntu/	ubuntu	2022-03-29 15:45:14
海外TECH	MakeUseOf	How to Install Two Copies of the Same App in Android	https://www.makeuseof.com/run-two-copies-of-app-android/	party	2022-03-29 15:45:15
海外TECH	MakeUseOf	8 Tips to Give Your Smartphone a Security Spring Clean	https://www.makeuseof.com/give-smartphone-security-spring-clean/	security	2022-03-29 15:30:14
海外TECH	MakeUseOf	11 File Explorer Tips and Tricks for Windows 11	https://www.makeuseof.com/windows-11-file-explorer-tips/	windows	2022-03-29 15:15:14
海外TECH	DEV Community	Create An Open Source Ecommerce Store with Gatsby and Medusa	https://dev.to/medusajs/create-an-open-source-ecommerce-store-with-gatsby-and-medusa-2h6b	Create An Open Source Ecommerce Store with Gatsby and MedusaMedusa is an open source headless commerce platform that s focused on providing a great developer experience while creating an ecommerce store Using Medusa s headless architecture you can create a storefront with any programming language or framework you prefer just by connecting it to the REST APIs Your storefront will be able to display products and will have the full ecommerce checkout workflow In this tutorial you ll learn how to connect our Gatsby starter to your Medusa server FeaturesThe Gatsby storefront comes with a lot of features out of the box here are some of them View all products and manage your cart Customer authentication Full checkout workflow Request swaps and returns using a customer s order ID and Email PrerequisitesThis article assumes you already have a Medusa server installed If you don t you can follow this quickstart guide to get you started in minutes You also need to have the Gatsby CLI tool installed as you ll be using it to create the Gatsby store You can install it with the following command npm install g gatsby cli Create Gatsby StorefrontIn your terminal run the following command to create the Medusa storefront with Gatsby gatsby new my medusa store This creates a Gatsby website using the already existing Medusa Gatsby starter Set Environment VariablesTo connect to the backend you need to set the URL that leads to the backend in the environment variables The Gatsby starter comes with the file env template that you can use to add your environment variables First rename the file mv env template env developmentThis changes the name to env development You can also rename it to env production for production use or just rename it to env This file includes environment variables GATSBY MEDUSA BACKEND URL http localhost GATSBY STRIPE KEY pk test somethingThe first one is the URL to the Medusa server By default the Medusa server runs on localhost So if you didn t change that you can leave it as is The second one is your Stripe public key if you have Stripe set up If you don t you can leave it as is Test it OutBefore you run your storefront make sure that your Medusa server is running Then in your terminal in your Gatsby storefront start the storefront with the following command npm startThe storefront should run on localhost Note About DevelopmentIt s worth noting that the Gatsby starter makes use of the gatsby source medusa plugin to source data from your Medusa server This data includes products collections and regions and as a result you are able to query this data in the starter using GraphQL queries You can also explore the data in your store on localhost graphql As a result of this you will need to rebuild the site every time you update any of this data for it to be reflected in your storefront We will soon be releasing a new version of the plugin which adds incremental builds which will improve build times but more on that at a later time ConclusionIn this tutorial you learned how to get a Gatsby store up and running Your storefront includes products listing cart management checkout workflow an option for customers to request returns and exchanges for their previous orders customer login and customer order history Should you have any issues or questions related to Medusa then feel free to reach out to the Medusa team via Discord	2022-03-29 15:56:19
海外TECH	DEV Community	Lasson 04 | Loops and errors	https://dev.to/ilosrim/lasson-04-loops-and-errors-4f5p	Lasson Loops and errors Types of LoopsLoops can be written as while loops do while loops and for loops while Loopswhile loops iterate until a condition is met int a while a lt a do while Loopsdo while loops are while loops that initially execute the body once before checking the condition do printf not true while for Loopsfor loops complete a set number of iterations before meeting a condition for int i i lt i printf Hello Loop KeywordsAll loops can utilize keywords like continue and break continue restarts the loop and break breaks out of or ends the loop Rewriting LoopsA for loop can always be re written as a while loop most while loops can be re written as a for loop	2022-03-29 15:40:55
海外TECH	DEV Community	css helper inspect alternative	https://dev.to/amreldessouki/css-helper-inspect-alternative-3g3i	css helper inspect alternativeWith this tool you can edit your CSS style and copy your code a powerful Chrome Extension was created to help web developers make CSS editing easy for anyone beginner or professional Copy and paste all your editing code into your project easily This tool does not work directly In your project it helps you to see changes in real time and copies all your code changes and also you can choose any element to modify by clicking inspect to choose your CSS class or ID to change styleI hope it helps youproject product hunt link	2022-03-29 15:34:34
Apple	AppleInsider - Frontpage News	How to clear the cache in Safari on Mac and iPhone - and what it does	https://appleinsider.com/inside/safari/tips/how-to-clear-the-cache-in-safari-on-mac-and-iphone---and-what-it-does?utm_medium=rss	How to clear the cache in Safari on Mac and iPhone and what it doesSafari s cache makes browsing quicker but it can go wrong it can slow down your Mac iPhone or iPad and it can take up a lot of storage space too Here s how and when to clear it Like every web browser Safari stores data on your Mac iPhone and iPad with the express purpose of speeding up your online life You may have the kind of internet speed that Apple Park takes for granted but even so it s better to spend that speed where it s needed So when you revisit a website Safari makes it so that as far as possible you only need to download what has changed Even with a news site that is being updated all day the basic layout the furniture of the page doesn t change all that often so Safari might as well download it once Read more	2022-03-29 15:24:44
Apple	AppleInsider - Frontpage News	Apple's MacBook Pro 14-inch is discounted to $1,849 ($150 off), AppleCare $60 off, units in stock	https://appleinsider.com/articles/22/02/28/apples-macbook-pro-14-inch-is-discounted-to-1849-150-off-applecare-60-off-units-in-stock?utm_medium=rss	Apple x s MacBook Pro inch is discounted to off AppleCare off units in stockApple s standard inch MacBook Pro is back in stock and off for March Plus save on AppleCare with our exclusive promo code inch MacBook Pro in stock nowRinging in as the lowest inch MacBook Pro price available according to our Price Guide the standard configuration is back in stock and on sale for with promo code APINSIDER using this activation link with the instructions below The M Pro model has an core CPU core GPU GB of memory and a GB SSD Read more	2022-03-29 15:21:11
海外TECH	Engadget	Kia's EV9 SUV will move from concept to reality in 2023	https://www.engadget.com/kia-ev9-suv-release-date-152842660.html?src=rss	Kia x s EV SUV will move from concept to reality in If you thought it wouldn t be long before Kia turned the Concept EV into a production model you guessed correctly The automaker has revealed that a road ready version of the electric SUV will be available in Europe in There was no mention of launches in North America or other regions but it s an SUV ーit may just be a matter of time before you see the EV cruising American and Canadian streets Kia didn t say what would change in the transition from concept to production However we d expect the badge to cut many of the more exotic features including the yoke giant wheels retractable roof rails and lounge like seating modes We wouldn t be surprised if Kia kept the inch display and even the hood mounted solar panel though The production EV might also preserve the claimed specs The concept promised up to miles of range and kW fast charging that could take it from a percent charge to percent in minutes Kia also recently detailed autonomous quot Automode quot technology for the EV that can take over from the driver on the highway There are still important unknowns like pricing Even so the EV could be one of Kia s most important all electric vehicles to date at least in some areas While the EV has been well received so far some markets particularly North America skew heavily toward crossovers and SUVs The EV could help Kia take on competitors like the Tesla Model Y and Volkswagen ID not to mention reel in buyers who haven t been thrilled by the Niro EV	2022-03-29 15:28:42
Cisco	Cisco Blog	Modern Transit Architecture with Cisco SD-WAN Cloud OnRamp for Azure Virtual WAN	https://blogs.cisco.com/networking/modern-transit-architecture-with-cisco-sd-wan-cloud-onramp-for-azure-virtual-wan	Modern Transit Architecture with Cisco SD WAN Cloud OnRamp for Azure Virtual WANThe Cisco SD WAN Cloud OnRamp with Azure Virtual WAN integration enables you to seamlessly extend your SD WAN overlay into Azure and improves cloud connectivity from branch offices	2022-03-29 15:00:45
海外TECH	CodeProject Latest Articles	How to easily map between entities and DTOs when using EntityFrameworkCore	https://www.codeproject.com/Tips/5328579/How-to-easily-map-between-entities-and-DTOs-when-u	entityframeworkcoreto	2022-03-29 15:37:00
海外科学	NYT > Science	How to Find a Therapist or Mental Health Provider	https://www.nytimes.com/2022/03/29/well/mind/therapist-mental-health-provider.html	providertips	2022-03-29 15:15:26
海外科学	NYT > Science	How 2 Industries Stymied Justice for Young Lead Paint Victims	https://www.nytimes.com/2022/03/29/us/lead-poisoning-insurance-landlords.html	How Industries Stymied Justice for Young Lead Paint VictimsThe U S insurance and real estate industries have waged a decades long campaign to avoid liability in lead cases helping to prolong an epidemic The cost for millions of children has been incalculable	2022-03-29 15:45:55
海外科学	NYT > Science	Review: ‘Vagina Obscura,’ by Rachel E. Gross	https://www.nytimes.com/2022/03/29/books/vagina-obscura-rachel-gross.html	bodies	2022-03-29 15:37:38
海外科学	NYT > Science	In Wisconsin: Stowing Mowers, Pleasing Bees	https://www.nytimes.com/2022/03/28/travel/no-mow-may-wisconsin.html	grass	2022-03-29 15:22:19
金融	◇◇　保険デイリーニュース　◇◇（損保担当者必携！）	保険デイリーニュース(03/30)	http://www.yanaharu.com/ins/?p=4873	三井住友海上	2022-03-29 15:43:37
金融	RSS FILE - 日本証券業協会	パブリックコメントの募集の結果について	https://www.jsda.or.jp/about/public/kekka/index.html	募集	2022-03-29 15:04:00
金融	金融庁ホームページ	「拠点開設サポートオフィス」について更新しました。	https://www.fsa.go.jp/policy/marketentry/index.html	開設	2022-03-29 17:00:00
金融	金融庁ホームページ	国際金融センター特設ページを更新しました。	https://www.fsa.go.jp/internationalfinancialcenter/index.html	alfinancialcenterjapan	2022-03-29 17:00:00
金融	金融庁ホームページ	「金融商品取引業等に関する内閣府令の一部を改正する内閣府令（案）」等に対するパブリックコメントの結果等について公表しました。	https://www.fsa.go.jp/news/r3/shouken/20220329/20220329.html	内閣府令	2022-03-29 17:00:00
金融	金融庁ホームページ	株式会社ゆうちょ銀行における新規業務の認可について公表しました。	https://www.fsa.go.jp/news/r3/yuusei/20220329.html	株式会社ゆうちょ銀行	2022-03-29 16:00:00
金融	金融庁ホームページ	OECDによる報告書「グループ会社のコーポレートガバナンスに関するアジア地域の優れた政策・取組み」について掲載しました。	https://www.fsa.go.jp/inter/etc/20220329/oecd_report_good_good_policies_and_practices_for_cg.html	政策	2022-03-29 16:00:00
ニュース	ジェトロ ビジネスニュース（通商弘報）	ベルリンでECバイヤーと連携し宮城県産「復興」日本酒をPR・販売	https://www.jetro.go.jp/biznews/2022/03/dd1045f4599e3e92.html	宮城県産	2022-03-29 15:40:00
ニュース	ジェトロ ビジネスニュース（通商弘報）	米テスラ、ドイツの「ギガファクトリー」を正式に開所	https://www.jetro.go.jp/biznews/2022/03/6f5dc3eeb0fd9ba7.html	正式	2022-03-29 15:30:00
ニュース	ジェトロ ビジネスニュース（通商弘報）	国連アフリカ経済委員会が域内FTAのレポートを発表、コストや手続き面での課題を指摘	https://www.jetro.go.jp/biznews/2022/03/4760445337658768.html	課題	2022-03-29 15:20:00
ニュース	ジェトロ ビジネスニュース（通商弘報）	コロナ規制緩和で実店舗へ回帰の動きも2021年のEC市場は堅調	https://www.jetro.go.jp/biznews/2022/03/4d77d7aeceab5f37.html	規制緩和	2022-03-29 15:10:00
ニュース	BBC News - Home	Shrewsbury and Telford Trust: Better care might have saved 201 babies	https://www.bbc.co.uk/news/uk-60919474?at_medium=RSS&at_campaign=KARANGA	telford	2022-03-29 15:44:40
ニュース	BBC News - Home	Twenty fines to be issued over No 10 lockdown parties	https://www.bbc.co.uk/news/uk-politics-60911798?at_medium=RSS&at_campaign=KARANGA	parties	2022-03-29 15:12:37
ニュース	BBC News - Home	University rugby player dies weeks after match injury	https://www.bbc.co.uk/news/uk-england-somerset-60905063?at_medium=RSS&at_campaign=KARANGA	march	2022-03-29 15:40:50
ニュース	BBC News - Home	Does video show Russian prisoners being shot?	https://www.bbc.co.uk/news/60907259?at_medium=RSS&at_campaign=KARANGA	prisoners	2022-03-29 15:30:58
ニュース	BBC News - Home	Barcelona to sign Chelsea's Andreas Christensen & AC Milan's Franck Kessie - Guillem Balague	https://www.bbc.co.uk/sport/football/60916864?at_medium=RSS&at_campaign=KARANGA	Barcelona to sign Chelsea x s Andreas Christensen amp AC Milan x s Franck Kessie Guillem BalagueBBC Sport s Guillem Balague says Barcelona have agreed deals to sign Chelsea defender Andreas Christensen and AC Milan midfielder Franck Kessie	2022-03-29 15:00:50
北海道	北海道新聞	岩手県で震度４　津波の心配なし	https://www.hokkaido-np.co.jp/article/662877/	津波の心配	2022-03-30 00:28:16
北海道	北海道新聞	国債利払い完了とロシア財務省　外貨建て３回目、デフォルト回避	https://www.hokkaido-np.co.jp/article/662876/	外貨建て	2022-03-30 00:20:00
GCP	Cloud Blog	Add preprocessing functions to Tensorflow models and deploy on Vertex AI	https://cloud.google.com/blog/topics/developers-practitioners/add-preprocessing-functions-tensorflow-models-and-deploy-vertex-ai/	Add preprocessing functions to Tensorflow models and deploy on Vertex AIOne does not simply productionalize machine learning models If you ve developed models before you know that most of the time input features need to be preprocessed before they are ready to be consumed by a model Often this preprocessing step is done by another application before sending the processed output to the prediction engine This adds a layer of complexity when productizing machine learning models and makes integrations difficult Fortunately TensorFlow supports preprocessing layers which can be attached to any model so that preprocessing and prediction can all be performed in the same application In this post we ll download a vision model from TensorFlow Hub attach an image preprocessing function and upload it to Vertex s prediction service which will host our model in the cloud and let us make predictions with it through a REST endpoint Not only does this make app development easier but it also provides more flexibility and ease of use Clients can send their data as is and let the model do the heavy lifting Furthermore Vertex s prediction service lets us take advantage of hardware like GPUs and performs model monitoring and autoscaling Prefer doing everything in code from a Jupyter Notebook Check out this colab Download a model from TensorFlow HubOn you ll find lots of free models that process audio text video and images In this post we ll grab the CenterNet Object and Keypoints detection model This model takes as input an image and returns object detection bounding boxes and detection keypoints Detection keypoints are used to detect object parts such as human body parts and joints On the CenterNet Object and Keypoints detection model page click “Download to grab the model in TensorFlow s SavedModel format You ll download a zipped file that contains a directory formatted like so code block StructValue u code u saved model pb r n variables r n variables data of r n variables index u language u Here the saved model pb file describes the structure of the saved neural network and the data in the variables folder contains the network s learned weights On the model s hub page you can see its example usage You feed the model an input Tensor and it spits out a dictionary with the number of detection objects the localized box and keypoint coordinates Unfortunately this model only supports Tensor like objects such as tf tensors and Numpy arrays This makes it quite difficult for clients as it requires them to create pre processing logic before sending the inputs to the model More so in other languages such as Javascript and Java TensorFlow preprocessing layersTensorFlow models contain a signature definition which defines the signature of a computation supported in a TensorFlow graph SignatureDefs aim to provide generic support to identify inputs and outputs of a function If you ve got TensorFlow installed on your computer in the directory of the Hub model you downloaded run saved model cli show dir tag set serve signature def serving defaultFor this model that command outputs We can modify this input layer with a preprocessing function so that clients can use base encoded images which is a standard way of sending images through RESTFUL APIs To do that we ll save a model with new serving signatures The new signatures use python functions to handle preprocessing the image from a JPEG to a Tensor code block StructValue u code u MODEL W PROCESSING model with processing r ndef preprocess bytes inputs r n decoded tf io decode jpeg bytes inputs channels r n resized tf image resize decoded size r n return tf cast resized dtype tf uint r n r ndef get serve image fn model r n tf function input signature tf TensorSpec None tf string r n def serve image fn bytes inputs r n decoded images tf map fn preprocess bytes inputs dtype tf uint r n return model decoded images r n return serve image fn r n r nsignatures r n serving default get serve image fn model get concrete function r n tf TensorSpec shape None dtype tf string r n r n r n r ntf saved model save model MODEL W PREPROCESSING signatures signatures u language u The base decoding is done natively by Vertex s prediction endpoint service More on that next Getting started with Vertex AIVertex AI is Google Cloud s new platform for training deploying and monitoring machine learning models and pipelines For this project we ll use the prediction service which will wrap our model in a convenient REST endpoint To get started you ll need a Google Cloud account with a GCP project set up Next you ll need to create a Cloud Storage Bucket which is where you ll upload the TensorFlow Hub model You can do this from the command line using gsutil code block StructValue u code u gcloud init Sign in to your google cloud account project r ngsutil mb mb l us central gs my model bucket Create a new storage bucket r ngsutil cp r model with preprocessing gs my model bucket model with preprocessing upload the model u language u If this model is big this could take a while In the side menu enable the Vertex AI API Once your Hub model is uploaded to Cloud Storage it s straightforward to import and deploy the model into Vertex AI using the Google Cloud aiplatform python SDK and gcloud cli commands  You can use python s standard package manager pip to install the SDK on your machine pip install upgrade google cloud aiplatformTo install the gcloud cli follow the steps outlined in this page based on your environment To create the model in Vertex AI run the following through command line code block StructValue u code u gcloud ai models upload r n region us central r n project PROJECT ID r n display name object detection r n container image uri us docker pkg dev vertex ai prediction tf cpu latest r n artifact uri BUCKET NAME model with preprocessing u language u Don t forget to change the PROJECT ID and BUCKET NAME values Once this is done you can create a Vertex AI endpoint using the gcloud cli code block StructValue u code u gcloud ai endpoints create r n project PROJECT ID r n region REGION r n display name object detection endpoint u language u Now that we have uploaded the model and created the Vertex AI endpoint let s grab the model and endpoint ids These will be used for deployment  The model id can be found from the Vertex AI console under Models Similarly the endpoint id is found under Endpoints Optionally you can run the following command to fetch these values from the command line code block StructValue u code u MODEL ID gcloud ai models list region project grep object detection r nMODEL ID MODEL ID cut d f tr d n r nENDPOINT ID gcloud ai endpoints list region project sed n p r nENDPOINT ID ENDPOINT ID cut d f tr d n u language u With these values set we deploy the model to the endpoint code block StructValue u code u gcloud ai endpoints deploy model ENDPOINT ID r n project PROJECT ID r n region REGION r n model MODEL ID r n display name object detection endpoint r n traffic split u language u This can take a while as Vertex AI provides a machine to serve the model Making predictionsWhen uploading is finished we can start making predictions against our model Vertex AI endpoints accept POST requests with a JSON body We ll base encode our image and send it to the model in the JSON body If the body contains the “b element inside then it knows to decode the image before passing it to the model  Let s try this out Download an image from the web and save it in your local environment Make sure the image is smaller than megabytes As of March Vertex AI public endpoints impose request limits of this size in order to keep containers from crashing during heavy load times Once you have an image ready you can create a request body using the command line code block StructValue u code u echo instances bytes inputs b base image jpg gt instances json u language u To test our new endpoint we can use curl to call the endpoint code block StructValue u code u curl POST r n H Authorization Bearer gcloud auth print access token r n H Content Type application json r n PROJECT ID locations us central endpoints ENDPOINT ID predict r n d instances json gt results json u language u The resulting JSON contains the response spec described in the CenterNet Object and Keypoints detection model page ​​Thetensorflow models github repository has Python libraries to visualize these results onto the images If you would like to please take a look this colab Now that we ve set our TensorFlow Hub model on Vertex we can use it in our app without having to think about most of the performance and ops challenges of using big machine learning models in production It s a nice serverless way to get building with AI fast Related ArticleServe a TensorFlow Hub model in Google Cloud with Vertex AIMake open source TensorFlow Hub models ready for production by hosting them with Google Cloud s Vertex AI Read Article	2022-03-29 16:00:00