Notes de version

Cette page répertorie les mises à jour en production d'AutoML Natural Language. Consultez-la régulièrement pour obtenir des informations sur les fonctionnalités nouvelles ou mises à jour, les corrections de bugs, les problèmes connus et les fonctionnalités obsolètes.

Vous pouvez consulter les dernières mises à jour de produits pour l'ensemble de Google Cloud sur la page Notes de version de Google Cloud.

Pour recevoir les dernières mises à jour concernant ce produit, ajoutez l'URL de cette page à votre lecteur de flux ou ajoutez l'URL du flux directement : https://cloud.google.com/feeds/automl-language-release-notes.xml

April 03, 2020

Integration with VPC Service Controls is now in beta stage.

March 20, 2020

AutoML Natural Language now supports TIFF files as training data and input for predictions.

AutoML Natural Language now supports classification and sentiment analysis in 20 languages.

January 30, 2020

AutoML Natural Language now supports entity analysis in over 100 languages.

December 12, 2019

AutoML Natural Language General Availability (GA) release

AutoML Natural Language enables you to train three types of custom machine learning models:

  • AutoML Text & Document Classification creates models to classify documents according to labels that you define.
  • AutoML Text & Document Entity Extraction creates models to identify a custom set of entities within documents.
  • AutoML Text & Document Sentiment Analysis creates models to analyze attitudes within text.

The GA release consolidates these features, which were documented as separate products during beta testing. See the AutoML Natural Language documentation for details about all features.

October 30, 2019

AutoML Natural Language offers you control over where your datasets and models are stored and where machine learning takes place. In particular, you can configure AutoML Natural Language to at rest and perform machine learning processing in the European Union only. See Locations for more information.

October 21, 2019

AutoML Natural Language Text Classification now supports batch prediction.

AutoML Natural Language Text Classification can now classify PDF files and also supports PDF files as training items.

September 04, 2019

AutoML Natural Language Entity Extraction offers support for document-based entity extraction, which enables you to identify entities in a format that recreates the basic layout of the training item. When you add annotations, AutoML Natural Language Entity Extraction captures the annotation's position on the page as a factor considered during training. See Managing datasets for more information.

April 10, 2019

AutoML Natural Language Entity Extraction Beta has been released.

AutoML Natural Language Sentiment Analysis Beta has been released.

Microsoft Edge and Microsoft Internet Explorer do not support all features of AutoML Natural Language Entity Extraction. If you are having problems, try Google Chrome, Safari, or Firefox.

Microsoft Edge and Microsoft Internet Explorer do not support all features of AutoML Natural Language Sentiment Analysis. If you are having problems, try Google Chrome, Safari, or Firefox.

January 22, 2019

Beginning January 22, 2019, inactive models are subject to automatic undeployment. An inactive model is one that has not been used for prediction in 60 days. An undeployed model is not available for use until you explicitly redeploy it using a method that will be made available in advance of any undeployments.

November 01, 2018

The AutoML Natural Language Text Classification service account requires additional permissions to access project resources when using the AutoML Natural Language Text Classification UI. You must add the role serviceusage.serviceUsageAdmin. See step 8 in Before You Begin.

August 24, 2018

The URL for the AutoML Natural Language Text Classification UI has changed to https://console.cloud.google.com/natural-language. The previous URL https://beta-dot-custom-vision.appspot.com/text/ will remain active.

July 24, 2018

AutoML Natural Language Text Classification Beta Release

Models created during the EAP will continue to be valid for the Beta. You do not need to recreate/retrain models. However, the format of resource IDs (dataset IDs, model IDs, and operation IDs) is changing. If you are using hardcoded resource IDs, you need to obtain new IDs for those resources (using the get or list method) by August 31, 2018.

Note: Enabling the AutoML API for your project automatically also enables the Cloud Vision API, Cloud Translation API, and Cloud Natural Language API. Disabling any of these three APIs will also disable the AutoML API.

June 07, 2018

The earlier version of the AutoML Natural Language Text Classification UI has been deprecated. Attempts to connect to the older UI are redirected to the new UI.

May 21, 2018

The refresh of the AutoML Natural Language Text Classification EAP release introduces a new UI design and a new API.

April 02, 2018

AutoML Natural Language Text Classification has changes to its modeling logic. You may notice an increase in training latency as well as potential improvements in training performance.

February 07, 2018

AutoML Natural Language Text Classification EAP release