Changelog¶
All notable changes in the Konfuzio Server will be documented according to the principles defined by Keep a Changelog.
The changelog adheres to Calendar Versioning and the release tag relates to the date and time when those changes have been released to app.konfuzio.com.
Self-hosted Konfuzio Server can be upgraded according to the documentation.
Planned¶
You can think of the Planned section as a Roadmap that lists Konfuzio Server features our team is actively working on. This list covers a planning horizon of 12 weeks.
Calculate and access Tokenizers via the web interface. The name Tokenizer might be revised. The SDK refers to this concept as Tokenizer. (Internal Ticket).
Add a filter to the list of Documents to find Documents that need to be revised by humans. (Internal Ticket).
Improved training time of Extraction AIs when using word detection mode (reduced up to 50%) (Internal Ticket).
Train custom Page-based Categorization AI via SDK (Internal Ticket).
First stable release of Capture Vue in version (Internal Ticket).
Suggest page breaks if one file contains multiple Documents (Internal Ticket).
Perfomance improvement of the Konfuzio Server interface (Internal Ticket).
Sort Labels in Label-Sets to allow users to customize the UI per Category (Internal Ticket).
Delta Training, Partial Fit an exisiting classifier, so that training documents used previously can be deleted (Internal Ticket).
Allow administrators of Konfuzio on-premise installations to run a speedtest (Internal Ticket).
Next Release (estimated release date: 6th February 2023)¶
Upcoming
released-2023-01-23_22-14-45¶
This version uses Konfuzio Trainer in version v.0.3.23, the Konfuzio Python SDK in version v.0.1.16 and Konfuzio Capture Vue in version 0.1.0.
Added¶
Changed¶
Fixed¶
Proper error handling when a Page number is used in the API which does not exist (Internal Ticket).
released-2023-01-23_14-32-08¶
This version uses Konfuzio Trainer in version v.0.3.23, the Konfuzio Python SDK in version v.0.1.16 and Konfuzio Capture Vue in version 0.1.0.
Added¶
Changed¶
released-2023-01-15_19-35-28¶
This version uses Konfuzio Trainer in version v.0.3.23, the Konfuzio Python SDK in version v.0.1.16 and Konfuzio Document Validation UI in version 0.1.0.
Added¶
Fixed¶
Restrict the maximum auto-deletion time of a Document to a maximum of 5 years (Internal Ticket).
The ordering of Annotations Sets in the API V3 (Internal Ticket).
released-2022-12-22_11-03-21¶
This version uses Konfuzio Trainer in version v.0.3.21, the Konfuzio Python SDK in version v.0.1.16 and Konfuzio Document Validation UI in version 0.0.11-pre-release-5.
Fixed¶
Re-running the categorization now also re-runs the extraction (Internal Ticket).
released-2022-12-20_11-23-04¶
This version uses Konfuzio Trainer in version v.0.3.21, the Konfuzio Python SDK in version v.0.1.16 and Konfuzio Document Validation UI in version 0.0.11-pre-release-5.
Added¶
Fixed¶
Show a message that informs a user if his account was deactivated (Internal Ticket).
Failed login attemps have not been shown in the web interface (Internal Ticket).
Missing German translation on the Member list page (Internal Ticket).
Improve performance of Document List Endpoint in API V3 by excluding large Document attributes (Internal Ticket).
released-2022-12-05_19-18-47¶
This version uses Konfuzio Trainer in version v.0.3.21, the Konfuzio Python SDK in version v.0.1.16 and Konfuzio Document Validation UI in version 0.0.11-pre-release-1.
Please note: When you upgrade to this version (or a newer one) we recommend to run “python manage.py init_email_templates” as the email templates have been updated. This needs to be run after the usual update actions.
Added¶
Access Service Desk Tickets which have been created via https://konfuzio.com/support (Internal Ticket).
API v3 endpoint to bulk accept/decline annotations (Internal Ticket).
API v3 Document endpoint can now be filtered by assignee (Internal Ticket).
Allow to configure a custom timeout for Document deletion Tasks in self-hosted environemnts (Internal Ticket).
Add ‘Reader’ as new Role for Project Members (Internal Ticket).
Changed¶
Improved Swagger Documentation for API V3 (Internal Ticket).
In the Label endpoint of API V3, rename “categories” to “label_sets” (Internal Ticket).
Detectron is now connected via API to uncouple its Python version and dependencies from Konfuzio Server (Internal Ticket).
The summarization functionality is now connected via API to uncouple its Python version and dependencies from Konfuzio Server (Internal Ticket).
Limit the number of objects returned from the API (Internal Ticket).
Fixed¶
In a very rare case text embeddings could not be extracted from Documents (Internal Ticket).
The error handling for invalid PDF Documents (Internal Ticket).
The notification email template for AI trainings was not considering errors in the training process (Internal Ticket).
callback_url is now called if re-extraction is triggered on a Document (for example, when the Category changes) (Internal Ticket).
Fix an issue that prevented the full deletion of on line of text in multiline Annotations (Internal Ticket).
Fix a missing placeholder in an email template (Internal Ticket).
Improved loading time of the Category list view (Internal Ticket).
The assignee filter for the Document List now requires Permissions to view the Members of a Project (Internal Ticket).
released-2022-11-16_12-13-49¶
Fixed¶
Speedup the Document List page for Superuser (Internal Ticket).
The Annotation creation on empty areas in a Document is now possible (Internal Ticket).
released-2022-11-11_13-19-29¶
This version uses Konfuzio Trainer in version v.0.3.21, the Konfuzio Python SDK in version v.0.1.16 and Konfuzio Document Validation UI in version 0.0.10-pre-release-7.
Please note: When you upgrade to this version (or a newer one) you need to run “python manage.py init_user_permissions”. This needs to be run after the usual update actions.
Added¶
Allow to invite Members with different Roles to Projects. Available Roles are “Reviewer” and “Manager”. All existing Members keep their current Permissions and will become Managers. (Internal Ticket).
Superusers can define custom Roles for Members: Inviting Users can select from those Roles when inviting new Members to a Project (Internal Ticket).
Add the property ‘has_multiple_top_candidates’ to the Label API V3 (Internal Ticket).
Add the property ‘has_multiple_annotation_sets’ to the Label Set API V3 (Internal Ticket).
Save feedback that there are no Annotations for a Label/Label-Set combination in a document (Internal Ticket).
Add the property ‘number’ to the Page API V3 (Internal Ticket).
Include the Label Set name in the API V3 even when no Label Set is present (Internal Ticket).
API v3 endpoint to sort and split Pages into Documents with different categories contained in one file (Internal Ticket, Internal Ticket).
Changed¶
When a user creates a new Project, this user will become the default assignee for new Documents (Internal Ticket).
If a user rejects an Annotation, this user is tracked in the ‘revised_by’ attribute of the Annotation (Internal Ticket).
The annotations endpoint is now top-level instead of being under the document endpoint. (Internal Ticket).
Fixed¶
The numbering of Annotation Sets in the SmartView does not consider deleted Annotation Sets anymore (Internal Ticket).
In some situations a Project could not be deleted via API (Internal Ticket).
In specific scenarios, the deletion of the last remaining Annotation in a Document was not possible (Internal Ticket).
The SmartView did not use rotated pages due to a caching problem (Internal Ticket).
The arrow in the Project- and language selector was not clickable (Internal Ticket).
On the Annotation list Page the Category filter was not showing all Annotations (Internal Ticket).
Fix an issue where the Category- and Document API V3 endpoint did not include all relevant Label-Sets (Internal Ticket).
Fix an issue that prevented specific SmartView messages to be dismissed (Internal Ticket).
Fix an issue that prevented null values to be passed to the API v3 annotation creation endpoint (Internal Ticket).
released-2022-10-28_07-23-39¶
Added¶
Fixed¶
Accepting an Annotation was overwritting already existing custom offset strings (Internal Ticket).
In some cases an Extraction AI training was failing when detecton mode ‘Character’ was selected (Internal Ticket).
released-2022-09-21_12-00-31¶
This version uses Konfuzio Trainer in version v.0.3.22, the Konfuzio Python SDK in version v.0.1.15 and Konfuzio Document Validation UI in version 0.0.8.
Fixed¶
Prevent an issue where a popup window could not be closed when using the SmartView (Internal Ticket).
Filtering for “feedback required” on Document overview (Internal Ticket).
released-2022-09-04_09-11-18¶
This version uses Konfuzio Trainer in version v.0.3.21, the Konfuzio Python SDK in version v.0.1.15 and Konfuzio Document Validation UI in version 0.0.8.
Added¶
Changed¶
Improve visibility of the left navigation bar.
Create sandwich PDFs on demand (until now they have been created on document upload).
Add content-type header (“Content-Type”: “application/json”) to the callback response.
Update Power Query Excel Documentation with animated screencasts
Fixed¶
Top annotation filter in the SmartView now considers unrevised annotations.
Fix an issue in which a Document cannot be processed because negative bounding boxes are detected.
Fix an issue which caused the processing time to be shown as negative.
released-2022-07-28_15-55-29¶
This version uses Konfuzio Trainer in version v.0.3.21, the Konfuzio Python SDK in version v.0.1.15 and Konfuzio Document Validation UI in version 0.0.6.
Changed¶
Speed up runtime of Extraction AIs
Fixed¶
Fix an issue which causes some Extraction AIs to crash on multipage documents.
Fix an issue that prevents the calculation of bounding boxes for small or slightly rotated characters.
released-2022-07-25_21-20-48¶
Added¶
Allow to set a default assignee for uploaded documents
Allow to notify users via email when they get assigned to documents
Fixed¶
Top annotation filter in the SmartView now takes accepted Annotations into account
Errors messages in case a document could not be processed are now displayed correctly
released-2022-07-19_16-30-46¶
Changed¶
New Extraction AIs are saved in a more efficient way
released-2022-07-05_19-35-21¶
This version uses Konfuzio Trainer in version v.0.3.15, the Konfuzio Python SDK in version v.0.1.15 and Konfuzio Document Validation UI in version 0.4.0.
Added¶
Show the user who started an AI training on the detail page of an AI
Allow to set a time (in days) after which documents are automatically deleted
Allow to rotate pages via API
Add thumbnail images for document pages
Changed¶
Links to deleted annotation will now redirect to the respective document
released-2022-06-10_15-32-19¶
This version uses Konfuzio Trainer in version v.0.3.15 and the Konfuzio Python SDK in version v.0.1.15.
Added¶
Option to enforce running OCR even if text embeddings are present
Improved error messages in case a document cannot be processed
Option to exclude email content when using the email-integration
Option to make document accessible via public link
Beta Version of APIV3
Beta Version of new document dashboard (bases on Konfuzio Document Validation UI)
Automatic rotation of pages (#8980)
Changed¶
For on-premise Users, now the Postgresql 10 is the minimum version
Improved Extraction AI
On-Premise container run now as non-root and using read-only fileystem
Improved mouse pointer in the Smartview
Fixed¶
An issue where empty Annotation Sets could appear on Documents
An issue where conflicting annotions could be created
An issue where negative annotations where not correctly being deleted (#9127)
Rare cases where OCR text included some characters mutiple times
released-2022-04-27_14-23-38¶
Added¶
Add assignee attribute of a Document to the API
released-2022-03-15_09-14-17¶
Changed¶
“Rerun extraction” via the user interface applies new annotations now also to training and test documents
released-2022-02-11_23-12-26¶
Added¶
Add option to filter for related annotation sets
Fixed¶
Sorting of annotation sets in the csv export
Document API endpoint returning declined annotations
released-2022-01-18_11-08-24¶
Added¶
Added api_name to Label API
Fixed¶
Link to documentation page
Missing translation on document list page
Evaluation did not complete for AIs with a large amount of training data
released-2021-12-11_14-33-57¶
Added¶
For on-premise installations, the OCR method for new projects is choosen based on the available OCR solutions.
For on-premise installations, the project import considers now declined annotations
For on-premise installations, Superusers can see the Konfuzio Server version and how many pages and documents have been processed.
released-2021-11-21_19-14-19¶
Added¶
Text summarization endpoint.
Categorization AI parameters in the Project view
Fixed¶
An issue where the reload after uploading new documents does not happen
released-2021-11-26_08-11-36¶
This version uses Konfuzio Trainer in version v.0.3.0. We recommend to use the Konfuzio Python SDK in version 0.1.15
Added¶
“Sentence” option to the available detection modes
Fixed¶
An error where an invalid date in the document text stoppped the training process
released-2021-11-23_18-14-28¶
Fixed¶
E-mails without an attachment have not been processed.
released-2021-11-16_23-02-22¶
Fixed¶
CSV export for ProRis by Inveos
released-2021-11-05_09-55-10¶
Added¶
Allow deletion of characters of an annotation without excluding it from the training process
released-2021-11-01_23-19-58¶
Added¶
An option to specify the category of a document when uploading it via API (and thereby skipping the categorization)
Changed¶
The GET document API endpoint now returns the annotation displayed in the SmartView (instead of only showing the extraction AI results)
released-2021-10-25_20-12-18¶
This version uses Konfuzio Trainer in version 2021-10-20_18-29-25. We recommend to use the Konfuzio Python SDK in version 0.1.10
Added¶
CSV export compatible with ProRis by Inveos
released-2021-10-16_13-20-12¶
Added¶
Improve detection of annotations which consist of multiple words
Date filtering for project documents API endpoint
Filtering of labels and label sets according to the category of a document (in the SmartView)
Fixed¶
Selection of characters in SmartView incomplete when editing an annotation
Dark Mode setting of browser not compatible with Konfuzio Server
Some case where the document list was not reloaded automatically
released-2021-10-07_11-42-29¶
Added¶
More advanced task priorities and improved worker ressource usage
Auto-reload of new uploaded documents
released-2021-09-28_09-29-43¶
Fixed¶
Evaluation does not complete if no test documents are specified
released-2021-09-24_13-53-32¶
Fixed¶
Incompleted evaluation
Formatting of the “Check your browser” page for logged out users.
released-2021-09-16_12-25-23¶
Fixed¶
Adding of categories to existing label sets
released-2021-09-08_16-01-46¶
Added¶
Migration scripts for user permissions and e-mail templates
released-2021-09-07_12-24-26¶
Added¶
Support for SMTP e-Mail backends via environment variables
Fixed¶
DOS protection prevents start of Konfuzio server
released-2021-09-05_20-57-31¶
Added¶
Autosave for any change on the document list page
German language support
Finetuning of exctraction AIs via parameters
New fields AI quality and data quality
More detailed evaluation
Description Field for extraction AI, label sets, categorization AI, categories
Changed¶
Rename project inviations to members
Rename the dataset status form “OCR Error” to “Excluded”
Start training per extraction AI
Get more insights via the document detail page
released-2021-08-10_17-08-11¶
Changed¶
Deactive adoption of template settings according to AI model if not explicitly allowed.
released-2021-08-10_11-19-33¶
Added¶
Maximum number of pages per document
Fixed¶
Slow processing of extraction tasks
Evaluation when multiple annotations are present
released-2021-07-28_18-53-12¶
Changed¶
Make word-based tokenizer the default for new projects
released-2021-07-23_09-33-20¶
Fixed¶
Usage of word-base tokenizer
Duplicated hints
released-2021-07-20_17-29-23¶
Fixed¶
Edited annotation were excluded from the training process
released-2021-07-15_17-29-25¶
Added¶
Support to reuse label sets across categories
Changed¶
Allow “rerun extraction” on test and training documents
Remove “project statistic csv export” as it is redundant to document csv export
Include evaluation for training data in the AI model evaluation report
Fixed¶
Fixed a bug where the EXIF attribute orientation corrupted the bounding boxes images
“accept top annotations” does not update human created annotations
released-2021-07-02_18-13-01¶
Changed¶
Rate limits for task system
released-2021-06-29_22-14-33¶
Added¶
HTTP codes to API interface
Fixed¶
Content type description for some API endpoints
released-2021-06-22_22-45-48¶
Added¶
A experimental version of a training health report
released-2021-06-20_15-14-31¶
Fixed¶
Failed retraninings for some projects
Increased disk usage due to an cache deletion issue
Filtering of project invotations according to currently selected project
Clarify return types in API documentation
released-2021-05-26_20-16-02¶
Added¶
Show confidence for categorization results
Show evaluation of categorization Ai models
Track version (number of retrainings) for all Ai models
Track project and template origin of AiModel
released-2021-05-24_13-42-45¶
Changed¶
Use business evaluation implementation from training package
Loading time for CSV export evaluation reduced by saving it in the database.
released-2021-05-18_16-37-25¶
Added¶
Global project switcher
“Top candidates” filter in SmartView
“Change dataset” functionality in SmartView
Landing page in case the user has no projects (i.e. just registered)
Language switcher (not enabled yet)
Initial support for German translations (not enabled yet)
Fixed¶
Label threshold is now limited from 0.0 to 1.0
Changed¶
New design for login/signup/reset password pages
Design improvements in the control panel and SmartView
New logo and favicon
API documentation has been improved with types and examples; is now based on OpenAPI 3
Updated frontend dependencies and tooling
Removed¶
admin_importer,copy_extraction_as_annotationand related functions have been removed
released-2021-05-04_12-37-16¶
Fixed¶
Calculation of true negative when using multiple templates.
released-2021-04-28_12-27-19¶
Added¶
Filter for top annotations in SmartView
Changed¶
Dont allow training if there are no training documents
released-2021-04-25_20-09-04¶
Added¶
Protect signup with captcha
Fixed¶
Editing of annotation if there are already declined annotations.
released-2021-04-19_22-32-19¶
Added¶
Add label creation endpoint
Token-based authentication for the API
released-2021-04-03_09-46-56¶
Added¶
Show Django sidebar in Smartview and template view.
Changed¶
Save extraction results in a more efficient way.
Show a warning if an annotation with a custom offset string is created
Shwo loading indicator in the smartview search
Fixed¶
Default template dropdown sometimes disabled when creating a Template
Rare case where the document list could not be loaded
released-2021-03-15_15-12-04¶
Added¶
Add option to accept all annotations.
released-2021-03-07_21-32-41¶
Added¶
Option to retrain project categorization model
Improved OCR settings
System check page https://app.konfuzio.com/check/
Fixed¶
Typo in privacy policy
Confirmation message when deleting labels
Performance of csv export
Changed¶
Delete old unrevised annotations when rerunning AiModel.
released-2021-02-25_09-28-07¶
Added¶
Option to select tokenizer for training (ProjectAdmin)
Option to add training parameters (SuperuserProjectAdmin)
Changed¶
Set a documents category_template on new documents if there is only one category_template available
Improved delete / accept performance of annotations
Fixed¶
Count of annotations on the LabelAdmin
released-2021-02-15_18-56-51¶
Changed¶
Show category template as empty when actual empty (instead of displaying the first available template)
Improved Smartview performance by changing entity loading
Added¶
Project name added to SectionLabel in the AiModelAdmin
Assign user to documents (“Assignee”). Can be enabled in the ProjectSuperuserAdmin
Add status field to the AiModel (“Training”, “Failed”, “Done”)
Dont allow new retraining if there is a training in progress AiModel.
released-2021-02-13_18-18-52¶
Changed¶
Use annotation permalink in LabelAdmin
Fixed¶
OCR Read API did not use text embeddings when available
Files with misssing fonts could not be processed
Creation of small annotations when accepting or declining
released-2021-02-10_13-52-15¶
Added¶
Admin action for Microsoft Graph API / Planner API
Fixed¶
SuperUserDocumentAdmin performance
OutOfMemory errors in the categorization
released-2021-02-03_17-07-23¶
Added¶
Permalink for annotations
Add an additional routine to fix corrupted pds
Improved frontend error tracking
Fixed¶
Validation when edting an annotation
Changed¶
Renamed option ‘priority_ocr’ to ‘priority_processing’
Allow rerun extraction for documents with revised annotations
Allow deletion default templates
released-2021-01-26_18-07-11¶
Added¶
Add column ‘category’ to csv export
released-2021-01-20_11-17-24¶
Added¶
Show selection bounding boxes for automtic created annotations
released-2021-01-14_22-06-52¶
Added¶
Visual annotations: images and area can now be annotate
Fixed¶
Loading time for Smartview
released-2021-01-13_23-26-03¶
Fixed¶
Retraining now assigns AIModels to templates even if they was no before
Added¶
Add Message when doing evaluation which tells the user if test set is empty.
released-2021-01-12_21-13-48¶
Fixed¶
Google Analytics integration
Empty Textextraction for ParagraphExtractions
released-2021-01-10_18-36-49¶
Fixed¶
Disable link formatting by sendgrid.
released-2021-01-08_22-30-10¶
Fixed¶
Bbox calculation in ParagraphModel
Evaluation sometimes not running
Speedup annotation creating
released-2021-01-05_11-53-22¶
Changed¶
Two column Annotation selection is now possible
Added¶
ParagraphModel introduced in addition to the Extraction- & CategoryModels, this is set per project via the SuperUserDocumentAdmin.
Option to update the document document text, this is set per project via the SuperUserDocumentAdmin.
Document Segmentation API Endpoint
released-2020-12-22_19-04-04¶
Changed¶
Email Template are now managed within the application.
Major improvement and refactor in the underlying training package.
Fixed¶
Link to imprint on SignUp
Smartview when scrolling horizontally
released-2020-12-16_20-17-30¶
Added¶
Search for Smartview
Fixed¶
TemplateCreationForm does not allow to select parent template
released-2020-12-16_09-44-30¶
Added¶
Searchbar for SuperuserProjectAdmin
Add link to flower (task monitoring) for superusers
Add support for GoogleTag Manager
Create Support Ticket for Retraining and Invitation of new Users
Changed¶
Increase SoftTimeLimit for extraction (necessary for large documents >500 pages).
Fixed¶
Fix bbox generation fox Paragraph Annotations
Fixed Evaluation not triggered for new AiModels
released-2020-12-10_13-15-14¶
Added¶
Sentry error reporting for Javascript Frontend (i.e. Smartview)
Allow to add Project specific document CategorizationModel
Changed¶
Document Search now considers filenames and shows links to Dashhboard, Labeling and Smartview
Allow deletion of Labels
Fixed¶
Allow “None” as confidence for rule-base ExtractionModels
released-2020-12-01_21-08-32¶
Added¶
Proof of Concept Microsoft Graph API connection (for logged in users): app.konfuzio.com/graph
Button to upload demo Documents
SuperuserProjectAdmin added (same like previous ProjectAdmin, however only accessible for Superusers only)
Google Analytics Tag for app.konfuzio.com
Changed¶
Default permission Group “CanReadProject” replaced with “CanCreateReadUpdateProject”. New users can now create new Projects.
Project Page for “normal” user does not show technical fields like “ocr” and “text_layout” anymore.
Dont show file endings like ‘.pkl’ for AiModels
released-2020-11-26_19-43-14¶
Fixed¶
Missing bbox attribute in Document API (prevents retraining via training package)
Running of proper ExtractionModel in Multi-Document-Template project
Loading time for the Document page (still room for improvements)
Added¶
Slightly better Categorization model.
released-2020-11-20_20-05-47¶
Added¶
A public registration page: https://app.konfuzio.com/accounts/signup
A Internal registration page to create users manually and faster:
https://app.konfuzio.com/register/(you need to be logged in to see this page)Users can invite new users to a project via “ProjectInvitations”
Password reset functionality
Fixed¶
The Smartview is much faster
Improved creation of Templates and additional validation logic template inconsistencies.
Changed¶
Save bbox and entity per page in order to improve performance
released-2020-11-09_18-04-28¶
Added¶
Support for more than one default Template in a project
Categorization for multi Template projects
Links to related models in the Project, AIModel, Label and Template view
Internal user registration form, app.konfuzio.com/register
Changed¶
AiModel belongs now to DefaultTemplates instead of project
released-2020-10-27_10-37-15¶
Changed¶
Documents are now soft-deleted. There is a hard delete option in the SuperuserDocumentAdmin.
AiModel are made active automatically for matching DefaultTemplates if the AIMode is better than before.
released-2020-10-21_08-53-42¶
Fixed¶
Loading time when updating a project.
released-2020-10-19_22-46-49¶
Changed¶
Increase max allowed workflow time from 90 to 180 seconds.
Fixed¶
sucess messages for ‘rerun_workflow’ admin action
loading time of AiModel
csv export
Added¶
add hocr fied to document api.
add a project option to hide the Smartview and Labeling tool.
released-2020-10-14_11-39-17¶
Changed¶
AIModel can be uploaded and evaluted before setting active for a project
released-2020-10-13_15-10-22¶
Added¶
Multilanguage Support (DE/EN) in the backend (actuall translation are not included yet)
Changed¶
‘create_labels_and_templates’ is now a project option (false by default).
Gunicorn workers restart after 500 requests.
Flower dashboard is running in separated container now
Fixed¶
Fix upload_ai_model to upload files larger than 2GB
Loading speed for SequenceAnnotation Admin
released-2020-10-03_15-18-47¶
Fixed¶
Recover tasks in case celery worker crashes
released-2020-10-01_12-02-37¶
Fixed¶
Internet Explorer warning badge
‘Not machine-readable’ was not detecting 0 as proper value for normalization.
Changed¶
Remove extraction count from AiModel admin.
Refactor annotation accept/delete buttons to separate components and SVG
released-2020-09-16_18-19-53¶
Added¶
Additional normalization formats
Sentry message if retraining is triggered.
Detectron (fully imlemented) and preparation for visual classification results in SuperUserDocumetAdmin
Changed¶
Dont raise sentry error if document got deleted during workflow
Fixed¶
released-2020-09-11_13-47-51¶
Added¶
Add sentry message if project retraining is triggered.
Fix cpu minute calculation.
released-2020-09-09_16-12-44¶
Added¶
Changed¶
Allow extractions which does not have an accuracy.
On the dashboard: Dont show section.position column if all extractions have the same. Dont show accuracy column if all extraction does not have one.
Dont show retraining webhook url (on the project detail page). Display is with **** like it is password.
released-2020-09-08_22-38-19¶
Added¶
Per-project measuring of cpu time.
Additional date-formats for normalization.
First draft of boolean-formats for normalization.
released-2020-09-08_09-17-00¶
Added¶
Document Filter added for ‘human feedback required’ and ‘100% machine readable.
Additional normalization formats for numbers.
Document Categorization Classifier added to DocumentSuperUserAdmin
Changed¶
For the document view and Smartview, rename ‘possibly incorrect’ to ‘not machine-readable’
For the document view and Smartview, rename ‘pending review’ to ‘require feedback’
For the document view, divide column NOTES into FEEDBACK REQUIRED and NOT MACHINE-READABLE
Fixed¶
Dont raise an error if ai_model predict section with a template that does not exist.
released-2020-09-07_17-48-22¶
Fixed¶
Filter for ‘possibly incorrect’ shows wrong number.
Sorting in csv is now correctly ordered by document_id and template position.