Releases: imixs/imixs-ml
Releases · imixs/imixs-ml
imixs-ml-1.1.6
Features
- Add JDK 11 support #84
- Update ml.status #83
- MLAdapter - provide optional regex patterns #77
- Improve logging - using the logging Python module #73
Enhancements
- TrainingService - generateTraingDataSet should ignore ml.definitions #82
- Training Service - add validation method #81
- Introduce a retrain_rate for spacy #80
- Support SpaCy displacy for testing #79
- Upgrade SpaCy version => 3.4 #75
- Multi-language support #65
Bugfixes
- DateAdapter - did not distinguish between year/month order #74
imixs-ml-1.1.5
imixs-ml-1.1.4
imixs-ml-1.1.3
Bugfixes
- MLEntity convert Object from XML failed #62
imixs-ml-1.1.2
Features
- ML - introduce new TrainingData Quality Levels #58
- ML - introduce MLEntity #57
- MLDateAdapter - support US Date Format "APR. 14, 2021" #59
- MLCurrencyAdapter - support USD with thousend separator . #60
- OCR Helper to fix bad scans #61
- ML - ml-config entity definition - support new Training flag 'required=true' #56
imixs-ml-1.1.1
imixs-ml-1.1.0
Features
- MLService - overwriting the ML Status flag #52
- MLAdapter - aggregate text content #47
- imixs-ml-spacy - provide health endpoint #51
- AnalyseText should return the categories including the score #49
- SpaCy - initialize model with categories #46
- Upgrade spaCy version 3.0 #43
- Change API Endpoints for a multi-model support #42
- Add Text Classification #40
Enhancements
- MLController - improve findMaches #50
- MLAdapter - interrupt processing life cycle on a ml api error #48
- Jakarta 8 support #41
- Upgrade Imixs-Workflow 5.2.7 #39
Bugfixes
- XMLTrainingData - build - wrong quality if no items found #45
imixs-ml-1.0.3
Bugfixes
- XMLTrainingData - build - wrong quality if no items found #44
imixs-ml-1.0.2
Enhancements
- Added Docker Hub Support #38
imixs-ml-1.0.1
Features
- MLService - support trainingdata quality level #36
Enhancements
Bugfixes
- Change default lcale to GERMANY, UK #35
- Training Data - US currency and German date not recognized correctly #32
- MLAdapter - ml items not defined in the workflow model must be ignored #34
- MLService - endpoint must not include the /analyzse/ resource #31
- MLTrainingScheduler does not remove events after processing #30
- MLTrainingScheduler - type in environment variables #28
- MLTraningService do not create training events if scheduler is disabled #29