Skip to content

release raiwidgets and responsibleai v0.30.0
Compare
Choose a tag to compare
@imatiach-msft imatiach-msft released this 05 Sep 14:05
· 209 commits to main since this release
v0.30.0
67169a8

v0.30.0

  • educational materials
  • new features

    • Responsible AI Dashboard

    • RAI Text

      • [Question Answering] Cache support for Model Overview Metrics by @Advitya17 in #2166
      • update responsibleai-text and responsibleai-vision packages to responsibleai 0.29.0 by @imatiach-msft in #2172
      • add base values and importances for ending tokens to QA explanations by @imatiach-msft in #2178

    • RAI Vision

      • [Question Answering] Cache support for Model Overview Metrics by @Advitya17 in #2166
      • update responsibleai-text and responsibleai-vision packages to responsibleai 0.29.0 by @imatiach-msft in #2172
      • [Object Detection] Model Overview Cache Support Extension for torchmetrics by @Advitya17 in #2170
      • OD model type support in raiutils for EA by @Advitya17 in #2227
      • GPU support for OD metrics by @Advitya17 in #2266

    • RAI Tabular

    • Error analysis

      • Error Analysis backend support for Object Detection by @Advitya17 in #2225
  • bug fixes and tests

    • Responsible AI Dashboard

    • RAI Vision

      • Require pillow>=10.0.0, catch drise failure due to low mask_res or num_masks by @romanlutz in #2181
      • Add retry logic to image reader by @kicha0 in #2206
      • fix error for automl object detection models when initializing error analysis and optimize explanation execution by @imatiach-msft in #2245
      • fix for aggregate parameter passed to torchmetrics by @imatiach-msft in #2235
      • pass transforms for optimized automl model explanations and improve error logging to include inner exception by @imatiach-msft in #2269

    • RAI Text

      • add RAITextInsights DBPedia notebook example and tests by @imatiach-msft in #2204
      • add RAITextInsights covid19 multilabel classification notebook example and tests by @imatiach-msft in #2221
      • add RAITextInsights blbooksgenre binary classification notebook example and tests by @imatiach-msft in #2226
      • Raise usererror on prediction failure. by @kicha0 in #2248
      • add RAITextInsights question answering notebook example and tests by @imatiach-msft in #2279
      • fix question answering error when running pipeline model on cuda by @imatiach-msft in #2291
      • add e2e UI notebook tests to covid19 event multilabel text classification notebook by @imatiach-msft in #2282

    • RAI Tabular

      • Change ValueError to UserConfigValidationException in casual_manager.py by @gaugup in #2209
      • Add input type validation for FeatureMetadata class by @gaugup in #2218
      • Move missing value validations before unique value validations by @gaugup in #2228
      • Pass categorical features after dropping dropped features in CounterfactualManager by @gaugup in #2241
      • RAIInsights: prevent failures on optional methods & fix feature range formatting for timestamps by @romanlutz in #2258
      • fix RAI explanation error for dropped categorical features after serialization by @imatiach-msft in #2261
      • Tabular Only: Address missing values as per each manager in responsibleai by @gaugup in #2263
      • responsibleai: Add prediction wrapper to missing value tests by @gaugup in #2270

    • Error Analysis

      • Update text insights and error_analysis_manager to handle pd.DataFrame predictions and prediction with type of string by @tongyu-microsoft in #2179
      • Error Analysis testing for IC & multilabel by @Advitya17 in #2205
      • change erroranalysis valueerror caused by invalid model prediction shape to user error by @imatiach-msft in #2254

    • Fairness Dashboard

      • Use only Fairlearn v0.7.0 due to breaking changes in v0.9.0 and lack of Python 3.7 support in v0.8.0 by @romanlutz in #2194
  • other
    • Localized file check-in by OneLocBuild Task: Build definition ID 21368: Build ID 99116202 by @RubyZ10 in #2177
    • release responsibleai-vision 0.2.4 by @imatiach-msft in #2192
    • fix responsibleai-vision pypi release script by installing pycocotools from conda by @imatiach-msft in #2197
    • fix build CI failing due to old pip-tools failing with latest pip by @imatiach-msft in #2200
    • release scikit-learn upper bound in responsibleai package by @imatiach-msft in #2173
    • improve notebook test logging and verbosity to diagnose timeouts by @imatiach-msft in #2202
    • Widget spacing for OD MO widgets by @Advitya17 in #2213
    • Update _NOTICE.md file by @romanlutz in #2231
    • LEGO: Pull request from JUNO/hb_a12a4630-4852-4e7d-9cbc-c0e1117da1ab_20230811114026921 to main by @csigs in #2237
    • Localized file check-in by OneLocBuild Task: Build definition ID 21368: Build ID 101597327 by @RubyZ10 in #2236
    • reduce PR execution time for tabular RAI dashboard changes by @imatiach-msft in #2238
    • Release new version of raiutils with OD support by @Advitya17 in #2244
    • Removed python 3.6 from python PR gate by @Advitya17 in #2249
    • fix ci-notebook-vision transient 6 hour CI timeouts by installing older pyarrow by @imatiach-msft in #2212
    • Update explainer_manager.py to change docs from MimicWrapper to MimicExplainer by @gaugup in #2240
    • refactor notebook examples into tabular, text and vision folders by @imatiach-msft in #2253
    • Localized file check-in by OneLocBuild Task: Build definition ID 21368: Build ID 101886729 by @RubyZ10 in #2255
    • Added additional paths for vision gates by @Advitya17 in #2256
    • move image folder for tabular responsibleaidashboard under tabular folder by @imatiach-msft in #2257
    • release erroranalysis 0.4.5 by @imatiach-msft in #2262
    • fix CI errors due to new mlflow and pydantic dependencies by @imatiach-msft in #2259
    • update responsibleai and raiwidgets to erroranalysis 0.4.5 by @imatiach-msft in #2264
    • convert CI-python to use conda and install pytorch dependencies by @imatiach-msft in #2272
    • update responsibleai to interpret-community 0.30.0 by @imatiach-msft in #2275
    • Remove deprecated module cohort.py from raiwidgets by @gaugup in #2278
    • Turn off flag should_construct_pandas_query in PredictionsModelWrapperClassification by @gaugup in #2277
    • Update CODEOWNERS to include romanlutz for responsibleai by @romanlutz in #2280
    • OD model and data util functions in rai_test_utils by @Advitya17 in #2246
    • Release new version of rai test utils by @Advitya17 in #2284
    • Updated release gate for rai test utils to install OD dependencies by @Advitya17 in #2285
    • Add torch installation support for rai test utils release by @Advitya17 in #2286
    • Release new version of erroranalysis with OD support by @Advitya17 in #2288
    • remove vision e2e notebook tests from raiwidgets and responsibleai release workflow by @imatiach-msft in #2292
    • Rename release-rai.yml to release-rai-tabular.yml by @gaugup in #2283

Contributors

romanlutz, csigs, and 7 other contributors
Assets 2
Loading