Search results

Inconsistent semantics of training_id in feature importance DataFrames

Status: Open.

#427 In emdgroup/octopus-automl;

· nihaase opened on Apr 13, 2026

AutoGluon group feature importance fails when features are dropped during preprocessing

Status: Open.

Bug

#426 In emdgroup/octopus-automl;

· nihaase opened on Apr 13, 2026

LOGLOSS and NEGBRIERSCORE metrics set higher_is_better=True

Metrics

PyCon

Status: Open.

Bug

#422 In emdgroup/octopus-automl;

· kalama-ai opened on Apr 9, 2026

penalty_factor in Octo does not scale with metric

Status: Open.

#419 In emdgroup/octopus-automl;

· kalama-ai opened on Apr 8, 2026

AutoGluon memory_limit not compatible with Ray parallelization

Dev Meeting

Modules

Parallelization

PyCon

Status: Open.

Bug

#412 In emdgroup/octopus-automl;

· kalama-ai opened on Apr 1, 2026

Move get_default_model_params from test helper into ModelConfig.get_default_params()

Status: Open.

#405 In emdgroup/octopus-automl;

· anwurl opened on Mar 27, 2026

Need tests for re-using an existing ray cluster and for distributed compute

Status: Open.

#403 In emdgroup/octopus-automl;

· dasmy opened on Mar 27, 2026

Refactor module/octo

Status: Open.

#401 In emdgroup/octopus-automl;

· anwurl opened on Mar 26, 2026

Add runtime assertions for datasplit

Status: Open.

#398 In emdgroup/octopus-automl;

· kalama-ai opened on Mar 25, 2026

Two sources of truth (ModelName enum vs Models registry)

Status: Open.

#394 In emdgroup/octopus-automl;

· anwurl opened on Mar 25, 2026

Ray workers write outersplit results to packaged working directory instead of project directory

Status: Open.

Bug

#393 In emdgroup/octopus-automl;

· nihaase opened on Mar 24, 2026

Make Tests faster - 2 test performance bottlenecks

Status: Open.

#367 In emdgroup/octopus-automl;

· anwurl opened on Mar 14, 2026