search

h2oai / h2o-3

H2O is an Open Source, Distributed, Fast & Scalable Machine Learning Platform: Deep Learning, Gradient Boosting (GBM) & XGBoost, Random Forest, Generalized Linear Modeling (GLM with Elastic Net), K-Means, PCA, Generalized Additive Models (GAM), RuleFit, Support Vector Machine (SVM), Stacked Ensembles, Automatic Machine Learning (AutoML), etc.
http://h2o.ai
Apache License 2.0
6.87k stars 1.99k forks source link

Grid hyper-parameter not being displayed correctly #10076

Open exalate-issue-sync[bot] opened 1 year ago

exalate-issue-sync[bot] commented 1 year ago

Running a GLM binomial grid where alpha is the only hyper-parameter gives me bad data in the output. E.g. "[D@6a542950" should have said "0.0"

alpha   model_ids   logloss
[D@6a542950 GLM_1_model_0   0.6356303595732312
[D@3b6479b9 GLM_1_model_2   0.6373182952263592
[D@41737f25 GLM_1_model_1   0.6374537364332546

This is with h2o 3.8.3.4. Bad data shown both in Flow and R API.

An almost identical grid worked fine with 3.8.2.2.

exalate-issue-sync[bot] commented 1 year ago

Darren Cook commented: http://stackoverflow.com/a/39213568/841830 shows an easily reproducible example of this (still affecting H2O as of 3.10.x)

The output of grid looks like:

Hyper-Parameter Search Summary: ordered by increasing logloss hidden hidden_dropout_ratios model_ids logloss 1 [I@131d3b74 [D@642a4679 stackoverflow_model_3 0.09508510431035037 2 [I@7ff228c9 [D@7353e5be stackoverflow_model_0 0.11160817371643475 3 [I@1378e54f [D@6616af85 stackoverflow_model_4 0.12268507664176394 4 [I@9105819 [D@1007d5f5 stackoverflow_model_1 0.13197547861601153 5 [I@5b5d791f [D@55bb1115 stackoverflow_model_5 0.14727184006516547 6 [I@726fd00b [D@ac3bcdf stackoverflow_model_2 0.16722886689804004

I.e. it is the list hyper-parameters that go wrong.

h2o-ops commented 1 year ago

JIRA Issue Migration Info

Jira Issue: PUBDEV-3153 Assignee: Tomas Nykodym Reporter: Darren Cook State: Open Fix Version: N/A Attachments: N/A Development PRs: N/A