FLAML

FLAML copied to clipboard

estimator function

2

Open #807 Isolate the ensemble part and expose it to users

007-WJR

Hi FLAML developers, We’re researchers at Cornell University focused on quality assurance for machine learning workflows ([our group](https://www.cs.cornell.edu/~saikatd/)). We’d love to introduce you to [**NBTest**](https://github.com/seal-research/NBTest), a new tool we’ve built...

ElaineYao

### Is your feature request related to a problem? Please describe. Need to support Python 3.12. ### Describe the solution you'd like _No response_ ### Additional context _No response_

thinkall

fix(automl): fix data leakage in classification holdout validation

4

>> >> - Stop duplicating X_first in both train/val sets >> - Implement dynamic class balancing >> - Preserve original dataset size ## Why are these changes needed? In `generic_task.py`,...

commint-tian

[Bug]: Feature importances not present in individual ensemble models

3

### Describe the bug After I fit an AutoML ensemble for a regression (`"ensemble": True` in AutoML settings), I get the best model of the ensamble in this way: `best_model...

lucazav

[Feature Request]: Update AutoML metric argument documentation

3

### Is your feature request related to a problem? Please describe. The `AutoML` instance argument `metric` documentation is currently: > metric - A string of the metric name or a...

bl3e967

[Bug]: Feature importances not present for an ensemble

1

### Describe the bug After I fit an AutoML ensemble for a regression (`"ensemble": True` in AutoML settings), trying to get the feature importances as follow I get an empty...

lucazav

[Bug]: It seems that longer time budgets result in worse outputs

1

### Describe the bug I have a data set where I have tried to optimise the hyperparameters on Flaml, and it seems that the model keeps getting worse, the longer...

kabeersvohra

[Bug]: Sometimes the optimal results of non-optimal estimators are not saved

10

### Describe the bug Hi: I've created two customized lightGBM estimators for automl: class MyMonotonicLightGBMGBDTClassifier(BaseEstimator): def __init__(self, task = 'binary:logistic', n_jobs = num_cores, **params): super().__init__(task, **params) self.estimator_class = LGBMClassifier #...

flippercy

[Bug]: Training and Test Set Overlap For Classification Tasks with Holdout Strategy

3

### Describe the bug Hi @thinkall, I think that I may have found an issue in FLAML, although it's possible it was a deliberate choice by the developers. Basically if...

dannycg1996