FLAML

FLAML copied to clipboard

skzhang1

dot_size argument not working for interaction values scatter plot

1

As the title says. The dot_size argument is simply missing where dependence_legacy is called recursively. `dependence_legacy( ind1, proj_shap_values, features, feature_names=feature_names, interaction_index=(None if ind1 == ind2 else ind2), display_features=display_features, ax=ax, show=False,...

zurKreuzigungLinks

Add an example using flaml to tune an AML pipeline.

ruizhuanguw

Skip transform

1

#660

jmrichardson

replaced set with ordered set to ensure reproduciblity

2

Replaced python 'set' with OrderedSets from "ordered-set" library to ensure reproducibility during runs when seed is set. Added ordered-set in setup.py as well

AdityaN1198

Update Research.md

liususan091219

Using Scikit-learn APIs directly

11

Almost yesterday (August 5 at 7 PM, Iran/Tehran time), I had a short [conversation](https://github.com/microsoft/FLAML/discussions/666) with @sonichi about this, and in general it is better to provide such features more easily...

sheikhartin

add a template for PR

sonichi

Ideas to improve the logging capabilities in `flaml.tune`: - [x] Take a `log_file_name` argument in `tune.run()`. Right now it takes `local_dir` as the directory for the log file, but the...

sonichi

is the final model fitted to the whole train set?

5

Dear, first of all thanks a lot and compliment! Then, in documentation is not clear one thing: after a fit the final model is fitted to the whole train set?...

luigif2000