config.yaml

# Project related
tracking_uri: http://127.0.0.1:3080
logging_precision: 5
train_dataset_name: train
test_dataset_name: test
datasets_file_format: csv
dvc_remote_name: s3_repo
local_datasets_dir: data
dt_col_name:
split_dt_thresh:
id_col_name:

# Model related
save_dir: models
name: breast_cancer_xgb
type_: classifier  # classifier/regressor/question-answering/text-summarization/text/retriever
champion_alias: champion
load_by_alias: False
mlflow_save_format: xgb  # non-sklearn libraries
local_save_format: ubj  # non-sklearn libraries
sklearn_save_format: pkl
sklearn_predict_fn: predict_proba  # predict/predict_proba/predict_log_proba/predict_joint_log_proba/score
target_name: target
default_test_size: 0.33  # specified in MLproject
params_tuning_n_trials: 10  # specified in MLproject
cv_n_folds: 3
params_eval_metrics:  # last will be used for early stopping
  - error
  - aucpr
  - auc
params_tuning_metric: auc
params_tuning_direction: maximize  # minimize/maximize
additional_metrics:
  gini:
    source: auc
    formula: "2 * {} - 1"
  accuracy:
    source: error
    formula: "1 - {}"
early_stopping_heuristic: 0.1
importance_types: # for linear models -- only "weight"
  - weight  # number of times a feature appears in a tree
  - gain  # average gain of splits which use the feature
  - cover  # average number of samples affected by the splits which use the feature