FlexMF BPR

This page analyzes the hyperparameter tuning results for the FlexMF scorer in implicit-feedback mode with pairwise loss (Bayesian Personalized Ranking).

Parameter Search Space

/home/mde48/lenskit/lenskit-codex/.venv/lib/python3.12/site-packages/ray/tune/search/sample.py:700: RayDeprecationWarning: The `base` argument is deprecated. Please remove it as it is not actually needed in this method.

Parameter	Type	Distribution	Values	Selected
embedding_size	Integer	LogUniform	4 ≤ \(x\) ≤ 512	19
regularization	Float	LogUniform	0.0001 ≤ \(x\) ≤ 10	0.0318
learning_rate	Float	LogUniform	0.001 ≤ \(x\) ≤ 0.1	0.00357
reg_method	Categorical	Uniform	L2, AdamW	AdamW
negative_count	Integer	Uniform	1 ≤ \(x\) ≤ 5	4
item_bias	Categorical	Uniform	True, False	False

Final Result

Searching selected the following configuration:

{
    'embedding_size': 19,
    'regularization': 0.031770202161679695,
    'learning_rate': 0.0035691437195705653,
    'reg_method': 'AdamW',
    'negative_count': 4,
    'item_bias': False,
    'epochs': 14
}

With these metrics:

{
    'RBP': 0.18413895566190336,
    'DCG': 11.50405688772486,
    'NDCG': 0.42494721030599253,
    'RecipRank': 0.35311611520891295,
    'Hit10': 0.5869742198100407,
    'max_epochs': 50,
    'epoch_train_s': 7.494721162998758,
    'epoch_measure_s': 12.775742112004082,
    'done': False,
    'training_iteration': 14,
    'trial_id': '2e52a817',
    'date': '2025-07-29_05-53-18',
    'timestamp': 1753782798,
    'time_this_iter_s': 20.274052619934082,
    'time_total_s': 367.2510471343994,
    'pid': 323395,
    'hostname': 'CCI-ws21',
    'node_ip': '10.248.127.152',
    'config': {
        'embedding_size': 19,
        'regularization': 0.031770202161679695,
        'learning_rate': 0.0035691437195705653,
        'reg_method': 'AdamW',
        'negative_count': 4,
        'item_bias': False,
        'epochs': 14
    },
    'time_since_restore': 367.2510471343994,
    'iterations_since_restore': 14
}