FlexMF Explicit on ML10M

This page analyzes the hyperparameter tuning results for the FlexMF scorer in explicit-feedback mode (a biased matrix factorization model trained with PyTorch).

Parameter Search Space

Parameter	Type	Distribution	Values
embedding_size	Integer	LogUniform	4 ≤ \(x\) ≤ 512
regularization	Float	LogUniform	0.0001 ≤ \(x\) ≤ 10
learning_rate	Float	LogUniform	0.001 ≤ \(x\) ≤ 0.1
reg_method	Categorical	Uniform	L2, AdamW

Final Result

Searching selected the following configuration:

{
    'embedding_size': 14,
    'regularization': 0.04977041557904842,
    'learning_rate': 0.002501599473184507,
    'reg_method': 'L2',
    'epochs': 7
}

With these metrics:

{
    'RBP': 0.1201926373103401,
    'NDCG': 0.4037863375695813,
    'RecipRank': 0.2678976522436296,
    'RMSE': 0.7612197947196359,
    'TrainTask': 'd8d9000f-2f57-4253-abfd-bf24a153a4ad',
    'TrainTime': None,
    'TrainCPU': None,
    'max_epochs': 50,
    'epoch': 7,
    'done': True,
    'training_iteration': 7,
    'trial_id': '3b368_00065',
    'date': '2025-04-02_22-14-42',
    'timestamp': 1743646482,
    'time_this_iter_s': 7.709898233413696,
    'time_total_s': 57.655722856521606,
    'pid': 323936,
    'hostname': 'CCI-ws21',
    'node_ip': '10.248.127.152',
    'config': {
        'embedding_size': 14,
        'regularization': 0.04977041557904842,
        'learning_rate': 0.002501599473184507,
        'reg_method': 'L2',
        'epochs': 7
    },
    'time_since_restore': 57.655722856521606,
    'iterations_since_restore': 7,
    'experiment_tag': '65_embedding_size=14,learning_rate=0.0025,reg_method=L2,regularization=0.0498'
}

Parameter Analysis

Embedding Size

The embedding size is the hyperparameter that most affects the model’s fundamental logic, so let’s look at performance as a fufnction of it:

Iteration Completion

How many iterations, on average, did we complete?

How did the metric progress in the best result?

How did the metric progress in the longest results?