Model selection failed

[lzw950905]

Description

When I run the Model selection process, even though I just use very small search space (just optimize one parameter), it always reports the bug: ValueError: too many values to unpack.

Actual Behavior

This is my code. X_dict = load_wn18rr()

model_class = ComplEx

# Use the template given below for doing grid search.
param_grid = {
                 "batches_count": [100],
                 "seed": 0,
                 "epochs": [10],
                 "k": [64,128],
                 "eta": [5],
                 "loss": ["pairwise"],
                 # We take care of mapping the params to corresponding classes

                 "embedding_model_params": {
                     # generate corruption using all entities during training
                     "negative_corruption_entities":"all"
                 },
                 "regularizer": [None, "LP"],
                 "regularizer_params": {
                     "p": [2],
                 },
                 "optimizer": ["adam"],

                 "verbose": False
             }

# Train the model on all possibile combinations of hyperparameters.
# Models are validated on the validation set.
# It returnes a model re-trained on training and validation sets.
best_model, best_params, best_mrr_train, \
ranks_test, mrr_test = select_best_model_ranking(model_class, # Class handle of the model to be used
                                                 # Dataset
                                                 X_dict['train'],
                                                 X_dict['valid'],
                                                 X_dict['test'],
                                                 # Parameter grid
                                                 param_grid,
                                                 max_combinations=3,
                                                 # Use filtered set for eval
                                                 use_filter=True,
                                                 # corrupt subject and objects separately during eval
                                                 use_default_protocol=True,
                                                 # Log all the model hyperparams and evaluation stats
                                                 verbose=False)
print(type(best_model).__name__, best_params, best_mrr_train, mrr_test)

Expected Behavior

No such error and run successfully.

Steps to Reproduce

[lukostaz]

Hey @lzw950905. select_best_model_ranking returns 6 values. You listed only 5 and missed experimental_history.

The snippet below will do:

    best_model, best_params, best_mrr_train, \
    ranks_test, mrr_test, experimental_history = select_best_model_ranking(ComplEx, 
                                                     # Dataset
                                                     X_dict['train'],
                                                     X_dict['valid'],
                                                     X_dict['test'],
                                                     # Parameter grid
                                                     param_grid,
                                                     max_combinations=3,
                                                     # Use filtered set for eval
                                                     use_filter=True,
                                                     # corrupt subject and objects separately during eval
                                                     use_default_protocol=True,
                                                     # Log all the model hyperparams and evaluation stats
                                                     verbose=False)