def test_xdeepfm_component_definition(deeprec_resource_path):
data_path = os.path.join(deeprec_resource_path, "xdeepfm")
yaml_file = os.path.join(data_path, "xDeepFM.yaml")
if not os.path.exists(yaml_file):
download_deeprec_resources(
"https://recodatasets.z20.web.core.windows.net/deeprec/",
data_path,
"xdeepfmresources.zip",
)
hparams = prepare_hparams(yaml_file)
model = XDeepFMModel(hparams, FFMTextIterator)
assert model.logit is not None
assert model.update is not None
assert model.iterator is not None
def test_model_xdeepfm(resource_path):
data_path = os.path.join(resource_path, '../resources/deeprec/xdeepfm')
yaml_file = os.path.join(data_path, r'xDeepFM.yaml')
data_file = os.path.join(data_path, r'sample_FFM_data.txt')
output_file = os.path.join(data_path, r'output.txt')
if not os.path.exists(yaml_file):
download_deeprec_resources(
r'https://recodatasets.blob.core.windows.net/deeprec/', data_path,
'xdeepfmresources.zip')
hparams = prepare_hparams(yaml_file, learning_rate=0.01)
assert hparams is not None
input_creator = FFMTextIterator
model = XDeepFMModel(hparams, input_creator)
assert model.run_eval(data_file) is not None
assert isinstance(model.fit(data_file, data_file), BaseModel)
assert model.predict(data_file, output_file) is not None
def test_model_xdeepfm(deeprec_resource_path):
data_path = os.path.join(deeprec_resource_path, "xdeepfm")
yaml_file = os.path.join(data_path, "xDeepFM.yaml")
data_file = os.path.join(data_path, "sample_FFM_data.txt")
output_file = os.path.join(data_path, "output.txt")
if not os.path.exists(yaml_file):
download_deeprec_resources(
"https://recodatasets.z20.web.core.windows.net/deeprec/",
data_path,
"xdeepfmresources.zip",
)
hparams = prepare_hparams(yaml_file, learning_rate=0.01)
assert hparams is not None
input_creator = FFMTextIterator
model = XDeepFMModel(hparams, input_creator)
assert model.run_eval(data_file) is not None
assert isinstance(model.fit(data_file, data_file), BaseModel)
assert model.predict(data_file, output_file) is not None
download_deeprec_resources(
r'https://recodatasets.blob.core.windows.net/deeprec/', data_path,
'xdeepfmresources.zip')
# set hyper-parameters
hparams = prepare_hparams(yaml_file,
FEATURE_COUNT=2300000,
FIELD_COUNT=39,
cross_l2=0.01,
embed_l2=0.01,
layer_l2=0.01,
learning_rate=0.002,
batch_size=BATCH_SIZE,
epochs=EPOCHS,
cross_layer_sizes=[20, 10],
init_value=0.1,
layer_sizes=[20, 20],
use_Linear_part=True,
use_CIN_part=True,
use_DNN_part=True)
# make model
model = XDeepFMModel(hparams, FFMTextIterator, seed=RANDOM_SEED)
# train model
model.fit(train_file, valid_file)
# profiling
model.train_timeliner.save('xDeepFM-timeliner.json')
# inference after training
#print(model.run_eval(test_file))
hparams = prepare_hparams(yaml_file,
FEATURE_COUNT=1000,
FIELD_COUNT=10,
cross_l2=0.0001,
embed_l2=0.0001,
learning_rate=0.001,
epochs=EPOCHS,
batch_size=BATCH_SIZE)
print("Hyper-parameters: ")
print(hparams)
# 2. create data loader
# designate a data iterator for xDeepFM model (FFMTextIterator)
input_creator = FFMTextIterator
# 3. create model
model = XDeepFMModel(hparams, input_creator, seed=RANDOM_SEED)
# we can also load a pre-trained model with model.load_model(r'model_path')
# untrained model's performance
print("Untrained model's performance: {}".format(model.run_eval(test_file)))
# 4. train model
print("Begin model training...")
model.fit(train_file, valid_file)
print("End model training...")
# 5. evaluate model
#res_syn = model.run_eval(test_file)
#print("Trained model's performance: {}".format(res_syn))
#pm.record("rest_syn", res_syn)
# we can also get full prediction scores rather than evaluation metrics with model.predict(test_file, output_file)