|
|
|
|
|
import sys
|
|
|
|
|
|
import os.path as osp
|
|
|
|
|
|
this_dir = osp.dirname(__file__)
|
|
|
|
|
|
lib_path = osp.join(this_dir, '..')
|
|
|
|
|
|
sys.path.insert(0, lib_path)
|
|
|
|
|
|
import torch
|
|
|
|
|
|
from DeepCrossing.trainer import Trainer
|
|
|
|
|
|
from DeepCrossing.network import DeepCrossing
|
|
|
|
|
|
from DeepCrossing.criteo_loader import getTestData, getTrainData
|
|
|
|
|
|
import torch.utils.data as Data
|
|
|
|
|
|
|
|
|
|
|
|
deepcrossing_config = \
|
|
|
|
|
|
{
|
|
|
|
|
|
'embed_dim': 8, # 用于控制稀疏特征经过Embedding层后的稠密特征大小
|
|
|
|
|
|
'min_dim': 256, # 稀疏特征维度小于min_dim的直接进入stack layer，不用经过embedding层
|
|
|
|
|
|
'hidden_layers': [512,256,128,64,32],
|
|
|
|
|
|
'num_epoch': 100,
|
|
|
|
|
|
'batch_size': 32,
|
|
|
|
|
|
'lr': 1e-3,
|
|
|
|
|
|
'l2_regularization': 1e-4,
|
|
|
|
|
|
'device_id': 0,
|
|
|
|
|
|
'use_cuda': True,
|
|
|
|
|
|
'train_file': '../Data/criteo/processed_data/train_set.csv',
|
|
|
|
|
|
'fea_file': '../Data/criteo/processed_data/fea_col.npy',
|
|
|
|
|
|
'validate_file': '../Data/criteo/processed_data/val_set.csv',
|
|
|
|
|
|
'test_file': '../Data/criteo/processed_data/test_set.csv',
|
|
|
|
|
|
'model_name': '../TrainedModels/DeepCrossing.model'
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
if __name__ == "__main__":
|
|
|
|
|
|
####################################################################################
|
|
|
|
|
|
# DeepCrossing 模型
|
|
|
|
|
|
####################################################################################
|
|
|
|
|
|
training_data, training_label, dense_features_col, sparse_features_col = getTrainData(deepcrossing_config['train_file'], deepcrossing_config['fea_file'])
|
|
|
|
|
|
train_dataset = Data.TensorDataset(torch.tensor(training_data).float(), torch.tensor(training_label).float())
|
|
|
|
|
|
test_data = getTestData(deepcrossing_config['test_file'])
|
|
|
|
|
|
test_dataset = Data.TensorDataset(torch.tensor(test_data).float())
|
|
|
|
|
|
|
|
|
|
|
|
deepCrossing = DeepCrossing(deepcrossing_config, dense_features_cols=dense_features_col, sparse_features_cols=sparse_features_col)
|
|
|
|
|
|
|
|
|
|
|
|
####################################################################################
|
|
|
|
|
|
# 模型训练阶段
|
|
|
|
|
|
####################################################################################
|
|
|
|
|
|
# 实例化模型训练器
|
|
|
|
|
|
trainer = Trainer(model=deepCrossing, config=deepcrossing_config)
|
|
|
|
|
|
# 训练
|
|
|
|
|
|
trainer.train(train_dataset)
|
|
|
|
|
|
# 保存模型
|
|
|
|
|
|
trainer.save()
|
|
|
|
|
|
|
|
|
|
|
|
####################################################################################
|
|
|
|
|
|
# 模型测试阶段
|
|
|
|
|
|
####################################################################################
|
|
|
|
|
|
deepCrossing.eval()
|
|
|
|
|
|
if deepcrossing_config['use_cuda']:
|
|
|
|
|
|
deepCrossing.loadModel(map_location=lambda storage, loc: storage.cuda(deepcrossing_config['device_id']))
|
|
|
|
|
|
deepCrossing = deepCrossing.cuda()
|
|
|
|
|
|
else:
|
|
|
|
|
|
deepCrossing.loadModel(map_location=torch.device('cpu'))
|
|
|
|
|
|
|
|
|
|
|
|
y_pred_probs = deepCrossing(torch.tensor(test_data).float().cuda())
|
|
|
|
|
|
y_pred = torch.where(y_pred_probs>0.5, torch.ones_like(y_pred_probs), torch.zeros_like(y_pred_probs))
|
|
|
|
|
|
print("Test Data CTR Predict...\n ", y_pred.view(-1))
|
|
|
|
|
|
|
