-
Notifications
You must be signed in to change notification settings - Fork 15
/
main.py
120 lines (101 loc) · 4.54 KB
/
main.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
import torch
from torch import nn, utils
import torch.nn.functional as F
import numpy as np
# Hyper-parameters
num_epochs = 100
learning_rate = 0.2
margin = 1.0 # margin in triplet loss
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
class Dataset(utils.data.Dataset):
def __init__(self, query, positive_document, negative_document):
self.query = query
self.positive_document = positive_document
self.negative_document = negative_document
def __getitem__(self, index):
return self.query[index], self.positive_document[index], self.negative_document[index]
def __len__(self):
return len(self.query)
class QueryEncoder(nn.Module):
def __init__(self, input_size):
super(QueryEncoder, self).__init__()
self.fc1 = nn.Linear(input_size, 16)
self.fc2 = nn.Linear(16, 10)
self.fc3 = nn.Linear(10, 8)
def forward(self, x):
out = F.relu(self.fc1(x))
out = F.relu(self.fc2(out))
out = F.relu(self.fc3(out))
return out
class DocumentEncoder(nn.Module):
def __init__(self, input_size, hidden_layer_sizes=(100,), activation=('relu',), solver='adam'):
super(DocumentEncoder, self).__init__()
self.fc1 = nn.Linear(input_size, 12)
self.fc2 = nn.Linear(12, 8)
def forward(self, x):
out = F.relu(self.fc1(x))
out = F.relu(self.fc2(out))
return out
class Trainer():
def __init__(self, query_encoder, document_encoder, criterion, query_optimizer, document_optimizer):
self.query_encoder = query_encoder
self.document_encoder = document_encoder
self.criterion = criterion
self.query_optimizer = query_optimizer
self.document_optimizer = document_optimizer
self.training_loss = []
def train(self, data_loader, epoch):
self.query_encoder.train()
self.document_encoder.train()
running_loss = 0
for _, (query_inputs, positive_document_inputs, negative_document_inputs) in enumerate(data_loader):
# Forward pass
anchor = query_encoder(query_inputs)
positive = document_encoder(positive_document_inputs)
negative = document_encoder(negative_document_inputs)
loss = triplet_loss(anchor, positive, negative)
# Backward and optimize
query_optimizer.zero_grad()
document_optimizer.zero_grad()
loss.backward()
query_optimizer.step()
document_optimizer.step()
running_loss += loss.item()
print('Epoch [{}], Loss: {:.4f}'.format(
epoch+1, running_loss / len(data_loader)))
self.training_loss.append(running_loss / len(data_loader))
def save_model(self, query_model_path, document_model_path):
torch.save(self.query_encoder, query_model_path)
torch.save(self.document_encoder, document_model_path)
def load_dummy_data(query_input_size: int, document_input_size: int) -> utils.data.DataLoader:
# Dummy Data
query_inputs = np.random.rand(100, query_input_size).astype(np.float32)
positive_document_inputs = np.random.rand(
100, document_input_size).astype(np.float32)
negative_document_inputs = np.random.rand(
100, document_input_size).astype(np.float32)
data_loader = utils.data.DataLoader(Dataset(torch.from_numpy(query_inputs), torch.from_numpy(positive_document_inputs), torch.from_numpy(negative_document_inputs)),
batch_size=50, shuffle=True, num_workers=2)
return data_loader
if __name__ == '__main__':
# number of features for query encoder and document encoder
query_input_size = 20
document_input_size = 15
# Encoder initialization
query_encoder = QueryEncoder(query_input_size).to(device)
document_encoder = DocumentEncoder(document_input_size).to(device)
# Optimizer initialization
query_optimizer = torch.optim.Adam(
query_encoder.parameters(), lr=learning_rate)
document_optimizer = torch.optim.Adam(
document_encoder.parameters(), lr=learning_rate)
# Triplet loss
triplet_loss = nn.TripletMarginWithDistanceLoss(
distance_function=lambda x, y: 1.0 - F.cosine_similarity(x, y), margin=margin)
# Trainer initialization
trainer = Trainer(query_encoder, document_encoder,
triplet_loss, query_optimizer, document_optimizer)
# load dummy data
data_loader = load_dummy_data(query_input_size, document_input_size)
for epoch in range(num_epochs):
trainer.train(data_loader, epoch)