Repository navigation
Expand file tree
/
Copy pathtrain.py
More file actions
215 lines (138 loc) · 5.72 KB
/
Copy pathtrain.py
File metadata and controls
215 lines (138 loc) · 5.72 KB
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
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
import click
import datetime
import pickle
import numpy as np
import torch
import time
import random
from torch.utils.data import random_split
from torch.utils.data import (DataLoader, RandomSampler, SequentialSampler,
TensorDataset, Dataset)
from transformers import BertForSequenceClassification, BertTokenizer, AdamW
from utils import format_time
def flat_accuracy(preds, labels):
""" Calculate the accuracy of our predictions vs labels
"""
pred_flat = np.argmax(preds, axis=1).flatten()
labels_flat = labels.flatten()
return np.sum(pred_flat == labels_flat) / len(labels_flat)
def train_model(model, train_dataloader, validation_dataloader, epochs, scheduler, optimizer, num_sample):
seed_val = 42
random.seed(seed_val)
np.random.seed(seed_val)
torch.manual_seed(seed_val)
torch.cuda.manual_seed_all(seed_val)
if torch.cuda.is_available():
device = torch.device("cuda")
else:
device = torch.device("cpu")
total_t0 = time.time()
# For each epoch...
for epoch_i in range(0, epochs):
print("")
print(f'======== Epoch {epoch_i + 1} / {epochs} ========')
print('Training...')
t0 = time.time()
total_train_loss = 0
model.train()
for step, batch in enumerate(train_dataloader):
if step % 40 == 0 and not step == 0:
elapsed = format_time(time.time() - t0)
print(f' Batch {step} of {len(train_dataloader)}. Elapsed: {elapsed}.')
b_input_ids = batch[0].to(device)
b_input_mask = batch[1].to(device)
b_token_type_ids = batch[2].to(device)
b_labels = batch[3].to(device)
model.zero_grad()
result = model(b_input_ids,
token_type_ids=b_token_type_ids,
attention_mask=b_input_mask,
labels=b_labels,
return_dict=True)
loss = result.loss
logits = result.logits
total_train_loss += loss.item()
loss.backward()
torch.nn.utils.clip_grad_norm_(model.parameters(), 1.0)
optimizer.step()
scheduler.step()
avg_train_loss = total_train_loss / len(train_dataloader)
training_time = format_time(time.time() - t0)
print("")
print(f" Average training loss: {avg_train_loss}")
print(f" Training epcoh took: {training_time}")
# ========================================
# Validation
# ========================================
print("")
print("Running Validation...")
t0 = time.time()
model.eval()
total_eval_accuracy = 0
total_eval_loss = 0
nb_eval_steps = 0
for batch in validation_dataloader:
b_input_ids = batch[0].to(device)
b_input_mask = batch[1].to(device)
b_token_type_ids = batch[2].to(device)
b_labels = batch[3].to(device)
with torch.no_grad():
result = model(b_input_ids,
token_type_ids=b_token_type_ids,
attention_mask=b_input_mask,
labels=b_labels,
return_dict=True)
loss = result.loss
logits = result.logits
total_eval_loss += loss.item()
logits = logits.detach().cpu().numpy()
label_ids = b_labels.to('cpu').numpy()
total_eval_accuracy += flat_accuracy(logits, label_ids)
avg_val_accuracy = total_eval_accuracy / len(validation_dataloader)
print(f" Accuracy: {avg_val_accuracy}")
avg_val_loss = total_eval_loss / len(validation_dataloader)
validation_time = format_time(time.time() - t0)
print(f" Validation Loss: {avg_val_loss}")
print(" Validation took: {validation_time}")
print("")
print("Training complete!")
model.save_pretrained(f"save_model/mode_{num_sample}")
@click.command()
@click.option('--num_sample', default=60000, help="Maximum number of sample from train data.")
def main(num_sample):
ts = time.time()
with open(f"train_dataset_{num_sample}.pkl", 'rb') as f:
train_dataset = pickle.load(f)
with open(f"val_dataset_{num_sample}.pkl", 'rb') as f:
val_dataset = pickle.load(f)
batch_size = 32
train_dataloader = DataLoader(
train_dataset,
sampler = RandomSampler(train_dataset),
batch_size = batch_size
)
validation_dataloader = DataLoader(
val_dataset,
sampler = SequentialSampler(val_dataset),
batch_size = batch_size
)
model = BertForSequenceClassification.from_pretrained(
'bert-base-uncased',
)
# model = BertForNextSentencePrediction.from_pretrained(
# 'bert-base-uncased',
# )
model.cuda()
optimizer = AdamW(model.parameters(),
lr = 1e-5,
eps = 1e-8
)
epochs = 1
total_steps = len(train_dataloader) * epochs
scheduler = get_linear_schedule_with_warmup(optimizer,
num_warmup_steps = 0,
num_training_steps = total_steps)
# Train Model
train_model(model, train_dataloader, validation_dataloader, epochs, scheduler, optimizer, num_sample)
if __name__=='__main__':
main()