File size: 6,913 Bytes
7587354
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
import csv
import torch
import random
import argparse
import numpy as np

from tqdm import tqdm
from types import SimpleNamespace
from torch.utils.data import Dataset, DataLoader
from sklearn.metrics import f1_score, accuracy_score

from bert import BertModel
from optimizer import AdamW
from classifier import seed_everything, tokenizer
from classifier import SentimentDataset, BertSentimentClassifier


TQDM_DISABLE = False


class TwitterDataset(Dataset):
    def __init__(self, dataset, args):
        self.dataset = dataset
        self.p = args

    def __len__(self):
        return len(self.dataset)
    
    def __getitem__(self, idx):
        return self.dataset[idx]
    
    def pad_data(self, sents):
        encoding = tokenizer(sents, return_tensors='pt', padding=True, truncation=True)
        token_ids = torch.LongTensor(encoding['input_ids'])
        attension_mask = torch.LongTensor(encoding['attention_mask'])

        return token_ids, attension_mask
    
    def collate_fn(self, sents):
        token_ids, attention_mask = self.pad_data(sents)

        batched_data = {
            'token_ids': token_ids,
            'attention_mask': attention_mask,
        }

        return batched_data


def load_data(filename, flag='train'):
    '''
    - for Twitter dataset: list of sentences
    - for SST/CFIMDB dataset: list of (sent, [label], sent_id)
    '''
    num_labels = set()
    data = []
    with open(filename, 'r') as fp:
        for record in csv.DictReader(fp, delimiter = ',', ):
            if flag == 'twitter':
                sent = record['clean_text'].lower().strip()
                data.append(sent)
            elif flag == 'test':
                sent = record['sentence'].lower().strip()
                sent_id = record['id'].lower().strip()
                data.append((sent,sent_id))
            else:
                sent = record['sentence'].lower().strip()
                sent_id = record['id'].lower().strip()
                label = int(record['sentiment'].strip())
                num_labels.add(label)
                data.append((sent, label, sent_id))
        print(f"load {len(data)} data from {filename}")

    if flag == 'train':
        return data, len(num_labels)
    else:
        return data


def save_model(model, optimizer, args, config, filepath):
    save_info = {
        'model': model.state_dict(),
        'optim': optimizer.state_dict(),
        'args': args,
        'model_config': config,
        'system_rng': random.getstate(),
        'numpy_rng': np.random.get_state(),
        'torch_rng': torch.random.get_rng_state(),
    }

    torch.save(save_info, filepath)
    print(f"save the model to {filepath}")


def train(args):
    '''
    Training Pipeline
    -----------------
    1. Load the Twitter Sentiment and SST Dataset.
    2. Determine batch_size (64) and number of batches (?).
    3. Initialize SentimentClassifier (including bert).
    4. Looping through 10 epoches.
    5. Finetune minBERT with SimCSE loss function.
    6. Finetune Classifier with cross-entropy function.
    7. Backpropagation using Adam Optimizer for both.
    8. Evaluating the model on dev dataset.
    9. If dev_acc > best_dev_acc: save_model(...)
    '''

    twitter_data = load_data(args.train_bert, 'twitter')
    train_data, num_labels = load_data(args.train, 'train')
    dev_data = load_data(args.dev, 'valid')

    twitter_dataset = TwitterDataset(twitter_data, args)
    train_dataset = SentimentDataset(train_data, args)
    dev_dataset = SentimentDataset(dev_data, args)

    twitter_dataloader = DataLoader(twitter_dataset, shuffle=True, batch_size=args.batch_size_cse,
                                    num_workers=args.num_cpu_cores, collate_fn=twitter_dataset.collate_fn)
    train_dataloader = DataLoader(train_dataset, shuffle=True, batch_size=args.batch_size_classifier,
                                  num_workers=args.num_cpu_cores, collate_fn=train_dataset.collate_fn)
    dev_dataloader = DataLoader(dev_dataset, shuffle=False, batch_size=args.batch_size_classifier,
                                num_workers=args.num_cpu_cores, collate_fn=dev_dataset.collate_fn)
    
    config = SimpleNamespace(
        hidden_dropout_prob=args.hidden_dropout_prob,
        num_labels=num_labels,
        hidden_size=768,
        data_dir='.',
        fine_tune_mode='full-model'
    )

    device = torch.device('cuda') if args.use_gpu else torch.device('cpu')
    model = BertSentimentClassifier(config)
    model = model.to(device)

    optimizer_cse = AdamW(model.bert.parameters(), lr=args.lr_cse)
    optimizer_classifier = AdamW(model.parameters(), lr=args.lr_classifier)
    best_dev_acc = 0

    for epoch in range(args.epochs):
        model.bert.train()
        train_loss = num_batches = 0
        for batch in tqdm(twitter_dataloader, f'train-twitter-{epoch}', leave=False, disable=TQDM_DISABLE):
            b_ids, b_mask = batch['token_ids'], batch['attention_mask']
            b_ids = b_ids.to(device)
            b_mask = b_mask.to(device)

            optimizer_cse.zero_grad()
            logits = model.bert.embed(b_ids)
            logits = model.bert.encode(logits, b_mask)




def get_args():
    parser = argparse.ArgumentParser()
    parser.add_argument("--seed", type=int, default=11711)
    parser.add_argument("--num-cpu-cores", type=int, default=4)
    parser.add_argument("--epochs", type=int, default=10)
    parser.add_argument("--use_gpu", action='store_true')
    parser.add_argument("--batch_size_cse", help="'unsup': 64, 'sup': 512", type=int)
    parser.add_argument("--batch_size_classifier", help="'sst': 64, 'cfimdb': 8", type=int)
    parser.add_argument("--hidden_dropout_prob", type=float, default=0.3)
    parser.add_argument("--lr_cse", default=2e-5)
    parser.add_argument("--lr_classifier", default=1e-5)

    args = parser.parse_args()
    return args


if __name__ == "__main__":
    args = get_args()
    seed_everything(args.seed)
    torch.set_num_threads(args.num_cpu_cores)

    print('Finetuning minBERT with Unsupervised SimCSE...')
    config = SimpleNamespace(
        filepath='contrastive-nli.pt',
        lr=args.lr,
        num_cpu_cores=args.num_cpu_cores,
        use_gpu=args.use_gpu,
        epochs=args.epochs,
        batch_size_cse=args.batch_size_cse,
        batch_size_classifier=args.batch_size_classifier,
        train_bert='data/twitter-unsup.csv',
        train='data/ids-sst-train.csv',
        dev='data/ids-sst-dev.csv',
        test='data/ids-sst-test-student.csv',
        dev_out = 'predictions/' + args.fine_tune_mode + '-sst-dev-out.csv',
        test_out = 'predictions/' + args.fine_tune_mode + '-sst-test-out.csv'
    )
    
    train(config)

    # model = BertModel.from_pretrained('bert-base-uncased')

    # model.eval()

    # s = set()
    # for param in model.parameters():
    #     s.add(param.requires_grad)
    
    # print(s)