File size: 8,010 Bytes
7900c16
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
"""
This script provides an example to use DeepSpeed for classification.
"""
import sys
import os
import random
import argparse
import torch
import torch.nn as nn
import deepspeed
import torch.distributed as dist

tencentpretrain_dir = os.path.abspath(os.path.join(os.path.dirname(__file__), ".."))
sys.path.append(tencentpretrain_dir)

from tencentpretrain.opts import deepspeed_opts
from finetune.run_classifier import *


def read_dataset(args, path, split):
    dataset, columns = [], {}
    if split:
        for i in range(args.world_size):
            dataset.append([])
        index = 0
    with open(path, mode="r", encoding="utf-8") as f:
        for line_id, line in enumerate(f):
            if line_id == 0:
                for i, column_name in enumerate(line.rstrip("\r\n").split("\t")):
                    columns[column_name] = i
                continue
            line = line.rstrip("\r\n").split("\t")
            tgt = int(line[columns["label"]])
            if args.soft_targets and "logits" in columns.keys():
                soft_tgt = [float(value) for value in line[columns["logits"]].split(" ")]
            if "text_b" not in columns:  # Sentence classification.
                text_a = line[columns["text_a"]]
                src = args.tokenizer.convert_tokens_to_ids([CLS_TOKEN] + args.tokenizer.tokenize(text_a) + [SEP_TOKEN])
                seg = [1] * len(src)
            else:  # Sentence-pair classification.
                text_a, text_b = line[columns["text_a"]], line[columns["text_b"]]
                src_a = args.tokenizer.convert_tokens_to_ids([CLS_TOKEN] + args.tokenizer.tokenize(text_a) + [SEP_TOKEN])
                src_b = args.tokenizer.convert_tokens_to_ids(args.tokenizer.tokenize(text_b) + [SEP_TOKEN])
                src = src_a + src_b
                seg = [1] * len(src_a) + [2] * len(src_b)

            if len(src) > args.seq_length:
                src = src[: args.seq_length]
                seg = seg[: args.seq_length]
            PAD_ID = args.tokenizer.convert_tokens_to_ids([PAD_TOKEN])[0]
            while len(src) < args.seq_length:
                src.append(PAD_ID)
                seg.append(0)
            if split:
                if args.soft_targets and "logits" in columns.keys():
                    dataset[index].append((src, tgt, seg, soft_tgt))
                else:
                    dataset[index].append((src, tgt, seg))
                index += 1
                if index == args.world_size:
                    index = 0
            else:
                if args.soft_targets and "logits" in columns.keys():
                    dataset.append((src, tgt, seg, soft_tgt))
                else:
                    dataset.append((src, tgt, seg))
    if split:
        max_data_num_rank_index = 0
        max_data_num = len(dataset[0])
        for i in range(args.world_size):
            if len(dataset[i]) > max_data_num:
                max_data_num_rank_index = i
                max_data_num = len(dataset[i])
        for i in range(args.world_size):
            if len(dataset[i]) < max_data_num:
                dataset[i].append(dataset[max_data_num_rank_index][-1])

    return dataset


def train_model(args, model, optimizer, scheduler, src_batch, tgt_batch, seg_batch, soft_tgt_batch=None):
    model.zero_grad()

    src_batch = src_batch.to(args.device)
    tgt_batch = tgt_batch.to(args.device)
    seg_batch = seg_batch.to(args.device)
    if soft_tgt_batch is not None:
        soft_tgt_batch = soft_tgt_batch.to(args.device)

    loss, _ = model(src_batch, tgt_batch, seg_batch, soft_tgt_batch)
    if torch.cuda.device_count() > 1:
        loss = torch.mean(loss)

    model.backward(loss)

    model.step()

    return loss


def main():
    parser = argparse.ArgumentParser(formatter_class=argparse.ArgumentDefaultsHelpFormatter)

    finetune_opts(parser)

    parser.add_argument("--world_size", type=int, default=1,
                        help="Total number of processes (GPUs) for training.")

    tokenizer_opts(parser)

    parser.add_argument("--soft_targets", action='store_true',
                        help="Train model with logits.")
    parser.add_argument("--soft_alpha", type=float, default=0.5,
                        help="Weight of the soft targets loss.")

    deepspeed_opts(parser)

    args = parser.parse_args()

    # Load the hyperparameters from the config file.
    args = load_hyperparam(args)

    set_seed(args.seed)

    # Count the number of labels.
    args.labels_num = count_labels_num(args.train_path)

    # Build tokenizer.
    args.tokenizer = str2tokenizer[args.tokenizer](args)

    # Build classification model.
    model = Classifier(args)

    # Load or initialize parameters.
    load_or_initialize_parameters(args, model)

    # Get logger.
    args.logger = init_logger(args)

    param_optimizer = list(model.named_parameters())
    no_decay = ["bias", "gamma", "beta"]
    optimizer_grouped_parameters = [
        {"params": [p for n, p in param_optimizer if not any(nd in n for nd in no_decay)], "weight_decay": 0.01},
        {"params": [p for n, p in param_optimizer if any(nd in n for nd in no_decay)], "weight_decay": 0.0},
    ]

    deepspeed.init_distributed()
    rank = dist.get_rank()
    args.rank = rank

    trainset = read_dataset(args, args.train_path, split=True)[args.rank]
    random.shuffle(trainset)
    instances_num = len(trainset)
    batch_size = args.batch_size
    args.train_steps = int(instances_num * args.epochs_num / batch_size) + 1

    custom_optimizer, custom_scheduler = build_optimizer(args, model)

    model, optimizer, _, scheduler = deepspeed.initialize(
        model=model,
        model_parameters=optimizer_grouped_parameters,
        args=args,
        optimizer=custom_optimizer,
        lr_scheduler=custom_scheduler,
        mpu=None,
        dist_init_required=False)

    src = torch.LongTensor([example[0] for example in trainset])
    tgt = torch.LongTensor([example[1] for example in trainset])
    seg = torch.LongTensor([example[2] for example in trainset])
    if args.soft_targets:
        soft_tgt = torch.FloatTensor([example[3] for example in trainset])
    else:
        soft_tgt = None

    args.model = model
    args.device = torch.device("cuda" if torch.cuda.is_available() else "cpu")

    total_loss, result, best_result, best_epoch = 0.0, 0.0, 0.0, 0

    result_tensor = torch.tensor(result).to(args.device)
    if args.rank == 0:
        args.logger.info("Batch size: {}".format(batch_size))
        args.logger.info("The number of training instances: {}".format(instances_num))
        args.logger.info("Start training.")

    for epoch in range(1, args.epochs_num + 1):
        model.train()
        for i, (src_batch, tgt_batch, seg_batch, soft_tgt_batch) in enumerate(batch_loader(batch_size, src, tgt, seg, soft_tgt)):
            loss = train_model(args, model, optimizer, scheduler, src_batch, tgt_batch, seg_batch, soft_tgt_batch)
            total_loss += loss.item()
            if (i + 1) % args.report_steps == 0 and args.rank == 0:
                args.logger.info("Epoch id: {}, Training steps: {}, Avg loss: {:.3f}".format(epoch, i + 1, total_loss / args.report_steps))
                total_loss = 0.0
        if args.rank == 0:
            result = evaluate(args, read_dataset(args, args.dev_path, split=False))
            result_tensor = torch.tensor(result[0]).to(args.device)
        dist.broadcast(result_tensor, 0, async_op=False)
        if result_tensor.float() >= best_result:
            best_result = result_tensor.float().item()
            best_epoch = epoch
        model.save_checkpoint(args.output_model_path, str(epoch))

    # Evaluation phase.
    if args.test_path is not None and args.rank == 0:
        args.logger.info("Test set evaluation.")
        model.load_checkpoint(args.output_model_path, str(best_epoch))
        evaluate(args, read_dataset(args, args.test_path, split=False), True)


if __name__ == "__main__":
    main()