-
Notifications
You must be signed in to change notification settings - Fork 1
/
engine_pretrain.py
158 lines (130 loc) · 6.99 KB
/
engine_pretrain.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
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
import math
import sys
from typing import Iterable
import torch
import torch.utils.checkpoint as cp
import util.misc as misc
import util.lr_sched as lr_sched
import numpy as np
from llama import LLaMA_adapter
def train_one_epoch(model: LLaMA_adapter,
data_loader: Iterable, optimizer: torch.optim.Optimizer,
device: torch.device, epoch: int, loss_scaler,
log_writer=None,
args=None):
model.train(True)
model.module.set_default_trainability()
metric_logger = misc.MetricLogger(delimiter=" ")
metric_logger.add_meter('lr', misc.SmoothedValue(window_size=1, fmt='{value:.6f}'))
header = 'Epoch: [{}]'.format(epoch)
print_freq = 10
use_checkpoint = args.use_checkpoint
accum_iter = args.accum_iter
optimizer.zero_grad()
if log_writer is not None:
print('log_dir: {}'.format(log_writer.log_dir))
# print(device)
for data_iter_step, (examples, labels, example_mask, imgs, Keyword) in enumerate(metric_logger.log_every(data_loader, print_freq, header)):
# we use a per iteration (instead of per epoch) lr scheduler
if data_iter_step % accum_iter == 0:
lr_sched.adjust_learning_rate(optimizer, data_iter_step / len(data_loader) + epoch, args)
imgs = imgs.to(device, non_blocking=True).half()
# for index in range(len(example_mask)):
coocurrence = np.array(
[[iDesc == iiDesc for iDesc in Keyword] for iiDesc in Keyword], np.float32)
target = torch.tensor(coocurrence / coocurrence.sum(-1)).to(device).to(torch.float32)
example_mask[0] = example_mask[0].type(torch.LongTensor).to(device)
example_mask[1] = example_mask[1].type(torch.LongTensor).to(device)
example_mask[2] = example_mask[2].type(torch.LongTensor).to(device)
example_mask[3] = example_mask[3].type(torch.LongTensor).to(device)
example_mask[4] = example_mask[4].type(torch.LongTensor).to(device)
example_mask[5] = example_mask[5].type(torch.LongTensor).to(device)
examples = examples.to(device, non_blocking=True)
labels = labels.to(device, non_blocking=True)
with torch.cuda.amp.autocast():
c_loss, m_loss, clip_loss = model(examples, labels, imgs, example_mask, Keyword, target)
loss = c_loss + m_loss + clip_loss
loss_value = loss.item()
c_loss_value = c_loss.item()
m_loss_value = m_loss.item()
clip_loss_value = clip_loss.item()
if not math.isfinite(loss_value):
print("Loss is {}, stopping training".format(loss_value))
sys.exit(1)
loss /= accum_iter
loss_scaler(loss, optimizer, clip_grad = 10, parameters=model.parameters(),
update_grad=(data_iter_step + 1) % accum_iter == 0)
if (data_iter_step + 1) % accum_iter == 0:
optimizer.zero_grad()
torch.cuda.synchronize()
# metric_logger.update(loss=loss_value)
metric_logger.update(closs=c_loss_value)
metric_logger.update(mloss=m_loss_value)
metric_logger.update(cliploss=clip_loss_value)
lr = optimizer.param_groups[0]["lr"]
metric_logger.update(lr=lr)
# loss_value_reduce = misc.all_reduce_mean(loss_value)
c_loss_value_reduce = misc.all_reduce_mean(c_loss_value)
m_loss_value_reduce = misc.all_reduce_mean(m_loss_value)
clip_loss_value_reduce = misc.all_reduce_mean(clip_loss_value)
if log_writer is not None and (data_iter_step + 1) % accum_iter == 0:
""" We use epoch_1000x as the x-axis in tensorboard.
This calibrates different curves when batch size changes.
"""
epoch_1000x = int((data_iter_step / len(data_loader) + epoch) * 1000)
# log_writer.add_scalar('train_loss', loss_value_reduce, epoch_1000x)
log_writer.add_scalar('c_train_loss', c_loss_value_reduce, epoch_1000x)
log_writer.add_scalar('m_train_loss', m_loss_value_reduce, epoch_1000x)
log_writer.add_scalar('clip_train_loss', clip_loss_value_reduce, epoch_1000x)
log_writer.add_scalar('lr', lr, epoch_1000x)
# gather the stats from all processes
metric_logger.synchronize_between_processes()
print("Averaged stats:", metric_logger)
return {k: meter.global_avg for k, meter in metric_logger.meters.items()}
def val_one_epoch(model: torch.nn.Module,
data_loader: Iterable, optimizer: torch.optim.Optimizer,
device: torch.device, epoch: int, loss_scaler,
log_writer=None,
args=None):
model.eval()
metric_logger = misc.MetricLogger(delimiter=" ")
metric_logger.add_meter('lr', misc.SmoothedValue(window_size=1, fmt='{value:.6f}'))
header = 'Epoch: [{}]'.format(epoch)
print_freq = 10
accum_iter = args.accum_iter
if log_writer is not None:
print('log_dir: {}'.format(log_writer.log_dir))
for data_iter_step, (examples, labels, example_mask, imgs) in enumerate(metric_logger.log_every(data_loader, print_freq, header)):
imgs = imgs.to(device, non_blocking=True)
example_mask = example_mask.type(torch.LongTensor).to(device)
with torch.no_grad():
# c_loss, m_loss, _, _ = model(examples, labels, imgs)
c_loss, m_loss = model(examples, labels, imgs, example_mask)
# loss = c_loss + m_loss * 0
loss = c_loss + m_loss
loss_value = loss.item()
c_loss_value = c_loss.item()
m_loss_value = m_loss.item()
if not math.isfinite(loss_value):
print("Loss is {}, stopping training".format(loss_value))
sys.exit(1)
metric_logger.update(closs=c_loss_value)
metric_logger.update(mloss=m_loss_value)
lr = optimizer.param_groups[0]["lr"]
metric_logger.update(lr=lr)
# loss_value_reduce = misc.all_reduce_mean(loss_value)
c_loss_value_reduce = misc.all_reduce_mean(c_loss_value)
m_loss_value_reduce = misc.all_reduce_mean(m_loss_value)
if log_writer is not None and (data_iter_step + 1) % accum_iter == 0:
""" We use epoch_1000x as the x-axis in tensorboard.
This calibrates different curves when batch size changes.
"""
epoch_1000x = int((data_iter_step / len(data_loader) + epoch) * 1000)
# log_writer.add_scalar('val_loss', loss_value_reduce, epoch_1000x)
log_writer.add_scalar('c_train_loss', c_loss_value_reduce, epoch_1000x)
log_writer.add_scalar('m_train_loss', m_loss_value_reduce, epoch_1000x)
log_writer.add_scalar('lr', lr, epoch_1000x)
# gather the stats from all processes
metric_logger.synchronize_between_processes()
print("Averaged stats:", metric_logger)
return {k: meter.global_avg for k, meter in metric_logger.meters.items()}