-
Notifications
You must be signed in to change notification settings - Fork 0
/
trainer.py
163 lines (135 loc) · 5.02 KB
/
trainer.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
159
160
161
162
163
import os
import yaml
import numpy as np
from tqdm import tqdm
from sklearn.metrics import f1_score, classification_report
import torch
import torch.nn as nn
from iteration import step, configure_optimizers
from utils.load_checkpoint import load_checkpoint
from transformer.model import TransformerClassifier
from utils.dataloader import get_dataloader_task1, get_dataloader_task2
with open("./config.yaml") as file:
config = yaml.safe_load(file)
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
model_dir = config["model"]["model_loc"]
if not os.path.exists(model_dir):
os.makedirs(model_dir)
model_dir = os.path.join(model_dir, config["dataset"]["file_name"])
if not os.path.exists(model_dir):
os.makedirs(model_dir)
model_dir = os.path.join(model_dir, config["model"]["model"].split("/")[-1])
if not os.path.exists(model_dir):
os.makedirs(model_dir)
_model = TransformerClassifier(
config["model"]["model"],
hidden_states=config["hyperparameters"]["hidden_layers"],
dropout=config["hyperparameters"]["dropout"],
).to(device)
train_dataloader, val_dataloader, class_wt = get_dataloader_task1(
config["dataset"]["data_dir"],
config["dataset"]["file_name"],
config["model"]["model"],
config["hyperparameters"]["batch_size"],
config["hyperparameters"]["max_len"],
)
criterion = nn.CrossEntropyLoss(
weight=class_wt.to(device) if config["hyperparameters"]["use_weights"] else None
)
optimizer, scheduler = configure_optimizers(
_model,
train_dataloader,
config["hyperparameters"]["lr"],
config["hyperparameters"]["epochs"],
)
_model, optimizer, scheduler, best_weighted_f1, start_epoch = load_checkpoint(
config["model"]["ckpt"],
config["model"]["model"],
model_dir,
device,
_model,
optimizer,
scheduler,
)
total_epochs = config["hyperparameters"]["epochs"] + start_epoch
for epoch in range(start_epoch, total_epochs):
train_loss, train_acc = [], []
val_loss, val_acc = [], []
#### TRAIN STEP ####
_model.train()
with tqdm(
train_dataloader, desc="train-{}/{}".format(epoch, total_epochs - 1)
) as tepoch:
tepoch.set_postfix(loss=0.0, acc=0.0)
for batch_idx, batch in enumerate(tepoch):
details, _, _, _, _ = step(_model, batch, criterion, device)
optimizer.zero_grad()
details["loss"].backward()
optimizer.step()
scheduler.step()
train_loss.append(details["loss"].item())
train_acc.append(details["accuracy"].item())
tepoch.set_postfix(
loss=details["loss"].item(), acc=np.array(train_acc).mean()
)
#### VAL STEP ####
_model.eval()
y_preds, y_test = np.array([]), np.array([])
with torch.set_grad_enabled(False):
with tqdm(
val_dataloader, desc="val-{}/{}".format(epoch, total_epochs - 1)
) as vepoch:
vepoch.set_postfix(loss=0.0, acc=0.0)
for batch_idx, batch in enumerate(vepoch):
details, ypred, ytrue, _, _ = step(_model, batch, criterion, device)
y_preds = np.hstack((y_preds, ypred.cpu().numpy()))
y_test = np.hstack((y_test, ytrue.to("cpu").numpy()))
val_loss.append(details["loss"].item())
val_acc.append(details["accuracy"].item())
vepoch.set_postfix(
loss=details["loss"].item(), acc=np.array(val_acc).mean()
)
weighted_f1 = f1_score(y_test, y_preds, average="weighted")
# avg_val_acc = np.array(val_acc).mean()
# if best_val_acc <= avg_val_acc:
# best_val_acc = avg_val_acc
# torch.save(
# {
# config["model"]["model"]: _model.state_dict(),
# "scheduler": scheduler.state_dict(),
# "optimizer": optimizer.state_dict(),
# "epoch": start_epoch + epoch,
# "val_acc": best_val_acc,
# },
# os.path.join(model_dir, f"{epoch}_{int(avg_val_acc*100.)}.pkl"),
# )
if best_weighted_f1 <= weighted_f1:
best_weighted_f1 = weighted_f1
torch.save(
{
config["model"]["model"]: _model.state_dict(),
"scheduler": scheduler.state_dict(),
"optimizer": optimizer.state_dict(),
"epoch": start_epoch + epoch,
"weighted_f1": weighted_f1,
},
os.path.join(model_dir, f"{epoch}_{int(weighted_f1*100.)}.pkl"),
)
print(
"Epoch {:.3f} - train loss: {:.3f}, train acc: {:.3f}, val loss: {:.3f}, val acc: {:.3f}, wted-f1: {:.3f}".format(
epoch,
np.array(train_loss).mean(),
np.array(train_acc).mean() * 100.0,
np.array(val_loss).mean(),
np.array(val_acc).mean() * 100.0,
weighted_f1 * 100.0,
)
)
print(
classification_report(
y_test,
y_preds,
target_names=["OFF", "NOT"],
)
)
print("--" * 30)