-
Notifications
You must be signed in to change notification settings - Fork 6
/
train_mem2seq_soccer.py
56 lines (47 loc) · 2 KB
/
train_mem2seq_soccer.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
from models.Mem2Seq import *
BLEU = False
if (args['decoder'] == "Mem2Seq"):
if args['dataset']=='kvr':
from utils.utils_kvr_mem2seq import *
BLEU = True
else:
print("You need to provide the --dataset information")
else:
if args['dataset']=='kvr':
from utils.utils_kvr import *
BLEU = True
else:
print("You need to provide the --dataset information")
avg_best,cnt,acc = 0.0,0,0.0
cnt_1 = 0
train, dev, test, testOOV, lang, max_len, max_r = prepare_data_seq(args['task'],batch_size=int(args['batch']),shuffle=True)
if args['decoder'] == "Mem2Seq":
model = globals()[args['decoder']](int(args['hidden']),
max_len,max_r,lang,args['path'],args['task'],
lr=float(args['learn']),
n_layers=int(args['layer']),
dropout=float(args['drop']),
unk_mask=bool(int(args['unk_mask']))
)
else:
model = globals()[args['decoder']](hidden_size=int(args['hidden']), max_r=max_r, n_words=lang.n_words, b_size=lang.n_words, emb_dim=lang.n_words, sos_tok=SOS_token, eos_tok=EOS_token, itos=5, kb_max_size=100
)
for epoch in range(300):
logging.info("Epoch:{}".format(epoch))
# Run the train function
pbar = tqdm(enumerate(train),total=len(train))
for i, data in pbar:
model.train_batch(data[0], data[1], data[2], data[3],data[4],data[5],
len(data[1]),10.0,0.5,i==0)
pbar.set_description(model.print_loss())
if((epoch+1) % int(args['evalp']) == 0):
acc = model.evaluate(dev,avg_best, BLEU)
if 'Mem2Seq' in args['decoder']:
model.scheduler.step(acc)
if(acc >= avg_best):
avg_best = acc
cnt=0
else:
cnt+=1
if(cnt == 5): break
if(acc == 1.0): break