-
Notifications
You must be signed in to change notification settings - Fork 0
/
globalvar.py
56 lines (42 loc) · 2.61 KB
/
globalvar.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
import argparse
import os
from os.path import join as oj
import torch
from utils import *
from pprint import pprint
from IPython import embed
def get_params():
parser = argparse.ArgumentParser(description='TagGNN')
parser.add_argument('--dataset', type = str, help='dataset name')
# parser.add_argument('--train_batch_size', type = int, default=1024, help='train_batch_size')
# parser.add_argument('--val_batch_size', type = int, default=4096, help='val_batch_size')
# parser.add_argument('--test_batch_size', type = int, default=4096, help='test_batch_size')
parser.add_argument('--train_num', type = int, default=4096, help='number of train items')
parser.add_argument('--test_num', type = int, default=4096, help='number of test items')
parser.add_argument('--val_num', type = int, default=4096, help='number of validation items')
# parser.add_argument('--label_num', type = int, default=4096, help='number of validation items')
parser.add_argument('--embedding_dim', type = int, default=100, help='node and word embedding dimension')
parser.add_argument('--drop_rate', type=float, default=0.5, help='dropout rate for GNN')
parser.add_argument('--initial_weight', type=float, default=0.01, help='initial weight for embeddings')
parser.add_argument('--lr', type = float, default=1e-3, help='learning rate')
parser.add_argument('--weight_decay', type = float, default=0, help='l2 norm')
parser.add_argument('--model', type = str, default='taggnn', choices=['taggnn-it', 'taggnn-qi', 'taggnn'], help='model name')
parser.add_argument('--max_epoch', type = int, default = 500, help='max training epoch')
parser.add_argument('--topk', type = str, default='[1,3,5]', help='evaluation topk')
parser.add_argument('--val_base_metric', type = str, default='recall', help='based on which metric for early stopping, e.g., p@10')
parser.add_argument('--early_stop_epoch', type = int, default=10, help='if the val metric does not increase in successive 10 epochs, the training will be stopped')
parser.add_argument('--only_eval', type = int, default=0, help='whether only evaluate the existing model.')
parser.add_argument('--eval_model_path', type = str, default=None, help='the load path of the trained model, if only_eval')
args, _ = parser.parse_known_args()
return args
args = get_params()
# other basic info
device = auto_gpu_setting()
args.device = device
args.topk = eval(args.topk)
args.topk.sort()
cur_time = now()
print('This Experiment started at {}'.format(cur_time))
print('Arguments Settings:')
pprint(args.__dict__)
# loading and preprocessing data