This repository has been archived by the owner on Feb 26, 2024. It is now read-only.
-
Notifications
You must be signed in to change notification settings - Fork 63
/
config.py
73 lines (52 loc) · 2.43 KB
/
config.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
import json
class Settings(dict):
"""Experiment configuration options.
Wrapper around in-built dict class to access members through the dot operation.
Experiment parameters:
"expt_name": Name/description of experiment, used for logging.
"gpu_id": Available GPU ID(s)
"train_filepath": Training set path
"val_filepath": Validation set path
"test_filepath": Test set path
"num_nodes": Number of nodes in TSP tours
"num_neighbors": Number of neighbors in k-nearest neighbor input graph (-1 for fully connected)
"node_dim": Number of dimensions for each node
"voc_nodes_in": Input node signal vocabulary size
"voc_nodes_out": Output node prediction vocabulary size
"voc_edges_in": Input edge signal vocabulary size
"voc_edges_out": Output edge prediction vocabulary size
"beam_size": Beam size for beamsearch procedure (-1 for disabling beamsearch)
"hidden_dim": Dimension of model's hidden state
"num_layers": Number of GCN layers
"mlp_layers": Number of MLP layers
"aggregation": Node aggregation scheme in GCN (`mean` or `sum`)
"max_epochs": Maximum training epochs
"val_every": Interval (in epochs) at which validation is performed
"test_every": Interval (in epochs) at which testing is performed
"batch_size": Batch size
"batches_per_epoch": Batches per epoch (-1 for using full training set)
"accumulation_steps": Number of steps for gradient accumulation (DO NOT USE: BUGGY)
"learning_rate": Initial learning rate
"decay_rate": Learning rate decay parameter
"""
def __init__(self, config_dict):
super().__init__()
for key in config_dict:
self[key] = config_dict[key]
def __getattr__(self, attr):
return self[attr]
def __setitem__(self, key, value):
return super().__setitem__(key, value)
def __setattr__(self, key, value):
return self.__setitem__(key, value)
__delattr__ = dict.__delitem__
def get_default_config():
"""Returns default settings object.
"""
return Settings(json.load(open("./configs/default.json")))
def get_config(filepath):
"""Returns settings from json file.
"""
config = get_default_config()
config.update(Settings(json.load(open(filepath))))
return config