-
Notifications
You must be signed in to change notification settings - Fork 0
/
search_params.py
309 lines (262 loc) · 14.1 KB
/
search_params.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
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
import argparse
import json
import logging
import os
import time
from datetime import datetime
from pathlib import Path
import yaml
from ray import tune
from ray.tune.schedulers import ASHAScheduler
import numpy as np
from libmultilabel.nn import data_utils
from libmultilabel.nn.nn_utils import set_seed
from libmultilabel.common_utils import AttributeDict, Timer
from torch_trainer import TorchTrainer
logging.basicConfig(level=logging.INFO,
format='%(asctime)s %(levelname)s:%(message)s')
def train_libmultilabel_tune(config, datasets, classes, word_dict):
"""The training function for ray tune.
Args:
config (AttributeDict): Config of the experiment.
datasets (dict): A dictionary of datasets.
classes(list): List of class names.
word_dict(torchtext.vocab.Vocab): A vocab object which maps tokens to indices.
"""
set_seed(seed=config.seed)
config.run_name = tune.get_trial_dir()
logging.info(f'Run name: {config.run_name}')
config.checkpoint_dir = os.path.join(config.result_dir, config.run_name)
config.log_path = os.path.join(config.checkpoint_dir, 'logs.json')
trainer = TorchTrainer(config=config,
datasets=datasets,
classes=classes,
word_dict=word_dict,
search_params=True,
save_checkpoints=False)
trainer.train()
def load_config_from_file(config_path):
"""Initialize the model config.
Args:
config_path (str): Path to the config file.
Returns:
AttributeDict: Config of the experiment.
"""
with open(config_path) as fp:
config = yaml.safe_load(fp)
# create directories that hold the shared data
os.makedirs(config['result_dir'], exist_ok=True)
if config['embed_cache_dir']:
os.makedirs(config['embed_cache_dir'], exist_ok=True)
# set relative path to absolute path (_path, _file, _dir)
for k, v in config.items():
if isinstance(v, str) and os.path.exists(v):
config[k] = os.path.abspath(v)
# find `train.txt`, `val.txt`, and `test.txt` from the data directory if not specified.
config['train_path'] = config['train_path'] or os.path.join(config['data_dir'], 'train.txt')
config['val_path'] = config['val_path'] or os.path.join(config['data_dir'], 'valid.txt')
config['test_path'] = config['test_path'] or os.path.join(config['data_dir'], 'test.txt')
return config
def init_search_params_spaces(config, parameter_columns, prefix):
"""Initialize the sample space defined in ray tune.
See the random distributions API listed here: https://docs.ray.io/en/master/tune/api_docs/search_space.html#random-distributions-api
Args:
config (dict): Config of the experiment.
parameter_columns (dict): Names of parameters to include in the CLIReporter.
The keys are parameter names and the values are displayed names.
prefix(str): The prefix of a nested parameter such as network_config/dropout.
Returns:
dict: Config with parsed sample spaces.
"""
search_spaces = ['choice', 'grid_search', 'uniform', 'quniform', 'loguniform',
'qloguniform', 'randn', 'qrandn', 'randint', 'qrandint']
for key, value in config.items():
if isinstance(value, list) and len(value) >= 2 and value[0] in search_spaces:
search_space, search_args = value[0], value[1:]
if isinstance(search_args[0], list) and any(isinstance(x, list) for x in search_args[0]) and search_space != 'grid_search':
raise ValueError(
"""If the search values are lists, the search space must be `grid_search`.
Take `filter_sizes: ['grid_search', [[2,4,8], [4,6]]]` for example, the program will grid search over
[2,4,8] and [4,6]. This is the same as assigning `filter_sizes` to either [2,4,8] or [4,6] in two runs.
""")
else:
config[key] = getattr(tune, search_space)(*search_args)
parameter_columns[prefix+key] = key
elif isinstance(value, dict):
config[key] = init_search_params_spaces(value, parameter_columns, f'{prefix}{key}/')
return config
def init_search_algorithm(search_alg, metric=None, mode=None):
"""Specify a search algorithm and you must pip install it first.
If no search algorithm is specified, the default search algorithm is BasicVariantGenerator.
See more details here: https://docs.ray.io/en/master/tune/api_docs/suggestion.html
Args:
search_alg (str): One of 'basic_variant', 'bayesopt', or 'optuna'.
metric (str): The metric to monitor for early stopping.
mode (str): One of 'min' or 'max' to determine whether to minimize or maximize the metric.
"""
if search_alg == 'optuna':
assert metric and mode, "Metric and mode cannot be None for optuna."
from ray.tune.suggest.optuna import OptunaSearch
return OptunaSearch(metric=metric, mode=mode)
elif search_alg == 'bayesopt':
assert metric and mode, "Metric and mode cannot be None for bayesian optimization."
from ray.tune.suggest.bayesopt import BayesOptSearch
return BayesOptSearch(metric=metric, mode=mode)
logging.info(f'{search_alg} search is found, run BasicVariantGenerator().')
def prepare_retrain_config(best_config, best_log_dir, merge_train_val):
"""Prepare the configuration for re-training.
Args:
best_config (AttributeDict): The best hyper-parameter configuration.
best_log_dir (str): The directory of the best trial of the experiment.
merge_train_val (bool): Whether to merge the training and validation data.
"""
if merge_train_val:
best_config.merge_train_val = True
log_path = os.path.join(best_log_dir, 'logs.json')
if os.path.isfile(log_path):
with open(log_path) as fp:
log = json.load(fp)
else:
raise FileNotFoundError("The log directory does not contain a log.")
# For re-training with validation data,
# we use the number of epochs at the point of optimal validation performance.
log_metric = np.array([l[best_config.val_metric] for l in log['val']])
optimal_idx = log_metric.argmax() if best_config.mode == 'max' else log_metric.argmin()
best_config.epochs = optimal_idx.item() + 1 # plus 1 for epochs
else:
best_config.merge_train_val = False
def load_static_data(config, merge_train_val=False):
"""Preload static data once for multiple trials.
Args:
config (AttributeDict): Config of the experiment.
merge_train_val (bool, optional): Whether to merge the training and validation data.
Defaults to False.
Returns:
dict: A dict of static data containing datasets, classes, and word_dict.
"""
datasets = data_utils.load_datasets(train_path=config.train_path,
test_path=config.test_path,
val_path=config.val_path,
val_size=config.val_size,
merge_train_val=merge_train_val,
tokenize_text='lm_weight' not in config['network_config'],
remove_no_label_data=config.remove_no_label_data
)
return {
"datasets": datasets,
"word_dict": None if config.embed_file is None else data_utils.load_or_build_text_dict(
dataset=datasets['train'],
vocab_file=config.vocab_file,
min_vocab_freq=config.min_vocab_freq,
embed_file=config.embed_file,
embed_cache_dir=config.embed_cache_dir,
silent=config.silent,
normalize_embed=config.normalize_embed
),
"classes": data_utils.load_or_build_label(datasets, config.label_file, config.include_test_labels)
}
def retrain_best_model(exp_name, best_config, best_log_dir, merge_train_val):
"""Re-train the model with the best hyper-parameters.
A new model is trained on the combined training and validation data if `merge_train_val` is True.
If a test set is provided, it will be evaluated by the obtained model.
Args:
exp_name (str): The directory to save trials generated by ray tune.
best_config (AttributeDict): The best hyper-parameter configuration.
best_log_dir (str): The directory of the best trial of the experiment.
merge_train_val (bool): Whether to merge the training and validation data.
"""
best_config.silent = False
checkpoint_dir = os.path.join(best_config.result_dir, exp_name, 'trial_best_params')
os.makedirs(checkpoint_dir, exist_ok=True)
with open(os.path.join(checkpoint_dir, 'params.yml'), 'w') as fp:
yaml.dump(dict(best_config), fp)
quit() # do not need re-training
best_config.run_name = '_'.join(exp_name.split('_')[:-1]) + '_best'
best_config.checkpoint_dir = checkpoint_dir
best_config.log_path = os.path.join(best_config.checkpoint_dir, 'logs.json')
prepare_retrain_config(best_config, best_log_dir, merge_train_val)
set_seed(seed=best_config.seed)
data = load_static_data(best_config, merge_train_val=best_config.merge_train_val)
logging.info(f'Re-training with best config: \n{best_config}')
trainer = TorchTrainer(config=best_config, **data)
trainer.train()
if 'test' in data['datasets']:
test_results = trainer.test()
logging.info(f'Test results after re-training: {test_results}')
logging.info(f'Best model saved to {trainer.checkpoint_callback.best_model_path or trainer.checkpoint_callback.last_model_path}.')
def main():
parser = argparse.ArgumentParser()
parser.add_argument(
'--config', help='Path to configuration file (default: %(default)s). Please specify a config with all arguments in LibMultiLabel/main.py::get_config.')
parser.add_argument('--cpu_count', type=int, default=4,
help='Number of CPU per trial (default: %(default)s)')
parser.add_argument('--gpu_count', type=int, default=1,
help='Number of GPU per trial (default: %(default)s)')
parser.add_argument('--num_samples', type=int, default=50,
help='Number of running trials. If the search space is `grid_search`, the same grid will be repeated `num_samples` times. (default: %(default)s)')
parser.add_argument('--mode', default='max', choices=['min', 'max'],
help='Determines whether objective is minimizing or maximizing the metric attribute. (default: %(default)s)')
parser.add_argument('--search_alg', default=None, choices=['basic_variant', 'bayesopt', 'optuna'],
help='Search algorithms (default: %(default)s)')
parser.add_argument('--no_merge_train_val', action='store_true',
help='Do not add the validation set in re-training the final model after hyper-parameter search.')
args, _ = parser.parse_known_args()
# Load config from the config file and overwrite values specified in CLI.
parameter_columns = dict() # parameters to include in progress table of CLIReporter
config = load_config_from_file(args.config)
config = init_search_params_spaces(config, parameter_columns, prefix='')
parser.set_defaults(**config)
config = AttributeDict(vars(parser.parse_args()))
config.merge_train_val = False # no need to include validation during parameter search
# Check if the validation set is provided.
val_path = config.val_path or os.path.join(config.data_dir, 'valid.txt')
assert config.val_size > 0 or os.path.exists(val_path), \
"You should specify either a positive `val_size` or a `val_path` defaults to `data_dir/valid.txt` for parameter search."
"""Run tune analysis.
- If no search algorithm is specified, the default search algorighm is BasicVariantGenerator.
https://docs.ray.io/en/master/tune/api_docs/suggestion.html#tune-basicvariant
- Arguments without search spaces will be ignored by `tune.run`
(https://github.com/ray-project/ray/blob/34d3d9294c50aea4005b7367404f6a5d9e0c2698/python/ray/tune/suggest/variant_generator.py#L333),
so we parse the whole config to `tune.run` here for simplicity.
"""
data = load_static_data(config)
reporter = tune.CLIReporter(metric_columns=[f'val_{metric}' for metric in config.monitor_metrics],
parameter_columns=parameter_columns,
metric=f'val_{config.val_metric}',
mode=args.mode,
sort_by_metric=True)
if config.scheduler is not None:
scheduler = ASHAScheduler(metric=f'val_{config.val_metric}',
mode=args.mode,
**config.scheduler)
else:
scheduler = None
exp_name = '{}_{}_{}'.format(
config.data_name,
Path(config.config).stem if config.config else config.model_name,
datetime.now().strftime('%Y%m%d%H%M%S'),
)
analysis = tune.run(
tune.with_parameters(
train_libmultilabel_tune,
**data),
search_alg=init_search_algorithm(
config.search_alg, metric=config.val_metric, mode=args.mode),
scheduler=scheduler,
local_dir=config.result_dir,
num_samples=config.num_samples,
resources_per_trial={
'cpu': args.cpu_count, 'gpu': args.gpu_count},
progress_reporter=reporter,
config=config,
name=exp_name,
)
# Save best model after parameter search.
best_config = analysis.get_best_config(f'val_{config.val_metric}', args.mode, scope='all')
best_log_dir = analysis.get_best_logdir(f'val_{config.val_metric}', args.mode, scope='all')
retrain_best_model(exp_name, best_config, best_log_dir, merge_train_val=not args.no_merge_train_val)
if __name__ == '__main__':
# calculate wall time.
wall_time = Timer()
main()
print(f'Wall time: {wall_time.time():.2f} (s)')