-
Notifications
You must be signed in to change notification settings - Fork 0
/
kd_utils.py
224 lines (177 loc) · 7.07 KB
/
kd_utils.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
"""
Tensorboard logger code referenced from:
https://github.com/yunjey/pytorch-tutorial/blob/master/tutorials/04-utils/
Other helper functions:
https://github.com/cs230-stanford/cs230-stanford.github.io
"""
import json
import logging
import os
import shutil
import torch
from collections import OrderedDict
#import tensorflow as tf
import numpy as np
import scipy.misc
from io import BytesIO # Python 3.x
class Params():
"""Class that loads hyperparameters from a json file.
Example:
```
params = Params(json_path)
print(params.learning_rate)
params.learning_rate = 0.5 # change the value of learning_rate in params
```
"""
def __init__(self, json_path):
with open(json_path) as f:
params = json.load(f)
self.__dict__.update(params)
def save(self, json_path):
with open(json_path, 'w') as f:
json.dump(self.__dict__, f, indent=4)
def update(self, json_path):
"""Loads parameters from json file"""
with open(json_path) as f:
params = json.load(f)
self.__dict__.update(params)
@property
def dict(self):
"""Gives dict-like access to Params instance by `params.dict['learning_rate']"""
return self.__dict__
class RunningAverage():
"""A simple class that maintains the running average of a quantity
Example:
```
loss_avg = RunningAverage()
loss_avg.update(2)
loss_avg.update(4)
loss_avg() = 3
```
"""
def __init__(self):
self.steps = 0
self.total = 0
def update(self, val):
self.total += val
self.steps += 1
def __call__(self):
return self.total/float(self.steps)
def set_logger(log_path):
"""Set the logger to log info in terminal and file `log_path`.
In general, it is useful to have a logger so that every output to the terminal is saved
in a permanent file. Here we save it to `model_dir/train.log`.
Example:
```
logging.info("Starting training...")
```
Args:
log_path: (string) where to log
"""
logger = logging.getLogger()
logger.setLevel(logging.INFO)
if not logger.handlers:
# Logging to a file
file_handler = logging.FileHandler(log_path)
file_handler.setFormatter(logging.Formatter('%(asctime)s:%(levelname)s: %(message)s'))
logger.addHandler(file_handler)
# Logging to console
stream_handler = logging.StreamHandler()
stream_handler.setFormatter(logging.Formatter('%(message)s'))
logger.addHandler(stream_handler)
def save_dict_to_json(d, json_path):
"""Saves dict of floats in json file
Args:
d: (dict) of float-castable values (np.float, int, float, etc.)
json_path: (string) path to json file
"""
with open(json_path, 'w') as f:
# We need to convert the values to float for json (it doesn't accept np.array, np.float, )
d = {k: float(v) for k, v in d.items()}
json.dump(d, f, indent=4)
def save_checkpoint(state, is_best, checkpoint):
"""Saves model and training parameters at checkpoint + 'last.pth.tar'. If is_best==True, also saves
checkpoint + 'best.pth.tar'
Args:
state: (dict) contains model's state_dict, may contain other keys such as epoch, optimizer state_dict
is_best: (bool) True if it is the best model seen till now
checkpoint: (string) folder where parameters are to be saved
"""
filepath = os.path.join(checkpoint, 'last.pth.tar')
if not os.path.exists(checkpoint):
print("Checkpoint Directory does not exist! Making directory {}".format(checkpoint))
os.mkdir(checkpoint)
else:
print("Checkpoint Directory exists! ")
torch.save(state, filepath)
if is_best:
shutil.copyfile(filepath, os.path.join(checkpoint, 'best.pth.tar'))
def load_checkpoint(checkpoint, model, optimizer=None):
"""Loads model parameters (state_dict) from file_path. If optimizer is provided, loads state_dict of
optimizer assuming it is present in checkpoint.
Args:
checkpoint: (string) filename which needs to be loaded
model: (torch.nn.Module) model for which the parameters are loaded
optimizer: (torch.optim) optional: resume optimizer from checkpoint
"""
if not os.path.exists(checkpoint):
raise("File doesn't exist {}".format(checkpoint))
if torch.cuda.is_available():
checkpoint = torch.load(checkpoint)
else:
# this helps avoid errors when loading single-GPU-trained weights onto CPU-model
checkpoint = torch.load(checkpoint, map_location=lambda storage, loc: storage)
model.load_state_dict(checkpoint['state_dict'])
if optimizer:
optimizer.load_state_dict(checkpoint['optim_dict'])
return checkpoint
# class Board_Logger(object):
# """Tensorboard log utility"""
# def __init__(self, log_dir):
# """Create a summary writer logging to log_dir."""
# self.writer = tf.summary.FileWriter(log_dir)
# def scalar_summary(self, tag, value, step):
# """Log a scalar variable."""
# summary = tf.Summary(value=[tf.Summary.Value(tag=tag, simple_value=value)])
# self.writer.add_summary(summary, step)
# def image_summary(self, tag, images, step):
# """Log a list of images."""
# img_summaries = []
# for i, img in enumerate(images):
# # Write the image to a string
# try:
# s = StringIO()
# except:
# s = BytesIO()
# scipy.misc.toimage(img).save(s, format="png")
# # Create an Image object
# img_sum = tf.Summary.Image(encoded_image_string=s.getvalue(),
# height=img.shape[0],
# width=img.shape[1])
# # Create a Summary value
# img_summaries.append(tf.Summary.Value(tag='%s/%d' % (tag, i), image=img_sum))
# # Create and write Summary
# summary = tf.Summary(value=img_summaries)
# self.writer.add_summary(summary, step)
# def histo_summary(self, tag, values, step, bins=1000):
# """Log a histogram of the tensor of values."""
# # Create a histogram using numpy
# counts, bin_edges = np.histogram(values, bins=bins)
# # Fill the fields of the histogram proto
# hist = tf.HistogramProto()
# hist.min = float(np.min(values))
# hist.max = float(np.max(values))
# hist.num = int(np.prod(values.shape))
# hist.sum = float(np.sum(values))
# hist.sum_squares = float(np.sum(values**2))
# # Drop the start of the first bin
# bin_edges = bin_edges[1:]
# # Add bin edges and counts
# for edge in bin_edges:
# hist.bucket_limit.append(edge)
# for c in counts:
# hist.bucket.append(c)
# # Create and write Summary
# summary = tf.Summary(value=[tf.Summary.Value(tag=tag, histo=hist)])
# self.writer.add_summary(summary, step)
# self.writer.flush()