-
Notifications
You must be signed in to change notification settings - Fork 87
/
functions.py
480 lines (398 loc) · 14.5 KB
/
functions.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
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
# -*- coding: utf-8 -*-
# @Date : 2019-07-25
# @Author : Xinyu Gong (xy_gong@tamu.edu)
# @Link : None
# @Version : 0.0
import logging
import operator
import os
from copy import deepcopy
import numpy as np
import torch
import torch.nn as nn
from imageio import imsave
from torchvision.utils import make_grid
from tqdm import tqdm
from utils.fid_score import calculate_fid_given_paths
from utils.inception_score import get_inception_score
logger = logging.getLogger(__name__)
def train_shared(
args,
gen_net: nn.Module,
dis_net: nn.Module,
g_loss_history,
d_loss_history,
controller,
gen_optimizer,
dis_optimizer,
train_loader,
prev_hiddens=None,
prev_archs=None,
):
dynamic_reset = False
logger.info("=> train shared GAN...")
step = 0
gen_step = 0
# train mode
gen_net.train()
dis_net.train()
# eval mode
controller.eval()
for epoch in range(args.shared_epoch):
for iter_idx, (imgs, _) in enumerate(train_loader):
# sample an arch
arch = controller.sample(
1, prev_hiddens=prev_hiddens, prev_archs=prev_archs
)[0][0]
gen_net.set_arch(arch, controller.cur_stage)
dis_net.cur_stage = controller.cur_stage
# Adversarial ground truths
real_imgs = imgs.type(torch.cuda.FloatTensor)
# Sample noise as generator input
z = torch.cuda.FloatTensor(
np.random.normal(0, 1, (imgs.shape[0], args.latent_dim))
)
# ---------------------
# Train Discriminator
# ---------------------
dis_optimizer.zero_grad()
real_validity = dis_net(real_imgs)
fake_imgs = gen_net(z).detach()
assert fake_imgs.size() == real_imgs.size(), print(
f"fake image size is {fake_imgs.size()}, "
f"while real image size is {real_imgs.size()}"
)
fake_validity = dis_net(fake_imgs)
# cal loss
d_loss = torch.mean(
nn.ReLU(inplace=True)(1.0 - real_validity)
) + torch.mean(nn.ReLU(inplace=True)(1 + fake_validity))
d_loss.backward()
dis_optimizer.step()
# add to window
d_loss_history.push(d_loss.item())
# -----------------
# Train Generator
# -----------------
if step % args.n_critic == 0:
gen_optimizer.zero_grad()
gen_z = torch.cuda.FloatTensor(
np.random.normal(0, 1, (args.gen_batch_size, args.latent_dim))
)
gen_imgs = gen_net(gen_z)
fake_validity = dis_net(gen_imgs)
# cal loss
g_loss = -torch.mean(fake_validity)
g_loss.backward()
gen_optimizer.step()
# add to window
g_loss_history.push(g_loss.item())
gen_step += 1
# verbose
if gen_step and iter_idx % args.print_freq == 0:
logger.info(
"[Epoch %d/%d] [Batch %d/%d] [D loss: %f] [G loss: %f]"
% (
epoch,
args.shared_epoch,
iter_idx % len(train_loader),
len(train_loader),
d_loss.item(),
g_loss.item(),
)
)
# check window
if g_loss_history.is_full():
if (
g_loss_history.get_var() < args.dynamic_reset_threshold
or d_loss_history.get_var() < args.dynamic_reset_threshold
):
dynamic_reset = True
logger.info("=> dynamic resetting triggered")
g_loss_history.clear()
d_loss_history.clear()
return dynamic_reset
step += 1
return dynamic_reset
def train(
args,
gen_net: nn.Module,
dis_net: nn.Module,
gen_optimizer,
dis_optimizer,
gen_avg_param,
train_loader,
epoch,
writer_dict,
schedulers=None,
):
writer = writer_dict["writer"]
gen_step = 0
# train mode
gen_net = gen_net.train()
dis_net = dis_net.train()
for iter_idx, (imgs, _) in enumerate(tqdm(train_loader)):
global_steps = writer_dict["train_global_steps"]
# Adversarial ground truths
real_imgs = imgs.type(torch.cuda.FloatTensor)
# Sample noise as generator input
z = torch.cuda.FloatTensor(
np.random.normal(0, 1, (imgs.shape[0], args.latent_dim))
)
# ---------------------
# Train Discriminator
# ---------------------
dis_optimizer.zero_grad()
real_validity = dis_net(real_imgs)
fake_imgs = gen_net(z).detach()
assert fake_imgs.size() == real_imgs.size()
fake_validity = dis_net(fake_imgs)
# cal loss
d_loss = torch.mean(nn.ReLU(inplace=True)(1.0 - real_validity)) + torch.mean(
nn.ReLU(inplace=True)(1 + fake_validity)
)
d_loss.backward()
dis_optimizer.step()
writer.add_scalar("d_loss", d_loss.item(), global_steps)
# -----------------
# Train Generator
# -----------------
if global_steps % args.n_critic == 0:
gen_optimizer.zero_grad()
gen_z = torch.cuda.FloatTensor(
np.random.normal(0, 1, (args.gen_batch_size, args.latent_dim))
)
gen_imgs = gen_net(gen_z)
fake_validity = dis_net(gen_imgs)
# cal loss
g_loss = -torch.mean(fake_validity)
g_loss.backward()
gen_optimizer.step()
# adjust learning rate
if schedulers:
gen_scheduler, dis_scheduler = schedulers
g_lr = gen_scheduler.step(global_steps)
d_lr = dis_scheduler.step(global_steps)
writer.add_scalar("LR/g_lr", g_lr, global_steps)
writer.add_scalar("LR/d_lr", d_lr, global_steps)
# moving average weight
for p, avg_p in zip(gen_net.parameters(), gen_avg_param):
avg_p.mul_(0.999).add_(0.001, p.data)
writer.add_scalar("g_loss", g_loss.item(), global_steps)
gen_step += 1
# verbose
if gen_step and iter_idx % args.print_freq == 0:
tqdm.write(
"[Epoch %d/%d] [Batch %d/%d] [D loss: %f] [G loss: %f]"
% (
epoch,
args.max_epoch,
iter_idx % len(train_loader),
len(train_loader),
d_loss.item(),
g_loss.item(),
)
)
writer_dict["train_global_steps"] = global_steps + 1
def train_controller(
args, controller, ctrl_optimizer, gen_net, prev_hiddens, prev_archs, writer_dict
):
logger.info("=> train controller...")
writer = writer_dict["writer"]
baseline = None
# train mode
controller.train()
# eval mode
gen_net.eval()
cur_stage = controller.cur_stage
for step in range(args.ctrl_step):
controller_step = writer_dict["controller_steps"]
archs, selected_log_probs, entropies = controller.sample(
args.ctrl_sample_batch, prev_hiddens=prev_hiddens, prev_archs=prev_archs
)
cur_batch_rewards = []
for arch in archs:
logger.info(f"arch: {arch}")
gen_net.set_arch(arch, cur_stage)
is_score = get_is(args, gen_net, args.rl_num_eval_img)
logger.info(f"get Inception score of {is_score}")
cur_batch_rewards.append(is_score)
cur_batch_rewards = torch.tensor(cur_batch_rewards, requires_grad=False).cuda()
cur_batch_rewards = (
cur_batch_rewards.unsqueeze(-1) + args.entropy_coeff * entropies
) # bs * 1
if baseline is None:
baseline = cur_batch_rewards
else:
baseline = (
args.baseline_decay * baseline.detach()
+ (1 - args.baseline_decay) * cur_batch_rewards
)
adv = cur_batch_rewards - baseline
# policy loss
loss = -selected_log_probs * adv
loss = loss.sum()
# update controller
ctrl_optimizer.zero_grad()
loss.backward()
ctrl_optimizer.step()
# write
mean_reward = cur_batch_rewards.mean().item()
mean_adv = adv.mean().item()
mean_entropy = entropies.mean().item()
writer.add_scalar("controller/loss", loss.item(), controller_step)
writer.add_scalar("controller/reward", mean_reward, controller_step)
writer.add_scalar("controller/entropy", mean_entropy, controller_step)
writer.add_scalar("controller/adv", mean_adv, controller_step)
writer_dict["controller_steps"] = controller_step + 1
def get_is(args, gen_net: nn.Module, num_img):
"""
Get inception score.
:param args:
:param gen_net:
:param num_img:
:return: Inception score
"""
# eval mode
gen_net = gen_net.eval()
eval_iter = num_img // args.eval_batch_size
img_list = list()
for _ in range(eval_iter):
z = torch.cuda.FloatTensor(
np.random.normal(0, 1, (args.eval_batch_size, args.latent_dim))
)
# Generate a batch of images
gen_imgs = (
gen_net(z)
.mul_(127.5)
.add_(127.5)
.clamp_(0.0, 255.0)
.permute(0, 2, 3, 1)
.to("cpu", torch.uint8)
.numpy()
)
img_list.extend(list(gen_imgs))
# get inception score
logger.info("calculate Inception score...")
mean, std = get_inception_score(img_list)
return mean
def validate(args, fixed_z, fid_stat, gen_net: nn.Module, writer_dict, clean_dir=True):
writer = writer_dict["writer"]
global_steps = writer_dict["valid_global_steps"]
# eval mode
gen_net = gen_net.eval()
# generate images
sample_imgs = gen_net(fixed_z)
img_grid = make_grid(sample_imgs, nrow=5, normalize=True, scale_each=True)
# get fid and inception score
fid_buffer_dir = os.path.join(args.path_helper["sample_path"], "fid_buffer")
os.makedirs(fid_buffer_dir, exist_ok=True)
eval_iter = args.num_eval_imgs // args.eval_batch_size
img_list = list()
for iter_idx in tqdm(range(eval_iter), desc="sample images"):
z = torch.cuda.FloatTensor(
np.random.normal(0, 1, (args.eval_batch_size, args.latent_dim))
)
# Generate a batch of images
gen_imgs = (
gen_net(z)
.mul_(127.5)
.add_(127.5)
.clamp_(0.0, 255.0)
.permute(0, 2, 3, 1)
.to("cpu", torch.uint8)
.numpy()
)
for img_idx, img in enumerate(gen_imgs):
file_name = os.path.join(fid_buffer_dir, f"iter{iter_idx}_b{img_idx}.png")
imsave(file_name, img)
img_list.extend(list(gen_imgs))
# get inception score
logger.info("=> calculate inception score")
mean, std = get_inception_score(img_list)
print(f"Inception score: {mean}")
# get fid score
logger.info("=> calculate fid score")
fid_score = calculate_fid_given_paths(
[fid_buffer_dir, fid_stat], inception_path=None
)
print(f"FID score: {fid_score}")
if clean_dir:
os.system("rm -r {}".format(fid_buffer_dir))
else:
logger.info(f"=> sampled images are saved to {fid_buffer_dir}")
writer.add_image("sampled_images", img_grid, global_steps)
writer.add_scalar("Inception_score/mean", mean, global_steps)
writer.add_scalar("Inception_score/std", std, global_steps)
writer.add_scalar("FID_score", fid_score, global_steps)
writer_dict["valid_global_steps"] = global_steps + 1
return mean, fid_score
def get_topk_arch_hidden(args, controller, gen_net, prev_archs, prev_hiddens):
"""
~
:param args:
:param controller:
:param gen_net:
:param prev_archs: previous architecture
:param prev_hiddens: previous hidden vector
:return: a list of topk archs and hiddens.
"""
logger.info(
f"=> get top{args.topk} archs out of {args.num_candidate} candidate archs..."
)
assert args.num_candidate >= args.topk
controller.eval()
cur_stage = controller.cur_stage
archs, _, _, hiddens = controller.sample(
args.num_candidate,
with_hidden=True,
prev_archs=prev_archs,
prev_hiddens=prev_hiddens,
)
hxs, cxs = hiddens
arch_idx_perf_table = {}
for arch_idx in range(len(archs)):
logger.info(f"arch: {archs[arch_idx]}")
gen_net.set_arch(archs[arch_idx], cur_stage)
is_score = get_is(args, gen_net, args.rl_num_eval_img)
logger.info(f"get Inception score of {is_score}")
arch_idx_perf_table[arch_idx] = is_score
topk_arch_idx_perf = sorted(
arch_idx_perf_table.items(), key=operator.itemgetter(1)
)[::-1][: args.topk]
topk_archs = []
topk_hxs = []
topk_cxs = []
logger.info(f"top{args.topk} archs:")
for arch_idx_perf in topk_arch_idx_perf:
logger.info(arch_idx_perf)
arch_idx = arch_idx_perf[0]
topk_archs.append(archs[arch_idx])
topk_hxs.append(hxs[arch_idx].detach().requires_grad_(False))
topk_cxs.append(cxs[arch_idx].detach().requires_grad_(False))
return topk_archs, (topk_hxs, topk_cxs)
class LinearLrDecay(object):
def __init__(self, optimizer, start_lr, end_lr, decay_start_step, decay_end_step):
assert start_lr > end_lr
self.optimizer = optimizer
self.delta = (start_lr - end_lr) / (decay_end_step - decay_start_step)
self.decay_start_step = decay_start_step
self.decay_end_step = decay_end_step
self.start_lr = start_lr
self.end_lr = end_lr
def step(self, current_step):
if current_step <= self.decay_start_step:
lr = self.start_lr
elif current_step >= self.decay_end_step:
lr = self.end_lr
else:
lr = self.start_lr - self.delta * (current_step - self.decay_start_step)
for param_group in self.optimizer.param_groups:
param_group["lr"] = lr
return lr
def load_params(model, new_param):
for p, new_p in zip(model.parameters(), new_param):
p.data.copy_(new_p)
def copy_params(model):
flatten = deepcopy(list(p.data for p in model.parameters()))
return flatten