-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathct_train.py
347 lines (303 loc) · 22.1 KB
/
ct_train.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
# ---------------------------------------------------------------
# Copyright (c) 2024, NVIDIA CORPORATION. All rights reserved.
#
# This file has been modified from EDM2 and ECT codebase, which built upon EDM.
#
# Source:
# https://github.com/NVlabs/edm/blob/main/train.py (EDM)
# https://github.com/locuslab/ect/blob/main/ct_train.py (ECT)
# https://github.com/NVlabs/edm2/blob/main/train_edm2.py (EDM2)
#
# The license for these can be found in license/ directory.
# The modifications to this file are subject to the same license.
# ---------------------------------------------------------------
import os
import re
import json
import click
import torch
import dnnlib
from torch_utils import distributed as dist
from training import ct_training_loop as training_loop
import warnings
warnings.filterwarnings('ignore', 'Grad strides do not match bucket view strides') # False warning printed by PyTorch 1.12.
#----------------------------------------------------------------------------
# Scripts for auto-resuming
def create_expt_folder_with_auto_resuming(output_dir, name):
exp_dir = os.path.join(output_dir, name)
checkpoint = None
if os.path.exists(exp_dir):
all_tags = os.listdir(exp_dir)
all_existing_tags = [tag for tag in all_tags if tag.startswith('tag')]
all_existing_tags.sort()
all_existing_tags = all_existing_tags[::-1]
for previous_tag in all_existing_tags:
potential_ckpt = os.path.join(exp_dir, previous_tag, 'training-state-latest.pt')
if os.path.exists(potential_ckpt):
checkpoint = potential_ckpt
if dist.get_rank() == 0:
print('auto-resuming ckpt found ' + potential_ckpt)
break
curr_tag = 'tag' + str(len(all_existing_tags)).zfill(2)
exp_dir = os.path.join(exp_dir, curr_tag) # output/name/tagxx
else:
exp_dir = os.path.join(exp_dir, 'tag00') # output/name/tag00
dist.synchronize()
if dist.get_rank() == 0:
os.makedirs(exp_dir)
return exp_dir, checkpoint
#----------------------------------------------------------------------------
# Parse a comma separated list of numbers or ranges and return a list of ints.
# Example: '1,2,5-10' returns [1, 2, 5, 6, 7, 8, 9, 10]
def parse_int_list(s):
if isinstance(s, list): return s
ranges = []
range_re = re.compile(r'^(\d+)-(\d+)$')
for p in s.split(','):
m = range_re.match(p)
if m:
ranges.extend(range(int(m.group(1)), int(m.group(2))+1))
else:
ranges.append(int(p))
return ranges
class CommaSeparatedList(click.ParamType):
name = 'list'
def convert(self, value, param, ctx):
_ = param, ctx
if value is None or value.lower() == 'none' or value == '':
return []
return value.split(',')
#----------------------------------------------------------------------------
@click.command()
# Main options.
@click.option('--outdir', help='Where to save the results', metavar='DIR', type=str, required=True)
@click.option('--data', help='Path to the dataset', metavar='ZIP|DIR', type=str, required=True)
@click.option('--cond', help='Train class-conditional model', metavar='BOOL', type=bool, default=False, show_default=True)
@click.option('--arch', help='Network architecture', type=click.Choice(['ddpmpp', 'ncsnpp', 'adm', 'edm2-cifar-s', 'edm2-cifar-m', 'edm2-img64-s', 'edm2-img64-m', 'edm2-img64-l', 'edm2-img64-xl' ]), default='ddpmpp', show_default=True)
@click.option('--precond', help='Preconditioning & loss function', metavar='ect', type=click.Choice(['ect']), default='ect', show_default=True)
# Hyperparameters.
@click.option('--duration', help='Training duration', metavar='MIMG', type=click.FloatRange(min=0, min_open=True), default=200, show_default=True)
@click.option('--batch', help='Total batch size', metavar='INT', type=click.IntRange(min=1), default=512, show_default=True)
@click.option('--batch-gpu', help='Limit batch size per GPU', metavar='INT', type=click.IntRange(min=1))
@click.option('--cbase', help='Channel multiplier [default: varies]', metavar='INT', type=int)
@click.option('--cres', help='Channels per resolution [default: varies]', metavar='LIST', type=parse_int_list)
@click.option('--optim', help='Name of Optimizer', metavar='Optimizer', type=str, default='Adam', show_default=True)
@click.option('--lr', help='Learning rate', metavar='FLOAT', type=click.FloatRange(min=0, min_open=True), default=10e-4, show_default=True)
@click.option('--beta1', help='Adam beta1', metavar='FLOAT', type=click.FloatRange(min=0, max=1), default=0.9, show_default=True)
@click.option('--beta2', help='Adam beta2', metavar='FLOAT', type=click.FloatRange(min=0, max=1), default=0.999, show_default=True)
@click.option('--eps',help='Adam epsilon', metavar='FLOAT', type=click.FloatRange(min=0), default=1e-8, show_default=True)
@click.option('--decay_iter', help='Decay learning rate after this many iterations, 0: no decay', metavar='INT', type=click.IntRange(min=0), default=0, show_default=True)
@click.option('--rampup_iter', help='Rampup learning rate for this many iterations, 0: no rampup', metavar='INT', type=click.IntRange(min=0), default=0, show_default=True)
@click.option('--lr_schedule_start_iter', help='When computing the current iteration for lr scheduling, start from this iteration', metavar='INT', type=click.IntRange(min=0), default=0, show_default=True)
@click.option('--ema_type', help='Type of EMA', metavar='STR', default='constant', show_default=True)
@click.option('--ema_beta', help='EMA decay rate', metavar='FLOAT', type=click.FloatRange(min=0), default=0.9999, show_default=True)
@click.option('--ema_gamma', help='EMA decay rate for the rampup period', metavar='FLOAT', type=click.FloatRange(min=0), default=None, show_default=True)
@click.option('--dropout', help='Dropout probability', metavar='FLOAT', type=click.FloatRange(min=0, max=1), default=0.13, show_default=True)
@click.option('--augment', help='Augment probability', metavar='FLOAT', type=click.FloatRange(min=0, max=1), default=0., show_default=True)
@click.option('--xflip', help='Enable dataset x-flips', metavar='BOOL', type=bool, default=False, show_default=True)
# Model Hyperparameters
@click.option('--mean', help='P_mean of Log Normal Distribution', metavar='FLOAT', type=click.FloatRange(), default=-1.1, show_default=True)
@click.option('--std', help='P_std of Log Normal Distribution', metavar='FLOAT', type=click.FloatRange(), default=2.0, show_default=True)
@click.option('--t_lower', help='Lower bound of t', metavar='FLOAT', type=click.FloatRange(), default=0.002, show_default=True)
@click.option('--t_upper', help='Upper bound of t', metavar='FLOAT', type=click.FloatRange(), default=80, show_default=True)
@click.option('--tdist', help='Type of t distribution', metavar='STR', default='normal', show_default=True)
@click.option('--df', help='Degrees of freedom for t distribution', metavar='FLOAT', type=click.FloatRange(), default=2.0, show_default=True)
@click.option('--double', help='How often to reduce dt', metavar='TICKS', type=click.IntRange(min=1), default=500, show_default=True)
@click.option('--start_stage', help='Start stage', metavar='INT', type=click.IntRange(min=-1), default=0, show_default=True)
@click.option('--w_boundary', help='Weight for boundary condition', metavar='FLOAT', type=click.FloatRange(), default=1, show_default=True)
@click.option('-q', help='Decay Factor', metavar='FLOAT', type=click.FloatRange(min=0, min_open=True), default=2.0, show_default=True)
@click.option('-c', help='Constant c for Huber Loss', metavar='FLOAT', type=click.FloatRange(), default=0.0, show_default=True)
@click.option('-k', help='Mapping fn hyperparams', metavar='FLOAT', type=click.FloatRange(), default=8.0, show_default=True)
@click.option('-b', help='Mapping fn hyperparams', metavar='FLOAT', type=click.FloatRange(), default=1.0, show_default=True)
@click.option('--cut', help='Cutoff value.', metavar='FLOAT', type=click.FloatRange(), default=4.0, show_default=True)
@click.option('--boundary_prob', help='With this probability, convert t to transition_t for boundary condition', metavar='FLOAT', type=click.FloatRange(), default=0.05, show_default=True)
@click.option('--weighting', help='Weighting function', metavar='STR', default='default', show_default=True)
@click.option('--ratio_limit', help='Limit the ratio', metavar='FLOAT', type=click.FloatRange(min=0, max=1), default=0.999, show_default=True)
@click.option('--sqrt', help='Take sqrt to the squared L2 loss', metavar='BOOL', type=bool, default=True, show_default=True)
@click.option('--gclip', help='Gradient clipping', metavar='FLOAT', type=click.FloatRange(min=0), default=1000000., show_default=True)
# Performance-related.
@click.option('--fp16', help='Enable mixed-precision training', metavar='BOOL', type=bool, default=False, show_default=True)
@click.option('--ls', help='Loss scaling', metavar='FLOAT', type=click.FloatRange(min=0, min_open=True), default=1, show_default=True)
@click.option('--bench', help='Enable cuDNN benchmarking', metavar='BOOL', type=bool, default=True, show_default=True)
@click.option('--cache', help='Cache dataset in CPU memory', metavar='BOOL', type=bool, default=True, show_default=True)
@click.option('--workers', help='DataLoader worker processes', metavar='INT', type=click.IntRange(min=1), default=1, show_default=True)
# I/O-related.
@click.option('--desc', help='String to include in result dir name', metavar='STR', type=str)
@click.option('--nosubdir', help='Do not create a subdirectory for results', is_flag=True)
@click.option('--tick', help='How often to print progress', metavar='KIMG', type=click.FloatRange(min=1), default=50, show_default=True)
@click.option('--snap', help='How often to save snapshots', metavar='TICKS', type=click.IntRange(min=1), default=500, show_default=True)
@click.option('--dump', help='How often to dump state', metavar='TICKS', type=click.IntRange(min=1), default=500, show_default=True)
@click.option('--ckpt', help='How often to save latest checkpoints', metavar='TICKS', type=click.IntRange(min=1), default=50, show_default=True)
@click.option('--seed', help='Random seed [default: random]', metavar='INT', type=int)
@click.option('--transfer', help='Transfer learning from network pickle', metavar='PKL|URL', type=str)
@click.option('--resume', help='Resume from previous training state', metavar='PT', type=str)
@click.option('--resume_tick', help='Number of tick from previous training state', metavar='INT', type=int)
@click.option('-n', '--dry_run', help='Print training options and exit', is_flag=True)
@click.option('--use_wandb', help='Use wandb or not', type=bool, default=False, show_default=True)
@click.option('--distill', help='Path to teacher net for distillation', metavar='PKL|URL', type=str)
@click.option('--stage_max', help='Maximum stage in step scheduling', metavar='INT', type=click.IntRange(min=1), default=100, show_default=True)
# TCM
@click.option('--tcm_teacher_pkl', help='Path to teacher net (stage-1 model) for stage-2 training', metavar='PKL|URL', type=str)
@click.option('--tcm_transition_t', help='Transition time t prime', metavar='FLOAT', type=click.FloatRange(min=0), default=1., show_default=True)
# Evaluation
@click.option('--mid_t', help='Sampler steps [default: 0.821]', multiple=True, default=[0.821])
@click.option('--metrics', help='Comma-separated list or "none" [default: fid50k_full]', type=CommaSeparatedList(), default='fid50k_full')
@click.option('--sample_every', help='How often to sample imgs', metavar='TICKS', type=click.IntRange(min=1), default=10, show_default=True)
@click.option('--eval_every', help='How often to evaluate metrics', metavar='TICKS', type=click.IntRange(min=1), default=50, show_default=True)
@click.option('--dfid_ts', help='Comma-separated list of t_max values for FID calculation', type=CommaSeparatedList(), default='1.,2.,3.,4.')
@click.option('--dout_resolutions', help='Resolutions at which to apply dropout', metavar='LIST', type=CommaSeparatedList(), default='16,8,4,2,1')
def main(**kwargs):
"""Train ECMs using the techniques described in the
blog "Consistency Models Made Easy".
"""
opts = dnnlib.EasyDict(kwargs)
opts.tick = opts.batch * 0.1 # 1 tick = 100 iterations
# current_date = datetime.datetime.now().strftime("%m.%d.%Y")
# opts.outdir = os.path.join(opts.outdir, current_date)
print(f"--------------------------------ticks = {opts.tick}--------------------------------")
torch.multiprocessing.set_start_method('spawn')
dist.init()
# Initialize config dict.
c = dnnlib.EasyDict()
c.dataset_kwargs = dnnlib.EasyDict(class_name='training.dataset.ImageFolderDataset', path=opts.data, use_labels=opts.cond, xflip=opts.xflip, cache=opts.cache)
c.data_loader_kwargs = dnnlib.EasyDict(pin_memory=True, num_workers=opts.workers, prefetch_factor=2)
c.network_kwargs = dnnlib.EasyDict()
c.loss_kwargs = dnnlib.EasyDict(P_mean=opts.mean, P_std=opts.std, t_lower=opts.t_lower, t_upper=opts.t_upper, tdist=opts.tdist, df=opts.df,
q=opts.q, c=opts.c, k=opts.k, b=opts.b, cut=opts.cut, ratio_limit=opts.ratio_limit, sqrt=opts.sqrt, weighting=opts.weighting, boundary_prob=opts.boundary_prob)
c.optimizer_kwargs = dnnlib.EasyDict(class_name=f'torch.optim.{opts.optim}', lr=opts.lr, betas=[opts.beta1, opts.beta2], eps=opts.eps)
c.lr_kawrgs = dnnlib.EasyDict(decay_iter=opts.decay_iter, rampup_iter=opts.rampup_iter, start_iter=opts.lr_schedule_start_iter)
if opts.ema_type == 'constant':
assert opts.ema_beta is not None, 'ema_beta must be specified for constant EMA'
if opts.ema_type == 'power':
assert opts.ema_gamma is not None, 'ema_gamma must be specified for power EMA'
c.ema_kwargs = dnnlib.EasyDict(ema_type=opts.ema_type, ema_beta=opts.ema_beta, ema_gamma=opts.ema_gamma)
# Validate dataset options.
try:
dataset_obj = dnnlib.util.construct_class_by_name(**c.dataset_kwargs)
dataset_name = dataset_obj.name
c.dataset_kwargs.resolution = dataset_obj.resolution # be explicit about dataset resolution
c.dataset_kwargs.max_size = len(dataset_obj) # be explicit about dataset size
if opts.cond and not dataset_obj.has_labels:
raise click.ClickException('--cond=True requires labels specified in dataset.json')
del dataset_obj # conserve memory
except IOError as err:
raise click.ClickException(f'--data: {err}')
# Network architecture.
if opts.arch == 'ddpmpp':
c.network_kwargs.update(model_type='SongUNet', embedding_type='positional', encoder_type='standard', decoder_type='standard')
c.network_kwargs.update(channel_mult_noise=1, resample_filter=[1,1], model_channels=128, channel_mult=[2,2,2])
elif opts.arch == 'adm':
c.network_kwargs.update(model_type='DhariwalUNet', model_channels=192, channel_mult=[1,2,3,4])
elif 'edm2-img64' in opts.arch:
nc_dict = {'s': 192, 'm': 256, 'l': 320, 'xl': 384}
c.network_kwargs.update(model_type='EDM2UNet', model_channels=nc_dict[opts.arch.split('-')[-1]])
c.network_kwargs.update(dout_resolutions=[int(res) for res in opts.dout_resolutions])
else:
raise ValueError(f"Unrecognized architecture: {opts.arch}")
# Preconditioning & loss function.
if opts.precond == 'ect':
c.network_kwargs.class_name = 'training.networks.ECMPrecond'
c.loss_kwargs.class_name = 'training.loss.ECMLoss'
else:
raise ValueError('Unrecognized Precond & Loss!')
# Network options.
if opts.cbase is not None:
c.network_kwargs.model_channels = opts.cbase
if opts.cres is not None:
c.network_kwargs.channel_mult = opts.cres
if opts.augment:
c.augment_kwargs = dnnlib.EasyDict(class_name='training.augment.AugmentPipe', p=opts.augment)
c.augment_kwargs.update(xflip=1e8, yflip=1, scale=1, rotate_frac=1, aniso=1, translate_frac=1)
c.network_kwargs.augment_dim = 9
c.network_kwargs.update(dropout=opts.dropout, use_fp16=opts.fp16)
# Progressive training options.
if opts.tcm_teacher_pkl is not None:
if opts.tcm_transition_t > opts.t_lower:
raise ValueError(f"tcm_transition_t must be less than t_lower, but got {opts.tcm_transition_t} > {opts.t_lower}")
c.tcm_kwargs = dnnlib.EasyDict(teacher_pkl=opts.tcm_teacher_pkl, transition_t=opts.tcm_transition_t)
# Training options.
c.total_kimg = max(int(opts.duration * 1000), 1)
c.update(batch_size=opts.batch, batch_gpu=opts.batch_gpu)
c.update(loss_scaling=opts.ls, cudnn_benchmark=opts.bench)
c.update(kimg_per_tick=opts.tick, snapshot_ticks=opts.snap, state_dump_ticks=opts.dump, ckpt_ticks=opts.ckpt, double_ticks=opts.double)
c.update(mid_t=opts.mid_t, metrics=opts.metrics, sample_ticks=opts.sample_every, eval_ticks=opts.eval_every)
c.update(gclip=opts.gclip, start_stage=opts.start_stage, w_boundary=opts.w_boundary)
# Random seed.
if opts.seed is not None:
c.seed = opts.seed
else:
c.seed = 101
# Description string.
# cond_str = 'cond' if c.dataset_kwargs.use_labels else 'uncond'
# dtype_str = 'fp16' if c.network_kwargs.use_fp16 else 'fp32'
desc = opts.desc
# desc = (f'{dataset_name:s}-{cond_str:s}-{opts.arch:s}-{opts.precond:s}-{opts.optim:s}-{opts.lr:f}'
# f'-mstage{opts.stage_max}-gpus{dist.get_world_size():d}-batch{c.batch_size:d}-{dtype_str:s}')
# if opts.desc is not None:
# desc += f'-{opts.desc}'
c.desc = desc
# Pick output directory.
if opts.nosubdir:
c.run_dir = opts.outdir
ckpt_auto_resume = None
else:
c.run_dir, ckpt_auto_resume = create_expt_folder_with_auto_resuming(opts.outdir, desc)
if ckpt_auto_resume is None:
# Transfer learning and resume from given checkpoints or training states.
if opts.transfer is not None:
if opts.resume is not None:
raise click.ClickException('--transfer and --resume cannot be specified at the same time')
c.resume_pkl = opts.transfer
c.resume_tick = 0 if opts.resume_tick is None else opts.resume_tick
elif opts.resume is not None:
match = re.fullmatch(r'training-state-(\d+|latest).pt', os.path.basename(opts.resume))
if not match or not os.path.isfile(opts.resume):
raise click.ClickException(f'--resume must point to training-state-*.pt from a previous training run, but got {opts.resume}')
c.resume_pkl = os.path.join(os.path.dirname(opts.resume), f'network-snapshot-{match.group(1)}.pkl')
c.resume_tick = 0 if opts.resume_tick is None else opts.resume_tick
c.resume_state_dump = opts.resume
else:
# Auto resuming form the last training state
match = re.fullmatch(r'training-state-(\d+|latest).pt', os.path.basename(ckpt_auto_resume))
if not match or not os.path.isfile(ckpt_auto_resume):
raise click.ClickException('Auto resuming must point to training-state-*.pt from a previous training run')
c.resume_pkl = os.path.join(os.path.dirname(ckpt_auto_resume), f'network-snapshot-{match.group(1)}.pkl')
c.resume_state_dump = ckpt_auto_resume
# teacher network path for distillation
if opts.distill is not None:
c.teacher_net_pkl = opts.distill
c.stage_max = opts.stage_max
c.use_wandb = opts.use_wandb
c.dfid_ts = [round(float(t), 3) for t in opts.dfid_ts]
# Print options.
dist.print0()
dist.print0('Training options:')
dist.print0(json.dumps(c, indent=2))
dist.print0()
dist.print0(f'Output directory: {c.run_dir}')
dist.print0(f'Dataset path: {c.dataset_kwargs.path}')
dist.print0(f'Class-conditional: {c.dataset_kwargs.use_labels}')
dist.print0(f'Network architecture: {opts.arch}')
dist.print0(f'Preconditioning & loss: {opts.precond}')
dist.print0(f'Enable distillation: {opts.distill is not None}')
dist.print0(f'Number of GPUs: {dist.get_world_size()}')
dist.print0(f'Batch size: {c.batch_size}')
dist.print0(f'Mixed-precision: {c.network_kwargs.use_fp16}')
dist.print0()
# Dry run?
if opts.dry_run:
dist.print0('Dry run; exiting.')
return
# Create output directory.
dist.print0('Creating output directory...')
if dist.get_rank() == 0:
os.makedirs(c.run_dir, exist_ok=True)
with open(os.path.join(c.run_dir, 'training_options.json'), 'wt') as f:
json.dump(c, f, indent=2)
dnnlib.util.Logger(file_name=os.path.join(c.run_dir, 'log.txt'), file_mode='a', should_flush=True)
# Train.
training_loop.training_loop(**c)
#----------------------------------------------------------------------------
if __name__ == "__main__":
main()
#----------------------------------------------------------------------------