DL-Art-School/codes/trainer/base_model.py

196 lines
8.0 KiB
Python
Raw Normal View History

2019-08-23 13:42:47 +00:00
import os
from collections import OrderedDict
import torch
import torch.nn as nn
2022-01-25 00:59:31 +00:00
from torch.distributed.optim import ZeroRedundancyOptimizer
from torch.nn.parallel.distributed import DistributedDataParallel
2020-06-01 21:09:52 +00:00
import utils.util
from utils.util import opt_get, optimizer_to, map_to_device
2019-08-23 13:42:47 +00:00
class BaseModel():
def __init__(self, opt):
self.opt = opt
if opt['dist']:
self.rank = torch.distributed.get_rank()
else:
self.rank = -1 # non dist training
2022-03-04 19:46:41 +00:00
self.device = torch.cuda.current_device() if opt['gpu_ids'] else torch.device('cpu')
self.amp_level = 'O0' if opt['amp_opt_level'] is None else opt['amp_opt_level']
2019-08-23 13:42:47 +00:00
self.is_train = opt['is_train']
self.opt_in_cpu = opt_get(opt, ['keep_optimizer_states_on_cpu'], False)
2019-08-23 13:42:47 +00:00
self.schedulers = []
self.optimizers = []
self.disc_optimizers = []
self.save_history = {}
2019-08-23 13:42:47 +00:00
def feed_data(self, data):
pass
def optimize_parameters(self):
pass
def get_current_visuals(self):
pass
def get_current_losses(self):
pass
def print_network(self):
pass
def save(self, label):
pass
def load(self):
pass
def _set_lr(self, lr_groups_l):
"""Set learning rate for warmup
lr_groups_l: list for lr_groups. each for a optimizer"""
for optimizer, lr_groups in zip(self.optimizers, lr_groups_l):
for param_group, lr in zip(optimizer.param_groups, lr_groups):
param_group['lr'] = lr
def _get_init_lr(self):
"""Get the initial lr, which is set by the scheduler"""
init_lr_groups_l = []
for optimizer in self.optimizers:
init_lr_groups_l.append([v['initial_lr'] for v in optimizer.param_groups])
return init_lr_groups_l
def update_learning_rate(self, cur_iter, warmup_iter=-1):
for scheduler in self.schedulers:
scheduler.last_epoch = cur_iter
2019-08-23 13:42:47 +00:00
scheduler.step()
# set up warm-up learning rate
if cur_iter < warmup_iter:
# get initial lr for each group
init_lr_g_l = self._get_init_lr()
# modify warming-up learning rates
warm_up_lr_l = []
for init_lr_g in init_lr_g_l:
warm_up_lr_l.append([v / warmup_iter * cur_iter for v in init_lr_g])
# set learning rate
self._set_lr(warm_up_lr_l)
def get_current_learning_rate(self):
return [param_group['lr'] for param_group in self.optimizers[0].param_groups]
def get_network_description(self, network):
"""Get the string and total parameters of the network"""
if isinstance(network, nn.DataParallel) or isinstance(network, DistributedDataParallel):
network = network.module
return str(network), sum(map(lambda x: x.numel(), network.parameters()))
def save_network(self, network, network_label, iter_label):
save_filename = '{}_{}.pth'.format(iter_label, network_label)
save_path = os.path.join(self.opt['path']['models'], save_filename)
if isinstance(network, nn.DataParallel) or isinstance(network, DistributedDataParallel):
network = network.module
state_dict = network.state_dict()
for key, param in state_dict.items():
state_dict[key] = param.cpu()
torch.save(state_dict, save_path)
if network_label not in self.save_history.keys():
self.save_history[network_label] = []
self.save_history[network_label].append(save_path)
# Also save to the 'alt_path' which is useful for caching to Google Drive in colab, for example.
if 'alt_path' in self.opt['path'].keys():
torch.save(state_dict, os.path.join(self.opt['path']['alt_path'], save_filename))
2020-06-01 21:09:52 +00:00
if self.opt['colab_mode']:
utils.util.copy_files_to_server(self.opt['ssh_server'], self.opt['ssh_username'], self.opt['ssh_password'],
save_path, os.path.join(self.opt['remote_path'], 'models', save_filename))
return save_path
2019-08-23 13:42:47 +00:00
def load_network(self, load_path, network, strict=True, pretrain_base_path=None):
2021-06-13 16:25:23 +00:00
# Sometimes networks are passed in as DDP modules, we want the raw parameters.
if hasattr(network, 'module'):
network = network.module
2022-01-24 22:08:13 +00:00
load_net = torch.load(load_path, map_location=utils.util.map_cuda_to_correct_device)
# Support loading torch.save()s for whole models as well as just state_dicts.
if 'state_dict' in load_net:
load_net = load_net['state_dict']
2019-08-23 13:42:47 +00:00
load_net_clean = OrderedDict() # remove unnecessary 'module.'
if pretrain_base_path is not None:
t = load_net
load_net = {}
for k, v in t.items():
if k.startswith(pretrain_base_path):
load_net[k[len(pretrain_base_path):]] = v
2019-08-23 13:42:47 +00:00
for k, v in load_net.items():
2020-10-26 17:13:01 +00:00
if k.startswith('module.'):
load_net_clean[k.replace('module.', '')] = v
2019-08-23 13:42:47 +00:00
else:
load_net_clean[k] = v
network.load_state_dict(load_net_clean, strict=strict)
2022-01-25 00:59:31 +00:00
2022-01-25 01:12:08 +00:00
def consolidate_state(self):
2022-01-25 00:59:31 +00:00
for o in self.optimizers:
if isinstance(o, ZeroRedundancyOptimizer):
2022-01-25 01:12:08 +00:00
o.consolidate_state_dict(to=0)
2022-01-25 00:59:31 +00:00
def save_training_state(self, state):
2019-08-23 13:42:47 +00:00
"""Save training state during training, which will be used for resuming"""
state.update({'schedulers': [], 'optimizers': []})
2019-08-23 13:42:47 +00:00
for s in self.schedulers:
state['schedulers'].append(s.state_dict())
for o in self.optimizers:
state['optimizers'].append(o.state_dict())
if 'amp_opt_level' in self.opt.keys():
state['amp'] = amp.state_dict()
save_filename = '{}.state'.format(utils.util.opt_get(state, ['iter'], 'no_step_provided'))
2019-08-23 13:42:47 +00:00
save_path = os.path.join(self.opt['path']['training_state'], save_filename)
torch.save(map_to_device(state, 'cpu'), save_path)
if '__state__' not in self.save_history.keys():
self.save_history['__state__'] = []
self.save_history['__state__'].append(save_path)
# Also save to the 'alt_path' which is useful for caching to Google Drive in colab, for example.
if 'alt_path' in self.opt['path'].keys():
torch.save(state, os.path.join(self.opt['path']['alt_path'], 'latest.state'))
2020-06-01 21:09:52 +00:00
if self.opt['colab_mode']:
utils.util.copy_files_to_server(self.opt['ssh_server'], self.opt['ssh_username'], self.opt['ssh_password'],
save_path, os.path.join(self.opt['remote_path'], 'training_state', save_filename))
2019-08-23 13:42:47 +00:00
def stash_optimizers(self):
"""
When enabled, puts all optimizer states in CPU memory, allowing forward and backward passes more memory
headroom.
"""
if not self.opt_in_cpu:
return
for opt in self.optimizers:
optimizer_to(opt, 'cpu')
def restore_optimizers(self):
"""
Puts optimizer states back into device memory.
"""
if not self.opt_in_cpu:
return
for opt in self.optimizers:
optimizer_to(opt, self.device)
def resume_training(self, resume_state, load_amp=True):
2019-08-23 13:42:47 +00:00
"""Resume the optimizers and schedulers for training"""
resume_optimizers = resume_state['optimizers']
resume_schedulers = resume_state['schedulers']
assert len(resume_optimizers) == len(self.optimizers), 'Wrong lengths of optimizers'
assert len(resume_schedulers) == len(self.schedulers), 'Wrong lengths of schedulers'
for i, o in enumerate(resume_optimizers):
self.optimizers[i].load_state_dict(o)
for i, s in enumerate(resume_schedulers):
self.schedulers[i].load_state_dict(s)
if load_amp and 'amp' in resume_state.keys():
from apex import amp
amp.load_state_dict(resume_state['amp'])
self.stash_optimizers()