-
Notifications
You must be signed in to change notification settings - Fork 2
/
train.py
163 lines (135 loc) · 8.3 KB
/
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
import copy
import math
import os
from functools import partial
import wandb
import torch
torch.multiprocessing.set_sharing_strategy('file_system')
#torch.autograd.set_detect_anomaly(True)
import resource
rlimit = resource.getrlimit(resource.RLIMIT_NOFILE)
resource.setrlimit(resource.RLIMIT_NOFILE, (64000, rlimit[1]))
from utils.diffusion_utils import t_to_sigma as t_to_sigma_compl
from datasets_utils.pdbbind import construct_loader
from utils.parsing import parse_train_args
from utils.training import train_epoch, test_epoch, loss_function, inference_epoch
from utils.utils import save_yaml_file, get_optimizer_and_scheduler, ExponentialMovingAverage
from utils.model_utils import get_model
def train(args, model, optimizer, scheduler, ema_weights, train_loader, val_loader, t_to_sigma, run_dir):
best_val_loss = math.inf
best_val_inference_value = math.inf if args.inference_earlystop_goal == 'min' else 0
best_epoch = 0
best_val_inference_epoch = 0
loss_fn = partial(loss_function, tr_weight=args.tr_weight, rot_weight=args.rot_weight,
tor_weight=args.tor_weight, no_torsion=args.no_torsion)
score_model = model.module.score_model if device.type == 'cuda' else model.score_model
print("Starting training...")
for epoch in range(args.n_epochs):
if epoch % 5 == 0: print("Run name: ", args.run_name)
logs = {}
train_losses = train_epoch(model, train_loader, optimizer, device, t_to_sigma, loss_fn, ema_weights,
use_latent=args.latent_dim>0, training_latent_temperature=args.training_latent_temperature)
print("Epoch {}: Training loss {:.4f} tr {:.4f} rot {:.4f} tor {:.4f}"
.format(epoch, train_losses['loss'], train_losses['tr_loss'], train_losses['rot_loss'],
train_losses['tor_loss']))
ema_weights.store(score_model.parameters())
if args.use_ema: ema_weights.copy_to(score_model.parameters()) # load ema parameters into model for running validation and inference
val_losses = test_epoch(model, val_loader, device, t_to_sigma, loss_fn, args.test_sigma_intervals, use_latent=args.latent_dim>0,
training_latent_temperature=args.training_latent_temperature)
print("Epoch {}: Validation loss {:.4f} tr {:.4f} rot {:.4f} tor {:.4f}"
.format(epoch, val_losses['loss'], val_losses['tr_loss'], val_losses['rot_loss'], val_losses['tor_loss']))
if args.val_inference_freq != None and (epoch + 1) % args.val_inference_freq == 0:
dataset = [val_loader.dataset.get(i) for i in range(min(args.num_inference_complexes, len(val_loader.dataset)))]
inf_metrics = inference_epoch(model, dataset, device, t_to_sigma, args, use_latent=args.latent_dim>0)
print("Epoch {}: Val inference oracle rmsds_lt2 {:.3f} rmsds_lt5 {:.3f}"
.format(epoch, inf_metrics['rmsds_lt2'], inf_metrics['rmsds_lt5']))
logs.update({'valinf_' + k: v for k, v in inf_metrics.items()}, step=epoch + 1)
if not args.use_ema: ema_weights.copy_to(score_model.parameters())
ema_state_dict = copy.deepcopy(score_model.state_dict())
ema_weights.restore(score_model.parameters())
if args.wandb:
logs.update({'train_' + k: v for k, v in train_losses.items()})
logs.update({'val_' + k: v for k, v in val_losses.items()})
logs['current_lr'] = optimizer.param_groups[0]['lr']
wandb.log(logs, step=epoch + 1)
state_dict = model.module.state_dict() if device.type == 'cuda' else model.state_dict()
if args.inference_earlystop_metric in logs.keys() and \
(args.inference_earlystop_goal == 'min' and logs[args.inference_earlystop_metric] <= best_val_inference_value or
args.inference_earlystop_goal == 'max' and logs[args.inference_earlystop_metric] >= best_val_inference_value):
best_val_inference_value = logs[args.inference_earlystop_metric]
best_val_inference_epoch = epoch
torch.save(state_dict, os.path.join(run_dir, 'best_inference_epoch_model.pt'))
torch.save(ema_state_dict, os.path.join(run_dir, 'best_ema_inference_epoch_model.pt'))
if val_losses['loss'] <= best_val_loss:
best_val_loss = val_losses['loss']
best_epoch = epoch
torch.save(state_dict, os.path.join(run_dir, 'best_model.pt'))
torch.save(ema_state_dict, os.path.join(run_dir, 'best_ema_model.pt'))
if scheduler:
if args.val_inference_freq is not None:
scheduler.step(best_val_inference_value)
else:
scheduler.step(val_losses['loss'])
torch.save({
'epoch': epoch,
'model': state_dict,
'optimizer': optimizer.state_dict(),
'ema_weights': ema_weights.state_dict(),
}, os.path.join(run_dir, 'last_model.pt'))
print("Best Validation Loss {} on Epoch {}".format(best_val_loss, best_epoch))
print("Best inference metric {} on Epoch {}".format(best_val_inference_value, best_val_inference_epoch))
def main_function():
args = parse_train_args()
assert (args.inference_earlystop_goal == 'max' or args.inference_earlystop_goal == 'min')
if args.val_inference_freq is not None and args.scheduler is not None:
assert (args.scheduler_patience > args.val_inference_freq) # otherwise we will just stop training after args.scheduler_patience epochs
if args.cudnn_benchmark:
torch.backends.cudnn.benchmark = True
# construct loader
t_to_sigma = partial(t_to_sigma_compl, args=args)
train_loader, val_loader = construct_loader(args, t_to_sigma)
model = get_model(args, device, t_to_sigma=t_to_sigma)
optimizer, scheduler = get_optimizer_and_scheduler(args, model, scheduler_mode=args.inference_earlystop_goal if args.val_inference_freq is not None else 'min')
ema_weights = ExponentialMovingAverage(model.module.score_model.parameters() if device.type == 'cuda' else model.score_model.parameters(), decay=args.ema_rate)
if args.restart_dir:
try:
dict = torch.load(f'{args.restart_dir}/last_model.pt', map_location=torch.device('cpu'))
dict['model'] = {k.replace('module.', ''):v for k,v in dict['model'].items()}
if device.type == 'cuda':
load_return = model.module.load_state_dict(dict['model'], strict=False)
else:
load_return = model.load_state_dict(dict['model'], strict=False)
print("Loaded model with", load_return)
if len(load_return.missing_keys) == 0 and len(load_return.unexpected_keys) == 0:
if args.restart_lr is not None: dict['optimizer']['param_groups'][0]['lr'] = args.restart_lr
optimizer.load_state_dict(dict['optimizer'])
if hasattr(args, 'ema_rate'):
ema_weights.load_state_dict(dict['ema_weights'], device=device)
print("Restarting from epoch", dict['epoch'])
else:
assert args.non_strict_loading, f"Issue with loading of the model: missing keys {load_return.missing_keys} unexpected_keys {load_return.unexpected_keys}"
except Exception as e:
raise e
print("Exception", e)
dict = torch.load(f'{args.restart_dir}/best_model.pt', map_location=torch.device('cpu'))
model.load_state_dict(dict, strict=True)
print("Due to exception had to take the best epoch and no optimiser")
numel = sum([p.numel() for p in model.parameters()])
print('Model with', numel, 'parameters in total')
if args.wandb:
wandb.init(
settings=wandb.Settings(start_method="fork"),
project=args.project,
name=args.run_name,
config=args
)
wandb.log({'numel': numel})
# record parameters
run_dir = os.path.join(args.log_dir, args.run_name)
yaml_file_name = os.path.join(run_dir, 'model_parameters.yml')
save_yaml_file(yaml_file_name, args.__dict__)
args.device = device
train(args, model, optimizer, scheduler, ema_weights, train_loader, val_loader, t_to_sigma, run_dir)
if __name__ == '__main__':
device = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')
main_function()