-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathmmaml.py
executable file
·191 lines (178 loc) · 8.05 KB
/
mmaml.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
import torch.cuda
from layers import *
from config import *
from taskset_wrapper import *
class MMAML(nn.Module):
def __init__(self, c, h, w, ways, shots, queries,
proto_hidden=64, conv_hidden=64,
n_layers=4, max_pool=False,
resnet=False,
save_dir='../artifact'):
super(MMAML, self).__init__()
self.c, self.h, self.w = c, h, w
self.ways, self.shots, self.queries = ways, shots, queries
self.proto_hidden = proto_hidden
self.conv_hidden = conv_hidden
self.n_layers = n_layers
# init learning networks
self.proto_net = ProtoNetEmbedding(
c, h, w,
hidden=proto_hidden,
max_pool=max_pool,
n_layers = n_layers
)
if resnet:
self.predictor_net = ResNetMMAML(ways)
self.modulator_net = nn.Sequential(
FFW([ways * proto_hidden, 36864]),
nn.Tanh()
)
else:
self.predictor_net = ConvBaseMMAML(
c, h, w, ways,
hidden=conv_hidden,
n_layers= n_layers,
max_pool=max_pool
)
self.modulator_net = nn.Sequential(
FFW([ways * proto_hidden, 9 * conv_hidden ** 2]),
nn.Tanh()
)
self.results = None
self.save_dir = save_dir
self.loss = nn.CrossEntropyLoss()
def compute_embedding(self, data, labels):
embedding = []
for i in range(self.ways):
xi = data[torch.where(labels==i)]
if xi.shape[0] == 0:
embedding.append(torch.zeros(self.task_embedding_dim).to(data.device))
else:
embedding.append(torch.mean(self.proto_net(xi), dim=0, keepdim=False))
return torch.cat(embedding)
def maml_fast_adapt(self, batch, learner, adaptation_steps=1, predict=True):
adapt_data, adapt_labels, eval_data, eval_labels = batch
task_embedding = self.compute_embedding(adapt_data, adapt_labels)
# modulation = [m(task_embedding) for m in self.modulator_net]
modulation = [
self.modulator_net(task_embedding), None, None, None
]
# Adapt the model
for step in range(adaptation_steps):
train_error = self.loss(learner(adapt_data, modulation), adapt_labels)
learner.adapt(train_error)
logits = learner(eval_data, modulation)
eval_loss = self.loss(logits, eval_labels)
if not predict:
return eval_loss, logits
else:
prediction = logits.argmax(dim=1).view(eval_labels.shape)
eval_acc = (prediction == eval_labels).sum().float() / eval_labels.shape[0]
return eval_loss, eval_acc
def protonet_loss(self, batch, logit=False):
adapt_data, adapt_labels, eval_data, eval_labels = batch
support = self.compute_embedding(adapt_data, adapt_labels).reshape(self.ways, -1)
query = self.proto_net(eval_data)
logits = torch.cdist(query, support)
if logit:
return logits
else:
return self.loss(logits, eval_labels)
def meta_train(self, task, tasksets, max_epoch=800, meta_batch_size=32, test_batch_size=5):
maml = l2l.algorithms.MAML(
self.predictor_net, lr=5e-1,
first_order=False,
allow_unused=True,
allow_nograd=True
)
maml_opt = optim.Adam(maml.parameters(), lr=3e-3)
routing_opt = optim.Adam([
{'params': self.proto_net.parameters(), 'lr': 3e-3},
{'params': self.modulator_net.parameters(), 'lr': 3e-3},
])
results = {
'mean_loss': [],
'std_loss': [],
'mean_acc': [],
'std_acc': [],
'time': []
}
bar = trange(max_epoch)
for epoch in bar:
start_time = time()
bar.set_description_str(f'Train Epoch {epoch}')
maml_opt.zero_grad()
for i in range(meta_batch_size):
# Compute meta-training loss
routing_opt.zero_grad()
learner = maml.clone()
train_batch = tasksets.split_batch(tasksets.sample_task('train'))
eval_loss, eval_acc = self.maml_fast_adapt(train_batch, learner)
eval_loss.backward()
routing_opt.step()
bar.set_postfix_str(f'Eval acc {i+1}/{meta_batch_size}={eval_acc}')
for p in maml.parameters():
if p.grad is not None:
p.grad.data.mul_(1.0 / meta_batch_size)
maml_opt.step()
results['time'].append(time() - start_time)
bar.set_description_str(f'Test Epoch {epoch}')
test_loss, test_acc = [], []
for _ in range(test_batch_size):
# Compute meta-testing loss
learner = maml.clone()
test_batch = tasksets.split_batch(tasksets.sample_task('test'))
eval_loss, logits = self.maml_fast_adapt(test_batch, learner, predict=False)
logits += self.protonet_loss(test_batch, logit=True)
adapt_data, adapt_labels, eval_data, eval_labels = test_batch
prediction = logits.argmax(dim=1).view(eval_labels.shape)
eval_acc = (prediction == eval_labels).sum().float() / eval_labels.shape[0]
test_loss.append(eval_loss.item())
test_acc.append(eval_acc.item())
results['mean_loss'].append(torch.tensor(test_loss).mean().item())
results['std_loss'].append(torch.tensor(test_loss).std().item())
results['mean_acc'].append(torch.tensor(test_acc).mean().item())
results['std_acc'].append(torch.tensor(test_acc).std().item())
torch.save(self.state_dict(), os.path.join(self.save_dir, f'mmaml-{task}-model.pt'))
torch.save(results, os.path.join(self.save_dir, f'mmaml-{task}-results.pt'))
bar.set_postfix_str(f'Test loss={results["mean_loss"][-1]:.3f} '
f'acc={results["mean_acc"][-1]:.3f}')
print('')
return results
def run_mmaml(task, max_epoch=10000, meta_batch_size=32, test_batch_size=5):
seed()
config['resnet'] = False
model = MMAML(**config[task])
model = model.to(device)
tasksets = TASKS[task](model.ways, model.shots, model.queries)
results = model.meta_train(task, tasksets, max_epoch, meta_batch_size, test_batch_size)
return model, results
def run_mmaml_resnet(task, max_epoch=2000, meta_batch_size=32, test_batch_size=5):
seed()
config['True'] = False
model = MMAML(**config[task])
model = model.to(device)
tasksets = TASKS[task](model.ways, model.shots, model.queries)
results = model.meta_train(task, tasksets, max_epoch, meta_batch_size, test_batch_size)
return model, results
def run_mmaml_measure_time(task, max_epoch=20, meta_batch_size=32, test_batch_size=5):
seed()
for c in [16, 32, 64, 128, 256]:
cfg = deepcopy(config[task])
cfg['conv_hidden'] = cfg['proto_hidden'] = c
cfg['save_dir'] = cfg['save_dir'] + f'-{c}c'
model = MMAML(**cfg)
model = model.to(device)
tasksets = TASKS[task](model.ways, model.shots, model.queries)
results = model.meta_train(task, tasksets, max_epoch, meta_batch_size, test_batch_size)
return model, results
if __name__ == '__main__':
torch.cuda.set_device(1)
if len(sys.argv) < 2:
# model, results = run_mmaml_measure_time('omniglot')
# model, results = run_mmaml('jigsaw-44-mini-imagenet')
# model, results = run_mmaml('faf-same-start')
# model, results = run_mmaml_resnet('jigsaw-mini-imagenet-hard')
model, results = run_mmaml_resnet('mini-imagenet-hard')
else:
model, results = run_mmaml(sys.argv[1])