-
Notifications
You must be signed in to change notification settings - Fork 6
/
engine_nus_first_stage.py
175 lines (129 loc) · 7.48 KB
/
engine_nus_first_stage.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
import os
import torch
import torch.nn.functional as F
from tqdm import tqdm
import utils.lr_sched as lrs
from models.rank_loss import ranking_lossT
from utils.misc import compute_F1, compute_AP
def train(model, clip_model, args, optimizer, dataloader, logger, label_emb, epoch):
logger.info("TRAINING MODE")
mean_dist_loss = 0
mean_rank_loss = 0
for i, (train_inputs, train_labels) in enumerate(tqdm(dataloader)):
lrs.adjust_learning_rate(optimizer, i / len(dataloader) + epoch, args)
# import pdb; pdb.set_trace() print(torch.nonzero(1+train_labels[2]))
optimizer.zero_grad()
### remove empty label images while training ###
temp_label = torch.clamp(train_labels,0,1)
temp_seen_labels = temp_label.sum(1)
temp_label = temp_label[temp_seen_labels>0]
train_labels = train_labels[temp_seen_labels>0]
train_inputs = train_inputs[temp_seen_labels>0]
train_inputs = train_inputs.cuda()
train_labels = train_labels.cuda()
label_embed = label_emb[:925]
logits, _, dist_feat = model(train_inputs, label_embed)
rank_loss = ranking_lossT(logits, train_labels.float())
with torch.no_grad():
_, tea_dist_feat = clip_model.encode_image(train_inputs)
dist_loss = F.l1_loss(dist_feat, tea_dist_feat.float())
loss = dist_loss + rank_loss
mean_dist_loss += dist_loss.item()
mean_rank_loss += rank_loss.item()
loss.requires_grad_()
loss.backward()
optimizer.step()
mean_dist_loss /= len(dataloader)
mean_rank_loss /= len(dataloader)
learning_rate = optimizer.param_groups[-1]['lr']
logger.info("------------------------------------------------------------------")
logger.info("FINETUNING Epoch: {}/{} \tRankLoss: {:.6f}\tDistLoss: {:.6f}\tLearningRate {}".format(epoch, args.epochs, mean_rank_loss, mean_dist_loss, learning_rate))
logger.info("------------------------------------------------------------------")
torch.save(model.state_dict(), os.path.join(args.record_path, "model_epoch_{}.pth".format(epoch)))
########### TEST FUNC ###########
def test(model, args, dataloader, logger, label_emb, len_testdataset, writer, epoch=-1):
logger.info("=======================EVALUATION MODE=======================")
prediction_81 = torch.empty(len_testdataset,81)
prediction_1006 = torch.empty(len_testdataset,1006)
lab_81 = torch.empty(len_testdataset,81)
lab_1006 = torch.empty(len_testdataset,1006)
test_batch_size = args.test_batch_size
cnt = 0
for features, labels_1006, labels_81, _ in tqdm(dataloader):
strt = cnt
endt = min(cnt + test_batch_size, len_testdataset)
cnt += test_batch_size
with torch.no_grad():
pred_feat, dist_feat = model.encode_img(features.cuda())
score1 = torch.topk(pred_feat @ label_emb[925:].t(),k=model.topk, dim=1)[0].mean(dim=1)
score2 = dist_feat @ label_emb[925:].t()
score1 = score1 / score1.norm(dim=-1, keepdim=True)
score2 = score2 / score2.norm(dim=-1, keepdim=True)
logits_81 = (score1 + score2) / 2
score1 = torch.topk(pred_feat @ label_emb.t(),k=model.topk, dim=1)[0].mean(dim=1)
score2 = dist_feat @ label_emb.t()
score1 = score1 / score1.norm(dim=-1, keepdim=True)
score2 = score2 / score2.norm(dim=-1, keepdim=True)
logits_1006 = (score1 + score2) / 2
prediction_81[strt:endt,:] = logits_81
prediction_1006[strt:endt,:] = logits_1006
lab_81[strt:endt,:] = labels_81
lab_1006[strt:endt,:] = labels_1006
logger.info("completed calculating predictions over all images")
logits_81_5 = prediction_81.clone()
ap_81 = compute_AP(prediction_81.cuda(), lab_81.cuda())
F1_3_81,P_3_81,R_3_81 = compute_F1(prediction_81.cuda(), lab_81.cuda(), 'overall', k_val=3)
F1_5_81,P_5_81,R_5_81 = compute_F1(logits_81_5.cuda(), lab_81.cuda(), 'overall', k_val=5)
logger.info('ZSL AP: %.4f',torch.mean(ap_81))
logger.info('k=3: %.4f,%.4f,%.4f',torch.mean(F1_3_81),torch.mean(P_3_81),torch.mean(R_3_81))
logger.info('k=5: %.4f,%.4f,%.4f',torch.mean(F1_5_81),torch.mean(P_5_81),torch.mean(R_5_81))
logits_1006_5 = prediction_1006.clone()
ap_1006 = compute_AP(prediction_1006.cuda(), lab_1006.cuda())
F1_3_1006,P_3_1006,R_3_1006 = compute_F1(prediction_1006.cuda(), lab_1006.cuda(), 'overall', k_val=3)
F1_5_1006,P_5_1006,R_5_1006 = compute_F1(logits_1006_5.cuda(), lab_1006.cuda(), 'overall', k_val=5)
logger.info('GZSL AP:%.4f',torch.mean(ap_1006))
logger.info('g_k=3:%.4f,%.4f,%.4f',torch.mean(F1_3_1006), torch.mean(P_3_1006), torch.mean(R_3_1006))
logger.info('g_k=5:%.4f,%.4f,%.4f',torch.mean(F1_5_1006), torch.mean(P_5_1006), torch.mean(R_5_1006))
def eval(model, args, dataloader, label_emb, len_testdataset):
txt_feat = label_emb
prediction_81 = torch.empty(len_testdataset,81)
prediction_1006 = torch.empty(len_testdataset,1006)
lab_81 = torch.empty(len_testdataset,81)
lab_1006 = torch.empty(len_testdataset,1006)
test_batch_size = args.test_batch_size
cnt = 0
for features, labels_1006, labels_81, _ in tqdm(dataloader):
strt = cnt
endt = min(cnt + test_batch_size, len_testdataset)
cnt += test_batch_size
with torch.no_grad():
pred_feat, dist_feat = model.encode_img(features.cuda())
# import pdb; pdb.set_trace() # logger.info(pred_feat[0][0][:10]) logger.info(txt_feat[0][:10])
score1 = torch.topk(pred_feat @ txt_feat[925:].t(),k=model.topk, dim=1)[0].mean(dim=1)
score2 = dist_feat @ txt_feat[925:].t()
score1 = score1 / score1.norm(dim=-1, keepdim=True)
score2 = score2 / score2.norm(dim=-1, keepdim=True)
logits_81 = (score1 + score2) / 2
score1 = torch.topk(pred_feat @ txt_feat.t(),k=model.topk, dim=1)[0].mean(dim=1)
score2 = dist_feat @ txt_feat.t()
score1 = score1 / score1.norm(dim=-1, keepdim=True)
score2 = score2 / score2.norm(dim=-1, keepdim=True)
logits_1006 = (score1 + score2) / 2
prediction_81[strt:endt,:] = logits_81
prediction_1006[strt:endt,:] = logits_1006
lab_81[strt:endt,:] = labels_81
lab_1006[strt:endt,:] = labels_1006
logits_81_5 = prediction_81.clone()
ap_81 = compute_AP(prediction_81.cuda(), lab_81.cuda())
F1_3_81,P_3_81,R_3_81 = compute_F1(prediction_81.cuda(), lab_81.cuda(), 'overall', k_val=3)
F1_5_81,P_5_81,R_5_81 = compute_F1(logits_81_5.cuda(), lab_81.cuda(), 'overall', k_val=5)
print('ZSL AP: %.4f',torch.mean(ap_81))
print('k=3: %.4f,%.4f,%.4f',torch.mean(F1_3_81),torch.mean(P_3_81),torch.mean(R_3_81))
print('k=5: %.4f,%.4f,%.4f',torch.mean(F1_5_81),torch.mean(P_5_81),torch.mean(R_5_81))
logits_1006_5 = prediction_1006.clone()
ap_1006 = compute_AP(prediction_1006.cuda(), lab_1006.cuda())
F1_3_1006,P_3_1006,R_3_1006 = compute_F1(prediction_1006.cuda(), lab_1006.cuda(), 'overall', k_val=3)
F1_5_1006,P_5_1006,R_5_1006 = compute_F1(logits_1006_5.cuda(), lab_1006.cuda(), 'overall', k_val=5)
print('GZSL AP:%.4f',torch.mean(ap_1006))
print('g_k=3:%.4f,%.4f,%.4f',torch.mean(F1_3_1006), torch.mean(P_3_1006), torch.mean(R_3_1006))
print('g_k=5:%.4f,%.4f,%.4f',torch.mean(F1_5_1006), torch.mean(P_5_1006), torch.mean(R_5_1006))