forked from estebandito22/NYUDeepLearningProject
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathpredict.py
188 lines (152 loc) · 6.59 KB
/
predict.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
import os
import sys
import json
import pickle
import argparse
import torch
import torch.nn as nn
from torch.autograd import Variable
import torch.nn.functional as functional
import torch.optim as optim
import input.dataloader as loader
import layers.utils as utils
from csal import CSAL
from stal import STAL
try:
from layers.wordspretrained import PretrainedEmbeddings
except:
from wordspretrained import PretrainedEmbeddings
USE_CUDA = False
if torch.cuda.is_available():
USE_CUDA = True
def _caption(hyp, videoid, vocab):
generatedstring = ' '.join([str(vocab.index2word[index.cpu().item()]) for index in hyp[1:-1]])
string_hyp = {'videoid': str(videoid), 'captions': [generatedstring]}
#print(string_hyp)
return string_hyp
def evaluate(dataloader, model_filepath, vocab, epoch, model_name, spatial, map_location, returntype = 'ALL'):
checkpoint = torch.load(model_filepath, map_location=map_location)
print(checkpoint['dict_args'])
if not spatial : model = CSAL(checkpoint['dict_args'])
else : model = STAL(checkpoint['dict_args'])
model = nn.DataParallel(model) if data_parallel else model
model.load_state_dict(checkpoint['state_dict'])
cur_dir = os.getcwd()
input_dir = 'input'
output_dir = 'output'
MSRVTT_dir = 'MSRVTT'
predcaptionsjson = 'epoch{}_predcaptions.json'.format(epoch)
valscoresjson = 'epoch{}_valscores.json'.format(epoch)
stringcaptions = []
criterion = nn.NLLLoss(reduce = False)
if USE_CUDA:
model = model.cuda()
criterion = criterion.cuda()
for i,batch in enumerate(dataloader):
#######Load Data
padded_imageframes_batch = Variable(torch.stack(batch[0]), volatile=True) #batch_size*num_frames*3*224*224
frame_sequence_lengths = Variable(torch.LongTensor(batch[1]), volatile=True) #batch_size
padded_inputwords_batch = Variable(torch.LongTensor([[vocab.word2index['<bos>']]]), volatile=True) #batch_size*num_words
dummy_input_sequence_lengths = Variable(torch.LongTensor([[0]]), volatile=True) #batch_size
video_ids_list = batch[2]
if USE_CUDA:
padded_imageframes_batch = padded_imageframes_batch.cuda()
frame_sequence_lengths = frame_sequence_lengths.cuda()
padded_inputwords_batch = padded_inputwords_batch.cuda()
dummy_input_sequence_lengths = dummy_input_sequence_lengths.cuda()
#######Forward
model.eval()
indexcaption = model(padded_imageframes_batch, frame_sequence_lengths, \
padded_inputwords_batch, dummy_input_sequence_lengths)
#######Captions
#print(indexcaption)
stringcaptions += [_caption(indexcaption, video_ids_list[0], vocab)]
#######Write predicted captions
if not os.path.isdir(os.path.join(cur_dir, output_dir, MSRVTT_dir, model_name)):
os.makedirs(os.path.join(cur_dir, output_dir, MSRVTT_dir, model_name))
with open(os.path.join(cur_dir, output_dir, MSRVTT_dir, model_name, predcaptionsjson), 'w') as predsout:
json.dump(stringcaptions, predsout)
return
if __name__=="__main__":
"""python predict.py -m baseline1 -e 100 -vf Alexnet -sp True"""
ap = argparse.ArgumentParser()
ap.add_argument("-bs", "--batch_size", default=1, required=False,
help="If predict is True, the batch size to use during predictions.")
ap.add_argument("-m", "--saved_model_dir", required=False,
help="If predict True, the directory of the model you wish to load.")
ap.add_argument("-e", "--saved_model_epoch", required=False,
help="If predict True, the epoch you want to load e.g. 'epoch0'.")
ap.add_argument("-vf", "--video_features", required=True,
help="Either Resnet or Alexnet")
ap.add_argument("-sp", "--spatial_features", action='store_true',
required=False, help="use spatial features")
args = vars(ap.parse_args())
EVAL_BATCH_SIZE = int(args['batch_size'])
#EPOCH = int(args['saved_model_epoch'])
SPATIAL = args['spatial_features']
VID_FEATS = args['video_features']
EPOCHS = args['saved_model_epoch'].split(',')
print(EPOCHS)
cur_dir = os.getcwd()
input_dir = 'input'
output_dir = 'output'
MSRVTT_dir = 'MSRVTT'
models_dir = 'models'
saved_model_dir = args['saved_model_dir']
#epoch_dir = 'epoch'+args['saved_model_epoch']
#csalfile = 'csal.pth'
csalfile = 'stal.pth'
glovefile = 'glove.pkl'
vocabfile = 'vocab.pkl'
captionsjson = 'captions.json'
#predcaptionsjson = 'epoch{}_predcaptions.json'.format(EPOCH)
#valscoresjson = 'epoch{}_valscores.json'.format(EPOCH)
glove_filepath = os.path.join(cur_dir, models_dir, saved_model_dir, glovefile)
vocab_filepath = os.path.join(cur_dir, models_dir, saved_model_dir, vocabfile)
#model_filepath = os.path.join(cur_dir, models_dir, saved_model_dir, epoch_dir, csalfile)
#captions_filepath = os.path.join(cur_dir, input_dir, MSRVTT_dir, captionsjson)
#preds_filepath = os.path.join(cur_dir, output_dir, MSRVTT_dir, saved_model_dir, predcaptionsjson)
print("Loading previously trained model...")
if USE_CUDA == True:
maplocation = None
else:
maplocation = 'cpu'
data_parallel = False
frame_trunc_length = 45
val_num_workers = 0
val_pretrained = True
val_pklexist = True
spatial = SPATIAL
#checkpoint = torch.load(model_filepath, map_location=maplocation)
'''if not spatial : model = CSAL(checkpoint['dict_args'])
else : model = STAL(checkpoint['dict_args'])
model = nn.DataParallel(model) if data_parallel else model
model.load_state_dict(checkpoint['state_dict'])'''
glovefile = open(glove_filepath, 'rb')
glove = pickle.load(glovefile)
glovefile.close()
vocabfile = open(vocab_filepath, 'rb')
vocab = pickle.load(vocabfile)
vocabfile.close()
print("Get validation data...")
if VID_FEATS == "Resnet":
val_pkl_file = 'MSRVTT/Pixel/Resnet51222/valvideo.pkl'
elif VID_FEATS == "Alexnet":
if not spatial : val_pkl_file = 'MSRVTT/Pixel/Alexnet1000/valvideo.pkl'
else: val_pkl_file = 'MSRVTT/Pixel/Alexnet25622/valvideo.pkl'
file_names = [('MSRVTT/captions.json', 'MSRVTT/valvideo.json', 'MSRVTT/Frames')]
files = [[os.path.join(cur_dir, input_dir, filetype) for filetype in file] for file in file_names]
val_pkl_path = os.path.join(cur_dir, input_dir, val_pkl_file)
val_dataloader = loader.get_val_data(files, val_pkl_path, vocab, glove,
batch_size=EVAL_BATCH_SIZE,
num_workers=val_num_workers,
pretrained=val_pretrained,
pklexist= val_pklexist,
data_parallel=data_parallel,
frame_trunc_length=frame_trunc_length)
for EPOCH in EPOCHS:
epoch_dir = 'epoch'+EPOCH
EPOCH = int(EPOCH)
model_filepath = os.path.join(cur_dir, models_dir, saved_model_dir, epoch_dir, csalfile)
evaluate(val_dataloader, model_filepath, vocab, epoch=EPOCH, model_name=saved_model_dir, spatial=spatial, map_location=maplocation, returntype='Bleu')
#python predict.py -m SpatTempstal -e 12 -vf Alexnet -sp