-
Notifications
You must be signed in to change notification settings - Fork 187
/
test_video_captioning_from_frames_mapper.py
232 lines (197 loc) · 10.2 KB
/
test_video_captioning_from_frames_mapper.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
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
import os
import unittest
from data_juicer.core.data import NestedDataset as Dataset
from data_juicer.ops.mapper.video_captioning_from_frames_mapper import \
VideoCaptioningFromFramesMapper
from data_juicer.utils.mm_utils import SpecialTokens
from data_juicer.utils.unittest_utils import (SKIPPED_TESTS,
DataJuicerTestCaseBase)
# Skip tests for this OP in the GitHub actions due to OOM on the current runner
# These tests have been tested locally.
@SKIPPED_TESTS.register_module()
class VideoCaptioningFromFramesMapperTest(DataJuicerTestCaseBase):
data_path = os.path.join(os.path.dirname(os.path.realpath(__file__)), '..',
'data')
vid1_path = os.path.join(data_path, 'video1.mp4')
vid2_path = os.path.join(data_path, 'video2.mp4')
hf_img2seq = 'Salesforce/blip2-opt-2.7b'
@classmethod
def tearDownClass(cls) -> None:
super().tearDownClass(cls.hf_img2seq)
def _run_mapper(self, ds_list, op, num_proc=1, caption_num=0):
dataset = Dataset.from_list(ds_list)
dataset = dataset.map(op.process, num_proc=num_proc, with_rank=True)
dataset_list = dataset.select_columns(column_names=['text']).to_list()
# assert the caption is generated successfully in terms of not_none
# as the generated content is not deterministic
self.assertEqual(len(dataset_list), caption_num)
def test_default_params_no_eoc(self):
ds_list = [{
'text': f'{SpecialTokens.video}白色的小羊站在一旁讲话。旁边还有两只灰色猫咪和一只拉着灰狼的猫咪',
'videos': [self.vid1_path]
}, {
'text': f'{SpecialTokens.video}身穿白色上衣的男子，拿着一个东西，拍打自己的胃',
'videos': [self.vid2_path]
}]
op = VideoCaptioningFromFramesMapper(hf_img2seq=self.hf_img2seq)
self._run_mapper(ds_list, op, caption_num=len(ds_list) * 2)
def test_default_params_with_eoc(self):
ds_list = [
{
'text':
f'{SpecialTokens.video}白色的小羊站在一旁讲话。旁边还有两只灰色猫咪和一只拉着灰狼的猫咪'
f'{SpecialTokens.eoc}',
'videos': [self.vid1_path]
},
{
'text':
f'{SpecialTokens.video}身穿白色上衣的男子，拿着一个东西，拍打自己的胃{SpecialTokens.eoc}', # noqa: E501
'videos': [self.vid2_path]
}
]
op = VideoCaptioningFromFramesMapper(hf_img2seq=self.hf_img2seq)
self._run_mapper(ds_list, op, caption_num=len(ds_list) * 2)
def test_multi_candidate_keep_random_any(self):
ds_list = [{
'text': f'{SpecialTokens.video}白色的小羊站在一旁讲话。旁边还有两只灰色猫咪和一只拉着灰狼的猫咪',
'videos': [self.vid1_path]
}, {
'text': f'{SpecialTokens.video}身穿白色上衣的男子，拿着一个东西，拍打自己的胃',
'videos': [self.vid2_path]
}]
caption_num = 4
op = VideoCaptioningFromFramesMapper(hf_img2seq=self.hf_img2seq,
caption_num=caption_num,
keep_candidate_mode='random_any')
self._run_mapper(ds_list, op, caption_num=len(ds_list) * 2)
def test_multi_candidate_keep_all(self):
ds_list = [{
'text': f'{SpecialTokens.video}白色的小羊站在一旁讲话。旁边还有两只灰色猫咪和一只拉着灰狼的猫咪',
'videos': [self.vid1_path]
}, {
'text': f'{SpecialTokens.video}身穿白色上衣的男子，拿着一个东西，拍打自己的胃',
'videos': [self.vid2_path]
}]
caption_num = 4
op = VideoCaptioningFromFramesMapper(hf_img2seq=self.hf_img2seq,
caption_num=caption_num,
keep_candidate_mode='all')
self._run_mapper(ds_list,
op,
caption_num=(1 + caption_num) * len(ds_list))
def test_multi_candidate_keep_similar_one(self):
ds_list = [{
'text': f'{SpecialTokens.video}白色的小羊站在一旁讲话。旁边还有两只灰色猫咪和一只拉着灰狼的猫咪',
'videos': [self.vid1_path]
}, {
'text': f'{SpecialTokens.video}身穿白色上衣的男子，拿着一个东西，拍打自己的胃',
'videos': [self.vid2_path]
}]
caption_num = 4
op = VideoCaptioningFromFramesMapper(
hf_img2seq=self.hf_img2seq,
caption_num=caption_num,
keep_candidate_mode='similar_one_simhash')
self._run_mapper(ds_list, op, caption_num=len(ds_list) * 2)
def test_remove_original_sample(self):
ds_list = [
{
'text':
f'{SpecialTokens.video}白色的小羊站在一旁讲话。旁边还有两只灰色猫咪和一只拉着灰狼的猫咪',
'videos': [self.vid1_path]
},
{
'text':
f'{SpecialTokens.video}身穿白色上衣的男子，拿着一个东西，拍打自己的胃', # noqa: E501
'videos': [self.vid2_path]
}
]
op = VideoCaptioningFromFramesMapper(hf_img2seq=self.hf_img2seq,
keep_original_sample=False)
self._run_mapper(ds_list, op, caption_num=len(ds_list))
def test_multi_candidate_remove_original_sample(self):
ds_list = [{
'text': f'{SpecialTokens.video}白色的小羊站在一旁讲话。旁边还有两只灰色猫咪和一只拉着灰狼的猫咪',
'videos': [self.vid1_path]
}, {
'text': f'{SpecialTokens.video}身穿白色上衣的男子，拿着一个东西，拍打自己的胃',
'videos': [self.vid2_path]
}]
caption_num = 4
op = VideoCaptioningFromFramesMapper(hf_img2seq=self.hf_img2seq,
caption_num=caption_num,
keep_original_sample=False)
self._run_mapper(ds_list, op, caption_num=len(ds_list))
def test_multi_process(self):
ds_list = [{
'text': f'{SpecialTokens.video}白色的小羊站在一旁讲话。旁边还有两只灰色猫咪和一只拉着灰狼的猫咪',
'videos': [self.vid1_path]
}] * 10
op = VideoCaptioningFromFramesMapper(hf_img2seq=self.hf_img2seq)
self._run_mapper(ds_list, op, num_proc=4, caption_num=len(ds_list) * 2)
def test_multi_process_remove_original_sample(self):
ds_list = [{
'text': f'{SpecialTokens.video}白色的小羊站在一旁讲话。旁边还有两只灰色猫咪和一只拉着灰狼的猫咪',
'videos': [self.vid1_path]
}] * 10
op = VideoCaptioningFromFramesMapper(hf_img2seq=self.hf_img2seq,
keep_original_sample=False)
self._run_mapper(ds_list, op, num_proc=4, caption_num=len(ds_list))
def test_frame_sampling_method(self):
ds_list = [{
'text': f'{SpecialTokens.video}白色的小羊站在一旁讲话。旁边还有两只灰色猫咪和一只拉着灰狼的猫咪',
'videos': [self.vid1_path]
}, {
'text': f'{SpecialTokens.video}身穿白色上衣的男子，拿着一个东西，拍打自己的胃',
'videos': [self.vid2_path]
}]
op = VideoCaptioningFromFramesMapper(hf_img2seq=self.hf_img2seq,
frame_sampling_method='uniform')
self._run_mapper(ds_list, op, caption_num=len(ds_list) * 2)
def test_frame_num(self):
ds_list = [{
'text': f'{SpecialTokens.video}白色的小羊站在一旁讲话。旁边还有两只灰色猫咪和一只拉着灰狼的猫咪',
'videos': [self.vid1_path]
}, {
'text': f'{SpecialTokens.video}身穿白色上衣的男子，拿着一个东西，拍打自己的胃',
'videos': [self.vid2_path]
}]
op = VideoCaptioningFromFramesMapper(hf_img2seq=self.hf_img2seq,
frame_sampling_method='uniform',
frame_num=5)
self._run_mapper(ds_list, op, caption_num=len(ds_list) * 2)
def test_horizontal_flip(self):
ds_list = [{
'text': f'{SpecialTokens.video}白色的小羊站在一旁讲话。旁边还有两只灰色猫咪和一只拉着灰狼的猫咪',
'videos': [self.vid1_path]
}, {
'text': f'{SpecialTokens.video}身穿白色上衣的男子，拿着一个东西，拍打自己的胃',
'videos': [self.vid2_path]
}]
op = VideoCaptioningFromFramesMapper(hf_img2seq=self.hf_img2seq,
horizontal_flip=True)
self._run_mapper(ds_list, op, caption_num=len(ds_list) * 2)
def test_vertical_flip(self):
ds_list = [{
'text': f'{SpecialTokens.video}白色的小羊站在一旁讲话。旁边还有两只灰色猫咪和一只拉着灰狼的猫咪',
'videos': [self.vid1_path]
}, {
'text': f'{SpecialTokens.video}身穿白色上衣的男子，拿着一个东西，拍打自己的胃',
'videos': [self.vid2_path]
}]
op = VideoCaptioningFromFramesMapper(hf_img2seq=self.hf_img2seq,
vertical_flip=True)
self._run_mapper(ds_list, op, caption_num=len(ds_list) * 2)
def test_multi_tag(self):
ds_list = [{
'text': f'{SpecialTokens.video}{SpecialTokens.video}'
'白色的小羊站在一旁讲话。旁边还有两只灰色猫咪和一只拉着灰狼的猫咪',
'videos': [
self.vid1_path,
self.vid1_path,
]
}]
op = VideoCaptioningFromFramesMapper(hf_img2seq=self.hf_img2seq)
self._run_mapper(ds_list, op, caption_num=len(ds_list) * 2)
if __name__ == '__main__':
unittest.main()