zhuohan-7 commited on
Commit
a8e955a
·
verified ·
1 Parent(s): 0b95ac3

Upload folder using huggingface_hub

Browse files
examples/ytb_asr_batch1/data-00000-of-00001.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6d980330172ab4c7b26d252d71508c7fc227ca9ab9b8fff34becbeaf455af814
3
+ size 2259320
examples/ytb_asr_batch1/dataset_info.json ADDED
@@ -0,0 +1,87 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "citation": "",
3
+ "description": "",
4
+ "features": {
5
+ "context": {
6
+ "text": {
7
+ "dtype": "string",
8
+ "_type": "Value"
9
+ },
10
+ "audio": {
11
+ "sampling_rate": 16000,
12
+ "_type": "Audio"
13
+ }
14
+ },
15
+ "instruction": {
16
+ "text": {
17
+ "dtype": "string",
18
+ "_type": "Value"
19
+ },
20
+ "audio": {
21
+ "sampling_rate": 16000,
22
+ "_type": "Audio"
23
+ }
24
+ },
25
+ "answer": {
26
+ "text": {
27
+ "dtype": "string",
28
+ "_type": "Value"
29
+ },
30
+ "audio": {
31
+ "sampling_rate": 16000,
32
+ "_type": "Audio"
33
+ }
34
+ },
35
+ "other_attributes": {
36
+ "YouTube": {
37
+ "feature": {
38
+ "dtype": "string",
39
+ "_type": "Value"
40
+ },
41
+ "_type": "Sequence"
42
+ },
43
+ "YouTube End Time": {
44
+ "dtype": "float64",
45
+ "_type": "Value"
46
+ },
47
+ "YouTube Start Time": {
48
+ "dtype": "float64",
49
+ "_type": "Value"
50
+ },
51
+ "audio_path": {
52
+ "dtype": "string",
53
+ "_type": "Value"
54
+ },
55
+ "index": {
56
+ "dtype": "string",
57
+ "_type": "Value"
58
+ },
59
+ "segments": [
60
+ {
61
+ "end_time": {
62
+ "dtype": "float64",
63
+ "_type": "Value"
64
+ },
65
+ "start_time": {
66
+ "dtype": "float64",
67
+ "_type": "Value"
68
+ },
69
+ "text": {
70
+ "dtype": "string",
71
+ "_type": "Value"
72
+ }
73
+ }
74
+ ]
75
+ },
76
+ "language": {
77
+ "dtype": "string",
78
+ "_type": "Value"
79
+ },
80
+ "audio_length": {
81
+ "dtype": "float64",
82
+ "_type": "Value"
83
+ }
84
+ },
85
+ "homepage": "",
86
+ "license": ""
87
+ }
examples/ytb_asr_batch1/state.json ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_data_files": [
3
+ {
4
+ "filename": "data-00000-of-00001.arrow"
5
+ }
6
+ ],
7
+ "_fingerprint": "7360fb5a1a8ab82b",
8
+ "_format_columns": [
9
+ "answer",
10
+ "audio_length",
11
+ "context",
12
+ "instruction",
13
+ "language",
14
+ "other_attributes"
15
+ ],
16
+ "_format_kwargs": {},
17
+ "_format_type": null,
18
+ "_output_all_columns": false,
19
+ "_split": null
20
+ }
examples/ytb_asr_batch2/data-00000-of-00001.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e5322820138a746bfae193ae0a4ed30bc0ebdbd507f1d0eed1204fac03ac49bd
3
+ size 2885640
examples/ytb_asr_batch2/dataset_info.json ADDED
@@ -0,0 +1,60 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "citation": "",
3
+ "description": "",
4
+ "features": {
5
+ "context": {
6
+ "text": {
7
+ "dtype": "string",
8
+ "_type": "Value"
9
+ },
10
+ "audio": {
11
+ "sampling_rate": 16000,
12
+ "_type": "Audio"
13
+ }
14
+ },
15
+ "instruction": {
16
+ "text": {
17
+ "dtype": "string",
18
+ "_type": "Value"
19
+ },
20
+ "audio": {
21
+ "sampling_rate": 16000,
22
+ "_type": "Audio"
23
+ }
24
+ },
25
+ "answer": {
26
+ "text": {
27
+ "dtype": "string",
28
+ "_type": "Value"
29
+ },
30
+ "audio": {
31
+ "sampling_rate": 16000,
32
+ "_type": "Audio"
33
+ }
34
+ },
35
+ "other_attributes": {
36
+ "emotion": {
37
+ "dtype": "string",
38
+ "_type": "Value"
39
+ },
40
+ "file": {
41
+ "dtype": "string",
42
+ "_type": "Value"
43
+ },
44
+ "index": {
45
+ "dtype": "string",
46
+ "_type": "Value"
47
+ }
48
+ },
49
+ "language": {
50
+ "dtype": "string",
51
+ "_type": "Value"
52
+ },
53
+ "audio_length": {
54
+ "dtype": "float64",
55
+ "_type": "Value"
56
+ }
57
+ },
58
+ "homepage": "",
59
+ "license": ""
60
+ }
examples/ytb_asr_batch2/state.json ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_data_files": [
3
+ {
4
+ "filename": "data-00000-of-00001.arrow"
5
+ }
6
+ ],
7
+ "_fingerprint": "6c4b1fab13123a29",
8
+ "_format_columns": [
9
+ "answer",
10
+ "audio_length",
11
+ "context",
12
+ "instruction",
13
+ "language",
14
+ "other_attributes"
15
+ ],
16
+ "_format_kwargs": {},
17
+ "_format_type": null,
18
+ "_output_all_columns": false,
19
+ "_split": null
20
+ }
examples/ytb_asr_batch3_chinese/data-00000-of-00001.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fca176caa14fa2d13d2fdf96c99a0e8dfec17e2b83af14b78a1d18a0d311ed1c
3
+ size 152264
examples/ytb_asr_batch3_chinese/dataset_info.json ADDED
@@ -0,0 +1,68 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "citation": "",
3
+ "description": "",
4
+ "features": {
5
+ "context": {
6
+ "text": {
7
+ "dtype": "string",
8
+ "_type": "Value"
9
+ },
10
+ "audio": {
11
+ "sampling_rate": 16000,
12
+ "_type": "Audio"
13
+ }
14
+ },
15
+ "instruction": {
16
+ "text": {
17
+ "dtype": "string",
18
+ "_type": "Value"
19
+ },
20
+ "audio": {
21
+ "sampling_rate": 16000,
22
+ "_type": "Audio"
23
+ }
24
+ },
25
+ "answer": {
26
+ "text": {
27
+ "dtype": "string",
28
+ "_type": "Value"
29
+ },
30
+ "audio": {
31
+ "sampling_rate": 16000,
32
+ "_type": "Audio"
33
+ }
34
+ },
35
+ "language": {
36
+ "dtype": "string",
37
+ "_type": "Value"
38
+ },
39
+ "audio_length": {
40
+ "dtype": "float64",
41
+ "_type": "Value"
42
+ },
43
+ "other_attributes": {
44
+ "youtube_name": {
45
+ "dtype": "string",
46
+ "_type": "Value"
47
+ },
48
+ "start_time_in_sec": {
49
+ "dtype": "float32",
50
+ "_type": "Value"
51
+ },
52
+ "end_time_in_sec": {
53
+ "dtype": "float32",
54
+ "_type": "Value"
55
+ },
56
+ "link": {
57
+ "dtype": "string",
58
+ "_type": "Value"
59
+ },
60
+ "percentage_of_chinese": {
61
+ "dtype": "string",
62
+ "_type": "Value"
63
+ }
64
+ }
65
+ },
66
+ "homepage": "",
67
+ "license": ""
68
+ }
examples/ytb_asr_batch3_chinese/state.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_data_files": [
3
+ {
4
+ "filename": "data-00000-of-00001.arrow"
5
+ }
6
+ ],
7
+ "_fingerprint": "ee34dce79289288d",
8
+ "_format_columns": null,
9
+ "_format_kwargs": {},
10
+ "_format_type": null,
11
+ "_output_all_columns": false,
12
+ "_split": null
13
+ }
examples/ytb_asr_batch3_malay/data-00000-of-00001.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9659b947c6b10c86c127690ac15403d40f9e71e541a4bb4391545d6329c7313f
3
+ size 702736
examples/ytb_asr_batch3_malay/dataset_info.json ADDED
@@ -0,0 +1,68 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "citation": "",
3
+ "description": "",
4
+ "features": {
5
+ "context": {
6
+ "text": {
7
+ "dtype": "string",
8
+ "_type": "Value"
9
+ },
10
+ "audio": {
11
+ "sampling_rate": 16000,
12
+ "_type": "Audio"
13
+ }
14
+ },
15
+ "instruction": {
16
+ "text": {
17
+ "dtype": "string",
18
+ "_type": "Value"
19
+ },
20
+ "audio": {
21
+ "sampling_rate": 16000,
22
+ "_type": "Audio"
23
+ }
24
+ },
25
+ "answer": {
26
+ "text": {
27
+ "dtype": "string",
28
+ "_type": "Value"
29
+ },
30
+ "audio": {
31
+ "sampling_rate": 16000,
32
+ "_type": "Audio"
33
+ }
34
+ },
35
+ "language": {
36
+ "dtype": "string",
37
+ "_type": "Value"
38
+ },
39
+ "audio_length": {
40
+ "dtype": "float64",
41
+ "_type": "Value"
42
+ },
43
+ "other_attributes": {
44
+ "youtube_name": {
45
+ "dtype": "string",
46
+ "_type": "Value"
47
+ },
48
+ "start_time_in_sec": {
49
+ "dtype": "float32",
50
+ "_type": "Value"
51
+ },
52
+ "end_time_in_sec": {
53
+ "dtype": "float32",
54
+ "_type": "Value"
55
+ },
56
+ "link": {
57
+ "dtype": "string",
58
+ "_type": "Value"
59
+ },
60
+ "percentage_of_malay": {
61
+ "dtype": "string",
62
+ "_type": "Value"
63
+ }
64
+ }
65
+ },
66
+ "homepage": "",
67
+ "license": ""
68
+ }
examples/ytb_asr_batch3_malay/state.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_data_files": [
3
+ {
4
+ "filename": "data-00000-of-00001.arrow"
5
+ }
6
+ ],
7
+ "_fingerprint": "23fffb8632becbbd",
8
+ "_format_columns": null,
9
+ "_format_kwargs": {},
10
+ "_format_type": null,
11
+ "_output_all_columns": false,
12
+ "_split": null
13
+ }
examples/ytb_pqa_batch1/data-00000-of-00001.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:87e88d6389fa09074fa4a5a918f114c28f675f34d3d2c3e70a1612d7f795ebe7
3
+ size 7184704
examples/ytb_pqa_batch1/dataset_info.json ADDED
@@ -0,0 +1,68 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "citation": "",
3
+ "description": "",
4
+ "features": {
5
+ "context": {
6
+ "text": {
7
+ "dtype": "string",
8
+ "_type": "Value"
9
+ },
10
+ "audio": {
11
+ "sampling_rate": 16000,
12
+ "_type": "Audio"
13
+ }
14
+ },
15
+ "instruction": {
16
+ "text": {
17
+ "dtype": "string",
18
+ "_type": "Value"
19
+ },
20
+ "audio": {
21
+ "sampling_rate": 16000,
22
+ "_type": "Audio"
23
+ }
24
+ },
25
+ "answer": {
26
+ "text": {
27
+ "dtype": "string",
28
+ "_type": "Value"
29
+ },
30
+ "audio": {
31
+ "sampling_rate": 16000,
32
+ "_type": "Audio"
33
+ }
34
+ },
35
+ "other_attributes": {
36
+ "End_Time": {
37
+ "dtype": "null",
38
+ "_type": "Value"
39
+ },
40
+ "Link": {
41
+ "dtype": "null",
42
+ "_type": "Value"
43
+ },
44
+ "Question_Type": {
45
+ "dtype": "null",
46
+ "_type": "Value"
47
+ },
48
+ "Start_Time": {
49
+ "dtype": "null",
50
+ "_type": "Value"
51
+ },
52
+ "Title": {
53
+ "dtype": "null",
54
+ "_type": "Value"
55
+ }
56
+ },
57
+ "language": {
58
+ "dtype": "string",
59
+ "_type": "Value"
60
+ },
61
+ "audio_length": {
62
+ "dtype": "float64",
63
+ "_type": "Value"
64
+ }
65
+ },
66
+ "homepage": "",
67
+ "license": ""
68
+ }
examples/ytb_pqa_batch1/state.json ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_data_files": [
3
+ {
4
+ "filename": "data-00000-of-00001.arrow"
5
+ }
6
+ ],
7
+ "_fingerprint": "51c95b1c01487908",
8
+ "_format_columns": [
9
+ "answer",
10
+ "audio_length",
11
+ "context",
12
+ "instruction",
13
+ "language",
14
+ "other_attributes"
15
+ ],
16
+ "_format_kwargs": {},
17
+ "_format_type": null,
18
+ "_output_all_columns": false,
19
+ "_split": null
20
+ }
examples/ytb_sds_batch1/data-00000-of-00001.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1f0221db597d7c925862295bfbe391ad23173cb823bc900f159b5d01c71c809f
3
+ size 2309240
examples/ytb_sds_batch1/dataset_info.json ADDED
@@ -0,0 +1,68 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "citation": "",
3
+ "description": "",
4
+ "features": {
5
+ "context": {
6
+ "text": {
7
+ "dtype": "string",
8
+ "_type": "Value"
9
+ },
10
+ "audio": {
11
+ "sampling_rate": 16000,
12
+ "_type": "Audio"
13
+ }
14
+ },
15
+ "instruction": {
16
+ "text": {
17
+ "dtype": "string",
18
+ "_type": "Value"
19
+ },
20
+ "audio": {
21
+ "sampling_rate": 16000,
22
+ "_type": "Audio"
23
+ }
24
+ },
25
+ "answer": {
26
+ "text": {
27
+ "dtype": "string",
28
+ "_type": "Value"
29
+ },
30
+ "audio": {
31
+ "sampling_rate": 16000,
32
+ "_type": "Audio"
33
+ }
34
+ },
35
+ "other_attributes": {
36
+ "End_Time": {
37
+ "dtype": "float64",
38
+ "_type": "Value"
39
+ },
40
+ "Link": {
41
+ "dtype": "string",
42
+ "_type": "Value"
43
+ },
44
+ "Question_Type": {
45
+ "dtype": "int64",
46
+ "_type": "Value"
47
+ },
48
+ "Start_Time": {
49
+ "dtype": "float64",
50
+ "_type": "Value"
51
+ },
52
+ "Title": {
53
+ "dtype": "string",
54
+ "_type": "Value"
55
+ }
56
+ },
57
+ "language": {
58
+ "dtype": "string",
59
+ "_type": "Value"
60
+ },
61
+ "audio_length": {
62
+ "dtype": "float64",
63
+ "_type": "Value"
64
+ }
65
+ },
66
+ "homepage": "",
67
+ "license": ""
68
+ }
examples/ytb_sds_batch1/state.json ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_data_files": [
3
+ {
4
+ "filename": "data-00000-of-00001.arrow"
5
+ }
6
+ ],
7
+ "_fingerprint": "001adf7ce82937e0",
8
+ "_format_columns": [
9
+ "answer",
10
+ "audio_length",
11
+ "context",
12
+ "instruction",
13
+ "language",
14
+ "other_attributes"
15
+ ],
16
+ "_format_kwargs": {},
17
+ "_format_type": null,
18
+ "_output_all_columns": false,
19
+ "_split": null
20
+ }
examples/ytb_sqa_batch1/data-00000-of-00001.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5b87dd0297fb72a8e5d3415aec986c7e40ea60e2a54a3a2f32f821073b7e5d46
3
+ size 2116664
examples/ytb_sqa_batch1/dataset_info.json ADDED
@@ -0,0 +1,68 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "citation": "",
3
+ "description": "",
4
+ "features": {
5
+ "context": {
6
+ "text": {
7
+ "dtype": "string",
8
+ "_type": "Value"
9
+ },
10
+ "audio": {
11
+ "sampling_rate": 16000,
12
+ "_type": "Audio"
13
+ }
14
+ },
15
+ "instruction": {
16
+ "text": {
17
+ "dtype": "string",
18
+ "_type": "Value"
19
+ },
20
+ "audio": {
21
+ "sampling_rate": 16000,
22
+ "_type": "Audio"
23
+ }
24
+ },
25
+ "answer": {
26
+ "text": {
27
+ "dtype": "string",
28
+ "_type": "Value"
29
+ },
30
+ "audio": {
31
+ "sampling_rate": 16000,
32
+ "_type": "Audio"
33
+ }
34
+ },
35
+ "other_attributes": {
36
+ "End_Time": {
37
+ "dtype": "float64",
38
+ "_type": "Value"
39
+ },
40
+ "Link": {
41
+ "dtype": "string",
42
+ "_type": "Value"
43
+ },
44
+ "Question_Type": {
45
+ "dtype": "int64",
46
+ "_type": "Value"
47
+ },
48
+ "Start_Time": {
49
+ "dtype": "float64",
50
+ "_type": "Value"
51
+ },
52
+ "Title": {
53
+ "dtype": "string",
54
+ "_type": "Value"
55
+ }
56
+ },
57
+ "language": {
58
+ "dtype": "string",
59
+ "_type": "Value"
60
+ },
61
+ "audio_length": {
62
+ "dtype": "float64",
63
+ "_type": "Value"
64
+ }
65
+ },
66
+ "homepage": "",
67
+ "license": ""
68
+ }
examples/ytb_sqa_batch1/state.json ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_data_files": [
3
+ {
4
+ "filename": "data-00000-of-00001.arrow"
5
+ }
6
+ ],
7
+ "_fingerprint": "f02f9612426fa5b1",
8
+ "_format_columns": [
9
+ "answer",
10
+ "audio_length",
11
+ "context",
12
+ "instruction",
13
+ "language",
14
+ "other_attributes"
15
+ ],
16
+ "_format_kwargs": {},
17
+ "_format_type": null,
18
+ "_output_all_columns": false,
19
+ "_split": null
20
+ }