Hugging Face's logo

Datasets:
dolphinteam
/
OpenWhistle-Detection-Finetuning

Modalities:
Audio
Text
Formats:
parquet
Size:
1K - 10K
ArXiv:
Tags:
arxiv:2609.34839
Libraries:
Datasets
pandas
Polars
Dataset card Data Studio Files
xet
 Community
1
OpenWhistle commited on
Commit
238392c
·
verified ·
1 Parent(s): 922426d

Add files using upload-large-folder tool

Browse files
Files changed (2) hide show
  1. artifacts/split_manifest.csv +0 -0
  2. artifacts/summary.json +163 -0
artifacts/split_manifest.csv ADDED
The diff for this file is too large to render. See raw diff
 
artifacts/summary.json ADDED
@@ -0,0 +1,163 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "assignment_method": "exact_milp_source_label_balanced",
3
+ "binary_counts_total": {
4
+ "noise": 2800,
5
+ "whistle": 2800
6
+ },
7
+ "binary_label_names": [
8
+ "noise",
9
+ "whistle"
10
+ ],
11
+ "features": [
12
+ {
13
+ "dtype": "audio(decode=false)",
14
+ "name": "audio"
15
+ },
16
+ {
17
+ "dtype": "class_label(noise, whistle)",
18
+ "name": "binary_label"
19
+ },
20
+ {
21
+ "dtype": "string",
22
+ "name": "name"
23
+ },
24
+ {
25
+ "dtype": "string",
26
+ "name": "source_label"
27
+ },
28
+ {
29
+ "dtype": "sequence(int64, length=7)",
30
+ "name": "label"
31
+ }
32
+ ],
33
+ "label_names": [
34
+ "SW_Neo",
35
+ "SW_Luna",
36
+ "SW_Nikita",
37
+ "SW_Nana",
38
+ "SW_Yosefa",
39
+ "SW_Dana",
40
+ "NSW_1"
41
+ ],
42
+ "leakage": {
43
+ "any_session_overlap_count": 0,
44
+ "pairwise_session_overlap": {
45
+ "train__test": 0,
46
+ "train__validation": 0,
47
+ "validation__test": 0
48
+ }
49
+ },
50
+ "output_dataset_dir": "openwhistle_detection_finetuning/output/OpenWhistle-1.0-Detection-Finetuning-session-disjoint",
51
+ "previous_split_leakage": {
52
+ "test_rows_total": 1120,
53
+ "test_rows_with_train_session": 1120,
54
+ "train_test_session_overlap": 232,
55
+ "train_test_session_overlap_examples": [
56
+ "13-11-19-1221",
57
+ "Exp_01_Feb_2020_0945am",
58
+ "Exp_01_Feb_2020_1045am",
59
+ "Exp_01_Feb_2020_1145am",
60
+ "Exp_01_Feb_2020_1345pm",
61
+ "Exp_01_Feb_2020_1545pm",
62
+ "Exp_01_Jan_2020_0845am",
63
+ "Exp_01_Jan_2020_0945am",
64
+ "Exp_01_Jan_2020_1045am",
65
+ "Exp_01_Jan_2020_1245pm",
66
+ "Exp_01_Jan_2020_1345pm",
67
+ "Exp_01_Jan_2020_1545pm",
68
+ "Exp_01_Mar_2020_1145am",
69
+ "Exp_01_Mar_2020_1545pm",
70
+ "Exp_01_Mar_2020_1745pm",
71
+ "Exp_02_Feb_2020_1245pm",
72
+ "Exp_02_Feb_2020_1545pm",
73
+ "Exp_02_Jan_2020_0845am",
74
+ "Exp_02_Jan_2020_1045am",
75
+ "Exp_02_Jan_2020_1345pm"
76
+ ]
77
+ },
78
+ "ratios": {
79
+ "test": 0.15,
80
+ "train": 0.7,
81
+ "validation": 0.15
82
+ },
83
+ "repo_id": "OpenWhistleNeurIPS26/OpenWhistle-Detection-Finetuning",
84
+ "rows_total": 5600,
85
+ "seed": 42,
86
+ "sessions_total": 241,
87
+ "source_dir": "/DOLPHIN1/Wh_detection_database_NEURIPS/OW_detection",
88
+ "source_label_counts_total": {
89
+ "Dana": 400,
90
+ "Luna": 400,
91
+ "NSW_1": 400,
92
+ "Nana": 400,
93
+ "Neo": 400,
94
+ "Nikita": 400,
95
+ "Yosefa": 400,
96
+ "noise": 2800
97
+ },
98
+ "splits": {
99
+ "test": {
100
+ "binary_counts": {
101
+ "noise": 420,
102
+ "whistle": 420
103
+ },
104
+ "row_ratio": 0.15,
105
+ "rows": 840,
106
+ "sessions": 38,
107
+ "source_label_counts": {
108
+ "Dana": 60,
109
+ "Luna": 60,
110
+ "NSW_1": 60,
111
+ "Nana": 60,
112
+ "Neo": 60,
113
+ "Nikita": 60,
114
+ "Yosefa": 60,
115
+ "noise": 420
116
+ },
117
+ "target_ratio": 0.15,
118
+ "target_rows": 840.0
119
+ },
120
+ "train": {
121
+ "binary_counts": {
122
+ "noise": 1960,
123
+ "whistle": 1960
124
+ },
125
+ "row_ratio": 0.7,
126
+ "rows": 3920,
127
+ "sessions": 165,
128
+ "source_label_counts": {
129
+ "Dana": 280,
130
+ "Luna": 280,
131
+ "NSW_1": 280,
132
+ "Nana": 280,
133
+ "Neo": 280,
134
+ "Nikita": 280,
135
+ "Yosefa": 280,
136
+ "noise": 1960
137
+ },
138
+ "target_ratio": 0.7,
139
+ "target_rows": 3919.9999999999995
140
+ },
141
+ "validation": {
142
+ "binary_counts": {
143
+ "noise": 420,
144
+ "whistle": 420
145
+ },
146
+ "row_ratio": 0.15,
147
+ "rows": 840,
148
+ "sessions": 38,
149
+ "source_label_counts": {
150
+ "Dana": 60,
151
+ "Luna": 60,
152
+ "NSW_1": 60,
153
+ "Nana": 60,
154
+ "Neo": 60,
155
+ "Nikita": 60,
156
+ "Yosefa": 60,
157
+ "noise": 420
158
+ },
159
+ "target_ratio": 0.15,
160
+ "target_rows": 840.0
161
+ }
162
+ }
163
+ }