RetentionLabs commited on
Commit
0be676a
·
verified ·
1 Parent(s): 2a4744f

Add files using upload-large-folder tool

Browse files
README.md ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ configs:
3
+ - config_name: default
4
+ data_files:
5
+ - split: train
6
+ path: metadata/train.parquet
7
+ - split: valid
8
+ path: metadata/valid.parquet
9
+ - split: test
10
+ path: metadata/test.parquet
11
+ ---
12
+
13
+ # stream_align
14
+
15
+ The table above is one row of metadata per row of the corpus. The rows themselves are
16
+ the tar shards in this repository, which carry the four streams as arrays and are
17
+ read by unpacking a snapshot rather than through a `datasets` builder.
18
+
19
+ | split | rows | shards |
20
+ | --- | ---: | --- |
21
+ | `train` | 120,000 | `train-*.tar` |
22
+ | `valid` | 5,730 | `valid-*.tar` |
23
+ | `test` | 4,830 | `test-*.tar` |
build_report.json CHANGED
@@ -1,156 +1,151 @@
1
  {
2
- "phase": "channel_align",
3
  "seed": 0,
4
- "rows": 87656,
5
  "rows_by_lang": {
6
- "ko": 42050,
7
- "en": 45606
8
  },
9
- "rows_with_voice_prompt": 52036,
10
- "rows_with_tool_prompt": 7714,
11
- "rows_with_voice_hint": 25660,
12
- "rows_prompt_withheld": 9080,
13
  "prompt_frames": {
14
  "median": 48,
15
- "p10": 42,
16
  "p90": 48,
17
- "below_full": 0.127
18
  },
19
  "by_lang": {
20
- "ko": {
21
- "rows": 42050,
22
- "rows_with_voice_prompt": 25673,
23
- "prompt_share": 0.6105,
24
- "rows_with_voice_hint": 12686,
25
- "rows_prompt_withheld": 4609,
26
- "prompt_frames": {
27
- "median": 48,
28
- "p10": 31,
29
- "p90": 48,
30
- "below_full": 0.222
31
- }
32
- },
33
  "en": {
34
- "rows": 45606,
35
- "rows_with_voice_prompt": 26363,
36
- "prompt_share": 0.5781,
37
- "rows_with_voice_hint": 12974,
38
- "rows_prompt_withheld": 4471,
39
  "prompt_frames": {
40
  "median": 48,
41
  "p10": 48,
42
  "p90": 48,
43
- "below_full": 0.0345
44
  }
45
  }
46
  },
47
- "rows_skipped": {
48
- "other_speaker_bleed": 12683
49
- },
50
- "voice_hint_share_of_prompted": 0.4931,
51
- "tool_prompt_share_realized": 0.088,
52
- "zone_a_frames_mean": 54.6,
53
- "zone_a_share_of_timeline": 0.0761,
54
  "opens": {
55
- "self": 43682,
56
- "user": 43974
57
  },
58
  "row_frames": {
59
- "p10": 156,
60
- "median": 796,
61
- "p90": 1222,
62
  "max": 1498
63
  },
64
- "rows_in_target_range": 0.5443,
65
  "rows_over_hard_cap": 0,
66
- "rows_supply_bound": 52225,
67
  "frames_by_lang": {
68
- "ko": 27739396,
69
- "en": 35174319
70
  },
71
  "frame_share_by_lang": {
72
- "en": 0.5591,
73
- "ko": 0.4409
 
 
 
 
 
 
 
 
 
 
74
  },
75
  "frame_share_declared": {
76
  "ko": 0.549,
77
  "en": 0.451
78
  },
79
  "turns": {
80
- "p10": 2,
81
- "median": 9,
82
  "p90": 15,
83
  "max": 17
84
  },
85
  "frames_by_self_supply": {
86
  "1": {
87
- "p10": 103,
88
- "median": 199,
89
- "p90": 388,
90
- "max": 921,
91
- "rows": 26540
92
  },
93
  "2-3": {
94
- "p10": 245,
95
- "median": 412,
96
- "p90": 662,
97
- "max": 1161,
98
- "rows": 7174
99
  },
100
  "4-7": {
101
- "p10": 717,
102
- "median": 976,
103
- "p90": 1254,
104
- "max": 1498,
105
- "rows": 39100
106
  },
107
  "8+": {
108
- "p10": 774,
109
- "median": 1056,
110
- "p90": 1327,
111
- "max": 1496,
112
- "rows": 14842
113
  }
114
  },
115
- "frames_total": 62913715,
116
- "hours_of_timeline": 1398.08,
117
- "sources": {
118
- "ko": [
119
- {
120
- "name": "kspon",
121
- "rows": 587185,
122
- "build_report": {
123
- "source": "kspon",
124
- "reader": "project_amnesty.datasets.kspon.KsponDataset",
125
- "repo_commit": "30870612497d014e4f32f4e266fb90df1501bb4c",
126
- "cache_version": 2,
127
- "rows": 594781,
128
- "rows_by_subset": {
129
- "train": 587185,
130
- "valid": 2398,
131
- "test": 5198
132
- },
133
- "stats": {
134
- "total": 627820,
135
- "cached": 589573,
136
- "encoded": 5208,
137
- "failed": 33039
138
- },
139
- "seconds": 4388.7,
140
- "rows_per_second": 135.53,
141
- "processor": "/home/work/voice_research/haan/trainer_output/init_scale_match",
142
- "align_model": "kresnik/wav2vec2-large-xlsr-korean",
143
- "mimi": "kyutai/mimi",
144
- "target_lufs": -20.0,
145
- "encode_batch_size": 32,
146
- "device": "cuda",
147
- "source_root": "data/corpora",
148
- "limit": null,
149
- "python": "3.12.3",
150
- "torch": "2.11.0+cu128"
151
- }
152
  }
153
- ],
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
154
  "en": [
155
  {
156
  "name": "libritts_r",
@@ -188,47 +183,6 @@
188
  "python": "3.12.3",
189
  "torch": "2.11.0+cu128"
190
  }
191
- },
192
- {
193
- "name": "msp_podcast",
194
- "rows": 157363,
195
- "build_report": {
196
- "source": "msp_podcast",
197
- "reader": "project_amnesty.datasets.msp_podcast.MspPodcastDataset",
198
- "repo_commit": "d257d9308d235c351e2c00f71ebf9e10b1e92de2",
199
- "cache_version": 2,
200
- "rows": 245006,
201
- "rows_by_subset": {
202
- "test": 55539,
203
- "train": 157363,
204
- "valid": 32104
205
- },
206
- "stats": {
207
- "total": 245273,
208
- "cached": 0,
209
- "encoded": 245006,
210
- "failed": 267,
211
- "drops": {
212
- "licence_licence": 19014,
213
- "licence_no_licence": 3470,
214
- "no_transcript": 7,
215
- "empty_transcript": 141,
216
- "align": 267
217
- }
218
- },
219
- "seconds": 7643.2,
220
- "rows_per_second": 32.06,
221
- "processor": "/home/work/voice_research/haan/trainer_output/init_scale_match",
222
- "align_model": "facebook/wav2vec2-base-960h",
223
- "mimi": "kyutai/mimi",
224
- "target_lufs": -20.0,
225
- "encode_batch_size": 8,
226
- "device": "cuda",
227
- "source_root": "data/corpora",
228
- "limit": null,
229
- "python": "3.12.3",
230
- "torch": "2.11.0+cu128"
231
- }
232
  }
233
  ]
234
  },
@@ -266,26 +220,12 @@
266
  "tool_prompt_share_declared": 0.125,
267
  "num_codebooks": 8,
268
  "eligibility": {
269
- "ko": [
270
- {
271
- "name": "kspon",
272
- "rows": 587185,
273
- "rows_label_covered": 442086,
274
- "rows_pairable_in_split": 377915
275
- }
276
- ],
277
  "en": [
278
  {
279
  "name": "libritts_r",
280
  "rows": 149291,
281
  "rows_label_covered": null,
282
  "rows_pairable_in_split": null
283
- },
284
- {
285
- "name": "msp_podcast",
286
- "rows": 157363,
287
- "rows_label_covered": null,
288
- "rows_pairable_in_split": null
289
  }
290
  ]
291
  },
@@ -295,8 +235,8 @@
295
  "test"
296
  ]
297
  },
298
- "repo_commit": "b4e5a02f5a9f4f0794dc86fbd139c25abf3b3940",
299
- "seconds": 2585.6,
300
  "python": "3.12.3",
301
  "pipeline_validation": true
302
  }
 
1
  {
2
+ "phase": "stream_align",
3
  "seed": 0,
4
+ "rows": 130560,
5
  "rows_by_lang": {
6
+ "en": 130560
 
7
  },
8
+ "rows_with_voice_prompt": 110917,
9
+ "rows_with_tool_prompt": 9216,
10
+ "rows_with_voice_hint": 55478,
11
+ "rows_prompt_withheld": 19633,
12
  "prompt_frames": {
13
  "median": 48,
14
+ "p10": 48,
15
  "p90": 48,
16
+ "below_full": 0.0696
17
  },
18
  "by_lang": {
 
 
 
 
 
 
 
 
 
 
 
 
 
19
  "en": {
20
+ "rows": 130560,
21
+ "rows_with_voice_prompt": 110917,
22
+ "prompt_share": 0.8495,
23
+ "rows_with_voice_hint": 55478,
24
+ "rows_prompt_withheld": 19633,
25
  "prompt_frames": {
26
  "median": 48,
27
  "p10": 48,
28
  "p90": 48,
29
+ "below_full": 0.0696
30
  }
31
  }
32
  },
33
+ "rows_skipped": null,
34
+ "voice_hint_share_of_prompted": 0.5002,
35
+ "tool_prompt_share_realized": 0.0706,
36
+ "zone_a_frames_mean": 50.0,
37
+ "zone_a_share_of_timeline": 0.0495,
 
 
38
  "opens": {
39
+ "self": 65358,
40
+ "user": 65202
41
  },
42
  "row_frames": {
43
+ "p10": 756,
44
+ "median": 1002,
45
+ "p90": 1289,
46
  "max": 1498
47
  },
48
+ "rows_in_target_range": 0.9102,
49
  "rows_over_hard_cap": 0,
50
+ "rows_supply_bound": 46877,
51
  "frames_by_lang": {
52
+ "en": 132073045
 
53
  },
54
  "frame_share_by_lang": {
55
+ "en": 1.0
56
+ },
57
+ "frame_share_by_subset": {
58
+ "test": {
59
+ "en": 1.0
60
+ },
61
+ "train": {
62
+ "en": 1.0
63
+ },
64
+ "valid": {
65
+ "en": 1.0
66
+ }
67
  },
68
  "frame_share_declared": {
69
  "ko": 0.549,
70
  "en": 0.451
71
  },
72
  "turns": {
73
+ "p10": 8,
74
+ "median": 12,
75
  "p90": 15,
76
  "max": 17
77
  },
78
  "frames_by_self_supply": {
79
  "1": {
80
+ "p10": 118,
81
+ "median": 245,
82
+ "p90": 326,
83
+ "max": 375,
84
+ "rows": 10
85
  },
86
  "2-3": {
87
+ "p10": 243,
88
+ "median": 396,
89
+ "p90": 689,
90
+ "max": 739,
91
+ "rows": 34
92
  },
93
  "4-7": {
94
+ "p10": 754,
95
+ "median": 989,
96
+ "p90": 1271,
97
+ "max": 1497,
98
+ "rows": 87125
99
  },
100
  "8+": {
101
+ "p10": 763,
102
+ "median": 1032,
103
+ "p90": 1318,
104
+ "max": 1498,
105
+ "rows": 43391
106
  }
107
  },
108
+ "ctc_target_ratio": {
109
+ "assistant": {
110
+ "frame_ratio": {
111
+ "rows": 130557,
112
+ "p10": 0.0808,
113
+ "median": 0.1225,
114
+ "p90": 0.1665,
115
+ "max": 0.2651,
116
+ "share_over_one": 0.0
117
+ },
118
+ "speech_ratio": {
119
+ "rows": 130557,
120
+ "p10": 0.3111,
121
+ "median": 0.3687,
122
+ "p90": 0.4349,
123
+ "max": 1.0,
124
+ "share_over_one": 0.0
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
125
  }
126
+ },
127
+ "user": {
128
+ "frame_ratio": {
129
+ "rows": 130556,
130
+ "p10": 0.0849,
131
+ "median": 0.1296,
132
+ "p90": 0.1723,
133
+ "max": 0.2708,
134
+ "share_over_one": 0.0
135
+ },
136
+ "speech_ratio": {
137
+ "rows": 130556,
138
+ "p10": 0.3256,
139
+ "median": 0.3631,
140
+ "p90": 0.4055,
141
+ "max": 1.0,
142
+ "share_over_one": 0.0
143
+ }
144
+ }
145
+ },
146
+ "frames_total": 132073045,
147
+ "hours_of_timeline": 2934.96,
148
+ "sources": {
149
  "en": [
150
  {
151
  "name": "libritts_r",
 
183
  "python": "3.12.3",
184
  "torch": "2.11.0+cu128"
185
  }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
186
  }
187
  ]
188
  },
 
220
  "tool_prompt_share_declared": 0.125,
221
  "num_codebooks": 8,
222
  "eligibility": {
 
 
 
 
 
 
 
 
223
  "en": [
224
  {
225
  "name": "libritts_r",
226
  "rows": 149291,
227
  "rows_label_covered": null,
228
  "rows_pairable_in_split": null
 
 
 
 
 
 
229
  }
230
  ]
231
  },
 
235
  "test"
236
  ]
237
  },
238
+ "repo_commit": "aa1915e249851610fa71b4bf7b99e53399d23b1f",
239
+ "seconds": 1994.7,
240
  "python": "3.12.3",
241
  "pipeline_validation": true
242
  }
metadata/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a213915aa91231ca662219181aa242d5563240c06db4b7fe13adf87c0173b890
3
+ size 18790938
metadata/train.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:07a185bf391422cdd9e9a156ae81eadaca691a676d850786b5b8348f329a29a7
3
+ size 468325924
metadata/valid.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:655b7c6023622b29a834e9399471e9c824d7c28dbf0fec52deb863e7c3ce2189
3
+ size 22520805
shards.json CHANGED
@@ -11,13 +11,13 @@
11
  1
12
  ],
13
  "rows": {
14
- "test": 9660,
15
- "train": 24486,
16
- "valid": 11460
17
  },
18
  "shards": {
19
  "test": 1,
20
- "train": 2,
21
  "valid": 1
22
  }
23
  }
 
11
  1
12
  ],
13
  "rows": {
14
+ "train": 120000,
15
+ "test": 4830,
16
+ "valid": 5730
17
  },
18
  "shards": {
19
  "test": 1,
20
+ "train": 9,
21
  "valid": 1
22
  }
23
  }
test-0000.tar CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8072bd228f803c8f106d9357dcdaad86f6d1501e8260b7d65f0798935aeecded
3
- size 224563200
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fc5c85bcf80a266db963c36b75ebdd962e010417ad13cc71283bc3506672091e
3
+ size 193167360
train-0000.tar CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:32afda7558dc9991e653680fbd1ce3dbe6de74383889f98699846d2c449399d0
3
- size 615168000
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:459cd602ccac81de3dbbffd07d2a8890dfdbeee37028916e077355191aec9097
3
+ size 589864960
train-0001.tar CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8d28fa174f3fdacf5da7c39c0534cdff3f08c2e9596c14351ce01db04aca9b64
3
- size 74158080
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8af592e3f3931da301e0b94daf2d0e768f03471bccf23fea1bba0ddbbe215483
3
+ size 589783040
train-0002.tar ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:76da3c3038ad3b1e91400c70385c19cfa8380040d56df25f338682e47897957b
3
+ size 589916160
train-0003.tar ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9a334fe71113a88bca2c1b8d8c7aefb0406fd1a2a36a0651a8c4c5e9ac616b1a
3
+ size 589772800
train-0004.tar ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:38c188fb4c47f8a1cb4a1c0499a3e8d5e2093231de735a5a03afe528ed2b19e1
3
+ size 589895680
train-0005.tar ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9d25d148515213f70883f0a79b1fe363fedff2d88307d25acfba98bdb46dfeae
3
+ size 589742080
train-0006.tar ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b162a717c2c932206d3b6192a40959f3363d59474c26ba3480604db650d77cf1
3
+ size 589967360
train-0007.tar ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b55cd59ff727d09919985e80c57fbdd1590ca674e4864f1ebaf5f2f820f017a0
3
+ size 589905920
train-0008.tar ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e77173b292d807c87bd6e6adac88ba1017226a7aa060f7d0618e0fb5bdacec3d
3
+ size 70184960
valid-0000.tar CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7bbccda397c61a48159a19e9c49070b4fc747f849b64f54d094e07da28459fb8
3
- size 262000640
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:756014c70599863bdcd2e11596e419de445a53a86ade3519f3f65b5a4edf4db2
3
+ size 229130240