Hugging Face's logo

Datasets:
VoxCPM-Kernel
/
VoxCPM2-Turbo-Benchmark-200-Requests

Tasks:
Text-to-Speech
Modalities:
Audio
Text
Formats:
parquet
Languages:
English
Size:
1K - 10K
Libraries:
Datasets
pandas
Polars
License:
Dataset card Data Studio Files
xet
 Community
Dataset Viewer
Auto-converted to Parquet Duplicate
request
int32
1
200
text
stringlengths
19
114
audio
audioduration (s)
1.28
7.04
ttfa_ms
float64
2.37
12.1
playable_ms
float64
6.89
81.2
done_s
float64
0.95
3.21
wer_pct
float64
0
25
utmos
float64
1.29
4.53
dnsmos_ovrl
float64
1.84
3.56
dnsmos_sig
float64
2.54
3.75
dnsmos_bak
float64
2.63
4.24
dnsmos_p808
float64
2.53
4.28
transcript
stringlengths
19
115
audio_s
float64
1.28
7.04
chunk2_ms
float64
136
241
submit_ms
float64
0
9.76
memo_hit
bool
1 class
hit_limit
bool
1 class
babble
bool
1 class
utt
stringlengths
41
49
1
Get the trust fund to the bank early.
12.13
12.13
1.589
0
4.406
2.939
3.208
3.992
3.528
Get the trust fund to the bank early.
2.24
141.21
0
true
false
false
common_voice_en_10119832-common_voice_en_10119840
2
One by one, the campfires were extinguished, and the oasis fell as quiet as the desert.
12.01
12.01
2.849
0
4.412
3.283
3.577
4.073
3.626
One by one the campfires were extinguished and the oasis fell as quiet as the desert.
4.48
141.08
0.13
true
false
false
common_voice_en_103675-common_voice_en_103676
3
The boy knew the desert sensed his fear.
11.95
11.95
1.717
0
4.406
3.257
3.504
4.089
3.618
The boy knew the desert, sensed his fear.
2.4
141.02
0.19
true
false
false
common_voice_en_103675-common_voice_en_103677
4
When it comes to the crunch, our company will become insolvent.
11.88
11.88
2.166
0
2.121
2.501
2.884
3.62
3.195
When it comes to the crunch, our company will become insolvent.
3.04
140.95
0.26
true
false
false
common_voice_en_10933823-common_voice_en_10933822
5
I'm never more aware of a room's acoustics than when I'm trying to enjoy a snack I have no intention of sharing.
11.77
11.77
2.994
0
3.691
3.323
3.646
3.968
3.419
I'm never more aware of a room's acoustics than when I'm trying to enjoy a snack I have no intention of sharing.
5.12
140.84
0.36
true
false
false
common_voice_en_120405-common_voice_en_120402
6
Thousands of people die every year as a consequence of air pollution.
11.7
11.7
2.679
0
4.294
3.235
3.473
4.096
3.819
Thousands of people die every year as a consequence of air pollution.
4
140.77
0.44
true
false
false
common_voice_en_15903802-common_voice_en_15903807
7
I was well, but I'm all the better for being here.
11.65
11.65
2.166
0
3.893
3.24
3.594
3.942
2.828
I was well, but I'm all the better for being here.
3.04
140.72
0.49
true
false
false
common_voice_en_16666041-common_voice_en_16666040
8
what is the forecast for California for rain
11.6
11.6
2.486
0
1.288
1.838
2.538
2.625
2.526
What is the forecast for California for rain?
3.52
140.68
0.53
true
false
false
common_voice_en_167249-common_voice_en_167246
9
Take these capsules over to Mrs. David's house.
11.56
11.56
2.165
0
4.147
3.18
3.452
4.116
3.489
Take these capsules over to Mrs. David's house.
3.04
140.63
0.57
true
false
false
common_voice_en_167249-common_voice_en_167247
10
Lions hunt their prey with sprints, whereas wolves take advantage of their stamina.
11.49
11.49
2.933
0
4.472
3.218
3.485
4.098
3.65
Lions hunt their prey with sprints, whereas wolves take advantage of their stamina.
4.8
140.56
0.65
true
false
false
common_voice_en_17147545-common_voice_en_17147546
11
I heard the land where the hobbits live, the Shire, has actually been filmed in New Zealand.
11.41
11.41
2.892
0
3.902
3.431
3.686
4.126
3.431
I heard the land where the hobbits live, the Shire, has actually been filmed in New Zealand.
4.64
140.49
0.72
true
false
false
common_voice_en_17249419-common_voice_en_17249428
12
It was such a thrill, her veins were pumped full of adrenaline.
11.35
11.35
2.549
0
3.833
3.147
3.493
3.927
3.626
It was such a thrill, her veins were pumped full of adrenaline.
3.68
140.43
0.78
true
false
false
common_voice_en_17271507-common_voice_en_17271510
13
Ninety five lines and no more, that's it.
11.31
11.31
2.679
25
2.86
2.9
3.499
3.325
3.14
95 lines and no more, that's it.
4
140.38
0.83
true
false
false
common_voice_en_17274522-common_voice_en_17274525
14
As he couldn't prove his innocence, he was brought to the detention zone, as if guilty.
11.23
11.23
2.892
0
4.253
3.126
3.372
4.089
3.672
As he couldn't prove his innocence, he was brought to the detention zone as if guilty.
4.64
140.3
0.9
true
false
false
common_voice_en_17281385-common_voice_en_17281386
15
Does the new splinter cell game come with a multiplayer mode?
11.17
11.17
2.165
0
3.292
3.305
3.566
4.125
3.728
Does the new Splinter Cell game come with a multiplayer mode?
3.04
140.24
0.96
true
false
false
common_voice_en_17369337-common_voice_en_17369336
16
Sooner or later, they'll find out how to read these hieroglyphics.
11.12
11.12
2.293
0
3.031
2.746
3.204
3.503
3.181
Sooner or later, they'll find out how to read these hieroglyphics.
3.2
140.19
1.02
true
false
false
common_voice_en_17380945-common_voice_en_17380947
17
They both stormed out, neither thinking to fill me in on where they were going.
11.06
11.06
2.892
0
4.004
2.92
3.519
3.377
3.913
They both stormed out, neither thinking to fill me in on where they were going.
4.64
140.13
1.08
true
false
false
common_voice_en_17406732-common_voice_en_17406734
18
He paused, looked back at the house, but then pocketed the keys, opened the gate and strode down the path quickly.
10.97
10.97
3.164
0
3.295
2.909
3.554
3.323
3.08
He paused, looked back at the house, but then pocketed the keys, opened the gate, and strode down the path quickly.
6.56
140.04
1.16
true
false
false
common_voice_en_17444666-common_voice_en_17444667
19
To summarize, our comprehensive survey discloses the true versatility of their method.
10.91
10.91
2.933
0
4.17
3.153
3.501
3.877
3.748
To summarize, our comprehensive survey discloses the true versatility of their method,
4.8
139.99
1.22
true
false
false
common_voice_en_17444666-common_voice_en_17444668
20
Songs are made to make you feel.
10.87
10.87
1.846
0
4.062
2.729
3.104
3.815
3.718
Songs are made to make you feel.
2.56
139.94
1.26
true
false
false
common_voice_en_17502-common_voice_en_17503
21
The logs fell and tumbled into the clear stream.
10.83
10.83
2.485
0
4.282
3.376
3.614
4.139
3.647
The logs fell and tumbled into the clear stream.
3.52
139.9
1.3
true
false
false
common_voice_en_17727889-common_voice_en_17727904
22
My grandmother has Type One diabetes.
10.8
10.8
1.846
16.67
3.775
3.172
3.533
3.981
2.845
My grandmother has type 1 diabetes.
2.56
139.87
1.34
true
false
false
common_voice_en_17727889-common_voice_en_17727905
23
The loss exposed the city to foreign attackers.
10.76
10.76
1.845
0
2.819
2.663
3.068
3.785
2.95
The loss exposed the city to foreign attackers.
2.56
139.83
1.38
true
false
false
common_voice_en_17729945-common_voice_en_17729944
24
A dog with a water pack is walking through clear green water.
10.71
10.71
2.798
0
4.226
2.967
3.394
3.779
3.503
A dog with a water pack is walking through clear green water.
4.32
139.78
1.42
true
false
false
common_voice_en_177895-common_voice_en_177894
25
Every cloud has a silver lining.
10.66
10.66
1.398
0
3.94
2.993
3.331
3.876
3.255
Every cloud has a silver lining.
1.92
139.73
1.47
true
false
false
common_voice_en_17879420-common_voice_en_17879418
26
Appearances are deceptive.
10.63
10.63
1.398
0
3.109
2.486
2.79
3.777
3.029
Appearances are deceptive.
1.92
139.7
1.51
true
false
false
common_voice_en_17920116-common_voice_en_17920112
27
One of the world's biggest problems is the growth of the human population.
10.57
10.57
2.932
0
3.786
3.08
3.501
3.723
2.777
One of the world's biggest problems is the growth of the human population.
4.8
139.64
1.56
true
false
false
common_voice_en_18012457-common_voice_en_18012447
28
Since I've gotten a dog, the regular visits of the fox have stopped.
10.52
10.52
2.548
0
4.124
3.263
3.551
4.109
3.599
Since I've gotten a dog, the regular visits of the fox have stopped.
3.68
139.59
1.62
true
false
false
common_voice_en_18012457-common_voice_en_18012464
29
The hinge on the door creaked with old age.
10.47
10.47
1.845
0
3.956
3.122
3.435
3.951
3.443
The hinge on the door creaked with old age.
2.56
139.54
1.66
true
false
false
common_voice_en_18070362-common_voice_en_18070351
30
Also there was a decrease in trade interest between east and west.
10.41
10.41
2.485
0
4.255
3.261
3.564
4.031
3.791
Also, there was a decrease in trade interest between East and West.
3.52
139.49
1.72
true
false
false
common_voice_en_18234333-common_voice_en_18234353
31
It seemed the ordained order of things that dogs should work.
10.37
10.37
2.612
0
4.361
3.26
3.571
4.077
3.807
it seemed the ordained order of things that dogs should work.
3.84
139.44
1.77
true
false
false
common_voice_en_18313326-common_voice_en_18313324
32
The history of the eighteenth century is written, Ernest prompted.
10.32
10.32
2.548
20
3.753
3.088
3.418
4.03
3.151
The history of the 18th century is written, earnest, prompted,
3.68
139.39
1.82
true
false
false
common_voice_en_18338543-common_voice_en_18338546
33
My mother has rheumatoid arthritis.
10.27
10.27
2.164
0
1.985
2.433
3.107
3.24
2.744
My mother has rheumatoid arthritis.
3.04
139.34
1.86
true
false
false
common_voice_en_18485041-common_voice_en_18485042
34
Only it must remain a secret between you and me.
10.23
10.23
1.973
0
4.428
3.276
3.574
4.061
3.575
Only it must remain a secret between you and me.
2.72
139.31
1.9
true
false
false
common_voice_en_18575865-common_voice_en_18575863
35
Indeed, he came in and turned our conversation.
10.2
10.2
2.358
0
2.699
2.888
3.146
4.024
3.44
indeed he came in and turned our conversation
3.36
139.27
1.94
true
false
false
common_voice_en_18647546-common_voice_en_18647550
36
In testimony to investigators, the page described the relationship as consensual and not intimidating.
10.13
10.13
3.134
0
3.934
3.232
3.588
3.929
3.596
In testimony to investigators, the page described the relationship as consensual and not intimidating.
6.24
139.2
2.01
true
false
false
common_voice_en_18672480-common_voice_en_18672484
37
The national monument surrounds the Castle Mountain Mine Area.
10.07
10.07
2.358
0
4.299
3.215
3.552
3.987
3.847
The national monument surrounds the Castle Mountain Mine area.
3.36
139.14
2.06
true
false
false
common_voice_en_18699078-common_voice_en_18699081
38
Smith is too shocked by Paine's betrayal to defend himself, and runs away.
10.01
10.01
2.932
7.69
4.447
3.122
3.423
3.915
4.044
Smith is too shocked by Payne's betrayal to defend himself and runs away.
4.8
139.08
2.12
true
false
false
common_voice_en_18707936-common_voice_en_18707937
39
A judge later found that Harrison had unintentionally plagiarized the earlier song.
9.95
9.95
3.047
0
4.089
3.087
3.539
3.718
3.12
A judge later found that Harrison had unintentionally plagiarized the earlier song,
5.44
139.03
2.18
true
false
false
common_voice_en_18786931-common_voice_en_18786946
40
His career took a significant turn in these years.
9.91
9.91
2.291
0
3.338
3.034
3.471
3.7
3.316
His career took a significant turn in these years.
3.2
138.98
2.22
true
false
false
common_voice_en_18828571-common_voice_en_18828573
41
Because of wartime secrecy, he had not been aware of the reactor's existence.
9.86
9.86
2.797
0
4.288
3.453
3.675
4.188
3.604
Because of wartime secrecy he had not been aware of the reactor's existence.
4.32
138.93
2.27
true
false
false
common_voice_en_18829843-common_voice_en_18829841
42
It borders five other Kent districts, as well as East Sussex to the south-west.
9.81
9.81
3.02
0
3.91
3.163
3.495
4.008
3.601
It borders five other Kent districts, as well as East Sussex to the southwest.
5.28
138.88
2.32
true
false
false
common_voice_en_18877786-common_voice_en_18877788
43
In the aftermath of this event, the National Aeronautics and Space Administration was formed.
9.75
9.75
2.797
0
3.485
3.308
3.576
4.114
3.706
In the aftermath of this event, the National Aeronautics and Space Administration was formed
4.32
138.82
2.39
true
false
false
common_voice_en_18902273-common_voice_en_18902274
44
He also make himself useful in a variety of ways to the military.
9.69
9.69
2.358
7.69
4.044
3.208
3.589
3.851
3.22
He also made himself useful in a variety of ways to the military.
3.36
138.76
2.44
true
false
false
common_voice_en_18993020-common_voice_en_18993067
45
There are two cemeteries, one across the street from the other.
9.65
9.65
2.1
0
4.162
3.141
3.46
3.947
3.85
There are two cemeteries, one across the street from the other.
2.88
138.72
2.49
true
false
false
common_voice_en_18994734-common_voice_en_18994736
46
Netarhat is famous for its glorious sunrises and sunsets during the summer months.
9.59
9.59
3.02
7.69
3.008
2.804
3.111
4.04
3.192
Netterhout is famous for its glorious sunrises and sunsets during the summer months.
5.28
138.66
2.54
true
false
false
common_voice_en_19002369-common_voice_en_19002367
47
However, he died in the fire and explosion.
9.55
9.55
1.909
0
3.436
2.704
3.133
3.596
3.33
However, he died in the fire and explosion.
2.56
138.63
2.58
true
false
false
common_voice_en_19017623-common_voice_en_19017624
48
The airport is served by Calm Air and Perimeter Aviation with flights to Winnipeg.
9.49
9.49
2.847
0
4.032
3.073
3.45
3.892
3.949
The airport is served by Calm Air and Perimeter Aviation with flights to Winnipeg.
4.48
138.56
2.64
true
false
false
common_voice_en_19211875-common_voice_en_19211870
49
She stars also in the sequel to "Mirrors", which is directed by Victor Garcia.
9.44
9.44
3.047
0
1.409
2.745
3.13
3.821
3.657
She stars also in the sequel to Mirrors, which is directed by Victor Garcia.
5.44
138.51
2.7
true
false
false
common_voice_en_19361636-common_voice_en_19361634
50
His interpretation of a doctrine of free will became known as Pelagianism.
9.39
9.39
2.547
0
3.828
3.086
3.616
3.643
3.757
His interpretation of a doctrine of free will became known as Pelagianism.
3.68
138.46
2.74
true
false
false
common_voice_en_19361636-common_voice_en_19361635
51
His wife Ginny survives him with their son, two grandchildren and a great grandchild.
9.34
9.34
3.047
14.29
3.532
3.163
3.473
4.009
3.866
His wife Ginny survives him with their son, two grandchildren, and a great-grandchild.
5.44
138.41
2.8
true
false
false
common_voice_en_19397397-common_voice_en_19397399
52
After the war, May returned to work in the textiles industry.
9.28
9.28
2.738
0
4.383
3.295
3.558
4.085
3.9
After the war, May returned to work in the textiles industry.
4.16
138.35
2.85
true
false
false
common_voice_en_19397397-common_voice_en_19397407
53
This marks the culmination of the Physical Education Program, in which every student participates.
9.22
9.22
3.103
0
4.485
3.562
3.747
4.236
4.078
This marks the culmination of the physical education program in which every student participates.
5.92
138.29
2.92
true
false
false
common_voice_en_19449066-common_voice_en_19449075
54
He is an elder in the Presbyterian Church in Ireland.
9.17
9.17
2.1
0
2.3
2.826
3.356
3.507
3.159
He is an elder in the Presbyterian Church in Ireland.
2.88
138.24
2.96
true
false
false
common_voice_en_19517720-common_voice_en_19517721
55
He is now President of Ramblers Scotland.
9.14
9.14
2.357
0
3.928
3.113
3.448
3.929
3.265
He is now president of Ramblers Scotland.
3.36
138.21
3
true
false
false
common_voice_en_19527999-common_voice_en_19528000
56
Eventually the nomenclature was standardized to the conventions used today.
9.09
9.09
2.738
0
4.107
3.206
3.51
3.993
3.982
Eventually, the nomenclature was standardized to the conventions used today.
4.16
138.16
3.05
true
false
false
common_voice_en_19527999-common_voice_en_19528002
57
He worked as a school janitor to pay his tuition.
9.04
9.04
2.1
0
3.849
2.908
3.405
3.597
3.215
He worked as a school janitor to pay his tuition.
2.88
138.11
3.09
true
false
false
common_voice_en_19619551-common_voice_en_19619550
58
Excessive efforts to control or exploit agents of influence can also have negative consequences.
8.99
8.99
3.162
0
4.108
3.07
3.418
3.858
3.958
Excessive efforts to control or exploit agents of influence can also have negative consequences,
6.56
138.06
3.15
true
false
false
common_voice_en_19619855-common_voice_en_19619858
59
The murals draw on the Gospel of John and the Book of Revelation.
8.93
8.93
2.546
0
3.603
3.261
3.568
4
3.661
The murals draw on the Gospel of John and the book of Revelation.
3.68
138
3.2
true
false
false
common_voice_en_19629167-common_voice_en_19629166
60
A statue of the king stands at the southwest entrance to the park.
8.89
8.89
2.611
0
4.181
3.219
3.54
3.966
3.251
A statue of the king stands at the southwest entrance to the park.
3.84
137.96
3.25
true
false
false
common_voice_en_19636954-common_voice_en_19636955
61
For her Friday nights and her French lessons were gone.
8.84
8.84
2.099
0
4.415
3.286
3.551
4.082
3.753
for her Friday nights and her French lessons were gone.
2.88
137.92
3.29
true
false
false
common_voice_en_19656513-common_voice_en_19656571
62
The Reagan administration also asked Congress to ease the rule.
8.8
8.8
2.991
0
3.534
2.961
3.57
3.524
3.452
The Reagan administration also asked Congress to ease the rule
5.12
137.87
3.33
true
false
false
common_voice_en_19665909-common_voice_en_19665910
63
And they're really good.
8.77
8.77
0.951
0
1.464
2.506
2.781
3.844
3
And they're really good.
1.28
137.84
3.36
true
false
false
common_voice_en_19670674-common_voice_en_19670675
64
Former Australian captain Steve Waugh referred to the practice as 'mental disintegration'.
8.71
8.71
3.199
16.67
3.276
3.34
3.576
4.166
3.85
Former Australian Captain Steve Waugh referred to the practice as mental disintegration.
7.04
137.78
3.42
true
false
false
common_voice_en_19690002-common_voice_en_19690003
65
The law has been replaced twice since then.
8.67
8.67
1.908
0
4.085
2.876
3.258
3.898
3.335
The law has been replaced twice since then.
2.56
137.74
3.46
true
false
false
common_voice_en_19691556-common_voice_en_19691557
66
The county is an alcohol prohibition or dry county.
8.63
8.63
1.972
22.22
3.49
3.127
3.423
3.987
3.663
The count E is an alcohol prohibition or dry county.
2.72
137.7
3.5
true
false
false
common_voice_en_19698719-common_voice_en_19698718
67
It presents weapons, vehicles and uniforms, as well as abundant documentation history.
8.58
8.58
2.93
0
4.366
3.272
3.533
4.071
3.908
It presents weapons, vehicles, and uniforms, as well as abundant documentation history.
4.8
137.65
3.55
true
false
false
common_voice_en_19724030-common_voice_en_19724028
68
I came to offer myself as escort to your mother.
8.54
8.54
2.099
0
3.23
2.481
3.039
3.313
3.493
I came to offer myself as escort to your mother.
2.88
137.62
3.59
true
false
false
common_voice_en_19726989-common_voice_en_19726990
69
All listings are with original names.
8.51
8.51
2.226
0
3.458
2.948
3.401
3.595
3.564
All listings are with original names.
3.04
137.58
3.62
true
false
false
common_voice_en_19727183-common_voice_en_19727187
70
Underwater demolition specialists may still be referred to as underwater demolition teams.
8.46
8.46
2.991
0
3.65
2.664
3.453
3.084
3.454
Underwater demolition specialists may still be referred to as underwater demolition teams.
5.12
137.53
3.68
true
false
false
common_voice_en_19730681-common_voice_en_19730680
71
In the past decade, the university has sought to develop a stronger international focus.
8.4
8.4
2.93
0
4.009
3.137
3.551
3.786
3.752
In the past decade, the university has sought to develop a stronger international focus.
4.8
137.47
3.73
true
false
false
common_voice_en_19734247-common_voice_en_19734252
72
Relief carving can be described as "carving pictures in wood".
8.35
8.35
2.796
0
4.523
3.225
3.475
4.107
3.794
Relief carving can be described as carving pictures in wood,
4.32
137.42
3.78
true
false
false
common_voice_en_19748971-common_voice_en_19748972
73
This tended to discourage further development of industry and enterprise.
8.3
8.3
2.546
0
3.993
3.119
3.431
3.984
3.767
This tended to discourage further development of industry and enterprise,
3.68
137.37
3.84
true
false
false
common_voice_en_19760248-common_voice_en_19760250
74
The airport is also the home to the National Weather Service Pocatello Office.
8.25
8.25
2.737
0
3.695
3.088
3.402
3.982
3.986
The airport is also the home to the National Weather Service Pocatello office.
4.16
137.32
3.89
true
false
false
common_voice_en_19767978-common_voice_en_19767980
75
It is a hermaphroditic fluke that requires two intermediate hosts.
8.2
8.2
2.796
0
4.158
3.181
3.464
4.022
3.65
It is a hermaphroditic fluke that requires two intermediate hosts,
4.32
137.27
3.94
true
false
false
common_voice_en_19793104-common_voice_en_19793103
76
For example, they both consist of predominately older, hence redder, stars.
8.15
8.15
2.846
9.09
4.382
3.289
3.571
4.034
3.947
For example, they both consist of predominantly older, hence redder, stars.
4.48
137.22
3.98
true
false
false
common_voice_en_19869780-common_voice_en_19869779
77
She was a pioneer in the sensory analysis of food.
8.11
8.11
2.225
0
4.42
3.124
3.45
3.862
3.845
She was a pioneer in the sensory analysis of food.
3.04
137.18
4.03
true
false
false
common_voice_en_19933652-common_voice_en_19933655
78
A wide array of topics on business ethics are discussed in the responsa literature.
8.06
8.06
2.93
14.29
3.713
3.131
3.515
3.881
3.633
A wide array of topics on business ethics are discussed in the response of literature.
4.8
137.13
4.08
true
false
false
common_voice_en_19934187-common_voice_en_19934317
79
The team colors were teal, red and black.
8.02
8.02
2.099
0
3.702
2.673
3.078
3.645
3.433
The team colors were teal, red, and black.
2.88
137.09
4.11
true
false
false
common_voice_en_19949855-common_voice_en_19949857
80
In Japan, he is currently under the Cutting Edge label.
7.98
7.98
2.225
0
3.27
3.01
3.34
3.856
3.55
In Japan he is currently under the cutting edge label.
3.04
137.05
4.15
true
false
false
common_voice_en_19961449-common_voice_en_19961451
81
Saba is known as "The Unspoiled Queen" of the Caribbean.
7.94
7.94
2.845
0
4.48
3.202
3.441
4.14
4.283
Saba is known as the unspoiled queen of the Caribbean.
4.48
137.01
4.2
true
false
false
common_voice_en_19961449-common_voice_en_19961453
82
The remaining singles failed to hit the dance chart.
7.9
7.9
2.035
0
4.27
3.238
3.517
4.062
3.584
The remaining singles failed to hit the dance chart,
2.72
136.97
4.23
true
false
false
common_voice_en_19967495-common_voice_en_19967497
83
Many reviewers compared its style to that of the "Mario Kart" series.
7.85
7.85
2.795
0
3.565
3.355
3.6
4.141
4.066
Many reviewers compared its style to that of the Mario Kart series.
4.32
136.92
4.28
true
false
false
common_voice_en_19972684-common_voice_en_19972807
84
A few content management implementations cover all these issues.
7.81
7.81
2.545
0
4.015
3.175
3.516
3.979
2.926
A few content management implementations cover all these issues.
3.68
136.88
4.32
true
false
false
common_voice_en_19974164-common_voice_en_19974166
85
That same year, a civilian executive version, Disponent, saw the light of day.
7.76
7.76
2.99
0
3.354
3.001
3.367
3.875
4.102
That same year, a civilian executive version, Disponent, saw the light of day.
5.12
136.83
4.37
true
false
false
common_voice_en_19977527-common_voice_en_19977526
86
The town of Norwich is located east of the center of the county.
7.72
7.72
2.545
0
3.328
3.013
3.399
3.902
3.651
The town of Norwich is located east of the center of the county.
3.68
136.79
4.41
true
false
false
common_voice_en_19979514-common_voice_en_19979516
87
Rayford starts to read his wife's Bible.
7.67
7.67
1.648
0
3.489
2.791
3.017
4.055
3.217
Rayford starts to read his wife's Bible.
2.24
136.75
4.46
true
false
false
common_voice_en_19984445-common_voice_en_19984446
88
The children of Gowanda had school that day.
7.64
7.64
2.034
12.5
4.138
3.298
3.568
4.085
3.46
The children of Gawanda had school that day.
2.72
136.71
4.49
true
false
false
common_voice_en_20030172-common_voice_en_20030173
89
Flood gates were erected and drainage systems improved.
7.6
7.6
2.482
25
3.936
3.1
3.464
3.863
3.662
Floodgates were erected and drainage systems improved.
3.52
136.67
4.54
true
false
false
common_voice_en_20065698-common_voice_en_20065701
90
In some cases, the entire subject has been dropped from the merit badge roster.
7.55
7.55
2.795
0
4.085
2.957
3.33
3.775
3.834
In some cases the entire subject has been dropped from the merit badge roster.
4.32
136.62
4.59
true
false
false
common_voice_en_20066900-common_voice_en_20066899
91
The community was also once home to two distilleries and a tannery.
7.5
7.5
2.545
0
4.082
3.2
3.504
3.978
3.855
The community was also once home to two distilleries and a tannery,
3.68
136.57
4.64
true
false
false
common_voice_en_20214898-common_voice_en_20214908
92
There are three colour combinations that are used on several flags in certain regions.
7.44
7.44
3.045
7.14
4.415
3.239
3.535
4.028
3.496
There are three color combinations that are used on several flags in certain regions,
5.44
136.52
4.69
true
false
false
common_voice_en_20252760-common_voice_en_20252766
93
Numerous ships were queried and boarded to verify their cargo manifests.
7.4
7.4
2.99
0
4.529
3.175
3.465
4.073
3.674
Numerous ships were queried and boarded to verify their cargo manifests.
5.12
136.47
4.73
true
false
false
common_voice_en_20273907-common_voice_en_20273911
94
Kaufman Lake is also a popular full day hiking destination.
7.35
7.35
2.545
20
4.413
3.256
3.501
4.099
3.651
Kaufman Lake is also a popular full-day hiking destination.
3.68
136.42
4.78
true
false
false
common_voice_en_20303932-common_voice_en_20303930
95
Byron was traditionally a rural area of dairy and vegetable farms.
7.3
7.3
2.675
0
2.652
2.624
3.421
2.957
3.631
Byron was traditionally a rural area of dairy and vegetable farms.
4
136.37
4.83
true
false
false
common_voice_en_20303932-common_voice_en_20303934
96
The society of the day held that reward without labour was misguided.
7.26
7.26
2.675
8.33
3.357
2.999
3.298
3.912
3.911
The society of the day held that reward without labor was misguided.
4
136.33
4.88
true
false
false
common_voice_en_20317297-common_voice_en_20317298
97
The suburb is distinctly but unofficially split into the north and the south.
7.21
7.21
2.959
0
4.374
2.882
3.198
3.868
3.781
The suburb is distinctly but unofficially split into the north and the south.
4.96
136.28
4.93
true
false
false
common_voice_en_20317297-common_voice_en_20317300
98
Some of the grey ash was falling off the circular edge.
7.16
7.16
2.418
9.09
3.845
3.278
3.507
4.106
3.782
Some of the gray ash was falling off the circular edge.
3.36
136.24
4.97
true
false
false
common_voice_en_203357-common_voice_en_203358
99
A few languages such as Haskell implement them as linked lists instead.
7.12
7.12
2.845
0
3.642
2.913
3.367
3.589
3.618
A few languages such as Haskell implement them as linked lists instead.
4.48
136.19
5.01
true
false
false
common_voice_en_20337838-common_voice_en_20337841
100
Two physicists pondered their revised quantum field theory together.
7.07
7.07
2.609
0
4.361
3.208
3.478
4.03
3.714
two physicists pondered their revised quantum field theory together.
3.84
136.14
5.06
true
false
false
common_voice_en_2072999-common_voice_en_2073000
End of preview. Expand in Data Studio

VoxCPM2-Turbo benchmark: 200 requests at once

200 English sentences sent to VoxCPM2 at the same moment, served by the VoxCPM2-Turbo engine on one NVIDIA A10. Four ways of choosing the voice, each in the default and the low-latency mode; every output can be played here next to its latency and quality scores.

At a glance

Model openbmb/VoxCPM2
Engine VoxCPM2-Turbo, main (4c61885)
GPU 1× NVIDIA A10 (24 GB, 150 W), torch 2.10 + CUDA 12.8
Load one burst of 200 requests, after a warm-up burst of 200 other lines
Sentences 200 lines of Seed-TTS-eval test-EN (random.Random(0).sample of 1,088); cloning uses each line's own prompt recording (179 speakers)
Modes default: the original algorithm (10 diffusion steps, CFG 2.0) with INT8 W8A8 + SmoothQuant 0.7; low_latency: the same plus opt-in, training-free settings that send the first audio sooner
Size 2.55 GB: 13 parquet splits and 2,800 WAV files
Related VoxCPM2-Turbo-Voice-Cloning-200-Requests (when the first word is heard), VoxCPM2-Turbo-Listening-Samples
Subset What each request gets Splits
auto_voice the text only; the model picks a voice low_latency, default, low_latency_previous
voice_clone the text + its prompt recording and transcript, sent with every request low_latency, default, original_voxcpm2, ground_truth
registered_voice the text + the name of a voice registered before the burst (recording only) low_latency, default, low_latency_previous
registered_clone the text + the name of a voice registered before the burst (recording and transcript: the same cloning as voice_clone) low_latency, default, low_latency_k4

original_voxcpm2: the original voxcpm 2.0.3, one request at a time (quality reference). ground_truth: the test set's human recordings. low_latency_previous / low_latency_k4: earlier versions of the low-latency mode.

Latency

Subset Mode TTFA mean p50 p90 max Playable without stalls, mean Request done, mean / max
auto_voice low-latency 7.0 ms 7.0 ms 10.8 ms 12.1 ms 42.7 ms 2.55 / 3.21 s
auto_voice default 188.3 ms 184.4 ms 275.2 ms 295.3 ms 727.4 ms 4.19 / 5.23 s
voice_clone low-latency 658.6 ms 690.5 ms 1028.9 ms 1037.6 ms 1080.1 ms 3.90 / 4.88 s
voice_clone default 701.1 ms 736.7 ms 1092.5 ms 1101.6 ms 1599.1 ms 5.52 / 6.76 s
registered_voice low-latency 7.8 ms 7.8 ms 11.5 ms 12.5 ms 49.9 ms 2.68 / 3.38 s
registered_voice default 188.7 ms 184.6 ms 274.6 ms 293.8 ms 731.9 ms 4.22 / 5.33 s
registered_clone low-latency 8.8 ms 8.8 ms 13.0 ms 14.3 ms 226.1 ms 3.18 / 4.02 s
registered_clone default 424.8 ms 416.3 ms 604.5 ms 659.9 ms 1160.2 ms 5.08 / 6.34 s
  • TTFA: from a request's submission to its first 160 ms audio chunk. In the low-latency mode of auto_voice, registered_voice and registered_clone that chunk is a cached, quiet start, so the voice begins later.
  • Playable without stalls: the earliest start from which the stream plays to the end without waiting.
  • Request done: time to the last chunk.
  • voice_clone is slow to start because every request's prompt recording must be processed first; registering the voice in advance (registered_clone) gives the same cloning with first audio after 8.8 ms.

Quality

Subset Mode WER SIM UTMOS DNSMOS
auto_voice low-latency 1.71 % - 3.84 3.08
auto_voice default 3.02 % - 3.89 3.05
voice_clone low-latency 1.47 % 75.2 % 3.82 3.11
voice_clone default 1.58 % 75.3 % 3.83 3.11
voice_clone original VoxCPM2, 1 at a time 1.60 % 75.5 % 3.85 3.12
voice_clone human recordings 2.53 % 73.7 % 3.61 3.11
registered_voice low-latency 1.59 % 67.8 % 3.81 3.07
registered_voice default 1.64 % 64.9 % 3.84 3.05
registered_clone low-latency 1.77 % 74.9 % 3.85 3.07
registered_clone default 1.92 % 75.3 % 3.82 3.12
  • WER: word error rate, official Seed-TTS-eval scoring (Whisper large-v3); lower is better.
  • SIM: speaker similarity to the prompt recording, official Seed-TTS-eval scoring (WavLM-large); higher is better.
  • UTMOS (UTMOS22 strong) and DNSMOS (P.835 overall): predicted naturalness and quality, 1-5; higher is better. Cloning copies the prompts' home-microphone conditions (Common Voice), so cloning scores sit below auto_voice.
  • Paired bootstrap over the same 200 lines (95 %): low-latency vs default is within noise for voice_clone; it lowers auto_voice WER by 1.31 points (-2.74 … -0.23) and raises registered_voice SIM by 2.89 points (+1.86 … +3.90).
  • One default auto_voice output babbled; no other run in these tables babbled or hit the length limit.

Contents

Path What it is
<subset>/<split>-*.parquet one row per request, with audio players next to the scores (Dataset Viewer)
wav/<subset>/<split>/NNN.wav every output as a WAV; NNN is the request number (001 = submitted first), the same in every folder
wav/prompt/NNN.wav the voice recording used for line NNN
per_request.csv the same columns without audio, plus subset and split
scripts/ generation and scoring scripts

Main columns: request, text, audio (48 kHz; ground truth 16 kHz), prompt_audio, prompt_text, ttfa_ms, playable_ms, done_s, wer_pct, transcript (Whisper), sim, utmos, dnsmos_*, memo_hit (first chunk from the cache), hit_limit, babble, utt (Seed-TTS-eval line id).

Load 

from datasets import load_dataset
ds = load_dataset("VoxCPM-Kernel/VoxCPM2-Turbo-Benchmark-200-Requests", "registered_clone", split="low_latency")
print(ds[0]["text"], ds[0]["ttfa_ms"], ds[0]["wer_pct"])

License

The generated audio follows VoxCPM2 (Apache-2.0); the prompt and human recordings are Mozilla Common Voice clips (CC0), as packaged by Seed-TTS-eval.
Downloads last month
692
Number of rows:
1,800
Total file size:
2.55 GB