VoXtream2: Full-stream TTS with dynamic speaking rate control
Paper • 2603.13518 • Published • 1
Dataset Viewer
The dataset viewer is not available for this split.
Cannot load the dataset split (in streaming mode) to extract the first rows.
Error code: StreamingRowsError
Exception: CastError
Message: Couldn't cast
tokens: list<item: int64>
child 0, item: int64
silence_db: double
min_count: int64
min_silent: double
calibration_entries: int64
frames_seen: int64
frames_called_silent: int64
vocab_sizes: list<item: int64>
child 0, item: int64
mimi_tps: double
plan_dir: string
num_frames: int64
spk_emb_dim: int64
sample_rate: int64
num_phonemes: int64
num_entries: int64
phone_streams: int64
to
{'num_entries': Value('int64'), 'num_frames': Value('int64'), 'mimi_tps': Value('float64'), 'sample_rate': Value('int64'), 'num_phonemes': Value('int64'), 'phone_streams': Value('int64'), 'vocab_sizes': List(Value('int64')), 'spk_emb_dim': Value('int64'), 'plan_dir': Value('string')}
because column names don't match
Traceback: Traceback (most recent call last):
File "/src/services/worker/src/worker/utils.py", line 149, in get_rows_or_raise
return get_rows(
dataset=dataset,
...<4 lines>...
column_names=column_names,
)
File "/src/libs/libcommon/src/libcommon/utils.py", line 272, in decorator
return func(*args, **kwargs)
File "/src/services/worker/src/worker/utils.py", line 129, in get_rows
rows_plus_one = list(itertools.islice(safe_iter(ds, dataset=dataset), rows_max_number + 1))
File "/src/services/worker/src/worker/utils.py", line 544, in safe_iter
yield from ds.decode(False) if ds.features else ds
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 3052, in __iter__
for key, example in ex_iterable:
^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2559, in __iter__
for key, pa_table in self._iter_arrow():
~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2584, in _iter_arrow
for key, pa_table in self.ex_iterable._iter_arrow():
~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 558, in _iter_arrow
for key, pa_table in iterator:
^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 440, in _iter_arrow
for key, pa_table in self.generate_tables_fn(**gen_kwags):
~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 361, in _generate_tables
self._cast_table(pa_table, json_field_paths=json_field_paths),
~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 132, in _cast_table
pa_table = table_cast(pa_table, self.info.features.arrow_schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2381, in table_cast
return cast_table_to_schema(table, schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2309, in cast_table_to_schema
raise CastError(
...<3 lines>...
)
datasets.table.CastError: Couldn't cast
tokens: list<item: int64>
child 0, item: int64
silence_db: double
min_count: int64
min_silent: double
calibration_entries: int64
frames_seen: int64
frames_called_silent: int64
vocab_sizes: list<item: int64>
child 0, item: int64
mimi_tps: double
plan_dir: string
num_frames: int64
spk_emb_dim: int64
sample_rate: int64
num_phonemes: int64
num_entries: int64
phone_streams: int64
to
{'num_entries': Value('int64'), 'num_frames': Value('int64'), 'mimi_tps': Value('float64'), 'sample_rate': Value('int64'), 'num_phonemes': Value('int64'), 'phone_streams': Value('int64'), 'vocab_sizes': List(Value('int64')), 'spk_emb_dim': Value('int64'), 'plan_dir': Value('string')}
because column names don't matchNeed help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
Dataset Card for VoXtream2.5 training dataset
This repository contains a training dataset for VoXtream2.5 TTS model.
The dataset contains 54k hours of speech:
- 39k hours of speech from the English part of Emilia dataset (1.91M entries).
- 14.5k hours of speech from a 15.6k-hour subset of HiFiTTS2 dataset (693k entries).
All entries are 80 seconds long (1000 Mimi frames). We concatenated utterances of the same speaker, ending each entry at a word boundary, and padded it with silence. Emilia speakers are clustered with speaker embeddings, HiFiTTS2 entries follow its speaker and session labels. Sampling rate: 24kHz.
Description
- mimi_codes, mimi_codes_16 - Tokens extracted by the Mimi audio codec: codebooks 0-15 and 16-31,
uint16 [entries, 16, 1000]. - phone_tokens - IPA phoneme tokens extracted with espeak-ng phonemizer, in five streams (root, stress, length, following punctuation, end of word):
valuesuint8 [phonemes, 5]andoffsetsof every entry. - frame_state - Alignment of phoneme tokens to Mimi audio frames extracted by ClapIPA forced aligner, as the phoneme shift state of every frame:
uint8 [entries, 1000]. - words, word_* - Word borders: first phoneme, first and last frame, pause before and after (frames) and flags (followed by punctuation, end of a concatenated utterance), with
offsetsof every entry. - num_speech_frames - Frames before the silence padding.
- spk_emb - Speaker embedding of the entry: median of the embeddings of its utterances, each from the first 3 seconds of audio, extracted by ReDimNet2 B3 (vb2+vox2+cnc2 lm) model.
- silence_tokens - Tokens of the first Mimi codebook that mark silent frames.
- cluster_id - Entry id in the crop plan of voxtream2.5-metadata, which lists the source utterances of every entry.
- meta - Number of entries and frames, frame rate and vocabulary sizes.
The phoneme vocabulary is phoneme_vocab.json of the model.
Sources
Usage
To download the dataset, use the following code:
from huggingface_hub import snapshot_download
local_dir = snapshot_download('herimor/voxtream2.5-train', repo_type='dataset')
Clone our repo and follow the instructions in the README file.
Citation
@inproceedings{torgashov2026voxtream,
title={Vo{X}tream: Full-Stream Text-to-Speech with Extremely Low Latency},
author={Torgashov, Nikita and Henter, Gustav Eje and Skantze, Gabriel},
booktitle={Proc. IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
year={2026},
note={to appear},
url={https://arxiv.org/abs/2509.15969}
}
@article{torgashov2026voxtream2,
author = {Torgashov, Nikita and Henter, Gustav Eje and Skantze, Gabriel},
title = {Vo{X}tream2: Full-stream TTS with dynamic speaking rate control},
journal = {arXiv:2603.13518},
year = {2026}
}
- Downloads last month
- -