Hugging Face's logo

instinct
non-profit
Activity Feed Request to join this org

AI & ML interests

None defined yet.

Recent Activity
instinct1912  updated a dataset 3 days ago
instinct-org/multilingual-tts-voice-dataset
instinct1912  updated a collection 4 days ago
Speech Sources — Unchunked
instinct1912  updated a collection 4 days ago
TTS — Current Training Targets
View all activity

instinct-org 's collections 16

STT — Current Training Targets
Latest published speech-restored leaf for each of the 14 accepted STT source families.
STT — Validated Upstream Inputs
The 14 upstream repositories used by the completed STT training run. These are not final immutable training releases.
TTS — Voice-Clone Training Targets
Packaged non-raw reference/target pair shards for public voice-clone training.
TTS Train Clone Pairs - Raw Pair Plans
Raw TTS speaker-reference/target clone-pair metadata plans split by source dataset.
Uzbek/Russian Audio Datasets - Speech Restorised
Speech-restorised audio variants, including restored 48 kHz chunks.
TTS — Raw Pair Plans (Not Train-Ready)
Raw clone-pair planning metadata. These repositories are inputs to consensus and packaging, not training releases.
STT — Aligned Upstream Candidates
Restored or aligned STT repositories that are not in the validated 14-repository upstream set.
TTS Train Clone Pairs - Tokenized
Tokenized TTS speaker-reference/target clone-pair training shards split by source dataset.
STT — Current Training Targets
Latest published speech-restored leaf for each of the 14 accepted STT source families.
TTS — Raw Pair Plans (Not Train-Ready)
Raw clone-pair planning metadata. These repositories are inputs to consensus and packaging, not training releases.
STT — Validated Upstream Inputs
The 14 upstream repositories used by the completed STT training run. These are not final immutable training releases.
TTS — Voice-Clone Training Targets
Packaged non-raw reference/target pair shards for public voice-clone training.
STT — Aligned Upstream Candidates
Restored or aligned STT repositories that are not in the validated 14-repository upstream set.
TTS Train Clone Pairs - Raw Pair Plans
Raw TTS speaker-reference/target clone-pair metadata plans split by source dataset.
TTS Train Clone Pairs - Tokenized
Tokenized TTS speaker-reference/target clone-pair training shards split by source dataset.
Uzbek/Russian Audio Datasets - Speech Restorised
Speech-restorised audio variants, including restored 48 kHz chunks.