DualTurn: Learning Turn-Taking from Dual-Channel Generative Speech Pretraining
Shangeth Rajaa
shangeth
AI & ML interests
Speech Representation Learning, Multi-Modal LLM, Spoken Dialogue Systems, Speech Synthesis
Recent Activity
updated a model 28 days ago
shangeth/Wren-ASR-0.5B-multi updated a Space about 1 month ago
shangeth/Wren-ASR-0.5B-multi-demo published a Space about 1 month ago
shangeth/Wren-ASR-0.5B-multi-demoOrganizations
Wren: A Family of Small Open-Weight Models for Unified Speech-Text Modelling
-
shangeth/Wren-TTS-0.5B-multi-expressive
Text-to-Speech • 0.5B • Updated • 33 -
Wren-TTS-0.5B-multi-expressive
🎭1Expressive multilingual voice-cloning TTS — 23 style tags
-
shangeth/Wren-TTS-0.5B-multi
Text-to-Speech • 0.5B • Updated • 33 • 1 -
Wren-TTS-0.5B-multi
🐦Multilingual voice-cloning TTS — 8 languages
SpeechLLM: Multi-Modal LLM for Speech Understanding
DualTurn: Learning Turn-Taking from Dual-Channel Generative Speech Pretraining
Wren: A Family of Small Open-Weight Models for Unified Speech-Text Modelling
-
shangeth/Wren-TTS-0.5B-multi-expressive
Text-to-Speech • 0.5B • Updated • 33 -
Wren-TTS-0.5B-multi-expressive
🎭1Expressive multilingual voice-cloning TTS — 23 style tags
-
shangeth/Wren-TTS-0.5B-multi
Text-to-Speech • 0.5B • Updated • 33 • 1 -
Wren-TTS-0.5B-multi
🐦Multilingual voice-cloning TTS — 8 languages
Semantic Turn Taking LLMs
SpeechLLM: Multi-Modal LLM for Speech Understanding