Endpoint Anticipation for Low-Latency Spoken Dialogue
Paper • 2606.13450 • Published
Pretrained checkpoint for the anticipation model from:
Endpoint Anticipation for Low-Latency Spoken Dialogue
Sathvik Udupa, Shinji Watanabe, Petr Schwarz, Jan Cernocky
Interspeech 2026 — arXiv:2606.13450
| File | Description |
|---|---|
best_val_acc.pt |
Model checkpoint (fc960 transformer, Mimi features, 12.5 Hz, data_mix_4) |
config.yaml |
Model and training config used to produce this checkpoint |
See the full codebase and instructions at:
github.com/bloodraven66/EndpointAnticipation