Memory efficient transducer loss computation
-
Updated
Jun 10, 2022 - CMake
Memory efficient transducer loss computation
A tutorial on how to train RNN-T from scratch with Whisper encoder
Streaming on-device speech recognition for Android — NEON-accelerated, encrypted FastConformer (32M params), ~150 ms latency, no cloud. Powered by the VoxRT runtime.
Streaming on-device speech recognition for iOS — NEON-accelerated, encrypted FastConformer (32M params), RTF 0.08–0.10 on iPhone 13 Pro Max. Built on the VoxRT custom Rust inference runtime. SwiftPM distribution.
Streaming 가능한 RNN Transducer 모델을 PyTorch Lightning으로 구현해본다.
PyTorch Implementation of RNN-Transducer
ASR (Speech to Text) in Bengali
Unofficial, From-scratch PyTorch replication of the Conformer paper (Gulati et al., 2020) — encoder, RNN-T decoder, training loop, and NeMo weight validation. Built block by block with documented maths.
Add a description, image, and links to the rnn-t topic page so that developers can more easily learn about it.
To associate your repository with the rnn-t topic, visit your repo's landing page and select "manage topics."