nttcslab-sp / torchain
WIP: pytorch FFI wrapper for Kaldi chain loss (a.k.a. Lattice Free MMI)
☆20Updated 5 years ago
Alternatives and similar repositories for torchain:
Users that are interested in torchain are comparing it to the libraries listed below
- The repo contains our code of ``Semantic Mask for Transformer based End-to-End Speech Recognition"☆37Updated 4 years ago
- Recurrent Neural Aligner☆49Updated 4 years ago
- mWER loss implementation in tensorflow☆31Updated 4 years ago
- streaming attention networks for end-to-end automatic speech recognition☆55Updated 4 years ago
- ☆41Updated 6 years ago
- ☆75Updated 2 years ago
- Kaldi extended by Kaituo XU with new features in nnet1.☆12Updated 6 years ago
- ☆27Updated 3 years ago
- ☆48Updated 4 years ago
- ESPnet extensions for semi-supervised end-to-end speech recognition. See also https://github.com/ShigekiKarita/espnet-semi-supervised/tre…☆38Updated 4 years ago
- it's a train acoustics model code lib☆26Updated 4 years ago
- ☆16Updated 5 years ago
- Memory efficient transducer loss computation☆68Updated 2 years ago
- An implementation of rnn transducer for sequence labeling problem☆22Updated 6 years ago
- ☆16Updated 5 years ago
- ☆22Updated 7 years ago
- A pytorch wrapper for LF-MMI training and parallel training in Kaldi☆73Updated 2 years ago
- ☆67Updated 2 years ago
- Custom decoders for Kaldi☆80Updated 5 years ago
- Script to simulate room impulse responses☆14Updated 8 years ago
- Python wrapper for OpenFST and its extensions from Kaldi. Also support reading/writing ark/scp files☆48Updated 6 months ago
- Lattice combination algorithm to combine inaccurate transcripts with hypothesis lattices☆16Updated 9 months ago
- Script for converting kaldi GMM/HMM models to HTK format☆11Updated 6 months ago
- C++ implementation of End to End TTS which combines both Tacatron2 and LPCNET Vocoder.☆32Updated 5 years ago
- Python wrapper for kaldi's arpa2fst☆38Updated last month
- ☆16Updated 2 years ago
- Fork of the official kaldi.☆22Updated 2 years ago
- ☆14Updated 6 years ago