daniilrobnikov / sherpa

Speech-to-text server framework with next-gen Kaldi

sherpa

Usage with kaldi-gstreamer-server and Estonian model

# go to the sherpa root directory
python ./sherpa/bin/streaming_server.py \
  --port=6006 \
  --encoder-model="path/to/icefall_pruned_transducer_stateless7_streaming_et/encoder_jit_trace.pt" \
  --decoder-model="path/to/icefall_pruned_transducer_stateless7_streaming_et/decoder_jit_trace.pt" \
  --joiner-model="path/to/icefall_pruned_transducer_stateless7_streaming_et/joiner_jit_trace.pt" \
  --tokens="path/to/icefall_pruned_transducer_stateless7_streaming_et/data/lang_bpe_1000/tokens.txt"

# go to the kaldi-gstreamer-server root directory
python kaldigstserver/client.py -r=32000 test/data/lause2.raw

sherpa

sherpa is an open-source speech-text-text inference framework using PyTorch, focusing exclusively on end-to-end (E2E) models, namely transducer- and CTC-based models. It provides both C++ and Python APIs.

This project focuses on deployment, i.e., using pre-trained models to transcribe speech. If you are interested in how to train or fine-tune your own models, please refer to icefall.

We also have other similar projects that don't depend on PyTorch:

• sherpa-onnx
• sherpa-ncnn

sherpa-onnx and sherpa-ncnn also support iOS, Android and embedded systems.

Installation and Usage

Please refer to the documentation at https://k2-fsa.github.io/sherpa/

Try it in your browser

Try sherpa from within your browser without installing anything: https://huggingface.co/spaces/k2-fsa/automatic-speech-recognition