sherpa-onnx
sherpa-onnx copied to clipboard
k2-fsa
Speech-to-text, text-to-speech, speaker diarization, speech enhancement, source separation, and VAD using next-gen Kaldi with onnxruntime without Internet connection. Support embedded systems, Android...
Hi, I am installing sherpa-onnx from src with `SHERPA_ONNX_CMAKE_ARGS="-DBUILD_SHARED_LIBS=ON -DSHERPA_ONNX_ENABLE_GPU=ON"` The build commands are: ``` python3 setup.py bdist_wheel pip install ./dist/sherpa_onnx-*.whl ``` or ``` pip install . ``` (both ways...
See https://onnxruntime.ai/docs/execution-providers/DirectML-ExecutionProvider.html and https://learn.microsoft.com/en-us/windows/ai/directml/dml-intro and https://www.nuget.org/packages/Microsoft.ML.OnnxRuntime.DirectML It would be nice to also support DirectML
Hi, In icefall, there are multiple decoding methods available, eg. greedy_search, beam_search, modified_beam_search, fast_beam_search, fast_beam_search_nbest. There are some other decoding methods for LM as well (modified_beam_search_lm_shallow_fusion, modified_beam_search_LODR, modified_beam_search_lm_rescore, modified_beam_search_lm_rescore_LODR). But...
When a `buid-ios.sh` trying to create headers for build/simulator/lib it encounters an error like: `fatal error: /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/lipo: can't open input file: build/simulator_arm64/lib/libkaldi-native-fbank-core.a (No such file or directory)` This happens because...
This creates a problem when generating headers for sherpa-onnx.xcframework. Solution: Set it in ios.toolchain.cmake like this set(CMAKE_XCODE_ATTRIBUTE_PRODUCT_BUNDLE_IDENTIFIER "Some bundle ID") set(CMAKE_XCODE_ATTRIBUTE_DEVELOPMENT_TEAM "Your team ID")
Here you use minimal buffer size, sometimes it can be 80 ms. If recognizer is slow and can't manage in 80 ms samples can get lost and accuracy will degrade....
I might be mistaken, but it seems like the segments is only push out when no speech is detected. However the audio might be cut off say in between a...
fix issue in upload and offlind-record pages of python streaming server
I found that the fast-beam-search decoding is currently not supported in sherpa-onnx. Is this activity is planned for future? If yes, when can this be expected (timeline)? In specific, do...
I am using zipformer with online model (streaming mode) using C APIs on android platform, the ASR functionality is working OK. In my code during constructor call, I am initializing...
