Skip to main content

kaldi-io for Tensorflow

Project description


A python package: provide a custom tensorflow dataset for kaldi io

Python is its wrapper, C++ is its backend implemention. It depends on two things:

Through [kaldi-io lib](https://github.com/open-speech/kaldi-io.git), it is able to:
- direct read from kaldi rspecifier(scp, ark, in text or binary, just as [kaldi](https://github.com/kaldi-asr/kaldi))
- support multiple feature transforms:
- delta
- cmvn
- splice
- sampling
- compute fast: kaldi Matrix|Vector with blas math lib is used

Through [tensorflow dataset](https://www.tensorflow.org/guide/datasets), it is able to:
- shuffling
- batching at frame or utt level
- bucketing with input sequence lengths
- and all other tensorflow native dataset manipulations and features (parellel, prefetch, ..)

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

tf_kaldi_io-0.1.0.tar.gz (13.6 kB view hashes)

Uploaded Source

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page