Info: Zenodo’s user support line is staffed on regular business days between Dec 23 and Jan 5. Response times may be slightly longer than normal.

There is a newer version of the record available.

Published May 11, 2022 | Version 1.1
Software Open

Lahjoita puhetta baseline Kaldi ASR model

  • 1. Aalto University

Description

Lahjoita puhetta baseline speech recognition model, built with the Kaldi toolkit. Trained on 1600 hours of Finnish speech. Described in more detail in the paper "Lahjoita puhetta: a large-scale corpus of spoken Finnish with some benchmarks". For details and instructions, see the Github page.

This upload includes the acoustic model in "lp_baseline_1600h.zip", the i-vector extractor in "extractor.zip", word-based decoding graph in "graph_word_lp_web_dsp.zip", and subword-based decoding graph in "graph_morfessor_lp_web_dsp.zip".

Files

graph_morfessor_lp_web_dsp.zip

Files (829.1 MB)

Name Size Download all
md5:b1ec09aa2879ec5acfdeba3c495ae48e
829.1 MB Preview Download