Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Phonetisaurus G2P
| Date | Stars |
|---|---|
| 2026-07-24 | 517 |
| 2026-07-25 | 517 |
| 2026-07-28 | 517 |
| 2026-07-30 | 517 |
| 2026-08-06 | 517 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
## Phonetisaurus G2P ## [](https://travis-ci.org/AdolfVonKleist/Phonetisaurus) This repository contains scripts suitable for training, evaluating and using grapheme-to-phoneme models for speech recognition using the OpenFst framework. The current build requires OpenFst version 1.6.0 or later, and the examples below use version 1.7.2. The repository includes C++ binaries suitable for training, compiling, and evaluating G2P models. It also some simple python bindings which may be used to extract individual multigram scores, alignments, and to dump the raw lattices in .fst format for each word. The python scripts and bindings were tested most recently with python v3.8.5. Standalone distributions related to previous INTERSPEECH papers, as well as the complete, exported final version of the old google-code repository are available via ```git-lfs``` in a separate repository: * https://github.com/AdolfVonKleist/phonetisaurus-downloads #### Contact: #### * [email protected] #### Scratch Build for OpenFst v1.7.2 and Ubuntu 20.04 #### This build was tested via AWS EC2 with a fresh Ubuntu 20.04 base, and m4.large instance. ``` $ sudo apt-get update # Basics $ sudo apt-get install git g++ autoconf-archive make libtool # Python bindings $ sudo apt-get install python-setuptools python-dev # mitlm (to build a quick play model) $ sudo apt-get install gfortran ``` Create a work directory of your choice: ``` $ mkdir g2p $ cd g2p/ ``` Next grab and install OpenFst-1.7.2: ``` $ wget http://www.openfst.org/twiki/pub/FST/FstDownload/openfst-1.7.2.tar.gz $ tar -xvzf openfst-1.7.2.tar.gz $ cd openfst-1.7.2 # Minimal configure, compatible with current defaults for Kaldi $ ./configure --enable-static --enable-shared --enable-far --enable-ngram-fsts $ make -j # Now wait a while... $ sudo make install # Extend your LD_LIBRARY_PATH .bashrc (assumes OpenFst installed to default location): $ echo 'export LD_LIBRARY_PATH=${LD_LIBRARY_PATH}:/usr/local/lib:/usr/local/lib/fst' \ >> ~/.bashrc $ source ~/.bashrc $ cd .. ``` Checkout the latest Phonetisaurus from master and compile without bindings: ``` $ git clone https://github.com/AdolfVonKleist/Phonetisaurus.git $ cd Phonetisaurus # if OpenFst is installed in the default location: $ ./configure # if OpenFst is installed in a special location: $ ./configure \ --with-openfst-includes=${OFST_PATH}/openfst-1.7.2/include \ --with-openfst-libs=${OFST_PATH}/openfst-1.7.2/lib $ make $ sudo make install $ cd .. ``` Checkout the latest Phonetisaurus from master and compile with python3 bindings: ``` $ git clone https://github.com/AdolfVonKleist/Phonetisaurus.git $ cd Phonetisaurus $ sudo pip3 install pybindgen # if OpenFst is installed in the default location: $ PYTHON=python3 ./configure --enable-python # if OpenFst is installed in a special location: $ PYTHON=python3 ./configure \ --with-openfst-includes=${OFST_PATH}/openfst-1.7.2/include \ --with-openfst-libs=${OFST_PATH}/openfst-1.7.2/lib \ --enable-python $ make $ sudo make install $ cd python $ cp ../.libs/Phonetisaurus.so . $ sudo python3 setup.py install $ cd ../.. ``` Grab and install mitlm to build a quick test model with the cmudict (5m): ``` $ git clone https://github.com/mitlm/mitlm.git $ cd mitlm/ $ ./autogen.sh $ make $ sudo make install $ cd .. ``` Grab a copy of the latest version of CMUdict and clean it up a bit: ``` $ mkdir example $ cd example $ wget https://raw.githubusercontent.com/cmusphinx/cmudict/master/cmudict.dict # Clean it up a bit and reformat: $ cat cmudict.dict \ | perl -pe 's/\([0-9]+\)//; s/\s+/ /g; s/^\s+//; s/\s+$//; @_ = split (/\s+/); $w = shift (@_); $_ = $w."\t".join (" ", @_)."\n";' \ > cmudict.formatted.dict ``` Train a complete model with default parameters using the wrapper script. NOTE: this assumes the tool was compiled wi
Excerpt of 11,096 characters
Read on GitHub
97
Giulio Paci
44
5
4
2
Jörg Thalheim · no
2
1
Jan "yenda" Trmal
1
Karel Vesely · Brno University of Technology · Czech Republic
1
1
1
1
1
1
1
Would you bet a product on this? Bounded 0–100 and slow moving.
matched fp:3121f24914a7e15e, topic:speech-recognition, topic:speech-to-text, readme:speech recognition