GitHub/rhasspy
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Compare commits

base: GitHub/rhasspy:v2.4
Branches Tags
GitHub/rhasspy:master GitHub/rhasspy:dependabot/npm_and_yarn/elliptic-6.5.3 GitHub/rhasspy:dependabot/npm_and_yarn/lodash-4.17.19 GitHub/rhasspy:fuzzyfix GitHub/rhasspy:mqtt-refactor GitHub/rhasspy:ipa
GitHub/rhasspy:v2.4.20 GitHub/rhasspy:v2.4.19 GitHub/rhasspy:v2.4.18 GitHub/rhasspy:v2.4.17 GitHub/rhasspy:2.4.16 GitHub/rhasspy:v2.4.15 GitHub/rhasspy:v2.4 GitHub/rhasspy:v2.0
..
compare: GitHub/rhasspy:ipa
Branches Tags
GitHub/rhasspy:dependabot/npm_and_yarn/elliptic-6.5.3 GitHub/rhasspy:dependabot/npm_and_yarn/lodash-4.17.19 GitHub/rhasspy:master GitHub/rhasspy:fuzzyfix GitHub/rhasspy:mqtt-refactor GitHub/rhasspy:ipa
GitHub/rhasspy:v2.4.20 GitHub/rhasspy:v2.4.19 GitHub/rhasspy:v2.4.18 GitHub/rhasspy:v2.4.17 GitHub/rhasspy:2.4.16 GitHub/rhasspy:v2.4.15 GitHub/rhasspy:v2.4 GitHub/rhasspy:v2.0

2 Commits
v2.4 .. ipa

Author SHA1 Message Date
Michael Hansen 6f4992e310 First pass at Dutch IPA 2019-05-19 20:52:43 -04:00
Michael Hansen 8245155fab Working for English 2019-05-19 15:12:15 -04:00
206 changed files with 9647 additions and 24692 deletions
Expand all files Collapse all files
.dockerignore
+4 -12
View File
@@ -1,26 +1,18 @@
.git/
.venv/
node_modules/
__pycache__/
test/
tools/
etc/test/
download/precise-engine/
download/kaldi/
opt/
etc/homeassistant/config/.storage
examples/typical/home-assistant/config/.storage
examples/typical-intent/home-assistant/config/.storage
examples/client-server/home-assistant/config/.storage
examples/mqtt-hermes/home-assistant/config/.storage
profiles/*/base_dictionary.txt
profiles/*/base_language_model.txt
profiles/*/acoustic_model/
profiles/*/g2p.fst
base_dictionary.txt
base_language_model.txt
acoustic_model/
profiles/en-kaldi/
profiles/en-zamia/
profiles/*/download/
profiles/*/download/
.gitignore
+1 -22
View File
@@ -4,7 +4,6 @@ __pycache__/
.venv/
.ipynb_checkpoints/
download/
build/
# QEMU
etc/qemu-arm-static
@@ -27,11 +26,6 @@ profiles/*/*.umdl
profiles/*/flair/
profiles/*/*.pb
profiles/*/*.pb.params
profiles/*/snowboy/
profiles/*/precise/
profiles/*/porcupine/
.doit.db
vocab.txt
adapt_config.json
tagged_sentences.md
@@ -47,11 +41,6 @@ grammars/
sentences/
record/
# Tools
tools/dist/
tools/etc/
opt/
# Third party
etc/*.tar.gz
etc/*.tar.xz
@@ -64,15 +53,5 @@ tts/
*.log
.HA_VERSION
# Tests
etc/test/pt/Google/
etc/test/pt/PT*
etc/test/pt/original/
etc/test/vi/2*
# Misc
workbench.xmi
# Debian
site/
debian/rhasspy-server*
workbench.xmi
LICENSE
-21
View File
@@ -1,21 +0,0 @@
MIT License
Copyright (c) 2019 Michael Hansen
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
Makefile
+1 -11
View File
@@ -22,7 +22,7 @@ docker-armhf:
-t synesthesiam/rhasspy-server:armhf
docker-aarch64:
docker build . -f docker/templates/dockerfiles/Dockerfile.prebuilt.alsa.all \
docker build . -f docker/templates/dockerfiles/Dockerfile.from-source.alsa.all \
--build-arg BUILD_ARCH=aarch64 \
--build-arg CPU_ARCH=arm64v8 \
--build-arg BUILD_FROM=arm64v8/ubuntu:bionic \
@@ -49,9 +49,6 @@ manifest:
web-dist:
yarn build
mkdir -p download
rm -f download/rhasspy-web-dist.tar.gz
tar -czf download/rhasspy-web-dist.tar.gz dist/
# -----------------------------------------------------------------------------
# Documentation
@@ -76,10 +73,3 @@ g2p: $(G2P_MODELS)
%/g2p.fst: %/base_dictionary.txt
./make-g2p.sh $< $@
# -----------------------------------------------------------------------------
# Testing
# -----------------------------------------------------------------------------
mypy:
mypy app.py rhasspy
README.md
+18 -21
View File
@@ -7,7 +7,7 @@ Rhasspy (pronounced RAH-SPEE) is an offline, [multilingual](#supported-languages
* [Hass.IO Add-On Repository](https://github.com/synesthesiam/hassio-addons)
* [Discussion](https://community.home-assistant.io/t/rhasspy-offline-voice-assistant-toolkit/60862)
Rhasspy transca voice commands into [JSON](https://json.org) events that can trigger actions in home automation software, like [Home Assistant automations](https://www.home-assistant.io/docs/automation/trigger/#event-trigger) or [Node-RED flows](https://rhasspy.readthedocs.io/en/latest/usage/#node-red). You define custom voice commands in a [profile](https://rhasspy.readthedocs.io/en/latest/profiles/) using a [specialized template syntax](https://rhasspy.readthedocs.io/en/latest/training/#sentencesini), and Rhasspy takes care of the rest.
Rhasspy transforms voice commands into [JSON](https://json.org) events that can trigger actions in home automation software, like [Home Assistant automations](https://www.home-assistant.io/docs/automation/trigger/#event-trigger) or [Node-RED flows](https://rhasspy.readthedocs.io/en/latest/usage/#node-red). You define custom voice commands in a [profile](profiles.md) using a [specialized template syntax](https://rhasspy.readthedocs.io/en/latest/training/#sentencesini), and Rhasspy takes care of the rest.
To run Rhasspy with the English (en) profile using Docker:
@@ -38,29 +38,26 @@ Rhasspy currently supports the following languages:
* Mandarin (`zh`)
* Vietnamese (`vi`)
* Portuguese (`pt`)
* Swedish (`sv`)
* Catalan (`ca`)
The table below summarizes language support across the various supporting technologies that Rhasspy uses:
| Category | Name | Offline? | en | de | es | fr | it | nl | ru | el | hi | zh | vi | pt | sv | ca |
| -------- | ------ | -------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- |
| **Wake Word** | [pocketsphinx](https://rhasspy.readthedocs.io/en/latest/wake-word/#pocketsphinx) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | | &#x2713; | | &#x2713; |
| | [porcupine](https://rhasspy.readthedocs.io/en/latest/wake-word.md#porcupine) | &#x2713; | &#x2713; | | | | | | | | | | | | | |
| | [snowboy](https://rhasspy.readthedocs.io/en/latest/wake-word/#snowboy) | *requires account* | &#x2713; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; |
| | [precise](https://rhasspy.readthedocs.io/en/latest/wake-word/#mycroft-precise) | &#x2713; | &#x2713; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; |
| **Speech to Text** | [pocketsphinx](https://rhasspy.readthedocs.io/en/latest/speech-to-text/#pocketsphinx) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | | &#x2713; | | &#x2713; |
| | [kaldi](https://rhasspy.readthedocs.io/en/latest/speech-to-text/#kaldi) | &#x2713; | | | | | | | | | | | &#x2713; | | &#x2713; | |
| **Intent Recognition** | [fsticuffs](https://rhasspy.readthedocs.io/en/latest/intent-recognition/#fsticuffs) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; |
| | [fuzzywuzzy](https://rhasspy.readthedocs.io/en/latest/intent-recognition/#fuzzywuzzy) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; |
| | [adapt](https://rhasspy.readthedocs.io/en/latest/intent-recognition/#mycroft-adapt) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; |
| | [flair](https://rhasspy.readthedocs.io/en/latest/intent-recognition/#flair) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | | &#x2713; | | | | | | &#x2713; | | &#x2713; |
| | [rasaNLU](https://rhasspy.readthedocs.io/en/latest/intent-recognition/#rasanlu) | *needs extra software* | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; |
| **Text to Speech** | [espeak](https://rhasspy.readthedocs.io/en/latest/text-to-speech/#espeak) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; |
| | [flite](https://rhasspy.readthedocs.io/en/latest/text-to-speech/#flite) | &#x2713; | &#x2713; | | | | | | | | &#x2713; | | | | | |
| | [picotts](https://rhasspy.readthedocs.io/en/latest/text-to-speech/#picotts) | &#x2713; | &#x2713; | | | | | | | | | | | | | |
| | [marytts](https://rhasspy.readthedocs.io/en/latest/text-to-speech/#marytts) | &#x2713; | &#x2713; | &#x2713; | | &#x2713; | &#x2713; | | &#x2713; | | | | | | | |
| | [wavenet](https://rhasspy.readthedocs.io/en/latest/text-to-speech/#google-wavenet) | | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | | &#x2713; | &#x2713; | | &#x2713; | &#x2713; | |
| Category | Name | Offline? | en | de | es | fr | it | nl | ru | el | hi | zh | vi | pt |
| -------- | ------ | -------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- |
| **Wake Word** | [pocketsphinx](https://rhasspy.readthedocs.io/en/latest/wake-word/#pocketsphinx) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | | |
| | [snowboy](https://rhasspy.readthedocs.io/en/latest/wake-word/#snowboy) | *requires account* | &#x2713; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; |
| | [precise](https://rhasspy.readthedocs.io/en/latest/wake-word/#mycroft-precise) | &#x2713; | &#x2713; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; |
| **Speech to Text** | [pocketsphinx](https://rhasspy.readthedocs.io/en/latest/speech-to-text/#pocketsphinx) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | | |
| | [kaldi](https://rhasspy.readthedocs.io/en/latest/speech-to-text/#kaldi) | &#x2713; | | | | | | | | | | | &#x2713; | &#x2713; |
| **Intent Recognition** | [fsticuffs](https://rhasspy.readthedocs.io/en/latest/intent-recognition/#fsticuffs) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; |
| | [fuzzywuzzy](https://rhasspy.readthedocs.io/en/latest/intent-recognition/#fuzzywuzzy) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; |
| | [adapt](https://rhasspy.readthedocs.io/en/latest/intent-recognition/#mycroft-adapt) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; |
| | [flair](https://rhasspy.readthedocs.io/en/latest/intent-recognition/#flair) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | | &#x2713; | | | | | | &#x2713; |
| | [rasaNLU](https://rhasspy.readthedocs.io/en/latest/intent-recognition/#rasanlu) | *needs extra software* | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; |
| **Text to Speech** | [espeak](https://rhasspy.readthedocs.io/en/latest/text-to-speech/#espeak) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; |
| | [flite](https://rhasspy.readthedocs.io/en/latest/text-to-speech/#flite) | &#x2713; | &#x2713; | | | | | | | | &#x2713; | | | |
| | [picotts](https://rhasspy.readthedocs.io/en/latest/text-to-speech/#picotts) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; |
| | [marytts](https://rhasspy.readthedocs.io/en/latest/text-to-speech/#marytts) | &#x2713; | &#x2713; | &#x2713; | | &#x2713; | &#x2713; | | &#x2713; | | | | | |
| | [wavenet](https://rhasspy.readthedocs.io/en/latest/text-to-speech/#google-wavenet) | | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | | &#x2713; | &#x2713; | | &#x2713; |
&bull; - yes, but requires training/customization
app.py
+304 -403
View File
File diff suppressed because it is too large Load Diff
bin/get-frequent-words.py
-2
View File
@@ -28,8 +28,6 @@ def main():
"hin": "hi",
"ell": "el",
"por": "pt",
"swe": "sv",
"cat": "ca",
}
for language in languages:
bin/get-phoneme-examples.py
+1 -1
View File
@@ -29,7 +29,7 @@ def main():
if len(line) == 0:
continue
parts = re.split(r"[\t ]+", line)
parts = re.split(r"[ ]+", line)
word = parts[0]
if "(" in word:
bin/guess-espeak.py
+1 -1
View File
@@ -34,7 +34,7 @@ def main():
if len(line) == 0:
continue
parts = re.split(r"[\t ]+", line)
parts = re.split(r"[ ]+", line)
word = parts[0].lower()
if ("(" in word) or (word in freq_phonemes):
bin/rhasspy
+5 -20
View File
@@ -1,29 +1,14 @@
#!/usr/bin/env bash
set -e
# Directory of *this* script
this_dir="$( cd "$( dirname "$0" )" && pwd )"
DIR="$( cd "$( dirname "$0" )" && pwd )"
venv="${this_dir}/../.venv"
if [[ ! -d "${venv}" ]]; then
echo "Missing virtual environment at ${venv}"
echo "Did you run create-venv.sh?"
exit 1
fi
source "${venv}/bin/activate"
# Force .venv/lib to be used
export LD_LIBRARY_PATH="${venv}/lib:${LD_LIBRARY_PATH}"
# Use local Kaldi
if [[ -d "${this_dir}/opt/kaldi" ]]; then
export KALDI_PREFIX="${this_dir}/opt"
fi
cd "$DIR/.."
source .venv/bin/activate
# Path to sphinxtrain tools
if [[ -d "/usr/lib/sphinxtrain" ]]; then
export PATH="/usr/lib/sphinxtrain:${PATH}"
export PATH="/usr/lib/sphinxtrain:$PATH"
fi
cd "${this_dir}/.." && python3 -m rhasspy "$@"
python3 -m rhasspy "$@"
create-venv.sh
+164 -221
View File
@@ -1,149 +1,16 @@
#!/usr/bin/env bash
this_dir="$( cd "$( dirname "$0" )" && pwd )"
# -----------------------------------------------------------------------------
# Command-line Arguments
# -----------------------------------------------------------------------------
. "${this_dir}/etc/shflags"
DEFINE_string 'venv' "${this_dir}/.venv" 'Path to create virtual environment'
DEFINE_string 'download-dir' "${this_dir}/download" 'Directory to cache downloaded files'
DEFINE_boolean 'system' true 'Install system dependencies'
DEFINE_boolean 'flair' false 'Install flair'
DEFINE_boolean 'precise' false 'Install Mycroft Precise'
DEFINE_boolean 'adapt' true 'Install Mycroft Adapt'
DEFINE_boolean 'google' false 'Install Google Text to Speech'
DEFINE_boolean 'kaldi' true 'Install Kaldi'
DEFINE_boolean 'offline' false "Don't download anything"
DEFINE_integer 'make-threads' 4 'Number of threads to use with make' 'j'
FLAGS "$@" || exit $?
eval set -- "${FLAGS_ARGV}"
# -----------------------------------------------------------------------------
# Default Settings
# -----------------------------------------------------------------------------
set -e
venv="${FLAGS_venv}"
download_dir="${FLAGS_download_dir}"
# Directory of *this* script
DIR="$( cd "$( dirname "$0" )" && pwd )"
# Place where downloaded artifacts are stored
download_dir="${DIR}/download"
mkdir -p "${download_dir}"
if [[ "${FLAGS_system}" -eq "${FLAGS_FALSE}" ]]; then
no_system='true'
fi
if [[ "${FLAGS_flair}" -eq "${FLAGS_FALSE}" ]]; then
no_flair='true'
fi
if [[ "${FLAGS_precise}" -eq "${FLAGS_FALSE}" ]]; then
no_precise='true'
fi
if [[ "${FLAGS_adapt}" -eq "${FLAGS_FALSE}" ]]; then
no_adapt='true'
fi
if [[ "${FLAGS_kaldi}" -eq "${FLAGS_FALSE}" ]]; then
no_kaldi='true'
fi
if [[ "${FLAGS_google}" -eq "${FLAGS_FALSE}" ]]; then
no_google='true'
fi
if [[ "${FLAGS_offline}" -eq "${FLAGS_TRUE}" ]]; then
offline='true'
fi
make_threads="${FLAGS_make_threads}"
# -----------------------------------------------------------------------------
# Create a temporary directory for building stuff
temp_dir="$(mktemp -d)"
function cleanup {
rm -rf "${temp_dir}"
}
trap cleanup EXIT
# -----------------------------------------------------------------------------
function maybe_download {
if [[ ! -f "$2" ]]; then
if [[ ! -z "${offline}" ]]; then
echo "Need to download $1 but offline."
exit 1
fi
mkdir -p "$(dirname "$2")"
curl -sSfL -o "$2" "$1"
echo "$1 => $2"
fi
}
# -----------------------------------------------------------------------------
# Debian dependencies
# -----------------------------------------------------------------------------
if [[ -z "${no_system}" ]]; then
echo "Installing system dependencies"
sudo apt-get update
sudo apt-get install --no-install-recommends --yes \
python3 python3-pip python3-venv python3-dev \
python \
build-essential autoconf autoconf-archive libtool automake bison \
sox espeak flite swig portaudio19-dev \
libatlas-base-dev \
gfortran \
sphinxbase-utils sphinxtrain pocketsphinx \
jq checkinstall unzip xz-utils \
curl
fi
# -----------------------------------------------------------------------------
# Python >= 3.6
# -----------------------------------------------------------------------------
if [[ ! -z "$(which python3.8)" ]]; then
PYTHON='python3.8'
elif [[ ! -z "$(which python3.7)" ]]; then
PYTHON='python3.7'
elif [[ ! -z "$(which python3.6)" ]]; then
PYTHON='python3.6'
else
echo "Installing Python 3.6 from source. This is going to take a LONG time."
sudo apt-get install --no-install-recommends --yes \
tk-dev libncurses5-dev libncursesw5-dev \
libreadline6-dev libdb5.3-dev libgdbm-dev \
libsqlite3-dev libssl-dev libbz2-dev \
libexpat1-dev liblzma-dev zlib1g-dev
python_file="${download_dir}/Python-3.6.8.tar.xz"
python_url='https://www.python.org/ftp/python/3.6.8/Python-3.6.8.tar.xz'
maybe_download "${python_url}" "${python_file}"
tar -C "${temp_dir}" -xf "${python_file}"
cd "${temp_dir}/Python-3.6.8" && \
./configure && \
make -j "${make_threads}" && \
sudo make altinstall
PYTHON='python3.6'
fi
# -----------------------------------------------------------------------------
# Download dependencies
# -----------------------------------------------------------------------------
# CPU architecture
CPU_ARCH="$(uname --m)"
case "${CPU_ARCH}" in
CPU_ARCH="$(lscpu | awk '/^Architecture/{print $2}')"
case $CPU_ARCH in
x86_64)
FRIENDLY_ARCH=amd64
;;
@@ -157,90 +24,113 @@ case "${CPU_ARCH}" in
;;
esac
# Create a temporary directory for building stuff
temp_dir="$(mktemp -d)"
function cleanup {
rm -rf "${temp_dir}"
}
trap cleanup EXIT
# -----------------------------------------------------------------------------
# Debian dependencies
# -----------------------------------------------------------------------------
echo "Installing system dependencies (${FRIENDLY_ARCH})"
sudo apt-get update
sudo apt-get install -y python3 python3-pip python3-venv python3-dev \
python \
build-essential autoconf autoconf-archive libtool automake bison \
sox espeak flite swig portaudio19-dev \
libatlas-base-dev \
gfortran \
sphinxbase-utils sphinxtrain pocketsphinx \
jq checkinstall unzip xz-utils \
curl
# Download dependencies
echo "Downloading dependencies"
download_args=()
if [[ ! -z "${offline}" ]]; then
download_args+=('--offline')
fi
bash download-dependencies.sh
if [[ ! -z "${no_precise}" ]]; then
download_args+=('--noprecise')
fi
# -----------------------------------------------------------------------------
# OpenFST
# -----------------------------------------------------------------------------
if [[ ! -z "${no_kaldi}" ]]; then
download_args+=('--nokaldi')
fi
case $CPU_ARCH in
armv7l|arm64v8)
# Build from source
openfst_file="${download_dir}/openfst-1.6.2.tar.gz"
echo "Building OpenFST (${openfst_file})"
tar -C "${temp_dir}" -xzf "${openfst_file}" && \
cd "${temp_dir}/openfst-1.6.2" && \
./configure --enable-static --enable-shared --enable-far --enable-ngram-fsts && \
make -j 4 && \
sudo make install
;;
bash download-dependencies.sh "${download_args[@]}"
*)
# Use pre-built packages
sudo apt-get install -y libfst-dev libfst-tools
esac
# -----------------------------------------------------------------------------
# Python 3.6
# -----------------------------------------------------------------------------
if [[ -z "$(which python3.6)" ]]; then
echo "Installing Python 3.6 from source. This is going to take a LONG time."
sudo apt-get install -y tk-dev libncurses5-dev libncursesw5-dev \
libreadline6-dev libdb5.3-dev libgdbm-dev \
libsqlite3-dev libssl-dev libbz2-dev \
libexpat1-dev liblzma-dev zlib1g-dev
python_file="${download_dir}/Python-3.6.8.tar.xz"
if [[ ! -f "${python_file}" ]]; then
python_url='https://www.python.org/ftp/python/3.6.8/Python-3.6.8.tar.xz'
curl -sSfL-o "${python_file}" "${python_url}"
fi
tar -C "${temp_dir}" -xf "${python_file}"
cd "${temp_dir}/Python-3.6.8" && \
./configure && \
make -j 4 && \
sudo make altinstall
fi
# -----------------------------------------------------------------------------
# Virtual environment
# -----------------------------------------------------------------------------
cd "${this_dir}"
cd "${DIR}"
echo "${venv}"
PYTHON="python3.6"
VENV_PATH="${DIR}/.venv"
echo "${VENV_PATH}"
if [[ -d "${venv}" ]]; then
echo "Removing existing virtual environment"
rm -rf "${venv}"
fi
echo "Removing existing virtual environment"
rm -rf "${VENV_PATH}"
echo "Creating new virtual environment"
mkdir -p "${venv}"
"${PYTHON}" -m venv "${venv}"
# Extract Rhasspy tools
rhasspy_tools_file="${download_dir}/rhasspy-tools_${FRIENDLY_ARCH}.tar.gz"
echo "Extracting tools (${rhasspy_tools_file})"
tar -C "${venv}" -xf "${rhasspy_tools_file}"
# Force .venv/lib to be used
export LD_LIBRARY_PATH="${venv}/lib:${LD_LIBRARY_PATH}"
mkdir -p "${VENV_PATH}"
"${PYTHON}" -m venv "${VENV_PATH}"
# shellcheck source=/dev/null
source "${venv}/bin/activate"
source "${VENV_PATH}/bin/activate"
"${PYTHON}" -m pip install wheel
echo "Installing Python requirements"
"${PYTHON}" -m pip install wheel setuptools
"${PYTHON}" -m pip install requests
# pytorch is not available on ARM
case "${CPU_ARCH}" in
case $CPU_ARCH in
armv7l|arm64v8)
no_flair="true" ;;
# Exclude flair
grep -v flair requirements.txt > "${temp_dir}/requirements-noflair.txt"
"${PYTHON}" -m pip install -r "${temp_dir}/requirements-noflair.txt"
;;
*)
# Install all requirements
"${PYTHON}" -m pip install -r requirements.txt
esac
requirements_file="${temp_dir}/requirements.txt"
cp "${this_dir}/requirements.txt" "${requirements_file}"
# Exclude requirements
if [[ ! -z "${no_flair}" ]]; then
echo "Excluding flair from virtual environment"
sed -i '/^flair/d' "${requirements_file}"
fi
if [[ ! -z "${no_precise}" ]]; then
echo "Excluding Mycroft Precise from virtual environment"
sed -i '/^precise-runner/d' "${requirements_file}"
fi
if [[ ! -z "${no_adapt}" ]]; then
echo "Excluding Mycroft Adapt from virtual environment"
sed -i '/^adapt-parser/d' "${requirements_file}"
fi
if [[ ! -z "${no_google}" ]]; then
echo "Excluding Google Text to Speech from virtual environment"
sed -i '/^google-cloud-texttospeech/d' "${requirements_file}"
fi
"${PYTHON}" -m pip install \
--global-option=build_ext \
--global-option="-I${venv}/include" \
--global-option="-L${venv}/lib" \
-r "${requirements_file}"
# -----------------------------------------------------------------------------
# Pocketsphinx for Python
# -----------------------------------------------------------------------------
@@ -252,9 +142,15 @@ pocketsphinx_file="${download_dir}/pocketsphinx-python.tar.gz"
# Snowboy
# -----------------------------------------------------------------------------
case "${CPU_ARCH}" in
case $CPU_ARCH in
x86_64|armv7l)
snowboy_file="${download_dir}/snowboy-1.3.0.tar.gz"
if [[ ! -f "${snowboy_file}" ]]; then
snowboy_url='https://github.com/Kitt-AI/snowboy/archive/v1.3.0.tar.gz'
echo "Downloading snowboy (${snowboy_url})"
curl -sSfL-o "${snowboy_file}" "${snowboy_url}"
fi
"${PYTHON}" -m pip install "${snowboy_file}"
;;
@@ -266,14 +162,13 @@ esac
# Mycroft Precise
# -----------------------------------------------------------------------------
if [[ -z "${no_precise}" && -z "$(which precise-engine)" ]]; then
case "${CPU_ARCH}" in
if [[ -z "$(which precise-engine)" ]]; then
case $CPU_ARCH in
x86_64|armv7l)
echo "Installing Mycroft Precise"
precise_file="${download_dir}/precise-engine_0.3.0_${CPU_ARCH}.tar.gz"
precise_install="${venv}/lib"
tar -C "${precise_install}" -xf "${precise_file}"
ln -s "${precise_install}/precise-engine/precise-engine" "${venv}/bin/precise-engine"
precise_install='/usr/lib'
sudo tar -C "${precise_install}" -xf "${precise_file}"
sudo ln -s "${precise_install}/precise-engine/precise-engine" '/usr/bin/precise-engine'
;;
*)
@@ -282,23 +177,71 @@ if [[ -z "${no_precise}" && -z "$(which precise-engine)" ]]; then
fi
# -----------------------------------------------------------------------------
# Kaldi
# Opengrm
# -----------------------------------------------------------------------------
if [[ -z "${no_kaldi}" ]]; then
kaldi_file="${download_dir}/kaldi_${FRIENDLY_ARCH}.tar.gz"
echo "Installing Kaldi (${kaldi_file})"
mkdir -p "${this_dir}/opt"
tar -C "${this_dir}/opt" -xf "${kaldi_file}"
if [[ -z "$(which ngramcount)" ]]; then
opengrm_file="${download_dir}/opengrm-ngram-1.3.3.tar.gz"
echo "Building Opengrm ${opengrm_file}"
tar -C "${temp_dir}" -xf "${opengrm_file}" && \
cd "${temp_dir}/opengrm-ngram-1.3.3" && \
./configure && \
make -j 4 && \
sudo make install && \
sudo ldconfig
fi
# -----------------------------------------------------------------------------
# Phonetisaurus
# -----------------------------------------------------------------------------
if [[ -z "$(which phonetisaurus-apply)" ]]; then
case $CPU_ARCH in
x86_64|armv7l|arm64v8)
# Install pre-built package
phonetisaurus_file="${download_dir}/phonetisaurus-2019_${FRIENDLY_ARCH}.deb"
echo "Installing phonetisaurus (${phonetisaurus_file})"
sudo dpkg -i "${phonetisaurus_file}"
;;
*)
# Build from source
phonetisaurus_file="${download_dir}/phonetisaurus-2019.zip"
echo "Building phonetisaurus (${phonetisaurus_file})"
unzip -d "${temp_dir}" "${phonetisaurus_file}" && \
cd "${temp_dir}/phonetisaurus" && \
./configure && \
make -j 4 && \
sudo make install
esac
fi
# Add /usr/local/lib to LD_LIBRARY_PATH
sudo ldconfig
# -----------------------------------------------------------------------------
# NodeJS / Yarn
# -----------------------------------------------------------------------------
if [[ -z "$(which node)" ]]; then
echo "Installing nodejs"
sudo apt-get install -y nodejs
fi
if [[ -z "$(which yarn)" ]]; then
echo "Installing yarn"
curl -o- -L https://yarnpkg.com/install.sh | bash
# Need to re-source .bashrc so yarn is in the path
source "${HOME}/.bashrc"
fi
# -----------------------------------------------------------------------------
# Web Interface
# -----------------------------------------------------------------------------
rhasspy_web_file="${download_dir}/rhasspy-web-dist.tar.gz"
echo "Extracting web interface (${rhasspy_web_file})"
tar -C "${this_dir}" -xf "${rhasspy_web_file}"
echo "Building web interface"
cd "${DIR}" && yarn && yarn build
# -----------------------------------------------------------------------------
debian/DEBIAN/control
Vendored
-8
View File
@@ -1,8 +0,0 @@
Package: rhasspy-server
Version: 2.4.8
Section: utils
Priority: optional
Depends: sox,alsa-utils,espeak,libstdc++6,jq,xz-utils,unzip,curl,sphinxbase-utils,sphinxtrain,flite,libatlas-base-dev,gfortran
Architecture: ${architecture}
Maintainer: Michael Hansen
Description: Offline voice assistant
debian/bin/rhasspy-server
Vendored
-11
View File
@@ -1,11 +0,0 @@
#!/usr/bin/env bash
if [[ -z "${RHASSPY_BASE_DIR}" ]]; then
export RHASSPY_BASE_DIR="/usr/lib/rhasspy"
fi
if [[ -z "${KALDI_PREFIX}" ]]; then
export KALDI_PREFIX="${RHASSPY_BASE_DIR}"
fi
cd "${RHASSPY_BASE_DIR}" && rhasspy/rhasspy "$@"
debianize.sh
-177
View File
@@ -1,177 +0,0 @@
#!/usr/bin/env bash
rhasspy_version="2.4.8"
this_dir="$( cd "$( dirname "$0" )" && pwd )"
# -----------------------------------------------------------------------------
# Command-line Arguments
# -----------------------------------------------------------------------------
. "${this_dir}/etc/shflags"
DEFINE_string 'architecture' '' 'Debian architecture'
DEFINE_string 'version' "${rhasspy_version}" 'Package version'
DEFINE_boolean 'package' true 'Create debian package (.deb)'
FLAGS "$@" || exit $?
eval set -- "${FLAGS_ARGV}"
# -----------------------------------------------------------------------------
# Settings
# -----------------------------------------------------------------------------
export architecture="${FLAGS_architecture}"
version="${FLAGS_version}"
debian_dir="${this_dir}/debian"
set -e
if [[ -z "${architecture}" ]]; then
# Guess architecture
architecture="$(dpkg-architecture | grep 'DEB_BUILD_ARCH=' | sed 's/^[^=]\+=//')"
fi
# -----------------------------------------------------------------------------
# Activate virtual environment
# -----------------------------------------------------------------------------
venv="${this_dir}/.venv"
if [[ ! -d "${venv}" ]]; then
echo "Missing virtual environment at ${venv}"
echo "Did you run create-venv.sh?"
exit 1
fi
cd "${this_dir}"
source "${venv}/bin/activate"
if [[ -z "$(which pyinstaller)" ]]; then
echo "Missing PyInstaller"
exit 1
fi
# -----------------------------------------------------------------------------
# Run PyInstaller
# -----------------------------------------------------------------------------
echo "Running PyInstaller"
package_name="rhasspy-server_${version}_${architecture}"
package_dir="${debian_dir}/${package_name}"
output_dir="${package_dir}/usr/lib/rhasspy"
share_dir="${package_dir}/usr/share/rhasspy"
pyinstaller\
-y \
--workpath "pyinstaller/build" \
--distpath "${output_dir}" \
"${this_dir}/rhasspy.spec"
# Remove all symbols (Liantian warning)
strip --strip-all "${output_dir}/rhasspy"/*.so* || true
# Remove executable bit from shared libs (Lintian warning)
chmod -x "${output_dir}/rhasspy"/*.so* || true
# -----------------------------------------------------------------------------
# Copy Rhasspy
# -----------------------------------------------------------------------------
# Profiles
mkdir -p "${output_dir}/profiles"
rsync -av \
--delete \
--exclude 'acoustic_model' \
--exclude 'download' \
--exclude 'flair' \
--exclude 'base_dictionary.txt' \
--exclude 'base_language_model.txt' \
--exclude 'g2p.fst' \
--exclude 'HCLG.fst' \
--exclude 'final.mdl' \
--exclude '*.umdl' \
"${this_dir}/profiles/" \
"${output_dir}/profiles/"
# Sounds
mkdir -p "${output_dir}/etc/wav"
rsync -av \
--delete \
"${this_dir}/etc/wav/" \
"${output_dir}/etc/wav/"
# Web
mkdir -p "${output_dir}/dist"
rsync -av \
--delete \
"${this_dir}/dist/" \
"${output_dir}/dist/"
# Documentation
mkdocs build
mkdir -p "${share_dir}/docs"
rsync -av \
--delete \
"${this_dir}/site/" \
"${share_dir}/docs/"
# Source code
mkdir -p "${share_dir}/src"
rsync -av \
--delete \
--exclude '.mypy_cache' \
--exclude '__pycache__' \
"${this_dir}/rhasspy/" \
"${share_dir}/src/rhasspy/"
cp "${this_dir}/app.py" "${share_dir}/src/"
# -----------------------------------------------------------------------------
# Copy Kaldi
# -----------------------------------------------------------------------------
echo "Copying Kaldi"
kaldi_src="${venv}/kaldi"
if [[ ! -d "${kaldi_src}" ]]; then
echo "Missing Kaldi at ${kaldi_src}"
exit 1
fi
kaldi_dest="${output_dir}/kaldi"
mkdir -p "${kaldi_dest}"
rsync -av --delete "${kaldi_src}/" "${kaldi_dest}/"
# Avoid link recursion
rm -f "${kaldi_dest}/egs/wsj/s5/utils/utils"
# Turn duplicate .so files into symbolic links
function fix_library_links() {
lib_dir="$1"
for lib in "${lib_dir}"/*.so; do
lib_base="$(basename ${lib})"
for lib_link in "${lib_dir}/${lib_base}".*; do
rm -f "${lib_link}"
ln -s "${lib_base}" "${lib_link}"
done
done
}
fix_library_links "${kaldi_dest}/tools/openfst/lib"
# -----------------------------------------------------------------------------
# Create Debian package
# -----------------------------------------------------------------------------
echo "Creating Debian package"
mkdir -p "${package_dir}/DEBIAN"
cat "${debian_dir}/DEBIAN/control" | \
envsubst > "${package_dir}/DEBIAN/control"
mkdir -p "${package_dir}/usr/bin"
cp "${debian_dir}/bin/rhasspy-server" "${package_dir}/usr/bin/"
if [[ "${FLAGS_package}" -eq "${FLAGS_TRUE}" ]]; then
# Actually build the package
cd 'debian' && fakeroot dpkg --build "${package_name}"
fi
docker/templates/Dockerfile.template
+5 -4
View File
@@ -14,7 +14,9 @@ QEMU
SYSTEM_DEPENDENCIES
RHASSPY_TOOLS
PHONETISAURUS
OPENGRM
PYTHON_REQUIREMENTS
@@ -22,6 +24,8 @@ PYTHON_POCKETSPHINX
SNOWBOY
MYCROFT_PRECISE
TTS
KALDI
@@ -32,8 +36,6 @@ RUN ldconfig
RUN find / -name bw -exec cp '{}' /usr/bin/ \;
RUN find / -name mllr_solve -exec cp '{}' /usr/bin/ \;
GSTREAMER
PULSEAUDIO
# Copy script to run
@@ -47,6 +49,5 @@ PROFILES
RHASSPY_CODE
ENV CONFIG_PATH /data/options.json
ENV KALDI_PREFIX /opt
ENTRYPOINT ["/run.sh"]
docker/templates/all_profiles/profiles.txt
+24 -31
View File
@@ -1,5 +1,7 @@
COPY profiles/zh/profile.json \
profiles/zh/custom_words.txt \
profiles/zh/download-profile.sh \
profiles/zh/check-profile.sh \
profiles/zh/espeak_phonemes.txt \
profiles/zh/phoneme_examples.txt \
profiles/zh/frequent_words.txt \
@@ -9,6 +11,8 @@ COPY profiles/zh/profile.json \
COPY profiles/hi/ \
profiles/hi/profile.json \
profiles/hi/custom_words.txt \
profiles/hi/download-profile.sh \
profiles/hi/check-profile.sh \
profiles/hi/espeak_phonemes.txt \
profiles/hi/phoneme_examples.txt \
profiles/hi/frequent_words.txt \
@@ -17,6 +21,8 @@ COPY profiles/hi/ \
COPY profiles/el/profile.json \
profiles/el/custom_words.txt \
profiles/el/download-profile.sh \
profiles/el/check-profile.sh \
profiles/el/espeak_phonemes.txt \
profiles/el/phoneme_examples.txt \
profiles/el/frequent_words.txt \
@@ -25,19 +31,18 @@ COPY profiles/el/profile.json \
COPY profiles/de/profile.json \
profiles/de/custom_words.txt \
profiles/de/download-profile.sh \
profiles/de/check-profile.sh \
profiles/de/espeak_phonemes.txt \
profiles/de/phoneme_examples.txt \
profiles/de/frequent_words.txt \
profiles/de/sentences.ini \
profiles/de/stop_words.txt ${RHASSPY_APP}/profiles/de/
COPY profiles/de/kaldi/custom_words.txt \
profiles/de/kaldi/espeak_phonemes.txt \
profiles/de/kaldi/phoneme_examples.txt \
${RHASSPY_APP}/profiles/de/kaldi/
COPY profiles/it/profile.json \
profiles/it/custom_words.txt \
profiles/it/download-profile.sh \
profiles/it/check-profile.sh \
profiles/it/espeak_phonemes.txt \
profiles/it/phoneme_examples.txt \
profiles/it/frequent_words.txt \
@@ -46,6 +51,8 @@ COPY profiles/it/profile.json \
COPY profiles/es/profile.json \
profiles/es/custom_words.txt \
profiles/es/download-profile.sh \
profiles/es/check-profile.sh \
profiles/es/espeak_phonemes.txt \
profiles/es/phoneme_examples.txt \
profiles/es/frequent_words.txt \
@@ -54,6 +61,8 @@ COPY profiles/es/profile.json \
COPY profiles/fr/profile.json \
profiles/fr/custom_words.txt \
profiles/fr/download-profile.sh \
profiles/fr/check-profile.sh \
profiles/fr/espeak_phonemes.txt \
profiles/fr/phoneme_examples.txt \
profiles/fr/frequent_words.txt \
@@ -62,6 +71,8 @@ COPY profiles/fr/profile.json \
COPY profiles/ru/profile.json \
profiles/ru/custom_words.txt \
profiles/ru/download-profile.sh \
profiles/ru/check-profile.sh \
profiles/ru/espeak_phonemes.txt \
profiles/ru/phoneme_examples.txt \
profiles/ru/frequent_words.txt \
@@ -70,19 +81,18 @@ COPY profiles/ru/profile.json \
COPY profiles/nl/profile.json \
profiles/nl/custom_words.txt \
profiles/nl/download-profile.sh \
profiles/nl/check-profile.sh \
profiles/nl/espeak_phonemes.txt \
profiles/nl/phoneme_examples.txt \
profiles/nl/frequent_words.txt \
profiles/nl/sentences.ini \
profiles/nl/stop_words.txt ${RHASSPY_APP}/profiles/nl/
COPY profiles/nl/kaldi/custom_words.txt \
profiles/nl/kaldi/espeak_phonemes.txt \
profiles/nl/kaldi/phoneme_examples.txt \
${RHASSPY_APP}/profiles/nl/kaldi/
COPY profiles/vi/profile.json \
profiles/vi/custom_words.txt \
profiles/vi/download-profile.sh \
profiles/vi/check-profile.sh \
profiles/vi/espeak_phonemes.txt \
profiles/vi/phoneme_examples.txt \
profiles/vi/frequent_words.txt \
@@ -91,37 +101,20 @@ COPY profiles/vi/profile.json \
COPY profiles/pt/profile.json \
profiles/pt/custom_words.txt \
profiles/pt/download-profile.sh \
profiles/pt/check-profile.sh \
profiles/pt/espeak_phonemes.txt \
profiles/pt/phoneme_examples.txt \
profiles/pt/frequent_words.txt \
profiles/pt/sentences.ini \
profiles/pt/stop_words.txt ${RHASSPY_APP}/profiles/pt/
COPY profiles/sv/profile.json \
profiles/sv/custom_words.txt \
profiles/sv/espeak_phonemes.txt \
profiles/sv/phoneme_examples.txt \
profiles/sv/frequent_words.txt \
profiles/sv/sentences.ini \
profiles/sv/stop_words.txt ${RHASSPY_APP}/profiles/sv/
COPY profiles/ca/profile.json \
profiles/ca/custom_words.txt \
profiles/ca/espeak_phonemes.txt \
profiles/ca/phoneme_examples.txt \
profiles/ca/frequent_words.txt \
profiles/ca/sentences.ini \
profiles/ca/stop_words.txt ${RHASSPY_APP}/profiles/ca/
COPY profiles/en/profile.json \
profiles/en/custom_words.txt \
profiles/en/download-profile.sh \
profiles/en/check-profile.sh \
profiles/en/espeak_phonemes.txt \
profiles/en/phoneme_examples.txt \
profiles/en/frequent_words.txt \
profiles/en/sentences.ini \
profiles/en/stop_words.txt ${RHASSPY_APP}/profiles/en/
COPY profiles/en/kaldi/custom_words.txt \
profiles/en/kaldi/espeak_phonemes.txt \
profiles/en/kaldi/phoneme_examples.txt \
${RHASSPY_APP}/profiles/en/kaldi/
docker/templates/dockerfiles/Dockerfile.from-source.alsa.all
+64 -36
View File
@@ -14,30 +14,48 @@ COPY etc/qemu-arm-static /usr/bin/
COPY etc/qemu-aarch64-static /usr/bin/
RUN apt-get update && \
apt-get install --no-install-recommends --yes \
bash jq unzip \
python3 python3-pip python3-dev \
build-essential portaudio19-dev swig \
libatlas-base-dev \
sox espeak flite alsa-utils \
git curl \
autoconf libtool automake bison \
sphinxbase-utils sphinxtrain
apt-get install -y bash jq unzip \
python3 python3-pip python3-dev \
build-essential portaudio19-dev swig \
libatlas-base-dev \
sox espeak flite alsa-utils \
git curl \
autoconf libtool automake bison \
sphinxbase-utils sphinxtrain
COPY download/openfst-1.6.2-1_${BUILD_ARCH}.deb /
RUN if [ "$BUILD_ARCH" != "aarch64" ]; then \
apt-get install --no-install-recommends --yes libfst-dev libfst-tools; \
apt-get install -y libfst-dev libfst-tools; \
else \
dpkg -i /openfst-1.6.2-1_${BUILD_ARCH}.deb; \
rm /openfst*.deb; \
fi
RHASSPY_TOOLS
COPY download/phonetisaurus-2019.tar.gz /phonetisaurus.tar.gz
RUN cd / && tar -xf phonetisaurus.tar.gz
RUN cd /phonetisaurus && \
./configure && \
make -j $MAKE_THREADS && \
make install && \
rm -rf /phonetisaurus*
# Install Opengrm
COPY download/opengrm-ngram-1.3.3.tar.gz /
RUN cd / && tar -xf opengrm-ngram-1.3.3.tar.gz && \
cd opengrm-ngram-1.3.3 && \
./configure && \
make -j $MAKE_THREADS && \
make install && \
ldconfig && \
rm -rf /opengrm*
# Install Python dependencies
RUN python3 -m pip install --no-cache-dir setuptools wheel
RUN python3 -m pip install --no-cache-dir wheel
RUN apt-get install -y libfreetype6-dev libpng-dev pkg-config libffi-dev libssl-dev
COPY download/jsgf2fst-0.1.0.tar.gz \
/download/
RUN apt-get install -y libfreetype6-dev libpng-dev pkg-config
COPY requirements.txt /requirements.txt
RUN if [ "$BUILD_ARCH" != "amd64" ]; then \
grep -v flair /requirements.txt > /requirements-noflair.txt; \
@@ -54,7 +72,15 @@ RUN python3 -m pip install --no-cache-dir /pocketsphinx-python.tar.gz && \
COPY download/snowboy-1.3.0.tar.gz /
RUN if [ "$BUILD_ARCH" != "aarch64" ]; then pip3 install --no-cache-dir /snowboy-1.3.0.tar.gz; fi
RUN apt-get install --no-install-recommends --yes flite libttspico-utils
# Install Mycroft Precise
COPY download/precise-engine_0.3.0_${CPU_ARCH}.tar.gz /precise-engine.tar.gz
RUN if [ "$BUILD_ARCH" != "aarch64" ]; then \
cd / && tar -xzf /precise-engine.tar.gz && \
ln -s /precise-engine/precise-engine /usr/bin/precise-engine && \
rm /precise-engine.tar.gz; \
fi
RUN apt-get install -y flite libttspico-utils
COPY download/kaldi_${BUILD_ARCH}.tar.gz /kaldi.tar.gz
RUN mkdir -p /opt && \
@@ -67,9 +93,6 @@ RUN ldconfig
RUN find / -name bw -exec cp '{}' /usr/bin/ \;
RUN find / -name mllr_solve -exec cp '{}' /usr/bin/ \;
# Install gstreamer and plugins
RUN apt-get install --no-install-recommends --yes gstreamer1.0-tools gstreamer1.0-plugins-good
ENV RHASSPY_APP /usr/share/rhasspy
# Copy script to run
@@ -80,6 +103,8 @@ RUN chmod +x /run.sh
COPY profiles/zh/profile.json \
profiles/zh/custom_words.txt \
profiles/zh/download-profile.sh \
profiles/zh/check-profile.sh \
profiles/zh/espeak_phonemes.txt \
profiles/zh/phoneme_examples.txt \
profiles/zh/frequent_words.txt \
@@ -89,6 +114,8 @@ COPY profiles/zh/profile.json \
COPY profiles/hi/ \
profiles/hi/profile.json \
profiles/hi/custom_words.txt \
profiles/hi/download-profile.sh \
profiles/hi/check-profile.sh \
profiles/hi/espeak_phonemes.txt \
profiles/hi/phoneme_examples.txt \
profiles/hi/frequent_words.txt \
@@ -97,6 +124,8 @@ COPY profiles/hi/ \
COPY profiles/el/profile.json \
profiles/el/custom_words.txt \
profiles/el/download-profile.sh \
profiles/el/check-profile.sh \
profiles/el/espeak_phonemes.txt \
profiles/el/phoneme_examples.txt \
profiles/el/frequent_words.txt \
@@ -105,6 +134,8 @@ COPY profiles/el/profile.json \
COPY profiles/de/profile.json \
profiles/de/custom_words.txt \
profiles/de/download-profile.sh \
profiles/de/check-profile.sh \
profiles/de/espeak_phonemes.txt \
profiles/de/phoneme_examples.txt \
profiles/de/frequent_words.txt \
@@ -113,6 +144,8 @@ COPY profiles/de/profile.json \
COPY profiles/it/profile.json \
profiles/it/custom_words.txt \
profiles/it/download-profile.sh \
profiles/it/check-profile.sh \
profiles/it/espeak_phonemes.txt \
profiles/it/phoneme_examples.txt \
profiles/it/frequent_words.txt \
@@ -121,6 +154,8 @@ COPY profiles/it/profile.json \
COPY profiles/es/profile.json \
profiles/es/custom_words.txt \
profiles/es/download-profile.sh \
profiles/es/check-profile.sh \
profiles/es/espeak_phonemes.txt \
profiles/es/phoneme_examples.txt \
profiles/es/frequent_words.txt \
@@ -129,6 +164,8 @@ COPY profiles/es/profile.json \
COPY profiles/fr/profile.json \
profiles/fr/custom_words.txt \
profiles/fr/download-profile.sh \
profiles/fr/check-profile.sh \
profiles/fr/espeak_phonemes.txt \
profiles/fr/phoneme_examples.txt \
profiles/fr/frequent_words.txt \
@@ -137,6 +174,8 @@ COPY profiles/fr/profile.json \
COPY profiles/ru/profile.json \
profiles/ru/custom_words.txt \
profiles/ru/download-profile.sh \
profiles/ru/check-profile.sh \
profiles/ru/espeak_phonemes.txt \
profiles/ru/phoneme_examples.txt \
profiles/ru/frequent_words.txt \
@@ -145,6 +184,8 @@ COPY profiles/ru/profile.json \
COPY profiles/nl/profile.json \
profiles/nl/custom_words.txt \
profiles/nl/download-profile.sh \
profiles/nl/check-profile.sh \
profiles/nl/espeak_phonemes.txt \
profiles/nl/phoneme_examples.txt \
profiles/nl/frequent_words.txt \
@@ -153,6 +194,8 @@ COPY profiles/nl/profile.json \
COPY profiles/vi/profile.json \
profiles/vi/custom_words.txt \
profiles/vi/download-profile.sh \
profiles/vi/check-profile.sh \
profiles/vi/espeak_phonemes.txt \
profiles/vi/phoneme_examples.txt \
profiles/vi/frequent_words.txt \
@@ -161,30 +204,18 @@ COPY profiles/vi/profile.json \
COPY profiles/pt/profile.json \
profiles/pt/custom_words.txt \
profiles/pt/download-profile.sh \
profiles/pt/check-profile.sh \
profiles/pt/espeak_phonemes.txt \
profiles/pt/phoneme_examples.txt \
profiles/pt/frequent_words.txt \
profiles/pt/sentences.ini \
profiles/pt/stop_words.txt ${RHASSPY_APP}/profiles/pt/
COPY profiles/sv/profile.json \
profiles/sv/custom_words.txt \
profiles/sv/espeak_phonemes.txt \
profiles/sv/phoneme_examples.txt \
profiles/sv/frequent_words.txt \
profiles/sv/sentences.ini \
profiles/sv/stop_words.txt ${RHASSPY_APP}/profiles/sv/
COPY profiles/ca/profile.json \
profiles/ca/custom_words.txt \
profiles/ca/espeak_phonemes.txt \
profiles/ca/phoneme_examples.txt \
profiles/ca/frequent_words.txt \
profiles/ca/sentences.ini \
profiles/ca/stop_words.txt ${RHASSPY_APP}/profiles/ca/
COPY profiles/en/profile.json \
profiles/en/custom_words.txt \
profiles/en/download-profile.sh \
profiles/en/check-profile.sh \
profiles/en/espeak_phonemes.txt \
profiles/en/phoneme_examples.txt \
profiles/en/frequent_words.txt \
@@ -196,12 +227,9 @@ COPY docker/rhasspy ${RHASSPY_APP}/bin/
COPY dist/ ${RHASSPY_APP}/dist/
COPY etc/wav/* ${RHASSPY_APP}/etc/wav/
COPY rhasspy/profile_schema.json ${RHASSPY_APP}/rhasspy/
COPY rhasspy/train/jsgf2fst/*.py ${RHASSPY_APP}/rhasspy/train/jsgf2fst/
COPY rhasspy/train/*.py ${RHASSPY_APP}/rhasspy/train/
COPY *.py ${RHASSPY_APP}/
COPY rhasspy/*.py ${RHASSPY_APP}/rhasspy/
ENV CONFIG_PATH /data/options.json
ENV KALDI_PREFIX /opt
ENTRYPOINT ["/run.sh"]
docker/templates/dockerfiles/Dockerfile.from-source.alsa.en
+40 -20
View File
@@ -14,30 +14,48 @@ COPY etc/qemu-arm-static /usr/bin/
COPY etc/qemu-aarch64-static /usr/bin/
RUN apt-get update && \
apt-get install --no-install-recommends --yes \
bash jq unzip \
python3 python3-pip python3-dev \
build-essential portaudio19-dev swig \
libatlas-base-dev \
sox espeak flite alsa-utils \
git curl \
autoconf libtool automake bison \
sphinxbase-utils sphinxtrain
apt-get install -y bash jq unzip \
python3 python3-pip python3-dev \
build-essential portaudio19-dev swig \
libatlas-base-dev \
sox espeak flite alsa-utils \
git curl \
autoconf libtool automake bison \
sphinxbase-utils sphinxtrain
COPY download/openfst-1.6.2-1_${BUILD_ARCH}.deb /
RUN if [ "$BUILD_ARCH" != "aarch64" ]; then \
apt-get install --no-install-recommends --yes libfst-dev libfst-tools; \
apt-get install -y libfst-dev libfst-tools; \
else \
dpkg -i /openfst-1.6.2-1_${BUILD_ARCH}.deb; \
rm /openfst*.deb; \
fi
RHASSPY_TOOLS
COPY download/phonetisaurus-2019.tar.gz /phonetisaurus.tar.gz
RUN cd / && tar -xf phonetisaurus.tar.gz
RUN cd /phonetisaurus && \
./configure && \
make -j $MAKE_THREADS && \
make install && \
rm -rf /phonetisaurus*
# Install Opengrm
COPY download/opengrm-ngram-1.3.3.tar.gz /
RUN cd / && tar -xf opengrm-ngram-1.3.3.tar.gz && \
cd opengrm-ngram-1.3.3 && \
./configure && \
make -j $MAKE_THREADS && \
make install && \
ldconfig && \
rm -rf /opengrm*
# Install Python dependencies
RUN python3 -m pip install --no-cache-dir setuptools wheel
RUN python3 -m pip install --no-cache-dir wheel
RUN apt-get install -y libfreetype6-dev libpng-dev pkg-config libffi-dev libssl-dev
COPY download/jsgf2fst-0.1.0.tar.gz \
/download/
RUN apt-get install -y libfreetype6-dev libpng-dev pkg-config
COPY requirements.txt /requirements.txt
RUN if [ "$BUILD_ARCH" != "amd64" ]; then \
grep -v flair /requirements.txt > /requirements-noflair.txt; \
@@ -54,7 +72,15 @@ RUN python3 -m pip install --no-cache-dir /pocketsphinx-python.tar.gz && \
COPY download/snowboy-1.3.0.tar.gz /
RUN if [ "$BUILD_ARCH" != "aarch64" ]; then pip3 install --no-cache-dir /snowboy-1.3.0.tar.gz; fi
RUN apt-get install --no-install-recommends --yes flite libttspico-utils
# Install Mycroft Precise
COPY download/precise-engine_0.3.0_${CPU_ARCH}.tar.gz /precise-engine.tar.gz
RUN if [ "$BUILD_ARCH" != "aarch64" ]; then \
cd / && tar -xzf /precise-engine.tar.gz && \
ln -s /precise-engine/precise-engine /usr/bin/precise-engine && \
rm /precise-engine.tar.gz; \
fi
RUN apt-get install -y flite libttspico-utils
COPY download/kaldi_${BUILD_ARCH}.tar.gz /kaldi.tar.gz
RUN mkdir -p /opt && \
@@ -67,9 +93,6 @@ RUN ldconfig
RUN find / -name bw -exec cp '{}' /usr/bin/ \;
RUN find / -name mllr_solve -exec cp '{}' /usr/bin/ \;
# Install gstreamer and plugins
RUN apt-get install --no-install-recommends --yes gstreamer1.0-tools gstreamer1.0-plugins-good
ENV RHASSPY_APP /usr/share/rhasspy
# Copy script to run
@@ -85,12 +108,9 @@ COPY docker/rhasspy ${RHASSPY_APP}/bin/
COPY dist/ ${RHASSPY_APP}/dist/
COPY etc/wav/* ${RHASSPY_APP}/etc/wav/
COPY rhasspy/profile_schema.json ${RHASSPY_APP}/rhasspy/
COPY rhasspy/train/jsgf2fst/*.py ${RHASSPY_APP}/rhasspy/train/jsgf2fst/
COPY rhasspy/train/*.py ${RHASSPY_APP}/rhasspy/train/
COPY *.py ${RHASSPY_APP}/
COPY rhasspy/*.py ${RHASSPY_APP}/rhasspy/
ENV CONFIG_PATH /data/options.json
ENV KALDI_PREFIX /opt
ENTRYPOINT ["/run.sh"]
docker/templates/dockerfiles/Dockerfile.from-source.pulseaudio.en
+40 -20
View File
@@ -14,30 +14,48 @@ COPY etc/qemu-arm-static /usr/bin/
COPY etc/qemu-aarch64-static /usr/bin/
RUN apt-get update && \
apt-get install --no-install-recommends --yes \
bash jq unzip \
python3 python3-pip python3-dev \
build-essential portaudio19-dev swig \
libatlas-base-dev \
sox espeak flite alsa-utils \
git curl \
autoconf libtool automake bison \
sphinxbase-utils sphinxtrain
apt-get install -y bash jq unzip \
python3 python3-pip python3-dev \
build-essential portaudio19-dev swig \
libatlas-base-dev \
sox espeak flite alsa-utils \
git curl \
autoconf libtool automake bison \
sphinxbase-utils sphinxtrain
COPY download/openfst-1.6.2-1_${BUILD_ARCH}.deb /
RUN if [ "$BUILD_ARCH" != "aarch64" ]; then \
apt-get install --no-install-recommends --yes libfst-dev libfst-tools; \
apt-get install -y libfst-dev libfst-tools; \
else \
dpkg -i /openfst-1.6.2-1_${BUILD_ARCH}.deb; \
rm /openfst*.deb; \
fi
RHASSPY_TOOLS
COPY download/phonetisaurus-2019.tar.gz /phonetisaurus.tar.gz
RUN cd / && tar -xf phonetisaurus.tar.gz
RUN cd /phonetisaurus && \
./configure && \
make -j $MAKE_THREADS && \
make install && \
rm -rf /phonetisaurus*
# Install Opengrm
COPY download/opengrm-ngram-1.3.3.tar.gz /
RUN cd / && tar -xf opengrm-ngram-1.3.3.tar.gz && \
cd opengrm-ngram-1.3.3 && \
./configure && \
make -j $MAKE_THREADS && \
make install && \
ldconfig && \
rm -rf /opengrm*
# Install Python dependencies
RUN python3 -m pip install --no-cache-dir setuptools wheel
RUN python3 -m pip install --no-cache-dir wheel
RUN apt-get install -y libfreetype6-dev libpng-dev pkg-config libffi-dev libssl-dev
COPY download/jsgf2fst-0.1.0.tar.gz \
/download/
RUN apt-get install -y libfreetype6-dev libpng-dev pkg-config
COPY requirements.txt /requirements.txt
RUN if [ "$BUILD_ARCH" != "amd64" ]; then \
grep -v flair /requirements.txt > /requirements-noflair.txt; \
@@ -54,7 +72,15 @@ RUN python3 -m pip install --no-cache-dir /pocketsphinx-python.tar.gz && \
COPY download/snowboy-1.3.0.tar.gz /
RUN if [ "$BUILD_ARCH" != "aarch64" ]; then pip3 install --no-cache-dir /snowboy-1.3.0.tar.gz; fi
RUN apt-get install --no-install-recommends --yes flite libttspico-utils
# Install Mycroft Precise
COPY download/precise-engine_0.3.0_${CPU_ARCH}.tar.gz /precise-engine.tar.gz
RUN if [ "$BUILD_ARCH" != "aarch64" ]; then \
cd / && tar -xzf /precise-engine.tar.gz && \
ln -s /precise-engine/precise-engine /usr/bin/precise-engine && \
rm /precise-engine.tar.gz; \
fi
RUN apt-get install -y flite libttspico-utils
COPY download/kaldi_${BUILD_ARCH}.tar.gz /kaldi.tar.gz
RUN mkdir -p /opt && \
@@ -67,9 +93,6 @@ RUN ldconfig
RUN find / -name bw -exec cp '{}' /usr/bin/ \;
RUN find / -name mllr_solve -exec cp '{}' /usr/bin/ \;
# Install gstreamer and plugins
RUN apt-get install --no-install-recommends --yes gstreamer1.0-tools gstreamer1.0-plugins-good
RUN apt-get install -y pulseaudio
COPY etc/pulseaudio.client.conf /etc/pulse/client.conf
@@ -102,12 +125,9 @@ COPY docker/rhasspy ${RHASSPY_APP}/bin/
COPY dist/ ${RHASSPY_APP}/dist/
COPY etc/wav/* ${RHASSPY_APP}/etc/wav/
COPY rhasspy/profile_schema.json ${RHASSPY_APP}/rhasspy/
COPY rhasspy/train/jsgf2fst/*.py ${RHASSPY_APP}/rhasspy/train/jsgf2fst/
COPY rhasspy/train/*.py ${RHASSPY_APP}/rhasspy/train/
COPY *.py ${RHASSPY_APP}/
COPY rhasspy/*.py ${RHASSPY_APP}/rhasspy/
ENV CONFIG_PATH /data/options.json
ENV KALDI_PREFIX /opt
ENTRYPOINT ["/run.sh"]
docker/templates/dockerfiles/Dockerfile.prebuilt.alsa.all
Regular → Executable
+60 -52
View File
@@ -14,24 +14,37 @@ COPY etc/qemu-arm-static /usr/bin/
COPY etc/qemu-aarch64-static /usr/bin/
RUN apt-get update && \
apt-get install --no-install-recommends --yes \
bash jq unzip \
python3 python3-pip python3-dev \
build-essential portaudio19-dev swig \
libatlas-base-dev \
sox espeak flite alsa-utils \
git curl \
autoconf libtool automake bison \
sphinxbase-utils sphinxtrain
apt-get install -y bash jq unzip \
python3 python3-pip python3-dev \
build-essential portaudio19-dev swig \
libfst-dev libfst-tools \
libatlas-base-dev \
sox espeak flite alsa-utils \
git curl \
autoconf libtool automake bison \
sphinxbase-utils sphinxtrain
COPY download/rhasspy-tools_${BUILD_ARCH}.tar.gz /
RUN tar -C /usr -xvf /rhasspy-tools_${BUILD_ARCH}.tar.gz && \
rm -f /rhasspy-tools_${BUILD_ARCH}.tar.gz
COPY download/phonetisaurus-2019_${BUILD_ARCH}.deb /phonetisaurus.deb
RUN dpkg -i /phonetisaurus.deb && \
rm /phonetisaurus.deb
# Install Opengrm
COPY download/opengrm-ngram-1.3.3.tar.gz /
RUN cd / && tar -xf opengrm-ngram-1.3.3.tar.gz && \
cd opengrm-ngram-1.3.3 && \
./configure && \
make -j $MAKE_THREADS && \
make install && \
ldconfig && \
rm -rf /opengrm*
# Install Python dependencies
RUN python3 -m pip install --no-cache-dir setuptools wheel
RUN python3 -m pip install --no-cache-dir wheel
RUN apt-get install -y libfreetype6-dev libpng-dev pkg-config libffi-dev libssl-dev
COPY download/jsgf2fst-0.1.0.tar.gz \
/download/
RUN apt-get install -y libfreetype6-dev libpng-dev pkg-config
COPY requirements.txt /requirements.txt
RUN if [ "$BUILD_ARCH" != "amd64" ]; then \
grep -v flair /requirements.txt > /requirements-noflair.txt; \
@@ -48,7 +61,15 @@ RUN python3 -m pip install --no-cache-dir /pocketsphinx-python.tar.gz && \
COPY download/snowboy-1.3.0.tar.gz /
RUN if [ "$BUILD_ARCH" != "aarch64" ]; then pip3 install --no-cache-dir /snowboy-1.3.0.tar.gz; fi
RUN apt-get install --no-install-recommends --yes flite libttspico-utils
# Install Mycroft Precise
COPY download/precise-engine_0.3.0_${CPU_ARCH}.tar.gz /precise-engine.tar.gz
RUN if [ "$BUILD_ARCH" != "aarch64" ]; then \
cd / && tar -xzf /precise-engine.tar.gz && \
ln -s /precise-engine/precise-engine /usr/bin/precise-engine && \
rm /precise-engine.tar.gz; \
fi
RUN apt-get install -y flite libttspico-utils
COPY download/kaldi_${BUILD_ARCH}.tar.gz /kaldi.tar.gz
RUN mkdir -p /opt && \
@@ -61,9 +82,6 @@ RUN ldconfig
RUN find / -name bw -exec cp '{}' /usr/bin/ \;
RUN find / -name mllr_solve -exec cp '{}' /usr/bin/ \;
# Install gstreamer and plugins
RUN apt-get install --no-install-recommends --yes gstreamer1.0-tools gstreamer1.0-plugins-good
ENV RHASSPY_APP /usr/share/rhasspy
# Copy script to run
@@ -74,6 +92,8 @@ RUN chmod +x /run.sh
COPY profiles/zh/profile.json \
profiles/zh/custom_words.txt \
profiles/zh/download-profile.sh \
profiles/zh/check-profile.sh \
profiles/zh/espeak_phonemes.txt \
profiles/zh/phoneme_examples.txt \
profiles/zh/frequent_words.txt \
@@ -83,6 +103,8 @@ COPY profiles/zh/profile.json \
COPY profiles/hi/ \
profiles/hi/profile.json \
profiles/hi/custom_words.txt \
profiles/hi/download-profile.sh \
profiles/hi/check-profile.sh \
profiles/hi/espeak_phonemes.txt \
profiles/hi/phoneme_examples.txt \
profiles/hi/frequent_words.txt \
@@ -91,6 +113,8 @@ COPY profiles/hi/ \
COPY profiles/el/profile.json \
profiles/el/custom_words.txt \
profiles/el/download-profile.sh \
profiles/el/check-profile.sh \
profiles/el/espeak_phonemes.txt \
profiles/el/phoneme_examples.txt \
profiles/el/frequent_words.txt \
@@ -99,19 +123,18 @@ COPY profiles/el/profile.json \
COPY profiles/de/profile.json \
profiles/de/custom_words.txt \
profiles/de/download-profile.sh \
profiles/de/check-profile.sh \
profiles/de/espeak_phonemes.txt \
profiles/de/phoneme_examples.txt \
profiles/de/frequent_words.txt \
profiles/de/sentences.ini \
profiles/de/stop_words.txt ${RHASSPY_APP}/profiles/de/
COPY profiles/de/kaldi/custom_words.txt \
profiles/de/kaldi/espeak_phonemes.txt \
profiles/de/kaldi/phoneme_examples.txt \
${RHASSPY_APP}/profiles/de/kaldi/
COPY profiles/it/profile.json \
profiles/it/custom_words.txt \
profiles/it/download-profile.sh \
profiles/it/check-profile.sh \
profiles/it/espeak_phonemes.txt \
profiles/it/phoneme_examples.txt \
profiles/it/frequent_words.txt \
@@ -120,6 +143,8 @@ COPY profiles/it/profile.json \
COPY profiles/es/profile.json \
profiles/es/custom_words.txt \
profiles/es/download-profile.sh \
profiles/es/check-profile.sh \
profiles/es/espeak_phonemes.txt \
profiles/es/phoneme_examples.txt \
profiles/es/frequent_words.txt \
@@ -128,6 +153,8 @@ COPY profiles/es/profile.json \
COPY profiles/fr/profile.json \
profiles/fr/custom_words.txt \
profiles/fr/download-profile.sh \
profiles/fr/check-profile.sh \
profiles/fr/espeak_phonemes.txt \
profiles/fr/phoneme_examples.txt \
profiles/fr/frequent_words.txt \
@@ -136,6 +163,8 @@ COPY profiles/fr/profile.json \
COPY profiles/ru/profile.json \
profiles/ru/custom_words.txt \
profiles/ru/download-profile.sh \
profiles/ru/check-profile.sh \
profiles/ru/espeak_phonemes.txt \
profiles/ru/phoneme_examples.txt \
profiles/ru/frequent_words.txt \
@@ -144,19 +173,18 @@ COPY profiles/ru/profile.json \
COPY profiles/nl/profile.json \
profiles/nl/custom_words.txt \
profiles/nl/download-profile.sh \
profiles/nl/check-profile.sh \
profiles/nl/espeak_phonemes.txt \
profiles/nl/phoneme_examples.txt \
profiles/nl/frequent_words.txt \
profiles/nl/sentences.ini \
profiles/nl/stop_words.txt ${RHASSPY_APP}/profiles/nl/
COPY profiles/nl/kaldi/custom_words.txt \
profiles/nl/kaldi/espeak_phonemes.txt \
profiles/nl/kaldi/phoneme_examples.txt \
${RHASSPY_APP}/profiles/nl/kaldi/
COPY profiles/vi/profile.json \
profiles/vi/custom_words.txt \
profiles/vi/download-profile.sh \
profiles/vi/check-profile.sh \
profiles/vi/espeak_phonemes.txt \
profiles/vi/phoneme_examples.txt \
profiles/vi/frequent_words.txt \
@@ -165,52 +193,32 @@ COPY profiles/vi/profile.json \
COPY profiles/pt/profile.json \
profiles/pt/custom_words.txt \
profiles/pt/download-profile.sh \
profiles/pt/check-profile.sh \
profiles/pt/espeak_phonemes.txt \
profiles/pt/phoneme_examples.txt \
profiles/pt/frequent_words.txt \
profiles/pt/sentences.ini \
profiles/pt/stop_words.txt ${RHASSPY_APP}/profiles/pt/
COPY profiles/sv/profile.json \
profiles/sv/custom_words.txt \
profiles/sv/espeak_phonemes.txt \
profiles/sv/phoneme_examples.txt \
profiles/sv/frequent_words.txt \
profiles/sv/sentences.ini \
profiles/sv/stop_words.txt ${RHASSPY_APP}/profiles/sv/
COPY profiles/ca/profile.json \
profiles/ca/custom_words.txt \
profiles/ca/espeak_phonemes.txt \
profiles/ca/phoneme_examples.txt \
profiles/ca/frequent_words.txt \
profiles/ca/sentences.ini \
profiles/ca/stop_words.txt ${RHASSPY_APP}/profiles/ca/
COPY profiles/en/profile.json \
profiles/en/custom_words.txt \
profiles/en/download-profile.sh \
profiles/en/check-profile.sh \
profiles/en/espeak_phonemes.txt \
profiles/en/phoneme_examples.txt \
profiles/en/frequent_words.txt \
profiles/en/sentences.ini \
profiles/en/stop_words.txt ${RHASSPY_APP}/profiles/en/
COPY profiles/en/kaldi/custom_words.txt \
profiles/en/kaldi/espeak_phonemes.txt \
profiles/en/kaldi/phoneme_examples.txt \
${RHASSPY_APP}/profiles/en/kaldi/
COPY profiles/defaults.json ${RHASSPY_APP}/profiles/
COPY docker/rhasspy ${RHASSPY_APP}/bin/
COPY dist/ ${RHASSPY_APP}/dist/
COPY etc/wav/* ${RHASSPY_APP}/etc/wav/
COPY rhasspy/profile_schema.json ${RHASSPY_APP}/rhasspy/
COPY rhasspy/train/jsgf2fst/*.py ${RHASSPY_APP}/rhasspy/train/jsgf2fst/
COPY rhasspy/train/*.py ${RHASSPY_APP}/rhasspy/train/
COPY *.py ${RHASSPY_APP}/
COPY rhasspy/*.py ${RHASSPY_APP}/rhasspy/
ENV CONFIG_PATH /data/options.json
ENV KALDI_PREFIX /opt
ENTRYPOINT ["/run.sh"]
docker/templates/dockerfiles/Dockerfile.prebuilt.alsa.en
Regular → Executable
+36 -22
View File
@@ -14,25 +14,37 @@ COPY etc/qemu-arm-static /usr/bin/
COPY etc/qemu-aarch64-static /usr/bin/
RUN apt-get update && \
apt-get install --no-install-recommends --yes \
bash jq unzip \
python3 python3-pip python3-dev \
build-essential portaudio19-dev swig \
libfst-dev libfst-tools \
libatlas-base-dev \
sox espeak flite alsa-utils \
git curl \
autoconf libtool automake bison \
sphinxbase-utils sphinxtrain
apt-get install -y bash jq unzip \
python3 python3-pip python3-dev \
build-essential portaudio19-dev swig \
libfst-dev libfst-tools \
libatlas-base-dev \
sox espeak flite alsa-utils \
git curl \
autoconf libtool automake bison \
sphinxbase-utils sphinxtrain
COPY download/rhasspy-tools_${BUILD_ARCH}.tar.gz /
RUN tar -C /usr -xvf /rhasspy-tools_${BUILD_ARCH}.tar.gz && \
rm -f /rhasspy-tools_${BUILD_ARCH}.tar.gz
COPY download/phonetisaurus-2019_${BUILD_ARCH}.deb /phonetisaurus.deb
RUN dpkg -i /phonetisaurus.deb && \
rm /phonetisaurus.deb
# Install Opengrm
COPY download/opengrm-ngram-1.3.3.tar.gz /
RUN cd / && tar -xf opengrm-ngram-1.3.3.tar.gz && \
cd opengrm-ngram-1.3.3 && \
./configure && \
make -j $MAKE_THREADS && \
make install && \
ldconfig && \
rm -rf /opengrm*
# Install Python dependencies
RUN python3 -m pip install --no-cache-dir setuptools wheel
RUN python3 -m pip install --no-cache-dir wheel
RUN apt-get install -y libfreetype6-dev libpng-dev pkg-config libffi-dev libssl-dev
COPY download/jsgf2fst-0.1.0.tar.gz \
/download/
RUN apt-get install -y libfreetype6-dev libpng-dev pkg-config
COPY requirements.txt /requirements.txt
RUN if [ "$BUILD_ARCH" != "amd64" ]; then \
grep -v flair /requirements.txt > /requirements-noflair.txt; \
@@ -49,7 +61,15 @@ RUN python3 -m pip install --no-cache-dir /pocketsphinx-python.tar.gz && \
COPY download/snowboy-1.3.0.tar.gz /
RUN if [ "$BUILD_ARCH" != "aarch64" ]; then pip3 install --no-cache-dir /snowboy-1.3.0.tar.gz; fi
RUN apt-get install --no-install-recommends --yes flite libttspico-utils
# Install Mycroft Precise
COPY download/precise-engine_0.3.0_${CPU_ARCH}.tar.gz /precise-engine.tar.gz
RUN if [ "$BUILD_ARCH" != "aarch64" ]; then \
cd / && tar -xzf /precise-engine.tar.gz && \
ln -s /precise-engine/precise-engine /usr/bin/precise-engine && \
rm /precise-engine.tar.gz; \
fi
RUN apt-get install -y flite libttspico-utils
COPY download/kaldi_${BUILD_ARCH}.tar.gz /kaldi.tar.gz
RUN mkdir -p /opt && \
@@ -62,9 +82,6 @@ RUN ldconfig
RUN find / -name bw -exec cp '{}' /usr/bin/ \;
RUN find / -name mllr_solve -exec cp '{}' /usr/bin/ \;
# Install gstreamer and plugins
RUN apt-get install --no-install-recommends --yes gstreamer1.0-tools gstreamer1.0-plugins-good
ENV RHASSPY_APP /usr/share/rhasspy
# Copy script to run
@@ -80,12 +97,9 @@ COPY docker/rhasspy ${RHASSPY_APP}/bin/
COPY dist/ ${RHASSPY_APP}/dist/
COPY etc/wav/* ${RHASSPY_APP}/etc/wav/
COPY rhasspy/profile_schema.json ${RHASSPY_APP}/rhasspy/
COPY rhasspy/train/jsgf2fst/*.py ${RHASSPY_APP}/rhasspy/train/jsgf2fst/
COPY rhasspy/train/*.py ${RHASSPY_APP}/rhasspy/train/
COPY *.py ${RHASSPY_APP}/
COPY rhasspy/*.py ${RHASSPY_APP}/rhasspy/
ENV CONFIG_PATH /data/options.json
ENV KALDI_PREFIX /opt
ENTRYPOINT ["/run.sh"]
docker/templates/dockerfiles/Dockerfile.prebuilt.pulseaudio.en
Regular → Executable
+36 -22
View File
@@ -14,25 +14,37 @@ COPY etc/qemu-arm-static /usr/bin/
COPY etc/qemu-aarch64-static /usr/bin/
RUN apt-get update && \
apt-get install --no-install-recommends --yes \
bash jq unzip \
python3 python3-pip python3-dev \
build-essential portaudio19-dev swig \
libfst-dev libfst-tools \
libatlas-base-dev \
sox espeak flite alsa-utils \
git curl \
autoconf libtool automake bison \
sphinxbase-utils sphinxtrain
apt-get install -y bash jq unzip \
python3 python3-pip python3-dev \
build-essential portaudio19-dev swig \
libfst-dev libfst-tools \
libatlas-base-dev \
sox espeak flite alsa-utils \
git curl \
autoconf libtool automake bison \
sphinxbase-utils sphinxtrain
COPY download/rhasspy-tools_${BUILD_ARCH}.tar.gz /
RUN tar -C /usr -xvf /rhasspy-tools_${BUILD_ARCH}.tar.gz && \
rm -f /rhasspy-tools_${BUILD_ARCH}.tar.gz
COPY download/phonetisaurus-2019_${BUILD_ARCH}.deb /phonetisaurus.deb
RUN dpkg -i /phonetisaurus.deb && \
rm /phonetisaurus.deb
# Install Opengrm
COPY download/opengrm-ngram-1.3.3.tar.gz /
RUN cd / && tar -xf opengrm-ngram-1.3.3.tar.gz && \
cd opengrm-ngram-1.3.3 && \
./configure && \
make -j $MAKE_THREADS && \
make install && \
ldconfig && \
rm -rf /opengrm*
# Install Python dependencies
RUN python3 -m pip install --no-cache-dir setuptools wheel
RUN python3 -m pip install --no-cache-dir wheel
RUN apt-get install -y libfreetype6-dev libpng-dev pkg-config libffi-dev libssl-dev
COPY download/jsgf2fst-0.1.0.tar.gz \
/download/
RUN apt-get install -y libfreetype6-dev libpng-dev pkg-config
COPY requirements.txt /requirements.txt
RUN if [ "$BUILD_ARCH" != "amd64" ]; then \
grep -v flair /requirements.txt > /requirements-noflair.txt; \
@@ -49,7 +61,15 @@ RUN python3 -m pip install --no-cache-dir /pocketsphinx-python.tar.gz && \
COPY download/snowboy-1.3.0.tar.gz /
RUN if [ "$BUILD_ARCH" != "aarch64" ]; then pip3 install --no-cache-dir /snowboy-1.3.0.tar.gz; fi
RUN apt-get install --no-install-recommends --yes flite libttspico-utils
# Install Mycroft Precise
COPY download/precise-engine_0.3.0_${CPU_ARCH}.tar.gz /precise-engine.tar.gz
RUN if [ "$BUILD_ARCH" != "aarch64" ]; then \
cd / && tar -xzf /precise-engine.tar.gz && \
ln -s /precise-engine/precise-engine /usr/bin/precise-engine && \
rm /precise-engine.tar.gz; \
fi
RUN apt-get install -y flite libttspico-utils
COPY download/kaldi_${BUILD_ARCH}.tar.gz /kaldi.tar.gz
RUN mkdir -p /opt && \
@@ -62,9 +82,6 @@ RUN ldconfig
RUN find / -name bw -exec cp '{}' /usr/bin/ \;
RUN find / -name mllr_solve -exec cp '{}' /usr/bin/ \;
# Install gstreamer and plugins
RUN apt-get install --no-install-recommends --yes gstreamer1.0-tools gstreamer1.0-plugins-good
RUN apt-get install -y pulseaudio
COPY etc/pulseaudio.client.conf /etc/pulse/client.conf
@@ -97,12 +114,9 @@ COPY docker/rhasspy ${RHASSPY_APP}/bin/
COPY dist/ ${RHASSPY_APP}/dist/
COPY etc/wav/* ${RHASSPY_APP}/etc/wav/
COPY rhasspy/profile_schema.json ${RHASSPY_APP}/rhasspy/
COPY rhasspy/train/jsgf2fst/*.py ${RHASSPY_APP}/rhasspy/train/jsgf2fst/
COPY rhasspy/train/*.py ${RHASSPY_APP}/rhasspy/train/
COPY *.py ${RHASSPY_APP}/
COPY rhasspy/*.py ${RHASSPY_APP}/rhasspy/
ENV CONFIG_PATH /data/options.json
ENV KALDI_PREFIX /opt
ENTRYPOINT ["/run.sh"]
docker/templates/from-source/system_dependencies.txt
+9 -10
View File
@@ -1,17 +1,16 @@
RUN apt-get update && \
apt-get install --no-install-recommends --yes \
bash jq unzip \
python3 python3-pip python3-dev \
build-essential portaudio19-dev swig \
libatlas-base-dev \
sox espeak flite alsa-utils \
git curl \
autoconf libtool automake bison \
sphinxbase-utils sphinxtrain
apt-get install -y bash jq unzip \
python3 python3-pip python3-dev \
build-essential portaudio19-dev swig \
libatlas-base-dev \
sox espeak flite alsa-utils \
git curl \
autoconf libtool automake bison \
sphinxbase-utils sphinxtrain
COPY download/openfst-1.6.2-1_${BUILD_ARCH}.deb /
RUN if [ "$BUILD_ARCH" != "aarch64" ]; then \
apt-get install --no-install-recommends --yes libfst-dev libfst-tools; \
apt-get install -y libfst-dev libfst-tools; \
else \
dpkg -i /openfst-1.6.2-1_${BUILD_ARCH}.deb; \
rm /openfst*.deb; \
docker/templates/generate.sh
+15 -15
View File
@@ -39,28 +39,28 @@ function set_variables {
#------------
# From source
#------------
# set_variables "$DIR/shared/" "$DIR/from-source/" \
# "$DIR/alsa/" "$DIR/en_profile/" \
# | cat - "$template" | m4 > "$out/Dockerfile.from-source.alsa.en"
set_variables "$DIR/shared/" "$DIR/from-source/" \
"$DIR/alsa/" "$DIR/en_profile/" \
| cat - "$template" | m4 > "$out/Dockerfile.from-source.alsa.en"
# set_variables "$DIR/shared/" "$DIR/from-source/" \
# "$DIR/pulseaudio/" "$DIR/en_profile/" \
# | cat - "$template" | m4 > "$out/Dockerfile.from-source.pulseaudio.en"
set_variables "$DIR/shared/" "$DIR/from-source/" \
"$DIR/pulseaudio/" "$DIR/en_profile/" \
| cat - "$template" | m4 > "$out/Dockerfile.from-source.pulseaudio.en"
# set_variables "$DIR/shared/" "$DIR/from-source/" \
# "$DIR/alsa/" "$DIR/all_profiles/" \
# | cat - "$template" | m4 > "$out/Dockerfile.from-source.alsa.all"
set_variables "$DIR/shared/" "$DIR/from-source/" \
"$DIR/alsa/" "$DIR/all_profiles/" \
| cat - "$template" | m4 > "$out/Dockerfile.from-source.alsa.all"
#-----------
# Pre-built
#-----------
# set_variables "$DIR/shared/" "$DIR/prebuilt/" \
# "$DIR/alsa/" "$DIR/en_profile/" \
# | cat - "$template" | m4 > "$out/Dockerfile.prebuilt.alsa.en"
set_variables "$DIR/shared/" "$DIR/prebuilt/" \
"$DIR/alsa/" "$DIR/en_profile/" \
| cat - "$template" | m4 > "$out/Dockerfile.prebuilt.alsa.en"
# set_variables "$DIR/shared/" "$DIR/prebuilt/" \
# "$DIR/pulseaudio/" "$DIR/en_profile/" \
# | cat - "$template" | m4 > "$out/Dockerfile.prebuilt.pulseaudio.en"
set_variables "$DIR/shared/" "$DIR/prebuilt/" \
"$DIR/pulseaudio/" "$DIR/en_profile/" \
| cat - "$template" | m4 > "$out/Dockerfile.prebuilt.pulseaudio.en"
set_variables "$DIR/shared/" "$DIR/prebuilt/" \
"$DIR/alsa/" "$DIR/all_profiles/" \
docker/templates/prebuilt/rhasspy_tools.txt
-3
View File
@@ -1,3 +0,0 @@
COPY download/rhasspy-tools_${BUILD_ARCH}.tar.gz /
RUN tar -C /usr -xvf /rhasspy-tools_${BUILD_ARCH}.tar.gz && \
rm -f /rhasspy-tools_${BUILD_ARCH}.tar.gz
docker/templates/prebuilt/system_dependencies.txt
+9 -9
View File
@@ -1,10 +1,10 @@
RUN apt-get update && \
apt-get install --no-install-recommends --yes \
bash jq unzip \
python3 python3-pip python3-dev \
build-essential portaudio19-dev swig \
libatlas-base-dev \
sox espeak flite alsa-utils \
git curl \
autoconf libtool automake bison \
sphinxbase-utils sphinxtrain
apt-get install -y bash jq unzip \
python3 python3-pip python3-dev \
build-essential portaudio19-dev swig \
libfst-dev libfst-tools \
libatlas-base-dev \
sox espeak flite alsa-utils \
git curl \
autoconf libtool automake bison \
sphinxbase-utils sphinxtrain
docker/templates/shared/gstreamer.txt
-2
View File
@@ -1,2 +0,0 @@
# Install gstreamer and plugins
RUN apt-get install --no-install-recommends --yes gstreamer1.0-tools gstreamer1.0-plugins-good
docker/templates/shared/mitlm.txt
+1 -1
View File
@@ -1,5 +1,5 @@
# Install mitlm
RUN apt-get install --no-install-recommends --yes gfortran
RUN apt-get install -y gfortran
COPY download/mitlm-0.4.2.tar.xz /
RUN cd / && tar -xf mitlm-0.4.2.tar.xz && cd mitlm-0.4.2/ && \
./configure && \
docker/templates/shared/python_requirements.txt
+5 -2
View File
@@ -1,7 +1,10 @@
# Install Python dependencies
RUN python3 -m pip install --no-cache-dir setuptools wheel
RUN python3 -m pip install --no-cache-dir wheel
RUN apt-get install -y libfreetype6-dev libpng-dev pkg-config libffi-dev libssl-dev
COPY download/jsgf2fst-0.1.0.tar.gz \
/download/
RUN apt-get install -y libfreetype6-dev libpng-dev pkg-config
COPY requirements.txt /requirements.txt
RUN if [ "$BUILD_ARCH" != "amd64" ]; then \
grep -v flair /requirements.txt > /requirements-noflair.txt; \
docker/templates/shared/rhasspy_code.txt
-2
View File
@@ -3,7 +3,5 @@ COPY docker/rhasspy ${RHASSPY_APP}/bin/
COPY dist/ ${RHASSPY_APP}/dist/
COPY etc/wav/* ${RHASSPY_APP}/etc/wav/
COPY rhasspy/profile_schema.json ${RHASSPY_APP}/rhasspy/
COPY rhasspy/train/jsgf2fst/*.py ${RHASSPY_APP}/rhasspy/train/jsgf2fst/
COPY rhasspy/train/*.py ${RHASSPY_APP}/rhasspy/train/
COPY *.py ${RHASSPY_APP}/
COPY rhasspy/*.py ${RHASSPY_APP}/rhasspy/
docker/templates/shared/tts.txt
+1 -1
View File
@@ -1 +1 @@
RUN apt-get install --no-install-recommends --yes flite libttspico-utils
RUN apt-get install -y flite libttspico-utils
docs/_config.yml
-1
View File
@@ -1 +0,0 @@
theme: jekyll-theme-cayman
docs/about.md
+1 -2
View File
@@ -28,13 +28,12 @@ The following tools/libraries help to support Rhasspy:
* [Phonetisaurus](https://github.com/AdolfVonKleist/Phonetisaurus) (word pronunciations)
* [PicoTTS](https://en.wikipedia.org/wiki/SVOX) (text to speech)
* [Pocketsphinx](https://github.com/cmusphinx/pocketsphinx) (speech to text, wake word)
* [porcupine](https://github.com/Picovoice/Porcupine) (wake word)
* [PyAudio](https://people.csail.mit.edu/hubert/pyaudio/) (microphone)
* [pyjsgf](https://github.com/Danesprite/pyjsgf) (JSGF grammar parsing)
* [Python 3](https://www.python.org)
* [OpenFST](http://www.openfst.org) (intent recognition)
* [Opengrm](http://www.opengrm.org/twiki/bin/view/GRM/NGramLibrary) (language modeling)
* [Rasa NLU](https://rasa.com/) (intent recognition)
* [RasaNLU](https://rasa.com/) (intent recognition)
* [sphinxtrain](https://github.com/cmusphinx/sphinxtrain) (acoustic model tuning)
* [snowboy](https://snowboy.kitt.ai) (wake word)
* [Sox](http://sox.sourceforge.net) (WAV conversion)
docs/audio-input.md
+1 -61
View File
@@ -54,7 +54,7 @@ See `rhasspy.audio_recorder.ARecordAudioRecorder` for details.
Listens to the `hermes/audioServer/<SITE_ID>/audioFrame` topic for WAV data ([Hermes protocol](https://docs.snips.ai/ressources/hermes-protocol)).
This allows Rhasspy to receive audio from [Snips.AI](https://snips.ai/).
Audio data is automatically converted to 16-bit, 16 kHz mono with [sox](http://sox.sourceforge.net).
Audio data is automatically converted to 16-bit, 16Khz mono with [sox](http://sox.sourceforge.net).
Add to your [profile](profiles.md):
@@ -78,66 +78,6 @@ Set `mqtt.site_id` to match your Snips.AI siteId.
See `rhasspy.audio_recorder.HermesAudioRecorder` for details.
## HTTP Stream
Accepts chunks of 16-bit 16Khz mono audio via an HTTP POST stream (assumes [chunked transfer encoding](https://en.wikipedia.org/wiki/Chunked_transfer_encoding)).
Add to your [profile](profiles.md):
```json
"microphone": {
"system": "http",
"http": {
"host": "127.0.0.1",
"port": 12333,
"stop_after": "never"
}
}
```
Set `microphone.http.stop_after` to one of "never", "text", or "intent". When set to "never", you can continously stream (chunked) audio into Rhasspy across multiple voice commands. When set to "text" or "intent", the stream will be closed when the first voice command has been transcribed ("text") or recognized ("intent"). Once closed, you can perform an HTTP GET request to the stream URL to retrieve the result (text for transcriptions or JSON for intent).
Note that `microphone.http.port` must be different than Rhasspy's webserver port (usually 12101).
See `rhasspy.audio_recorder.HTTPAudioRecorder` for details.
## GStreamer
Receives audio chunks via stdout from a [GStreamer](https://gstreamer.freedesktop.org/) pipeline.
Add to your [profile](profiles.md):
```json
"microphone": {
"system": "gstreamer",
"gstreamer": {
"pipeline": "...",
}
}
```
Set `microphone.gstreamer.pipeline` to your GStreamer pipeline **without a sink** (this will be added by Rhasspy). By default, the pipeline is:
```
udpsrc port=12333 ! rawaudioparse use-sink-caps=false format=pcm pcm-format=s16le sample-rate=16000 num-channels=1 ! queue ! audioconvert ! audioresample
```
which "simply" receives raw 16-bit 16khz audio chunks via UDP port 12333. You could stream microphone audio to Rhasspy from another machine by running the following terminal command:
```bash
gst-launch-1.0 \
autoaudiosrc ! \
audioconvert ! \
audioresample ! \
audio/x-raw, rate=16000, channels=1, format=S16LE ! \
udpsink host=RHASSPY_SERVER port=12333
```
where `RHASSPY_SERVER` is the hostname of your Rhasspy server (e.g., `localhost`).
The Rhasspy Docker images contains the ["good" plugin](https://gstreamer.freedesktop.org/data/doc/gstreamer/head/gst-plugins-good-plugins/html/) set for GStreamer, which includes a wide variety of ways to stream/transform audio.
See `rhasspy.audio_recorder.GStreamerAudioRecorder` for details.
## Dummy
docs/audio-output.md
+1 -1
View File
@@ -27,7 +27,7 @@ See `rhasspy.audio_player.APlayAudioPlayer` for details.
Publishes WAV data to the `hermes/audioServer/<SITE_ID>/playBytes/<REQUEST_ID>` topic ([Hermes protocol](https://docs.snips.ai/ressources/hermes-protocol)).
This allows Rhasspy to send audio to [Snips.AI](https://snips.ai/).
Rhasspy will always try to send 16 kHz, 16-bit mono audio.
Rhasspy will always try to send 16Khz, 16-bit mono audio.
The request id is generated each time a sound is played using `uuid.uuid4`.
Add to your [profile](profiles.md):
docs/command-listener.md
+1 -1
View File
@@ -124,7 +124,7 @@ Add to your [profile](profiles.md):
When awake, Rhasspy normally listens for voice commands from the microphone and waits for silence by using [webrtcvad](https://github.com/wiseman/py-webrtcvad). You can call a custom program that will listen for a voice command and simply return the recorded WAV audio data to Rhasspy.
When Rhasspy wakes up, your program will be called with the given arguments. The program's output should be WAV data with the recorded voice command (Rhasspy will automatically convert this to 16-bit 16 kHz mono if necessary).
When Rhasspy wakes up, your program will be called with the given arguments. The program's output should be WAV data with the recorded voice command (Rhasspy will automatically convert this to 16-bit 16Khz mono if necessary).
The following environment variables are available to your program:
docs/hardware.md
-21
View File
@@ -5,27 +5,6 @@ Rhasspy is designed to be run on different kinds of hardware, such as:
* Raspberry Pi 2-3 B/B+ (`armhf`/`aarch64`)
* Desktop/laptop/server (`amd64`)
The table below summarizes architecture compatibility with Rhasspy's components:
| Category | Name | amd64 | armhf | aarch64 |
| -------- | ------ | ------- | ------- | ------- |
| **Wake Word** | [pocketsphinx](wake-word.md#pocketsphinx) | &#x2713; | &#x2713; | &#x2713; |
| | [snowboy](wake-word.md#snowboy) | &#x2713; | &#x2713; | |
| | [precise](wake-word.md#mycroft-precise) | &#x2713; | &#x2713; | |
| | [porcupine](wake-word.md#porcupine) | &#x2713; | &#x2713; | &#x2713; |
| **Speech to Text** | [pocketsphinx](speech-to-text.md#pocketsphinx) | &#x2713; | &#x2713; | &#x2713; |
| | [kaldi](speech-to-text.md#kaldi) | &#x2713; | &#x2713; | &#x2713; |
| **Intent Recognition** | [fsticuffs](intent-recognition.md#fsticuffs) | &#x2713; | &#x2713; | &#x2713; |
| | [fuzzywuzzy](intent-recognition.md#fuzzywuzzy) | &#x2713; | &#x2713; | &#x2713; |
| | [adapt](intent-recognition.md#mycroft-adapt) | &#x2713; | &#x2713; | &#x2713; |
| | [flair](intent-recognition.md#flair) | &#x2713; | | |
| | [rasaNLU](intent-recognition.md#rasanlu) | &#x2713; | &#x2713; | &#x2713; |
| **Text to Speech** | [espeak](text-to-speech.md#espeak) | &#x2713; | &#x2713; | &#x2713; |
| | [flite](text-to-speech.md#flite) | &#x2713; | &#x2713; | &#x2713; |
| | [picotts](text-to-speech.md#picotts) | &#x2713; | &#x2713; | &#x2713; |
| | [marytts](text-to-speech.md#marytts) | &#x2713; | &#x2713; | &#x2713; |
| | [wavenet](text-to-speech.md#google-wavenet) | &#x2713; | &#x2713; | &#x2713; |
## Raspberry Pi
To run Rhasspy on a Raspberry Pi, you'll need at least a 4 GB SD card and a good power supply. I highly recommend the [CanaKit Starter Kit](https://www.amazon.com/CanaKit-Raspberry-Starter-Premium-Black/dp/B07BCC8PK7), which includes a 32 GB SD card, a 2.5 A power supply, and a case.
docs/index.md
+21 -28
View File
@@ -1,6 +1,6 @@
![Rhasspy logo](img/rhasspy.svg)
Rhasspy (pronounced RAH-SPEE) is an [open source](https://github.com/synesthesiam/rhasspy), fully offline voice assistant toolkit for [many languages](#supported-languages) that works well with [Home Assistant](https://www.home-assistant.io/), [Hass.io](https://www.home-assistant.io/hassio/), and [Node-RED](https://nodered.org).
Rhasspy (pronounced RAH-SPEE) is an offline, [multilingual](#supported-languages) voice assistant toolkit inspired by [Jasper](https://jasperproject.github.io/) that works well with [Home Assistant](https://www.home-assistant.io/), [Hass.io](https://www.home-assistant.io/hassio/), and [Node-RED](https://nodered.org).
Rhasspy transforms voice commands into [JSON](https://json.org) events that can trigger actions in home automation software, like [Home Assistant automations](https://www.home-assistant.io/docs/automation/trigger/#event-trigger) or [Node-RED flows](usage.md#node-red). You define custom voice commands in a [profile](profiles.md) using a [specialized template syntax](training.md), and Rhasspy takes care of the rest.
@@ -20,7 +20,6 @@ Rhasspy provides **offline, private solutions** to problems 1-4 using off-the-sh
* [Pocketsphinx keyphrase](https://cmusphinx.github.io/wiki/tutoriallm/#using-keyword-lists-with-pocketsphinx)
* [Mycroft Precise](https://github.com/MycroftAI/mycroft-precise)
* [snowboy](https://snowboy.kitt.ai)
* [porcupine](https://github.com/Picovoice/Porcupine)
* **Command listener**
* [webrtcvad](https://github.com/wiseman/py-webrtcvad)
* **Speech to text**
@@ -30,16 +29,13 @@ Rhasspy provides **offline, private solutions** to problems 1-4 using off-the-sh
* [OpenFST](https://www.openfst.org)
* [fuzzywuzzy](https://github.com/seatgeek/fuzzywuzzy)
* [Mycroft Adapt](https://github.com/MycroftAI/adapt)
* [flair](http://github.com/zalandoresearch/flair)
* [Rasa NLU](https://rasa.com/)
* [RasaNLU](https://rasa.com/)
For problem 5 (fulfilling the speaker's intent), Rhasspy works with external home automation software, such as Home Assistant's built-in [automation capability](https://www.home-assistant.io/docs/automation/) or a [Node-RED flow](https://nodered.org).
For each intent you define, Rhasspy emits a JSON event that can do anything Home Assistant can do (toggle switches, call REST services, etc.). This means that Rhasspy will do very little out of the box compared to other voice assistants, but there are also be *no limits* to what can be done.
For problem 5 (fulfilling the speaker's intent), Rhasspy works with external home automation software, such as Home Assistant's built-in [automation capability](https://www.home-assistant.io/docs/automation/) or a [Node-RED flow](https://nodered.org). For each intent you define, Rhasspy emits a JSON event that can, for example, be used to do anything Home Assistant can do (toggle switches, call REST services, etc.). This means that Rhasspy will do very little out of the box compared to other voice assistants, but there are also be *no limits* to what can be done.
## Supported Languages
Rhasspy supports the following languages:
Rhasspy currently supports the following languages:
* English (`en`)
* German (`de`)
@@ -53,8 +49,6 @@ Rhasspy supports the following languages:
* Mandarin (`zh`)
* Vietnamese (`vi`)
* Portuguese (`pt`)
* Swedish (`sv`)
* Catalan (`ca`)
Support for these languages comes directly from existing [CMU Sphinx](https://sourceforge.net/projects/cmusphinx/files/Acoustic%20and%20Language%20Models/) and [Kaldi](https://montreal-forced-aligner.readthedocs.io/en/latest/pretrained_models.html) acoustic models.
@@ -66,24 +60,23 @@ It is possible to extend Rhasspy to new languages with only:
The table below summarizes language support across the various supporting technologies that Rhasspy uses:
| Category | Name | Offline? | en | de | es | fr | it | nl | ru | el | hi | zh | vi | pt | sv | ca |
| -------- | ------ | -------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- |
| **Wake Word** | [pocketsphinx](wake-word.md#pocketsphinx) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | | | | |
| | [porcupine](wake-word.md#porcupine) | &#x2713; | &#x2713; | | | | | | | | | | | | | |
| | [snowboy](wake-word.md#snowboy) | *requires account* | &#x2713; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; |
| | [precise](wake-word.md#mycroft-precise) | &#x2713; | &#x2713; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; |
| **Speech to Text** | [pocketsphinx](speech-to-text.md#pocketsphinx) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | | &#x2713; | | &#x2713; |
| | [kaldi](speech-to-text.md#kaldi) | &#x2713; | | | | | | | | | | | &#x2713; | | &#x2713; | |
| **Intent Recognition** | [fsticuffs](intent-recognition.md#fsticuffs) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; |
| | [fuzzywuzzy](intent-recognition.md#fuzzywuzzy) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; |
| | [adapt](intent-recognition.md#mycroft-adapt) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; |
| | [flair](intent-recognition.md#flair) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | | &#x2713; | | | | | | &#x2713; | | &#x2713; |
| | [rasaNLU](intent-recognition.md#rasanlu) | *needs extra software* | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; |
| **Text to Speech** | [espeak](text-to-speech.md#espeak) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; |
| | [flite](text-to-speech.md#flite) | &#x2713; | &#x2713; | | | | | | | | &#x2713; | | | | | |
| | [picotts](text-to-speech.md#picotts) | &#x2713; | &#x2713; | | | | | | | | | | | | | |
| | [marytts](text-to-speech.md#marytts) | &#x2713; | &#x2713; | &#x2713; | | &#x2713; | &#x2713; | | &#x2713; | | | | | | | |
| | [wavenet](text-to-speech.md#google-wavenet) | | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | | &#x2713; | &#x2713; | | &#x2713; | &#x2713; | |
| Category | Name | Offline? | en | de | es | fr | it | nl | ru | el | hi | zh | vi | pt |
| -------- | ------ | -------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- |
| **Wake Word** | [pocketsphinx](wake-word.md#pocketsphinx) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | | |
| | [snowboy](wake-word.md#snowboy) | *requires account* | &#x2713; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; |
| | [precise](wake-word.md#mycroft-precise) | &#x2713; | &#x2713; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; | &bull; |
| **Speech to Text** | [pocketsphinx](speech-to-text.md#pocketsphinx) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | | |
| | [kaldi](speech-to-text.md#kaldi) | &#x2713; | | | | | | | | | | | &#x2713; | &#x2713; |
| **Intent Recognition** | [fsticuffs](intent-recognition.md#fsticuffs) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; |
| | [fuzzywuzzy](intent-recognition.md#fuzzywuzzy) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; |
| | [adapt](intent-recognition.md#mycroft-adapt) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; |
| | [flair](intent-recognition.md#flair) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | | &#x2713; | | | | | | &#x2713; |
| | [rasaNLU](intent-recognition.md#rasanlu) | *needs extra software* | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; |
| **Text to Speech** | [espeak](text-to-speech.md#espeak) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; |
| | [flite](text-to-speech.md#flite) | &#x2713; | &#x2713; | | | | | | | | &#x2713; | | | |
| | [picotts](text-to-speech.md#picotts) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; |
| | [marytts](text-to-speech.md#marytts) | &#x2713; | &#x2713; | &#x2713; | | &#x2713; | &#x2713; | | &#x2713; | | | | | |
| | [wavenet](text-to-speech.md#google-wavenet) | | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | | &#x2713; | &#x2713; | | &#x2713; |
&bull; - yes, but requires training/customization
docs/installation.md
+4 -4
View File
@@ -3,7 +3,7 @@
Rhasspy should run in a variety of software environments, including:
* Within a [Docker](https://www.docker.com/) container
* As a [Hass.io add-on](https://www.home-assistant.io/addons/)
* As a [Hass.IO add-on](https://www.home-assistant.io/addons/)
* Inside a [Python virtual environment](https://docs.python-guide.org/dev/virtualenvs/)
### Docker
@@ -45,9 +45,9 @@ If you're using [docker compose](https://docs.docker.com/compose/), add the foll
- "/dev/snd:/dev/snd"
command: --user-profiles /profiles --profile en
### Hass.io
### Hass.IO
The second easiest was to install Rhasspy is as a [Hass.io add-on](https://www.home-assistant.io/addons/). Following the [installation instructions for Hass.io](https://www.home-assistant.io/hassio/installation/) before proceeding.
The second easiest was to install Rhasspy is as a [Hass.IO add-on](https://www.home-assistant.io/addons/). Following the [installation instructions for Hass.IO](https://www.home-assistant.io/hassio/installation/) before proceeding.
To install the add-on, add my [Hass.IO Add-On Repository](https://github.com/synesthesiam/hassio-addons) in the Add-On Store, refresh, then install the "Rhasspy Assistant" under “Synesthesiam Hass.IO Add-Ons” (all the way at the bottom of the Add-On Store screen).
@@ -59,7 +59,7 @@ Watch the system log for a message like `Build 8e35c251/armhf-addon-rhasspy:2.11
Before starting the add-on, make sure to give it access to your microphone and speakers:
![Audio settings for Hass.io](img/hass-io-audio.png)
![Audio settings for Hass.IO](img/hass-io-audio.png)
### Virtual Environment
docs/intent-recognition.md
+4 -30
View File
@@ -24,14 +24,11 @@ Add to your [profile](profiles.md):
"system": "fsticuffs",
"fsticuffs": {
"intent_fst": "intent.fst",
"ignore_unknown_words": true,
"fuzzy": true
"ignore_unknown_words": true
}
}
```
By default, fuzzy mathing is enabled (`fuzzy` is true). This allows `fsticuffs` to be less strict when matching text, skipping over any words in `stop_words.txt`, and handling repeated words gracefully. Words must still appear in the correct order according to `sentences.ini`, but additional words will not cause a recognition failure.
When `ignore_unknown_words` is true, any word outside of `sentences.ini` is simply ignored. This allows a lot more sentences to be accepted, but may cause unexpected results when used with arbitrary input from text chat.
See `rhasspy.intent.FsticuffsRecognizer` for details.
@@ -98,7 +95,7 @@ See `rhasspy.intent.FlairRecognizer` for details.
## RasaNLU
Recognizes intents **remotely** using a [Rasa NLU](https://rasa.com/) server. You must [install a Rasa NLU server](https://rasa.com/docs/rasa/user-guide/installation/) somewhere that Rhasspy can access. Works well when you have a large number of sentences (thousands to hundreds of thousands) and need to handle sentences *and* words not seen during training. This needs Rasa 1.0 or higher.
Recognizes intents **remotely** using a [rasaNLU](https://rasa.com/) server. You must [install a rasaNLU server](https://rasa.com/docs/nlu/installation) somewhere that Rhasspy can access. Works well when you have a large number of sentences (thousands to hundreds of thousands) and need to handle sentences *and* words not seen during training.
Add to your [profile](profiles.md):
@@ -108,7 +105,7 @@ Add to your [profile](profiles.md):
"rasa": {
"examples_markdown": "intent_examples.md",
"project_name": "rhasspy",
"url": "http://localhost:5005/"
"url": "http://localhost:5000/"
}
}
```
@@ -132,29 +129,6 @@ Add to your [profile](profiles.md):
See `rhasspy.intent.RemoteRecognizer` for details.
## Home Assistant Conversation
Sends transcriptions from [speech to text](speech-to-text.md) to [Home Assistant's conversation API](https://www.home-assistant.io/integrations/conversation/). If the response contains speech, Rhasspy can optionally speak it.
Add to your [profile](profiles.md):
```json
"intent": {
"system": "conversation",
"conversation": {
"handle_speech": true
}
}
```
When `handle_speech` is `true`, Rhasspy will forward the returned speech to your [text to speech](text-to-speech.md) system.
The settings from your profile's `home_assistant` section are automatically used (URL, access token, etc.).
Because Home Assistant will already handle your intent (probably using an [intent script](https://www.home-assistant.io/integrations/intent_script/)), Rhasspy will always generate an empty intent with this recognizer.
See `rhasspy.intent.HomeAssistantConversationRecognizer` for details.
## Command
Recognizes intents from text using a custom external program.
@@ -171,7 +145,7 @@ Add to your [profile](profiles.md):
}
```
Rhasspy recognizes intents from text using one of several systems, such as [fuzzywuzzy](https://github.com/seatgeek/fuzzywuzzy) or [Rasa NLU](https://rasa.com/). You can call a custom program that does intent recognition from a text command.
Rhasspy recognizes intents from text using one of several systems, such as [fuzzywuzzy](https://github.com/seatgeek/fuzzywuzzy) or [rasaNLU](https://rasa.com/). You can call a custom program that does intent recognition from a text command.
When a voice command is successfully transcribed, your program will be called with the text transcription printed to standard in. Your program should return JSON on standard out, something like:
docs/profiles.md
+5 -42
View File
@@ -23,7 +23,7 @@ The default location for each of these directories is:
* User profile location is `$HOME/.config/rhasspy/profiles`
* Docker
* System profile location is either `/usr/share/rhasspy/profiles` (ALSA) or `/home/rhasspy/profiles` (PulseAudio)
* User profile location **must** be explicitly set and mapped to a volume:
* User profile location **must** be explicity set and mapped to a volume:
* `docker run ... -v /path/to/profiles:/profiles synesthesiam/rhasspy-server --user-profiles /profiles`
### Example
@@ -73,8 +73,6 @@ All available profile sections and settings are listed below:
* `dictionary` - text file with all words/pronunciations needed for example sentences
* `unknown_words` - small text file with guessed word pronunciations (from phonetisaurus)
* `language_model` - text file with trigram [ARPA language model](https://cmusphinx.github.io/wiki/arpaformat/) built from example sentences
* `open_transcription` - true if general language model should be used (custom voices commands ignored)
* `base_language_model` - large general language model (read only)
* `mllr_matrix` - MLLR matrix from [acoustic model tuning](https://cmusphinx.github.io/wiki/tutorialtuning/)
* `mix_weight` - how much of the base language model to [mix in during training](training.md#language-model-mixing) (0-1)
* `mix_fst` - path to save mixed ngram FST model
@@ -83,11 +81,9 @@ All available profile sections and settings are listed below:
* `kaldi_dir` - absolute path to Kaldi root directory
* `model_dir` - directory where Kaldi model is stored (relative to profile directory)
* `graph` - directory where HCLG.fst is located (relative to `model_dir`)
* `base_graph` - directory where large general HCLG.fst is located (relative to `model_dir`)
* `base_dictionary` - large text file with word pronunciations (read only)
* `custom_words` - small text file with words/pronunciations added by user
* `dictionary` - text file with all words/pronunciations needed for example sentences
* `open_transcription` - true if general language model should be used (custom voices commands ignored)
* `unknown_words` - small text file with guessed word pronunciations (from phonetisaurus)
* `mix_weight` - how much of the base language model to [mix in during training](training.md#language-model-mixing) (0-1)
* `mix_fst` - path to save mixed ngram FST model
@@ -107,14 +103,13 @@ All available profile sections and settings are listed below:
* `fsticuffs` - configuration for [OpenFST-based](https://www.openfst.org) intent recognizer
* `intent_fst` - path to generated finite state transducer with all intents combined
* `ignore_unknown_words` - true if words not in the FST symbol table should be ignored
* `fuzzy` - true if text is matching in a fuzzy manner, skipping words in `stop_words.txt`
* `fuzzywuzzy` - configuration for simplistic [Levenshtein distance](https://en.wikipedia.org/wiki/Levenshtein_distance) based intent recognizer
* `examples_json` - JSON file with intents/example sentences
* `min_confidence` - minimum confidence required for intent to be converted to a JSON event (0-1)
* `remote` - configuration for remote Rhasspy server
* `url` - URL to POST text to for intent recognition (e.g., `http://your-rhasspy-server:12101/api/text-to-intent`)
* `rasa` - configuration for [Rasa NLU](https://rasa.com/) based intent recognizer
* `url` - URL of remote Rasa NLU server (e.g., `http://localhost:5005/`)
* `rasa` - configuration for [rasaNLU](https://rasa.com/) based intent recognizer
* `url` - URL of remote rasaNLU server (e.g., `http://localhost:5000/`)
* `examples_markdown` - Markdown file to generate with intents/example sentences
* `project_name` - name of project to generate during training
* `adapt` - configuration for [Mycroft Adapt](https://github.com/MycroftAI/adapt) based intent recognizer
@@ -134,15 +129,6 @@ All available profile sections and settings are listed below:
* `url` - address:port of MaryTTS server (port is usually 59125)
* `voice` - name of voice to use (e.g., `cmu-slt`). Default if not present.
* `locale` - name of locale to use (e.g., `en-US`). Default if not present.
* `wavenet` - configuration for Google's [WaveNet](https://cloud.google.com/text-to-speech/docs/wavenet)
* `cache_dir` - path to directory in your profile where WAV files are cached
* `credentials_json` - path to the JSON credentials file (generated online)
* `gender` - gender of speaker (`MALE` `FEMALE`)
* `language_code` - language/locale e.g. `en-US`,
* `sample_rate` - WAV sample rate (default: 22050)
* `url` - URL of WaveNet endpoint
* `voice` - voice to use (e.g., `Wavenet-C`)
* `fallback_tts` - text to speech system to use when offline or error occurs (e.g., `espeak`)
* `phoneme_examples` - text file with examples for each CMU phoneme
* `training` - training speech/intent recognizers
* `dictionary_number_duplicates` - true if duplicate words in dictionary should be suffixed by `(2)`, `(3)`, etc.
@@ -164,7 +150,7 @@ All available profile sections and settings are listed below:
* `program` - path to executable
* `arguments` - list of arguments to pass to program
* `wake` - waking Rhasspy up for speech input
* `system` - wake word recognition system (`pocketsphinx`, `snowboy`, `precise`, `porcupine`, `command`, or `dummy`)
* `system` - wake word recognition system (`pocketsphinx`, `snowboy`, `precise`, `command`, or `dummy`)
* `pocketsphinx` - configuration for Pocketsphinx wake word recognizer
* `keyphrase` - phrase to wake up on (3-4 syllables recommended)
* `threshold` - sensitivity of detection (recommended range 1e-50 to 1e-5)
@@ -180,28 +166,17 @@ All available profile sections and settings are listed below:
* `sensitivity` - model sensitivity (0-1, default 0.5)
* `trigger_level` - number of events to trigger activation (default 3)
* `chunk_size` - number of bytes per chunk to feed to Precise (default 2048)
* `porcupine` - configuration for [PicoVoice's Porcupine](https://github.com/Picovoice/Porcupine)
* `library_path` - path to `libpv_porcupine.so` for your platform/architecture
* `model_path` - path to the `porcupine_params.pv` (lib/common)
* `keyword_path` - path to the `.ppn` keyword file
* `sensitivity` - model sensitivity (0-1, default 0.5)
* `command` - configuration for external speech-to-text program
* `program` - path to executable
* `arguments` - list of arguments to pass to program
* `microphone` - configuration for audio recording
* `system` - audio recording system (`pyaudio`, `arecord`, `hermes`, `http`, or `dummy`)
* `system` - audio recording system (`pyaudio`, `arecord`, `hermes`, or `dummy`)
* `pyaudio` - configuration for [PyAudio](https://people.csail.mit.edu/hubert/pyaudio/) microphone
* `device` - index of device to use or empty for default device
* `frames_per_buffer` - number of frames to read at a time (default 480)
* `arecord` - configuration for ALSA microphone
* `device` - name of ALSA device (see `arecord -L`) to use or empty for default device
* `chunk_size` - number of bytes to read at a time (default 960)
* `http` - configuration for HTTP audio stream
* `host` - hostname or IP address of HTTP audio server (default 127.0.0.1)
* `port` - port to receive audio stream on (default 12333)
* `stop_after` - one of "never", "text", or "intent" ([see documentation](audio-input.md#http-stream))
* `gstreamer` - configuration for GStreamer audio recorder
* `pipeline` - GStreamer pipeline (e.g., `FILTER ! FILTER ! ...`) without sink
* `hermes` - configuration for MQTT "microphone" ([Hermes protocol](https://docs.snips.ai/ressources/hermes-protocol))
* Subscribes to WAV data from `hermes/audioServer/<SITE_ID>/audioFrame`
* Requires MQTT to be enabled
@@ -251,15 +226,3 @@ All available profile sections and settings are listed below:
* `system` - system for tuning (currently only `sphinxtrain`)
* `sphinxtrain` - configuration for [sphinxtrain](https://github.com/cmusphinx/sphinxtrain) based acoustic model tuning
* `mllr_matrix` - name of generated MLLR matrix (should match `speech_to_text.pocketsphinx.mllr_matrix`)
* `download` - configuration for profile file downloading
* `cache_dir` - directory in your profile where downloaded files are cached
* `conditions` - profile settings that will trigger file downloads
* keys are profile setting paths (e.g., `wake.system`)
* values are dictionaries whose keys are profile settings values (e.g., `snowboy`)
* settings may have the form `<=N` or `!X` to mean "less than or equal to N" or "not X"
* leaf nodes are dictionaries whose keys are destination file paths and whose values reference the `files` dictionary
* `files` - locations, etc. of files to download
* keys are names of files
* values are dictionaries with:
* `url` - URL of file to download
* `cache` - `false` if file should be downloaded directly into profile (skipping cache)
docs/speech-to-text.md
+12 -13
View File
@@ -4,10 +4,10 @@ Rhasspy's primary function is convert voice commands to JSON events. The first s
The following table summarizes language support for the various speech to text systems:
| System | en | de | es | fr | it | nl | ru | el | hi | zh | vi | pt | ca |
| ------ | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- |
| [pocketsphinx](speech-to-text.md#pocketsphinx) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | | &#x2713; | &#x2713; |
| [kaldi](speech-to-text.md#kaldi) | &#x2713; | &#x2713; | | | | &#x2713; | | | | | &#x2713; | | |
| System | en | de | es | fr | it | nl | ru | el | hi | zh | vi | pt |
| ------ | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- |
| [pocketsphinx](speech-to-text.md#pocketsphinx) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | | |
| [kaldi](speech-to-text.md#kaldi) | | | | | | | | | | | &#x2713; | &#x2713; |
## Pocketsphinx
@@ -24,7 +24,9 @@ Add to your [profile](profiles.md):
"base_dictionary": "base_dictionary.txt",
"custom_words": "custom_words.txt",
"dictionary": "dictionary.txt",
"language_model": "language_model.txt"
"language_model": "language_model.txt",
"unknown_words": "unknown_words.txt",
"mllr_matrix": "acoustic_model_mllr"
}
}
```
@@ -36,10 +38,11 @@ When Rhasspy starts, it creates a pocketsphinx decoder with the following attrib
* `hmm` - `speech_to_text.pocketsphinx.acoustic_model` (directory)
* `dict` - `speech_to_text.pocketsphinx.dictionary` (file)
* `lm` - `speech_to_text.pocketsphinx.language_model` (file)
* `mllr` - `speech_to_text.pocketsphinx.mllr_matrix` (file, optional)
### Open Transcription
The `mllr_matrix` file is intended for advanced users who want to [tune/adapt their acoustic models](https://cmusphinx.github.io/wiki/tutorialadapt). This can increase the performance of Rhasspy's speech recognition for a specific user/microphone/acoustic environment.
If you just want to use Rhasspy for general speech to text, you can set `speech_to_text.pocketsphinx.open_transcription` to `true` in your profile. This will use the included general language model (much slower) and ignore any custom voice commands you've specified. For English, German, and Dutch, you may want to use [Kaldi](#kaldi) instead for better results.
Pocketsphinx allows Rhasspy to support Enlgish (en), German (de), Dutch (nl), Spanish (es), Italian (it), French (fr), Greek (el), Russian (ru), Hindi (hi), and Mandarin (zh).
See `rhasspy.stt.PocketsphinxDecoder` for details.
@@ -73,10 +76,6 @@ This requires Kaldi to be installed, which is...challenging. The [Docker image o
Rhasspy expects a Kaldi-compatible profile to contain a `model` directory with a `train.sh` and `decode.sh` script. See the Vietnamese (vi) or Portuguese (pt) [profile](https://github.com/synesthesiam/rhasspy-profiles/releases) for an example.
### Open Transcription
If you just want to use Rhasspy for general speech to text, you can set `speech_to_text.kaldi.open_transcription` to `true` in your profile. This will use the included general language model (much slower) and ignore any custom voice commands you've specified.
## Remote HTTP Server
Uses a remote HTTP server to transform speech (WAV) to text.
@@ -94,7 +93,7 @@ Add to your [profile](profiles.md):
}
```
During speech recognition, 16-bit 16 kHz mono WAV data will be POST-ed to the endpoint with the `Content-Type` set to `audio/wav`. A `text/plain` response with the transcription is expected back. An additional `profile` query argument is sent with the current profile name, so the POST URL is effectively something like `http://remote-server:12101/api/speech-to-text?profile=en`.
During speech recognition, 16-bit 16Khz mono WAV data will be POST-ed to the endpoint with the `Content-Type` set to `audio/wav`. A `text/plain` response with the transcription is expected back. An additional `profile` query argument is sent with the current profile name, so the POST URL is effectively something like `http://remote-server:12101/api/speech-to-text?profile=en`.
See `rhasspy.stt.RemoteDecoder` for details.
@@ -116,7 +115,7 @@ Add to your [profile](profiles.md):
}
```
When a voice command is received, Rhasspy will call your program and push the recorded WAV data (16-bit 16 kHz mono) to standard in. Your program should print the text transcription to standard out.
When a voice command is received, Rhasspy will call your program and push the recorded WAV data (16-bit 16 Khz mono) to standard in. Your program should print the text transcription to standard out.
The following environment variables are available to your program:
docs/text-to-speech.md
+8 -8
View File
@@ -4,13 +4,13 @@ After you voice command has been [handled](intent-handling.md), it's common to p
The following table summarizes language support for the various text to speech systems:
| System | en | de | es | fr | it | nl | ru | el | hi | zh | vi | pt | ca |
| ------ | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- |
| [espeak](text-to-speech.md#espeak) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; |
| [flite](text-to-speech.md#flite) | &#x2713; | | | | | | | | &#x2713; | | | | |
| [picotts](text-to-speech.md#picotts) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | | | | | | | | |
| [marytts](text-to-speech.md#marytts) | &#x2713; | &#x2713; | | &#x2713; | &#x2713; | | &#x2713; | | | | | | |
| [wavenet](text-to-speech.md#google-wavenet) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | | &#x2713; | &#x2713; | | &#x2713; | |
| System | en | de | es | fr | it | nl | ru | el | hi | zh | vi | pt |
| ------ | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- | ------- |
| [espeak](text-to-speech.md#espeak) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; |
| [flite](text-to-speech.md#flite) | &#x2713; | | | | | | | | &#x2713; | | | |
| [picotts](text-to-speech.md#picotts) | &#x2713; | | | | | | | | | | | |
| [marytts](text-to-speech.md#marytts) | &#x2713; | &#x2713; | | &#x2713; | &#x2713; | | &#x2713; | | | | | |
| [wavenet](text-to-speech.md#google-wavenet) | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | &#x2713; | | &#x2713; | &#x2713; | | &#x2713; |
## eSpeak
@@ -159,7 +159,7 @@ Add to your [profile](profiles.md):
}
```
For compatibility with other services and Rhasspy components, it's best to return 16 kHz, 16-bit mono WAV data.
For compatibility with other services and Rhasspy components, it's best to return 16 Khz, 16-bit mono WAV data.
See `rhasspy.tts.CommandSentenceSpeaker` for details.
docs/training.md
+7 -7
View File
@@ -1,8 +1,8 @@
# Training
Rhasspy is designed to recognize voice commands that [you provide](#sentencesini). These commands are categorized by **intent**, and may contain variable **slots** or **entities**, such as the color and name of a light.
Rhasspy is designed to recognize only the specific set of voice commands that [you provide](#sentencesini). These commands are categorized by **intent**, and may contain variable **slots** or **entities**, such as the color and name of a light.
During the training process, Rhasspy simultaneously trains *both* a speech and intent recognizer. The speech recognizer converts voice commands to text, and the intent recognizer converts text to JSON events. Combined, they enable a low power, offline system like a Raspberry Pi to understand and respond to your voice commands.
During the training process, Rhasspy simulataneously trains *both* a speech and intent recognizer. The speech recognizer converts voice commands to text, and the intent recognizer converts text to JSON events. Combined, they enable a low power, offline system like a Raspberry Pi to understand and respond to your voice commands.
## How It Works
@@ -38,7 +38,7 @@ For step (4), Rhasspy can use a [variety of intent recognition systems](intent-r
and provided as training material to the intent recognition system. The [fuzzywuzzy](intent-recognition.md#fuzzywuzzy) system, for example, simply saves the JSON file and, during recognition, finds the closest matching sentence according to the [Levenshtein distance](https://en.wikipedia.org/wiki/Levenshtein_distance). The [default intent recognizer](intent-recognition.md#fsticuffs) interacts directly with the finite state transducer(s) generated in step (1) and, while less tolerant of errors than `fuzzywuzzy`, is significantly faster for large sets of voice commands (i.e., millions).
More sophisticated systems like [Rasa NLU](intent-recognition.md#rasanlu) use machine learning techniques to classify sentences by intent and assign slot (entity) values. These systems are much better at recognizing sentences not seen during training, but can take minutes to hours to train.
More sophisticated systems like [rasaNLU](intent-recognition.md#rasanlu) use machine learning techniques to classify sentences by intent and assign slota (entity) values. These systems are much better at recognizing sentences not seen during training, but can take minutes to hours to train.
## sentences.ini
@@ -65,7 +65,7 @@ Compared to JSON, YAML, etc., there is minimal syntactic overhead for the purpos
3. You cannot share commonly *repeated phrases* across sentences or intents.
4. There is no way to *tag phrases* so the intent recognizer knows the values for an intent's slots (e.g., color).
Each of these shortcomings are addressed by considering the space between intent headings (`[Intent 1]`, etc.) as a **grammar** that represent many possible voice commands. The possible sentences, stripped of their tags, are used as input to [opengrm](https://www.opengrm.org) to produce a standard ARPA language model for [pocketsphinx](https://github.com/cmusphinx/pocketsphinx) or [Kaldi](https://kaldi-asr.org). The tagged sentences are then used to train an intent recognizer.
Each of these shortcomings are addressed by considering the space between intent headings (`[Intent 1]`, etc.) as a **grammar** that will *generate* tagged sentences in [rasaNLU's training data format](https://rasa.com/docs/nlu/dataformat/#markdown-format). The generated sentences, stripped of their tags, are used as input to [opengrm](https://www.opengrm.org) to produce a language model for [pocketsphinx](https://github.com/cmusphinx/pocketsphinx) or [Kaldi](https://kaldi-asr.org). The tagged sentences are then used to train an intent recognizer.
### Optional Words
@@ -180,7 +180,7 @@ In the `SetLightColor` example above, the color names are stored in `sentences.i
colors = (red | green | blue)
This is convenient when the list of colors is small, changes infrequently, and does not depend on an external service.
Ths is convenient when the list of colors is small, changes infrequently, and does not depend on an external service.
But what if this was a list of movie names that were stored on your [Kodi Home Theater](https://kodi.tv)?
movies = ("Primer" | "Moon" | "Chronicle" | "Timecrimes" | "Mulholland Drive" | ... )
@@ -399,7 +399,7 @@ During training, Rhasspy can **mix** this large, open ended language model with
![Diagram of Rhasspy's training process](img/training.svg)
To see the effect of language model mixing, consider a simple `sentences.ini` file:
To see the effect of language model mixing, consder a simple `sentences.ini` file:
```
[ChangeLightState]
@@ -433,7 +433,7 @@ $ rhasspy-cli --profile en \
OK
```
Note that training will take **significantly** longer because of the size of the base language model. Now, let's test our two WAV files:
Note that training will take **significantly** longer because of the size of the base langauge model. Now, let's test our two WAV files:
```
$ rhasspy-cli --profile en wav2text < turn_on_living_room_lamp.wav
docs/usage.md
+6 -8
View File
@@ -77,14 +77,12 @@ Rhasspy features a comprehensive HTTP API available at `/api`, documented with [
* POST text and have Rhasspy process it as command
* `/api/text-to-speech`
* POST text and have Rhasspy speak it
* `/api/slots`
* POST JSON to update [slot values](training.md#slots-lists)
See `public/swagger.yaml` in Rhasspy's repository for all available endpoints, or visit `/api` on your Rhasspy web server (e.g., [http://localhost:12101/api](http://localhost:12101/api)).
See `public/swagger.yaml` in Rhasspy's repository for all available endpoints, or visit `/api` on your Rhasspy web server (e.g., [http://locahost:12101/api](http://localhost:12101/api)).
## Secure Hosting with HTTPS
If you need to access Rhasspy's web interface/API through HTTPS (formally SSL), you can provide a certificate and key file via command-line parameters or the Hass.io configuration.
If you need to access Rhasspy's web interface/API through HTTPS (formally SSL), you can provide a certificate and key file via command-line parameters or the Hass.IO configuration.
If you're running Rhasspy via Docker or in a virtual environment, add `--ssl <CERT_FILE> <KEY_FILE>` to the command-line arguments where `<CERT_FILE>` is your SSL certificate and `<KEY_FILE>` is your SSL key file.
@@ -96,9 +94,9 @@ After answering the series of questions, you should have `cert.pem` and `key.pem
<RHASSPY COMMAND> --ssl cert.pem key.pem
The web interface will now be available at [https://localhost:12101](https://localhost:12101) and the web socket events at `wss://localhost:12101/api/events/intent`
The web interface will now be available at [https://localhost:12101](https://locahost:12101) and the web socket events at `wss://localhost:12101/api/events/intent`
In Hass.io, you will need to set the following options via the web interface or in your JSON configuration:
In Hass.IO, you will need to set the following options via the web interface or in your JSON configuration:
* `ssl`: `true`
* `certfile`: `cert.pem`
@@ -191,7 +189,7 @@ If you run Rhasspy through [Docker](installation.md#docker), the [rhasspy-cli](h
Put this script in your `~/bin` directory so that you can refer to it as `rhasspy-cli` from any directory.
By default, it will look for profiles in `$XDG_CONFIG_FILE/rhasspy/profiles`, which is probably `~/.config/rhasspy/profiles` (see [XDG specification](https://specifications.freedesktop.org/basedir-spec/basedir-spec-latest.html) for more information).
**Beware**: the `rhasspy-cli` script runs under your user account and grants Rhasspy **write access to your home directory**.
**Beware**: the `rhasspy-cli` script run under your user accout and grants Rhasspy **write access to your home directory**.
This is needed to save files during the training process, and to avoid those files being owned by `root`.
The [rhasspy-cli-ro](https://github.com/synesthesiam/rhasspy/blob/master/bin/rhasspy-cli-ro) script can be used for read only operations, such as speech to text or intent handling, but cannot make any changes to your file system.
@@ -252,7 +250,7 @@ For `rhasspy-cli --profile <PROFILE_NAME> <COMMAND> <ARGUMENTS>`, `<COMMAND>` ca
### Profile Operations
Print the complete JSON for the English profile with:
Print the complete JSON for the Enlgish profile with:
rhasspy-cli --profile en info
docs/wake-word.md
+36 -65
View File
@@ -6,71 +6,11 @@ You can also wake Rhasspy up using the [HTTP API](usage.md#http-api) by POST-ing
The following table summarizes the key characteristics of each wake word system:
| System | Performance | Training to Customize | Online Sign Up |
| ------ | ----------- | ----------------- | ----------------------- |
| [porcupine](wake-word.md#porcupine) | excellent | yes, offline | no |
| [snowboy](wake-word.md#snowboy) | good | yes, online | yes |
| [pocketsphinx](wake-word.md#pocketsphinx) | poor | no | no |
| [precise](wake-word.md#mycroft-precise) | moderate | yes, offline | no |
## Porcupine
Listens for a wake word with [porcupine](https://github.com/Picovoice/Porcupine). This system has the best performance out of the box. If you want a custom wake word, however, you will need to re-run their optimizer tool every 30 days.
Add to your [profile](profiles.md):
```json
"wake": {
"system": "porcupine",
"porcupine": {
"library_path": "porcupine/libpv_porcupine.so",
"model_path": "porcupine/porcupine_params.pv",
"keyword_path": "porcupine/porcupine.ppn",
"sensitivity": 0.5
}
},
"rhasspy": {
"listen_on_start": true
}
```
There are a lot of [keyword files](https://github.com/Picovoice/Porcupine/tree/master/resources/keyword_files) available for download. Use the `linux` platform if you're on desktop/laptop (`amd64`) and the `raspberrypi` platform if you're using a Raspberry Pi (`armhf`/`aarch64`). The `.ppn` files should go in the `porcupine` directory inside your profile (referenced by `keyword_path`).
If you want to create a custom wake word, you will need to run the [Porcupine Optimizer](https://github.com/Picovoice/Porcupine/tree/master/tools/optimizer). **NOTE**: the generated keyword file is only valid for 30 days, though you can always just re-run the optimizer.
See `rhasspy.wake.PorcupineWakeListener` for details.
## Snowboy
Listens for a wake word with [snowboy](https://snowboy.kitt.ai). This system has the good performance out of the box, but requires an online service to train.
Add to your [profile](profiles.md):
```json
"wake": {
"system": "snowboy",
"hermes": {
"wakeword_id": "default"
},
"snowboy": {
"model": "model-name-in-profile.(u|p)mdl",
"audio_gain": 1,
"sensitivity": 0.5,
"chunk_size": 960
}
},
"rhasspy": {
"listen_on_start": true
}
```
Visit [the snowboy website](https://snowboy.kitt.ai) to train your own wake word model (requires linking to a GitHub/Google/Facebook account). This *personal* model with end with `.pmdl`, and should go in your profile directory. Then, set `wake.snowboy.model` to the name of that file.
You also have the option of using a pre-train *universal* model (`.umdl`) from [Kitt.AI](https://github.com/Kitt-AI/snowboy/tree/master/resources/models). I've received errors using anything but `snowboy.umdl`, but YMMV.
See `rhasspy.wake.SnowboyWakeListener` for details.
| System | Performance | Requires Training | Requires Online Sign Up |
| ------ | ----------- | ----------------- | ----------------------- |
| [pocketsphinx](wake-word.md#pocketsphinx) | poor | no | no |
| [precise](wake-word.md#mycroft-precise) | moderate | yes, offline | no |
| [snowboy](wake-word.md#snowboy) | good | yes, online | yes |
## Pocketsphinx
@@ -125,6 +65,37 @@ Follow [the instructions from Mycroft AI](https://github.com/MycroftAI/mycroft-p
See `rhasspy.wake.PreciseWakeListener` for details.
## Snowboy
Listens for a wake word with [snowboy](https://snowboy.kitt.ai). This system has the best performance out of the box, but requires an online service to train.
Add to your [profile](profiles.md):
```json
"wake": {
"system": "snowboy",
"hermes": {
"wakeword_id": "default"
},
"snowboy": {
"model": "model-name-in-profile.(u|p)mdl",
"audio_gain": 1,
"sensitivity": 0.5,
"chunk_size": 960
}
},
"rhasspy": {
"listen_on_start": true
}
```
Visit [the snowboy website](https://snowboy.kitt.ai) to train your own wake word model (requires linking to a GitHub/Google/Facebook account). This *personal* model with end with `.pmdl`, and should go in your profile directory. Then, set `wake.snowboy.model` to the name of that file.
You also have the option of using a pre-train *universal* model (`.umdl`) from [Kitt.AI](https://github.com/Kitt-AI/snowboy/tree/master/resources/models). I've received errors using anything but `snowboy.umdl`, but YMMV.
See `rhasspy.wake.SnowboyWakeListener` for details.
## MQTT/Hermes
Subscribes to the `hermes/hotword/<WAKEWORD_ID>/detected` topic, and wakes Rhasspy up when a message is received ([Hermes protocol](https://docs.snips.ai/ressources/hermes-protocol)). This allows Rhasspy to use the wake word functionality in [Snips.AI](https://snips.ai/).
download-dependencies.sh
+91 -100
View File
@@ -1,139 +1,130 @@
#!/usr/bin/env bash
this_dir="$( cd "$( dirname "$0" )" && pwd )"
cpu_arch=$(uname --m)
# -----------------------------------------------------------------------------
# Command-line Arguments
# -----------------------------------------------------------------------------
. "${this_dir}/etc/shflags"
DEFINE_string 'download-dir' "${this_dir}/download" 'Directory to cache downloaded files'
DEFINE_boolean 'precise' true 'Install Mycroft Precise'
DEFINE_boolean 'kaldi' true 'Install Kaldi'
DEFINE_boolean 'offline' false "Don't download anything"
DEFINE_boolean 'all-cpu' false 'Download dependencies for all CPU architectures'
DEFINE_string 'cpu-arch' "${cpu_arch}" 'CPU architecture (x86_64, armv7l, arm64v8)'
FLAGS "$@" || exit $?
eval set -- "${FLAGS_ARGV}"
# -----------------------------------------------------------------------------
# Default Settings
# -----------------------------------------------------------------------------
set -e
cpu_arch="${FLAGS_cpu_arch}"
download_dir="${FLAGS_download_dir}"
# Directory of *this* script
DIR="$( cd "$( dirname "$0" )" && pwd )"
# Place where downloaded artifacts are stored
download_dir="${DIR}/download"
mkdir -p "${download_dir}"
if [[ "${FLAGS_offline}" -eq "${FLAGS_TRUE}" ]]; then
offline='true'
fi
if [[ "${FLAGS_all_cpu}" -eq "${FLAGS_TRUE}" ]]; then
all_cpu='true'
fi
if [[ "${FLAGS_precise}" -eq "${FLAGS_FALSE}" ]]; then
no_precise='true'
fi
if [[ "${FLAGS_kaldi}" -eq "${FLAGS_FALSE}" ]]; then
no_kaldi='true'
fi
# -----------------------------------------------------------------------------
function maybe_download {
if [[ ! -f "$2" ]]; then
if [[ ! -z "${offline}" ]]; then
echo "Need to download $1 but offline."
exit 1
fi
mkdir -p "$(dirname "$2")"
curl -sSfL -o "$2" "$1"
echo "$1 => $2"
fi
}
# -----------------------------------------------------------------------------
# CPU architecture
CPU_ARCHS=("x86_64" "armv7l" "arm64v8")
FRIENDLY_ARCHS=("amd64" "armhf" "aarch64")
declare -A CPU_TO_FRIENDLY
CPU_TO_FRIENDLY["x86_64"]="amd64"
CPU_TO_FRIENDLY["armv7l"]="armhf"
CPU_TO_FRIENDLY["arm64v8"]="aarch64"
# CPU architecture
if [[ ! -z "${all_cpu}" ]]; then
CPU_ARCHS=("x86_64" "armv7l" "arm64v8")
FRIENDLY_ARCHS=("amd64" "armhf" "aarch64")
else
CPU_ARCHS=("${cpu_arch}")
FRIENDLY_ARCHS=("${CPU_TO_FRIENDLY[${cpu_arch}]}")
# -----------------------------------------------------------------------------
# OpenFST
# -----------------------------------------------------------------------------
openfst_file="${download_dir}/openfst-1.6.2.tar.gz"
if [[ ! -f "${openfst_file}" ]]; then
openfst_url='http://www.openfst.org/twiki/pub/FST/FstDownload/openfst-1.6.2.tar.gz'
echo "Downloading OpenFST source (${openfst_url})"
curl -sSfL -o "${openfst_file}" "${openfst_url}"
fi
# -----------------------------------------------------------------------------
# Rhasspy
# -----------------------------------------------------------------------------
for FRIENDLY_ARCH in "${FRIENDLY_ARCHS[@]}";
do
rhasspy_files=("rhasspy-tools_${FRIENDLY_ARCH}.tar.gz" "rhasspy-web-dist.tar.gz")
for rhasspy_file_name in "${rhasspy_files}"; do
rhasspy_file="${download_dir}/${rhasspy_file_name}"
rhasspy_file_url="https://github.com/synesthesiam/rhasspy/releases/download/v2.0/${rhasspy_file_name}"
maybe_download "${rhasspy_file_url}" "${rhasspy_file}"
done
done
# -----------------------------------------------------------------------------
# Pocketsphinx for Python
# -----------------------------------------------------------------------------
pocketsphinx_file="${download_dir}/pocketsphinx-python.tar.gz"
pocketsphinx_url='https://github.com/synesthesiam/pocketsphinx-python/releases/download/v1.0/pocketsphinx-python.tar.gz'
maybe_download "${pocketsphinx_url}" "${pocketsphinx_file}"
if [[ ! -f "${pocketsphinx_file}" ]]; then
pocketsphinx_url='https://github.com/synesthesiam/pocketsphinx-python/releases/download/v1.0/pocketsphinx-python.tar.gz'
echo "Downloading pocketsphinx (${pocketsphinx_url})"
curl -sSfL -o "${pocketsphinx_file}" "${pocketsphinx_url}"
fi
# -----------------------------------------------------------------------------
# jsgf2fst
# -----------------------------------------------------------------------------
jsgf2fst_file="${download_dir}/jsgf2fst-0.1.0.tar.gz"
if [[ ! -f "${jsgf2fst_file}" ]]; then
jsgf2fst_url='https://github.com/synesthesiam/jsgf2fst/releases/download/v0.1.0/jsgf2fst-0.1.0.tar.gz'
echo "Downloading jsgf2fst (${jsgf2fst_url})"
curl -sSfL -o "${jsgf2fst_file}" "${jsgf2fst_url}"
fi
# -----------------------------------------------------------------------------
# Snowboy
# -----------------------------------------------------------------------------
snowboy_file="${download_dir}/snowboy-1.3.0.tar.gz"
snowboy_url='https://github.com/Kitt-AI/snowboy/archive/v1.3.0.tar.gz'
maybe_download "${snowboy_url}" "${snowboy_file}"
if [[ ! -f "${snowboy_file}" ]]; then
snowboy_url='https://github.com/Kitt-AI/snowboy/archive/v1.3.0.tar.gz'
echo "Downloading snowboy (${snowboy_url})"
curl -sSfL -o "${snowboy_file}" "${snowboy_url}"
fi
# -----------------------------------------------------------------------------
# Mycroft Precise
# -----------------------------------------------------------------------------
if [[ -z "${no_precise}" ]]; then
for CPU_ARCH in "${CPU_ARCHS}";
do
case $CPU_ARCH in
x86_64|armv7l)
precise_file="${download_dir}/precise-engine_0.3.0_${CPU_ARCH}.tar.gz"
precise_url="https://github.com/MycroftAI/mycroft-precise/releases/download/v0.3.0/precise-engine_0.3.0_${CPU_ARCH}.tar.gz"
maybe_download "${precise_url}" "${precise_file}"
esac
done
for CPU_ARCH in "x86_64" "armv7l"
do
precise_file="${download_dir}/precise-engine_0.3.0_${CPU_ARCH}.tar.gz"
if [[ ! -f "${precise_file}" ]]; then
precise_url="https://github.com/MycroftAI/mycroft-precise/releases/download/v0.3.0/precise-engine_0.3.0_${CPU_ARCH}.tar.gz"
echo "Downloading Mycroft Precise (${precise_url})"
curl -sSfL -o "${precise_file}" "${precise_url}"
fi
done
# -----------------------------------------------------------------------------
# Opengrm
# -----------------------------------------------------------------------------
if [[ -z "$(which ngramcount)" ]]; then
opengrm_file="${download_dir}/opengrm-ngram-1.3.3.tar.gz"
if [[ ! -f "${opengrm_file}" ]]; then
opengrm_url='https://www.opengrm.org/twiki/pub/GRM/NGramDownload/opengrm-ngram-1.3.3.tar.gz'
echo "Download Opengrm (${opengrm_url})"
curl -sSfLk -o "${opengrm_file}" "${opengrm_url}"
fi
fi
# -----------------------------------------------------------------------------
# Phonetisaurus
# -----------------------------------------------------------------------------
for FRIENDLY_ARCH in "${FRIENDLY_ARCHS[@]}"
do
# Install pre-built package
phonetisaurus_file="${download_dir}/phonetisaurus-2019_${FRIENDLY_ARCH}.deb"
if [[ ! -f "${phonetisaurus_file}" ]]; then
phonetisaurus_url="https://github.com/synesthesiam/phonetisaurus-2019/releases/download/v1.0/phonetisaurus-2019_${FRIENDLY_ARCH}.deb"
echo "Downloading phonetisaurus (${phonetisaurus_url})"
curl -sSfL -o "${phonetisaurus_file}" "${phonetisaurus_url}"
fi
done
# Build from source
phonetisaurus_file="${download_dir}/phonetisaurus-2019.zip"
if [[ ! -f "${phonetisaurus_file}" ]]; then
phonetisaurus_url="https://github.com/synesthesiam/phonetisaurus-2019/releases/download/v1.0/phonetisaurus-2019_${FRIENDLY_ARCH}.deb"
echo "Downloading phonetisaurus source (${phonetisaurus_url})"
curl -sSfL -o "${phonetisaurus_file}" "${phonetisaurus_url}"
fi
# -----------------------------------------------------------------------------
# Kaldi
# -----------------------------------------------------------------------------
if [[ -z "${no_kaldi}" ]]; then
for FRIENDLY_ARCH in "${FRIENDLY_ARCHS}"
do
# Install pre-built package
kaldi_file="${download_dir}/kaldi_${FRIENDLY_ARCH}.tar.gz"
for FRIENDLY_ARCH in "${FRIENDLY_ARCHS[@]}"
do
# Install pre-built package
kaldi_file="${download_dir}/kaldi_${FRIENDLY_ARCH}.tar.gz"
if [[ ! -f "${kaldi_file}" ]]; then
kaldi_url="https://github.com/synesthesiam/kaldi-docker/releases/download/v1.0/kaldi_${FRIENDLY_ARCH}.tar.gz"
maybe_download "${kaldi_url}" "${kaldi_file}"
done
fi
echo "Downloading kaldi (${kaldi_url})"
curl -sSfL -o "${kaldi_file}" "${kaldi_url}"
fi
done
# -----------------------------------------------------------------------------
etc/shflags
-1206
View File
File diff suppressed because it is too large Load Diff
etc/test/el/test.json
+3 -3
View File
@@ -1,11 +1,11 @@
{
"text": "σβήσε το φως του γκαράζ",
"text": "ενεργοποιήστε τη λάμπα του καθιστικού",
"intent": {
"name": "ChangeLightState"
},
"entities": {
"name": "γκαράζ",
"state": "σβήσε"
"name": "καθιστικού",
"state": "ενεργοποιήστε"
},
"words": {
"known": {
etc/test/pt/test.json
+2 -2
View File
@@ -10,11 +10,11 @@
"words": {
"known": {
"word": "tempo",
"phonemes": "t e~ p u"
"phonemes": "T E~+ P UX"
},
"unknown": {
"word": "tempoada",
"phonemes": "t e~ p o a d a"
"phonemes": "T E~ P O A+ D AX"
}
}
}
etc/test/sv/test.json
-18
View File
@@ -1,18 +0,0 @@
{
"text": "är garage dörren öppen",
"intent": {
"name": "GetGarageState"
},
"entities": {
},
"words": {
"known": {
"word": "klockan",
"phonemes": "k l ol s k abl n"
},
"unknown": {
"word": "klockanklacken",
"phonemes": "k l ol s k abl n k l a k el n"
}
}
}
etc/test/sv/test.wav
BIN
View File
Binary file not shown.
examples/client-server/client/profiles/en/profile.json
+2 -2
View File
@@ -27,7 +27,7 @@
"intent": {
"system": "remote",
"rasa": {
"url": "http://localhost:5005/"
"url": "http://localhost:5000/"
},
"remote": {
"url": "http://server:12101/api/text-to-intent"
@@ -70,4 +70,4 @@
"username": "",
"site_id": "default"
}
}
}
examples/client-server/server/profiles/en/profile.json
+2 -2
View File
@@ -27,7 +27,7 @@
"intent": {
"system": "fuzzywuzzy",
"rasa": {
"url": "http://localhost:5005/"
"url": "http://localhost:5000/"
},
"remote": {
"url": "http://my-server:12101/api/text-to-intent"
@@ -73,4 +73,4 @@
"username": "",
"site_id": "default"
}
}
}
examples/typical-intent/rhasspy/profiles/en/profile.json
+1 -1
View File
@@ -27,7 +27,7 @@
"intent": {
"system": "fuzzywuzzy",
"rasa": {
"url": "http://localhost:5005/"
"url": "http://localhost:5000/"
},
"remote": {
"url": "http://my-server:12101/api/text-to-intent"
examples/typical/rhasspy/profiles/en/profile.json
+1 -1
View File
@@ -27,7 +27,7 @@
"intent": {
"system": "fuzzywuzzy",
"rasa": {
"url": "http://localhost:5005/"
"url": "http://localhost:5000/"
},
"remote": {
"url": "http://my-server:12101/api/text-to-intent"
mypy.ini
-6
View File
@@ -52,10 +52,4 @@ ignore_missing_imports = True
ignore_missing_imports = True
[mypy-cerberus.*]
ignore_missing_imports = True
[mypy-flair.*]
ignore_missing_imports = True
[mypy-google.*]
ignore_missing_imports = True
porcupine.py
-168
View File
@@ -1,168 +0,0 @@
#
# Copyright 2018 Picovoice Inc.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
#
import os
from ctypes import *
from enum import Enum
class Porcupine(object):
"""Python binding for Picovoice's wake word detection (aka Porcupine) library."""
class PicovoiceStatuses(Enum):
"""Status codes corresponding to 'pv_status_t' defined in 'include/picovoice.h'"""
SUCCESS = 0
OUT_OF_MEMORY = 1
IO_ERROR = 2
INVALID_ARGUMENT = 3
_PICOVOICE_STATUS_TO_EXCEPTION = {
PicovoiceStatuses.OUT_OF_MEMORY: MemoryError,
PicovoiceStatuses.IO_ERROR: IOError,
PicovoiceStatuses.INVALID_ARGUMENT: ValueError
}
class CPorcupine(Structure):
pass
def __init__(
self,
library_path,
model_file_path,
keyword_file_path=None,
sensitivity=None,
keyword_file_paths=None,
sensitivities=None):
"""
Loads Porcupine's shared library and creates an instance of wake word detection object.
:param library_path: Absolute path to Porcupine's shared library.
:param model_file_path: Absolute path to file containing model parameters.
:param keyword_file_path: Absolute path to keyword file containing hyper-parameters. If not present then
'keyword_file_paths' will be used.
:param sensitivity: Sensitivity parameter. A higher sensitivity value lowers miss rate at the cost of increased
false alarm rate. For more information regarding this parameter refer to 'include/pv_porcupine.h'. If not
present then 'sensitivities' is used.
:param keyword_file_paths: List of absolute paths to keyword files. Intended to be used for multiple keyword
scenario. This parameter is used only when 'keyword_file_path' is not set.
:param sensitivities: List of sensitivity parameters. Intended to be used for multiple keyword scenario. This
parameter is used only when 'sensitivity' is not set.
"""
if not os.path.exists(library_path):
raise IOError("Could not find Porcupine's library at '%s'" % library_path)
library = cdll.LoadLibrary(library_path)
if not os.path.exists(model_file_path):
raise IOError("Could not find model file at '%s'" % model_file_path)
if sensitivity is not None and keyword_file_path is not None:
if not os.path.exists(keyword_file_path):
raise IOError("Could not find keyword file at '%s'" % keyword_file_path)
keyword_file_paths = [keyword_file_path]
if not (0 <= sensitivity <= 1):
raise ValueError('Sensitivity should be within [0, 1]')
sensitivities = [sensitivity]
elif sensitivities is not None and keyword_file_paths is not None:
if len(keyword_file_paths) != len(sensitivities):
raise ValueError("Different number of sensitivity and keyword file path parameters are provided.")
for x in keyword_file_paths:
if not os.path.exists(os.path.expanduser(x)):
raise IOError("Could not find keyword file at '%s'" % x)
for x in sensitivities:
if not (0 <= x <= 1):
raise ValueError('Sensitivity should be within [0, 1]')
else:
raise ValueError("Sensitivity and/or keyword file path is missing")
self._num_keywords = len(keyword_file_paths)
init_func = library.pv_porcupine_multiple_keywords_init
init_func.argtypes = [
c_char_p,
c_int,
POINTER(c_char_p),
POINTER(c_float),
POINTER(POINTER(self.CPorcupine))]
init_func.restype = self.PicovoiceStatuses
self._handle = POINTER(self.CPorcupine)()
status = init_func(
model_file_path.encode(),
self._num_keywords,
(c_char_p * self._num_keywords)(*[os.path.expanduser(x).encode() for x in keyword_file_paths]),
(c_float * self._num_keywords)(*sensitivities),
byref(self._handle))
if status is not self.PicovoiceStatuses.SUCCESS:
raise self._PICOVOICE_STATUS_TO_EXCEPTION[status]('Initialization failed')
self.process_func = library.pv_porcupine_multiple_keywords_process
self.process_func.argtypes = [POINTER(self.CPorcupine), POINTER(c_short), POINTER(c_int)]
self.process_func.restype = self.PicovoiceStatuses
self._delete_func = library.pv_porcupine_delete
self._delete_func.argtypes = [POINTER(self.CPorcupine)]
self._delete_func.restype = None
self._sample_rate = library.pv_sample_rate()
self._frame_length = library.pv_porcupine_frame_length()
@property
def sample_rate(self):
"""Audio sample rate accepted by Porcupine library."""
return self._sample_rate
@property
def frame_length(self):
"""Number of audio samples per frame expected by C library."""
return self._frame_length
def process(self, pcm):
"""
Monitors incoming audio stream for given wake word(s).
:param pcm: An array (or array-like) of consecutive audio samples. For more information regarding required audio
properties (i.e. sample rate, number of channels encoding, and number of samples per frame) please refer to
'include/pv_porcupine.h'.
:return: For a single wake-word use cse True if wake word is detected. For multiple wake-word use case it
returns the index of detected wake-word. Indexing is 0-based and according to ordering of input keyword file
paths. It returns -1 when no keyword is detected.
"""
result = c_int()
status = self.process_func(self._handle, (c_short * len(pcm))(*pcm), byref(result))
if status is not self.PicovoiceStatuses.SUCCESS:
raise self._PICOVOICE_STATUS_TO_EXCEPTION[status]('Processing failed')
keyword_index = result.value
if self._num_keywords == 1:
return keyword_index == 0
else:
return keyword_index
def delete(self):
"""Releases resources acquired by Porcupine's library."""
self._delete_func(self._handle)
profiles/ca/custom_words.txt
View File
profiles/ca/espeak_phonemes.txt
View File
profiles/ca/frequent_words.txt
-2884
View File
File diff suppressed because it is too large Load Diff
profiles/ca/phoneme_examples.txt
-37
View File
@@ -1,37 +0,0 @@
A abby A B BV I
AE adéu AE DH E W
AO això AE Y SH AO
B baix B A Y SH
BV alba A L BV AE
C guia C I AE
CH boig B O CH
D andy A N D I
DH cada K A DH AE
E cent S E N
EA acte A K T EA
EE aneu AE N EE W
F cafè K AE F EE
G èxit EE G Z I T
GH agut AE GH U T
I ahir AE I
J boja B O J AE
K amic AE M I K
L ales A L EA S
LY allí AE LY I
M amor AE M O R
N anam A N AE M
NG banc B A NG K
NY anys A NY S
O avió AE BV Y O
P capa K A P AE
R aire A Y R EA
RR arma A RR M AE
S avís AE BV I S
SH així AE Y SH I
T alta A L T AE
U algú AE L GH U
UO alto A L T UO
V vagi V A GH I
W aqui A K W I
Y avui AE BV U Y
Z base B A Z EA
profiles/ca/profile.json
-28
View File
@@ -1,28 +0,0 @@
{
"language": "ca",
"name": "ca",
"speech_to_text": {
"system": "pocketsphinx",
"dictionary_casing": "lower"
},
"download": {
"conditions": {
"speech_to_text.system": {
"pocketsphinx": {
"acoustic_model": "ca-es-models-v0.4.0.zip:ca-es/acoustic-model",
"base_dictionary.txt": "ca-es-models-v0.4.0.zip:ca-es/pronounciation-dictionary.dict",
"g2p.fst": "ca-g2p.tar.gz:g2p.fst"
}
}
},
"files": {
"ca-es-models-v0.4.0.zip": {
"url": "https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-ca/ca-es-models-v0.4.0.zip"
},
"ca-g2p.tar.gz": {
"url": "https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-ca/ca-g2p.tar.gz"
}
}
}
}
profiles/ca/sentences.ini
-15
View File
@@ -1,15 +0,0 @@
[GetTime]
quina hora es
[GetTemperature]
quina és la temperatura
[GetGarageState]
la porta del garatge està oberta
està la porta del garatge tancada
[ChangeLightState]
light_name = (làmpada de la sala | llum del garatge) {name}
light_state = (encén:on | apagueu:on | engegueu:off) {state}
<light_state> [la] <light_state>
profiles/ca/stop_words.txt
-701
View File
@@ -1,701 +0,0 @@
a
abans
abans-d'ahir
abintestat
ací
adesiara
adés
adéu
adàgio
ah
ahir
ai
aitambé
aitampoc
aitan
aitant
aitantost
aixà
això
així
aleshores
algun
alguna
algunes
alguns
algú
alhora
allà
allèn
allò
allí
almenys
alto
altra
altre
altres
altresí
altri
alça
al·legro
amargament
amb
ambdues
ambdós
amunt
amén
anc
andante
andantino
anit
ans
antany
apa
aprés
aqueix
aqueixa
aqueixes
aqueixos
aqueixs
aquell
aquella
aquelles
aquells
aquest
aquesta
aquestes
aquests
aquèn
aquí
ara
arran
arrera
arrere
arreu
arri
arruix
atxim
au
avall
avant
aviat
avui
açò
bah
baix
baldament
ballmanetes
banzim-banzam
bastant
bastants
ben
bis
bitllo-bitllo
bo
bé
ca
cada
cal
cap
car
caram
catorze
cent
centes
cents
cerca
cert
certa
certes
certs
cinc
cinquanta
cinquena
cinquenes
cinquens
cinquè
com
comsevulla
contra
cordons
corrents
cric-crac
d
daixonses
daixò
dallonses
dallò
dalt
daltabaix
damunt
darrera
darrere
davall
davant
de
debades
dedins
defora
dejorn
dejús
dellà
dementre
dempeus
demés
demà
des
desena
desenes
desens
després
dessobre
dessota
dessús
desè
deu
devers
devora
deçà
diferents
dinou
dins
dintre
disset
divers
diversa
diverses
diversos
divuit
doncs
dos
dotze
dues
durant
ecs
eh
el
ela
elis
ell
ella
elles
ells
els
em
emperò
en
enans
enant
encara
encontinent
endalt
endarrera
endarrere
endavant
endebades
endemig
endemés
endemà
endins
endintre
enfora
engir
enguany
enguanyasses
enjús
enlaire
enlloc
enllà
enrera
enrere
ens
ensems
ensota
ensús
entorn
entre
entremig
entretant
entrò
envers
envides
environs
enviró
ençà
ep
ep
era
eren
eres
ergo
es
escar
essent
esser
est
esta
estada
estades
estan
estant
estar
estaran
estarem
estareu
estaria
estarien
estaries
estaré
estarà
estaràs
estaríem
estaríeu
estat
estats
estava
estaven
estaves
estem
estes
esteu
estic
estiguem
estigueren
estigueres
estigues
estiguessis
estigueu
estigui
estiguin
estiguis
estigué
estiguérem
estiguéreu
estigués
estiguí
estos
està
estàs
estàvem
estàveu
et
etc
etcètera
ets
excepte
fins
fora
foren
fores
força
fos
fossin
fossis
fou
fra
fui
fóra
fórem
fóreu
fóreu
fóssim
fóssiu
gaire
gairebé
gaires
gens
girientorn
gratis
ha
hagi
hagin
hagis
haguda
hagudes
hagueren
hagueres
haguessin
haguessis
hagut
haguts
hagué
haguérem
haguéreu
hagués
haguéssim
haguéssiu
haguí
hala
han
has
hauran
haurem
haureu
hauria
haurien
hauries
hauré
haurà
hauràs
hauríem
hauríeu
havem
havent
haver
haveu
havia
havien
havies
havíem
havíeu
he
hem
heu
hi
ho
hom
hui
hàgim
hàgiu
i
igual
iguals
inclusive
ja
jamai
jo
l
la
leri-leri
les
li
lla
llavors
llevat
lluny
llur
llurs
lo
los
ls
m
ma
mai
mal
malament
malgrat
manco
mant
manta
mantes
mantinent
mants
massa
mateix
mateixa
mateixes
mateixos
me
mentre
mentrestant
menys
mes
meu
meua
meues
meus
meva
meves
mi
mig
mil
mitges
mitja
mitjançant
mitjos
moixoni
molt
molta
moltes
molts
mon
mos
més
n
na
ne
ni
ningú
no
nogensmenys
només
noranta
nos
nosaltres
nostra
nostre
nostres
nou
novena
novenes
novens
novè
ns
nòs
nós
o
oh
oi
oidà
on
onsevulga
onsevulla
onze
pas
pengim-penjam
per
perquè
pertot
però
piano
pla
poc
poca
pocs
poques
potser
prest
primer
primera
primeres
primers
pro
prompte
prop
prou
puix
pus
pàssim
qual
quals
qualsevol
qualsevulla
qualssevol
qualssevulla
quan
quant
quanta
quantes
quants
quaranta
quart
quarta
quartes
quarts
quasi
quatre
que
quelcom
qui
quin
quina
quines
quins
quinze
quisvulla
què
ran
re
rebé
renoi
rera
rere
res
retruc
s
sa
salvament
salvant
salvat
se
segon
segona
segones
segons
seguida
seixanta
sempre
sengles
sens
sense
ser
seran
serem
sereu
seria
serien
series
seré
serà
seràs
seríem
seríeu
ses
set
setanta
setena
setenes
setens
setze
setè
seu
seua
seues
seus
seva
seves
si
sia
siau
sic
siguem
sigues
sigueu
sigui
siguin
siguis
sinó
sis
sisena
sisenes
sisens
sisè
sobre
sobretot
sol
sola
solament
soles
sols
som
son
sos
sota
sots
sou
sovint
suara
sí
sóc
són
t
ta
tal
tals
també
tampoc
tan
tanmateix
tant
tanta
tantes
tantost
tants
te
tercer
tercera
terceres
tercers
tes
teu
teua
teues
teus
teva
teves
ton
tos
tost
tostemps
tot
tota
total
totes
tothom
tothora
tots
trenta
tres
tret
tretze
tu
tururut
u
uf
ui
uix
ultra
un
una
unes
uns
up
upa
us
va
vagi
vagin
vagis
vaig
vair
vam
van
vares
vas
vau
vem
verbigràcia
vers
vet
veu
vint
vora
vos
vosaltres
vostra
vostre
vostres
vostè
vostès
vuit
vuitanta
vuitena
vuitenes
vuitens
vuitè
vés
vàreig
vàrem
vàreu
vós
xano-xano
xau-xau
xec
érem
éreu
és
ésser
àdhuc
àlies
ça
ço
òlim
ídem
últim
última
últimes
últims
únic
única
únics
úniques
profiles/de/check-profile.sh
Executable
+67
View File
@@ -0,0 +1,67 @@
#!/usr/bin/env bash
set -e
if [[ -z "$1" ]]; then
echo "Directory required as first argument"
exit 1
fi
DIR="$1"
#------------------------------------------------------------------------------
# Acoustic Model
#------------------------------------------------------------------------------
acoustic_output="${DIR}/acoustic_model"
if [[ ! -d "${acoustic_output}" ]]; then
echo "Missing acoustic model"
exit 1
fi
#------------------------------------------------------------------------------
# G2P
#------------------------------------------------------------------------------
g2p_output="${DIR}/g2p.fst"
if [[ ! -s "${g2p_output}" ]]; then
echo "Missing g2p model"
exit 1
fi
#------------------------------------------------------------------------------
# Dictionary
#------------------------------------------------------------------------------
dict_output="${DIR}/base_dictionary.txt"
if [[ ! -s "${dict_output}" ]]; then
echo "Missing base dictionary"
exit 1
fi
#------------------------------------------------------------------------------
# Language Model
#------------------------------------------------------------------------------
lm_output="${DIR}/base_language_model.txt"
if [[ ! -s "${lm_output}" ]]; then
echo "Missing base language model"
exit 1
fi
#------------------------------------------------------------------------------
# Flair Embeddings
#------------------------------------------------------------------------------
flair_dir="${DIR}/flair/cache/embeddings"
flair_files=("lm-mix-german-forward-v0.2rc.pt" "lm-mix-german-backward-v0.2rc.pt")
for file_name in "${flair_files[@]}"; do
file_output="${flair_dir}/${file_name}"
if [[ ! -s "${file_output}" ]]; then
echo "Missing flair embedding (${file_name})"
exit 1
fi
done
profiles/de/download-profile.sh
Executable
+119
View File
@@ -0,0 +1,119 @@
#!/usr/bin/env bash
set -e
if [[ -z "$1" ]]; then
echo "Directory required as first argument"
exit 1
fi
DIR="$1"
download_dir="${DIR}/download"
if [[ "$2" = "--delete" ]]; then
rm -rf "${download_dir}"
fi
mkdir -p "${download_dir}"
echo "Downloading German (de) profile (sphinx)"
#------------------------------------------------------------------------------
# Acoustic Model
#------------------------------------------------------------------------------
acoustic_url='https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-de/cmusphinx-de-voxforge-5.2.tar.gz'
acoustic_file="${download_dir}/cmusphinx-de-voxforge-5.2.tar.gz"
acoustic_output="${DIR}/acoustic_model"
if [[ ! -s "${acoustic_file}" ]]; then
echo "Downloading acoustic model (${acoustic_url})"
curl -sSfL -o "${acoustic_file}" "${acoustic_url}"
fi
echo "Extracting acoustic model (${acoustic_file})"
rm -rf "${acoustic_output}"
tar -C "${DIR}" -xf "${acoustic_file}" "cmusphinx-cont-voxforge-de-r20171217/model_parameters/voxforge.cd_cont_6000/" && mv "${DIR}/cmusphinx-cont-voxforge-de-r20171217/model_parameters/voxforge.cd_cont_6000" "${acoustic_output}" && rm -rf "${DIR}/cmusphinx-cont-voxforge-de-r20171217"
#------------------------------------------------------------------------------
# G2P
#------------------------------------------------------------------------------
g2p_url='https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-de/de-g2p.tar.gz'
g2p_file="${download_dir}/de-g2p.tar.gz"
g2p_output="${DIR}/g2p.fst"
if [[ ! -s "${g2p_file}" ]]; then
echo "Downloading g2p model (${g2p_url})"
curl -sSfL -o "${g2p_file}" "${g2p_url}"
fi
echo "Extracting g2p model (${g2p_file})"
tar --to-stdout -xzf "${g2p_file}" 'g2p.fst' > "${g2p_output}"
#------------------------------------------------------------------------------
# Dictionary
#------------------------------------------------------------------------------
dict_output="${DIR}/base_dictionary.txt"
echo "Extracting dictionary (${acoustic_file})"
tar --to-stdout -xf "${acoustic_file}" "cmusphinx-cont-voxforge-de-r20171217/etc/voxforge.dic" > "${dict_output}"
#------------------------------------------------------------------------------
# Language Model
#------------------------------------------------------------------------------
lm_url='https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-de/cmusphinx-voxforge-de.lm.gz'
lm_file="${download_dir}/cmusphinx-voxforge-de.lm.gz"
lm_output="${DIR}/base_language_model.txt"
if [[ ! -s "${lm_file}" ]]; then
echo "Downloading language model (${lm_url})"
curl -sSfL -o "${lm_file}" "${lm_url}"
fi
echo "Extracting language model (${lm_file})"
zcat "${lm_file}" > "${lm_output}"
#------------------------------------------------------------------------------
# Snowboy
#------------------------------------------------------------------------------
snowboy_models=("snowboy.umdl" "computer.umdl")
for model_name in "${snowboy_models[@]}"; do
model_output="${DIR}/${model_name}"
if [[ ! -s "${model_output}" ]]; then
model_url="https://github.com/Kitt-AI/snowboy/raw/master/resources/models/${model_name}"
echo "Downloading ${model_output} (${model_url})"
curl -sSfL -o "${model_output}" "${model_url}"
fi
done
#------------------------------------------------------------------------------
# Mycroft Precise
#------------------------------------------------------------------------------
precise_files=("hey-mycroft-2.pb" "hey-mycroft-2.pb.params")
for file_name in "${precise_files[@]}"; do
file_output="${DIR}/${file_name}"
if [[ ! -s "${file_output}" ]]; then
file_url="https://github.com/MycroftAI/precise-data/raw/models/${file_name}"
echo "Downloading ${file_output} (${file_url})"
curl -sSfL -o "${file_output}" "${file_url}"
fi
done
#------------------------------------------------------------------------------
# Flair Embeddings
#------------------------------------------------------------------------------
flair_dir="${DIR}/flair/cache/embeddings"
mkdir -p "${flair_dir}"
flair_files=("lm-mix-german-forward-v0.2rc.pt" "lm-mix-german-backward-v0.2rc.pt")
for file_name in "${flair_files[@]}"; do
file_output="${flair_dir}/${file_name}"
if [[ ! -s "${file_output}" ]]; then
file_url="https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-de/${file_name}"
echo "Downloading ${file_output} (${file_url})"
curl -sSfL -o "${file_output}" "${file_url}"
fi
done
profiles/de/kaldi/custom_words.txt
-2
View File
@@ -1,2 +0,0 @@
garagenlicht g a R 'a: Z @ ? 'a l 'I C t
wohnzimmerlampe v 'o: n ts I m 6 l 'a m p @
profiles/de/kaldi/espeak_phonemes.txt
-101
View File
@@ -1,101 +0,0 @@
'2 3:
'2: 3:
'3 3:
'3: ?
'6 3:
'9 3:
'@ k
'A 0
'A: ?
'E E
'E: E
'I I
'O 0
'O: ?
'OI E
'OY j
'U V
'V V
'Y u:
'a a
'a: a
'aI eI
'aU O:
'e d
'e: E
'i i:
'i: i:
'o aa
'o: oU
'u l
'u: V
'y ?
'y: u:
'{ a
2 ?
2: r
3 3
3: ?
6 3
9 ?
: ?
? 3:
A ?
A: ?
C tS
D D
E 3:
E: h
H ?
I I
N N
O 0
O: ?
OI ?
OY U@
R r
S S
SIL ?
T T
U V
V ?
Y u:
Z ?
a a
a: @
aI aI
aU f
b b
d d
dZ dZ
e E
e: h
f f
g dZ
h h
i i
i: i
j dZ
k k
l l
m m
n n
nC ?
o 0
o: oU
p p
pf f
r r
s s
t t
tS tS
ts z
u l
u: j
v w
w w
x tS
y ?
y: b
z s
{ ?
profiles/de/kaldi/phoneme_examples.txt
-82
View File
@@ -1,82 +0,0 @@
'2 götter g '2 t 6
'2: blöd b l '2: t
'3 wir w '3
'6 konzern k O n ts '6 n
'9 hölle h '9 l
'@ tony t '@ U n i
'A high h 'A I
'E berg b 'E 6 k
'E: fair f 'E: 6
'I bild b 'I l t
'O bord b 'O 6 t
'OI fräulein f R 'OI l aI n
'OY euch ? 'OY C
'U dumm d 'U m
'V love l 'V v
'Y fünf f 'Y n f
'a akte ? 'a k t
'a: bahn b 'a: n
'aI beim b 'aI m
'aU blau b l 'aU
'e erstmals ? 'e 6 s t m a: l s
'e: dreh d R 'e:
'i deal d 'i l
'i: bier b 'i: 6
'o chance S 'o s
'o: bloß b l 'o: s
'u louis l 'u i z
'u: blut b l 'u: t
'y: früh f R 'y:
'{ adam '{ d m
2: aufhören ? 'aU f h 2: R n
3 roger r 'A dZ 3
6 arme ? 'a 6 m
? aber ? 'a b 6
C dich d 'I C
D the D '@
E eddie ? E d 'i:
E: erzähle ? 'E 6 ts E: l
I dave d 'e I f
N bank b 'a N k
O okay ? O k 'e I
OY europa ? OY R 'o: p a:
R brot b R 'o: t
S chef S 'E f
T smith s m 'I T
U code k '@ U d
Y system z Y s t 'e: m
a amen ? a m 'E n
a: anna ? 'a n a:
aI anteil ? 'a n t aI l
aU august ? aU g 'U s t
b baby b 'e: b i
d dach d 'a x
dZ jake dZ 'E I k
e egal ? e g 'a: l
e: daher d 'a: h e: 6
f aufs ? 'aU f s
g auge ? 'aU g
h haar h 'a: 6
i joey dZ '@ U i
i: city s 'I t i:
j jahr j 'a: 6
k burg b 'U 6 k
l alle ? 'a l
m baum b 'aU m
n alan '{ l n
o kino k 'i: n o
o: auto ? 'aU t o:
p club k l 'U p
pf kopf k 'O pf
r mary m 'E r i:
s alex ? a l 'E k s
t acht ? 'a x t
tS jack tS '{ k
ts arzt ? 'a 6 ts t
u yahoo j 'A h u
u: anruf ? 'a n R u: f
v etwa ? 'E t v a
w william w 'I l j m
x auch ? 'aU x
y: büro b y: R 'o:
z also ? 'a l z o:
profiles/de/profile.json
+16 -87
View File
@@ -1,94 +1,23 @@
{
"language": "de",
"name": "de",
"speech_to_text": {
"system": "pocketsphinx",
"dictionary_casing": "lower",
"kaldi": {
"base_dictionary": "kaldi/base_dictionary.txt",
"base_language_model": "kaldi/base_language_model.txt",
"compatible": true,
"custom_words": "kaldi/custom_words.txt",
"dictionary": "kaldi/dictionary.txt",
"graph": "graph",
"language_model": "kaldi/language_model.txt",
"model_dir": "kaldi/model",
"unknown_words": "kaldi/unknown_words.txt",
"mix_fst": "kaldi/mixed.fst",
"g2p_model": "kaldi/g2p.fst",
"phoneme_examples": "kaldi/phoneme_examples.txt",
"phoneme_map": "kaldi/espeak_phonemes.txt"
}
},
"intent": {
"flair": {
"embeddings": [
"lm-mix-german-forward-v0.2rc.pt",
"lm-mix-german-backward-v0.2rc.pt"
]
}
},
"text_to_speech": {
"wavenet": {
"language_code": "de-DE"
"language": "de",
"name": "de",
"wake": {
"system": "snowboy"
},
"marytts": {
"locale": "de-DE"
}
},
"download": {
"conditions": {
"speech_to_text.system": {
"pocketsphinx": {
"acoustic_model": "cmusphinx-de-voxforge-5.2.tar.gz:cmusphinx-cont-voxforge-de-r20171217/model_parameters/voxforge.cd_cont_6000",
"base_dictionary.txt": "cmusphinx-de-voxforge-5.2.tar.gz:cmusphinx-cont-voxforge-de-r20171217/etc/voxforge.dic",
"g2p.fst": "de-g2p.tar.gz:g2p.fst"
},
"kaldi": {
"kaldi": "de_kaldi-zamia.tar.gz:kaldi"
}
},
"speech_to_text.kaldi.open_transcription": {
"True": {
"kaldi/model/base_graph": "de_kaldi-zamia-base_graph.tar.gz:base_graph"
}
},
"speech_to_text.pocketsphinx.mix_weight": {
">0": {
"base_language_model.txt": "cmusphinx-voxforge-de.lm.gz:cmusphinx-voxforge-de.lm"
}
},
"intent.system": {
"intent": {
"flair": {
"flair/cache/embeddings/lm-mix-german-forward-v0.2rc.pt": "lm-mix-german-forward-v0.2rc.pt",
"flair/cache/embeddings/lm-mix-german-backward-v0.2rc.pt": "lm-mix-german-backward-v0.2rc.pt"
"embeddings": [
"lm-mix-german-forward-v0.2rc.pt",
"lm-mix-german-backward-v0.2rc.pt"
]
}
}
},
"files": {
"cmusphinx-de-voxforge-5.2.tar.gz": {
"url": "https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-de/cmusphinx-de-voxforge-5.2.tar.gz"
},
"cmusphinx-voxforge-de.lm.gz": {
"url": "https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-de/cmusphinx-voxforge-de.lm.gz"
},
"de-g2p.tar.gz": {
"url": "https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-de/de-g2p.tar.gz"
},
"lm-mix-german-forward-v0.2rc.pt": {
"url": "https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-de/lm-mix-german-forward-v0.2rc.pt",
"cache": false
},
"lm-mix-german-backward-v0.2rc.pt": {
"url": "https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-de/lm-mix-german-backward-v0.2rc.pt",
"cache": false
},
"de_kaldi-zamia.tar.gz": {
"url": "https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-de/de_kaldi-zamia.tar.gz"
},
"de_kaldi-zamia-base_graph.tar.gz": {
"url": "https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-de/de_kaldi-zamia-base_graph.tar.gz"
}
"text_to_speech": {
"wavenet": {
"language_code": "de-DE"
},
"marytts": {
"locale": "de-DE"
}
}
}
}
profiles/de/stop_words.txt
-231
View File
@@ -1,231 +0,0 @@
aber
alle
allem
allen
aller
alles
als
also
am
an
ander
andere
anderem
anderen
anderer
anderes
anderm
andern
anderr
anders
auch
auf
aus
bei
bin
bis
bist
da
damit
dann
der
den
des
dem
die
das
daß
derselbe
derselben
denselben
desselben
demselben
dieselbe
dieselben
dasselbe
dazu
dein
deine
deinem
deinen
deiner
deines
denn
derer
dessen
dich
dir
du
dies
diese
diesem
diesen
dieser
dieses
doch
dort
durch
ein
eine
einem
einen
einer
eines
einig
einige
einigem
einigen
einiger
einiges
einmal
er
ihn
ihm
es
etwas
euer
eure
eurem
euren
eurer
eures
für
gegen
gewesen
hab
habe
haben
hat
hatte
hatten
hier
hin
hinter
ich
mich
mir
ihr
ihre
ihrem
ihren
ihrer
ihres
euch
im
in
indem
ins
ist
jede
jedem
jeden
jeder
jedes
jene
jenem
jenen
jener
jenes
jetzt
kann
kein
keine
keinem
keinen
keiner
keines
können
könnte
machen
man
manche
manchem
manchen
mancher
manches
mein
meine
meinem
meinen
meiner
meines
mit
muss
musste
nach
nicht
nichts
noch
nun
nur
ob
oder
ohne
sehr
sein
seine
seinem
seinen
seiner
seines
selbst
sich
sie
ihnen
sind
so
solche
solchem
solchen
solcher
solches
soll
sollte
sondern
sonst
über
um
und
uns
unsere
unserem
unseren
unser
unseres
unter
viel
vom
von
vor
während
war
waren
warst
was
weg
weil
weiter
welche
welchem
welchen
welcher
welches
wenn
werde
werden
wie
wieder
will
wir
wird
wirst
wo
wollen
wollte
würde
würden
zu
zum
zur
zwar
zwischen
profiles/defaults.json
+228 -312
View File
@@ -1,330 +1,246 @@
{
"command": {
"command": {
"arguments": [],
"program": ""
"command": {
"arguments": [],
"program": ""
},
"hermes": {
"timeout_sec": 30
},
"oneshot": {
"timeout_sec": 30
},
"system": "webrtcvad",
"webrtcvad": {
"chunk_size": 960,
"min_sec": 2,
"sample_rate": 16000,
"silence_sec": 0.5,
"speech_buffers": 5,
"throwaway_buffers": 10,
"timeout_sec": 30,
"vad_mode": 0
}
},
"hermes": {
"timeout_sec": 30
"handle": {
"command": {
"arguments": [],
"program": ""
},
"forward_to_hass": true,
"system": "hass"
},
"oneshot": {
"timeout_sec": 30
},
"system": "webrtcvad",
"webrtcvad": {
"chunk_size": 960,
"min_sec": 2,
"sample_rate": 16000,
"silence_sec": 0.5,
"speech_buffers": 5,
"throwaway_buffers": 10,
"timeout_sec": 30,
"vad_mode": 0
}
},
"handle": {
"command": {
"arguments": [],
"program": ""
},
"forward_to_hass": false,
"system": "dummy"
},
"home_assistant": {
"access_token": "",
"api_password": "",
"event_type_format": "rhasspy_{0}",
"pem_file": "",
"url": "http://hassio/homeassistant/"
},
"intent": {
"adapt": {
"stop_words": "stop_words.txt"
},
"command": {
"arguments": [],
"program": ""
},
"conversation": {
"handle_speech": true
},
"fuzzywuzzy": {
"examples_json": "intent_examples.json",
"min_confidence": 0
},
"fsticuffs": {
"intent_fst": "intent.fst",
"ignore_unknown_words": true,
"fuzzy": true
},
"flair": {
"cache_dir": "flair/cache",
"data_dir": "flair/data",
"embeddings": [],
"max_epochs": 25,
"do_sampling": true,
"num_samples": 10000,
"compatible": true
},
"rasa": {
"examples_markdown": "intent_examples.md",
"project_name": "rhasspy",
"url": "http://localhost:5005/"
},
"remote": {
"url": "http://my-server:12101/api/text-to-intent"
},
"replace_numbers": true,
"system": "fsticuffs"
},
"language": "",
"microphone": {
"arecord": {
"chunk_size": 960,
"device": ""
},
"pyaudio": {
"device": "",
"frames_per_buffer": 480
},
"stdin": {
"auto_start": true,
"chunk_size": 960
},
"http": {
"host": "127.0.0.1",
"port": 12333,
"stop_after": "never"
},
"gstreamer": {
"pipeline": "udpsrc port=12333 ! rawaudioparse use-sink-caps=false format=pcm pcm-format=s16le sample-rate=16000 num-channels=1 ! queue ! audioconvert ! audioresample"
},
"system": "pyaudio"
},
"mqtt": {
"enabled": false,
"host": "localhost",
"password": "",
"port": 1883,
"publish_intents": true,
"reconnect_sec": 5,
"site_id": "default",
"username": ""
},
"rhasspy": {
"listen_on_start": true,
"load_timeout_sec": 15,
"preload_profile": true
},
"sounds": {
"recorded": "${RHASSPY_BASE_DIR}/etc/wav/beep_lo.wav",
"system": "aplay",
"wake": "${RHASSPY_BASE_DIR}/etc/wav/beep_hi.wav"
},
"speech_to_text": {
"command": {
"arguments": [],
"program": ""
},
"dictionary_casing": "",
"g2p_casing": "",
"grammars_dir": "grammars",
"fsts_dir": "fsts",
"kaldi": {
"base_dictionary": "base_dictionary.txt",
"base_language_model": "base_language_model.txt",
"compatible": false,
"custom_words": "custom_words.txt",
"dictionary": "dictionary.txt",
"graph": "graph",
"base_graph": "base_graph",
"kaldi_dir": "${KALDI_PREFIX}/kaldi",
"language_model": "language_model.txt",
"model_dir": "model",
"unknown_words": "unknown_words.txt",
"mix_weight": 0,
"mix_fst": "mixed.fst",
"g2p_model": "g2p.fst",
"phoneme_examples": "phoneme_examples.txt",
"phoneme_map": "espeak_phonemes.txt",
"open_transcription": false
},
"pocketsphinx": {
"acoustic_model": "acoustic_model",
"base_dictionary": "base_dictionary.txt",
"base_language_model": "base_language_model.txt",
"compatible": true,
"custom_words": "custom_words.txt",
"dictionary": "dictionary.txt",
"language_model": "language_model.txt",
"min_confidence": 0,
"mllr_matrix": "acoustic_model_mllr",
"unknown_words": "unknown_words.txt",
"mix_weight": 0,
"mix_fst": "mixed.fst",
"g2p_model": "g2p.fst",
"phoneme_examples": "phoneme_examples.txt",
"phoneme_map": "espeak_phonemes.txt",
"open_transcription": false
},
"remote": {
"url": "http://my-server:12101/api/speech-to-text"
},
"sentences_ini": "sentences.ini",
"slots_dir": "slots",
"system": "dummy"
},
"text_to_speech": {
"command": {
"arguments": [],
"program": ""
},
"espeak": {},
"flite": {
"voice": "kal16"
},
"marytts": {
"url": "http://localhost:59125"
},
"picotts": {},
"system": "espeak",
"wavenet": {
"cache_dir": "tts/googlewavenet/cache",
"credentials_json": "tts/googlewavenet/credentials.json",
"gender": "FEMALE",
"language_code": "en-US",
"sample_rate": 22050,
"url": "https://texttospeech.googleapis.com/v1/text:synthesize",
"voice": "Wavenet-C",
"fallback_tts": "espeak"
}
},
"training": {
"dictionary_number_duplicates": true,
"grammars": {
"delete_before_training": true
"home_assistant": {
"access_token": "",
"api_password": "",
"event_type_format": "rhasspy_{0}",
"pem_file": "",
"url": "http://hassio/homeassistant/"
},
"intent": {
"command": {
"arguments": [],
"program": ""
},
"intent_map": "intent_map.json",
"system": "auto"
"adapt": {
"stop_words": "stop_words.txt"
},
"command": {
"arguments": [],
"program": ""
},
"fuzzywuzzy": {
"examples_json": "intent_examples.json",
"min_confidence": 0
},
"fsticuffs": {
"intent_fst": "intent.fst",
"ignore_unknown_words": true
},
"flair": {
"cache_dir": "flair/cache",
"data_dir": "flair/data",
"embeddings": [],
"max_epochs": 25,
"do_sampling": true,
"num_samples": 10000,
"compatible": true
},
"rasa": {
"examples_markdown": "intent_examples.md",
"project_name": "rhasspy",
"url": "http://localhost:5000/"
},
"remote": {
"url": "http://my-server:12101/api/text-to-intent"
},
"system": "fsticuffs"
},
"regex": {
"split": "\\s+"
"language": "",
"microphone": {
"arecord": {
"chunk_size": 960,
"device": ""
},
"pyaudio": {
"device": "",
"frames_per_buffer": 480
},
"stdin": {
"auto_start": true,
"chunk_size": 960
},
"system": "pyaudio"
},
"mqtt": {
"enabled": false,
"host": "localhost",
"password": "",
"port": 1883,
"publish_intents": true,
"reconnect_sec": 5,
"site_id": "default",
"username": ""
},
"rhasspy": {
"listen_on_start": true,
"load_timeout_sec": 15,
"preload_profile": true
},
"sounds": {
"recorded": "etc/wav/beep_lo.wav",
"system": "aplay",
"wake": "etc/wav/beep_hi.wav"
},
"speech_to_text": {
"command": {
"arguments": [],
"program": ""
},
"system": "auto"
"command": {
"arguments": [],
"program": ""
},
"dictionary_casing": "",
"g2p_casing": "",
"g2p_model": "g2p.fst",
"grammars_dir": "grammars",
"fsts_dir": "fsts",
"kaldi": {
"base_dictionary": "base_dictionary.txt",
"base_language_model": "base_language_model.txt",
"compatible": false,
"custom_words": "custom_words.txt",
"dictionary": "dictionary.txt",
"graph": "graph",
"kaldi_dir": "/opt/kaldi",
"language_model": "language_model.txt",
"model_dir": "model",
"unknown_words": "unknown_words.txt",
"mix_weight": 0.0,
"mix_fst": "mixed.fst"
},
"pocketsphinx": {
"acoustic_model": "acoustic_model",
"base_dictionary": "base_dictionary.txt",
"base_language_model": "base_language_model.txt",
"compatible": true,
"custom_words": "custom_words.txt",
"dictionary": "dictionary.txt",
"language_model": "language_model.txt",
"min_confidence": 0,
"mllr_matrix": "acoustic_model_mllr",
"unknown_words": "unknown_words.txt",
"mix_weight": 0.0,
"mix_fst": "mixed.fst"
},
"remote": {
"url": "http://my-server:12101/api/speech-to-text"
},
"sentences_ini": "sentences.ini",
"slots_dir": "slots",
"system": "pocketsphinx"
},
"tokenizer": "regex",
"unknown_words": {
"fail_when_present": true,
"guess_pronunciations": true
}
},
"wake": {
"command": {
"arguments": [],
"program": ""
"text_to_speech": {
"command": {
"arguments": [],
"program": ""
},
"espeak": {
"phoneme_map": "espeak_phonemes.txt",
"ipa_map": "ipa_phonemes.txt"
},
"flite": {
"voice": "kal16"
},
"marytts": {
"url": "http://localhost:59125"
},
"phoneme_examples": "phoneme_examples.txt",
"picotts": {},
"system": "espeak",
"wavenet": {
"cache_dir": "tts/googlewavenet/cache",
"credentials_json": "tts/googlewavenet/credentials.json",
"gender": "FEMALE",
"language_code": "en-US",
"sample_rate": 22050,
"url": "https://texttospeech.googleapis.com/v1/text:synthesize",
"voice": "Wavenet-C",
"fallback_tts": "espeak"
}
},
"hermes": {
"wakeword_id": "default"
"training": {
"dictionary_number_duplicates": true,
"grammars": {
"delete_before_training": true
},
"intent": {
"command": {
"arguments": [],
"program": ""
},
"intent_map": "intent_map.json",
"system": "auto"
},
"regex": {
"split": "\\s+"
},
"speech_to_text": {
"command": {
"arguments": [],
"program": ""
},
"system": "auto"
},
"tokenizer": "regex",
"unknown_words": {
"fail_when_present": true,
"guess_pronunciations": true
}
},
"pocketsphinx": {
"chunk_size": 960,
"compatible": true,
"keyphrase": "okay rhasspy",
"mllr_matrix": "wake_mllr",
"threshold": 1e-30
},
"precise": {
"chunk_delay": 0.005,
"chunk_size": 2048,
"engine_path": "precise-engine",
"model": "precise/hey-mycroft-2.pb",
"sensitivity": 0.5,
"trigger_level": 3
},
"snowboy": {
"apply_frontend": false,
"audio_gain": 1,
"chunk_size": 960,
"model": "snowboy/snowboy.umdl",
"sensitivity": 0.5
},
"porcupine": {
"library_path": "porcupine/libpv_porcupine.so",
"model_path": "porcupine/porcupine_params.pv",
"keyword_path": "porcupine/porcupine.ppn",
"sensitivity": 0.5
},
"system": "dummy"
},
"download": {
"cache_dir": "download",
"conditions": {
"wake.system": {
"snowboy": {
"snowboy/snowboy.umdl": "snowboy.umdl"
"wake": {
"command": {
"arguments": [],
"program": ""
},
"hermes": {
"wakeword_id": "default"
},
"pocketsphinx": {
"chunk_size": 960,
"compatible": true,
"keyphrase": "okay rhasspy",
"mllr_matrix": "wake_mllr",
"threshold": 1e-30
},
"precise": {
"precise/hey-mycroft-2.pb": "hey-mycroft-2.pb",
"precise/hey-mycroft-2.pb.params": "hey-mycroft-2.pb.params"
"chunk_delay": 0.005,
"chunk_size": 2048,
"engine_path": "precise-engine",
"model": "hey-mycroft-2.pb",
"sensitivity": 0.5,
"trigger_level": 3
},
"porcupine": {
"porcupine/porcupine_params.pv": "porcupine_params.pv",
"porcupine/libpv_porcupine.so": "libpv_porcupine.so",
"porcupine/porcupine.ppn": "porcupine.ppn"
}
}
},
"files": {
"snowboy.umdl": {
"url": "https://github.com/Kitt-AI/snowboy/raw/master/resources/models/snowboy.umdl",
"cache": false
},
"hey-mycroft-2.pb": {
"url": "https://github.com/MycroftAI/precise-data/raw/models/hey-mycroft-2.pb",
"cache": false
},
"hey-mycroft-2.pb.params": {
"url": "https://github.com/MycroftAI/precise-data/raw/models/hey-mycroft-2.pb.params",
"cache": false
},
"porcupine_params.pv": {
"url": "https://github.com/Picovoice/Porcupine/raw/master/lib/common/porcupine_params.pv",
"cache": false
},
"porcupine.ppn": {
"cache": false,
"x86_64": {
"url": "https://github.com/Picovoice/Porcupine/raw/master/resources/keyword_files/linux/porcupine_linux.ppn"
"snowboy": {
"apply_frontend": false,
"audio_gain": 1,
"chunk_size": 960,
"model": "snowboy.umdl",
"sensitivity": 0.5
},
"armv7l": {
"url": "https://github.com/Picovoice/Porcupine/raw/master/resources/keyword_files/raspberrypi/porcupine_raspberrypi.ppn"
},
"aarch64": {
"url": "https://github.com/Picovoice/Porcupine/raw/master/resources/keyword_files/raspberrypi/porcupine_raspberrypi.ppn"
}
},
"libpv_porcupine.so": {
"cache": false,
"x86_64": {
"url": "https://github.com/Picovoice/Porcupine/raw/master/lib/linux/x86_64/libpv_porcupine.so"
},
"armv7l": {
"url": "https://github.com/Picovoice/Porcupine/raw/master/lib/raspberry-pi/cortex-a53/libpv_porcupine.so"
},
"aarch64": {
"url": "https://github.com/Picovoice/Porcupine/raw/master/lib/raspberry-pi/cortex-a53/libpv_porcupine.so"
}
}
"system": "pocketsphinx"
}
}
}
profiles/el/check-profile.sh
Executable
+53
View File
@@ -0,0 +1,53 @@
#!/usr/bin/env bash
set -e
if [[ -z "$1" ]]; then
echo "Directory required as first argument"
exit 1
fi
DIR="$1"
#------------------------------------------------------------------------------
# Acoustic Model
#------------------------------------------------------------------------------
acoustic_output="${DIR}/acoustic_model"
if [[ ! -d "${acoustic_output}" ]]; then
echo "Missing acoustic model"
exit 1
fi
#------------------------------------------------------------------------------
# G2P
#------------------------------------------------------------------------------
g2p_output="${DIR}/g2p.fst"
if [[ ! -s "${g2p_output}" ]]; then
echo "Missing g2p model"
exit 1
fi
#------------------------------------------------------------------------------
# Dictionary
#------------------------------------------------------------------------------
dict_output="${DIR}/base_dictionary.txt"
if [[ ! -s "${dict_output}" ]]; then
echo "Missing base dictionary"
exit 1
fi
#------------------------------------------------------------------------------
# Language Model
#------------------------------------------------------------------------------
lm_output="${DIR}/base_language_model.txt"
if [[ ! -s "${lm_output}" ]]; then
echo "Missing base language model"
exit 1
fi
profiles/el/download-profile.sh
Executable
+95
View File
@@ -0,0 +1,95 @@
#!/usr/bin/env bash
set -e
if [[ -z "$1" ]]; then
echo "Directory required as first argument"
exit 1
fi
DIR="$1"
download_dir="${DIR}/download"
if [[ "$2" = "--delete" ]]; then
rm -rf "${download_dir}"
fi
mkdir -p "${download_dir}"
echo "Downloading Greek (el) profile (sphinx)"
#------------------------------------------------------------------------------
# Acoustic Model
#------------------------------------------------------------------------------
acoustic_url='https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-el/cmusphinx-el-5.2.tar.gz'
acoustic_file="${download_dir}/cmusphinx-el-5.2.tar.gz"
acoustic_output="${DIR}/acoustic_model"
if [[ ! -s "${acoustic_file}" ]]; then
echo "Downloading acoustic model (${acoustic_url})"
curl -sSfL -o "${acoustic_file}" "${acoustic_url}"
fi
echo "Extracting acoustic model (${acoustic_file})"
rm -rf "${acoustic_output}"
tar -C "${DIR}" -xf "${acoustic_file}" "cmusphinx-el-gr-5.2/el-gr.cd_cont_5000/" && mv "${DIR}/cmusphinx-el-gr-5.2/el-gr.cd_cont_5000" "${acoustic_output}" && rm -rf "${DIR}/cmusphinx-el-gr-5.2"
#------------------------------------------------------------------------------
# G2P
#------------------------------------------------------------------------------
g2p_url='https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-el/el-g2p.tar.gz'
g2p_file="${download_dir}/el-g2p.tar.gz"
g2p_output="${DIR}/g2p.fst"
if [[ ! -s "${g2p_file}" ]]; then
echo "Downloading g2p model (${g2p_url})"
curl -sSfL -o "${g2p_file}" "${g2p_url}"
fi
echo "Extracting g2p model (${g2p_file})"
tar --to-stdout -xzf "${g2p_file}" 'g2p.fst' > "${g2p_output}"
#------------------------------------------------------------------------------
# Dictionary
#------------------------------------------------------------------------------
dict_output="${DIR}/base_dictionary.txt"
echo "Extracting dictionary (${acoustic_file})"
tar --to-stdout -xzf "${acoustic_file}" 'cmusphinx-el-gr-5.2/el-gr.dic' > "${dict_output}"
#------------------------------------------------------------------------------
# Language Model
#------------------------------------------------------------------------------
lm_output="${DIR}/base_language_model.txt"
echo "Extracting language model (${acoustic_file})"
tar --to-stdout -xzf "${acoustic_file}" 'cmusphinx-el-gr-5.2/el-gr.lm.gz' | zcat > "${lm_output}"
#------------------------------------------------------------------------------
# Snowboy
#------------------------------------------------------------------------------
snowboy_models=("snowboy.umdl" "computer.umdl")
for model_name in "${snowboy_models[@]}"; do
model_output="${DIR}/${model_name}"
if [[ ! -s "${model_output}" ]]; then
model_url="https://github.com/Kitt-AI/snowboy/raw/master/resources/models/${model_name}"
echo "Downloading ${model_output} (${model_url})"
curl -sSfL -o "${model_output}" "${model_url}"
fi
done
#------------------------------------------------------------------------------
# Mycroft Precise
#------------------------------------------------------------------------------
precise_files=("hey-mycroft-2.pb" "hey-mycroft-2.pb.params")
for file_name in "${precise_files[@]}"; do
file_output="${DIR}/${file_name}"
if [[ ! -s "${file_output}" ]]; then
file_url="https://github.com/MycroftAI/precise-data/raw/models/${file_name}"
echo "Downloading ${file_output} (${file_url})"
curl -sSfL -o "${file_output}" "${file_url}"
fi
done
profiles/el/profile.json
+4 -30
View File
@@ -1,41 +1,15 @@
{
"language": "el",
"name": "el",
"wake": {
"system": "snowboy"
},
"speech_to_text": {
"g2p_casing": "lower",
"system": "pocketsphinx",
"dictionary_casing": "lower"
"g2p_casing": "lower"
},
"intent": {
"flair": {
"compatible": false
}
},
"download": {
"conditions": {
"speech_to_text.system": {
"pocketsphinx": {
"acoustic_model": "cmusphinx-el-5.2.tar.gz:cmusphinx-el-gr-5.2/el-gr.cd_cont_5000",
"base_dictionary.txt": "cmusphinx-el-5.2.tar.gz:cmusphinx-el-gr-5.2/el-gr.dic",
"g2p.fst": "el-g2p.tar.gz:g2p.fst"
}
},
"speech_to_text.pocketsphinx.mix_weight": {
">0": {
"base_language_model.txt": "cmusphinx-el-5.2.tar.gz:cmusphinx-el-gr-5.2/el-gr.lm.gz:"
}
}
},
"files": {
"cmusphinx-el-5.2.tar.gz": {
"url": "https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-el/cmusphinx-el-5.2.tar.gz"
},
"el-g2p.tar.gz": {
"url": "https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-el/el-g2p.tar.gz"
}
}
}
}
profiles/el/sentences.ini
+3 -2
View File
@@ -1,9 +1,10 @@
[GetTemperature]
τι θερμοκρασία έχει
ποια είναι η θερμοκρασία
πόσο (ζεστό | κρύο) είναι
[GetGarageState]
είναι (ανοιχτή | κλειστή) η πόρτα του γκαράζ
[ChangeLightState]
(σβήσε | άναψε){state} το φως του (γκαράζ){name}
(ενεργοποιήστε | κλείστε){state} τη λάμπα του (καθιστικού){name}
(απενεργοποιήστε | ανάψτε){state} το φως του (γκαράζ){name}
profiles/el/stop_words.txt
-153
View File
@@ -1,153 +0,0 @@
ἀλλ’
αλλα
ἀλλά
ἄλλοσ
αν
ἄν
αντι
απο
ἀπό
ἄρα
αυτα
αυτεσ
αυτη
αυτο
αυτοι
αυτοσ
αὐτόσ
αυτουσ
αυτων
γα^
γάρ
γε
για
δ’
δαί
δαίσ
δε
δέ
δεν
δή
διά
εαν
ἐάν
ἑαυτοῦ
ἐγώ
εἰ
ειμαι
ειμαστε
εἰμί
εἴμι
ειναι
εἰσ
εισαι
ειστε
ἐκ
εκεινα
εκεινεσ
εκεινη
εκεινο
εκεινοι
εκεινοσ
εκεινουσ
εκεινων
ἐμόσ
ἐν
ενω
επι
ἐπί
ἔτι
η
ἡ
ἤ
θα
ισωσ
κ
και
καί
κατα
κατά
κι
μα
με
μέν
μετα
μετά
μη
μή
μην
να
ο
ὁ
ὅδε
οι
οἱ
ομωσ
οπωσ
ὅσ
οσο
ὅστισ
οτι
ὅτι
οὐ
οὐδέ
οὐδείσ
οὐκ
οὖν
οὔτε
οὗτοσ
οὕτωσ
παρα
παρά
περί
ποια
ποιεσ
ποιο
ποιοι
ποιοσ
ποιουσ
ποιων
που
προσ
πρόσ
πωσ
σε
σόσ
στη
στην
στο
στον
σύ
σύν
τα
τά
τε
τῇ
την
τήν
τησ
τῆσ
τι
τί
τισ
τίσ
το
τό
τοί
τοιοῦτοσ
τον
τόν
τοτε
του
τοῦ
τούσ
τῷ
των
τῶν
ὑμόσ
ὑπέρ
ὑπό
ὦ
ωσ
ὡσ
ὥστε
profiles/en/check-profile.sh
Executable
+67
View File
@@ -0,0 +1,67 @@
#!/usr/bin/env bash
set -e
if [[ -z "$1" ]]; then
echo "Directory required as first argument"
exit 1
fi
DIR="$1"
#------------------------------------------------------------------------------
# Acoustic Model
#------------------------------------------------------------------------------
acoustic_output="${DIR}/acoustic_model"
if [[ ! -d "${acoustic_output}" ]]; then
echo "Missing acoustic model"
exit 1
fi
#------------------------------------------------------------------------------
# G2P
#------------------------------------------------------------------------------
g2p_output="${DIR}/g2p.fst"
if [[ ! -s "${g2p_output}" ]]; then
echo "Missing g2p model"
exit 1
fi
#------------------------------------------------------------------------------
# Dictionary
#------------------------------------------------------------------------------
dict_output="${DIR}/base_dictionary.txt"
if [[ ! -s "${dict_output}" ]]; then
echo "Missing base dictionary"
exit 1
fi
#------------------------------------------------------------------------------
# Language Model
#------------------------------------------------------------------------------
lm_output="${DIR}/base_language_model.txt"
if [[ ! -s "${lm_output}" ]]; then
echo "Missing base language model"
exit 1
fi
#------------------------------------------------------------------------------
# Flair Embeddings
#------------------------------------------------------------------------------
flair_dir="${DIR}/flair/cache/embeddings"
flair_files=("news-forward-0.4.1.pt" "news-backward-0.4.1.pt")
for file_name in "${flair_files[@]}"; do
file_output="${flair_dir}/${file_name}"
if [[ ! -s "${file_output}" ]]; then
echo "Missing flair embedding (${file_name})"
exit 1
fi
done
profiles/en/computer.umdl
BIN
View File
Binary file not shown.
profiles/en/dictionary.txt
Executable
+43
View File
@@ -0,0 +1,43 @@
bedroom B EH D R UW M
blue B L UW
closed K L OW Z D
cold K OW L D
door D AO R
garage G ER AA ZH
green G R IY N
hot HH AA T
how HH AW
is IH Z
it IH T
lamp L AE M P
light L AY T
living L IH V IH NG
make M EY K
me M IY
off AO F
okay OW K EY
on AA N
on(2) AO N
open OW P AH N
orange AO R AH N JH
orange(2) AO R IH N JH
purple P ER P AH L
red R EH D
rhasspy R AE S P IY
room R UW M
set S EH T
tell T EH L
temperature T EH M P R AH CH ER
temperature(2) T EH M P ER AH CH ER
the DH AH
the(2) DH IY
time T AY M
to T UW
to(2) T IH
to(3) T AH
turn T ER N
what W AH T
what(2) HH W AH T
whats W AH T S
whats(2) HH W AH T S
yellow Y EH L OW
profiles/en/download-profile.sh
Executable
+122
View File
@@ -0,0 +1,122 @@
#!/usr/bin/env bash
set -e
if [[ -z "$1" ]]; then
echo "Directory required as first argument"
exit 1
fi
DIR="$1"
download_dir="${DIR}/download"
if [[ "$2" = "--delete" ]]; then
rm -rf "${download_dir}"
fi
mkdir -p "${download_dir}"
echo "Downloading English (en) profile (sphinx)"
#------------------------------------------------------------------------------
# Acoustic Model
#------------------------------------------------------------------------------
acoustic_url='https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-en/cmusphinx-en-us-5.2.tar.gz'
acoustic_file="${download_dir}/cmusphinx-en-us-5.2.tar.gz"
echo "${acoustic_file}"
acoustic_output="${DIR}/acoustic_model"
if [[ ! -s "${acoustic_file}" ]]; then
echo "Downloading acoustic model (${acoustic_url})"
curl -sSfL -o "${acoustic_file}" "${acoustic_url}"
fi
echo "Extracting acoustic model (${acoustic_file})"
rm -rf "${acoustic_output}"
tar -C "${DIR}" -xzf "${acoustic_file}" && mv "${DIR}/cmusphinx-en-us-5.2" "${acoustic_output}"
#------------------------------------------------------------------------------
# G2P
#------------------------------------------------------------------------------
g2p_url='https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-en/en-g2p.tar.gz'
g2p_file="${download_dir}/en-g2p.tar.gz"
g2p_output="${DIR}/g2p.fst"
if [[ ! -s "${g2p_file}" ]]; then
echo "Downloading g2p model"
curl -sSfL -o "${g2p_file}" "${g2p_url}"
fi
echo "Extracting g2p model (${g2p_file})"
tar --to-stdout -xzf "${g2p_file}" 'g2p.fst' > "${g2p_output}"
#------------------------------------------------------------------------------
# Dictionary
#------------------------------------------------------------------------------
dict_output="${DIR}/base_dictionary.txt"
echo "Extracting dictionary (${g2p_file})"
tar --to-stdout -xzf "${g2p_file}" 'base_dictionary.txt' > "${dict_output}"
#------------------------------------------------------------------------------
# Language Model
#------------------------------------------------------------------------------
lm_url='https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-en/en-70k-0.2-pruned.lm.gz'
lm_file="${download_dir}/en-70k-0.2-pruned.lm.gz"
lm_output="${DIR}/base_language_model.txt"
if [[ ! -s "${lm_file}" ]]; then
echo "Downloading language model (${lm_url})"
curl -sSfL -o "${lm_file}" "${lm_url}"
fi
echo "Extracting language model (${lm_file})"
zcat "${lm_file}" > "${lm_output}"
#------------------------------------------------------------------------------
# Snowboy
#------------------------------------------------------------------------------
snowboy_models=("snowboy.umdl" "computer.umdl")
for model_name in "${snowboy_models[@]}"; do
model_output="${DIR}/${model_name}"
if [[ ! -s "${model_output}" ]]; then
model_url="https://github.com/Kitt-AI/snowboy/raw/master/resources/models/${model_name}"
echo "Downloading ${model_output} (${model_url})"
curl -sSfL -o "${model_output}" "${model_url}"
fi
done
#------------------------------------------------------------------------------
# Mycroft Precise
#------------------------------------------------------------------------------
precise_files=("hey-mycroft-2.pb" "hey-mycroft-2.pb.params")
for file_name in "${precise_files[@]}"; do
file_output="${DIR}/${file_name}"
if [[ ! -s "${file_output}" ]]; then
file_url="https://github.com/MycroftAI/precise-data/raw/models/${file_name}"
echo "Downloading ${file_output} (${file_url})"
curl -sSfL -o "${file_output}" "${file_url}"
fi
done
#------------------------------------------------------------------------------
# Flair Embeddings
#------------------------------------------------------------------------------
flair_dir="${DIR}/flair/cache/embeddings"
mkdir -p "${flair_dir}"
flair_files=("news-forward-0.4.1.pt" "news-backward-0.4.1.pt")
for file_name in "${flair_files[@]}"; do
file_output="${flair_dir}/${file_name}"
if [[ ! -s "${file_output}" ]]; then
file_url="https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-en/${file_name}"
echo "Downloading ${file_output} (${file_url})"
curl -sSfL -o "${file_output}" "${file_url}"
fi
done
profiles/en/kaldi/custom_words.txt
-1
View File
@@ -1 +0,0 @@
vlc v l 'E s
profiles/en/kaldi/espeak_phonemes.txt
-41
View File
@@ -1,41 +0,0 @@
3 3
A A@
D D
E E
I I
N N
O O:
OI OI
S S
T T
U U
V @
Z Z
aI aI
aU aU
b b
d d
dZ dZ
e e
f f
g g
h h
i i
j j
k k
l l
m m
n n
o o
p p
pf f
r r
s s
t t
tS tS
ts ts
u u:
v v
w w
z z
{ a
profiles/en/kaldi/phoneme_examples.txt
-41
View File
@@ -1,41 +0,0 @@
3 bird b 3 d
A arts A r ts
D than D { n
E bear b E r
I ages e I dZ V z
N bang b { N
O also O l s o U
OI boys b OI z
S bush b U S
T bath b { T
U boat b o U t
V able e I b V l
Z usual j u Z V w V l
aI bike b aI k
aU down d aU n
b babe b e I b
d andy { n d i
dZ edge E dZ
e amen e I m E n
f face f e I s
g bags b { g z
h hadn h { d V n
i area E r i V
j cute k j u t
k back b { k
l alex { l V k s
m army A r m i
n alan { l V n
o blow b l o U
p camp k { m p
pf pfft pf t
r aren A r V n
s asks { s k s
t anti { n t i
tS each i tS
ts gets g E ts
u blue b l u
v dave d e I v
w away V w e I
z arms A r m z
{ adam { d V m
profiles/en/language_model.txt
Executable
+241
View File
@@ -0,0 +1,241 @@
\data\
ngram 1=35
ngram 2=83
ngram 3=110
\1-grams:
-0.725667 </s>
-99 <s> -2.913284
-1.098857 bedroom -1.670246
-1.880682 blue -0.151015
-1.898802 closed -0.143907
-1.898802 cold -0.143907
-1.898802 door -0.143907
-1.255349 garage -0.771521
-1.880682 green -0.151015
-1.898802 hot -0.143907
-1.898802 how -0.143907
-1.255349 is -0.674611
-1.898802 it -0.143907
-1.898802 lamp -0.143907
-1.255349 light -0.231380
-1.255349 living -1.494155
-1.898802 make -0.143907
-1.898802 me -0.143907
-1.361536 off -0.146263
-1.361536 on -0.146263
-1.898802 open -0.143907
-1.880682 orange -0.151015
-1.880682 purple -0.151015
-1.880682 red -0.151015
-1.898802 room -0.143907
-1.898802 set -0.143907
-1.898802 tell -0.143907
-1.898802 temperature -0.143907
-0.984060 the -0.885934
-1.880682 time -0.143907
-1.898802 to -0.143907
-1.898802 turn -0.143907
-1.898802 what -0.143907
-1.898802 whats -0.143907
-1.880682 yellow -0.151015
\2-grams:
-0.875508 <s> how -1.271067
-0.699154 <s> is -1.748188
-1.176986 <s> make -0.970037
-0.875508 <s> set -1.271067
-1.000596 <s> tell -1.447158
-0.699268 <s> turn -0.970037
-1.000596 <s> what -1.447158
-1.176986 <s> whats -1.271067
-0.008854 bedroom light -0.831734
-0.370037 blue </s>
-0.379781 closed </s>
-0.492238 cold is -1.271067
-0.823651 door closed -1.447158
-0.823651 door open -1.447158
-1.232498 garage door -1.447158
-0.105817 garage light -0.793946
-0.370037 green </s>
-0.492238 hot is -1.271067
-0.823651 how cold -1.271067
-0.823651 how hot -1.271067
-0.142294 is it -1.815135
-1.034058 is the -1.748188
-0.379781 it </s>
-0.640077 lamp </s>
-0.902236 lamp off -0.669007
-0.902236 lamp on -0.669007
-0.832360 light </s>
-0.727664 light blue -1.124939
-1.595955 light green -0.726999
-1.299018 light light -0.726999
-1.205894 light off -0.970037
-1.205894 light on -0.970037
-1.595955 light orange -0.726999
-1.595955 light purple -0.726999
-1.595955 light red -0.726999
-1.601397 light to -0.425969
-1.595955 light yellow -0.726999
-0.013966 living room -1.271067
-0.702886 make bedroom -0.970037
-0.666544 make the -0.970037
-0.447904 me the -1.447158
-0.634137 off </s>
-0.983476 off bedroom -0.367977
-1.062123 off garage -0.367977
-1.062123 off living -0.367977
-0.916912 off the -0.367977
-0.634137 on </s>
-0.983476 on bedroom -0.367977
-1.062123 on garage -0.367977
-1.062123 on living -0.367977
-0.916912 on the -0.367977
-0.379781 open </s>
-0.370037 orange </s>
-0.370037 purple </s>
-0.370037 red </s>
-0.535936 room lamp -0.793946
-0.702886 set bedroom -1.271067
-0.666544 set the -1.271067
-0.535936 tell me -1.447158
-0.379781 temperature </s>
-0.445703 the bedroom -1.572097
-0.547019 the garage -1.514105
-0.672945 the living -0.970037
-1.661781 the temperature -1.271067
-1.660389 the time -1.447158
-0.559005 time </s>
-0.742550 time is -1.447158
-1.303271 to blue -0.425969
-1.303271 to green -0.425969
-1.095977 to light -0.425969
-1.303271 to orange -0.425969
-1.303271 to purple -0.425969
-1.303271 to red -0.425969
-1.303271 to yellow -0.425969
-0.982184 turn bedroom -0.669007
-1.061043 turn garage -0.669007
-1.061043 turn living -0.669007
-1.106585 turn off -0.544068
-1.106585 turn on -0.544068
-0.915469 turn the -0.669007
-0.535360 what time -1.447158
-0.447904 whats the -1.271067
-0.370037 yellow </s>
\3-grams:
-0.317625 <s> how cold
-0.317625 <s> how hot
-0.007096 <s> is the
-0.330065 <s> make bedroom
-0.328347 <s> make the
-0.315305 <s> set bedroom
-0.314474 <s> set the
-0.011137 <s> tell me
-1.115549 <s> turn bedroom
-1.126182 <s> turn garage
-1.126182 <s> turn living
-0.618839 <s> turn off
-0.618839 <s> turn on
-0.610552 <s> turn the
-0.011131 <s> what time
-0.015235 <s> whats the
-0.875237 bedroom light </s>
-1.067444 bedroom light blue
-1.209216 bedroom light green
-1.184183 bedroom light light
-1.234501 bedroom light off
-1.234501 bedroom light on
-1.209216 bedroom light orange
-1.209216 bedroom light purple
-1.209216 bedroom light red
-0.619254 bedroom light to
-1.209216 bedroom light yellow
-0.006549 cold is it
-0.009137 door closed </s>
-0.009137 door open </s>
-0.312023 garage door closed
-0.312023 garage door open
-0.327836 garage light </s>
-0.685222 garage light off
-0.685222 garage light on
-0.006549 hot is it
-0.016069 how cold is
-0.016069 how hot is
-0.003892 is it </s>
-0.005590 is the garage
-0.078068 lamp off </s>
-0.078068 lamp on </s>
-0.019092 light blue </s>
-0.049403 light green </s>
-0.071808 light light blue
-0.037282 light off </s>
-0.037282 light on </s>
-0.049403 light orange </s>
-0.049403 light purple </s>
-0.049403 light red </s>
-0.966821 light to blue
-0.966821 light to green
-0.923177 light to light
-0.966821 light to orange
-0.966821 light to purple
-0.966821 light to red
-0.966821 light to yellow
-0.049403 light yellow </s>
-0.016814 living room lamp
-0.000940 make bedroom light
-0.030933 make the bedroom
-0.015443 me the time
-0.003772 off bedroom light
-0.042237 off garage light
-0.005931 off living room
-0.463377 off the bedroom
-0.505712 off the garage
-0.550547 off the living
-0.003772 on bedroom light
-0.042237 on garage light
-0.005931 on living room
-0.463377 on the bedroom
-0.505712 on the garage
-0.550547 on the living
-0.315838 room lamp </s>
-0.664427 room lamp off
-0.664427 room lamp on
-0.000470 set bedroom light
-0.015191 set the bedroom
-0.010097 tell me the
-0.000235 the bedroom light
-0.073849 the garage door
-0.819462 the garage light
-0.001475 the living room
-0.013778 the temperature </s>
-0.011376 the time </s>
-0.004355 time is it
-0.105156 to blue </s>
-0.105156 to green </s>
-0.157887 to light blue
-0.105156 to orange </s>
-0.105156 to purple </s>
-0.105156 to red </s>
-0.105156 to yellow </s>
-0.001882 turn bedroom light
-0.020605 turn garage light
-0.002955 turn living room
-0.903377 turn off bedroom
-0.920814 turn off garage
-0.920814 turn off living
-0.334262 turn off the
-0.903377 turn on bedroom
-0.920814 turn on garage
-0.920814 turn on living
-0.334262 turn on the
-0.470195 turn the bedroom
-0.491181 turn the garage
-0.512284 turn the living
-0.012894 what time is
-0.023377 whats the temperature
\end\
profiles/en/okay-rhasspy.pb
Executable
BIN
View File
Binary file not shown.
profiles/en/okay-rhasspy.pb.params
Executable
+1
View File
@@ -0,0 +1 @@
{"window_t": 0.1, "hop_t": 0.05, "buffer_t": 1.5, "sample_rate": 16000, "sample_depth": 2, "n_mfcc": 13, "n_filt": 20, "n_fft": 512}
profiles/en/profile.json
+16 -87
View File
@@ -1,94 +1,23 @@
{
"language": "en",
"name": "en",
"speech_to_text": {
"system": "pocketsphinx",
"dictionary_casing": "lower",
"kaldi": {
"base_dictionary": "kaldi/base_dictionary.txt",
"base_language_model": "kaldi/base_language_model.txt",
"compatible": true,
"custom_words": "kaldi/custom_words.txt",
"dictionary": "kaldi/dictionary.txt",
"graph": "graph",
"language_model": "kaldi/language_model.txt",
"model_dir": "kaldi/model",
"unknown_words": "kaldi/unknown_words.txt",
"mix_fst": "kaldi/mixed.fst",
"g2p_model": "kaldi/g2p.fst",
"phoneme_examples": "kaldi/phoneme_examples.txt",
"phoneme_map": "kaldi/espeak_phonemes.txt"
}
},
"intent": {
"flair": {
"embeddings": [
"news-forward-0.4.1.pt",
"news-backward-0.4.1.pt"
]
}
},
"text_to_speech": {
"wavenet": {
"language_code": "en-US"
"language": "en",
"name": "en",
"wake": {
"system": "snowboy"
},
"marytts": {
"locale": "en-US"
}
},
"download": {
"conditions": {
"speech_to_text.system": {
"pocketsphinx": {
"acoustic_model": "cmusphinx-en-us-5.2.tar.gz:cmusphinx-en-us-5.2",
"base_dictionary.txt": "en-g2p.tar.gz:base_dictionary.txt",
"g2p.fst": "en-g2p.tar.gz:g2p.fst"
},
"kaldi": {
"kaldi": "en_kaldi-zamia.tar.gz:kaldi"
}
},
"speech_to_text.kaldi.open_transcription": {
"True": {
"kaldi/model/base_graph": "en_kaldi-zamia-base_graph.tar.gz:base_graph"
}
},
"speech_to_text.pocketsphinx.mix_weight": {
">0": {
"base_language_model.txt": "en-70k-0.2-pruned.lm.gz:en-70k-0.2-pruned.lm"
}
},
"intent.system": {
"intent": {
"flair": {
"flair/cache/embeddings/news-forward-0.4.1.pt": "news-forward-0.4.1.pt",
"flair/cache/embeddings/news-backward-0.4.1.pt": "news-backward-0.4.1.pt"
"embeddings": [
"news-forward-0.4.1.pt",
"news-backward-0.4.1.pt"
]
}
}
},
"files": {
"cmusphinx-en-us-5.2.tar.gz": {
"url": "https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-en/cmusphinx-en-us-5.2.tar.gz"
},
"en-70k-0.2-pruned.lm.gz": {
"url": "https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-en/en-70k-0.2-pruned.lm.gz"
},
"en-g2p.tar.gz": {
"url": "https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-en/en-g2p.tar.gz"
},
"news-forward-0.4.1.pt": {
"url": "https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-en/news-forward-0.4.1.pt",
"cache": false
},
"news-backward-0.4.1.pt": {
"url": "https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-en/news-backward-0.4.1.pt",
"cache": false
},
"en_kaldi-zamia.tar.gz": {
"url": "https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-en/en_kaldi-zamia.tar.gz"
},
"en_kaldi-zamia-base_graph.tar.gz": {
"url": "https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-en/en_kaldi-zamia-base_graph.tar.gz"
}
"text_to_speech": {
"wavenet": {
"language_code": "en-US"
},
"marytts": {
"locale": "en-US"
}
}
}
}
profiles/en/slots/other_colors.txt
Executable
+4
View File
@@ -0,0 +1,4 @@
purple
yellow
orange
light blue
profiles/en/snowboy.umdl
Executable
BIN
View File
Binary file not shown.
profiles/es/check-profile.sh
Executable
+67
View File
@@ -0,0 +1,67 @@
#!/usr/bin/env bash
set -e
if [[ -z "$1" ]]; then
echo "Directory required as first argument"
exit 1
fi
DIR="$1"
#------------------------------------------------------------------------------
# Acoustic Model
#------------------------------------------------------------------------------
acoustic_output="${DIR}/acoustic_model"
if [[ ! -d "${acoustic_output}" ]]; then
echo "Missing acoustic model"
exit 1
fi
#------------------------------------------------------------------------------
# G2P
#------------------------------------------------------------------------------
g2p_output="${DIR}/g2p.fst"
if [[ ! -s "${g2p_output}" ]]; then
echo "Missing g2p model"
exit 1
fi
#------------------------------------------------------------------------------
# Dictionary
#------------------------------------------------------------------------------
dict_output="${DIR}/base_dictionary.txt"
if [[ ! -s "${dict_output}" ]]; then
echo "Missing base dictionary"
exit 1
fi
#------------------------------------------------------------------------------
# Language Model
#------------------------------------------------------------------------------
lm_output="${DIR}/base_language_model.txt"
if [[ ! -s "${lm_output}" ]]; then
echo "Missing base language model"
exit 1
fi
#------------------------------------------------------------------------------
# Flair Embeddings
#------------------------------------------------------------------------------
flair_dir="${DIR}/flair/cache/embeddings"
flair_files=("lm-es-forward-fast.pt" "lm-es-backward-fast.pt")
for file_name in "${flair_files[@]}"; do
file_output="${flair_dir}/${file_name}"
if [[ ! -s "${file_output}" ]]; then
echo "Missing flair embedding (${file_name})"
exit 1
fi
done
profiles/es/download-profile.sh
Executable
+111
View File
@@ -0,0 +1,111 @@
#!/usr/bin/env bash
set -e
if [[ -z "$1" ]]; then
echo "Directory required as first argument"
exit 1
fi
DIR="$1"
download_dir="${DIR}/download"
if [[ "$2" = "--delete" ]]; then
rm -rf "${download_dir}"
fi
mkdir -p "${download_dir}"
echo "Downloading Spanish (es) profile (sphinx)"
#------------------------------------------------------------------------------
# Acoustic Model
#------------------------------------------------------------------------------
acoustic_url='https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-es/cmusphinx-es-5.2.tar.gz'
acoustic_file="${download_dir}/cmusphinx-es-5.2.tar.gz"
acoustic_output="${DIR}/acoustic_model"
if [[ ! -s "${acoustic_file}" ]]; then
echo "Downloading acoustic model (${acoustic_url})"
curl -sSfL -o "${acoustic_file}" "${acoustic_url}"
fi
echo "Extracting acoustic model (${acoustic_file})"
rm -rf "${acoustic_output}"
tar -C "${DIR}" -xf "${acoustic_file}" "cmusphinx-es-5.2/model_parameters/voxforge_es_sphinx.cd_ptm_4000/" && mv "${DIR}/cmusphinx-es-5.2/model_parameters/voxforge_es_sphinx.cd_ptm_4000/" "${acoustic_output}" && rm -rf "${DIR}/cmusphinx-es-5.2"
#------------------------------------------------------------------------------
# G2P
#------------------------------------------------------------------------------
g2p_url='https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-es/es-g2p.tar.gz'
g2p_file="${download_dir}/es-g2p.tar.gz"
g2p_output="${DIR}/g2p.fst"
if [[ ! -s "${g2p_file}" ]]; then
echo "Downloading g2p model (${g2p_url})"
curl -sSfL -o "${g2p_file}" "${g2p_url}"
fi
echo "Extracting g2p model (${g2p_file})"
tar --to-stdout -xzf "${g2p_file}" 'g2p.fst' > "${g2p_output}"
#------------------------------------------------------------------------------
# Dictionary
#------------------------------------------------------------------------------
dict_output="${DIR}/base_dictionary.txt"
echo "Extracting dictionary (${acoustic_file})"
tar --to-stdout -xf "${acoustic_file}" "cmusphinx-es-5.2/etc/voxforge_es_sphinx.dic" > "${dict_output}"
#------------------------------------------------------------------------------
# Language Model
#------------------------------------------------------------------------------
lm_output="${DIR}/base_language_model.txt"
echo "Extracting language model (${acoustic_file})"
tar --to-stdout -xf "${acoustic_file}" "cmusphinx-es-5.2/etc/es-20k.lm.gz" | zcat > "${lm_output}"
#------------------------------------------------------------------------------
# Snowboy
#------------------------------------------------------------------------------
snowboy_models=("snowboy.umdl" "computer.umdl")
for model_name in "${snowboy_models[@]}"; do
model_output="${DIR}/${model_name}"
if [[ ! -s "${model_output}" ]]; then
model_url="https://github.com/Kitt-AI/snowboy/raw/master/resources/models/${model_name}"
echo "Downloading ${model_output} (${model_url})"
curl -sSfL -o "${model_output}" "${model_url}"
fi
done
#------------------------------------------------------------------------------
# Mycroft Precise
#------------------------------------------------------------------------------
precise_files=("hey-mycroft-2.pb" "hey-mycroft-2.pb.params")
for file_name in "${precise_files[@]}"; do
file_output="${DIR}/${file_name}"
if [[ ! -s "${file_output}" ]]; then
file_url="https://github.com/MycroftAI/precise-data/raw/models/${file_name}"
echo "Downloading ${file_output} (${file_url})"
curl -sSfL -o "${file_output}" "${file_url}"
fi
done
#------------------------------------------------------------------------------
# Flair Embeddings
#------------------------------------------------------------------------------
flair_dir="${DIR}/flair/cache/embeddings"
mkdir -p "${flair_dir}"
flair_files=("lm-es-forward-fast.pt" "lm-es-backward-fast.pt")
for file_name in "${flair_files[@]}"; do
file_output="${flair_dir}/${file_name}"
if [[ ! -s "${file_output}" ]]; then
file_url="https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-es/${file_name}"
echo "Downloading ${file_output} (${file_url})"
curl -sSfL -o "${file_output}" "${file_url}"
fi
done
profiles/es/ipa_examples.txt
+25
View File
@@ -0,0 +1,25 @@
a abre əbɹˈɛ
b baja bˈɑhɑ
ch ocho ˈɒtʃəʊ
d anda ˈændə
e aire ˈɛə
f café kæfˈeɪ
g agua ˈæɡjʉːə
gn años sˈænsəːn
i allí ˈælɪ
j bajo bˈɑdʒəʊ
k acto ˈæktəʊ
l alex ˈælɪks
ll allá əlˈɑ
m alma ˈɒlmə
n andy ˈændɪ
o algo ˈælɡəʊ
p copa kˈəʊpə
r amor ˈæmɚ
rr raro ɹˈæɹəʊ
s base bˈeɪs
t alta ˈɔːltə
u auto ˈɔːtəʊ
x sexo sˈiksəʊ
y ayer ˈaɪɚ
z azul æzˈʉːl
profiles/es/ipa_phonemes.txt
+25
View File
@@ -0,0 +1,25 @@
a abre a
b baja b
ch ocho tʃ
d anda d
e aire e
f café f
g agua g
gn años ɲ
i allí i
j bajo j
k acto k
l alex l
ll allá ʎ
m alma m
n andy n
o algo o
p copa p
r amor r
rr raro ɾ
s base s
t alta t
u auto u
x sexo x
y ayer ʝ
z azul z
profiles/es/phoneme_examples.txt
+25 -22
View File
@@ -1,22 +1,25 @@
A A A
B VEZ B E Z
CH MUCHO M U CH O
D DE D E
E EL E L
F FUE F U E
G ALGO A L G O
GN AñOS A GN O S
I Y I
K CON K O N
L AL A L
LL ELLA E LL A
M ME M E
N EN E N
O O O
P POR P O R
R ERA E R A
S ES E S
T TE T E
U SU S U
Y YA Y A
Z HACE A Z E
a abre a b r e
b baja b a j a
ch ocho o ch o
d anda a n d a
e aire a i r e
f café k a f e
g agua a g u a
gn años a gn o s
i allí a ll i
j bajo b a j o
k acto a k t o
l alex a l e x
ll allá a ll a
m alma a l m a
n andy a n d i
o algo a l g o
p copa k o p a
r amor a m o r
rr raro rr a r o
s base b a s e
t alta a l t a
u auto a u t o
x sexo s e x o
y ayer a y e r
z azul a z u l
profiles/es/profile.json
+2 -44
View File
@@ -1,9 +1,8 @@
{
"language": "es",
"name": "es",
"speech_to_text": {
"system": "pocketsphinx",
"dictionary_casing": "lower"
"wake": {
"system": "snowboy"
},
"text_to_speech": {
"wavenet": {
@@ -17,46 +16,5 @@
"lm-es-backward-fast.pt"
]
}
},
"download": {
"conditions": {
"speech_to_text.system": {
"pocketsphinx": {
"acoustic_model": "cmusphinx-es-5.2.tar.gz:cmusphinx-es-5.2/model_parameters/voxforge_es_sphinx.cd_ptm_4000/",
"base_dictionary.txt": "cmusphinx-es-5.2.tar.gz:cmusphinx-es-5.2/etc/voxforge_es_sphinx.dic",
"g2p.fst": "es-g2p.tar.gz:g2p.fst"
}
},
"speech_to_text.pocketsphinx.mix_weight": {
">0": {
"base_language_model.txt": "cmusphinx-es-5.2.tar.gz:cmusphinx-es-5.2/etc/es-20k.lm.gz:"
}
},
"intent.system": {
"flair": {
"flair/cache/embeddings/lm-es-forward-fast.pt": "lm-es-forward-fast.pt",
"flair/cache/embeddingslm-es-backward-fast.pt": "lm-es-backward-fast.pt"
}
}
},
"files": {
"cmusphinx-es-5.2.tar.gz": {
"url": "https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-es/cmusphinx-es-5.2.tar.gz"
},
"es-g2p.tar.gz": {
"url": "https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-es/es-g2p.tar.gz"
},
"lm-es-forward-fast.pt": {
"url": "https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-es/lm-es-forward-fast.pt",
"cache": false
},
"lm-es-backward-fast.pt": {
"url": "https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-es/lm-es-backward-fast.pt",
"cache": false
}
}
}
}
profiles/es/stop_words.txt
-313
View File
@@ -1,313 +0,0 @@
de
la
que
el
en
y
a
los
del
se
las
por
un
para
con
no
una
su
al
lo
como
más
pero
sus
le
ya
o
este
sí
porque
esta
entre
cuando
muy
sin
sobre
también
me
hasta
hay
donde
quien
desde
todo
nos
durante
todos
uno
les
ni
contra
otros
ese
eso
ante
ellos
e
esto
mí
antes
algunos
qué
unos
yo
otro
otras
otra
él
tanto
esa
estos
mucho
quienes
nada
muchos
cual
poco
ella
estar
estas
algunas
algo
nosotros
mi
mis
tú
te
ti
tu
tus
ellas
nosotras
vosostros
vosostras
os
mío
mía
míos
mías
tuyo
tuya
tuyos
tuyas
suyo
suya
suyos
suyas
nuestro
nuestra
nuestros
nuestras
vuestro
vuestra
vuestros
vuestras
esos
esas
estoy
estás
está
estamos
estáis
están
esté
estés
estemos
estéis
estén
estaré
estarás
estará
estaremos
estaréis
estarán
estaría
estarías
estaríamos
estaríais
estarían
estaba
estabas
estábamos
estabais
estaban
estuve
estuviste
estuvo
estuvimos
estuvisteis
estuvieron
estuviera
estuvieras
estuviéramos
estuvierais
estuvieran
estuviese
estuvieses
estuviésemos
estuvieseis
estuviesen
estando
estado
estada
estados
estadas
estad
he
has
ha
hemos
habéis
han
haya
hayas
hayamos
hayáis
hayan
habré
habrás
habrá
habremos
habréis
habrán
habría
habrías
habríamos
habríais
habrían
había
habías
habíamos
habíais
habían
hube
hubiste
hubo
hubimos
hubisteis
hubieron
hubiera
hubieras
hubiéramos
hubierais
hubieran
hubiese
hubieses
hubiésemos
hubieseis
hubiesen
habiendo
habido
habida
habidos
habidas
soy
eres
es
somos
sois
son
sea
seas
seamos
seáis
sean
seré
serás
será
seremos
seréis
serán
sería
serías
seríamos
seríais
serían
era
eras
éramos
erais
eran
fui
fuiste
fue
fuimos
fuisteis
fueron
fuera
fueras
fuéramos
fuerais
fueran
fuese
fueses
fuésemos
fueseis
fuesen
sintiendo
sentido
sentida
sentidos
sentidas
siente
sentid
tengo
tienes
tiene
tenemos
tenéis
tienen
tenga
tengas
tengamos
tengáis
tengan
tendré
tendrás
tendrá
tendremos
tendréis
tendrán
tendría
tendrías
tendríamos
tendríais
tendrían
tenía
tenías
teníamos
teníais
tenían
tuve
tuviste
tuvo
tuvimos
tuvisteis
tuvieron
tuviera
tuvieras
tuviéramos
tuvierais
tuvieran
tuviese
tuvieses
tuviésemos
tuvieseis
tuviesen
teniendo
tenido
tenida
tenidos
tenidas
tened
profiles/fr/check-profile.sh
Executable
+67
View File
@@ -0,0 +1,67 @@
#!/usr/bin/env bash
set -e
if [[ -z "$1" ]]; then
echo "Directory required as first argument"
exit 1
fi
DIR="$1"
#------------------------------------------------------------------------------
# Acoustic Model
#------------------------------------------------------------------------------
acoustic_output="${DIR}/acoustic_model"
if [[ ! -d "${acoustic_output}" ]]; then
echo "Missing acoustic model"
exit 1
fi
#------------------------------------------------------------------------------
# G2P
#------------------------------------------------------------------------------
g2p_output="${DIR}/g2p.fst"
if [[ ! -s "${g2p_output}" ]]; then
echo "Missing g2p model"
exit 1
fi
#------------------------------------------------------------------------------
# Dictionary
#------------------------------------------------------------------------------
dict_output="${DIR}/base_dictionary.txt"
if [[ ! -s "${dict_output}" ]]; then
echo "Missing base dictionary"
exit 1
fi
#------------------------------------------------------------------------------
# Language Model
#------------------------------------------------------------------------------
lm_output="${DIR}/base_language_model.txt"
if [[ ! -s "${lm_output}" ]]; then
echo "Missing base language model"
exit 1
fi
#------------------------------------------------------------------------------
# Flair Embeddings
#------------------------------------------------------------------------------
flair_dir="${DIR}/flair/cache/embeddings"
flair_files=("lm-fr-charlm-forward.pt" "lm-fr-charlm-backward.pt")
for file_name in "${flair_files[@]}"; do
file_output="${flair_dir}/${file_name}"
if [[ ! -s "${file_output}" ]]; then
echo "Missing flair embedding (${file_name})"
exit 1
fi
done
profiles/fr/download-profile.sh
Executable
+119
View File
@@ -0,0 +1,119 @@
#!/usr/bin/env bash
set -e
if [[ -z "$1" ]]; then
echo "Directory required as first argument"
exit 1
fi
DIR="$1"
download_dir="${DIR}/download"
if [[ "$2" = "--delete" ]]; then
rm -rf "${download_dir}"
fi
mkdir -p "${download_dir}"
echo "Downloading French (fr) profile (sphinx)"
#------------------------------------------------------------------------------
# Acoustic Model
#------------------------------------------------------------------------------
acoustic_url='https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-fr/cmusphinx-fr-5.2.tar.gz'
acoustic_file="${download_dir}/cmusphinx-fr-5.2.tar.gz"
acoustic_output="${DIR}/acoustic_model"
if [[ ! -s "${acoustic_file}" ]]; then
echo "Downloading acoustic model (${acoustic_url})"
curl -sSfL -o "${acoustic_file}" "${acoustic_url}"
fi
echo "Extracting acoustic model (${acoustic_file})"
rm -rf "${acoustic_output}"
tar -C "${DIR}" -xf "${acoustic_file}" && mv "${DIR}/cmusphinx-fr-5.2" "${acoustic_output}"
#------------------------------------------------------------------------------
# G2P
#------------------------------------------------------------------------------
g2p_url='https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-fr/fr-g2p.tar.gz'
g2p_file="${download_dir}/fr-g2p.tar.gz"
g2p_output="${DIR}/g2p.fst"
if [[ ! -s "${g2p_file}" ]]; then
echo "Downloading g2p model (${g2p_url})"
curl -sSfL -o "${g2p_file}" "${g2p_url}"
fi
echo "Extracting g2p model (${g2p_file})"
tar --to-stdout -xzf "${g2p_file}" 'g2p.fst' > "${g2p_output}" || exit 1
#------------------------------------------------------------------------------
# Dictionary
#------------------------------------------------------------------------------
dict_output="${DIR}/base_dictionary.txt"
echo "Extracting dictionary (${g2p_file})"
tar --to-stdout -xzf "${g2p_file}" 'base_dictionary.txt' > "${dict_output}" || exit 1
#------------------------------------------------------------------------------
# Language Model
#------------------------------------------------------------------------------
lm_url='https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-fr/fr-small.lm.gz'
lm_file="${download_dir}/fr-small.lm.gz"
lm_output="${DIR}/base_language_model.txt"
if [[ ! -s "${lm_file}" ]]; then
echo "Downloading language model (${lm_url})"
curl -sSfL -o "${lm_file}" "${lm_url}"
fi
echo "Extracting language model (${lm_file})"
zcat "${lm_file}" > "${lm_output}" || exit 1
#------------------------------------------------------------------------------
# Snowboy
#------------------------------------------------------------------------------
snowboy_models=("snowboy.umdl" "computer.umdl")
for model_name in "${snowboy_models[@]}"; do
model_output="${DIR}/${model_name}"
if [[ ! -s "${model_output}" ]]; then
model_url="https://github.com/Kitt-AI/snowboy/raw/master/resources/models/${model_name}"
echo "Downloading ${model_output} (${model_url})"
curl -sSfL -o "${model_output}" "${model_url}"
fi
done
#------------------------------------------------------------------------------
# Mycroft Precise
#------------------------------------------------------------------------------
precise_files=("hey-mycroft-2.pb" "hey-mycroft-2.pb.params")
for file_name in "${precise_files[@]}"; do
file_output="${DIR}/${file_name}"
if [[ ! -s "${file_output}" ]]; then
file_url="https://github.com/MycroftAI/precise-data/raw/models/${file_name}"
echo "Downloading ${file_output} (${file_url})"
curl -sSfL -o "${file_output}" "${file_url}"
fi
done
#------------------------------------------------------------------------------
# Flair Embeddings
#------------------------------------------------------------------------------
flair_dir="${DIR}/flair/cache/embeddings"
mkdir -p "${flair_dir}"
flair_files=("lm-fr-charlm-forward.pt" "lm-fr-charlm-backward.pt")
for file_name in "${flair_files[@]}"; do
file_output="${flair_dir}/${file_name}"
if [[ ! -s "${file_output}" ]]; then
file_url="https://github.com/synesthesiam/rhasspy-profiles/releases/download/v1.0-fr/${file_name}"
echo "Downloading ${file_output} (${file_url})"
curl -sSfL -o "${file_output}" "${file_url}"
fi
done

Some files were not shown because too many files have changed in this diff Show More