Age | Commit message (Collapse) | Author | Files | Lines |
|
V1:
- Removed the following recipes from meta-offline-voiceagent:
- python3-scipy
- python3-scikit-learn
- python3-openblas
- python3-lapack
- python3-numpy
- Updated to use meta-python-ai layer for:
- python3-scipy
- python3-scikit-learn
- python3-openblas
- python3-lapack
- python3-numpy
- python3-pytorch
- Added all the required recipes for Whisper AI:
- Included recipes for Whisper AI base and tiny model
- Updated recipes for:
- python3-uvloop
- python3-python-crfsuite
V2:
- Remove vulkan from DISTRO_FEATURES as it has become coditional
at upstream
- Fix flutter_voiceassistant service
- Update recipe of llvmlite to include the wheel for aarch64 also.
V3:
- Fix llvmlite recipe and some patches
- Removes rasa and its dependencies
- Fix agl-service-voiceagent.service
Bug-AGL: SPEC-5200
Change-Id: I429e7d4b98fbbe1a4f8c1341ac8918d7d177b382
Signed-off-by: Anuj Solanki <anuj603362@gmail.com>
|
|
Replace 'virtual/vosk-model' RPROVIDES use with just 'vosk-model'
to future-proof for Yocto Project 5.0 / Scarthgap, where it triggers
a warning. Also update checksums in a couple of the model recipes
to fix errors seen during test builds.
Bug-AGL: SPEC-5067
Change-Id: I6d39342059c93d7bef77e9b65ff11a9b2aee51aa
Signed-off-by: Scott Murray <scott.murray@konsulko.com>
|
|
Change the SRCURI to use gerrit repos. Update agl-service-voicegent
recipe to correctly package all the required content. Add recipe for
another en-us vosk model. Add service files for RASA and Voice Agent
server.
Bug-AGL: SPEC-4906
Change-Id: Ic9f77386ff5c545c2a80433dd538c94bd8763e9e
Signed-off-by: Malik Talha <talhamalik727x@gmail.com>
|
|
To speed up the build process, increase the default values for BAZEL_JOBS to HOST_CPUS*.4
and for BAZEL_MEM to HOST_RAM*0.4 .
Bug-AGL: SPEC-4856
Change-Id: I3a167df914373979ea39b7ccbd508de15861c115
Signed-off-by: Jan-Simon Moeller <jsmoeller@linuxfoundation.org>
|
|
This adds the libraries for the RASA machine learning framework.
Bug-AGL: SPEC-4856
Change-Id: I477b09cae8ff51e284b51238cb5940464ea57a51
Signed-off-by: Malik Talha <talhamalik727x@gmail.com>
|
|
This brings the voice agent to the state shown at AMM and EW.
Change-Id: I31ac4adc874a206e6e598ba116523ecc7eb99fc9
Signed-off-by: amanarora_09 <aman.arora9848@gmail.com>
Signed-off-by: Jan-Simon Moeller <jsmoeller@linuxfoundation.org>
|
|
- This enables offline voice recognition in AGL
- Can successfully run vosk API and vosk websocket server
Changes:
- Create a new layer: meta-offline-voice-recognition
- Create recipes for the the Vosk library, API, websocket server and model
- Create feature template for the layer (agl-offline-voice-agent)
- Add all required packages to IMAGE_INSTALL
- ptest enable recipe python3-vosk-api
- Few fixes
Bug-AGL: SPEC-4497
Change-Id: Icb290ceee22e2c3a51cbd67f7431701795f1201b
Signed-off-by: amanarora_09 <aman.arora9848@gmail.com>
|