Closed Kirollos-Henry closed 2 years ago
thud
is not included in the confirmed version of meta-tensorflow-lite.
The version (LAYERSERIES_COMPAT) is below.
Perhaps adding thud
to the list will allow the build.
thanks for your reply.
I have already noticed that and I already did this before, but I have got an error that "python3-wheel-native" is not provided.
I think this is native in the other distributions or do you have any idea to add and build it.
Thank you in advance,
Best Regards, Kirollos Henry
Are you adding a meta-python
layer?
e.g)
bitbake-layers add-layer _PATH_TO_/meta-openembedded/meta-python/
unfortunately it is already added
meta-processor-sdk /home/ave-lap-99/processor-sdk-yocto/tisdk/sources/meta-processor-sdk 15 meta-aws /home/ave-lap-99/processor-sdk-yocto/tisdk/sources/meta-aws 6 meta-ros /home/ave-lap-99/processor-sdk-yocto/tisdk/sources/meta-ros 7 meta-arago-distro /home/ave-lap-99/processor-sdk-yocto/tisdk/sources/meta-arago/meta-arago-distro 10 meta-arago-extras /home/ave-lap-99/processor-sdk-yocto/tisdk/sources/meta-arago/meta-arago-extras 10 meta-browser /home/ave-lap-99/processor-sdk-yocto/tisdk/sources/meta-browser 7 meta-qt5 /home/ave-lap-99/processor-sdk-yocto/tisdk/sources/meta-qt5 7 meta-virtualization /home/ave-lap-99/processor-sdk-yocto/tisdk/sources/meta-virtualization 8 meta-networking /home/ave-lap-99/processor-sdk-yocto/tisdk/sources/meta-openembedded/meta-networking 5 meta-python /home/ave-lap-99/processor-sdk-yocto/tisdk/sources/meta-openembedded/meta-python 7 meta-oe /home/ave-lap-99/processor-sdk-yocto/tisdk/sources/meta-openembedded/meta-oe 6 meta-gnome /home/ave-lap-99/processor-sdk-yocto/tisdk/sources/meta-openembedded/meta-gnome 7 meta-multimedia /home/ave-lap-99/processor-sdk-yocto/tisdk/sources/meta-openembedded/meta-multimedia 6 meta-filesystems /home/ave-lap-99/processor-sdk-yocto/tisdk/sources/meta-openembedded/meta-filesystems 6 meta-ti /home/ave-lap-99/processor-sdk-yocto/tisdk/sources/meta-ti 6 meta-linaro-toolchain /home/ave-lap-99/processor-sdk-yocto/tisdk/sources/meta-linaro/meta-linaro-toolchain 30 meta-optee /home/ave-lap-99/processor-sdk-yocto/tisdk/sources/meta-linaro/meta-optee 8 meta /home/ave-lap-99/processor-sdk-yocto/tisdk/sources/oe-core/meta 5 meta-custom /home/ave-lap-99/processor-sdk-yocto/tisdk/meta-custom 6 meta-scipy /home/ave-lap-99/processor-sdk-yocto/tisdk/sources/meta-scipy 6 meta-tensorflow-lite /home/ave-lap-99/processor-sdk-yocto/tisdk/sources/meta-tensorflow-lite 6
and off course there is no other way to build and install python3 tensorflow on yocto right?
This is because there is no python3-wheel recipe in the thud branch of meta-openembedded (after warrior).
If possible, build with warrior or later.
or You may be able to use python3-wheel-native in the meta-sca layer (I have never confirmed).
The problem that I am not using native yocto it is arago project a yocto project based on thud a product of Texas instruments for their hardware.
I will check with Texas instruments if there is way build different distro for the same board and I will also check the layer and recipes you mentioned. I will check and try all this and update you.
Thank you so much for your help.
The problem that I am not using native yocto it is arago project a yocto project based on thud a product of Texas instruments for their hardware.
I will check with Texas instruments if there is way build different distro for the same board and I will also check the layer and recipes you mentioned. I will check and try all this and update you.
Thank you so much for your help.
@Kirollos-Henry Did you find the solution?
Hello all,
I am using arago ti distribution on TDA2Px-EVM and I was able to build the file system and created custom layers to build and install new packages required for some python packages and also added meta-scipy layer. When I am trying to add meta-tensorflow-lite layer I always gets that error
"ERROR: Layer meta-tensorflow-lite is not compatible with the core layer which only supports these series: thud".
The Arago project of Texas instruments that is compatible with my board is based on thud so is there anyway to build and install tensorflow lite on the that arago project base on thud.
Thanks in Advance.
Best Regards, Kirollos Henry