v2.0.0-rc.6 #280
decahedron1
announced in
Announcements
v2.0.0-rc.6
#280
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
ort::Errorrefactorort::Erroris no longer an enum, but rather an opaque struct with a message and a newErrorCodefield.ort::Errorstill implementsstd::error::Error, so this change shouldn't be too breaking; however, if you were previouslymatching onort::Errors, you'll have to refactor your code to instead match on the error's code (acquired with theError::code()function).AllocationDevicerefactorThe
AllocationDevicetype has also been converted from an enum to a struct. Common devices like CUDA or DirectML are accessible via associated constants likeAllocationDevice::CUDA&AllocationDevice::DIRECTML.Features
ModelMetadata::custom_keys()to get a Vec of all custom keys.SessionBuilderoptions affecting compute & graph optimizations.AllocatorAPI. You can now allocate & free buffers acquired from a session or operator kernel context.ValueType::Optional.KernelContext::par_for, allowing operator kernels to use ONNX Runtime's thread pool without needing an extra dependency on a crate like rayon.Fixes
Tensors from&CowArrays.tracing'sattributesfeature - a--no-default-featuresbuild ofortnow only builds 9 crates!operator-librariesfeature - you can still useSessionBuilder::with_operator_library, it's just no longer gated behind the feature!If you have any questions about this release, we're here to help:
#💬|ort-discussionsLove
ort? Consider supporting us on Open Collective 💖❤️💚💙💛
This discussion was created from the release v2.0.0-rc.6.
Beta Was this translation helpful? Give feedback.
All reactions