Adding iree_vm_async_invoke API for loop-based invocation. This reuses the low-level iree_vm_*_invoke APIs to provide a fire-and-forget callback-style interface. Storage is externalized such that it can be embedded in higher-level binding data structures and avoid allocations. The API here is unopinionated with respect to overlapping invocations within the same context. Callers are expected to set the IREE_VM_CONTEXT_FLAG_CONCURRENT if they will overlap (mostly just to make tracing work) but otherwise there's no difference in the implementation between sequential and concurrent invocation execution.
IREE (Intermediate Representation Execution Environment, pronounced as “eerie”) is an MLIR-based end-to-end compiler and runtime that lowers Machine Learning (ML) models to a unified IR that scales up to meet the needs of the datacenter and down to satisfy the constraints and special considerations of mobile and edge deployments.
See our website for project details, user guides, and instructions on building from source.
IREE is still in its early phase. We have settled down on the overarching infrastructure and are actively improving various software components as well as project logistics. It is still quite far from ready for everyday use and is made available without any support at the moment. With that said, we welcome any kind of feedback on any communication channels!
See our website for more information.
IREE is licensed under the terms of the Apache 2.0 License with LLVM Exceptions. See LICENSE for more information.