commit | 8dc07e642a6ac644cea2b5e795445b020aca6fbc | [log] [tgz] |
---|---|---|
author | Ben Vanik <benvanik@google.com> | Wed Aug 25 14:03:46 2021 -0700 |
committer | Ben Vanik <benvanik@google.com> | Wed Aug 25 14:30:11 2021 -0700 |
tree | 132031f2ae74cc9e42125287f1610ada2c759297 | |
parent | 151b2d9ccf0784c2cb51d055af7db6a5023a5722 [diff] |
Adding HAL allocator statistics interface. Each HAL backend gets the ability to track statistics as it pleases. This is useful for allocators that can adjust allocation sizes or are performing suballocating from larger block pools, etc. The default heap allocator is the only one populating the stats today. Future extensions would be to track things like imported/mapped memory or reach into VMA in Vulkan to query its statistics. Adding --print_statistics to iree-run-module prints: ``` [[ iree_hal_allocator_t memory statistics ]] HOST_LOCAL: 0b peak / 0b allocated / 0b freed / 0b live DEVICE_LOCAL: 106340864b peak / 106340864b allocated / 106340864b freed / 0b live ``` Fixes #6845.
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!
IREE is licensed under the terms of the Apache 2.0 License with LLVM Exceptions. See LICENSE for more information.