commit | cc44a855465ec783880275ff2168cd753c40b384 | [log] [tgz] |
---|---|---|
author | saienduri <77521230+saienduri@users.noreply.github.com> | Fri Aug 23 17:20:18 2024 -0700 |
committer | GitHub <noreply@github.com> | Sat Aug 24 00:20:18 2024 +0000 |
tree | 789420a9cee4946030c1b5c60cd38ade0b045a70 | |
parent | eb694f10eae82dde5f694d674ccf34fdfb562ba4 [diff] |
Rework special model testing to avoid shared cache interference. (#18344) This commit reworks the special model testing for sdxl and sd3 so that different instances generating compiler artifacts don't write to the same shared cache. This is a major problem because the things we are running for each test are no longer consistent. On top of that, benchmarking was picking up from this same cache, so there were a bunch of failures leading to chaos. Now, the output artifacts are written to a place local to each instance running, so there should be no attempts to write to the same cache in between jobs. We still could use a better longterm solution to avoid this problem for our input artifacts (mlir, weights), but for now we are good as they rarely change and we can be strategic with when we update them. Huggingface seems like a good option to move toward Fixes https://github.com/iree-org/iree/issues/18336 --------- Signed-off-by: saienduri <saimanas.enduri@amd.com>
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
Package | Release status |
---|---|
GitHub release (stable) | |
GitHub release (nightly) | |
Python iree-compiler | |
Python iree-runtime |
Host platform | Build status |
---|---|
Linux | |
macOS | |
Windows |
For the full list of workflows see https://iree.dev/developers/general/github-actions/.
See our website for more information.
Community meeting recordings: IREE YouTube channel
IREE is licensed under the terms of the Apache 2.0 License with LLVM Exceptions. See LICENSE for more information.