autoware.universe
autoware.universe copied to clipboard
feat(tvm_utility): port tvm_utility
Description
This package adds tvm utility, used for ML inference. It has been ported from Autoware.Auto. It depends on https://github.com/autowarefoundation/autoware.universe/pull/892
Related links
https://github.com/autowarefoundation/autoware.universe/issues/628 https://github.com/autowarefoundation/autoware/discussions/226
Tests performed
Notes for reviewers
It's probably easier to wait for https://github.com/autowarefoundation/autoware.universe/pull/892 to be merged before reviewing this.
Pre-review checklist for the PR author
The PR author must check the checkboxes below when creating the PR.
- [x] I've confirmed the contribution guidelines.
- [x] The PR follows the pull request guidelines.
In-review checklist for the PR reviewers
The PR reviewers must check the checkboxes below before approval.
- [ ] The PR follows the pull request guidelines.
- [ ] The PR has been properly tested.
- [ ] The PR has been reviewed by the code owners.
Post-review checklist for the PR author
The PR author must check the checkboxes below before merging.
- [ ] There are no open discussions or they are tracked via tickets.
- [ ] The PR is ready for merge.
After all checkboxes are checked, anyone who has write access can merge the PR.
Awesome
@LucaFos Pls fix some CI errors :pray:
@yukkysaito I have a problem with fixing the CI on that patch: the linting from colcon test
disagrees with the linting from pre-commit
.
pre-commit
chooses style like this: ci(pre-commit): autofix but colcon test
doesn't: feat(tvm_utility): port tvm_utility build-and-test-differential.
What should I do?
I see. CI is updated recently, so I closed and reopen this PR to re-run CI
@kenji-miyake can you help about CI erros? :pray:
@yukkysaito I've fixed the linting/formatting issues.
The tvm_vendor package has been updated last week with a humble release, but it hasn't had time to make it in the ros repos, so the humble CI build is failing. Does it need to pass for this patch to be merged?
Note: I can't close Kenji's comment as I am not the PR author.
@ambroise-arm I cannot merge on the system if all required CI tasks are passed. :pray:
But it is blocked by build-and-test-differential / build-and-test-differential (humble)
for humble migration.
@kenji-miyake
pls advise something :pray:
It seems we have to release tvm_vendor to ROS buildfarm. https://github.com/autowarefoundation/autoware.universe/runs/6671634960?check_suite_focus=true
updated cache in /github/home/.ros/rosdep/sources.cache
E: Unable to locate package ros-humble-tvm-vendor
ERROR: the following rosdeps failed to install
apt: command [apt-get install -y -qq ros-humble-tvm-vendor] failed
executing command [apt-get install -y -qq ros-humble-tvm-vendor]
Or, please add the source dependency to both of the following files:
- https://github.com/autowarefoundation/autoware/blob/humble/autoware.repos
- https://github.com/autowarefoundation/autoware.universe/blob/main/build_depends.humble.repos
Probably @wep21 -san is working on the release now? :thinking: https://github.com/autowarefoundation/tvm_vendor/pull/10
Yes, I faced the issue which caused only in the ros2 ci environment. I am trying to reproduce the issue in the local environment by Steven's advice.
Now it's released? I'll re-run the CI. https://discourse.ros.org/t/new-packages-for-ros-2-humble-hawksbill-2022-06-10/25995
-> Fixed! Thank you @wep21 for your great work!
Codecov Report
Merging #893 (6cf2e58) into main (15f5b57) will decrease coverage by
1.29%
. The diff coverage isn/a
.
:exclamation: Current head 6cf2e58 differs from pull request most recent head 842b99a. Consider uploading reports for the commit 842b99a to get more accurate results
@@ Coverage Diff @@
## main #893 +/- ##
=========================================
- Coverage 10.37% 9.08% -1.30%
=========================================
Files 1135 1031 -104
Lines 79579 70063 -9516
Branches 18548 14640 -3908
=========================================
- Hits 8258 6366 -1892
+ Misses 63224 58242 -4982
+ Partials 8097 5455 -2642
Flag | Coverage Δ | |
---|---|---|
total | 9.06% <0.00%> (-1.30%) |
:arrow_down: |
Impacted Files | Coverage Δ | |
---|---|---|
...ator/nodes/costmap_generator/points_to_costmap.cpp | 0.00% <0.00%> (-82.36%) |
:arrow_down: |
...wer/include/trajectory_follower/mpc_trajectory.hpp | 0.00% <0.00%> (-50.00%) |
:arrow_down: |
...tor/nodes/costmap_generator/objects_to_costmap.cpp | 0.00% <0.00%> (-33.34%) |
:arrow_down: |
...odel/vehicle_model_bicycle_kinematics_no_delay.cpp | 73.91% <0.00%> (-17.40%) |
:arrow_down: |
...vehicle_model/sim_model_ideal_steer_acc_geared.cpp | 71.69% <0.00%> (-11.33%) |
:arrow_down: |
...vehicle_model/sim_model_delay_steer_acc_geared.cpp | 71.26% <0.00%> (-6.90%) |
:arrow_down: |
...ory_follower/src/longitudinal_controller_utils.cpp | 76.56% <0.00%> (-4.69%) |
:arrow_down: |
...ehicle/vehicle_info_util/src/vehicle_info_util.cpp | 27.27% <0.00%> (-4.55%) |
:arrow_down: |
...ulator/vehicle_model/sim_model_ideal_steer_acc.cpp | 91.66% <0.00%> (-4.17%) |
:arrow_down: |
...ulator/vehicle_model/sim_model_ideal_steer_vel.cpp | 92.30% <0.00%> (-3.85%) |
:arrow_down: |
... and 405 more |
Continue to review full report at Codecov.
Legend - Click here to learn more
Δ = absolute <relative> (impact)
,ø = not affected
,? = missing data
Powered by Codecov. Last update 05f6a6b...842b99a. Read the comment docs.
I've rebased this branch onto the new main since https://github.com/autowarefoundation/autoware.universe/pull/892 is now merged.
We should resolve this so the blocked issue can be resolved: https://github.com/autowarefoundation/autoware.universe/issues/628
This is a port of an Autoware.Auto package with only minor modifications to pass the linting of .universe. I don't expect there will be a problem to having it merged. If none of the assigned reviewers have time to review it, can someone else be assigned?
@xmfcx Since you seems to be busy with the bus integration, I will do take over this review task.
I have approved the PR.
The code was ported from Autoware.Auto, and I'm not worried about the functionality.
One remaining concern that I have is having artifacts as a part of tvm_utility. In the readme, it says place any test artifacts, but are we expecting more artifacts to be added in the future as more models are added? If yolo_v2_tiny is good enough for test purpose, then we might want to clarify that in the README. This is just a suggestion and I'm okay with merging with current code.
We definitely don't want to have one test-case/artifact per supported model. And yolo_v2_tiny is good enough for now. But we may want to add some in the future, and the test procedure has been written to support multiple test cases. So I would suggest keeping it as it is.