Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Skills benchmarking without pose estimation #184

Open
wants to merge 19 commits into
base: skills_benchmarking
Choose a base branch
from

Conversation

Achuthankrishna
Copy link
Contributor

Updated local_only.sh to not run pose estimation and semantic segmentation modules while running mobile gaze and static pick. This prevents from causing any cuDNN error in its dependencies.

tusharsangam and others added 19 commits June 5, 2024 11:58
* load the newly trained checkpoints

port semantic pick

tidy

appraching

update checkpoint and make the MAX_JOINT_MOVEMENT_SEMANTIC_GAZE large and it works well

smooth arm actions

pose estimation

pose correction with discretization

Orientation Correction Working

pose correction finetuning

adding more grasp orientation & object correction

pose estimation with object facing camera

creating new pick api

force control for object grasping

PR ready

to pr

PR ready changes

Starting pose estimation service from spot-sim2real

PR hot fixes

submodule init instruction update

add todo and fix pre-commit issue

PR review 1

PR fixes 2

fix precommit

fix precommit v2

* add trest srcipt for nn basew

* PR review final

---------

Co-authored-by: Jimmy Yang <yangtsungyen@gmail.com>
* load the newly trained checkpoints

* port semantic pick

* tidy

* appraching

* update checkpoint and make the MAX_JOINT_MOVEMENT_SEMANTIC_GAZE large and it works well

* smooth arm actions

* Chris Cox Demo test2

lsc_demo.py fix

keep_alive change

parameter update for metaailocal api

Reducing time for Meta AI Local & occassional format crashing issues

loading whisper on cuda:1

fixing precommit

* uncommenting pose estimation

* restoring keyboard teleop

* deleting test_semantic_gaze.py

* polish

---------

Co-authored-by: Jimmy Yang <yangtsungyen@gmail.com>
* load the newly trained checkpoints

* port semantic pick

* tidy

* appraching

* update checkpoint and make the MAX_JOINT_MOVEMENT_SEMANTIC_GAZE large and it works well

* smooth arm actions

* first commit

* force control first commit

* parameter adjustment

* Generalization to grasp

* some changes remaining

* arm command follow

* pose estimation fix for grasping

* video shoot

* body follow fix

* some changes

* caliberation fix

* Resolving calibartaion error with minimal changes

* gitignore

* backup

* benchmark files

* ros_communications (#173)

reducing file size

Bug fixes n support for Float32MultiArray

bug fixes after PPG with Jimmy

fixing pre-commit

precommit fix 2

* gripper_T_intel

* adding new objects in config

* patch fix for ros tcp

* gitignore

* cleanup

* removing unnecessary

* ros tcp fix

* ros test update

* pose correction + BD grasp, small patch

* clean up

---------

Co-authored-by: Jimmy Yang <yangtsungyen@gmail.com>
Co-authored-by: Jimmy Yang <jimmytyyang@meta.com>
* load the newly trained checkpoints

* port semantic pick

* tidy

* appraching

* update checkpoint and make the MAX_JOINT_MOVEMENT_SEMANTIC_GAZE large and it works well

* smooth arm actions

* grasp mode loading from config

* reverting test_pick_with_pose.py

* polish

---------

Co-authored-by: Jimmy Yang <yangtsungyen@gmail.com>
Co-authored-by: Jimmy Yang <jimmytyyang@meta.com>
* File add for contrained place point estimation

* nyc and fre placement

* nyc and fre placement

* tidy

* add support of proposition

* generate the image services automatically

* move rospy

* tidy

* tidy

---------

Co-authored-by: Jimmy Yang <yangtsungyen@gmail.com>

Integrate meta ai, add llama3 local support, read reset_home params at undock, add safety-net to language parsing (#163)

pose estimation (#166)

* load the newly trained checkpoints

port semantic pick

tidy

appraching

update checkpoint and make the MAX_JOINT_MOVEMENT_SEMANTIC_GAZE large and it works well

smooth arm actions

pose estimation

pose correction with discretization

Orientation Correction Working

pose correction finetuning

adding more grasp orientation & object correction

pose estimation with object facing camera

creating new pick api

force control for object grasping

PR ready

to pr

PR ready changes

Starting pose estimation service from spot-sim2real

PR hot fixes

submodule init instruction update

add todo and fix pre-commit issue

PR review 1

PR fixes 2

fix precommit

fix precommit v2

* add trest srcipt for nn basew

* PR review final

---------

Co-authored-by: Jimmy Yang <yangtsungyen@gmail.com>

first commit

force control first commit

parameter adjustment

Generalization to grasp

some changes remaining

arm command follow

pose estimation fix for grasping

video shoot

body follow fix

some changes

caliberation fix

Resolving calibartaion error with minimal changes

gitignore

saving hanges

running episode tests

joannes apt changes

joanna changes

remaining changes

running pose estimation correction with BD grasp

pose correction + BD grasp

force control after BD grasp

rebasing error fix

rebase fix

pose estimation all testing

removing semantic pick testing artifacts
@facebook-github-bot facebook-github-bot added the CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. label Aug 13, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
Projects
None yet
Development

Successfully merging this pull request may close these issues.

4 participants