18bba97cd6
change default to 256 latent video
2026-01-25 19:48:54 +01:00
9c14524470
add video backbone to pi05
2026-01-25 19:38:39 +01:00
5ab3dfd762
add videoprism example
2026-01-25 15:51:50 +01:00
bbe9407ead
add videoprism
2026-01-25 15:51:21 +01:00
6d34a986de
feat(ci): trigger manually documentation release version ( #2841 )
2026-01-22 12:26:17 +01:00
961277d86e
chore(dependencies): Bump lerobot to 0.4.4 ( #2840 )
2026-01-22 12:24:12 +01:00
0b067df57d
feat(robots): add context managers ( #2828 )
2026-01-20 18:02:38 +01:00
9ca680dce2
Update README.md ( #2827 )
...
Add Chinese doc link.
Signed-off-by: Tommy in Tongji <36354458+TommyZihao@users.noreply.github.com >
2026-01-20 17:54:24 +01:00
9919b16b36
fix: ensure action tensors are moved to client_device in async training ( #2792 )
...
* feat(async_inference): server always sends CPU tensors, client handles device conversion
* fix:fix the type annotation of RawObservation in src/lerobot/async_inference/helpers.py
* update the import of robot_client
---------
Co-authored-by: Sato shinji <wwwsatoshinji@gmail.com >
Co-authored-by: Steven Palma <imstevenpmwork@ieee.org >
Co-authored-by: KB <kevin-brian.n-diaye@epita.fr >
2026-01-20 15:17:38 +01:00
d36dfcdf71
fix(discord link): fixing discord link in CONTRIBUTING.md ( #2826 )
...
Signed-off-by: Caroline Pascal <caroline8.pascal@gmail.com >
2026-01-20 15:00:45 +01:00
13bfee1aa4
Set 10 direction bit for Current Load attribute ( #1014 )
2026-01-20 11:20:30 +01:00
79688a09f2
improve(dataset-tools): image2video editing tools : Multiple episodes per video file ( #2811 )
...
* improve image2video
* add episodes video encoding
* fix mypy failing
* iterate on review
* nit
* remove max, and let it be optional
* iterate more
* update docs
* fix test
---------
Co-authored-by: Michel Aractingi <michel.aractingi@huggingface.co >
2026-01-20 11:04:22 +01:00
b2ff219624
Fixes aggregation of image datasets ( #2717 )
...
* fix: use features when aggregating image based datasets
* add: test asserting for data type
* add: features param to writing dataset
---------
Co-authored-by: Steven Palma <imstevenpmwork@ieee.org >
2026-01-19 23:36:41 +01:00
66929c5935
feat: add async server-client streaming support for Groot policy ( #2812 )
2026-01-19 22:13:48 +01:00
5286ef8439
feat(utils): extend import check util ( #2820 )
...
* refactor(utils): is_package_available now differentiate between pkg name and module name
* refactor(tests): update require_package decorator
2026-01-19 16:43:11 +01:00
fe068df711
fix(train): eval env initialization on train script ( #2818 )
...
* fix: eval env initialization on train script
Signed-off-by: bigmbigk <bigmbigk@gmail.com >
* fix: eval env creation condition
---------
Signed-off-by: bigmbigk <bigmbigk@gmail.com >
2026-01-19 14:14:10 +01:00
da41646073
fix libero reset logic for correct resetting ( #2817 )
2026-01-19 13:18:52 +01:00
46e19ae579
feat: is connect checks decorators ( #2813 )
2026-01-16 18:52:06 +01:00
77dc49b3a3
Fix delta timestamps with episodes filter and add tests ( #2612 )
2026-01-16 18:14:54 +01:00
33910673ec
Bugfix: Add tests for image deletion and fix mixed image-video deletion ( #2592 )
...
* Add tests for image deletion and fix mixed-image-video deletion
* Fix docstring whitespace
* Remove debug print
Signed-off-by: Alex Tyshka <atyshka15@gmail.com >
* Remove inaccurate comment
* Remove batched video test
---------
Signed-off-by: Alex Tyshka <atyshka15@gmail.com >
2026-01-16 18:14:15 +01:00
19dce78457
Refactor: Move PEFT config from training script to policy level ( #2806 )
...
* move peft config from `lerobot_train` to policy level
* Update src/lerobot/scripts/lerobot_train.py
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com >
Signed-off-by: Michel Aractingi <michel.aractingi@huggingface.co >
* copilot response
* Change the polciy function to return targets rather than peft config.`_get_default_peft_targets()` override in PI0, PI0.5, SmolVLA
* remove none check when building config dict
---------
Signed-off-by: Michel Aractingi <michel.aractingi@huggingface.co >
2026-01-16 17:14:28 +01:00
112b2d173a
chore(ci): deactivates cron job on unbound dep tests ( #2810 )
2026-01-16 14:39:00 +01:00
b825880c40
chore: add security policy ( #2809 )
...
* chore: add security policy
* pre-commit style
2026-01-16 14:38:42 +01:00
76d6b71b0a
Correct Frodobots Earth Rover SDK link and add setup instructions ( #2671 )
...
* Fix SDK link and enhance setup instructions
Updated the Frodobots SDK link and added credential setup instructions.
Signed-off-by: ./c² <cagataycali@icloud.com >
* Update docs/source/earthrover_mini_plus.mdx
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com >
Signed-off-by: ./c² <cagataycali@icloud.com >
* Update docs/source/earthrover_mini_plus.mdx
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com >
Signed-off-by: Steven Palma <imstevenpmwork@ieee.org >
---------
Signed-off-by: ./c² <cagataycali@icloud.com >
Signed-off-by: Steven Palma <imstevenpmwork@ieee.org >
Co-authored-by: Steven Palma <imstevenpmwork@ieee.org >
2026-01-16 02:39:58 +01:00
5de38813d9
Add small context to the envHub doc page ( #2807 )
...
* Add small context to the envHub doc page
* Add the cfg: EnvConfig on the main function explaination.
2026-01-15 18:31:17 +01:00
6797ce615e
chore(deps): bump wandb & protobuf ( #2800 )
2026-01-15 10:51:42 +01:00
a17df523e0
chore(ci): merge annoying section in PR template ( #2802 )
...
* chore(ci): merge annoying section in PR template
* pre-commit
2026-01-14 17:17:56 +01:00
1c61b43b15
fix(teleop): add is_connected check to get_action ( #2801 )
2026-01-14 17:14:12 +01:00
15724826dd
chore: use alias & constants ( #2785 )
...
* chore: use alias and constants
* fix(rl): solve circular dependecy
* chore: nit right constant
* chore: pre-commit
* chore(script): conflict tokenizer train
---------
Signed-off-by: Steven Palma <imstevenpmwork@ieee.org >
2026-01-13 09:49:46 +01:00
2cdd9f43f7
fix: train tokenizer CLI entry point ( #2784 )
2026-01-13 01:42:53 +01:00
d0f57f58d1
Move cfg.validate() earlier to fix NoneType error with --policy.path ( #2782 )
2026-01-12 19:24:19 +01:00
b8ec1152d4
fix(robots): add reachy2 fixes ( #2783 )
...
* fix(robots): add reachy2 fixes
* tests(robots): remove reachy sdk stub
2026-01-12 18:05:16 +01:00
6b8d4c75a6
Feat/g1 improvements record sim ( #2765 )
...
This PR extends the integration of Unitree g1 with the LeRobot codebase. By converting robot state to a flat dict we can now record and replay episodes (example groot/holosoma scripts need to be adjusted as well). We also improve the simulation integration by calling .step @ _subscribe_motor_state instead of it running in a separate thread. We also add ZMQ camera to lerobot, streaming base64 images over json
2026-01-12 17:31:39 +01:00
d791a431fe
feat(robots): consolidates bi SO setups ( #2780 )
...
* feat(robots): consolidates bi SO setups
* fix(robots): solve circular dependecy
* fix(robots): teleop & record working
* feat(robots): only one SO
* fix(utils): rename bi so
* fix(scripts): bi so import
* fix(rl): remove imports
2026-01-12 16:01:22 +01:00
473f1bd0e0
docs: improve assets ( #2777 )
...
* add assets
* add libero results pifast:
* update
* update
* update size
* update naems:
:
* update training tokenizer
2026-01-12 13:33:28 +01:00
91ff9c4975
Fix: Respect policy.device=cpu config in training ( #2778 )
...
* fix cpu training in lerobot_train
* Update src/lerobot/scripts/lerobot_train.py
Signed-off-by: Michel Aractingi <michel.aractingi@huggingface.co >
2026-01-12 12:19:02 +01:00
1d86c9b7f2
feat(policies): add autoregressive VLAs with tokenization PiFast ( #2734 )
2026-01-09 23:08:37 +01:00
ba3d2148a3
skip peft cmd test in cli ( #2776 )
...
* skip peft cmd test in cli
* pre commit
* update desc
2026-01-09 19:10:02 +01:00
8b6fc0ae05
feat(datasets): expose video codec option for dataset recording ( #2771 )
...
* expose codec options + add tests
* pre-commit run -a
2026-01-08 18:06:39 +01:00
242b65d2df
chore(docs): update code block syntax to specify python for clarity ( #2770 )
2026-01-08 14:45:07 +01:00
ccfd609ece
feat(robots): consolidate SO arms implementation ( #2763 )
...
* feat(robots): consolidate SO arms implementation
* chore(robots): delete unnecessary init modules
2026-01-08 13:04:30 +01:00
fbe4c8b94f
Feat/remote rerunviz encoded images ( #2767 )
...
* feat(visualization): allow remote viewer + compress rerun images
* fix(tests): allow named argument in mocked rerun
* feat(visualization): ip instead or url & cli arg for compressing images
---------
Co-authored-by: J4nn1K <jannik@grothusen.de >
2026-01-07 17:38:13 +01:00
4f7cd8d369
Revert "feat(visualization): allow remote viewer + compress rerun images ( #2756 )" ( #2766 )
...
This reverts commit f844c7a458
2026-01-07 17:33:36 +01:00
f844c7a458
feat(visualization): allow remote viewer + compress rerun images ( #2756 )
...
* feat(visualization): allow remote viewer + compress rerun images
* fix(tests): allow named argument in mocked rerun
* feat(visualization): ip instead or url & cli arg for compressing images
2026-01-07 17:30:45 +01:00
7e9d05a799
add holosoma locomotion ( #2669 )
...
Add holosoma locomotion from Amazon-FAR
Add reset method to unitree_g1
Format actions as dict
Update docs
2026-01-07 16:05:31 +01:00
ecd8cd23d2
chore(dependencies): bound new dependecies ( #2759 )
2026-01-07 11:04:21 +01:00
a9d81e7f67
refactor(ci): Docker Hub image env ( #2755 )
...
* Refactor Docker Hub image env
Updated environment variable usage for Docker Hub credentials and corrected image tag extraction.
Signed-off-by: Pauline Bailly-Masson <155966238+paulinebm@users.noreply.github.com >
* same
Signed-off-by: Pauline Bailly-Masson <155966238+paulinebm@users.noreply.github.com >
* Apply suggestions from code review
Signed-off-by: Steven Palma <imstevenpmwork@ieee.org >
* chore(ci): remove duplicated IMAGE_FULL variable definition
---------
Signed-off-by: Pauline Bailly-Masson <155966238+paulinebm@users.noreply.github.com >
Signed-off-by: Steven Palma <imstevenpmwork@ieee.org >
Co-authored-by: Steven Palma <imstevenpmwork@ieee.org >
2026-01-07 00:21:03 +01:00
e2957d7783
fix: precise_sleep is never called with negative value ( #2757 )
2026-01-06 20:09:43 +01:00
963a3482fa
typo LW ( #2758 )
2026-01-06 18:17:29 +01:00
603d44434f
fix a bug for kwargs in wallx ( #2714 )
...
* support wallx
* fix bugs in flow
* incorporate wallx model into lerobot
* update the policy methods
* reduce to least config and params & pass lerobot basic test
* fixed dtype bugs
* add wallx dependencies
* update
* remove flash-attn requirement && fix bug in inference and fast mode
* fix bug for inference
* add some small modifications
* fix pre-commit errors
* remove lerobot[wallx]
* fix ci
* fix precommit issues
* fix: exclude wallx extra properly in CI workflows
* fix: add uv conflicts for wallx transformers version
* fix: peft test import
* pre-commit
* only export WallXConfig from wall_x package to avoid peft import in CI
* remove torch dep
* precommit
* add import
* update doc files
* fix minor errors
* fix a bug for kwargs
* fix precommit issue
---------
Signed-off-by: Pepijn <138571049+pkooij@users.noreply.github.com >
Co-authored-by: vincentchen <chenlufang@x2robot.com >
Co-authored-by: Geoffrey19 <sympathischmann35@gmail.com >
Co-authored-by: Pepijn <138571049+pkooij@users.noreply.github.com >
Co-authored-by: Pepijn <pepijn@huggingface.co >
Co-authored-by: geoffrey <geoffrey@x2robot.com >
2026-01-06 15:13:35 +01:00
Jade Choghari