real-stanford/diffusion_policy issues and pull requests

#139 - How to implement pick&place through a real robot

Issue - State: open - Opened by yechan519 10 days ago

#138 - About pose_trajectory_interpolation & Rotation representation

Issue - State: open - Opened by YanJiaHuan 24 days ago

#137 - Is it necessary to get another abs dataset or can I use my image. hdf5 as well?

Issue - State: open - Opened by carloskiyoto 27 days ago

#136 - rotation transformation error

Issue - State: closed - Opened by johnnevins about 1 month ago

#135 - Improve the definition of the saved simulation result video

Issue - State: open - Opened by yihang-xdu about 1 month ago

#134 - Error with env.reset() in robomimic enviroment (in google colab vision-based)

Issue - State: open - Opened by carloskiyoto about 1 month ago - 2 comments

#133 - What is the difference between the diffusion_unet_hybrid_image_policy and diffusion_unet_image_policy ?

Issue - State: open - Opened by KIM-HYEDO about 2 months ago - 1 comment

#132 - AssertionError in Fixed Workers in Robomimic

Issue - State: closed - Opened by mr20010112 about 2 months ago

#130 - Robomimic Version Incompatability

Issue - State: open - Opened by Benjamin-So 2 months ago

#129 - problem when reproducing square task

Issue - State: open - Opened by lucywang720 3 months ago

#128 - Why max_train_episodes = 90 in pusht

Issue - State: open - Opened by c0g 3 months ago

#127 - About velocity control and position control

Issue - State: open - Opened by 2292685327 3 months ago

#126 - reasoning resources to real robots

Issue - State: open - Opened by robot1lyj 3 months ago

#125 - Debug

Pull Request - State: closed - Opened by Hiromasa56 3 months ago - 1 comment

#125 - Debug

Pull Request - State: closed - Opened by Hiromasa56 3 months ago - 1 comment

#124 - why when evaluating a pretrained checkpoint, the test/mean_score in eval_log.json is not equal to the test_mean_score in the checkpoint

Issue - State: open - Opened by zack5582 3 months ago - 2 comments

#124 - why when evaluating a pretrained checkpoint, the test/mean_score in eval_log.json is not equal to the test_mean_score in the checkpoint

Issue - State: open - Opened by zack5582 3 months ago

#123 - Adding support for pre-trained dinov2 base and large models

Pull Request - State: closed - Opened by giricme 3 months ago

#123 - Adding support for pre-trained dinov2 base and large models

Pull Request - State: closed - Opened by giricme 3 months ago

#122 - Adding support for pre-trained dinov2 base and large models

Pull Request - State: closed - Opened by giricme 3 months ago - 1 comment

#122 - Adding support for pre-trained dinov2 base and large models

Pull Request - State: closed - Opened by giricme 3 months ago - 1 comment

#121 - Can I use it in Window System?

Issue - State: open - Opened by ZisongXu 3 months ago - 2 comments

#120 - Pre-trained Checkpoints for TrainRobomimicLowdimWorkspace

Issue - State: closed - Opened by mr20010112 3 months ago

#119 - import error

Issue - State: open - Opened by fengluying 3 months ago - 2 comments

#118 - RobomimicLowdimRunner: ConnectionResetError [Errno 104]

Issue - State: closed - Opened by Luca-Wiehe 4 months ago - 1 comment

#117 - Different kitchen datasets for different methods

Issue - State: open - Opened by ziyadsheeba 4 months ago

#116 - ConnectionResetError(104) in KitchenLowdimRunner during environment initialization

Issue - State: closed - Opened by mr20010112 4 months ago - 5 comments

#115 - The 7 action dimension is for relative position between the two positions that is how the robomimic dataset is collected. Whereas the action dimension is the absolute position of the robot, they use it because they find that DP performs better with abs position (or absolute configuration of the robot) compared to relative position.

Issue - State: closed - Opened by wudier483 4 months ago

#114 - Sensor fusion

Pull Request - State: open - Opened by KomK2 4 months ago

#113 - Question related to how it has been trained the architecture. Trained from video of human doing a task with hands or moving a robot by jogging?

Issue - State: open - Opened by kappa95 5 months ago

#112 - Normalising the images twice

Issue - State: open - Opened by KomK2 5 months ago

#111 - What are p1 p2 p3 p4 of kitchen? How can I replicate it?

Issue - State: open - Opened by bluerand4 5 months ago

#110 - how to train one policy for multi tasks?

Issue - State: open - Opened by huangxu1991 5 months ago - 1 comment

#109 - What is the meaning of the content of eval_log.json?

Issue - State: open - Opened by Shi-jiayu 6 months ago - 2 comments

#108 - arx5 collects data

Issue - State: closed - Opened by fighting-j 6 months ago

#107 - arx5 collects data

Issue - State: open - Opened by fighting-j 6 months ago

#106 - What is the difference between the robomimic_image and robomimic_lowdim?

Issue - State: open - Opened by Shi-jiayu 6 months ago - 2 comments

#105 - module 'jax.random' has no attribute 'KeyArray'

Issue - State: open - Opened by monk279 6 months ago - 1 comment

#104 - Prediction Horizon in IBC implementation.

Issue - State: open - Opened by migooll 6 months ago

#103 - Some Problems About GroupNorm

Issue - State: open - Opened by Selen-Suyue 6 months ago - 2 comments

#102 - ImportError: ImportError: cannot import name 'IGNORE_GIT_FOLDER_PATTERNS' from 'huggingface_hub.utils' ....

Issue - State: open - Opened by CodingCatMountain 6 months ago - 3 comments

#101 - Available Checkpoints and Paper Reults

Issue - State: open - Opened by andrearosasco 7 months ago

#100 - .

Issue - State: closed - Opened by MaruGreen 7 months ago

#99 - Finetuning diffusion policy

Issue - State: closed - Opened by RuiyuWANG 7 months ago

#98 - My picking task is not picking with image or hybrid mode.

Issue - State: open - Opened by RochMollero 7 months ago

#97 - Bump pytorch version

Pull Request - State: closed - Opened by knicked 7 months ago

#96 - full_key: task.dataset

Issue - State: open - Opened by coderpracticer 7 months ago - 8 comments

#95 - pip package version inconsistency in 2024 july

Issue - State: open - Opened by kangnil 7 months ago

#94 - Clarification on Global Condition vs. Local Condition in Diffusion Policy

Issue - State: open - Opened by Gear420 8 months ago - 2 comments

#93 - Why action dimension of DP is 10 but action dimension of lstm-gmm is 7 in tool_hang task? What's the difference?

Issue - State: open - Opened by Vhyrsd 8 months ago - 3 comments

#92 - DOES DP REALLY WORK ON REAL TASK?

Issue - State: closed - Opened by huangxu1991 8 months ago - 4 comments

#91 - chat group

Issue - State: open - Opened by XPLearner 8 months ago

#90 - Is it possible to release the code for the mid-level controller used in Franka?

Issue - State: open - Opened by Kimagure7 8 months ago - 1 comment

#89 - dataset format

Issue - State: open - Opened by deavn2236 8 months ago

#88 - Why does that model need the environment?

Issue - State: open - Opened by liops-seungyeob 8 months ago

#87 - Unable to Access Project Webpage and Training Data

Issue - State: closed - Opened by renyu2016 9 months ago - 2 comments

#86 - Maybe ther is a bug in the code of diffusion_policy.model.diffusion.conditional_unet1d.py line 143:

Issue - State: closed - Opened by Davi-carto 9 months ago

#85 - feat: inital commit of training working for LeTO

Pull Request - State: closed - Opened by darshanhegde 9 months ago

#84 - why when evaluating a pretrained checkpoint, the test/mean_score in eval_log.json is not equal to the test_mean_score in the checkpoint

Issue - State: open - Opened by MaximaMaSun 9 months ago - 2 comments

#83 - how to use multiGPU

Issue - State: open - Opened by deavn2236 9 months ago - 3 comments

#82 - Question on pusht real data?

Issue - State: open - Opened by johnnevins 9 months ago - 1 comment

#81 - Why we must to use MultiStepWrapper?

Issue - State: closed - Opened by CAI23sbP 9 months ago - 2 comments

#80 - Training loss is relatively large

Issue - State: closed - Opened by COST-97 9 months ago

#79 - code for teleoperation using Quest?

Issue - State: open - Opened by johnnevins 9 months ago - 2 comments

#78 - how to visualize the trajectory

Issue - State: open - Opened by seagul1 9 months ago - 2 comments

#77 - Can you help explain what are the two dimension of parameter 'robot0_gripper_qpos' stand for?

Issue - State: open - Opened by YanJiaHuan 10 months ago

#76 - Segmentation fault probably associated with OpenGL or multithreading?

Issue - State: open - Opened by sys-shutdown 10 months ago - 1 comment

#75 - Metadata info about the dataset?

Issue - State: closed - Opened by johnnevins 10 months ago

#74 - About collecting data in real robot scene

Issue - State: open - Opened by Yet-ian 10 months ago

#73 - Failure to eval with other checkpoints

Issue - State: open - Opened by yanghu819 10 months ago - 2 comments

#72 - about eval results of pushT, is this normal?

Issue - State: open - Opened by yanghu819 10 months ago - 1 comment

#71 - Error running robosuite Lift-ph with abs_action=false

Issue - State: open - Opened by NagisaZj 10 months ago - 1 comment

#70 - Relative or Absolute Actions

Issue - State: closed - Opened by andrearosasco 10 months ago

#69 - conda failed to create environment

Issue - State: open - Opened by duidui6666 10 months ago - 1 comment

#68 - create environment error

Issue - State: open - Opened by Chenrf1121 10 months ago - 1 comment

#67 - AttributeError: module 'jax.random' has no attribute 'KeyArray'

Issue - State: open - Opened by BE-A-HERO 10 months ago - 1 comment

#66 - Fixed bug that prevented to use center crop

Pull Request - State: open - Opened by chisarie 10 months ago

#66 - Fixed bug that prevented to use center crop

Pull Request - State: open - Opened by chisarie 10 months ago

#65 - wrong rotation of the predicted actions

Issue - State: open - Opened by yl-wang996 10 months ago - 9 comments

#64 - Any advice on perturbation in simulator?

Issue - State: open - Opened by johnnevins 10 months ago

#63 - Is EMA used to update the training weights?

Issue - State: open - Opened by StarCycle 10 months ago

#62 - Running eval.py "main workspace = cls(cfg, output_dir=output_dir) TypeError: init() got an unexpected keyword argument 'output_dir'?

Issue - State: closed - Opened by johnnevins 10 months ago - 2 comments

#61 - Possible Issue with Data Organization in Buffer Addition of kitchen_lowdim_dataset.py

Issue - State: open - Opened by hoost14 10 months ago

#60 - How to access to the depth data in the simulation?

Issue - State: closed - Opened by yl-wang996 10 months ago

#59 - Visualization of the trajectories

Issue - State: closed - Opened by yl-wang996 11 months ago

#58 - Goal conditioning via FiLM?

Issue - State: open - Opened by robertmash2 11 months ago - 1 comment

#57 - How much computing power is required at least?

Issue - State: open - Opened by mayeRobot 11 months ago

#56 - fixed val_ratio dose not function as expected in dataset

Pull Request - State: open - Opened by LYouC 11 months ago

#55 - about the keypoint?

Issue - State: open - Opened by songhat 11 months ago

#54 - Transformer based vision colab example

Issue - State: open - Opened by StephenYangjz 11 months ago - 3 comments

#53 - Fix variable assignment typo in multi_image_obs_encoder.py

Pull Request - State: open - Opened by devinluo27 12 months ago

#52 - shapely.errors.TopologicalError: The operation 'GEOSIntersection_r' could not be performed.

Issue - State: closed - Opened by MaximaMaSun 12 months ago - 3 comments

#51 - train config

Issue - State: open - Opened by Nimingez 12 months ago

#50 - real robot

Issue - State: closed - Opened by Nimingez 12 months ago

#49 - how to prepare a custom skill dataset in the simulator?

Issue - State: open - Opened by zacario-li 12 months ago

#48 - config yaml for the pusht_real experiment?

Issue - State: closed - Opened by MohammedAlJaff about 1 year ago - 1 comment

#47 - Questions on Velocity Control Implementation

Issue - State: open - Opened by chadwick-yao about 1 year ago

#46 - Mujoco eq_active attribute not found

Issue - State: open - Opened by BarisYazici about 1 year ago - 1 comment

#45 - Incorrect Image Normalization

Issue - State: open - Opened by Yingdong-Hu about 1 year ago - 1 comment

#44 - [Question] Why a_{t-1}, a_{t-2}, ... also contribute to diffusion loss?

Issue - State: open - Opened by ManUtdMoon about 1 year ago

GitHub / real-stanford/diffusion_policy issues and pull requests