Skip to content

Conversation

@rojas70
Copy link

@rojas70 rojas70 commented Jun 11, 2025

Commit 422d207 fixes a bug in the way the Mujoco Renderer is instantiated so that there is no crush when calling. test_gym_env_human.py.

Commit 7e23bbd includes a critic_actor_ratio that is missing when agent.high_utd_ratio is called.

Others...

rojas70 and others added 28 commits August 21, 2025 22:32
…s well using: a,d,w,s,q,e to change azimuth, elevation, and distance.

- code checks first for cam if so, then checks for action. if no cam, then just checks for action.
- correctly updates OXEEnvLogger generated files to be read by TFDS tfds.builder_from_directory in launcher.py
- When running outside debugger still get an exception that I need to work out, but can be used to move forward.
- Changing name of env to sparse reach version.
- In demos branch, fixing jax.tree.leaves.
- Reward is 1 for success cases based on the error norm proximity.
- Slight change to step function.
- Increased max-steps to 200 for this environment.
… using a format similar to the following:

└── franka_reach_drq_demo_script
    ├── session_20250914_210112_num_demos_2
    │   └── PandaReachSparseCube-v0
    │       └── 0.1.0
    │           ├── dataset_info.json
    │           ├── features.json
    │           └── PandaReachSparseCube-v0-train.tfrecord-00000

Succeeded in recording 2,5,10 demos for Panda Reach Sparse with Cube.
Copy link
Author

@rojas70 rojas70 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Ultimately what name are we using for the environment. I thought we had agreed on a 'Vision' named environment.

Currently I see: PandaReachSparseCube

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants