Skip to content

[Bug Report] Isaac Lab RSL RL framework doesn't work, others do. #1824

Description

@MuammerBay

Describe the bug

When running any training scripts on the RSL RL Framework, such as the one from the installation guide
python scripts/reinforcement_learning/rsl_rl/train.py --task=Isaac-Ant-v0 --headless
it errors ValueError: Reference at 'refs/heads/master' does not exist (see complete errors below).
Other libraries such as RL_Games, SKRL and SB3 work just fine.

Steps to reproduce

Following the pip installation on https://isaac-sim.github.io/IsaacLab/main/source/setup/installation/pip_installation.html

  • conda env with python 3.10 (always activated)
  • pip install torch==2.5.1 for CUDA 11.8 with GPU Driver: 537.58 on RTX 3090
  • latest pip version
  • pip install isaac sim 4.5
  • git clone repo (no fork)
  • isaac.bat --install
  • python scripts\tutorials\00_sim\create_empty.py
(isaaclab_2) C:\IsaacLab2\IsaacLab>python scripts/reinforcement_learning/rsl_rl/train.py --task=Isaac-Ant-v0 --headless
[INFO][AppLauncher]: Loading experience file: C:\IsaacLab2\IsaacLab\apps\isaaclab.python.headless.kit
[Warning] [simulation_app.simulation_app] Modules: ['omni.kit_app'] were loaded before SimulationApp was started and might not be loaded correctly.
[Warning] [simulation_app.simulation_app] Please check to make sure no extra omniverse or pxr modules are imported before the call to SimulationApp(...)
Loading user config located at: 'c:/users/murme/miniconda3/envs/isaaclab_2/lib/site-packages/omni/data/Kit/Isaac-Sim/4.5/user.config.json'
[Info] [carb] Logging to file: c:/users/murme/miniconda3/envs/isaaclab_2/lib/site-packages/omni/logs/Kit/Isaac-Sim/4.5/kit_20250209_155810.log
2025-02-09 14:58:10 [0ms] [Warning] [omni.kit.app.plugin] No crash reporter present, dumps uploading isn't available.
2025-02-09 14:58:10 [405ms] [Warning] [omni.usd_config.extension] Enable omni.materialx.libs extension to use MaterialX
2025-02-09 14:58:11 [985ms] [Warning] [omni.isaac.dynamic_control] omni.isaac.dynamic_control is deprecated as of Isaac Sim 4.5. No action 
is needed from end-users.

|---------------------------------------------------------------------------------------------|
| Driver Version: 537.58        | Graphics API: D3D12
|=============================================================================================|
| GPU | Name                             | Active | LDA | GPU Memory | Vendor-ID | LUID       |
|     |                                  |        |     |            | Device-ID | UUID       |
|     |                                  |        |     |            | Bus-ID    |            |
|---------------------------------------------------------------------------------------------|
| 0   | NVIDIA GeForce RTX 3090          | Yes: 0 |     | 24340   MB | 10de      | 0bde0000.. |
|     |                                  |        |     |            | 2204      | 0          |
|     |                                  |        |     |            | 2b        |            |
|=============================================================================================|
| OS: Windows 10 Home, Version: 10.0 (22H2), Build: 19045, Kernel: 10.0.19041.5369
| Processor: AMD Ryzen 5 5600X 6-Core Processor
| Cores: 6 | Logical Cores: 12
|---------------------------------------------------------------------------------------------|
| Total Memory (MB): 32694 | Free Memory: 15726
| Total Page/Swap (MB): 50090 | Free Page/Swap: 15773
|---------------------------------------------------------------------------------------------|
[INFO]: Parsing configuration from: isaaclab_tasks.manager_based.classic.ant.ant_env_cfg:AntEnvCfg
[INFO]: Parsing configuration from: isaaclab_tasks.manager_based.classic.ant.agents.rsl_rl_ppo_cfg:AntPPORunnerCfg
[INFO] Logging experiment in directory: C:\IsaacLab2\IsaacLab\logs\rsl_rl\ant
Exact experiment name requested from command line: 2025-02-09_15-58-16
Setting seed: 42
[INFO]: Base environment:
        Environment device    : cuda:0
        Environment seed      : 42
        Physics step-size     : 0.008333333333333333
        Rendering step-size   : 0.016666666666666666
        Environment step-size : 0.016666666666666666
[INFO]: Time taken for scene creation : 2.988643 seconds
[INFO]: Scene manager:  <class InteractiveScene>
        Number of environments: 4096
        Environment spacing   : 5.0
        Source prim name      : /World/envs/env_0
        Global prim paths     : ['/World/ground']
        Replicate physics     : True
[INFO]: Starting the simulation. This may take a few seconds. Please wait...
[INFO]: Time taken for simulation start : 1.622173 seconds
[INFO] Command Manager:  <CommandManager> contains 0 active terms.
+------------------------+
|  Active Command Terms  |
+--------+-------+-------+
| Index  | Name  |  Type |
+--------+-------+-------+
+--------+-------+-------+

[INFO] Recorder Manager:  <RecorderManager> contains 0 active terms.
+---------------------+
| Active Recorder Terms |
+-----------+---------+
|   Index   | Name    |
+-----------+---------+
+-----------+---------+

[INFO] Action Manager:  <ActionManager> contains 1 active terms.
+----------------------------------+
|  Active Action Terms (shape: 8)  |
+-------+--------------+-----------+
| Index | Name         | Dimension |
+-------+--------------+-----------+
|   0   | joint_effort |         8 |
+-------+--------------+-----------+

[INFO] Observation Manager: <ObservationManager> contains 1 groups.
+-----------------------------------------------------------+
| Active Observation Terms in Group: 'policy' (shape: (60,)) |
+-----------+-----------------------------------+-----------+
|   Index   | Name                              |   Shape   |
+-----------+-----------------------------------+-----------+
|     0     | base_height                       |    (1,)   |
|     1     | base_lin_vel                      |    (3,)   |
|     2     | base_ang_vel                      |    (3,)   |
|     3     | base_yaw_roll                     |    (2,)   |
|     4     | base_angle_to_target              |    (1,)   |
|     5     | base_up_proj                      |    (1,)   |
|     6     | base_heading_proj                 |    (1,)   |
|     7     | joint_pos_norm                    |    (8,)   |
|     8     | joint_vel_rel                     |    (8,)   |
|     9     | feet_body_forces                  |   (24,)   |
|     10    | actions                           |    (8,)   |
+-----------+-----------------------------------+-----------+

[INFO] Event Manager:  <EventManager> contains 1 active terms.
+-------------------------------------+
| Active Event Terms in Mode: 'reset' |
+---------+---------------------------+
|  Index  | Name                      |
+---------+---------------------------+
|    0    | reset_base                |
|    1    | reset_robot_joints        |
+---------+---------------------------+

[INFO] Termination Manager:  <TerminationManager> contains 2 active terms.
+---------------------------------+
|     Active Termination Terms    |
+-------+--------------+----------+
| Index | Name         | Time Out |
+-------+--------------+----------+
|   0   | time_out     |   True   |
|   1   | torso_height |  False   |
+-------+--------------+----------+

[INFO] Reward Manager:  <RewardManager> contains 7 active terms.
+---------------------------------+
|       Active Reward Terms       |
+-------+----------------+--------+
| Index | Name           | Weight |
+-------+----------------+--------+
|   0   | progress       |    1.0 |
|   1   | alive          |    0.5 |
|   2   | upright        |    0.1 |
|   3   | move_to_target |    0.5 |
|   4   | action_l2      | -0.005 |
|   5   | energy         |  -0.05 |
|   6   | joint_limits   |   -0.1 |
+-------+----------------+--------+

[INFO] Curriculum Manager:  <CurriculumManager> contains 0 active terms.
+----------------------+
| Active Curriculum Terms |
+-----------+----------+
|   Index   | Name     |
+-----------+----------+
+-----------+----------+

[INFO]: Completed setting up the environment...
Actor MLP: Sequential(
  (0): Linear(in_features=60, out_features=400, bias=True)
  (1): ELU(alpha=1.0)
  (2): Linear(in_features=400, out_features=200, bias=True)
  (3): ELU(alpha=1.0)
  (4): Linear(in_features=200, out_features=100, bias=True)
  (5): ELU(alpha=1.0)
  (6): Linear(in_features=100, out_features=8, bias=True)
)
Critic MLP: Sequential(
  (0): Linear(in_features=60, out_features=400, bias=True)
  (1): ELU(alpha=1.0)
  (2): Linear(in_features=400, out_features=200, bias=True)
  (3): ELU(alpha=1.0)
  (4): Linear(in_features=200, out_features=100, bias=True)
  (5): ELU(alpha=1.0)
  (6): Linear(in_features=100, out_features=1, bias=True)
)
################################################################################
                       Learning iteration 0/1000 

                       Computation: 51070 steps/s (collection: 2.371s, learning 0.196s)
               Value function loss: 0.0397
                    Surrogate loss: -0.0088
             Mean action noise std: 0.99
                       Mean reward: -0.46
               Mean episode length: 19.68
           Episode_Reward/progress: 0.0016
              Episode_Reward/alive: 0.0086
            Episode_Reward/upright: 0.0016
     Episode_Reward/move_to_target: 0.0085
          Episode_Reward/action_l2: -0.0007
             Episode_Reward/energy: -0.0406
       Episode_Reward/joint_limits: -0.0025
      Episode_Termination/time_out: 3.8125
  Episode_Termination/torso_height: 0.0312
--------------------------------------------------------------------------------
                   Total timesteps: 131072
                    Iteration time: 2.57s
                        Total time: 2.57s
                               ETA: 2566.5s

Error executing job with overrides: []
Traceback (most recent call last):
  File "c:/isaaclab2/isaaclab/source/isaaclab_tasks/isaaclab_tasks/utils/hydra.py", line 101, in hydra_main
    func(env_cfg, agent_cfg, *args, **kwargs)
  File "C:\IsaacLab2\IsaacLab\scripts\reinforcement_learning\rsl_rl\train.py", line 145, in main
    runner.learn(num_learning_iterations=agent_cfg.max_iterations, init_at_random_ep_len=True)
  File "C:\Users\murme\miniconda3\envs\isaaclab_2\lib\site-packages\rsl_rl\runners\on_policy_runner.py", line 165, in learn
    git_file_paths = store_code_state(self.log_dir, self.git_status_repos)
  File "C:\Users\murme\miniconda3\envs\isaaclab_2\lib\site-packages\rsl_rl\utils\utils.py", line 76, in store_code_state
    t = repo.head.commit.tree
  File "C:\Users\murme\miniconda3\envs\isaaclab_2\lib\site-packages\git\refs\symbolic.py", line 297, in _get_commit
    obj = self._get_object()
  File "C:\Users\murme\miniconda3\envs\isaaclab_2\lib\site-packages\git\refs\symbolic.py", line 288, in _get_object
    return Object.new_from_sha(self.repo, hex_to_bin(self.dereference_recursive(self.repo, self.path)))
  File "C:\Users\murme\miniconda3\envs\isaaclab_2\lib\site-packages\git\refs\symbolic.py", line 168, in dereference_recursive
    hexsha, ref_path = cls._get_ref_info(repo, ref_path)
  File "C:\Users\murme\miniconda3\envs\isaaclab_2\lib\site-packages\git\refs\symbolic.py", line 278, in _get_ref_info
    return cls._get_ref_info_helper(repo, ref_path)
  File "C:\Users\murme\miniconda3\envs\isaaclab_2\lib\site-packages\git\refs\symbolic.py", line 257, in _get_ref_info_helper
    raise ValueError("Reference at %r does not exist" % ref_path)
ValueError: Reference at 'refs/heads/master' does not exist

Set the environment variable HYDRA_FULL_ERROR=1 for a complete stack trace.
2025-02-09 14:58:27 [16,888ms] [Warning] [omni.fabric.plugin] gFabricState->gUsdStageToSimStageWithHistoryMap had 1 outstanding SimStageWithHistory(s) at shutdown
2025-02-09 14:58:27 [17,062ms] [Warning] [carb] Recursive unloadAllPlugins() detected!

System Info

Describe the characteristic of your environment:

  • Commit: c4bec8f
  • Isaac Sim Version: 4.5
  • OS: Windows 10 Home
  • GPU: RTX 3090
  • CUDA: 11.8
  • GPU Driver: 537.58

Additional context

I have tried other libraries and bunch of examples, and only all the training scripts containing RSL RL don't work. This was also the case before the update on the 08.02.25 - "Switches to RSL-RL install from PyPI".
I have also tried to switch to master branch as the error indicated "ValueError: Reference at 'refs/heads/master' does not exist" and the same happened on CUDA 12.8 with driver version 552.86 before downgrading.

Checklist

  • I have checked that there is no similar issue in the repo (required)
  • I have checked that the issue is not in running Isaac Sim itself and is related to the repo

Acceptance Criteria

Add the criteria for which this task is considered done. If not known at issue creation time, you can add this once the issue is assigned.

  • Making RSL RL Framework work

Metadata

Metadata

Assignees

No one assigned

    Labels

    bugSomething isn't workingduplicateThis issue or pull request already exists

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions