Adds loggers and git state logging for RSL-RL workflow (#42)

# Description

Previously, we did not add the Orbit repo to the RSL-RL git repositories
list. This resulted in the git state of the runs not being logged when
using RSL-RL. This MR allows RSL-RL to log the git state of the orbit
repository.

If the argument `--run_name` is passed with the command line interface,
it overrides the default `run_name` in the agent cfg.

Additionally, the MR adds support for different logging types. These
changes were suggested in #36.

If the argument `--logger` is passed with the command line interface, it
overrides the default `logger` in the agent cfg to be one of
tensorboard, wandb or neptune.

## Type of change

- New feature (non-breaking change which adds functionality)

## Checklist

- [x] I have run the [`pre-commit` checks](https://pre-commit.com/) with
`./orbit.sh --format`
- [ ] I have made corresponding changes to the documentation
- [x] My changes generate no new warnings
- [ ] I have added tests that prove my fix is effective or that my
feature works
- [ ] I have updated the changelog and the corresponding version in the
extension's `config/extension.toml` file
- [x] I have added my name to the `CONTRIBUTORS.md` or my name already
exists there

---------
Co-authored-by: Mayank Mittal <mittalma@leggedrobotics.com>

.gitignore

@@ -32,6 +32,8 @@
 **/output/*
 **/outputs/*
 **/videos/*
+**/wandb/*
+**/.neptune/*
 docker/artifacts/
 *.tmp
 

source/extensions/omni.isaac.orbit_tasks/omni/isaac/orbit_tasks/utils/wrappers/rsl_rl/rl_cfg.py

@@ -6,6 +6,7 @@
 from __future__ import annotations
 
 from dataclasses import MISSING
+from typing_extensions import Literal
 
 from omni.isaac.orbit.utils import configclass
 
@@ -93,6 +94,16 @@ class RslRlOnPolicyRunnerCfg:
     empirical_normalization: bool = MISSING
     """Whether to use empirical normalization."""
 
+    policy: RslRlPpoActorCriticCfg = MISSING
+    """The policy configuration."""
+
+    algorithm: RslRlPpoAlgorithmCfg = MISSING
+    """The algorithm configuration."""
+
+    ##
+    # Checkpointing parameters
+    ##
+
     save_interval: int = MISSING
     """The number of iterations between saves."""
 
@@ -100,7 +111,29 @@ class RslRlOnPolicyRunnerCfg:
     """The experiment name."""
 
     run_name: str = ""
-    """The run name. Default is empty string."""
+    """The run name. Default is empty string.
+
+    The name of the run directory is typically the time-stamp at execution. If the run name is not empty,
+    then it is appended to the run directory's name, i.e. the logging directory's name will become
+    ``{time-stamp}_{run_name}``.
+    """
+
+    ##
+    # Logging parameters
+    ##
+
+    logger: Literal["tensorboard", "neptune", "wandb"] = "tensorboard"
+    """The logger to use. Default is tensorboard."""
+
+    neptune_project: str = "orbit"
+    """The neptune project name. Default is "orbit"."""
+
+    wandb_project: str = "orbit"
+    """The wandb project name. Default is "orbit"."""
+
+    ##
+    # Loading parameters
+    ##
 
     resume: bool = False
     """Whether to resume. Default is False."""
@@ -116,9 +149,3 @@ class RslRlOnPolicyRunnerCfg:
 
     If regex expression, the latest (alphabetical order) matching file will be loaded.
     """
-
-    policy: RslRlPpoActorCriticCfg = MISSING
-    """The policy configuration."""
-
-    algorithm: RslRlPpoAlgorithmCfg = MISSING
-    """The algorithm configuration."""

source/extensions/omni.isaac.orbit_tasks/omni/isaac/orbit_tasks/utils/wrappers/rsl_rl/vecenv_wrapper.py

@@ -87,6 +87,11 @@ class RslRlVecEnvWrapper(VecEnv):
     Properties -- Gym.Wrapper
     """
 
+    @property
+    def cfg(self) -> object:
+        """Returns the configuration class instance of the environment."""
+        return self.unwrapped.cfg
+
     @property
     def render_mode(self) -> str | None:
         """Returns the :attr:`Env` :attr:`render_mode`."""

source/standalone/workflows/rsl_rl/cli_args.py

@@ -26,6 +26,12 @@ def add_rsl_rl_args(parser: argparse.ArgumentParser):
     arg_group.add_argument("--resume", type=bool, default=None, help="Whether to resume from a checkpoint.")
     arg_group.add_argument("--load_run", type=str, default=None, help="Name of the run folder to resume from.")
     arg_group.add_argument("--checkpoint", type=str, default=None, help="Checkpoint file to resume from.")
+    arg_group.add_argument(
+        "--logger", type=str, default=None, choices={"wandb", "tensorboard", "neptune"}, help="Logger module to use."
+    )
+    arg_group.add_argument(
+        "--log_project_name", type=str, default=None, help="Name of the logging project when using wandb or neptune."
+    )
 
 
 def parse_rsl_rl_cfg(task_name: str, args_cli: argparse.Namespace) -> RslRlOnPolicyRunnerCfg:
@@ -52,5 +58,13 @@ def parse_rsl_rl_cfg(task_name: str, args_cli: argparse.Namespace) -> RslRlOnPol
         rslrl_cfg.load_run = args_cli.load_run
     if args_cli.checkpoint is not None:
         rslrl_cfg.load_checkpoint = args_cli.checkpoint
+    if args_cli.run_name is not None:
+        rslrl_cfg.run_name = args_cli.run_name
+    if args_cli.logger is not None:
+        rslrl_cfg.logger = args_cli.logger
+    # set the project name for wandb and neptune
+    if rslrl_cfg.logger in {"wandb", "neptune"} and args_cli.log_project_name:
+        rslrl_cfg.wandb_project = args_cli.log_project_name
+        rslrl_cfg.neptune_project = args_cli.log_project_name
 
     return rslrl_cfg

source/standalone/workflows/rsl_rl/train.py

@@ -27,6 +27,7 @@ parser.add_argument("--video_interval", type=int, default=2000, help="Interval b
 parser.add_argument("--cpu", action="store_true", default=False, help="Use CPU pipeline.")
 parser.add_argument("--num_envs", type=int, default=None, help="Number of environments to simulate.")
 parser.add_argument("--task", type=str, default=None, help="Name of the task.")
+parser.add_argument("--run_name", type=str, default=None, help="Name of the run.")
 parser.add_argument("--seed", type=int, default=None, help="Seed used for the environment")
 # append RSL-RL cli arguments
 cli_args.add_rsl_rl_args(parser)
@@ -81,7 +82,7 @@ def main():
     log_root_path = os.path.join("logs", "rsl_rl", agent_cfg.experiment_name)
     log_root_path = os.path.abspath(log_root_path)
     print(f"[INFO] Logging experiment in directory: {log_root_path}")
-    # specify directory for logging runs
+    # specify directory for logging runs: {time-stamp}_{run_name}
     log_dir = datetime.now().strftime("%Y-%m-%d_%H-%M-%S")
     if agent_cfg.run_name:
         log_dir += f"_{agent_cfg.run_name}"
@@ -105,6 +106,8 @@ def main():
 
     # create runner from rsl-rl
     runner = OnPolicyRunner(env, agent_cfg.to_dict(), log_dir=log_dir, device=agent_cfg.device)
+    # write git state to logs
+    runner.add_git_repo_to_log(__file__)
     # save resume path before creating a new log_dir
     if agent_cfg.resume:
         # get path to previous checkpoint