Language

unilab.training¶

Shared training helpers for entrypoint scripts.

class unilab.training.BackendAdapter[source]¶

Bases: object

Build env/play overrides from the final composed config.

Parameters:

cfg (DictConfig)
root_dir (str | Path)
algo_name (str | None)
scene_materializer (Callable[..., str])

__init__(cfg, *, root_dir, algo_name=None, scene_materializer=<function materialize_scene_visual_override>)[source]¶

Parameters:

cfg (DictConfig)
root_dir (str | Path)
algo_name (str | None)
scene_materializer (Callable[..., str])

build_task_env_cfg_override()[source]¶

Build env_cfg_override from the resolved reward + env sections.

Return type:: dict[str, Any]

build_play_env_cfg_override()[source]¶

Build play-mode overrides from an optional backend-agnostic play profile.

Return type:: dict[str, Any]

class unilab.training.ExperimentTracker[source]¶

Bases: object

Tracks experiment metadata locally and optionally in Weights & Biases.

Parameters:

root_dir (str | Path)
log_dir (str | Path)
algo_name (str)
task_name (str)
sim_backend (str)
training_cfg (Any)
full_cfg (Any)
device (str | None)
collector_device (str | None)
seed_info (Any | None)

__init__(*, root_dir, log_dir, algo_name, task_name, sim_backend, training_cfg, full_cfg, device=None, collector_device=None, seed_info=None)[source]¶

Parameters:

root_dir (str | Path)
log_dir (str | Path)
algo_name (str)
task_name (str)
sim_backend (str)
training_cfg (Any)
full_cfg (Any)
device (str | None)
collector_device (str | None)
seed_info (Any | None)

property run: Any | None¶

property run_url: str | None¶

property wandb_settings: dict[str, Any]¶

start()[source]¶

Return type:: None

update_summary(summary=None)[source]¶

Parameters:: summary (dict[str, Any] | None)
Return type:: None

log_video(video_path, key='media/play_video')[source]¶

Parameters:

video_path (str | Path | None)
key (str)

Return type:

None

finish()[source]¶

Return type:: None

class unilab.training.HardwareMonitor[source]¶

Bases: object

Monitor CPU, GPU, memory usage.

__init__()[source]¶

get_metrics()[source]¶

Get current hardware metrics.

Return type:: Dict[str, float]

unilab.training.assert_offpolicy_task_choice_matches_algo(cfg, *, algo_name=None)[source]¶

Reject offpolicy configs whose task owner path does not match the selected algo.

Parameters:

cfg (DictConfig)
algo_name (str | None)

Return type:

None

unilab.training.create_env(cfg, *, num_envs, env_cfg_override=None, sim_backend=None, task_name=None)[source]¶

Construct an environment via the registry using the current Hydra config.

Parameters:

cfg (DictConfig)
num_envs (int)
env_cfg_override (dict[str, Any] | None)
sim_backend (str | None)
task_name (str | None)

unilab.training.ensure_registries()[source]¶

Import env modules so registry-based entrypoints can instantiate tasks.

Return type:: None

unilab.training.get_entrypoint_log_root(root_dir, *, algo_log_name, log_root=None)[source]¶

Resolve the log root for non-Hydra entrypoints using training helper semantics.

Parameters:

root_dir (str | Path)
algo_log_name (str)
log_root (str | Path | None)

Return type:

Path

unilab.training.get_hydra_runtime_choice(cfg, group)[source]¶

Return a selected Hydra config-group choice when runtime metadata is available.

Parameters:

cfg (DictConfig)
group (str)

Return type:

str | None

unilab.training.get_latest_checkpoint(run_dir, *, suffix='.pt')[source]¶

Return the latest model checkpoint inside a run directory.

Parameters:

run_dir (str | Path)
suffix (str)

Return type:

Path | None

unilab.training.get_latest_run(log_dir)[source]¶

Return the lexicographically latest run directory under a task log root.

Parameters:: log_dir (str | Path)
Return type:: Path | None

unilab.training.get_log_root(root_dir, cfg)[source]¶

Resolve the algorithm log root, honoring optional training.log_root overrides.

Parameters:

root_dir (str | Path)
cfg (DictConfig)

Return type:

Path

unilab.training.log_playback_plan(plan, *, prefix='')[source]¶

Print user-facing playback status for a resolved backend plan.

Parameters:

plan (BackendPlayRenderPlan)
prefix (str)

Return type:

None

unilab.training.parse_checkpoint_path(cfg, *, root_dir, load_run=None, task_name=None, checkpoint=None, suffix='.pt')[source]¶

Resolve a checkpoint path from Hydra config and repository root.

Parameters:

cfg (DictConfig)
root_dir (str | Path)
load_run (str | int | PathLike[str] | None)
task_name (str | None)
checkpoint (str | int | None)
suffix (str)

Return type:

tuple[Path | None, Path | None]

unilab.training.resolve_checkpoint_path(base_log_dir, load_run, *, suffix='.pt')[source]¶

Resolve a latest or explicit checkpoint path from a task log root.

Parameters:

base_log_dir (str | Path)
load_run (str | int | PathLike[str])
suffix (str)

Return type:

tuple[Path | None, Path | None]

unilab.training.resolve_task_checkpoint_path(root_dir, *, task_name, load_run, algo_log_name, checkpoint=None, suffix='.pt', log_root=None)[source]¶

Resolve checkpoint paths for auxiliary entrypoints through shared training semantics.

Parameters:

root_dir (str | Path)
task_name (str)
load_run (str | int | PathLike[str])
algo_log_name (str)
checkpoint (str | None)
suffix (str)
log_root (str | Path | None)

Return type:

tuple[Path | None, Path | None]

unilab.training.should_run_playback(*, play_only, no_play, play_render_mode)[source]¶

Return whether train/eval should enter playback for the configured mode.

Parameters:

play_only (bool)
no_play (bool)
play_render_mode (str | None)

Return type:

bool

class unilab.training.TrainingSeedInfo[source]¶

Bases: object

Configured and effective seed metadata for a training run.

Parameters:

configured_seed (int | None)
configured_seed_source (str | None)
effective_seed (int | None)

configured_seed: int | None¶

configured_seed_source: str | None¶

effective_seed: int | None¶

to_dict()[source]¶

Return type:: dict[str, int | str | None]

__init__(configured_seed, configured_seed_source, effective_seed)¶

Parameters:

configured_seed (int | None)
configured_seed_source (str | None)
effective_seed (int | None)

unilab.training.apply_configured_training_seed(cfg, *, torch_runtime=True, cuda=True, mlx_runtime=False)[source]¶

Resolve and apply the configured training seed before runtime construction.

Parameters:

cfg (Any)
torch_runtime (bool)
cuda (bool)
mlx_runtime (bool)

Return type:

TrainingSeedInfo

unilab.training.apply_training_seed(seed, *, torch_runtime=True, cuda=True, mlx_runtime=False)[source]¶

Apply a seed to the runtimes used by training entrypoints.

Parameters:

seed (int | None)
torch_runtime (bool)
cuda (bool)
mlx_runtime (bool)

Return type:

int | None

unilab.training.derive_worker_seed(base_seed, worker_index=0)[source]¶

Derive deterministic subprocess seeds from the effective run seed.

Parameters:

base_seed (int | None)
worker_index (int)

Return type:

int | None

unilab.training.resolve_training_seed(cfg)[source]¶

Resolve the configured seed, preferring the algorithm-level contract.

Parameters:: cfg (Any)
Return type:: TrainingSeedInfo

unilab.training.setup_logger(log_dir, algo_name, *, echo=True, filename='train.log')[source]¶

Create a simple file-backed logger for script-local progress messages.

Parameters:

log_dir (str | Path)
algo_name (str)
echo (bool)
filename (str)

Return type:

Logger

Modules

`backend_adapter`	Resolved config adaptation for training entrypoints.
`common`	Shared helpers for training entrypoints.
`experiment`	Shared experiment tracking utilities for local files and W&B.
`monitoring`	Hardware monitoring utilities for performance profiling.
`reward`	Utility functions for reward config handling.
`rsl_rl`	RSL-RL-specific training helpers.
`run`	Run directory and checkpoint resolution helpers.
`seed`	Shared training seed contract helpers.