Repository navigation
feat: backend iterations param, step timing, and bench script - #203
Merged
Merged
Conversation
added 2 commits
April 10, 2026 17:09
motrixsim methods return untyped (Any) values; wrapping with np.asarray() provides the concrete ndarray type mypy expects.
Collaborator
|
TATP-233
added a commit
that referenced
this pull request
May 4, 2026
…ming feat: backend iterations param, step timing, and bench script
TATP-233
added a commit
that referenced
this pull request
May 30, 2026
…ming feat: backend iterations param, step timing, and bench script
TATP-233
added a commit
that referenced
this pull request
Sep 4, 2026
…ming feat: backend iterations param, step timing, and bench script
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
iterationsfield toEnvCfgand pass through to both MuJoCo (opt.iterations) and Motrix (options.max_iterations) backendsSimBackend.step()return type todict | None, returning timing breakdown under{"timing": {...}}key (set_ctrl_ms,physics_ms,refresh_cache_ms)step_n()for multi-substep, remove redundantnp.asarray()wrappers, fixset_stateto use correctdata_sliceforforward_kinematicNpEnv.step()timing instrumentation withbackend_prefixed keysscripts/bench_env_step.py: runs all task × backend combinations by default (go1/go2/g1 × mujoco/motrix) with box-drawing comparison tableTest plan
make testpassesuv run python scripts/bench_env_step.py task=g1_joystick training.sim_backend=mujocouv run python scripts/bench_env_step.pyuv run python scripts/train_offpolicy.py task=go2_joystick training.sim_backend=motrix