Add headless EGL pixel rendering benchmark - #63
Open
eust-w wants to merge 1 commit into
Open
Conversation
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
Why
BiGym documents pixel observations but does not currently provide a reproducible way to validate headless GPU rendering. This makes blank frames, software-rendering fallbacks and container device-permission problems difficult to distinguish from environment or policy failures.
The new example exercises a real ReachTarget environment with rgb_head observations, reports physics-plus-rendering steps per second, and produces a small machine-readable result that can be compared across hosts.
AMD and portability boundary
On AMD Linux hosts, MuJoCo normally reaches the GPU through Mesa EGL/OpenGL and /dev/dri; installing ROCm is not required for BiGym's built-in pixel observations. External policies and renderers may still require ROCm, but those dependencies remain outside BiGym.
The same example works for other Linux EGL devices. MUJOCO_GL=osmesa remains documented as a CPU fallback, not a GPU performance result. MUJOCO_EGL_DEVICE_ID follows the device-selection support in the official MuJoCo Python EGL implementation.
Downstream Radeon W7900D integration evidence exists at https://github.com/eust-w/amd-bigym-openpi-eval/blob/main/evidence/amd-full-recorder-smoke-20260805.json, but this PR does not present that forked/custom-renderer run as an execution of this exact canonical example.
Validation
Supersedes #58 after the source branch was renamed to remove the agent/ prefix.