Repository navigation
Conversation
Measure ActiveVLN and StreamVLN navigation through the versioned HTTP model-service contract with fixed manifests, primitive history, per-episode evidence and explicit tensor-batch validation. Pin the published model-only EmbodiInfer optimization revision. Keep physical-robot deployment and asynchronous navigation changes on their existing branches. Signed-off-by: yufoo1 <yufoo1.cs@gmail.com>
Signed-off-by: yufoo1 <yufoo1.cs@gmail.com>
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Superseded by #11 after renaming the source branch to
benchmark/vln-navigation. The commit remains03983a7dae682749754db7f10445e0313683b453.Summary
Add Habitat navigation benchmarks for ActiveVLN and StreamVLN using the versioned model-service HTTP API. Preserve StreamVLN primitive-frame history and independent tensor-batch episodes; record executed actions, model/transport timing, SR/SPL evidence and resumable per-episode results. Validate manifests, checkpoint action grammar, batch identity and failure cleanup.
Pin
third_party/embodiinferto the published model-only optimization revision in EmbodiInfer PR 12. Merge that dependency first. Inference code and device-specific first-128 summaries remain in EmbodiInfer; this change contains no Go2 deployment or asynchronous physical-navigation changes.Type of change
Verification
CUDA_VISIBLE_DEVICES='' PYTHONPATH=src NO_PROXY=localhost,127.0.0.1 no_proxy=localhost,127.0.0.1 python -m pytest -q tests/test_activevln_navigation_benchmark.py tests/test_streamvln_navigation_benchmark.py58 passed. Ruff check/format passed for all 13 changed Python files; diff whitespace and Markdown links passed. No GPU, Habitat, real-weight or physical-robot workloads were started for this integration; published measurements refer to their preserved frozen source revisions.
Process boundaries
src/embodirun.Checklist