Repository navigation
feat(vllm): add ActiveVLN action service and simulation demo - #14
Merged
pppppppppper merged 1 commit intoNov 2, 2026
Merged
Conversation
Signed-off-by: yufoo1 <yufoo1.cs@gmail.com>
yufoo1
force-pushed
the
feat/vllm-activevln-integration
branch
from
November 2, 2026 08:17
cc0b9ca to
4d82f3e
Compare
yufoo1
marked this pull request as ready for review
November 2, 2026 08:28
ZhouAo-ZA
approved these changes
Nov 2, 2026
pppppppppper
approved these changes
Nov 2, 2026
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Add
provider: vllmto EmbodiRun with an optional native vLLM ActiveVLN action service, a Habitat action binding and one example configuration. Reuse the existing HTTP session/action protocol and addvllmto the simulator's existing backend choices. Core runtime code imports no inference framework.Include the bilingual simulation demo pages and README links. This PR leaves
tests/identical to main.Demo
EmbodiRun with EmbodiInfer / SGLang / vLLM — 1×, 4.9 seconds
The video shows elapsed time and distance to goal for one recorded RTX 4090 episode. Media is in BUAA-CI-LAB/misc#3; merge it before publishing the canonical documentation links. Recording conditions and SGLang rerun selection are described on the demo page.
Validation: existing provider/simulation/boundary checks (52 passed), Ruff and both strict documentation builds passed. The packaged vLLM entrypoint has not been rerun on a GPU; the video uses the earlier experimental service wrappers. Existing EmbodiInfer and SGLang providers retain their contracts.