Skip to content

Pull requests: EvolvingLMMs-Lab/lmms-eval

New pull request New
40 Open 909 Closed
40 Open 909 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

fix(mathvision): extract the last \\boxed{...} with balanced braces
#1512 opened Sep 4, 2026 by feiiiiii5 Contributor Loading…
1 task done
fix(mathvision): unwrap \\text{...}/\\mbox{...} instead of leaving braces behind
#1510 opened Sep 4, 2026 by feiiiiii5 Contributor Loading…
1 task done
fix(mathvision): call _fix_fracs unconditionally in _strip_string
#1508 opened Sep 4, 2026 by feiiiiii5 Contributor Loading…
1 task done
fix(mathvision): preserve optional root indices in _fix_sqrt
#1506 opened Sep 4, 2026 by feiiiiii5 Contributor Loading…
1 task done
1
feat: add group-level postprocess hook for custom result aggregation
#1499 opened Aug 31, 2026 by lorenzozanee Contributor Loading…
7 tasks
1
feat: add EgoLifeQA benchmark
#1498 opened Aug 31, 2026 by lorenzozanee Contributor Loading…
1 task done
1
fix(llava): handle transformers compat for apply_chunking_to_forward import
#1497 opened Aug 31, 2026 by lorenzozanee Contributor Loading…
1 task done
1
fix(mmvp): apply ground truth corrections in code for robust evaluation
#1495 opened Aug 31, 2026 by lorenzozanee Contributor Loading…
7 tasks
1
fix: make opencv optional on linux ppc64le to allow uv sync
#1494 opened Aug 31, 2026 by lorenzozanee Contributor Loading…
1 of 7 tasks
1
1
fix(qwen2_5_vl): propagate sampled fps to processor for correct second_per_grid_ts
#1491 opened Aug 30, 2026 by lorenzozanee Contributor Loading…
1 task done
1
perf: (PR 4/4) add reproducible evaluator benchmark harness
#1460 opened Aug 23, 2026 by Luodian Contributor • 4/4 • Draft
1 of 7 tasks
perf: (PR 3/4) instrument evaluation phases
#1459 opened Aug 23, 2026 by Luodian Contributor • 3/4 • Draft
1 of 7 tasks
perf: (PR 2/4) atomically write baseline records
#1457 opened Aug 23, 2026 by Luodian Contributor • 2/4 • Draft
1 of 7 tasks
perf: (PR 1/4) add baseline phase recorder
#1456 opened Aug 23, 2026 by Luodian Contributor • 1/4 • Draft
1 of 7 tasks
perf: (PR 4/4) capture baseline provenance
#1455 opened Aug 23, 2026 by Luodian Contributor • 4/4 • Draft
1 of 7 tasks
perf: (PR 3/4) add secret-safe baseline invocation
#1454 opened Aug 23, 2026 by Luodian Contributor • 3/4 • Draft
1 of 7 tasks
perf: (PR 2/4) add fail-closed legacy redaction
#1453 opened Aug 23, 2026 by Luodian Contributor • 2/4 • Draft
1 of 7 tasks
perf: (PR 1/4) add RFC 8785 JSON primitives
#1452 opened Aug 23, 2026 by Luodian Contributor • 1/4 • Draft
1 of 7 tasks
test: (PR 2/2) remove mirrored request-layout checks
#1451 opened Aug 23, 2026 by Luodian Contributor • 2/2 • Draft
1 of 7 tasks
refactor(models): (collector) make ModelRegistryV2 authoritative
#1450 opened Aug 23, 2026 by Luodian Contributor • Draft
3 of 7 tasks
docs(models): (PR 4/4) align registry contribution guidance
#1449 opened Aug 23, 2026 by Luodian Contributor • 4/4 • Draft
1 of 7 tasks
refactor(models): (PR 3/4) route model discovery through V2
#1448 opened Aug 23, 2026 by Luodian Contributor • 3/4 • Draft
3 of 7 tasks
refactor(models): (PR 2/4) derive legacy views from V2
#1447 opened Aug 23, 2026 by Luodian Contributor • 2/4 • Draft
1 of 7 tasks
refactor(models): (PR 1/4) isolate plugin registration failures
#1446 opened Aug 23, 2026 by Luodian Contributor • 1/4 • Draft
1 of 7 tasks
security(tui): (collector) land structured execution and origin guards
#1445 opened Aug 23, 2026 by Luodian Contributor • Draft
2 of 7 tasks
ProTip! no:milestone will show everything without a milestone.