rrrepo-raterRATINGS AGENCY
search 2.1M repos…/

Categories / huggingface/transformers / evidence

huggingface/transformersframework

Transformers centralizes model definitions for text, vision, audio and multimodal models so training and inference frameworks across the ecosystem can share a single implementation.

Pythonstars 163,268license Apache-2.0pushed today
audiodeep-learningdeepseekgemmaglmhacktoberfestllmmachine-learningmodel-hubnatural-language-processingnlppretrained-modelspythonpytorchpytorch-transformersqwenspeech-recognitiontransformervlm
7.5COMPOSITE
REPO-RATERADEEP AUDIT · 07/2026

Dossier compilation

This audit's judge read a single 13.6k-token evidence dossier, compiled mechanically under per-section budgets. Below: each section, what went in, and what was cut. Citations on the audit page resolve here.

#Section / contentsModeSpent / budget
[28]Vitals + README (distilled)SECTION1.4k / 1.5k
[1]README.md · distilled: badges/TOC strippedDOSSIER1.4k
[2]Structure (dir-aggregated)SECTION738 / 1.0k
[29]Health + Static digestSECTION1.3k / 2.0k
[3](github api)PROBE625
[4](static analysis)PROBE625
[30]Manifests + lockfiles + CI (verbatim)SECTION1.2k / 1.2k
[5]pyproject.tomlDOSSIER294
[6].github/workflows/dependents-smoke-test.ymlDOSSIER294
[7].github/workflows/self-scheduled-amd-mi250-caller.ymlDOSSIER294
[8].github/workflows/release-conda.ymlDOSSIER294
[31]Core files (complete / whole-function)SECTION5.0k / 5.0k
[9]src/transformers/modeling_utils.py · partial: 1/132 whole functionsDOSSIER2.5k
[10]src/transformers/testing_utils.py · partial: 1/297 whole functionsDOSSIER2.1k
[11]src/transformers/image_processing_utils.py · partial: 1/21 whole functionsDOSSIER310
cut: src/transformers/configuration_utils.py — over budget (17507 tok), functions too large to extract
cut: src/transformers/training_args.py — over budget (35405 tok), functions too large to extract
cut: src/transformers/processing_utils.py — over budget (29581 tok), functions too large to extract
cut: src/transformers/modeling_outputs.py — over budget (26422 tok), functions too large to extract
[32]Test realitySECTION1.4k / 1.5k
[12]tests/models/markuplm/test_feature_extraction_markuplm.pyDOSSIER1.4k
[33]Probe answersSECTION1.4k / 1.5k
[13] probe-deps[14] probe-license[15] probe-ci[16] probe-todo[17] probe-assertions[18] probe-top-tested[19] probe-cve[20] probe-claim-1[21] probe-claim-2[22] probe-claim-3[23] probe-claim-4[24] probe-claim-5[25] probe-claim-6
[26]Outlines (files not read above)SECTION323 / 500
[34]Coverage notes (what was cut & why)SECTION257 / 400
[27](composer)DOSSIER257
6320 files mapped · 1 complete, 3 whole-function · 13 probes13.6k / 15k cap

Measured properties

Raw signals behind the scores. Source tags: COMPUTED derived from the clone · API GitHub/registry data · JUDGE model estimate

loc_source1253655 COMPUTED
loc_test448187 COMPUTED
test_to_source_ratio0.36 COMPUTED
files_tracked6320 COMPUTED
language_mix[object Object] COMPUTED
deps_direct0 COMPUTED
lockfileCOMPUTED
ci_workflows60 COMPUTED
security_mdtrue COMPUTED
total_commits23450 COMPUTED
commits_last_90d747 COMPUTED
top_author_share_pct6 COMPUTED
bus_factor_80pct20 COMPUTED
first_commit2018-10-29 COMPUTED
last_commit2026-07-23 COMPUTED
gh_stars162854 API
gh_forks33996 API
gh_open_issues2400 API
gh_pushed_at2026-07-23T05:37:18Z API
gh_created_at2018-10-29T13:56:00Z API
gh_archivedfalse API
gh_licenseApache-2.0 API
gh_topicsaudio,deep-learning,deepseek,gemma,glm,hacktoberfest,llm,machine-learning,model-hub,natural-language-processing,nlp,pretrained-models,python,pytorch,pytorch-transformers,qwen,speech-recognition,transformer,vlm API
gh_default_branchmain API
issue_open958 API
issue_stale_ratio_pct85 API
issue_median_first_response_days0.2 API
issue_median_close_days0.8 API
issue_bot_first_response_pct0 API
pr_open1442 API
pr_median_merge_days0.9 API
pr_external_merge_share_pct50 API
pr_reviewed_share_pct95 API
releases_total100 API
releases_last_24mo100 API
release_latest2026-07-16T09:41:36Z API
release_notes_share_pct100 API
release_latest_automatedfalse API
contributors_100cap100 API
contributor_top10_share_pct53 API
community_health_pct100 API
discussions_enabledfalse API
complexity_p9044 STATIC
hotspot_topsrc/transformers/modeling_utils.py,src/transformers/trainer.py,src/transformers/generation/utils.py,src/transformers/models/auto/tokenization_auto.py,src/transformers/testing_utils.py,src/transformers/tokenization_utils_base.py,src/transformers/utils/import_utils.py,src/transformers/processing_utils.py,src/transformers/training_args.py,utils/modular_model_converter.py,src/transformers/cache_utils.py,utils/check_docstrings.py,src/transformers/configuration_utils.py,utils/check_repo.py,src/transformers/models/moshi/modeling_moshi.py STATIC
dup_pct36.2 STATIC
dup_worst_pair.github/ISSUE_TEMPLATE/bug-report.yml ↔ .github/ISSUE_TEMPLATE/migration.yml (26 lines) STATIC
cve_count_critical17 STATIC
cve_count_high83 STATIC
cve_count_medium59 STATIC
secrets_found79 STATIC
secret_scan_depthworking tree + last 50 commits STATIC
smell_empty_catch0 STATIC
smell_dynamic_exec4 STATIC
smell_long_functions244 STATIC
smell_deep_nesting759 STATIC
smell_todo_density0.52 STATIC
dossier_tokens13562 COMPUTED
graph_coverage_pct100 COMPUTED

Claim traceability

Every judgment on the audit page, mapped to what backs it.

ClaimBacked byKind
pipeline() quickstart runs a working text-generation call after pip installvitals-readmeOBSERVATION
Dependency scan found 17 critical and 83 high-severity CVEs with no lockfileprobe-cve probe-depsMEASUREMENT
85% of open issues are stale beyond 180 dayshealth-staticMEASUREMENT
36.2% of the codebase is duplicated, concentrated in high-churn hotspot filesstaticMEASUREMENT

Coverage, provenance & bloat

How much of the repository this audit actually covered, who probably wrote it, and how heavy it is — informational, never part of the grade.

Coverage

structure mapped: 6320 files · dossier: 13562 tok · complete files: 0 · whole-function: 3 · probes: 13

Provenance

AI-likelihoodUNLIKELY

large long-running multi-contributor project with human commit history and governance files

Bloat signal (lean 0 → heavy 10, descriptive only)

Index6.0
deps_direct0
deps_transitive
install_footprint
bundle_vs_category_median
loc_to_api_ratio
boilerplate_density