{"contract_version":"1.4.0","rubric_version":"1.0.0","archetype_version":"1.1.0","skill_version":"0.2.0","tier":"deep","judge_model":"sonnet","judge_effort":"medium","commit_sha":"c0bd9e62468ec3e20b88b91ddf20375777409331","filed_at":"2026-07-23T06:15:02.302Z","repo":{"owner":"huggingface","name":"transformers","default_branch":"main"},"purpose":{"what_its_for":["unified model-definition layer for ML models","inference pipelines","training integration across ecosystem tools","1M+ hosted checkpoints access"],"purpose_statement":"Transformers centralizes model definitions for text, vision, audio and multimodal models so training and inference frameworks across the ecosystem can share a single implementation.","audience":"ML engineers and researchers building on pretrained models, and downstream framework/tooling authors (vLLM, DeepSpeed, TGI, etc.)","category_candidates":["ml-model-hub-client","nlp-framework","model-inference-library","ml-training-framework"],"archetype":"framework"},"scores":{"axes":{"concept":{"score":8,"grade":"AA","rationale_ref":"src-src-transformers-modeling-utils-py"},"usefulness":{"score":9.5,"grade":"AAA","rationale_ref":"probe-claim-4"},"ergonomics":{"score":7,"grade":"A","rationale_ref":"vitals-readme"},"maturity":{"score":7,"grade":"A","rationale_ref":"static"},"longevity":{"score":9.5,"grade":"AAA","rationale_ref":"health-static"},"openness":{"score":8,"grade":"AA","rationale_ref":"health-static"},"setup_friction":{"score":8,"grade":"AA","rationale_ref":"vitals-readme"},"issue_health":{"score":5.5,"grade":"BB","rationale_ref":"health-static"},"supply_chain":{"score":4,"grade":"B","rationale_ref":"probe-cve"},"hype_substance":{"score":7.5,"grade":"A","rationale_ref":"probe-claim-3"}},"rollups":{"dependability":6.7,"craftsmanship":7.3,"docs":7.6,"trust":6.6},"composite":7.5,"grade":"A","recommendation":"TRIAL","confidence":"high"},"summary":{"context_line":"Audited from a 0k-token evidence dossier; 4 files read, 13 probes answered.","intro":["Transformers is the model-definition layer HuggingFace positions as the shared pivot across training frameworks (Axolotl, DeepSpeed, FSDP) and inference engines (vLLM, TGI, llama.cpp) [vitals-readme]. The pipeline() quickstart and pip install path work as described [vitals-readme], and the codebase's core files back the README's model-definition claims closely [probe-claim-1][probe-claim-2].","The dependency scan surfaces 17 critical and 83 high-severity CVEs across the resolved dependency tree with no lockfile to pin against [probe-cve][probe-deps], and 85% of open issues sit stale beyond 180 days against a backlog of 958 [health-static]."],"knocks":[{"lead":"The dependency scan turned up a large unresolved CVE surface.","text":"Osv-scanner found 17 critical and 83 high-severity vulnerabilities in the resolved dependency tree, and the repo ships no lockfile to pin against [probe-cve][probe-deps]. Nothing in the dossier engages these findings.","refs":["probe-cve","probe-deps"]},{"lead":"The issue tracker is dominated by stale tickets.","text":"85% of 958 open issues are stale beyond 180 days [health-static], which for a repo this size means most reported problems go unresolved rather than triaged shut.","refs":["health-static"]},{"lead":"Code duplication and complexity concentrate in the same hot files.","text":"36.2% of the codebase is duplicated and the highest-churn files (modeling_utils.py, trainer.py, generation/utils.py) also carry the highest complexity scores [static], raising the cost of touching the parts of the API most consumers depend on.","refs":["static"]}],"holds_up":[{"lead":"The install-to-first-call path is short and works as documented.","text":"pip install transformers[torch] followed by the pipeline() quickstart example runs a real text-generation call with no extra configuration [vitals-readme].","refs":["vitals-readme"]},{"lead":"Release and review discipline is consistent at scale.","text":"100 releases in the last 24 months all carry notes, and 95% of sampled merged PRs went through at least one review with a 50% external-author merge share [health-static].","refs":["health-static"]},{"lead":"Core internals are exercised by an extensive test suite.","text":"983 test files contain real assertions and the most graph-central module is referenced by over a thousand test files [probe-assertions][probe-top-tested].","refs":["probe-assertions","probe-top-tested"]}],"bottom_line":"The de facto standard for pretrained models, but adopt with an explicit plan for its unresolved CVE backlog and thin issue triage."},"use_cases":[{"use_case":"run inference on a pretrained text, vision, or audio model quickly","fit":"strong","note":"pipeline() API handles preprocessing and works out of the box after a single pip install [vitals-readme]."},{"use_case":"share a single model implementation across training and inference frameworks","fit":"strong","note":"positioned and integrated as the model-definition layer consumed by vLLM, DeepSpeed, TGI and others [vitals-readme]."},{"use_case":"fine-tune or train a transformer model with a standard Trainer loop","fit":"strong","note":"trainer.py is a heavily used, heavily tested core module [static][probe-assertions]."},{"use_case":"build a security-sensitive production pipeline with strict dependency pinning","fit":"poor","note":"no lockfile ships and the dependency scan shows 17 critical, 83 high CVEs unaddressed [probe-cve][probe-deps]","instead":"pin a vetted subset of dependencies yourself and audit the resolved tree before deploying"},{"use_case":"expect fast triage on a filed GitHub issue","fit":"poor","note":"85% of the 958 open issues are stale beyond 180 days [health-static]","instead":"use HuggingFace forums/Discord or file targeted, minimal repro PRs instead"},{"use_case":"extend or modify a hot-path internal like modeling_utils.py","fit":"partial","note":"high complexity and churn concentrate in the same files, raising the cost of safe changes [static]"}],"alternatives_note":"For narrower needs, single-framework libraries (e.g. sentence-transformers, timm) trade breadth for a smaller dependency and CVE surface."}