# Horizon-Labs/content-safety-guard-small

Model from Hugging Face for text classification: Horizon-Labs/content-safety-guard-small · transformers

## Facts

- Type: LLM
- Task: text-classification
- Framework: transformers
- License: apache-2.0 — Permissive open license: commercial use, modification, and redistribution are permitted. (https://spdx.org/licenses/Apache-2.0.html)
- Source: https://huggingface.co/Horizon-Labs/content-safety-guard-small
- Hosting: hosted (5 files)
- Files: 5 hosted of 27 listed
- File counts: "listed" is every file in the source repository; only hosted files download from Hugging Bay, and the download plan lists the files it selects for one download, so its count can differ.
- Hosted here: 5 files, 597 MB
- Full repo upstream: 1.43 GB
- Recommended file: 1.39 GB (3 files)
- Download: https://huggingbay.xyz/artifact/hf-model-horizon-labs-content-safety-guard-small/files (download plan: https://huggingbay.xyz/api/artifacts/hf-model-horizon-labs-content-safety-guard-small/download-plan)
- Per-file SHA-256: available for 5 of 5 hosted files via https://huggingbay.xyz/api/artifacts/hf-model-horizon-labs-content-safety-guard-small/files?limit=100
- Run-ready: no (Hugging Bay has not certified a run path for this model; hosted bytes are downloadable and hash-checkable only)
- Peer-assisted fallbacks: 0 verified, 0 pending review
- Downloads: 113
- Stars: 0
- Card completeness: 96%

## Description

External Hugging Face catalog record for Horizon-Labs/content-safety-guard-small. Task: text classification. Library: transformers. Languages: multilingual, en, ar, de, es, fr, hi, it. Datasets referenced by card metadata: nvidia/Nemotron-Safety-Guard-Dataset-v3, allenai/WildChat-1M, OpenAssistant/oasst2, CohereForAI/aya_dataset, OpenSafetyLab/Salad-Data, JailbreakBench/JBB-Behaviors. Base model: jhu-clsp/mmBERT-small. Files we host have a published SHA-256.

## Model Card Facts

- Declared license: apache-2.0
- Pipeline: text-classification
- Library: transformers
- Base model: jhu-clsp/mmBERT-small
- Task categories: text-classification
- Languages: multilingual, en, ar, de, es, fr, hi, it
- Referenced datasets: nvidia/Nemotron-Safety-Guard-Dataset-v3, allenai/WildChat-1M, OpenAssistant/oasst2, CohereForAI/aya_dataset, OpenSafetyLab/Salad-Data, JailbreakBench/JBB-Behaviors, google/civil_comments
- File hints: 16 upstream file samples

## Referenced Data

- Dataset: nvidia/Nemotron-Safety-Guard-Dataset-v3
- Dataset: allenai/WildChat-1M
- Dataset: OpenAssistant/oasst2
- Dataset: CohereForAI/aya_dataset
- Dataset: OpenSafetyLab/Salad-Data
- Dataset: JailbreakBench/JBB-Behaviors
- Dataset: google/civil_comments

## File Hints

- .gitattributes (1.5 KB)
- README.md (14 KB)
- code/safety/build_pool2.py (1.2 KB)
- code/safety/build_prompt_pool.py (3.7 KB)
- code/safety/build_safety_evals.py (5.0 KB)
- code/safety/build_safety_v0.py (4.4 KB)
- code/safety/build_safety_v2.py (2.9 KB)
- code/safety/build_safety_v3.py (2.3 KB)
- code/safety/cat_thresholds.py (2.6 KB)
- code/safety/eval_safety.py (6.3 KB)
- code/safety/label_teacher.py (1.1 KB)
- code/safety/qwenguard.py (3.3 KB)

## Quality Checklist

- Summary: passed
- Description: passed
- Model card metadata: passed
- License: passed
- Provenance: passed
- File inventory: passed
- Hosted bytes: passed
- Scan result: passed
- Manifest signature: passed

## Safety Notes

- External catalog metadata imported from Hugging Face public API on 2026-10-01T09:43:15.350Z.
- Open license detected as apache-2.0.
- Model card references dataset context: nvidia/Nemotron-Safety-Guard-Dataset-v3, allenai/WildChat-1M, OpenAssistant/oasst2, CohereForAI/aya_dataset.
- Not a run kit: hosted, hash-checked bytes only; run readiness is not certified for this model yet.

## Actions

- Web: https://huggingbay.xyz/artifact/hf-model-horizon-labs-content-safety-guard-small
- Files API: https://huggingbay.xyz/api/artifacts/hf-model-horizon-labs-content-safety-guard-small/files?limit=100
- Metadata: https://huggingbay.xyz/api/artifacts/hf-model-horizon-labs-content-safety-guard-small/metadata.json
- Distribution: https://huggingbay.xyz/api/artifacts/hf-model-horizon-labs-content-safety-guard-small/distribution
- Manifest: https://huggingbay.xyz/api/manifests/hf-model-horizon-labs-content-safety-guard-small
- Download plan (per-file SHA-256 + signed download URLs): https://huggingbay.xyz/api/artifacts/hf-model-horizon-labs-content-safety-guard-small/download-plan
- Hosted download rows: https://huggingbay.xyz/api/downloadable-files?artifactId=hf-model-horizon-labs-content-safety-guard-small&limit=100
- Request mirror: not applicable.
- Submit peer fallback for review: POST https://huggingbay.xyz/api/artifacts/hf-model-horizon-labs-content-safety-guard-small/decentralized-fallbacks