Sentence Similarity
sentence-transformers
Safetensors
bert
feature-extraction
compliance
nist-800-53
hipaa
cybersecurity
text-embeddings-inference
Instructions to use stetteh/regmap-embedder with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- sentence-transformers
How to use stetteh/regmap-embedder with sentence-transformers:
from sentence_transformers import SentenceTransformer model = SentenceTransformer("stetteh/regmap-embedder") sentences = [ "That is a happy person", "That is a happy dog", "That is a very happy person", "Today is a sunny day" ] embeddings = model.encode(sentences) similarities = model.similarity(embeddings, embeddings) print(similarities.shape) # [4, 4] - Notebooks
- Google Colab
- Kaggle
model card: add public availability (HF/Docker/Release)
Browse files
README.md
CHANGED
|
@@ -24,6 +24,12 @@ control set onto HIPAA without manual, line-by-line mapping.
|
|
| 24 |
- **Fine-tuning:** `MultipleNegativesRankingLoss` on curated NIST↔HIPAA control/provision pairs
|
| 25 |
- **Task:** semantic retrieval (embed a control, cosine-rank against the HIPAA corpus, return top-k)
|
| 26 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 27 |
## Intended use — an *assistive* retriever, not an authoritative classifier
|
| 28 |
RegMap returns the **top-k most similar HIPAA provisions** for a human to review and confirm. It is
|
| 29 |
designed to accelerate an expert's mapping work, not to make a final compliance determination on its
|
|
|
|
| 24 |
- **Fine-tuning:** `MultipleNegativesRankingLoss` on curated NIST↔HIPAA control/provision pairs
|
| 25 |
- **Task:** semantic retrieval (embed a control, cosine-rank against the HIPAA corpus, return top-k)
|
| 26 |
|
| 27 |
+
## Where to get it
|
| 28 |
+
- **Hugging Face:** `stetteh/regmap-embedder` — `SentenceTransformer("stetteh/regmap-embedder")`
|
| 29 |
+
- **Docker (serving API):** `docker run -p 8080:8080 ghcr.io/samuelgtetteh/regmap-embedder:0.1`
|
| 30 |
+
then `POST /map {"control": "..."}` → top-k HIPAA provisions
|
| 31 |
+
- **GitHub Release:** `v0.1-regmap` — a self-contained archive (model + corpus + wrapper)
|
| 32 |
+
|
| 33 |
## Intended use — an *assistive* retriever, not an authoritative classifier
|
| 34 |
RegMap returns the **top-k most similar HIPAA provisions** for a human to review and confirm. It is
|
| 35 |
designed to accelerate an expert's mapping work, not to make a final compliance determination on its
|