Update README.md

README.md

@@ -1,30 +1,35 @@
 ---
-pretty_name: NIST CSF 2.0 Q&A (Sample)
+language:
+- en
+license: apache-2.0
 tags:
-- nist
-- cybersecurity
+- nist-csf
 - grc
-- question-answering
+- cybersecurity
+- policy
+- qna
 task_categories:
 - text-generation
-license: cc-by-4.0
-language:
-- en
+pretty_name: GRC NIST CSF 2.0 QA (v1)
+configs:
+- config_name: alpaca
+  data_files: nist_csf2_qa_alpaca.jsonl
+- config_name: chatml
+  data_files: nist_csf2_qa_chat.jsonl
 ---
 
-# NIST CSF 2.0 Q&A (Sample)
-Small starter dataset derived from public NIST CSF 2.0 materials. Each row is a JSON object with fields:
-- `question`: user question
-- `answer`: reference-style answer
-
-## Format
-JSONL — one object per line.
+# GRC NIST CSF 2.0 QA (v1)
 
-## Intended use
-Bootstrapping GRC assistants. Not authoritative; verify against official sources.
+This dataset contains two JSONL files for training/evaluating LLMs on Governance, Risk & Compliance focused on **NIST CSF 2.0**.
 
-## Citation / Sources
-Link to original public sources you used.
+## Files
+- **nist_csf2_qa_alpaca.jsonl** — Alpaca format:
+  - `instruction`, `input` (optional), `output`
+- **nist_csf2_qa_chat.jsonl** — ChatML-style turns:
+  - `messages`: list of `{role, content}`
 
-## Limitations
-May omit context; verify before production use.
+## Example: Load with `datasets`
+```python
+from datasets import load_dataset
+ds = load_dataset("USERNAME/grc-nist-csf2-qa-v1")  # replace USERNAME
+print(ds)