Upload README.md with huggingface_hub

README.md

@@ -75,4 +75,86 @@ configs:
   data_files:
   - split: train
     path: sft/train-*
+language:
+- code
+- en
+license: other
+source_datasets:
+- bigcode/commitpackft
+task_categories:
+- text-generation
+pretty_name: CodeAlign curated CommitPackFT
+license_name: per-sample-permissive
+tags:
+- code
+- sft
+- commitpackft
+- code-quality
 ---
+
+# CodeAlign — curated CommitPackFT (8 languages)
+
+Instruction/code pairs from [`bigcode/commitpackft`](https://huggingface.co/datasets/bigcode/commitpackft), filtered for syntax validity (tree-sitter), per-language lint errors, cyclomatic complexity, internal duplication and cross-sample near-duplicates (MinHash/LSH). Built as the SFT set of [CodeAlign](https://github.com/Sergasgr/codealign); the pipeline, thresholds and the full curation report live in that repository.
+
+| config | rows | contents |
+|---|---|---|
+| `sft` (default) | 122,018 | accepted samples — the SFT training set minus the rows dropped for secrets (see Privacy) |
+| `annotated` | 145,058 | every processed sample with its curation verdict (`status`, `error`) |
+
+```python
+from datasets import load_dataset
+sft = load_dataset("Sergasgr/codealign-commitpackft", "sft", split="train")
+```
+
+## Format
+
+`messages` is ChatML (`user` prompt, `assistant` target file). Two prompt types, derived from the commit itself: `new_file` (write-from-spec; commit created the file) and `edit` (existing file + commit message as instruction). 84.6% of `sft` rows are `edit`.
+
+## Languages (`sft`)
+
+| language | rows |
+|---|---|
+| javascript | 45,084 |
+| python | 39,725 |
+| java | 15,779 |
+| c_sharp | 8,062 |
+| typescript | 4,339 |
+| cpp | 3,668 |
+| go | 3,034 |
+| rust | 2,327 |
+
+## Licensing and provenance
+
+Only samples whose upstream `license` is one of apache-2.0, bsd-2-clause, bsd-3-clause, cc0-1.0, isc, mit, unlicense are included; the licence of every sample is in its `license` column and applies to that sample's code.
+
+| license | `sft` rows |
+|---|---|
+| mit | 76,499 |
+| apache-2.0 | 26,849 |
+| bsd-3-clause | 11,913 |
+| bsd-2-clause | 3,887 |
+| isc | 1,423 |
+| unlicense | 985 |
+| cc0-1.0 | 462 |
+
+As in CommitPackFT, each row keeps its source `commit` and `repos` so the copyright holder can be identified (100.0% of `sft` rows carry provenance). Code authors who want their code removed can open an issue on the GitHub repository.
+
+## Known issues in the quality columns
+
+`lint_errors` holds the values computed by the v1.0 curation run, which had linter bugs (the data was not re-curated):
+- **C++, JavaScript, TypeScript:** `lint_errors` is always 0, so these languages were filtered on syntax, complexity and duplication only. cpplint's total was parsed from the wrong output stream (fixed in the repository afterwards); ESLint ≥ 9 rejects the `--no-eslintrc` command line (still open).
+- **Python:** every value includes ruff's summary line (+1, or +2 when ruff also printed a fix hint), so the actual number of violations is 1–2 lower (fixed in the repository afterwards).
+
+## Privacy
+
+- Rows containing a high-confidence secret (private key blocks, AWS / GitHub / Slack / Google / Stripe credentials) were dropped: 48 from `sft`, 56 from `annotated`.
+- E-mail addresses (other than documentation and GitHub no-reply domains) were replaced with `<EMAIL>` in 8,272 `sft` rows.
+- Detection is pattern-based and will miss some personal data; do not use this dataset to identify individuals.
+
+## Decontamination
+
+13-gram overlap of every HumanEval and MBPP problem (prompt + canonical solution) against the Python samples is reported in `src/notebooks/01_curation_report.ipynb` of the GitHub repository.
+
+## Citation
+
+Upstream data: Muennighoff et al., *OctoPack: Instruction Tuning Code Large Language Models* (2023), arXiv:2308.07124.