Release pass: licence classes, recipe links, fixed snippet

README.md

@@ -1,9 +1,10 @@
 ---
 license: other
 license_name: mixed
-pretty_name: Curation decision mix (draft)
+pretty_name: Curation decision mix
 task_categories: [text-classification, zero-shot-classification]
-tags: [data-curation, decision-model, gliclass, provenance]
+language: [en, zh, pt, ja, id, ar, ko, fr, el, es, pl, uk, cs, hi, tr, sv, it, de, da, nl, "no", fi]
+tags: [data-curation, decision-model, decision-models, system-one, gliclass, provenance, hf-jobs]
 size_categories: [100K<n<1M]
 configs:
   - {config_name: all, data_files: "data/*/train.parquet", default: true}
@@ -18,15 +19,18 @@ configs:
   - {config_name: procpile, data_files: data/procpile/train.parquet}
 ---
 
-<!-- DRAFT 2026-10-28, private review PR. -->
+# Curation decision mix
 
-# Curation decision mix — draft
-
-The training data of `davanstrien/ultrabert-380m-curation`, a small classifier that answers curation questions about web pages, prompts,
+The training data of [davanstrien/ultrabert-380m-curation](https://huggingface.co/davanstrien/ultrabert-380m-curation), a small decision model that answers curation questions about web pages, prompts,
 maths problems and library records ("what topic is this page?", "how educational is it?", "how hard is this
 problem?"). Each row is a text, a question, a set of options and a probability for each option. Every row records
 where it came from, who made its label and under which licence.
 
+This card is also the data recipe: it says which sources were used, why, and which choices moved the scores (see
+[How the mix was designed](#how-the-mix-was-designed)). The scripts are in
+[github.com/davanstrien/curation-decision-recipe](https://github.com/davanstrien/curation-decision-recipe). If you
+train decision models, please share how you made your data too.
+
 | | |
 |---|---|
 | Rows | 516,964 released, from a 535,730-row training mix; 18,766 base rows without a verified open licence are left out (see Licences) |
@@ -34,22 +38,23 @@ where it came from, who made its label and under which licence.
 | Row shapes | decision rows `{state, question, target}` and GLiClass rows `{text, all_labels, true_labels, prompt}` |
 | Labels made by | dataset gold, library cataloguers, programs, and LLMs (named per source and per row) |
 | Licence | mixed, per row; web page text is **not** included (labels plus a pointer to the source row; rebuild snippet below) |
-| Trains | `davanstrien/ultrabert-380m-curation` (UltraBERT-380M + GLiClass head; 12,000 steps over this mix) |
-| Build | mix builder on Hugging Face Jobs (CPU), then `scripts/release_build.py` (in this repo) for the release; file checksum (sha256) `3a1d0fa84e7342f7…`; test sets fixed before the mix was built, and every row checked against them |
+| Trains | [davanstrien/ultrabert-380m-curation](https://huggingface.co/davanstrien/ultrabert-380m-curation) (UltraBERT-380M + GLiClass head; 12,000 steps over this mix) |
+| Build | mix builder on Hugging Face Jobs (CPU), then [`scripts/release_build.py`](scripts/release_build.py) for the release (a record of how this repo was made: it reads the mix from a private bucket); test sets fixed before the mix was built, and every row checked against them |
 
 ```python
 import json
 from datasets import load_dataset
-ds = load_dataset("davanstrien/curation-decision-mix", "all", split="train")          # or one config per source, e.g. "decisionbank"
-ok = ds.filter(lambda r: r["licence_class"] == "permissive")        # filter by licence before reuse
-q = json.loads(ds[0]["question"])                                   # nested fields are JSON strings
+ds = load_dataset("davanstrien/curation-decision-mix", "all", split="train")   # or one config per source, e.g. "decisionbank"
+ok = ds.filter(lambda r: r["licence_class"] == "permissive")                 # filter by licence before reuse
+decisions = ok.filter(lambda r: r["row_type"] == "decision")                 # decision rows; the rest are GLiClass rows
+q = json.loads(decisions[0]["question"])                                     # nested fields are JSON strings
 ```
 
 ---
 
 ## Layout
 
-One config per source, plus `all`. Splitting by source keeps licence terms together, so a user can take only the parts
+One config per source, plus `all`. `web` rows have an empty `state` by design (see Web text). Splitting by source keeps licence terms together, so a user can take only the parts
 whose terms suit them.
 
 | Config | Rows (training mix) | Shape |
@@ -64,6 +69,9 @@ whose terms suit them.
 | `arxiv` | 13,597 | GLiClass |
 | `procpile` | 10,997 | GLiClass |
 
+`manifest/training_mix_manifest.parquet` lists all 535,730 rows of the training mix (including the 18,766 rows not
+released here): row id, source, licence class, labeller and a SHA-256 hash of each row's text.
+
 Columns (union of both shapes; unused fields are null):
 
 | Column | Meaning |
@@ -72,7 +80,7 @@ Columns (union of both shapes; unused fields are null):
 | `config`, `src` | source group and the finer origin (e.g. `db_bitext_support`, a tasksource task) |
 | `row_type` | `decision` or `gliclass` |
 | `teacher` | who made the label (e.g. `Llama-3.1-405B`, `cataloguer`, `program`, `50% source label + 50% Qwen3.8-27B`) |
-| `licence_class`, `licence` | `permissive` / `sa` (share-alike) / `publisher` (text not redistributed), and the licence id |
+| `licence_class`, `licence` | `permissive` / `sa` (share-alike, e.g. CC BY-SA, CDLA-Sharing) / `other_open` (MPL-2.0, AFL-3.0) / `publisher` (text not redistributed), and the licence id |
 | `state` | the input (decision rows): a string, or a record as a JSON string |
 | `question` | JSON string: `{type: noul|choice, instructions, criteria}`; `noul` = yes/no; `choice` options are `{key: description}`; graded scales use `level_<k>` keys, lowest first |
 | `target` | JSON string: probability per option (soft targets allowed; sums to 1 for single-answer rows) |
@@ -104,6 +112,8 @@ source dataset's `train` split; the training text was the first 3,000 characters
 (do not redistribute it):
 
 ```python
+from datasets import load_dataset
+
 src = {}
 def rebuild(r):
     if r["source_dataset"] not in src:
@@ -112,7 +122,7 @@ def rebuild(r):
 web = load_dataset("davanstrien/curation-decision-mix", "web", split="train").map(rebuild)
 ```
 
-Row indices assume the source datasets are unchanged since October 2026.
+Row indices assume the source datasets are unchanged since October 2026. The snippet loads each source dataset in full; it has not been tested end to end.
 
 ---
 
@@ -121,11 +131,11 @@ Row indices assume the source datasets are unchanged since October 2026.
 | Source | Rows | What the rows ask | Labels made by | Licence |
 |---|---:|---|---|---|
 | base (tasksource + knowledgator GLiClass data) | 173,859 (192,625 trained on) | many public classification tasks | dataset gold; knowledgator gliclass-v2.0 / v3-logic (LLM-generated) | per source repo, in `licence`; only rows with a verified permissive, share-alike or other open licence are released (see Licences) |
-| DecisionBank-260k (jgeuter/DecisionBank) | 165,277 | public tasks, written and programmatic decision questions | 50 % source label + 50 % Qwen3.8-27B; Qwen3.8-27B with thinking; 50 % Jeff 1.0 teachers (Qwen3.6-27B, Qwen3.5-122B-A10B) + 50 % Qwen3.8-27B (4,188 rows); code | per row: CC BY / CC BY-SA, Apache, MIT, CC0 (public domain), CDLA-Sharing only (non-commercial, research-only, unknown and other terms dropped) |
+| DecisionBank-260k (jgeuter/DecisionBank) | 165,277 | public tasks, written and programmatic decision questions | 50 % source label + 50 % Qwen3.8-27B; Qwen3.8-27B with thinking; 50 % Jeff 1.0 teachers (Qwen3.6-27B, Qwen3.5-122B-A10B) + 50 % Qwen3.8-27B (4,188 rows); code | per row: CC BY / CC BY-SA, Apache, MIT, CC0 (public domain), CDLA-Sharing-1.0; non-commercial, research-only, unknown and other terms dropped |
 | prompt slice | 52,491 | NuminaMath-1.5 question type / problem type / validity; DeepMath-103K difficulty / topic; WebInstruct-verified answer type / category / difficulty; LMArena criteria and tags | dataset columns (NuminaMath, WebInstruct: labeller not documented); GPT-4o (DeepMath); LMArena classifiers | Apache-2.0; MIT; Apache-2.0; CC-BY-4.0 |
 | web slice (**ids and labels only**) | 41,890 | WebOrganizer topic and format; FineWeb-Edu educational value | Llama-3.1-405B (WebOrganizer; the annotation datasets state no licence); Llama-3-70B (FineWeb-Edu) | labels: none stated; ODC-By. Page text: publisher copyright, not included (rebuild snippet above) |
 | National Library of Scotland catalogue | 23,846 | subject areas of library records | National Library of Scotland cataloguers | CC0 |
-| tool lane | 21,089 | tool selection | program-computed | permissive |
+| tool lane (code decisions) | 21,089 | linter findings (classification and yes/no) and complexity comparisons on functions from Python packages | program-computed (linter and complexity tools) | permissive package licences (MIT / BSD / Apache-2.0 / PSF); the package is not recorded per row |
 | Harvard modern catalogue | 13,918 | Library of Congress subject class, literary form, government publication | Harvard Library cataloguers (from their catalogue records) | CC0 |
 | arXiv | 13,597 | arXiv subject category | author-chosen categories | arXiv metadata (CC0) |
 | procedural pile | 10,997 | program-checkable reasoning | program-computed | CC BY 4.0 (attribute) |
@@ -136,7 +146,7 @@ Row indices assume the source datasets are unchanged since October 2026.
 
 The mix is the result of a series of screened experiments (2026-10-23 → 10-27). Each choice below states what it rests
 on: earlier decision-model reports, published curation work, or our own measurements. "Curation score" is the
-benchmark used for selection: the mean of 16 curation test sets, balanced by family (see the model card).
+benchmark used for selection: the mean of 16 curation test units (see the model card for the grouping and all 16 scores).
 
 **1. Start from a broad public base, cleaned by task identity.** The base follows the GLiClass v4 recipe
 (knowledgator): tasksource classification tasks plus knowledgator's GLiClass training sets, so the model keeps general
@@ -178,7 +188,7 @@ over the other rows (3 seeds each, all 9 pairs positive), mostly on evals whose
 labeller. Only rows with allowed licences were kept; the HelpSteer families were dropped because they are held-out test sets.
 
 **7. Program-computed lanes for reasoning and tools.** Procedural reasoning tasks, tool selection and arXiv
-categories have labels that are computed or author-chosen, so they carry no labeller bias. They help general decision
+categories have labels that are computed or author-chosen, so no LLM labeller is involved (the task design still shapes them). They help general decision
 tasks more than the curation score.
 
 ### How train and eval were split
@@ -193,8 +203,8 @@ tasks more than the curation score.
 - **Separate same-labeller from independent gold.** Where an eval's gold comes from the same LLM that labelled
   training rows (WebOrganizer, FineWeb-Edu, DeepMath), it partly measures agreement with that labeller. The fair
   comparison uses independent gold (Omni-MATH human ratings, Essential-Web), reported apart in the model card.
-- **Seeds and review.** Each data choice was screened with ≥ 2 training seeds and claimed with ≥ 3; every claimed result was reviewed by
-  two independent reviewers before adoption.
+- **Seeds and review.** Each data choice was screened with ≥ 2 training seeds and claimed with ≥ 3; every claimed result was checked by
+  two independent automated reviews (separate AI models) before adoption.
 
 ## Decontamination
 
@@ -206,9 +216,10 @@ DecisionBank row. The catalogue test sets were kept out by id and title when the
 
 ## Licences
 
-Licence classes in the training mix: permissive 396,026 rows, share-alike 97,814, publisher terms 41,890 (web text;
-trained on, never redistributed). DecisionBank's own parts are CC BY-SA 4.0; share-alike rows keep their licence in
-this dataset. The model is released under Apache-2.0 (share-alike data may train a permissively licensed model).
+Licence classes in this release: permissive 375,637 rows, share-alike 98,941, other open 496, publisher terms 41,890
+(`web` labels only; page text trained on, never redistributed). DecisionBank's own parts are CC BY-SA 4.0;
+share-alike rows keep their licence in this dataset. The model is released under Apache-2.0. These are the licences
+stated by each source; this is not legal advice, so check the terms of the sources you reuse.
 
 Base rows, by the licence on the source repo's card (or, for tasksource tasks, the origin dataset's card; not legal
 advice): of 200,000 rows before decontamination, 178,084 verified permissive, 1,127 share-alike (CC BY-SA), 497 under
@@ -229,11 +240,12 @@ and test sets in Hugging Face storage buckets mounted into each job.
 
 ## Known issues
 
-- Several label sources are LLM outputs (named above and per row); the WebOrganizer annotation datasets carry no
-  licence statement; the Llama 3 / 3.1 licences apply to the Llama-made labels.
+- Several label sources are LLM outputs (named above and per row in `teacher`). The WebOrganizer annotation datasets
+  carry no licence statement. The Llama 3 and Llama 3.1 licences cover the Llama-made labels (WebOrganizer, FineWeb-Edu);
+  read them before you reuse those rows.
 - Use terms of the original sources apply to their rows; check `licence` / `src` per row before reuse.
 - Multilingual rows (DecisionBank) are present; the model was evaluated in English only.
-- Maths difficulty labels are the weakest question type (LLM-made; human ratings agree only moderately).
+- Difficulty labels are subjective and depend on the labeller (LLM-made here; human ratings agree only moderately).
 
 ## Notes for agents
 

scripts/release_build.py

@@ -34,6 +34,8 @@ CONFIG_LICENCE = {"nls": "cc0-1.0", "harvard": "cc0-1.0", "arxiv": "cc0-1.0", "p
 SRC_LICENCE = {"numina_": "apache-2.0", "wi_": "apache-2.0", "dm_": "mit", "arena_": "cc-by-4.0",
                "fineweb_edu": "odc-by", "weborganizer_": None,
                "tool": "per package: mit / bsd / apache-2.0 / psf"}
+# Base rows: the class follows the source repo's resolved licence, not the mix-level "permissive" default.
+BASE_CLASS = {"permissive": "permissive", "sa": "sa", "other_open": "other_open"}
 # Web rows: the id suffix is the row index in the source dataset's train split (scripts/cur_build.py).
 WEB_SOURCE = {"cur-wo-topic-": "WebOrganizer/TopicAnnotations-Llama-3.1-405B-FP8",
               "cur-wo-format-": "WebOrganizer/FormatAnnotations-Llama-3.1-405B-FP8",
@@ -78,6 +80,11 @@ def main():
                     stats[f"base:drop_{entry['status']}"] += 1
                     continue
                 licence = entry["licence"]
+                licence_class = BASE_CLASS[entry["status"]]
+            if config != "base":
+                licence_class = r.get("licence_class")
+            if licence and ("-sa" in licence or "sharing" in licence):
+                assert licence_class == "sa", (r["id"], licence, licence_class)
             row_type = "gliclass" if "all_labels" in r else "decision"
             state = r.get("state")
             source_dataset = source_row = None
@@ -87,7 +94,7 @@ def main():
                 source_dataset, source_row = WEB_SOURCE[prefix], int(r["id"][len(prefix):])
             out = {
                 "id": r["id"], "config": config, "src": r.get("src"), "row_type": row_type,
-                "teacher": r.get("teacher"), "licence_class": r.get("licence_class"), "licence": licence,
+                "teacher": r.get("teacher"), "licence_class": licence_class, "licence": licence,
                 "state": state if state is None or isinstance(state, str) else as_json(state),
                 "question": as_json(r.get("question")), "target": as_json(r.get("target")),
                 "text": r.get("text"), "all_labels": r.get("all_labels"),
@@ -102,6 +109,7 @@ def main():
                 no_licence[(config, out["src"])] += 1
             rows[config].append(out)
             stats[f"{config}:out"] += 1
+            stats[f"class:{licence_class}"] += 1
 
     schema = {c: pl.Utf8 for c in COLUMNS}
     schema["all_labels"] = pl.List(pl.Utf8)