text2sql: hard negatives and teacher scores (jina35-u2)

README.md

@@ -50,7 +50,7 @@ A seeded sample of [`gretelai/synthetic_text_to_sql`](https://huggingface.co/dat
 | Layout | `queries` · `corpus` · `qrels` · `hard-negatives` · `teacher-scores`, split `train` |
 | Splits | `corpus`: train · `hard-negatives`: train · `qrels`: train · `queries`: train · `teacher-scores`: train |
 | Hard negatives | sources: `dense` · 2,970,047 rows |
-| Teacher scores | none yet (0 rows): `jinaai/jina-reranker-v3.5` scores come next |
+| Teacher scores | `jinaai/jina-reranker-v3.5` · 3,000,047 rows (positives included) |
 | Ids | `sha1(text)[:20]`; identical texts collapse to one document / query |
 | License | `apache-2.0` |
 
@@ -78,14 +78,26 @@ Files are Parquet, sorted by id, zstd-compressed, sharded at 500 MB. Every rule
 
 ## Hard negatives and teacher scores
 
-Filled by the collection's annotation pipeline (`annotation=jina35`). **Interim**: the candidates are final, the teacher scores are still to come.
+Filled by the owner's annotation pipeline (`annotation=jina35-u2`) for the `train` split of the
+query set(s) below; queries without a labelled positive are left out.
 
-- **Candidates**: dense retrieval with `jinaai/jina-embeddings-v5-text-small` over this corpus to depth 1,000; 100 candidates per query drawn from the rank windows 1–30 (30), 31–100 (30), 101–300 (20), 301–1000 (20), the query's labelled positives excluded. `rank` is the dense rank; `source` is `dense` for a mined row and `dataset` for a negative the source labels itself.
-- **Teacher scores**: none yet. `teacher-scores` holds 0 rows until the `jinaai/jina-reranker-v3.5` scores (listwise, as in the other repositories) are filled in; `datasets` cannot return a 0-example split, so read that file with `pyarrow` / `pandas` meanwhile. The candidates stay.
+- **Candidates**: dense retrieval with `jinaai/jina-embeddings-v5-text-small` over the full corpus to depth 1,000;
+  100 candidates per query drawn from the rank windows 1–30 (30), 31–100 (30), 101–300 (20), 301–1000 (20), the query's labelled positives excluded.
+  `rank` is the dense rank; `source` is `dense` for a mined row and `dataset` for a negative the source labels itself
+  (those are kept for every query of the split, sampled or not).
+- **Teacher**: `jinaai/jina-reranker-v3.5`, listwise: a query's positive and all of its candidates are scored together in one
+  context of up to 32,768 tokens. `score` is the raw cosine score, one row per (query, positive) and per
+  (query, candidate); a labelled negative that was also mined is scored once. No filtering is applied to the tables.
 
 | configs | queries | hard negatives | teacher scores |
 |---|---:|---:|---:|
-| `hard-negatives` · `teacher-scores` | 30,000 (all) | 2,970,047 (2,970,047 dense) | 0 |
+| `hard-negatives` · `teacher-scores` | 30,000 (all) | 2,970,047 (2,970,047 dense) | 3,000,047 |
+
+```python
+from datasets import load_dataset
+negatives = load_dataset("Hyukkyu/train-text2sql", "hard-negatives", split="train")
+scores    = load_dataset("Hyukkyu/train-text2sql", "teacher-scores", split="train")
+```
 
 ## Load it
 

teacher-scores/train-00000-of-00001.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8316266e7cbe6867227eefc104eb93ee2252b885a6127a24b8e5c1982716406a
-size 815
+oid sha256:6312edb98cefb6ef67c2a1def65c55230814ef28561d814d3fc1f96c8a9c4c29
+size 19622737