Datasets:
File size: 1,297 Bytes
90e79ee | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 | # Reproducing Pointerbench-Text
The benchmark is generated programmatically from the
`text-coordinate-grounding` generator in the source repository.
```bash
# from the generator directory in the source repo:
python3 build_textbench.py --n 500 --seed 90211 --out /path/to/pointerbench-text
```
- **Seed `90211`** is disjoint from the training-data seed (`7`), so every
benchmark page is unseen during training.
- The build renders one distinct page per example, saves it as
`data/test/<i>.png`, and writes one `data/test/metadata.jsonl` row with the
instruction, the pixel `bbox`, a reference `point`, `answer_type`, `eval`, and
the data_type / category / surface / language / difficulty tags.
- Target kinds follow the generator's native `KIND_WEIGHTS`; bbox-answer kinds
are preserved instead of being filtered out.
- Language mix is fixed by the builder: 50% English, then 10% each of German,
French, Spanish, Italian, Dutch.
- Rendering is deterministic given the seed (pure PIL), and each instruction is
re-resolved by an independent verifier to confirm a single matching target,
so the box coordinates are pixel-exact.
Regenerating with the same seed reproduces the identical 500 examples. To grow
or refresh the set, change `--n` / `--seed` and bump the dataset version.
|