Datasets:
re-chunk to <=384-token windows
Browse files- README.md +24 -0
- holdout.jsonl +0 -0
- train.jsonl +0 -0
- validation.jsonl +0 -0
README.md
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: cc-by-4.0
|
| 3 |
+
task_categories:
|
| 4 |
+
- token-classification
|
| 5 |
+
tags:
|
| 6 |
+
- ner
|
| 7 |
+
- gliner2
|
| 8 |
+
- data-use
|
| 9 |
+
configs:
|
| 10 |
+
- config_name: default
|
| 11 |
+
data_files:
|
| 12 |
+
- split: train
|
| 13 |
+
path: train.jsonl
|
| 14 |
+
- split: validation
|
| 15 |
+
path: validation.jsonl
|
| 16 |
+
- split: holdout
|
| 17 |
+
path: holdout.jsonl
|
| 18 |
+
---
|
| 19 |
+
|
| 20 |
+
# datause-dataset (re-chunked)
|
| 21 |
+
|
| 22 |
+
Re-chunk of `ai4data/datause-dataset` into <=384-token windows (max_tokens=384, overlap=50) so GLiNER's window is never truncated during evaluation.
|
| 23 |
+
|
| 24 |
+
`{"train": {"orig": 1779, "rechunked": 1822}, "validation": {"orig": 415, "rechunked": 458}, "holdout": {"orig": 1149, "rechunked": 1166}}`
|
holdout.jsonl
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
train.jsonl
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
validation.jsonl
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|