Copy of DatologyAI/drop_silver @ a02643946413
Browse files- README.md +18 -0
- drop/train-00000-of-00001.parquet +3 -0
README.md
ADDED
|
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
dataset_info:
|
| 3 |
+
- config_name
|
| 4 |
+
- features
|
| 5 |
+
- splits
|
| 6 |
+
- download_size
|
| 7 |
+
- dataset_size
|
| 8 |
+
configs:
|
| 9 |
+
- config_name: drop
|
| 10 |
+
data_files:
|
| 11 |
+
- split: train
|
| 12 |
+
path: drop/train-*
|
| 13 |
+
---
|
| 14 |
+
**Worked-answer continuations, not the bare gold.** Each row's `silver` field is a short explained answer (a justification or worked solution) that ends in the gold answer; BPB is scored over that text, whereas the OLMES `<eval>:bpb` cloze heads score only the gold label or number.
|
| 15 |
+
|
| 16 |
+
> Copied from `DatologyAI/drop_silver` @ `a02643946413`.
|
| 17 |
+
|
| 18 |
+
|
drop/train-00000-of-00001.parquet
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4e7791a651d7e3b053a877a92a835fec675ea26ae758b9a933dfc51f84295d7d
|
| 3 |
+
size 8855718
|