Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
|
@@ -30,4 +30,20 @@ configs:
|
|
| 30 |
path: data/validation-*
|
| 31 |
- split: test
|
| 32 |
path: data/test-*
|
|
|
|
|
|
|
| 33 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 30 |
path: data/validation-*
|
| 31 |
- split: test
|
| 32 |
path: data/test-*
|
| 33 |
+
license: other
|
| 34 |
+
source_datasets: []
|
| 35 |
---
|
| 36 |
+
|
| 37 |
+
# evidence-inference
|
| 38 |
+
|
| 39 |
+
Evidence Inference v2: full article + intervention/comparator/outcome → effect direction.
|
| 40 |
+
|
| 41 |
+
Join annotations by PromptID AND PMCID, never by article alone. Use verified
|
| 42 |
+
labels, exclude tied votes and upstream known-bad prompts. Evidence spans in
|
| 43 |
+
metadata refer only to the same prompt and winning label. Full article text
|
| 44 |
+
stays in the input so isolated snippets need not identify treatment arms.
|
| 45 |
+
Native article train/validation/test splits are preserved. The upstream code
|
| 46 |
+
and annotations have MIT terms; article-text rights vary and are not relabeled
|
| 47 |
+
MIT. Mirror text/data licensing remains unspecified.
|
| 48 |
+
|
| 49 |
+
Repackaged as parquet for [tasksource](https://github.com/sileod/tasksource) by [scripts/repackage_dataset/](https://github.com/sileod/tasksource/tree/main/scripts/repackage_dataset).
|