Set evaluation_framework to inspect-ai in eval.yaml
Browse files
eval.yaml
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
name: olmOCR-bench
|
| 2 |
description: olmOCR-bench is a benchmark to evaluate how well VLLMs can parse PDF and scanned documents into clean, semantic markdown/HTML.
|
| 3 |
-
evaluation_framework:
|
| 4 |
|
| 5 |
tasks:
|
| 6 |
- id: arxiv_math
|
|
|
|
| 1 |
name: olmOCR-bench
|
| 2 |
description: olmOCR-bench is a benchmark to evaluate how well VLLMs can parse PDF and scanned documents into clean, semantic markdown/HTML.
|
| 3 |
+
evaluation_framework: inspect-ai
|
| 4 |
|
| 5 |
tasks:
|
| 6 |
- id: arxiv_math
|