Download eval.yaml from llamaindex/ExtractBenchPro: direct link, hf CLI and curl.
- Browser
- Download file 723 Bytes
-
https://huggingface.co/datasets/llamaindex/ExtractBenchPro/resolve/refs%2Fpr%2F7/eval.yaml
- Command line
-
hf download hf://datasets/llamaindex/ExtractBenchPro@refs/pr/7/eval.yaml
-
curl -L -o eval.yaml https://huggingface.co/datasets/llamaindex/ExtractBenchPro/resolve/refs%2Fpr%2F7/eval.yaml
723 Bytes
| name: ExtractBench Pro | |
| description: ExtractBench Pro evaluates schema-guided extraction across six challenging | |
| document categories. | |
| evaluation_framework: extractbench | |
| tasks: | |
| - id: mean | |
| config: extract-bench-pro | |
| split: mean | |
| - id: interpretation_rules | |
| config: extract-bench-pro | |
| split: interpretation_rules | |
| - id: custom_terminology | |
| config: extract-bench-pro | |
| split: custom_terminology | |
| - id: conditional_aggregation | |
| config: extract-bench-pro | |
| split: conditional_aggregation | |
| - id: relational_diagrams | |
| config: extract-bench-pro | |
| split: relational_diagrams | |
| - id: charts_and_maps | |
| config: extract-bench-pro | |
| split: charts_and_maps | |
| - id: long_single_pages | |
| config: extract-bench-pro | |
| split: long_single_pages | |