From a6d7e6ec1791698b84d692445e3692d743ac0dac Mon Sep 17 00:00:00 2001 From: morgan-pulse Date: Fri, 16 Oct 2026 10:15:13 -0700 Subject: [PATCH] Add Pulse effort ExtractBench Pro result submission --- README.md | 2 ++ 1 file changed, 2 insertions(+) diff --git a/README.md b/README.md index 759bdd71..40876eba 100644 --- a/README.md +++ b/README.md @@ -210,6 +210,8 @@ uv run extract-bench status [ExtractBench Pro](https://huggingface.co/datasets/llamaindex/ExtractBenchPro) contains 100 documents across six challenges: document conventions, house vocabulary, house counting rules, drawn relations, charts inside documents, and long chart sheets. It uses the same evaluator as ExtractBench. Read the [ExtractBench Pro launch post](https://www.llamaindex.ai/blog/introducing-extractbench-pro) for the challenges and results. +Pulse (Effort) submission: **27% perfect documents (27/100)**. + Download the PDFs, schemas, ground truth, and scoring rules into the layout the harness reads: ```bash