Upload README.md with huggingface_hub

README.md

@@ -23,16 +23,16 @@ Benchmark results for [DocLD](https://docld.com) table extraction on the
 
 | Metric | Value |
 |--------|-------|
-| **Mean Accuracy** | 80.6% |
-| **Median** | 81.2% |
-| **P25 / P75** | 75.2% / 85.8% |
-| **Min / Max** | 49.2% / 100.0% |
-| **Scored Samples** | 68 |
-| **Total Samples** | 500 |
+| **Mean Accuracy** | 82.2% |
+| **Median** | 82.3% |
+| **P25 / P75** | 72.2% / 96.8% |
+| **Min / Max** | 41.4% / 100.0% |
+| **Scored Samples** | 451 |
+| **Total Samples** | 1000 |
 
 ## Methodology
 
-- **Dataset**: [FinTabNet_OTSL](https://huggingface.co/datasets/docling-project/FinTabNet_OTSL) — 500 samples from the test split
+- **Dataset**: [FinTabNet_OTSL](https://huggingface.co/datasets/docling-project/FinTabNet_OTSL) — 1000 samples from the test split
 - **Extraction**: DocLD agentic table extraction (VLM-based, gpt-5-mini)
 - **Scoring**: Needleman-Wunsch hierarchical alignment (same as [RD-TableBench](https://github.com/reductoai/rd-tablebench))
 - **Output**: HTML tables with rowspan/colspan for merged cells
@@ -41,7 +41,7 @@ Benchmark results for [DocLD](https://docld.com) table extraction on the
 
 | Provider | FinTabNet Accuracy |
 |----------|-------------------|
-| **DocLD** | **80.6%** |
+| **DocLD** | **82.2%** |
 | GTE (IBM) | ~78% |
 | TATR (Microsoft) | ~65% |