Update README.md

README.md

@@ -80,6 +80,41 @@ The goal is to measure whether models can reliably perceive the critical **Visua
 
 > **Note:** The files in the `data/` directory are included solely to enable previewing examples in the Hugging Face Dataset Viewer. They are not part of the recommended usage pipeline, and users may safely ignore them when using EPIC-Bench.
 
+
+## Download
+
+EPIC-Bench contains ~35,000 small files across three task categories. To work around HuggingFace's per-file rate limit (1,000 API requests / 5 min) and to keep download speed reasonable, the annotations are distributed as **three task-level tarballs** instead of raw folders.
+
+Download the tarballs from [HuggingFace](https://huggingface.co/datasets/rxc205/EPIC-Bench) (or [ModelScope](https://www.modelscope.cn/datasets/macarich/EPIC-Bench)):
+
+| Archive | Size | # Files | Contents |
+|---------|------|---------|----------|
+| `Manipulation.tar.gz`       | 1.70 GB | 7,061   | AffordanceRegion / ContactRelationship / PlacementRegion |
+| `Navigation.tar.gz`         | 2.71 GB | 10,868  | FeasiblePath / GroundDetection / VisualMatching |
+| `TargetLocalization.tar.gz` | 3.34 GB | 17,665  | BasicAttributes / EmbodiedCompositionalAttributes / SpatialRelatedAttributes |
+
+Place them under `dataset/annotation/` and extract:
+
+```bash
+mkdir -p dataset/annotation/EPIC_Bench
+cd dataset/annotation/EPIC_Bench
+
+# Option 1: download via huggingface-cli (recommended; supports resume)
+hf download rxc205/EPIC-Bench \
+    Manipulation.tar.gz Navigation.tar.gz TargetLocalization.tar.gz \
+    --repo-type dataset --local-dir .
+
+# Option 2: direct wget
+wget https://huggingface.co/datasets/rxc205/EPIC-Bench/resolve/main/Manipulation.tar.gz
+wget https://huggingface.co/datasets/rxc205/EPIC-Bench/resolve/main/Navigation.tar.gz
+wget https://huggingface.co/datasets/rxc205/EPIC-Bench/resolve/main/TargetLocalization.tar.gz
+
+# Extract all three (preserves the original folder layout)
+for f in Manipulation.tar.gz Navigation.tar.gz TargetLocalization.tar.gz; do
+    tar -xzf "$f" && rm "$f"
+done
+```
+
 ## ✨ Highlights
 
 -  **Embodied-Scenario** evaluation of VLM **Visual Perception** capability.