Upload README.md with huggingface_hub

README.md

@@ -35,19 +35,20 @@ SpatialWorld is a comprehensive benchmark designed to evaluate spatial reasoning
 
 ## 📊 Dataset Statistics
 
-| Platform | Scenarios | Tasks | Task Types |
-|----------|-----------|-------|------------|
-| AI2Thor | 120 | ~2,500 | Object manipulation, navigation |
-| CARLA | 5 Towns | 80 | Urban navigation, traffic scenarios |
-| ProcTHOR | Procedural | ~1,200 | Indoor navigation, household tasks |
-| VirtualHome | 7 Houses | ~800 | Multi-agent household activities |
-| EmbodiedCity | City-scale | ~400 | Real-world urban scenarios |
-| Game Tasks | Various | 100 | Rubik's cube, maze solving, etc. |
+This Hugging Face repository contains 630 unified tasks:
+
+| Platform | Tasks | Task Types |
+|----------|-------|------------|
+| AI2Thor | 343 | Object manipulation, navigation |
+| CARLA | 80 | Urban navigation, traffic scenarios |
+| ProcTHOR | 127 | Indoor navigation, household tasks |
+| VirtualHome | 80 | Multi-agent household activities |
+| **Total** | **630** | Multi-platform spatial reasoning |
 
 ## 🗂️ Repository Structure
 
 ```
-benchmark/            # Full benchmark data (all platforms and tasks)
+benchmark/            # Full benchmark data with original JSON structures
 ├── ai2thor/
 ├── carla/
 ├── procthor/
@@ -56,10 +57,13 @@ benchmark/            # Full benchmark data (all platforms and tasks)
 ├── game/
 └── shared/
 
-data/                 # Sample data for Hugging Face Dataset Viewer
-└── spatialworld_samples.parquet   # 12 sample tasks (2 per platform)
+data/                 # Unified Hugging Face dataset
+├── spatialworld_unified.jsonl     # 630 tasks in unified schema
+└── spatialworld_samples.parquet  # Sample for quick preview
 ```
 
+The Hugging Face dataset (`spatialworld_unified.jsonl`) provides a unified schema across all platforms. Original platform-specific JSON files remain in `benchmark/` for reference.
+
 ## 🎮 Unified Action Space
 
 All tasks use a standardized action space:
@@ -109,7 +113,30 @@ Each task contains:
 
 ## 🔧 Usage
 
-### Loading Tasks
+### Loading from Hugging Face
+
+```python
+from datasets import load_dataset
+
+# Load the unified dataset (all 630 tasks from all platforms)
+dataset = load_dataset("Spatialworld/Spatialworld-bench", split="train")
+
+# Access a task
+task = dataset[0]
+print(task["task_id"])           # e.g., "ai2thor00000"
+print(task["platform"])         # e.g., "ai2thor"
+print(task["instruction"])
+
+# Parse JSON fields
+import json
+golden_actions = json.loads(task["golden_actions_json"])
+success_conditions = json.loads(task["success_conditions_json"])
+platform_specific = json.loads(task["platform_specific_json"])
+```
+
+### Loading Original Tasks (Local)
+
+For full benchmark with original JSON structures:
 
 ```python
 import json
@@ -126,6 +153,25 @@ print(task["instruction"])
 print(task["golden_actions"]["actions"])
 ```
 
+### Data Format
+
+The Hugging Face dataset provides a unified schema across all platforms:
+
+| Field | Type | Description |
+|-------|------|-------------|
+| `task_id` | string | Unique identifier (e.g., "ai2thor00000") |
+| `platform` | string | Platform name (ai2thor/carla/procthor/virtualhome) |
+| `task_name` | string | Human-readable task name |
+| `instruction` | string | Natural language instruction |
+| `scene` | string | Scene identifier (FloorPlan, Town, etc.) |
+| `max_steps` | int | Maximum allowed steps |
+| `golden_actions_json` | string | JSON-encoded golden action sequence |
+| `success_conditions_json` | string | JSON-encoded success conditions |
+| `target_object_types_json` | string | JSON-encoded target objects |
+| `success_logic` | string | Logic for combining success conditions (AND/OR) |
+| `target_description` | string | Detailed target description |
+| `platform_specific_json` | string | Platform-specific fields (scene_index, executor, etc.) |
+
 ### Action Parsing
 
 ```python