Upload README.md with huggingface_hub

README.md

@@ -1,52 +1,13 @@
----
-configs:
-- config_name: default
-  data_files:
-  - split: train
-    path: data/train-*
-  - split: validation
-    path: data/validation-*
-  - split: test
-    path: data/test-*
-dataset_info:
-  features:
-  - name: board
-    dtype: string
-  - name: size
-    dtype: int32
-  - name: moves
-    dtype: int32
-  - name: solution
-    sequence: string
-  - name: difficulty
-    dtype: string
-  - name: messages
-    sequence:
-    - name: role
-      dtype: string
-    - name: content
-      dtype: string
-  splits:
-  - name: train
-    num_bytes: 10448525
-    num_examples: 13917
-  - name: validation
-    num_bytes: 1307028
-    num_examples: 1740
-  - name: test
-    num_bytes: 1303967
-    num_examples: 1737
-  download_size: 995271
-  dataset_size: 13059520
----
-# Rush Hour 4×4, 5×5 & 6×6 — Balanced (solvable-only, A present)
-Enumerates boards (4×4/5×5), samples for selected sizes (e.g., 6×6), BFS-solves to optimal sequences, and stratifies by difficulty.
+# Rush Hour 4×4, 5×5 & 6×6 — Balanced (solvable-only, fixed thresholds)
+Enumerates boards (4×4/5×5) and/or samples for selected sizes (e.g., 6×6), BFS-solves to optimal sequences, and stratifies by difficulty.
 
 ## Policies
 - sizes: 4, 5, 6
 - solvable_only: True
 - exclude_already_solved: True
-- difficulty_quantiles: 0.33,0.66 (per size)
+- difficulty_scheme: fixed
+- fixed_thresholds: 3,5  # easy < t1, medium < t2, hard ≥ t2
+- balance_per_size: True
 
 ## Schema
 - board (str) | size (int) | moves (int) | solution (list[str]) | difficulty (easy/medium/hard) | messages (chat)