Manually convert and push optimized Parquet format (overwriting) from local cache

README.md

@@ -6,7 +6,33 @@ tags:
 - LeRobot
 configs:
 - config_name: default
-  data_files: data/*/*.parquet
+  data_files:
+  - split: train
+    path: data/train-*
+dataset_info:
+  features:
+  - name: action
+    list: float32
+    length: 6
+  - name: observation.state
+    list: float32
+    length: 6
+  - name: timestamp
+    dtype: float32
+  - name: frame_index
+    dtype: int64
+  - name: episode_index
+    dtype: int64
+  - name: index
+    dtype: int64
+  - name: task_index
+    dtype: int64
+  splits:
+  - name: train
+    num_bytes: 2459100
+    num_examples: 29275
+  download_size: 782664
+  dataset_size: 2459100
 ---
 
 This dataset was created using [LeRobot](https://github.com/huggingface/lerobot).

data/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a41ff468cffe522f775ed88e1c8695be1119c8b26aff288963162302f233ea77
+size 782664