README.md

---
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train*.parquet
  - split: validation
    path: data/val*.parquet
  - split: train_positive
    path: data/train_positive*
  - split: train_negative
    path: data/train_negative*
  - split: val_positive
    path: data/val_positive*
  - split: val_negative
    path: data/val_negative*
dataset_info:
  description: "train_pos:234345;val_pos:41396"
  features:
  - name: pcd_rf
    sequence:
      sequence: float32
  - name: pcd_of
    sequence:
      sequence: float32
  - name: preshape_q
    sequence: float32
  - name: grasp_q
    sequence: float32
  - name: palm_xyz_xyzw_rf
    sequence: float32
  - name: palm_T_of
    sequence:
      sequence: float32
  - name: class
    dtype: string
  - name: object
    dtype: string
  - name: filename
    dtype: string
  - name: config_idx
    dtype: int64
---