---
dataset_info:
  features:
  - name: data
    sequence:
      sequence: float32
  - name: label
    dtype: int64
  splits:
  - name: train
    num_bytes: 531730928
    num_examples: 4324
  - name: val
    num_bytes: 5164824
    num_examples: 42
  - name: test
    num_bytes: 5164824
    num_examples: 42
  download_size: 207795149
  dataset_size: 542060576
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: val
    path: data/val-*
  - split: test
    path: data/test-*
license: odc-by
---

The [EEG Motor Movement/Imagery (MMI) Dataset](https://physionet.org/content/eegmmidb/1.0.0/) preprocessed with [DN3](https://github.com/SPOClab-ca/dn3/) to be used for downstream fine-tuning with [BENDR](https://github.com/SPOClab-ca/BENDR). 

The labels correspond to Task 4 (imagine opening and closing both fists or both feet) from experimental runs 4, 10 and 14.


## Creating dataloaders
```python
from datasets import load_dataset
from torch.utils.data import DataLoader

dataset = load_dataset("rasgaard/mmi-bendr-preprocessed")
dataset.set_format("torch")

train_loader = DataLoader(dataset["train"], batch_size=8)
val_loader = DataLoader(dataset["val"], batch_size=8)
test_loader = DataLoader(dataset["test"], batch_size=8)

batch = next(iter(train_loader))
batch["data"].shape, batch["label"].shape
>>> (torch.Size([8, 20, 1536]), torch.Size([8]))
```