sajjad5221/eeg2text-unique-sentences

EEG-to-Text RawNet Model

Brain-Computer Interface model for translating EEG signals to text.

Model Details

• Architecture: RawNet (Multi-scale Conv1D + Transformer)
• Best Validation Loss: 1.4613
• Parameters: 33,956,593

Usage

from huggingface_hub import hf_hub_download
import torch

model_path = hf_hub_download(repo_id="sajjad5221/eeg2text-unique-sentences", filename="best_rawnet_model.pt")
checkpoint = torch.load(model_path)
model.load_state_dict(checkpoint['model_state_dict'])

Dataset: sajjad5221/eeg2text-emmt-dataset-unique-sentences