MAGNET: Multi-Label Text Classification using Attention-based Graph Neural Network
Under development to implement this paper https://arxiv.org/abs/2003.11644 Using PyTorch
This is the result of what i am doing so far:
| Embedding | Dataset | Seq Length | Train Val Split | Micro-F1 Score (Validation) |
|---|---|---|---|---|
| GloVe | Reuters-21578 | 128 | 7769 / 3019 | 0.8523± |