Modelpublic

AISE-TUDelft/Custom-Activations-BERT-Swish

sourceHugging Faceupdated 2y agoView on Hugging Face

0likes11downloads

Model Card

Basemodel: roBERTa

Configs: Vocab size: 10,000 Hidden size: 512 Max position embeddings: 512 Number of layers: 2 Number of heads: 4 Window size: 256 Intermediate-size: 1024

Results:

—Task: glue Score: 57.89 Confidence Interval: [57.28, 58.48]
—Task: blimp Score: 58.26 Confidence Interval: [57.71, 58.75]