This is electra base tuned with optimal lr, lr scheduler and weight decay on dnd-style-intents dataset.

parametrs value
learning rate 6.6e-5
lr scheduler cosine with restarts
weight decay 0

Model has next metrics on test data from dataset

metric value
accuracy 0.978
Macro F1 0.976
Micro F1 0.978
Downloads last month
11
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Dataset used to train neurae/electra-dnd-intents