Fine-tuning a small six-head Nepali BERT on a mixed news and social media corpus improved average intrinsic purity from 0.65 to 0.78 and downstream F1 from 0.74 to 0.81, while still trailing the larger NepaliBERT model.
Enriching word vectors with subword information
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Fine-Tuning Small Embeddings for Elevated Performance
Fine-tuning a small six-head Nepali BERT on a mixed news and social media corpus improved average intrinsic purity from 0.65 to 0.78 and downstream F1 from 0.74 to 0.81, while still trailing the larger NepaliBERT model.