Search papers, labs, and topics across Lattice.
1
0
2
A 135M parameter model for Bangla outperforms larger counterparts, proving efficiency can rival scale in low-resource NLP tasks.
