Search papers, labs, and topics across Lattice.
University of California San Diego;
1
0
3
Training language models with just 100 million tokens could redefine efficiency benchmarks in Chinese NLP.
