chunked/nltk_chunking/_model_summary/chunk_46.txt

However, most Transformer models continued to trend towards more parameters, leading to new models focused on improving training efficiency.