8 Trillion: Liang Wenfeng Also Starts Competing in the 'Ultra-Large' Arena
11 hour ago / Read about 0 minute
Author:小编   

DeepSeek has joined the parameter competition. According to a report by The Information, it is currently training a new model with approximately 2 trillion parameters. Recently, Liang Wenfeng revealed at an investor conference that the company plans to push the model parameters to 8 trillion—a figure that was hard to imagine just a year ago.