Meta releases Llama 2 model, cooperates with Microsoft and Qualcomm

According to a report by the Financial Associated Press on July 19, after half a year, Meta AI released the latest generation of open source large model Llama 2 on Tuesday. Compared with Llama 1 released in February this year, the token used for training has doubled to 2 trillion. At the same time, Llama 2 has also doubled the most important context length limit for using large models. Llama 2 includes models with 7 billion, 13 billion, and 70 billion parameters.

Meta also announced on Tuesday that it will cooperate with Microsoft's cloud service Azure to launch a cloud service based on the Llama 2 model to developers around the world. In addition, Meta also announced with Qualcomm that Llama 2 will be able to run on Qualcomm chips, breaking the monopoly of Nvidia and AMD processors on the AI industry in the market.