Menlopark: Facebook, Instagram and WhatsApp parent Meta has announced a massive language model called Llama to Llama 2.
Metta claims that it has passed several important tests in which it has been proven better and more effective than ChatGPT and Google Bard. Metta has said that this model is very powerful in usability, security and other areas and has also signed a contract with Microsoft.
Meta is expected to provide this facility for free, which is slightly better than OpenAI (ChatGPT) and Google (Bard). Meta offers three different models of Llama.
One of them has been trained on 7 billion, another 13 billion and the third 70 billion, data parameters. Llama to chat service is being started on this basis.
However, efforts are still being made to fix its flaws. Meta and Microsoft have built LamaTo on a large language model that will work under Azure and Windows.
Thanks to this, individuals and institutions will be able to build and experiment with generative AI tools. Both Microsoft and Meta believe in making AI publicly accessible and democratized.
And Llama Two is presented in the same context. It should be clear that Microsoft has previously invested ten billion dollars in OpenAI.
Secondly, now he will also try Lama Tu. Its misuse Adequate measures are being taken to prevent it. Among them is to prevent prejudice, falsehood and consequences based on any discrimination.
