Google Introduces PaLM 2, The Most Advanced Language Model

Google included many artificial intelligence announcements at its I/O event today. PaLM 2, the newest language model with support for over 100 languages, was one of them.
 Google Introduces PaLM 2, The Most Advanced Language Model
READING NOW Google Introduces PaLM 2, The Most Advanced Language Model

Google held its highly anticipated developer conference I/O. Introducing many new products and services, the technology giant also gave a great place to artificial intelligence, which is one of the most curious subjects, at the event.

The company introduced its newest language model, PaLM2, at its presentation. Let’s take a look at the details Google shared about the multilingual, coding and reasoning model.

PaLM 2 already powers 25 services

It was stated that PaLM 2, which will be presented in four dimensions named Gecko, Otter, Bison and Unicorn, has been opened for preview. Sundar Pichai adds that Gecko is light enough to run on mobile devices, but fast enough even when offline. It is added that with this versatility, PaLM 2 will help more people.

PaLM2 powers Bard, Google’s ChatGPT rival chatbot, and serves as the base model for the company’s AI moves. Google added that the language model is already included in 25 products.

Google says its most advanced language model is good at things like logic, math. For example, the model; It has features such as writing code, correcting errors. It adds that it has support for more than 20 programming languages, including Google JavaScript and Python.

Like other major language models that take a lot of time and resources to create, PaLM 2 is not a single product, but rather a product family with different versions to be used in consumer and corporate environments.

System; Available in four sizes, from smallest to largest, named Gecko, Otter, Bison and Unicorn. Google says Gecko, the lightest version of PaLM 2, is small enough to run on mobile phones.

PaLM 2 supports more than 100 languages

In addition, the model is said to be trained on a structure containing more than 100 languages. In this way, the model becomes more advanced in understanding and translating texts such as idioms and poems.

Comments
Leave a Comment

Details
169 read
okunma44959