Second only to GPT-4,Inflection AI to release the Inflection-2 model
CTOnews.com, November 23, Inflection AI released Inflection-2, which officially claims to be the second generation of the world's Big language Model (LLM), second only to OpenAI's GPT-4.
Inflection AI indicates that Inflection-2 carries out FP8 mixed precision training on 5000 Nvidia H100 GPU, with ~ 10 ²FLOP.
Inflection-2 and Google's PaLM 2 Large model are in the same training computing category and outperform PaLM 2 in most AI benchmarks such as MMLU, TriviaQA, HellaSwag and GSM8k.
Inflection-2 also takes into account the service efficiency in the design, and can provide real-time application support for chatbots and virtual assistants.
CTOnews.com encloses the Inflection-2 press release here, which can be read in depth by interested users.
