iFLYTEK announces 2024 big model goals: 1 million ecosystems, 10 million hardware terminals, 100 million software users

On February 6, the theme of the event was “Standing tall and firm, setting the field ablaze with fire”.iFLYTEK The 2023-2024 annual meeting of the company was held at the University of Science and Technology of China. In their internal speeches, Liu Qingfeng, Chairman of iFLYTEK, and Wu Xiaoru, President of iFLYTEK, systematically summarized the development of iFLYTEK in 2023 and clarified the development goals for 2024.

Looking ahead to 2024,Liu Qingfeng proposed iFlytek Spark CognitionLarge ModelFour annual strategic goals:

  • The Spark large model is the leading domestic benchmark against GPT-4;

  • Empowering 10 million smart hardware terminals and over 100 million independent software users;

  • No. 1 in education, medical care, finance, automobiles, central state-owned enterprises, and urban industry markets;

  • Bringing together 1 million large model developers to create the industry's first ecosystem.

Liu Qingfeng said that Spark will continue to maintain rapid upgrades in 2024.It is expected to reach the current best level of GPT-4 Turbo in the first half of the year.

At the iFlytek Spark Cognitive Big Model V3.5 upgrade launch conference on January 30, iFlytek officially announced the iFlytek Spark Cognitive Big Model V3.5.

iFLYTEK announces 2024 big model goals: 1 million ecosystems, 10 million hardware terminals, 100 million software users

It is reported that Spark Cognitive Big Model V3.5 brings improvements in seven major capabilities, including text generation improvement of 7.3%, language comprehension improvement of 7.6%, knowledge question and answer improvement of 4.7%, logical reasoning improvement of 9.5%, mathematical ability improvement of 9.8%, coding ability improvement of 8.0%, and multimodal ability improvement of 6.6%.

also,Spark Cognitive Model V3.5’s language comprehension and mathematical capabilities are said to have surpassed GPT-4 Turbo, coding ability reaches GPT-4 Turbo 96%, and multimodal understanding reaches GPT-4V 91%.

statement:The content is collected from various media platforms such as public websites. If the included content infringes on your rights, please contact us by email and we will deal with it as soon as possible.
Information

OpenAI forms child safety team to prevent misuse of AI tools

2024-2-9 11:46:07

Information

Google announces that it will rename its AI chat assistant Bard to Gemini and launch an independent Android app

2024-2-9 11:55:29

Search