Robin Lee's latest voice: the open source model will fall behind
芊芊551
发表于 2024-4-16 12:06:37
244
0
0
On April 16, Robin Lee, the founder, chairman and CEO of Baidu, said at the Create 2024 Baidu AI Developers Conference that the open source model will become more and more backward.
Based on the basic model Wenxin 4.0, Robin Lee said that users can tailor smaller models suitable for various scenarios according to needs, taking into account effects, corresponding speed, reasoning costs and other considerations. Small models are significantly better than models directly called out from open source models under the same size; Under the same effect, the cost is significantly lower.
Robin Lee interacted with Xiaodu on the spot in real time, demonstrating that Xiaodu performed different tasks by combining multiple models. For example, using the small model ERNIE Tiny to perform model routing work, while the best performing Wenxin 4.0 is used to perform complex requirements such as scheduling. Compared to the flagship version that uses the entire Wenxin model, Xiaodu can achieve a response speed increase of 2 times and a cost reduction of 99%.
CandyLake.com is an information publishing platform and only provides information storage space services.
Disclaimer: The views expressed in this article are those of the author only, this article does not represent the position of CandyLake.com, and does not constitute advice, please treat with caution.
Disclaimer: The views expressed in this article are those of the author only, this article does not represent the position of CandyLake.com, and does not constitute advice, please treat with caution.
You may like
- The delivery fee pricing has been lowered to 6 yuan, and McDonald's has adjusted the McDonald's delivery fee model
- Ideal Automobile implements a limited time zero interest policy for all models for the first time
- OpenAI launches full health version of the o1 big model and $200 per month ChatGPT Pro
- OpenAI has Rocket again! Officially launched Sora, an AI video generation model
- Google releases its most powerful model to attack OpenAI, shifting focus to AI agents
- Challenge OpenAI, Google's new move! Significantly updated generative AI, launching video model VEO 2 and the latest version Imagen3
- Is it increasingly difficult to distinguish between truth and falsehood? Google launches new generation video generation model Veo 2
- Microsoft is reportedly committed to adding non OpenAI models to its 365 Copilot product
- The most expensive and cheapest models will be launched on the same day! Li Bin: NIO will double its efforts to become one of the top ten global car companies
- How will Google respond under stricter regulation in a more competitive track? CEO: Focus on Gemini model next year