China's technology company Baidu releases Wenxin large model 4.5 series with complete open-source pre-trained weights and inference codes

China's technology company Baidu releases Wenxin large model 4.5 series with complete open-source pre-trained weights and inference codes

Baidu has made a significant stride in the artificial intelligence landscape by unveiling the Wenxin Large Model 4.5 series. The company has fully open-sourced the model, which includes ten new models with varying parameter configurations. These models have been trained and inferred using the PaddlePaddle framework, achieving a FLOPs utilization rate of 47%. The models have demonstrated exceptional performance in text multimodal benchmark tests and come with a comprehensive one-stop usage guide and tools, making it easier for developers to fine-tune and deploy the models [1].

The Wenxin Large Model 4.5 series is available on platforms such as Hugging Face and GitHub, allowing developers worldwide to access and utilize the models. This open-source initiative is a testament to Baidu's commitment to fostering innovation and collaboration within the AI community. By providing complete pre-trained weights and inference codes, Baidu is enabling developers to integrate the models into their applications seamlessly [1].

The release of the Wenxin Large Model 4.5 series is part of Baidu's broader strategy to advance AI technology and applications. The company's efforts to promote open-source models are aligned with the global trend of democratizing AI development. By making these powerful tools accessible, Baidu is empowering developers to push the boundaries of what is possible in AI, potentially leading to new products and services that drive economic growth and innovation [1].

References:
[1] https://www.aibase.com/news/19353