社交巨头Meta近期宣布，对其开源代码生成AI模型Code Llama 70B进行了重大更新。据悉，新版Code Llama 70B是目前“最大、性能最优”的模型，能够处理更多的查询。在HumanEval基准测试中，其准确率达到了53%，超越了GPT-3.5的48.1%，但与OpenAI的GPT-4 67%的准确率相比仍有差距。
Code Llama工具去年8月发布，并可供全球开发者免费使用。此次更新的Code Llama 70B模型，不仅在性能上有所提升，还进一步降低了开发门槛，有望为人工智能领域带来更多的创新应用。
此次Meta发布的Code Llama 70B模型，是Meta在人工智能领域的重要进展之一。人工智能作为当前科技领域的热点，Meta的这一举动，无疑进一步加大了其在人工智能领域的竞争力度。
英文标题：Meta Unveils Open Source Large Model Code Llama 70B With Higher Accuracy Than GPT-3.5
英文关键词：Meta, Code Llama, Artificial Intelligence
英文新闻内容：
Social media giant Meta recently announced a significant update to its open-source code-generating AI model, Code Llama 70B. The updated version of Code Llama 70B is currently the “largest and most performant” model, capable of handling more queries. In the HumanEval benchmark test, its accuracy reached 53%, surpassing the 48.1% accuracy of GPT-3.5, but it still lags behind the 67% accuracy of OpenAI’s GPT-4.
Released in August last year, the Code Llama tool is available for free to all developers worldwide. This update to the Code Llama 70B model not only improves performance but also further lowers the entry barrier for developers, promising more innovative applications in the field of artificial intelligence.
Meta’s release of the Code Llama 70B model is another significant step in the company’s efforts in the field of artificial intelligence. As artificial intelligence remains a hotspot in the tech industry, Meta’s move further intensifies its competition in the AI sector.
【来源】https://www.ithome.com/0/747/917.htm
Views: 1