PulseAugur
实时 23:42:45
English(EN) Baidu's Ernie 5.1 cuts 94 percent of pre-training costs while competing with top models

百度Ernie 5.1预训练成本降低94%

百度发布了其大语言模型的新迭代Ernie 5.1,该模型将预训练成本显著降低了94%。这种效率是通过“Once-For-All”方法实现的,允许从单一训练过程中派生出更小的子模型。尽管与前代模型相比参数数量有所减少,Ernie 5.1仍表现出竞争力,在Search Arena排行榜上名列第四。 AI

影响 为成本效益高的LLM训练树立了新标杆,可能降低模型开发的门槛。

排序理由 百度发布前沿模型并附带系统卡。[lever_c_降级自frontier_release: ic=1 ai=1.0]

在 The Decoder 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →

百度Ernie 5.1预训练成本降低94%

报道来源 [1]

  1. The Decoder TIER_1 English(EN) · Jonathan Kemper ·

    Baidu's Ernie 5.1 cuts 94 percent of pre-training costs while competing with top models

    <p><img alt="The Chinese search and Internet company Baidu presents the latest version of its chatbot Ernie, which is supposed to deliver significantly better results than before. It also gets support for plugins." class="attachment-full size-full wp-post-image" height="816" src=…