{"id":26491,"date":"2025-01-09T19:12:35","date_gmt":"2025-01-09T11:12:35","guid":{"rendered":"https:\/\/www.1ai.net\/?p=26491"},"modified":"2025-01-09T19:12:35","modified_gmt":"2025-01-09T11:12:35","slug":"%e5%be%ae%e8%bd%af%e5%bc%80%e6%ba%90-140-%e4%ba%bf%e5%8f%82%e6%95%b0%e5%b0%8f%e8%af%ad%e8%a8%80-ai%e6%a8%a1%e5%9e%8b-phi-4%ef%bc%8c%e6%80%a7%e8%83%bd%e6%af%94%e8%82%a9-gpt-4o-mini","status":"publish","type":"post","link":"https:\/\/www.1ai.net\/en\/26491.html","title":{"rendered":"Microsoft open-sources Phi-4, a 14 billion parameter small language AI model that outperforms GPT-4o Mini"},"content":{"rendered":"<p>January 9, 2024 - After the December 12, 2024 release of the<a href=\"https:\/\/www.1ai.net\/en\/tag\/%e5%be%ae%e8%bd%af\" title=\"[View articles tagged with [Microsoft]]\" target=\"_blank\" >Microsoft<\/a>Yesterday (Jan. 8) on the Hugging Face platform.<strong><a href=\"https:\/\/www.1ai.net\/en\/tag\/%e5%bc%80%e6%ba%90\" title=\"[View articles tagged with [open source]]\" target=\"_blank\" >Open Source<\/a>The small language model, Phi-4, is available for interested developers and testers to download, fine-tune, and deploy the AI model.<\/strong><\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-26492\" title=\"e56c33b1j00sptj44006bd000qk00scp\" src=\"https:\/\/www.1ai.net\/wp-content\/uploads\/2025\/01\/e56c33b1j00sptj44006bd000qk00scp.jpg\" alt=\"e56c33b1j00sptj44006bd000qk00scp\" width=\"956\" height=\"1020\" \/><\/p>\n<p>Note: With only 14 billion parameters, the model outperforms even the much larger Llama 3.3 70B (nearly five times as many as Phi-4) and OpenAI's GPT-4o Mini in several benchmarks; Phi-4 even outperforms Gemini 1.5 Pro and OpenAI's GPT-4o in math competition problems.<\/p>\n<p>The main reason behind Phi-4's strong performance is due to the fact that Microsoft has selected high-quality datasets for training, but has not yet optimized the inference, which can be further optimized and quantified by developers in the future to allow it to run locally on devices such as PCs and laptops.<\/p>","protected":false},"excerpt":{"rendered":"<p>On 9 January, following its release on 12 December 2024, Microsoft yesterday (8 January) on the Hugging Face platform, the Open Source Small Language Model Phi-4 can be downloaded, fine-tuned and deployed by interested developers and tasters. Note: The model parameter was only 14 billion, but performed excellently in multiple benchmark tests, even surpassing the larger Llama 3.3 70B (almost five times more than Phi-4) and OpenAI GPT-4o Mini; in mathematics competitions, Phi-4 even outperformed Gemini 1.5 Pro and OpenAI GPT-4o. Phi<\/p>","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[146],"tags":[167,219,280],"collection":[],"class_list":["post-26491","post","type-post","status-publish","format-standard","hentry","category-news","tag-ai","tag-219","tag-280"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/26491","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/comments?post=26491"}],"version-history":[{"count":0,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/26491\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/media?parent=26491"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/categories?post=26491"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/tags?post=26491"},{"taxonomy":"collection","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/collection?post=26491"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}