{"id":9093,"date":"2026-06-14T02:22:26","date_gmt":"2026-06-13T18:22:26","guid":{"rendered":"https:\/\/infernews.com\/blog\/?p=9093"},"modified":"2026-06-14T02:22:28","modified_gmt":"2026-06-13T18:22:28","slug":"gemma-4-12b-qat-q4-%e6%af%94%e8%bc%83","status":"publish","type":"post","link":"https:\/\/infernews.com\/blog\/gemma-4-12b-qat-q4-%e6%af%94%e8%bc%83\/","title":{"rendered":"Gemma 4 12b, QAT, q4 \u6bd4\u8f03"},"content":{"rendered":"<figure class=\"wp-block-embed-youtube wp-block-embed is-type-video is-provider-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"lyte-wrapper\" title=\"Gemma 4 12B QAT vs non-QAT - 16GB VRAM Local LLM setup\" style=\"width:853px;max-width:100%;margin:5px auto;\"><div class=\"lyMe\" id=\"WYL_NeVLMl632OE\" itemprop=\"video\" itemscope itemtype=\"https:\/\/schema.org\/VideoObject\"><div><meta itemprop=\"thumbnailUrl\" content=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FNeVLMl632OE%2Fhqdefault.jpg\" \/><meta itemprop=\"embedURL\" content=\"https:\/\/www.youtube.com\/embed\/NeVLMl632OE\" \/><meta itemprop=\"duration\" content=\"PT13M26S\" \/><meta itemprop=\"uploadDate\" content=\"2026-06-08T15:17:30Z\" \/><\/div><div id=\"lyte_NeVLMl632OE\" data-src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FNeVLMl632OE%2Fhqdefault.jpg\" class=\"pL\"><div class=\"tC\"><div class=\"tT\" itemprop=\"name\">Gemma 4 12B QAT vs non-QAT - 16GB VRAM Local LLM setup<\/div><\/div><div class=\"play\"><\/div><div class=\"ctrl\"><div class=\"Lctrl\"><\/div><div class=\"Rctrl\"><\/div><\/div><\/div><noscript><a href=\"https:\/\/youtu.be\/NeVLMl632OE\" rel=\"nofollow noopener\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FNeVLMl632OE%2F0.jpg\" alt=\"Gemma 4 12B QAT vs non-QAT - 16GB VRAM Local LLM setup\" width=\"853\" height=\"460\" \/><br \/>Watch this video on YouTube<\/a><\/noscript><meta itemprop=\"description\" content=\"In this video I am testing the QAT version of the Gemma 4 12B model from Google and comparing the quality of the QAT from Unsloth (which is q4) vs the regular q4 GGUF from Unsloth. The model is running on a local AI PC I have built with 16GB VRAM and 32GB DDR4 RAM. I run the model through a few tests which are: 1. Adherence 2. Agency 3. Coding 4. Memory If you&#039;re interested in local LLMs, AI and homelabs from the perspective of a software engineer with many years of professional experience working with LLMs in production - feel free to subscribe! Models - \u2022 QAT: https:\/\/huggingface.co\/unsloth\/gemma-4-12B-it-qat-GGUF \u2022 non-QAT: https:\/\/huggingface.co\/unsloth\/gemma-4-12b-it-GGUF GitHub: https:\/\/github.com\/lukesdevlab\/youtube Patreon: https:\/\/www.patreon.com\/cw\/LukesDevLab #localllm #localai #homelab #llamacpp #homelab #gemma4 #quantization #qat Chapters: 0:00 Coming up 0:08 Intro 0:55 Models 1:16 Tests 1:39 System Specs 1:50 Adherence - q4 2:53 Adherence - QAT 3:35 Agency 5:56 Coding - q4 7:55 Coding - QAT 10:55 Memory 12:40 Conclusion\"><\/div><\/div><div class=\"lL\" style=\"max-width:100%;width:853px;margin:5px auto;\"><\/div><figcaption><\/figcaption><\/figure>\n","protected":false},"excerpt":{"rendered":"","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"ai_generated_summary":"","footnotes":""},"categories":[76,178],"tags":[],"class_list":["post-9093","post","type-post","status-publish","format-standard","hentry","category-76","category-google"],"_links":{"self":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/posts\/9093","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/comments?post=9093"}],"version-history":[{"count":1,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/posts\/9093\/revisions"}],"predecessor-version":[{"id":9094,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/posts\/9093\/revisions\/9094"}],"wp:attachment":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/media?parent=9093"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/categories?post=9093"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/tags?post=9093"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}