1
谷歌开源端侧模型家族 Gemma 总下载量破 10 亿次，派生超 10 万个变体
谷歌开源端侧模型家族Gemma下载量破10亿，衍生10万变体，生态爆发力惊人
IT之家 8 月 21 日消息，Google（谷歌）当地时间 20 日宣布，其开源端侧模型家族 Gemma 的总下载量已突破 10 亿次 ；开发者在过去两年间 发布了超过 10 万个 Gemma 模型变体 ，构成了一个被谷歌称为 "Gemmaverse" 的创新生态系统。 谷歌列举了多个 Gemma…
2
From Manuals to Maintenance: Fine-Tuning MedGemma for Multi-Modal Imaging System Support in Low-Resource Settings
用微调MedGemma助力低资源地区成像系统维护，医疗AI落地新思路。
arXiv:2608.08896v1 Announce Type: new Abstract: Imaging device downtime is a major barrier to healthcare delivery in low- and middle-income countries …
3
Investigating first-language bias in LLM-based automated essay scoring: A cross-prompt evaluation of an open-weight AI-model on TOEFL essays
研究揭示LLM自动评分中存在的第一语言偏见，基于LoRA微调的开源模型在托福作文跨提示评估中的表现实验。
arXiv:2607.14605v1 Announce Type: new Abstract: This study examines the cross-prompt generalization and first-language (L1) scoring effects of a LoRA-…
4
ZML/LLMD alpha – cross platform LLM server
跨平台LLM推理服务器，原生支持Qwen、Gemma、Mistral、LLaMa等主流架构，一键Docker部署并集成HuggingFace与S3/GCS存储
Article URL: https://zml.ai/posts/llmd/ Comments URL: https://news.ycombinator.com/item?id=48829412 Points: 3 # Comments: 2
5
Master Local Fine-Tuning with "gemma-trainer"
手把手教你用 "gemma-trainer" 在本地微调 Gemma 4 E2B 模型，快速上手大模型定制！
Remember back in May when I introduced the gemma-skills repository? It's been rewarding to see how many of you have used my previous post to streamlin…
6
Show HN: Gemma 3 inference in pure C++ with Metal acceleration
纯C++实现Gemma 3推理，Metal加速，让本地大模型运行更轻快。
Article URL: https://github.com/ybubnov/metalchat Comments URL: https://news.ycombinator.com/item?id=48786298 Points: 3 # Comments: 1
7
Ornith-1.0: Self-Scaffolding LLMs for Agentic Coding
看 Ornith-1.0 用“自脚手架”思路玩转智能体编码，底层可选 Apache 2.0 的 Gemma 4，许可省心又开源友好。
Ornith-1.0: Self-Scaffolding LLMs for Agentic Coding This is an interesting new open weights (MIT licensed) model, the first model release from DeepRe…
8
Where Do Models Find Happiness? Emotion Vectors in Open-Source LLMs
从Claude到开源模型：研究发现LLM内部存在类似人类心理结构的情绪向量，并揭示其因果影响行为。
arXiv:2606.26987v1 Announce Type: cross Abstract: Recent work identified emotion vectors in Claude Sonnet 4.5, which are internal representations that…
9
LLM APIs with built-in chatbot in 1 line of code
一行命令部署LLM API，Flama 2.0自带聊天界面，轻松搞定Gemma等模型服务。
Article URL: https://flama.dev/blog/serving_llms_with_flama_cli/ Comments URL: https://news.ycombinator.com/item?id=48671740 Points: 17 # Comments: 0
10
Lean Zig by building an LLM from scratch
从乘法函数到生成文本，用 Zig 手搓 GPT-2 与 Gemma，硬核实战带你彻底吃透 LLM 底层。
Article URL: https://hamanlp.org/small-kernels.html Comments URL: https://news.ycombinator.com/item?id=48659158 Points: 1 # Comments: 1
11
Evaluation of Medical Vision Language Models HuluMed and MedGemma, and general purpose chatbots Gemma 3, ChatGPT Plus, and Claude Pro on real previously unseen wound images
真实伤口图像上评估医学视觉语言模型与通用聊天机器人，HuluMed、MedGemma、Gemma 3、ChatGPT Plus、Claude Pro谁更强？
arXiv:2606.20723v1 Announce Type: new Abstract: Chronic wound assessment remains a clinically challenging task that requires accurate interpretation o…
12
EP219: 12 Open-source LLMs
盘点12款主流开源大模型，覆盖Meta、阿里、谷歌等头部玩家，一文掌握选型要点。
Twelve models worth knowing in 2026, each with one standout strength.
13
Serving any LLM using a single command line with Flama
一行命令部署任何LLM，Flama让Gemma 4等模型在Apple Silicon上极速运行
Flama 2.0 brings first-class support for generative AI: downloading, packaging, and serving large language models (LLMs) is now as simple as running a…
14
Solstice Eternal – AI Powered Adventure Game with Gemma 4
Gemma 4赋能《Solstice Eternal》，用AI生成创意与强化叙事，打造沉浸式冒险体验。
This is a submission for the Gemma 4 Challenge: Build with Gemma 4 What I Built Solstice Eternal – AI Adventure Game Solstice Eternal is an interactiv…
15
Running local models is good now
本地模型性能跃升：M2 Mac实测达到前沿模型75%效果，低成本也能玩转Agentic Coding
Article URL: https://vickiboykis.com/2026/06/15/running-local-models-is-good-now/ Comments URL: https://news.ycombinator.com/item?id=48555993 Points: …
16
Can Editing 1 Neuron Fix Repetition Loops in LLMs?
编辑单个神经元，就能修复大模型在长枚举任务中的重复循环崩溃？这项研究揭示了 Gemma 4 模型的深层问题与解法。
arXiv:2606.13705v1 Announce Type: cross Abstract: Yes. Can it cure doom loops? Probably not. The Gemma 4 instruction-tuned models share a reproducible…
17
谷歌推出 DiffusionGemma 文本扩散模型：本地 AI 推理速度提升 4 倍
基于文本扩散的开放AI模型，本地推理速度提升4倍，高效处理文本生成任务
IT之家 6 月 11 日消息，谷歌今天（6 月 11 日）发布公告，宣布推出 DiffusionGemma，是基于文本扩散机制的开放 AI 模型， 相比较自回归模型在本地推理速度上提升了 4 倍。 IT之家注：自回归模型（Autoregressive Model）是当前主流的大语言模型架构（如 G…
18
Gemma 4 QAT on 10GB Laptop: Local AI with 6.7GB VRAM
用Ollama配合QAT量化，10GB显存笔记本也能跑12B的Gemma 4大模型，内存仅需6.7GB。
This stack uses Ollama with Gemma 4 QAT to run a 12B model on a 10GB VRAM laptop GPU. The latest Gemma 4 QAT checkpoints reduce memory usage and enabl…
19
Bonsai: Human->LLM->Web with LLM interface using Gemma4 12B locally on Windows
本地运行Gemma4实现人-LLM-Web交互，Bonsai项目让AI操作浏览器门槛大降
Article URL: https://drive.google.com/drive/folders/1QsdOvsBKIavXDhwikouzQMOc8H_94bQd Comments URL: https://news.ycombinator.com/item?id=48461245 Poin…
20
Gemma 4 E4B as a primary local LLM (replaced Qwen)
从Qwen换到Gemma 4 E4B：本地大模型的新选择，性能与易用性如何？
Article URL: https://digg.com/ai/bfr4bqhh Comments URL: https://news.ycombinator.com/item?id=48437537 Points: 2 # Comments: 0