在 Hugging Face 上部署語音對語音 (Speech-to-Speech) 模型★ 75
Hugging Face Blog645 days agoTutorialAs real-time voice interaction technologies like GPT-4o become more widespread, the open-source community is also actively developing speech-to-speech (S2S)…
修正梯度累積：解決 LLM 微調中常被忽視的數學偏差★ 85
Hugging Face Blog651 days agoTutorial### The Mathematical Flaw in Traditional Gradient Accumulation Gradient accumulation is an extremely common technique in deep learning. When VRAM is limited…
介紹 AMD 第 5 代 EPYC™ 處理器：Hugging Face 攜手 AMD 釋放 CPU 的 AI 推論潛能★ 75
Hugging Face Blog657 days agoReleaseAMD has officially launched its 5th-generation EPYC processor, codenamed "Turin," and Hugging Face has promptly published a blog post detailing the deep…
透過動態投機（Dynamic Speculation）加速 Hugging Face 輔助生成（Assisted Generation）★ 75
Hugging Face Blog659 days agoReleaseHugging Face has published a technical blog post on "Dynamic Speculation," aimed at optimizing the inference speed of large language models (LLMs)…
Hugging Face 推出 Open FinLLM 排行榜：專為金融領域大語言模型打造的開源評測基準★ 75
Hugging Face Blog663 days agoReleaseHugging Face has officially launched the "Open FinLLM Leaderboard" — a new platform dedicated to evaluating and tracking the performance of large language…
中國 AI 全球擴張簡析：開源模型如何席捲 Hugging Face 與全球市場★ 75
Hugging Face Blog664 days agoCommentaryThis article from the Hugging Face blog takes an in-depth look at how China's artificial intelligence forces have successfully gone global in recent years…
🇨🇿 BenCzechMark：你的 LLM 能聽懂捷克語嗎？全新捷克語基準測試發布
Hugging Face Blog666 days agoReleaseThe Hugging Face team and its collaborators have jointly launched a new benchmark called "BenCzechMark," designed to evaluate the understanding and generation…
FineVideo 幕後秘辛：Hugging Face 如何打造高品質開源影片資料集★ 75
Hugging Face Blog674 days agoReleaseWith the explosion of video generation and understanding models such as Sora and Gen-3, high-quality video training data has become a key battleground for…
使用合成訓練數據提升 Flux 微調效果的實用指南★ 75
Replicate Blog677 days agoTutorialFine-tuning the open-source image generation model Flux.1 has become a highly sought-after capability for creators and developers alike. However, relying…
使用 Optimum-Intel 與 OpenVINO GenAI 進行模型優化與部署★ 75
Hugging Face Blog677 days agoTutorialThis article provides a detailed look at how to use Hugging Face's `optimum-intel` library and Intel's OpenVINO GenAI toolkit to optimize and deploy generative…
微調 LLM 至 1.58-bit：讓極限模型量化變得簡單★ 85
Hugging Face Blog679 days agoTutorialThe deployment of large language models (LLMs) has long faced a dual bottleneck of VRAM capacity and memory bandwidth. Microsoft previously introduced the…
HuggingChat 推出「社群工具 (Community Tools)」：賦予開源 AI 助理強大的 Agent 擴充能力★ 75
Hugging Face Blog681 days agoReleaseHugging Face has officially introduced the "Community Tools" feature to its open-source chat platform, HuggingChat. This major update injects powerful Agent…
Hugging Face 釋出 Accelerate 1.0.0 正式版：邁向分散式訓練與大模型推理的全新里程碑★ 80
Hugging Face Blog684 days agoReleaseHugging Face has officially released version 1.0.0 of its core open-source library, Accelerate. This is a milestone update, signifying that since the project's…
透過 API 微調 FLUX.1：Replicate 推出程式化微調服務★ 75
Replicate Blog688 days agoNew ToolReplicate has officially announced support for fine-tuning the popular open-source image generation model FLUX.1 [dev] via its HTTP API. FLUX.1, developed by…
Hugging Face 官方點名：5 個被低估的開源生態系「隱藏神兵」★ 75
Hugging Face Blog706 days agoTutorialHugging Face, as the world's largest open-source AI community, has developed many powerful tools beyond its well-known Model Hub that often go unnoticed by…
透過 Flash Attention 2 的 Packing 技術提升 Hugging Face 訓練效率★ 80
Hugging Face Blog707 days agoTutorialWhen fine-tuning or pre-training large language models (LLMs), the sequence lengths of input data are typically uneven. The traditional approach is to use…
Replicate 推出 FLUX.1 微調功能：用一行程式碼即可訓練專屬的 LoRA 圖像生成模型★ 80
Replicate Blog713 days agoReleaseReplicate, the well-known cloud AI execution platform, has announced official fine-tuning support for FLUX.1, the image generation model that has taken the…
GGML 基礎入門介紹：讓大語言模型在消費級硬體上高效運行的關鍵技術★ 80
Hugging Face Blog715 days agoTutorialGGML is a lightweight, zero-dependency C/C++ tensor library developed by Georgi Gerganov. It was originally designed to enable efficient local inference of the…
如何構建具備擴展性的 AI 應用程式：Vercel 的架構與最佳實踐★ 75
Vercel Changelog715 days agoTutorialAs generative AI applications proliferate, developers face challenges fundamentally different from traditional web development: high model invocation costs…
歡迎 Falcon Mamba：首款強大的無注意力機制（Attention-Free）7B 語言模型★ 85
Hugging Face Blog716 days agoReleaseThe Technology Innovation Institute (TII) of Abu Dhabi has officially released Falcon Mamba 7B, a significant milestone in the evolution of AI architectures…
Hugging Face 統一工具調用（Tool Use）標準：簡化開源 LLM Agent 開發流程★ 85
Hugging Face Blog716 days agoRelease### Background and Pain Points In AI agent development, "tool use" (also known as function calling) is the core capability that allows large language models…
Replicate Intelligence #10：Flux 模型最新進展、Minecraft AI 機器人與 Streamlit 實作指南
Replicate Blog719 days agoReleaseThis edition of Replicate Intelligence #10 brings developers and AI enthusiasts several useful technical updates and community projects: 1. **Latest…
Hugging Face 2024 年安全功能亮點：全面提升開源 AI 生態系的防禦力★ 75
Hugging Face Blog722 days agoReleaseAs AI models have proliferated, AI supply chain security has become a top priority that developers and enterprises can no longer afford to overlook. Hugging…
Hugging Face 推出適用於文件圖像的 TextImage 數據增強技術 (TextImage Augmentation)★ 75
Hugging Face Blog722 days agoNew Tool### Solving Real-World Document AI Pain Points In the fields of Document AI and OCR (Optical Character Recognition), datasets used in academic research or…
Replicate Intelligence #9：開源影像模型新巔峰 FLUX.1、影片物件分割 SAM 2 與全新 Python 網頁框架 FastHTML★ 85
Replicate Blog726 days agoReleaseThis edition of Replicate Intelligence highlights three of the most noteworthy open-source and developer tool developments from early August 2024: 1. **The new…
FLUX.1 首發實測：開源圖像生成的新巔峰與第一手體驗★ 85
Replicate Blog726 days agoReleaseFLUX.1 is a new family of text-to-image models developed by Black Forest Labs — a startup founded by the original core research team behind Stable Diffusion…
Google 發布 Gemma 2 2B、安全分類器 ShieldGemma 與可解釋性工具 Gemma Scope★ 85
Hugging Face Blog728 days agoReleaseGoogle released a major update to the Gemma 2 family in late July 2024, comprising three core components: 1. **Gemma 2 2B**: A lightweight model with just 2.6B…
使用 Quanto 與 Diffusers 打造記憶體高效的 Diffusion Transformers (DiT)★ 80
Hugging Face Blog729 days agoRelease### Background and Challenges As generative AI technology evolves, image and video generation models are increasingly transitioning from traditional UNet…
Hugging Face 聯手 NVIDIA NIM 推出無伺服器推論服務 (Serverless Inference)★ 82
Hugging Face Blog730 days agoReleaseHugging Face and NVIDIA announced a major partnership in late July 2024, officially launching a serverless inference service powered by NVIDIA NIM (NVIDIA…
Hugging Face 推出 Docmatix：用於文件視覺問答（DocVQA）的超大型開源數據集★ 75
Hugging Face Blog741 days agoReleaseThe Hugging Face official blog has announced the release of a new, massive dataset called "Docmatix," specifically designed for training and fine-tuning…