介紹 Chatbot Guardrails Arena：評估大語言模型安全防護網的全新競技場★ 75
Hugging Face Blog816 days agoReleaseAs large language models (LLMs) have been widely adopted across industries, ensuring AI systems remain safe and compliant while preventing harmful outputs has…
GaLore：在消費級硬體上訓練大型語言模型的突破性技術★ 85
Hugging Face Blog817 days agoReleaseAs the parameter counts of large language models (LLMs) have skyrocketed, the hardware requirements for training and fine-tuning these models have risen…
Hugging Face 推出 Red-Teaming 抗性排行榜：評估 LLM 抵禦惡意越獄與對抗性攻擊的能力★ 75
Hugging Face Blog843 days agoRelease### Background: The Shortcomings of Static Safety Evaluations As large language models (LLMs) are widely adopted across industries, AI safety has become an…
開源合成數據：如何幫你省錢、省時並減少碳排放★ 75
Hugging Face Blog850 days agoOpinionThis article takes an in-depth look at the critical role of "synthetic data" in the open-source ecosystem, and explains how it helps enterprises and developers…
Hugging Face 推出 NPHardEval 排行榜：透過計算複雜度與動態更新揭示大型語言模型的推理能力★ 75
Hugging Face Blog864 days agoReleaseHugging Face has announced the launch of the new **NPHardEval** leaderboard — a benchmark specifically designed to evaluate the reasoning capabilities of large…
Hugging Face TGI (Text Generation Inference) 正式支援 AWS Inferentia2 晶片★ 75
Hugging Face Blog865 days agoReleaseHugging Face has partnered with AWS to officially bring its widely popular open-source LLM inference optimization framework, Text Generation Inference (TGI)…
Hugging Face 推出「企業情境排行榜」：專為真實世界應用設計的 LLM 評測基準★ 75
Hugging Face Blog866 days agoReleaseHugging Face has partnered with Patronus AI — a startup focused on LLM evaluation and defense — to officially launch the **Enterprise Scenarios Leaderboard**…
使用 API 運行 Code Llama 70B：Replicate 雲端部署教學
Replicate Blog867 days agoTutorialMeta's Code Llama 70B is a powerful open-source large language model designed specifically for code generation and understanding. However, running such a…
Hugging Face 推出「幻覺排行榜」，開源量化評估大型語言模型的幻覺率★ 75
Hugging Face Blog868 days agoReleaseWhile large language models (LLMs) have demonstrated remarkable generative capabilities across many domains, "hallucination" — where a model confidently…
Hugging Face 推出 AI Secure LLM 安全排行榜：基於 DecodingTrust 框架深度評估大模型信任度★ 75
Hugging Face Blog871 days agoRelease### Introduction: Capability Is Not Safety — A New Benchmark for LLM Safety Evaluation As large language models (LLMs) are adopted more deeply across…
使用開源 LLM 作為 LangChain Agent：Hugging Face 實作指南★ 80
Hugging Face Blog873 days agoTutorialThis official Hugging Face blog post provides a detailed guide on how to use open-source large language models (LLMs) as intelligent agents within LangChain…
使用 Unsloth 與 🤗 TRL 讓 LLM 微調速度提升 2 倍★ 80
Hugging Face Blog887 days agoReleaseHugging Face's official blog announced a partnership with the Unsloth team to integrate Unsloth's efficient fine-tuning technology directly into Hugging Face's…
2023 年：開源大語言模型（Open LLMs）爆發之年★ 75
Hugging Face Blog910 days agoCommentaryLooking back on 2023, the most notable trend in the AI landscape was the explosive growth of open-source large language models (Open LLMs). In this annual…
Optimum-NVIDIA：只需一行程式碼，即可解鎖極速 LLM 推理★ 80
Hugging Face Blog923 days agoReleaseHugging Face announced the launch of a new open-source library called "Optimum-NVIDIA," the result of a deep collaboration with NVIDIA, aimed at seamlessly…
Replicate 完成由 a16z 領投的 4,000 萬美元 B 輪融資，加速企業基於開源 AI 進行開發★ 75
Replicate Blog923 days agoBusinessAI infrastructure startup Replicate announced the successful completion of a $40 million Series B funding round. This round was led by prominent Silicon Valley…
告別冷啟動：Hugging Face 如何將 LoRA 推論速度提升 300%★ 85
Hugging Face Blog923 days agoReleaseIn real-world generative AI applications, fine-tuning for specific tasks or clients is a common requirement. However, deploying a full base model for every…
讓你的 Llama 生成速度飛起來：使用 AWS Inferentia2 進行加速★ 72
Hugging Face Blog951 days agoTutorialAs large language models (LLMs) such as Llama 2 become more widely adopted, achieving efficient and cost-effective inference in production environments has…
比較大語言模型性能：深入探討使用 LoRA 微調 RoBERTa、Llama 2 與 Mistral 進行災難推特分析★ 75
Hugging Face Blog951 days agoTutorialThis Hugging Face blog post takes an in-depth look at how to use LoRA (Low-Rank Adaptation) to fine-tune three models of different architectures and scales for…
使用 Llama 2 與 Grammar 進行結構化資訊萃取 (Jet-setting with Llama 2 + Grammars)★ 75
Replicate Blog980 days agoTutorialIn practical natural language processing (NLP) applications, converting unstructured text (such as emails or conversation logs) into structured data (such as…
Chat Templates：終結默默蠶食模型效能的隱形殺手★ 85
Hugging Face Blog986 days agoReleaseAs large language models (LLMs) shift toward conversational (Chat/Instruct) applications, correctly formatting and feeding a user's conversation history —…
使用 Inference API 部署 AI Comic Factory（AI 漫畫工廠）
Hugging Face Blog987 days agoTutorial"AI Comic Factory" is one of the most popular open-source AI applications on the Hugging Face platform. It allows users to simply enter a brief plot…
非工程師指南：如何訓練專屬的 LLaMA 2 聊天機器人
Hugging Face Blog991 days agoTutorialThis official guide from Hugging Face is designed for readers without a technical background. It provides a detailed walkthrough of how to use Hugging Face's…
Llama 2 在 Amazon SageMaker 上的部署效能基準測試
Hugging Face Blog993 days agoTutorialThis Hugging Face blog post presents detailed performance benchmarks for deploying Meta's open-source large language models — Llama 2 (covering 7B, 13B, and…
Hugging Face 為 PRO 訂閱者推出專屬推理服務：更高速率、支援大型開源模型★ 70
Hugging Face Blog997 days agoReleaseThe Hugging Face official blog has announced a new "Inference for PROs" upgraded service for PRO subscribers (at $9 per month). This service is designed to…
使用 PyTorch FSDP 高效微調 Llama 2 70B：解決 CPU 記憶體不足的實務指南★ 72
Hugging Face Blog1,006 days agoTutorialWhen fine-tuning massively large open-source models like Llama 2 70B — with its 70 billion parameters — developers frequently encounter a bottleneck that goes…
Replicate 大幅優化微調模型冷啟動時間，現在只需不到一秒即可完成載入★ 75
Replicate Blog1,013 days agoReleaseAI cloud hosting platform Replicate has announced a major technical breakthrough for fine-tuned models: the "cold boot" time for fine-tuned models has been…
Meta 推出 Code Llama：基於 Llama 2 的開源程式碼生成模型，支援 100k 上下文★ 85
Hugging Face Blog1,025 days agoReleaseMeta has officially launched Code Llama, a family of state-of-the-art open-source code generation models fine-tuned on Llama 2. Code Llama achieves leading…
使用 AutoGPTQ 與 transformers 讓大型語言模型更輕量化★ 85
Hugging Face Blog1,027 days agoReleaseThis Hugging Face official blog post introduces a major update that integrates AutoGPTQ into the `transformers` and `optimum` libraries. GPTQ (Generalized…
Hugging Face 推出 IDEFICS：開源重現 SOTA 多模態視覺語言模型 Flamingo★ 78
Hugging Face Blog1,028 days agoReleaseHugging Face has officially launched IDEFICS (Image-supervised Decoder-Encoder-Few-shot-In-Context-Shorthand), an open-source multimodal vision-language model…
Llama 2 提示詞指南：掌握官方 Prompt 格式與特殊標記
Replicate Blog1,036 days agoTutorialThis is a practical prompting guide for Llama 2, written by the Replicate team. Unlike OpenAI's GPT series models, the open-source model Llama 2 has extremely…