AI要約 GoogleはエッジデバイスでのエージェントAI向けに設計されたGemma 4を発表。最先端のエージェントスキルをオンデバイスで実現し、ローカル環境で高度なツール利用や自律的タスク処理を可能にする。
EN Google announced Gemma 4, bringing state-of-the-art agentic capabilities to edge devices, enabling advanced tool use and autonomous task handling directly on-device.
AI要約 NVIDIAがHuggingFaceブログで、合成ペルソナデータセット「Nemotron-Personas-Korea」を活用して韓国の実際の人口統計に基づくAIエージェントを構築する方法を紹介。地域・年齢・職業分布を反映したペルソナで、韓国向けLLMアプリの現実的な評価やデータ生成を可能にする。
EN NVIDIA's Hugging Face blog post shows how to build Korean AI agents grounded in real demographics using the Nemotron-Personas-Korea synthetic dataset, enabling realistic evaluation and data generation for Korea-targeted LLM applications.
EN Hermes Agent ollama launch hermes Hermes learns with you, automatically creating skills to better serve your workflows. Great for research and engineering tasks. What's Changed Gemma 4 on MLX. Added s
media EN Ecom-RLVE: Adaptive Verifiable Environments for E-Commerce Conversational Agents
EN The PR you would have opened yourself
EN Training and Finetuning Multimodal Embedding & Reranker Models with Sentence Transformers
EN Inside VAKRA: Reasoning, Tool Use, and Failure Modes of Agents
EN Meet HoloTab by HCompany. Your AI browser companion.
EN What's Changed Fix quality of gemma:e2b and gemma:e4b when thinking is disabled ROCm: Update to ROCm 7.2.1 on Linux by @saman-amd in #15483 Full Changelog : v0.20.6...v0.20.7
media EN gemma4: implement Gemma 4 model for MLX (text-only runtime) gemma4: two MoE + SWA prefill perf fixes Two performance optimizations in the gemma4 forward pass Memoize the sliding-window prefill mask ac
media EN What's Changed Gemma 4 tool calling ability is improved and updated to use Google's latest post-launch fixes Parallel tool calling improved for streaming responses Hermes agent Ollama integration guid
media EN Waypoint-1.5: Higher-Fidelity Interactive Worlds for Everyday GPUs
EN Multimodal Embedding & Reranker Models with Sentence Transformers
EN Safetensors is Joining the PyTorch Foundation
EN Open models are driving a new wave of on-device AI, extending innovation beyond the cloud to everyday devices. As these models advance, their value increasingly depends on access to local, real-time c
EN Gemma 4: Our most intelligent open models to date, purpose-built for advanced reasoning and agentic workflows.
media EN Welcome Gemma 4: Frontier multimodal intelligence on device
EN Any Custom Frontend with Gradio's Backend
EN Granite 4.0 3B Vision: Compact Multimodal Intelligence for Enterprise Documents
EN Training mRNA Language Models Across 25 Species for $165
EN TRL v1.0: Post-Training Library Built to Move with the Field
EN What's Changed fix: remove Llama 3.1 405B from Groq provider by @RomneyDa in #11804 fix: add keepAlive to YAML completion options schema by @RomneyDa in #11801 fix(llama.cpp): include model name in co
media EN What's Changed fix: remove Llama 3.1 405B from Groq provider by @RomneyDa in #11804 fix: add keepAlive to YAML completion options schema by @RomneyDa in #11801 fix(llama.cpp): include model name in co
media