Archive month
月内の記事を日付順に並べています。カテゴリと source の偏りを見てから、下のタイムラインで詳細を確認できます。
原文EN Google releases Gemma 2 2B, ShieldGemma and Gemma Scope
EN Google releases Gemma 2 2B, ShieldGemma and Gemma Scope
原文EN Memory-efficient Diffusion Transformers with Quanto and Diffusers
EN Memory-efficient Diffusion Transformers with Quanto and Diffusers
原文EN Serverless Inference with Hugging Face and NVIDIA NIM
EN Serverless Inference with Hugging Face and NVIDIA NIM
原文EN We’re testing SearchGPT, a temporary prototype of new search features that give you fast and timely answers with clear and relevant sources.
EN We’re testing SearchGPT, a temporary prototype of new search features that give you fast and timely answers with clear and relevant sources.
原文EN LAVE: Zero-shot VQA Evaluation on Docmatix with LLMs - Do We Still Need Fine-Tuning?
EN LAVE: Zero-shot VQA Evaluation on Docmatix with LLMs - Do We Still Need Fine-Tuning?
原文EN We've developed and applied a new method leveraging Rule-Based Rewards (RBRs) that aligns models to behave safely without extensive human data collection.
EN We've developed and applied a new method leveraging Rule-Based Rewards (RBRs) that aligns models to behave safely without extensive human data collection.
原文EN Llama 3.1 - 405B, 70B & 8B with multilinguality and long context
EN Llama 3.1 - 405B, 70B & 8B with multilinguality and long context
原文EN WWDC 24: Running Mistral 7B with Core ML
EN WWDC 24: Running Mistral 7B with Core ML
原文EN GPT-4o mini: advancing cost-efficient intelligence
EN GPT-4o mini: advancing cost-efficient intelligence
原文EN Compliance API integrations, SCIM, and GPT controls to support compliance programs, data security, and user access at scale
EN Compliance API integrations, SCIM, and GPT controls to support compliance programs, data security, and user access at scale
原文EN Docmatix - a huge dataset for Document Visual Question Answering
EN Docmatix - a huge dataset for Document Visual Question Answering
原文EN TGI Multi-LoRA: Deploy Once, Serve 30 Models
EN TGI Multi-LoRA: Deploy Once, Serve 30 Models
原文EN Discover how prover-verifier games improve the legibility of language model outputs, making AI solutions clearer, easier to verify, and more trustworthy for both humans and machines.
EN Discover how prover-verifier games improve the legibility of language model outputs, making AI solutions clearer, easier to verify, and more trustworthy for both humans and machines.
原文EN SmolLM - blazingly fast and remarkably powerful
EN SmolLM - blazingly fast and remarkably powerful
原文EN How we leveraged distilabel to create an Argilla 2.0 Chatbot
EN How we leveraged distilabel to create an Argilla 2.0 Chatbot
原文EN How NuminaMath Won the 1st AIMO Progress Prize
EN How NuminaMath Won the 1st AIMO Progress Prize
原文EN OpenAI and Los Alamos National Laboratory are working to develop safety evaluations to assess and measure biological capabilities and risks associated with frontier models.
EN OpenAI and Los Alamos National Laboratory are working to develop safety evaluations to assess and measure biological capabilities and risks associated with frontier models.
AI要約 Hugging FaceとKerasがKerasHubで統合され、Hubに公開されたLlamaやGemma、Mistral等のTransformers形式モデルをKerasHubから直接ロード・微調整・再公開できるようになった。JAX・TensorFlow・PyTorchの全バックエンドに対応する。
EN Hugging Face and KerasHub now integrate so users can load, fine-tune and re-upload Transformers-format checkpoints (Llama, Gemma, Mistral, etc.) from the Hub directly into Keras across JAX, TensorFlow and PyTorch backends.
原文EN Experimenting with Automatic PII Detection on the Hub using Presidio
EN Experimenting with Automatic PII Detection on the Hub using Presidio
原文EN Preference Optimization for Vision Language Models
EN Preference Optimization for Vision Language Models
原文EN Google Cloud TPUs made available to Hugging Face users
EN Google Cloud TPUs made available to Hugging Face users
原文EN Banque des Territoires (CDC Group) x Polyconseil x Hugging Face: Enhancing a Major French Environmental Program with a Sovereign Data Solution
EN Banque des Territoires (CDC Group) x Polyconseil x Hugging Face: Enhancing a Major French Environmental Program with a Sovereign Data Solution
AI要約 Hugging Faceがデータセットハブの検索機能を刷新。モダリティやサイズ、フォーマットなどによる絞り込みが可能になり、目的に合ったデータセットを効率的に発見できるようになった。
EN Hugging Face has rolled out new dataset search and filtering features on its Hub, letting users narrow datasets by modality, size, format and other facets to find suitable training data faster.
原文EN Accelerating Protein Language Model ProtST on Intel Gaudi 2
EN Accelerating Protein Language Model ProtST on Intel Gaudi 2
原文EN Our Transformers Code Agent beats the GAIA benchmark 🏅
EN Our Transformers Code Agent beats the GAIA benchmark 🏅