The best of dev and AI, scored every day by an agent. Filtered, summarized, ranked. No color, no noise — just the substance.
3LM est un benchmark multidomaine pour évaluer les LLM arabes en STEM et en code, déployant trois jeux de données (Native STEM MCQs, Synthetic STEM et Arabic Code Benchmarks) et des métriques comme pass@1 via EvalPlus. Le pipeline combine OCR, génération par LLM et vérifications humaines, et propose l'accès aux jeux sur HuggingFace et le code sur GitHub.
The article demonstrates how to implement an MCP server with Gradio to plug an LLM into Hugging Face models. It covers automatic conversion of Python functions into MCP tools, real-time progress notifications, and automatic file uploads, illustrated by an AI shopping assistant using IDM-VTON for virtual try-on.
Trackio est une bibliothèque légère d’expérimentation open-source qui remplace wandb et offre un tableau de bord local Gradio, avec synchronisation vers Hugging Face Spaces pour partager les métriques. Elle vise simplicité, traçabilité des métriques et énergie GPU via nvidia-smi, et peut être utilisée comme drop-in wandb.
HF renames the Hugging Face CLI from huggingface-cli to hf to reorganize commands around a clear hf <resource> <action> pattern, improving ergonomics and discoverability. The migration keeps the legacy CLI functional with warnings and documents how to install (pip install -U huggingface_hub) and verify (hf version) and explore commands (hf --help). It also introduces the hf jobs service to run scripts or Docker images on HF infrastructure with pay-as-you-go billing.
Parquet Content-Defined Chunking (CDC) is now available in PyArrow and Pandas, enabling deduplication of Parquet files on Hugging Face’s Xet storage layer. CDC reduces data transfer and storage costs by uploading or downloading only changed data chunks. The article shows how to enable CDC with use_content_defined_chunking and outlines several deduplication scenarios with code examples.
TimeScope is an open-source benchmark that tests long-video understanding by inserting short needle clips into base videos (1 min to 8 hours). It evaluates localized retrieval, information synthesis, and fine-grained temporal perception, revealing that many SOTA models struggle with true temporal comprehension.
LoRA adapters enable customizing diffusion models but can slow or complicate inference. This post shows how to speed up LoRA inference for Flux using a recipe based on FA3, FP8 quantization, and torch.compile, while staying hot-swapping–ready. It includes a concise code example that applies quantization, sets the attn processor, and compiles the transformer for faster inference.
NVIDIA NIM offre un conteneur unique pour deployer rapidement une large gamme de LLM via Hugging Face, en automatisant l’adaptation, l’analyse du modele et le choix du backend (TensorRT-LLM, vLLM, SGLang). Il prend en charge Hugging Face, GGUF et TensorRT-LLM et illustre le deployment avec Codestral-22B via une commande Docker et tokens API.
Arc Institute lance le Virtual Cell Challenge, qui vise à entraîner un modèle capable de prédire l’effet du silençage d’un gène sur une cellule, même dans des types cellulaires inédits (contexte généralisation). Le jeu de données réunit environ 300k profils RNA‑seq d’une cellule unique et propose un cadre mathématique pour séparer le signal du perturbation, l’hétérogénéité et le bruit technique, avec des architectures comme le State Transition Model et le State Embedding Model.
Consilium est une plateforme qui fait débattre et faire consensus entre plusieurs LLMs via des discussions structurées, avec des modes comme consensus, vote majoritaire ou classement par choix. Déployée comme interface Gradio et serveur MCP, elle visualise une table ronde et les échanges des experts. L'article relie l'idée au MAI-DxO de Microsoft pour démontrer l'efficacité du multi-LLM.
Gradio has released version 5.38.0 to enhance MCP servers with Seamless Local File Support via a new File Upload endpoint, real-time progress streaming for MCP clients, and a one-line OpenAPI-to-MCP conversion using gr.load_openapi. The update also improves authentication by allowing server arguments to be declared as gr.Header and surfaced in docs.
AI evaluation should shift from recalling facts to forecasting future events. FutureBench uses real-world prediction markets and live news to test agents' ability to reason under uncertainty and synthesize information to predict outcomes.
Ettin introduces the first SoTA paired encoder-only and decoder-only models (17M–1B params) trained identically for apples-to-apples comparisons. It extends the ModernBERT recipe to both architectures, with encoders beating ModernBERT and decoders beating Llama 3.2 and SmolLM2, while preserving architecture-specific advantages.
Hugging Face has deployed Xet as the Hub’s storage backend, migrating hundreds of petabytes and millions of repos with minimal disruption. The migration relies on a Git LFS Bridge and background content migrations to support both LFS and Xet, allowing a no-hard-cutover transition. The approach aims to scale storage with AI workloads while preserving existing workflows.
Kimina-Prover-72B uses a Test-Time Reinforcement Learning (TTRL) search to autonomously discover and reuse lemmas for long-horizon Lean proofs, plus an error-fixing module that interprets Lean messages for targeted corrections. It achieves a state-of-the-art miniF2F performance (92.2% pass rate) and shows pass@1/32/1024 of 63.9/84.0/87.7, with two distilled variants released (8B and 1.7B).
Asynchronous inference decouples action prediction from execution in robotic policies, reducing runtime lag and enabling replanning with action chunks. The article describes a two-component architecture (PolicyServer and RobotClient) using gRPC to achieve ~2× speedups and continuous operation, and explains why sequential inference falls short.
Building the Hugging Face MCP Server enables customized AI Assistants to access the Hub and thousands of apps through a single URL. The article compares MCP transports (Streamable HTTP vs SSE) and outlines three patterns: Direct Response, Request Scoped Streams, and Server Push Streams, with their trade-offs and needed connection management. It also covers making the server dynamic and remotely configurable, plus client-side usage hints (TypeScript/Python).
ScreenEnv is a Python library that runs isolated Ubuntu desktop environments in Docker to test and deploy GUI agents with full desktop control, including window management and file operations. It supports both the Model Context Protocol (MCP) and a Direct Sandbox API, enabling flexible integration with existing backends or AI systems. The article provides setup examples and a quick path to build a custom Desktop Agent using smolagents.
Hugging Face and AMD optimize custom kernels for MI300X to boost Llama 3.1 405B inference in FP8 on 8 GPUs. They combine fused residual/RMS norm/FP8 conversion, SwiGLU, and a Skinny GEMM kernel, with benchmarks and open-source tooling in hf-rocm-kernels.
The article introduces the Model Context Protocol (MCP) and explains how MCP servers can extend LLMs with new abilities via an app-store-like ecosystem. It guides readers to find MCP-compatible spaces in Hugging Face Spaces and demonstrates wiring a Flux.1 Kontext[dev] MCP server to an LLM to edit images from text prompts.