AINOVAT
Source
60 articles from this source
Sentence Transformers is a Python library for using and training embedding and reranker models for a wide range of applications, such as retrieval augmented generation, semantic search, semantic textual similarity, and…
A technical walkthrough of how we built the Granite 4.2 reasoning model family.
Making a large language model smaller almost always comes with a cost.
Most interesting AI apps are pipelines.
3 months ago, we started a revival of Papers with Code (see also the announcement tweet ).
Public voice AI benchmarks increasingly suggest that models are performing at human levels.
Today, we release DSpark draft model checkpoints for three models from our LFM2.5 family: LFM2.5-1.2B-Instruct, LFM2.5-2.6B, and LFM2.5-8B-A1B.
Today, we release QAD Q4_0 GGUFs.
In our previous post , we compared ALTK-Evolve with ACE and showed that how you deliver an agent's self-distilled guidelines — a few retrieved per task vs.
Sentence Transformers is a Python library for using and training embedding and reranker models for applications like retrieval augmented generation, semantic search, and more.
The previous post argued that utilization, not intelligence, is where the next real constraint in enterprise AI is forming, and it closed by noting that no playbook has emerged yet for what a mature GPU Management pract…
In the AI world, time feels compressed.
A walkthrough of the streaming data loop in Strands Robots, one agent loop that records robot demonstrations, trains on them by reading straight from the Hub, and deploys the policy back to hardware, with the dataset in…
Back in July, we ran a hackathon where more than 1,200 community members brought their own coding agents and tried to reproduce the papers published at ICML 2026, claim by claim.
📄 Tech Report: https://allenai.org/papers/olmoearth | 📊 Documentation: https://docs.olmoearth.allenai.org/embeddings | 💻 Learn more about OlmoEarth: https://allenai.org/olmoearth OlmoEarth Studio , our platform for buil…
LFM2.5-VL-3B is our most capable vision-language model you can run on your own hardware.
ALTK-Evolve and ACE both let an agent learn from its own trajectories.
Every voice interaction has a latency budget.
Knowledge distillation , training a smaller student model to match the performance of a larger teacher, is a well-known technique in Machine Learning.
Great news from the OGs of open source LLMs!
📄 Tech Report: https://tutormoments.allen.ai/static/paper/tutormoments-preview.pdf | 📊 Data: https://huggingface.co/datasets/allenai/tutormoments-preview | 💻 Code: https://github.com/allenai/tutormoments Today we're int…
We're thrilled to share that Baseten is now a supported Inference Provider on the Hugging Face Hub!
LFM2.5-2.6B is built to power capable agents entirely on-device.
Aviation learned this the hard way.
🌍 Learn more about OlmoEarth Platform: https://allenai.org/olmoearth The OlmoEarth models are our family of Earth observation foundation models, pretrained on roughly 10 terabytes of multimodal satellite data.
Today, we release two new encoder models on Hugging Face: LFM2.5-Encoder-230M and LFM2.5-Encoder-350M .
Surgical robotics is moving quickly from teleoperation toward increasingly capable vision-language-action policies.
A companion technical writeup to our incident disclosure .
Large diffusion transformers can create stunning images (or even videos, audio snippets, and now text), but loading a modern text-to-image model in BF16 precision often requires 20-30 GB of VRAM, which puts these models…
Figure 1 : Humanoid robot locomotion simulation.
Record your own manipulation tasks in minutes with a handheld gripper, turn them into robot-ready datasets automatically, and help grow an open, collaborative dataset for robot learning.
The real world is vast and to operate in it physical AI systems need to understand how a scene is changing, anticipate what could happen next, and determine how an action will affect the world.
A joint post from NVIDIA and Hugging Face.
Retrieval is critical in multi-step agentic workflows where poor retrieval can cause agents to fetch irrelevant context, re-query, waste token budget, and carry noise into later reasoning steps.
Three months ago, we published a paper on DharmaOCR and open-sourced one of the models .
Earlier this week, we detected and responded to an intrusion into part of our production infrastructure.
Shippy is a maritime AI agent built for high-stakes decisions, where the wrong answer has real impacts.
Building a router into your agent sounds like an easy win.
Inkling is a large (1T params!) open model to natively accept image, text, and audio inputs.
Existing benchmarks suggest voice AI is nearing human-level performance but real-world conversations tell a different story.
This is the third post of Profiling in PyTorch, a series where we slowly build the skill of reading profiler traces and use it to drive optimization: The series "Profiling in PyTorch" is meant to make you comfortable re…
Why agentic AI needs open data, and why synthetic data is how we scale it.
TL;DR : The transformers vLLM backend is now as fast (or faster) than custom vLLM implementations for many LLM architectures.
Today, we’re excited to announce a deep-link integration between Hugging Face and Amazon SageMaker AI .
At Microsoft Build 2026, we announced Foundry Managed Compute and Hugging Face models on Foundry — a curated catalog of open-weight models from the Hugging Face ecosystem, refreshed weekly, deployable in one click onto…
For most teams, models and datasets live in a bucket in one region of one cloud.
LeRobot v0.6.0 introduces world model policies, reward models, and deployment tools to improve robot learning and adaptation.
Photoroom reveals the data strategy behind PRX, its 7B text-to-image model, including data assembly, captioning, and filtering.
Hugging Face's Kernels project introduces new repository type, trusted publishers, and code signing for enhanced security and discoverability.
For voice AI, latency is a critical parameter.
Modernizing enterprise applications is one of the largest and most expensive software engineering activities organizations undertake.
Those who follow Dharma AI already know that we view specialization as one of the defining principles of effective AI systems, shaping everything from cost and performance to reliability and sovereignty.
Every Eval Ever (EEE) and Hugging Face Community Evals are now intercompatible.
Many problems in machine learning and the sciences come down to the same task: you have a collection of data points and want to recover the distribution they came from—which values are common, and which are rare.
You can spin up a private, OpenAI-compatible LLM endpoint on Hugging Face infrastructure with a single command — no servers to provision, no Kubernetes, pay-per-second.
HuggingFace Transformers has become the foundation of the open-source AI ecosystem, and the recent Transformers v5 release strengthened it with first-class support for Mixture-of-Experts (MoE) models, now the dominant a…
🚀 First open far-field ASR benchmark: community-driven evaluation across 14 simulated rooms, validated against real-world measurements: https://huggingface.co/spaces/treble-technologies/ffasr 📉 The gap is real and it is…
TL;DR — Building an agent is mostly plumbing: tools, state, guardrails, scaling from one agent to many.
(This is a guest post by Developer Relations Engineer Thomas Steiner from the Chrome team at Google.) Transformers.js provides Web developers with a simple way to use the power of transformers in their Web apps through…
huggingface_hub is the Python client at the base of the Hugging Face ecosystem.