Mistral 24B is Apache 2! 🔥 Already available for llama.cpp and MLX, what are you waiting for? Thank you Mistral AI 🙌, let's go open source!

I would actually guess that any models are gonna get smaller and smaller. On device models that can act as assistant will be the norm.

🔥 New Speech Recognition Dataset Released! 🔥 The SpeechBrain Team has released 25,000 hours of transcribed and diverse English speech data for both research and commercial use. This dataset is a unified, normalized, and cleaned super set of existing datasets, with tools

Super happy that we're releasing hibiki today, our first speech 🇫🇷 to speech 🇬🇧 translation model. It works in real-time while preserving the voice of the speaker, and best of all it can run on your phone. Code on GitHub (pytorch, mlx, swift, rust), and weights on HF 🚀

Today we're releasing Ultravox v0.5, the next iteration of our open-weight speech language model With this release, we've closed the gap with proprietary models. Ultravox now outperforms GPT-4o Realtime & Gemini 1.5 Flash on key benchmarks for speech understanding 🧵

We're doing a more flexible github release schedule now in transformers: as soon as your model is merge you get a tag, and soon a pip install transformers[your-model]! !

For anyone that wants to integrate a model to transformers, here's my talk from yesterday's gemma3 launch event Google ! Gives you all the MUST have and some of the common issues we face!

Today, we're launching Orpheus Multilingual, a family of open-source models that makes state-of-the-art TTS accessible to billions of new people! 🌎🌎 (1/5)

We've just released Helium 1, a 2B model that is best of its class on multi-lingual benchmarks and supports the 24 EU languages!

A quick update on the future of the `transformers` library! In order to provide a source of truth for all models, we are working with the rest of the ecosystem to make the modeling code the standard. A joint effort with vLLM, LlamaCPP, SGLang, Mlx, Qwen, Glm, Unsloth, Axoloth,

The Transformers library is undergoing it's largest pivot to date 🙌 It now cements its role as the central model definition, irrespective of the backend and runner. One ground truth to bring more reliability across the ecosystem. Why is this important?

Excited to expand our collaboration with Microsoft Azure, thanks for the shoutout Satya Nadella!

If AI stays closed-source, proprietary and monopolistic like it is now, it will destroy lots of jobs and just make the richest companies richer and more powerful! If we open it up thanks to open science and open-source, foster competition and decentralization of value and

🚀 Our paper **MDMP** has been accepted at CVPR’25 - HuMoGen 🚀 We propose a multi-modal diffusion model that fuses textual action descriptions and 3D skeletal data to generate long-term human motion predictions, with interpretable uncertainty — paving the way for safer and

Digging through cryptic researcher codebases stacked on top of each other with incorrect/missing/misleading variable names and papers that do not reflect them is such pain... BUT it perfectly confirms our vision for Transformers: a source of truth. Less time questioning