Skip to content

Organizations

Hugging Face · Latest

Open-source AI research

5 selected reportsIn the last 30 days 5 reportsIndexed 14 reports

RSS
updated

Selected: Hugging Face

Oct 29ThuItems 1–5
  1. Hugging Face BlogFirst-party

    Project ProtoStar launches to make publicly funded science AI-ready

    Project ProtoStar, an initiative led by Carnegie Mellon University and partners including Hugging Face and ORCA, aims to make publicly funded scientific datasets, models, and code more accessible and reusable for AI research. The project addresses the gap between public availability and usability by encouraging systematic sharing in AI-ready formats through open infrastructure and challenges, starting with a climate risk agent challenge hosted by Carnegie Mellon University on November 2.

    Read source huggingface.co
Oct 21Wed
  1. Hugging Face BlogFirst-party

    Hugging Face Launches TorchTPU, Making TPUs a Native PyTorch Backend

    Hugging Face has released TorchTPU, a native PyTorch backend that lets users run Transformers, Diffusers, and Kernels directly on Google’s Tensor Processing Units (TPUs). By specifying device_map="tpu", models from the Hub load onto the TPU without extra conversion steps, and the library includes Pallas kernels, distributed training support, and torch.compile integration for speed. This removes the need for the older torch_xla stack and Optimum TPU for TPU inference and training.

    Read source huggingface.co
Oct 20Tue
  1. Hugging Face BlogFirst-party

    Hugging Face Launches TorchNeuron, a Native PyTorch Backend for AWS Trainium

    Hugging Face has released TorchNeuron, a new PyTorch plugin that adds a native backend for AWS Trainium. The backend supports eager execution, torch.compile, and distributed APIs such as FSDP and DTensor, allowing Transformers, Diffusers, Accelerate, PEFT, and TRL to run on Trainium with a simple device_map=neuron argument. From October 20, 2026, TorchNeuron will be available as a public beta via pip install torch-neuronx, and it already supports 89% of the top 100 text‑generation model architectures compared with 46% under Optimum Neuron.

    Read source huggingface.co
Oct 15Thu
  1. Hugging Face BlogFirst-party

    FLUX.3 Image Klein Debuts in Hugging Face's Diffusers

    Hugging Face introduces FLUX.3 Image Klein, a model capable of text-to-image, multi-reference image-to-image, and layout control, now available in Diffusers. The model supports bounding-box layout manipulation and offers a 1024x1024 image generation in approximately 2 seconds on an H100 GPU.

    Read source huggingface.co
  2. Hugging Face BlogFirst-party

    HiRO-ACEv2: Cost-Effective Extreme Weather Simulation

    Hugging Face introduces HiRO-ACEv2, a climate emulator that simulates extreme weather with high accuracy at 3 km resolution, at a fraction of the cost of X-SHiELD. HiROv2, the new downscaling model, improves prediction of rare events and matches X-SHiELD's statistics within 3% accuracy.

    Read source huggingface.co