# Hugging Face — TorchNeuron: running HF models on AWS Trainium is easier than ever

- Company: Hugging Face (huggingface.co)
- Announced: 2026-10-19
- Category: not stated
- Coverage: 1 outlet
- Announcement: yes
- Group: covered
- Source: https://huggingface.co/blog/hf-torchneuron
- Record: https://forck.live/items/24704-torchneuron-running-hf-models-on-aws-trainium-is-easier-than-ever
- Subject: Platform

TorchNeuron provides a new native backend for [PyTorch](https://pytorch.org/blog/introducing-torchneuron-native-pytorch-backend-for-aws-trainium) on AWS Trainium. Transformers, Diffusers and Kernels natively support it, making it easier than ever to run models on Trainium and to share them on the HF hub. TorchNeuron TorchNeuron is a [PyTorch](https://pytorch.org/blog/introducing-torchneuron-native-pytorch-backend-for-aws-trainium) plugin that brings a new backend for Trainium enabling support for eager mode, torch.compile and native distributed APIs like FSDP and TP. Read the [Pytorch](https://pytorch.org/blog/introducing-torchneuron-native-pytorch-backend-for-aws-trainium) blog to learn more about TorchNeuron. Starting Oct. 20, 2026, TorchNeuron will be available as a public beta. Researchers and developers will be able to install it with pip install torch-neuronx . New backend, native interface in HF libraries To date, the Hugging Face experience on AWS Trainium and Inferentia — AI chips designed respectively for cost-effective training and inference at scale — has been delivered through optimum-neuron . This library relies on the torch-neuronx package, which is based on [PyTorch](https://pytorch.org/blog/introducing-torchneuron-native-pytorch-backend-for-aws-trainium) XLA. It was the natural choice at the time, but it meant thinking like a compiler: lazy tensors instead of eager execution, static shapes only, a recompilation for every graph break. …

## Around this story

Outlets this record can name and link:

- PyTorch — Introducing TorchNeuron: Native PyTorch backend for AWS Trainium: https://pytorch.org/blog/introducing-torchneuron-native-pytorch-backend-for-aws-trainium

---

Record: https://forck.live/items/24704-torchneuron-running-hf-models-on-aws-trainium-is-easier-than-ever
Catalogue: https://forck.live/llms.txt
Current issue: https://forck.live/feed.md