Transformers

Join the Hugging Face community

and get access to the augmented documentation experience

Collaborate on models, datasets and Spaces

Faster examples with accelerated inference

Switch between documentation themes

to get started

Training on Specialized Hardware

Note: Most of the strategies introduced in the single GPU section (such as mixed precision training or gradient accumulation) and multi-GPU section are generic and apply to training models in general so make sure to have a look at it before diving into this section.

This document will be completed soon with information on how to train on specialized hardware.

←Training on TPU with TensorFlow Custom hardware for training→