TorchNeuron Makes It Easier Than Ever to Run Hugging Face Models on AWS Trainium
TorchNeuron runs Hugging Face models natively on Trainium via PyTorch, supporting eager execution, torch.compile, sharded data and tensor parallelism.
TorchNeuron runs Hugging Face models natively on Trainium via PyTorch, supporting eager execution, torch.compile, sharded data and tensor parallelism.