Riva NMT NIM Overview

Riva NMT NIM APIs provide easy access to state-of-the-art neural machine translation (NMT) models, capable of translating text from one language to another with exceptional accuracy. It is based on the Transformer “Big” architecture having 24 layers in the encoder and decoder, and uses the SentencePiece tokenizer. Riva NMT NIM models are built on the NVIDIA software platform, incorporating CUDA, TensorRT, and Triton to offer out-of-the-box GPU acceleration.

Architecture

Model architecture can be found from the Attention is all you need. paper.

Enterprise-Ready Features

Riva NMT NIM comes with enterprise-ready features, such as a high-performance inference server, flexible integration, and enterprise-grade security.

State-of-the-art accuracy: Superior performance across diverse sources and domains.
Open-source and extensibility: Built on NVIDIA NeMo, allowing for seamless integration and customization.
Pre-trained checkpoints: Ready-to-use model for inference or fine-tuning.
Permissive license: Released under CC-BY-4.0 license, model checkpoints can be used in any commercial application.

Try It Out

Riva NMT NIM can be tried out at this link.

Previous Riva NMT NIM

Next Getting Started