Serve, optimize and scale PyTorch models in production

AWS Deep Learning Containers (DLCs) are a set of Docker images for training and serving models in TensorFlow, TensorFlow 2, PyTorch, and MXNet.

A universal scalable machine learning model deployment solution

A high-throughput and memory-efficient inference and serving engine for LLMs

Demo applications showcasing DJL

An Engine-Agnostic Deep Learning Framework in Java

🤗 PEFT: State-of-the-art Parameter-Efficient Fine-Tuning.

Transformer related optimization, including BERT, GPT

Serve machine learning models within a 🐳 Docker container using 🧠 Amazon SageMaker.

Toolkit for allowing inference and serving with PyTorch on SageMaker. Dockerfiles used for building SageMaker Pytorch Containers are at https://github.com/aws/deep-learning-containers.

DeepSpeed is a deep learning optimization library that makes distributed training and inference easy, efficient, and effective.

Example 📓 Jupyter notebooks that demonstrate how to build, train, and deploy machine learning models using 🧠 Amazon SageMaker.

The Triton Inference Server provides an optimized cloud and edge inferencing solution.

Tensors and Dynamic neural networks in Python with strong GPU acceleration

A PyTorch Extension: Tools for easy mixed precision and distributed training in Pytorch

Deep Learning Examples

🤗 Transformers: State-of-the-art Natural Language Processing for TensorFlow 2.0 and PyTorch.

An implementation of a deep learning recommendation model (DLRM)

Project starter code for the Garbage Collector project.

The official Caffe2 port on AMD platform, rocm-caffe2 is

HIP back-end for Thrust

Caffe2 is a lightweight, modular, and scalable deep learning framework.