- Notifications
You must be signed in to change notification settings - Fork881
Serve, optimize and scale PyTorch models in production
License
NotificationsYou must be signed in to change notification settings
pytorch/serve
Folders and files
| Name | Name | Last commit message | Last commit date | |
|---|---|---|---|---|
Repository files navigation
This project is no longer actively maintained. While existing releases remain available, there are no planned updates, bug fixes, new features, or security patches. Users should be aware that vulnerabilities may not be addressed.
TorchServe now enforces token authorization enabled and model API control disabled by default. These security features are intended to address the concern of unauthorized API calls and to prevent potential malicious code from being introduced to the model server. Refer the following documentation for more information:Token Authorization,Model API control
TorchServe is a flexible and easy-to-use tool for serving and scaling PyTorch models in production.
Requires python >= 3.8
curl http://127.0.0.1:8080/predictions/bert -T input.txt
# Install dependenciespython ./ts_scripts/install_dependencies.py# Include dependencies for accelerator support with the relevant optional flagspython ./ts_scripts/install_dependencies.py --rocm=rocm61python ./ts_scripts/install_dependencies.py --cuda=cu121# Latest releasepip install torchserve torch-model-archiver torch-workflow-archiver# Nightly buildpip install torchserve-nightly torch-model-archiver-nightly torch-workflow-archiver-nightly
# Install dependenciespython ./ts_scripts/install_dependencies.py# Include depeendencies for accelerator support with the relevant optional flagspython ./ts_scripts/install_dependencies.py --rocm=rocm61python ./ts_scripts/install_dependencies.py --cuda=cu121# Latest releaseconda install -c pytorch torchserve torch-model-archiver torch-workflow-archiver# Nightly buildconda install -c pytorch-nightly torchserve torch-model-archiver torch-workflow-archiver
# Latest releasedocker pull pytorch/torchserve# Nightly builddocker pull pytorch/torchserve-nightly
Refer totorchserve docker for details.
# Make sure to install torchserve with pip or conda as described above and login with `huggingface-cli login`python -m ts.llm_launcher --model_id meta-llama/Llama-3.2-3B-Instruct --disable_token_auth# Try it outcurl -X POST -d'{"model":"meta-llama/Llama-3.2-3B-Instruct", "prompt":"Hello, my name is", "max_tokens": 200}' --header"Content-Type: application/json""http://localhost:8080/predictions/model/1.0/v1/completions"
# Make sure to install torchserve with python venv as described above and login with `huggingface-cli login`# pip install -U --use-deprecated=legacy-resolver -r requirements/trt_llm.txtpython -m ts.llm_launcher --model_id meta-llama/Meta-Llama-3.1-8B-Instruct --engine trt_llm --disable_token_auth# Try it outcurl -X POST -d'{"prompt":"count from 1 to 9 in french ", "max_tokens": 100}' --header"Content-Type: application/json""http://localhost:8080/predictions/model"
#export token=<HUGGINGFACE_HUB_TOKEN>docker build --pull. -f docker/Dockerfile.vllm -t ts/vllmdocker run --rm -ti --shm-size 10g --gpus all -e HUGGING_FACE_HUB_TOKEN=$token -p 8080:8080 -v data:/data ts/vllm --model_id meta-llama/Meta-Llama-3-8B-Instruct --disable_token_auth# Try it outcurl -X POST -d'{"model":"meta-llama/Meta-Llama-3-8B-Instruct", "prompt":"Hello, my name is", "max_tokens": 200}' --header"Content-Type: application/json""http://localhost:8080/predictions/model/1.0/v1/completions"
Refer toLLM deployment for details and other methods.
- Write once, run anywhere, on-prem, on-cloud, supports inference on CPUs, GPUs, AWS Inf1/Inf2/Trn1, Google Cloud TPUs,Nvidia MPS
- Model Management API: multi model management with optimized worker to model allocation
- Inference API: REST and gRPC support for batched inference
- TorchServe Workflows: deploy complex DAGs with multiple interdependent models
- Default way to serve PyTorch models in
- Sagemaker
- Vertex AI
- Kubernetes with support forautoscaling, session-affinity, monitoring using Grafana works on-prem, AWS EKS, Google GKE, Azure AKS
- Kserve: Supports both v1 and v2 API,autoscaling and canary deployments for A/B testing
- Kubeflow
- MLflow
- Export your model for optimized inference. Torchscript out of the box,PyTorch Compiler preview,ORT and ONNX,IPEX,TensorRT,FasterTransformer, FlashAttention (Better Transformers)
- Performance Guide: builtin support to optimize, benchmark, and profile PyTorch and TorchServe performance
- Expressive handlers: An expressive handler architecture that makes it trivial to support inferencing for your use case withmany supported out of the box
- Metrics API: out-of-the-box support for system-level metrics withPrometheus exports, custom metrics,
- Large Model Inference Guide: With support for GenAI, LLMs including
- SOTA GenAI performance using
torch.compile - Fast Kernels with FlashAttention v2, continuous batching and streaming response
- PyTorchTensor Parallel preview,Pipeline Parallel
- MicrosoftDeepSpeed,DeepSpeed-Mii
- Hugging FaceAccelerate,Diffusers
- Running large models on AWSSagemaker andInferentia2
- RunningMeta Llama Chatbot locally on Mac
- SOTA GenAI performance using
- Monitoring using Grafana andDatadog
- Model Server for PyTorch Documentation: Full documentation
- TorchServe internals: How TorchServe was built
- Contributing guide: How to contribute to TorchServe
- Serving Meta Llama with TorchServe
- Chatbot with Meta Llama on Mac 🦙💬
- 🤗 HuggingFace Transformers with aBetter Transformer Integration/ Flash Attention & Xformer Memory Efficient
- Stable Diffusion
- Model parallel inference
- MultiModal models with MMF combining text, audio and video
- Dual Neural Machine Translation for a complex workflow DAG
- TorchServe Integrations
- TorchServe Internals
- TorchServe UseCases
We welcome all contributions!
To learn more about how to contribute, see the contributor guidehere.
- High performance Llama 2 deployments with AWS Inferentia2 using TorchServe
- Naver Case Study: Transition From High-Cost GPUs to Intel CPUs and oneAPI powered Software with performance
- Run multiple generative AI models on GPU using Amazon SageMaker multi-model endpoints with TorchServe and save up to 75% in inference costs
- Deploying your Generative AI model in only four steps with Vertex AI and PyTorch
- PyTorch Model Serving on Google Cloud TPU v5
- Monitoring using Datadog
- Torchserve Performance Tuning, Animated Drawings Case-Study
- Walmart Search: Serving Models at a Scale on TorchServe
- 🎥 Scaling inference on CPU with TorchServe
- 🎥 TorchServe C++ backend
- Grokking Intel CPU PyTorch performance from first principles: a TorchServe case study
- Grokking Intel CPU PyTorch performance from first principles( Part 2): a TorchServe case study
- Case Study: Amazon Ads Uses PyTorch and AWS Inferentia to Scale Models for Ads Processing
- Optimize your inference jobs using dynamic batch inference with TorchServe on Amazon SageMaker
- Using AI to bring children's drawings to life
- 🎥 Model Serving in PyTorch
- Evolution of Cresta's machine learning architecture: Migration to AWS and PyTorch
- 🎥 Explain Like I’m 5: TorchServe
- 🎥 How to Serve PyTorch Models with TorchServe
- How to deploy PyTorch models on Vertex AI
- Quantitative Comparison of Serving Platforms
- Efficient Serverless deployment of PyTorch models on Azure
- Deploy PyTorch models with TorchServe in Azure Machine Learning online endpoints
- Dynaboard moving beyond accuracy to holistic model evaluation in NLP
- A MLOps Tale about operationalising MLFlow and PyTorch
- Operationalize, Scale and Infuse Trust in AI Models using KFServing
- How Wadhwani AI Uses PyTorch To Empower Cotton Farmers
- TorchServe Streamlit Integration
- Dynabench aims to make AI models more robust through distributed human workers
- Announcing TorchServe
Made withcontrib.rocks.
This repository is jointly operated and maintained by Amazon, Meta and a number of individual contributors listed in theCONTRIBUTORS file. For questions directed at Meta, please send an email toopensource@fb.com. For questions directed at Amazon, please send an email totorchserve@amazon.com. For all other questions, please open up an issue in this repositoryhere.
TorchServe acknowledges theMulti Model Server (MMS) project from which it was derived
About
Serve, optimize and scale PyTorch models in production
Topics
Resources
License
Code of conduct
Security policy
Uh oh!
There was an error while loading.Please reload this page.