Easy, advanced inference platform for large language models on Kubernetes
[](https://github.com/mkenney/software-guides/blob/master/STABILITY-BADGES.md#alpha)
[![GoReport Widget]][GoReport Status]
[](https://github.com/inftyai/llmaz/releases/latest)
[GoReport Widget]: https://goreportcard.com/badge/github.com/inftyai/llmaz
[GoReport Status]: https://goreportcard.com/report/github.com/inftyai/llmaz
**llmaz** (pronounced `/lima:z/`), aims to provide a **Production-Ready** inference platform for large language models on Kubernetes. It closely integrates with the state-of-the-art inference backends to bring the leading-edge researches to cloud.
> 🌱 llmaz is alpha now, so API may change before graduating to Beta.
## Architecture
## Features Overview
- **Easy of Use**: People can quick deploy a LLM service with minimal configurations.
- **Broad Backend Support**: llmaz supports a wide range of advanced inference backends for different scenarios, like [vLLM](https://github.com/vllm-project/vllm), [Text-Generation-Inference](https://github.com/huggingface/text-generation-inference), [SGLang](https://github.com/sgl-project/sglang), [llama.cpp](https://github.com/ggerganov/llama.cpp). Find the full list of supported backends [here](./docs/support-backends.md).
- **Scaling Efficiency (WIP)**: llmaz works smoothly with autoscaling components like [Cluster-Autoscaler](https://github.com/kubernetes/autoscaler/tree/master/cluster-autoscaler) or [Karpenter](https://github.com/kubernetes-sigs/karpenter) to support elastic scenarios.
- **Accelerator Fungibility (WIP)**: llmaz supports serving the same LLM with various accelerators to optimize cost and performance.
- **SOTA Inference**: llmaz supports the latest cutting-edge researches like [Speculative Decoding](https://arxiv.org/abs/2211.17192) or [Splitwise](https://arxiv.org/abs/2311.18677)(WIP) to run on Kubernetes.
- **Various Model Providers**: llmaz supports a wide range of model providers, such as [HuggingFace](https://huggingface.co/), [ModelScope](https://www.modelscope.cn), ObjectStores(aliyun OSS, more on the way). llmaz automatically handles the model loading requiring no effort from users.
- **Multi-hosts Support**: llmaz supports both single-host and multi-hosts scenarios with [LWS](https://github.com/kubernetes-sigs/lws) from day 1.
## Quick Start
### Installation
Read the [Installation](./docs/installation.md) for guidance.
### Deploy
Here's a toy sample for deploying `facebook/opt-125m`, all you need to do
is to apply a `Model` and a `Playground`.
Please refer to **[examples](/docs/examples/README.md)** to learn more.
> Note: if your model needs Huggingface token for weight downloads, please run `kubectl create secret generic modelhub-secret --from-literal=HF_TOKEN=` ahead.
#### Model
```yaml
apiVersion: llmaz.io/v1alpha1
kind: OpenModel
metadata:
name: opt-125m
spec:
familyName: opt
source:
modelHub:
modelID: facebook/opt-125m
inferenceFlavors:
- name: t4 # GPU type
requests:
nvidia.com/gpu: 1
```
#### Inference Playground
```yaml
apiVersion: inference.llmaz.io/v1alpha1
kind: Playground
metadata:
name: opt-125m
spec:
replicas: 1
modelClaim:
modelName: opt-125m
```
### Test
#### Expose the service
```cmd
kubectl port-forward pod/opt-125m-0 8080:8080
```
#### Get registered models
```cmd
curl http://localhost:8080/v1/models
```
#### Request a query
```cmd
curl http://localhost:8080/v1/completions \
-H "Content-Type: application/json" \
-d '{
"model": "opt-125m",
"prompt": "San Francisco is a",
"max_tokens": 10,
"temperature": 0
}'
```
### More than quick-start
If you want to learn more about this project, please refer to [develop.md](./docs/develop.md).
## Roadmap
- Gateway support for traffic routing
- Metrics support
- Serverless support for cloud-agnostic users
- CLI tool support
- Model training, fine tuning in the long-term
## Contributions
🚀 All kinds of contributions are welcomed ! Please follow [CONTRIBUTING.md](./CONTRIBUTING.md).
**🎉 Thanks to all these contributors !**
