OpenLLM: Operating LLMs in production

Navigation

Verified details

These details have been verified by PyPI
Maintainers
Avatar for aar0npham from gravatar.com aar0npham Avatar for parano from gravatar.com parano

Unverified details

These details have not been verified by PyPI
Project links
GitHub Statistics

View statistics for this project via Libraries.io, or by using our public dataset on Google BigQuery

Meta

License: Apache Software License

Author: Aaron Pham

Tags AI, Alpaca, BentoML, Falcon, Fine tuning, Generative AI, LLMOps, Large Language Model, Llama 2, MLOps, Model Deployment, Model Serving, PyTorch, Serverless, StableLM, Transformers, Vicuna

Requires: Python >=3.8

Classifiers

Project description

Banner for OpenLLM

๐Ÿฆพ OpenLLM

pypi_status test_pypi_status Twitter Discord ci pre-commit.ci status
python_version Hatch code style Ruff types - mypy types - pyright

An open platform for operating large language models (LLMs) in production.
Fine-tune, serve, deploy, and monitor any LLMs with ease.

๐Ÿ“– Introduction

OpenLLM is an open-source platform designed to facilitate the deployment and operation of large language models (LLMs) in real-world applications. With OpenLLM, you can run inference on any open-source LLM, deploy them on the cloud or on-premises, and build powerful AI applications.

Key features include:

๐Ÿš‚ State-of-the-art LLMs: Integrated support for a wide range of open-source LLMs and model runtimes, including but not limited to Llama 2, StableLM, Falcon, Dolly, Flan-T5, ChatGLM, and StarCoder.

๐Ÿ”ฅ Flexible APIs: Serve LLMs over a RESTful API or gRPC with a single command. You can interact with the model using a Web UI, CLI, Python/JavaScript clients, or any HTTP client of your choice.

โ›“๏ธ Freedom to build: First-class support for LangChain, BentoML and Hugging Face, allowing you to easily create your own AI applications by composing LLMs with other models and services.

๐ŸŽฏ Streamline deployment: Automatically generate your LLM server Docker images or deploy as serverless endpoints via โ˜๏ธ BentoCloud, which effortlessly manages GPU resources, scales according to traffic, and ensures cost-effectiveness.

๐Ÿค–๏ธ Bring your own LLM: Fine-tune any LLM to suit your needs. You can load LoRA layers to fine-tune models for higher accuracy and performance for specific tasks. A unified fine-tuning API for models (LLM.tuning()) is coming soon.

โšกย Quantization: Run inference with less computational and memory costs with quantization techniques such as LLM.int8, SpQR (int4), AWQ,ย GPTQ, and SqueezeLLM.

๐Ÿ“กย Streaming: Support token streaming through server-sent events (SSE). You can use the /v1/generate_streamย endpoint for streaming responses from LLMs.

๐Ÿ”„ย Continuous batching: Support continuous batching via vLLM for increased total throughput.

OpenLLM is designed for AI application developers working to build production-ready applications based on LLMs. It delivers a comprehensive suite of tools and features for fine-tuning, serving, deploying, and monitoring these models, simplifying the end-to-end deployment workflow for LLMs.

Gif showing OpenLLM Intro

๐Ÿƒ Get started

To quickly get started with OpenLLM, follow the instructions below or try this OpenLLM tutorial in Google Colab: Serving Llama 2 with OpenLLM.

Prerequisites

You have installed Python 3.8 (or later) andย pip. We highly recommend using a Virtual Environment to prevent package conflicts.

Install OpenLLM

Install OpenLLM by using pip as follows:

pip install openllm

To verify the installation, run:

$ openllm -h

Usage: openllm [OPTIONS] COMMAND [ARGS]...

   โ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ•— โ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ•— โ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ•—โ–ˆโ–ˆโ–ˆโ•—   โ–ˆโ–ˆโ•—โ–ˆโ–ˆโ•—     โ–ˆโ–ˆโ•—     โ–ˆโ–ˆโ–ˆโ•—   โ–ˆโ–ˆโ–ˆโ•—
  โ–ˆโ–ˆโ•”โ•โ•โ•โ–ˆโ–ˆโ•—โ–ˆโ–ˆโ•”โ•โ•โ–ˆโ–ˆโ•—โ–ˆโ–ˆโ•”โ•โ•โ•โ•โ•โ–ˆโ–ˆโ–ˆโ–ˆโ•—  โ–ˆโ–ˆโ•‘โ–ˆโ–ˆโ•‘     โ–ˆโ–ˆโ•‘     โ–ˆโ–ˆโ–ˆโ–ˆโ•— โ–ˆโ–ˆโ–ˆโ–ˆโ•‘
  โ–ˆโ–ˆโ•‘   โ–ˆโ–ˆโ•‘โ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ•”โ•โ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ•—  โ–ˆโ–ˆโ•”โ–ˆโ–ˆโ•— โ–ˆโ–ˆโ•‘โ–ˆโ–ˆโ•‘     โ–ˆโ–ˆโ•‘     โ–ˆโ–ˆโ•”โ–ˆโ–ˆโ–ˆโ–ˆโ•”โ–ˆโ–ˆโ•‘
  โ–ˆโ–ˆโ•‘   โ–ˆโ–ˆโ•‘โ–ˆโ–ˆโ•”โ•โ•โ•โ• โ–ˆโ–ˆโ•”โ•โ•โ•  โ–ˆโ–ˆโ•‘โ•šโ–ˆโ–ˆโ•—โ–ˆโ–ˆโ•‘โ–ˆโ–ˆโ•‘     โ–ˆโ–ˆโ•‘     โ–ˆโ–ˆโ•‘โ•šโ–ˆโ–ˆโ•”โ•โ–ˆโ–ˆโ•‘
  โ•šโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ•”โ•โ–ˆโ–ˆโ•‘     โ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ•—โ–ˆโ–ˆโ•‘ โ•šโ–ˆโ–ˆโ–ˆโ–ˆโ•‘โ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ•—โ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ–ˆโ•—โ–ˆโ–ˆโ•‘ โ•šโ•โ• โ–ˆโ–ˆโ•‘
   โ•šโ•โ•โ•โ•โ•โ• โ•šโ•โ•     โ•šโ•โ•โ•โ•โ•โ•โ•โ•šโ•โ•  โ•šโ•โ•โ•โ•โ•šโ•โ•โ•โ•โ•โ•โ•โ•šโ•โ•โ•โ•โ•โ•โ•โ•šโ•โ•     โ•šโ•โ•.

  An open platform for operating large language models in production.
  Fine-tune, serve, deploy, and monitor any LLMs with ease.

Options:
  -v, --version  Show the version and exit.
  -h, --help     Show this message and exit.

Commands:
  build       Package a given models into a BentoLLM.
  import      Setup LLM interactively.
  models      List all supported models.
  prune       Remove all saved models, (and optionally bentos) built with OpenLLM locally.
  query       Query a LLM interactively, from a terminal.
  start       Start a LLMServer for any supported LLM.
  start-grpc  Start a gRPC LLMServer for any supported LLM.

Extensions:
  build-base-container  Base image builder for BentoLLM.
  dive-bentos           Dive into a BentoLLM.
  get-containerfile     Return Containerfile of any given Bento.
  get-prompt            Get the default prompt used by OpenLLM.
  list-bentos           List available bentos built by OpenLLM.
  list-models           This is equivalent to openllm models...
  playground            OpenLLM Playground.

Start an LLM server

OpenLLM allows you to quickly spin up an LLM server using openllm start. For example, to start anย OPTย server, run the following:

openllm start facebook/opt-1.3b

This starts the server atย http://0.0.0.0:3000/. OpenLLM downloads the model to the BentoML local Model Store if they have not been registered before. To view your local models, run bentoml models list.

To interact with the server, you can visit the web UI atย http://0.0.0.0:3000/ or send a request usingย curl. You can also use OpenLLMโ€™s built-in Python client to interact with the server:

import openllm
client = openllm.client.HTTPClient('http://localhost:3000')
client.query('Explain to me the difference between "further" and "farther"')

Alternatively, use theย openllm queryย command to query the model:

export OPENLLM_ENDPOINT=http://localhost:3000
openllm query 'Explain to me the difference between "further" and "farther"'

OpenLLM seamlessly supports many models and their variants. You can specify different variants of the model to be served by providing theย --model-id option. For example:

openllm start facebook/opt-2.7b

[!NOTE] OpenLLM supports specifying fine-tuning weights and quantized weights for any of the supported models as long as they can be loaded with the model architecture. Use theย openllm modelsย command to see the complete list of supported models, their architectures, and their variants.

๐Ÿงฉ Supported models

OpenLLM currently supports the following models. By default, OpenLLM doesn't include dependencies to run all models. The extra model-specific dependencies can be installed with the instructions below.

Mistral

Quickstart

Run the following commands to quickly spin up a Llama 2 server and send a request to it.

openllm start HuggingFaceH4/zephyr-7b-beta
export OPENLLM_ENDPOINT=http://localhost:3000
openllm query 'What are large language models?'

[!NOTE] Note that any Mistral variants can be deployed with OpenLLM. Visit the Hugging Face Model Hub to see more Mistral compatible models.

Supported models

You can specify any of the following Mistral models by using --model-id.

Supported backends

  • PyTorch (Default):

    openllm start HuggingFaceH4/zephyr-7b-beta --backend pt

  • vLLM (Recommended):

    pip install "openllm[vllm]"
openllm start HuggingFaceH4/zephyr-7b-beta --backend vllm

[!NOTE] Currently when using the vLLM backend, adapters is yet to be supported.

Llama

Installation

To run Llama models with OpenLLM, you need to install the llama dependency as it is not installed by default.

pip install "openllm[llama]"

Quickstart

Run the following commands to quickly spin up a Llama 2 server and send a request to it.

openllm start meta-llama/Llama-2-7b-chat-hf
export OPENLLM_ENDPOINT=http://localhost:3000
openllm query 'What are large language models?'

[!NOTE] To use the official Llama 2 models, you must gain access by visiting the Meta AI website and accepting its license terms and acceptable use policy. You also need to obtain access to these models on Hugging Face. Note that any Llama 2 variants can be deployed with OpenLLM if you donโ€™t have access to the official Llama 2 model. Visit the Hugging Face Model Hub to see more Llama 2 compatible models.

Supported models

You can specify any of the following Llama models by using --model-id.

Supported backends

  • PyTorch (Default):

    openllm start meta-llama/Llama-2-7b-chat-hf --backend pt

  • vLLM (Recommended):

    pip install "openllm[llama, vllm]"
openllm start meta-llama/Llama-2-7b-chat-hf --backend vllm

[!NOTE] Currently when using the vLLM backend, adapters is yet to be supported.

ChatGLM

Installation

To run ChatGLM models with OpenLLM, you need to install the chatglm dependency as it is not installed by default.

pip install "openllm[chatglm]"

Quickstart

Run the following commands to quickly spin up a ChatGLM server and send a request to it.

openllm start thudm/chatglm2-6b
export OPENLLM_ENDPOINT=http://localhost:3000
openllm query 'What are large language models?'

Supported models

You can specify any of the following ChatGLM models by using --model-id.

Supported backends

  • PyTorch (Default):

    openllm start thudm/chatglm2-6b --backend pt
Dolly-v2

Installation

Dolly-v2 models do not require you to install any model-specific dependencies once you have openllm installed.

pip install openllm

Quickstart

Run the following commands to quickly spin up a Dolly-v2 server and send a request to it.

openllm start databricks/dolly-v2-3b
export OPENLLM_ENDPOINT=http://localhost:3000
openllm query 'What are large language models?'

Supported models

You can specify any of the following Dolly-v2 models by using --model-id.

Supported backends

  • PyTorch (Default):

    openllm start databricks/dolly-v2-3b --backend pt

  • vLLM:

    openllm start databricks/dolly-v2-3b --backend vllm

[!NOTE] Currently when using the vLLM backend, adapters is yet to be supported.

Falcon

Installation

To run Falcon models with OpenLLM, you need to install the falcon dependency as it is not installed by default.

pip install "openllm[falcon]"

Quickstart

Run the following commands to quickly spin up a Falcon server and send a request to it.

openllm start tiiuae/falcon-7b
export OPENLLM_ENDPOINT=http://localhost:3000
openllm query 'What are large language models?'

Supported models

You can specify any of the following Falcon models by using --model-id.

Supported backends

  • PyTorch (Default):

    openllm start tiiuae/falcon-7b --backend pt

  • vLLM:

    pip install "openllm[falcon, vllm]"
openllm start tiiuae/falcon-7b --backend vllm

[!NOTE] Currently when using the vLLM backend, adapters is yet to be supported.

Flan-T5

Installation

To run Flan-T5 models with OpenLLM, you need to install the flan-t5 dependency as it is not installed by default.

pip install "openllm[flan-t5]"

Quickstart

Run the following commands to quickly spin up a Flan-T5 server and send a request to it.

openllm start google/flan-t5-large
export OPENLLM_ENDPOINT=http://localhost:3000
openllm query 'What are large language models?'

Supported models

You can specify any of the following Flan-T5 models by using --model-id.

Supported backends

  • PyTorch (Default):

    openllm start google/flan-t5-large --backend pt

[!NOTE] Currently when using the vLLM backend, adapters is yet to be supported.

GPT-NeoX

Installation

GPT-NeoX models do not require you to install any model-specific dependencies once you have openllm installed.

pip install openllm

Quickstart

Run the following commands to quickly spin up a GPT-NeoX server and send a request to it.

openllm start eleutherai/gpt-neox-20b
export OPENLLM_ENDPOINT=http://localhost:3000
openllm query 'What are large language models?'

Supported models

You can specify any of the following GPT-NeoX models by using --model-id.

Supported backends

  • PyTorch (Default):

    openllm start eleutherai/gpt-neox-20b --backend pt

  • vLLM:

    openllm start eleutherai/gpt-neox-20b --backend vllm

[!NOTE] Currently when using the vLLM backend, adapters is yet to be supported.

MPT

Installation

To run MPT models with OpenLLM, you need to install the mpt dependency as it is not installed by default.

pip install "openllm[mpt]"

Quickstart

Run the following commands to quickly spin up a MPT server and send a request to it.

openllm start mosaicml/mpt-7b-chat
export OPENLLM_ENDPOINT=http://localhost:3000
openllm query 'What are large language models?'

Supported models

You can specify any of the following MPT models by using --model-id.

Supported backends

  • PyTorch (Default):

    openllm start mosaicml/mpt-7b-chat --backend pt

  • vLLM (Recommended):

    pip install "openllm[mpt, vllm]"
openllm start mosaicml/mpt-7b-chat --backend vllm

[!NOTE] Currently when using the vLLM backend, adapters is yet to be supported.

OPT

Installation

To run OPT models with OpenLLM, you need to install the opt dependency as it is not installed by default.

pip install "openllm[opt]"

Quickstart

Run the following commands to quickly spin up an OPT server and send a request to it.

openllm start facebook/opt-2.7b
export OPENLLM_ENDPOINT=http://localhost:3000
openllm query 'What are large language models?'

Supported models

You can specify any of the following OPT models by using --model-id.

Supported backends

  • PyTorch (Default):

    openllm start facebook/opt-2.7b --backend pt

  • vLLM:

    pip install "openllm[opt, vllm]"
openllm start facebook/opt-2.7b --backend vllm

[!NOTE] Currently when using the vLLM backend, adapters is yet to be supported.

StableLM

Installation

StableLM models do not require you to install any model-specific dependencies once you have openllm installed.

pip install openllm

Quickstart

Run the following commands to quickly spin up a StableLM server and send a request to it.

openllm start stabilityai/stablelm-tuned-alpha-7b
export OPENLLM_ENDPOINT=http://localhost:3000
openllm query 'What are large language models?'

Supported models

You can specify any of the following StableLM models by using --model-id.

Supported backends

  • PyTorch (Default):

    openllm start stabilityai/stablelm-tuned-alpha-7b --backend pt

  • vLLM:

    openllm start stabilityai/stablelm-tuned-alpha-7b --backend vllm

[!NOTE] Currently when using the vLLM backend, adapters is yet to be supported.

StarCoder

Installation

To run StarCoder models with OpenLLM, you need to install the starcoder dependency as it is not installed by default.

pip install "openllm[starcoder]"

Quickstart

Run the following commands to quickly spin up a StarCoder server and send a request to it.

openllm start bigcode/starcoder
export OPENLLM_ENDPOINT=http://localhost:3000
openllm query 'What are large language models?'

Supported models

You can specify any of the following StarCoder models by using --model-id.

Supported backends

  • PyTorch (Default):

    openllm start bigcode/starcoder --backend pt

  • vLLM:

    pip install "openllm[startcoder, vllm]"
openllm start bigcode/starcoder --backend vllm

[!NOTE] Currently when using the vLLM backend, adapters is yet to be supported.

Baichuan

Installation

To run Baichuan models with OpenLLM, you need to install the baichuan dependency as it is not installed by default.

pip install "openllm[baichuan]"

Quickstart

Run the following commands to quickly spin up a Baichuan server and send a request to it.

openllm start baichuan-inc/baichuan-13b-base
export OPENLLM_ENDPOINT=http://localhost:3000
openllm query 'What are large language models?'

Supported models

You can specify any of the following Baichuan models by using --model-id.

Supported backends

  • PyTorch (Default):

    openllm start baichuan-inc/baichuan-13b-base --backend pt

  • vLLM:

    pip install "openllm[baichuan, vllm]"
openllm start baichuan-inc/baichuan-13b-base --backend vllm

[!NOTE] Currently when using the vLLM backend, adapters is yet to be supported.

More models will be integrated with OpenLLM and we welcome your contributions if you want to incorporate your custom LLMs into the ecosystem. Check out Adding a New Model Guide to learn more.

๐Ÿ’ป Run your model on multiple GPUs

OpenLLM allows you to start your model server on multiple GPUs and specify the number of workers per resource assigned using the --workers-per-resource option. For example, if you have 4 available GPUs, you set the value as one divided by the number as only one instance of the Runner server will be spawned.

openllm start facebook/opt-2.7b --workers-per-resource 0.25

[!NOTE] The amount of GPUs required depends on the model size itself. You can use the Model Memory Calculator from Hugging Face to calculate how much vRAM is needed to train and perform big model inference on a model and then plan your GPU strategy based on it.

When using the --workers-per-resource option with the openllm build command, the environment variable is saved into the resulting Bento.

For more information, see Resource scheduling strategy.

๐Ÿ›ž Runtime implementations (Experimental)

Different LLMs may support multiple runtime implementations. Models that have vLLM (vllm) supports will use vLLM by default, otherwise it fallback to use PyTorch (pt).

To specify a specific runtime for your chosen model, use the --backend option. For example:

openllm start meta-llama/Llama-2-7b-chat-hf --backend vllm

Note:

  1. To use the vLLM backend, you need a GPU with at least the Ampere architecture or newer and CUDA version 11.8.
  2. To see the backend options of each model supported by OpenLLM, see the Supported models section or run openllm models.

๐Ÿ“ Quantization

Quantization is a technique to reduce the storage and computation requirements for machine learning models, particularly during inference. By approximating floating-point numbers as integers (quantized values), quantization allows for faster computations, reduced memory footprint, and can make it feasible to deploy large models on resource-constrained devices.

OpenLLM supports the following quantization techniques

PyTorch backend

With PyTorch backend, OpenLLM supports int8, int4, gptq

For using int8 and int4 quantization through bitsandbytes, you can use the following command:

openllm start opt --quantize int8

To run inference withย gptq, simply passย --quantize gptq:

openllm start TheBloke/Llama-2-7B-Chat-GPTQ --quantize gptq

[!NOTE] In order to run GPTQ, make sure you runย pip install "openllm[gptq]" --extra-index-url https://huggingface.github.io/autogptq-index/whl/cu118/ first to install the dependency. From the GPTQ paper, it is recommended to quantized the weights before serving. Seeย AutoGPTQย for more information on GPTQ quantization.

vLLM backend

With vLLM backend, OpenLLM supports awq, squeezellm

To run inference withย awq, simply passย --quantize awq:

openllm start mistral --model-id TheBloke/zephyr-7B-alpha-AWQ --quantize awq

To run inference with squeezellm, simply pass --quantize squeezellm:

openllm start squeeze-ai-lab/sq-llama-2-7b-w4-s0 --quantize squeezellm --serialization legacy

[!IMPORTANT] Since both squeezellm and awq are weight-aware quantization methods, meaning the quantization is done during training, all pre-trained weights needs to get quantized before inference time. Make sure to fine compatible weights on HuggingFace Hub for your model of choice.

๐Ÿ› ๏ธ Serving fine-tuning layers

PEFT, or Parameter-Efficient Fine-Tuning, is a methodology designed to fine-tune pre-trained models more efficiently. Instead of adjusting all model parameters, PEFT focuses on tuning only a subset, reducing computational and storage costs. LoRA (Low-Rank Adaptation) is one of the techniques supported by PEFT. It streamlines fine-tuning by using low-rank decomposition to represent weight updates, thereby drastically reducing the number of trainable parameters.

With OpenLLM, you can take advantage of the fine-tuning feature by serving models with any PEFT-compatible layers using the --adapter-id option. For example:

openllm start opt --model-id facebook/opt-6.7b --adapter-id aarnphm/opt-6-7b-quotes:default

OpenLLM also provides flexibility by supporting adapters from custom file paths:

openllm start opt --model-id facebook/opt-6.7b --adapter-id /path/to/adapters:local_adapter

To use multiple adapters, use the following format:

openllm start opt --model-id facebook/opt-6.7b --adapter-id aarnphm/opt-6.7b-lora:default --adapter-id aarnphm/opt-6.7b-french:french_lora

By default, all adapters will be injected into the models during startup. Adapters can be specified per request via adapter_name:

curl -X 'POST' \
  'http://localhost:3000/v1/generate' \
  -H 'accept: application/json' \
  -H 'Content-Type: application/json' \
  -d '{
  "prompt": "What is the meaning of life?",
  "stop": [
    "philosopher"
  ],
  "llm_config": {
    "max_new_tokens": 256,
    "temperature": 0.75,
    "top_k": 15,
    "top_p": 1
  },
  "adapter_name": "default"
}'

To include this into the Bento, you can specify theย --adapter-idย option when using theย openllm build command:

openllm build facebook/opt-6.7b --adapter-id ...

If you use a relative path for --adapter-id, you need to add --build-ctx.

openllm build facebook/opt-6.7b --adapter-id ./path/to/adapter_id --build-ctx .

[!IMPORTANT] Fine-tuning support is still experimental and currently only works with PyTorch backend. vLLM support is coming soon.

๐Ÿฅ… Playground and Chat UI

The following UIs are currently available for OpenLLM:

UI Owner Type Progress
Clojure @GutZuFusss Community-maintained ๐Ÿ”ง
TS BentoML Team ๐Ÿšง

๐Ÿ Python SDK

Each LLM can be instantiated with openllm.LLM:

import openllm

llm = openllm.LLM('facebook/opt-2.7b')

The main inference API is the streaming generate_iterator method:

async for generation in llm.generate_iterator('What is the meaning of life?'): print(generation.outputs[0].text)

[!NOTE] The motivation behind making llm.generate_iterator an async generator is to provide support for Continuous batching with vLLM backend. By having the async endpoints, each prompt will be added correctly to the request queue to process with vLLM backend.

There is also a one-shot generate method:

await llm.generate('What is the meaning of life?')

This method is easy to use for one-shot generation use case, but merely served as an example how to use llm.generate_iterator as it uses generate_iterator under the hood.

[!IMPORTANT] If you need to call your code in a synchronous context, you can use asyncio.run that wraps an async function:

import asyncio
async def generate(prompt, **attrs): return await llm.generate(prompt, **attrs)
asyncio.run(generate("The meaning of life is", temperature=0.23))

โš™๏ธ Integrations

OpenLLM is not just a standalone product; it's a building block designed to integrate with other powerful tools easily. We currently offer integration with BentoML, LangChain, and Transformers Agents.

BentoML

OpenLLM LLM can be integrated as a Runner in your BentoML service. Simply call await llm.generate to generate text. Note that llm.generate uses runner under the hood:

import bentoml
import openllm

llm = openllm.LLM('facebook/opt-2.7b')

svc = bentoml.Service(name="llm-opt-service", runners=[llm.runner])

@svc.api(input=bentoml.io.Text(), output=bentoml.io.Text())
async def prompt(input_text: str) -> str:
  generation = await llm.generate(input_text)
  return generation.outputs[0].text

LangChain

To quickly start a local LLM with langchain, simply do the following:

from langchain.llms import OpenLLM

llm = OpenLLM(model_name="llama", model_id='meta-llama/Llama-2-7b-hf')

llm("What is the difference between a duck and a goose? And why there are so many Goose in Canada?")

[!IMPORTANT] By default, OpenLLM use safetensors format for saving models. If the model doesn't support safetensors, make sure to pass serialisation="legacy" to use the legacy PyTorch bin format.

langchain.llms.OpenLLM has the capability to interact with remote OpenLLM Server. Given there is an OpenLLM server deployed elsewhere, you can connect to it by specifying its URL:

from langchain.llms import OpenLLM

llm = OpenLLM(server_url='http://44.23.123.1:3000', server_type='grpc')
llm("What is the difference between a duck and a goose? And why there are so many Goose in Canada?")

To integrate a LangChain agent with BentoML, you can do the following:

llm = OpenLLM(
    model_name='flan-t5',
    model_id='google/flan-t5-large',
    embedded=False,
    serialisation="legacy"
)
tools = load_tools(["serpapi", "llm-math"], llm=llm)
agent = initialize_agent(
    tools, llm, agent=AgentType.ZERO_SHOT_REACT_DESCRIPTION
)
svc = bentoml.Service("langchain-openllm", runners=[llm.runner])
@svc.api(input=Text(), output=Text())
def chat(input_text: str):
    return agent.run(input_text)

[!NOTE] You can find out more examples under the examples folder.

Transformers Agents

OpenLLM seamlessly integrates with Transformers Agents.

[!WARNING] The Transformers Agent is still at an experimental stage. It is recommended to install OpenLLM with pip install -r nightly-requirements.txt to get the latest API update for HuggingFace agent.

import transformers

agent = transformers.HfAgent("http://localhost:3000/hf/agent")  # URL that runs the OpenLLM server

agent.run("Is the following `text` positive or negative?", text="I don't like how this models is generate inputs")

Gif showing Agent integration

๐Ÿš€ Deploying models to production

There are several ways to deploy your LLMs:

๐Ÿณ Docker container

  1. Building a Bento: With OpenLLM, you can easily build a Bento for a specific model, like mistralai/Mistral-7B-Instruct-v0.1, using the build command.:

    openllm build mistralai/Mistral-7B-Instruct-v0.1

    A Bento, in BentoML, is the unit of distribution. It packages your program's source code, models, files, artefacts, and dependencies.

  2. Containerize your Bento

    bentoml containerize <name:version>

    This generates a OCI-compatible docker image that can be deployed anywhere docker runs. For best scalability and reliability of your LLM service in production, we recommend deploy with BentoCloudใ€‚

โ˜๏ธ BentoCloud

Deploy OpenLLM with BentoCloud, the serverless cloud for shipping and scaling AI applications.

  1. Create a BentoCloud account: sign up here for early access

  2. Log into your BentoCloud account:

    bentoml cloud login --api-token <your-api-token> --endpoint <bento-cloud-endpoint>

[!NOTE] Replace <your-api-token> and <bento-cloud-endpoint> with your specific API token and the BentoCloud endpoint respectively.

  1. Bulding a Bento: With OpenLLM, you can easily build a Bento for a specific model, such as mistralai/Mistral-7B-Instruct-v0.1:

    openllm build mistralai/Mistral-7B-Instruct-v0.1

  2. Pushing a Bento: Push your freshly-built Bento service to BentoCloud via the push command:

    bentoml push <name:version>

  3. Deploying a Bento: Deploy your LLMs to BentoCloud with a single bentoml deployment create command following the deployment instructions.

๐Ÿ‘ฅ Community

Engage with like-minded individuals passionate about LLMs, AI, and more on our Discord!

OpenLLM is actively maintained by the BentoML team. Feel free to reach out and join us in our pursuit to make LLMs more accessible and easy to use ๐Ÿ‘‰ Join our Slack community!

๐ŸŽ Contributing

We welcome contributions! If you're interested in enhancing OpenLLM's capabilities or have any questions, don't hesitate to reach out in our discord channel.

Checkout our Developer Guide if you wish to contribute to OpenLLM's codebase.

๐Ÿ‡ Telemetry

OpenLLM collects usage data to enhance user experience and improve the product. We only report OpenLLM's internal API calls and ensure maximum privacy by excluding sensitive information. We will never collect user code, model data, or stack traces. For usage tracking, check out the code.

You can opt out of usage tracking by using the --do-not-track CLI option:

openllm [command] --do-not-track

Or by setting the environment variable OPENLLM_DO_NOT_TRACK=True:

export OPENLLM_DO_NOT_TRACK=True

๐Ÿ“” Citation

If you use OpenLLM in your research, we provide a citation to use:

@software{Pham_OpenLLM_Operating_LLMs_2023,
author = {Pham, Aaron and Yang, Chaoyu and Sheng, Sean and  Zhao, Shenyang and Lee, Sauyon and Jiang, Bo and Dong, Fog and Guan, Xipeng and Ming, Frost},
license = {Apache-2.0},
month = jun,
title = {{OpenLLM: Operating LLMs in production}},
url = {https://github.com/bentoml/OpenLLM},
year = {2023}
}

Release Information

Features

  • OpenLLM server now provides a helpers endpoint to help easily create new prompt and other utilities in the future

    /v1/helpers/messages will format a list of messages into the correct chat messages given the chat model #613

  • client now have an additional helpers attribute class to work with helpers endpoint

    client = openllm.HTTPClient()

prompt = client.helpers.messages(
  add_generation_prompt=False,
  messages=[
    {'role': 'system', 'content': 'You are acting as Ernest Hemmingway.'},
    {'role': 'user', 'content': 'Hi there!'},
    {'role': 'assistant', 'content': 'Yes?'},
  ],
)

    Async variant

    client = openllm.AsyncHTTPClient()

prompt = await client.helpers.messages(
  add_generation_prompt=False,
  messages=[
    {'role': 'system', 'content': 'You are acting as Ernest Hemmingway.'},
    {'role': 'user', 'content': 'Hi there!'},
    {'role': 'assistant', 'content': 'Yes?'},
  ],
)

    #615

Click me for full changelog

Project details

Verified details

These details have been verified by PyPI
Maintainers
Avatar for aar0npham from gravatar.com aar0npham Avatar for parano from gravatar.com parano

Unverified details

These details have not been verified by PyPI
Project links
GitHub Statistics

View statistics for this project via Libraries.io, or by using our public dataset on Google BigQuery

Meta

License: Apache Software License

Author: Aaron Pham

Tags AI, Alpaca, BentoML, Falcon, Fine tuning, Generative AI, LLMOps, Large Language Model, Llama 2, MLOps, Model Deployment, Model Serving, PyTorch, Serverless, StableLM, Transformers, Vicuna

Requires: Python >=3.8

Classifiers

Release history Release notifications | RSS feed

0.5.0a12 pre-release

0.5.0a11 pre-release

0.5.0a10 pre-release

0.5.0a9 pre-release

0.5.0a8 pre-release

0.5.0a7 pre-release

0.5.0a6 pre-release

0.5.0a5 pre-release

0.5.0a4 pre-release

0.5.0a3 pre-release

0.5.0a2 pre-release

0.5.0a1 pre-release

0.5.0a0 pre-release

0.4.44

0.4.43

0.4.42

0.4.41

0.4.40

0.4.39

0.4.38

0.4.37

0.4.36

0.4.35

0.4.34

0.4.33

0.4.32

0.4.31

0.4.30

0.4.29

0.4.28

0.4.27

0.4.26

0.4.25

0.4.24

0.4.23

0.4.22

0.4.21

0.4.20

0.4.19

0.4.18

0.4.17

0.4.16

0.4.15

0.4.14

0.4.13

0.4.12

0.4.11

0.4.10

0.4.9

0.4.8

0.4.7

0.4.6

0.4.5

0.4.4

This version

0.4.3

0.4.2

0.4.1

0.4.0

0.3.14

0.3.13

0.3.12

0.3.10

0.3.9

0.3.8

0.3.7

0.3.6

0.3.5

0.3.4

0.3.3

0.3.2

0.3.1

0.3.0

0.2.27

0.2.26

0.2.25

0.2.24

0.2.23

0.2.22

0.2.21 yanked

Reason this release was yanked:

broken client

0.2.20

0.2.19 yanked

Reason this release was yanked:

broken imports from compiled init

0.2.18

0.2.17

0.2.16

0.2.15 yanked

Reason this release was yanked:

include a regression with vllm

0.2.14 yanked

Reason this release was yanked:

include a regression with vllm

0.2.13

0.2.12

0.2.11

0.2.10

0.2.9

0.2.8

0.2.7

0.2.6

0.2.5

0.2.4

0.2.3

0.2.2

0.2.1 yanked

Reason this release was yanked:

Broken installation with openllm[llama]

0.2.0

0.1.20

0.1.19

0.1.18

0.1.17

0.1.16

0.1.15

0.1.14

0.1.13

0.1.12

0.1.11

0.1.10

0.1.9

0.1.8

0.1.7

0.1.6

0.1.5

0.1.4

0.1.3

0.1.2

0.1.1

0.1.0

0.0.34

0.0.33

0.0.32

0.0.31

0.0.30

0.0.29

0.0.28

0.0.27

0.0.26

0.0.25

0.0.24

0.0.23

0.0.22

0.0.21

0.0.19

0.0.18

0.0.17

0.0.16

0.0.15

0.0.14

0.0.13

0.0.12

0.0.11

0.0.10

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

openllm-0.4.3.tar.gz (114.3 kB view hashes)

Uploaded Source

Built Distribution

openllm-0.4.3-py3-none-any.whl (118.5 kB view hashes)

Uploaded Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page