Myla: MY Local Assistants
Self-hosting AI Assistants compatible with OpenAI
Myla stands for MY Local Assistants and is designed and optimized for deploying AI assistants based on large language models (LLMs) in a private environment. Myla provides an API compatible with the OpenAI assistants API, with support for multiple LLM backends. Whether on a laptop or a production server, you can quickly develop and run AI assistants.
Key Features:
- Support for OpenAI API and compatible LLM services
- Assistant API compatible with OpenAI
- Vector retrieval (FAISS/LanceDB)
- sentence_transformers
- WebUI
- Tool extensions
- Document Q&A (in progress)
Quick Start
Installation
Python version requirement: >= 3.9
Myla can be installed from PyPI using pip. It is recommended to create a new virtual environment before installation to avoid conflicts.
pip install myla
If you need Retrieval, please install all dependencies:
pip install "myla[all]"
Configuration
Myla supports using an OpenAI API-compatible LLM service as the backend. You can use the OpenAI API directly or deploy your own local LLM. If you want to deploy a local LLM, it is recommended to use Xorbits Inference.
Create a .env file in the current directory with the following content:
# LLM configuration
LLM_ENDPOINT=https://api.openai.com/v1/
LLM_API_KEY=sk-xx
DEFAULT_LLM_MODEL_NAME=gpt-3.5-turbo
More configurations can be found in: env-example.txt
ChatGLM as backend for your MacBook
Myla supports running ChatGLM locally using chatglm.cpp as the backend. To install the Python Binding, refer to: https://github.com/li-plus/chatglm.cpp#python-binding
.env configuration example:
DEFAULT_LLM_MODEL_NAME=chatglm@/Users/shellc/Workspaces/chatglm.cpp/chatglm-ggml.bin
Start
myla
or
python -m myla
For more startup options:
myla --help
WebUI
Myla provides a simple web interface that makes it easy to develop and debug assistants.
Access from your browser: http://localhost:2000/
API
You can directly use the OpenAI python SDK to access the assistants API.
- API Docs: http://localhost:2000/api/docs
- Swagger: http://localhost:2000/api/swagger
Community
Myla is still under rapid development, and community contributions are welcome.
Metadata
Release files for myla 0.2.40
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| myla-0.2.40-py3-none-any.whl | Python 3 | none | any | Details for myla-0.2.40-py3-none-any.whl |
Release files / myla-0.2.40-py3-none-any.whl
| Download URL | myla-0.2.40-py3-none-any.whl |
|---|---|
| Size | 588.5 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
daf541cce63fbe303de8cee96d1c1994ecd7dcc1fd9ba883f191b95835ba599b
|
|
BLAKE2b-256 checksum How to use checksums |
363e2ba7b674cab48b73023619fd43be82376615d55a913bc11a00cec03b4840
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/5.0.0 CPython/3.11.7
|
Security for myla 0.2.40
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.