This release is a pre-release and may not be stable for production use.
AI Runtime
A versatile sidecar enabling metric standardization, model downloading, and management.
Quick Start
Installation
AI Runtime can be installed by pip.
pip install aibrix
Model download
The AI Runtime supports model downloading from the following storage backends:
- HuggingFace
- S3
- TOS
For more details on model downloading, please refer to our Runtime docs.
Integrate with inference engines
The AI Runtime hides various implementation details on the inference engine side, providing a universal method to guide model management, as well as expose inference monitoring metrics.
At present, vLLM engine is supported, and in the future, SGLang and other inference engines will be supported.
For more details on integration with vLLM, please refer to our Runtime docs.
Contributing
We welcome contributions from the community! Check out our contributing guidelines to see how you can make a difference.
Build from source
# This may take several minutes
pip install -e .
Lint, Format and Type Check
Before contributing your code, please run the following commands to ensure that your code passes the tests and linting checks.
# install dependencies
poetry install --no-root --with dev --extras gpu-optimizer
# linting, formatting and type checking
bash ./scripts/format.sh
License
AI Runtime is licensed under the APACHE License.
1 maintainer
Metadata
1 maintainer
Release files for aibrix 0.8.0rc2
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| aibrix-0.8.0rc2.tar.gz | 449.0 kB | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| aibrix-0.8.0rc2-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 1.0 MB
Release files / aibrix-0.8.0rc2.tar.gz
| Download URL | aibrix-0.8.0rc2.tar.gz |
|---|---|
| Size | 449.0 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
a1c11bb741bfb2b25e18c90c9a842ceaf1cb090f4e635e8301f3e056a92ffa76
|
|
BLAKE2b-256 checksum How to use checksums |
4a0b45f58847e7a6910424da364b26a462b3fc6234549b642babd0b3cf210e5c
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
poetry/2.5.1 CPython/3.12.15 Linux/6.17.0-1022-azure
|
Release files / aibrix-0.8.0rc2-py3-none-any.whl
| Download URL | aibrix-0.8.0rc2-py3-none-any.whl |
|---|---|
| Size | 598.7 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
846a0b16990788797d6887f1398abbe64a2f13bfe002224bff52cccbd41b64d3
|
|
BLAKE2b-256 checksum How to use checksums |
e008661494eaf3febc2fcde986b1276a0ceada39ffe8a29e404c0afea0345044
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
poetry/2.5.1 CPython/3.12.15 Linux/6.17.0-1022-azure
|
