Quickstart
Install the Zeno Python package.
pip install zenoml
Getting Started with Zeno Build
The best way to get started with Zeno is with Zeno Build, a Python library that makes it easy to set up Zeno projects for common AI and ML tasks.
Follow the Zeno Build Tutorial to learn the basic Zeno concepts on an example LLM task.
You can follow the Using Zeno tutorial to learn more about what you can accomplish with Zeno!
Running Zeno
Command Line
Zeno includes a command line helper to set up a new project which guides you through creating the TOML configuration file:
zeno init
You can then run Zeno with
zeno config.toml
Jupyter Notebooks
Import zeno with
import zeno from zeno
Create a Pandas DataFrame with your metadata.
Run zeno with a dict of options:
zeno({
"view": "image-classification",
"metadata": my_dataframe,
"data_path": "path/to/data",
"data_column": "image_path",
})
Adding Models and Metrics
To add model predictions and metrics, you can write functions with the Python API wrappers.
If using the command line, create a folder with a Python file including these functions and add the folder as an option to the TOML file:
functions = "path/to/functions"
In Jupyter Notebooks, you can pass a list of functions to the functions
option:
zeno({
"functions": [model_function, accuracy_function]
})