Metadata-Version: 2.1
Name: refuel
Version: 0.1.2
Summary: Library to log your Machine Learning datasets to Refuel Platform
Home-page: UNKNOWN
Author: Refuel.ai
Author-email: support@refuel.ai
Maintainer: Refuel.ai
Maintainer-email: support@refuel.ai
License: BSD 3-clause
Description: 
        # Refuel Python API
        
        <p align="center"><img height="250" src="https://user-images.githubusercontent.com/1568137/172486199-f4eddb09-7c58-4841-8f3c-60f647079073.png"></p>
        
        This is the [Refuel.ai](https://www.refuel.ai/) Python client library. The primary use for this library is to log your machine learning datasets to the Refuel platform. More features coming soon!
        
        # Installation
        You can install this library using `pip`:
        
        ```bash
        pip install refuel
        ```
        
        # Usage
        
        Make sure you have a valid API key from Refuel (shared with your team during onboarding). Contact us at support@refuel.ai if you need help. 
        
        ## Initialization:
        
        Initialize the Refuel client with your API key. This can be done in one of two ways:
        
        ```python
        import refuel
        
        # Assuming you've set `REFUEL_API_KEY` in your env,
        # init() will pick it up automatically
        refuel_client = refuel.init()
        ```
        
        Alternatively, specify it as an explicit option:
        ```python
        import refuel
        
        options = {
            "api_key": "<YOUR_API_KEY>"
        }
        
        refuel_client = refuel.init(**options)
        ```
        
        ## Logging data
        
        ### Log a single prediction event (no ground truth)
        
        ```python
        refuel_client.log(
            model_name='my-model-name',
            x={'id': 'id1', 'image_url': 's3://<bucket>/<path>', 'embedding': [0.42, -0.13, ...]},
            y_pred={'scores': {'cat': 0.92, 'dog': 0.08}, 'label': 'cat'},
            metadata={'camera_id': 'camera1'}
        )
        ```
        
        ### Log a single prediction event (with ground truth)
        
        ```python
        refuel_client.log(
            model_name='my-model-name',
            x={'id': 'id2', 'image_url': 's3://<bucket>/<path>', 'embedding': [0.35, -0.27, ...]},
            y_pred={'scores': {'cat': 0.12, 'dog': 0.88}, 'label': 'dog'},
            y_true={'label': 'cat'},
            metadata={'camera_id': 'camera1'}
        )
        ```
        
        ### Log a single prediction event (ground truth becomes available at a later point)
        
        ```python
        # This will be joined with the rest of the event logged previously
        refuel_client.log(
            model_name='my-model-name',
            x={'id': 'id1'},
            y_true={'label': 'cat'},
            metadata={'camera_id': 'camera1'}
        )
        ```
        
        ### Log a batch of prediction events
        
        ```python
        # List of events to be logged. 
        events = [{'x': ..., 'y': ..., 'y_true': ..., 'metadata': ...}]
        
        refuel_client.log_batch(
            model_name='my-model-name',
            events=events
        )
        ```
        # Questions?
        
        Reach out to us at support@refuel.ai with any questions!
        
Platform: UNKNOWN
Classifier: Development Status :: 4 - Beta
Classifier: Topic :: Software Development :: Libraries :: Python Modules
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
Classifier: Operating System :: OS Independent
Classifier: Programming Language :: Python
Requires-Python: >=3.6
Description-Content-Type: text/markdown
