An open API service indexing awesome lists of open source software.

https://github.com/lkuich/coach-python

Coach Client Library for Python
https://github.com/lkuich/coach-python

computer-vision image-classification image-recognition ml python tensorflow

Last synced: 8 days ago
JSON representation

Coach Client Library for Python

Awesome Lists containing this project

README

          

# Coach Python SDK

Coach is an end-to-end Image Recognition platform, we provide the tooling to do effective data collection, training, and on-device parsing of Image Recognition models.

See https://coach.lkuich.com for more information!

## Installing
Install and update using [pip](https://pip.pypa.io/en/stable/quickstart/):
```bash
pip install coach-ml
```

## Usage
Coach can be initialized 2 different ways. If you are only using the offline model parsing capabilities and already have a model package on disk, you can initialize like so:

```python
coach = CoachClient()

# We already had the `flowers` model on disk, no need to authenticate:
result = coach.get_model('flowers').predict('rose.jpg')
```

However, in order to download your trained models, you must authenticate with your API key:
```python
coach = CoachClient().login('myapikey')

# Now that we're authenticated, we can cache our models for future use:
coach.cache_model('flowers')

# Evaluate with our cached model:
result = coach.get_model('flowers').predict('rose.jpg')
```

## API Breakdown

### CoachClient
`__init__(is_debug=False)`
Optional `is_debug`, if `True`, additional logs will be displayed

`login(apiKey) -> CoachClient`
Authenticates with Coach service and allows for model caching. Accepts API Key as its only parameter. Returns its own instance.

`cache_model(model_name, path='.', skip_match=False, model_type='frozen')`
Downloads model from Coach service to disk. Specify the name of the model, and the path to store it. This will create a new directory in the specified path and store any model related documents there.
By default, if a model already exists with the same version, in the same path, caching will be skipped. Set `skip_match` to `False` to override this behaviour.
`model_type` can be one of: `frozen`, `unity`, `mobile`, and can be useful if you're interested in caching a specific version of your model.

`get_model(path='.') -> CoachModel`
Loads model into memory. Specify the path of the cached models directory. Returns a `CoachModel`

`get_model_remote(model_name, path='.') -> CoachModel`
Downloads and loads model into memory. Specify the path of the cached models directory. Returns a `CoachModel`

### CoachModel
`__init__(graph, labels, base_module)`
Initializes a new instance of `CoachModel`, accepts a loaded `tf.Graph()`, array of `labels`, and the `base_module` the graph was trained off of.

`predict(image, input_name="input", output_name="output") -> dict`
Specify the directory of an image file or the image as a byte array. Parses the specified image into memory and runs it through the loaded model. Returns a dict of its predictions in order of confidence.
If you have a pretrained frozen graph with different Tensor input/output names, you can specify them with `input_name` and `output_name` respectfully.