Public Profiles
TensorBoards

Models Client

Importing

from gradient import ModelsClient
api_key='YOUR_API_KEY'
models_client = ModelsClient(api_key)

Models List

List the models created by an experiment or within a project

models = models_client.list(project_id = project_id)
print(models[0])

The model object

Model(id='mo301q57sawrvw', name=None, project_id='prlvf0u05',
experiment_id='esxfbgzu7xanst', tags=None, model_type='Tensorflow',
url='s3://ps-projects/prlvf0u05/esxfbgzu7xanst/model/',
model_path='/storage/models/sdk_test', deployment_state='Stopped',
summary={'accuracy': {'result': {'max': 0.8986999988555908, 'median': 0.8986999988555908,
'min': 0.8986999988555908, 'stddev': 0, 'var': 0, 'mean': 0.8986999988555908},
'scalar': 'accuracy'}, 'loss': {'result': {'max': 0.3866434097290039, 'median': 0.3866434097290039,
'min': 0.3866434097290039, 'stddev': 0, 'var': 0, 'mean': 0.3866434097290039}, 'scalar': 'loss'}},
detail=None)

Sorting models by accuracy is really easy since these are just python objects