Create

Hub¶

Hub provide a unified interface for training, evaluating, exporting and benchmarking. You can create a hub by Hub.new function.

Select Backend¶

In [1]:

from waffle_hub.hub import Hub

Hub.get_available_backends()

from waffle_hub.hub import Hub Hub.get_available_backends()

/home/lhj/anaconda3/envs/waffle/lib/python3.9/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html
  from .autonotebook import tqdm as notebook_tqdm

Out[1]:

['ultralytics', 'autocare_dlt', 'transformers']

In [2]:

Hub.get_available_tasks("ultralytics")

Hub.get_available_tasks("ultralytics")

Out[2]:

[OBJECT_DETECTION, CLASSIFICATION, INSTANCE_SEGMENTATION]

In [4]:

Hub.get_available_model_types("ultralytics", "OBJECT_DETECTION")

Hub.get_available_model_types("ultralytics", "OBJECT_DETECTION")

Out[4]:

['yolov8']

In [5]:

Hub.get_available_model_sizes("ultralytics", "OBJECT_DETECTION", "yolov8")

Hub.get_available_model_sizes("ultralytics", "OBJECT_DETECTION", "yolov8")

Out[5]:

['n', 's', 'm', 'l', 'x']

Create Hub¶

In [6]:

Hub.new(
    name="detector",
    backend="ultralytics",
    task="OBJECT_DETECTION",
    model_type="yolov8",
    model_size="n",
    categories=["1", "2"]
)

Hub.new( name="detector", backend="ultralytics", task="OBJECT_DETECTION", model_type="yolov8", model_size="n", categories=["1", "2"] )

Out[6]:

ModelConfig(name='detector', backend='ultralytics', version='8.0.112', task='OBJECT_DETECTION', model_type='yolov8', model_size='n', categories=[{'supercategory': 'object', 'name': '1'}, {'supercategory': 'object', 'name': '2'}])