cayleypy.models.ModelConfig

class cayleypy.models.ModelConfig

Configuration used to describe ML model.

__init__(model_type: str, input_size: int, num_classes_for_one_hot: int, layers_sizes: list[int], weights_kaggle_id: str | None = None, weights_path: str | None = None) → None

Methods

__init__(model_type, input_size, ...[, ...])
from_dict(cfg)	Creates config from Python dict.
load([device])	Creates model described by this config and loads weights.

Attributes

weights_kaggle_id
weights_path
model_type
input_size
num_classes_for_one_hot
layers_sizes

static from_dict(cfg: dict[str, Any])

Creates config from Python dict.

load(device='cpu') → Module

Creates model described by this config and loads weights.