Dummy Torch Model

A small dummy model specifically for unit and integration testing purposes

class flood_forecast.transformer_xl.dummy_torch.DummyTorchModel(forecast_length: int)

__init__(forecast_length: int) → None

A dummy model that will return a tensor of ones (batch_size, forecast_len).

Parameters:

forecast_length (int) – The length to forecast

forward(x: Tensor) → Tensor

The forward pass for the dummy model

Parameters:

• x (torch.Tensor) – Here the data is irrelvant. Only batch_size is grabbed

• mask (torch.Tensor, optional) – [description], defaults to None

Returns:

A tensor with fixed data of one

Return type:

torch.Tensor