How to use bidirectional LSTM

[verina01]

Hi,

Thank you for this wonderful package.
I'm new to programming and fastai. Because of my project, I had to jump into time series.

I want to use Bidirectional LSTM for time series forecasting using this package. I have tried and got this error:

len = 3
models = LSTMPlus(c_in=n_input, c_out=n_output, seq_len=len, bidirectional = True)

batch_tfms = TSStandardize()
model = TSForecaster(X, y, splits=splits, batch_tfms=batch_tfms, bs=512, arch=models, metrics=[mse, rmse])
TypeError : 'LSTMPlus' object has no attribute '__name__'

Can you help me to use Bidirectional LSTM for time series forecasting correctly and predict 'targets' with this model ?

[oguiza]

Sorry for my late reply, @verina01.
This issue is fixed now.