Why is number of samples in the forecast limited to 100?

[Sharma96]

I am using DeepAREstimator to predict a time series, the training_data has 900 target values, the training code:

model = DeepAREstimator(
    prediction_length=1, freq="D", trainer_kwargs={"max_epochs": 5}
).train(training_data)

But, while predicting the values for the same dataset:

forecasts = list(model.predict(training_data))

The length of samples in forecasts[0] is just 100, how can I get the forecasts for each value in the training set(I was expecting 900 as the length of samples).