Scalable and user friendly neural forecasting algorithms.
APACHE-2.0 License
Bot releases are hidden (Show)
Published by FedericoGarza about 1 year ago
horizon_weight
parameter to losses and BasePointLoss
in https://github.com/Nixtla/neuralforecast/pull/704
horizon_weight
in https://github.com/Nixtla/neuralforecast/pull/706
Full Changelog: https://github.com/Nixtla/neuralforecast/compare/v1.6.1...v1.6.2
Published by FedericoGarza over 1 year ago
Full Changelog: https://github.com/Nixtla/neuralforecast/compare/v1.4.0...v1.5.0
Published by FedericoGarza over 1 year ago
Recurrent models (RNN, LSTM, GRU, DilatedRNN) can now take static, historical, and future exogenous variables. These variables are combined with lags to produce "context" vectors based on MLP decoders, based on the MQ-RNN model (https://arxiv.org/pdf/1711.11053.pdf).
The new DistributionLoss
class allows for producing probabilistic forecasts with all available models. By changing the loss
hyperparameter to one of these losses, the model will learn and output the distribution parameters:
predict
method can return samples, quantiles, or distribution parameters.sCRPS loss in PyTorch to minimize errors generating prediction intervals.
We included new optimization features commonly used to train neural models:
torch.optim.lr_scheduler.StepLR
scheduler. The new num_lr_decays
hyperparameter controls the number of decays (evenly distributed) during training.early_stop_patience_steps
controls the number of validation steps with no improvement after which training will be stopped.Training, scheduler, validation loss computation, and early stopping are now defined in steps (instead of epochs) to control the training procedure better. Use max_steps
to define the number of training iterations. Note: max_epochs
will be deprecated in the future.
Published by cchallu almost 2 years ago
Full Changelog: https://github.com/Nixtla/neuralforecast/compare/v1.2.0...v1.3.0
Published by FedericoGarza almost 2 years ago
languages
in https://github.com/Nixtla/neuralforecast/pull/355
num_samples
to Distribution's initialization in https://github.com/Nixtla/neuralforecast/pull/359
Full Changelog: https://github.com/Nixtla/neuralforecast/compare/v1.1.0...v1.2.0
Published by FedericoGarza almost 2 years ago
torch.nn.module
classes in https://github.com/Nixtla/neuralforecast/pull/311
Full Changelog: https://github.com/Nixtla/neuralforecast/compare/v1.0.0...v1.1.0
Published by FedericoGarza about 2 years ago
Full Changelog: https://github.com/Nixtla/neuralforecast/compare/v0.1.0...v1.0.0
Published by FedericoGarza over 2 years ago
Full Changelog: https://github.com/Nixtla/neuralforecast/compare/v0.0.9...v0.1.0
Published by FedericoGarza over 2 years ago
Full Changelog: https://github.com/Nixtla/neuralforecast/compare/v0.0.8...v0.0.9
Published by FedericoGarza over 2 years ago
Full Changelog: https://github.com/Nixtla/neuralforecast/compare/v0.0.7...v0.0.8
Published by FedericoGarza over 2 years ago
auto
ml pipeline.RNN
model.