Lorenzo Stella

**DO NOT MERGE: work in progress** This PR cleans and reorganizes the code into **packages**, rather than scattered m-files in `private/` and `library/`. The idea is to organize the code...

Add call to freeze_support in training script

2

*Issue #, if available:* May resolve #155 *Description of changes:* See [here](https://docs.python.org/2/library/multiprocessing.html#windows), and answer [here](https://stackoverflow.com/questions/24374288/where-to-put-freeze-support-in-a-python-script). TODO: - [ ] validate script running fine on windows - [ ] validate script...

How to reproduce training and evaluation as done in the paper?

5

Please check the updated [README](https://github.com/amazon-science/chronos-forecasting). We have also released an evaluation script and backtest configs to compute the WQL and MASE numbers as reported in the paper. The scripts for...

*Issue #, if available:* *Description of changes:* By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice....

Add missing features to PyTorch implementation of DeepAR

1

Pull request #1460 is adding a PyTorch implementation of DeepAR. Some features which were recently added to the equivalente MXNet implementation are not part of the PR: these should be...

*Description of changes:* Backporting fixes - #3117 - #3175 - #3179 By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the...