7 Pull requests merged by 6 people

• [ENH] two-way linkage of model package classes and neural network classes

  #1888 merged Jul 15, 2025

• [DOC] fix minor typo in changelog

  #1917 merged Jul 11, 2025

• [BUG] Updated base_model.py to account for importing error

  #1488 merged Jul 10, 2025

• [DOC] 1.1.0 changelog - missing entries

  #1512 merged Jul 10, 2025

• [ENH] Suppress PyTorch deprecation warning: UserWarning: nn.init.constant is now deprecated in favor of nn.init.constant_

  #1915 merged Jul 10, 2025

• [BUG] Correcly set lagged variables to known when lag >= horizon

  #1910 merged Jul 10, 2025

• [DOC] Correct documentation for N-BEATS

  #1914 merged Jul 10, 2025

2 Pull requests opened by 2 people

• [BUG] Device inconstency in `MQF2DistributionLoss` raising: RuntimeError: Expected all tensors to be on the same device

  #1916 opened Jul 10, 2025

• [ENH] Move `BaseFixtureGenerator` to `pytorch-forecasting/tests/_base`

  #1919 opened Jul 14, 2025

2 Issues closed by 1 person

• [BUG] Lagged variables not properly set to known variables when lag >= horizon

  #1909 closed Jul 10, 2025

• [DOC] I think the doc is incorrect in a misleading way - need to confirm

  #1881 closed Jul 10, 2025

2 Issues opened by 1 person

• [ENH] How should Two-way Linkage of `model_pkg` and `model_cls` look like?

  #1920 opened Jul 14, 2025

• [ENH] Add test skip mechanism to test framework model v1

  #1918 opened Jul 13, 2025

11 Unresolved conversations

Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.

• [ENH] Improve test framework for v1 metrics

  #1907 commented on Jul 16, 2025 • 14 new comments

• [ENH] Kolmogorov Arnold Block for NBeats

  #1751 commented on Jul 13, 2025 • 3 new comments

• [BUG] MQF2DistributionLoss - RuntimeError: Expected all tensors to be on the same device

  #1182 commented on Jul 10, 2025 • 0 new comments

• [ENH] relocate testing modules in codebase

  #1913 commented on Jul 10, 2025 • 0 new comments

• [BUG] TFT + categorical features seems not to be compatible with DDP in some situations.

  #1825 commented on Jul 11, 2025 • 0 new comments

• Why is length of test_dataloader 2, but the equal size is 200

  #1726 commented on Jul 13, 2025 • 0 new comments

• TFT predict() with output_dir() running out of memory on large validation sets

  #1461 commented on Jul 15, 2025 • 0 new comments

• [ENH] Improve the test framework of model testing in v1 and make it more robust

  #1906 commented on Jul 16, 2025 • 0 new comments

• [ENH] Precompute data to massively accelerate training by ~2000% in GPU

  #1850 commented on Jul 13, 2025 • 0 new comments

• [ENH] Add warning for multi-target forecasting support in `BaseModel`

  #1880 commented on Jul 15, 2025 • 0 new comments

• [ENH] Improve test framework for v1 models

  #1908 commented on Jul 16, 2025 • 0 new comments