README.md

📊 Time series analysis leaderboard is now available on our 🤗 Hugging Face space. Discover the performance of different models!

---

This is also the official repository for the following paper:

• Learning Recursive Multi-Scale Representations for Irregular Multivariate Time Series Forecasting (ICLR 2026) [poster] [OpenReview] [arXiv]

  @inproceedings{li_LearningRecursiveMultiScale_2026,
      author = {Li, Boyuan and Liu, Zhen and Luo, Yicheng  and Ma, Qianli},
      booktitle = {International Conference on Learning Representations},
      title = {Learning Recursive Multi-Scale Representations for Irregular Multivariate Time Series Forecasting},
      year = {2026}
  }
  

• HyperIMTS: Hypergraph Neural Network for Irregular Multivariate Time Series Forecasting (ICML 2025) [poster] [OpenReview] [arXiv]

  @inproceedings{li_HyperIMTSHypergraphNeural_2025,
      author = {Li, Boyuan and Luo, Yicheng and Liu, Zhen and Zheng, Junhao and Lv, Jianming and Ma, Qianli},
      booktitle = {Forty-Second International Conference on Machine Learning},
      title = {HyperIMTS: Hypergraph Neural Network for Irregular Multivariate Time Series Forecasting},
      year = {2025}
  }
  

1. ✨ Hightlighted Features

• Extensibility: Adapt your model/dataset once, train almost any combination of "model" $\times$ "dataset" $\times$ "loss function".
• Compatibility: Accept models with any number/type of arguments in forward; Accept datasets with any number/type of return values in __getitem__; Accept tailored loss calculation for specific models.
• Maintainability: No need to worry about breaking the training codes of existing models/datasets/loss functions when adding new ones.
• Reproducibility: Minimal library dependencies for core components. Try the best to get rid of fancy third-party libraries (e.g., PyTorch Lightning, EasyTorch).
• Efficiency: Multi-GPU parallel training; Python built-in logger; structured experimental result saving (json)...
• Transferability: Even if you don't like our framework, you can still easily find and copy the models/datasets you want. No overwhelming encapsulation.

2. 🧭 Documentation

Checkout the new documentation website.

Using 🦞agent? Check out our official PyOmniTS skill on clawhub. Your agent will understand the essentials of our framework, and even automate the code replication process by adapting other papers' codes into PyOmniTS!

3. 🤖 Models

53 models, covering regular, irregular, pretrained, and traffic models, have been included in PyOmniTS, and more are coming.

Model classes can be found in models/, and their dependencies can be found in layers/

• ✅: supported
• ❌: not supported
• '-': not implemented
• MTS: regularly sampled multivariate time series
• IMTS: able to handle irregularly sampled multivariate time series

Model	Venue	Type	Forecasting	Classification	Imputation	Available Versions
Ada-MSHyper	NeurIPS 2024	MTS	✅	✅	✅	v1.0.0+
APN	AAAI 2026	IMTS	✅	-	-	v1.3.2+
ASTGI	ICLR 2026	IMTS	✅	-	-	v2.0.1+
Autoformer	NeurIPS 2021	MTS	✅	✅	✅	v1.0.0+
Scaleformer	ICLR 2023	MTS	✅	-	✅	v2.0.0+
BigST	VLDB 2024	MTS	✅	✅	✅	v1.0.0+
Crossformer	ICLR 2023	MTS	✅	✅	✅	v1.0.0+
CRU	ICML 2022	IMTS	✅	❌	✅	v1.0.0+
DLinear	AAAI 2023	MTS	✅	✅	✅	v1.0.0+
ETSformer	arXiv 2022	MTS	✅	✅	✅	v1.0.0+
FEDformer	ICML 2022	MTS	✅	✅	✅	v1.0.0+
FiLM	NeurIPS 2022	MTS	✅	✅	✅	v1.0.0+
FourierGNN	NeurIPS 2023	MTS	✅	✅	✅	v1.0.0+
FreTS	NeurIPS 2023	MTS	✅	✅	✅	v1.0.0+
GNeuralFlow	NeurIPS 2024	IMTS	✅	❌	✅	v1.0.0+
GraFITi	AAAI 2024	IMTS	✅	✅	✅	v1.0.0+
GRU-D	Scientific Reports 2018	IMTS	✅	✅	✅	v1.0.0+
HD-TTS	ICML 2024	IMTS	✅	-	✅	v2.0.0+
Hi-Patch	ICML 2025	IMTS	✅	✅	✅	v1.0.0+
higp	ICML 2024	MTS	✅	✅	✅	v1.0.0+
HyperIMTS (Ours)	ICML 2025	IMTS	✅	-	✅	v1.0.0+
Informer	AAAI 2021	MTS	✅	✅	✅	v1.0.0+
iTransformer	ICLR 2024	MTS	✅	✅	✅	v1.0.0+
Koopa	NeurIPS 2023	MTS	✅	❌	✅	v1.0.0+
Latent_ODE	NeurIPS 2019	IMTS	✅	❌	✅	v1.0.0+
Leddam	ICML 2024	MTS	✅	✅	✅	v1.0.0+
LightTS	arXiv 2022	MTS	✅	✅	✅	v1.0.0+
Mamba	Language Modeling 2024	MTS	✅	✅	✅	v1.0.0+
MICN	ICLR 2023	MTS	✅	✅	✅	v1.0.0+
MOIRAI	ICML 2024	Any	✅	-	✅	v1.0.0+
mTAN	ICLR 2021	IMTS	✅	✅	✅	v1.0.0+
NeuralFlows	NeurIPS 2021	IMTS	✅	❌	✅	v1.0.0+
NHITS	AAAI 2023	MTS	✅	-	✅	v2.0.0+
Nonstationary Transformer	NeurIPS 2022	MTS	✅	✅	✅	v1.0.0+
PatchTST	ICLR 2023	MTS	✅	✅	✅	v1.0.0+
Pathformer	ICLR 2024	MTS	✅	-	✅	v2.0.0+
PrimeNet	AAAI 2023	IMTS	✅	✅	✅	v1.0.0+
Pyraformer	ICLR 2022	MTS	✅	✅	✅	v2.0.0+
Raindrop	ICLR 2022	IMTS	✅	✅	✅	v1.0.0+
Reformer	ICLR 2020	MTS	✅	✅	✅	v1.0.0+
ReIMTS (Ours)	ICLR 2026	IMTS	✅	✅	-	v2.0.0+
SeFT	ICML 2020	IMTS	✅	✅	✅	v1.0.0+
SegRNN	arXiv 2023	MTS	✅	✅	✅	v1.0.0+
Temporal Fusion Transformer	arXiv 2019	MTS	✅	-	-	v1.0.0+
TFMixer	ICML 2026	IMTS	✅	-	-	v2.0.1+
TiDE	TMLR 2023	MTS	✅	✅	✅	v1.0.0+
TimeCHEAT	AAAI 2025	MTS	✅	✅	✅	v1.0.0+
TimeMixer	ICLR 2024	MTS	✅	✅	✅	v1.0.0+
TimesNet	ICLR 2023	MTS	✅	✅	✅	v1.0.0+
tPatchGNN	ICML 2024	IMTS	✅	✅	✅	v1.0.0+
Transformer	NeurIPS 2017	MTS	✅	✅	✅	v1.0.0+
TSMixer	TMLR 2023	MTS	✅	✅	✅	v1.0.0+
Warpformer	KDD 2023	IMTS	✅	✅	✅	v1.0.0+

4. 💾 Datasets

Dataest classes are put in data/data_provider/datasets, and dependencies can be found in data/dependencies:

11 datasets, covering regular and irregular ones, have been included in PyOmniTS, and more are coming.

• ✅: supported
• ❌: not supported
• '-': not implemented
• MTS: regularly sampled multivariate time series
• IMTS: irregularly sampled multivariate time series

Dataset	Type	Field	Forecasting
ECL	MTS	electricity	✅
ETTh1	MTS	electricity	✅
ETTm1	MTS	electricity	✅
Human Activity	IMTS	biomechanics	✅
ILI	MTS	healthcare	✅
MIMIC III	IMTS	healthcare	✅
MIMIC IV	IMTS	healthcare	✅
PhysioNet'12	IMTS	healthcare	✅
Traffic	MTS	traffic	✅
USHCN	IMTS	weather	✅
Weather	MTS	weather	✅

Datasets for classification and imputation have not released yet.

5. 📉 Loss Functions

The following loss functions are included under loss_fns/:

Loss Function	Task	Note
CrossEntropyLoss	Classification	-
MAE	Forecasting/Imputation	-
ModelProvidedLoss	-	Some models prefer to calculate loss within forward(), such as GNeuralFlows.
MSE_Dual	Forecasting/Imputation
MSE	Forecasting/Imputation	-

6. 🚧 Roadmap

PyOmniTS is continously evolving:

• More tutorials.
• Classification support in core components.
• Imputation support in core components.
• Optional python package management via uv.

Yet Another Code Framework?

We encountered the following problems when using existing ones:

• Argument & return value chaos for models' forward():

  Different models usually take varying number and shape of arguments, especially ones from different domains. Changes to training logic are needed to support these differences.

• Return value chaos for datasets' __getitem__():

  datasets can return a number of tensors in different shapes, which have to be aligned with arguments of models' forward() one by one. Changes to training logic are also needed to support these differences.

• Argument & return value chaos for loss functions' forward():

  loss functions take different types of tensors as input, require aligning with return values from models' forward().

• Overwhelming dependencies:

  some existing pipelines use fancy high-level packages in building the pipeline, which can lower the flexibility of code modification.

Contributors

Ladbaby 💻 🐛

Acknowledgement

• Time Series Library: Models and datasets for regularly sampled time series are mostly adapted from it.
• BasicTS: Documentation design reference.
• Google Gemini: Icon creation.