File size: 3,470 Bytes
dd2cc16 05a82ef 9d8ea8d aca2b5c 7ddaa6a aca2b5c cfe0f8d aca2b5c cfe0f8d dd2cc16 6307ba7 aca2b5c dd2cc16 79dc62b dd2cc16 bba110f 5251b05 e9982ea 99bde61 bba110f 5d83163 bba110f 35a991e 6307ba7 6a31c14 79303e2 6a31c14 99bde61 a792e0e 99bde61 386b658 456cf7a 99bde61 6176caa 5251b05 99bde61 5d83163 d091651 99bde61 5592854 99bde61 630c277 99bde61 dd2cc16 023bcf2 dd2cc16 d2a36be dd2cc16 97a938f dbe7a49 97a938f |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 |
---
license: apache-2.0
library_name: timer
pipeline_tag: time-series-forecasting
datasets:
- thuml/UTSD
- Salesforce/lotsa_data
metrics:
- mae
- mse
tags:
- time series
- time-series
- forecasting
- foundation models
- pretrained models
- generative models
- time series foundation models
---
# Time-Series Transformer (Timer)
Large time-series model introduced in this [paper](https://arxiv.org/abs/2402.02368) and enhanced with our [further work](https://arxiv.org/abs/2410.04803).
This version is univariate pre-trained on **260B** time points with **84M** parameters, a lightweight generative Transformer for zero-shot point forecasting.
We evaluate the model on the following benchmark: [TSLib Dataset](https://cdn-uploads.huggingface.co/production/uploads/64fbe24a2d20ced4e91de38a/AXhZLVGR8Cnuxe8CVK4Fu.png).
For more information, please see the [GitHub Repo](https://github.com/thuml/Large-Time-Series-Model).
There's indeed room for improvement on this model. We are actively working around it and are glad to see constructive suggestions and noteworthy cases :)
## Quickstart
```
pip install transformers==4.40.1 # Use this version and Python 3.10 for stable compatibility
```
```
import torch
from transformers import AutoModelForCausalLM
# load pretrain model
model = AutoModelForCausalLM.from_pretrained('thuml/timer-base-84m', trust_remote_code=True)
# prepare input
batch_size, lookback_length = 1, 2880
seqs = torch.randn(batch_size, lookback_length)
# generate forecast
prediction_length = 96
output = model.generate(seqs, max_new_tokens=prediction_length)
print(output.shape)
```
A notebook example is also provided [here](https://github.com/thuml/Large-Time-Series-Model/blob/main/examples/quickstart_zero_shot.ipynb). Try it out!
## Specification
* Architecture: Causal Transformer (Decoder-only)
* Pre-training Scale: 260B time points
* Context Length: up to 2880
* Parameter Count: 84M
* Patch Length: 96
* Number of Layers: 8
## Adaptation
For developers interest in fine-tune this model, we provide model checkpoint and code implementation in this [repo](https://github.com/thuml/OpenLTM).
## Acknowledgments
This work was supported by the National Natural Science Foundation of China (62022050 and U2342217), the BNRist Innovation Fund (BNR2024RC01010), and the National Engineering Research Center for Big Data Software.
The model is mostly built from the Internet public time series dataset, which comes from different research teams and providers. We sincerely thank all individuals and organizations who have contributed the data. Without their generous sharing, this model would not have existed.
## Citation
```
@inproceedings{liutimer,
title={Timer: Generative Pre-trained Transformers Are Large Time Series Models},
author={Liu, Yong and Zhang, Haoran and Li, Chenyu and Huang, Xiangdong and Wang, Jianmin and Long, Mingsheng},
booktitle={Forty-first International Conference on Machine Learning}
}
@article{liu2024timer,
title={Timer-XL: Long-Context Transformers for Unified Time Series Forecasting},
author={Liu, Yong and Qin, Guo and Huang, Xiangdong and Wang, Jianmin and Long, Mingsheng},
journal={arXiv preprint arXiv:2410.04803},
year={2024}
}
```
## Contact
If you have any questions or want to use the code, feel free to contact:
* Yong Liu ([email protected])
* Guo Qin ([email protected])
## License
This model is licensed under the Apache-2.0 License. |