File size: 3,024 Bytes
bbe123c
 
 
 
 
 
 
 
 
 
 
 
8aa33a2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
bbe123c
 
 
 
8aa33a2
 
 
 
 
bbe123c
 
8aa33a2
bbe123c
 
 
 
 
 
 
8aa33a2
bbe123c
 
8aa33a2
bbe123c
 
8aa33a2
 
bbe123c
 
 
 
8aa33a2
bbe123c
 
 
 
 
 
 
 
 
 
 
 
 
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
---
license: apache-2.0
pipeline_tag: time-series-forecasting
tags:
  - time series
  - forecasting
  - pretrained models
  - foundation models
  - time series foundation models
  - time-series
---

# Chronos⚡️-Base

Chronos⚡️ (read: Chronos-Bolt) is a family of pretrained time series forecasting models which can be used for zero-shot forecasting. Chronos⚡️ models are based on the [T5 architecture](https://arxiv.org/abs/1910.10683) and are available in the following sizes.


<div align="center">

| Model                                                                  | Parameters | Based on                                                               |
| ---------------------------------------------------------------------- | ---------- | ---------------------------------------------------------------------- |
| [**chronos-bolt-tiny**](https://huggingface.co/autogluon/chronos-bolt-tiny)   | 9M         | [t5-efficient-tiny](https://huggingface.co/google/t5-efficient-tiny)   |
| [**chronos-bolt-mini**](https://huggingface.co/autogluon/chronos-bolt-mini)   | 21M        | [t5-efficient-mini](https://huggingface.co/google/t5-efficient-mini)   |
| [**chronos-bolt-small**](https://huggingface.co/autogluon/chronos-bolt-small) | 48M        | [t5-efficient-small](https://huggingface.co/google/t5-efficient-small) |
| [**chronos-bolt-base**](https://huggingface.co/autogluon/chronos-bolt-base)   | 205M       | [t5-efficient-base](https://huggingface.co/google/t5-efficient-base)   |

</div>


## Usage

> [!WARNING]  
> Chronos⚡️ models will be available in the next stable release of AutoGluon, so the following instructions will only work once AutoGluon 1.2 has been released.


A minimal example showing how to perform zero-shot inference using Chronos⚡️ with AutoGluon:

```
pip install autogluon
```

```python
from autogluon.timeseries import TimeSeriesPredictor, TimeSeriesDataFrame

df = TimeSeriesDataFrame("https://autogluon.s3.amazonaws.com/datasets/timeseries/m4_hourly/train.csv")

predictor = TimeSeriesPredictor(prediction_length=48).fit(
    df,
    hyperparameters={
        "Chronos": {"model_path": "autogluon/chronos-bolt-base"},
    },
)

predictions = predictor.predict(df)
```

## Citation

If you find Chronos or Chronos⚡️ models useful for your research, please consider citing the associated [paper](https://arxiv.org/abs/2403.07815):

```
@article{ansari2024chronos,
  author  = {Ansari, Abdul Fatir and Stella, Lorenzo and Turkmen, Caner and Zhang, Xiyuan, and Mercado, Pedro and Shen, Huibin and Shchur, Oleksandr and Rangapuram, Syama Syndar and Pineda Arango, Sebastian and Kapoor, Shubham and Zschiegner, Jasper and Maddix, Danielle C. and Mahoney, Michael W. and Torkkola, Kari and Gordon Wilson, Andrew and Bohlke-Schneider, Michael and Wang, Yuyang},
  title   = {Chronos: Learning the Language of Time Series},
  journal = {arXiv preprint arXiv:2403.07815},
  year    = {2024}
}
```

## License

This project is licensed under the Apache-2.0 License.