Taming Pre-trained LLMs for Generalized Time Series Forecasting through Cross-modal Knowledge Distillation
Before proceeding, ensure Python 3.9 is installed. Install the required dependencies with the following command:
pip install -r requirements.txt
Acquire datasets from Autoformer. Organize them in the ./datasets
directory as shown below:
datasets
├── electricity
│ └── electricity.csv
├── ETT-small
│ ├── ETTh1.csv
│ ├── ETTh2.csv
│ ├── ETTm1.csv
│ └── ETTm2.csv
├── traffic
│ └── traffic.csv
└── weather
└── weather.csv
For short-term forecasting, download the M4 datasets from Time-Series-Library. Place the m4
folder within ./datasets
.
Execute the command below to extract principal components from the word token embeddings:
python pca.py
These embeddings will be saved in ./wte_pca_500.pt
.
Training scripts are located in the ./scripts
folder. For instance, to train the LLaTA model on the ETTh2 dataset for long-term forecasting, execute:
sh scripts/long_term_forecasting/ETTh2.sh
For short-term forecasting, use:
sh scripts/short_term_forecasting/m4.sh
Post-Training:
- Trained models will be saved in
./checkpoints
. - Numerical results are available in
.npy
format under./results
. - Detailed summaries of performance metrics can be found in
./results_{task_name}.txt
.
If this repository contributes to your research, please consider citing our work:
@article{liu2024taming,
title={Taming Pre-trained LLMs for Generalised Time Series Forecasting via Cross-modal Knowledge Distillation},
author={Liu, Peiyuan and Guo, Hang and Dai, Tao and Li, Naiqi and Bao, Jigang and Ren, Xudong and Jiang, Yong and Xia, Shu-Tao},
journal={arXiv preprint arXiv:2403.07300},
year={2024},
arxiv={2403.07300}
}
Our gratitude extends to the authors of the following repositories for their foundational model implementations:
For inquiries or further assistance, contact us at [email protected] or open an issue on this repository.