metadata
tags:
- traffic-forecasting
- time-series
- graph-neural-network
- stgformer_depthwise
datasets:
- metr-la
Spatial-Temporal Graph Transformer (Depthwise) - METR-LA
Spatial-Temporal Graph Transformer (Depthwise) (STGFORMER_DEPTHWISE) trained on METR-LA dataset for traffic speed forecasting.
Model Description
STGFormer with Depthwise Separable Conv temporal processing (fast alternative to Transformer)
Dataset
METR-LA: Traffic speed data from highway sensors.
Usage
from utils.stgformer import load_from_hub
# Load model from Hub
model, scaler = load_from_hub("METR-LA", hf_repo_prefix="STGFORMER_DEPTHWISE")
# Get predictions
from utils.stgformer import get_predictions
predictions = get_predictions(model, scaler, test_dataset)
Training
Model was trained using the STGFORMER_DEPTHWISE implementation with default hyperparameters.
Citation
If you use this model, please cite the original STGFORMER_DEPTHWISE paper:
@inproceedings{lan2022stgformer,
title={STGformer: Spatial-Temporal Graph Transformer for Traffic Forecasting},
author={Lan, Shengnan and Ma, Yong and Huang, Weijia and Wang, Wanwei and Yang, Hui and Li, Peng},
booktitle={IEEE Transactions on Neural Networks and Learning Systems},
year={2022}
}
License
This model checkpoint is released under the same license as the training code.