Leezekun / ViTST

[NeurIPS 2023] The official repo for the paper: "Time Series as Images: Vision Transformer for Irregularly Sampled Time Series"."

https://arxiv.org/abs/2303.12799

Leezekun/ViTST Issues

How do I convert data into images?
Updated 2 months ago
Reported metrics for RAINDROP method is very different from that in the original paper RAINDROP
Closed 3 months ago2
Hyperparameters for regular time series dataset
Updated 4 months ago
Can not reproduce regular time series datasets results
Closed 4 months ago
Getting Error: ImportError: cannot import name 'color_detailed_description' from 'configs'
Updated 5 months ago1
A code error for translating static features to text features
Closed 5 months ago
The performance of baselines are different from the paper
Closed 5 months ago
How to adapt framework for different type of data?
Updated 8 months ago2
Can't run the script"run_VisionTextCLS.py"
Updated 8 months ago1
Why does it keep hogging GPU memory constantly?
Closed 10 months ago
Hello , I'm here again. In readme, You can download the datasets at http://www.timeseriesclassification.com/Downloads/Archives/Multivariate2018_ts.zip. When I enter the page,it coms many options, which should I choose?
Updated a year ago2
how many pictures will P12 produce? I have run the code for about an hour and it has generated 4800 pictures.
Updated a year ago1
The tokenizer is just for analyzing the length of static features. It is not necessary. You can remove the code related to the tokenizer if you have trouble loading it.
Updated a year ago1
ValueError: Unrecognized model identifier in allenai/longformer-base-4096. Should contains one of 'bert', 'openai-gpt', 'gpt2', 'transfo-xl', 'xlnet', 'xlm', 'roberta', 'ctrl'
Updated a year ago5
patient ids in P19 dataset
Closed a year ago1