SiMBA: Simplified Mamba-based Architecture for Vision and Multivariate Time series

---

Implementation of:

• https://arxiv.org/pdf/2403.15360.pdf

This repository contains the code for the SiMBA (Simplified Mamba-based Architecture for Vision and Multivariate Time series), a transformer-based model that can process both visual and time series inputs.

Installation

To install the required dependencies, run the following command:

coming soon...

Usage

To train the model, run the following command:

python train.py --config config.json

where config.json is the path to the configuration file.

To evaluate the trained model, run the following command:

python evaluate.py --model_path best_model.pth --data_path data/test.json

where best_model.pth is the path to the trained model and data/test.json is the path to the test data.

Configuration

The configuration file is a JSON file that specifies the hyperparameters for the model. The following options are available: coming soon...

Data

coming soon...

Evaluation

coming soon...

License

This code is licensed under the MIT License.