Published January 16, 2024
| Version v1.0
Software
Open
An actor-model framework for visual sensory encoding
Description
This record contains the source code used in the article:
"Leong, F., Rahmani, B., Psaltis, D. et al. An actor-model framework for visual sensory encoding. Nat Commun 15, 808 (2024).
https://doi.org/10.1038/s41467-024-45105-5"
To train the forward model, run train_fwd.py. To train the actor network, run train_actor.py. The optimized model is provided to generate the figures found in the article. To understand the data analysis pipeline and generate the figures presented in the article, look at gen_fig_X.py.
Software-requirements: The pipeline is written fully in python. The requirements file for the libraries used is provided. To install the necessary libraries, run pip install -r requirements.txt
Hardware-requirements: Our customed-scripts have been executed on an NVIDIA RTX 3090 GPU. Be advised that execution of the code without GPU-backed machines (CPU only) can considerably increases the training time.
Files
lne-lab/actor-retina-v1.0.zip
Additional details
Related works
- Is supplemented by
- Dataset: 10.5281/zenodo.10515301 (DOI)