explainable_embedding

Experiments with regard to explanation of embedded spaces and multi modal models.

Setup

Clone the explainable_embedding repository, cd into it and install the additional dependencies for these experiments:

pip install -r requirements.txt

Choose experiments

In the benchmarking folder, you can edit distance_benchmark.py to set which experiments to run. Since we are currently still actively using this repo for running experiments, the file you downloaded may not be set to run all experiments. The code at the bottom of the file should speak for itself. To run all experiments used for the paper, import all_configs from .distance_benchmark_configs and modify the loop to go over all_configs.

Run experiments

Run experiments with

python -m benchmarking.distance_benchmark

A directory called output will be created with the results.