Where is the full replicate code?

[ieellee]

It seem to be a long time, will the replicate code be published?

[alejandro-lozano-dev]

If you would like to run evaluation experiments this is exactly the code base we used: https://github.com/Ale9806/eVLLM/tree/main

We originally ran Micro-Bench on a locally saved dataset. Given that the dataset has been published as HF, I have updated the code to work directly with HF datasets (and leverage the functionalities, reducing overhead). This should be easier to run