Open mmolinar-stanford opened 4 months ago
Thank you for your attention. We have fully open-sourced it and provided tutorials. You are welcome to try it out!
Thank you for your response. Could you please guide me on where I can find tutorials for running the code?
I noticed that the pretrain section seems to only include a basic Transformer network without any attention weights exploration, as mentioned in the paper. I was wondering if the absence of attention weights in the provided code was intentional or if there's an alternative implementation I might have missed.
Additionally, I couldn't find the code for DistriXplorer and SCM, which play a crucial role according to your paper. Are these implementations available separately or integrated within the current codebase? If I overlooked something, I'd appreciate any guidance or instructions on running the code.
Also, it would be awesome if you could share the model and data files required to reproduce the results mentioned in your paper. Having access to these resources would greatly aid in understanding and replicating your work.
Thanks for your time. I'm looking forward to your response and any help you can provide in addressing these queries.