Hi,
Your batched pdist implementation looks really nice. I tried building your repository from source but I am getting errors. In your opinion, how easy would it be to integrate as a separate CUDA operation (to avoid building from scratch) ? If it's possible, do you have any recommendations on how to do it ?
Thanks in advance!
Lucas
Hi,
Your batched pdist implementation looks really nice. I tried building your repository from source but I am getting errors. In your opinion, how easy would it be to integrate as a separate CUDA operation (to avoid building from scratch) ? If it's possible, do you have any recommendations on how to do it ?
Thanks in advance!
Lucas