optimizing-attention-flash

Speed up long-sequence transformer training and inference.

Reference it in any coding agent with:

@skills nousresearch/optimizing-attention-flash