Hello fellow Zindians

I’m seeking clarification about the following requirement

*"To reflect the realities of deploying AI in low-resource environments, your solution must run on a single NVIDIA T4 GPU with ≤16 GB RAM, such as what’s available through Google Colab’s free tier."*

Question: Does this constraint apply to:

1. Both fine-tuning/training and inference? (i.e., the entire pipeline must run on a T4), or
2. Only inference (after training on more powerful hardware)?

I want to ensure my approach aligns with the rules. If anyone has insights or has asked organizers directly, I’d appreciate your input!