Hi, I have two questions regarding Task 3: 1. The submission guidelines mention that a single unified model is allowed. However, in the baseline, there are three separate YAML files under Task 3 (one for each contrast). Could you clarify whether we are expected to use a unified model or separate configurations per contrast? 2. In the submission folder example structure: ``` └── task3/ # Task 3: Any -> Any (single unified model) — pred/ only, no seg/ ├── 0.1T_to_1.5T/ │ └── pred/ │ ├── T1W/ │ ├── T2W/ │ └── T2FLAIR/ ├── 0.1T_to_3T/ ├── 0.1T_to_5T/ ├── 0.1T_to_7T/ ├── 1.5T_to_7T/ ├── 3T_to_7T/ └── 5T_to_7T/ ``` Does Task 3 only cover low-field to high-field translation (i.e., “any to any” in that direction), or does it also include high-field to low-field cases? Thanks!