X-Plane evaluation kit
Data
│
└──
│ │
│ └─Airport
│ │
│ └─flight228
│ │
│ └─ X.csv
| | data_0.jpg
| | ...
│ │
│ └─turbulence
│ └─ X.csv
│ │ target.csv
│ │ data_0.jpg
| | ...
│ │ ...
│
└──splits.json
Launch scripts/preparator.py
to:
- get dataset feature size
- get supervised tasks names
- get action inds
- update configs:
- update dataset path
- generate benchmarks config
- Train - 45%
- Validation - 15%
- Test - 40%:
- Train benchmarks 66%
- Validation benchmarks 33%