https://github.com/advimman/xeval

Benchmark kit
https://github.com/advimman/xeval

Last synced: about 2 months ago
JSON representation

Benchmark kit

• Host: GitHub
• URL: https://github.com/advimman/xeval
• Owner: advimman
• License: mit
• Created: 2018-11-06T16:20:05.000Z (almost 8 years ago)
• Default Branch: master
• Last Pushed: 2018-11-06T16:24:10.000Z (almost 8 years ago)
• Last Synced: 2025-02-26T10:46:31.599Z (over 1 year ago)
• Language: Python
• Size: 51.8 KB
• Stars: 3
• Watchers: 3
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
# xeval

X-Plane evaluation kit

## DATA Structure

```

Data   

│

└──

│  │

│  └─Airport

│    │

│    └─flight228

│      │

│      └─ X.csv

|      |  data_0.jpg

|      | ...

│      │

│      └─turbulence

│        └─ X.csv

│        │  target.csv

│        │  data_0.jpg

|        |  ...

│    │  ...

│

└──splits.json

```

## Before Evaluation

Launch `scripts/preparator.py` to:

* get dataset feature size

* get supervised tasks names

* get action inds

* update configs:

  * update dataset path

  * generate benchmarks config

## Splits

* Train - 45%

* Validation - 15%

* Test - 40%:

  * Train benchmarks 66%

  * Validation benchmarks 33%