Skip to content

Commit

Permalink
Update Readme.md
Browse files Browse the repository at this point in the history
  • Loading branch information
alanent authored Sep 30, 2023
1 parent 59ecd26 commit f91d238
Showing 1 changed file with 4 additions and 2 deletions.
6 changes: 4 additions & 2 deletions Readme.md
Original file line number Diff line number Diff line change
Expand Up @@ -48,13 +48,15 @@ Exécutez `./scripts/convert_segformers_onnx.py`.
Notre solution est l'inférence simultanée des modèles : `SegFormer-B5 IGB Norm Aerial pseudo labeled` + `SegFormer-B5 RGB Norm Sentinel2`. Une sélection des résultats par classe est effectué dans un deuxième temps. Le temps total d'inférence des modèles sous l'environnement Colab décrit ci-dessous est de 46 minutes et le score obtenu est de 63.55 IoU.


[![Ouvrir dans Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://drive.google.com/file/d/1mM4oTfXj6wthzVneihG-BHHI800BOz_M/view?usp=sharing)


### Environnement d'Exécution

- **Environnement**: Colab Pro
- **GPU**: NVIDIA Tesla T4
- **GPU**: NVIDIA Tesla T4 16 Go
- **RAM**: Haute capacité

[![Ouvrir dans Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://drive.google.com/file/d/1mM4oTfXj6wthzVneihG-BHHI800BOz_M/view?usp=sharing)

### Execution locale

Expand Down

0 comments on commit f91d238

Please sign in to comment.