Closed sync-by-unito[bot] closed 1 year ago
➤ Thabied Majal commented:
A baseline one-step image classifier with no hyperparameter tuning was trained, using 2 models: MobileNetv2 and ViT small. The MobileNetv2 model had a stronger performance during the training process.
Below are some notable metrics.
!Screenshot 2023-07-26 at 10.34.07.png|width=1754,height=908!
!Screenshot 2023-07-26 at 11.56.09.png|width=1751,height=905!
!Screenshot 2023-07-26 at 11.56.57.png|width=1770,height=918!
!Screenshot 2023-07-26 at 11.57.37.png|width=1773,height=918!
┆Issue is synchronized with this Jira Task by Unito ┆Attachments: Screenshot 2023-07-26 at 10.34.07.png | Screenshot 2023-07-26 at 11.56.09.png | Screenshot 2023-07-26 at 11.56.57.png | Screenshot 2023-07-26 at 11.57.37.png