All-Plants-18-Epochs-Model

This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the image_folder dataset. It achieves the following results on the evaluation set:

• Loss: 0.0888
• Accuracy: 0.9848
• F1: 0.9849

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 2e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 18

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
0.9212	1.0	407	0.3931	0.9501	0.9579
0.2659	2.0	814	0.2176	0.9668	0.9674
0.137	3.0	1221	0.1481	0.9723	0.9731
0.0865	4.0	1628	0.1043	0.9834	0.9836
0.0557	5.0	2035	0.0888	0.9848	0.9849
0.0408	6.0	2442	0.0839	0.9848	0.9848
0.0289	7.0	2849	0.0920	0.9848	0.9849
0.0229	8.0	3256	0.0817	0.9834	0.9837
0.0175	9.0	3663	0.0890	0.9820	0.9823
0.0156	10.0	4070	0.0966	0.9820	0.9823
0.0121	11.0	4477	0.0809	0.9834	0.9837
0.0102	12.0	4884	0.0875	0.9820	0.9823
0.0086	13.0	5291	0.0873	0.9820	0.9823
0.0077	14.0	5698	0.0860	0.9820	0.9823
0.0068	15.0	6105	0.0876	0.9820	0.9823
0.0062	16.0	6512	0.0896	0.9820	0.9823
0.0059	17.0	6919	0.0890	0.9820	0.9823
0.0056	18.0	7326	0.0894	0.9820	0.9823

Framework versions

• Transformers 4.35.0
• Pytorch 2.0.0
• Datasets 2.1.0
• Tokenizers 0.14.1