CGIAR-Crop-disease

This model is a fine-tuned version of gianlab/swin-tiny-patch4-window7-224-finetuned-plantdisease on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.7438
• Accuracy: 0.6964

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: 0.001
• train_batch_size: 32
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 50
• num_epochs: 40

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.0386	1.0	652	0.9385	0.5669
0.9619	2.0	1304	0.9422	0.5811
0.9193	3.0	1956	0.8806	0.6348
0.8876	4.0	2608	0.8703	0.6488
0.8777	5.0	3260	0.8361	0.6607
0.863	6.0	3912	0.8543	0.6417
0.8316	7.0	4564	0.8101	0.6607
0.8301	8.0	5216	0.8197	0.6609
0.8264	9.0	5868	0.8111	0.6720
0.8283	10.0	6520	0.8065	0.6669
0.816	11.0	7172	0.8115	0.6578
0.8263	12.0	7824	0.8029	0.6753
0.8017	13.0	8476	0.7929	0.6707
0.8005	14.0	9128	0.8025	0.6661
0.7989	15.0	9780	0.8153	0.6594
0.7961	16.0	10432	0.8033	0.6720
0.7769	17.0	11084	0.7879	0.6682
0.7757	18.0	11736	0.7868	0.6732
0.7713	19.0	12388	0.7773	0.6747
0.7638	20.0	13040	0.7678	0.6811
0.7645	21.0	13692	0.7826	0.6795
0.7497	22.0	14344	0.7931	0.6807
0.761	23.0	14996	0.7719	0.6820
0.7486	24.0	15648	0.7641	0.6895
0.7446	25.0	16300	0.7686	0.6832
0.7418	26.0	16952	0.7683	0.6904
0.7344	27.0	17604	0.7549	0.6895
0.7369	28.0	18256	0.7501	0.6891
0.7238	29.0	18908	0.7454	0.6933
0.7264	30.0	19560	0.7565	0.6876
0.7185	31.0	20212	0.7524	0.6880
0.7112	32.0	20864	0.7712	0.6807
0.7073	33.0	21516	0.7532	0.6897
0.7102	34.0	22168	0.7457	0.6960
0.7053	35.0	22820	0.7438	0.6964
0.6979	36.0	23472	0.7449	0.6933
0.6973	37.0	24124	0.7477	0.6929
0.6967	38.0	24776	0.7508	0.6926
0.6939	39.0	25428	0.7481	0.6933
0.6936	40.0	26080	0.7460	0.6968

Framework versions

• Transformers 4.37.1
• Pytorch 2.0.0
• Datasets 2.16.1
• Tokenizers 0.15.0