convnext-base-224_finetuned_on_ImageIn_annotations

This model is a fine-tuned version of facebook/convnext-base-224 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0749
• Precision: 0.9722
• Recall: 0.9811
• F1: 0.9765
• Accuracy: 0.9824

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: 50
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	83	0.1368	0.9748	0.9632	0.9688	0.9772
No log	2.0	166	0.0734	0.9750	0.9727	0.9739	0.9807
No log	3.0	249	0.0693	0.9750	0.9727	0.9739	0.9807
No log	4.0	332	0.0698	0.9750	0.9727	0.9739	0.9807
No log	5.0	415	0.0688	0.9750	0.9727	0.9739	0.9807
No log	6.0	498	0.0690	0.9729	0.9751	0.9740	0.9807
0.0947	7.0	581	0.0666	0.9689	0.9800	0.9743	0.9807
0.0947	8.0	664	0.0642	0.9689	0.9800	0.9743	0.9807
0.0947	9.0	747	0.0790	0.9763	0.9763	0.9763	0.9824
0.0947	10.0	830	0.0813	0.9750	0.9727	0.9739	0.9807
0.0947	11.0	913	0.0797	0.9750	0.9727	0.9739	0.9807
0.0947	12.0	996	0.0791	0.9763	0.9763	0.9763	0.9824
0.0205	13.0	1079	0.0871	0.9750	0.9727	0.9739	0.9807
0.0205	14.0	1162	0.0716	0.9722	0.9811	0.9765	0.9824
0.0205	15.0	1245	0.0746	0.9776	0.9799	0.9787	0.9842
0.0205	16.0	1328	0.0917	0.9738	0.9692	0.9714	0.9789
0.0205	17.0	1411	0.0694	0.9776	0.9799	0.9787	0.9842
0.0205	18.0	1494	0.0697	0.9768	0.9859	0.9812	0.9859
0.0166	19.0	1577	0.0689	0.9702	0.9835	0.9766	0.9824
0.0166	20.0	1660	0.0995	0.9738	0.9692	0.9714	0.9789
0.0166	21.0	1743	0.0847	0.9776	0.9799	0.9787	0.9842
0.0166	22.0	1826	0.0843	0.9776	0.9799	0.9787	0.9842
0.0166	23.0	1909	0.0869	0.9750	0.9727	0.9739	0.9807
0.0166	24.0	1992	0.0762	0.9789	0.9835	0.9811	0.9859
0.0125	25.0	2075	0.0778	0.9789	0.9835	0.9811	0.9859
0.0125	26.0	2158	0.0834	0.9763	0.9763	0.9763	0.9824
0.0125	27.0	2241	0.0818	0.9776	0.9799	0.9787	0.9842
0.0125	28.0	2324	0.0756	0.9684	0.9859	0.9768	0.9824
0.0125	29.0	2407	0.1150	0.9591	0.9824	0.9700	0.9772
0.0125	30.0	2490	0.0781	0.9748	0.9883	0.9813	0.9859
0.0111	31.0	2573	0.0793	0.9716	0.9871	0.9790	0.9842
0.0111	32.0	2656	0.0713	0.9748	0.9883	0.9813	0.9859
0.0111	33.0	2739	0.0802	0.9748	0.9883	0.9813	0.9859
0.0111	34.0	2822	0.0636	0.9802	0.9870	0.9835	0.9877
0.0111	35.0	2905	0.0702	0.9789	0.9835	0.9811	0.9859
0.0111	36.0	2988	0.0773	0.9748	0.9883	0.9813	0.9859
0.0145	37.0	3071	0.0663	0.9781	0.9894	0.9836	0.9877
0.0145	38.0	3154	0.0721	0.9789	0.9835	0.9811	0.9859
0.0145	39.0	3237	0.0708	0.9789	0.9835	0.9811	0.9859
0.0145	40.0	3320	0.0729	0.9748	0.9883	0.9813	0.9859
0.0145	41.0	3403	0.0760	0.9748	0.9883	0.9813	0.9859
0.0145	42.0	3486	0.0771	0.9716	0.9871	0.9790	0.9842
0.0106	43.0	3569	0.0713	0.9748	0.9883	0.9813	0.9859
0.0106	44.0	3652	0.0721	0.9748	0.9883	0.9813	0.9859
0.0106	45.0	3735	0.0732	0.9768	0.9859	0.9812	0.9859
0.0106	46.0	3818	0.0783	0.9789	0.9835	0.9811	0.9859
0.0106	47.0	3901	0.0770	0.9789	0.9835	0.9811	0.9859
0.0106	48.0	3984	0.0744	0.9735	0.9847	0.9789	0.9842
0.0082	49.0	4067	0.0752	0.9722	0.9811	0.9765	0.9824
0.0082	50.0	4150	0.0749	0.9722	0.9811	0.9765	0.9824

Framework versions

• Transformers 4.22.1
• Pytorch 1.12.1+cu113
• Datasets 2.5.1
• Tokenizers 0.12.1