ViT-Base-Document-Classifier

This model is a fine-tuned version of google/vit-base-patch16-224 on the None dataset. It achieves the following results on the evaluation set:

Loss: 0.0415
Accuracy: 0.9889
F1: 0.9888
Precision: 0.9888
Recall: 0.9888

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: 5e-05
train_batch_size: 512
eval_batch_size: 512
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 2048
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
training_steps: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
0.0696	1.25	50	0.0566	0.9852	0.9851	0.9852	0.9852
0.0673	2.0	51	0.0549	0.9870	0.9870	0.9870	0.9870
0.0599	2.02	52	0.0545	0.9864	0.9863	0.9863	0.9864
0.0639	2.02	53	0.0551	0.9876	0.9875	0.9875	0.9875
0.0694	2.04	54	0.0539	0.9864	0.9863	0.9863	0.9864
0.0655	2.04	55	0.0528	0.9879	0.9878	0.9878	0.9879
0.0629	2.06	56	0.0519	0.9877	0.9876	0.9876	0.9876
0.0761	2.06	57	0.0532	0.9872	0.9871	0.9871	0.9871
0.0741	2.08	58	0.0524	0.9865	0.9864	0.9864	0.9865
0.0585	2.08	59	0.0519	0.9879	0.9878	0.9878	0.9878
0.0534	2.1	60	0.0504	0.9881	0.9880	0.9880	0.9880
0.056	2.1	61	0.0497	0.9876	0.9875	0.9875	0.9875
0.0588	2.12	62	0.0485	0.9878	0.9877	0.9877	0.9877
0.0554	2.12	63	0.0482	0.9872	0.9871	0.9871	0.9872
0.0674	2.13	64	0.0491	0.9870	0.9870	0.9870	0.9869
0.0613	2.15	65	0.0480	0.9877	0.9876	0.9876	0.9876
0.0688	2.15	66	0.0468	0.9877	0.9876	0.9876	0.9876
0.0677	2.17	67	0.0476	0.9874	0.9874	0.9873	0.9874
0.0598	2.17	68	0.0471	0.9874	0.9873	0.9873	0.9873
0.0658	2.19	69	0.0462	0.9877	0.9876	0.9876	0.9876
0.051	2.19	70	0.0467	0.9880	0.9879	0.9879	0.9879
0.0601	2.21	71	0.0456	0.9881	0.9880	0.9880	0.9880
0.0619	2.21	72	0.0460	0.9879	0.9878	0.9878	0.9879
0.0459	2.23	73	0.0458	0.9883	0.9882	0.9882	0.9883
0.0705	2.23	74	0.0447	0.9884	0.9883	0.9883	0.9883
0.0606	2.25	75	0.0447	0.9878	0.9878	0.9878	0.9878
0.0599	3.0	76	0.0441	0.9887	0.9886	0.9887	0.9886
0.0489	3.01	77	0.0438	0.9886	0.9885	0.9885	0.9885
0.0533	3.02	78	0.0442	0.9883	0.9882	0.9882	0.9883
0.0573	3.03	79	0.0438	0.9880	0.9879	0.9879	0.9880
0.0622	3.04	80	0.0439	0.9886	0.9885	0.9885	0.9885
0.0625	3.05	81	0.0434	0.9881	0.9880	0.9880	0.9880
0.0577	3.06	82	0.0431	0.9886	0.9885	0.9885	0.9885
0.0688	3.07	83	0.0435	0.9887	0.9886	0.9886	0.9887
0.0478	3.08	84	0.0434	0.9889	0.9888	0.9888	0.9888
0.0516	3.09	85	0.0436	0.9888	0.9887	0.9887	0.9887
0.0588	3.1	86	0.0426	0.9889	0.9888	0.9888	0.9888
0.0563	3.11	87	0.0422	0.9889	0.9888	0.9888	0.9888
0.0463	3.12	88	0.0422	0.9886	0.9886	0.9885	0.9886
0.0582	3.13	89	0.0421	0.9887	0.9886	0.9886	0.9887
0.0643	3.14	90	0.0419	0.9891	0.9890	0.9890	0.9891
0.0706	3.15	91	0.0417	0.9892	0.9891	0.9891	0.9891
0.0554	3.16	92	0.0417	0.9892	0.9891	0.9891	0.9891
0.0644	3.17	93	0.0416	0.9890	0.9890	0.9890	0.9890
0.0624	3.18	94	0.0415	0.9893	0.9892	0.9892	0.9892
0.0555	3.19	95	0.0416	0.9886	0.9886	0.9885	0.9886
0.0507	3.2	96	0.0415	0.9889	0.9888	0.9888	0.9888
0.0443	3.21	97	0.0415	0.9889	0.9888	0.9888	0.9888
0.0527	3.22	98	0.0415	0.9889	0.9888	0.9888	0.9888
0.0589	3.23	99	0.0415	0.9889	0.9888	0.9888	0.9888
0.0647	3.24	100	0.0415	0.9889	0.9888	0.9888	0.9888

Framework versions

Transformers 4.39.3
Pytorch 2.2.2
Datasets 2.18.0
Tokenizers 0.15.2

amaye15/ViT-Base-Document-Classifier

ViT-Base-Document-Classifier

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

为此模型做种