convnext-tiny-224-finetuned-barkley

This model is a fine-tuned version of facebook/convnext-tiny-224 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0128
• Precision: 1.0
• Recall: 1.0
• F1: 1.0
• Accuracy: 1.0
• Top1 Accuracy: 1.0
• Error Rate: 0.0

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: 0.0002
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 30
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy	Top1 Accuracy	Error Rate
1.6288	1.0	38	1.6005	0.2133	0.2697	0.2043	0.2371	0.2697	0.7629
1.6059	2.0	76	1.5802	0.2384	0.2763	0.2243	0.2473	0.2763	0.7527
1.5808	3.0	114	1.5570	0.2778	0.3026	0.2595	0.2744	0.3026	0.7256
1.5555	4.0	152	1.5291	0.3831	0.375	0.3491	0.3511	0.375	0.6489
1.5232	5.0	190	1.4933	0.4252	0.4408	0.4154	0.4147	0.4408	0.5853
1.4784	6.0	228	1.4484	0.5076	0.5197	0.4926	0.4972	0.5197	0.5028
1.4242	7.0	266	1.3902	0.6857	0.6382	0.6307	0.6249	0.6382	0.3751
1.3586	8.0	304	1.3186	0.7728	0.7171	0.7166	0.7134	0.7171	0.2866
1.276	9.0	342	1.2236	0.8547	0.8026	0.8109	0.8060	0.8026	0.1940
1.1778	10.0	380	1.1122	0.8899	0.8553	0.8609	0.8601	0.8553	0.1399
1.0543	11.0	418	0.9839	0.9064	0.8947	0.8958	0.9005	0.8947	0.0995
0.921	12.0	456	0.8418	0.9541	0.9539	0.9537	0.9575	0.9539	0.0425
0.773	13.0	494	0.6935	0.9624	0.9605	0.9605	0.9652	0.9605	0.0348
0.6204	14.0	532	0.5515	0.9688	0.9671	0.9672	0.9708	0.9671	0.0292
0.4835	15.0	570	0.4146	0.9704	0.9671	0.9676	0.9697	0.9671	0.0303
0.3641	16.0	608	0.3043	0.9805	0.9803	0.9802	0.9830	0.9803	0.0170
0.2706	17.0	646	0.2247	0.9805	0.9803	0.9802	0.9830	0.9803	0.0170
0.1998	18.0	684	0.1705	0.9873	0.9868	0.9868	0.9889	0.9868	0.0111
0.1446	19.0	722	0.1271	0.9937	0.9934	0.9934	0.9944	0.9934	0.0056
0.1106	20.0	760	0.1047	0.9873	0.9868	0.9868	0.9889	0.9868	0.0111
0.0872	21.0	798	0.0780	0.9937	0.9934	0.9934	0.9944	0.9934	0.0056
0.0614	22.0	836	0.0739	0.9873	0.9868	0.9868	0.9889	0.9868	0.0111
0.0491	23.0	874	0.0517	0.9937	0.9934	0.9934	0.9944	0.9934	0.0056
0.0365	24.0	912	0.0401	0.9871	0.9868	0.9868	0.9878	0.9868	0.0122
0.0255	25.0	950	0.0336	0.9937	0.9934	0.9934	0.9944	0.9934	0.0056
0.0212	26.0	988	0.0377	0.9873	0.9868	0.9868	0.9889	0.9868	0.0111
0.0175	27.0	1026	0.0195	0.9937	0.9934	0.9934	0.9944	0.9934	0.0056
0.0125	28.0	1064	0.0214	0.9936	0.9934	0.9934	0.9933	0.9934	0.0067
0.0155	29.0	1102	0.0128	1.0	1.0	1.0	1.0	1.0	0.0
0.0104	30.0	1140	0.0159	0.9937	0.9934	0.9934	0.9944	0.9934	0.0056

Framework versions

• Transformers 4.44.2
• Pytorch 2.3.1+cu121
• Datasets 3.0.1
• Tokenizers 0.19.1