convnextv2-tiny-1k-224-finetuned-topwear

This model is a fine-tuned version of facebook/convnextv2-tiny-1k-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.6478
• Accuracy: 0.8389

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: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 120

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.7006	0.9412	12	2.6782	0.1167
2.6863	1.9608	25	2.6272	0.1611
2.6437	2.9804	38	2.5389	0.2889
2.4851	4.0	51	2.4116	0.4111
2.3732	4.9412	63	2.2707	0.4889
2.2546	5.9608	76	2.0710	0.5722
2.1023	6.9804	89	1.8371	0.6167
1.7115	8.0	102	1.6161	0.6111
1.5295	8.9412	114	1.4381	0.6278
1.3366	9.9608	127	1.2540	0.65
1.0556	10.9804	140	1.1632	0.6611
0.9657	12.0	153	1.0600	0.7
0.8703	12.9412	165	0.9983	0.7222
0.8007	13.9608	178	0.9474	0.7278
0.6398	14.9804	191	0.8634	0.75
0.6023	16.0	204	0.8527	0.7278
0.583	16.9412	216	0.7928	0.7667
0.5279	17.9608	229	0.7897	0.7833
0.4643	18.9804	242	0.7886	0.7667
0.4296	20.0	255	0.7329	0.7833
0.41	20.9412	267	0.7317	0.7611
0.3674	21.9608	280	0.7171	0.7667
0.3285	22.9804	293	0.7005	0.7833
0.2978	24.0	306	0.6576	0.7889
0.293	24.9412	318	0.6450	0.8
0.2724	25.9608	331	0.6765	0.7889
0.2494	26.9804	344	0.6826	0.8056
0.2504	28.0	357	0.6710	0.8056
0.2332	28.9412	369	0.6667	0.7778
0.2012	29.9608	382	0.7399	0.7944
0.1866	30.9804	395	0.7311	0.7833
0.2031	32.0	408	0.7077	0.7944
0.1969	32.9412	420	0.7769	0.7667
0.1968	33.9608	433	0.7666	0.7833
0.1712	34.9804	446	0.6796	0.8
0.1813	36.0	459	0.6654	0.8111
0.1678	36.9412	471	0.6851	0.7889
0.1461	37.9608	484	0.7054	0.7833
0.1244	38.9804	497	0.7013	0.8056
0.1329	40.0	510	0.6785	0.8
0.1186	40.9412	522	0.7500	0.7778
0.1397	41.9608	535	0.6819	0.8167
0.1324	42.9804	548	0.6257	0.8111
0.111	44.0	561	0.5939	0.8278
0.1228	44.9412	573	0.6379	0.8222
0.1085	45.9608	586	0.6789	0.8222
0.1234	46.9804	599	0.6241	0.8278
0.1129	48.0	612	0.7503	0.7889
0.1197	48.9412	624	0.6862	0.7944
0.0898	49.9608	637	0.6764	0.7889
0.1057	50.9804	650	0.6339	0.8167
0.0893	52.0	663	0.5828	0.85
0.0736	52.9412	675	0.6573	0.8111
0.0752	53.9608	688	0.6806	0.7944
0.1127	54.9804	701	0.6222	0.8111
0.1126	56.0	714	0.6305	0.8167
0.0874	56.9412	726	0.6593	0.8111
0.0806	57.9608	739	0.7006	0.8167
0.0978	58.9804	752	0.6680	0.8056
0.0875	60.0	765	0.6739	0.8167
0.0722	60.9412	777	0.6341	0.8333
0.0942	61.9608	790	0.6428	0.8
0.0957	62.9804	803	0.6758	0.8
0.0814	64.0	816	0.6104	0.8167
0.077	64.9412	828	0.6226	0.8111
0.1004	65.9608	841	0.6899	0.8056
0.0697	66.9804	854	0.7105	0.8167
0.0754	68.0	867	0.6751	0.8111
0.0842	68.9412	879	0.6912	0.7833
0.0684	69.9608	892	0.7235	0.8167
0.0684	70.9804	905	0.5840	0.8278
0.0705	72.0	918	0.6636	0.8222
0.0681	72.9412	930	0.6787	0.8
0.0906	73.9608	943	0.6243	0.8389
0.0453	74.9804	956	0.6787	0.8222
0.0874	76.0	969	0.6259	0.8278
0.051	76.9412	981	0.6590	0.8278
0.0858	77.9608	994	0.6307	0.8278
0.0601	78.9804	1007	0.6042	0.8444
0.0601	80.0	1020	0.5875	0.8389
0.067	80.9412	1032	0.6078	0.8389
0.0556	81.9608	1045	0.6007	0.8444
0.0661	82.9804	1058	0.6062	0.8333
0.0651	84.0	1071	0.6387	0.8111
0.0546	84.9412	1083	0.6861	0.8167
0.0827	85.9608	1096	0.6073	0.8389
0.052	86.9804	1109	0.5935	0.85
0.0524	88.0	1122	0.5899	0.8389
0.066	88.9412	1134	0.5954	0.8444
0.0617	89.9608	1147	0.6145	0.8444
0.0572	90.9804	1160	0.6176	0.8444
0.0719	92.0	1173	0.6406	0.8278
0.0734	92.9412	1185	0.6485	0.8333
0.0616	93.9608	1198	0.6198	0.8333
0.0557	94.9804	1211	0.6167	0.8389
0.0494	96.0	1224	0.6480	0.8444
0.0587	96.9412	1236	0.6076	0.85
0.052	97.9608	1249	0.6512	0.8389
0.0383	98.9804	1262	0.6782	0.8333
0.0499	100.0	1275	0.6542	0.8278
0.0511	100.9412	1287	0.6795	0.8389
0.0452	101.9608	1300	0.6740	0.8333
0.0475	102.9804	1313	0.6616	0.8389
0.0455	104.0	1326	0.6490	0.8278
0.0486	104.9412	1338	0.6331	0.8333
0.0585	105.9608	1351	0.6299	0.8333
0.0549	106.9804	1364	0.6398	0.8278
0.0436	108.0	1377	0.6338	0.8444
0.0429	108.9412	1389	0.6459	0.8389
0.0449	109.9608	1402	0.6470	0.8444
0.0559	110.9804	1415	0.6463	0.8389
0.0378	112.0	1428	0.6480	0.8389
0.0476	112.9412	1440	0.6478	0.8389

Framework versions

• Transformers 4.44.0
• Pytorch 2.4.0
• Datasets 2.21.0
• Tokenizers 0.19.1