convnextv2-tiny-1k-224-finetuned-pattern-edge

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: 0.9202
• Accuracy: 0.81

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
2.302	0.9912	28	2.2666	0.1575
2.2226	1.9823	56	2.1654	0.315
2.0639	2.9735	84	1.9970	0.445
1.8559	4.0	113	1.7373	0.56
1.5966	4.9912	141	1.4823	0.605
1.3967	5.9823	169	1.2925	0.6125
1.204	6.9735	197	1.0512	0.68
1.0206	8.0	226	0.9307	0.7025
0.9408	8.9912	254	0.8286	0.7425
0.8501	9.9823	282	0.8590	0.6975
0.7545	10.9735	310	0.7702	0.7475
0.7484	12.0	339	0.7739	0.765
0.6909	12.9912	367	0.7344	0.75
0.6558	13.9823	395	0.6874	0.775
0.5923	14.9735	423	0.6641	0.7675
0.5764	16.0	452	0.6110	0.7925
0.5235	16.9912	480	0.6806	0.76
0.4883	17.9823	508	0.7903	0.76
0.4682	18.9735	536	0.6469	0.7825
0.441	20.0	565	0.6694	0.7825
0.4201	20.9912	593	0.7145	0.7625
0.387	21.9823	621	0.6505	0.7775
0.4034	22.9735	649	0.6169	0.7875
0.3041	24.0	678	0.6416	0.795
0.3021	24.9912	706	0.6992	0.775
0.2853	25.9823	734	0.6566	0.7975
0.27	26.9735	762	0.6970	0.7825
0.2722	28.0	791	0.6863	0.785
0.2143	28.9912	819	0.6794	0.795
0.2238	29.9823	847	0.6782	0.7975
0.2387	30.9735	875	0.6945	0.81
0.223	32.0	904	0.7377	0.7825
0.2211	32.9912	932	0.7431	0.7775
0.1882	33.9823	960	0.7029	0.815
0.1562	34.9735	988	0.6887	0.815
0.1689	36.0	1017	0.7190	0.7975
0.1886	36.9912	1045	0.7678	0.795
0.1887	37.9823	1073	0.7334	0.81
0.1531	38.9735	1101	0.7359	0.7925
0.1662	40.0	1130	0.7594	0.8075
0.1273	40.9912	1158	0.7342	0.81
0.1986	41.9823	1186	0.7781	0.805
0.1891	42.9735	1214	0.7376	0.8225
0.1573	44.0	1243	0.7304	0.815
0.1536	44.9912	1271	0.7773	0.8
0.1562	45.9823	1299	0.7623	0.8
0.1264	46.9735	1327	0.8314	0.7925
0.1596	48.0	1356	0.7831	0.8175
0.1237	48.9912	1384	0.7949	0.8
0.1355	49.9823	1412	0.7813	0.795
0.1251	50.9735	1440	0.7647	0.81
0.1181	52.0	1469	0.7552	0.8175
0.1224	52.9912	1497	0.8346	0.795
0.1201	53.9823	1525	0.7741	0.7975
0.1109	54.9735	1553	0.7724	0.785
0.1084	56.0	1582	0.7904	0.805
0.1187	56.9912	1610	0.7424	0.8125
0.0935	57.9823	1638	0.7411	0.815
0.1023	58.9735	1666	0.7476	0.81
0.1166	60.0	1695	0.7742	0.8175
0.099	60.9912	1723	0.7697	0.815
0.1157	61.9823	1751	0.8538	0.8
0.1137	62.9735	1779	0.8545	0.8125
0.094	64.0	1808	0.8463	0.7925
0.1161	64.9912	1836	0.8351	0.81
0.08	65.9823	1864	0.8610	0.7925
0.0799	66.9735	1892	0.8593	0.8075
0.0783	68.0	1921	0.8423	0.815
0.0851	68.9912	1949	0.8265	0.82
0.0775	69.9823	1977	0.8708	0.805
0.0902	70.9735	2005	0.8181	0.81
0.0904	72.0	2034	0.8297	0.82
0.0898	72.9912	2062	0.8464	0.82
0.1013	73.9823	2090	0.8325	0.81
0.0726	74.9735	2118	0.8772	0.8
0.0745	76.0	2147	0.8505	0.8125
0.0891	76.9912	2175	0.8694	0.81
0.0791	77.9823	2203	0.8766	0.81
0.0639	78.9735	2231	0.8462	0.8125
0.0676	80.0	2260	0.8991	0.8075
0.0904	80.9912	2288	0.8551	0.815
0.0788	81.9823	2316	0.9302	0.795
0.0787	82.9735	2344	0.8706	0.8025
0.0918	84.0	2373	0.8680	0.805
0.0681	84.9912	2401	0.8481	0.8125
0.115	85.9823	2429	0.8553	0.8025
0.0599	86.9735	2457	0.8887	0.805
0.0774	88.0	2486	0.9255	0.81
0.0701	88.9912	2514	0.8795	0.81
0.074	89.9823	2542	0.8634	0.8175
0.0497	90.9735	2570	0.8793	0.82
0.0569	92.0	2599	0.9007	0.7925
0.0722	92.9912	2627	0.8701	0.815
0.0674	93.9823	2655	0.8880	0.8225
0.0643	94.9735	2683	0.8855	0.8075
0.0583	96.0	2712	0.8918	0.815
0.0558	96.9912	2740	0.8736	0.8275
0.0622	97.9823	2768	0.9058	0.815
0.0689	98.9735	2796	0.9007	0.8075
0.0782	100.0	2825	0.9216	0.8025
0.0696	100.9912	2853	0.9159	0.8075
0.0554	101.9823	2881	0.9195	0.8125
0.0585	102.9735	2909	0.9314	0.8125
0.0541	104.0	2938	0.8939	0.825
0.0636	104.9912	2966	0.9045	0.8025
0.0684	105.9823	2994	0.8892	0.8075
0.0608	106.9735	3022	0.8999	0.8075
0.0663	108.0	3051	0.9033	0.8075
0.054	108.9912	3079	0.9249	0.805
0.0538	109.9823	3107	0.9065	0.81
0.0696	110.9735	3135	0.9002	0.8175
0.0585	112.0	3164	0.9106	0.8025
0.0641	112.9912	3192	0.9088	0.81
0.0611	113.9823	3220	0.9152	0.8075
0.0528	114.9735	3248	0.9140	0.8125
0.0631	116.0	3277	0.9184	0.81
0.0744	116.9912	3305	0.9216	0.8125
0.0407	117.9823	3333	0.9211	0.8125
0.0573	118.9381	3360	0.9202	0.81

Framework versions

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