convnextv2-tiny-1k-224-finetuned-print

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: 0.1811
• Accuracy: 0.9263

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: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	0.8889	6	1.6653	0.1158
1.6915	1.9259	13	1.5942	0.2316
1.5895	2.9630	20	1.4918	0.3895
1.5895	4.0	27	1.3637	0.5053
1.433	4.8889	33	1.2241	0.6737
1.2158	5.9259	40	1.0279	0.7474
1.2158	6.9630	47	0.8303	0.8105
0.9438	8.0	54	0.6756	0.8211
0.727	8.8889	60	0.5751	0.8421
0.727	9.9259	67	0.4679	0.8632
0.5516	10.9630	74	0.4432	0.8526
0.4337	12.0	81	0.3815	0.8526
0.4337	12.8889	87	0.3641	0.8842
0.3757	13.9259	94	0.3233	0.8737
0.3017	14.9630	101	0.3230	0.9158
0.3017	16.0	108	0.3018	0.8842
0.2495	16.8889	114	0.3445	0.9053
0.2177	17.9259	121	0.2987	0.8947
0.2177	18.9630	128	0.2727	0.8947
0.1738	20.0	135	0.2865	0.8842
0.1572	20.8889	141	0.2646	0.9263
0.1572	21.9259	148	0.3100	0.9053
0.1165	22.9630	155	0.3039	0.9263
0.1057	24.0	162	0.3023	0.9053
0.1057	24.8889	168	0.2254	0.9158
0.0825	25.9259	175	0.3308	0.8737
0.0795	26.9630	182	0.2040	0.9368
0.0795	28.0	189	0.2148	0.9263
0.072	28.8889	195	0.3450	0.8632
0.0701	29.9259	202	0.2418	0.9263
0.0701	30.9630	209	0.2495	0.9263
0.0635	32.0	216	0.3267	0.8947
0.0537	32.8889	222	0.3728	0.9158
0.0537	33.9259	229	0.2852	0.9053
0.0607	34.9630	236	0.2386	0.9474
0.052	36.0	243	0.2070	0.9158
0.052	36.8889	249	0.1860	0.9474
0.049	37.9259	256	0.3069	0.8947
0.0578	38.9630	263	0.4477	0.8737
0.0533	40.0	270	0.2612	0.8947
0.0533	40.8889	276	0.2649	0.8842
0.0505	41.9259	283	0.1950	0.9263
0.0433	42.9630	290	0.2903	0.8842
0.0433	44.0	297	0.2526	0.9368
0.0395	44.8889	303	0.3016	0.8842
0.035	45.9259	310	0.3509	0.8947
0.035	46.9630	317	0.2943	0.8842
0.0335	48.0	324	0.2613	0.8842
0.0408	48.8889	330	0.2165	0.9158
0.0408	49.9259	337	0.2872	0.9263
0.0244	50.9630	344	0.3134	0.8842
0.0323	52.0	351	0.3006	0.9158
0.0323	52.8889	357	0.3758	0.8737
0.0241	53.9259	364	0.3033	0.9263
0.0193	54.9630	371	0.2741	0.9368
0.0193	56.0	378	0.1684	0.9368
0.0273	56.8889	384	0.2403	0.9474
0.0244	57.9259	391	0.1500	0.9474
0.0244	58.9630	398	0.1377	0.9368
0.0268	60.0	405	0.1898	0.9158
0.0405	60.8889	411	0.1756	0.9053
0.0405	61.9259	418	0.1907	0.9263
0.0219	62.9630	425	0.1790	0.9053
0.0329	64.0	432	0.1885	0.9368
0.0329	64.8889	438	0.1550	0.9368
0.019	65.9259	445	0.1811	0.9158
0.0205	66.9630	452	0.2166	0.9263
0.0205	68.0	459	0.1701	0.9053
0.0232	68.8889	465	0.2153	0.9158
0.0269	69.9259	472	0.2229	0.9263
0.0269	70.9630	479	0.2237	0.9263
0.0306	72.0	486	0.1828	0.9368
0.0298	72.8889	492	0.1448	0.9368
0.0298	73.9259	499	0.1948	0.9158
0.0154	74.9630	506	0.2570	0.9158
0.0193	76.0	513	0.2462	0.9263
0.0193	76.8889	519	0.2194	0.9158
0.0188	77.9259	526	0.2254	0.9158
0.0198	78.9630	533	0.1924	0.9263
0.0147	80.0	540	0.1525	0.9368
0.0147	80.8889	546	0.1314	0.9474
0.0282	81.9259	553	0.1381	0.9368
0.0168	82.9630	560	0.1496	0.9158
0.0168	84.0	567	0.1806	0.9263
0.018	84.8889	573	0.2080	0.9263
0.0172	85.9259	580	0.2199	0.9158
0.0172	86.9630	587	0.1939	0.9263
0.0117	88.0	594	0.1815	0.9263
0.0149	88.8889	600	0.1811	0.9263

Framework versions

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