resnet50_fold_4

This model is a fine-tuned version of microsoft/resnet-50 on the imagefolder dataset. It achieves the following results on the evaluation set:

Loss: 1.1418
Accuracy: 0.9689
F1 Score: 0.9722
Recall: 0.9716

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: 0.0001
train_batch_size: 64
eval_batch_size: 64
seed: 42
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: cosine
lr_scheduler_warmup_steps: 150
num_epochs: 100
label_smoothing_factor: 0.15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1 Score	Recall
2.7671	1.0	19	2.7678	0.2318	0.1882	0.2166
2.7503	2.0	38	2.7545	0.3426	0.3034	0.3082
2.7213	3.0	57	2.7309	0.3668	0.2407	0.2890
2.6794	4.0	76	2.6964	0.4325	0.3279	0.3555
2.5996	5.0	95	2.6350	0.4291	0.3107	0.3550
2.4985	6.0	114	2.5509	0.5398	0.4577	0.5065
2.3794	7.0	133	2.4167	0.6471	0.6204	0.6289
2.2129	8.0	152	2.2265	0.7509	0.7466	0.7396
1.9935	9.0	171	1.9928	0.7958	0.8015	0.7969
1.7990	10.0	190	1.6869	0.8616	0.8718	0.8666
1.6583	11.0	209	1.4996	0.8927	0.9026	0.8976
1.4808	12.0	228	1.4386	0.8962	0.9024	0.9029
1.3637	13.0	247	1.3121	0.9239	0.9314	0.9314
1.3261	14.0	266	1.2538	0.9308	0.9374	0.9386
1.2501	15.0	285	1.2300	0.9412	0.9465	0.9460
1.2732	16.0	304	1.2137	0.9481	0.9532	0.9509
1.1900	17.0	323	1.1924	0.9550	0.9598	0.9581
1.1741	18.0	342	1.1971	0.9550	0.9595	0.9584
1.1796	19.0	361	1.1691	0.9619	0.9656	0.9645
1.2468	20.0	380	1.1596	0.9585	0.9628	0.9617
1.1665	21.0	399	1.1725	0.9585	0.9630	0.9642
1.1717	22.0	418	1.1619	0.9516	0.9565	0.9559
1.1338	23.0	437	1.1772	0.9412	0.9476	0.9507
1.1823	24.0	456	1.1483	0.9619	0.9660	0.9654
1.1496	25.0	475	1.1418	0.9689	0.9722	0.9716
1.1255	26.0	494	1.1412	0.9619	0.9660	0.9643
1.1738	27.0	513	1.1333	0.9619	0.9661	0.9656
1.1220	28.0	532	1.1453	0.9654	0.9692	0.9680
1.1044	29.0	551	1.1490	0.9654	0.9692	0.9680
1.2009	30.0	570	1.1470	0.9619	0.9660	0.9643
1.0877	31.0	589	1.1416	0.9654	0.9686	0.9680
1.0595	32.0	608	1.1410	0.9654	0.9689	0.9666
1.1055	33.0	627	1.1317	0.9619	0.9656	0.9656
1.1170	34.0	646	1.1390	0.9654	0.9686	0.9680
1.0753	35.0	665	1.1486	0.9654	0.9686	0.9680
1.0905	36.0	684	1.1484	0.9619	0.9656	0.9656
1.0513	37.0	703	1.1434	0.9654	0.9686	0.9680
1.0587	38.0	722	1.1543	0.9654	0.9686	0.9680
1.0972	39.0	741	1.1305	0.9654	0.9686	0.9680
1.1102	40.0	760	1.1498	0.9654	0.9685	0.9668
1.0822	41.0	779	1.1352	0.9619	0.9656	0.9656
1.0650	42.0	798	1.1278	0.9654	0.9686	0.9680
1.0829	43.0	817	1.1359	0.9689	0.9716	0.9705
1.0483	44.0	836	1.1264	0.9619	0.9656	0.9656
1.0582	45.0	855	1.1225	0.9654	0.9686	0.9691
1.0203	46.0	874	1.1401	0.9654	0.9685	0.9668
1.0798	47.0	893	1.1233	0.9689	0.9716	0.9705
1.0345	48.0	912	1.1356	0.9654	0.9686	0.9680
1.0466	49.0	931	1.1199	0.9689	0.9721	0.9727
1.0794	50.0	950	1.1331	0.9654	0.9686	0.9680
1.0989	51.0	969	1.1303	0.9619	0.9656	0.9656
1.0443	52.0	988	1.1192	0.9654	0.9686	0.9680
1.0481	53.0	1007	1.1297	0.9654	0.9686	0.9680
1.0449	54.0	1026	1.1304	0.9654	0.9686	0.9680
1.0478	55.0	1045	1.1142	0.9654	0.9686	0.9680
1.0319	56.0	1064	1.1359	0.9689	0.9716	0.9705
1.0219	57.0	1083	1.1302	0.9689	0.9716	0.9705
1.0564	58.0	1102	1.1215	0.9654	0.9686	0.9680
1.0231	59.0	1121	1.1234	0.9689	0.9716	0.9705
1.1116	60.0	1140	1.1276	0.9585	0.9625	0.9631
1.0473	61.0	1159	1.1263	0.9654	0.9686	0.9680
1.0148	62.0	1178	1.1277	0.9654	0.9686	0.9680
1.0230	63.0	1197	1.1273	0.9654	0.9686	0.9680
1.0277	64.0	1216	1.1233	0.9654	0.9686	0.9680
1.0271	65.0	1235	1.1232	0.9689	0.9716	0.9705

Framework versions

Transformers 5.0.0
Pytorch 2.10.0+cu128
Datasets 5.0.0
Tokenizers 0.22.2

nqvii/resnet50_fold_4

resnet50_fold_4

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions