swin-tiny-patch4-window7-224-swinnn

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

• Loss: 0.0883
• Accuracy: 0.8232

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: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.2802	0.9979	351	0.2783	0.3222
0.2702	1.9986	703	0.2652	0.376
0.2565	2.9993	1055	0.2474	0.431
0.2448	4.0	1407	0.2358	0.4558
0.2433	4.9979	1758	0.2223	0.4994
0.2095	5.9986	2110	0.2058	0.5434
0.2197	6.9993	2462	0.1963	0.568
0.2093	8.0	2814	0.1906	0.5764
0.2047	8.9979	3165	0.1888	0.5874
0.1952	9.9986	3517	0.1743	0.6192
0.1926	10.9993	3869	0.1740	0.6234
0.1838	12.0	4221	0.1667	0.6448
0.1822	12.9979	4572	0.1629	0.6468
0.1838	13.9986	4924	0.1587	0.6638
0.1689	14.9993	5276	0.1563	0.675
0.1697	16.0	5628	0.1472	0.6916
0.1643	16.9979	5979	0.1435	0.6912
0.1655	17.9986	6331	0.1395	0.706
0.1555	18.9993	6683	0.1371	0.714
0.1577	20.0	7035	0.1321	0.7258
0.1575	20.9979	7386	0.1318	0.7284
0.141	21.9986	7738	0.1228	0.7438
0.151	22.9993	8090	0.1260	0.7392
0.1403	24.0	8442	0.1178	0.7558
0.1434	24.9979	8793	0.1185	0.7534
0.1465	25.9986	9145	0.1162	0.759
0.1362	26.9993	9497	0.1121	0.769
0.138	28.0	9849	0.1099	0.769
0.1293	28.9979	10200	0.1094	0.7754
0.1273	29.9986	10552	0.1091	0.7768
0.1363	30.9993	10904	0.1078	0.7766
0.1293	32.0	11256	0.1091	0.7736
0.1275	32.9979	11607	0.1068	0.7806
0.1263	33.9986	11959	0.1040	0.7888
0.1243	34.9993	12311	0.1019	0.7954
0.1237	36.0	12663	0.1016	0.7958
0.1243	36.9979	13014	0.0993	0.7988
0.1194	37.9986	13366	0.1011	0.7986
0.1213	38.9993	13718	0.0959	0.8064
0.1155	40.0	14070	0.0942	0.8108
0.1179	40.9979	14421	0.0950	0.8072
0.1057	41.9986	14773	0.0924	0.8166
0.1042	42.9993	15125	0.0924	0.8152
0.1151	44.0	15477	0.0928	0.8132
0.1122	44.9979	15828	0.0920	0.8146
0.11	45.9986	16180	0.0906	0.8152
0.1096	46.9993	16532	0.0894	0.82
0.1082	48.0	16884	0.0885	0.821
0.108	48.9979	17235	0.0886	0.8204
0.112	49.8934	17550	0.0883	0.8232

Framework versions

• Transformers 4.46.3
• Pytorch 2.1.0a0+32f93b1
• Datasets 3.1.0
• Tokenizers 0.20.3