swin-tiny-patch4-window7-224-dmae-va-da2-80

This model is a fine-tuned version of microsoft/swin-tiny-patch4-window7-224 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.2098
• Accuracy: 0.9483

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: 80

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	0.94	4	1.3020	0.3793
No log	1.88	8	1.2779	0.3966
1.3725	2.82	12	1.3148	0.3276
1.3725	4.0	17	1.3035	0.3793
1.202	4.94	21	1.1813	0.4828
1.202	5.88	25	1.0447	0.5172
1.202	6.82	29	1.0149	0.5172
1.0137	8.0	34	0.9113	0.5345
1.0137	8.94	38	0.8100	0.5690
0.7855	9.88	42	0.7788	0.6552
0.7855	10.82	46	0.6669	0.6724
0.6434	12.0	51	0.6208	0.7586
0.6434	12.94	55	0.5012	0.8276
0.6434	13.88	59	0.5190	0.8276
0.5249	14.82	63	0.5358	0.7759
0.5249	16.0	68	0.4111	0.8448
0.4479	16.94	72	0.4384	0.8103
0.4479	17.88	76	0.3855	0.8448
0.3549	18.82	80	0.4300	0.8103
0.3549	20.0	85	0.3456	0.8793
0.3549	20.94	89	0.2936	0.8966
0.3078	21.88	93	0.3196	0.9138
0.3078	22.82	97	0.3426	0.8966
0.3084	24.0	102	0.2944	0.8966
0.3084	24.94	106	0.3100	0.8966
0.2515	25.88	110	0.2952	0.8793
0.2515	26.82	114	0.2951	0.8621
0.2515	28.0	119	0.3316	0.8966
0.23	28.94	123	0.3052	0.8793
0.23	29.88	127	0.3799	0.8621
0.2139	30.82	131	0.3239	0.8793
0.2139	32.0	136	0.3369	0.8793
0.2048	32.94	140	0.2550	0.9310
0.2048	33.88	144	0.1913	0.8966
0.2048	34.82	148	0.3013	0.9138
0.189	36.0	153	0.2098	0.9483
0.189	36.94	157	0.1898	0.9310
0.1821	37.88	161	0.3412	0.8621
0.1821	38.82	165	0.3009	0.8793
0.1815	40.0	170	0.1856	0.9310
0.1815	40.94	174	0.1899	0.9310
0.1815	41.88	178	0.2124	0.9310
0.1396	42.82	182	0.2113	0.8966
0.1396	44.0	187	0.1331	0.9483
0.1571	44.94	191	0.2087	0.9310
0.1571	45.88	195	0.2447	0.9310
0.1571	46.82	199	0.1935	0.9483
0.1454	48.0	204	0.2409	0.9310
0.1454	48.94	208	0.2427	0.9310
0.1389	49.88	212	0.2864	0.9138
0.1389	50.82	216	0.3099	0.9138
0.1182	52.0	221	0.2566	0.9138
0.1182	52.94	225	0.2442	0.9138
0.1182	53.88	229	0.2263	0.9310
0.1411	54.82	233	0.2366	0.9138
0.1411	56.0	238	0.1764	0.9483
0.1023	56.94	242	0.1281	0.9310
0.1023	57.88	246	0.1593	0.9483
0.1198	58.82	250	0.2230	0.9310
0.1198	60.0	255	0.1710	0.9310
0.1198	60.94	259	0.1543	0.9483
0.0966	61.88	263	0.1757	0.9310
0.0966	62.82	267	0.1813	0.9138
0.1209	64.0	272	0.1941	0.9138
0.1209	64.94	276	0.1982	0.9310
0.107	65.88	280	0.2108	0.9310
0.107	66.82	284	0.2435	0.9310
0.107	68.0	289	0.2253	0.9310
0.0981	68.94	293	0.2033	0.9310
0.0981	69.88	297	0.1985	0.9310
0.1205	70.82	301	0.1905	0.9310
0.1205	72.0	306	0.1732	0.9483
0.129	72.94	310	0.1672	0.9483
0.129	73.88	314	0.1616	0.9483
0.129	74.82	318	0.1606	0.9483
0.1066	75.29	320	0.1608	0.9483

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu118
• Datasets 2.15.0
• Tokenizers 0.15.0