mobilevitv2-1.0-imagenet1k-256-finetuned-swin-tiny

This model is a fine-tuned version of apple/mobilevitv2-1.0-imagenet1k-256 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.3595
• Accuracy: 0.5468

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: 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: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
3.6149	0.96	20	3.6094	0.0363
3.601	1.98	41	3.5936	0.0544
3.5892	2.99	62	3.5643	0.1057
3.5556	4.0	83	3.5195	0.1752
3.505	4.96	103	3.4422	0.2870
3.4072	5.98	124	3.2947	0.3172
3.2477	6.99	145	3.0629	0.3233
3.0508	8.0	166	2.8124	0.3444
2.8381	8.96	186	2.6019	0.3867
2.6407	9.98	207	2.4012	0.4018
2.5312	10.99	228	2.2300	0.4441
2.3687	12.0	249	2.0957	0.4411
2.2963	12.96	269	1.9972	0.4653
2.1898	13.98	290	1.9019	0.4743
2.0632	14.99	311	1.8381	0.4834
2.0279	16.0	332	1.7724	0.4955
1.998	16.96	352	1.7243	0.5015
1.9156	17.98	373	1.6919	0.5015
1.8914	18.99	394	1.6483	0.4985
1.8466	20.0	415	1.6211	0.5045
1.853	20.96	435	1.5899	0.5166
1.8124	21.98	456	1.5613	0.5015
1.7247	22.99	477	1.5355	0.5227
1.7034	24.0	498	1.5121	0.5287
1.6678	24.96	518	1.5000	0.5317
1.6832	25.98	539	1.4876	0.5287
1.6727	26.99	560	1.4796	0.5287
1.5744	28.0	581	1.4712	0.5227
1.5842	28.96	601	1.4492	0.5166
1.5416	29.98	622	1.4345	0.5347
1.5757	30.99	643	1.4229	0.5257
1.5574	32.0	664	1.4138	0.5378
1.5665	32.96	684	1.4077	0.5438
1.4837	33.98	705	1.3861	0.5438
1.5114	34.99	726	1.3956	0.5529
1.5207	36.0	747	1.3883	0.5468
1.4879	36.96	767	1.3750	0.5378
1.4547	37.98	788	1.3817	0.5408
1.4668	38.99	809	1.3643	0.5529
1.457	40.0	830	1.3669	0.5408
1.4604	40.96	850	1.3653	0.5498
1.4556	41.98	871	1.3621	0.5438
1.4852	42.99	892	1.3549	0.5498
1.4198	44.0	913	1.3461	0.5498
1.3824	44.96	933	1.3495	0.5498
1.4035	45.98	954	1.3495	0.5589
1.4586	46.99	975	1.3476	0.5529
1.4265	48.0	996	1.3481	0.5498
1.4563	48.19	1000	1.3595	0.5468

Framework versions

• Transformers 4.37.0
• Pytorch 2.1.2
• Datasets 2.16.1
• Tokenizers 0.15.1