swinv2-tiny-patch4-window8-256-finetuned-microbes

This model is a fine-tuned version of microsoft/swinv2-tiny-patch4-window8-256 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 1.0939
• Accuracy: 0.7051

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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
4.1153	0.97	16	3.8048	0.1239
3.2047	2.0	33	2.9123	0.2949
2.6979	2.97	49	2.2162	0.4231
1.9422	4.0	66	1.8476	0.5043
1.5677	4.97	82	1.6194	0.5684
1.3485	6.0	99	1.4825	0.5855
1.146	6.97	115	1.4073	0.5983
1.0408	8.0	132	1.2730	0.6325
0.9334	8.97	148	1.2782	0.6282
0.8702	10.0	165	1.1758	0.6752
0.8589	10.97	181	1.1652	0.6838
0.7607	12.0	198	1.2129	0.6795
0.7676	12.97	214	1.1509	0.6795
0.7359	14.0	231	1.1327	0.6966
0.7491	14.97	247	1.1059	0.6966
0.6664	16.0	264	1.1413	0.6923
0.618	16.97	280	1.0954	0.7009
0.6504	18.0	297	1.1030	0.7009
0.6241	18.97	313	1.0956	0.7009
0.6258	19.39	320	1.0939	0.7051

Framework versions

• Transformers 4.32.1
• Pytorch 2.0.1+cpu
• Datasets 2.14.4
• Tokenizers 0.13.3