0.50-200Train-100Test-swinv2-large

This model is a fine-tuned version of microsoft/swinv2-large-patch4-window12-192-22k on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.7669
• Accuracy: 0.8233

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: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• 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
2.4602	0.9825	14	1.7254	0.4318
1.7105	1.9649	28	0.8579	0.7047
0.6096	2.9474	42	0.7268	0.7562
0.3983	4.0	57	0.6706	0.7852
0.1083	4.9825	71	0.7051	0.7897
0.0952	5.9649	85	0.8423	0.7696
0.1106	6.9474	99	0.6406	0.8121
0.0357	8.0	114	0.8410	0.7897
0.0522	8.9825	128	0.8197	0.7987
0.0274	9.9649	142	0.8788	0.8098
0.0203	10.9474	156	0.8037	0.8233
0.0361	12.0	171	0.7932	0.8076
0.0204	12.9825	185	0.7503	0.8210
0.0165	13.9649	199	0.7416	0.8098
0.0129	14.9474	213	0.8474	0.8277
0.0062	16.0	228	0.7788	0.8233
0.0028	16.9825	242	0.7687	0.8255
0.001	17.9649	256	0.7730	0.8255
0.0019	18.9474	270	0.7681	0.8255
0.0014	19.6491	280	0.7669	0.8233

Framework versions

• Transformers 4.41.2
• Pytorch 2.1.2
• Datasets 2.19.2
• Tokenizers 0.19.1