Train-Test-Augmentation-V5-beit-base

This model is a fine-tuned version of microsoft/beit-base-patch16-224-pt22k-ft22k on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.6899
• Accuracy: 0.8442

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: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.0473	1.0	55	0.8312	0.7759
0.3767	2.0	110	0.5476	0.8336
0.176	3.0	165	0.5248	0.8256
0.07	4.0	220	0.5597	0.8527
0.043	5.0	275	0.5707	0.8472
0.0272	6.0	330	0.6225	0.8264
0.0168	7.0	385	0.5721	0.8553
0.0076	8.0	440	0.5967	0.8608
0.006	9.0	495	0.7036	0.8272
0.007	10.0	550	0.7167	0.8400
0.0048	11.0	605	0.6734	0.8506
0.0023	12.0	660	0.7424	0.8332
0.0032	13.0	715	0.7283	0.8340
0.002	14.0	770	0.6805	0.8502
0.0021	15.0	825	0.6899	0.8442

Framework versions

• Transformers 4.39.3
• Pytorch 2.1.2
• Datasets 2.19.1
• Tokenizers 0.15.2