elpv-vit

This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.7687
• Accuracy: 0.7259

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: 14

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	69	1.1146	0.5787
No log	2.0	138	0.9812	0.5787
No log	3.0	207	0.8885	0.6472
No log	4.0	276	0.7930	0.7081
No log	5.0	345	0.8019	0.6929
No log	6.0	414	0.8009	0.6878
No log	7.0	483	0.7984	0.6853
0.8194	8.0	552	0.7714	0.7107
0.8194	9.0	621	0.7667	0.7081
0.8194	10.0	690	0.7303	0.7234
0.8194	11.0	759	0.7321	0.7284
0.8194	12.0	828	0.7373	0.7335
0.8194	13.0	897	0.8051	0.6904
0.8194	14.0	966	0.7687	0.7259

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.1+cu121
• Datasets 2.15.0
• Tokenizers 0.15.0