vit-large-patch16-224-dungeons-001

This model is a fine-tuned version of google/vit-large-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.6325
• Accuracy: 0.75

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: 1e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine_with_restarts
• lr_scheduler_warmup_steps: 100
• num_epochs: 85
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.8218	6.6667	10	1.8564	0.1667
1.4325	13.3333	20	1.7325	0.3333
0.8869	20.0	30	1.5186	0.4167
0.3717	26.6667	40	1.1131	0.6667
0.0945	33.3333	50	0.8408	0.75
0.0175	40.0	60	0.7224	0.75
0.0051	46.6667	70	0.6674	0.75
0.0024	53.3333	80	0.6325	0.75

Framework versions

• Transformers 4.44.2
• Pytorch 2.5.0+cu121
• Datasets 3.1.0
• Tokenizers 0.19.1