vit-base-patch16-224-in21k-finetuned-mobile-eye-tracking-dataset-v2

This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.2615
• Accuracy: 0.9231

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
No log	1.0	2	0.6641	0.6154
No log	2.0	4	0.6343	0.6154
No log	3.0	6	0.5990	0.6154
No log	4.0	8	0.5438	0.8462
No log	5.0	10	0.5108	0.9231
No log	6.0	12	0.4413	0.8462
No log	7.0	14	0.3947	0.8462
No log	8.0	16	0.3568	0.9231
No log	9.0	18	0.3297	0.9231
0.4923	10.0	20	0.3110	0.9231
0.4923	11.0	22	0.2988	0.9231
0.4923	12.0	24	0.2836	0.9231
0.4923	13.0	26	0.2702	0.9231
0.4923	14.0	28	0.2636	0.9231
0.4923	15.0	30	0.2615	0.9231

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu121
• Datasets 2.16.1
• Tokenizers 0.15.0