vit-large-patch32-384-limb-person-crop

This model is a fine-tuned version of google/vit-large-patch32-384 on the c14kevincardenas/beta_caller_284_person_crop dataset. It achieves the following results on the evaluation set:

• Loss: 0.8342
• Accuracy: 0.6733

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: 2014
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 10.0
• mixed_precision_training: Native AMP
• label_smoothing_factor: 0.005

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.3972	1.0	214	1.3778	0.2977
1.3229	2.0	428	1.2778	0.3433
1.1861	3.0	642	1.0543	0.5240
1.0712	4.0	856	0.9633	0.5829
1.0372	5.0	1070	0.9747	0.5788
0.9167	6.0	1284	0.9168	0.6186
0.9135	7.0	1498	0.9436	0.6128
0.7999	8.0	1712	0.8458	0.6667
0.7583	9.0	1926	0.8397	0.6725
0.8234	10.0	2140	0.8342	0.6733

Framework versions

• Transformers 4.41.0.dev0
• Pytorch 2.0.1+cu117
• Datasets 2.19.1
• Tokenizers 0.19.1