distill-vit

This model is a fine-tuned version of on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.4751
• Accuracy: 0.7656

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
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 30
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.8986	1.0	65	0.7788	0.4286
0.8556	2.0	130	0.9774	0.4812
0.7581	3.0	195	0.6150	0.6541
0.6434	4.0	260	0.6455	0.6090
0.609	5.0	325	0.5329	0.7143
0.5503	6.0	390	0.5829	0.6466
0.5492	7.0	455	0.6716	0.6917
0.504	8.0	520	0.5342	0.6917
0.4966	9.0	585	0.5668	0.6617
0.4978	10.0	650	0.5347	0.6767
0.4535	11.0	715	0.5580	0.6090
0.4415	12.0	780	0.5085	0.7444
0.4308	13.0	845	0.5131	0.7068
0.4247	14.0	910	0.4808	0.7068
0.4307	15.0	975	0.5542	0.6917
0.4165	16.0	1040	0.5410	0.6992
0.3975	17.0	1105	0.5944	0.6015
0.3942	18.0	1170	0.4730	0.6917
0.3932	19.0	1235	0.4806	0.6917
0.3437	20.0	1300	0.5341	0.6842
0.3628	21.0	1365	0.5836	0.6692
0.3483	22.0	1430	0.6234	0.6316
0.3318	23.0	1495	0.4950	0.7143
0.3189	24.0	1560	0.4590	0.7068
0.3243	25.0	1625	0.5789	0.6692
0.3169	26.0	1690	0.5702	0.7218
0.3031	27.0	1755	0.4415	0.7519
0.2928	28.0	1820	0.4680	0.7368
0.3117	29.0	1885	0.5384	0.6842
0.3127	30.0	1950	0.5148	0.6767

Framework versions

• Transformers 4.45.0.dev0
• Pytorch 2.2.1+cu118
• Datasets 2.21.0
• Tokenizers 0.19.1