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

Loss: 0.0605
Accuracy: 0.9868
Confusion Matrix: [[64, 2], [1, 160]]

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: 32
eval_batch_size: 32
seed: 42
optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
training_steps: 1820
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Confusion Matrix
0.1028	0.2747	100	0.1222	0.9736	[[60, 6], [0, 161]]
0.0829	0.5495	200	0.0828	0.9824	[[62, 4], [0, 161]]
0.0207	0.8242	300	0.0638	0.9868	[[63, 3], [0, 161]]
0.0136	1.0989	400	0.0714	0.9824	[[63, 3], [1, 160]]
0.011	1.3736	500	0.0754	0.9824	[[63, 3], [1, 160]]
0.0089	1.6484	600	0.0643	0.9824	[[63, 3], [1, 160]]
0.0078	1.9231	700	0.0626	0.9824	[[63, 3], [1, 160]]
0.007	2.1978	800	0.0633	0.9824	[[63, 3], [1, 160]]
0.0064	2.4725	900	0.0608	0.9824	[[63, 3], [1, 160]]
0.0059	2.7473	1000	0.0607	0.9868	[[64, 2], [1, 160]]
0.0056	3.0220	1100	0.0594	0.9824	[[63, 3], [1, 160]]
0.0053	3.2967	1200	0.0584	0.9868	[[63, 3], [0, 161]]
0.0051	3.5714	1300	0.0596	0.9868	[[64, 2], [1, 160]]
0.0049	3.8462	1400	0.0597	0.9868	[[64, 2], [1, 160]]
0.0047	4.1209	1500	0.0607	0.9868	[[64, 2], [1, 160]]
0.0046	4.3956	1600	0.0607	0.9868	[[64, 2], [1, 160]]
0.0046	4.6703	1700	0.0615	0.9868	[[64, 2], [1, 160]]
0.0045	4.9451	1800	0.0605	0.9868	[[64, 2], [1, 160]]

Framework versions

Transformers 4.46.2
Pytorch 2.4.0
Datasets 2.20.0
Tokenizers 0.20.3

jerlawson13/vit-base-gpu-aug

vit-base-gpu-aug

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

jerlawson13/
vit-base-gpu-aug