vit-base-patch16-224-in21k-datascience8

dwiedarioo/vit-base-patch16-224-in21k-datascience8

vit-base-patch16-224-in21k-datascience8 is a image classification model from dwiedarioo. Use it when you need a label for an image. It is set up for transformers. The card lists the license as apache-2.0.

probably proofread and complete it, then remove this comment. --

Downloads · 30 days

9% of all-time downloads

All-time downloads

383

Public

Repo size

18.7 GB

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.h5344 MB · 98%

At a glance

Task: Image Classification
Library: transformers
License: apache-2.0
Model type: vit
Access: Public
Created: Nov 18, 2023
Updated: Nov 19, 2023
SHA: b80a40fe

Base models

google/vit-base-patch16-224-in21k

Model card

From the Hugging Face model README

dwiedarioo/vit-base-patch16-224-in21k-datascience8

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:

Train Loss: 0.0061
Train Accuracy: 1.0
Train Top-3-accuracy: 1.0
Validation Loss: 0.1289
Validation Accuracy: 0.9633
Validation Top-3-accuracy: 0.9935
Epoch: 53

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:

optimizer: {'inner_optimizer': {'module': 'transformers.optimization_tf', 'class_name': 'AdamWeightDecay', 'config': {'name': 'AdamWeightDecay', 'learning_rate': {'module': 'keras.optimizers.schedules', 'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 3e-05, 'decay_steps': 8200, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}, 'registered_name': None}, 'decay': 0.0, 'beta_1': 0.8999999761581421, 'beta_2': 0.9990000128746033, 'epsilon': 1e-08, 'amsgrad': False, 'weight_decay_rate': 0.01}, 'registered_name': 'AdamWeightDecay'}, 'dynamic': True, 'initial_scale': 32768.0, 'dynamic_growth_steps': 2000}
training_precision: mixed_float16

Training results

Train Loss	Train Accuracy	Train Top-3-accuracy	Validation Loss	Validation Accuracy	Validation Top-3-accuracy	Epoch
2.2657	0.4219	0.6250	1.9041	0.5875	0.8121	0
1.5469	0.7006	0.8771	1.3444	0.7322	0.9136	1
1.0263	0.8519	0.9553	0.9408	0.8769	0.9719	2
0.6814	0.9412	0.9893	0.6752	0.9244	0.9827	3
0.4663	0.9779	0.9966	0.5106	0.9460	0.9935	4
0.3372	0.9927	0.9981	0.4127	0.9503	0.9892	5
0.2526	0.9958	0.9989	0.3468	0.9546	0.9914	6
0.2015	0.9973	1.0	0.3072	0.9568	0.9914	7
0.1663	0.9981	1.0	0.2609	0.9611	0.9935	8
0.1391	0.9989	0.9996	0.2353	0.9654	0.9957	9
0.1186	0.9992	1.0	0.2889	0.9438	0.9914	10
0.1201	0.9954	0.9996	0.3820	0.9006	0.9762	11
0.1402	0.9905	1.0	0.2185	0.9546	0.9892	12
0.0812	1.0	1.0	0.1898	0.9590	0.9914	13
0.0697	1.0	1.0	0.1757	0.9611	0.9935	14
0.0618	1.0	1.0	0.1698	0.9611	0.9914	15
0.0554	1.0	1.0	0.1625	0.9611	0.9935	16
0.0500	1.0	1.0	0.1592	0.9611	0.9935	17
0.0454	1.0	1.0	0.1526	0.9611	0.9935	18
0.0415	1.0	1.0	0.1494	0.9611	0.9935	19
0.0380	1.0	1.0	0.1473	0.9590	0.9935	20
0.0350	1.0	1.0	0.1443	0.9590	0.9935	21
0.0323	1.0	1.0	0.1403	0.9611	0.9935	22
0.0299	1.0	1.0	0.1408	0.9590	0.9935	23
0.0277	1.0	1.0	0.1368	0.9590	0.9935	24
0.0258	1.0	1.0	0.1369	0.9611	0.9935	25
0.0241	1.0	1.0	0.1361	0.9590	0.9935	26
0.0225	1.0	1.0	0.1355	0.9590	0.9935	27
0.0211	1.0	1.0	0.1349	0.9611	0.9935	28
0.0197	1.0	1.0	0.1312	0.9590	0.9935	29
0.0185	1.0	1.0	0.1317	0.9590	0.9935	30
0.0175	1.0	1.0	0.1328	0.9611	0.9935	31
0.0165	1.0	1.0	0.1318	0.9611	0.9935	32
0.0155	1.0	1.0	0.1320	0.9611	0.9935	33
0.0147	1.0	1.0	0.1294	0.9611	0.9935	34
0.0139	1.0	1.0	0.1306	0.9611	0.9935	35
0.0132	1.0	1.0	0.1291	0.9611	0.9935	36
0.0125	1.0	1.0	0.1295	0.9611	0.9935	37
0.0119	1.0	1.0	0.1306	0.9611	0.9935	38
0.0113	1.0	1.0	0.1275	0.9633	0.9935	39
0.0107	1.0	1.0	0.1282	0.9633	0.9935	40
0.0102	1.0	1.0	0.1272	0.9633	0.9935	41
0.0097	1.0	1.0	0.1282	0.9633	0.9935	42
0.0093	1.0	1.0	0.1269	0.9633	0.9935	43
0.0089	1.0	1.0	0.1286	0.9633	0.9935	44
0.0085	1.0	1.0	0.1278	0.9633	0.9935	45
0.0081	1.0	1.0	0.1285	0.9633	0.9935	46
0.0078	1.0	1.0	0.1291	0.9633	0.9935	47
0.0074	1.0	1.0	0.1290	0.9633	0.9935	48
0.0071	1.0	1.0	0.1283	0.9633	0.9935	49
0.0068	1.0	1.0	0.1292	0.9633	0.9935	50
0.0066	1.0	1.0	0.1295	0.9633	0.9935	51
0.0063	1.0	1.0	0.1290	0.9633	0.9935	52
0.0061	1.0	1.0	0.1289	0.9633	0.9935	53

Framework versions

Transformers 4.35.2
TensorFlow 2.14.0
Tokenizers 0.15.0

Task: Image Classification
Library: transformers
Type: vit
License: apache-2.0
Created: Nov 18, 2023
Updated: Nov 19, 2023