ahishamm/vit-base-HAM-10000-sharpened-large-patch-16 - AI Model Zoo

vit-base-HAM-10000-sharpened-large-patch-16

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

Loss: 0.5504
Accuracy: 0.8075
Recall: 0.8075
F1: 0.8075
Precision: 0.8075

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: 0.0002
train_batch_size: 16
eval_batch_size: 8
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 4

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Recall	F1	Precision
0.9294	0.2	100	1.0377	0.6733	0.6733	0.6733	0.6733
1.0067	0.4	200	0.8976	0.6813	0.6813	0.6813	0.6813
1.0081	0.6	300	0.9345	0.6773	0.6773	0.6773	0.6773
0.9326	0.8	400	0.8494	0.6883	0.6883	0.6883	0.6883
0.8243	1.0	500	0.7481	0.7267	0.7267	0.7267	0.7267
0.7408	1.2	600	0.7277	0.7317	0.7317	0.7317	0.7317
0.6844	1.4	700	0.7114	0.7392	0.7392	0.7392	0.7392
0.7411	1.6	800	0.6772	0.7416	0.7416	0.7416	0.7416
0.7138	1.8	900	0.7136	0.7377	0.7377	0.7377	0.7377
0.5838	2.0	1000	0.6625	0.7521	0.7521	0.7521	0.7521
0.5315	2.2	1100	0.6104	0.7776	0.7776	0.7776	0.7776
0.6391	2.4	1200	0.6317	0.7591	0.7591	0.7591	0.7591
0.6903	2.59	1300	0.6098	0.7656	0.7656	0.7656	0.7656
0.5798	2.79	1400	0.6211	0.7751	0.7751	0.7751	0.7751
0.5448	2.99	1500	0.5824	0.7820	0.7820	0.7820	0.7820
0.4523	3.19	1600	0.5951	0.7776	0.7776	0.7776	0.7776
0.4485	3.39	1700	0.6114	0.7815	0.7815	0.7815	0.7815
0.487	3.59	1800	0.5730	0.7950	0.7950	0.7950	0.7950
0.4104	3.79	1900	0.5597	0.7965	0.7965	0.7965	0.7965
0.4468	3.99	2000	0.5504	0.8075	0.8075	0.8075	0.8075

Framework versions

Transformers 4.30.2
Pytorch 2.0.1+cu118
Datasets 2.13.1
Tokenizers 0.13.3