hushem_conflu_deneme_fold2

This model is a fine-tuned version of facebook/deit-tiny-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 1.9900
• Accuracy: 0.5333

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: 0.001
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	6	1.5124	0.2444
2.1014	2.0	12	1.4172	0.2667
2.1014	3.0	18	1.3682	0.2667
1.3494	4.0	24	1.5568	0.3333
1.1794	5.0	30	1.1703	0.3778
1.1794	6.0	36	1.1853	0.5333
0.9962	7.0	42	0.9960	0.5778
0.9962	8.0	48	0.9911	0.5778
0.7941	9.0	54	1.7710	0.4444
0.6504	10.0	60	1.0188	0.5111
0.6504	11.0	66	1.3899	0.4889
0.3424	12.0	72	1.3633	0.5333
0.3424	13.0	78	1.6911	0.4667
0.1576	14.0	84	1.8405	0.5556
0.0563	15.0	90	1.8925	0.5333
0.0563	16.0	96	2.0167	0.5333
0.0162	17.0	102	1.9900	0.5333
0.0162	18.0	108	1.9900	0.5333
0.009	19.0	114	1.9900	0.5333
0.0088	20.0	120	1.9900	0.5333

Framework versions

• Transformers 4.35.0
• Pytorch 2.1.0+cu118
• Datasets 2.14.6
• Tokenizers 0.14.1