real_vs_fake_image_model_vit_base

This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0189
• Accuracy: 0.9953

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: 3e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.0094	0.1883	100	0.0243	0.9941
0.0165	0.3766	200	0.0351	0.9901
0.0239	0.5650	300	0.0470	0.9876
0.0179	0.7533	400	0.0678	0.9856
0.0166	0.9416	500	0.0296	0.9920
0.0138	1.1299	600	0.0337	0.9926
0.0574	1.3183	700	0.1020	0.9772
0.0256	1.5066	800	0.0612	0.9847
0.0327	1.6949	900	0.0616	0.9846
0.0086	1.8832	1000	0.0272	0.9923
0.008	2.0716	1100	0.0329	0.9920
0.0014	2.2599	1200	0.0250	0.9939
0.0132	2.4482	1300	0.0248	0.9937
0.0189	2.6365	1400	0.0266	0.9936
0.0034	2.8249	1500	0.0225	0.9948
0.009	3.0132	1600	0.0240	0.9942
0.0009	3.2015	1700	0.0244	0.9942
0.0054	3.3898	1800	0.0339	0.9928
0.0046	3.5782	1900	0.0248	0.9945
0.0135	3.7665	2000	0.0245	0.9945
0.0274	3.9548	2100	0.0241	0.9947
0.0031	4.1431	2200	0.0225	0.9947
0.0121	4.3315	2300	0.0210	0.9952
0.0055	4.5198	2400	0.0209	0.9953
0.0183	4.7081	2500	0.0197	0.9955
0.0077	4.8964	2600	0.0189	0.9953

Framework versions

• Transformers 4.47.0
• Pytorch 2.5.1+cu121
• Datasets 3.2.0
• Tokenizers 0.21.0