swin-tiny-patch4-window7-224-finetuned-mobile-eye-tracking-dataset-v2

This model is a fine-tuned version of microsoft/swin-tiny-patch4-window7-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.2851
• Accuracy: 0.9094

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
• 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: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.0421	0.9941	127	1.9090	0.3566
1.2943	1.9922	254	1.2209	0.5679
0.8757	2.9902	381	0.8373	0.7108
0.7362	3.9961	509	0.6260	0.7967
0.6011	4.9941	636	0.5550	0.8049
0.5267	5.9922	763	0.4797	0.8339
0.4969	6.9902	890	0.4508	0.8583
0.4418	7.9961	1018	0.4243	0.8571
0.4564	8.9941	1145	0.4035	0.8618
0.4199	9.9922	1272	0.3678	0.8804
0.3929	10.9902	1399	0.3479	0.8897
0.3578	11.9961	1527	0.3355	0.8931
0.3405	12.9941	1654	0.3173	0.8955
0.3399	13.9922	1781	0.3177	0.9048
0.3219	14.9902	1908	0.3354	0.8943
0.3296	15.9961	2036	0.3175	0.8978
0.3259	16.9941	2163	0.2908	0.9048
0.3201	17.9922	2290	0.2883	0.9082
0.2922	18.9902	2417	0.2945	0.9013
0.3204	19.9961	2545	0.2961	0.9013
0.2926	20.9941	2672	0.2934	0.9048
0.3083	21.9922	2799	0.2960	0.9024
0.3174	22.9902	2926	0.2876	0.9059
0.2693	23.9961	3054	0.2896	0.9036
0.2807	24.9941	3181	0.2852	0.8990
0.289	25.9922	3308	0.2847	0.9048
0.2829	26.9902	3435	0.2888	0.9048
0.2749	27.9961	3563	0.2898	0.9059
0.2726	28.9941	3690	0.2780	0.9094
0.2744	29.9374	3810	0.2851	0.9094

Framework versions

• Transformers 4.46.2
• Pytorch 2.5.1+cu121
• Datasets 3.1.0
• Tokenizers 0.20.3