emotion_recog

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

• Loss: 1.4314
• Accuracy: 0.5188

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: 64
• eval_batch_size: 64
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	10	2.0378	0.2812
No log	2.0	20	1.9741	0.325
No log	3.0	30	1.8878	0.4188
No log	4.0	40	1.7969	0.4188
No log	5.0	50	1.6954	0.4375
No log	6.0	60	1.6114	0.5062
No log	7.0	70	1.5550	0.5125
No log	8.0	80	1.5190	0.5312
No log	9.0	90	1.4752	0.5125
No log	10.0	100	1.4542	0.5563
No log	11.0	110	1.4416	0.5125
No log	12.0	120	1.4155	0.55
No log	13.0	130	1.3733	0.5437
No log	14.0	140	1.3943	0.5062
No log	15.0	150	1.3682	0.5375
No log	16.0	160	1.3847	0.5188
No log	17.0	170	1.3590	0.525
No log	18.0	180	1.3557	0.5375
No log	19.0	190	1.3619	0.525
No log	20.0	200	1.3239	0.55

Framework versions

• Transformers 4.41.1
• Pytorch 2.3.0+cu121
• Datasets 2.19.2
• Tokenizers 0.19.1