emotion_classification_v1

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: 1.1905
• Accuracy: 0.575

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: 5e-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: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	10	2.0278	0.2437
No log	2.0	20	1.8875	0.3875
No log	3.0	30	1.6890	0.4313
No log	4.0	40	1.5484	0.5
No log	5.0	50	1.4799	0.5125
No log	6.0	60	1.4148	0.5375
No log	7.0	70	1.3529	0.5375
No log	8.0	80	1.3120	0.5312
No log	9.0	90	1.2790	0.5813
No log	10.0	100	1.2498	0.575
No log	11.0	110	1.2610	0.525
No log	12.0	120	1.1896	0.5938
No log	13.0	130	1.2251	0.5312
No log	14.0	140	1.2019	0.575
No log	15.0	150	1.1797	0.5563
No log	16.0	160	1.2484	0.5437
No log	17.0	170	1.1766	0.5875
No log	18.0	180	1.2401	0.4938
No log	19.0	190	1.1977	0.5312
No log	20.0	200	1.1839	0.5875
No log	21.0	210	1.2028	0.5687
No log	22.0	220	1.2048	0.5625
No log	23.0	230	1.2637	0.5375
No log	24.0	240	1.2371	0.5375
No log	25.0	250	1.2777	0.5687
No log	26.0	260	1.2544	0.525
No log	27.0	270	1.2104	0.5625
No log	28.0	280	1.1372	0.5938
No log	29.0	290	1.2405	0.575
No log	30.0	300	1.1624	0.6062
No log	31.0	310	1.2376	0.5875
No log	32.0	320	1.1794	0.5875
No log	33.0	330	1.2156	0.5563
No log	34.0	340	1.1725	0.55
No log	35.0	350	1.2394	0.55
No log	36.0	360	1.1886	0.5938
No log	37.0	370	1.1760	0.6188
No log	38.0	380	1.2757	0.525
No log	39.0	390	1.1703	0.6062
No log	40.0	400	1.2734	0.575
No log	41.0	410	1.2265	0.5563
No log	42.0	420	1.2651	0.5687
No log	43.0	430	1.2419	0.5813
No log	44.0	440	1.1871	0.6
No log	45.0	450	1.2542	0.575
No log	46.0	460	1.1910	0.5813
No log	47.0	470	1.1990	0.6
No log	48.0	480	1.2097	0.5813
No log	49.0	490	1.2226	0.5875
0.699	50.0	500	1.2793	0.5375

Framework versions

• Transformers 4.33.2
• Pytorch 2.0.1+cu118
• Datasets 2.14.5
• Tokenizers 0.13.3