emotion_classification

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: 1.3578
• Accuracy: 0.5125

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: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.0796	1.0	10	2.0709	0.1562
2.0631	2.0	20	2.0496	0.225
2.0242	3.0	30	2.0148	0.2875
1.9387	4.0	40	1.9268	0.325
1.789	5.0	50	1.7454	0.3812
1.6216	6.0	60	1.5996	0.3937
1.4795	7.0	70	1.5577	0.375
1.3735	8.0	80	1.5090	0.4062
1.2889	9.0	90	1.4418	0.4313
1.2092	10.0	100	1.4209	0.425
1.1127	11.0	110	1.3828	0.4437
1.032	12.0	120	1.3507	0.4562
0.9616	13.0	130	1.3556	0.4875
0.9099	14.0	140	1.3204	0.5188
0.8425	15.0	150	1.3490	0.4688
0.806	16.0	160	1.3690	0.5062
0.7377	17.0	170	1.3344	0.5563
0.677	18.0	180	1.4178	0.4625
0.6071	19.0	190	1.3305	0.4875
0.5581	20.0	200	1.3070	0.5
0.5599	21.0	210	1.3245	0.4938
0.5222	22.0	220	1.3765	0.4562
0.4856	23.0	230	1.3345	0.5
0.458	24.0	240	1.2938	0.5188
0.4393	25.0	250	1.3380	0.5062
0.4239	26.0	260	1.3756	0.525
0.4443	27.0	270	1.4586	0.4813
0.4374	28.0	280	1.2996	0.55
0.3917	29.0	290	1.3222	0.5062
0.3986	30.0	300	1.4486	0.4813
0.353	31.0	310	1.5204	0.4562
0.3598	32.0	320	1.3027	0.5625
0.3538	33.0	330	1.6122	0.4313
0.3246	34.0	340	1.5237	0.4437
0.3089	35.0	350	1.4717	0.5125
0.3278	36.0	360	1.5666	0.45
0.2865	37.0	370	1.4377	0.5
0.2958	38.0	380	1.4766	0.4938
0.3036	39.0	390	1.5345	0.4375
0.286	40.0	400	1.4174	0.5062
0.3099	41.0	410	1.4087	0.4625
0.2801	42.0	420	1.4439	0.4813
0.2973	43.0	430	1.4712	0.4938
0.2892	44.0	440	1.4099	0.5188
0.2835	45.0	450	1.3011	0.5563
0.261	46.0	460	1.6512	0.4188
0.2589	47.0	470	1.5651	0.4375
0.2806	48.0	480	1.5194	0.4938
0.2749	49.0	490	1.4519	0.525
0.2482	50.0	500	1.4127	0.5188

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.0+cu121
• Datasets 2.21.0
• Tokenizers 0.19.1