my_awesome_pokemon_model

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

• Loss: 7.3838
• Accuracy: 0.0755

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: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
4.926	1.0	76	5.4705	0.0007
3.7521	1.99	152	5.9651	0.0129
1.9692	2.99	228	5.8631	0.0144
0.7605	4.0	305	5.9688	0.0482
0.4163	5.0	381	6.1329	0.0655
0.3085	5.99	457	6.2311	0.0806
0.2155	6.99	533	6.4040	0.0683
0.2188	8.0	610	6.4869	0.0748
0.2241	9.0	686	6.6527	0.0763
0.1505	9.99	762	6.7076	0.0755
0.1429	10.99	838	6.7627	0.0719
0.1378	12.0	915	6.8740	0.0712
0.1335	13.0	991	6.9456	0.0741
0.1335	13.99	1067	6.8821	0.0748
0.1131	14.99	1143	6.9655	0.0763
0.1041	16.0	1220	7.0660	0.0763
0.0844	17.0	1296	7.1479	0.0770
0.086	17.99	1372	7.1182	0.0748
0.1028	18.99	1448	7.1395	0.0734
0.0456	20.0	1525	7.2099	0.0748
0.0617	21.0	1601	7.2512	0.0734
0.0711	21.99	1677	7.3157	0.0813
0.0623	22.99	1753	7.2590	0.0791
0.0419	24.0	1830	7.3413	0.0712
0.0924	25.0	1906	7.3051	0.0784
0.0471	25.99	1982	7.3136	0.0763
0.0654	26.99	2058	7.3667	0.0734
0.0836	28.0	2135	7.4039	0.0770
0.06	29.0	2211	7.3998	0.0799
0.0694	29.9	2280	7.3838	0.0755

Framework versions

• Transformers 4.29.2
• Pytorch 2.0.0+cu118
• Datasets 2.12.0
• Tokenizers 0.13.3