ProteinClassificationModelV2

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

• Loss: 0.4631
• Accuracy: 0.92

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: 2e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 40

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.707	1.0	113	0.3861	0.845
0.3382	2.0	226	0.3137	0.87
0.2635	3.0	339	0.2881	0.885
0.2469	4.0	452	0.3090	0.88
0.1995	5.0	565	0.2577	0.92
0.1869	6.0	678	0.2783	0.91
0.1799	7.0	791	0.2146	0.925
0.1376	8.0	904	0.3013	0.89
0.1079	9.0	1017	0.4004	0.885
0.1207	10.0	1130	0.5334	0.85
0.089	11.0	1243	0.3145	0.93
0.1056	12.0	1356	0.3250	0.9
0.1087	13.0	1469	0.4588	0.89
0.0746	14.0	1582	0.3866	0.895
0.081	15.0	1695	0.3356	0.905
0.0539	16.0	1808	0.5468	0.89
0.0772	17.0	1921	0.4060	0.895
0.0729	18.0	2034	0.5682	0.875
0.0525	19.0	2147	0.4364	0.905
0.0654	20.0	2260	0.3860	0.895
0.0461	21.0	2373	0.5711	0.895
0.0539	22.0	2486	0.6325	0.89
0.0309	23.0	2599	0.5805	0.885
0.0452	24.0	2712	0.4380	0.905
0.0406	25.0	2825	0.4316	0.9
0.0404	26.0	2938	0.4540	0.905
0.0288	27.0	3051	0.5276	0.915
0.0493	28.0	3164	0.6025	0.895
0.0244	29.0	3277	0.6028	0.9
0.025	30.0	3390	0.5895	0.915
0.0277	31.0	3503	0.4587	0.915
0.0284	32.0	3616	0.4466	0.915
0.029	33.0	3729	0.4343	0.905
0.021	34.0	3842	0.4058	0.915
0.012	35.0	3955	0.4433	0.92
0.0299	36.0	4068	0.4442	0.925
0.0206	37.0	4181	0.4736	0.92
0.0253	38.0	4294	0.5221	0.915
0.0147	39.0	4407	0.4488	0.92
0.0261	40.0	4520	0.4631	0.92

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu121
• Datasets 2.16.0
• Tokenizers 0.15.0