nucleotide-transformer-v2-500m-multi-species_ft_BioS2_1kbpHG19_DHSs_H3K27AC

This model is a fine-tuned version of InstaDeepAI/nucleotide-transformer-v2-500m-multi-species on the None dataset. It achieves the following results on the evaluation set:

• Loss: 2.0128
• F1 Score: 0.8753
• Precision: 0.8186
• Recall: 0.9403
• Accuracy: 0.8603
• Auc: 0.9383
• Prc: 0.9326

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: 1e-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: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	F1 Score	Precision	Recall	Accuracy	Auc	Prc
0.534	0.1681	500	0.4570	0.7800	0.8228	0.7414	0.7820	0.8740	0.8666
0.3972	0.3361	1000	0.3836	0.8574	0.8086	0.9126	0.8418	0.9181	0.9115
0.3583	0.5042	1500	0.3394	0.8617	0.8135	0.9158	0.8467	0.9321	0.9322
0.3405	0.6723	2000	0.3418	0.8664	0.8789	0.8542	0.8627	0.9379	0.9360
0.323	0.8403	2500	0.3204	0.8565	0.9085	0.8101	0.8585	0.9453	0.9453
0.3109	1.0084	3000	0.3163	0.8774	0.8875	0.8675	0.8736	0.9471	0.9477
0.2315	1.1765	3500	0.3899	0.8803	0.8255	0.9429	0.8664	0.9473	0.9447
0.2265	1.3445	4000	0.3476	0.8816	0.8476	0.9184	0.8714	0.9483	0.9489
0.2286	1.5126	4500	0.3797	0.8587	0.9147	0.8091	0.8612	0.9445	0.9471
0.23	1.6807	5000	0.3251	0.8845	0.8714	0.8981	0.8778	0.9486	0.9497
0.2271	1.8487	5500	0.3160	0.8836	0.8579	0.9110	0.8750	0.9489	0.9489
0.2199	2.0168	6000	0.4896	0.8836	0.8300	0.9445	0.8703	0.9477	0.9459
0.2465	2.1849	6500	2.0128	0.8753	0.8186	0.9403	0.8603	0.9383	0.9326

Framework versions

• Transformers 4.42.3
• Pytorch 2.3.0+cu121
• Datasets 2.18.0
• Tokenizers 0.19.0