hyenadna-medium-160k-seqlen-hf_ft_BioS45_1kbpHG19_DHSs_H3K27AC

This model is a fine-tuned version of LongSafari/hyenadna-medium-160k-seqlen-hf on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.5104
• F1 Score: 0.8100
• Precision: 0.8002
• Recall: 0.8202
• Accuracy: 0.7993
• Auc: 0.8765
• Prc: 0.8748

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
0.5145	0.4205	500	0.4663	0.7945	0.8003	0.7887	0.7871	0.8575	0.8420
0.4678	0.8410	1000	0.4667	0.8075	0.7334	0.8984	0.7766	0.8691	0.8628
0.4586	1.2616	1500	0.4759	0.8157	0.7463	0.8992	0.7880	0.8674	0.8598
0.4467	1.6821	2000	0.4349	0.8134	0.7961	0.8315	0.8010	0.8798	0.8751
0.4298	2.1026	2500	0.4558	0.8170	0.7480	0.9	0.7897	0.8749	0.8679
0.4068	2.5231	3000	0.4374	0.7982	0.8139	0.7831	0.7934	0.8830	0.8793
0.4299	2.9437	3500	0.4358	0.8117	0.8033	0.8202	0.8014	0.8816	0.8787
0.3994	3.3642	4000	0.4543	0.8120	0.8287	0.7960	0.8077	0.8815	0.8781
0.3742	3.7847	4500	0.4463	0.8139	0.7948	0.8339	0.8010	0.8796	0.8733
0.345	4.2052	5000	0.5104	0.8100	0.8002	0.8202	0.7993	0.8765	0.8748

Framework versions

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