hyenadna-medium-160k-seqlen-hf_ft_BioS73_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: 0.5049
• F1 Score: 0.8498
• Precision: 0.8092
• Recall: 0.8946
• Accuracy: 0.8312
• Auc: 0.9050
• Prc: 0.8919

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
0.4658	0.3726	500	0.4353	0.8344	0.7760	0.9022	0.8088	0.8792	0.8707
0.4465	0.7452	1000	0.4110	0.8362	0.8066	0.8680	0.8185	0.8881	0.8822
0.4114	1.1177	1500	0.3973	0.8398	0.7786	0.9113	0.8144	0.8952	0.8901
0.3962	1.4903	2000	0.4114	0.8288	0.8245	0.8331	0.8163	0.8981	0.8925
0.3888	1.8629	2500	0.3946	0.8529	0.7811	0.9392	0.8271	0.9034	0.8962
0.3713	2.2355	3000	0.3950	0.8505	0.7984	0.9099	0.8293	0.9022	0.8959
0.357	2.6080	3500	0.3787	0.8476	0.8081	0.8911	0.8289	0.9044	0.8980
0.3466	2.9806	4000	0.3876	0.8538	0.7936	0.9239	0.8312	0.9075	0.8995
0.3164	3.3532	4500	0.3935	0.8513	0.8097	0.8973	0.8327	0.9059	0.8937
0.33	3.7258	5000	0.4684	0.8245	0.8426	0.8073	0.8166	0.9083	0.9012
0.31	4.0984	5500	0.4464	0.8520	0.7870	0.9288	0.8278	0.9049	0.8975
0.2786	4.4709	6000	0.4071	0.8506	0.8159	0.8883	0.8334	0.9028	0.8933
0.2874	4.8435	6500	0.3797	0.8640	0.8040	0.9337	0.8431	0.9050	0.8913
0.2567	5.2161	7000	0.4544	0.8517	0.7925	0.9204	0.8289	0.8983	0.8855
0.2388	5.5887	7500	0.4834	0.8415	0.8340	0.8492	0.8293	0.9032	0.8919
0.2386	5.9613	8000	0.4892	0.8533	0.8033	0.9099	0.8330	0.9003	0.8847
0.1956	6.3338	8500	0.5595	0.8511	0.8348	0.8680	0.8379	0.8992	0.8811
0.2257	6.7064	9000	0.5049	0.8498	0.8092	0.8946	0.8312	0.9050	0.8919

Framework versions

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