nucleotide-transformer-v2-50m-multi-species_ft_BioS2_1kbpHG19_DHSs_H3K27AC

This model is a fine-tuned version of InstaDeepAI/nucleotide-transformer-v2-50m-multi-species on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.4513
• F1 Score: 0.8606
• Precision: 0.7946
• Recall: 0.9386
• Accuracy: 0.8428
• Auc: 0.9312
• Prc: 0.9283

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: 8
• eval_batch_size: 8
• 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
0.5632	0.0839	500	0.5934	0.7793	0.6466	0.9805	0.7129	0.8537	0.8277
0.4814	0.1679	1000	0.5125	0.7531	0.8452	0.6791	0.7698	0.8800	0.8632
0.4422	0.2518	1500	0.4025	0.8289	0.8203	0.8376	0.8212	0.9001	0.8868
0.4233	0.3358	2000	0.4586	0.8106	0.8643	0.7632	0.8156	0.9044	0.8958
0.4191	0.4197	2500	0.3875	0.8464	0.8236	0.8704	0.8366	0.9077	0.8961
0.4044	0.5037	3000	0.3871	0.8515	0.8122	0.8948	0.8387	0.9134	0.9067
0.4218	0.5876	3500	0.3585	0.8514	0.8380	0.8652	0.8439	0.9211	0.9158
0.3903	0.6716	4000	0.3746	0.8562	0.7974	0.9243	0.8395	0.9215	0.9164
0.384	0.7555	4500	0.3564	0.8527	0.8263	0.8808	0.8427	0.9206	0.9164
0.3728	0.8395	5000	0.3657	0.8508	0.8471	0.8545	0.8450	0.9226	0.9188
0.3862	0.9234	5500	0.3616	0.8619	0.8376	0.8876	0.8529	0.9281	0.9248
0.3829	1.0074	6000	0.3860	0.8595	0.8243	0.8977	0.8482	0.9282	0.9248
0.3051	1.0913	6500	0.4836	0.8485	0.7533	0.9711	0.8207	0.9287	0.9238
0.3191	1.1753	7000	0.4513	0.8606	0.7946	0.9386	0.8428	0.9312	0.9283

Framework versions

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