nucleotide-transformer-v2-100m-multi-species_ft_BioS2_1kbpHG19_DHSs_H3K27AC

This model is a fine-tuned version of InstaDeepAI/nucleotide-transformer-v2-100m-multi-species on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.3735
• F1 Score: 0.8587
• Precision: 0.8425
• Recall: 0.8754
• Accuracy: 0.8487
• Auc: 0.9238
• Prc: 0.9216

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
0.5667	0.0840	500	0.5015	0.7975	0.7265	0.8838	0.7643	0.8328	0.8048
0.4953	0.1681	1000	0.4806	0.8157	0.7288	0.9260	0.7803	0.8607	0.8304
0.4666	0.2521	1500	0.4297	0.8321	0.7856	0.8844	0.8126	0.8818	0.8659
0.4404	0.3362	2000	0.4224	0.8267	0.8060	0.8485	0.8132	0.8860	0.8693
0.4366	0.4202	2500	0.4054	0.8393	0.7599	0.9372	0.8116	0.9012	0.8936
0.4145	0.5043	3000	0.3880	0.8276	0.8491	0.8072	0.8235	0.9060	0.8967
0.408	0.5883	3500	0.3836	0.8360	0.8481	0.8242	0.8302	0.9105	0.9063
0.3924	0.6724	4000	0.4161	0.8444	0.7698	0.9350	0.8191	0.9053	0.8972
0.3858	0.7564	4500	0.4237	0.8432	0.7774	0.9212	0.8201	0.9114	0.9041
0.3781	0.8405	5000	0.3678	0.8446	0.8435	0.8457	0.8366	0.9169	0.9082
0.3915	0.9245	5500	0.4415	0.8158	0.8908	0.7525	0.8217	0.9192	0.9156
0.3862	1.0086	6000	0.4456	0.8584	0.8201	0.9004	0.8440	0.9106	0.8937
0.3313	1.0926	6500	0.3869	0.8593	0.8335	0.8866	0.8475	0.9234	0.9158
0.3144	1.1767	7000	0.4080	0.8581	0.8527	0.8636	0.8501	0.9266	0.9210
0.3239	1.2607	7500	0.3974	0.8515	0.8587	0.8444	0.8454	0.9239	0.9188
0.3446	1.3448	8000	0.3735	0.8587	0.8425	0.8754	0.8487	0.9238	0.9216

Framework versions

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