nucleotide-transformer-500m-1000g_ft_BioS2_1kbpHG19_DHSs_H3K27AC

This model is a fine-tuned version of InstaDeepAI/nucleotide-transformer-500m-1000g on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.8664
• F1 Score: 0.8336
• Precision: 0.8251
• Recall: 0.8424
• Accuracy: 0.8245
• Auc: 0.9047
• Prc: 0.9003

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
0.529	0.0841	500	0.4714	0.8056	0.7472	0.8740	0.7799	0.8519	0.8346
0.4706	0.1683	1000	0.4616	0.7843	0.8182	0.7531	0.7837	0.8702	0.8570
0.4663	0.2524	1500	0.4523	0.8218	0.7276	0.9439	0.7863	0.8782	0.8664
0.4404	0.3366	2000	0.4665	0.8218	0.7254	0.9478	0.7854	0.8827	0.8716
0.4456	0.4207	2500	0.4759	0.8275	0.7481	0.9259	0.7986	0.8924	0.8871
0.4422	0.5049	3000	0.4230	0.8282	0.7707	0.8949	0.8061	0.8861	0.8819
0.4275	0.5890	3500	0.4123	0.8286	0.8132	0.8446	0.8176	0.8946	0.8883
0.4234	0.6732	4000	0.4592	0.8256	0.7345	0.9426	0.7922	0.8961	0.8929
0.4296	0.7573	4500	0.3919	0.8369	0.8014	0.8756	0.8218	0.9019	0.8968
0.4189	0.8415	5000	0.4052	0.8345	0.7781	0.8997	0.8137	0.8981	0.8948
0.4233	0.9256	5500	0.3965	0.8389	0.8040	0.8769	0.8241	0.9024	0.8995
0.4089	1.0098	6000	0.4514	0.8382	0.7861	0.8978	0.8191	0.9044	0.9015
0.3368	1.0939	6500	0.4123	0.8428	0.8032	0.8865	0.8273	0.9067	0.9039
0.3194	1.1781	7000	0.5789	0.8284	0.7382	0.9436	0.7959	0.9006	0.8994
0.3444	1.2622	7500	0.4602	0.8283	0.8278	0.8288	0.8206	0.9007	0.8993
0.3405	1.3463	8000	0.4591	0.8375	0.7816	0.9020	0.8172	0.9008	0.8986
0.335	1.4305	8500	0.5358	0.8430	0.8141	0.8740	0.8300	0.9036	0.9015
0.3228	1.5146	9000	0.6466	0.7698	0.8828	0.6825	0.7869	0.9052	0.9035
0.3409	1.5988	9500	0.5102	0.8326	0.8362	0.8291	0.8260	0.9077	0.9055
0.339	1.6829	10000	0.4643	0.8373	0.8178	0.8578	0.8260	0.9076	0.9054
0.3345	1.7671	10500	0.4526	0.8456	0.7977	0.8997	0.8285	0.9091	0.9062
0.3325	1.8512	11000	0.5876	0.8356	0.7666	0.9181	0.8113	0.9020	0.8999
0.344	1.9354	11500	0.4975	0.8424	0.8131	0.8740	0.8294	0.9081	0.9068
0.3019	2.0195	12000	0.7725	0.8352	0.8406	0.8298	0.8290	0.9123	0.9114
0.2254	2.1037	12500	0.7338	0.7948	0.8647	0.7353	0.8018	0.9051	0.9042
0.2171	2.1878	13000	0.8664	0.8336	0.8251	0.8424	0.8245	0.9047	0.9003

Framework versions

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