README.md

metadata

license: cc-by-nc-sa-4.0
base_model: InstaDeepAI/nucleotide-transformer-2.5b-multi-species
tags:
  - generated_from_trainer
metrics:
  - precision
  - recall
  - accuracy
model-index:
  - name: nucleotide-transformer-2.5b-multi-species_ft_Hepg2_1kbpHG19_DHSs_H3K27AC
    results: []

nucleotide-transformer-2.5b-multi-species_ft_Hepg2_1kbpHG19_DHSs_H3K27AC

This model is a fine-tuned version of InstaDeepAI/nucleotide-transformer-2.5b-multi-species on the None dataset. It achieves the following results on the evaluation set:

• Loss: 2.2344
• F1 Score: 0.8878
• Precision: 0.8690
• Recall: 0.9074
• Accuracy: 0.8836

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: 24
• eval_batch_size: 24
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	F1 Score	Precision	Recall	Accuracy
0.3737	0.5593	500	0.3067	0.8745	0.8377	0.9147	0.8668
0.292	1.1186	1000	0.3043	0.8852	0.8332	0.9441	0.8757
0.1865	1.6779	1500	0.3219	0.8854	0.8324	0.9456	0.8757
0.1744	2.2371	2000	2.0896	0.8683	0.8712	0.8654	0.8668
0.2341	2.7964	2500	1.9718	0.8887	0.8383	0.9456	0.8799
0.1357	3.3557	3000	2.2402	0.8787	0.8268	0.9375	0.8687
0.0779	3.9150	3500	2.0910	0.8857	0.8664	0.9059	0.8813
0.0426	4.4743	4000	2.2457	0.8724	0.8911	0.8544	0.8731
0.0549	5.0336	4500	1.8462	0.8952	0.8718	0.9199	0.8907
0.0331	5.5928	5000	1.9918	0.8803	0.8790	0.8816	0.8784
0.0206	6.1521	5500	2.0727	0.8847	0.8811	0.8882	0.8825
0.0101	6.7114	6000	2.0343	0.8882	0.8853	0.8912	0.8862
0.0099	7.2707	6500	2.1056	0.8831	0.8914	0.875	0.8825
0.0058	7.8300	7000	2.3964	0.8673	0.9036	0.8338	0.8705
0.0	8.3893	7500	2.2680	0.8808	0.8814	0.8801	0.8791
0.0047	8.9485	8000	2.1908	0.8863	0.8730	0.9	0.8828
0.0	9.5078	8500	2.2344	0.8878	0.8690	0.9074	0.8836

Framework versions

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