README.md

metadata

tags:
  - generated_from_trainer
metrics:
  - accuracy
model-index:
  - name: DNABert_K6_G_quad
    results: []

DNABert_K6_G_quad

This model is a fine-tuned version of armheb/DNA_bert_6 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.2424
• Accuracy: 0.9737

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: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 25
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.0927	1.0	9375	0.0818	0.9719
0.0681	2.0	18750	0.0714	0.9756
0.0607	3.0	28125	0.0863	0.9734
0.055	4.0	37500	0.0787	0.9757
0.0496	5.0	46875	0.0882	0.9758
0.0445	6.0	56250	0.0968	0.9752
0.0391	7.0	65625	0.1024	0.9755
0.0345	8.0	75000	0.1108	0.9739
0.0304	9.0	84375	0.1235	0.9745
0.0261	10.0	93750	0.1348	0.9730
0.023	11.0	103125	0.1427	0.9733
0.0197	12.0	112500	0.1462	0.9738
0.0182	13.0	121875	0.1570	0.9730
0.0145	14.0	131250	0.1757	0.9729
0.0122	15.0	140625	0.1911	0.9735
0.0108	16.0	150000	0.1977	0.9736
0.01	17.0	159375	0.1993	0.9732
0.0083	18.0	168750	0.2172	0.9736
0.0074	19.0	178125	0.2242	0.9740
0.0059	20.0	187500	0.2245	0.9732
0.0058	21.0	196875	0.2306	0.9733
0.0043	22.0	206250	0.2414	0.9737
0.0044	23.0	215625	0.2394	0.9735
0.0039	24.0	225000	0.2420	0.9736
0.0032	25.0	234375	0.2424	0.9737

Framework versions

• Transformers 4.22.1
• Pytorch 1.12.1
• Datasets 2.4.0
• Tokenizers 0.12.1