Modelpublic

LiukG/nucleotide-transformer-finetuned-lora-NucleotideTransformer

sourceHugging Facecc-by-nc-sa-4.0updated 3y agoView on Hugging Face

0likes25downloads

Model Card

nucleotide-transformer-finetuned-lora-NucleotideTransformer

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: 0.6652
—F1: 0.8473
—Mcc Score: 0.5631
—Accuracy: 0.7930

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: 0.0005
—trainbatchsize: 3
—evalbatchsize: 64
—seed: 42
—optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
—lrschedulertype: linear
—training_steps: 1000

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Mcc Score	Accuracy
1.4035	0.05	100	0.6460	0.7279	0.3679	0.6875
1.0673	0.11	200	0.4946	0.8437	0.5583	0.7930
0.7808	0.16	300	0.6190	0.7766	0.2749	0.6719
0.8938	0.21	400	2.1858	0.0	0.0	0.3906
0.9329	0.26	500	0.8452	0.8352	0.5179	0.7656
0.8721	0.32	600	0.7470	0.5286	0.2993	0.5820
0.6548	0.37	700	0.6967	0.8242	0.4769	0.75
0.6719	0.42	800	0.9450	0.7913	0.4425	0.7383
0.8265	0.47	900	0.5426	0.8328	0.5234	0.7773
0.5561	0.53	1000	0.6652	0.8473	0.5631	0.7930

Framework versions

—Transformers 4.38.1
—Pytorch 2.1.0+cu121
—Datasets 2.18.0
—Tokenizers 0.15.2