metadata
base_model: ProTrekHub/Protein_Encoder_35M
library_name: peft
Model Card for Model-Demo-ProTrek-35M
Base model
ProTrekHub/Protein_Encoder_35M
Task description
The task is to predict the protein solubility.
Label meaning
Label 1 - soluble Lable 0 - unsoluble
Task type
Protein-level Classification
Model input type
AA Sequence
LoRA config
- r: 8
- lora_dropout: 0.0
- lora_alpha: 16
- target_modules: ['value', 'intermediate.dense', 'query', 'output.dense', 'key']
- modules_to_save: ['classifier']
Training config
- optimizer:
- class: AdamW
- betas: (0.9, 0.98)
- weight_decay: 0.01
- learning rate: 0.0005
- epoch: 5
- batch size: 1
- precision: 16-mixed