README.md

metadata

base_model: westlake-repl/SaProt_35M_AF2
library_name: peft

Model Card for Model ID

This model is used to predict fitness of GB1 protein variants.

Task type

protein level regression

Dataset description

The dataset is from:

Nicholas C Wu, Lei Dai, C Anders Olson, James O Lloyd-Smith, Ren Sun (2016) Adaptation in protein fitness landscapes is facilitated by indirect paths eLife 5:e16965 https://doi.org/10.7554/eLife.16965

Label is the fitness of mutant protein.

Model input type

Amino acid sequence

Performance

test_spearman: 0.54

test_pearson: 0.98

LoRA config

lora_dropout: 0.0

lora_alpha: 16

target_modules: ["query", "key", "value", "intermediate.dense", "output.dense"]

modules_to_save: ["classifier"]

Training config

class: AdamW

betas: (0.9, 0.98)

weight_decay: 0.01

learning rate: 1e-3

epoch: 20

batch size: 1000

precision: 16-mixed