IDiom-20M

The 20M variant of IDiom, an autoregressive protein language model for generating and studying intrinsically disordered protein regions (IDRs).

See the IDiom-300M model card for more information. Use jxliu2/idiom-20M in place of jxliu2/idiom-300M when loading this model.

IDiom Github   |   IDiom Models   |   IDiom-DB dataset

Paper: Generative modeling of intrinsically disordered protein regions by reinforcing sparse autoencoder features

License: MIT.

Downloads last month
202
Safetensors
Model size
18.9M params
Tensor type
F32
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Dataset used to train jxliu2/idiom-20M

Space using jxliu2/idiom-20M 1

Collection including jxliu2/idiom-20M

Paper for jxliu2/idiom-20M