multimolecule/carp-640m

CARP

Pre-trained convolutional protein language model using a masked language modeling (MLM) objective.

Disclaimer

This is an UNOFFICIAL implementation of Convolutions are competitive with transformers for protein sequence pretraining by Kevin K. Yang, et al.

The OFFICIAL repository of CARP is at microsoft/protein-sequence-models.

The MultiMolecule team has confirmed that the provided model and checkpoints are producing the same intermediate representations as the original implementation.

The team releasing CARP did not write this model card for this model so this model card has been written by the MultiMolecule team.

Model Details

CARP is a family of ByteNet-style convolutional protein language models. It uses learned token embeddings, a stack of residual dilated 1D convolution blocks, and a final layer normalization before the masked-language-model decoder. The models were pre-trained on the March 2020 release of UniRef50 using the same masked language modeling task as BERT and ESM-1b.

Variants

• multimolecule/carp-600k: The CARP model with about 600 thousand parameters.
• multimolecule/carp-38m: The CARP model with about 38 million parameters.
• multimolecule/carp-76m: The CARP model with about 76 million parameters.
• multimolecule/carp-640m: The CARP model with about 640 million parameters.

Model Specification

Variant	Num Layers	Hidden Size	Intermediate Size	Num Parameters (M)	FLOPs (G)	MACs (G)	Max Num Tokens
CARP-600k	16	128	64	0.61	1.25	0.61	1024
CARP-38M	16	1024	512	37.90	77.68	38.70
CARP-76M	32			75.74	155.26	77.36
CARP-640M	56	1280	1280	642.96	1317.22	657.73

Links

• Code: multimolecule.carp
• Data: UniRef50
• Paper: Convolutions are competitive with transformers for protein sequence pretraining
• Developed by: Kevin K. Yang, Nicolo Fusi, Alex X. Lu
• Model type: ByteNet-style convolutional protein masked language model
• Original Repository: microsoft/protein-sequence-models

Usage

The model file depends on the multimolecule library. You can install it using pip:

pip install multimolecule

Direct Use

Masked Language Modeling

You can use this model directly with a pipeline for masked language modeling:

import multimolecule  # you must import multimolecule to register models
from transformers import pipeline

predictor = pipeline("fill-mask", model="multimolecule/carp-640m")
output = predictor("MVLSPADKTNVKAAW<mask>KVGAHAGEYGAEALER")

Downstream Use

Extract Features

Here is how to use this model to get the features of a given sequence in PyTorch:

from multimolecule import ProteinTokenizer, CarpModel


tokenizer = ProteinTokenizer.from_pretrained("multimolecule/carp-640m")
model = CarpModel.from_pretrained("multimolecule/carp-640m")

text = "MVLSPADKTNVKAAWGKVGAHAGEYGAEALER"
input = tokenizer(text, return_tensors="pt")

output = model(**input)

Sequence Classification / Regression

This model is not fine-tuned for any specific task. You will need to fine-tune the model on a downstream task to use it for sequence classification or regression.

Here is how to use this model as backbone to fine-tune for a sequence-level task in PyTorch:

import torch
from multimolecule import ProteinTokenizer, CarpForSequencePrediction


tokenizer = ProteinTokenizer.from_pretrained("multimolecule/carp-640m")
model = CarpForSequencePrediction.from_pretrained("multimolecule/carp-640m")

text = "MVLSPADKTNVKAAWGKVGAHAGEYGAEALER"
input = tokenizer(text, return_tensors="pt")
label = torch.tensor([1])

output = model(**input, labels=label)

Token Classification / Regression

This model is not fine-tuned for any specific task. You will need to fine-tune the model on a downstream task to use it for token classification or regression.

Here is how to use this model as backbone to fine-tune for a residue-level task in PyTorch:

import torch
from multimolecule import ProteinTokenizer, CarpForTokenPrediction


tokenizer = ProteinTokenizer.from_pretrained("multimolecule/carp-640m")
model = CarpForTokenPrediction.from_pretrained("multimolecule/carp-640m")

text = "MVLSPADKTNVKAAWGKVGAHAGEYGAEALER"
input = tokenizer(text, return_tensors="pt")
label = torch.randint(2, (len(text), ))

output = model(**input, labels=label)

Contact Classification / Regression

This model is not fine-tuned for any specific task. You will need to fine-tune the model on a downstream task to use it for contact classification or regression.

Here is how to use this model as backbone to fine-tune for a contact-level task in PyTorch:

import torch
from multimolecule import ProteinTokenizer, CarpForContactPrediction


tokenizer = ProteinTokenizer.from_pretrained("multimolecule/carp-640m")
model = CarpForContactPrediction.from_pretrained("multimolecule/carp-640m")

text = "MVLSPADKTNVKAAWGKVGAHAGEYGAEALER"
input = tokenizer(text, return_tensors="pt")
label = torch.randint(2, (len(text), len(text)))

output = model(**input, labels=label)

Training Details

CARP was trained with Masked Language Modeling (MLM) as the pre-training objective. Masked residues are predicted from the surrounding protein sequence using bidirectional dilated convolution blocks rather than self-attention layers.

Training Data

CARP was pre-trained on the March 2020 release of UniRef50.

Training Procedure

Preprocessing

The released CARP checkpoints use the protein alphabet from the official sequence_models package. During conversion, equivalent amino-acid and special-token rows are mapped into the MultiMolecule protein tokenizer vocabulary.

Pre-training

The model was trained with masked language modeling over a ByteNet-style residual dilated convolution stack. Please refer to the original paper for details on the training setup.

Citation

@article{yang2024convolutions,
  author  = {Yang, Kevin K. and Fusi, Nicolo and Lu, Alex X.},
  title   = {Convolutions are competitive with transformers for protein sequence pretraining},
  journal = {Cell Systems},
  volume  = {15},
  number  = {3},
  pages   = {286--294.e2},
  year    = {2024},
  doi     = {10.1016/j.cels.2024.01.008},
  url     = {https://doi.org/10.1016/j.cels.2024.01.008},
}

The artifacts distributed in this repository are part of the MultiMolecule project. If MultiMolecule supports your research, please cite the MultiMolecule project as follows:

@software{chen_2024_12638419,
  author    = {Chen, Zhiyuan and Zhu, Sophia Y.},
  title     = {MultiMolecule},
  doi       = {10.5281/zenodo.12638419},
  publisher = {Zenodo},
  url       = {https://doi.org/10.5281/zenodo.12638419},
  year      = 2024,
  month     = may,
  day       = 4
}

Contact

Please use GitHub issues of MultiMolecule for any questions or comments on the model card.

Please contact the authors of the CARP paper for questions or comments on the paper/model.

License

This model implementation is licensed under the GNU Affero General Public License.

For additional terms and clarifications, please refer to our License FAQ.

SPDX-License-Identifier: AGPL-3.0-or-later