|
|
--- |
|
|
library_name: transformers |
|
|
license: mit |
|
|
base_model: |
|
|
- facebook/esm2_t33_650M_UR50D |
|
|
--- |
|
|
|
|
|
## ft-ESM |
|
|
ft-ESM is a finetuned version of the [650M-parameter ESM2 protein language model](https://huggingface.co/facebook/esm2_t33_650M_UR50D), finetuned on paired antibody sequences from [Jaffe et al.](https://www.nature.com/articles/s41586-022-05371-z) |
|
|
Datasets used for pre-training are available on [Zenodo](https://doi.org/10.5281/zenodo.8237395) and code is available on [GitHub](https://github.com/brineylab/BALM-paper). More details can be found in [our paper](https://doi.org/10.1016/j.patter.2024.100967) published in Patterns. |
|
|
|
|
|
### Use |
|
|
Load the model and tokenizer as follows: |
|
|
```python |
|
|
from transformers import EsmTokenizer, EsmForMaskedLM |
|
|
|
|
|
model = EsmForMaskedLM.from_pretrained("brineylab/ft-ESM") |
|
|
tokenizer = EsmTokenizer.from_pretrained("brineylab/ft-ESM") |
|
|
``` |
|
|
|
|
|
The tokenizer expects sequences formatted as: `HEAVY_CHAIN<cls><cls>LIGHT_CHAIN`. |
