| import os |
| import sys |
| sys.path.append(os.path.dirname(os.path.dirname(__file__))) |
|
|
| import torch |
| import os |
| |
| from omegaconf import OmegaConf |
| import pytorch_lightning as pl |
|
|
| class MInterface(pl.LightningModule): |
| def __init__(self, model_name=None, loss=None, lr=None, **kargs): |
| super().__init__() |
| self.save_hyperparameters() |
| self.load_model() |
| self.cross_entropy = torch.nn.NLLLoss(reduction='none') |
|
|
| def forward(self, batch): |
| batch = self.model._get_features(batch) |
| results = self.model(batch) |
| log_probs, mask = results['log_probs'], batch['mask'] |
| cmp = log_probs.argmax(dim=-1) == batch['S'] |
| recovery = (cmp * mask).sum() / (mask.sum()) |
| return recovery |
|
|
| def load_model(self): |
| params = self.hparams.configs |
| params.update(self.hparams) |
|
|
| if params.dataset == 'ligand': |
| from src.model.modules.invfold.src.models.odesign_ligand import ODesign_Ligand_Model |
| self.model = ODesign_Ligand_Model(params) |
| else: |
| from src.model.modules.invfold.src.models.odesign_model import ODesign_Model |
| self.model = ODesign_Model(params) |
| |
