63 lines
2.2 KiB
Python
63 lines
2.2 KiB
Python
import os
|
|
from types import MethodType
|
|
import hydra
|
|
from hydra.utils import instantiate
|
|
from omegaconf import DictConfig
|
|
from torch.optim.lr_scheduler import ReduceLROnPlateau
|
|
from pytorch_lightning.callbacks import ModelCheckpoint, EarlyStopping
|
|
from pytorch_lightning.loggers import MLFlowLogger
|
|
os.environ["HYDRA_FULL_ERROR"] = "1"
|
|
JOB_ID = os.environ.get("SLURM_JOBID")
|
|
|
|
@hydra.main(config_path="train_config",config_name="config")
|
|
def main(config: DictConfig):
|
|
|
|
callbacks = []
|
|
logger = MLFlowLogger(experiment_name=config.mlflow.experiment_name,
|
|
run_name=config.mlflow.run_name, tags={"JOB_ID":JOB_ID})
|
|
|
|
|
|
parameters = config.hyperparameters
|
|
|
|
dataset = instantiate(config.dataset)
|
|
model = instantiate(config.model,dataset=dataset,lr=parameters.get("lr"),
|
|
loss=parameters.get("loss"), metric = parameters.get("metric"))
|
|
|
|
direction = model.valid_monitor
|
|
checkpoint = ModelCheckpoint(
|
|
dirpath="./model",filename=f"model_{JOB_ID}",monitor="val_loss",verbose=False,
|
|
mode=direction,every_n_epochs=1
|
|
)
|
|
callbacks.append(checkpoint)
|
|
early_stopping = EarlyStopping(
|
|
monitor="val_loss",
|
|
mode=direction,
|
|
min_delta=0.0,
|
|
patience=parameters.get("EarlyStopping_patience",10),
|
|
strict=True,
|
|
verbose=False,
|
|
)
|
|
callbacks.append(early_stopping)
|
|
|
|
def configure_optimizer(self):
|
|
optimizer = instantiate(config.optimizer,lr=parameters.get("lr"),parameters=self.parameters())
|
|
scheduler = ReduceLROnPlateau(
|
|
optimizer=optimizer,
|
|
mode=direction,
|
|
factor=parameters.get("ReduceLr_factor",0.1),
|
|
verbose=True,
|
|
min_lr=parameters.get("min_lr",1e-6),
|
|
patience=parameters.get("ReduceLr_patience",3)
|
|
)
|
|
return {"optimizer":optimizer, "lr_scheduler":scheduler}
|
|
|
|
model.configure_parameters = MethodType(configure_optimizer,model)
|
|
|
|
trainer = instantiate(config.trainer,logger=logger,callbacks=callbacks)
|
|
trainer.fit(model)
|
|
logger.experiment.log_artifact(logger.run_id,f"./model/model_{JOB_ID}")
|
|
|
|
|
|
|
|
if __name__=="__main__":
|
|
main() |