Add experiment name

src/batdetect2/cli/train.py

@@ -26,6 +26,7 @@ __all__ = ["train_command"]
 @click.option("--config-field", type=str)
 @click.option("--train-workers", type=int)
 @click.option("--val-workers", type=int)
+@click.option("--experiment-name", type=str)
 @click.option(
     "-v",
     "--verbose",
@@ -42,6 +43,7 @@ def train_command(
     config_field: Optional[str] = None,
     train_workers: int = 0,
     val_workers: int = 0,
+    experiment_name: Optional[str] = None,
     verbose: int = 0,
 ):
     logger.remove()
@@ -87,6 +89,7 @@ def train_command(
         model_path=model_path,
         train_workers=train_workers,
         val_workers=val_workers,
+        experiment_name=experiment_name,
         log_dir=log_dir,
         checkpoint_dir=ckpt_dir,
     )

src/batdetect2/train/logging.py

@@ -61,6 +61,7 @@ LoggerConfig = Annotated[
 def create_dvclive_logger(
     config: DVCLiveConfig,
     log_dir: Optional[data.PathLike] = None,
+    experiment_name: Optional[str] = None,
 ) -> Logger:
     try:
         from dvclive.lightning import DVCLiveLogger  # type: ignore
@@ -73,7 +74,9 @@ def create_dvclive_logger(
 
     return DVCLiveLogger(
         dir=log_dir if log_dir is not None else config.dir,
-        run_name=config.run_name,
+        run_name=experiment_name
+        if experiment_name is not None
+        else config.run_name,
         prefix=config.prefix,
         log_model=config.log_model,
         monitor_system=config.monitor_system,
@@ -83,12 +86,13 @@ def create_dvclive_logger(
 def create_csv_logger(
     config: CSVLoggerConfig,
     log_dir: Optional[data.PathLike] = None,
+    experiment_name: Optional[str] = None,
 ) -> Logger:
     from lightning.pytorch.loggers import CSVLogger
 
     return CSVLogger(
         save_dir=str(log_dir) if log_dir is not None else config.save_dir,
-        name=config.name,
+        name=experiment_name if experiment_name is not None else config.name,
         version=config.version,
         flush_logs_every_n_steps=config.flush_logs_every_n_steps,
     )
@@ -97,12 +101,13 @@ def create_csv_logger(
 def create_tensorboard_logger(
     config: TensorBoardLoggerConfig,
     log_dir: Optional[data.PathLike] = None,
+    experiment_name: Optional[str] = None,
 ) -> Logger:
     from lightning.pytorch.loggers import TensorBoardLogger
 
     return TensorBoardLogger(
         save_dir=str(log_dir) if log_dir is not None else config.save_dir,
-        name=config.name,
+        name=experiment_name if experiment_name is not None else config.name,
         version=config.version,
         log_graph=config.log_graph,
     )
@@ -111,6 +116,7 @@ def create_tensorboard_logger(
 def create_mlflow_logger(
     config: MLFlowLoggerConfig,
     log_dir: Optional[data.PathLike] = None,
+    experiment_name: Optional[str] = None,
 ) -> Logger:
     try:
         from lightning.pytorch.loggers import MLFlowLogger
@@ -122,7 +128,9 @@ def create_mlflow_logger(
         ) from error
 
     return MLFlowLogger(
-        experiment_name=config.experiment_name,
+        experiment_name=experiment_name
+        if experiment_name is not None
+        else config.experiment_name,
         run_name=config.run_name,
         save_dir=str(log_dir) if log_dir is not None else config.save_dir,
         tracking_uri=config.tracking_uri,
@@ -142,6 +150,7 @@ LOGGER_FACTORY = {
 def build_logger(
     config: LoggerConfig,
     log_dir: Optional[data.PathLike] = None,
+    experiment_name: Optional[str] = None,
 ) -> Logger:
     """
     Creates a logger instance from a validated Pydantic config object.
@@ -157,7 +166,11 @@ def build_logger(
 
     creation_func = LOGGER_FACTORY[logger_type]
 
-    return creation_func(config, log_dir=log_dir)
+    return creation_func(
+        config,
+        log_dir=log_dir,
+        experiment_name=experiment_name,
+    )
 
 
 def get_image_plotter(logger: Logger):

src/batdetect2/train/train.py

@@ -55,6 +55,7 @@ def train(
     val_workers: Optional[int] = None,
     checkpoint_dir: Optional[data.PathLike] = None,
     log_dir: Optional[data.PathLike] = None,
+    experiment_name: Optional[str] = None,
 ):
     config = config or FullTrainingConfig()
 
@@ -107,6 +108,7 @@ def train(
         targets=targets,
         checkpoint_dir=checkpoint_dir,
         log_dir=log_dir,
+        experiment_name=experiment_name,
     )
 
     logger.info("Starting main training loop...")
@@ -135,10 +137,14 @@ def build_trainer_callbacks(
     preprocessor: PreprocessorProtocol,
     config: EvaluationConfig,
     checkpoint_dir: Optional[data.PathLike] = None,
+    experiment_name: Optional[str] = None,
 ) -> List[Callback]:
     if checkpoint_dir is None:
         checkpoint_dir = "outputs/checkpoints"
 
+    if experiment_name is not None:
+        checkpoint_dir = f"{checkpoint_dir}/{experiment_name}"
+
     return [
         ModelCheckpoint(
             dirpath=str(checkpoint_dir),
@@ -164,13 +170,18 @@ def build_trainer(
     targets: TargetProtocol,
     checkpoint_dir: Optional[data.PathLike] = None,
     log_dir: Optional[data.PathLike] = None,
+    experiment_name: Optional[str] = None,
 ) -> Trainer:
     trainer_conf = conf.train.trainer
     logger.opt(lazy=True).debug(
         "Building trainer with config: \n{config}",
         config=lambda: trainer_conf.to_yaml_string(exclude_none=True),
     )
-    train_logger = build_logger(conf.train.logger, log_dir=log_dir)
+    train_logger = build_logger(
+        conf.train.logger,
+        log_dir=log_dir,
+        experiment_name=experiment_name,
+    )
 
     train_logger.log_hyperparams(
         conf.model_dump(
@@ -187,6 +198,7 @@ def build_trainer(
             config=conf.evaluation,
             preprocessor=build_preprocessor(conf.preprocess),
             checkpoint_dir=checkpoint_dir,
+            experiment_name=train_logger.name,
         ),
     )