Use the run timestamp as run name on WandB (#1619)

Makes it way more convenient to run many experiments by splitting by experiment name, and then by timestamp.

Co-authored-by: Matthew Tancik <tancik@berkeley.edu>

nerfstudio/utils/writer.py

@@ -283,7 +283,7 @@ class WandbWriter(Writer):
     """WandDB Writer Class"""
 
     def __init__(self, log_dir: Path):
-        wandb.init(project="nerfstudio-project", dir=str(log_dir), reinit=True)
+        wandb.init(project="nerfstudio-project", dir=str(log_dir), reinit=True, name=log_dir.name)
 
     def write_image(self, name: str, image: TensorType["H", "W", "C"], step: int) -> None:
         image = torch.permute(image, (2, 0, 1))