Use UPath in helios/model.py

rslp/helios/model.py

@@ -1,7 +1,6 @@
 """Helios model wrapper for fine-tuning in rslearn."""
 
 import json
-import os
 from contextlib import nullcontext
 from typing import Any
 
@@ -12,6 +11,7 @@
 from helios.train.masking import MaskedHeliosSample, MaskValue
 from olmo_core.config import Config
 from olmo_core.distributed.checkpoint import load_model_and_optim_state
+from upath import UPath
 
 from rslp.log_utils import get_logger
 
@@ -63,6 +63,7 @@ def __init__(
             autocast_dtype: which dtype to use for autocasting, or set None to disable.
         """
         super().__init__()
+        _checkpoint_path = UPath(checkpoint_path)
         self.forward_kwargs = forward_kwargs
         self.embedding_size = embedding_size
         self.patch_size = patch_size
@@ -75,16 +76,16 @@ def __init__(
         # Load the model config and initialize it.
         # We avoid loading the train module here because it depends on running within
         # olmo_core.
-        with open(f"{checkpoint_path}/config.json") as f:
+        with (_checkpoint_path / "config.json").open() as f:
             config_dict = json.load(f)
             model_config = Config.from_dict(config_dict["model"])
 
         model = model_config.build()
 
         # Load the checkpoint.
         if not random_initialization:
-            train_module_dir = os.path.join(checkpoint_path, "model_and_optim")
-            if os.path.exists(train_module_dir):
+            train_module_dir = _checkpoint_path / "model_and_optim"
+            if train_module_dir.exists():
                 load_model_and_optim_state(train_module_dir, model)
                 logger.info(f"loaded helios encoder from {train_module_dir}")
             else: